Full text of "Vascular-Surgery-Books_1aim.net"

See other formats

Vascular Surgery

Dedication

To our families

Vascular Surgery

Basic Science and
Clinical Correlations

Second Edition

Edited by

Rodney A. White, md

Professor of Surgery
UCLA School of Medicine
Chief, Division of Vascular Surgery
Harbor-UCLA Medical Center
Torrance, California

Larry H. Hollier, md

Dean

Louisiana State University School of Medicine

in New Orleans

New Orleans, Louisiana

^-Blackwell

€r Futura

Blackwell Futura is an imprint of Blackwell Publishing

Blackwell Publishing, Inc., 350 Main Street, Maiden, Massachusetts 02148-5020, USA

Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK

Blackwell Science Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia

All rights reserved. No part of this publication may be reproduced in any form or by any
electronic or mechanical means, including information storage and retrieval systems, without
permission in writing from the publisher, except by a reviewer who may quote brief passages in
a review.

First edition 1994 by J.B. Lippincott Company
Second edition 2005

ISBN: 1-4051-2202-1

Library of Congress Cataloging-in-Publication Data

Vascular surgery : basic science and clinical correlations /edited by Rodney A. White and Larry
H. Hollier.-2nd ed.

p. ; cm.
Includes bibliographical references and index.
ISBN 1-4051-2202-1 (hardback : alk. paper)

I. Blood-vessels-Surgery 2. Blood-vessels-Pathophysiology 3. Blood-
vessels-Physiology

[DNLM: 1. Vascular Surgical Procedures. WG 170 V33132 2004] I. White, Rodney A.

II. Hollier, Larry H.

RD598.5.V37452004
617.4'13-dc22

A catalogue record for this title is available from the British Library

Acquisitions: Steven Korn

Production: Lindsey Williams, Prepress Projects Ltd

Typesetter: SNP Best-set Typesetter Ltd., Hong Kong, in Palatino 9 / 12 pt

Printed and bound in India by Gopsons Papers Limited, New Delhi

For further information on Blackwell Publishing, visit our website:
www.blackwellfutura.com

The publisher's policy is to use permanent paper from mills that operate a sustainable forestry
policy, and which has been manufactured from pulp processed using acid-free and elementary
chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board
used have met acceptable environmental accreditation standards.

Notice: The indications and dosages of all drugs in this book have been recommended in the
medical literature and conform to the practices of the general community. The medications
described do not necessarily have specific approval by the Food and Drug Administration for
use in the diseases and dosages for which they are recommended. The package insert for each
drug should be consulted for use and dosage as approved by the FDA. Because standards for
usage change, it is advisable to keep abreast of revised recommendations, particularly those
concerning new drugs.

Contents

Contributors, vii
Preface, xiii
Acknowledgments, xiv

I Vascular pathology and physiology

1 Embryology and development of the vascular system, 3
C. Phifer Nicholson and Peter Gloviczki

2 Vascular wall physiology, 19
Christian C. Haudenschild

3 Hemostasis and coagulation, 27
Donald L. Jacobs and Jonathan B. Towne

4 Molecular aspects of atherosclerosis, 43
/. Jeffrey Alexander and John A. Moawad

5 Localization of atherosclerotic lesions, 55
Christopher K. Zarins, Chengpei Xu, Charles A. Taylor,
and Seymour Glagov

6 Pathogenesis of arterial fibrodysplasia, 66
James C. Stanley

7 Physiology of vasospastic disorders, 80

Scott E. Musicant, Jean-Baptiste Roullet, James M. Edwards,
and Gregory L. Moneta

8 Buerger's disease, 92

John Blebea and Richard F. Kempczinski

9 Ergotism, 101
Roger F.J. Shepherd

10 Arteritis, 114
Francis J. Kazmier

11 Adventitial cystic disease, 119
Carlos E.Donayre

12 Entrapment syndromes, 126
Carlos E.Donayre

13 Intimal hyperplasia, 135
TedR.Kohler

14 Thoracic outlet syndrome, 146
Herbert I. Machleder

15 Aneurysmal disease, 162

Juan Carlos Jimenez and Samuel Eric Wilson

16 Pathophysiology of renovascular hypertension, 180
David L. Robaczewski, Richard H. Dean,

and KimberleyJ. Hansen

17 Pathophysiology, hemodynamics, and complications of
venous disease, 192

Harold J. Welch, Kevin B. Raftery, and Thomas F. 0'Donnell,Jr.

18 Physiologic changes in lymphatic dysfunction, 207
Peter Gloviczki

19 Physiologic changes in visceral ischemia, 215
Tina R. Desai, Joshua A. Tepper, and Bruce L. Gewertz

20 Natural history of atherosclerosis in the lower extremity,
carotid, and coronary circulations, 225

Daniel B. Walsh

21 Neurologic basis for sympathetically maintained pain:
causalgia and reflex sympathetic dystrophy, 233
Marco Scoccianti and Rodney A. White

22 Compartment syndromes physiology, 241
Malcolm O. Perry

23 Physiology of reperfusion injury, 245
Shervanthi Homer -Vanniasinkam and D. Neil Granger

24 Cerebral ischemia, 251

Hao Bui and Christian deVirgilio

25 Pathophysiology of spinal cord ischemia, 257
Larry H. Hollier

26 Vascular erectile dysfunction: mechanisms and current
approaches, 228

Ralph G. DePalma

Contents

27 Portal hypertension: pathophysiology and clinical
correlates, 275
David Rigberg and Hugh A. Gelabert

II Noninvasive vascular diagnostics

28 Physiologic basis of hemodynamic measurement, 295
R. Eugene Zierler

29 Spectral analysis, 306
Christopher R.B. Merritt

30 Ultrasound imaging, 315
Christopher R.B. Merritt

31 Radionuclide scanning, 325
Robert E. Sonnemaker

32 Computed tomography, 348
Anton Mlikotic and Irwin Walot

33 Magnetic resonance imaging, 371

David Saloner, Rem van Tyen, Charles M. Anderson,
and Gary R. Caputo

III Invasive vascular diagnostics

34 Angiography, 385

Anton Mlikotic and C. Mark Mehringer

35 Intravascular ultrasound, 401

James T. Lee, George Kopchok, and Rodney A. White

36 Angioscopy in peripheral vascular surgery, 423
Arnold Miller and Thomas J. Holzenbein

IV Medical management

37 Atherosclerosis: risk factors and medical
management, 441

Ralph G. DePalma and Virginia W. Hayes

38 Pharmacologic intervention: thrombolytic therapy, 454
Anthony J. Comerota, A. KonetiRao,

and Mohammad H. Eslami

39 Pharmacologic intervention: vasodilation therapy and
rheologic agents, 468

George Johnson, Jr.

40 Pharmacologic intervention: lipid-lowering agents, 473
Ralph G. DePalma

41 Infections and antibiotics in vascular surgery, 477
Martin R. Back

V Endovascular interventions for vascular
disease

42 Catheter-based approaches to the treatment of
atheroembolic disease, 495

Frank R. Arko, Christine Newman, and Thomas J. Fogarty

43 Balloon angioplasty and transluminal recanalization
devices, 503

Rajesh Subramanian and Stephen R. Ramee

44 Endovascular stents, 516

Frank J. Criado, YoussefRizk, Gregory S. Domer,
and Hilde Jerius

45 Endovascular prostheses for repair of abdominal aortic
aneurysms, 520

Carlos E.Donay re

VI Comparison of conventional vascular
reconstruction and endovascular techniques

46 Surgical and endovascular treatment of chronic ischemia
of the lower limbs, 533

Jean-Paul P.M. de Vries, Frans L. Moll, and Jos C. van den Berg

47 Aortoiliac endovascular recanalization compared with
surgical reconstruction, 543

Peter L. Paries and Michael L. Marin

48 Endovascular stent-graft repair of thoracic aortic
aneurysms and dissections, 554

Jason T Lee and Rodney A. White

49 Brachiocephalic vascular reconstructions compared with
endovascular repair, 567

Edward B. Diethrich

50 Carotid endarterectomy compared with carotid
angioplasty and stenting, 575

Mark R. Harrigan, RicardoA. Hanel, Elad I. Levy,
Lee R. Guterman, and L. Nelson Hopkins

5 1 Endovascular intervention for venous occlusion
compared with surgical reconstruction, 587
Patricia E. Thorpe and Francisco J. Osse

Index, 609

Colour plate section follows p. 370

Contributors

J. Jeffrey Alexander, MD

Associate Professor of Surgery
Case Western Reserve University
MetroHealth Medical Center
Cleveland, Ohio

Charles M. Anderson, MD, PhD

Clinical Professor of Radiology
VA Medical Center

University of California, San Francisco
San Francisco, California

Frank R. Arko, MD

Director, Endovascular Surgery
Assistant Professor of Surgery
Stanford University Medical Center
Stanford, California

Martin R. Back, MD

Assistant Professor of Surgery
University of South Florida;
Chief, Vascular Surgery
James A. Haley Veterans Hospital
Tampa, Florida

John Blebea, MD

Professor of Surgery

Department of Surgery

Temple University School of Medicine

Philadelphia, Pennsylvania

HaoBui, MD

Senior Resident
Department of Surgery
Harbor-UCLA Medical Center
Torrance, California

Anthony J. Comerota, MD, FACS

Director, Jobst Vascular Center
Toledo, Ohio

Frank J. Criado, MD

Director, Center for Vascular Intervention
Chief, Division of Vascular Surgery
Union Memorial Hospital/MedStar Health
Baltimore, Maryland

Richard H. Dean, MD

President and CEO

Wake Forest University Health Sciences

Winston-Salem, North Carolina

Ralph G. DePalma, MD, FACS

National Director of Surgery
Professor of Surgery

Uniformed Services of the Armed Forces;
National Director of Surgery
Department of Veterans Affairs
Washington, District of Columbia

Tina R. Desai, MD, FACS

Assistant Professor of Surgery
Department of Surgery
The University of Chicago
Chicago, Illinois

Christian deVirgilio, MD

Vice Chair, Education
Director, General Surgery Residency
Harbor-UCLA Medical Center;
Associate Professor of Surgery
UCLA School of Medicine
Torrance, California

Gary R. Caputo, MD

Associate Professor of Radiology
University of California, San Francisco
San Francisco, California

Jean-Paul P.M. de Vries, MD, PhD

Vascular Surgeon
St. Antonius Hospital
Nieuwegein
The Netherlands

VII

Contributors

Edward B. Diethrich, MD

Medical Director

Arizona Heart Institute and Arizona Heart Hospital

Phoenix, Arizona

Gregory S. Domer, MD

Center for Vascular Intervention and Division of Vascular Surgery
Union Memorial Hospital-MedStar Health
Baltimore, Maryland

Carlos E. Donayre, MD

Associate Professor of Surgery
Harbor-UCLA Medical Center
Torrance, California

James M. Edwards, MD

Chief of Surgery

Portland VAMC;

Associate Professor of Surgery

Division of Vascular Surgery

Oregon Health and Science University

Portland, Oregon

Mohammad H. Eslami, MD

Assistant Professor of Surgery
Temple University School of Medicine
Philadelphia, Pennsylvania

Peter L. Faries, MD, FACS

Chief of Endovascular Surgery
New York Presbyterian Hospital
Weill Cornell Medical School
New York, New York

Thomas J. Fogarty, MD

Clinical Professor of Surgery
Stanford University Medical Center
Stanford, California

HughA.Gelabert, MD

Assistant Professor of Surgery
Section of Vascular Surgery
UCLA School of Medicine
Los Angeles, California

Bruce L. Gewertz, MD, FACS

The Dallas B. Phemister Professor
Chairman, Department of Surgery
The University of Chicago
Chicago, Illinois

Seymour Glagov, MD

Professor Emeritus of Pathology and Surgery
Department of Surgery
Section of Vascular Surgery
The University of Chicago
Chicago, Illinois

Peter Gloviczki, MD

Associate Professor of Surgery
Mayo Clinic
Rochester, Minnesota

D. Neil Granger, PhD

Boyd Professor

Head, Department of Molecular and Cellular Physiology

LSU Health Sciences Center

Shrieveport, Louisiana

Lee R. Guterman, MD, PhD

Assistant Professor, Department of Neurosurgery

Co-Director, Toshiba Stroke Research Center

School of Medicine and Biomedical Sciences

University at Buffalo

State University of New York

Buffalo, New York

Ricardo A. Hanel, MD

Assistant Clinical Instructor of Neurosurgery

Neuroendovascular Fellow

Department of Neurosurgery and Toshiba Stroke Research Center

School of Medicine and Biomedical Sciences

University at Buffalo

State University of New York

Buffalo, New York

Kimberley J. Hansen, MD

Professor of Surgery

Department of General Surgery;

Head, Section of Vascular Surgery

Division of Surgical Sciences

Wake Forest University School of Medicine

Winston-Salem, North Carolina

Mark R. Harrigan, MD

Assistant Clinical Instructor of Neurosurgery and

Neuroendovascular Fellow

Department of Neurosurgery and Toshiba Stroke Research Center

School of Medicine and Biomedical Sciences

University at Buffalo

State University of New York

Buffalo, New York

Christian C. Haudenschild, MD

Professor of Pathology and Medicine

George Washington University Medical Center

Washington, District of Columbia

Virginia W. Hayes, RN, MS, CFNP, CVN

Nurse Practitioner for Primary Care and Surgical Research
VA Sierra Nevada Health Care System
Reno, Nevada

Larry H. Hollier, MD

Dean, Louisiana State University School of Medicine in New Orleans
New Orleans, Louisiana

VII

Contributors

Thomas J. Holzenbein, MD

Fellow in Vascular Research
Harvard Medical School
Division of Vascular Surgery
Harvard-Deaconess Surgical Service
New England Deaconess Hospital
Boston, Massachusetts

Shervanthi Homer-Vanniasinkam,
IBSc, MD, FRCSEd, FRCS

Professor, Consultant Vascular Surgeon

Vascular Surgery Unit

Leeds General Infirmary

Leeds

United Kingdom

L. Nelson Hopkins, MD

Professor and Chairman, Department of Neurosurgery;

Professor, Department of Radiology; and Director

Toshiba Stroke Research Center;

School of Medicine and Biomedical Sciences

University at Buffalo

State University of New York

Buffalo, New York

Donald L. Jacobs, MD, MS

Associate Professor of Surgery
St. Louis University
St. Louis, Missouri

Hilde Jerius, MD

Center for Vascular Intervention and Division of Vascular Surgery
Union Memorial Hospital-MedStar Health
Baltimore, Maryland

Juan Carlos Jimenez, MD

Department of Surgery
University of California, Irvine
Irvine, California

George Johnson, Jr., MD

Roscoe B.G. Cowper Distinguished Professor of Surgery
Vice Chairman, Department of Surgery

University of North Carolina at Chapel Hill School of Medicine
Chapel Hill, North Carolina

Francis J. Kazmier, MD, FACC

Ochsner Clinic Foundation
New Orleans, Louisiana

Richard F. Kempczinski, MD

Professor of Surgery Emeritus

University of Cincinnati School of Medicine

Cincinnati, Ohio

Ted R. Kohler, MD

Chief of Vascular Surgery

Surgical Service of the Veterans Affairs Puget Sound Health Care

System;

Professor of Surgery
Department of Surgery
University of Washington
Seattle, Washington

George Kopchok, BS

Biomedical Engineering
Research and Education Institute
Division of Vascular Surgery
Harbor-UCLA Medical Center
Torrance, California

James T. Lee, MD

Clinical Faculty, Surgical Services
UCLA School of Medicine
Harbor-UCLA Medical Center Campus;
Peripheral Vascular and Endovascular Surgery
Southern California Permanente Medical Group
Bellflower Medical Center
Bellflower, California

Jason T. Lee, MD

Vascular Surgery Fellow
Division of Vascular Surgery
Stanford University Medical Center
Stanford, California

Elad I. Levy, MD

Assistant Clinical Instructor of Neurosurgery and

Neuroendovascular Fellow

Department of Neurosurgery and Toshiba Stroke Research Center

School of Medicine and Biomedical Sciences

University at Buffalo

State University of New York

Buffalo, New York

Herbert I. Machleder, MD

Department of Surgery
UCLA Medical Center
Los Angeles, California

Michael L. Marin, MD, FACS

Chief, Division of Vascular Surgery
Mount Sinai School of Medicine
New York, New York

C. Mark Mehringer, MD

Professor of Radiological Sciences
David Geffen School of Medicine at UCLA
Harbor-UCLA Medical Center
Torrance, California

Christopher R.B. Merritt, MD, FACR

Professor of Radiology
Department of Radiology
Thomas Jefferson University Hospital
Philadelphia, Pennsylvania

Contributors

Arnold Miller, MD

Attending Vascular Surgeon

Department of Surgery

Metro West Medical Center

Framingham-Natick, Massachusetts;

Assistant Clinical Professor of Surgery, Harvard Medical School,

Boston, Massachusetts

Anton Mlikotic, MD

Assistant Professor of Radiological Sciences
David Geffen School of Medicine at UCLA
Harbor-UCLA Medical Center
Torrance, California

John A. Moawad, MD

Assistant Professor of Surgery
Case Western Reserve University
MetroHealth Medical Center
Cleveland, Ohio

FransL. Moll, MD, PhD

Professor of Vascular Surgery

Head of the Department of Vascular Surgery

University Medical Center Utrecht

Utrecht

The Netherlands

Gregory L. Moneta, MD

Professor of Surgery
Chief, Division of Vascular Surgery
Oregon Health and Science University
Portland, Oregon

Scott E. Musicant, MD

Research Fellow in Vascular Surgery
Division of Vascular Surgery
Oregon Health and Science University
Portland, Oregon

Christine Newman, RIM

Fogarty Research
Portola Valley, California

C. Phifer Nicholson, MD

Surgical Consultants, PA.
Edina, Minnesota

Thomas F. O'Donnell, Jr., MD

Professor of Surgery

President and CEO

New England Medical Center

Tufts University School of Medicine

Boston, Massachusetts

Francisco J. Osse, MD

Associate Professor of Radiology

Division of Vascular and Interventional Radiology

University of Iowa
Iowa City, Iowa

Malcolm O. Perry, MD

Professor Emeritus

The University of Texas Southwestern Medical School

Dallas, Texas

Kevin B. Raftery, MD

Lahey Clinic Medical Center
Burlington, Massachusetts

Stephen R. Ramee, MD, FACC

Section Head, Interventional Cardiology
Ochsner Clinic Foundation
New Orleans, Louisiana

A. Koneti Rao, MD

Professor of Medicine

Temple University School of Medicine

Philadelphia, Pennsylvania

David Rigberg, MD

Clinical Fellow
Section of Vascular Surgery
UCLA School of Medicine
Los Angeles, California

Youssef Rizk, DO

Center for Vascular Intervention and Division of Vascular Surgery
Union Memorial Hospital-MedStar Health
Baltimore, Maryland

David L. Robaczewski, MD

Bradshaw Fellow of Surgical Research
Department of General Surgery
Division of Surgical Sciences
Wake Forest University School of Medicine
Winston-Salem, North Carolina

Jean-Baptiste Roullet, PhD

Director, Basic Science Research
Division of Vascular Surgery
Oregon Health and Science University
Portland, Oregon

David Saloner, PhD

Professor of Radiology

VA Medical Center

University of California, San Francisco

San Francisco, California

Marco Scoccianti, MD, EBSQ (vase)

Head, Endovascular Surgery Unit

Division of Vascular Surgery

S. Giovanni-Addolorata Hospital Complex

Rome

Italy

Contributors

Roger F.J. Shepherd, MB, BCh

Assistant Professor of Medicine
Mayo Clinic College of Medicine
Mayo Clinic
Rochester, Minnesota

Robert E. Sonnemaker, MD

Medical Director

PET Imaging

Department of Nuclear Medicine

St. John's Health System

Springfield, Missouri

James C. Stanley, MD

Professor of Surgery
Head, Section of Vascular Surgery
University of Michigan Medical Center
Ann Arbor, Michigan

Rajesh Subramanian, MD, FACC

Ochsner Clinic Foundation
New Orleans, Louisiana

Charles A. Taylor, PhD

Assistant Professor of Mechanical Engineering,
Surgery and Pediatrics (by courtesy)
Stanford University
Stanford, California

Joshua A. Tepper, MD

Resident in General Surgery
Department of Surgery
The University of Chicago
Chicago, Illinois

Patricia E. Thorpe, MD

Professor of Radiology
University of Iowa
Iowa City, Iowa

Jonathan B. Towne, MD

Professor of Surgery
Chairman, Division of Vascular Surgery
Medical College of Wisconsin
Milwaukee, Wisconsin

Jos C. van den Berg, MD, PhD

Interventional Radiology
San Antonio Hospital
Nieuwegein
The Netherlands

Rem van Tyen, PhD

Assistant Research Physicist

VA Medical Center

University of California, San Francisco

San Francisco, California

Irwin Walot, MD

Associate Professor of Radiological Sciences
Chief, Cardiovascular/Interventional Radiology
Harbor-UCLA Medical Center
Torrance, California

Daniel B. Walsh, MD

Section of Vascular Surgery
Dartmouth-Hitchcock Medical Center
Dartmouth Medical School
Lebanon, New Hampshire

Harold J. Welch, MD

Lahey Clinic Medical Center
Burlington, Massachusetts

Rodney A. White, MD

Professor of Surgery
UCLA School of Medicine;
Chief, Division of Vascular Surgery
Harbor-UCLA Medical Center
Torrance, California

Samuel Eric Wilson, MD, FACS

Professor and Chair
Department of Surgery;
Associate Dean
University of California, Irvine
Irvine, California

Chengpei Xu, MD, PhD

Senior Research Scientist
Division of Vascular Surgery
Stanford University School of Medicine
Stanford, California

Christopher K. Zarins, MD

Professor of Surgery
Division of Vascular Surgery
Stanford University School of Medicine
Stanford, California

R. Eugene Zierler, MD

Professor of Surgery

Medical Director

Vascular Diagnostic Services

University of Washington Medical Center

Seattle, Washington

Preface

This revised edition of Vascular Surgery: Basic Science and Clini-
cal Correlations was developed in order to address significant
changes that have occurred in contemporary vascular surgery
and to highlight new information that has developed regard-
ing vascular imaging and interventional and endovascular
procedures. The overall length of the text is slightly shorter
than the first edition with relevant core chapters being
retained to emphasize the basic science nature of the text,
with approximately 60 percent of the material undergoing
major revisions or being new chapters.

The significant change from the first text is an emphasis on
vascular pathology and physiology that is relevant to current
practice, including information that is currently included
on the vascular board examinations. A new emphasis on
endovascular therapies has been added by including five
chapters on endovascular techniques and an additional
section with six chapters comparing conventional vascular

reconstruction with endovascular methods. These new
chapters address the most important issue in contemporary
vascular surgery, i.e. the role of endovascular methods in
treating vascular lesions and the impact that this has on
training and credentialing. A unique aspect of this book
differentiating it from other texts is a comparison of conven-
tional methods with the endovascular techniques.

Overall, the text provides a comprehensive approach to
contemporary vascular surgery and future perspectives. The
authors are preeminent in the field and are most capable for
addressing the assigned topics, with the goals being to provide
an updated and forward-looking text that accommodates the
needs of practicing and training vascular surgeons.

Rodney A. White
Larry H. Hollier

XII

Acknowledgments

We would like to acknowledge the efforts of Blackwell Pub- Strauss, and the invaluable expertise of Joanna Bellhouse,
lishing, Futura Division, for the timely preparation of this text. Development Editor, who has meticulously and efficiently
In particular, we appreciate the efforts of Steve Korn, Jacques organized materials and prepared the text for publication.

XIV

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Vascular pathology
and physiology

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Embryology and development of
the vascular system

C. Phifer Nicholson
Peter Gloviczki

The vascular system develops between the third and eighth
weeks of gestation. In the middle of the third week, the embryo
is no longer able to meet its nutritional requirements by diffu-
sion alone, thus prompting differentiation of extraembryonic
mesodermal cells (angioblasts) located in the wall of the yolk
sac. These angioblasts form angiogenic cell clusters, which
canalize to form early blood vessels. Cells that are centrally lo-
cated in these clusters differentiate into blood cells, while
those at the periphery flatten and form endothelial cells. 1
Similarly, during this same period, intraembryonic mesoder-
mal cells differentiate to form the heart tube, paired dorsal aor-
tae, visceral arteries, and axial arteries of the developing limb
buds. Woollard 2 described the above events in the develop-
ment of the vascular system in three stages: (1) the capillary
network stage, an undifferentiated network of primitive blood
lakes; (2) the retiform stage, when separation of the primitive
arterial and venous channels occurs; and (3) the gross differen-
tiation phase with the appearance of mature vascular chan-
nels. By the end of the eighth week of gestation, development
of the vascular system is virtually complete with only minor
changes occurring after this time.

Arterial system

Aortic arch and great vessels

The aortic arch and its major branches develop from the six
embryologic aortic arches, which, in turn, originate from the
aortic sac. Each branchial arch is supplied by one of the aortic
arches. The fifth aortic arch is often not formed at all (Fig. 1.1).
In the 4-mm embryo (end of fourth week), the first aortic arch
has nearly disappeared with only a small portion persisting on
the maxillary artery (Fig. 1.2). The second aortic arch also
regresses with portions persisting as the hyoid and stapedial
arteries. 1

In the 1 0-mm embryo (beginning of sixth week), the first and
second aortic arches have disappeared and the third, fourth,
and sixth aortic arches enlarge (Fig. 1.3). The third aortic arch is

the anlage of the common carotid artery and the first portion of
the internal carotid artery with the remainder of the internal
carotid artery formed by the dorsal aorta (Fig. 1.4). 1 The proxi-
mal right subclavian artery develops from the right fourth aor-
tic arch. Its distal portion is formed by a portion of the right
dorsal aorta and the seventh intersegmental artery (see Fig.
1.4). The embryologic left fourth aortic arch forms the arch of
the aorta between the left common carotid and left subclavian
arteries.

The fifth aortic arch is transient and never well developed.
No portion persists in the extrauterine life.

The sixth aortic arch (pulmonary arch) gives off branches
that grow toward the developing lung bud. The right sixth aor-
tic arch forms the proximal segment of the right pulmonary
artery, while the distal left sixth aortic arch persists as the duc-
tus arteriosus; it later becomes the ligamentum anteriosum
(see Fig. 1.4).

Formation of the neck causes the heart to descend from its
initial cervical position into the thoracic cavity. This results in
elongation of the innominate and carotid arteries and a shift of
the origin of the left subclavian artery from the level of the sev-
enth intersegmental artery to a point closer to the origin of the
left common carotid artery (Fig. 1.5). In embryologic develop-
ment, the recurrent laryngeal nerves supply the sixth
branchial arches. With the caudal shift of the heart and disap-
pearance of portions of the right fifth and sixth aortic arches,
the right recurrent laryngeal nerve moves up to hook around
the fourth aortic arch while the left recurrent laryngeal nerve
hooks around the ligamentum anteriosum (see Figs. 1.4 and
1.5).

Visceral arteries

Most of the differentiation of the arterial supply to the abdomi-
nal viscera has occurred by the end of the eighth week. The pri-
mordium of the celiac artery is represented by the paired
cephalic roots of the vitelline arteries at the level of the 10th ven-
tral segmental artery. The superior mesenteric artery originates
by fusion of the paired vitelline arteries at the level of the 13th

pa rt I Vascular pathology and physiology

Pharyngeal pouches

aortic arch

■

aortic arch

Esophagus
Dorsal aorta

Trachea
and lung bud

Figure 1.1 Aortic arches supplying branchial
clefts and pharyngeal pouches.

Obliterated
aortic arch I

--III

-'IV

Primitive
pulmonary artery

Right dorsal aorta-

- - - Left dorsal aorta

SVma

fMAYO

VI-

Ascending aorta

' %/*

Septum between

aorta and
pulmonary artery

Pulmonary
trunk

Primitive
pulmonary artery

Left 7th

intersegmental

artery

Figure 1 .2 Aortic arches at the end of fourth week of development.

MAYO
©1^32

ventral segmental artery Fusion of the vitelline arteries in a
more caudal location forms the inferior mesenteric artery.

Renal arteries

The adult kidney (metanephros) begins to develop in the
fifth week of gestation and is initially located in the
pelvis. With diminution of the body curvature and growth

Figure 1 .3 Aortic arches at the beginning of sixth week of development
with early pulmonary arteries.

of the body in the lumbar and sacral regions, the kidney
ascends into the abdomen. The metanephros receives its origi-
nal blood supply from a pelvic branch of the aorta but as it
ascends, arteries originating from successively higher levels of
the abdominal aorta supply the kidney while the lower vessels
degenerate. 1

chapter 1 Embryology and development of the vascular system

Aortic
arches

Figure 1 .4 Transformation of aortic arches
into adult configuration.

Right

dorsal aorta - \

Internal carotid
artery

Right vagus ■
nerve

Common

carotid artery - ?.- -f -

Right
subclavian

artery

Right recurrent -
nerve

7th intersegmental _ *
artery

External

carotid
arteries

Left
f vagus
1 nerve

,Arch of

t aorta
Si

_Left

recurrent
nerve

.Ductus
arteriosus

* Pulmonary
artery

MAVO

IfNKJ

Right external .
carotid artery

Right vagus v
nerve

Right subclavian _
artery

Brachiocephalic'
artery

Ascending aorta

Pulmonary artery

Left interna!
carotid artery

Left common
carotid artery

Left subclavian
artery

Ugamentum
arteriosum

- -Descending aorta

mayo

Figure 1.5 Adult configuration of great vessels. Note position of recurrent
laryngeal nerves.

Arteries to the lower extremity

During the fifth week of development (6-mm embryo), the
umbilical artery gives rise to the sciatic artery The sciatic
artery is a continuation of the internal iliac artery, which devel-
ops with the lower limb bud as its axial artery The femoral
artery, an extension of the external iliac artery, replaces the sci-
atic artery and its branches to the thigh during the eighth week
of development. 3 Adult derivatives of the sciatic system
include the popliteal, anterior tibial, and peroneal arteries.

Proximal portions of the umbilical arteries persist to form
the internal iliac and superior vesical arteries. 1

Venous system

During the fifth week of gestation, three major pairs of veins
are present in the embryo: (1) vitelline or omphalomesenteric
veins between the yolk sac and the sinus venosus; (2) umbilical
veins, which course between the chorionic villi and the
embryo; and (3) cardinal veins, which drain the body of the
embryo (Fig. 1.6).

Vitelline vein derivatives

The vitelline veins pass from the yolk sac to the venous plexus
surrounding the duodenum prior to passing into the septum
transversum (Fig. 1.7). Liver cords budding from the duode-
num grow into the septum transversum, interrupting the
course of the vitelline veins to form the hepatic sinusoids. The
left and right hepatocardiac channels drain the hepatic sinu-
soids into the sinus venosus (Fig. 1.8). With obliteration of the
left hepatocardiac channel, the right hepatocardiac channel
becomes the posthepatic (suprahepatic) inferior vena cava.
The portal vein forms as the venous plexus surrounding the
duodenum coalesces into a single vein. The superior mesen-
teric vein develops from the distal right vitelline vein.

Umbilical vein derivatives

The entire right umbilical vein and the proximal portion of the
left umbilical vein disappear, while the distal left umbilical
vein persists to carry blood to the liver from the placenta. A
communication, the ductus venosus, later forms between the

pa rt I Vascular pathology and physiology

Common cardinal vein

Anterior cardinal vein

Aortic arches

Internal
carotid '
artery

Dorsal aorta

Posterior cardinal vein

Vitelline vein

Vitelline artery

Chorionic villus

■

- -Chorion

MAYO
8 '99?

Figure 1 .6 Venous system at end of fifth week
of gestation.

Sinus venosus

He pato- cardiac

channels

Cardinal vein

Liver buds

-- 'Left vitelline
vein

Duodenum

--Umbilical
vein

MAYO

©'39?

Figure 1.7 Vitelline veins forming venous plexusaround duodenum.

left umbilical vein and the right hepatocardiac channel, by-
passing the sinusoids of the liver (Fig. 1.9). After birth, the left
umbilical vein and the ductus venosus are obliterated to
form the ligamentum teres hepatis and ligamentum venosum,
respectively.

Cardinal vein derivatives

In early embryologic development, the cardinal venous sys-
tem is composed of three pairs of veins: (1) the anterior cardi-
nal veins, which drain the cephalic embryo; (2) the posterior
cardinal veins, which drain the remainder of the embryo; and

Cardinal vein

Right umbilical

vem

-Hepatic
sinusoids

Left umbilical
vein

Duodenum

Figure 1.8 Liver cords interrupting course of vitelline veins.

(3) the common cardinal veins, which are formed by the junc-
tion of the anterior and posterior cardinal veins (see Fig. 1.6).
During the fifth to seventh weeks of gestation, the following
veins form: (1) the subcardinal veins, which drain the kidneys;
(2) the sacrocardinal veins, which drain the lower extremities;
and (3) the supracardinal veins, which drain the body wall via
intercostal veins (Fig. 1.10).

In the formation of the vena cava, anastomoses develop be-
tween the left and right sides of the cardinal system, channel-
ing blood from left to right. The communication between the
anterior cardinal veins develops into the left brachiocephalic
vein. The right common cardinal vein and the proximal por-
tion of the right anterior cardinal vein form the superior vena
cava.

chapter 1 Embryology and development of the vascular system

Duodenum -

Right

hepatocardiac
channel

Figure 1 .9 Formation of hepatic veins, hepatic
portion of inferior vena cava, and portal vein.

Portal vein

Vitelline veins

--■Left umbilical
vein

Hepatic portion of
inferior vena cava

Hepatic vein

Superior - -
mesenteric "

vein

Splenic vein

r /- Left umbilical
vein

Anterior-"
cardinal vein

Superior - -
vena cava

Azygos vein ^

Posterior- —
cardinal vein

Subcardial vein

Renal segment - -
inferior vena cava

Supracardinal
vein

-Hemiazygos vein

- Hepatic segment
inferior vena cava

h- Left renal vein

Left gonadal vein

MAYO

Sacrocardinal vein

Figure 1.10 Development of the venous system. (A) In seventh week. (B) At birth.

Superior- -
vena cava

Azygos vein

Hepatic segment

Renal segment

Sacrocardinal -
segment

Left

brachiocephalic

vein

- Coronary
sinus

* Hemiazygos
vein

MAYO

Left common

iliac vein

The communication between the subcardinal veins forms
the left renal vein. After development of this communication,
the proximal left subcardinal vein disappears with its distal
portion persisting as the left gonadal vein. 1 Hence, the right
subcardinal vein becomes the renal segment of the inferior
vena cava (see Fig. 1.10).

The communication between the sacrocardinal veins be-

comes the left common iliac vein. The left sacrocardinal vein
then involutes while the right sacrocardinal vein persists to
become the sacrocardinal segment of the inferior vena cava. 1

As portions of the posterior cardinal veins disappear, the
supracardinal veins become more important. The azygos vein,
into which the 4th through 11th intercostal veins empty, forms
from the right supracardinal vein and a portion of the right

pa rt I Vascular pathology and physiology

posterior cardinal vein (see Fig. 1.10). The hemiazygous vein,
into which the fourth through seventh intercostal veins empty,
develops from the left supracardinal vein. 1

Lymphatic system

Disagreement remains as to the origin of the lymphatics but
the leading theories are the centrifugal theory proposed by
Lewis 4 and Sabin 5 and the centripetal theory proposed by
Huntington. 6 According to the centrifugal theory, the lym-
phatics are believed to arise by proliferation from the venous
system. The centripetal theory, however, suggests that lym-
phatics form from coalescence of mesenchymal spaces into a
system of vessels.

By the sixth week of gestation, paired jugular lymph sacs are
identifiable in the vicinity of the anterior cardinal veins. The
cisterna chyli dorsal to the aorta and retroperitoneal lymph
sacs at the root of the mesentery are present by the end of the
eighth week of development. Communications between the
jugular lymph sacs and the cisterna chyli develop, forming a
paired system of lymphatic trunks with numerous anasto-
moses across the midline. Portions of the right and left systems
will involute so that in adults the major lymphatic system
consists of left and right lumbar lymphatic trunks, which
drain into the cisterna chyli and then the thoracic duct. The
thoracic duct has an inferior right portion, then crosses the
midline at the level of the fourth to sixth thoracic vertebrae
to eventually empty into the left subclavian vein at its
junction with the left internal jugular vein (Fig. 1.11). The
thoracic duct, therefore, provides lymph drainage for the
left upper extremity, the chest, abdomen, and the lower
extremities. Lymph from the head, neck, and right upper
extremity drains into the right subclavian vein via the right
cervical lymphatic trunk.

Embryologic derangements in
vascular pathology

Arterial anomalies

Anomalies of the aortic arch

True anomalies of aortic arch are rare; they occur in less than
2% of adults.

Right aortic arch results from obliteration of the left fourth
aortic arch and the left dorsal aorta, which are replaced by cor-
responding vessels on the right side.

Double aortic arch or aortic ring results from persistence of
the right dorsal aorta between the seventh intersegmental
artery and its junction with the left dorsal aorta (Fig. 1.12). The
aortic ring thus formed surrounds the trachea and the esopha-
gus, compressing these structures.

Thoracic
duct

H Cisterna
chyli

Figure 1.11 Adult configuration of major lymphatic channels.

Interrupted aortic arch is also a relatively rare anomaly,
resulting from obliteration of the left fourth aortic arch
(Fig. 1.13). The ductus arteriosus remains widely patent,
supplying blood of low oxygen content to the systemic circula-
tion while the aortic trunk supplies the two common carotid
arteries.

Anomalies of the aortic arch branches

Common ostial origin of the innominate and left common carotid ar-
teries, the most common anomaly of the arch branches, occurs
in approximately 10% of patients. Origin of the left vertebral
artery from the aortic arch proximal to the left subclavian artery
occurs in 5% of patients.

Aberrant right subclavian artery (arteria lusoria) occurs in ap-
proximately 2% of patients, resulting from obliteration of the
right fourth aortic arch and proximal right dorsal aorta (Fig.
1.14). In this anomaly, the right subclavian artery arises from
the aortic arch just distal to the left subclavian artery, passing
behind the esophagus to the right arm, frequently compress-
ing the esophagus (dysphagia lusoria). Absence of the normal
origin of the right subclavian artery results in a nonrecurrent
right recurrent laryngeal nerve.

Coarctation of the aorta

Coarctation of the aorta may be congenital or acquired and

chapter 1 Embryology and development of the vascular system

Figure 1.12 Persistent right dorsal aorta,
which forms double aortic arch (aortic ring)
surrounding trachea and esophagus.

Persistent
portion of *
right dorsal
aorta

Right common

Subclavian carotid arter *

artery

Trachea

Left common
carotid artery

Brachiocephalic

artery

Ascending
aorta

, Esophagus

Left subclavian
artery

Right aortic arch
Left aortic arch

Descending

aorta

@'»;

Abnormal
obliteration l

Figure 1.13 (A) Interrupted aortic arch.
Abnormal obliteration of right and left fourth
aortic arches with persistence of portion of
right dorsal aorta. (B) Aorta supplies head while
pulmonary artery via patent ductus arteriosus
supplies remainder of body.

Persistent - "
portion of right
dorsal aorta

Right common
carotid artery

Right

subclavian

artery-.

Abnormal
obliteration

Left common
carotid artery

Left

subclavian

artery

- Patent
ductus

Aorta ■

Pulmonary \
artery

MAVO

Interrupted Aortic Arch

may occur in the descending thoracic aorta or the abdominal
aorta. Our discussion will focus on congenital coarctation.

Several hypotheses have been proposed as causes of
congenital coarctation of the aorta. According to Dean and
coworkers, 7 congenital coarctations result from either failure
of maturation of the mesenchymal cell component or arrested
development of the artery during the period of gross differen-
tiation. If arrest occurs during the mesenchymal cell stage, the
artery may appear as a fibrous cord. With developmental
arrest during the gross differentiation phase, the aorta may
appear normal in early childhood, but later may be recognized
as a nonexpanding portion of aorta adjacent to a normally
growing segment.

With aortic coarctation from anomalous mesenchymal cell
maturation, luminal fibrous clefts and ridges causing partial
obstruction may be noted on arteriography. Microscopically,
dysplastic mesenchymal cell layers compose a disorganized
media.

Coarctation of the thoracic aorta may be preductal or postduc-
tal. In preductal aortic coarctation, the ductus arteriosus per-
sists supplying poorly oxygenated blood to the lower body. In
the postductal type, this channel is obliterated and numerous
collaterals from the subclavian and axillary arteries supply the
lower body.

Coarctation of the abdominal aorta is rare, accounting for 0.5%
to 2% of clinically recognized coarctations of the thoracic and

pa rt I Vascular pathology and physiology

Trachea

Esophagus

i nterseg mental
artery

Right dorsal aorta -
(abnormal right
subclavian artery)

Common

carotid * 3
arteries

Ascending
aorta

Left
/subclavian
t artery

Right

subclavian

artery

Descending
aorta

MAYO

Figure 1.14 (A) Aberrant right subclavian
artery. Abnormal obliteration of right fourth
aortic arch and proximal right dorsal aorta. (B)
Aberrant right subclavian artery passing
posterior to trachea and esophagus.

"Normal"
celiac axis

Gastrosplenic
trunk

Hepatosplenic
trunk

Hepatogastric Single celiac

trunk superior mesenteric

arterial trunk

Figure 1.15 Celiacarteryanomalies.

abdominal aorta. Reconstruction may be challenging because
the stenosis may extend from the celiac axis to the infrarenal
abdominal aorta. In about 80% of patients, renal artery steno-
sis with renovascular hypertension is present. Untreated ab-
dominal coarctation may eventually result in cardiac failure or
cerebral hemorrhage, the major causes of death from this
anomaly 8 Repair often requires renal revascularization and
bypass or replacement of the narrowed aorta in the second or
third decade of life. 9

Anomalies of the visceral arteries

Congenital anomalies of the visceral arteries are not uncom-
mon; however, visceral arterial anomalies requiring vascular
surgical intervention are rare. We define a visceral artery anom-
aly as a difference in number or origin of the arterial supply to
an organ from the accepted normal. The normal arterial sup-
ply of an organ is that pattern of arteries to a viscus that occurs
most commonly. Celiac, hepatic, and renal arterial anomalies
of importance to the vascular surgeon are described.

Celiac artery anomalies are found in 11% to 40% of patients.

The typical celiac axis, which branches into left gastric, splenic,
and common hepatic arteries, is found in 60% to 89% of pa-
tients. The most common variation is a gastrosplenic trunk
with the common hepatic artery arising from the aorta or the
superior mesenteric artery occurring in 5% to 8% of patients. 10
Hepatosplenic and hepatogastric trunks occur less frequently
and, rarely, the celiac axis may be combined with the superior
mesenteric artery (Fig. 1.15).

Hepatic artery anomalies may be of two types: replaced or ac-
cessory. A replaced hepatic substitutes for a normal hepatic
artery that is absent, while an accessory hepatic is an addition
to the normal one that is present. Michels, 11 from 200 anatomic
dissections, found one or more hepatic artery anomalies in 83
cases (41%). The four most common variations in the arterial
supply to the liver were (1) replaced right hepatic artery, 17%;
(2) replaced left hepatic artery, 16%; (3) accessory left hepatic
artery, 12%; and (4) accessory right hepatic artery, 8% (Fig.
1.16). In 2.5% of his dissections, Michels noted the common he-
patic artery originated from the superior mesenteric artery.

As previously described, during embryologic develop-
ment, the kidney arterial supply originates from the aorta at

chapter 1 Embryology and development of the vascular system

M.H.

/— l.h.

Figure 1.16 Hepaticartery anomalies (C. A.,
celiac axis; L.G., left gastric; H, hepatic; M.H.,
middle hepatic; R.H., right hepatic; L.H., left
hepatic).

58%

Access R.H

M - H - L-Access L.H. %

L.H. *

M.H.

16%

C.A.

12%

MAVO
©199Z

successively higher levels as the kidney ascends from the
pelvis. Failure of lower vessels to degenerate results in multi-
ple renal arteries, present in 25% to 33% of adults. Multiple
renal arteries are slightly more common on the left than the
right and may enter the renal hilum or directly into the
parenchyma of one of the poles of the kidney. Supernumerary
arteries most commonly enter the upper pole of the kidney
and are more common in ectopic kidneys. Lower pole super-
numerary arteries to the right kidney typically cross anterior
to the inferior vena cava. 12

As the kidneys ascend from the pelvis, they must pass be-
tween the umbilical arteries. The kidneys are closely opposed
and may come into contact with each other as they ascend be-
tween the umbilical arteries. If they come into contact, their
lower poles may fuse, resulting in a horseshoe kidney, which is
found in 1 in 600 persons. Similarly, one or the other kidney
may fail to ascend, resulting in a pelvic kidney. Usually, these
ectopic kidneys are located in the pelvis close to the common
iliac artery. 1 Multiple renal arteries often supply horseshoe
and pelvic kidneys, commonly arising from the aorta near the
aortic bifurcation or from the common iliac arteries.

The Arc ofBuhler is represented in intrauterine life as a longi-
tudinal anastomosis that connects the 10th through 13th
ventral segmental arteries. The 10th ventral segmental artery
contributes to the formation of the celiac artery; the 11th and
the 12th segmental arteries regress; and the 13th ventral seg-
mental artery contributes to the development of the superior
mesenteric artery. Normally, this longitudinal communication
regresses by the eighth week of embryonic life; however, if it
persists, the Arc of Buhler forms a communication between the
celiac and superior mesenteric arteries. Discovered in 2% of
autopsy cases and usually found in the location of the pancre-
aticoduodenal arteries, the Arc of Buhler may undergo
aneurysmal degeneration and rupture, probably related to in-
herent weakness in the persistent embryonic artery 13 (Fig.

Celiac artery -\ ra
Arc of Buhler <■

Aneurysm - y -

Superior -

mesenteric

artery

Figure 1.17 Persistent arc of Buhler with associated aneurysm.

1.17). If an aneurysm of this artery is identified, recommenda-
tions pertinent to other visceral artery aneurysms should be
followed.

Persistent sciatic artery

Persistent sciatic artery is a congenital anomalous continuation
of the internal iliac artery, which in 63% of these cases serves as
the major blood supply to the lower extremity 3 If the sciatic
artery is the major artery of the lower extremity, the superficial
femoral artery is hypoplastic or absent. Following the course
of the inferior gluteal artery, the sciatic artery passes with the
sciatic nerve through the greater sciatic foramen below the pir-
iformis muscle and enters the thigh (Fig. 1.18). 14 The artery
then courses along the posterior aspect of the adductor

pa rt I Vascular pathology and physiology

magnus muscle to the popliteal fossa, where it continues as
the popliteal artery Early atheromatous degeneration and
aneurysm formation are common. Due to its proximity to the
sciatic nerve, a sciatic artery aneurysm may present as a
painful buttock mass or with sciatic pain. Sciatic artery
aneurysms are bilateral in 12% of the cases. Palpable popliteal
and pedal pulses without palpable femoral pulses are clinical
findings highly suggestive of persistent sciatic artery. Mag-
netic resonance imaging (MRI) and arteriography provide a
definitive diagnosis. Proximal and distal ligation of the
aneurysm and femoropopliteal bypass graft 3 is the preferred
treatment.

Venous anomalies

Anomalies of the superior vena cava

Anomalies of the superior vena cava of importance to the vas-
cular surgeon include left superior vena cava and double
superior vena cava.

Persistence of the left anterior cardinal vein and obliteration
of the right common cardinal and proximal right anterior car-
dinal veins after the eighth week of gestation results in a left-

Superior

gluteal

artery

Inferior. _

gluteal

artery

Sciatic --
nerve

Aneurysm "

Sciatic artery -

Deep femoral
artery

sided superior vena cava (Fig. 1.19). 10 Blood from the right upper
extremity and right side of the head drains into the brachio-
cephalic vein and then into the left superior vena cava, which
courses anterolateral to the aortic arch and anterior to the
hilum of the left lung. 1 The left-sided superior vena cava then
drains into the coronary sinus.

Persistence of the left anterior cardinal vein and failure of
the left brachiocephalic vein to form results in double superior
vena cava (Fig. 1.20). The left superior vena cava drains into the
coronary sinus as previously described.

Right brachiocephalic vein

Left superior -
vena cava

MAYO

■ Pulmonary
veins

- Coronary
sinus

Inferior
vena cava

Figure 1.19 Left superior vena cava draining into coronary sinus.

Left superior
vena cava

I^RiQht superior
vena cava

Pulmonary

vems

Coronary
"sinus

Inferior
' vena cava

MAYO

Figure 1.18 Persistent sciatic artery and sciatic artery aneurysm.

Figure 1.20 Double superiorvena cava.

chapter 1 Embryology and development of the vascular system

Anomalies of the inferior vena cava

Embryologic abnormalities of the inferior vena cava and renal
veins pose potentially difficult problems for the vascular sur-
geon during abdominal aortic surgery Important anomalies
of the inferior vena cava include double inferior vena cava and
left inferior vena cava.

Double inferior vena cava results when the left sacrocardinal
vein fails to lose its communication with the left subcardinal
vein. With this anomaly, the left iliac vein may or may not be
present but the left gonadal vein is found in its normal loca-
tion 1 (Fig. 1.21).

Left inferior vena cava results from regression of the right
sacrocardinal vein, the normal precursor of the lower in-
frarenal inferior vena cava, and persistence of the left sacrocar-
dinal vein, which maintains its communication with the left
subcardinal vein 1 (Fig. 1.22).

If the right subcardinal vein fails to make communication
with the liver, absence of the suprarenal inferior vena cava results.
Blood from the caudal part of the body is shunted directly into
the right supracardinal (azygous) vein (Fig. 1.23). The hepatic
veins enter the right atrium at the site normally occupied by
the inferior vena cava. 15

Renal vein anomalies

Important renal vein anomalies include a circumaortic renal

collar and a posterior (retroaortic) left renal vein. In utero,
communications between the subcardinal and supracardinal
veins form a venous ring around the aorta at the level of the
renal veins. Failure of the dorsal portion of the ring to regress
results in either a posterior renal vein if the ventral portion of

, -Aorta

Right renal vein

Left renal vein

Aneurysm -

-Left inferior
vena cava

MAYO
©1992

Figure 1.22 Left inferior vena cava.

Left IVC

Superior vena cava

Hepatic segment

Azygos vein ^ m

Renal segment -

Sacrocardinal -
segment

MAYO

s " Hepatic veins

Figure 1.21 Double inferiorvena cava.

Figure 1.23 Absent inferiorvena cava. Suprarenal inferiorvena cava drains
into axygos vein.

pa rt I Vascular pathology and physiology

the ring regresses, or a circumaortic venous collar if the ventral
portion persists (Fig. 1.24).

Brener and colleagues 16 reviewed venous anomalies found
during abdominal aortic reconstructions at the Massachusetts
General Hospital between 1959 and 1973. During that period,
31 anomalies of the inferior vena cava or renal veins were
found and 11 of these resulted in complications. The most com-
mon venous anomaly was posterior left renal vein, followed
by duplication of the inferior vena cava. In their review of the
literature, the most frequent major venous anomaly was the
circumaortic renal collar (1.5% to 8.7%) (Table 1.1). Of the
above anomalies, the circumaortic renal collar and the posteri-
or left renal vein pose the greatest threat since the posterior
veins may be easily injured during dissection prior to
placement of an aortic cross clamp. Meticulous attention to de-
tail during dissection of the infrarenal aorta and common iliac
arteries is essential to avoid potentially disastrous hemor-
rhage from anomalous veins.

A rare, congenital venous anomaly is an aneurysm of the in-
ferior vena cava. In Sweeny et al.'s review, 17 only three cases
had been reported before 1990: two patients had aneurysms of
the supradiaphragmatic inferior vena cava and one patient

Figure 1.24 Circumaortic renal collar.

Table 1.1 Incidence of major inferior vena caval and renal vein anomalies

Venous anomaly

Incidence percentage

Circumaortic renal collar
Double inferior vena cava
Posterior left renal vein
Left inferior vena cava

1.5-8.7
2.2-3.0
1.8-2.4
0.2-0.5

had an aneurysm of the retrohepatic vena cava. The patient
Sweeny et ah reported presented with thrombosis of an in-
frarenal vena cava aneurysm following strenuous exercise.

Arteriovenous malformations

Congenital arteriovenous malformations (AVMs) result from
anomalous development of the primitive vascular system. 18
AVMs are usually present at birth although signs and symp-
toms may not be manifest until later in life. 19 Associated with
many different syndromes, AVMs have multiple clinical pre-
sentations (Table 1.2). Progression is usually the result of
hemodynamic factors because tumor-like behavior with en-
dothelial proliferations is not characteristic. 20

In AVMs, the pathologic vasculature is mixed arteriove-
nous. The amount of blood shunted through the abnormal
vessels and the resultant hemodynamic factors determine the
secondary morphologic changes in the feeding arteries and
draining veins. 19

Although multiple classifications have been suggested, the
accepted classification by Szilagyi and coworkers 18 ' 21 is based
on the developmental stages of the vascular system. As previ-
ously noted, the developmental stages of the vascular system
are the capillary network phase, the retiform stage, and the
gross differentiation phase. Hemangiomas result from devel-
opmental abnormalities in the capillary network stage, while
congenital arteriovenous fistulas result from arrest in devel-
opment in the retiform stage. Arteriovenous fistulas have been
further subdivided into microfistulous or macrofistulous
AVMs, depending on the size of the abnormal communicating
vessel and whether or not angiography can demonstrate the
site of the arteriovenous connections (Fig. 1.25). According
to Mulliken and Glowacki, 20 the term hemangioma applies
to those lesions that clinically undergo growth and usually
resolution with endothelial hyperplasia present during the
proliferative phase. In the proliferative phase, hemangiomas
incorporate [ 3 H] thymidine and have an increased mast cell
count. 22 The term vascular malformation (such as arteriovenous,
venous or lymphatic malformations, and port wine stains) ap-
plies to clinically and cellularly adynamic lesions. Seventy
percent of congenital AVMs, however, include not only
microfistulous or macrofistulous communications but also
include hemangiomatous lesions. 21

Congenital AVMs may be located anywhere on the body;
however, lesions involving the upper extremity are most fre-
quent, followed by lesions of the head and neck. AVMs of the
head and neck are classified as intraaxial if arising from arter-
ies supplying brain tissue (carotid artery or vertebral arteries)
or extraaxial if arising from arteries supplying dura, bone, or
muscle. 23 Other locations of AVMs include the lower extrem-
ity, the pelvis, and the viscera (lung, gastrointestinal tract, kid-
neys, and liver).

Schwartz and colleagues 24 reviewed 185 patients at the
Mayo Clinic with AVMs of the extremities and pelvis. Lesions

chapter 1 Embryology and development of the vascular system

E
o

M—

_fO

u
oo

"to

c
o

CD
-i— '
ro

~fO
U

U
CM

■

-i^

+-<

'oo

'cn

"+■'
J/J

to
>

"s_

-=?

oo
-I— 1

u
'cl

-i— 1

c
o

"+■»
ro
u
O

£
O

3
u

vn ,3

ro ro

- £

o O

a; ^

a ro

^ E

E
o

:_
T3
C
>»

oo .^J

_CD
CD

Q.
CD
CD
Q

O
O
Q.

CD
>

JO
CD

ro
N

.^ CD

CD -^

CD
Q_

£ °

O +-"

"-P CD

ro 15

£ ■£

M—

±: O

00
00

O
to

CD
>

CD
-1— '

CD
-1—'

_ro

~fO
CD
00

JZ
Q.

'q.

ro
-1— >
00

O
Q_

to
Q_

o
en

"TO
=3
00

-1— '

'00

'-1— '

_ro

to
CD
00

t ^

O Q-

R- 'q-

Q- CD

O
TD

C?>

Q.
ro

CD
Q.

00
CD

CD
5

00
I

CD
>

ro
00
CD

CL L -
J=" QJ

.E ro

"a -a

CD Qj

CD =

_a P

O u

O CD
O _Q

00
CD

O
'cn

CD
-i—'
tO
-1— '
U

O
00

-t—>

-1— '

"to

ro
£

-I— 1

" — ""

M—

'cn

-1— '

'cn

00
00

'00

!—

ro"

'-P>

-1— >

M—
O

H — '

CD
Q.

00
'o.

CD
CD

iyt

LJJ

^ ro

'00

CD
-I— 1
00

O
>

ro
-1— '
00
CD

s y.

cn ^]

CD .y

£ CT

o o

■to 2

CL Z3

CD CD

"ro

ro
_Q

—

"ro

'-I— '

CD
-1— '

C
'O

c
3

'cn

JO
.CD

Q- CD
CD —

CD
"D

u
ro

O .E

-i—'

■I— 1

'cn

-1— 1

1—

—

'00

00
JO

Z3
CD

CD
U
O

ro
'cn

a.
£

c
o

o
u

00
CD

CD
"D

_fO

Z3
U

ro
"D

ro
-1— •
CD

"ro

1 1

Q.
QJ

00
O
u

ro
CD

"ro

—

ro
-1— >
00

_Q
CD

CD
U

ro .^

o
u

to
u

br - J u
CD ro -S

t cn u

'c
to

-1— '

to
2

-1— 1

M—

CD
J>

M—

"+- '
u

c
to

JZ
-1—'

CD
Q.

C
O

CD
CL

M—

c5^

'00

'j_

_CD

00
-I— 1

c
o

'00

CD
U

00
-1—'
00
>^

CD
"O

00
CD

CL J^.

o
'cn

c
to

"to

M—
00

■1— '
to

M—
O

to
O
u

'00
00

'00

00
-1 — •

M—

_CD

Cl
O

cn
to

■1— 1
u

'00

H — '

-1— '

_CD
_Q

?= O

00
00
CD

Cl O

O
u

00
CD

CD
JZ>
JZ!

ro .^ — ~=i -^

to
£

TD
QJ
QJ

u
O

.£5 ro

CD~

cn
to

cd ro
> c=

to
u

"to

O
00

'cn

c
to

to
u

cn
to

O
"to

o
'cn

c
to

'cn

O
-1—'
to

JZ
CD

^-^

_f0

JZ
CD

' 1

~CD

'00
to
-i—'
u

_o
to

'00

O
-1— •

'cn

c
to

CD
Cl
CL

JZ
CD

CD
U

Z3
U

'cn

CD
JZ
JZ

Z)
"D
C
CD
Cd

'cn

c
ro

Z3
-I— 1

c
to

00
Z3
>

u
to

c
to

Cl
00

00
to

V 1

CD
"D

00
CD

QJ
QJ

_Z3

~a3

-t-^"

!—

CD
U

-1— >

to"

CD
>

C
Z3

CD
5

00
CD

CU
u
00

iy>

_ra

TD^

o
cn

>
£

-1—'
X

CD
-1—'

CD
>

'cn

"ro

-1—'

-1— >

'00

_iy ro

■=r ro

pa rt I Vascular pathology and physiology

Figure 1.25 (A) Capillary hemangiomas. (B)MicrofistulousAVM. (C)
MacrofistulousAVM.

were first noted at a median age of 1.9 years with the median
age at onset of symptoms 11 years. Presenting signs and symp-
toms included skin discoloration (43%), pain (37%), a palpable
mass (35%), and limb hypertrophy (34%). On physical exami-
nation, the most frequent abnormality was a capillary heman-
gioma (34%). An audible bruit was present in 26% of patients,
while ulceration and skin necrosis were found in 20% of
patients.

The etiology of soft tissue and bone hypertrophy in associa-
tion with congenital AVMs is not well understood. Hypothe-
ses include increased arterial flow in the area of the epiphyseal
plates, venous stasis, a tissue growth factor, and an anomaly in
the development of mesenchymal tissue.

The diagnosis of a congenital AVM frequently can be made
by history and physical examination. Noninvasive studies in-
cluding sequential limb systolic measurements, pulse volume
recording, and Doppler examination may also be useful. 25
With angiography, the size of the feeding arteries and the size
of the shunts can be estimated based on the time of appearance
of contrast medium in the veins. 26 Complemented by com-
puted tomography or MRI, angiography should be performed
before a treatment plan is formulated.

Contrast-enhanced computed tomography scanning delin-
eates congenital AVMs from surrounding tissue and, because
of its easy availability, is an important diagnostic test. 19 MRI
has become the main technique for diagnosis and follow-up of
congenital AVMs. It defines the relationship of AVMs to mus-
cle groups, fascial planes, nerves, tendons, and bones without
radiation and without contrast. 26 It is especially helpful to
evaluate children.

Asymptomatic or minimally symptomatic AVMs require
observation only since any intervention may stimulate
growth. Large lesions producing significant disfigurement or
overgrowth of an extremity should be treated, as should AVMs
with complications such as ulcers, bleeding, infection, tissue
necrosis, or congestive heart failure. 19

Nonsurgical treatment modalities include elastic compres-
sion, laser treatment (argon, carbon diozide, and neodymium :
yttrium aluminum garnet) and sclerotherapy with 3% sodium
tetradecyl sulfate. Treatment with laser or with sclerotherapy
may be of benefit with smaller, low shunt lesions. 26 Emboliza-
tion may be used alone or rarely in combination with surgery
to decrease shunting at the precapillary or capillary level. Em-
bolization materials may be temporary, such as blood clot,
gelatin sponge, or microfibrillar collagen, or permanent, such
as silicon spheres, polyvinyl alcohol particles, stainless steel
coils, or detachable balloons. 26

In general, a conservative attitude toward surgical resection
of AVMs is warranted. In a series of 80 patients with congenital
AVMs of the extremities reported by Gomes and Bernatz, 27
surgical resection was attempted in only 10 patients. If surgery
is indicated, complete extirpation of the AVM in one stage with
or without embolization should be attempted. Curative resec-
tion can be performed in only about 20% of all AVMs. 28 In
Schwartz et al.'s 2A retrospective review from our institution, 18
of 82 patients in the surgical group required amputation of the
extremity at various levels.

Visceral congenital AVMs may be found in the gastrointesti-
nal tract, kidney, spleen, liver, or lung. Congenital gastroin-
testinal AVMs are found primarily in the upper portion of the
small bowel in younger patients. If bleeding occurs and per-
sists after correction of coagulation abnormalities, emboliza-
tion, endoscopic treatment, or surgical excision are options.
Because of the risk of bowel necrosis following embolization,
endoscopic therapy is becoming more popular. Renal AVMs,
usually located beneath the mucosa of the renal collecting
system, should be embolized or surgically removed if
symptomatic.

Hepatic, splenic, and pulmonary AVMs may be associated
with hereditary hemorrhagic telangiectasia (Rendu-
Osler- Weber syndrome). 29 Hepatic lesions may present
with jaundice, hepatomegaly, or cardiac failure. Pulmonary
AVMs may cause dyspnea, hemoptysis, or palpitations with
60% of patients having a bruit. 30 Hepatic, splenic, or
pulmonary AVMs may be managed with either surgery or
embolization.

chapter 1 Embryology and development of the vascular system

BBdBV •*'**, i 1^ .^^BBBfl

B j "" * ^ ,i| Kr H

H ff

K- u ft 1

bH rsH
bH>< W ObbbI

BBB* ; 11 ' BBbI

■ 1 bbb!

BBBft ll ♦ ^XBbI

^B 1! bH

■l ■ I

1 BJ

1 . ■

bbbS:

[i • 4 1

bbbBwI

F- . ■ ( J

1 bbL ' J 3? bbI

LJ i M

Lr 1 1

I dBv

1 F | Va

1* V

^BBM <B

■■ F JM BUyfl

Figure 1.26 (A) Eighteen-year-old man with KTS involving the right lower extremity. (B)MRI of the extremity.

Klippel-Trenaunay syndrome, a rare congenital malforma-
tion, is one of the more common syndromes with associated
AVMs. At the Mayo Clinic, we have observed 144 patients with
this syndrome. 31 Characteristic findings included heman-
gioma in 137 patients (95.1%); varicose veins in 110 (76.4%);
and hypertrophy of the soft tissues or bones in 134 (93.1%)
(Fig. 1.26). Only one lower extremity was involved in 71.5% of
patients. Atresia or hypoplasia of the deep veins may be pre-
sent (Fig. 1.27). Most patients did well with observation or
with elastic compression only. Surgical treatment was under-
taken in nine patients with lower extremity vascular malfor-
mations. Of seven patients who underwent resection of
varicose veins or hemangiomas, none was cured but six im-
proved. Two patients became worse after resection of varicose
veins at another institution. One patient underwent deep ve-
nous reconstruction for atresia of the superficial femoral veins
using contralateral saphenous vein. A patent graft with
competent valves was noted at follow-up 6 months after the
operation. Although patients with severe chronic venous in-
sufficiency, with complications from hemangioma, or with
cosmetic disfigurement may benefit from surgery, preopera-
tive imaging of the extremity with MRI and contrast veno-
graphy is important to prevent complications. Rarely,
reconstruction for atresia or hypoplasia of the deep veins may
be needed.

Figure 1.27 Venogram demonstrating agenesis of the iliofemoral vein and
large suprapubic venous collaterals (arrow).

pa rt I Vascular pathology and physiology

References

1. Sadler TW. hangman's Medical Embryology. 6th ed. Baltimore:
Williams & Wilkins, 1990:179.

2. Woollard HH. The development of the principal arterial stems in
the forelimb of the pig. Cont Embryol 1922; 14:139.

3. Martin KW, Hyde GL, McCready RA, Hill DA. Sciatic artery
aneurysms: report of three cases and review of the literature. / Vase
Surg 1986; 4:365.

4. Lewis FT. The development of the lymphatic system in rabbits.
Am JAnat 1921)5:95.

5. Sabin FR. On the origin of the lymphatic system from the veins and
the development of lymph hearts and thoracic duct in the pig. Am
} Anat 1902; 1:367.

6. Huntington GS. The anatomy and development of the jugular
lymph sacs in the domestic cat. Am} Anat 1910; 10:177.

7. Dean RH, Turner CS, Hansen KJ. Aortic lesions in children. In:
Bergan JJ, Yao JST, eds. Aortic Surgery. Philadelphia: WB Saunders,
1989:441.

8. DeBakey MF, Garrett E, Howell JF, Morris GC. Coarctation of the
abdominal aorta with renal artery stenosis: surgical considera-
tions. Ann Surg 1967; 165:830.

9. Hallett JW Jr, Brewster DC, Darling RC, O'Hara PJ. Coarctation of
the abdominal aorta: current options in surgical management.
Ann Surg 1980; 191:430.

10. Wind CG, Valentine RJ. Anatomic variation of the blood vessels.
In: Anatomic Exposures in Vascular Surgery. Baltimore: Williams &
Wilkins, 1992:445.

11. Michels NA. Blood Supply and Anatomy of the Upper Abdominal
Organs with a Descriptive Atlas. Philadelphia: JB Lippincott,
1955:152.

12. Kabalin JN. Anatomy of the retroperitoneum and kidney.
In: Walsh PC, Retic AB, Stamey TA, Vaughan ED Jr, eds.
Campbell's Urology. 6th ed. Philadelphia: WB Saunders, 1992:
30.

13. Nicholson CP, Cherry KJ, Frazee RC, Stanson AQ. Aneurysm of
the Arc of Buhler: a case report. Unpublished data.

14. Mandell VS, Jaques PF, Dekeny DJ, Oberheu V Persistent sciatic
artery. Ann Surg 1984; 199:69.

15. Anderson RC, Herlig W, Novick R, Jarvic C. Anomalous
inferior vena cava with azygous drainage. Am Heart } 1955: 49:
318.

16. Brener BJ, Darling RC, Frederick PL, Linton RR. Major venous

anomalies complicating abdominal aortic surgery. Arch Surg 1974;
108:159.

17. Sweeny JP, Turner K, Harris KA. Aneurysms of the inferior vena
cava. / Vase Surg 1990; 12:25.

18. Szilagyi DE, Smith RF, Elliott JP, Hageman JH. Congenital arteri-
ovenous anomalies of the limbs. Arch Surg 1976; 111:423.

19. Gloviczki P, Hollier LH. Arteriovenous fistulas. In: Haimovici H,
Callow AD, DePalma RG, Ernst GB, Hollier LH, eds. Vascular
Surgery: Principles and Techniques. 3rd ed. Norwalk, CT: Appleton
&Lange, 1989:698.

20. Mulliken JB, Glowacki J. Hemangiomas and vascular malforma-
tions in infants and children: a classification based on endothelial
characteristics. Plast Reconstr Surg 1982; 69:412.

21. Szilagyi DE, Elliott JP, DeRusso FJ, Smith RF. Peripheral congeni-
tal arteriovenous fistulas. Surgery 1965; 57:61.

22. Glowacki J, Mulliken JB. Mast cells in hemangiomas and vascular
malformations. Surgery 1982; 92:348.

23. Forbes G, Earnest F IV, Jackson IT, Marsh WR, Jack CR, Cross SA.
Therapeutic embolization angiography for extra-axial lesions in
the head. Mayo Clin Proc 1986; 61:427.

24. Schwartz RS, Osmundson PJ, Hollier LH. Treatment and progno-
sis in congenital arteriovenous malformation of the extremity.
Phlebologie 1986)1:171.

25. Haimovici H, Sprayregen S. Congenital microarteriovenous
shunts: angiographic and Doppler ultrasonographic identifica-
tion. Arch Surg 1986; 121:1065.

26. Jackson IT, Forbes G, May GT. Vascular anomalies. In: Mustarde J,
Jackson IT, eds. Plastic Surgery in Infancy and Childhood. 3rd ed.
New York: Churchill Livingstone, 1988:691.

27. Gomes MMR, Bernatz PE. Arteriovenous fistulas: a review and
ten-year experience at the Mayo Clinic. Mayo Clin Proc 1970; 45:81.

28. Szilagyi DE. Vascular malformations (with special emphasis on
peripheral arteriovenous lesion). In: Moore W, ed. Vascular
Surgery: A Comprehensive Review. 2nd ed. New York: Grune &
Stratton, 1986:773.

29. Burckhardt D, Stalder GA, Ludin H, Bianch L. Hyperdynamic
circulatory state due to Oster-Weber-Rendu disease with intra-
hepatic arteriovenous fistulas. Am Heart J 1973; 85:797.

30. Klimberg I, Wilson J, Davis K, Finlayson B. Hemorrhage from con-
genital renal arteriovenous malformation in pregnancy. Urology
1984; 23:381.

31. Gloviczki P, Stanson AW, Stickler GB, Davis K, Finlayson B.
Klippel-Trenaunay syndrome: the risks and benefits of vascular
interventions. Surgery 1990; 110:469.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Vascular wall physiology

Christian C. Haudenschild

The vasculature not only is the prime concern for the vascular
surgeon, it is also the crucial determinant for success or failure
of all general surgery. No matter how complex the current
knowledge about vascular wall pathophysiology has become,
the basic principles remain clear and simple. All physiologic
mechanisms involving the vascular wall have only three
purposes— to keep the lumen patent, adequate in size, and
without leakage. These three demands for appropriate tissue
support are so fundamental that nature has developed an al-
most infinite number of cellular and humoral mechanisms, en-
zymatic cascades, and interacting biochemical and molecular
loops with multiple backups, reserves, and emergency func-
tions to guarantee patency, flow control, and hemostasis in al-
most every situation except one— surgery. This is where the
physician assumes temporary partial or total control with
mechanical and pharmacologic tools; these tools are sophis-
ticated but cannot approach the finesse of physiologic
autoregulation of the vascular lumen through mechanisms
residing in the cells of the vascular wall and in the interacting
circulating cells.

The problem is that the mechanisms controlling hemostasis
are exactly opposed to those maintaining patency, and that the
margin of error on either side is small in most surgical patients.
Naturally, in the mammalian closed circulation, the hemo-
static and repair mechanisms prevail over those maintaining
patency, especially after injury. There are more clotting factors
than fibrinolytic ones, more natural vasoconstrictors than
dilators, more growth factors than inhibitors, and more
known stimulators than blockers. This may be related to the
fact that the stimulators are easier to study than the inhibitors,
but in the vasculature, mechanisms to stop bleeding are so
dominant that it seems that nature attempts to maintain the
integrity of the closed vascular circulation almost at any price.

For example, contracting vascular cells can act only to
narrow the lumen, but dilation is entirely passive— that is,
through stretching of relaxed vascular wall cells by hemostat-
ic and hemodynamic forces. Whereas every flexor in the skele-
tal muscle system has its opposite active extensor, and the
smooth musculature of the gastrointestinal tract can use

multiple layers in different directions and peristaltic coordina-
tion for active control of the lumen, the huge, contractile,
smooth muscle cell apparatus of the vasculature has no active
muscular antagonist. While the quiescent state of vascular
cells in terms of growth is maintained through a number of
subtle mechanisms acting in concert at the levels of cellular
and nuclear membranes and the extracellular matrix, the cel-
lular migration and proliferation response to vascular injury is
governed by the rapid release of a few, powerful growth
factors that are readily available in both residing and circulat-
ing cells, accompanied by the expression of their respective
receptors.

This prevalence of hemostatic and repair functions is of
course helpful during surgery itself, and probably makes pos-
sible the infliction of a deliberate wound for the purpose of
healing and repair. In the postoperative patient, however,
where the surgeon has to take over nature's controls and bal-
ances temporarily, and where hemostasis is ensured by appro-
priate surgical ligation and coagulation techniques,
overreactive hemostatic mechanisms easily can cause life- or
tissue-threatening thromboembolic events, and excessive re-
pair can cause adhesions, hyperplastic scars, and intimal
hyperplasia at vascular anastomoses.

Knowledge of these pathophysiologic mechanisms— their
prevalence, timing, and relative balance— and the ability to
exploit them practically when they work in the patient's favor
are therefore of great advantage for the management of the
surgical patient. In the past two decades, a number of discov-
eries in the field of vascular pathophysiology have enriched
this knowledge extensively, and a few have made it into prac-
tical clinical application. The downside of this abundance of
detailed information is that it is almost impossible to keep
up with the newest developments, even for the basic re-
searcher, and certainly for the busy clinician. Evaluating
the impact of the information is further complicated by the
hunger for publicity on the part of the researchers, who, almost
by definition, think that their discovery is the key and solution
to every known problem, and by the flood of noneditorial
literature that tends to emphasize the virtues of the mecha-

pa rt I Vascular pathology and physiology

nisms that can be influenced by commercially available
agents.

In dealing with the confusing abundance of new factors,
mediators, and mechanisms, it is reassuring to realize that
there are major and minor regulators, and that the major regu-
lators tend to be the best known, because they have more gen-
eral effects and therefore were discovered earlier, whereas
many of the more recently described mechanisms are of pre-
dominantly local importance and are involved in the fine-
tuning of vascular functions. It is also true that whereas many
of the new mediators of vascular reactivity have been fully
characterized in vitro, their relative importance and some-
times even their presence or activity in the reactive, wounded,
or stimulated vascular wall of the living organism have not yet
been proven.

Endothelium

The traditional understanding of the vascular endothelium,
and its most general definition, is that of a monolayer of cells
lining the luminal side of the entire cardiovascular system. In
spite of the discovery of new and exciting endothelial func-
tions, this unique, strategic position at the blood-tissue inter-
face still constitutes the basis of almost everything that is
special about these cells. The second fundamental fact about
the endothelial cell layer is its heterogeneity: capillary en-
dothelial cells differ from those lining the large vessels,
arterial endothelium is different from venous endothelium,
and special vascular beds, such as the brain with its especially
tight blood-brain barrier, or the organs with sinusoidal vascu-
lature that facilitates the exchange of cells and fluids, all have
their own, special types of endothelia that differ both in their
morphology and their functions.

The modern understanding of the endothelial cell, incorpo-
rating most of the new findings on endothelial cell function
and dysfunction, is that of a controlling or regulating cell. The
regulation is mostly over a short range, on a local basis (which
allows the cells to function differently in different locations),
and it is exercised through the cell's strategic position, its con-
nections with adjacent vascular cells, its membrane properties
in terms of passive surface and expression of receptors and
adhesion molecules, its capability of active and directional
resorption and secretion, and its synthesis and controlled re-
lease of powerful, but short-ranging vasoactive agents such as
prostaglandins and cytokines. 1 Using these capabilities in
combination, the endothelium exercises mostly local control
over (in order of importance in conditions of injury): hemosta-
sis, vascular tone, vascular cell growth, and vascular perme-
ability. For a long time, endothelial cells have been considered
"good" or "favorable" for vascular patency as well as for vas-
cular quiescence; the endothelial functions discovered first
were anticoagulant and vasodilatory ones, and the removal of
endothelium, an obvious pathologic situation, triggers a

Figure 2.1 Scanning electron micrograph of the site of a fresh vascular
suture. Endothelial discontinuity and cellular compression are obvious, but
subtle functional changes are not readily visible (x1 00).

number of undesirable responses such as platelet adhesion
and aggregation, sometimes followed by excessive smooth
muscle cell growth. 2

It has become clear, however, that for every known endothe-
lial function there is at least one opposite mechanism, usually
located in the same cell, but often controlled by different sig-
naling pathways, and expressed with different timing. This
has led to the somewhat inappropriate term of dysfunctional
endothelium for cells that express a combination of physiologic
functions that we happen not to like in a given condition. 3 The
term still is practical, since it has replaced the formerly preva-
lent idea of "absent" or "denuded" endothelium as the cause
of all vascular evil; it probably can be defined best as the
temporary lack of balance between promoting and inhibiting
activities with respect to paired mechanisms such as hemosta-
sis-thrombosis, vasodilation-contraction, cellular growth-
differentiation, or secretion-resorption (Figs. 2.1 and 2.2).

Control of hemostasis and thrombosis

The primary mechanism of endothelial coagulation control is
that of a physical barrier that covers the highly thrombogenic
subendothelial components such as von Willebrand factor,
basement membrance, fibrillar collagen (types IV and III), and
other extracellular matrix constituents that, if exposed, signal
the presence of a real injury. The second mechanism is an en-
dothelial surface composed of proteoglycans that prohibit the
adhesion of platelets in a wide range of normal flow condi-
tions. Platelets, however, can stick to dysfunctional endotheli-
um— for example, to the incomplete endothelial cover of
already existing and advanced atherosclerotic plaques —
although most of the time some platelet pseudopods are seen
extending between endothelial cells to the subendothelium.

CHAPTER2 Vascular wall physiology

Figure 2.2 After vascular balloon injury, abundant platelets adhere to the
subendothelial layer but not to the remaining endothelial cells at the edge of
the wound. As these endothelial cells respond with migration and
proliferation, however, they change their phenotype and become, at least
temporarily, dysfunctional during the process of reendothelialization
(scanning electron micrograph, x300).

Secretion of prostacyclin, a powerful inhibitor of platelet ag-
gregation as well as a vasodilator, is the third endothelial anti-
coagulant mechanism. With regard to fine-tuning functions,
endothelial cell expression of thrombomodulin is notable be-
cause this molecule, interacting with protein C, can cause
thrombin to inactivate activated factor Va, resulting in a para-
doxical anticoagulant effect. True to the principle of biologic
balance, at least one major procoagulant factor (von
Willebrand factor or factor VIII) also is produced and
secreted by normal endothelium, but most of it remains
dormant in the subendothelial space. 4

The endothelium also exercises effective hemostasis-
thrombosis control when a clot already has formed; it secretes
plasminogen activator and the corresponding plasminogen
activator inhibitor. The timing and control of this fibrinolytic
cascade are only slightly less complicated than those of the
clotting cascade; its extensive study in the past few years has
led to one of the more important practical applications of
basic research, in the form of the direct use of recombinant
plasminogen activators to remove thrombotic coronary
obstructions. 5

Control of vascular tone

Overall control of vascular tone is exercised by the sym-
pathetic and parasympathetic nervous systems and the

rennin-angiotensin-aldosterone and related humoral sys-
tems. Endothelium contributes significantly to this control
through the activity of angiotensin-converting enzyme, which
converts a less active decapeptide into the active octapeptide
angiotensin II, a powerful vasoconstrictor and upregulator of
blood pressure. 6 In addition, this enzyme removes active
bradykinin. Systemic inhibition of this enzyme is a widely ac-
cepted treatment of hypertension. Although it is possible to
lower plasma levels of converting enzyme to below the thresh-
old of detection, the vascular wall tissue concentrations of this
enzyme can remain high, most likely because of the continued
synthetic and secretory activity of local endothelial cells. Inhi-
bition of this enzyme also has some antiproliferative effect in
injured arteries of a few experimental animal species, but large
clinical trials have not showed any improvement of the angio-
graphically defined restenosis rate in atherosclerotic human
coronary arteries after angioplasty.

Much research is being devoted to endothelial control of
vascular tone through endothelium-derived relaxing factors.
One of these powerful dilating factors is nitric oxide (NO),
which is derived from L-arginine through the enzyme nitric
oxide synthetase. Because not all experimental vasodilation
can be explained solely by the action of NO, additional factors
have been proposed, most notably an endothelium-derived
hyperpolarizing factor that acts in cooperation with NO,
opening potassium channels and closing voltage-dependent
calcium channels, and thus contributing to smooth muscle cell
relaxation. 7 The typical test for these factors is the exposure of
a previously contracted vessel to acetylcholine, which can be
done in vitro using a vessel ring carrying a weight, or in vivo
(even in patients) through a catheter. Functionally intact
endothelium produces these rapidly acting relaxing factors in
response to stimuli in a dose-dependent fashion; the vasodila-
tion force is measured by the attached weight in vitro, or by
perfusion pressure in vivo. Endothelial-derived relaxing
factors also inhibit certain platelet functions.

Prostacyclin is the other short-lived, potent vasodilator
locally produced by endothelium that has some platelet-
inhibiting action as well. Prostacyclin is one of many products
of arachidonic acid metabolism mediated by cyclooxygenases;
some of them, especially thromboxane A 2 , which is derived
mostly from platelets but is also produced by endothelial cells,
have an effect exactly opposite to that of prostacyclin.
Somewhat higher doses of the cyclooxygenase inhibitor,
aspirin, are needed for the suppression of prostacyclin produc-
tion than for the inhibition of thromboxane A 2 formation.

Keeping the controlling balance intact, endothelium
also produces the powerful vasoconstrictors endothelin-1,
prostaglandin H 2 , and some endoperoxides. Local overpro-
duction of such vasoconstrictors has been implicated in
vasospasm in irritated or injured vessels with dysfunctional
endothelium; practically speaking, such conditions may occur
near vascular anastomoses and may not be entirely control-
lable by sympathetic blockade.

pa rt I Vascular pathology and physiology

Growth control of and by endothelium

Most of our knowledge about vascular endothelium is derived
from work with endothelial cells in tissue culture. When
human umbilical cord vein endothelial cells first became avail-
able, followed by bovine aortic endothelium and tube-
forming capillary endothelial cells from a variety of species,
one common characteristic seen was the rigorous growth con-
trol of these cells in vitro. In vitro, endothelial cell growth
depends on the presence of both optimal growth factor combi-
nations, usually achieved with high serum concentrations,
and on the presence of favorable growth substrates, often
gelatin or fibronectin. With the formation of a confluent mono-
layer or a complete network of tubes, the growth effectively
is arrested despite addition of more growth factors. In
vivo, quiescent endothelium shows low rates of replication;
when stimulated, however, endothelium can replicate quick-
ly. Thus, unlike nerve cells, which virtually never grow in
adults, and unlike bowel epithelial cells, which almost always
grow, vascular endothelium shows a wide range of growth re-
sponses, governed by more or less specific growth factors and
their respective receptors.

Some of the most prominent endothelial growth factors be-
long to the rapidly growing family of heparin-binding fibro-
blast growth factors (FGFs), which are produced by almost all
mesenchymal cells, including the endothelial cells them-
selves. The naturally occurring prototypes acid FGF and basic
FGF lack a signal sequence necessary for secretion, but they
can be released readily by, for example, cellular heat shock and
other conditions found in wound, ischemic, or inflammatory
environments, including, apparently, cell death. Low-affinity
binding sites (probably heparan sulfate proteoglycans) and
high-affinity cell surface receptors (FGFR-1/flg, FGFR-2/bek
and others, including many isoforms) regulate access of the
growth factors to the target cells. Together with other intracell-
ular signaling mechanisms, growth factors inside the target
cell and respective receptors on the cell nucleus are responsi-
ble for the growth signal finally reaching the nuclear synthetic
and dividing components. 8 Many other growth factors, in-
cluding the more specific vascular endothelial growth factors,
function with similar receptor- or double-receptor-regulated
pathways, with the noticeable exception of platelet-derived
growth factor, for which large-vessel endothelium lacks the
appropriate receptors.

Endothelial cell growth control is treated here in greater de-
tail for two reasons. First, in vascular surgery as well as in
transplanted vessels and grafts made from biomaterials, rapid
covering with viable endothelium is clearly desirable, as long
as the lining rapidly turns into functional rather than dysfunc-
tional endothelium, with anticoagulant, vasodilating, and
growth-inhibiting properties prevailing. The availability of
specific endothelial growth factors and of genetically
engineered cells that continuously produce and secrete such
factors has revitalized interest in seeding biomaterials with

endothelial cells, before their use as vascular grafts. 9 Progress
has been made in the knowledge of cellular adhesion mole-
cules, and further advances were made possible with the real-
ization that both desired cells (endothelium) and others
(platelets and leukocytes) adhere to a biomaterial-modified
film of proteins rather than to the biomaterials themselves.

The second reason for expanding on growth factors in the
context of vascular endothelium is that the development of the
early vasculature, the reactivation of vascular growth in
wound healing 10 and inflammation, 11 and the vascular sup-
port of some malignant tumors, 12 all are under the control of
vascular growth factors. Angiogenesis usually is defined as
the sprouting of new vessels from existing ones, whereas vas-
culogenesis usually is understood as the assembly of new
tubes from dispersed individual cells, as often is observed in
embryonal development. Both processes can be initiated as
well as supported by the action of growth factors, which
form directional concentration gradients, or are sometimes re-
tained and concentrated in the extracellular matrix. The for-
mation of new vessels by either mechanism involves
phenotypic endothelial cell changes, cell migration, cell divi-
sion, cell attachment, synthesis of basement membrane com-
ponents, and endothelial cell redifferentiation into a
quiescent, functional state. This rather complex sequence of
events is triggered by growth factors, but is sustained and
completed with the help of many other cell-to-cell and cell-to-
matrix mediators, including those derived from circulation
and inflammatory cells.

In addition to being readily responsive to growth factors,
endothelial cells produce their own growth factors, as well as
some growth inhibitors, by which they assume control over
the migration, growth, and phenotype of their associated
vascular smooth muscle cells. Failure of this control is thought
to be pathogenetic in atherogenesis 13 and in the development
of intimal hyperplasia 14 at anastomosis sites and after
angioplasty. Endothelial cell-derived growth promoters
include molecules analogous to basic fibroblast growth
factor, platelet-derived growth factor, and possibly endothe-
lin, whereas heparin-like molecules and transforming
growth factor-(3 1 represent typical, endothelium-derived,
growth-inhibiting agents.

Endothelial control of vascular permeability

Like most other mesenchymal cells, endothelial cells can syn-
thesize and secrete a variety of molecules, especially extracell-
ular matrix components such as laminin, collagen type IV,
and others. Furthermore, they can pass molecules by a variety
of pathways from the vascular lumen into the wall and sur-
rounding tissue, and vice versa. In specialized endothelia
in various capillary exchange regions, there are passages
through the endothelium, between endothelial cells, along en-
dothelial channels, by ways of active vesicular transport, and,
for lipids, along the endothelial cell membrane from one side

CHAPTER2 Vascular wall physiology

to the other. Most of these transport mechanisms are active,
selective, and capable of modification of the transported
molecules. Most notable is the possible modification of
lipoproteins such as low-density lipoprotein, which develops
its highest atherogenic potential when it is modified into a
mildly oxidized low-density lipoprotein. By mechanisms of
membrane incorporation, endothelial cells also act as antigen-
presenting cells, assuming a role similar to that of monocyte-
macrophages; this function is thought to be important in graft
rejection and also may play a role in immune-mediated
intimal thickening of coronary arteries in heart transplants.

in this book. Likewise, the lymphocytes have their own set of
activating molecules, including y-globulins and complements
in addition to the ones they share with other inflammatory
cells. In general, the IL are the initiators and are active in the
early phases of the interactions, whereas the various cellular
adhesion molecules are expressed secondarily and are more
involved in facilitating the transendothelial migration. Many
of these interaction sequences are based on in vitro observa-
tions of simplified and relatively well controlled culture
systems; the relative importance of any of these steps and
mediators in vivo often remain to be elucidated.

Interactions with other blood-borne cells

Because of its strategic position at the blood-tissue interface,
endothelium is the first cell layer that comes into contact with
white blood cells. There are specialized endothelia, especially
in the lymphatic system, that facilitate the exit of white cells
out of the blood circulation under normal conditions. For
larger conduit blood vessels such as arteries and veins,
adhesion and penetration of any blood cells are abnormal
events, although it has been speculated that platelets, passing
along but not sticking to normal endothelium, play a suppor-
tive role in maintaining the integrity of this cell layer.

Under pathologic conditions (i.e. inflammation injury, and
immune reactions), white cells interact with and eventually
penetrate endothelium within minutes. For each type of white
cell, there is a set of stimuli, adhesion molecules, and chemo-
tactic gradients that control, in sequence, the events of mar-
gination, adhesion, and vascular wall penetration. For the
neutrophils, for example, "rolling" over endothelium is medi-
ated by selectins (E and P selectin derived from endothelium
after cytokine stimulation, and L selectin expressed by the
neutrophils). Subsequent adhesion is mediated by integrins
(CDlla/CD18 and CDllb/CD18), which in turn interact with
intercellular adhesion molecule-1 to allow, first, the migration
of the white cell into the vessel wall, and then, along other
chemotactic gradients, into the tissue. Many of the cytokines
and inflammatory mediators are expressed by the injured en-
dothelial cells themselves, whereas others come from the
white cells or are stored in some inactive form. Some represen-
tative members of a number of cytokines too great to be dis-
cussed here in detail are tumor necrosis factor, the interleukins
IL-1, IL-6, and IL-8, and interferon-y, which mediate tissue
damage and expression of adhesion molecules, and also have
procoagulant properties. 15 Longer known inflammatory me-
diators such as histamine and reactive radicals also are in-
volved in these complex events.

Monocyte-macrophage adhesion and migration are
associated with their own set of mediators, 16 in particular
macrophage colony-stimulating factor and macrophage
chemotactic protein-1. The recruitment of monocyte-
macrophages into the arterial wall is a crucial event in the
pathogenesis of atherosclerosis, which is discussed elsewhere

Preservation of functional endothelium by gentle
surgical technique

Some amount of vascular trauma is unavoidable in any kind of
surgery; but, given the properties of endothelial cells de-
scribed earlier, it is clear that the preservation of endothelial
integrity should be a primary concern in general surgery, and
even more so in vascular surgery. It is not enough merely to
have a few scattered endothelial cells sticking around; the goal
is to leave large areas of minimally irritated endothelium as
near as possible to the locations where damage is unavoidable.
With the inherent rapid response of endothelium to growth
factors, endothelial regeneration will start from these areas
within 12 h of damage. More important, the smaller the area
that needs to be covered, and the lower the number of
endothelial cells that initially were rendered dysfunctional,
the faster the regenerated endothelium resumes its favorable,
fully functional state. Functionally intact endothelium can
greatly reduce the incidences of thrombosis, a major cause of
early occlusion, and of intimal hyperplasia and contracture,
major causes of late narrowing and occlusion.

In larger vessels, endothelium detaches more easily than in
smaller ones; on the other hand, a 1-mm layer of thrombotic
material or intimal hyperplasia is much more devastating in a
vessel of 3 mm total diameter than in a vessel that is two or
three times larger. Endothelium can be damaged without
being touched; clamping, stretching, and prolonged stasis of
the entire vessel are the most common causes of endothelial in-
jury, even in closed vessels that are handled roughly from the
outside. Sharp dissection rather than blunt undermining and
stretching should be used for the isolation of vascular seg-
ments, especially for those that are used for grafting and those
near anastomoses. Specially padded plastic clamping is avail-
able to replace metal hemostats for better control of the forces
needed to interrupt blood flow temporarily. Spasms can be
prevented with the local application of vasodilators (e.g.
0.012% papaverine in saline) to prevent shear forces that
invariably remove endothelium and damage other vascular
cells when spastic vessels are dilated or stretched by mechani-
cal force. Endothelial loss in graft vessels can be minimized by
reducing the extracorporeal time, avoiding overstretching as
well as collapse, and using organ culture conditions rather

pa rt I Vascular pathology and physiology

than plain saline during the extracorporeal interval, when
much damage can take place. 17 Perhaps most important are
the surgeon's knowledge and continual awareness of the
fragility of endothelium, which will lead intuitively to the
moves that preserve optimal endothelial integrity in any
situation.

Smooth muscle cells

For the 6-year period from 1966 to 1974, the Index Medicus lists
a grand total of six papers dealing with vascular smooth
muscle cells; for the same period 20 years later, there are 11 440
papers on this topic (and almost 32000 on endothelium). As
with endothelium, interest in vascular smooth muscle cells
rose steadily once these cells became available in tissue cul-
ture, and also once it was demonstrated that many cells in ath-
erosclerotic plaques were derived from smooth muscle cells.
Although the nature and lineage of atherosclerotic plaque
cells, especially of lipid-laden foam cells, remain controver-
sial, the cells forming the intimal hyperplasia at vascular graft
anastomosis sites, and those in accelerated intimal lesions in
arteries of transplanted organs, almost certainly are of smooth
muscle cell origin. Cardiovascular medications often affect
smooth muscle cells intentionally or as a side effect; a
large number of papers deal with these pharmacologic effects,
including many in the field of hypertension research and
treatment. 18

The greater interest in smooth muscle cells has been gener-
ated by the fact that they display several strikingly different
phenotypes. Most cells can change appearance or function
in different environments and under different stimuli, but
only in vascular smooth muscle cells (and maybe in mono-
cyte-macrophages) is the phenotypic change so readily visible
and of so much practical importance. This phenotypic varia-
tion has made the smooth muscle cell a major research tool for
the study of growth control and differentiation.

The quiescent phenotype

This appearance of smooth muscle cells in the tunica media of
normal vessels (Fig. 2.3) has been called the contractile pheno-
type, because electron micrographs of undisturbed, well
preserved smooth muscle cells display a cytoplasm almost en-
tirely filled with contractile filaments, predominantly actin,
with membranous anchorage points and a few other patches
of greater electron density, and only sparse additional
organelles. The cytoplasmic membranes of these cells appear
smooth, and they are surrounded by a continuous basement
membrane. Contraction often is reflected in folding of the nu-
clear and cytoplasmic membranes. The turnover rate of these
cells is low; if tested for cell replication by a variety of different
techniques, most of these cells are found to be in the G phase,
that is, out of the cell division cycle. The relatively uniform

Figure 2.3 Transmission electron micrograph of normal endothelium and
quiescent smooth muscle cells of a rabbit aorta. As in human arteries,
including the coronary arteries, smooth muscle cells can be present both in
the intimal and the medial layers. In this example, both the intimal and the
medial smooth muscle cells (separated by the internal elastic lamina [IEL])
show a morphologically identical, filament-rich phenotype (bar = 1 jam).

appearance of these quiescent cells in most conventional light
and electron microscopic preparations is deceiving, however;
in fact, there is great heterogeneity in the function and respon-
siveness of these cells. If challenged with a stimulus such as
mechanical stretching, or with the loss of an inhibiting influ-
ence such as the loss of endothelium, almost every one of these
apparently uniform cells will react differently: some will
divide, some will migrate, some will migrate and then
divide, some will merely change their phenotype, some will
just enlarge, and a few will do nothing at all. 19

The most responsive cells have been compared with stem
cells, and may be the origin of clonal growth in atherosclerotic
plaque formation. More cells are found at a higher level of re-
sponsiveness in the undiseased portions of arterial walls of
people with the most prevalent risk factors for atherosclerosis
(hyperlipidemia, hypertension, diabetes, smoking). Highly
responsive smooth muscle cells are more likely to migrate onto
tissue culture dishes, or to survive enzymatic cell separation
methods and thrive under tissue culture conditions. Because
of this selection bias and the questionable reversibility of
changed phenotypes under culture conditions, most smooth
muscle cell culture systems are more representative of altered
phenotypes than of quiescent, untouched cells found in the
normal vascular wall.

CHAPTER2 Vascular wall physiology

The activated phenotype

These smooth muscle cells (Fig. 2.4) look strikingly different in
that they have lost almost all of their contractile filaments, as
well as many of the other smooth muscle cell-specific cell
markers. Because many of these cells instead display a large
number of other organelles such as ribosomes, endoplastic
reticulum, Golgi apparatus, and mitochondria, they have been
termed synthetic smooth muscle cells. Some may indeed syn-
thesize cellular proteins in preparation for their own division;
others may be quiescent in terms of proliferation, but may be
extremely active in synthesizing extracellular matrix proteins.
It is clear that the altered phenotypes of smooth muscle cells
also constitute a heterogeneous population; in contrast to
the unaltered phenotypes, this heterogeneity is readily visible
as a wide variation of cell sizes and shapes, variable organelle
contents, pseudopods and cytoplasmic extensions with
swellings, incomplete basement membranes, and a generally
activated appearance by light and electron microscopy. Such
cells can be found in the vascular tunica media, but they are
located more often in a thickened tunica intima, where they are
surrounded by increased amounts of extracellular matrix.
This abnormal matrix can vary in appearance from loose, indi-
cating the presence of many proteoglycans, to dense, fibrillar,
and collagenous, resembling scar tissue. Elastin also may
appear around phenotypically altered smooth muscle cells,

Figure 2.4 Eight days after balloon injury, this rabbit aorta displays highly
activated smooth muscle cells showing an organelle-rich phenotype. The
luminal cell (top) is indistinguishable from the one underneath, but no
platelets adhere to it. Freshly synthesized extracellular matrix also is visible
(transmission electron micrograph, bar= 1 jam).

but it always shows a less lamellar organization than the
elastic layers of the untouched normal vessel wall.

Smooth muscle cell hyperplasia

When tested for cell replication, many of the phenotypically
altered smooth muscle cells show active cell division and in-
creased cell turnover. Smooth muscle cells explanted into cul-
ture from previously balloon-injured arteries can divide every
8-10 h in the presence of abundant growth factors and serum,
and they keep growing faster than cells explanted from unin-
jured vessels even when most growth factors are withdrawn.
On the other hand, some cells found in restenosis specimens
after balloon angioplasty display a different morphologic ap-
pearance, and show no replicative activity at all when tested
with antibodies against a cell cycle-specific antigen. Further-
more, cells found in old atherosclerotic lesions, which may
look so different from a typical smooth muscle cell that they
might more appropriately be called fibrocytes, show no cell
division. In these extreme grades of phenotypic alteration, the
cell lineage cannot be determined positively, and it is possible
that monocyte-macrophages, histiocytes, or even real adven-
titial fibroblasts rather than smooth muscle cells have given
rise to these cell phenotypes found in diseased and injured
vascular walls. If tested with an antibody against smooth
muscle cell oc-actin, which is commonly used to prove smooth
muscle cell lineage, these extremely altered phenotypes react
negatively; they also are negative when tested for the RAM-11
antigen, which is commonly used to demonstrate the
monocyte-macrophage cellular lineage.

Differentiation and its significance for
vascular wound healing

The reversibility of phenotypic changes is undisputed in most
cells; however, for the highly differentiated vascular smooth
muscle cells, there is some doubt whether the phenotypic
changes are completely reversible in vivo. This doubt also
applies to the total reversibility of phenotypic changes in
endothelium, where the dysfunctional endothelial cell can be
regarded as a different phenotypic expression. The pheno-
types of vascular cells that feature increased replication are
associated with a loss of differentiation, and resemble embry-
onal cells ("pup" cells). 20 Partial, if not complete redifferentia-
tion is possible in tissue culture, usually under conditions that
closely mimic three-dimensional growth with matrix anchor-
age. In vivo, smooth muscle cells after vascular injury can reas-
sume a filament-rich, differentiated morphology, but a slightly
altered architecture of the extracellular matrix can be detected
up to a year after injury.

The smooth muscle cell that displays a phenotypic change
and then redifferentiates again represents the most important
repair mechanism after any kind of vascular surgery and other
interventions such as angioplasty. The wound-healing aspects

pa rt I Vascular pathology and physiology

of these phenotypic changes, the well known growth factors
that initiate them, and the lesser known differentiation influ-
ences that bring everything back to normal have important
clinical significance. There is a difference between desired vas-
cular healing and an excessive, hyperplastic response (e.g. at
the sites of anastomosis or angioplasty), but this difference lies
less in the involved factors themselves than in their timing, co-
ordination, and sequence. It is possible effectively to suppress
the response of smooth muscle cells to growth signals with,
for example, antibodies to the growth factors, with receptor
blockage, and, more recently, with antisense oligomere nu-
cleotides that block the synthesis of components of the intra-
cellular signal pathway that makes the nucleus respond to the
fact that the membrane receptor has been occupied by the ap-
propriate growth factor. Such inhibition of excessive growth,
however, probably needs to be limited in time and location,
because the basic growth response of smooth muscle cells is
the major mechanism by which injured blood vessels heal.

References

1. Lefer AM, Lefer DJ. Pharmacology of the endothelium in is-
chemia-reperfusion and circulatory shock. Annu Rev Pharmacol
Toxicol 1993; 33:71.

2. Zilla P, von Oppell U, Deutsch M. The endothelium: a key to the
future. / Cardiac Surg 1993; 8:32.

3. Liischer TF, Tanner FC, Tschudi MR, Noll G. Endothelial dysfunc-
tion in coronary artery disease. Annu Rev Med 1993; 44:395.

4. Buchanan MR, Brister SJ, Ofosu F. Prevention and treatment of
thrombosis: novel strategies arising from our understanding of
the healthy endothelium. Wien Klin Wochenschr 1993; 105:309.

5. Gertler JP, Abbott WM. Prothrombotic and fibrinolytic function of
normal and perturbed endothelium. / Surg Res 1992; 52:89.

6. Hahn AW, Resnik TJ, Mackie E, Scott-Burden T, Biihler FR. Effects

of peptide vasoconstrictors on vessel structure. Am J Med 1993;
94:13S.

7. Nagao T, Vanhoutte PM. Endothelium-derived hyperpolarizing
factor and endothelium-dependent relaxations. Am J Respir Cell
Mol Biol 1993; 8:1.

8. Nabel EG, Yang Z, Plautz G et ah Recombinant fibroblast growth
factor-1 promotes intimal hyperplasia and angiogenesis in arteries
in vivo. Nature 1993; 362:844.

9. Welch M, Durrans D, Carr HM et ah Endothelial cell seeding:
a review. Ann Vase Surg 1992; 6:473.

10. Gerritsen ME, Bloor CM. Endothelial gene expression in response
to injury. FASEB J 1993; 7:523.

11. DinarelloCA,GelfandJA, Wolff SM.Anticytokine strategies in the
treatment of the systemic inflammatory response syndrome.
JAMA 1993; 269:1829.

12. Denekamp J. Review article: angiogenesis, neovascular prolifera-
tion and vascular pathophysiology as targets for cancer therapy.
Br } Radiol 1993; 66:181.

13. Clinton SK, Libby P. Cytokines and growth factors in atherogene-
sis. Arch Pathol Lab Med 1992; 116:1292.

14. Clowes AW, Reidy MA. Prevention of stenosis after vascular
reconstruction: pharmacologic control of intimal hyperplasia —
a review. / Vase Surg 1991; 13:885.

15. Smith CW. Endothelial adhesion molecules and their role in
inflammation. Can J Physiol Pharmacol 1993; 71:76.

16. Farugi RM, DiCorietto PE. Mechanisms of monocyte recruitment
and accumulation. Br Heart J 1993; 69(Suppl. 1) :S19.

17. LoGerfo FW, Haudenschild CC, Quist WC. A clinical technique
for prevention of spasm and preservation of endothelium in
saphenous vein grafts. Arch Surg 1984; 119:1212.

18. Jackson CL, Schwartz SM. Pharmacology of smooth muscle cell
replication. Hypertension 1992; 20:713.

19. Reidy MA. Factors controlling smooth-muscle cell proliferation.
Arch Pathol Lab Med 1992; 116:1276.

20. Schwartz SM, Liaw L. Growth control and morphogenesis in the
development and pathology of arteries. / Cardiovasc Pharmacol
1993;21(Suppl.l):S31.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Hemostasis and coagulation

Donald L. Jacobs
Jonathan B. Towne

Formation of a hemostatic clot requires the coordinated action
of the vessel wall, endothelium, platelets, and plasma coagula-
tion factors. The process can be divided into primary and sec-
ondary hemostasis. Primary hemostasis involves adherence
of platelets to the site of injury to form a hemostatic plug. This
plug facilitates the formation and stabilization of fibrin at the
site of injury, which is called secondary hemostasis. Character-
istic of both primary and secondary hemostatic mechanisms is
the sequential activation of the factors (the coagulation cas-
cade) that allows for amplification of a small stimulus to result
in an adequate hemostatic response. In addition to amplifica-
tion, the sequential steps allow for many points of feedback in-
hibition to control the formation of clot. Fibrinolysis or clot
dissolution also involves activation of a series of factors that
can amplify and inhibit their own activity and interact with the
coagulation factors. This permits a balance to exist between
simultaneous thrombus formation and dissolution in the nor-
mal physiologic response. Throughout the hemostatic system
there are varying degrees of redundancy at different steps, re-
sulting in a wide range in the relative importance of various
factors in the hemostatic process. Clinically, this results in a
range in the severity of dysfunction seen with the inherited or
acquired deficiency of various coagulation or fibrinolytic
factors.

Normal hemostatic mechanisms

Response of the vasculature

When a vessel is injured the local smooth muscle of the vessel
wall contracts. This contraction is mediated by several factors
generated by the injured endothelium, the surrounding tis-
sues, and the forming thrombus. Thromboxane is a potent
vasoconstrictor that can arise from injured endothelium and
from activated platelets in an evolving thrombus. Endothelin,
a small peptide, is released from injured endothelium and also
may cause intense local vasoconstriction. Bradykinin, a vaso-
constrictor usually associated with an inflammatory response,

is produced from high-molecular-weight (HMW) kininogen
by the action of kallikrein, an activated protease involved in
the initiation of the intrinsic pathway of coagulation. Throm-
bin and the fibrinopeptide B released by thrombin's action on
fibrinogen also stimulate arterial smooth muscle contraction.
Degeneration and calcification of the arterial wall in athero-
sclerotic vessels results in an impaired ability to vasoconstrict.
Therefore, atherosclerotic vessels and other abnormal blood
vessels may not vasoconstrict appropriately when injured,
resulting in impaired hemostasis.

Endothelium

Endothelium has the obvious role as a barrier between the
blood and the thrombogenic subendothelial constituents. The
nonthrombogenic nature of the endothelium is reflected in
several metabolic functions of the endothelial cell that go
beyond the maintenance of a structural barrier. Endothelial
cells produce antiplatelet, anticoagulant, and fibrinolytic
substances that modulate the hemostatic response both locally
and systemically.

Prostacyclin (PGI 2 ) is a potent vasodilator and antiplatelet
compound produced by endothelial cell metabolism of arachi-
donic acid. Thromboxane A 2 (TBX A 2 ), a potent vasoconstric-
tor and platelet activator usually associated with platelet
granules, also is produced in the endothelium from the metab-
olism of arachidonic acid. The ratio of PGI 2 to TBX A 2 produc-
tion in cultured endothelial cells is of the order of 10 : 1 to 100 :
1. The role of endothelial cell-derived TBX A 2 in vivo is uncer-
tain; however, the relative levels of these compounds, regard-
less of their source, may be an important determinant of the
endothelium's antithrombotic state. 1-3 PGI 2 release is stimu-
lated by endothelial cell contact with thrombin, platelets, and
adenosine diphosphate (ADP). 4/5 Decreased PGI 2 release can
occur with denuding and nondenuding endothelial cell in-
jury. 6-8 Nitric oxide (NO), an endothelium-dependent relax-
ation factor produced by endothelial cells, is important not
only for the maintenance of normal vasorelaxation but also as
an inhibitor of platelet adhesion and activation. 9 Even minor

pa rt I Vascular pathology and physiology

endothelial cell injury can result in impaired production
of NO, and may cause increased vasospasm and platelet
adherence.

ADP is a potent platelet-aggregating agent released by acti-
vated platelets. Endothelial cells appear to have an enzyme
that hydrolyzes ADP to adenosine on their cell surface (ecto-
ADPase). 10-12 This may function to limit ADP-promoted
platelet aggregation on normal endothelium.

Thrombomodulin is an endothelial cell membrane receptor
for thrombin that binds thrombin and decreases its ability to
generate fibrin while increasing its ability to activate the
protein C anticoagulant pathway. 13 Endothelial cells have
heparin-like glycosaminoglycans on their cell surface that can
combine with anti thrombin III (AT-III), which also is produced
in the cell, to inhibit thrombin activity on normal endothe-
lium. 14 ' 15 These anticoagulant pathways are discussed in later
sections.

Another function of the endothelial cell is to modulate the
fibrinolytic system by producing tissue plasminogen activator
(tPA) and plasminogen activator inhibitor-1 (PAI-1). The net
balance of tPA and PAI-1 activities probably determines the
fibrinolytic activity of the vessel wall. 16 ' 17

Normal endothelium has some procoagulant properties, in-
cluding production of platelet-activating factor, von Wille-
brand factor (vWF), factor V, and PAI-1. The procoagulant
activity, however, is minimal in normal compared with injured
endothelium. 18 ' 19 The perturbation needed to affect the antico-
agulant properties of the endothelial cell may be minimal
and may result from manipulations as minor as a venous or
arterial cannulation or the vessel dissection and distention
commonly done in vein graft preparation. 7 ' 18-21 Also, the
inflammatory mediators interleukin-1, tumor necrosis factor,
and endotoxin can injure endothelial cells and inhibit their
anticoagulant properties. 22

Platelets

Platelets are involved in the initial phase of hemostasis by
forming a hemostatic plug at the site of vessel injury. Platelets
adhere to the exposed subendothelium by binding a surface
membrane glycoprotein (GPIb) to polymeric vWF, which has
been bound to the collagen in the subendothelium. In the ab-
sence of a high shear rate, the platelets can adhere to collagen
in the absence of vWF.

Once adherence has occurred, the platelets are activated,
resulting in at least three processes:

1 The platelets release dense granules and a granules. 23 Dense
granule contents include ADP, a potent platelet-aggregating
agent; serotonin, a vasoconstrictor and adrenergic agonist;
and ionized calcium. a-Granule contents include platelet
factor 4, a neutralizer of heparin; (3-thromboglobulin, a
platelet-specific protein of unknown activity that is used as
a marker for platelet activation; platelet-derived growth fac-
tor, a mitogen for smooth muscle cells and fibroblasts;

thrombospondin, a protein involved in platelet adhesion;
factor V; and small amounts of vWF.

2 Mobilization of arachidonic acid from the platelet phospho-
lipid results in synthesis of TBX A 2 , a potent vasoconstrictor
and platelet aggregator. 23

3 Conformational change in the platelet surface membrane
lipoproteins causes the platelet to become spherical. This
activated form of the platelet membrane is referred to as
platelet factor 3 and functions as a platform for formation of
the prothrombinase complex (prothrombin activation) and
the VHI-IXa-X complex (factor X activation by the intrinsic
pathway). 24 ' 25 These pathways are detailed further in subse-
quent sections.

Platelet aggregation can be stimulated by the platelet-
derived ADP or by epinephrine, collagen, platelet-activating
factor, thrombin, or immune complexes. All of these agents
seem to mediate platelet activation by increasing intracellular
calcium ion concentration. 23 Aggregation requires fibrinogen,
which links the platelets by binding to the GPIIb-IIIa receptor
expressed on the surface membrane after activation. Stabiliza-
tion of the platelet plug occurs when fibrin is formed and
cross-linked in the platelet mass. Platelets then mediate retrac-
tion of the mature clot by their constituent cytoskeletal con-
tractile proteins and microtubules.

Thrombin generation

Thrombin is the central enzyme in the hemostatic mechanism.
In addition to its primary role in fibrin formation, thrombin
has effects on the platelets, endothelium, and anticoagulant
pathways. The cascades involved in thrombin generation
have been divided into extrinsic and intrinsic pathways (Fig.
3.1). This is another way of differentiating between initiation
by a negatively charged surface (the intrinsic pathway) and
initiation by tissue factor (the extrinsic pathway). The two
pathways interact at two points (Xlla can activate factor VII,
and Vila-tissue factor complex can activate factor IX), and
they converge at the step of factor X activation. It is unlikely
that the two pathways ever function individually in vivo. The
extrinsic pathway seems relatively more important than the
intrinsic because people with a deficiency in the factors of
the early steps of the intrinsic pathway have no significant
bleeding problems. Division into the intrinsic and extrinsic
pathways does have practical value in that the prothrombin
time reflects the efficiency of the extrinsic pathway and the
partial thromboplastin time reflects the efficiency of the
intrinsic pathway. Consequently, these tests can help in local-
izing an abnormality in the coagulation cascade.

Extrinsic pathway

The extrinsic pathway is initiated by the expression of tissue
factor (tissue thromboplastin) on the cells of injured tissue or
from activated endothelial cells and monocytes. Tissue factor

chapter 3 Hemostasis and coagulation

Figure 3.1 Thrombin generation, HMK, high-
molecular-weight kininogen; TF, tissue factor;
PF3, platelet factor 3; Ca 2+ , calcium ion.

INTRINSIC PATHWAY

XII

prekallikrein

HMK

Xla

Xla — Vlla-TF

IXa

IXa Villa Ca
Q PF3

EXTRINSIC PATHWAY

Xlla £1 Xa

Vila-tissue factor

Vlla-TF

/Prothrombinase\

V complex / Villa Xa Ca

PF3

Prothombin

THROMBIN

is a transmembrane protein that is not fully expressed until the
cell is activated or injured. When blood is exposed to injured
tissue, a complex forms between factor VII and tissue factor.
Factor VII is a vitamin K-dependent protein that is able to con-
vert factor X, another vitamin K-dependent protein, to factor
Xa, using tissue factor as a cofactor. Once a small amount of
factor Xa is produced, it amplifies the response by cleaving fac-
tor VII to Vila, which is able to complex with tissue factor and
convert factor X to Xa much more efficiently (100-fold increase
in the activation rate). Factor Xa is the primary junction point
of the extrinsic and intrinsic pathways, and cleaves prothrom-
bin to thrombin via the prothrombinase complex.

Intrinsic pathway

The intrinsic pathway is the cascade of coagulation factors that
is activated by contact with a negatively charged surface such
as glass or a biologic activator such as basement membrane,
insoluble collagen, or endotoxin. It involves six clotting factors

and is divided into the contact phase and the factor IX and X
activation phase.

Contact phase

The contact phase involves three precursors, factors XII and XI
and prekallikrein, which are converted to their active serine
protease forms by surface contact and HMW kininogen as a
cofactor. The surface must be negatively charged. In contrast
to most other complex formations in the coagulation cascades,
the contact phase requires no calcium and involves no vitamin
K-dependent factors. The exact mechanisms involved in the
contact activation of these factors are not fully understood.
The contact reaction is involved not only in coagulation but
also in fibrinolysis (kallikrein and Xlla convert plasminogen to
plasmin), kinin generation (kallikrein releases bradykinin
from HMW kininogen), complement activation (kallikrein
and factor Xlla activate the first component of complement),
and activation of the renin-angiotensin system. 26 ' 27 In the co-
agulation cascades, factor Xla proceeds in the intrinsic path-

pa rt I Vascular pathology and physiology

way to factor IX activation, and factor Xlla primes the extrinsic
pathway by activating factor VII. Deficiency of factor XII,
prekallikrein, or HMW kininogen is an asymptomatic condi-
tion. 28 As noted, this may indicate the relative unimportance of
contact activation of coagulation in vivo.

Factor IX and X activation

The second phase of the intrinsic pathway is the activation of
factor IX by factor XIa, which combines with cofactors (includ-
ing factor VIII) to convert factor X to factor Xa. Factor IX, a vit-
amin K-dependent protein also known as Christmas factor, is
deficient in hemophilia B. Factor IX is converted to the serine
protease factor IXa by factor XIa in the presence of calcium ion
or by the tissue factor-Vila complex of the extrinsic pathway.
Factor VIII, also known as antihemophilic factor, is deficient in
hemophilia A and is present in plasma as a stable complex with
vWF. Factor VIII requires modification by thrombin for its full
activity. This is an example of a feedback amplification process
in the cascade. Factor VIII, along with phospholipid (the
platelet factor 3 of activated platelets) and calciumion, all act as
cofactors in a complex with factor IXa that converts factor X to
factor Xa. Factor Xa is the point of final convergence of the ex-
trinsic and intrinsic pathways, and goes on to participate in the
prothrombinase complex.

Prothrombinase complex

The final step in thrombin generation is the cleavage of pro-
thrombin to thrombin by the action of factor Xa in the presence
of factor Va, calcium, and phospholipid (platelet factor 3). The
association of these factors in the prothrombinase complex ac-
celerates the activity of factor Xa and also protects it from inhi-
bition by AT-III. 29 ' 30 The phospholipid platform is provided
most typically by activated platelets but also can be provided
by injured tissue, endothelium, or white blood cells, thus lo-
calizing the process to a site of injury. As noted, thrombin has
effects on the coagulation system at many points. It not only
cleaves fibrinogen to fibrin, but activates factors XIII, V, and

VIII, prothrombin, and protein C. Thrombin also can aggre-
gate platelets and stimulate endothelial cell release of PGI 2 ,
vWF,andPAI-l. 31 - 33

Fibrin formation

Fibrin is the cohesive substance of the mature clot whose for-
mation is central to secondary hemostasis. Fibrin formation
occurs in three steps (Fig. 3.2): (1) splitting of fibrinogen by
thrombin into fibrin monomer and fibrinopeptides A and B; (2)
polymerization of monomers to fibrin strands; and (3) action
of factor XIII (activated by thrombin) and calcium ion to cross-
link the fibrin strands. Fibrinogen is an acute-phase reactant
whose concentration in the plasma is a relatively high 200 to
400mg/dl. Thrombin cleavage of fibrinogen to fibrin appears
to be a minor part of the overall catabolism of fibrinogen. Fib-
rinopeptide A release is essential for the polymerization of the
fibrin monomers to occur. Fibrinopeptide B release appears to
be important before cross-linking of the fibrin can occur. 34 Fib-
rinopeptide B release also may result in vasoconstriction
owing to its action on smooth muscle. As the fibrin monomers
associate into strands, other proteins, including plasminogen,
tPA, and oc 2 -antiplasmin (oc 2 -AP) are incorporated into the
forming thrombus and are therefore localized to the thrombus
to facilitate the balance of thrombosis vs. fibrinolysis. 35
Thrombin has the dual actions in fibrin formation of
fibrinogen cleavage and activation of the fibrin-stabilizing
enzyme factor XIII. Activated factor XIII can covalently
cross-link fibrin, making it mechanically more rigid and more
resistant to the action of plasmin. 36 Factor XIII is able to cross-
link other proteins like actin, a 2 -macroglobulin, fibronectin,
and collagen, and may play a part in the processes of tissue
repair.

Fibrinolysis: plasmin generation

Plasmin is the protease active in the degradation of fibrin, and
facilitates the lysis of clot. Plasmin not only acts to lyse clot but

Fibrinogen

Fibrin monomer + fibrinopeptides a, b

Fibrin polymer

XIII

Xllla
Ca ++

Insoluble fibrin

Figure 3.2 Three steps of fibrin formation:
(1) cleavage of fibrinogen by thrombin; (2)
spontaneous polymerization of fibrin
monomers; and (3) cross-linking of fibrin
strands.

chapter 3 Hemostasis and coagulation

limits clot formation by digestion of the coagulation factors V,
VIII, and XII, and prekallikrein.

Plasmin is derived from plasminogen by the proteolytic
action of many different activators. Intrinsic activators are
present in the plasma and can act when blood comes into con-
tact with a foreign surface. These include the contact-phase
proteins factor XII, prekallikrein, and HMW kininogen. 26 ' 27
Extrinsic activators include tPA, urokinase, and streptokinase.
tPA is found in many tissues, including vascular endothelium,
and is incorporated into thrombus by a high affinity for fibrin,
placing it in position to activate plasminogen at the site of the
thrombus. 37 ' 38 tPAhas minimal effect on circulating plasmino-
gen. Urokinase is the principal pharmacologic plasminogen
activator used in patients undergoing lytic therapy. Uroki-
nase's role in the blood under normal conditions is not clear;
however, prourokinase is present in plasma and may be
converted to urokinase by the activated contact system. 39
Streptokinase is derived from the bacterium (3-hemolytic
streptococcus and used to be an important lytic therapy agent,
but its potential for causing hypersensitivity reactions and the
difficulties in controlling the dose response now limit its use.

Inhibition of plasmin activity occurs by two mechanisms.
PAI-1, which is released by endothelial cells, acts to inhibit the
cleavage of plasminogen by all activators. 16 ' 40 Plasmin activity
also is inhibited by another endothelial cell product, oc 2 -AP,
which forms a stable complex with plasmin. 41 oc 2 -AP also in-
hibits the incorporation of plasminogen into the fibrin strands
and becomes cross-linked to fibrin during clot formation,
thereby inhibiting the subsequent degradation of fibrin by
plasmin. 35

Control of the fibrinolytic system depends on the relative
activities of the plasminogen activators and inhibitors. As
noted in the discussion of the endothelium's role in hemo-
stasis, the balance of tPA vs. PAI-1 is central to the control of
fibrinolysis at the endothelial surface. Modulation of the
release of tPA and PAI-1 occurs in response to shear stress on
the endothelium, venous distention, and the presence of
thrombin, fibrin, endotoxin, and interleukin-1. 37 ' 38 The impor-
tance of oc 2 -AP in the regulation of fibrinolysis is evident in the
severe hemorrhagic problems seen in people with an inherited
oc 2 -AP deficiency.

Anticoagulant pathways

Control of hemostasis occurs by balancing thrombosis and
fibrinolysis through multiple feedback mechanisms. This
involves the action of several natural anticoagulants that both
inhibit the coagulation pathways and activate the fibrinolytic
system.

AT-III, also known as heparin cofactor, is the major inhi-
bitor of thrombin and is the most important natural anti-
coagulant. 14 ' 15 AT-III also is an inhibitor of the activated
clotting factors Xlla, XIa, Xa, and IXa, and kallikrein. Con-
versely, AT-III can decrease fibrinolytic activity by inhibiting

plasmin. Heparin functions as an anticoagulant by binding
to AT-III and increasing its activity greatly. 15 Natural heparin-
like glycosaminoglycans on the surface of the endothelium
also appear to increase AT-III activity, contributing to the
nonthrombogenic nature of the endothelium. 15 ' 42 The heredi-
tary deficiency of AT-III results in recurrent thrombotic
episodes. A patient deficient in AT-III often is diagnosed due to
an inability to achieve adequate anticoagulation while on
heparin.

The protein C pathway is a major anticoagulant pathway
consisting of two vitamin K-dependent plasma proteins,
protein C and protein S, and the thrombin receptor thrombo-
modulin on the endothelium. 13 ' 43 ' 44 Thrombin combines with
thrombomodulin and becomes less active at catalyzing fibrin
formation and more active at converting protein C to the acti-
vated form. Activated protein C combines with protein S to
form a complex on the endothelial or platelet surface that has
two actions. One is to increase the rate of fibrinolysis by neu-
tralizing PAI-1, and the second is to decrease clot formation by
selectively degrading factors Va and Villa. Protein C and S
deficiencies both result in thrombotic tendencies, primarily
venous thromboembolism. These deficiencies are detailed
further in the discussion of hypercoagulable states.

Other circulating anticoagulants include heparin cofac-
tor II, which inhibits thrombin in a manner similar to that
of AT-III but does not inhibit other activated factors, and oc 2 -
macroglobulin, a nonspecific plasma protease that can inhibit
thrombin but whose physiologic role is not defined.

Hepatic metabolism

Hepatic biosynthetic function is required for synthesis of most
of the coagulation proteins and some of the fibrinolytic factors
and natural anticoagulants. Hepatic clearance of activated
coagulants or plasminogen activators also influences hemo-
stasis. The liver appears to be able to clear selectively only the
activated forms of the coagulation factors and leave the inac-
tive forms in the circulation. 45 The liver reticuloendothelial
macrophages selectively clear fibrin but not fibrinogen. Alter-
ation in hepatic function by intrinsic hepatic disease, metasta-
tic disease, or as a result of hemorrhagic shock can result in
decreased or increased levels of various clotting factors. In
general, the loss of this homeostatic function results in a
hemorrhagic rather than a thrombotic dysfunction.

Vitamin K is needed for synthesis of the coagulation factors
VII, IX, and X, and prothrombin, and the anticoagulants pro-
teins C and S in the liver. Vitamin K is required for the insertion
of a second carboxyl group to the y-carbon of certain glutamic
acid residues in the polypeptide precursors of these proteins.
These carboxylated poly glutamic acid regions are essential for
the binding of calcium ions and allow these proteins to interact
with phospholipids. These interactions are the key to facilitat-
ing the formation of the prothrombinase complex and other
membrane-bound processes of the hemostatic and fibrinolytic

pa rt I Vascular pathology and physiology

systems. 29 ' 30,43,46 The carboxylation reaction requires the me-
tabolism of vitamin K to an epoxide. This usually is recycled by
a vitamin K epoxide reductase. The oral anticoagulant drug
warfarin inhibits the vitamin K epoxide reductase and can de-
plete vitamin K by preventing its recycling to the active form. 47
This results in the decreased synthesis of functional vitamin In-
dependent coagulation factors containing the carboxylated
poly glutamic acid regions. Severe malnutrition and fat malab-
sorption impairing fat-soluble vitamin absorption can result
in a decrease in the vitamin K-dependent clotting factors. This
is rare, however, unless concurrent antibiotic therapy further
reduces vitamin K levels by inhibiting the gut flora production
of vitamin K-like compounds. 48 The cephalosporin antibiotic
cefamandole and its structural analogues also can inhibit vita-
min K carboxylase directly. 49

Hypercoagulable states

Dysfunctions of the hemostatic mechanisms that result
in hypercoagulable states are becoming more frequently
recognized clinical problems. In vascular surgery, these pro-
thrombotic states have particular importance in unexpected
arterial as well as venous thrombosis. Many hypercoagulable
states are difficult to diagnose by history and laboratory find-
ings, and a high index of suspicion is helpful in evaluating
patients with unexpected thrombosis.

AT-III deficiency also can be acquired. Low levels of AT-III
can result from acute thrombosis, disseminated intravascular
coagulation (DIC), and liver disease. Postoperative measure-
ments have demonstrated that AT-III levels can drop 63% after
major vascular procedures, with a nadir around the third post-
operative day. 56 Also, the administration of heparin can de-
crease AT-III levels, presumably owing to increased clearance
of AT-III after complexing with heparin. 57 ' 58 The possibility of
an acutely acquired decrease in AT-III should be considered
when diagnosing inherited or chronically acquired AT-III defi-
ciency. Conversely, oral anticoagulants have been reported to
increase the level of AT-III into the normal range in patients
with known inherited AT-III deficiency. 59 ' 60

Treatment of patients with symptomatic AT-III deficiency
requires life-long warfarin anticoagulation. Warfarin not only
achieves anticoagulation by decreasing vitamin K-dependent
factors, but increases AT-III levels, as noted earlier. Patients
with asymptomatic AT-III deficiency identified after the
diagnosis of a symptomatic relative are not prophylactically
treated unless they are to be placed at increased risk (e.g.
pregnancy or operation). Treatment of an acute thrombosis in
a patient with AT-III deficiency requires replacement of AT-III
using fresh frozen plasma and heparin anticoagulation until
adequate oral anticoagulation is achieved. AT-III concentrates
may prove useful in treatment of the inherited or acquired AT-
III deficiency and other prothrombotic states, including DIC
and postoperative deep venous thrombosis. 61 ' 62

Antithrombin-lll deficiency

Inherited AT-III deficiency is an autosomal dominant trait
with a prevalence of 1 in 2000 to 5000 people. Homozygous
inheritance is fatal in infancy. It is reported that 55% of bio-
chemically affected people have a thrombotic event. 50 This
event is associated most frequently with an inciting circum-
stance such as pregnancy, hormonal therapy, or an operation.
About 40% of the thrombotic events are spontaneous. A com-
mon presentation of AT-III deficiency is that of a patient who
has a thrombotic event and manifests an inability to anticoag-
ulate adequately in response to heparin. Patients typically pre-
sent in their third or fourth decade. AT-III deficiency results
primarily in venous thrombosis; however, acute arterial
thrombosis and thrombosis after arterial reconstruction can
occur, and AT-III deficiency should be suspected in any patient
with unexpected thrombosis. 51-53

Diagnosis of AT-III deficiency can be confirmed by measure-
ment of plasma AT-III. The most common type of inherited
AT-III deficiency results in a decreased amount of AT-III, and
an immunologic assay of AT-III levels reveals levels below
70% of normal. 54 A much less frequent type of inherited AT-III
deficiency results in the production of an abnormal AT-III, and
an immunologic assay will show normal levels of AT-III, but
the functional assay will show markedly decreased biologic
activity. 55

Protein C deficiency

Inherited protein C deficiency is an autosomal dominant trait
that results in a clinical syndrome very similar to AT-III defi-
ciency. The prevalence of heterozygosity for protein C defi-
ciency is 1 in 200 to 300 people. Homozygosity is associated
with a severe deficiency and a condition termed neonatal
purpura fulminans, which usually is fatal. Heterozygotes
have protein C levels of 50% of normal or less. About 75%
of heterozygotes experience a thrombotic event, 70% spon-
taneously and 30% in association with some risk factor such as
pregnancy, hormonal therapy, or an operation. 63-65 Nearly
all of these thrombotic events are venous; however, it has been
reported that in a small series of patients under 51 years of age
who required arterial reconstruction, 15% had a protein C
deficiency. 66

As with AT-III deficiency, there are different forms of protein
C deficiency. The most common type manifests as a decrease in
the amount of protein C. In the second type, normal amounts
of a dysfunctional protein C are produced. This second type re-
quires the use of a functional assay to diagnose the deficiency
correctly. 67

Acquired protein C deficiency is associated primarily with
liver disease. Decreased levels have been reported in DIC,
adult respiratory distress syndrome, and in the postopera-
tive period. 68 ' 69 Diagnosis of protein C deficiency requires

chapter 3 Hemostasis and coagulation

consideration of any condition that may impair hepatic syn-
thetic function because the liver is the source of protein C. Pro-
tein C is a vitamin K-dependent protein, and administration of
warfarin results in its decreased synthesis. Hence, the mea-
surement of protein C must be performed while patients are
not on oral anticoagulants.

When warfarin therapy is initiated in a patient with low
levels of protein C, a paradoxical hypercoagulable state may
be transiently induced. 70 ' 71 Because the half-life of protein C is
a relatively short 6-8 h, its levels fall before the depression of
the vitamin K-dependent procoagulant factors IX and X, and
prothrombin, which have half-lives of 5-7 days. This results in
the loss of the protein C anticoagulant effect before inhibition
of the coagulation pathway is achieved, and for a short time
the patient becomes paradoxically more hypercoagulable. The
association of protein C deficiency with warfarin-induced
skin necrosis is probably due to thrombosis in the microcircu-
lation during this transient hypercoagulable state. 72 ' 73 By fully
heparinizing the patient before initiation of warfarin and
maintaining heparin until oral anticoagulation is adequate,
this transient hypercoagulable state is prevented.

Protein S deficiency

Inherited protein S deficiency is nearly identical to protein C
deficiency in its mode of inheritance and clinical presentation.
Thrombotic events generally are venous but, in the small se-
ries cited previously, 20% of the patients under 51 years of age
who required arterial reconstruction manifested a protein
S deficiency. 66 The diagnosis of protein S deficiency is more
difficult owing to the variability in normal levels seen in the
population. Also, approximately 60% of protein S is bound in
normal plasma to the C4b-binding protein of the complement
system. Only the free portion of the protein S in the plasma is
active, 74 which further complicates the meaningful measure-
ment of protein S. Acquired protein S deficiency is associated
with pregnancy, hormonal therapy, and DIC. 75,76 Total and free
concentrations of protein S are only moderately decreased
in liver disease. 76 Any inflammatory response that releases
the acute-phase reactant C4b-binding protein can result in a
decrease in the activity of protein S. Because protein S is a
vitamin K-dependent protein, its level is decreased in warfarin
therapy, and this could contribute to a transient hypercoagula-
ble state during initiation of warfarin anticoagulation by the
same mechanism as seen with protein C.

Treatment of patients with protein S deficiency also is simi-
lar to the treatment recommended for protein C deficiency; he-
parin is effective therapy for an acute thrombotic event, with
oral anticoagulation initiated after the patient is heparinized.
As with AT-III and protein C deficiencies, treatment is re-
quired only for symptomatic patients. Prophylactic treatment
is reserved for asymptomatic people who are placed in cir-
cumstances of increased risk, such as an operation or obstetric
event.

Antiphospholipid antibodies

Antibodies to the phospholipid components of cell mem-
branes can react with the phospholipid components of the
phospholipid-dependent coagulation tests and prolong the
coagulation time in vitro. Of patients with systemic lupus
erythematosus, 10-30% may exhibit such plasma inhibitors
of in-vitro coagulation— hence their designation as lupus anti-
coagulants. 77-79 These antibodies, however, are associated
clinically with a hypercoagulable state. About 30% of patients
with these antibodies have histories of a thrombotic event. 80,81
Both venous and arterial thrombosis can occur, with a
50% incidence of thrombosis after vascular procedures
reported in a series of patients positive for antiphospholipid
antibodies. 82 Recurrent spontaneous abortions in patients
with antiphospholipid antibodies are common.

The mechanism by which these antibodies produce throm-
bosis is not clear. It has been proposed that there is an inhibi-
tion of PGI 2 production by endothelial cells and an increase in
TBX production by platelets. Inhibition of protein C activation
and inhibition of prekallikrein with subsequent impaired
plasminogen activation also have been proposed. Although
the antibodies have been shown to bind to platelets in a specif-
ic manner, activation of platelets has not been consistently
demonstrated.

Diagnosis of the presence of antiphospholipid antibodies
(lupus anticoagulants) can be made by demonstration of a pro-
longation of coagulation times [i.e. activated partial thrombo-
plastin time (aPTT), Russell's viper venom time, or the kaolin
clotting time] that do not correct with the addition of normal
plasma. The platelet neutralization assay uses the anti-
phospholipid antibodies' ability to inhibit platelet binding
to collagen as evidence of their presence in a patient's plasma.
Reactivity of a patient's plasma with cardiolipin by an
enzyme-linked immunosorbent assay is a useful screening
test; however, it is not a specific test, and must be confirmed by
one of the aforementioned assays.

Prophylactic therapy in asymptomatic patients with a
known lupus anticoagulant who are undergoing vascular re-
construction is recommended. This consists of preoperative
antiplatelet therapy, intraoperative dextran 40 and heparin,
and postoperative warfarin. 83 Therapy for symptomatic pa-
tients with antiphospholipid antibodies consists of heparin
for the acute event and chronic oral anticoagulation subse-
quently. 84 Steroids also have been used to decrease the level
of the anticoagulant antibodies. 85,86 Monitoring of the partial
thromboplastin time can be problematic owing to the antico-
agulant effect of the antibody on the test. Measurement of the
thrombin time or measurement of heparin levels are alterna-
tive methods of monitoring heparin therapy.

Heparin-induced thrombosis

Paradoxical thrombotic complications of heparin sodium

pa rt I Vascular pathology and physiology

anticoagulant therapy are uncommon but potentially limb
threatening and occasionally fatal. Several investigators have
identified a chemically induced, immune thrombocytopenia
as the cause of heparin-induced intravascular thrombosis,
which usually occurs after 4-10 days of continued exposure to
the drug. 87-91 The immune factor that triggers the thrombocy-
topenia has been identified as an IgG antibody, which pro-
duces agglutination of normal platelets when either porcine
gut or beef lung heparin is added. The IgG protein is stimulat-
ed by the heparin /platelet factor 4 complex and activates the
platelet via the platelet F c receptor. 92 The thrombi that occur
with heparin-induced thrombosis have an unusual grayish
white appearance in contradistinction to the red color of most
thrombi. The white color is secondary to the creation of fibrin-
platelet aggregates, which can be clearly identified on electron
microscopy. 93

Rhodes, Dixon and Silver 94 found a heparin-dependent IgG
antibody in the serum of several patients by means of the com-
plement lysis inhibition test. They also demonstrated a resid-
ual heparin platelet-aggregating effect 12 days to 2 months
after patient recovery from the initial exposure to heparin. In
these patients a 24-hour infusion of heparin caused a mean
reduction of platelet count of 197000/mm. Since heparin pre-
parations are not pure substances, it is also possible that a
high-molecular contaminant not eliminated by the extraction
procedure may cause the antiplatelet defect.

Up to 30% of patients may manifest a decrease in their
platelet count after starting heparin therapy, but the incidence
of significant thrombocytopenia and resulting thrombotic or
hemorrhagic complications is approximately 5%. 95 Two types
of heparin-induced thrombocytopenia are described. Type I,
or the acute form, occurs relatively early and results in a be-
nign course with improvement in the platelet count during
continued heparin therapy. Type II, or the delayed form, oc-
curs 5-14 days after the institution of heparin therapy in a pa-
tient not previously exposed to heparin and after 3-9 days in
patients with a history of previous heparin therapy. Type II
heparin-induced thrombocytopenia is reported to have a
23-60% thrombotic or hemorrhagic complication rate and a
12-18% mortality rate. Early recognition and treatment results
in a significant improvement in the associated morbidity and
mortality. 9697 In type I heparin-induced thrombocytopenia,
the mechanism of action is thought to be a nonimmune-
mediated direct effect of heparin on platelets that causes ag-
gregation. Type II heparin-induced thrombocytopenia is due
to an immune-mediated (IgG and IgM) platelet aggregation.

Clinical presentation

Heparin-induced intravascular thrombosis can occur follow-
ing a wide variety of indications for heparin administration,
including thrombophlebitis with and without pulmonary em-
bolus, perioperative heparin prophylaxis in patients at risk for
thrombophlebitis, cardiac surgery, and vascular reconstruc-

tion. Platelet aggregation induced by heparin can result from
both porcine gut and bovine lung heparin and can affect either
the arterial or venous circulation. Both subcutaneous and
intravenous heparin administration can produce this phe-
nomenon. 98 Even heparin-coated catheters can cause heparin-
induced thrombocytopenia. Laster and Silver 98 reported the
development of heparin-induced thrombocytopenia in 10
patients in whom heparin-coated pulmonary artery catheters
were inserted. Despite discontinuation of all other sources
of heparin, the thrombocytopenia persisted. Although all the
patients also were given heparin, it is theoretically possible
that heparin-coated catheters alone could have caused ab-
normal platelet aggregation.

The clinical features of this syndrome are often dramatic. In
any patient who has had thrombotic complications while
receiving heparin therapy, heparin-induced aggregation of
platelets should be considered. This is especially important
in patients with arterial occlusions who do not have any other
evidence of atherosclerotic vascular disease. At operation, the
finding of a white clot at thrombectomy should alert the sur-
geon to the possibility of heparin-induced thrombosis. In
contrast to several reports in the literature, increased heparin
sensitivity rather than increased heparin resistance was noted
in several of our patients. 93 The cause of this is uncertain, but it
is presently believed that it is unrelated to the heparin-induced
aggregative immunoglobulin.

It was initially felt that arterial thrombosis was more preva-
lent than venous thrombosis with this complication. However,
some prospective studies by Warkentin et al. demonstrated
that actually there is a prevalence of venous to arterial emboli
at a 4 : 1 ratio. 92 Many of the venous thromboses are not detect-
ed unless studies such as duplex scanning search these out.
The most common arterial location of thrombosis is the ex-
tremities, primarily the lower extremities, followed by the
cerebral circulation and finally manifesting as myocardial in-
farctions. Exact cause of this distribution of prevalence of
thrombosis is uncertain, but it is certainly the authors' view
that the thrombosis occurs more commonly on diseased ves-
sels, and the incidence of diseased vessels in the extremities,
particularly the lower extremities, is much higher than many
other vessels. This is followed by carotid bifurcation disease and
coronary artery disease.

Skin lesions have been noted in patients with heparin-
induced thrombosis. These are often seen at the site of the sub-
cutaneous injection. They can present as painful erythematous
plaques which can progress to skin necrosis. These can be un-
accompanied by thrombocytopenia. With the prevalence of
subcutaneous injection, the incidence of these findings has
increased."

Diagnosis

Definitive diagnosis of heparin-induced intravascular throm-
bosis is obtained by performing platelet aggregation tests.

chapter 3 Hemostasis and coagulation

Two patterns of response have been noted. The more common
pattern is for the patient's platelet-poor plasma to aggregate
donor platelets on the addition of heparin, indicating the pres-
ence of a relative nonspecific platelet-aggregating factor in the
patient's plasma. The less common pattern is for the patient's
plasma to be active against only the patient's platelets and
have no effect on donor platelets. Other more sensitive tests
include C 100 serotonin release testing and Elisa testing for the
antibody to the heparin PF 4 complex.

Other clotting factors are usually normal: fibrinogen level
is normal, fibrin split product level may be mildly elevated
but not in the range seen with intravascular coagulation, and
prothrombin time is normal or slightly prolonged. All patients
have a marked reduction in platelet count of less than
100000/mm 3 or a 50% decrease from admission level. In our
series, 93 the platelet count averaged 37 500 /mm 3 with a range
of6000to73000/mm 3 .

Patients with arterial thrombosis often present with unique
angiographic findings. These lesions consist of broad-based,
isolated, lobulated excrescences that produce a variable
amount of narrowing of the arterial lumen. Usually these find-
ings have an abrupt appearance, with prominent luminal
contour deformities in arterial segments that are otherwise
normal. This distribution of disease is unusual and distinct
from findings commonly seen with atherosclerosis. These
changes occur in both the suprarenal and infrarenal portions
of the abdominal aorta and represent adherent mural thrombi
composed of aggregates of platelets and fibrin incorporating
varying amounts of leukocytes and erythrocytes. Platelet ag-
gregation tests also should be performed on any patient in
whom recurrent pulmonary embolism developed while re-
ceiving adequate heparin therapy.

The diagnosis of heparin-induced thrombosis is primarily
a clinical diagnosis. All the current laboratory tests have a
relatively high percentage of false-negative rates and the more
difficult to perform serotonin release and Elisa tests are not
available in all hospitals. A patient in whom the clinical syn-
drome of low platelet count and abnormal thrombosis is
noted, and in whom the tests are negative, should be treated
with a presumptive diagnosis of heparin-induced thrombosis
and the tests repeated. Although false-negative tests are re-
ported, the incidence of false-positive tests is quite unusual.

Treatment

Currently there are three approaches to treating patients with
heparin-induced thrombosis. 92 The first is the use of Dana-
poroid, a mixture of glycosaminoglycans (heparin sulfate)
and dermatin sulfate. This is quite effective except that this
has a 10-40% cross-reactivity with patients having heparin-
induced platelet aggregation. Therefore, prior to instituting
this therapy, patients need to have their platelets tested
against Danaporoid to make sure it does not cause platelet
aggregation. The second course of therapy is the use of

Lepirudin, a recombinant form of the medicinal lead salivary
protein hirudin, which is a direct thrombin inhibitor which
can be quite effective. Patients are monitored by obtaining an
activated thromboplastin time (A-PTT), which is kept in the
1.5 to 3 times normal level. Following an adequate therapeutic
response, the patient can be converted to coumarin for long-
term anticoagulation. The third choice is Argatroban, a syn-
thetic direct thrombin inhibitor, derived from L-arginine. Like
Lepirudin this is monitored by following A-PTT levels.

When heparin-induced thrombocytopenia is diagnosed,
heparin treatment should be reversed immediately with
protamine sulfate, and dextran 40 should be administered for
its antiaggre gating and rheologic effects. Warfarin therapy
also should be initiated and continued for several months.
In patients with arterial occlusive manifestations of heparin-
induced thrombosis, long-term warfarin therapy is recom-
mended because of the possibility of coexisting latent venous
occlusive disease.

The response of the platelet count to discontinuation of
heparin therapy is usually prompt, often resulting in thrombo-
cytosis, with a platelet count of 500000 to 6000000/mm 3
being achieved in several days.

Coagulation tests distinguish heparin-induced platelet ag-
gregation from other clotting disorders. The fibrinogen level
and prothrombin time are usually normal. The fibrin split
products level and prothrombin time are normal or slightly el-
evated. The sole patient in our series with a noticeable elevated
fibrin split products level was the initial patient, in whom the
diagnosis was not made antemortem. Heparin therapy was
not stopped, and before her death (caused by an intracerebral
hemorrhage), she had massive venous thrombosis involving
both upper and lower extremities, which resulted in an elevat-
ed fibrin split products level. Early identification of heparin-
induced thrombosis is necessary to minimize the catastrophic
complications of major limb amputation and death.

This experience suggests that it is imperative that all pa-
tients receiving heparin therapy have serial platelet counts
done from the fourth day of heparin therapy onward. It is our
policy to perform platelet counts every other day starting on
the fourth day of heparin therapy. If thrombocytopenia devel-
ops, platelet aggregation studies should be performed imme-
diately. With early recognition of complications, the mortality
and morbidity of major amputation can be prevented. Mor-
bidity and mortality rates reported in the literature vary from
22% to 61% and 12% to 33% respectively. 100 ' 101

Strategies for patients with heparin-induced
platelet aggregation

Patients who require subsequent heparin therapy for other
vascular or cardiac surgery procedures require special man-
agement. In patients in whom heparin-induced platelet aggre-
gation develops, the platelet aggregation tests usually revert
to normal from 6 weeks to 3 months. Preferably vascular or

pa rt I Vascular pathology and physiology

cardiac surgery procedures are delayed until these tests revert
to normal. We test the patient at 6 weeks and then every 2
weeks thereafter to determine when the platelet aggregation
tests are negative. When they are negative, the patient is then
admitted to the hospital for surgery. Cardiac catheterization or
angiography is done as required without the use of heparin
flush solutions, since even small amounts of heparin in the
flush solutions can stimulate the development of heparin-
induced antiplatelet antibodies. The vascular or cardiac
surgery procedure is then performed with the usual admin-
istration of heparin. At the conclusion of the procedure, the
heparin is reversed with protamine, and care is taken during
the postoperative period to ensure the patient does not re-
ceive heparin inadvertently through the flushing of either cen-
tral venous catheters or arterial lines. By using this procedure,
we have not had any difficulty with reexposure to heparin.

However, for those patients who require an additional vas-
cular or cardiac surgery procedure and who cannot wait until
the results from heparin-induced platelet aggregation tests are
negative, a different strategy is necessary. In patients requiring
procedures that can be done without the use of heparin, such
as resection of abdominal aortic aneurysm, heparin is not
used. However, in patients who require complex lower ex-
tremity revascularization or cardiopulmonary bypass, some
sort of anticoagulation is necessary. There are basically two
approaches. That favored by Laster, Elfrink, and Silver 102 in-
volves administering aspirin and dipyridamole (Persantine)
preoperatively and then using heparin for the operative proce-
dure as is customary. In addition to aspirin and dipyridamole,
we prefer to also use low-molecular-weight dextran, which
in addition to its rheologic properties coats the platelets and
interferes with platelet adhesion. In some patients, however,
as noted by Kappa et al., 103 the administration of aspirin has no
effect on heparin-induced platelet aggregation. Makhoul,
Greenberg, and McCann 104 noted that although aspirin
abolished platelet aggregation in 9 of 16 patients with heparin-
induced platelet aggregation, it only decreased platelet aggre-
gation in the remaining seven, suggesting that aspirin is not
able to reverse abnormal platelet aggregation in all patients.
Based on these reports, our procedure is to administer aspirin
and dipyridamole for several days before the operative proce-
dure. On the day of operation, the platelet aggregation tests are
performed with the addition of heparin. If the heparin causes
abnormal platelet aggregation, iloprost can then be used to
prevent heparin-induced platelet aggregation during the pro-
cedure. The use of iloprost can be complicated, particularly
since it is a very potent vasodilator, rather large doses of ad-
renergic agents are often required to support blood pressure.
Also it has been approved for this use by the FDA and may be
used off label.

Sobel et al. 105 reported an alternative technique in which pa-
tients received warfarin anticoagulant combined with dextran
as a means of preventing intraoperative thrombosis during
reconstruction. This is a reasonable alternative for peripheral

vascular reconstructions but is not possible for cardiopul-
monary bypass. In the future, different substances may be
available to allow for adequate anticoagulation. Makhoul,
Greenberg, and McCann 104 noted in vitro that heparinoids did
not cause platelet aggregation. These new anticoagulant
agents are being developed in Europe and may in the future be
available in the United States. Latham et al. have described the
use of recombinant hirudin for treatment of a patient who had
heparin-induced platelet aggregation who require cardio-
pulmonary bypass. 106 They were successfully able to anti-
coagulate the patient and place him on bypass without any
untoward results.

Cole and Bormanis 107 have reported the use of ancrod,
which is made from the venom of the Malaysian pit viper
(Agkistrudon rhodastoma), as an anticoagulant in patients who
have heparin-induced platelet aggregation. Ancrod acts enzy-
matically on the fibrinogen molecule to form a product that
cannot be clotted by physiologic thrombin.

Fibrinolytic dysfunction

Impaired fibrinolytic activity results in a hypercoagulable
state that can result in a spontaneous thrombotic event
or thrombosis in response to a minimal stimulus such as an
arterial puncture. A functional, genetically determined
abnormality in plasminogen has been described. 108-110 This is
detected as an abnormal plasminogen band on electrophore-
sis. This abnormal band is detectable in 10% of the population,
yet the incidence of the clinically significant hypercoagulable
state associated with it is much lower. The role of other factors
in the etiology of the prothrombotic state seen in symptomatic
patients with abnormal plasminogen electrophoresis is not
fully understood.

Abnormalities in the fibrinolytic system from an alteration
in the balance between plasminogen activators and inhibitors
have been reported. 111,112 A group of patients who had a
myocardial infarction at a young age were found to have a ge-
netic variation in the PAI-1 gene locus that appeared to influ-
ence the level of PAI-1, resulting in an inhibition of fibrinolytic
activity relative to control subjects. This variation is thought to
be important in the pathogenesis of myocardial ischemia. Im-
paired fibrinolysis due to changes in levels of tPA, PAI-1, or
both is described in the postoperative period and may con-
tribute to the development of postoperative deep venous
thrombosis. As noted, the release of tPA and PAI-1 from the en-
dothelial cell can be modulated by multiple factors, including
endotoxin and interleukin-1, which may explain the imbal-
ance and tendency to thrombosis seen in acute inflammatory
states and sepsis. 16 ' 17

Disseminated intravascular coagulation

Disseminated intravascular coagulation is the syndrome of
systemically activated coagulation and fibrinolysis that con-

chapter 3 Hemostasis and coagulation

sumes the platelets, fibrinogen, and coagulation factors, re-
sulting in a consumptive coagulopathy. Two forms of DIC are
described, an acute and a chronic form. The acute form is a pro-
found disturbance of the hemostatic process resulting from
the massive release of procoagulants such as tissue factor into
the circulation. Initiating circumstances can include sepsis,
major trauma, prolonged shock, or placental abruption. Surgi-
cal patients who have a chronic or low-grade DIC process be-
fore surgery can progress to fulminant acute DIC during or
after surgery. 113 ' 114 The diagnosis of acute DIC in a patient with
severe bleeding in the appropriate setting usually is not diffi-
cult. Laboratory confirmation is made by documenting deple-
tion of platelets and the coagulation-sensitive factors (factors
V and VIII, prothrombin, and fibrinogen). Activation of the
fibrinolytic system results in release of fibrin-fibrinogen split
products that inhibit platelet aggregation and fibrin polymer-
ization. Measurement of these products offers further
confirmation of ongoing DIC.

Chronic DIC is a more common clinical entity associated
with collagen diseases, autoimmune diseases, and malig-
nancy. Chronic DIC also has been described in 4% of patients
with extensive abdominal aortic aneurysms. 115 The diagnosis
of chronic DIC is more difficult because it often is a compensat-
ed state with no overt clinical bleeding and no significant
depletion of platelets or coagulation factors. A more specific
test of both thrombin and plasmin activity is the measure-
ment of fibrin D-dimer, which is formed when fibrinogen
is cleaved to fibrin and cross-linked before digestion by
plasmin. The release of the cross-linked D-dimer portion is
specific for plasmin degradation of fibrin but not fibrinogen,
therefore indicating ongoing intravascular coagulation and
fibrinolysis.

Therapy for DIC syndromes consists of correction of the
primary illness and replacement of plasma factors and
platelets. Anticoagulation or use of antiplatelet drugs have
been described in various settings, but have shown proven
value primarily in chronic DIC syndromes. 116

Recognition and management of
hypercoagulable states in patients with peripheral
vascular disorders

A thorough medical history remains the most important
means of identifying patients with potential hypercoagulable
disorders. The clinician must inquire directly about any his-
tory of previously unexplained thromboses in the patients or
family members. Patients with hypercoagulable syndromes
often report episodes of thrombophlebitis as young adults. Of
particular importance are those episodes of thrombophlebitis
that occurred in the absence of any associated risk factors such
as long bone fractures, or prolonged immobilization or bed
rest due to illness. Even more significant is a history of recur-
rent episodes of thrombophlebitis. Likewise, a history of arter-
ial thrombosis, particularly if it occurs at a young age, should

alert the clinician to the possibility of a hypercoagulable
state. 66 ' 83

To recognize patients whose initial manifestation of a
hypercoagulable state is at the time of their presentation with a
peripheral vascular problem requires the recognition of atypi-
cal manifestations of atherosclerotic disease and the skill to
predict the expected outcome or success of particular recon-
structions. Unusual or unexplained thrombosis such as a
thrombosed suprarenal aorta, upper extremity thrombosis, or
total tibial artery occlusion in a patient who is neither diabetic
nor has any evidence of any atherosclerotic occlusive disease
elsewhere, should alert the surgeon to look for hypercoagula-
ble disorders. Unusual appearance or patterns of atheroscle-
rotic disease seen on angiograms, such as occlusions seen in
one extremity when the other extremity has no evidence of any
disease, should trigger an investigation into the coagulation
system.

The role of screening vascular surgery patients for hyperco-
agulable states is difficult to ascertain. It has been reported that
9.5% of patients undergoing a variety of vascular surgery
procedures have abnormal laboratory test results indicating
potential hypercoagulability 117 The three most common
entities these patients demonstrated were heparin-induced
platelet aggregation, lupus anticoagulants, and protein C defi-
ciency. The incidence of inf rainguinal graft occlusion within 30
days was 27% among those patients who were in the hyperco-
agulable group, compared with 1.6% in those patients that
were not. Routine screening for the wide variety of hypercoag-
ulable states is not recommended; rather, a complete history
and clinical evaluation with selection of those patients who
warrant further laboratory evaluation probably is more effi-
cient and cost effective.

The most difficult experience for a vascular surgeon is to en-
counter an unexplained intraoperative thrombosis. Often, this
occurs during late evening or night-time hours when support
from the coagulation laboratory often is not available. If,
indeed, heparin has been given, the first step is to observe if
any clotting is present in the operative field and perform
laboratory evaluation of the heparin anticoagulant effect (i.e.
aPTT or activated clotting time) and measure a platelet count.
If the platelet count is higher than 100000/ml and there is no
prolongation of the clotting time, the problem is presumed to
be the anti thrombin system. The patient is then given 2 units of
fresh frozen plasma, with continued administration of 2 units
every 12 h for 5 days. Confirmation of an AT-III deficiency is
made in the postoperative period by measuring AT-III levels
just before the administration of a dose of fresh frozen plasma.
Patients with AT-III deficiency are maintained on long-term
warfarin therapy.

If the platelet count is below 100 000 /ml, it is presumed that
the patient has heparin-induced platelet aggregation. The
patient's history should be examined carefully to document
the administration of heparin at some time in the past. The ini-
tiation of antiplatelet therapy with dextran 40 (50-ml bolus fol-

pa rt I Vascular pathology and physiology

lowed by a continuous drip of 25 ml/h) and complete reversal
of heparin with protamine is accomplished. Postoperative
confirmation of the platelet aggregation abnormality is ob-
tained using the heparin-induced platelet aggregation tests
described in the section on heparin-induced thrombocytope-
nia. Patients with a confirmed diagnosis should receive war-
farin treatment for 3 weeks to 6 months after surgery. The
evaluation and therapy for patients with heparin-induced
thrombocytopenia who need subsequent exposure to heparin
also is described in the earlier section.

If the platelet count is above 100000/ml and the heparin
effect is present as manifested by a prolongation of the
appropriate clotting times, AT-III deficiency and heparin-
induced platelet aggregation are ruled out, and other hyperco-
agulable states such as fibrinolytic abnormalities, protein C or
S deficiencies, or lupus-type anticoagulants must be consid-
ered. It is important to draw blood specimens in the operating
room for subsequent testing before initiation of any therapeu-
tic measures to avoid erroneous results by the false elevation
of protein C or S or replacement of the patient's plasminogen
with autologous blood products. After blood specimens are
obtained, continuous heparin therapy is initiated and contin-
ued postoperatively, and fresh frozen plasma is administered
to treat potential protein C or S deficiencies or plasminogen
abnormalities. Long-term therapy is determined by the results
of the laboratory testing and any further clinical manifesta-
tions of hypercoagulation.

One of the problems in accurately diagnosing coagulation
abnormalities is that in the process of clotting, plasma factors
can be consumed and abnormalities may be the result of
clotting and not the cause of it. As noted, perioperative de-
creases in AT-III have been described that do not recover
until 7 days postoperatively 56 ' 118 This requires that any
intraoperative or perioperative tests positive for abnormal
coagulation be confirmed at 5-7 days and again at 1 month
before a patient is labeled as truly hypercoagulable. Most
patients who experience complications of hypercoagulable
states are placed on warfarin in the perioperative and
postoperative period. In patients with heparin-induced
platelet aggregation, this usually can be stopped in 3 months;
however, prolonged anticoagulation is recommended in pa-
tients with protein C and S deficiency, AT-III deficiency, and
plasminogen abnormalities because of the risk of recurrent
thrombosis.

Bleeding disorders

Perioperative abnormalities

Hemorrhagic disorders in surgical patients frequently are the
result of the dilution of platelets, although rarely they are
caused by dilution of coagulation factors during resuscitation
with large volumes of crystalloid or packed red cells. The

need for correction of the platelet and factor deficiencies
with platelet transfusions and fresh frozen plasma should
be determined objectively. Patients with platelet counts
greater than 60 000 /ml are not likely to benefit from platelet
transfusions unless there is a functional abnormality of the
platelets. Only a documented elevation in the coagulation
time justifies the use of fresh frozen plasma transfusions in
such circumstances.

Hypothermia is a frequent complication of resuscitation
and operations and is a common cause of perioperative coagu-
lopathy. Temperatures of less than 35°C are associated with
impaired coagulation despite the presence of normal levels of
coagulation proteins. 119 ' 120 A hypothermic patient's coagula-
tion times may be normal when tested in vitro at 37°C, yet hy-
pothermia may still impair in-vivo hemostasis. Vigilance at
maintaining the patient's core temperature is critical to avoid
this problem.

Particular to vascular operations, placement of a prosthetic
bypass graft or extensive endarterectomy can result in seques-
tration of platelets in the area of arterial reconstruction. This
also may result in systemic abnormalities in the hemostatic
process. 121 ' 122 This rarely is a significant problem requiring any
specific therapy. In instances where the patient has a preopera-
tive chronic DIC, however, further stimulation of the coagula-
tion process could result in progression to the full DIC
syndrome noted earlier. Rare anaphylactic reactions to place-
ment of a vascular graft have been reported, with abnormal
bleeding as an early finding and activation of the coagulation,
fibrinolytic, and kinin systems noted. 123 Other situations that
may result in a consumptive coagulopathy in a vascular
surgery patient include reperfusion of limbs after prolonged
ischemia, reperfusion of intestinal tissue after relatively short
periods of ischemia, and vascular trauma.

Coagulation factor deficiencies

Isolated coagulation factor deficiencies rarely are a vascular
surgical problem. The most common types of isolated factor
deficiencies are the inherited hemophilia A and B and type I
von Willebrand disease. Hemophilia A (classic hemophilia) is
inherited as an X-linked recessive deficiency of factor VIII. In
20% of the cases the deficiency is the result of a spontaneous
mutation. Hemophilia B (Christmas disease) is also an X-
linked recessive trait resulting in a deficiency of factor IX.
Hemophilia A occurs in 1 in 1 000 births and hemophilia B in 1
in 100000 births. Clinical manifestations of both hemophilias
include spontaneous hemarthrosis, epistaxis, hematuria, gas-
trointestinal bleeding, and intracranial hemorrhage. Bleeding
after minimal trauma and persistent bleeding also are charac-
teristic of hemophilia A and B. A variability in the severity of
these defects is a manifestation of the variability in the de-
crease seen in the level of the affected factor. Laboratory abnor-
malities seen with both hemophilias are a prolonged aPTT and
normal thrombin time, protime, and platelet function. Pre-

chapter 3 Hemostasis and coagulation

operative supplementation with factor VIII to near normal
levels and maintenance of factor VIII levels greater than 50% of
normal for 14 postoperative days are recommended for
patients with hemophilia A. 124/125 Prophylaxis of hemophilia
B patients consists of factor IX transfusion and vitamin K.
Perioperative and postoperative factor IX levels of 50% of
normal are considered adequate. 124

Deficiency of vWF, termed von Willebrand's disease (vWD),
occurs in a multitude of subtypes and may be transmitted as
an autosomal dominant or recessive trait. The different sub-
types are known to have different molecular defects. A domi-
nant inheritance pattern should result in 50% of the offspring
having the disease, yet owing to variable penetrance and ex-
pression, only 67% of the carriers of the gene for type I vWD are
symptomatic. In the recessive traits, a heterozygous state may
not manifest a clinical problem, yet a true homozygote or the
combination of various subtypes in a doubly heterozygous
state can result in a clinically significant bleeding disorder. The
true prevalence of vWD is not known because of the variable
clinical and laboratory manifestations of the disease, with
many cases probably unrecognized or misdiagnosed. The
incidence of the dominant form of vWD (type I) is approxi-
mately 1 in 10 000.

Clinical manifestations are widely variable, with the most
common being mucocutaneous bleeding. Epistaxis, easy
bruising, menorrhagia, and gingival bleeding are characteris-
tic. Abnormalities in bleeding time, aPTT, factor VIII coagulant
activity, vWF antigen, and ristocetin-induced platelet aggre-
gation may be manifest. Usually one or more of these test
results are abnormal; however, all may be negative in some
patients. One mechanism of the dysfunctional bleeding seen in
vWD relates to a decrease in factor VIII activity. An increased
turnover of factor VIII results from a decrease in the normal
vWF function of complexing with factor VIII and protecting
it from rapid clearance. Another manifestation of vWD is the
impaired adhesion of platelets to collagen and other platelets,
a function dependent on the normal binding of multimeric
vWF to collagen and the GPlb glycoprotein on the platelet sur-
face. This results in prolongation of the bleeding time.

Treatment of vWD is varied depending on the severity and
particular subtype of disease present. In general, the goal of
therapy is to normalize the factor VIII activity and the bleeding
time. Cryoprecipitate is the preferred source of vWF, although
fresh frozen plasma can be substituted in its absence. The
amount of cryoprecipitate given is empiric, and it is adminis-
tered until the coagulation time and bleeding time normalize.
It is recommended that therapy be continued for 7-10 days
after a major surgical procedure. 124 Alternative therapy to spe-
cific blood product replacement lies in the use of DDAVP, an
analogue of antidiuretic hormone. The mechanism of DDAVP
is not known, but it is presumed to increase the release of vWF
from endothelial cells. Subtypes of vWD in which there is no
normal vWF in the endothelial cells do not respond to DDAVP
therapy.

References

1. Ingerman-Wojenski C, Silver JM, Smith JB et al. Bovine endothe-
lial cells in culture produce thromboxane as well as prostacyclin.
/ Clin Invest 1981; 67:1292.

2. Goldsmith J, Neddleman SW. A comparative study of throm-
boxane and prostacyclin release from ex vivo and cultured
bovine vascular endothelium. Prostaglandins 1982; 24:173.

3. Watkins MT, Sharefkin JB, Maciag TM et al. Adult human saphe-
nous vein endothelial cells: assessment of their reproductive ca-
pacity for endothelial seeding of vascular prostheses. / Surg Res
1984; 36:588.

4. Mehta J, Roberts A. Human vascular tissue produce throm-
boxane as well as prostacyclin. Am } Physiol 1983; 244:R839.

5. Bagyalakshmi A, Frangos J A. Mechanism of shear-induced
prostacyclin production in endothelial cells. Biochem Biovhys Res
Commun 1989; 158:31.

6. Bush HJ. Favorable balance of prostacyclin and thromboxane A2
improves early patency of human in situ vein grafts. / Vase Surg
1984; 1:149.

7. Bush HL, McCabe ME, Nabseth DC. Functional injury of vein
graft endothelium. Arch Surg 1984; 119:770.

8. Eldor A, Facone DJ, Hajjar DP et al. Recovery of prostacyclin pro-
duction by de-endothelialized rabbit aorta: critical role of neo-
intimal smooth muscle cells. / Clin Invest 1981; 67:1292.

9. Radomski MW, Palmer, Mocada S. The anti-aggregatory proper-
ties of vascular endothelium: interactions between prostacyclin
and nitric oxide. Br J Pharmacol 1987; 92:639.

10. Cooper DR, Lewis GP, Lieberman GE et al. ADP metabolism
in vascular tissue, a possible thromboregulatory mechanism.
ThrombRes 1979; 14:901.

11. Lieberman G, Leake S, Peters TJ. Subcellular localization of
adenosine diphosphatase in cultured pig arterial endothelial
cells. Thromb Haemost 1982; 47:249.

12. Emms H, Lewis GP The anatomic distribution of ADP-ase
activity in the rabbit aorta. Artery 1982; 10:150.

13. Esmon CT. The regulation of natural anticoagulant pathways.
Science 1987; 235:1348.

14. Rosenburg RD, Rosenburg JS. Natural anticoagulant mecha-
nisms. / Clin Invest 1984; 74:1.

15. Rosenburg RD. Biochemistry of heparin antithrombin inter-
actions, and the physiologic role of this natural anticoagulant
mechanism. Am J Med 1989; 87(Suppl. 3B):2S.

16. Erickson LA, Schleef RR, Ny T et al. The fibrinolytic system of the
vascular wall. Clin Haematol 1985; 14:513.

17. Loskutoff DJ, Curriden SA. The fibrinolytic system of the vessel
wall and its role in the control of thrombosis. Ann NY Acad Sci
1990; 598:238.

18. Rodgers GM. Hemostatic properties of normal and perturbed
vascular cells. FASEB J 1988; 2:116.

19. Palombo JD, Blackburn GL. Endothelial cell factors and response
to injury. Surg Gynecol Obstet 1991; 173:505.

20. Krupski W, Thai ER, Gewertz BL. Endothelial response to venous
injury. Arch Surg 1979; 114:1240.

21. Guidoin R, Doyon B, Marios M et al. A SEM investigation of the
trauma to prostheses and arteries during vascular reconstruc-
tion. Artery 1980; 8:244.

pa rt I Vascular pathology and physiology

22. Manotovani A, Bussolino F, Dejana E. Cytokine regulation of
endothelial cell function. FASEB J 1992; 6:2591.

23. Crawford N, Scutton MC. Biochemistry of the blood platelets. In:
Bloom AL, Thomas DP, eds. Haemostasis and Thrombosis, 2nd edn.
Edinburgh: Churchill Livingstone, 1987:47.

24. Miletich JP, Jackson CM, Majerus PW. Interaction of coagulation
factor Xa with human platelets. Proc Natl Acad Sci USA 1977;
74:4033.

25. Walsh PN, Griffin JH. Contribution of human platelets to the pro-
teolytic activation of blood coagulation factors XII and XL Blood
1981; 57:106.

26. Ratnoff OD, Saito H. Surface-mediated reactions. Curr Topics
Hematol 1979; 2:1.

27. Colman RW. Surface-mediated defense reactions: the plasma
contact activation system. / Clin Invest 1984; 73:1249.

28. Saito H. Contact factors in health and disease. Semin Thromb
Hemost 1987; 13:36.

29. Furie B, Furie BC. The molecular basis of blood coagulation. Cell
1988; 53:505.

30. Mann KG, Jenny RJ, Krishnaswamy S. Cofactor proteins in the
assembly and expression of blood clotting enzyme complexes.
Annu Rev Biochem 1988; 57:915.

31. Weksler BB, Ley CW, Jaffe EA. Stimulation of endothelial cell
prostacyclin production by thrombin, trypsin, and the ionophore
A23187. / Clin Invest 1978; 62:923.

32. DeGroot PG, Gonsalves MD, Loesburg C et ah Thrombin-
induced release of von Willebrand factor from endothelial cells
is mediated by phospholipid methylation. / Biol Chem 1984;
259:13329.

33. Gelehrter TD, Synycer-Laszuk R. Thrombin induction of plas-
minogen activator-inhibitor in cultured human endothelial cells.
} Clin Invest 1986; 77:165.

34. Olexa S A, Budzynski AZ. Localization of a fibrin polymerization
site. } Biol Chem 1981; 256:3544.

35. Sakata Y, Aoki N. Cross-linking of oc 2 -plasmin inhibitor to fibrin
by fibrin-stabilizing factor. / Clin Invest 1980; 65:290.

36. Schartz ML, Pizzo SV, Hill RL et ah Human factor XIII from plas-
ma and platelets: molecular weights, subunit structure, pro-
teolytic activation and cross-linking of fibrinogen and fibrin. /
Biol Chem 1973; 248:1395.

37. Collen D. On the regulation and control of fibrinolysis. Thromb
Haemost 1980; 43:77.

38. Hanss M, Collen D. Secretion of tissue-type plasminogen acti-
vator and plasminogen activator inhibitor by cultured human
endothelial cells: modulation by thrombin, endotoxin and hista-
mine. J Lab Clin Med 1987; 109:97.

39. Ichinose A, Fugikawa K, Suyama T. The activation of prouroki-
nase by plasma kallikrein and its inactivation by thrombin. J Biol
Chem 1986; 261:3486.

40. Sprengers ED, Kluft C. Plasminogen activator inhibitors. Blood
1987; 69:381.

41. AokiN, Harpel PC. Inhibitors of the fibrinolytic enzyme system.
Semin Thromb Hemost 1984; 10:24.

42. Marcum JA, McKenney JB, Rosenburg RD. The acceleration of
thrombin-antithrombin complex formation in rat hindquarters
via naturally occurring heparin-like molecules bound to the
endothelium. J Clin Invest 1984; 74:341.

43. Stenflo J. Structure and function of protein C. Semin Thromb
Hemost 1984; 10:109.

44. Esmon CT. The roles of protein C and thrombomodulin in the
regulation of blood coagulation. /B/o/C/zem 1989; 264:4743.

45. Deykin D, Cochios F, DeCamp G et ah Hepatic removal of ac-
tivated factor X by the perfused rabbit liver. Am J Physiol 1968;
214:414.

46. Stenflo J, Fernlund P, Egan W et ah Vitamin K-dependent modifi-
cations of glutamic acid residues in prothrombin. Proc Natl Acad
Sci USA 1974; 71:2730.

47. Suttie JW. Oral anticoagulant therapy: the biosynthetic basis.
Semin Hematol 1977; 14:365.

48. Frick PG, Reidler G, Brogli H. Dose response and minimal daily
requirement for vitamin K in man. JAppl Physiol 1967; 23:387.

49. Uotila L, Suttie JW. Inhibition of vitamin K-dependent carboxy-
lase in vitro by cefamandole and its structural analogs. } Infect Dis
1983; 148:571.

50. Thaler E, Lechner K. Antithrombin III deficiency and throm-
boembolism. Clin Haematol 1981; 10:369.

51. Flinn WR, McDaniel MD, Yao JST et ah Antithrombin III deficien-
cy as a reflection of dynamic protein metabolism in patients
undergoing vascular reconstruction. / Vase Surg 1984; 1:888.

52. Towne JB, Bernhard VM, Hussey C et ah Antithrombin defi-
ciency: a cause of unexplained thrombosis in vascular surgery.
Surgery 1981; 89:735.

53. Lynch DM, Leff LK, Howe SE. Preoperative values and clinical
postoperative thrombosis: a comparison of three antithrombin
III assays. Thromb Haemast 1984; 52:42.

54. Chan V, Chan TK, Wong V et ah The determination of antithrom-
bin III by radioimmunoassay and its clinical application. Br J
Haematol 1979; 41:563.

55. Sas G, Blasko G, Banghogyi D et ah Abnormal antithrombin III
(antithrombin Budapest) as a cause of familial thrombophilia.
Thromb Diath Haemorrh 1974; 32:105.

56. Buller HR, TenCate JW. Acquired antithrombin III deficiency:
laboratory diagnosis, incidence, clinical implications and
treatment with antithrombin III concentrate. Am } Med 1989;
87:(Suppl.3B):44S.

57. Marciniak E, Gockeman JP. Heparin-induced clearance of circu-
lating antithrombin III. Lancet 1978; 2:581.

58. Conrad J, Lecompte T, Horrelou MH et ah Antithrombin III in
patients treated with subcutaneous and intravenous heparin.
Thromb Res 1981; 22:507.

59. Marciniack E, Farley CH, DeSimone PA. Familial thrombosis due
to antithrombin III deficiency. Blood 1974; 43:219.

60. Kitchens CS. Amelioration of antithrombin III deficiency by
coumarin administration. Am} Med Sci 1987; 293:403.

61. Hoffman DL. Purification and large-scale production of anti-
thrombin III. Am J Med 1989; 87(Suppl. 3B):23S.

62. Schwartz RS, Bauer KA, Rosenburg RD et ah Clinical experience
with antithrombin III concentrate in treatment of congenital and
acquired deficiency of antithrombin. Am } Med 1989; 87(Suppl.
3B):53S.

63. Griffin JH, Evatt B, Zimmerman TS et ah Deficiency of protein C
in congenital thrombotic disease. / Clin Invest 1981; 68:1370.

64. Broekmans AW, Velykamp JJ, Bertina RM. Congenital protein C
deficiency and venous thromboembolism: a study of three Dutch
families. N Engl} Med 1983; 309:340.

65. Tollefson DFJ, Bandyk DF, Towne JB et ah Protein C deficiency: a
cause of unusual or unexplained thrombosis. Arch Surg 1988;
123:881.

chapter 3 Hemostasis and coagulation

66. Eldrup-Jorgensen J, Flanigan DP, Brace L et ah Hypercoagulable
states and lower limb ischemia in young adults. / Vase Surg 1989;
9:334.

67. Bertina RM, Broekmans AW, Krommenhoek-van Es C et ah The
use of a functional and immunologic assay for plasma protein
C on the study of the heterogeneity of congenital protein C
deficiency. Thromb Haemost 1984; 51:1.

68. D'Angelo SV, Comp PC, Esmon ST et ah Relationship between
protein C antigen and anticoagulant activity during oral
anticoagulation and in selected disease states. / Clin Invest 1986;
77:416.

69. Mannucci PM, Vigano S. Deficiencies of protein C, an inhibitor of
blood coagulation. Lancet 1982; 2:463.

70. Conway EM, Bauer KA, Barzegar S et ah Suppression of
hemostatic system activation by oral anticoagulants in the blood
of patients with thrombotic diatheses. / Clin Invest 1987; 80:
1535.

71. Peterson CE, Kwann HC. Current concepts of warfarin therapy.
Arch Intern Med 1986; 146:581.

72. Kazmier FJ. Thromboembolism, coumarin necrosis and protein
C. Mayo Clin Proc 1985; 60:673.

73. Broekmans AW, Teepe RCG, von der Meer FJM et ah Protein C
and coumarin-induced skin necrosis. Thromb Res 1986; 6:137.
Abstract.

74. Comp PC, Nixon R, Cooper R et ah Familial protein S deficiency
is associated with recurrent thrombosis. / Clin Invest 1984;
74:2082.

75. Malm J, Laurell M, Dahlback B. Changes in the plasma levels of
vitamin K-dependent protein C and S and of C4b-binding pro-
tein during pregnancy and oral contraception. Br J Haematol 1988;
68:437.

76. D'Angelo A, Vigano-D'Angelo S, Esmon CT et ah Acquired
deficiencies of protein S: protein S activity during oral anti-
coagulation, in liver disease, and in disseminated intravascular
coagulation. J Clin Invest 1988; 81:1445.

77. Petri M, Rheinschmidt M, Whiting-O'Keefe Q et ah The frequen-
cy of lupus anticoagulant in systemic lupus erythematosus. Ann
Intern Med 1987; 106:524.

78. Espinoza LR, Hartman RC. Significance of the lupus anti-
coagulant. Am J Hematol 1986; 22:33 1 .

79. Shi W, Krilis SA, Chong BH et ah Prevalence of lupus anticoagu-
lant and anticardiolipin antibodies in a healthy population. Aust
NZJMed 1990; 20:231.

80. Greenfield LJ. Lupus-like anticoagulants and thrombosis. / Vase
Surg 1988; 7:818.

81. Gastineau DA, Kazmier FJ, Nichols WL et ah Lupus anticoagu-
lant: an analysis of the clinical and laboratory features of 219
cases. Am J Hematol 1985; 19:265.

82. Ahn SS, Kalunian L, Rosove M et ah Postoperative thrombotic
complications in patients with the lupus anticoagulant: in-
creased risk after vascular procedures. / Vase Surg 1988; 7:749.

83. Towne JB. Hypercoagulable states and unexplained vascular
thrombosis. In: Bernhard VM, Towne JB, eds. Complications in
Vascular Surgery, 3rd ed. St Louis: Quality Medical Publishing,
1991:101.

84. Elias M. Thromboembolism in patients with the "lupus"-type
circulating anticoagulant. Arch Intern Med 1984; 144:510.

85. Harris EN, Gharavi AE, Hughes GRV. Antiphospholipid anti-
bodies. Clin Rheum Dis 1985; 11:591.

86. Branch DW, Scott JR, Kochenour NK et ah Obstetric complica-
tions associated with the lupus anticoagulant. N Engl} Med 1985;
313:1322.

87. Babcock RB, Dumper CW, Scharfman WB. Heparin-induced im-
mune thrombocytopenia. N Engl J Med 1976; 295:237.

88. Baird RA, Convery RE Arterial thromboembolism in patients
receiving systemic heparin therapy. / Bone Joint Surg 1977; 59:
1061.

89. Bell WR, Romasulo PA, Alving BM et ah Thrombocytopenia oc-
curring during the administration of heparin. Ann Intern Med
1976; 87:155.

90. Fratantoni JC, Pollet R, Gralnick HR. Heparin-induced thrombo-
cytopenia: Confirmation of diagnosis with in vitro methods.
Blood 1975; 45:395.

91 . Nelson JC, Lerner RG, Goldstein R et ah Heparin-induced throm-
bocytopenia. Arch Intern Med 1978; 138:548.

92. Warkentin TE, Levina MN, Hirsh J et ah Heparin-induced
thrombocytopenia in patients treated with low molecular weight
heparin or unfractionated heparin. N Engl J Med 1995; 332:1330.

93. Towne JB, Bernhard VM, Hussey C et ah White clot syndrome.
Arch Surg 1979; 114:372.

94. Rhodes GR, Dixon RH, Silver D. Heparin-induced thrombocy-
topenia. Ann Surg 1977; 186:752.

95. Silver D. Heparin-induced thrombocytopenia. Semin Vase Surg
1988; 1:228.

96. Laster J, Cikrit D, Walder N et ah The heparin-induced thrombo-
cytopenia syndrome update. Surgery 1987; 102:763.

97. Kapsch DN, Adelstein EH, Rhodes GR et ah Heparin-induced
thrombocytopenia, thrombosis, and hemorrhage. Surgery 1979;
86:148.

98. Laster J, Silver D. Heparin-coated catheters and heparin-induced
thrombocytopenia. / Vase Surg 1988; 7:667.

99. Rosenthal F. Risk factors for venous thrombosis: Prevalence, risk
and interaction. Semin Hematol 1997; 34:171.

100. Silver D, Kapsch DN, Tsoi EKM. Heparin-induced throm-
bocytopenia, thrombosis, and hemorrhage. Ann Surg 1983; 198:
301.

101. Laster J, Cikrit D, Walker N et ah The heparin-induced thrombo-
cytopenia syndrome: An update. Surgery 1987; 102:763.

102. Laster J, Elfrink R, Silver D. Reexposure to heparin of patients
with heparin-associated antibodies. / Vase Surg 1989; 9:677.

103. Kappa JR, Fisher CA, Berkowitz HD et ah Heparin-induced
platelet activation in sixteen surgical patients: Diagnosis and
management. / Vase Surg 1987; 5:101.

104. Makhoul RG, Greenberg CS, McCann RL. Heparin-associated
thrombocytopenia and thrombosis: A serious clinical problem
and potential solution. / Vase Surg 1986; 4:522.

105. Sobel M, Adelman B, Szaboles S et ah Surgical management
of heparin-associated thrombocytopenia. / Vase Surg 1988; 8:
395.

106. Latham P, Revelis AF, Joshi GP, DiMaiio JM, Jessen ME. Use of re-
combinant hirudin in patients with heparin-induced thrombo-
cytopenia with thrombosis requiring cardiopulmonary bypass.
Anesthesiology 2000; 92:263.

107. Cole CW, Bormanis J. Ancrod: Apractical alternative to heparin. /
Vase Surg 1988; 8:59.

108. Aoki N, Moroi M, Sakata Y et ah Abnormal plasminogen: a he-
reditary molecular abnormality found in patients with recurrent
thrombosis. J Clin Invest 1978; 61:1186.

pa rt I Vascular pathology and physiology

109. Soria J, Soria C, Bertrand O et al. Plasminogen Paris I: congenital
abnormal plasminogen and its incidence in thrombosis. Thromb
Res 1983; 32:229.

110. Towne JB, Bandyk DF, Hussey CV et al. Abnormal plasminogen: a
genetically determined cause of hypercoagulability. / Vase Surg
1984; 1:896.

111. Dawson S, Hamsten A, Wiman B et al. Genetic variation at the
plasminogen activator inhibitor-1 locus is associated with al-
tered levels of plasma plasminogen activator inhibitor-1 activity.
Arterioscler Thromb 1991; 11:183.

112. Wiman B, Hamsten A. Correlations between fibrinolytic function
and acute myocardial infarction. Am } Cardiol 1990; 66:54G.

113. Bick RL. Alterations of hemostasis associated with surgery, car-
diovascular surgery prosthetic devices and transplantation. In:
Ratnoff OD, Forbes CD, eds. Disorders of Hemostasis. Philadel-
phia: WB Saunders, 1991:382.

114. Bick RL. Disseminated intravascular coagulation and related
syndromes: a clinical review. Semin Thromb Hemost 1988; 14:299.

115. Fisher DF, Yawn DH, Crawford ES. Preoperative disseminated
intravascular coagulation associated with aortic aneurysms: a
prospective study of 76 cases. Arch Surg 1983; 188:1252.

116. Ratnoff OD. Disseminated intravascular coagulation. In: Ratnoff
OD, Forbes CD, eds. Disorders of Hemostasis. Philadelphia: WB
Saunders, 1991:292.

117. Donaldson MC, Weinberg DS, Belkin M et al. Screening for

hypercoagulable states in a vascular surgery practice: a prelimi-
nary study. / Vase Surg 1990; 11:825.

118. McDaniel MD, Pearce WH, Yao JST et al. Sequential changes in co-
agulation and platelet function following femoro-tibial bypass. /
Vase Surg 1984; 1:261.

119. Reed RL, Bracey AW Jr, Hudson JD et al. Hypothermia and blood
coagulation: dissociation between enzyme activity and clotting
factor levels. Circ Shock 1990; 32:141.

120. Rohrer MJ, Natale AM. Effect of hypothermia on the coagulation
cascade. Crit Care Med 1992; 20:1402.

121. Wakefield TW, Shulkin BL, Fellows EP et al. Platelet reactivity in
human aortic grafts: a prospective randomized midterm study of
platelet adherence and release products in Dacron and ePTFE
grafts. / Vase Surg 1989; 9:234.

122. Mulcare RJ, Royster TS, Phillips LL. Intravascular coagulation in
surgical procedures on the abdominal aorta. Surg Gynecol Obstet
1977; 143:730.

123. Roizen MF, Rodgers GM, Valone FH et al. Anaphylactoid reac-
tions to vascular graft material presenting with vasodilation and
subsequent disseminated intravascular coagulation. Anesthesiol-
ogy 1989; 71:331.

124. Kasper CK. Hematologic care. In: Boone DC, ed. Comprehensive
Management of Hemophilia. Philadelphia: FA Davis, 1976:3.

125. Kasper CK, Boylen AL, Ewing NP et al. Hematologic manage-
ment of hemophilia A for surgery. JAMA 1985; 253:1279.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Molecular aspects of atherosclerosis

J. Jeff rey Alexander
John A. Moawad

The initiation of atherosclerosis is thought to involve penetra-
tion of the endothelial layer by circulating monocytes,
migration and proliferation of intimal smooth muscle cells,
and accumulation of lipids and matrix elements within
the subendothelial space. The pathogenesis of this disorder
is complex, and it has become apparent that factors such
as injury, hemodynamic stress, lipoprotein metabolism,
thrombosis, and inflammation all may play a role. The multi-
factorial nature of atherogenesis is further supported by
epidemiologic studies that have implicated hypertension,
smoking, hyperlipidemia, homocysteinemia, and diabetes
mellitus as causative agents. An improved understanding of
the mechanisms by which these seemingly disparate factors
contribute to the development of mature plaque will require
greater knowledge of the cellular and molecular processes in-
volved in plaque formation and maturation. This knowledge,
in turn, may lead to more specific and effective therapeutic
modalities.

The fatty streak is believed to represent the precursor lesion
of atherosclerotic plaque. It appears as a yellow, raised lesion
on the luminal surface of the arterial wall, and has been found
as early as the third year of life. Histologically, its most distin-
guishing feature is the presence of lipid-filled cells adjacent to
the endothelial layer. 1 These "foam cells" develop when circu-
lating lipoproteins infiltrate the endothelial barrier and are
avidly taken up by resident macrophages or myointimal
smooth muscle cells. Although free lipid may be found in asso-
ciation with the internal elastic lamina or entrapped within the
subendothelial matrix, most is intracellular. Collagen, elastin,
and large amounts of proteoglycan are also frequently present.
At this early stage, there is no apparent loss of cell viability, and
the endothelium remains intact. Platelet adherence is not evi-
dent, but circulating monocytes continue to migrate across the
endothelial layer and convert to macrophages. Although
spontaneous regression of these lesions has been reported, a
cascade of cellular events more often results in the develop-
ment of mature plaque. This progression is marked by cellular
necrosis within the central portion of the lesion, peripheral
smooth muscle cell proliferation, and the accelerated produc-

tion of matrix elements including collagen, elastin, and pro-
teoglycans by these cells. Release of intracellular lipid from
lysed foam cells leads to the deposition of free cholesteryl es-
ters within the subendothelial space. Inflammatory changes
also occur, with the histologic appearance of a granulomatous
reaction. Continuation of this process is characterized by frag-
mentation of the internal elastic lamina, endothelial loss,
medial thinning, and adventitial fibrosis. Later changes
include calcification, plaque degeneration and remodeling,
and thrombosis. There is increasing knowledge regarding the
molecular events that trigger and sustain this histologic se-
quence leading to plaque maturation and eventual arterial oc-
clusion. This chapter will focus on the molecular responses of
the arterial wall to factors which promote atherogenesis, and
the means by which these responses may be modulated
through the activity of both intracellular and intercellular
signaling systems.

Response to injury hypothesis

Although the morphologic evolution of atherosclerosis has
been described in detail, its pathophysiologic processes are
less clear. The most pervasive theory of atherogenesis is the
response-to-injury hypothesis introduced by Virchow and
more recently promoted by Ross. 2 This theory suggests that
endothelial damage induced by mechanical, chemical,
viral, rheologic, or immunologic injury can lead to increased
permeability of the endothelial barrier to circulating lipopro-
teins, and the subsequent adherence of monocytes and
platelets (Fig. 4.1). Numerous studies using animal models
have shown that denuding injury of the arterial intima in con-
junction with an atherogenic diet can reliably produce arterial
lesions characteristic of early plaque. 3 It is believed that expo-
sure of the subendothelial substrate enables the attachment
and activation of platelets, resulting in the elaboration of a va-
riety of mitogens, including platelet-derived growth factor
(PDGF), fibroblast growth factor (FGF), and transforming
growth factor (TGF)-p. These peptides can then stimulate and

pa rt I Vascular pathology and physiology

Injury

{mrcJionical "\
chemical >
Jmrnunctogic J

i-ljlnk^i

L«rt«yt#i

Figure 4.1 Injury model of atherosclerosis
demonstrating loss of endothelial integrity
resulting in platelet and leukocyte adherence
followed by mitogenic stimulation of smooth
muscle cell and monocyte migration, foam eel
formation and overlying thrombosis.

regulate proliferation, migration, and matrix element secre-
tion by local fibroblasts and smooth muscle cells, and so alter
both the structure and function of the arterial intima. Platelet
activation also may have a more direct effect on the endothelial
cell, influencing vasoreactivity and thrombogenesis, as well
as permeability to both circulating lipoproteins and cellular
elements including monocytes, which are recognized as
the progenitors of foam cells. Once formed, these lipid-filled
macrophages can then release substances such as the oxidized
products of lipid metabolism that can be toxic to the overlying
endothelial cells. Ultimately, migration and conversion of
smooth muscle cells to a synthetic phenotype results in the
production of collagen, fibronectin, and glycoproteins which
are necessary for tissue repair after injury but which, if not
controlled, can contribute to the formation of hyperplastic
atheroma.

It has become apparent that endothelial denudation is not
necessary to initiate these early atherogenic processes. In hu-
mans, increased lipoprotein uptake and fatty streak formation
have been observed in areas of intact or regenerated endothe-
lium suggesting that, while endothelial loss may be a feature
of more advanced plaque, it is not requisite to the formation of
early lesions. It has also been observed that alterations in en-
dothelial activity due to a wide array of chemoattractants in-
cluding inflammatory cytokines [interleukin (IL)-l, tumor
necrosis factor (TNF), interferon (IFN)], bacterial lipopolysac-
charides (LPS), superoxides, modified lipoproteins, mitogens
(PDGF, TGF-(3, FGF), thrombin, fibrinopetides, homocysteine,
matrix elements (collagen, elastin, fibronectin), and biome-
chanical forces can all promote the transendothelial migration
of monocytes and leukocytes. The cellular response can in-
clude changes of vascular reactivity and altered permeability,

enhanced monocyte recruitment with the accumulation of
foam cells, changes in cell growth regulation and survival, and
altered hemostasis.

In the absence of these factors, leukocytes do not normally
interact with the vascular endothelium, as their surface adhe-
sion molecules [LFA-1 (leukocyte function associated antigen)
and MAC-1] remain in a nonadhesive conformation. How-
ever, in regions of inflammation or injury, leukocytes tether
and roll along the endothelial surface (Fig. 4.2). This initial in-
teraction occurs via selectins— transmembrane glycoproteins
on endothelial cells that recognize carbohydrate ligands on
leukocytes, and whose surface expression appears to be in re-
sponse to atherogenic stimuli. 4 Selectins comprise a family of
three distinct molecules which mediate intercellular adhesive
interactions. E-selectins [i.e. ELAM-1 (endothelial-leukocyte
adhesion molecule)] are synthesized by activated endothelial
cells and leukocytes in response to cytokines IL-1 and TNF,
and allow the adherence of neutrophils, monocytes, and lym-
phocytes to the endothelium. P-selectins [PADGEM (platelet
activation-dependent granule external membrane protein)]
are produced by endothelial cells and megakaryocytes in re-
sponse to histamine, thrombin, and other secretagogues, and
are thought to initiate cell rolling across the endothelial sur-
face, while L-selectins, found on most leukocytes and en-
dothelial cells, are primarily responsible for cell tethering to
the endothelium.

Once tethering has occurred, leukocytes roll along the en-
dothelial surface, providing contact with chemokines in the
vicinity of the endothelial cell membrane. These peptides pro-
duce signals to enable leukocyte activation which, through L-
selectins, leads to further leukocyte recruitment. In addition,
this activation promotes the binding of (3 2 -integrins on the en-

chapter 4 Molecular aspects of atherosclerosis

Tethering Rolling

Adhesion Migration

Integrins and IgG superfamily

Figure 4.2 Tethering and rolling of circulating
mononuclear leukocytes along the endothelial
surface in response to the expression of
adhesion molecules. This results in the
transendothelial migration of these cells and
their conversion to macrophages which
elaborate cytokines and growth factors,
establish an oxidative environment through the
production of superoxides, and ultimately form
lipid-laden foam cells.

Ox-LDL

Macrophage

PDGF
\* FGF

» TT 1

Foam Cell

JL-l
TNF

SMC

dothelial cell surface to ligands of the leukocyte cell mem-
brane, such as ICAM-1 (intercellular adhesion molecule-1),
VCAM-1 (vascular cell adhesion molecule-1), and ELAM-1.
These provide shear-resistant attachments both to the en-
dothelium and to the underlying matrix elements, including
collagen, laminin, and fibronectin. Surface integrins may also
have a role in cellular signaling, which is the primary means by
which extracellular stimuli result in alterations of cellular
function and growth, gene expression, and apoptosis. Cell
attachment through integrin activity has been shown to lead
to active phosphorylation at focal adhesion sites by means of
protein kinases, including focal adhesion kinase ppl25FAK,
which can activate signal transduction pathways. 5

Integrins are heterodimeric molecules containing coval-
ently bound a and (3 subunits, with functionally different
subfamilies based on the configuration of the (3 subunit. (3 1 -
integrins [VLA (very late appearing antigens)] appear to me-
diate cell adhesion principally to extracellular matrix proteins
including collagen, laminin, and fibronectin. p 2 -integrins
(LFA-1, Mac-1, pl50) are exclusive to leukocytes and include
ICAM-1, VCAM-1, and PECAM-1 (platelet-endothelial cell
adhesion molecule). These molecules interact with the im-
munoglobulin superfamily, which contains corresponding
ligands on the endothelial cell. Finally, the p 3 -integrin subfam-
ily includes cytoadhesion molecules such as Gllb-IIIa found
on megakaryocytes and platelets, and which are operative in
fibrinogen binding, as well as platelet adhesion and aggrega-
tion. Integrins are essential for the transendothelial migration
of leukocytes, as occurs in the course of endothelial injury, and
their presence has been noted at sites of atherosclerotic lesion
formation (Fig. 4.3). Integrin activity can be stimulated by
lysophosphatidylcholine (lyso-PC), a major component of
atherosclerotic plaque, as well as by cytokines and hemody-

Figure 4.3 Transmission electron micrograph demonstrating monocyte
movement across an endothelial monolayerwith pseudopod extension
above the basal lamina (arrow). Bar= 1 .1 jam. (From Migliorski G, FolkesE,
Pawlowski N, Cramer EB. In vitro studies of human monocyte migration
across endothelium in response to leukotriene B4andf-met-leu-phe./\/77j
Pathol 1987; 127: 157, with permission.)

pa rt I Vascular pathology and physiology

namic forces via gene transcription. 6 Similarly, integrins may
be negatively regulated by nitric oxide (NO), which inhibits
transcription through NFkB (nuclear factor kappa B) inhibi-
tion. Penetration of the arterial intima by mononuclear leuko-
cytes is now seen as a defining event in the formation of early
atheroma. While the exact mechanism of leukocyte transmi-
gration is not known, it appears to involve chemokines [i.e.
MCP-1 (monocyte chemotactic peptide)], p : - and p 2 -integrins,
and leukocyte interaction with PECAM-1 at the endothelial
cell junction. This is believed to result in the disruption of tight
junction adhesion molecules through the synthesis and re-
lease of proteolytic enzymes including elastase, collagenase,
and plasminogen activator, thereby influencing the perme-
ability of the endothelium to macromolecules and leukocytes.
The activation and migration of leukocytes has been shown to
result in the release of lysosomal and granule contents, and in
the "respiratory burst" production of reactive oxygen meta-
bolites. This, in turn, stimulates the secretion by resident
macrophages of cytokines, growth factors, proteases, lipases,
coagulation and complement factors, and reactive oxygen in-
termediates which markedly change the subendothelial envi-
ronment, affect the function of the vascular cells exposed to
these factors, and further propagate this effect through mono-
cyte chemotaxis. This respiratory activity can be further
enhanced by exposure to inflammatory cytokines including
TNF-oc, IFN-y, and GM-CSF (granulocyte-monocyte colony-
stimulating factor).

Shear effect

Hemodynamic forces, including vascular wall shear stress
(frictional force), hydrostatic pressure, and cyclic strain, have
been shown to alter endothelial cell function and structure.
This has been associated with changes in intimal permeability
and lipoprotein accumulation within the subendothelial
space, endothelial damage and repair, and the expression of
adhesion molecules, growth factors, matrix components, va-
soactive mediators, and fibrinolytic peptides by these cells. Al-
though increased shear was initially thought to be atherogenic
owing to shear-related endothelial injury, subsequent studies
have failed to demonstrate intimal disruption or other histo-
logic evidence of injury. It is now believed that plaque forma-
tion may occur preferentially at sites of low shear due to
increased expression of adhesion molecules, increased
platelet adherence, reduced NO production, increased low-
density lipoprotein (LDL) uptake, and altered smooth muscle
cell (SMC) response. 7 Cellular reaction to hemodynamic forces
can be acute and of limited duration, occurring through ion
conductance, inositol triphosphate (ITP) generation, G-
protein activation, and cytosolic calcium flux. These are simi-
lar to early signaling responses generated by agonist-receptor
binding, and this similarity suggests a common pathway
of mechanical and chemical excitation. 8 More delayed and

sustained responses are dependent on protein synthesis,
which requires the transcriptional upregulation of gene ex-
pression for the elaboration of cellular adhesion molecules,
growth factors, and fibrinolytic peptides. It has been shown
that spatial gradients in shear stress do alter transcriptional
factors [shear stress response elements (SSRE)], including
NFkB, AP-1, Egr-1 (early growth response protein 1), and the
early response genes c-jun, c-myc, and c-fos. 9 These factors
have been specifically linked to the transcription of eNOS
(endothelial derived nitric oxide synthase), COX-2 (cyclooxy-
genase enzyme 2), and Mn-SOD (manganese-dependent
superoxide dismutase), which influence vasoreactivity,
arachidonate metabolism, and oxidative activity. 10

Cellular signaling

There are probably several means by which shear or other ex-
ternal stimuli (i.e. cytokines, growth factors) induce cellular
responses, including the stimulation of cell membrane-based
mechanotransducer elements and the initiation of intracellu-
lar signaling through receptor-ligand binding. 11 ' 12 Second
messenger signaling systems represent the biochemical path-
ways which result in a cellular response. Initial signaling relat-
ed to environmental factors ("outside-in" signaling) involves
the binding of ligand to receptor resulting in the primary acti-
vation of second messenger proteins such as G-proteins or
tyrosine kinase to yield an effect. A multiplicity of pathways
exists to allow variation in response. In addition, the cell can
further regulate its response through "inside-out" signaling,
in which the translation of intracellular signals can alter recep-
tor number or affinity.

A primary signaling pathway involves fluctuations in intra-
cellular calcium levels which can occur rapidly through the re-
lease of Ca 2+ from within intracellular organelles such as the
endoplasmic reticulum via inositol triphosphate-activated
channels or ryanodine receptor-like channels. Alternatively,
more sustained responses may be dependent on calcium flux
through the cytoplasmic membrane, mediated by voltage-
dependent Ca 2+ channels (primarily "L-type"), by receptor-
mediated channels, and by Na 2+ -Ca 2+ exchange. Shear stress
has been shown to increase intracellular calcium relative to the
magnitude and the type of the flow stimulus, i.e. laminar, os-
cillatory, or pulsatile. 13 An early rise in cytosolic Ca 2+ can me-
diate NO release from the endothelial cell (EC) through the
serine /threonine kinase regulation of ecNOS and through
stimulation of ecNOS gene expression. This can have multiple
effects on the function of the endothelium.

Calcium-regulated signaling depends, as well, on the func-
tion of the plasma membrane pump (PMCA), which removes
excess calcium from the cytoplasm and restores homeostasis.
This pump has been shown to be mediated by protein kinase C
(PKC) through active phosphorylation of the pump protein
and through control of pump protein mRNA activity. 14 PKC,

chapter 4 Molecular aspects of atherosclerosis

pip

PLC

Figure 4.4 Inositol phosphate pathway. Receptor (R) activation of
regulatory G protein (G) stimulates the conversion of inositol diphosphate
(PIP) to inositol triphosphate (IP 3 ) and diacylglycerol (DAG) through
phospholipase C (PLC). DAG converts protein kinase C (PKC) to its active
form, which then promotes protein phosphorylation. IP 3 may stimulate Ca 2+ -

PKCj — ► PKC a — ► I

1?^ DAG

X-P*

mediated signal transduction. (Adapted from Stadler J, Simmons RL.
Molecular basis of cell signaling: physiology at the cellular level. In: Simmons
RL, Steed DL, eds. Basic Science Review for Surgeons. Philadelphia: WB
Saunders, 1 992: 1 1 , with permission from Elsevier.)

Signal

ip-

Ca +2

>PKC

Cell Response

(+/-)

Signal

> cAMP ► PKA

Cell Response

Figure 4.5 Interactions of the protein kinase C (PKC) and adenylate cyclase
(cAMP) signaling systems. Both can be modulated by Ca 2+ or by each other to
determine the cellular response. (Adapted from Nishizuka Y. The role of

protein kinase C in cell surface signal transduction and tumour production.
Nature 1984; 308:693.)

an end-product of the inositol phosphate pathway, appears to
be regulated by G-proteins in response to external ligand-re-
ceptor binding and other cell-stimulus interactions such as
shear (Fig. 4.4). A variety of PKC isoforms can determine the
nature of the response. These include conventional subspecies
cPKC (a, (3 1 , (3 2 , y) which are activated by Ca 2+ , diacylglycerol,
phosphatidylserine, cis-unsaturated fatty acid and lyso-PC.
Other subspecies are novel PKC (nPKC: 5, e, r|, 9) and atypical
PKC (aPKC: cj, X). These lack a calcium binding domain but are
associated with other stimuli. The activation of PKC can occur
through phospholipase C and diacylglycerol (DAG), which
has been linked to delayed cellular responses such as prolifer-
ation and differentiation. Alternatively, activation of phos-
pholipase A may modulate transient and sustained responses
through the activation of different subspecies of PKC. PKC is
known to result in downstream signaling through MAP (mi to-
gen-activated protein)-kinase activity. Included in this cas-
cade are ERK 1/2 (extracellular signal-related kinases), a
calcium-independent kinase which can affect other protein ki-
nases (i.e. p90rsk, MAPKAP, c-raf, MEK), transcription factors

{c-myc, c-jun, c-fos, p62TCF), and cell surface substrates (EGF-
R, cPLA 2 ). Additional MAP kinases which may be involved in
signal activation are stress-activated protein kinases (SAPK or
JNK), p38, and Big MAPK 1 (BMK 1), although their activities
are less well known. 7 These effects may result in alterations of
such varied functions as endothelial permeability to leuko-
cytes and macromolecules, and smooth muscle cell contrac-
tion, proliferation, differentiation, and secretion.

A second major signaling pathway is the adenylate cyclase
(cAMP) messenger system. This is activated following G-
protein-coupled receptor binding, with the release of the a
from the (3y subunit of Gs. The a subunit activates the adeny-
late cyclase effector, resulting in the increase of cAMP which,
in turn, increases activity of protein kinase A (PKA) with the
subsequent phosphorylation of cytosolic, membrane, and nu-
clear substrates. This pathway may interact directly with the
PKC pathway to modulate the cellular response further
(Fig. 4.5). Activation of cAMP has been associated with smooth
muscle cell relaxation through myosin light chain kinase phos-
phorylation, reduced cell proliferation and migration, in-

pa rt I Vascular pathology and physiology

i nes

Shear
Stress

Leukocyte Platelet

^ cyclase

GTP -
/

phosphodiesterase

#>cGMP

rotcin
inase

Relaxation

SMC

Figure 4.6 Stimulation of nitric oxide
synthase (NOS) by shear stress and cytokine
activity, resulting in the production of nitric
oxide (NO) which inhibits leukocyte and platelet
adhesion and promotes smooth muscle cell
relaxation through cyclic guanosine
monophosphate (cGMP).

creased cellular differentiation and altered lipoprotein metab-
olism. Similarly, NO provides a means of intracellular com-
munication leading to a defined cellular response 15 (Fig. 4.6).
Phospholipase C activation and calcium influx can result in
early and rapid NOS production of NO, while G-protein bind-
ing can lead to a later and more sustained release of NO
through MAP kinase activity. NO binds to guanyl cyclase to in-
crease cyclic guanosine monophosphate (cGMP) which, in
turn, can alter gene transcription through NFkB. 16 NO has
been shown to be the molecule principally responsible for the
antiatherogenic properties of the endothelium, as it reduces
the expression of surface adhesion molecules, increases en-
dothelial cell proliferation, inhibits smooth muscle cell repli-
cation, reduces monocyte and platelet adhesion, suppresses
the oxidation of LDL, and promotes vasodilatation.

Lipoprotein processing

Hyperlipidemia is a recognized causative factor in epidemio-
logic as well as experimental studies of atherosclerosis. Both
LDL, the predominant carrier of cholesterol to peripheral cells,
and very-low-density lipoprotein, a carrier of triglycerides
and a progenitor of LDL, have been implicated in plaque for-
mation. More recently, lipoprotein particles such as lipopro-
tein^) have also been shown to be highly atherogenic. In
contrast, high-density lipoprotein (HDL) appears to have a
negative correlation with the development of atherosclerosis.
The synthesis, conversion, and catabolism of lipoproteins is
regulated by an array of transfer proteins, catalytic enzymes,

and cell surface lipoprotein receptors that are under genetic
control. Genetic abnormalities affecting cholesterol transport,
uptake, storage, and clearance have been identified, and can
be separated into disorders that primarily alter lipoprotein
formation and cholesterol transport [i.e. abetalipoproteine-
mia, Tangier disease, lecithin cholesterol acyl transferase
(LCAT) deficiency], and those that affect lipoprotein catabo-
lism (i.e. familial hypercholesterolemia, familial dysbetal-
ipoproteinemia, Wolman disease, cholesteryl ester storage
disease). 17 Genetic variation of these control mechanisms can
lead to elevated levels of serum cholesterol and its carrier
lipoproteins. Although it is clear that the detrimental effects of
hyperlipoproteinemia are modulated by other risk factors
such as smoking, hypertension, and diabetes, it is also appar-
ent that cholesterol is the primary agent of plaque formation,
and is found to constitute 70% of the total lipid found in ma-
ture atherosclerotic plaque.

In 1976, Brown and Goldstein 18 described the means by
which cells regulate the uptake and utilization of cholesterol.
Through the activity of high-affinity LDL receptors on the
plasma membrane, LDL is taken up by the cell through
the process of endocytosis. It is then hydrolyzed within
the membrane-bound lysosomes (Fig. 4.7). Free cholesterol
is transported directly to the plasma membrane or, if not
required, is stored as an ester after conversion by acyl-
coenzyme A cholesterol acy transferase. Cholesterol is also
endogenously produced in the endoplasmic reticulum via
hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase,
the rate-limiting step in the synthetic pathway. The plasma
membrane therefore acts as the end-point of both intrinsic and

chapter 4 Molecular aspects of atherosclerosis

Figure 4.7 LDL receptor pathway. (Adapted
from Brown MS, Goldstein JL. How LDL
receptors influence cholesterol and
atherosclerosis. SciA m 1984; 251 ;61.)

— TS55SSJ5SSJ5SJ'SSD^

i* HX^SrtiJiXXWOOOOtM :

is JfKJWWOiXiOOWWOtK ;

Golgi

LDL

LDL
Receptor

g& Cholesteryl Ester
1111 Apoprotein B

Coated Pit

Coated Vesicle

Cholesterol
Synthesis

Receptor . \

Synthesis r ^

DNA V Ch

;g coa

Reductase

Cholesterol

Inhibits

ACAT

Endoplasmic Reticulum

Cholesteryl Ester

Lysosome

Amino Acids

Cell Membrane

extrinsic free cholesterol production. As free cholesterol con-
centration rises in the plasma membrane, feedback inhibition
of both intrinsic synthesis (via HMG-CoA reductase) and ex-
trinsic cholesterol uptake (via downregulation of LDL recep-
tor synthesis) occurs. Free cholesterol may also be extracted
from the cellular membrane through adsorption to an acceptor
molecule, such as HDL. Once it is taken up by the HDL mole-
cule, it can either be transferred directly to other plasma accep-
tor molecules by means of specific transfer protein, or be
transported directly to the liver for catabolism and excretion.

Although the study of genetic abnormalities, particularly
those involving the LDL receptor pathway, has yielded insight
into lipoprotein metabolism, it is clear that spontaneous
human atherosclerosis is rarely a monogenic disease. Geneti-
cally aberrant lipoprotein transport and receptor regulation
may have little significance in plaque formation. This is evi-
denced by the natural history of homozygous familial hyper-
cholesterolemia or its experimental analogue, the Watanabe
heritable hyperlipidemic (WHHL) rabbit, where the absence
or dysfunction of the LDL receptor and of receptor-mediated
LDL uptake fails to prevent lipid uptake and plaque formation
within the arterial wall. This observation has argued for the
presence of an alternative means of uptake by the arterial
intima that is not receptor dependent, and which may not be
subject to normal feedback control.

Endothelial permeability

Vascular permeability is controlled primarily by endothelial
cellular junctions, including tight junctions, gap junctions, ad-
herens junctions, and complex adherens junctions. Addition-

ally, molecules are present which influence junctional perme-
ability including occludin, promoting the attachment of cells
in tight junctions, connexons, to facilitate the passage of ions
and small-molecular-weight molecules, cadherins, to pro-
mote calcium-dependent intercellular adhesion, and PECAM,
whose role remains obscure but which appears to be necessary
for the migration of leukocytes. 19 The mechanisms for opening
of the intercellular junctions are not clear, but it is clear that the
cells have the ability to change the nature of these junctions
rapidly to alter permeability to plasma particles and circulat-
ing leukocytes. Retraction and separation of the endothelial
cells through alteration of their intercellular connections is
also thought to expose macrophages and matrix molecules to
circulating blood cells, allowing interaction with platelets,
with adherence and degranulation resulting in the release of
growth factors to promote a fibroproliferative response. 4 Al-
though the transport of LDL to the subendothelial space is be-
lieved to occur predominantly by means of energy-dependent
transcellular transport, the permeability of the endothelial
barrier can be increased by these paracellular routes in re-
sponse to external stimuli. The accessory channels provided
by widening of intercellular bridges permits the bulk move-
ment of macromolecules across the vascular endothelium
despite the absence of frank endothelial denudation.
Experimental models indicate that endothelial permeability
may increase significantly owing to leaky intercellular junc-
tions as cell turnover increases. This observation would be
consistent with enhanced lipid infiltration seen at sites of
hemodynamic stress or in conjunction with nondenuding en-
dothelial injury. A similar endothelial leak phenomenon has
been found in response to oxidants, lipoproteins, platelets,
thrombin, and a variety of vasoactive peptides, where it has

pa rt I Vascular pathology and physiology

been attributed to reversible changes in cell morphology due
to reconfiguration of the calcium-dependent F-actin cy-
toskeleton. Alterations in both cytosolic and extracellular cal-
cium with calcium antagonists, calcium channel blockade, or
chelators have led to changes in the permeability of endothe-
lial monolayers to macromolecules, and have influenced the
development and progression of atherosclerosis in experi-
mental and clinical studies. 20

There is evidence that, through calcium-dependent regula-
tion, PKC may modulate the effect of extracellular mediators
on macromolecular permeability through the endothelium.
This might occur through PKC-induced phosphorylation of
specific cytoskeletal proteins (vinculin, vimentin, actin, and
myosin light chain) to alter intercellular contacts and induce a
transient disruption of endothelial junction complexes. 21 It
has been shown that PKC and cAMP-dependent protein ki-
nase A may have opposing effects on endothelial permeabil-
ity 22 In addition, reduced cAMP levels have been associated
with lipid-laden arterial lesions, leading to the supposition
that cAMP may influence the process of lipid accumulation
within the arterial wall. These studies strongly suggest that
endothelial junctions are important to the integrity of the
endothelial barrier. Alterations of paracellular transport
mechanisms through the activity of inositol phosphate, cAMP,
and Ca 2+ second messenger systems may occur in response to
such external stimuli as vasoactive peptides, thrombogenic
agents, and hemodynamic stress, or to the release of specific
mitogens or cytokines. 23 LDL may directly recruit these intra-
cellular signaling systems and, through their activity, modify
endothelial cell structure, lipoprotein uptake and metabolism,
and mitogen secretion. In so doing, LDL may secondarily
affect functional changes in adjacent smooth muscle cells or
macrophages, which are integral to the pathophysiologic
process of atherosclerotic plaque.

Abnormal lipoprotein processing

Although it is possible that increases in endothelial permeabil-
ity due to exogenous stimuli may contribute to direct lipopro-
tein infiltration of the subendothelial space, it has become
apparent that native LDL is unable to enhance smooth muscle
cell and monocyte migration or secretion, which are associ-
ated with early plaque formation. Perhaps more important,
LDL is not rapidly ingested by monocyte-derived
macrophages to form foam cells. This latter finding was
addressed by Goldstein and colleagues, 24 who demonstrated
that chemical modification of the LDL molecule by acetylation
or malondialdehyde modification could lead to its rapid up-
take by macrophages in vitro, yielding foam cells. Since then,
other forms of modification, including oxidation, glycosyla-
tion, and aggregation have been shown to have a similar effect.
Modification of LDL can occur through its incubation with
endothelial or smooth muscle cells. Subsequent uptake by

monocytes is increased 3- to 10-fold over that of native LDL.
This uptake can be inhibited by chemically acetylated LDL, in-
dicating the presence of either a single or shared receptor.
Preincubation of LDL with cultured monocytes has yielded
the same result, enabling this cell to augment its own uptake of
lipoprotein. The receptor for this enhanced uptake, the scav-
enger receptor, appears to display both specificity and satura-
bility and, by its activity, is thought to be responsible for the
formation of foam cells. It can be found in homozygous famil-
ial hypercholesterolemia as well as in the WHHL rabbit model,
attesting to the fact that it is functionally independent of the
normal LDL receptor. Distinct scavenger receptors have been
isolated. Macrophages have been found to have at least six
membrane proteins including class A receptors, which recog-
nize the oxidized apoprotein portion of the lipoprotein parti-
cle, and class B receptors (CD36, SRBI) which bind to the lipid
moiety 25,26 CD36, unlike the classic LDL receptor, is found to
be upregulated by LDL through activation of the transcription
factor PPAR-y (peroxisome proliferator activated receptor-y),
such that the binding and uptake of Ox-LDL perpetuates both
lipid accumulation and receptor expression. The endothelial
cell also has scavenger receptors, although these are different
from those of the macrophage. 27 One such receptor, LOX-1
(lectin-like oxidized LDL receptor-1), is a membrane glycopro-
tein which can bind, internalize, and degrade oxidized LDL.
Its activity can be upregulated by TNF-oc and by shear stress,
and there is evidence that this receptor may also be involved in
the engulfment of apoptotic cells. 28 Although this receptor
does not result in massive uptake of lipoprotein by the cell as it
does in the macrophage, it has been shown to result in
endothelial dysfunction, with impaired production of NO,
increased leukocyte adhesion molecule expression, and
increased growth factor release. 6,29

There is no direct evidence that acetylation of LDL occurs in
vivo. However, there is considerable evidence that oxidation of
LDL represents its biologic equivalent. 30 Oxidized LDL can be
extracted directly from atherosclerotic lesion. In addition, im-
munohistochemical staining of human and animal plaque has
shown the presence of antigens similar to oxidized LDL. There
also has been documentation of circulating autoantibodies to
oxidized LDL in lesion-prone humans and animals. Finally,
use of the antioxidant, probucol, can lead to the stabilization
and regression of induced plaque in animal models. 31 Al-
though these findings do not confirm the pathophysiologic
role of oxidized LDL in atherogenesis, they do provide strong
support for its occurrence. Oxidation of LDL is characterized
by loss of cholesterol ester, hydrolysis of phosphatidylcholine,
and modification of its lysine residues. These lysine residues
are essential to receptor recognition of LDL, and modification
of 5-10% of these groups has been shown to be sufficient to
prevent the cellular binding and uptake of LDL through the
LDL receptor pathway. Acetylation of LDL also affects its ly-
sine groups, and acetylated LDL can compete with oxidized
LDL for uptake via the scavenger receptor.

chapter 4 Molecular aspects of atherosclerosis

Action of cellular
Oxygenase (lipoxygenase)

Cellular lipids

LDL lipids

Generation of reactive
oxygen in the cell

Release into the medium

Transfer of oxidized
cell lipids to LDL

Figure 4.8 The generation and action of
oxidized lipids leading to the formation of foam
cells. (Adapted from Steinberg D,
ParathasarathyS, CarewTE, KhooJC, Witzum
JL. Beyond cholesterol: modifications of low
density lipoprotein that increase its
atherogenicity. N Engl J Med 1 989; 320:91 5.)

Generation of LDL-con taming
oxidized lipids

Breakdown of lecithin to lysolecithin
Propagation of peroxidation

Degradation of apo-protein B

Conjugation of oxidized fatty
acids with apo-protein B

Generation of new epitopes on apoB
recognized by macrophage receptors

Foam Cells

In vivo, oxidation is believed to occur, in part, through the
peroxidation of fatty acids, of which polyunsaturated fats are
the most susceptible. This process may depend on low concen-
trations of iron or copper, and can be inhibited by metallo-
chelators such as EDTA, or by natural or chemical
antioxidants such as vitamin E or probucol. Potential path-
ways for lipid oxidation could include oxidation within the
cell, with transfer of oxidized lipid subspecies to the surround-
ing medium, cellular modification of intracellular or mem-
brane lipids, or indirect lipid oxidation through the cellular
release of superoxides (Fig. 4.8). Cell-induced oxidation prob-
ably occurs through the lipoxygenase rather than the cy-
clooxygenase system since exposure of LDL to phospholipase
A 2 can mimic modification induced by endothelial cells, and
inhibition of lipoxygenase activity can effectively block cell-
induced oxidative modification of LDL. 32 Neither aspirin nor
indomethacin is effective in blocking this reaction. Once this
process is initiated, its amplification and propagation are
made possible through a peroxidation chain reaction that gen-
erates additional free radicals. The creation of an oxidative en-
vironment can have additional effects on cellular function.
Although it is likely that LDL in plasma is relatively free from
oxidative metabolism owing to the presence of natural antiox-
idants within the plasma and within the lipoprotein molecule
itself, this protection may be self-limited. After the entrapment
and metabolic alteration of LDL in the subendothelial space,
such protective mechanisms may be inactive or inadequate to
control this reaction. The potential significance of lipoprotein
oxidation within the arterial wall is related to its effect on cell-
ular structure, function, and viability. Oxidized LDL has cyto-
toxic properties that can affect the endothelial and smooth
muscle cell as well as the resident macrophage. Rapid uptake
of oxidized LDL by the macrophage can result in the release of

a variety of cellular toxins, including the oxidized remnants of
lipoprotein metabolism. This may then lead to a loss of the
overlying endothelial layer and to the adhesion and activation
of circulating platelets, as proposed by the injury theory of
atherogenesis.

Other than its toxic effects, oxidized LDL can cause non-
lethal changes within the arterial intima that are characteristic
of, and possibly responsible for the formation of atherosclero-
tic plaque. Oxidized LDL or lipid peroxidation within the
intima may stimulate the secretion of specific monocyte-
directed chemoattractants such as monocyte chemotactic pro-
tein by the endothelial or smooth muscle cell. 33 Adherence of
the circulating monocyte with its subsequent transendothelial
migration is an invariable finding in early plaque. Oxidized
LDL is itself a potent monocyte attractant that can concurrent-
ly inhibit the basal and stimulated motility of the resident
macrophage. These properties would facilitate infiltration, re-
tention, and lipid ingestion by the macrophage within the
subendothelium, promoting the formation of foam cells and
the further release of macrophage-related mitogens and cy-
tokines. The presence of oxidized compounds may affect the
composition and ultrastructure of the endothelial plasma
membrane. An increase in measured membrane fluidity can
change its permeability to lipoproteins, either by changing
receptor movement on the cell surface or by influencing
nonreceptor-mediated fluid endocytosis. Oxidized LDL has
been shown to influence the phenotypic expression of the
monocyte and smooth muscle cell, affecting their secretion of
mitogenic peptides and of matrix materials, including colla-
gen and glycoproteins. Collagen production can result in
plaque build-up and stabilization, whereas glycoprotein
accumulation may lead to lipid aggregation and entrapment
within the subendothelial space. Glycolated LDL or LDL

pa rt I Vascular pathology and physiology

aggregates also may stimulate foam cell production through
their rapid ingestion by the resident macrophage through
nonregulated, receptor-independent phagocytosis.

Immune mechanisms

Evidence suggests that atherosclerosis may be mediated, in
part, by components of the immune system. Modification of
the LDL molecule renders it immunogenic. Autoantibodies
to oxidized LDL can be found in human plasma, and there
appears to be a positive correlation between elevated
immunoglobulin levels and clinical atherosclerotic coronary
artery disease. 34 It has been postulated that LDL immune com-
plexes may contribute to atherosclerosis by enhancing
lipoprotein uptake by macrophages, possibly by means of the
Fc receptor. This would result in the generation of foam cells,
but may also contribute to further cellular infiltration and acti-
vation through initiation of a humoral cascade and cytokine
secretion.

Both immunoglobulins and activated complement factors
have been found within the arterial intima, providing addi-
tional evidence of an immune response. Seifert and Hugo 35
demonstrated the presence of terminal C5a-9 complexes in
rabbit aortas in response to high-fat diets. The appearance of
complement fractions was followed by monocyte accumula-
tion and foam cell formation. Although this observation does
not prove a direct cause-effect relationship, it has been shown
that certain complement peptides and immunoglobulins are
chemotactic for monocytes. These may be produced in re-
sponse to the oxidation of LDL, or they may occur in associa-
tion with monocyte activation or cellular degeneration. In this
manner, the macrophage may act as an effector cell, secreting
interleukins, complement factors, and other peptides to stim-
ulate further adhesion and migration of chemotoxins by circu-
lating monocytes. Ultimately, the deposition of immune
complexes within the subendothelial space may be injurious
to the arterial endothelium. The infiltration of T lymphocytes,
which is a common finding in both early and mature athero-
sclerotic plaque, may further modulate the response of the ar-
terial wall through the stimulated production of additional
lymphokines and growth factors by smooth muscle cells and
macrophages. In addition, the release of matrix metallopro-
teinases by activated T cells may also contribute to the thin-
ning and rupture of the fibrous cap, thereby triggering an
acute thrombotic event. 36

In recent years, there has been increasing evidence of
infection as a cause of arterial wall inflammation and plaque
development. Pathogens implicated in this process include
Chlamydia pneumoniae, cytomegalovirus, and Helicobacter py-
lori, although a true causative role has not been firmly estab-
lished. Chlamydia pneumoniae, an intracellular pathogen, has
been found both in macrophages and in atherosclerotic le-
sions, where it has been identified by electron microscopy and

immunocytochemical assays in plaque derived from coro-
nary, carotid, and femoral plaque. 37-39 Patients with athero-
sclerosis have also been shown to have high titers of antibodies
to C. pneumoniae compared with age-matched controls. 38 It has
been suggested that Chlamydia may accelerate the develop-
ment of foam cells by promoting the uptake of LDL. 40 Others
have demonstrated that the infection of human endothelial
cells in culture resulted in increased platelet adhesion, indicat-
ing that C. pneumoniae may also precipitate acute thrombotic
events associated with unstable plaque. 41

Plaque regression

It is apparent from both animal and human studies that ma-
nipulation of the factors known to contribute to atherosclero-
sis, including serum cholesterol levels, can result in plaque
stabilization or regression. Knowledge of the cellular and mol-
ecular events involved in the initiation or regulation of plaque
formation may provide new therapeutic avenues to the pre-
vention and treatment of this disease. Of particular interest is
the phenomenon of reverse cholesterol transport, which al-
lows the peripheral cell to remove free cholesterol. This
process is mediated by HDL, which can function as an accep-
tor molecule for free cholesterol derived either from the cellu-
lar membrane or from other lipoprotein molecules. HDL is
composed of a phospholipid bilayer surrounded by apo-
lipoprotein Al protein and containing a cholesterol ester core.
Free cholesterol taken up by HDL can be esterified by circulat-
ing LCAT. This cholesterol ester may then be stored in the core
region of the molecule, permitting the uptake of additional
free cholesterol.

The transfer of cellular cholesterol to HDL can occur
through several different mechanisms. Free cholesterol can
passively diffuse from the plasma membrane to the acceptor
lipoprotein through the aqueous-serum interface. This form
of transfer would not require specific cell surface binding or
energy expenditure. Its rate of movement would depend on
the availability of the acceptor molecule, the relative choles-
terol and phospholipid content of the membrane and acceptor
molecule, the nature of the fluid interface, and the ability of the
cell to hydrolyze cytoplasmic cholesterol esters and provide
free cholesterol to the membrane for its removal. Alternatively,
cholesterol exchange may involve the binding of HDL to spe-
cific membrane receptors, which may then facilitate the trans-
fer process. The rate of transfer could then be regulated by
factors that affect the binding kinetics of HDL, including the
cellular content of cholesterol. Finally, HDL may incorporate
apolipoprotein E, which is secreted by arterial endothelial and
smooth muscle cells, as well as by monocytes. This association
with apolipoprotein E further increases the ability of HDL to
accept cholesterol, and may subsequently enhance its uptake
and secretion by the liver. Cholesterol ester, once contained
within the HDL core, may then transfer to apoprotein B

chapter 4 Molecular aspects of atherosclerosis

lipoproteins (LDL) by means of cholesterol ester transfer
protein (CETP), and be processed by the liver through LDL
receptor-mediated uptake. It may also be taken up directly by
the liver with reprocessing of the HDL molecule or, if asso-
ciated with apolipoprotein E, it may be removed though the
hepatic apolipoprotein E receptor. Cholesterol transfer via
these pathways is essential to reverse transport, and may be
altered by such factors as gender, dietary cholesterol, and
lipoprotein composition.

Treatment strategies

Treatment strategies have focused primarily on lowering
serum cholesterol levels by reducing dietary intake, blocking
its absorption through the intestine, or inhibiting its synthesis
by the liver. 42 A second approach has involved the dietary sub-
stitution of unsaturated fatty acids for cholesterol-elevating
saturated fatty acids. However, the response to these various
methods of lipid reduction has been variable and, in some
cases, associated with undesired effects. Recognition of the po-
tential role of lipid peroxidation in the process of atherogenesis
has resulted in an interest in the use of antioxidants such as
probucol for plaque inhibition. Vitamins A and C may have
similar value in inhibiting oxidation and interfering with the
peroxidation cascade. Alteration of intracellular signal trans-
duction and cellular response to atherogenic stimuli may also
prove to be useful, although specific transduction pathways
linked to atherosclerosis have not yet been identified. Most
promising to date is the benefit of calcium channel blockade,
which appears to control experimental plaque formation pos-
sibly through stabilization of cellular membranes or through
the effect of calcium on cellular secretion, communication, and
genetic expression. 17 In addition, calcium channel blockade
has been shown to have direct effects on smooth muscle cell
migration and proliferation, lipoprotein metabolism, matrix
production, platelet aggregation, and resistance to oxidant
injury. °

Augmentation of NOS may have similar benefits. Further
targets for intervention could include enhancement of reverse
cholesterol transport through chemical or genetic manipula-
tion, alteration of monocyte recruitment through the inhibi-
tion of either cellular chemoattractant secretion or adhesion
molecule expression, suppression of platelet activity and,
finally, alteration of the immunologic response associated
with atherogenesis.

References

1. Schwartz CJ, Valente AJ, Sprague EA, Kelley JL, Nerem RM. The
pathogenesis of atherosclerosis: an overview. Clin Cardiol 1991; 14
(Suppl. 2):11.

2. Ross R. The pathogenesis of atherosclerosis: an update. N Engl }
Med 1986; 314:488.

3. Ip JH, Fuster V, Badimon L, Badimon J, Taubmen MB, Chesebro
JH. Syndromes of accelerated atherosclerosis: role of vascular in-
jury and smooth muscle cell proliferation. / Am Coll Cardiol 1990;
15:1667.

4. Price DP, Loscalzo J. Cellular adhesion molecules and atherogene-
sis. Am J Med 1999; 107:85.

5. Schaller MD, Borgman CA, Cobb BS et al. ppl25FAK a structurally
distinctive protein-tyrosine kinase associated with focal adhe-
sions. Proc Natl Acad Sci USA 1992; 89:5192.

6. Kume N, Cybulsky MI, Gimbrone MA Jr. Lysophosphatidyl
choline, a component of atherogenic lipoproteins, induces
mononuclear leukocyte adhesion molecules in cultured human
and rabbit arterial endothelial cells. / Clin Invest 1992; 90:1138.

7. Ishida T, Takahashi M, Corson MA, Berk BC. Fluid shear stress
mediated signal transduction: how do endothelial cells transduce
mechanical force into biological responses? Ann NY Acad Sci 1997;
811:12.

8. Davies PF, Barbee KA, Lai R, Robotewskyj A, Griem ML. Hemo-
dynamics and atherogenesis. Ann NY Acad Sci 1995; 748:86.

9. Gimbrone MA Jr, Resnick N, Nagel T et al. Hemodynamics, en-
dothelial gene expression and atherogenesis. Ann NY Acad Sci
1997; 811:1.

10. Gimbrone MA Jr, Topper JN, Nagel T, Anderson KR, Garcia-
Cardena G. Endothelial dysfunction, hemodynamic forces and
atherogenesis. Ann NY Acad Sci 2000; 902:230.

11. Davies PF Flow mediated endothelial mechanotransduction.
Physiol Rev 75; 1995:519.

12. Resnick N, Yahav H, Schubert S, Wolfovitz E, Shay A. Signalling
pathways in vascular endothelium activated by shear stress: rele-
vance to atherosclerosis. Curr Opin Lipidol 2000; 11:167.

13. Hemlinger G, Berk BC, Nerem RM. The calcium responses of en-
dothelial cell monolayers subjected to pulsatile and steady lami-
nar flow differ. Am J Physiol 1995; 269:C367.

14. Kuo TH, Wang KK, Carlock L, Diglio C, Tsang W. Phorbol ester in-
duces both gene expression and phosphorylation of the plasma
membrane Ca(2+)pump. J Biol Chem 1991; 266:2520.

15. Lloyd-Jones DM, Bloch KD. The vascular biology of nitric oxide
and its role in atherogenesis. Ann Rev Med 1996; 47:365.

16. Libby P, Sukhova G, Lee RT, Liao JK. Molecular biology of athero-
sclerosis. Int J Cardiol 1997; 62 (Suppl. 2):823.

17. Kane JP, Havel RJ. Disorders of the biogenesis and secretion
of lipoproteins containing the B apolipoproteins. In: Scriver CR,
Beaudet Al, Sly WS, Valle D, eds. The Metabolic Basis of Inherited
Disease. New York: McGraw-Hill, 1989:1139.

18. Brown MS, Goldstein JL. Receptor-mediated control of choles-
terol metabolism. Science 1976; 191:150.

19. Weinbaum S, Tzeghai G, Ganatos P, Pfeffer R, Chien S. Effect of cell
turnover and leaky junctions on arterial macromolecular trans-
port. Am } Physiol 1985; 248:H945.

20. Weinstein DB, Heider JG. Antiatherogenic properties of calcium
antagonists. Am J Cardiol 1987; 59:B163.

21. Lynch JJ, Ferro TJ, Blumenstock FA, Brockenauer AM, Malik AB.
Increased endothelial albumin permeability mediated by protein
kinase C activation. / Clin Invest 1990; 85:1991.

22. Oliver JA. Adenylate cyclase and protein kinase C mediate oppo-
site actions on endothelial junctions. J Cell Physiol 1990; 145:536.

23. Smirnov VN, Voyno-Yasenetskaya TA, Antonov AS et al. Vascular
signal transduction and atherosclerosis. Ann NY Acad Sci 1990;
598:167.

pa rt I Vascular pathology and physiology

24. Goldstein JL, Ho YK, Basu SK, Brown MS. Binding site on
macrophages that mediates uptake and degradation of acetylated
low density lipoprotein, producing massive cholesterol deposi-
tion. Proc Natl Acad Sci USA 1979: 76:333.

25. Parathasarathy S, Fong L, Otero D, Steinberg D. Recognition of
solubilized apoproteins from delipidated, oxidized low density
lipoprotein (LDL) by the acetyl-LDL receptor. Proc Natl Acad Sci
USA 1987; 84:537.

26. Yla-Herttuala S. Is oxidized low-density lipoprotein present in
vivo? Curr Opin Lipidol 1998; 9:337.

27. Kume N, Arai H, Kawai C, Kita T. Receptors for modified
low-density lipoproteins on human endothelial cells: different
recognition for acetylated low-density lipoprotein and oxidized
low-density lipoprotein. Biochim Biophys Acta 1991; 1091:63.

28. Oka K, Sawamura T, Kikuta K et ah Lectin-like oxidized low-
density lipoprotein receptor 1 mediated phagocytosis of
aged/apoptotic cells in endothelial cells. Proc Natl Acad Sci USA
1998; 95:9535.

29. Kugyama K, Kerns SA, Morrisett JD, Roberts R, Henry PD. Im-
pairment of endothelium-dependent arterial relaxation by
lysolethicin in modified low-density lipoproteins. Nature 1990;
344:160.

30. Palinski W, Rosenfeld ME, Yla-Herttuala S et al. Low density
lipoprotein undergoes oxidative modification in vivo. Proc Natl
Acad Sci USA USA 1989: 86:1372.

31. Kita T, Nagano Y, Yokode M et al. Probucol prevents the progres-
sion of atherosclerosis in Watanabe heritable hyperlipidemc rab-
bit, an animal model for familial hypercholesterolemia. Proc Natl
Acad Sci USA 1987; 84:5928.

32. Parathasarathy S, Wieland E, Steinberg D. A role for endothelial

cell lipoxygenase in the oxidative modification of low density
lipoprotein. Proc Natl Acad Sci USA 1989; 86:1046.

33. Mazzone T, Jensen M, Chait A. Human arterial wall cells secrete
factors that are chemotactic for monocytes. Proc Natl Acad Sci USA
1983; 80:5094.

34. Hansson GK, Jonasson L, Seifert PS, Stemme S. Immune mecha-
nisms in atherosclerosis. Arteriosclerosis 1989; 9:567.

35. Seifert PS, Hugo F. Prelesional complement activation in experi-
mental atherosclerosis. Lab Invest 1989: 60:747.

36. Hansson GK. Immune mechanisms in atherosclerosis. Arterioscl
Thromb Vase Biol 2001; 21:1876.

37. Kuo C, Coulson AS, Campbell LA et al. Detection of Chlamydia
pneumoniae in atherosclerotic plaques in the walls of arteries of
lower extremities from patients undergoing bypass operation for
arterial obstruction. / Vase Surg 1997; 26:29.

38. Saikku P, Leinonen M, Mattila K et al. Serological evidence of an as-
sociation of a novel Chlamydia, TWAR, with chronic coronary
heart disease and acute myocardial infarction. Lancet 1988; 2
(8618):983.

39. Wong YK, Gallagher PJ, Ward ME. Chlamydia pneumoniae and ath-
erosclerosis. Heart 1999; 81:232.

40. Muhlestein JB. Chronic infection and coronary artery disease. Med
ClinNA 2000; 84:123.

41. Fryer RH. Chlamydia species infect human vascular endothelial
cells and induce procoagulant activity. / Invest Med 1997; 45:168.

42. Grundy SM. Cholesterol and coronary heart disease: future direc-
tions. JAMA 1990; 264:3053.

43. Schmitz G, Hankowitz J, Kovacs EM. Cellular processes in athero-
genesis: potential targets of Ca(2+) channel blockers. Atherosclero-
sis 1991; 88:109.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Localization of atherosclerotic lesions

Christopher K. Zarins
Chengpei Xu
Charles A. Taylor
Seymour Glagov

Arteries are now recognized as a distinct organ system, with
biosynthetic and biomechanical functions that maintain nor-
mal physiologic function across a wide range of conditions.
Under certain circumstances, artery wall cellular function is
altered, allowing the accumulation of atherosclerotic plaque
in the intima. Characteristically, atherosclerotic plaque is not
uniformally distributed throughout the vascular system, but
is localized to distinct and reproducible areas such as the
carotid bifurcation, the coronary arteries, the abdominal aorta,
and the lower extremity arteries. Hemodynamic forces, cellu-
lar responses, and other factors are thought to play an impor-
tant role in determining the location of these lesions. This
chapter reviews the etiologic factors responsible for the local-
ization of atherosclerotic lesions.

Arterial smooth muscle and endothelial cell responses to
physiologic and pathologic stimuli promote the induction and
progression of atherosclerotic plaque. These stimuli include (i)
blood flow and wall shear stress (flow fields) in proximity to
the endothelial surface; (ii) blood-borne substances, particu-
larly elevated concentrations of certain lipoprotein fractions;
(iii) the structure, proliferation, and biosynthetic reactivity of
cells comprising the arterial wall; and (iv) the reactivity of
migratory cells that enter the arterial wall and participate in
the evolving pathologic process. Although the relative contri-
bution of each of these stimuli to plaque localization may vary,
there is a close integration between mechanical and metabolic
arterial functions such that alteration of one stimulus will
affect other aspects of the pathogenic process.

Arterial structure and function

Intima

The intima comprises the innermost arterial layer, extending
from the luminal surface to the internal elastic lamina. An
endothelial cell monolayer overlies the internal elastic lamina,
with few leukocytes, smooth muscle cells, or connective fibers
present under normal circumstances. The basal lamina pro-

vides a continuous, pliable substrate for the endothelial cell
monolayer, as well as focal attachments to the internal elastic
lamina. Junctional overlap among adjacent endothelial cells,
and cell deformability in response to pulsatile wall motion,
bending, or stretching help prevent the development of dis-
continuities in the endothelial lining. Focal attachments to the
underlying elastic lamina and adjacent cells 1 prevent slippage,
telescoping, or endothelial cell detachment by elevations of
shear stress or other mechanical forces. Endothelial cells sense
changes in blood flow, pressure, and ambient oxygen ten-
sion 2 ' 3 through specific, receptor-activated cellular metabolic
and biosynthetic events. Agonists such as thrombin, platelet-
activating factor, and bradykinin increase endothelial cell
intracellular calcium through receptor-coupled changes in
phosphoinositol metabolism. Activation of ion channels in-
duces cellular events through a second messenger system that
enable endothelial cells to regulate tone, inflammation, and
hemostasis. 4-11 In addition, endothelial cells produce biologic
mediators 12-16 that influence hemostasis, immunogenicity,
and vascular remodeling, as well as vasoreactivity.

Response-to-injury hypothesis
(endothelial denudation)

The endothelial surface is exposed to shear stress and potential
mechanical injury by the force of luminal blood flow. In vivo,
experimentally induced endothelial cell denudation is tran-
sient, and is rapidly restored by regeneration. Despite this
vigorous and rapid response, endothelial disruption had been
proposed as a critical and essential first step in atherosclerotic
plaque formation. 17

In the response-to-injury hypothesis, the loss of endo-
thelium, with subsequent platelet adherence, the release of
platelet-derived growth factor, and the induction of smooth
muscle cell proliferation is considered to be the initial step in
atherogenesis. According to this hypothesis, local, repeated
endothelial denudation and the ensuing response to injury
determine the location of plaque formation. There is, how-
ever, little evidence to suggest that endothelial disruption or

pa rt I Vascular pathology and physiology

removal results in eventual sustained lesion formation, 18 even
in the presence of hyperlipidemia. Endothelial denudation
and platelet adherence in an animal model have not resulted in
smooth muscle proliferation or intimal lesions. 19 Strong evi-
dence suggests formation of intimal plaque requires the
presence of an intact endothelium. 20-22

Endothelium and atherogenesis

The endothelial lining regulates the movement of cells from
the arterial lumen to sites within the artery wall and into the
surrounding tissues. Adherence of circulating cells to the
endothelium and passage through the vessel wall are func-
tions of cell surface receptors and matrix proteins. 23-26 These
include the antigen-specific receptors of T and B lymphocytes,
the selectins, and the integrins. Selectins control lympho-
cytes and neutrophil interactions with the endothelium.
Integrins are responsible for platelet adhesion and cell
migration. 23-26 Inflammatory chemotactic factors such as
platelet-activating factor, leukotriene B 4 , complement C5a,
and formyl-methionyl-leucyl-phenylalanine stimulate in-
flammatory cell binding to endothelial cells via altered ex-
pression of the CD11/CD18 integrins. 23 ' 24 The cytokines
interleukin-1 and tumor necrosis factor increase leukocyte-
endothelial binding by upregulating surface expression of the
intracellular cell adhesion molecule and the endothelial leuko-
cyte adhesion molecule. Specific receptor interactions also
regulate lymphocyte binding (via so-called addressins).

Monocyte adhesion and infiltration into the vessel wall may
play a significant role in atherogenesis. Several specialized
receptors control monocyte-endothelial cell adhesion. 24 ' 25 The
early development of atherosclerosis (as induced by hyper-
cholesterolemia) is associated with the expression of adhesion
molecules on endothelial cells that specifically promote mono-
cyte adhesion. 27

Basic endothelial cell biology, shear stress, and atherogene-
sis are reviewed extensively in a monograph by Dzau and
colleagues. 28 Knowledge of endothelial cell function and
receptor-mediated changes in vitro and in vivo is rapidly
accumulating. Evidence clearly linking in vitro endothelial cell
functional alterations with in vivo plaque induction in humans
is lacking, however. The precise role of the endothelial cell
in the pathogenesis of plaque formation remains to be
determined.

Media

The media extends from the internal elastic lamina to the
adventitia. Although an external elastic lamina demarcates
the boundary between the media and adventitia in many
vessels, a distinct external elastic lamina may not be present,
particularly in vessels with a thick media and a fibrous adven-
titial layer. The media represents closely packed layers of
smooth muscle cells in association with elastin and collagen

Figure 5.1 Diagrammatic representation of the pericellular basal
lamina-collagen fibril matrix. Medial cell surfaces are invested by interlacing
fine collagen fibrils closely associated with and partially embedded in basal
lamina. At normal distending pressures (top), the interlacing fibrils are nearly
perpendicular to the long axes of the cells. Hyperdistention (bottom) results
in elongation of the cells and in increased obliquity and approximation of the
fibrils. Cells are apparently kept together by tightening of the interlacing
collagen network, in the manner of afingertrap. (From Clark JM, GlagovS.
Structural integration of the arterial wall, I: relationships and attachments of
medial smooth muscle cells in normally distended and hyperdistended
aortas. Lab Invest 1 979; 40:587.)

fibers (Fig. 5.1). Groups of similarly oriented cells are sur-
rounded by a common basal lamina of type IV collagen, and
closely associated, interlacing type III collagen fibrils. Me-
chanical stretch, with cyclic or sudden changes in diameter, re-
inforces fascicle cohesion. 29 Each cellular subgroup or fascicle
is surrounded by similarly oriented elastic fibers. Abundant,
focal, tight attachment sites exist between smooth muscle cells
and elastic fibers, 30 evenly distributing tension and recoil, and
preventing disruption.

The musculoelastic fascicles are the structural units of the
media (see Fig. 5.1). 29-31 The fascicles vary in size, orientation,
and matrix composition depending on location within the
arterial vasculature, the transmural distribution of tension,
and redistribution of tensile stress about zones of transition at
branches and bifurcations. On transverse section of larger ves-
sels, this structure appears as a series of layers. Thick, undulat-
ing collagen bundles (type I) are distributed among adjacent
fascicles, 32 and provide the major tensile support in large ves-
sels, preventing overdistention at elevated pressures.

Axial gradients of matrix composition exist along the
aorta, 33 and vary with media penetration by vasa vasorum,
wall thickness, and architecture. 34 Acute luminal pressure
elevation may disrupt fascicles, fracturing cell bodies and in-

chapter 5 Localization of atherosclerotic lesions

terrupting the basal lamina sheaths, 35 whereas the tight cell
insertions on elastic fibers tend to resist disruption. Gradual
increases in mural tension, such as those associated with
growth, result in increased cellular biosynthesis, with propor-
tional increases in collagen and elastin accumulation. 36 ' 37

Diffusion into the media from the lumen sustains the inner
0.5 mm of the adult mammalian aortic media, corresponding
to approximately 30 medial fibrocellular lamaellar units. 35
Thicker arteries, with more than 30 layers, are nourished by
penetrating adventitial vasa vasorum. Vasa vasorum arise
from the parent artery at branch junctions, arborizing in the
adventitia, and penetrating the media in thicker walled arter-
ies. Mural stresses and deformations may impair vasal flow. 38
Intimal plaque formation increases the diffusion barrier from
the lumen to the smooth muscle cells of the media. This
increase in wall thickness is accompanied by an ingrowth
of vasa vasorum, which also are identifiable in atherosclerotic
lesions. Both intraplaque hemorrhage and plaque disruption
may be potentiated by changes in the vascular supply of the
artery wall and plaque.

Adventitia

The adventitia is a framework of fibrocellular connective tis-
sue. This framework contains a network of vasa vasorum and
nerves that mediate smooth muscle tone and contraction. In
smaller arteries, the adventitia is indistinct or poorly devel-
oped. In large visceral arteries, the adventitia is a layered com-
position of collagen and elastic fibers. The adventitia in these
vessels may be more prominent than the associated media. In
atherosclerotic arteries, increasing intimal plaque thickness is
associated with underlying medial atrophy and adventitial
thickening. 39 Under these circumstances, the adventitia may
provide considerable tensile support for the vessel wall. In-
deed, after carotid or aortoiliac endarterectomy, removal of
the entire intima and most or all of the media leaves only the
adventitia to maintain integrity of the arterial wall.

Physiologic adaptation of the arterial wall

To maintain functional integrity, arteries adapt to changing
hemodynamic conditions with alterations in the dimensions,
structure, and composition of the arterial wall. Arterial tan-
gential wall tension is approximated closely by the product of
the lumen radius and the distending intraluminal pressure.
This tension is distributed and supported by the full thickness
of the vessel wall. Chronic changes in tangential vessel wall
tension significantly influence arterial wall thickness and
composition.

The relationship between human arterial wall thickness and
tangential wall tension is demonstrated effectively during the
early postnatal period. 36 Under normal circumstances, blood
pressures in the pulmonary trunk and the ascending aorta are

very nearly equal at birth, at about half the normal adult value.
At this time, the length, radius, wall thickness, and morphol-
ogy of the pulmonary trunk and ascending aorta are similar.
At birth, aortic blood pressure rises to approximately twice the
prenatal value, whereas pulmonary pressure falls by half. This
results in a marked increase in tangential tension in the aorta
and stimulates a corresponding increase in aortic wall thick-
ness (Fig. 5.2). Matrix fiber accumulation for the aorta and the
pulmonary trunk vessels parallels the increase in wall tensile
stress, and the rate of production of matrix per cell is markedly
different for the two artery segments. This accumulation
accounts for the difference in wall thickness between the two
vessels. Despite differences in tension and matrix fiber content
between the two vessels, cell proliferation continues at the
same rate in each. This phenomenon demonstrates the capa-
bility of smooth muscle cells to modulate their biosynthetic
metabolism in response to alterations in imposed tensile
stress.

A smooth muscle cell biosynthetic response to cyclic stretch
also has been demonstrated in cell culture. 40 Rabbit aortic
smooth muscle cells were grown on purified elastin mem-
branes, then subjected to cyclic stretching at 52 cycles /min for
4-8 days. Compared with cells grown without stretching, cells
on cyclical stretching showed a two- to fourfold increase in rates
of collagen, hyaluronate, and chondroitin 6-sulfate synthesis.

Fibrous plaques usually are eccentric, and are covered by an
intact endothelial surface. Although considerable variation
exists in plaque composition and configuration, a characteris-
tic architecture prevails. The immediate subendothelial region
of the plaque consists of a compact and well organized, strati-
fied layer of smooth muscle cells and connective tissue fibers
known as the fibrous cap. This structure may mimic medial
architecture, including the formation of a subendothelial elas-
tic lamina, which may function to sequester the underlying
necrotic and thrombogenic plaque core from the luminal sur-
face. This surface usually is regular, with a concave contour
corresponding to the circular or oval cross-sectional lumen of
the uninvolved vessel wall segment. The stable necrotic core
occupies the deeper plaque. The core contains amorphous,
crystalline, and droplet forms of lipid. Cells of undetermined
origin, with morphologic, functional, and cell surface receptor
characteristics of smooth muscle cells or macrophages are
noted beneath the core. These cells also may contain lipid
vacuoles. Calcium and myxoid deposits, collagen and elastin
matrix fibers, basal lamina, and amorphous ground substance
also are evident. Atherosclerotic plaques grow in an episodic
fashion, demonstrating dense fibrocellular regions adjacent to
organizing thrombus and atheromatous debris. Intermittent
ulceration and healing occur, with thrombi being incorporat-
ed into the lesion.

Vasa vasorum may nourish the plaque, facilitating the
organization of thrombotic deposits and the remodeling of the
plaque and artery wall. 41 Attenuation of the subadjacent
media promotes outward bulging of the plaque toward the

pa rt I Vascular pathology and physiology

30r

* 4

Figure 5.2 Relation between tension, age, and elastin and collagen deposition in the ascending aorta (AA) and pulmonary trunk (PT) in a rabbit model. Aortic
wall tension rises rapidly after birth with increasing blood pressure, and is accompanied by a significant increase in elastin and collagen. (From Leung DYM,
Glagov S, Mathews MB. Elastin and collagen accumulation in rabbit ascending aorta and pulmonary trunk during postnatal growth: correlation of cellular
synthetic response with medial tension. CircRes 1977; 41 :3 16, with permission from Lippincott, Williams & Wilkins.)

adventitia. Although this attenuation sequesters plaque,
enlarges the artery, and stabilizes the wall, a predominant lytic
reaction may result in excessive arterial dilation or aneurys-
mal degeneration. Experimental evidence suggesting such
a mechanism for aneurysm formation has been obtained
in nonhuman primates in our laboratory 42 and by other
investigators. 43

Tissue between the necrotic core and the media, however,
usually is densely fibrotic. Arterial wall support may thus be
maintained by the integrity of fibrous cap or thickened adven-
titia. Advanced lesions, particularly those associated with
aneurysms, may appear to be atrophic and relatively acellular,
consisting of dense fibrous tissue and a minimal necrotic cen-
ter. Calcification is a prominent feature, involving the superfi-
cial and deeper layers. Terms such as fibrocalcific, lipid-rich,
fibrocellular , necrotic, and myxomatous describe various pre-
dominant aspects of advanced plaques. Calcific deposits are
most prominent in plaques in older people and in the abdomi-
nal aorta or coronary arteries, where the earliest plaques form
in animal models and in humans. 44

Angiographic luminal narrowing often is perceived as
plaque protrusion into the lumen. This perception is sup-
ported by gross observations of vascular surgeons and pa-
thologists who examine collapsed atherosclerotic arteries
en face or on cross-section. Without distending intraluminal
pressure, elastic recoil causes the eccentric plaque to appear as
a protrusion or bulge. Pressure fixation 45 restores the cross-
sectional luminal contour to its usually regular, round, or oval
configuration, even with large and extensive raised athero-

sclerotic lesions. 46 Fixed in this manner, the usual eccentric
atherosclerotic plaque bulges outward from the lumen; the
external cross-sectional contour of an atherosclerotic artery
becomes oval while retaining a circular lumen. This character-
istic also is demonstrated in vivo in aortic cross-sectional
images obtained by computed tomographic aortography.
Protrusion of plaque or its contents into the membranes
produced two to four times more collagen. Cell proliferation
was not differentially altered by any of these procedures.
Furthermore, the cyclically stretched cells showed fewer
degenerative changes, and their cytoplasmic features
confirmed the proposed level of biosynthesis.

Chronically elevated adult arterial transmural tension
increases the cross-sectional area of the media without a sig-
nificant structural change in the lamellar architecture. Matrix
protein deposition increases, with a proportionally greater
increase in collagen compared with elastin fibers. 37 In patients
with hypertension, arterial and arteriolar intimal thickening
also may develop as an adaptive response to the increase in
wall tension.

Adaptive changes in artery luminal diameter are deter-
mined by changes in blood flow. During embryologic growth
and development, lumen diameter is determined by the
volume of blood flow. After birth, increases in artery diameter
continue as a response to increases in blood flow. 47 This
phenomenon also is demonstrated in mature arteries after
cessation of growth, with enlargement of arteries proximal to
arteriovenous fistulas, and a decrease in the size of arteries
proximal to amputated limbs. 48

chapter 5 Localization of atherosclerotic lesions

Luminal diameter adaptation is responsive to wall shear
stress, as determined by the effective velocity gradient at the
endothelial-blood interface. 49 In mammals, wall shear stress
normally ranges between 10 and 20 dynes /cm 2 at all locations
throughout the arterial vasculature. In arteriovenous fistulas,
the afferent artery enlarges enough to restore shear stress to
this physiologic range. 50 This response depends on the pres-
ence of an intact endothelial surface, 51 and may be mediated
by the release of endothelial-derived relaxing factors, includ-
ing nitric oxide, or other vasoactive agents 50 ( Ying H, Harris EJ
Jr, Dalman RL, unpublished observations, 1992).

Human atherosclerotic plaque morphology

Although atherosclerotic plaques are distinguished by the
presence of lipid, it is unclear whether all lesions containing
lipids are necessarily precursors of clinically significant
atherosclerotic plaques. A prime example of this uncertainty is
demonstrated by the questionable significance of the so-called
fatty streak lesion. This term describes a flat, yellow, focal
luminal patch or streak, representing an accumulation of
lipid-laden foam cells in the intima, evident in most people
older than 3 years. They are identified with increasing
frequency between the ages of 8 and 18 years, after which
many apparently resolve, despite the frequent presence of
matrix materials among the characteristic cells. Fatty streaks
exist at any age, often adjacent to or even superimposed on ad-
vanced atherosclerotic plaques. Fatty streaks and atheromata,
however, do not have identical patterns of localization, and
fatty streaks do not compromise the lumen or ulcerate. 52 In
experimental animals, diet-induced lesions resembling fatty
streaks occur early, before characteristic atherosclerotic le-
sions prevail. Although this subject remains controversial, the
link and transition between fatty streak and fibrous plaque
formation remains to be clarified.

The term fibrous plaque identifies the characteristic and
unequivocal atherosclerotic lesion. These intimal deposits
appear in the second decade of life, becoming predominant
or clinically significant only during or after lumen pressure
fixation signifies plaque ulceration, hemorrhage, dissection,
or thrombosis.

Mechanical determinants of
plaque localization

Near-wall properties of arterial flow fields and the distribu-
tion of mural wall shear stress correspond closely to athero-
sclerotic plaque localization. 53-59 Plaques develop where
shear stress is reduced, 55 ' 56 not elevated, with an intact en-
dothelial surface, even in the absence of platelet deposition. 60
The revised response-to-injury hypothesis now stresses meta-
bolic or functional changes sustained by intact endothelial

Cross-section
of carotid sinus

Low
shear

region

Figure 5.3 Flow field at the human carotid bifurcation. Plaque forms in the
region of low wall shear stress, and not in the region of high wall shear. The
low-shear area is also characterized by flow separation, oscillation shear
stress, and prolonged particle residence time. (From Zarins CK, Giddens DP,
Bharadvaj BK, Sottiurai VS, Mabon RF, Glagov S. Carotid bifurcation
atherosclerosis. Quantitative correlation of plaque localization with flow
velocity profiles and wall shear stress. CircRes 1983; 53:502, with permission
from Lippincott, Williams & Wilkins.)

cells that alter binding or metabolism of lipid molecules or
modify transendothelial transport, rather than denudation of
the endothelium itself. 21

Atherosclerosis tends to occur principally in three locations
within the arterial vasculature: the carotid-cerebral (Fig. 5.3),
coronary, and aortic-peripheral systems. Within these predis-
posed regions, lesions form in predictable geometric configu-
rations, demonstrating the influence of shear stress and flow
patterns. Size, as well as localization, closely correlate with
low wall shear stress and departures from unidirectional
flow. 55 ' 56 Plaque initiation and localization is the result of low,
rather than high, shear stress, low flow velocity, flow separa-
tion, and oscillation in wall shear direction. 61

Regions of increased mural tensile stress about branches, 53
pulsatile wall motion, 62 and wall thickness and density 63 ' 64
also are associated with selective plaque localization. Con-
versely, regions of relatively elevated wall shear or reduced
tensile stress, at flow dividers and along the outer or convex
aspects of curved arterial segments, generally are spared. 65
Hemodynamics and tensile influences also are important in
plaque progression and evolution, 66,67 and influence potential
plaque regression. 68 As an example of this influence on regres-
sion, hypertension was found to sustain experimental plaque
progression in a hypercholesterolemic cynomolgus monkey

pa rt I Vascular pathology and physiology

model, despite a reduction in serum cholesterol level. 69 Re-
duced flow and consequent reduction in wall shear stress also
tend to induce intimal thickening. An increase in wall volume,
including cell enlargement, cell proliferation, and net matrix
accumulation is demonstrated in long-term reactions. 70

A sieving effect related to these changes in wall composi-
tion 71 ' 72 and porosity 63 has been proposed. Wall thickening,
including intimal thickening, may retard transmural mass
transport, providing the basis for intimal lipid deposition. 73
The accumulation of matrix fibers with affinity for lipid mole-
cules 74-78 and the fusion or accretion of lipid particles on these
components also may be responsible.

Susceptible regions of
the arterial vasculature

Carotid artery bifurcation

The carotid bifurcation is particularly prone to plaque forma-
tion, with focal plaque deposition occurring principally at the
origin of the internal carotid artery (Fig. 5.4). The proximal
common and distal internal carotid arterial segments are rela-
tively spared. Plaque formation is thought to be the result of
hemodynamic conditions created by the geometry of the
bifurcation region.

The cross-sectional area of the sinus is twice that of the

Figure 5.4 Carotid arteriogram demonstrating plaque formation
predominantly along the outerwallofthe internal carotid sinus, resulting in
severe localized stenosis of the proximal internal carotid artery. Plaque also
forms along the outer wall of the external carotid artery. Both areas are
regions of lowwall shear stress, flow separation, and increased particle
residence time.

immediately distal internal carotid segment. This relation-
ship, in addition to the branching angle, results in a large area
of flow separation and low shear stress along the outer wall of
the sinus. Wall shear stress in this region oscillates in both mag-
nitude and direction during the cardiac cycle. 56 A region of
laminar flow and high unidirectional shear stress exists along
the relatively spared, flow-divider side inner wall of the sinus
(Fig. 5.3). 56 ' 79

The oscillations in flow direction in the region of greatest
plaque formation occur primarily during the downstroke of
systole. 80 If low and oscillating wall shear stresses favor
atherogenesis, then modification of heart rate could affect
atherogenesis, particularly in the proximal segments of the
coronary arteries, 58,81 the distal aorta, 79 as well as the carotid
bifurcation. 82

Outer wall plaque enlargement at the carotid bifurcation
modifies the geometric configuration of the lumen, favoring
subsequent plaque formation on the side and inner walls. In its
most advanced and stenotic form, carotid bifurcation athero-
sclerotic disease thus involves the entire circumference of
the sinus, including the region of the flow divider (Fig. 5.5).
Nonetheless, carotid bifurcation plaques remain largest and
most complicated at the outer and side walls of the sinus.
Characteristic hemodynamic conditions at the carotid bifurca-
tion, including the turbulence responsible for the characteris-
tic bruit, also may compromise integrity of existing carotid
plaques and contribute to their tendency to fissure, ulcerate,
and form thromboemboli.

Previous studies of hemodynamic factors in the region of
plaque formation in the carotid bifurcation were characterized
in a glass model of the carotid bifurcation. 55 More recent
studies of plaque localization, the extent of compensatory

9 2

6 382

METRIC 1 2 3 4

Figure 5.5 Complex carotid bifurcation plaque removed at the time of
carotid endarterectomy.

chapter 5 Localization of atherosclerotic lesions

artery enlargement, and the effect of heart rate have been
characterized in experimental atherosclerosis at the carotid bi-
furcation of the cynomolgus monkey. Heart rate was altered
by sinoatrial node ablation. The animals were fed an athero-
genic diet for 6 months and then killed. Sham-operated mon-
keys demonstrated no change in heart rate and served as
controls. Axial and circumferential plaque distribution about
cynomolgus monkey carotid bifurcation was similar to that
observed in humans, plaque formation induced compen-
satory artery enlargement, and plaque progression was re-
tarded by a lowered heart rate. 82

The finding that heart rate correlated positively with plaque
formation in experimental atherosclerosis at the carotid bifur-
cation lends further support to a role for heart rate as a deter-
minant of the severity of experimental atherogenesis, as we
have demonstrated previously in the coronary arteries. 83
These observations are in agreement with the experimental
findings of others, 84,85 and are in accord with epidemiologic re-
ports that elevated heart rate is associated with an increased
occurrence of clinical cardiovascular events. 86-88

We also have suggested that the combination of flow separa-
tion, low wall shear stress, and oscillation in shear stress direc-
tion during the cardiac cycle, which occurs at the lateral wall
opposite the flow divider about the carotid bifurcation, results
in regions of recirculation and increased particle residence
time. 55/56/80 Affected regions, such as the lateral wall of the
internal carotid artery at the carotid bifurcation, are there-
fore subjected to delayed clearance of putative blood-
borne atherogenic factors, and are thereby predisposed to
atherogenesis.

Abdominal aorta

Clinically significant aortic plaque generally is most promi-
nent below the level of the renal arteries. Plaque complications
include obstruction, ulceration, thrombus formation, and,
potentially, aneurysmal degeneration. Putative explanations
for the focal nature of these complications include flow differ-
ences in the infrarenal compared with the suprarenal aorta,
differences in mural architecture, or vasa vasorum distribu-
tion and aortic wall nutrition. Reduced physical activity
results in an overall reduction in flow volume and velocity in
the infrarenal segment, whereas suprarenal flow volume is
largely independent of skeletal muscular activity. The long-
term effect of reduced flow velocity may be accentuated by the
tendency of the aorta to enlarge with age. The frequency of
vasa vasorum present within the media drops precipitously
from the thoracic to abdominal and infrarenal segments of the
aorta, potentially contributing to the relatively avascular
nature of the abdominal aorta. 89

Hemodynamic forces in the abdominal aorta are rather
complex compared with those in the thoracic aorta. In the
abdominal aorta there are "adverse hemodynamic condi-

tions" such as low shear stress and high particle residence time
resulting from the complex flow patterns. These conditions
may account in part for plaque development preferentially in
the infrarenal abdominal aorta. 90-96 The pulsatile flow in the
abdominal aorta is particularly complex and recirculating
under normal resting conditions as a result of the multiple
branches, which deliver blood to the organs in the abdomen.
The abdominal aorta also experiences significantly different
hemodynamic forces compared with the thoracic aorta. Local-
ized differences in hemodynamic conditions include differ-
ences in velocity profiles, wall shear stress, and recirculation
zones. Utilizing a stabilized, time accurate, finite element
method, Taylor et al 97,98 have solved the equations governing
blood flow in a model of a normal human abdominal aorta
under simulated rest, pulsatile, flow conditions. They have
demonstrated that low time-averaged wall shear stress and
high shear stress temporal oscillations, as measured by an os-
cillatory shear index, were present in this location, along the
posterior wall opposite the superior mesenteric artery and
along the anterior wall between the superior and inferior
mesenteric arteries (Fig. 5.6). These regions were noted to coin-
cide with a high probability-of-occurrence of sudanophilic
lesions as reported by Cornhill et al 99 The reduction of
cross-sectional area and stiffening of the vessel wall in the ab-
dominal aorta compared with the thoracic aorta increases the
pressure pulse and contributes to a greater load on the aortic
wall. Further, the pressure pulse gets reflected off peripheral
vessels, thus contributing to a further increase in wall stress.
The pressure wave reflection of the iliac bifurcation would be
minimized if the sum of the cross-sectional areas of the iliac ar-
teries were 1.1 times that of the aortic cross-sectional area.
While this is the case in infants, this ratio decreases with age,
and reaches approximately 0.75 by age 50. This would be
further reduced with diffuse lower extremity disease. Thus the

Figure 5.6 Mean surface traction vectors along the posteriorwallofthe
abdominal aorta. Note the circumferential orientation of the mean surface
traction vectors along the posterior wall of the aorta in the neighborhood of
the renal arteries.

pa rt I Vascular pathology and physiology

pressure wave reflection is increased further with age and vas-
cular disease, further increasing the loads that the abdominal
aorta must bear. Concomitant with the increases in pressure
loads, the collagen content of the abdominal aorta increases
with age making the aorta less able to absorb pulsatile stress. It
may be that higher cumulative biomechanical stress of the ab-
dominal aorta in combination with the stiffening of the aortic
wall predisposes the abdominal aorta to atherosclerosis and
aneurysmal disease. 97 ' 98

The infrarenal abdominal aorta in humans is therefore
prone to the development of atherosclerotic plaque develop-
ment and its complications, while the thoracic aorta is rela-
tively spared of these consequences. It is not clear whether the
human aorta responds to atherosclerosis by enlarging and
whether the selective involvement of the abdominal aorta
reflects differences in response between the thoracic and
abdominal segments. Studies have demonstrated that aortic
aneurysms usually develop in the atherosclerosis-prone in-
frarenal abdominal aorta. Atherosclerotic plaque formation
in the infrarenal abdominal aorta in humans is associated
with aortic enlargement and decreased media thickness.
These changes may be predisposing factors for the preferential
development of subsequent aneurysmal dilation in the ab-
dominal aorta. 100 However, atherosclerosis in the abdominal
aorta may not necessarily result in aneurysmal dilatation.
There may be different local responses to atherosclerosis in the
abdominal aorta in human beings. Plaque deposition assoc-
iated with localized dilation, thinning of the media, and loss
of medial elastic lamellae may predispose that segment of
aorta to subsequent aneurysm formation. Plaque deposits
without media thinning, without loss of elastic lamellae, and
without artery wall dilation may predispose the aorta, in the
event of continuing plaque accumulation, to the development
of lumen stenosis. 101

Superficial femoral artery

No widely accepted explanation for the discrepancy between
the incidence of upper and lower extremity arterial atheroscle-
rotic plaque exists. Recognized differences in the two areas in-
clude hydrostatic pressure and activity-dependent variations
in volume flow. As in the abdominal aorta, relative inactivity,
and subsequently diminished shear stress may lead to in-
creased rates of plaque deposition in these arteries.

Cigarette smoking and diabetes mellitus are the risk factors
most closely associated with atherosclerotic disease of the
lower extremities. 102 The specific mechanism through which
these risk factors act is unknown. Lower extremity arterial me-
dial density, however, may be augmented by the chronically
increased smooth muscle tone characteristic of nicotine use, 103
interfering with the transluminal transfer of materials enter-
ing the intima. Speculation on the etiology of the predominant
incidence of occlusive plaque of the superficial femoral artery
at the adductor canal has centered on the likelihood of re-

peated mechanical trauma, limitations on vessel compliance,
or restrictions on compensatory enlargement due to the
closely applied adductor magnus tendon.

Management of arterial disorders in the low extremities in-
cludes control of risk factors important in the progression of
generalized atherosclerosis, exercise programs to develop col-
lateral flow, pharmacotherapy, and diet interventions with en-
dovascular therapy, or surgery to remedy the lower extremity
symptoms. 104 Surgical revascularization is appropriate thera-
py for patients with chronic critical limb ischemia, directed at
the prevention of limb loss and its accompanying disability. By
contrast, surgical intervention is rarely indicated in patients
with intermittent claudication alone, since the risk of major
amputation is very low. Only patients whose symptoms are
limiting to their lifestyle or performance of an occupation
are considered for endovascular or surgical revascularization.
There are two basic choices when surgery is considered for
chronic lower extremity disease, endarterectomy and bypass
grafting. 105 Despite these observations, the mechanism of the
pervasive and highly predictable localization of peripheral
vascular occlusive disease in the lower extremities remains to
be determined.

Conclusion

Arterial structural characteristics, the response of endothelial
and smooth muscle cells to tensile and shear stress, the infiltra-
tion of monocytes and other inflammatory cells, physiologic
adaptation and remodeling, in addition to pathogenic attenu-
ation and plaque formation, all play an important role in the
localization of atherosclerotic plaque to a few widely recog-
nized areas of the arterial system. Although the molecular
mechanisms responsible for this localization remain obscure,
much has been learned about the physical and cellular forces
responsible for this phenomenon. Further investigations
using real-time flow and plaque imaging modalities in vivo,
smooth muscle cell culture and proliferation studies in vitro,
and the development of sophisticated in vivo models of flow
separation, wall shear stress, and particle residence time
promise to provide more specific mechanistic clues regarding
the phenomenon of arterial atherosclerotic plaque localiza-
tion, as well as the related clinical problems of anastomotic
fibrointimal hyperplasia and vein graft stenosis.

References

1. Tsao CH, Glagov S. Basal endothelial attachment: tenacity at
cytoplasmic dense zone in the rabbit aorta. Lab Invest 1970;
23:520.

2. Rubanyi GM, Freay AD, Kauser K et ah Mechanoreception by the
endothelium: mediators and mechanisms of pressure and flow
induced vascular responses. Blood Vessels 1990; 27:246.

3. Kourembanas S, Marsden PA, McQuillan LP et ah Hypoxia

chapter 5 Localization of atherosclerotic lesions

induces endothelin gene expression and secretion in cultured
endothelium. / Clin Invest 1991; 88:1054.

4. Cooke JP, Rossitch E Jr, Andon NA, Loscalzo J, Dzau VJ. Flow ac-
tivates an endothelial potassium channel to release an endoge-
nous nitrovasodilator. / Clin Invest 1991; 88:1663.

5. Kojda G, Cheng YC, Burchfield J, Harrison DC Dysfunctional
regulation of endothelial nitric oxide synthase (eNOS) expres-
sion in response to exercise in mice lacking one eNOS gene. Cir-
culation 2001; 103:2839.

6. Davies PF, Shi C, Depaola N, Helmke BP, Polacek DC. Hemody-
namics and the focal origin of atherosclerosis: a spatial approach
to endothelial structure, gene expression, and function. Ann NY
Acad Sci 2001; 947:7; discussion 16.

7. Berk BC, Abe JI, Min W, Surapisitchat J, Yan C. Endothelial
atheroprotective and anti-inflammatory mechanisms. Ann NY
Acad Sci 2001; 947:93; discussion 109.

8. Fisher AB, Chien S, Barakat AI, Nerem RM. Endothelial cellular
response to altered shear stress. Am J Physiol Lung Cell Mol
Physiol 2001; 281:L529.

9. Nilius B, Droogmans G. Ion channels and their functional role in
vascular endothelium. Physiol Rev 2001; 81:1415.

10. Stamatas GN, Mclntire LV. Rapid flow-induced responses in en-
dothelial cells. Biotechnol Prog 2001; 17:383.

11. van Nieuw Amerongen GP, Draijer R, Vermeer MA, van
Hinsbergh VW. Transient and prolonged increase in endothelial
permeability induced by histamine and thrombin: role of protein
kinases, calcium, and RhoA. Circ Res 1998; 83:1115.

12. Ehringer WD, Wang OL, Haq A, Miller FN. Bradykinin and
alpha-thrombin increase human umbilical vein endothelial
macromolecular permeability by different mechanisms. Inflam-
mation 2000; 24:175.

13. Jiang L, Jha V, Dhanabal M, Sukhatme VP, Alper SL. Intracellular
Ca(2+) signaling in endothelial cells by the angiogenesis in-
hibitors endostatin and angiostatin. Am J Physiol Cell Physiol
2001; 280:0140.

14. Cooke JP, Rossitch E, Andon NA et ah Flow activates an endothe-
lial potassium channel to release an endogenous nitrovasodila-
tor. / Clin Invest 1991; 88:1663.

15. Dull RO, Davies PF. Flow modulation of agonist (ATP) response
(Ca ++ ) coupling in vascular endothelial cells. Am } Physiol 1991;
261:H149.

16. Glagov S, Zarins C, Giddens DP et ah Hemodynamics and
atherosclerosis: insights and perspectives gained from study
of human arteries. Arch Pathol Lab Med 1998; 152:1019.

17. Ross R. Cellular and molecular studies of atherogenesis. Athero-
sclerosis 1997; 131(Suppl.):S3.

18. Koyama H, Olson NE, Dastvan FF, Reidy MA. Cell replication
in the arterial wall: activation of signaling pathway following
in vivo injury. Circ Res 1998; 82:713.

19. Linder V, Reidy MA, Fingerle J. Regrowth of arterial endo-
thelium: denudation with minimal trauma leads to complete
endothelial cell regrowth. Lab Invest 1989; 61 :556.

20. Mason RP Mechanisms of atherosclerotic plaque stabilization
for a lipophilic calcium antagonist amlodipine. Am } Cardiol 2001;
88:2M.

21. Huo Y, Ley K. Adhesion molecules and atherogenesis. Acta
Physiol Scand 2001; 173:35.

22. Gibbons GH. Endothelial function as a determinant of vascular
function and structure: a new therapeutic target. Am J Cardiol

1997; 79:3.

23. Gorlatov S, Medved L. Interaction of fibrin(ogen) with the en-
dothelial cell receptor VE-cadherin: mapping of the receptor-
binding site in the NH(2)-terminal portions of the fibrin beta
chains. Biochemistry 2002; 41:4107.

24. Stupack DG, Cheresh DA. ECM remodeling regulates angio-
genesis: endothelial integrins look for new ligands. Sci STKE
2002;110:PE7.

25. Kliche S, Waltenberger J. VEGF receptor signaling and endothe-
lial function. IUBMB Life 2001; 52:61.

26. Ma L, Wang X, Zhang Z, Zhou X, Chen A, Yao L. Identification of
the ligand-binding domain of human vascular-endothelial-
growth-f actor receptor Flt-1. Biotechnol Appl Biochem 2001; 34:199.

27. Witzum JL, Steinberg D. Role of oxidized low density lipoprotein
in atherogenesis. / Clin Invest 1991; 88:1785.

28. Dzau VJ, Gibbons GH, Cooke JP et ah Vascular biology and
medicine in the 1990s: scope, concepts, potentials, and perspec-
tives. Circulation 1993; 87:705.

29. Clark JM, Glagov S. Structural integration of the arterial wall. I.
Relationships and attachments of medial smooth muscle cells in
normally distended and hyperdistended aortas. Lab Invest 1979;
40:587.

30. Clark JM, Glagov S. Transmural organization of the arterial wall:
the lamellar unit revisited. Arteriosclerosis 1985; 5:19.

31 . Wolinsky H, Glagov S. A lamellar unit of aortic medial structure
and function in mammals. Circ Res 1967; 20:99.

32. Wolinsky H, Glagov S. Structural basis for the static mechanical
properties of the aortic media. Circ Res 1964; 14:400.

33. Mayne R. Collagenous proteins of blood vessels. Arteriosclerosis
1986; 5:585.

34. Cantini C, Kieffer P, Corman B, Liminana P, Atkinson J, Lartaud-
Idjouadiene I. Aminoguanidine and aortic wall mechanics,
structure, and composition in aged rats. Hypertension 2001;
38:943.

35. Wolinsky H, Glagov S. Nature of species differences in the
medial distribution of aortic vasa vasorum in mammals. Circ Res
1967; 20:409.

36. Leung DYM, Glagov S., Mathews MB. Elastin and collagen accu-
mulation in rabbit ascending aorta and pulmonary trunk during
postnatal growth: correlation of cellular synthetic response with
medial tension. Circ Res 1977; 41:316.

37. Wolinsky H. Long-term effects of hypertension on the rat aortic
wall and their relation to concurrent aging changes. Circ Res 1972;
30:301.

38. Taber LA. A model for aortic growth based on fluid shear and
fiber stresses. / Biomech Eng 1998; 120:348.

39. Crawford T, Levene CI. Medial thinning in atheroma. / Pathol
1953; 66:19.

40. Leung DYM, Glagov S, Mathews MB. A new in vitro system for
studying cell response to mechanical stimulation: different ef-
fects of cyclic stretching and agitation on smooth muscle cell
biosynthesis. Exp Cell Res 1977; 109:285.

41. Paterson JC. Vascularization and haemorrhage of the intima of
arteriosclerotic coronary arteries. Arch Pathol 1936; 22:312.

42. Zarins CK, Glagov S, Vesselinovitch D et ah Aneurysm formation
in experimental atherosclerosis: relationship to plaque evolu-
tion. / Vase Surg 1990; 12:246.

43. Strickland HL, Bond MG. Aneurysms in a large colony of squirrel
monkeys (Saimirisciures). LabAnim Sci 1983; 33:589.

pa rt I Vascular pathology and physiology

44. Burke AP, Weber DK, Kolodgie FD, Farb A, Taylor AJ, Virmani R.
Pathophysiology of calcium deposition in coronary arteries.
Herz 2001; 26:239.

45. Glagov S, Eckner FAO, Lev M. Controlled pressure fixation ap-
paratus for hearts. Arch Pathol 1963; 76:640.

46. Zarins CK, Zatina MA, Glagov S. Correlation of postmortem
angiography with pathologic anatomy: quantitation of athero-
sclerotic lesions. In: Bond MG, Insull W Jr, Glagov S et al., eds.
Clinical Diagnosis of Atherosclerosis. New York, Springer- Verlag,
1983:283.

47. Mulvihill DA, Harvey SC. The mechanism of the development
of collateral circulation. N Engl] Med 1931; 104:1032.

48. Holman E. Problems in the dynamics of blood flow. I. Condition
controlling collateral circulation in the presence of an arteriove-
nous fistula, following ligation of an artery. Surgery 1949;26:889.

49. Kamiya A, Togawa T. Adaptive regulation of wall shear stress
to flow change in the canine carotid artery. Am } Physiol 1980;
239:H14.

50. Furchgott RF. The 1996 Albert Lasker Medical Research Awards.
The discovery of endothelium-derived relaxing factor and its
importance in the identification of nitric oxide. JAMA 1996;
276:1186.

51. Langille BL, O'Donnell F. Reductions in arterial diameter pro-
duced by chronic decreases in blood flow are endothelium de-
pendent. Science 1986; 231:405.

52. McGill HC Jr. Atherosclerosis: problems in pathogenesis. In:
Paoletti R, Gotto AM, eds. Atherosclerosis Reviews. New York,
Raven Press, 1977:27.

53. Thubrikar M, Baker J, Nolan S. Inhibition of atherosclerosis asso-
ciated with reduction of arterial intramural stenosis in rabbits.
Arteriosclerosis 1988; 8:410.

54. Friedman MH. Some atherosclerosis maybe a consequence of the
normal adaptive vascular response to shear. Atherosclerosis 1990;
82:193.

55. Zarins CK, Giddens DP, Bharadvaj BK et al. Carotid bifurcation
atherosclerosis: quantitative correlation of plaque localization
with flow velocity profiles and wall shear stress. Circ Res 1983;
53:502.

56. Ku DN, Zarins CK, Giddens DP et al. Pulsatile flow and athero-
sclerosis in the human carotid bifurcation: positive correlation
between plaque localization and low and oscillating shear stress.
Arteriosclerosis 1985; 5:292.

57. Karino T. Microscopic structure of disturbed flows in the arterial
and venous systems and its implication in the localization of vas-
cular disease. Int Angiol 1986; 5:297.

58. Glagov S, Rowley DA, Kohut R. Atherosclerosis of human aorta
and its coronary and renal arteries. Arch Pathol 1961; 72:558.

59. Svindland A. The localization of sudanophilic and fibrous
plaques in the main left coronary arteries. Atherosclerosis 1983;
48:139.

60. Fingerle J, Johnson R, Clowes AW. Role of platelets in smooth
muscle cell proliferation and migration after vascular injury in
rat carotid artery. Proc Natl Acad Sci USA 1989; 86:8412.

61. Bassiouny HS, Lieber BB, Giddens DP et al. Quantitative inverse
correlation of wall shear stress with experimental intimal thick-
ening. Surg Forum 1988; 39:328.

62. Lyon RT, Hass A, Davis HR. Protection from atherosclerotic
lesion formation by reduction of artery wall motion. / Vase Surg
1987; 5:59.

63. Caro GG, Fish PJ, Ja M et al. Influence of vasoactive agents on
arterial hemodynamics: possible relevance to atherogenesis.
Biorheology 1986; 23:197.

64. Glagov S. Microarchitecture of arteries and veins. In: Abramson
D, Dobrin P, eds. Blood Vessels and Lymphatics. Orlando, Academ-
ic Press, 1984:3.

65. Glagov S, Zarins CK, Giddens DP et al. Hemodynamics and
atherosclerosis: insights and perspectives gained from studies
of human arteries. Arch Pathol Lab Med 1988; 112:1018.

66. Glagov S, Zarins CK, Giddens DP et al. Establishing the hemody-
namic determinants of human plaque configuration, composi-
tion and complication. In: Yoshida Y, Yamaguchi T, Caro CG et al.,
eds. Role of Blood Flow in Atherogenesis. New York, Springer-
Verlag, 1988:3.

67. Born VRG, Richardson PD. Mechanical properties of human
atherosclerotic lesions. In: Glagov S, Newman WP, Schaffer SA,
eds. Pathobiology of the Human Atherosclerotic Plaque. New York,
Springer- Verlag, 1990:413.

68. Zarins CK, Bomberger RA, Taylor KE et al. Artery stenosis in-
hibits regression of diet-induced atherosclerosis. Surgery 1980;
88:86.

69. Xu C, Glagov S, 7,&\mdL M. Hypertension sustains plaque pro-
gression despite reduction of hypercholesterolemia. Hyperten-
sion 1991; 18:123.

70. Zarins CK, Zatina MA, Giddens DP et al. Shear stress regulation
of artery lumen diameter in experimental atherogenesis. / Vase
Surg 1987; 5:413.

71. Fry DL. Problems and progress in understanding "endothelial
permeability" and mass transport in human arteries. In:
Glagov S. Newman WP, Schaffer SA, eds. Pathobiology of the
Human Atherosclerotic Plaque. New York, Springer- Verlag, 1990:
271.

72. Smith EB. Accumulating evidence from human artery studies of
what is transported and what accumulates relative to atherogen-
esis. In: Glagov S, Newman WP, Schaffer SA, eds. Pathobiology of
the Human Atherosclerotic Plaque. New York, Springer-Verlag,
1990.

73. Tracy RE, Kissling GE. Comparisons of human populations for
histologic features of atherosclerosis. Arch Pathol Lab Med 1988;
112:156.

74. Frank JS, Fogelman AM. Ultrastructure of the intima in WHHL
and cholesterol-fed rabbit aortas prepared by ultra-rapid freez-
ing and heeze-etching. J Lipid Res 1989; 30:967.

75. Berenson GS, Radhakrishnamurthy B, Srinivasan SR et al.
In: Glagov S, Newman WP, Schaffer SA, eds. Pathobiology of
the Human Atherosclerotic Plaque. New York, Springer-Verlag,
1990:189.

76. Libby P, Schoenbeck U, Mach F, Selwyn AP, Ganz P. Current con-
cepts in cardiovascular pathology: the role of LDL cholesterol in
plaque rupture and stabilization. Am] Med 1998; 104:14S.

77. Wagner WD, Edwards IJ, St. Clair RW et al. Low density lipopro-
tein interaction with artery derived proteoglycan: the influence
of LDL particle size and the relationship to atherosclerosis sus-
ceptibility. Atherosclerosis 1989; 75:49.

78. Grande J, Davis HR, Bates S et al. Effect of an elastin growth
substrate on cholesteryl ester synthesis and foam cell formation
by cultured aortic smooth muscle cells. Atherosclerosis 1987;
68:87.

79. Friedman MH, Henderson JM, Aukerman JA, Clingan PA. Effect

chapter 5 Localization of atherosclerotic lesions

of periodic alterations in shear on vascular macromolecular up-
take. Biorheology 2000; 37:265.

Ku DN, Giddens DP. Pulsatile flow in a model carotid bifurca-
tion. Arteriosclerosis 1983; 3:31.

Svindland A. The localization of sudanophilic and fibrous
plaques in the main left coronary arteries. Atherosclerosis 1983;
48:139.

82. Beere PA, Glagov S, Zarins CK. Experimental atherosclerosis at
the carotid bifurcation of the cynomolgus monkey. Arteriosclero-
sis and Thrombosis 1992; 12:1245.

83. Beere PA, Glagov S, Zarins CK. Retarding effect of lowered heart
rate on coronary atherosclerosis. Science 1984; 226:180.

84. Manuck SB, Kaplan JR, Clarkson TB. Behaviorally induced heart
rate activity and atherosclerosis in cynomolgus monkeys. Psy-
chosom Med 1983; 56:27.

85. Kaplan JR, Clarkson TB. Social instability and coronary artery
atherosclerosis in cynomolgus monkey. Neurosci Biobehav Rev
1983; 7:485.

86. Dyer A, Persky V, Stamler J et ah Heart rate as a prognostic factor
for coronary heart disease and mortality: findings in three
Chicago epidemiologic studies. Am] Epidemiol 1980; 112:736.

87. Gillum RF. The epidemiology of resting heart rate in a national
sample of men and women: associations with hypertension,
coronary heart disease, blood pressure and other cardiovascular
risk factors. Am Heart] 1988; 116:163.

Kannel W, Kannel C, Paffenbarger R et ah Heart rate and cardio-
vascular mortality: the Framingham study. Am Heart J 1987;
113:1489.

Glagov S. Hemodynamic risk factors: mechanical stress, mural
architecture, medial nutrition and the vulnerability of arteries to
atherosclerosis. In: Wissler RW, Geer JC, eds. The Pathogenesis of
Atherosclerosis. Baltimore, Williams & Wilkins, 1972:164.
Ku DN, Glagov S, Moore JE, Zarins CK. Flow patterns in the ab-
dominal aorta under simulated postprandial and exercise condi-
tions: an experimental study. / Vase Surg 1989; 9:309.

91. Moore JE, Ku DN, Zarins CK, Glagov S. Pulsatile flow visualiza-
tion in the abdominal aorta under differing physiological condi-
tions: implications for increased susceptibility to atherosclerosis.
JBiomech Eng 1992; 114:391.

92. Moore JE, Ku DN. Pulsatile velocity measurements in a model of
the human abdominal aorta under resting conditions. / Biomech
Eng 1994; 116:337.

93. Moore JE, Ku DN. Pulsatile velocity measurements in a model of
the human abdominal aorta under simulated exercise and post-
prandial conditions. JBiomech Eng 1994; 116:107.

94. Moore JE, Maier SE, Ku DN, Boesiger P. Hemodynamics in the
abdominal aorta: a comparison of in vitro and in vivo measure-
ments. JAppl Physiol 1994; 76:1520.

95. Moore JE, Xu C, Glagov S, Zarins CK, Ku DN. Fluid wall shear
stress measurements in a model of the human abdominal aorta:
oscillatory behavior and relationship to atherosclerosis. Athero-
sclerosis 1994; 110:225.

96. Zarins CK, Taylor CA. Hemodynamic factors in atherosclerosis.
In: Moore W, ed. Vascular Surgery: A Comprehensive Review.
Philadelphia: W.B. Saunders Co., 1998:97.

97. Taylor CA, Hughes TJR, Zarins CK. Finite element modeling of
3-dimensional pulsatile flow in the abdominal aorta: relevance
to atherosclerosis. Ann Biomed Eng 1998; 26:975.

98. Taylor C A, Hughes TJR, Zarins CK. Effect of exercise on hemody-
namic conditions in the abdominal aorta. / Vase Surg 1998;
29:1077.

99. Cornhill JF, Herderick EE, Stary HC. Topography of human
aortic sudanophilic lesions. Monogr Atheroscler 1990; 15:13.

100. Zarins CK, Xu C, Glagov S. Atherosclerotic enlargement of the
human abdominal aorta. Atherosclerosis 2001; 155:157.

101. Xu C, Zarins CK, Glagov S. Aneurysmal and occlusive athero-
sclerosis of the human abdominal aorta. / Vase Surg 2001; 33:
91.

102. Gordon T, Kannel WB. Predisposition to atherosclerosis in the
head, heart and legs: the Framingham study. JAMA 1972;
221:661.

103. Winniford MD, Wheelan KR, Kremers MS et ah Smoking in-
duced coronary vasoconstriction in patients with atherosclerotic
coronary artery disease: evidence for adrenergically medicated
alterations in coronary artery tone. Circulation 1986; 73:662.

104. Schainfeld RM. Management of peripheral arterial disease and
intermittent claudication. J Am Board Earn Pract 2001; 14:443.

105. Ouriel K. Peripheral arterial disease. Eancet 2001; 358:1257.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Pathogenesis of arterial f ibrodysplasia

James C.Stanley

Arterial fibrodysplasia is a term used to describe a heterogeneic
group of nonarteriosclerotic, noninflammatory occlusive
and aneurysmal diseases. 1 ' 2 Dysplastic lesions have been
observed in most medium size muscular arteries, occasionally
in smaller arteries in many tissues, and, rarely, in the
aorta. In certain instances, the dysplastic lesion is not an
isolated disease but appears to represent a secondary pro-
cess affecting vessels exhibiting an underlying primary
disease. Nevertheless, three distinct forms of arterial fibrodys-
plasia have been extensively reported: medial fibroplasia,
perimedial dysplasia, and intimal fibroplasia. A fourth
lesion, developmental arterial dysplasia, has also been
recognized. These specific types of arterial dysplasia deserve
particular notice to those interested in vessel wall biology and
pathology.

Medial fibroplasia and perimedial dysplasia

Medial fibrodysplasia is the most common dysplastic arterial
disease encountered in muscular arteries. 2 This is a unique dis-
ease that in its classic form invariably affects women after the
onset of their reproductive years. This subgroup of dysplastic
diseases is most frequently encountered in whites, is uncom-
mon among Asians, and is rare among blacks. It represents a
systemic arteriopathy in certain patients, and is most evident
in the renal, extracranial internal carotid, and external iliac ar-
teries. Medial fibroplasia accounts for nearly 85% of dysplastic
renovascular disease and 90% of similar disease affecting the
internal carotid arteries.

The morphologic character of renal artery medial fibrodys-
plasia ranges from a single focal stenosis to the more common
series of stenoses with intervening aneurysmal outpouchings
causing a string-of-beads appearance (Fig. 6.1). 3 The mural
aneurysms affecting these arteries are usually grossly evident,
although the stenotic webs projecting internally are more ob-
vious on arteriographic studies than by direct inspection. Me-
dial fibroplasia affects the middle or distal main renal artery in
75% of cases; extensions into first order segmental branches

occur in approximately 25% of cases. Proximal lesions are
uncommon.

Progression of renal artery medial fibroplasia has been re-
ported to affect as few as one-eighth to as many as two-thirds
of patients with main renal artery lesions. 4-6 Progression is
more apt to affect premenopausal women, yet some have
noted no differences related to age. 7 Among potential kidney
donors with incidental angiographic diagnosis of renal artery
fibrodysplasia, hypertension developed in 26% over an aver-
age follow-up of 7.5 years. 8 In contrast, hypertension devel-
oped in only 6% of an age- and sex-matched group of control
patients. Blood pressure increases in these instances were as-
sumed to be a reflection of progressive renal artery disease. Re-
gression of renal artery dysplastic stenoses has been reported,
although such is open to question in that catheter-induced
spasm in certain cases may have led to an overestimation of the
initial lesion's severity.

Medial fibrodysplasia of the extracranial internal carotid
artery (ECICA) typically involves a 2- to 6-cm segment of the
carotid artery as a series of stenoses and mural dilations adja-
cent to the second and third cervical vertebrae (Fig. 6.2). 2 ' 9
Bilateral disease has been reported to affect 35-85% of pa-
tients. 9 ' 10 Medial fibrodysplasia isolated to the origin of the
ECICA has not been described. Carotid arteries affected by
medial fibrodysplasia are often elongated. This may occur as
a consequence of vessel stretching or may occur as sequelae
of the dysplastic disease itself. Kinking of moderate or severe
degrees occurs in approximately 5% of these cases. Typical
medial fibrodysplastic lesions of the intracranial arteries are
uncommon. Similar lesions of the external carotid artery or
its branches have been reported but are exceedingly rare.
Progression of ECICA medial fibrodysplasia may occur in as
many as one-third of affected patients. 9 ' 10

Extracerebrovascular medial fibrodysplasia is common
among patients with ECICA lesions. This is particularly true of
renal artery involvement, where as many as 25% of these indi-
viduals have been reported to exhibit medial dysplasia. 9 The
frequency of simultaneous ECICA and renal artery dysplasia
may be even higher, in that few series have reported patients

chapter 6 Pathogenesis of arterial fibrodysplasia

Figure 6.1 Medial fibroplasia. Serial stenoses alternating with mural
aneurysms, producing a string-of-beads appearance in the middle and distal
main renal artery. (From Stanley JC, Graham LM. Renovascular hypertension.
In: Miller DC, Roon AJ, eds. Diagnosis and Management of Peripheral
Vascular Disease. Menlo Park, CA: Addison-Wesley, 1981 :23 1-235.)

subjected to arteriography assessments of both cerebral and
renal vessels. Similar medial fibrodysplastic lesions have also
been observed in the external iliac and superior mesenteric
arteries of patients with lesions of the ECICA.

Coexistent intracranial aneurysms have been documented
in one-eighth to one-quarter of patients with ECICA medial
fibrodysplasia. 9 Although intracranial arteries are occasion-
ally the site of dysplastic disease, aneurysms do not necessari-
ly develop in the diseased artery. Instead, they appear to
evolve at arterial branches similar to usual berry aneurysm
formation. 2 Systemic hypertension may contribute to the
development of these aneurysms. A propensity for these
aneurysms to occur ipsilateral to the ECICA disease is of
dubious importance but has been reported. 11 In fact, the distri-
bution of intracranial aneurysms in patients with medial
fibrodysplasia is the same as in patients not affected with dys-
plastic ECICA. 9

The external iliac artery is the third most common vessel
to exhibit medial fibrodysplasia. 12 Serial stenoses with inter-
vening mural aneurysms are characteristically evident in the
proximal third of affected vessels (Fig. 6.3). Like other lesions
of this group, fibroproliferative stenoses appear adjacent to
areas of relative medial thinning. Occasional solitary dilations
have been attributed to medial fibrodysplasia of the iliac ves-
sels. Other extremity lesions reflecting the systemic nature
of medial fibroplasia have been reported in the femoral,
popliteal, and tibial vessels. 2 ' 13 Splanchnic arterial medial
fibrodysplasia is rare. 2 Splanchnic lesions are often associated

Figure 6.2 Medial fibrodysplasia of the extracranial internal carotid artery
adjacent to the second and third cervical vertebrae, with characteristic serial
stenoses alternating with mural aneurysms. (From Stanley JC, Fry WJ, Seeger
JF, Hoffman GL, Gabrielson TO. Extracranial internal carotid and vertebral
artery fibrodysplasia. Arch Surg 1974; 109:2 15.)

with similar renal or carotid lesions. Histologic evidence
of medial dysplasia is common in splenic arteries with
aneurysms. Similar aneurysms have been noted in other dys-
plastic splanchnic vessels. The proximal superior mesenteric
artery may exhibit medial fibrodysplastic occlusive disease a
few centimeters beyond its origin as it exits beneath the pan-
creas over the top of the duodenum. The basis for these latter
lesions has not been established, although unusual stretch
forces at the root of the mesentery may contribute to dysplastic
changes.

Two histologic forms of medial fibroplasia are well recog-
nized (Fig. 6.4), with disease either limited to the outer media
(peripheral form) or affecting the entire media (diffuse form).
The latter occurs twice as often as the former, and gradations
between these extremes have been observed in the same ves-
sel, supporting the tenet that they represent variations of the
same disease process.

The peripheral form of medial fibroplasia appears to occur
first, and in time progresses to more diffuse disease affecting
the entire media. Thus, multiple severe stenoses with inter-

pa rt I Vascular pathology and physiology

Figure 6.3 Medial fibrodysplasia of external iliac artery. Multiple stenoses
with intervening mural dilations. (From Walter JF, Stanley JC, Mehigan JT,
RueterSR, GuthanerDF. External iliacartery fibrodysplasia. Am J Roentgenol
1978; 131:125.)

vening mural aneurysms have evolved in certain patients who
initially had a solitary lesion or a few stenoses of minimal
severity. 3

Peripheral medial fibroplasia exhibits fibrous connective
tissue replacing normal smooth muscle in the outer one-third
or one-half of the media. 3 Moderate accumulations of pro-
teinaceous ground substances may be evident between disor-
ganized smooth muscle cells of the inner media. The intima,
internal elastic lamina, and adventitia are usually normal
in the peripheral form of this disease. In earlier reports,
peripheral forms of medial fibroplasia were considered, per-
haps erroneously, to represent perimedial or subadventitial
disease. 14 ' 15 The fact is these lesions are limited to medial tissue
in almost all cases.

Diffuse medial fibrodysplasia exhibits more severe disorga-
nization and disruption of the normal medial smooth muscle. 3
Accumulations of fibrous tissue alternate with areas of
marked medial thinning (Fig. 6.5). The media may be nearly
absent in regions of mural aneurysms. Internal elastic lamina
fragmentation and subendothelial fibrosis may occur as
secondary events in more advanced disease. However, even in
the more extensive lesions, adventitial tissues are relatively
uninvolved.

Perimedial dysplasia appears to be the dominant abnormality
in approximately 10% of dysplastic renal arteries. 2 ' 3 This le-
sion, as an isolated entity, has not been observed in extrarenal

muscular arteries. It may coexist with medial fibrodysplasia
and appears to share some of the same etiologies with the
former.

Most patients exhibiting perimedial dysplasia have been
women in their forties or fifties. 3 Focal stenoses or multiple
constrictions without mural aneurysms involving the mid-
portion of the main renal artery characterize perimedial dys-
plasia (Fig. 6.6). Excess elastic tissue at the junction of the
media and adventitia is the distinguishing feature of these le-
sions (Fig. 6.7). Increases in medial ground substances among
medial smooth muscle cells are not uncommon within the
inner medial tissues.

Certain similar ultrastructural features related to accumula-
tions of ground substance and fibrous elements are apparent
in medial fibroplasia and perimedial dysplasia. 16 Perimedial
dysplasia is differentiated from medial fibrodysplasia by
fewer and less obvious changes in the inner media, and
accumulations of amorphous proteinaceous material and
elastic tissue at the adventitial-medial border. The earliest
ultrastructural changes in smooth muscle of these dysplastic
vessels include focal myofilament reductions, as well as peri-
nuclear sublemmal and cytoplasmic vacuolations (Fig. 6.8). 16

Smooth muscle cells in more advanced disease exhibit
either extreme deterioration or a fibroblast-like appearance.
The former cells become islated from surrounding cells by ex-
cess ground substances (Fig. 6.9 A). 16 Cell membranes are often
indistinct, and the nucleus is usually pyknotic, containing
dense chromatin material. Combined with sparse subcellular
organelles, it appears that these cells are nearing death and are
certainly not functioning in a normal manner. These changes
are noted in vessel wall segments that are aneurysmal.

Modification of medial smooth muscle cells to fibroblast
type cells, or so-called myofibroblasts, also occurs in these tis-
sues (see Fig. 6.9B). Loss of myofilaments and increases in free
ribosomes, rough endoplasmic reticulum, Golgi complexes,
and mitochondria are compatible with altered cellular
function from one of contractility to one of secretion.
Myofibroblasts in this setting appear to be the end-product of
smooth muscle transformation and are characteristically
found in the stenotic regions of these lesions. 16 These cells ex-
hibit a convoluted nucleus with numerous indentations and
evaginations characteristic of smooth muscle (see Fig. 6.9B),
with marked increases in organelles and the presence of
peripherally located cytoplasmic filaments characteristic of
myofibroblasts.

Exopinocytotic deposition of proteinaceous matter may be
evident in these highly secretory cells (Fig. 6.10). Vasa vaso-
rum within the media of diseased arteries are usually widely
separated from adjacent cellular tissue by homogeneous
mucoid ground substances and tubular fibrous elements.
Vasa vasorum within medial fibroplasia are surrounded
predominantly by collagen fibrous bundles, whereas those
in perimedial dysplasia are usually surrounded by more
amorphous substances including elastic tissue. 16

chapter 6 Pathogenesis of arterial fibrodysplasia

Figure 6.4 (A) Peripheral form, medial fibrodysplasia; dense fibrous
connective tissue in the outer media, with disordered inner medial smooth
muscle, and normal intimal tissue. (B) Diffuse form, medial fibroplasia; total
replacement of media by disorganized cellular tissue (myofibroblasts)

surrounded by fibrous connective tissue (Masson stain, x1 20). (From Stanley
JC. Morphologic, histopathologic and clinical characteristics of renovascular
fibrodysplasia and arteriosclerosis. In: Bergan JJ, Yao JST, eds. Surgery of the
Aorta and Its Body Branches. New York: Grune&Stratton, 1979:355-376.)

Figure 6.5 Diffuse form, medial fibroplasia.
Regions of excessive fibroproliferation with
intervening area of medial thinning (Masson
stain, x60 longitudinal section). (From Stanley
JC. Morphologic, histopathologic, and clinical
characteristics of renovascular fibrodysplasia
and arteriosclerosis. In: Bergan JJ, Yao JST, eds.
Surgery of the Aorta and Its Body Branches.
New York: Grune & Stratton, 1 979:355-376.)

pa rt I Vascular pathology and physiology

Figure 6.6 Perimedial dysplasia. Multiple stenoses without mural
aneurysms in the midportion of the renal artery are characteristic of this
dysplastic lesion. (From Stanley JC. Morphologic, histopathologic and clinical
characteristics of renovascular fibrodysplasia and arteriosclerosis. In: Bergan
JJ, Yao JST, eds. Surgery of the Aorta and Its Body Branches. New York:
Grune&Stratton, 1979:355-376.)

The causes of medial fibroplasia and perimedial dysplasia
are poorly understood but certain contributing factors have
been suggested to be important. 3 Hormonal effects, mechani-
cal stresses, and potential ischemic events affecting vascular
smooth muscle may all play important roles in the etiology
of these lesions. To date, there have been no investigations
regarding the presence of commonly recognized mitogenic
growth factors in these lesions, as might be studied by conven-
tional immunocytochemical means or molecular analyses for
differing gene expression. Because of the occasional familial
nature of this disease, a genetic-related autosomal dominant
etiology with incomplete penetrance has been proposed, yet
evidence to establish such a contention convincingly has not
been forthcoming. 17-20 In fact, the female sex predilection for
this disease has not been explained by data generated by those
supporting a primary genetic etiology for this form of arterial
dysplasia.

Hormonal influences on vascular smooth muscle may ex-
plain medial and perimedial arterial dysplasia's unusual
female predilection. 3 Certain smooth muscle cells and
fibroblasts exposed to estrogens are known to increase synthe-
sis of many proteinaceous substances. It is reasonable to pre-
sume that physiologic preconditioning of vascular smooth

muscle cells to a secretory state by estrogens may account for
the more frequent occurrence of medial dysplastic disease in
ovulating women. This is in accord with the absence of these
lesions in patients prior to menarche and their lesser progres-
sion following menopause. Pregnancy does not appear to be
an obvious etiologic factor in arterial fibrodysplasia in that the
reproductive histories of patients in a large series of patients
with this arteriopathy did not reveal gravidity or parity rates
different from the general population. 3 Similarly, use of
progestin-based antiovulants by less than half the female pa-
tients in a retrospective study of this disease does not support
a major role for progesterins in arterial dysplasia. 3 A lack of
any association between oral contraceptives and these lesions
has also been reported in a case-control study. 20

Unusual physical stresses due to stretching of the renal
artery and ECICA may be associated with fibrodysplastic
changes (Fig. 6.11). Comparable stretch or traction forces are
less likely to occur in similar size muscular arteries not mani-
festing this disease. In this regard, ptotic kidneys are more
common among patients with renal artery medial fibrodys-
plasia. 21-23 The fact that the right kidney is usually more ptotic
than the left may explain why 80% of unilateral disease
involves the right kidney, and may account for the greater
severity of right-sided disease in the majority of adults with
bilateral disease. However, in one case-control study, renal
mobility was not greater in patients with renovascular
fibrodysplasia. 20 This has not been the impression of most
others, but firm data on this topic have not settled this issue.

Cyclic stretching of smooth muscle cells in tissue culture is
known to result in greater synthesis of collagen and certain
acid mucopolysaccharides. 24 Similar mechanisms in vivo are
certainly within the realm of possibility. The predilection for
dysplastic disease to occur most often in vessels subjected to
repetitive mechanic stretching may reflect such a pathogenic
process. Mural ischemia may also be a contributing factor
to arterial dysplasia. Vasa vasorum of muscular arteries are
known to originate usually from branchings of the parent ves-
sel and relatively few branches exist in the ECICA and external
iliac artery compared with similar size vessels elsewhere.
Compromise of vasa vasorum in these vessels where a
sparsity of usual nutrient vessels already exists may lead to
significant mural ischemia and eventually fibrous changes.
Vasospasm may further exacerbate vessel wall ischemia in
these cases. 25,26

The potential for impaired blood supply of the arterial wall
to cause dysplastic changes is supported by the peculiar in-
volvement of the outer media in peripheral medial fibroplasia.
It is precisely this region where ischemia would be expected
to be greatest from inadequate vasa vasorum blood flow. Fi-
brodysplasia limited to the inner part of the media has never
been reported. In fact this portion of the vessel wall may be
protected from vasa vasoral insufficiency by transluminal
diffusion of needed oxygen and nutriments. Vasa vasorum in
medial dysplastic vessels have exhibited both dilation and

chapter 6 Pathogenesis of arterial fibrodysplasia

4 .

fjrrr.4

^h*™ • r <%.

1 , ■ . m.

■

ii,
: -+ -

~**/r

■

■^

l ■^^*i?^ r '

Figure 6.7 Perimedial dysplasia. (A) Homogeneous collar of elastic tissue
adjacent to the outer media is the dominant feature of this lesion
(hematoxylin and eosin, x80). (B) Excessive accumulations of elastic tissue at
medial-adventitial junction are apparent with special staining (Verhoeff
stain, x1 20). (A from Stanley JC. Pathologic basis of macrovascular renal
artery disease. In: Stanley JC, Ernst CB, FryWJ, eds. Renovascular
Hypertension. Philadelphia: WB Saunders; 1984:46-74. Bfrom Stanley JC.
Morphologic, histopathologic and clinical characteristics of renovascular
fibrodysplasia and arteriosclerosis. In: Bergan JJ, YaoJST, eds. Surgery of the
Aorta and Its Body Branches. New York: Grune&Stratton, 1979:355-376.)

pa rt I Vascular pathology and physiology

Figure 6.8 Smooth muscle cell. (A) In the
region of minimal fibrodysplasia, there is a
relatively normal ultrastructure except for focal
reduction in myofilaments and the appearance
of perinuclear, sublemmal, as well as
cytoplasmic vacuoles (TEM, x1 8 000). (B) In the
region of moderate fibrodysplasia, more
extensive perinuclear and peripheral
vacuolation is evident. Loss of organelles,
basement membrane, and indistinct
myofilaments characterize this type cell (TEM,
x12 000). (From Stanley JC. Pathologic basis of
macrovascular renal artery disease. In: Stanley
JC, Ernst CB, Fry WJ, eds. Renovascular
Hypertension. Philadelphia: WB Saunders,
1984:46-74.)

isolation from adjacent medial smooth muscle cells. 16 ' 27 The
importance of these observations is poorly understood.

On the other hand, experimental occlusion of the vasa vaso-
rum produces fibrodysplastic changes and supports the tenet
that mural ischemia may contribute to arterial dysplasia. 28 Al-
tered tissue pH, accumulation of metabolites, or factors other
than hypoxia may also be important in the pathogenesis of
arterial dysplasia. In this regard, cigarette smoking has been
implicated as a potentially important contributing factor in
this disease, although the mechanisms surrounding this have
not been defined. 20

Intimal fibroplasia

Intimal fibroplasia of the renal artery affects men and women
equally, and is observed in infants, adolescents, and young
adults more often than among the elderly 2 This lesion ac-
counts for approximately 5% of all dysplastic renal artery
stenoses. It accounts for a much greater proportion of cere-
brovascular, extremity, and splanchnic dysplastic lesions. Inti-
mal lesions appear to progress at a slower rate than do medial
fibroplastic stenoses. 5 Progression of intimal fibrodysplasia,

chapter 6 Pathogenesis of arterial fibrodysplasia

Figure 6.9 (A) Smooth muscle cell in area of
advanced fibrodysplasia. Isolation of slender
cytoplasmic processes by excesses in ground
substances, and pyknotic nuclei were typical of
these markedly abnormal cells (TEM,x6000).
(B) Myofibroblast associated with medial
fibroplasia. Convoluted nucleus is typical of
smooth muscle but increased numbers of
centrally located organelles reflect change in
function from one of contractility to secretion
(TEM, x8000). SM, smooth muscle; GS, ground
substance; CP, cytoplasmic processes; mf,
myofilament; DB, dense body; RER, rough
endoplasmic reticulum; MF, myofibroblast; mf,
myofibril; GC, Golgi complex; BM, basement
membrane. (From SottiuraiVS, Fry WJ, Stanley
JC. Ultrastructure of medial smooth muscle and
myofibroblasts in human arterial dysplasia. Arch
5t/rg 1978;1 13:1280.)

once a hemodynamically important arterial stenosis develops,
is a likely consequence of abnormal blood flow, even if other
etiologic factors have resolved. The specific cellular messen-
gers responsible for this tissue proliferation have not been
identified.

Primary intimal fibroplasia occurs most often as a smooth
focal stenosis (Fig. 6.12A). Segmental renal artery or internal
carotid artery involvement is a more uncommon manifesta-
tion of intimal disease, usually presenting as a web-like lesion
(see Fig. 6.12B). Irregularly arranged subendothelial mes-
enchymal cells within a loose connective tissue matrix charac-
terize primary intimal fibroplasia (Fig. 6.13). 3 The internal
elastic lamina, although occasionally discontinuous, is usu-

ally intact. Primary intimal disease is usually circumferential.
Its cause is unknown but in certain cases it may represent
persistent neonatal arterial musculoelastic cushions. Lipid-
containing cells or inflammatory cells are not a factor in the
primary form of this disease.

Secondary intimal fibroplasia may be difficult to distin-
guish from primary intimal disease, although medial and ad-
ventitial tissues are more likely to be abnormal in the former
than in the latter. 2 In this regard, certain secondary lesions
accompany developmental ostial lesions or advanced medial
dysplasia, perhaps as a sequela of altered blood flow through
these vessels. This may be the basis for many intimal lesions
occurring in association with elongation, kinking, or coiling of

pa rt I Vascular pathology and physiology

Figure 6.10 Myofibroblast in region of
extensive fibroplasia, exhibiting exopinocytotic
secretion of proteinaceous matter (arrow) (TEM,
x25 000). (From Stanley JC. Pathologic basis of
macrovascular renal artery disease. In: Stanley
JC, Ernst CB, Fry WJ, eds. Renovascular
Hypertension. Philadelphia: WB Saunders,
1984:46-74.)

Figure 6.1 1 Medial fibrodysplasia manifest as irregular narrowings to
ptotic kidneys, affecting midportion of main renal arteries that appear
stretched during upright aortography. (From Stanley JC, Wakefield TW.
Arterial fibrodysplasia. In: Rutherford RB, ed. Vascular Surgery. 3rd edn.
Philadelphia: WB Saunders, 1 989:245-265.)

the carotid artery. 2 Vascular trauma or intraluminal thrombo-
sis may contribute to other focal secondary lesions. Long tubu-
lar stenoses may occur as a consequence of recanalization of a
previously thrombosed artery. Vessel wall inflammation may
also play a role in these cases, and, in some instances of intimal
fibroplasia, have been suggested to represent a resolved arteri-

tis, as might occur with rubella (Fig. 6.14). 29 An infectious-
immunologic etiology in certain lesions is supported by
immunoglobulin deposition within intimal tissues of the
affected vessels. 30

Intimal fibroplasia of the external iliac, femoral, popliteal,
and tibial vessels of the lower extremity is usually considered
a secondary phenomenon, rather than representing a primary
process. 2 Intimal disease affecting these vessels probably fol-
lows prior trauma, thromboembolism with recanalization of
intraluminal clot, or the sequelae of an earlier arteritis.

The most common form of upper extremity arterial dyspla-
sia is intimal fibroplasia, usually manifest by smooth focal or
long tubular stenoses. 2 The most likely cause of these lesions is
an arteritis, frequently affecting all mural elements. Difficul-
ties may exist in differentiating some of these lesions from
Takayasu's arteritis. However, this diagnosis is less likely in
the absence of aortic arch, brachiocephalic, or more distal ab-
dominal aortic disease. Other intimal dysplastic lesions of the
upper extremity vessels may be a consequence of repetitive
trauma, such as accompanying thoracic outlet entrapment, or
a consequence of blunt trauma, becoming apparent many
years after the actual vascular injury.

Intimal fibroplasia may also affect the origins of the celiac,
superior mesenteric, or inferior mesenteric arteries. This usu-
ally occurs as a secondary event in developmentally narrowed
vessels. Intimal fibrodysplasia in these circumstances tends
to affect women more often than men. A prior arteritis
or resolved thrombosis may account for some lesions,
especially those of the distal celiac or superior mesenteric
artery branches.

chapter 6 Pathogenesis of arterial fibrodysplasia

Figure 6.13 Primary intimal fibroplasia. Subendothelial mesenchymal cells
within a loose fibrous connective tissue matrix are noted above an intact
internal elastic lamina, normal media, and normal adventitial tissues
(hematoxylin and eosin, x1 00). (From Stanley JC, Graham LM. Renovascular
hypertension. In: Miller DC, RoonAJ,eds. Diagnosis and Management of
Peripheral Vascular Disease. Menlo Park, CA: Addison-Wesley,
1981:231-235.)

Figure 6.1 2 (left) Primary intimal fibroplasia. (A) Focal stenosis of main
renal artery midportion in a young adult. ( B) Intraparenchymal web-like
stenosisof a segmental artery in a child. (A from StanleyJC, Fry WJ.
Renovascular hypertension secondary to arterial fibrodysplasia in adults.
Criteria for operation and results of surgical therapy. Arch Surg 1 975;
1 10:922. Bfrom StanleyJC, Fry WJ. Pediatric renal artery occlusive disease
and renovascular hypertension. Etiology, diagnosis and operative treatment.
Arch Surg 1 984; 1 1 6:669-676.)

pa rt I Vascular pathology and physiology

Figure 6.14 Secondary intimal fibroplasia.
Long tubular stenoses in the main distal renal
arteries of an infant who had recovered from a
severe systemic arteritis of unknown etiology.
(From Whitehouse WM Jr, Cho KJ, Coran AS,
StanleyJC. Pediatric arterial disease. In: Neiman
HL, Yao JST, eds. Angiography of Vascular
Disease. New York: Churchill Livingstone,
1985:289-306.)

Developmental arterial dysplasia

Developmental arterial dysplasia represents a unique form of
vascular disease. 2 There is no apparent sex predilection for this
entity and its exact frequency is unknown. The most thor-
oughly studied vessel exhibiting this disease is the renal
artery. Nearly 40% of children with renovascular hyper-
tension have developmental renal artery lesions, 31 and among
adults with intimal fibrodysplastic renal artery disease
approximately 20% appear to have underlying growth or
developmental defects. In addition, nearly 80% of patients
with developmental abdominal aortic narrowings have co-
existing dysplastic splanchnic and renal arterial stenoses. 32

Developmental stenoses are invariably hypoplastic in char-
acter, with an external hour-glass appearance. Most develop-
mental lesions occur at the aortic origin of the vessel (Fig. 6.15).
These developmental lesions usually exhibit abnormalities in
all three principal vessel wall layers. 31 ' 33 ' 34 Intimal fibroplasia,
fragmentation and duplication of the internal elastic lamina,
excesses in adventitial elastic tissue, and irregular deficiencies
in medial tissue are characteristic of these diminutive vessels
(Fig. 6.16).

Developmental renal artery narrowings in certain patients
appear related to in utero developmental events. During the
same period of embryonic development, the paired dorsal
aortas fuse and all but one of the multiple lateral metanephric
branches usually regress, leaving a solitary renal artery.

Abnormal transition of mesenchyme to medial smooth mus-
cle tissue at this embryonic time, or its later condensation and
growth, may result in a dysplastic narrowed aorta, as well as
stenotic splanchnic and renal arteries.

Several hypotheses exist regarding these lesions, including
a proposal that constrictive lesions follow a lack of or unequal
fusion of the two dorsal aorta, 35 with subsequent obliteration
of one of these channels and constriction of the associated
splanchnic or renal arteries. This may reflect an acquired insult
in utero that arrests growth of the aorta. Such may be caused by
a virus, a tenet that is supported by the fact that certain viruses
appearing to be associated with these lesions, including
rubella, are cytocidal and inhibitory to cell replication. 36,37

A second observation relating to the process by which
renal arteries normally originate within mesenchymal tissue
about the two dorsal aortas supports the developmental na-
ture of these lesions. The renal vessels are initially represented
by a caudally located group of mesonephric arteries that are
replaced during fetal development by a more cephalic group
of metanephric arteries. A solitary artery to the primitive
kidney evolves from each of these lateral vessel groups in
65-75% of normal individuals. This evolution of a single
dominant vessel occurs because of its obligate hemodynamic
advantage over adjacent channels. Flow changes due to an
evolving aortic coarctation may give other developing renal
arteries hemodynamic advantages that cause their persis-
tence. In support of such a hypothesis of developmental
renal artery occlusive disease is the fact that central abdominal

chapter 6 Pathogenesis of arterial fibrodysplasia

Figure 6.1 5 Developmental renal artery stenoses. (A) Proximal lesion in a
patient with neurofibromatosis. (B) Proximal right upper renal artery stenosis
in a patient with multiple renal arteries and midabdominal coarctation. (C)
Multiple vessel stenoses in a patient with aortic hypoplasia. (A from Stanley
JC, Fry WJ. Pediatric renal artery occlusive disease and renovascular
hypertension. Etiology, diagnosis and operative treatment. Arch Surg
1984:1 16:669. Band Cfrom Graham LM,ZelenockGB, Erlandson EE, Coran
AG, LindenauerSM, Stanley JC. Abdominal aortic coarctation and segmental
hypoplasia. Surgery 1 979; 86:51 9.)

pa rt I Vascular pathology and physiology

Figure 6.16 Developmental, hypoplastic renal artery. Marked
fragmentation and duplication of the internal elastic lamina and attenuation
of medial tissues characterize this vessel. Intimal fibroplasia encroaches on
the vessel lumen, which is less than 1 mm in diameter. Adventitial elastic
tissues appear excessive (Movat stain, x1 00). (From Stanley JC, Graham LM,
Whitehouse WM jretal. Developmental occlusive disease of the abdominal
aorta, splanchnicand renal arteries. Am J Surg 1 981 ;1 42:1 90.)

aortic coarctations with their attending flow abnormalities
are associated with multiple stenotic renal arteries in 60-80%

of cases. 32/34/38

References

1. Stanley JC. Pathologic basis of macrovascular renal artery disease.
In: Stanley JC, Ernst CB, Fry WJ, eds. Renovascular Hypertension.
Philadelphia: WB Saunders, 1984:46.

2. Stanley JC, Wakefield TW. Arterial fibrodysplasia. In: Rutherford
RB, ed. Vascular Surgery, 5th edn. Philadelphia: WB Saunders,
2000:387.

3. Stanley JC, Gewertz BC, Bove EL, Sottiurai V, Fry WJ. Arterial
fibrodysplasia: histopathologic character and current etiologic
concepts. Arch Surg 1975; 110:561.

4. Goncharenko V, Gerlock AJ, Shaff MI, Hollifield SW. Progression
of renal artery fibromuscular dysplasia in 42 patients as seen on
angiography. Radiology 1981; 139:45.

5. Meaney TF, Dustan HF, McCormack LJ. Natural history of renal
arterial disease. Radiology 1968; 91:881.

6. Sheps SG, Kincaid OW, Hunt JC. Serial renal function and angio-
graphic observations in idiopathic fibrous and fibromuscular
stenoses of the renal arteries. Am } Cardiol 1972; 30:55.

7. Schreiber MJ, Pohl MA, Novick AC. The natural history of athero-
sclerotic and fibrous renal artery disease. Urol Clin North Am 1984;
11:383.

8. Cragg AH, Smith TP, Thompson BH et ah Incidental fibromuscular
dysplasia in potential renal donors: long-term clinical follow-up.
Radiology 1989; 172:145.

9. Stanley JC, Fry WJ, Seeger JF, Hoffman GL, Gabrielson TO.
Extracranial internal carotid and vertebral artery fibrodysplasia.
Arch Surg 1974; 109:215.

10. Stewart MT, Moritz MW, Smith RB III, Fulenwider JT, Perdue GD.
The natural history of carotid fibromuscular dysplasia. / Vase Surg
1986; 3:305.

11. Mettinger KL, Ericson K. Fibromuscular dysplasia and the brain:
observations on angiographic, clinical and genetic characteristics.
Stroke 1982; 13:46.

12. Walter JF, Stanley JC, Mehigan JJ, Rueter SR, Guthaner D. Iliac
artery fibroplasia. Am} Roentgenol 1978; 131:125.

13. van den Dungen JJAM, Boontje AH, Oosterhuis JW. Femoro-
popliteal arterial fibrodysplasia. Br J Surg 1990; 77:396.

14. Harrison EG, McCormack LJ. Pathologic classification of renal
artery disease in renovascular hypertension. Mayo Clin Proc 1971;
46:161.

15. McCormick LJ, Noto TJ Jr, Meaney TF, Poutasse EF, Dustan HP.
Subadventitial fibroplasia of the renal artery: a disease of young
women. Am Heart J 1967; 73:602.

16. Sottiurai VS, Fry WJ, Stanley JC. Ultrastructure of medial smooth
muscle and myofibroblasts in human arterial dysplasia. Arch Surg
1978; 113:1280.

17. Gladstien K, Rushton AR, Kidd KK. Penetrance estimates and
recurrence risks for fibromuscular dysplasia. Clin Genet 1980;
17:115.

18. Major P, Genest J, Cartier P, Kuchel O. Hereditary fibromuscular
dysplasia with renovascular hypertension. Ann Intern Med 1977;
86:583.

19. Rushton AR. The genetics of fibromuscular dysplasia. Arch Intern
Med 1980; 140:233.

20. Sang CN, Whelton PK, Hamper UM et al. Etiologic factors in
renovascular fibromuscular dysplasia. Hypertension 1989; 14:472.

21 . de Deeuw D, Donker AJM, Burema J, van der Hem GK, Mandema
E. Nephroptosis and hypertension. Lancet 1977; 1:213.

22. Kaufman JJ, Maxwell MH. Upright aortography in the study of
nephroptosis, stenotic lesions of the renal artery, and hyper-
tension. Surgery 1963; 53:736.

23. Tsukamoto Y, Komuro Y, Akutsu F et al. Orthostatic hypertension
due to coexistence of renal fibromuscular dysplasia and nephrop-
tosis. Jpn Circ } 1988; 52:1408.

24. Leung DYM, Glagov S, Matthews MB. Cyclic stretching stimu-
lates synthesis of matrix components by arterial smooth muscle
cells in vitro. Science 1976; 191:475.

25. Fievez ML. Fibromuscular dysplasia of arteries: a spastic
phenomenon? Med Hypotheses 1984; 13:341.

26. Paulson GW. Fibromuscular dysplasia, antiovulant drugs, and
ergot preparations. Stroke 1978; 9:172.

27. Hata J-I, Hosoda Y Perimedial fibroplasia of the renal artery: a
light and electron microscopy study. Arch Pathol Lab Med 1979;
103:220.

28. Sottiurai VS, Fry WJ, Stanley JC. Ultrastructural characteristics
of experimental arterial medial fibrodysplasia induced by vasa
vasorum occlusion. / Surg Res 1978; 24:169.

chapter 6 Pathogenesis of arterial fibrodysplasia

29. Stewart DR, Price RA, Nebesar R, Schuster SR. Progressing
peripheral fibromuscular hyperplasia in an infant: a possible
manifestation of the rubella syndrome. Surgery 1973; 73:374.

30. Dornfeld L, Kaufman JJ. Immunologic considerations in renovas-
cular hypertension. Urol Clin North Am 1975; 2:285.

31. Stanley JC, Fry WJ. Pediatric renal artery occlusive disease and
renovascular hypertension: etiology diagnosis and operative
treatment. Arch Surg 1981; 116:669.

32. Graham LM, Zelenock GB, Erlandson EE, Coran AG, Lindenauer
SM, Stanley JC. Abdominal aortic coarctation and segmental
hypoplasia. Surgery 1979; 86:519.

33. Devaney K, Kapur SP, Patterson K, Chandra RS. Pediatric renal
artery dysplasia: a morphologic study. Pediatr Pathol 1991; 11:609.

34. Stanley JC, Graham LM, Whitehouse WM Jr et al. Developmental
occlusive disease of the abdominal aorta, splanchnic and renal
arteries. Am J Surg 1981; 142:190.

35. Maycock Wd'A. Congenital stenosis of the abdominal aorta. Am
Heart J 1937; 13:633.

36. Plotkin SA, Boue A, Boue JG. The in vitro growth of rubella virus in
human embryo cells. Am] Epidemiol 1965; 81:71.

37. Siassi B, Glyman G, Emmanouilides GC. Hypoplasia of the ab-
dominal aorta associated with the rubella syndrome. Am } Dis
Child 1970; 120:476.

38. Stanley JC, Zelenock GB, Messina LM, Wakefield TW. Pediatric
renovascular hypertension: a thirty-year experience of operative
treatment. / Vase Surg 1995;21:212.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Physiology of vasospastic disorders

Scott E. Musicant
Jean-Baptiste Roullet
James M. Edwards
Gregory L. Moneta

There are many disease states where vasospasm occurs as a re-
sult of disturbances in the normal control of vascular wall tone
caused by either intrinsic or extrinsic factors. In addition to
vasospasm occurring in association with other diseases, there
are at least three primary vasospastic diseases, Raynaud's syn-
drome, migraine, and variant angina. 1 ' 2 Only in these diseases
does vasospasm appear as a primary event.

In this chapter we will discuss the pathophysiology of va-
sospasm in the vasospastic diseases as determined from clini-
cal studies in humans, and then describe the pathophysiology
of vasospasm using data obtained by laboratory studies utiliz-
ing both human and animal tissue. Most of the information
presented will relate to Raynaud's syndrome as this disease
is the most widely studied of the three primary vasospastic
diseases.

Clinical studies

Raynaud's syndrome

In the past three decades, the vascular surgery unit at the
Oregon Health & Science University has conducted detailed
prospective evaluations of over 1300 patients with Raynaud's
syndrome. This represents, by far, the largest prospectively
studied group of Raynaud's patients under continuous
observation in the world. While our group has made numer-
ous, widely published demographic observations, the under-
lying pathophysiology of Raynaud's syndrome has proved
elusive. Raynaud's syndrome has traditionally been divided
into two groups, Raynaud's phenomenon and Raynaud's
disease, based on the presumed presence or absence of an
associated disease. We, however, have used a different method
of categorization. We divide patients with Raynaud's syn-
drome into two categories: vasospastic and obstructive. Pa-
tients with vasospastic Raynaud's syndrome have normal
digital artery pressure at rest, but in response to cold or
emotional stress have an abnormally forceful vasospastic
response causing digital artery closure and digital ischemia.

Patients with obstructive Raynaud's syndrome have fixed
obstruction of the subclavian, brachial, radial, ulnar, or palmar
and digital arteries and diminished digital artery pressures
at room temperature. These patients, in response to cold, have
a normal vasoconstrictive response which, because of the
reduced intraluminal pressure, results in digital artery
closure.

Raynaud originally thought the defect underlying va-
sospasm resided in the sympathetic nervous system. 3 Lewis 4
demonstrated that blockade of digital nerve conduction did
not prevent vasospasm and suggested the existence of "a local
vascular fault." Research into the pathophysiologic mecha-
nisms of Raynaud's syndrome has continued over the past
century, but little definitive information has emerged. Several
proposed pathophysiologic mechanisms of vasospasm
have been suggested in Raynaud's syndrome: alterations
in the sympathetic nervous system; alteration in a- and /or
(3-adrenergic or serotoninergic receptor number or density;
alterations in circulating catecholamines; and most recently,
alteration in levels of the vasoactive peptides endothelin and
calcitonin gene-related peptide (CGRP). We will not further
discuss alterations in sympathetic nervous system function
since, although they may be present, the effector mechanism of
these changes is probably one of the other proposed abnormal-
ities (receptor or vasoactive peptide).

Similarly, we will not dwell on alterations in circulating cat-
echolamine levels. The available data are conflicting, a result
that is not surprising. 5-7 The measurement of catecholamines
is notoriously difficult in every aspect from drawing the blood
without alarming or otherwise exciting the patient, account-
ing for possible concentration or metabolism in low flow
states, as well as the actual laboratory measurement.

One of the earliest pathophysiologic mechanisms pro-
posed for Raynaud's syndrome suggested alterations in oc-
adrenergic receptor function and number. The basis for the
proposal of this mechanism was probably related to the
success of oc-blockers such as reserpine in the treatment of
Raynaud's syndrome in the 1950s. Early work with reserpine
demonstrated that a-adrenergic blockade resulted in in-

chapter 7 Physiology of vasospastic disorders

creased finger blood flow. 8 Intraarterial injection of either re-
serpine or phentolamine has been shown to abolish cold-in-
duced vasospasm. 9-11 Alpha-blockade with oral agents has
also been shown to be effective in the treatment of
Raynaud's syndrome. 9/12/13

Because of evidence implicating oc 2 -adrenoceptors in vaso-
constriction and as sympathetic nervous system mediators,
their role in Raynaud's syndrome has been investigated in
many laboratories. Administration of adrenoceptor antago-
nists to the digital skin in patients with Raynaud's syndrome
demonstrated abolishment of cold-induced vasospasm by oc 2 -
but not o^-antagonists. 14 A relatively pure population of oc 2 -
adrenoceptors exists on platelets. Because of the difficulty in
obtaining digital arteries from patients with Raynaud's syn-
drome and the observations that levels of receptors on circu-
lating cells mirror tissue levels, most investigators have
measured platelet levels of oc 2 -adrenoceptors. 15/16 We have
shown that platelets from patients with Raynaud's syndrome
have elevated levels of oc 2 -adrenoceptors, a finding confirmed
by others. 17-19

Further support for the oc-adrenergic hypothesis comes
from the work of Freedman and others, 20 who demonstrated
increased digital blood flow in response to intraarterial infu-
sions of both a t - and oc 2 -agonists. While preliminary evidence
indicates an increased number of oc 2 -adrenoceptor sites in pa-
tients with Raynaud's syndrome, other interpretations of
these data are possible. Alteration of receptor sensitivity or a
change in the number of receptors exposed at any one time
(rather than an absolute increase in number) is possible. 21-23
These alterations will be discussed in greater detail later in this
chapter.

The role of (3-adrenoceptors has also been studied in Ray-
naud's syndrome. While abnormalities in a-adrenoceptors
consisting of increased number or sensitivity have been postu-
lated as causing vasospasm, abnormalities in (3-adrenoceptors
have been implicated in vasospasm in a negative role. Since (3-
adrenoceptors are thought to cause vasodilation, abnormali-
ties may be responsible for decreased vasodilation which
results in vasoconstriction by unopposed oc-adrenergic action.
The earliest work in this area was triggered by the observation
that (3-blocking drugs occasionally resulted in digital va-
sospasm, and other data that some observers interpreted as in-
dicating active (3-adrenergic-mediated vasodilation which
countered normal vasoconstrictive tone. 24,25 No (3-adrenergic
vasodilating mechanism has been observed in the forearm, 26
although its presence was detected in the finger. 27 Further
work on (3-adrenergic digital vasodilation has demonstrated
that it appears independent of the nervous system in that digi-
tal nerve blockade does not affect vasodilation. 28 This led to
the suggestion that a circulating vasoactive substance may be
responsible for the vasoconstriction occurring in Raynaud's
attacks. This experimental work has been confirmed clinically
by several investigators who have demonstrated that treat-
ment with either (3-blockers with intrinsic sympathomimetic

activity or combining a- and (3-blocker therapy avoids the
induction of digital vasospasm. 29 ' 30 However, more recent
research investigating the influence of different types of (3-
blocking drugs on the peripheral circulation in patients with
Raynaud's syndrome has indicated neither beneficial nor
detrimental effects. 31

Other roles for the (3-adrenoceptor in Raynaud's syndrome
have been postulated. A group from the University of Cagliari
in Italy has argued that adrenoceptor alterations in Raynaud's
syndrome are pre-, rather than postsynaptic. At the presynap-
tic level, oc 2 -adrenoceptors inhibit and (3-receptors facilitate
noradrenaline release. According to their hypothesis, (3-
stimulation causes the release of noradrenaline which in turn
causes digital vasoconstriction. 32 These authors have conduct-
ed several clinical studies in which patients with Raynaud's
syndrome were treated with low dose (3-blockers in combina-
tion with a calcium channel blocker with marked relief of
symptoms. 33 ' 34 This work remains to be confirmed by others.

Other substances, with actions not associated with the
a- and (3-adrenergic receptors, have been implicated in
Raynaud's syndrome. 35 These include neurotransmitters such
as dopamine, serotonin, histamine, and acetylcholine, as
well as vasoactive peptides such as CGRP, endothelin, and
vasoactive intestinal peptide. 36-38

Serotonin (5-hydroxytryptamine) is perhaps the most
studied neurotransmitter in Raynaud's syndrome. 39 Because
of the availability of ketanserin, a serotonin 2 (S 2 )-blocking
agent, a number of experimental and clinical studies have
been performed. 40-42 Interestingly, ketanserin appears to be
effective in the relief of Raynaud's symptoms only in those
patients with obstructive Raynaud's syndrome due to
scleroderma, and not vasospastic Raynaud's syndrome. The
explanation for this selective benefit is unknown.

The most recent work on the pathophysiology of Raynaud's
syndrome has focused on the possible role of two vasoactive
peptides, endothelin, a potent vasoconstrictor, and CGRP, a
vasodilator. 43-47 Shawket and coworkers 48 demonstrated a su-
persensitivity of skin blood flow to CGRP infusion in patients
with Raynaud's syndrome. They attributed this reaction to a
baseline deficiency of CGRP in patients with Raynaud's syn-
drome. While the results of this work have been disputed by
others, 49 ' 50 further work has demonstrated a deficiency of
CGRP in skin neuronal terminals in patients with Raynaud's
syndrome. 51 ' 52 Correlation of this deficiency with circulating
CGRP levels has not been performed because of a myriad of
technical difficulties including concentration with reduced
flow during Raynaud's attacks, the difficulties of CGRP mea-
surement, and the unknown relationship between circulating
CGRP and arterial vasodilation. However, certain patients
with Raynaud's syndrome who receive infusions of CGRP
have shown improvement in thermographically measured
parameters when compared with infusion with prostacyclin.
This improvement persists for days after the infusion is termi-
nated, despite the short half-life of CGRP.

pa rt I Vascular pathology and physiology

Endothelin is a potent vasoconstrictor produced by
endothelial cells. In the first study of endothelin levels in
Raynaud's syndrome, baseline endothelin levels in patients
were three times higher than in controls, and cold-stimulated
values increased by approximately a factor of two in both
groups. 53 Leppert and colleagues 54 showed significant in-
creases in endothelin-1 levels in Raynaud's patients after
whole body cooling compared with controls. Further work by
other authors has yielded conflicting results, and indeed,
some work has suggested that endothelin does not play a role
in cold-induced vasoconstriction. 55-60 As noted above, the de-
termination of circulating vasoactive peptide levels is difficult.
The normal levels of these substances are miniscule, making
measurement difficult. The possibility of hemoconcentration
during a Raynaud's attack has not been addressed, nor is it
clear that venous forearm blood samples are representative of
digital arterial levels. Until these problems are addressed in
greater detail, the determination of the role of vasoactive pep-
tides will remain unclear. While intraarterial infusion of these
vasoactive substances would appear an ideal way to study
their physiologic effects, potentially serious adverse reactions
to endothelin were reported, 61 and experience with the other
substances is limited. Side-effects of agent administration, risk
of brachial artery injury in primarily young patients and con-
trols, and difficulties with the reproducibility of venous occlu-
sion plethysmography, all call into question the proper role of
brachial artery infusions with these substances. We currently
have no plans to use brachial artery infusion of vasoactive sub-
stances because of these safety and ethical issues. A possible
safer alternative may be Bier block delivery of vasoactive
substances. 62

Recently there has been interest in the effects of nitric oxide
(NO) on the digital circulation of patients with Raynaud's syn-
drome. Ringqvist and colleagues 63 were able to show a sea-
sonal variation in the plasma levels of NO in women with
Raynaud's syndrome and in healthy controls where higher
levels were present in the winter compared with summer.
However, they were unable to show a change in plasma levels
of NO with cold exposure and they were unable to show an in-
crease in NO levels in the patients with Raynaud's syndrome.
The results of a study at Stanford University suggested that
venodilation in patients with Raynaud's was impaired due to
a diminished release of NO; however, their data did not reach
significance. 64 Further studies are required to characterize the
vasoreactive effects of NO to determine their role in the patho-
physiology of Raynaud's syndrome.

Migraine

The mechanisms of migraine are poorly understood due, in
part, to the relative inaccessibility of the brain and also to the
lack of an animal model. Migraine appears related to alter-
ations in brain blood flow and changes in intracerebral and
perhaps extracerebral vessels. 65,66 Changes in cerebral metab-

olism or changes similar to the spreading depression of Leao,
which is a progressive depression of electrical activity of the
cerebral cortex seen in animal experiments, 67 have been postu-
lated as possible mechanisms for the changes seen in
migraine. 68 The three basic theories of the pathogenesis of
migraine are vascular, biochemical, and neuronal. 69 Neither
the vascular nor the biochemical theories adequately explain
all aspects of migraine (aura, pain, neurologic deficits, changes
in regional cerebral blood flow), so attention has more recently
focused on neural causes of migraine. These include
neurotransmitter abnormalities, the spreading depression
of Leao, and neuronal abnormalities. Alterations in central
catecholamine levels have also been postulated in the patho-
physiology of migraine, and decreased levels of platelet oc 2 -
adrenoceptor binding has been reported in patients with
migraine. 70 While the discussion of the pathophysiology of
migraine is beyond the scope of this chapter, we want to pre-
sent the results of one study which may link the pathophysiol-
ogy of migraine with that of Raynaud's syndrome, since the
two syndromes have themselves been linked. The linkage of
migraine and Raynaud's syndrome to date has been epidemi-
ologic, with multiple reports demonstrating that the incidence
of either migraine patients who also have Raynaud's syn-
drome or patients with Raynaud's syndrome who also have
migraine is greater than the incidence seen in the general
population. 1/2/71 Many works propose a common pathophy-
siologic mechanism for the two diseases.

Goadsby and coworkers 72 have examined external jugular
venous blood in patients during migraine attacks and demon-
strated significant elevation of levels of CGRP when compared
to antecubital venous blood. Levels of other vasoactive pep-
tides (neuropeptide Y, vasoactive intestinal polypeptide, and
substance P) were unchanged. They argue that one possible
pathophysiologic mechanism for migraine may be an
abnormality of trigeminal-cerebrovascular control which is
mediated by CGRP. 73

Variant angina

As early as the late 1800s it was widely accepted that
vasospasm may be responsible for some cases of angina. 74
Prinzmetal in 1959 described vasospasm of the coronary
arteries as a cause of angina. 75 In the 1970s, many authors
demonstrated arteriographically coronary artery vasospasm
in patients with angina and otherwise normal coronary
vessels. 76 ' 77 Since then, variant angina has variously been both
attributed to and been felt unrelated to oc-adrenergic and
sympathetically mediated mechanisms. 78-80

More recent work, as in Raynaud's syndrome and migraine,
has focused on the vasoactive peptides. Neuropeptide Y infu-
sion in coronary arteries results in vasoconstriction which can
be reversed by concomitant nitrate infusion. 81 Acetylcholine
causes coronary vasodilation at low doses but, interestingly,
causes coronary vasoconstriction while also increasing coro-

chapter 7 Physiology of vasospastic disorders

nary blood flow at higher doses. This is probably secondary to
vasodilation of the coronary arteriolar bed. 82 CGRP is a potent
nonendothelium-dependent vasodilator of coronary arteries
which has dose-dependent effects on myocardial perfusion in
dogs. 83,84 The coronary endothelium appears functionally
normal at sites of arteriographically proven vasospasm as
determined by its response to infusions of substance P and
acetylcholine, which led the authors of one study to conclude
that the abnormality in variant angina was in the vascular
smooth muscle rather than the endothelium. 85

In contrast, Teragawa and coworkers showed that peri-
pheral endothelial function is impaired in patients with
vasospastic angina. 86 They studied brachial artery diameter
responses to hyperemic flow using ultrasound and found
that flow-mediated diameter was lower in patients with
vasospastic angina compared with controls. Recently,
Tomimura and colleagues 87 studied the effects of inflammato-
ry cytokines on the induction of vasospastic angina. They eval-
uated the plasma levels of macrophage colony-stimulating
factor (M-CSF) in patients with vasospastic angina and found
that levels of M-CSF were significantly higher in patients with
active vasospastic angina compared with patients with inac-
tive vasospastic angina. They also showed that patients with
multivessel vasospasm had higher levels of M-CSF than those
with single-vessel vasospasm, which led them to conclude
that coronary vasoreactivity is affected by plasma M-CSF
concentration. 87

Interestingly, there is a report of cold-induced myocardial
ischemia. 88 In a group of patients with scleroderma, all of
whom had Raynaud's syndrome, cold provocation with body
cooling led to the development of myocardial ischemia as de-
tected by thallium imaging in 12 of 21 patients. The authors did
not postulate a mechanism, although they noted that platelet-
mediated vasospasm could be one explanation.

In summary, the pathophysiologic mechanisms of
Raynaud's syndrome, migraine, and variant angina remain
unknown, as does the relationship between the three dis-
eases. While abnormalities in the adrenergic receptors, and
more recently the vasoactive peptides, have been demon-
strated in each of these disease states, we await a unifying
pathophysiologic mechanism, if it exists.

Laboratory studies

As noted above, clinical studies have yielded an abundance of
conflicting results which can only support one conclusion: no
single treatment consistently provides relief to all patients
with symptomatic vasospasm. Sympathetic blockade works
in some, but not all, patients. Alpha-adrenergic antagonists
provide relief to some, but not to others. 89 The same dicho-
tomy has been observed with ketanserin, an antagonist of
5-HT 2 receptors. 41 ' 90 There is evidence that the peptide endo-
thelin is elevated in Raynaud's syndrome but there is present-

ly no antagonist to block this agent. 55 Currently, the pharmaco-
logic therapy of choice for Raynaud's syndrome for those few
patients who require it is the calcium channel blocker nifedip-
ine. 91 This treatment, of course, is empiric and is directed to-
ward the lowest common denominator of vascular
contraction and reflects our continued inability to develop a
specific treatment for this disease. Our inability to define pre-
cisely a single vascular wall defect precipitating Raynaud's at-
tacks suggests the cause of cold-induced vasospasm may be
multifactorial. The remainder of this chapter will consider the
laboratory-derived evidence suggesting a causal role for nor-
epinephrine, serotonin, and endothelin in precipitating the va-
sospasm observed in primary Raynaud's syndrome.

Norepinephrine and adrenergic receptors

Norepinephrine is an agonist of both a- and (3-adrenoceptors.
Indeed, activation of these receptors by the sympathetic ner-
vous system comprises the major mechanism responsible for
the regulation of blood flow and arterial pressure. Both a and (3
receptors have been subclassified pharmacologically through
the use of increasingly selective agonists and antagonists.
However, since at this writing no available data preferentially
implicate one subtype or another below the level of a t vs. oc 2 in
Raynaud's syndrome, further subdivisions will not be consid-
ered. For a detailed summary of pertinent data, refer to a
review by Ruffolo and colleagues. 92 There are, however, abun-
dant data suggesting important distinctions at the level of
a 1 - vs. a 2 -adrenoceptors which may well relate to Raynaud's
syndrome.

Alpha-adrenoceptors have been divided into two principal
subtypes, a t and oc 2 , based on their relative sensitivity to selec-
tive antagonists. Initially it was believed that a-adrenoceptor
types could also be distinguished anatomically, with a 1 -
adrenoceptors being found postjunctionally and (^-adreno-
ceptors occurring only on prejunctional or presynaptic
membranes. However, it is now well established that norepi-
nephrine causes contractions of vascular smooth muscle by
stimulating both postjunctional a 1 - and oc 2 -adrenoceptors. 93
Several studies demonstrated the existence, and the inner-
vation of postjunctional oc 2 -adrenoceptors in veins 94-96 and
arteries. 97-100 The density and distribution of the two oc-
adrenoceptor types varies significantly between different
beds. 96 ' 101 Stimulation of either a t - or oc 2 -adrenoceptors on vas-
cular smooth muscle results in contraction of those smooth
muscle cells. While elevation of intracellular calcium and
muscle contraction are the common endpoints resulting
from stimulation of these two receptor types, the events link-
ing excitation to contraction differ with the receptor type. For
example, oc 2 -mediated contractions are predominantly depen-
dent on influx of extracellular calcium, whereas release of Ca 2+
ions from an intracellular pool has a key role in activating the
response to a 1 -stimulation. These differences are mentioned
because of the very selective effects that cooling has upon the

pa rt I Vascular pathology and physiology

responsiveness of these two receptor types, and thus on their
potential involvement in cold-induced spasm.

Much of the work determining the receptor-specific influ-
ence of cooling was first conducted using organ chambers to
study isolated segments of canine veins. These studies
involved the addition of selective agonists and antagonists
to individual rings and quantitation of the differences in
contractile force generated by the rings. In the canine saphe-
nous vein, exogenous or nerve-released norepinephrine
causes contractions by stimulating both a 1 - and oc 2 -
adrenoceptors. 95 Contractions evoked by norepinephrine
may be blocked by antagonists selective for either a 1 - or
a 2 -adrenoceptors (prazosin or rauwolscine respectively).
Additionally, exposure of the rings to agonists selective for
either receptor type (phenylephrine for a 1 -adrenoreceptors,
B-HT 920 for oc 2 -adrenoreceptors) will evoke contractions.

In this same vein, norepinephrine-induced contractions are
potentiated by moderate cooling. 102 The augmentations pro-
duced by cooling of such adrenergically contracted rings does
not occur when the contractile stimulus is a depolarizing solu-
tion of potassium chloride. Further, by loading the perivascu-
lar nerves with tritiated norepinephrine and measuring
its presence in a superfusate of the vessel being studied,
Vanhoutte and Verbeuren were able to show that the amount
of norepinephrine released in response to electrical stimula-
tion is actually decreased by cooling. 103 Taken together, the
data seem to indicate that the cold-induced enhancement of
norepinephrine's vasoconstrictor effect is due neither to a di-
rect effect on the contractile apparatus of the vascular smooth
muscle, nor to an increase in transmitter release.

It is important, however, to bear in mind the complexity of
the molecular machinery which is activated by occupation of
the a x - or oc 2 -adrenergic receptors in smooth muscle cells and
triggers vasoconstriction. It is well established, for example,
that both types of receptors activate GTP-binding proteins (G-
proteins), to mediate their effect on plasma membrane ion
channels and inositol-l,4,5-triphosphate-dependent calcium
signaling. It is also well established that GTP increases the sen-
sitivity of the contractile apparatus to intracellular calcium, re-
sulting in greater contraction of the smooth muscle for given
intracellular calcium concentrations. 104 ' 105 Such sensitization
of the contractile apparatus to calcium, sensitization which is
not seen when contraction is evoked with potassium chloride,
could be enhanced by cold and further enhanced in patients
with Raynaud's syndrome. This theory has not been explored
yet, and remains speculative at the time of this writing. How-
ever, it illustrates the complexity of fully assessing the molecu-
lar mechanisms which underlie cold-induced vasospasm and
Raynaud's syndrome pathophysiology.

Flavahan and colleagues investigated the influence of acute
cooling on oc 2 - vs. ^-mediated responses. 106 Cooling en-
hanced contractions evoked by either norepinephrine or ago-
nists of oc 2 -adrenoceptors (UK 14,304 and B-HT 920); such
augmentations were blocked by rauwolscine, an antagonist of

oc 2 -adrenoceptors. In contrast, blockade of (^-adrenoceptors
did not inhibit the effect of cooling on norepinephrine-
induced contractions. Stimulation of (^-adrenoceptors
produced contractions which were either unaffected
(phenylephrine) or decreased (St 587) by cooling. These data
demonstrate clearly that the potentiation is in fact due to a
selective augmentation of oc 2 -, but not a 1 -, adrenoceptor-
mediated responses. The differential effect of cooling on
contractions evoked by the two a 1 -selective agonists, phenyle-
phrine and St 587, resulted from differences in the relative effi-
cacies of the two compounds. The fact that norepinephrine is a
full agonist and that spare receptors, or a receptor reserve,
exist for (^-adrenoceptors in this vessel is critical to the cold-
induced augmentation of contractions evoked by norepineph-
rine. Cold actually inhibits a 1 -mediated contractions, as was
seen with St 587. However, if a large enough reserve exists
(as it does for norepinephrine) the a x component will be
unaltered, providing a stable base upon which the enhanced
oc 2 activity may be observed.

To determine how the canine-derived data apply to human
vessels, our laboratory conducted similar studies in greater
saphenous veins obtained from patients at surgery 107 Exoge-
nous norepinephrine-induced contractions, which were
sensitive to both prazosin and rauwolscine, indicated the
presence of both a 1 - and oc 2 -adrenoceptors as was reported by
Docherty and Hyland. 108 Cooling enhanced contractions to
norepinephrine and B-HT 920, with the effect being most pro-
nounced at lower concentrations. Phenylephrine-mediated
contractions were not enhanced by cooling. The overall re-
sponse pattern was nearly identical to that observed in the ca-
nine saphenous vein after treatment with phenoxybenzamine
to lower the receptor reserve. We conclude that human saphe-
nous vein contractions mediated by oc 2 -adrenoceptors are po-
tentiated by cooling and that this vessel lacks the reserve for
^-agonists present in the canine vein. Eskinder et al. 109 have
also noted a lower a 1 -receptor reserve in the human vein.
More recently we investigated the influence of cooling on neu-
rogenic contractions in human saphenous veins. 110 Here, too,
we found cooling consistently enhances contractions evoked
by stimulating the perivascular nerves, and that the augmen-
tation is blocked by rauwolscine but not by prazosin. Taken as
a whole, contractile data collected on venous tissues support
the hypothesis of oc 2 -adrenoceptors contributing to cold-
induced vasospasm.

Of course, the spasms characteristic of Raynaud's syn-
drome are not venous but arterial in nature. The contributions
of oc 2 -receptors to adrenergic contractions and how they are in-
fluenced by cooling are less well documented in arteries than
in veins. However, where this has been studied, the data sup-
port the venous findings. Using intravital microscopy, Faber
addressed the issue of receptor subtype distribution in differ-
ent branches of resistance arteries of rat cremaster muscle. 111 In
feeder arterioles of 100 |im diameter, blockade of either a 1 - or
a 2 -adrenoceptors decreased contractions evoked by norepi-

chapter 7 Physiology of vasospastic disorders

nephrine. However, in smaller (25 jam diameter) arterioles,
only oc 2 -blockade was effective.

Human arteries possess both a 1 - and oc 2 -adrenoceptors.
Flavahan et al. 112 demonstrated that agonists and antagonists
selective for each subtype were effective in arteries obtained
from amputated arms and legs. Further, oc 2 -selective agents (B-
HT 920, rauwolscine) were more effective in the digital arteries
of the hands and feet than they were in more proximal vessels
(dorsalis pedis, superficial palmar arch). In contrast, a 1 -
selective compounds (phenylephrine, prazosin) were either
less or equally effective in the distal vs. the proximal vessels.
Nielsen et al. also observed oc 2 -adrenoceptors to be more
prominent in human subcutaneous resistance arteries than in
larger, more proximal arteries. 113

The influence of local cooling on arterial oc 2 -adrenoceptors
has been evaluated in the tail artery 114 and the cremaster mus-
cle resistance arteries 115 of the rat. In both cases, acute cooling
has preferentially enhanced oc 2 -mediated vasoconstriction.
Importantly, it remains undetermined if human subcutaneous
resistance arteries exhibit the same response to acute cooling.
Such studies will probably yield data even more applicable to
Raynaud's syndrome.

Ekenvall et al. u used laser-Doppler measurements of digital
cutaneous blood flow in conscious humans to assess the in-
volvement of oc 2 -adrenoceptors in the vascular response to
cooling. Blood flow measurements were made with a probe
which also regulated local temperature. The investigators
used electrical iontophoresis to apply either a 1 - or a 2 -selective
agonists (phenylephrine or B-HT 933 respectively) and antag-
onists (prazosin or rauwolscine respectively) to the areas
being studied. Phenylephrine and B-HT 933 decreased blood
flow to comparable degrees (29% and 24% respectively), indi-
cating that both a 1 - and oc 2 -receptors were present on the beds
being measured. In control fingers (no agonists or antagonists
present), lowering local temperature from 35°C to 20°C for 30 s
produced marked reductions (to 38% of initial) in cutaneous
flow. When digits were treated with prazosin (a 1 -blockade)
prior to cooling, lowering the temperature still produced a
marked decrease in flow (to 45% of initial). By contrast, the
presence of rauwolscine (oc 2 -blockade) nearly eliminated the
cold-induced vasoconstriction and blood flow remained at
96% of control.

The authors concluded that cold-induced vasoconstriction
is mediated by oc 2 -adrenoceptors in human finger skin. The di-
rect applicability of this observation to Raynaud's syndrome is
uncertain, however. The primary limitation is that the laser-
Doppler used to measure flow in this study reads only the flow
occurring through superficial vessels, and not through the
digital artery where the spasms of Raynaud's syndrome occur.
The authors mention that (i) iontophoresis is unable to deliver
drugs to deep vessels, and (ii) periods of cooling longer than
the 30 s used in this study produced decreases in blood flow
which persisted in the presence of rauwolscine. It may be that
the prolonged cooling lowered the temperature of the deeper

digital artery and an oc 2 -mediated vasoconstriction took place
since rauwolscine was unable to provide blockade of those
deeper receptors. Alternatively, the rauwolscine-resistant
constriction may be mediated by a different mechanism, such
as a myogenic or vasogenic response. Such intrinsic responses
to cooling have been reported in two other cutaneous vessels,
the facial vein 116 and central ear artery 117 of the rabbit.

Another obvious limitation of the Ekenvall study's rele-
vance to Raynaud's syndrome is that the data were obtained
in control patients only. In a recent study, Cooke and
coworkers 118 infused a x - or oc 2 -selective antagonists (prazosin
and yohimbine respectively) into the brachial artery of
patients with Raynaud's syndrome and control subjects, and
measured changes in finger blood flow (FBF) strain-gauge
venous occlusion plethysmography. Raynaud's patients
showed reduced basal FBF compared with control subjects at
22°C (room temperature), and greater reduction in FBF upon
local cooling. However, both groups had a similar sensitivity
to prazosin and yohimbine, whether tested at room tem-
perature or after cooling. The authors concluded that a
nonadrenergic mechanism contributes to local cold-induced
vasoconstriction.

A significant role of oc 2 -adrenergic receptors in cold-induced
vasospasm and Raynaud's syndrome is, however, probably
based on the clinical evidence that oc 2 - but not a 1 -adrenergic
antagonists block vasospastic attack in idiopathic Raynaud's
syndrome. 119 This issue was recently reevaluated by Chotani
and coworkers 120 in their attempt to assess the relative contri-
butions of oc 2A -, oc 2B -, oc 2C -adrenergic receptor subtypes to
thermoregulation in the mouse. Using isolated distal tail
arteries, these investigators determined that oc 2 - but not a 1 -
adrenoreceptor-mediated vasoconstriction was enhanced
during cold exposure (28°C), and that the oc 2 -adrenoreceptor
response was blocked by MK-912, a selective oc 2C -adrenergic
receptor antagonist. Interestingly, the Western blot analysis of
the tail arteries indicated that the oc 2C -adrenoreceptor was ex-
pressed predominantly as a low-molecular-weight, glycosy-
lated form, a potentially inactive form of the receptor. The
authors concluded that cold exposure may activate otherwise
silent oc 2C -adrenergic receptors, and that selective blockade
of these receptors might provide an effective treatment of
Raynaud's syndrome. 120

Sympathetic nerve activity is the primary means of stimu-
lating vascular adrenergic receptors in vivo. In addition to nor-
epinephrine, synaptic vesicles contain other cotransmitters
that are simultaneously released at the neuronal varicosi-
ties. 121 Specifically, both purinergic (ATP) and serotonergic
co-transmission have been documented with sympathetic
stimulation. 121 ' 122 In canine saphenous veins, a component
of neurogenic contractions is mediated by a nonadrenergic
mechanism, as indicated by its insensitivity to phento-
lamine. 123 This component is augmented by cooling, but is
blocked at both high and low temperatures by purinergic de-
sensitization with a,(3-methylene adenosine triphosphate. 123

pa rt I Vascular pathology and physiology

In these same vessels, serotonin-induced contractions are
augmented by cooling. 102

Platelet-derived vasoactive compounds

Aggregating platelets release 5-hydroxytryptamine, or sero-
tonin, which evokes contractions of vascular smooth muscle
from many species by stimulating S 2 -receptors. 124 The com-
pound also is released, at least in some beds, during sympa-
thetic nerve stimulation. 121 Evidence suggesting the
involvement of serotonin in cold-induced vasospasm has
been derived from a variety of preparations. Serum levels
have been reported elevated in Raynaud's syndrome. 42 In
some Raynaud's patients, the S 2 -antagonist, ketanserin, pro-
vides effective relief. 40

Coffman and Cohen 40 investigated the role of serotonergic
vasoconstriction in the normal sympathetic reflex response to
whole-body cooling. Total FBF was measured by air plethys-
mography and venous occlusion; capillary blood flow was
measured by clearance of a radioisotope injected in the finger-
tip. Selective stimulation (serotonin) and blockade (ke-
tanserin) of forearm S 2 -receptors was achieved by infusion
through a brachial catheter. Interactions between serotonin
and oc-adrenoceptors were excluded by pretreatment with
prazosin. Serotonin significantly decreased FBF in a dose-
dependent manner; this vasoconstriction was blocked by
ketanserin (50 jig/min). Importantly, the same concentration
of ketanserin was also able to block the reflex decrease
in FBF elicited when these patients were subjected to
whole-body cooling. The authors conclude that innervated S 2 -
receptors occur in the human finger and that they are activated
during the sympathetic vasoconstriction evoked by exposure
to cold.

In addition to increasing the amount of serotonin released
during sympathetic vasoconstriction, cooling also augments
the force of contractions evoked by a given concentration of
the compound. Data obtained with isolated vascular prepara-
tions show clearly that contractions elicited by serotonin are
augmented by cooling. This was demonstrated in both venous
and arterial tissues from human and nonhuman species. In ex-
periments very similar in design to those described above, a
comparable degree of cooling augmented serotonin-induced
contractions in canine saphenous veins and arteries. 102 ' 125
Human vessels shown to contract in vitro in response to
serotonin include the internal mammary artery, 126 veins and
arteries of the hands, 127 and digital arteries. 128 ' 129 Although
none of these studies on isolated human tissue evaluated
the effect of temperature on serotonin-induced contractions,
our data indicate that in the human saphenous vein they
are indeed augmented by cooling. 130 In rings of vessels ob-
tained at surgery, we observed profound enhancement of sero-
tonin-induced contractions when the temperature was
lowered to 24°C . We were able to block this augmentation with
ketanserin in a concentration-dependent manner.

Endothelium-derived vasoactive substances

Prior to 1980, the function of the vascular endothelial layer was
viewed almost exclusively as that of a selective-permeability
barrier to diffusion of substances into or out of the vascular
space. In that year, Furchgott and Zawadzki 131 first reported
the ability of the endothelium to modulate smooth muscle
contractility. Their observations that acetylcholine caused
contractions in vessels devoid of endothelium but relaxed
those with intact endothelial layers opened a new field of in-
vestigation into endothelium-derived vasoactive factors —
work which has radically altered our perceptions of both the
endothelium and local vascular control. A large majority of
this work focused on the relaxant factors (EDRFs), one of
which is almost certainly the nitric oxide radical, NO 132 or a ni-
trosothiol such as S-nitrosocysteine. 133 Alterations in endothe-
lial structure and function occur in Raynaud's patients. 134 ' 135
Studies evaluating the effects of cooling on endothelial func-
tion are scarce. De Mey and Vanhoutte 136 reported that while
acetylcholine-induced relaxations in the canine femoral artery
were unaffected by temperature changes in the 37°C range,
further cooling abolished the endothelium-dependent
response.

Recently, increasing attention has been directed toward
endothelium-derived contractile factors (EDCFs). This is
largely attributable to the discovery, sequencing, and cloning
of endothelin, a 21-amino acid peptide produced by cultured
endothelial cells. 137 It is the most potent constrictor of vascular
smooth muscle known, causing half-maximal contractions at
4 x 10~ 10 M in isolated segments of porcine coronary artery and
showing comparable potency in a variety of other arterial
preparations. Endothelin-induced contractions are resistant
to antagonism by a-adrenergic, H 1 -histaminergic and sero-
tonergic blockers, and appear entirely dependent upon extra-
cellular calcium. 137

Because of the extreme potency of endothelin as a contrac-
tile agonist, and because it is produced by endothelial cells lin-
ing much if not all of the vasculature, it is a prime candidate for
investigation in many vasospastic diseases. Intraarterial ad-
ministration of endothelin caused a pronounced and long-
lived reduction in coronary blood flow accompanied by
evidence of myocardial ischemia. 138 In isolated perfused kid-
ney preparations, endothelin decreases both renal blood flow
and glomerular filtration rate at lower concentrations than
does angiotensin II. The decreases are also of longer duration,
mimicking patterns associated with acute renal failure. 139 A
causative role in cerebral vasospasm has been suggested for
endothelin. 140 Two recent preliminary reports have suggested
the involvement of endothelin in Raynaud's phenomenon.
Both indicate that plasma endothelin levels rise following the
cold-pressor test. 53,141 One group found the increase was
greater in Raynaud's patients than in controls. 53 ' 141

Only two in-vitro studies to date have investigated the influ-
ence of cooling on endothelin-induced contractions of vascu-

chapter 7 Physiology of vasospastic disorders

lar preparations. Dalman et al. 142 obtained segments of greater
saphenous veins from patients at surgery and prepared them
as rings for organ chamber studies. As expected, endothelin
evoked concentration-dependent contractions in these rings.
When the temperature of the tissue was acutely lowered from
37°C to 24°C, a modest augmentation of the contractions was
observed. This cold-induced augmentation was less pro-
nounced than was observed with norepinephrine in the same
study.

The thermosensitivity of endothelin-evoked contractions
has also been addressed in the central artery of the rabbit ear
by Monge and colleagues. 143 These investigators observed a
slight but significant diminution of the contractions when
the preparation was cooled to 24°C. The combined results
of these studies do not provide much support for the hypothe-
sized role of endothelin in Raynaud's syndrome. Additional
doubt arises from the observed differences in the relative time
courses of the two types of contraction: vasospasm associated
with Raynaud's syndrome is much more rapid in onset than
vasospasm evoked by endothelin. However, since clinical
studies have reported elevated levels of endothelin in
Raynaud's patients as described above, 53 ' 141 a causal role
for the peptide can not be definitively ruled out at present.
One unaddressed possibility is that endothelin might enhance
contractions evoked by other agonists in Raynaud's syndrome
as reported in other vessels. 144 Further experiments are
required to determine the role, if any, of endothelin in
Raynaud's syndrome.

Summary

In this chapter we have presented current information on the
physiologic mechanisms of vasospasm using insights ob-
tained from both clinical and laboratory studies. Raynaud's
syndrome has been emphasized both because we believe this
disease is closely related to other vasospastic disorders, and
because it is the best studied clinical vasospastic condition.
Clinical studies have, for the most part, documented the
vasospastic event but not the responsible mechanisms.
Laboratory-based studies have shed light on the specific ef-
fects of cold on contractile processes, individual receptor
types, and vasoactive substances. As a result, several likely
mechanisms have been identified but conclusive evidence
demonstrating alterations of any single mechanism in a va-
sospastic disease state has not been detected. Such data contin-
ue to prove elusive due in large part to problems inherent in
either (i) obtaining the target vessels (e.g. normal and spastic
digital, cerebral, or coronary arteries) for isolated study or (ii)
delivering selective agents to the site of spasm.

Acknowledgment

Supported by grant number 5-M01-RR-00334, General

Clinical Research Centers, Division of Research Resources,
National Institutes of Health, Bethesda, MD, USA.

References

1. Miller D, Waters DD, Warnica W, Szlachcic J, Kreeft J, Theroux P.
Is variant angina the coronary manifestation of a generalized va-
sospastic disorder? N Engl] Med 1981; 304:763.

2. O'Keeffe ST, Tsapatsaris NP, Beetham WP Jr. Increased preva-
lence of migraine and chest pain in patients with primary Ray-
naud disease. Ann Intern Med 1992; 116:985.

3. Raynaud M. Nouvelles recherches sur la nature et la traitment de
l'aspyxie locule des extremites. Arch Gen Med 1874; 1:85.

4. Lewis T. Experiments relating to the peripheral mechanism in-
volved in spastic arrest of the circulation in the fingers, a variety
of Raynaud's disease. Heart 1929; 15:7.

5. Peacock J. Peripheral venous blood concentrations of epineph-
rine and norepinephrine in primary Raynaud's disease. Circ Res
1959; 7:821.

6. Sapira JD, Rodnan GP, Scheib ET, Klaniecki T, Rizk M. Studies of
endogenous catecholamines in patients with Raynaud's phe-
nomenon secondary to progressive systemic sclerosis (scleroder-
ma). Am J Med 1972; 52:330.

7. Freedman RR, Keegan D, Migaly P, Galloway MP, Mayes M.
Plasma catecholamines during behavioral treatments for
Raynaud's disease. Psychosom Med 1991; 53:433.

8. Coffman JD, Cohen AS. Total and capillary fingertip blood flow
in Raynaud's phenomenon. N Engl J Med 1971; 285:259.

9. Porter JM, Snider RL, Bardana EJ, Rosch J, Eidemiller LR. The di-
agnosis and treatment of Raynaud's phenomenon. Surgery 1975;
77:11.

10. Rosch J, Porter JM, Gralino BJ. Cryodynamic hand angiography
in the diagnosis and management of Raynaud's syndrome. Cir-
culation 1977; 55:807.

11. Arneklo-Nobin B, Edvinsson L, Eklof B, Haffajee D, Owman C,
Thylen U. Analysis of vasospasm in hand arteries by in vitro
pharmacology, hand angiography and finger plethysmography.
Gen Pharmacol 1983; 14:65.

12. Porter JM, Bardana EJ Jr, Baur GM, Wesche DH, Andrasch RH,
Rosch J. The clinical significance of Raynaud's syndrome.
Surgery 1976; 80:756.

13. Cleophas TJ, van Lier HJ, Fennis JF, van 't Laar A. Treatment of
Raynaud's syndrome with adrenergic alpha-blockade with or
without beta-blockade. Angiology 1984; 35:29.

14. Ekenvall L, Lindblad LE, Norbeck O, Etzell BM. Alpha-
adrenoceptors and cold-induced vasoconstriction in human
finger skin. Am J Physiol 1988; 255:H1000.

15. Williams LT, Snyderman R, Lefkowitz RJ. Identification of beta-
adrenergic receptors in human lymphocytes by (-) (3H) al-
prenolol binding. / Clin Invest 1976; 57:149.

16. Brodde OE, Seher U, Nohlen M, Fischer WM, Michel MC.
Correlation between human myometrial and platelet alpha 2-
adrenoceptor density. Eur J Pharmacol 1988; 150:403.

17. Keenan EJ, Porter JM. Alpha-2 adrenergic receptors in
platelets from patients with Raynaud's syndrome. Surgery 1983;
94:204.

18. Edwards JM, Phinney ES, Taylor LM Jr, Keenan EJ, Porter JM.

pa rt I Vascular pathology and physiology

Alpha 2-adrenergic receptor levels in obstructive and spastic
Raynaud's syndrome. / Vase Surg 1987; 5:38.

19. Graafsma SJ, Wollersheim H, Droste HT et ah Adrenoceptors on
blood cells from patients with primary Raynaud's phenomenon.
Clin Sci (Lond) 1991; 80:325.

20. Freedman RR, Sabharal SC, Desai N, Wenig P, Mayes M.
Increased alpha-adrenergic responsiveness in idiopathic
Raynaud's disease. Arthritis Rheum 1989; 32:61.

21. Motulsky HJ, Insel PA. Adrenergic receptors in man: direct iden-
tification, physiologic regulation, and clinical alterations. N Engl
J Med 1982; 307:18.

22. Lefkowitz RJ. Clinical physiology of adrenergic receptor regula-
tion. Am J Physiol 1982;243:E43.

23. Jones CR, Giembcyz M, Hamilton CA et ah Desensitization of
platelet alpha 2-adrenoceptors after short term infusions of
adrenoceptor agonist in man. Clin Sci (Lond) 1986; 70:147.

24. Marshall AJ, Roberts CJ, Barritt DW. Raynaud's phenomenon
as side effect of beta-blockers in hypertension. Br Med } 1976; 1:
1498.

25. Eliasson K, Lins LE, Sundqvist K. Vasospastic phenomena in pa-
tients treated with beta-adrenoceptor blocking agents. Acta Med
ScandSuppl 1979; 628:39.

26. Cobbold A, Ginsburg J, Paton A. Circulatory, respiratory, and
metabolic responses to isopropylnoradrenaline in man. / Physiol
1960; 151:539.

27. Cohen RA, Coffman JD. Beta-adrenergic vasodilator mechanism
in the finger. Circ Res 1981; 49:1196.

28. Freedman RR, Sabharwal SC, Ianni P, Desai N, Wenig P, Mayes
M. Nonneural beta-adrenergic vasodilating mechanism in tem-
perature biofeedback. Psychosom Med 1988; 50:394.

29. Ohlsson O, Lindell SE. The effects of pindolol and prazosin on
hand blood flow in patients with cold extremities and on treat-
ment with beta-blockers. Acta Med Scand 1981; 210:217.

30. Eliasson K, Danielson M, Hylander B, Lindblad LE. Raynaud's
phenomenon caused by beta-receptor blocking drugs. Improve-
ment after treatment with a combined alpha- and beta-blocker.
Acta Med Scand 1984; 215:333.

31. Franssen C, Wollersheim H, de Haan A, Thien T. The influence of
different beta-blocking drugs on the peripheral circulation in
Raynaud's phenomenon and in hypertension. / Clin Pharmacol
1992; 32:652.

32. Giovanni B, Giuseppina CM, Susanna F, Roberto M. Altered reg-
ulator mechanisms of presynaptic adrenergic nerve: a new phys-
iopathological hypothesis in Raynaud's disease. Microvasc Res
1984; 27:110.

33. Brotzu G, Falchi S, Mannu B, Montisci R, Petruzzo P, Staico R. The
importance of presynaptic beta receptors in Raynaud's disease. /
Vase Surg 1989; 9:767.

34. Brotzu G, Susanna F, Roberto M, Palmina P. Beta-blockers: a new
therapeutic approach to Raynaud's disease. Microvasc Res 1987;
33:283.

35. Bevan JA, Brayden JE. Nonadrenergic neural vasodilator mecha-
nisms. Circ Res 1987; 60:309.

36. VIP and the skin. Lancet 1991; 337:886.

37. Coffman JD, Cohen RA. Cholinergic vasodilator mechanism in
human fingers. Am ] Physiol 1987; 252:H594.

38. Bunker CB, Foreman JC, Dowd PM. Digital cutaneous vascular
responses to histamine and neuropeptides in Raynaud's phe-
nomenon. Jin vest Dermatol 1991; 96:314.

39. Halpern A, Kuhn P, Shaftel H et ah Raynaud's disease, Raynaud's
phenomenon, and serotonin. Angiology 1960; 11:151.

40. Coffman JD, Cohen RA. Serotonergic vasoconstriction in human
fingers during reflex sympathetic response to cooling. Am J Phys-
iol 1988; 254:H889.

41. Seibold JR, Jageneau AH. Treatment of Raynaud's phenomenon
with ketanserin, a selective antagonist of the serotonin2 (5-HT2)
receptor. Arthritis Rheum 1984; 27:139.

42. Stranden E, Roald OK, Krohg K. Treatment of Raynaud's phe-
nomenon with the 5-HT2-receptor antagonist ketanserin. Br Med
J (Clin Res Ed) 1982; 285:1069.

43. Dowd PM, Bunker CB, Bull HA et ah Raynaud's phenomenon,
calcitonin gene-related peptide, endothelin, and cutaneous vas-
culature. Lancet 1990; 336:1014.

44. McEwan JR, Benjamin N, Larkin S, Fuller RW, Dollery CT,
Maclntyre I. Vasodilatation by calcitonin gene-related peptide
and by substance P: a comparison of their effects on resistance
and capacitance vessels of human forearms. Circulation 1988;
77:1072.

45. Clarke JG, Benjamin N, Larkin SW, Webb DJ, Davies GJ, Maseri
A. Endothelin is a potent long-lasting vasoconstrictor in men. Am
J Physiol 1989; 257:H2033.

46. Hughes AD, Thorn SA, Woodall N et ah Human vascular re-
sponses to endothelin-1: observations in vivo and in vitro. / Car-
diovasc Pharmacol 1989; 13:S225.

47. Bunker CB, Goldsmith PC, Leslie TA, Hayes N, Foreman JC,
Dowd PM. Calcitonin gene-related peptide, endothelin-1, the cu-
taneous microvasculature and Raynaud's phenomenon. Br J Der-
matol 1996; 134:399.

48. Shawket S, Dickerson C, Hazleman B, Brown MJ. Prolonged
effect of CGRP in Raynaud's patients: a double-blind
randomised comparison with prostacyclin. Br J Clin Pharmacol
1991; 32:209.

49. Bunker CB, Foreman JC, Dowd PM. Calcitonin gene-related pep-
tide and Raynaud's phenomenon. Lancet 1990; 335:239.

50. Brain SD, Petty RG, Lewis JD, Williams TJ. Cutaneous blood flow
responses in the forearms of Raynaud's patients induced by local
cooling and intradermal injections of CGRP and histamine. Br J
Clin Pharmacol 1990; 30:853.

51. Bunker CB, Terenghi G, Springall DR, Polak JM, Dowd PM. Defi-
ciency of calcitonin gene-related peptide in Raynaud's phenom-
enon. Lancet 1990; 336:1530.

52. Terenghi G, Bunker CB, Liu YF et ah Image analysis quantifica-
tion of peptide-immunoreactive nerves in the skin of patients
with Raynaud's phenomenon and systemic sclerosis. / Pathol
1991; 164:245.

53. Zamora MR, O'Brien RF, Rutherford RB, Weil JV Serum endothe-
lin-1 concentrations and cold provocation in primary Raynaud's
phenomenon. Lancet 1990; 336:1144.

54. Leppert J, Ringqvist A, Karlberg BE, Ringqvist I. Whole-body
cooling increases plasma endothelin-1 levels in women with pri-
mary Raynaud's phenomenon. Clin Physiol 1998; 18:420.

55. Kanno K, Hirata Y, Shichiri M, Numano F, Miyasaka N, Marumo
F. Raised circulating endothelin-1 in vascular disease and its
pathogenic role in Raynaud's phenomenon (abstract). Circula-
tion 1990; 82 (Suppl. III):226.

56. Smits P, Hofman H, Rosmalen F, Wollersheim H, Thien T.
Endothelin-1 in patients with Raynaud's phenomenon. Lancet
1991; 337:236.

chapter 7 Physiology of vasospastic disorders

57. Hynynen M, Ilmarinen R, Saijonmaa O, Tikkanen I, Fyhrquist F.
Plasma endothelin-1 concentration during cold exposure (let-
ter). Lancet 1991; 337:1104.

58. Harker C, Edwards J, Taylor L, Porter J. Plasma endothelin-1 con-
centration during cold exposure. Lancet 1991; 337:1104.

59. Cimminiello C, Milani M, Uberti T, Arpaia G, Perolini S,
Bonfardeci G. Endothelin, vasoconstriction, and endothelial
damage in Raynaud's phenomenon. Lancet 1991; 337:114.

60. Smyth AE, Bell AL, Bruce IN, McGrann S, Allen JA. Digital vas-
cular responses and serum endothelin-1 concentrations in pri-
mary and secondary Raynaud's phenomenon. Ann Rheum Dis
2000; 59:870.

61. Dahlof B, Gustafsson D, Hedner T, Jern S, Hansson L. Regional
haemodynamic effects of endothelin-1 in rat and man: un-
expected adverse reaction. JHypert ens 1990; 8:811.

62. Taylor LM Jr, Rivers SP, Keller FS, Baur GM, Porter JM. Treatment
of finger ischemia with Bier block reserpine. Surg Gynecol Obstet
1982; 154:39.

63. Ringqvist A, Leppert J, Myrdal U, Ahlner J, Ringqvist I,
Wennmalm A. Plasma nitric oxide metabolite in women with
primary Raynaud's phenomenon and in healthy subjects. Clin
Physiol 1997; 17:269.

64. Bedarida G, Kim D, Blaschke TF, Hoffman BB. Venodilation in
Raynaud's disease. Lancet 1993; 342:1451.

65. Skyhoj Olsen T. Migraine with aura: onset of the attack. In:
Olesen J, ed. Migraine and Other Headaches. New York: Raven
Press, 1991:79.

66. Drummond PD, Lance JW. Extracranial vascular changes and the
source of pain in migraine headache. Ann Neurol 1983; 13:32.

67. Leao A. Spreading depression of activity in the cerebral cortex. /
Neurophysiol 1944; 7:359.

68. Lauritzen M, Hansen A. Spreading depression of Leao. Possible
relation to migraine pathophysiology. In: Olesen J, Edvinsson L,
eds. Basic Mechanisms of Headache. Amsterdam: Elsevier, 1988:
439.

69. Blau JN. Migraine: theories of pathogenesis. Lancet 1992;
339:1202.

70. Hasselmark L, Malmgren R, Hannerz J. Platelet alpha 2-
adrenoceptor binding in migraine. Headache 1988; 28:587.

71. Downey J A, Frewin DB. Vascular responses in the hands of pa-
tients suffering from migraine. / Neurol Neurosurg Psychiatry
1972; 35:258.

72. Goadsby PJ, Edvinsson L, Ekman R. Vasoactive peptide release
in the extracerebral circulation of humans during migraine
headache. Ann Neurol 1990; 28:183.

73. Edvinsson L, Goadsby PJ. Extracerebral manifestations in mi-
graine. A peptidergic involvement? } Intern Med 1990; 228:299.

74. Osier W. The Principles and Practice of Medicine. New York: D.
Appleton & Co., 1893:655.

75. Prinzmetal M, Keimainer R, Merlais R. A variant form of angina
pectoris: a preliminary report. Am J Med 1959; 27:375.

76. Oliva PB, Potts DE, Pluss RG. Coronary arterial spasm in
Prinzmetal angina. Documentation by coronary arteriography.
N Engl J Med 1973; 288:745.

77. Leon-Sontmayor L. Cardiac migraine: report of twelve cases. An-
giology 1974; 25:161.

78. Robertson D, Robertson RM, Nies AS, Oates JA, Friesinger GC.
Variant angina pectoris: investigation of indexes of sympathetic
nervous system function. Am } Cardiol 1979; 43:1080.

79. Ricci DR, Orlick AE, Cipriano PR, Guthaner DF, Harrison DC. Al-
tered adrenergic activity in coronary arterial spasm: insight into
mechanism based on study of coronary hemodynamics and the
electrocardiogram. Am J Cardiol 1979; 43:1073.

80. Chierchia S, Davies G, Berkenboom G, Crea F, Crean P, Maseri A.
alpha- Adrenergic receptors and coronary spasm: an elusive link.
Circulation 1984; 69:8.

81. Clarke JG, Davies GJ, Kerwin R et al. Coronary artery infusion of
neuropeptide Y in patients with angina pectoris. Lancet 1987;
1:1057.

82. Horio Y, Yasue H, Okumura K et al. Effects of intracoronary injec-
tion of acetylcholine on coronary arterial hemodynamics and di-
ameter. Am J Cardiol 1988; 62:887.

83. Greenberg B, Rhoden K, Barnes P. Calcitonin gene-related pep-
tide (CGRP) is a potent non-endothelium-dependent inhibitor of
coronary vasomotor tone. Br J Pharmacol 1987; 92:789.

84. Joyce CD, Prinz RA, Thomas JX et al. Calcitonin gene-related pep-
tide increases coronary flow and decreases coronary resistance. /
Surg Res 1990; 49:435.

85. Egashira K, Inou T, Yamada A, Hirooka Y, Takeshita A. Preserved
endothelium-dependent vasodilation at the vasospastic site in
patients with variant angina. / Clin Invest 1992; 89:1047.

86. Teragawa H, Kato M, Kurokawa J, Yamagata T, Matsuura
H, Chayama K. Endothelial dysfunction is an independent
factor responsible for vasospastic angina. Clin Sci (Lond) 2001;
101:707.

87. Tomimura M, Saitoh T, Kishida H, Kusama Y, Takano T. [Clinical
significance of plasma concentration of macrophage colony-
stimulating factor in patients with vasospastic angina]. / Cardiol
2002; 39:19.

88. Gustafsson R, Mannting F, Kazzam E, Waldenstrom A, Hallgren
R. Cold-induced reversible myocardial ischaemia in systemic
sclerosis. Lancet 1989; 2:475.

89. Porter J, Edwards J, Taylor LJ. Upper extremity vasospastic dis-
ease. In: Ernst C, Stanley J, eds. Current Therapy in Vascular
Surgery. Toronto: BC Decker, 1987:81.

90. Arosio E, Montesi G, Zannoni M, Paluani F, Lechi A. Compara-
tive efficacy of ketanserin and pentoxiphylline in treatment of
Raynaud's phenomenon. Angiology 1989; 40:633.

91. Edwards J, Porter J. Update on Raynaud's syndrome. Semin Vase
Surg 1990; 3:227.

92. Ruffolo RR Jr, Nichols AJ, Stadel JM, Hieble JP Structure
and function of alpha-adrenoceptors. Pharmacol Rev 1991; 43:
475.

93. McGrath JC. Evidence for more than one type of post-junctional
alpha-adrenoceptor. Biochem Pharmacol 1982; 31:467.

94. De Mey J, Vanhoutte PM. Uneven distribution of postjunctional
alpha 1- and alpha 2-like adrenoceptors in canine arterial and ve-
nous smooth muscle. Circ Res 1981; 48:875.

95. Flavahan NA, Rimele TJ, Cooke JP, Vanhoutte PM. Characteri-
zation of postjunctional alpha-1 and alpha-2 adrenoceptors
activated by exogenous or nerve-released norepinephrine in
the canine saphenous vein. / Pharmacol Exp Ther 1984; 230:699.

96. Milnor WR, Stone DN, Sastre A. Contributions of alpha 1- and
alpha 2-adrenoceptors to contractile response in canine blood
vessels. Blood Vessels 1988; 25:199.

97. Kawai Y, Kobayashi S, Ohhashi T. Existence of two types of
postjunctional alpha-adrenoceptors in the isolated canine inter-
nal carotid artery. Can J Physiol Pharmacol 1988; 66:655.

pa rt I Vascular pathology and physiology

98. Kiowski W, Hulthen UL, Ritz R, Buhler FR. Alpha 2
adrenoceptor-mediated vasoconstriction of arteries. Clin
Pharmacol Ther 1983; 34:565.

99. Ruff olo RR Jr, Waddell JE, Yaden EL. Postsynaptic alpha adrener-
gic receptor subtypes differentiated by yohimbine in tissues from
the rat. Existence of alpha-2 adrenergic receptors in rat aorta. /
Pharmacol Exp Ther 1981; 217:235.

100. Harker CT, Vanhoutte PM. Cooling and alpha adrenergic re-
sponses in the saphenous vein of the rabbit. / Pharmacol Exp Ther
1989; 249:56.

101. Langer SZ, Hicks PE. Alpha-adrenoreceptor subtypes in blood
vessels: physiology and pharmacology. / Cardiovasc Pharmacol
1984; 6:S547.

102. Vanhoutte PM, Shepherd JT. Effect of temperature on reactivity
of isolated cutaneous veins of the dog. Am } Physiol 1970; 218:
187.

103. Vanhoutte PM, Verbeuren TJ. Depression by local cooling of 3 H-
norepinephrine evoked by nerve stimulation in cutaneous veins.
Bloodvessels 1976; 13:92.

104. Kitazawa T, Kobayashi S, Horiuti K, Somlyo AV, Somlyo AP
Receptor-coupled, permeabilized smooth muscle. Role of the
phosphatidylinositol cascade, G-proteins, and modulation of the
contractile response to Ca2+. J Biol Chem 1989; 264:5339.

105. Kitazawa T, Gaylinn BD, Denney GH, Somlyo AP. G-protein-
mediated Ca2+ sensitization of smooth muscle contraction
through myosin light chain phosphorylation. / Biol Chem 1991;
266:1708.

106. Flavahan NA, Lindblad LE, Verbeuren TJ, Shepherd JT,
Vanhoutte PM. Cooling and alpha 1- and alpha 2-adrenergic
responses in cutaneous veins: role of receptor reserve. Am } Phys-
z'o/1985;249:H950.

107. Harker CT, Ousley PJ, Harris EJ, Edwards JM, Taylor LM, Porter
JM. The effects of cooling on human saphenous vein reactivity to
adrenergic agonists. / Vase Surg 1990; 12:45.

108. Docherty JR, Hyland L. Evidence for neuro-effector transmission
through postjunctional alpha 2-adrenoceptors in human saphe-
nous vein. Br J Pharmacol 1985; 84:573.

109. Eskinder H, Hillard CJ, Olinger GN et ah Alpha adrenoceptor
subtypes and receptor reserve in human versus canine saphe-
nous vein: sensitivity to blockade by nitroglycerin. / Pharmacol
Exp Ther 1988; 247:941.

110. Harker CT, Bowman CJ, Taylor LM Jr, Porter JM. Cooling aug-
ments human saphenous vein reactivity to electrical stimulation.
/ Cardiovasc Pharmacol 1994; 23:453.

111. Faber JE. In situ analysis of alpha-adrenoceptors on arteriolar
and venular smooth muscle in rat skeletal muscle micro-
circulation. Circ Res 1988; 62:37.

112. Flavahan NA, Cooke JP, Shepherd JT, Vanhoutte PM. Human
postjunctional alpha-1 and alpha-2 adrenoceptors: differential
distribution in arteries of the limbs. / Pharmacol Exp Ther 1987;
241:361.

113. Nielsen H, Thorn SM, Hughes AD, Martin GN, Mulvany MJ,
Sever PS. Postjunctional alpha 2-adrenoceptors mediate
vasoconstriction in human subcutaneous resistance vessels. Br J
Pharmacol 1989; 97:829.

114. Weiss RJ, Webb RC, Smith CB. Alpha-2 adrenoreceptors on arter-
ial smooth muscle: selective labeling by [3H]clonidine./P/zflrmfl-
col Exp Ther 1983; 225:599.

115. Faber JE. Effect of local tissue cooling on microvascular smooth

muscle and postjunctional alpha 2-adrenoceptors. Am J Physiol
1988;255:H121.

116. Winquist RJ, Bevan JA. Temperature sensitivity of tone in the rab-
bit facial vein: myogenic mechanism for cranial thermoregula-
tion. Science 1980; 207:1001.

117. Harker CT, Vanhoutte PM. Cooling the central ear artery of the
rabbit: myogenic and adrenergic responses. / Pharmacol Exp Ther
1988; 245:89.

118. Cooke JP, Creager SJ, Scales KM et al. Role of digital artery
adrenoceptors in Raynaud's disease. Vase Med 1997; 2:1.

119. Freedman RR, Baer RP, Mayes MD. Blockade of vasospastic at-
tacks by alpha 2-adrenergic but not alpha 1-adrenergic antago-
nists in idiopathic Raynaud's disease. Circulation 1995; 92:1448.

120. Chotani MA, Flavahan S, Mitra S, Daunt D, Flavahan NA. Silent
alpha(2C)-adrenergic receptors enable cold-induced vasocon-
striction in cutaneous arteries. Am } Physiol Heart Circ Physiol
2000;278:H1075.

121. Griffith SG, Lincoln J, Burnstock G. Serotonin as a neuro-
transmitter in cerebral arteries. Brain Res 1982; 247:388.

122. Burnstock G, Sneddon P. Evidence for ATP and noradrenaline
as cotransmitters in sympathetic nerves. Clin Sci (Lond) 1985;
68:89s.

123. Flavahan NA, Vanhoutte PM. Sympathetic purinergic vasocon-
striction and thermosensitivity in a canine cutaneous vein./ Phar-
macol Exp Ther 1986; 239:784.

124. Van Nueten JM, Janssen PA, Van Beek J, Xhonneux R, Verbeuren
TJ, Vanhoutte PM. Vascular effects of ketanserin (R 41 468), a
novel antagonist of 5-HT2 serotonergic receptors. / Pharmacol Exp
Ther 1981; 218:217.

125. Lindblad LE, Shepherd JT, Vanhoutte PM. Cooling augments
platelet-induced contraction of peripheral arteries of the dog.
Proc Soc Exp Biol Med 1984; 176:119.

126. Conti A, Monopoli A, Forlani A, Ongini E, Antona C, Biglioli P.
Role of 5-HT2 receptors in serotonin-induced contraction in the
human mammary artery. Eur } Pharmacol 1990; 176:207.

127. Arneklo-Nobin B, Owman C. Adrenergic and serotoninergic
mechanisms in human hand arteries and veins studied by fluo-
rescence histochemistry and in vitro pharmacology. Blood Vessels
1985; 22:1.

128. Moulds RF, Iwanov V, Medcalf RL. The effects of platelet-derived
contractile agents on human digital arteries. Clin Sci (Eond) 1984;
66:443.

129. Young MS, Iwanov V, Moulds RF. Interaction between platelet-
released serotonin and thromboxane A2 on human digital
arteries. Clin Exp Pharmacol Physiol 1986; 13:143.

130. Harker CT, Taylor LM Jr, Porter JM. Vascular contractions to sero-
tonin are augmented by cooling. / Cardiovasc Pharmacol 1991;
18:791.

131. Furchgott RF, Zawadzki JV The obligatory role of endothelial
cells in the relaxation of arterial smooth muscle by acetylcholine.
Nature 1980; 288:373.

132. Palmer R, Ferrigo A, Moncada S. Nitric oxide release accounts for
the biological activity of endothelium-derived relaxing factor.
Nature 1987; 327:524.

133. Myers PR, Minor RL Jr, Guerra R Jr, Bates JN, Harrison DG. Va-
sorelaxant properties of the endothelium-derived relaxing factor
more closely resemble S-nitrosocysteine than nitric oxide. Nature
1990; 345:161.

134. Kahaleh MB, Osborn I, LeRoy EC. Increased factor VHI/von

chapter 7 Physiology of vasospastic disorders

Willebrand factor antigen and von Willebrand factor activity in
scleroderma and in Raynaud's phenomenon. Ann Intern Med
1981; 94:482.

135. Belch JJ, Zoma AA, Richards IM, McLaughlin K, Forbes CD,
Sturrock RD. Vascular damage and factor-VIII-related antigen in
the rheumatic diseases. Rheumatol Int 1987; 7:107.

136. De Mey JG, Vanhoutte PM. Interaction between Na+,K+ ex-
changes and the direct inhibitory effect of acetylcholine on ca-
nine femoral arteries. Circ Res 1980; 46:826.

137. Yanagisawa M, Kurihara H, Kimura S et ah A novel potent vaso-
constrictor peptide produced by vascular endothelial cells. Na-
ture 1988; 332:411.

138. Kurihara H, Yamaoki K, Nagai R et ah Endothelin: a potent vaso-
constrictor associated with coronary vasospasm. Life Sci 1989;
44:1937.

139. Firth JD, Ratcliffe PJ, Raine AE, Ledingham JG. Endothelin: an
important factor in acute renal failure? Lancet 1988; 2:1179.

140. Vanhoutte PM, Auch-Schwelk W, Boulanger C et ah Does
endothelin-1 mediate endothelium-dependent contractions
during anoxia? / Cardiovasc Pharmacol 1989; 13:S124; discussion
S142.

141. Fyhrquist F, Saijonmaa O, Metsarinne K, Tikkanen I, Rosenlof K,
Tikkanen T. Raised plasma endothelin-I concentration following
cold pressor test. Biochem Biophys Res Commun 1990; 169:217.

142. Dalman R, Harker C, Taylor LJ, Porter J. Contractile response
of human vascular tissue to endothelin. Surg Forum 1990;
41:332.

143. Monge L, Garcia-Villalon AL, Montoya JJ, Garcia JL, Gomez B,
Dieguez G. Response of rabbit ear artery to endothelin-1 during
cooling. Br] Pharmacol 1991; 104:609.

144. Yang ZH, Richard V, von Segesser L et ah Threshold concentra-
tions of endothelin-1 potentiate contractions to norepinephrine
and serotonin in human arteries. A new mechanism of va-
sospasm? Circulation 1990; 82:188.

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Buerger's disease

John Blebea

Richard F. Kempczinski

Buerger's disease (thromboangiitis obliterans) is an inflam-
matory occlusive disorder of small to medium-sized arteries
and veins which occurs primarily in the extremities of young
male smokers. It was first described by von Winiwarter in 1879
who called it "endarteritis obliterans/ 7 believing that it repre-
sented a proliferation of intimal cells which lead to the occlu-
sion of the involved artery. 1 The clinical characteristics of this
disorder were more fully elucidated in 1908 by Leo Buerger, 2
who described the pathologic findings in eleven amputated
limbs and proposed that the disease be renamed "throm-
boangiitis obliterans" to reflect an intrinsic thrombotic process
within the lumen of the artery and associated vessel wall in-
flammation. His seminal insight into the underlying patho-
physiology, which is largely unchanged from our present day
understanding, prompted the eponym, Buerger's disease, for
this disorder.

An emphasis on the pathohistologic findings, with insuffi-
cient attention to the characteristic clinical presentation, led
to the disease being overdiagnosed during the subsequent
decades and little progress was made towards its understand-
ing. By 1960, the very existence of Buerger's disease as an
entity distinct from atherosclerosis, systemic embolization, or
idiopathic arterial thrombosis was questioned. 3,4 This contro-
versy has now been laid to rest but there are still no precise di-
agnostic criteria, or definitive pathognomonic characteristics,
for patient classification. These shortcomings plague even the
recent literature on the subject and make meaningful analysis
of the etiology, diagnosis, and treatment of Buerger's disease
more difficult. 5

Epidemiology

Buerger first suggested that thromboangiitis obliterans
(TAO) occurred more frequently in Jews of Eastern European
origin. However, this reflected the population that he was
treating at Mt Sinai Hospital in New York City. In fact,
Buerger's disease affects all races and ethnic groups, although
it is more common in the Middle and Far East than in Europe

and the United States. Its prevalence in Japan, where periodic
population studies have been performed, is approximately
5perl00000. 6

The incidence of TAO seems to have decreased. Using the
same clinical criteria for diagnosis in a well defined and stable
population, there has been an eightfold decrease in the inci-
dence of patients with Buerger's disease seen at the Mayo
Clinic over a 40-year period. In 1947, there were 104.3 patients
diagnosed per 100 000 population. This decreased consecu-
tively in every 5-year period until 1976 when the incidence
was 9.9/100 000. It has remained at approximately the same
level with the last reported figures being 13.5/100 000 for
1987. 7 Although the clinical criteria were unchanged, this de-
creasing incidence is probably due to the more widespread use
of angiography and diagnostic evaluation of possible vasculi-
tides and hypercoagulable states. These additional investiga-
tions have identified other causes for distal ischemia in
patients with similar clinical findings.

A possible explanation for this decreasing incidence is the
fall in the prevalence of male smokers in the United States from
52.6% in 1955 to only 25.7% in 2000. 8 ' 9 The higher number of
patients seen in Japan with TAO, and a similar decreasing inci-
dence, may be partially explained by the greater number of
smokers among men there, 83.7% in 1965, and a similar drop in
smoking rates to 57.5% by 1996. 7 ' 10 ' 11

Etiology

The cause of Buerger's disease remains unknown. No
causative bacterial or viral agent has ever been identified.
However, smoking is almost universally associated with the
initiation and progression of TAO. When objective measure-
ments of the amount of smoking such as serum carboxyhemo-
globin levels among these patients have been performed,
patients with Buerger's smoked significantly more than simi-
lar groups with atherosclerosis obliterans. 12 A similar objec-
tive parameter using the stable urinary metabolite of nicotine,
cotinine, documented the significant association between con-

chapter 8 Buerger's disease

tinued smoking and higher rates of disease progression. 13 True
nonsmokers virtually never develop the disease. In those rare
patients who are not smokers themselves, passive smoking in
the home or the work place may explain the onset of TAO. The
epidemiologic association with tobacco does not, however, es-
tablish the exact cause, nor clarify the mechanism of vascular
injury in Buerger's disease.

Becker et al. u purified a tobacco glycoprotein (TGP) and
suggested that it may act as an antigen and produce immuno-
logically mediated endothelial injury. Papa et al. 15 examined
13 patients with Buerger's disease for cellular and humoral
sensitivity to TGP and found no difference between healthy
smokers and smokers with Buerger's. It does not appear that
TGP alone is a pathogenic immunologic factor, although this
does not exclude other tobacco components, or TPG in com-
bination with other agents, from such a role. The specific
presenting antigen, however, has yet to be discovered. With
the present availability of nicotine patches, the potential
etiologic role of nicotine vs. the other constituents of cigarette
smoke could more easily be evaluated. If nicotine patches
can assist patients to give up smoking, yet not lead to sympto-
matic exacerbations of the disease, it would be both an
additional aid in treatment and provide a further clue as to
the causal relationship between tobacco components and
Buerger's disease.

There may be important autoimmune mechanisms active in
TAO which are not directly linked to cigarette smoking. Gulati
et ah, using homogenized human arteries as antigens, found
evidence of specific cellular immunity, increased serum im-
munoglobulins, antiarterial antibodies, and immune com-
plexes in the diseased vessels of patients with Buerger's. 16 ' 17
Increased levels of circulating IgG-containing immune
complexes have been described in these patients by deAlbu-
querque et al 18 but not confirmed by Smolen et al. 19 Higher de-
grees of cell-mediated sensitivity and antibodies to collagen I
and III, constituents of human arteries, have been found in pa-
tients with Buerger's compared with those with atherosclerot-
ic disease. 19 ' 20 There was a positive correlation between the
degree of immunologic response and the severity of initial
symptoms. Eichhorn et al. demonstrated significantly in-
creased serum antiendothelial-cell antibody titers in patients
with active TAO compared with normal subjects or patients in
remission. 21 Kobayashi etal. 22 proposed that Buerger's disease
is a vasculitis induced by an antigen in the intimal layer. This
may be initiated by T-cell-mediated cellular immunity as re-
flected by the increased number of CD4+ T cells found acutely
next to the intima. There was also found an infiltration of HLA-
DR+ cells and immunoglobulins and complement factors de-
posited in a linear manner along the internal elastic lamina.
Such findings were not seen in atherosclerotic specimens. The
significance of these immunologic findings awaits further
elucidation to determine what specific antigens may be re-
sponsible and to what degree immunologic mechanisms are
causal factors in the development of the disease or merely

reflect injury to the vessel wall and are secondary immune
responses.

A genetic predisposition may also play a role in the develop-
ment of Buerger's disease. Human lymphocyte antigen typing
(HLA) has found a significantly increased frequency of the A9
and B5, 23 DR4, 15 Aw24, Bw40, Bw54, Cwl, DR2, 24 Al, B8, 19 and
BwlO 25 antigens in patients with TAO. Unfortunately, there is
no consistent pattern to these findings among the various re-
ported series and others have not been able to confirm some of
these results. 26 Nonetheless, there may be a gene, linked to the
presence or absence of some of these HLA, that can control the
susceptibility of a particular individual to the disease. This
predisposition, in combination with tobacco products and
environmental factors, may induce the disease in selected
individuals.

Pathology

There are no specific pathognomonic histologic findings for
Buerger's disease. However, there are some highly character-
istic changes that are seen with the disease, especially if the
specimens are obtained during the acute stage. Nevertheless,
the diagnosis of Buerger's should be made only when the clin-
ical presentation is consistent with TAO and the histologic
findings are confirmatory. To make a diagnosis solely on the
basis of histology is to repeat the errors of the past.

Buerger's disease is a segmental inflammatory occlusive
process involving primarily the medium and small-sized ar-
teries of the extremities, although involvement of the visceral
vessels has also been described. 27 ' 28 The inflammatory
changes may affect all three layers of the vessel wall but the
architecture of the wall is preserved (Fig. 8.1A). Commonly,
thrombosis of the lumen occurs in the involved segments. This
becomes organized, cellular infiltration takes place and later
recanalization may be seen. The inflammatory process may
extend to involve the adjacent vein and nerve. These changes
may be separated into an early acute and a later chronic stage,
although the clinical spectrum of the disease represents a con-
tinuum between these two categories.

Acute stage

Occasionally, an acute lesion can be clinically recognized in the
superficial radial or posterior tibial arteries when the overly-
ing skin reddens and becomes tender, indicating a marked pe-
riarterial inflammation. 29 Microscopically, the occluding fresh
or organizing thrombus within the lumen demonstrates evi-
dence of intense focal inflammation. The thrombus is very cel-
lular and contains lymphocytes, including neutrophils, and
multinucleated giant cells. Microabscesses consisting of poly-
morphonuclear leukocytes and giant cells may be present but
there are no microorganisms (Fig. 6.1B). 22 ' 30 There is intimal
thickening with proliferation of endothelial cells and lympho-

pa rt I Vascular pathology and physiology

■■*-*;.

Figure 8.1 Photomicrographs of an inferior
mesenteric artery with occlusion of the lumen
with recanalized thrombus and intimal
proliferation. (A) The three layers of the vessel
wall are well preserved and the internal elastic
lamina is intact despite the intense
inflammatory cell infiltration (Verhoeff's
elastic stain, x 250). (B) Collections of
neutrophils form microabscesses in the media
(upper arrows) and within the cellular
thrombus (lower arrow). Polymorphonuclear
leukocytes are present within all layers of the
vessel. There is no calcification or lipid
deposition present (hematoxylin and eosin
stain, x 500).

cytic infiltration but the internal elastic lamina remains intact.
The media is less infiltrated by lymphocytes and occasional
macrophages but there are no necrotizing lesions or giant cells
and no thinning of this layer. Extensive fibroblastic prolifera-
tion and foci of lymphocytes are seen within the adventitial
layer. This inflammatory process can extend into the sur-
rounding tissue and incorporate the associated vein and
nerve.

The granulomatous reaction with microabscesses and giant
cells within the thrombus is characteristic of the acute stage of
Buerger's. It is not seen in atherosclerotic occlusions or arterial
thrombosis. Unfortunately, few clinical specimens are sec-
tioned in a sufficient number of locations to find the character-
istic changes.

Chronic stage

In the chronic stage, the occluding thrombus is well organized
and recanalization may have taken place. There is fibrinous
thickening of the vessel wall, particularly of the intima and
media, with variable degrees of nonspecific inflammatory
changes, although the general architecture and integrity of the
elastic laminae is preserved. 22 ' 31 These are the lesions most
often seen in pathologic specimens. By the time an amputation
is performed, the acute lesion usually has resolved. Further-
more, the acute lesions, even if present, may be missed by ran-
dom pathologic sectioning because of the segmental character
of the disease. 29 ' 30

Unlike arterial atherosclerosis, there are no lipid or calcium
deposits in the intima or media. Atherosclerotic arterial occlu-
sions usually contain a relatively acellular thrombus without a

prominent inflammatory reaction of the media or adventitia. 22
There is calcification and characteristic plaque formation with
the elastic laminae being fragmented. Unlike the vasculi tides,
in Buerger 's there is no fibrinoid necrosis of the vessel wall and
the internal elastic lamina is intact. Further differentiation
from giant cell arteritis, Takayasu's arteritis, or polyarteritis
can be made on the basis of the anatomical distribution of
the lesions, the size of the involved vessels, the absence of
medial degeneration, aneurysms, and the lack of intimal
proliferation. 32

Inflammatory reactions similar to the chronic stage of
Buerger's disease may be seen in some patients with athero-
sclerotic arterial thrombosis or embolization. 4 Although in-
consistent with the histologic changes found in the acute stage,
these observations serve to emphasize the fact that a possible
pathohistologic diagnosis of Buerger's disease must be corre-
lated with the patient's clinical presentation.

Veins

Veins adjacent to the involved arteries demonstrate similar
changes although giant cells and lymphocytes are seen more
frequently 32 The media has infiltration with lymphocytes and
fibroblasts while the adventitia has an extensive fibroblastic
proliferation.

Superficial phlebitis in patients with Buerger's disease
is histologically comparable to idiopathic venous thrombo-
phlebitis. There is intense lymphocytic and fibroblastic in-
filtration of all layers of the vessel wall and a cellular occluding
thrombus.

chapter 8 Buerger's disease

Clinical presentation

Certain demographic characteristics of patients with distal ex-
tremity arterial occlusions should immediately raise suspi-
cion of a possible diagnosis of Buerger's disease. In the past,
female gender precluded this diagnosis because it was felt that
only males could have TAO. More recently, females have been
recognized to be affected as well. In reviewing 17 series, Papa
and Adar 5 found that 73 of 1735 patients (4.2%) with TAO
reported in the literature were women. This proportion, how-
ever, varied from 1% to 23% in individual reports, reflecting
both the heterogeneity of the patient populations and the dif-
ferences in diagnostic criteria utilized. 33 The increasing num-
bers of women diagnosed with Buerger 's may be secondary to
the increased prevalence of smoking. Escalating with the en-
trance of women into the work force during World War II, the
prevalence of smoking in females reached 34.1% in 1965 and
thereafter has declined to 21.0% by 2000. 9 Assuming a long la-
tency period in the onset of Buerger's disease, the increasing
diagnosis among women appears to parallel the increased use
of cigarettes in this segment of the population. It may also re-
flect the slow acceptance by physicians of TAO into the differ-
ential diagnosis in young women with extremity ischemia.
Although still atypical, female sex should not exclude consid-
eration of Buerger's disease. 34

Symptoms of TAO first appear during the third or fourth
decade of life. If a careful history is taken, only rarely will onset
of symptoms occur after 40 years of age, even if the patient
comes to the physician for treatment later in life. Onset after
the age of 50 should effectively exclude the diagnosis. 6 ' 35 The
mean age at time of diagnosis is 34 years. 5

Historically, smoking has been the sine qua non of this
disease. Evolution in our appreciation of the pervasiveness of
cigarette smoke in our environment, and the phenomenon
of passive smoking, has lead to the acceptance that "non-
smokers" may suffer from this disease. 12 Approximately 5% of
patients felt to have Buerger's disease in recent series were
not smokers. 5 Lack of use of tobacco products is therefore not
an absolute exclusionary criterion.

The earliest symptom of TAO, as with atherosclerosis, is in-
termittent claudication. However, TAO commonly starts in
the arch of the foot and only later involves the calf. This is a re-
flection of the initiation of the occlusive process in the most
distal portion of the extremity with later proximal extension.
Foot or instep claudication is caused by ischemia of the plantar
muscles during ambulation after the posterior tibial or plantar
arteries have become occluded. 35 Hirai and Shionoya 36 believe
that initial foot claudication is virtually pathognomonic for
Buerger's disease. With involvement of the crural or popliteal
arteries, calf claudication may be noted.

Clinical deterioration proceeds rapidly if smoking con-
tinues. Rest pain occurs early and is usually a severe ache or
burning, associated with numbness of the involved digits,

persisting without amelioration. Part of this severe pain may
be ischemic neuritis from encasement of the nerve in the in-
flammatory process. 37 Ulceration and gangrene of the digit
soon follows. On physical examination, the affected fingers or
toes have a purplish-red color that is characteristic of the dis-
ease and has been called "Buerger's color." 38 The fingers or
toes are cold and damp to the touch. Ischemic neuropathy may
cause marked sympathetic overactivity and may be misdiag-
nosed as Raynaud's phenomenon although a true vasospastic
component is not generally seen. The digital arteries are
usually chronically dilated and do not respond to cold or emo-
tional stimuli. Gangrene and ulceration may follow local trau-
ma and develop most commonly at the tip of the digit. It is rare
that only one extremity is involved although the presenting
symptoms may be limited to just one digit or limb. More than
three-quarters of all patients have involvement of three or four
limbs if they are all evaluated angiographically 6/7/29/33

Most patients with Buerger's disease have loss of pedal
pulses in one or both of the legs. The popliteal and femoral
pulses are generally preserved until late in the course of
the disease. Lack of a palpable popliteal pulse at initial
presentation should prompt one to question seriously a pre-
sumptive diagnosis of TAO. There is frequently asymmetry
between the paired extremities and marked differences in
appearance and temperature between digits may be evident.
Because the initial lesions are found below the ankle, a
Doppler ankle-brachial pressure index is not a good indicator
of the severity of ischemia early on. Toe or finger systolic pres-
sure measurements and pulse volume recordings are more
useful.

If even asymptomatic patients are evaluated by angiogra-
phy, more than half demonstrate upper extremity involve-
ment which is one of the characteristics of this disease. 26 ' 29,37
The ulnar pulse is usually the first to be lost. 39 Allen 40 first de-
scribed the clinical examination for the patency of the palmar
arch in patients with TAO. The process in the upper limb is
analogous to that seen in the leg.

The other systemic manifestation of Buerger's disease is
migratory phlebitis. It is a focal phlebitis of the small veins
of the foot or ankle occurring in approximately 45% of
patients. 35 It is less commonly seen in the leg or arm and does
not usually involve the greater or lesser saphenous veins.
Idiopathic deep venous thrombosis has no relationship to
Buerger's disease.

The clinical diagnosis of Buerger's disease is not precisely
defined and the literature is full of reports with a wide variety
of criteria utilized with most of them poorly described. 5 The
criteria of Shionoya 35 are most frequently cited and include all
of the following: (i) smoking history; (ii) onset of symptoms
before the age of 50 years; (iii) infrapopliteal arterial occlusive
disease; (iv) either upper limb involvement or migratory
phlebitis; and (v) exclusion of other diseases and demon-
strated by the absence of atherosclerotic risk factors other
than smoking at the time of presentation. A point system has

pa rt I Vascular pathology and physiology

Table 8.1 Clinical differential diagnoses for Buerger's disease

Table 8.2 Comparison of the clinical features of thromboangiitis obliterans
(TAO) and arteriosclerosis obliterans (ASO)

Atherosclerotic disease
In situ thrombosis
Emboli

—Arterial origin

— Cardiacorigin
Risk factors

— Hyperlipidemia

— Hypertension

— Diabetes mellitus

Upper extremity
Innominate artery stenosis
Subclavian stenosis/aneurysm
Thoracic outlet syndrome
Occupational injury

Popliteal artery lesions
Entrapment

Adventitial cystic degeneration
Aneurysm

Embolus or thrombosis in situ
Trauma

Autoimmune disease
Scleroderma

Systemic lupus erythematosus
Rheumatoid arthritis
Mixed connective tissue disease

Vasculitides
Polyarteritis
Giant cell arteritis
Takayasu's arteritis
Hypersensitivity angiitis

Hematologic disorders
Polycythemia
Thrombocytosis
Dysproteinemias
1 ° and 2° hypercoagulable states

TAO

ASO

been advocated by Papa et al., A1 while Mills and Porter suggest
categorization into major and minor criteria. 42 The clinical
differential diagnosis must exclude a variety of lesions that
may mimic TAO (Table 8.1). The most frequent diagnosis to ex-
clude is that of atherosclerotic disease in a young patient
(Table 8.2). The more classic the findings (male sex, age less
than 35 years, heavy smoking history, foot claudication), the
more likely the clinical diagnosis. The essence of Buerger 's dis-
ease is peripheral ischemia of an inflammatory nature and
with a self-limiting course upon the cessation of smoking or
smoke exposure. The diagnosis is dependent upon a clinical
evaluation with corroborative angiographic or histopatholog-
ic studies.

There are no specific laboratory tests to assist in the diagno-
sis of Buerger's disease. Testing should be performed to ex-
clude other connective tissue disorders or hypercoagulable
states. Laboratory tests should therefore include a complete
blood count with a differential count, blood urea nitrogen and
creatinine levels, fasting blood glucose, liver-function tests,
urinalysis, erythrocyte sedimentation rate, C-reactive protein,
complement levels, rheumatoid factor, antinuclear antibody,
markers for the CREST syndrome and scleroderma (anticen-
tromere antibody and Scl-70), homocysteine, and screening
for hypercoagulable states: prothrombin time, activated par-
tial thromboplastin time, protein-C, protein-S, antithrombin
III, factor V Leiden, prothrombin abnormalities, antiphospho-
lipid antibodies. A proximal source of emboli can be ruled out
by cardiac echocardiography and arteriography.

Clinical features
Age of onset, years
Sex- male, %
Smoking, %

Migratory thrombophlebitis, %
Raynaud's phenomenon, %
Upper extremity involvement, %
Foot claudication
Multilimb involvement (3 or more)
Diabetes, %

Angiography
Aorta/iliac/femoral arteries
Stenoses/plaques
"Tree root" collaterals
"Corkscrew" collaterals

Pathology
Atheromatous plaque
Lipid/calcium deposits
Microabscesses
Diffuse inflammation
Vein/nerve involvement

Present

Absent

76%

Rare

Normal

Diseased

Absent

Present

Common

Occasional

Common

Rare

Absent

Present

Absent

Present

Absent

Present

Absent

Present

Absent

Modified from Lie JT The rise and fall and resurgence of thromboangiitis
obliterans (Buerger's Disease). Acta Pathol Jpn 1 989; 39: 1 53.

Angiography

Careful angiography is indicated in all patients with a pre-
sumed diagnosis of Buerger's disease. Although the arterio-
graphic findings are not pathognomonic and must be
interpreted in the context of the clinical presentation, angio-
graphy is useful for several reasons. First, it can exclude other
possible diagnoses being considered: large and medium ves-
sel disease, proximal atherosclerotic plaques and stenoses,
and aneurysms. Second, characteristic changes of Buerger's
disease, if found, would provide confirmatory evidence for
the clinical diagnosis. Finally, it would identify the specific site
and extent of the occlusions and, in a minority of patients, pro-
vide the anatomic road map for surgical intervention.

Characteristic changes found in Buerger's disease include
normal appearing aorta and iliac vessels with even caliber and
without any luminal irregularity. Most frequently, the profun-
da and superficial femoral arteries are uninvolved. Later in the
disease process, 10% of patients will display angiographic
evidence of disease in the aorta and iliac segments. 6 A "corru-
gated" or standing-wave pattern of the contrast within the
larger vessels was first thought by Szilagi and colleagues 43 to
be diagnostic of Buerger's disease. Subsequent experience has
shown this to be a nonspecific finding. 44 Multiple distal seg-

chapter 8 Buerger's disease

Figure 8.2 Collateral vessels in a "tree root" or "spider leg" configuration
are seen beyond the occlusion of the radial (curved arrow) and ulnar (straight
arrow) arteries.

Figure 8.3 "Corkscrew" appearance of vessel is seen following the path of
the posterior tibial artery with all the other vessels occluded.

mental occlusions are seen, usually beyond the popliteal and
brachial arteries. The collateral vessels around the area of oc-
clusion have a "tree root" or "spider leg" configuration
(Fig. 8.2). The more distal vessels have an abnormal tortuosity
giving it a "corkscrew" appearance, probably reflecting either
recanalization of the occluded vessel or the prominent disten-
sion of the vasa vasorum (Fig. 8.3). 30,43,45/46 The distal portions
of the tibial vessels and the ulnar artery above the wrist are
often occluded at the time of presentation. The peroneal and
interosseous arteries are the least likely of the distal vessels to
be involved. 39 Analogous to the pathologic descriptions, the
diagnosis of Buerger 's disease cannot be made on angiograph-
ic findings alone, no matter how characteristic.

Treatment

Medical

Cessation of smoking and the use of all tobacco products is the
mainstay of treatment for Buerger's disease. Patients must ab-

solutely avoid cigarettes, cigars, pipes, snuff, and chewing to-
bacco. Those who stop smoking show improvement and will
usually have no further progression of their disease although
claudication will persist. 26 ' 33 With resumption of smoking,
there is an invariable clinical recurrence and more extensive
limb loss becomes unavoidable. In a series of 2468 patients,
Barlas et alF found that of the patients who stopped smoking,
only 6% had subsequent amputations whereas 46% of patients
who continued to smoke required further amputations. Even
with repeated admonitions, and in the face of limb loss, some
patients cannot control their addiction and continue to smoke.
This may reflect denial of the severity of their illness and a ten-
dency toward self-destructive behavior in these patients. 48
The recent availability of nicotine patches, in conjunction with
group support therapy, may be helpful in assisting more of
these patients to stop smoking. Some have suggested that
nicotine replacement therapy may continue the process
although this has not yet been proven. 49 Local measures to
protect the involved extremity should be undertaken. Patients
should avoid trauma to their digits, extreme care should be
exercised in trimming nails, and emollients used to soften the

pa rt I Vascular pathology and physiology

skin and avoid drying and cracking. Infection should be
aggressively treated with wide-spectrum intravenous
antibiotics.

A wide variety of medications have been utilized in the past
to ameliorate the injury already sustained and to prevent pro-
gression of the disease. These have included anticoagulants,
steroids, vasodilators, prostaglandins, oc-adrenergic blockers,
antiplatelet agents, and hemorrheologic agents, but none of
these has been found to have any beneficial durable effects. A
prospective double-blind trial with the prostacyclin analogue,
iloprost, found that 18 of 52 (35%) patients treated intra-
venously had ulcer healing compared with only six of 46 (13%)
treated with aspirin. 50 Although there was only a short 6-
month follow-up period, these differences appeared to persist.
However, oral iloprost demonstrated no difference in healing
of ischemic lesions in 319 patients. 51 Calcium channel blockers
may be tried in patients with vasospastic episodes.

Surgical

Surgical intervention does nothing to address the underlying
pathology. It therefore is, at best, a palliative temporizing mea-
sure. Recurrence is assured unless the patient refrains from to-
bacco. Direct arterial reconstruction is usually not feasible
because of the distal extent of involvement or the presence of
multiple occluded vessels. The rare cases of suprapopliteal or
popliteal occlusions are the best candidates for bypass. Auto-
logous vein is the material of choice in such circumstances be-
cause of its better patency below the knee and in situations
with poor distal runoff. Only 15% of 266 patients had vascular
reconstructive procedures performed in Nagoya, with an
overall patency rate of only 24%. 52 Sasajima etal., 53 in a series of
71 bypasses in 61 patients, reported a 5-year secondary paten-
cy rate of 63%. However, this rate was only 35% for those
patients who continued to smoke. Nonetheless, bypasses are
useful if they remain patent long enough for ulcerations or
digit amputations to heal. If the bypass later occludes, ulcera-
tion recurrence will generally not occur if the patient refrains
from smoking.

Spinal cord stimulation occasionally has been found useful
in anecdotal case reports. 54 In earlier series, surgical sympa-
thectomy has been used in more than half the patients in Japan
with Buerger's disease. 6 This is approximately twice the fre-
quency reported in the United States. 33 Nakata et al. 55 found
overall good results in only 34% of patients but most of the
improvement took place in patients without ulcers or
gangrene and in those who subsequently stopped smoking. Its
primary use is in patients with limited superficial skin lesions
or patients with severe vasospastic symptoms. Because vaso-
motor tone returns to baseline within a few weeks to months
after sympathectomy, the temporary increase in blood flow
would need to be sufficient to heal superficial lesions during
this period of time. 56 It is of minimal benefit for intermittent
claudication or in the presence of deep gangrene because

muscle blood flow is not increased by sympathectomy. Few
patients in the United States fit these clinical criteria and
more recently the number of sympathectomies performed in
Japan has also decreased. 57 There has been no prospective
controlled trial that has documented a significant long-term
benefit of sympathectomy in these patients. As in non-
Buerger 's patients, we would first advocate a therapeutic trial
with chemical sympathectomy before surgical intervention
and use this as a last resort in an attempt to heal superficial
ischemic ulcers and prevent the requirement for a proximal
amputation.

More recently, gene therapy has been used in six patients
with Buerger's disease. Isner and colleagues 58 reported on the
use of naked plasmid DNA encoding for vascular endothelial
growth factor intramuscularly to stimulate angiogenesis for
critical limb ischemia. Improved blood flow in the limbs was
shown by magnetic resonance angiography and an increase in
the ankle brachial index in half the patients. These preliminary
anecdotal results, however, will need to be replicated and con-
firmed in prospective clinical trials before such innovative
therapy can be recommended.

Ulceration and gangrene should be treated conservatively
for as long as possible. The hope is for healing of ulcerations if
the patient stops smoking and eventual autoamputation with
maximal tissue preservation if gangrene is already present.
Surgical debridement should be avoided as long as possible.
Unlike the Japanese experience where they had only a 3.6%
major amputation rate in 193 patients, 6 Olin et al. 33 found 16
of 89 patients (18%) to have required either a below- or
above-knee amputation, with 26% requiring digital or
transmetatarsal amputation. Others have found a higher 36%
major amputation rate and combined major and minor rates
of up to 75%. 26/59 Although the outlook for limb salvage is grim
in patients who continue to smoke, their life expectancy is
not adversely affected. They have no increased mortality
compared with an age-adjusted normal population. Their 5-
year survival rate of 98% is much better than the 76% found in
similar patients with atherosclerotic disease. 60 In a contempo-
raneous series of 328 patients with Buerger's disease and 515
with atherosclerotic disease, the mortality rate was 2.8% and
37.8%, respectively, at a mean follow-up period of almost
5 years. 61

Conclusion

Buerger's disease was described almost 100 years ago. We still
do not know either the etiology or the underlying pathophysi-
ology of this fascinating disorder. There is no effective medical
or surgical therapy. Abstinence from tobacco, repeatedly
shown to stop the progression of the disease, appears to be dif-
ficult for most patients.

chapter 8 Buerger's disease

References

1. Von Winiwarter F. Ueber eine eigentumliche Form von Endarteri-
tis und Endophlebitis mit Gangran des Fusses. Arch Klin Chir 1879;
23:202.

2. Buerger L. Thrombo-angiitis obliterans: a study of the vascular
lesions leading to presenile spontaneous gangrene. Am J Med
Sci 1908; 136:567.

3. Fisher CM. Cerebral thromboangiitis obliterans (including a criti-
cal review of the literature). Medicine 1957; 36:169.

4. Wessler S, Ming SC, Gurewich V et ah A critical evaluation of
thromboangiitis obliterans. N Engl J Med 1960; 262:1149.

5. Papa MZ, Adar R. A critical look at thromboangiitis obliterans
(Buerger's disease). Perspect Vase Surg 1992; 5:1.

6. Shionoya S. Buerger's disease (thromboangiitis obliterans). In:
Rutherford RB, ed. Vascular Surgery, 4th edn. Philadelphia: WB
Saunders, 1995:235.

7. Lie JT. The rise and fall and resurgence of thromboangiitis obliter-
ans (Buerger's disease) Acta Pathol Jpn 1989; 39:153.

8. US Department of Health, Education, and Welfare. Smoking and
health: A report of the Surgeon General. DHEW (PHS) Pub. No.
79-50066. Washington, DC: Govt Printing Office, 1979: A-9.

9. Giovino G. Epidemiology of tobacco use in the United States.
Oncogene 2002; 21:7326.

10. Honjo K, Kawachi I. Effects of market liberalisation on smoking in
Japan. Tobacco Control 2000; 9:193.

11. Matsushita M, Nishikimi N, Sakurai T, Nimura Y. Decrease
in prevalence of Buerger's disease in Japan. Surgery 1998; 124:
498.

12. Kjeldsen K, Mozes M. Buerger's disease in Israel: investigations
on carboxy-hemoglobin and serum cholesterol levels after smok-
ing. Acta Chir Scand 1969; 135:495.

13. Matsushita M, Shionoya S, Matsumoto T. Urinary cotinine
measurement in patients with Buerger 's disease — effects of active
and passive smoking on the disease process. / Vase Surg 1991;
14:53.

14. Becker CG, Dublin T, Wiedman H. Hypersensitivity to tobacco
antigen. Proc Natl Acad Sci USA 1976; 73:1712.

15. Papa M, Bass A, Adar R et ah Autoimmune mechanisms in
thromboangiitis obliterans (Buerger's disease): the role of tobacco
antigen and the major histocompatibility complex. Surgery 1992;
111:527.

16. Gulati SM, Singh KS, Thusoo TK et ah Immunological studies in
thromboangiitis obliterans. / Surg Res 1979; 27:287.

17. Gulati SM, Saha K, Kant L et ah Significance of circulatory immune
complexes in thromboangiitis obliterans. Angiology 1984; 35:
276.

18. deAlbuquerque RR, Delgado L, Correia P et ah Circulating im-
mune complexes in Buerger's disease — endarteritis obliterans in
young men. / Cardiovasc Surg 1989; 30:821.

19. Smolen JS, Youngchaiyud U, Weidinger P et ah Autoimmunologi-
cal aspects of thromboangiitis obliterans. Clin Immunol Im-
munopathol 1978; 11:168.

20. Adar R, Papa MZ, Halpern Z et ah Cellular sensitivity to collagen
in thromboangiitis obliterans. N Engl J Med 1983; 308:1113.

21. Eichhorn J, Sima D, Lindschau C et ah Antiendothelial cell
antibodies in thromboangiitis obliterans. Am J Med Sci 1998;
315:17.

22. Kobayashi M, Ito M, Nakagawa A, Nishikimi N, Nimura Y. Im-
munohistochemical analysis of arterial wall cellular infiltration in
Buerger's disease (endarteritis obliterans). / Vase Surg 1999;
29:451.

23. McLaughlin GA, Helsby CR, Evans CC et ah Association of HLA-
A9 and HLA-B5 with Buerger's disease. Br Med J 1976; 2:1165.

24. Numano F, Sasazuki T, Koyama T et ah HLA in Buerger's disease.
Exp Clin Immunogenet 1986; 3:195.

25. Ohtawa T, Juji T, Kawano N et ah HLA antigen in thromboangiitis
obliterans. JAMA 1974; 230:1126.

26. Mills J, Taylor LM, Porter JM. Buerger's disease in the modern era.
Am J Surg 1987; 154:123.

27. Rosen N, Sommer I, Knode B. Intestinal Buerger's disease. Arch
Pathol Lab Med 1985; 109:962.

28. Kempczinski RF, Clark SM, Blebea J, Koeliker D, Fenoglio-Preiser
CM. Intestinal ischemia secondary to thromboangiitis obliterans:
case report and review of the literature. Ann Vase Surg 1993; 7:
354.

29. Shionoya S. What is Buerger's disease? World J Surg 1983; 7:544.

30. McKusick VA, Harris WS, Ottesen OE, Goodman RM, Shelley
WM, Bloodwell RD. Buerger's disease: a distinct clinical and
pathologic entity. JAMA 1962; 181:5.

31. Tanaka K. Pathology and pathogenesis of Buerger's disease. Int J
Cardiol 1998; 66:S237.

32. Hollier L. Thromboangiitis obliterans. In: Kempczinski RF, ed. The
Ischemic Leg. Chicago: Year Book, 1985:71.

33. Olin JW, Young JR, Graor RA, Ruschhaupt WF, Bartholomew JR.
The changing clinical spectrum of thromboangiitis obliterans
(Buerger's disease). Circulation 1990; 82 (Suppl. IV):IV3.

34. Lie JT. Thromboangiitis obliterans (Buerger's disease) in women.
Medicine 1986; 65:65.

35. Shionoya S. Diagnostic criteria of Buerger's disease. Int J Cardiol
1998;66:S243.

36. Hirai M, Shionoya S. Intermittent claudication in the foot and
Buerger's disease. Br J Surg 1978; 65:210.

37. Mishima Y Thromboangiitis obliterans (Buerger's disease): The
Japanese experience. In: HaimoviciH, ed. Vascular Surgery— Prin-
ciples and Techniques, 3rd edn. Norwalk, CT: Appleton & Lange,
1989:441.

38. Kimura T, Yoshizaki S, Tsushima N et ah Buerger 's color. Br J Surg
1990; 77:1299.

39. Sasaki S, Sakuma M, Kunihara T, Yasuda K. Distribution of
arterial involvement in thromboangiitis obliterans (Buerger's
disease): results of a study conducted by the Intractable Vasculitis
Syndromes Research Group in Japan. Surg Today 2000; 30:600.

40. Allen EV Thromboangiitis obliterans. Methods of diagnosis of
chronic occlusive arterial lesions distal to the wrist with illustra-
tive cases. Am J Med Sci 1929; 178:237.

41. Papa MZ, Rabi I, Adar R. A point scoring system for the clinical di-
agnosis of Buerger 's disease. Eur J Vase Endovasc Surg 1996; 11:335.

42. Mills JL, Porter JM. Buerger's disease: a review and update. Semin
Vase Surg 1993; 6:14.

43. Szilagyi DE, DeRusso FJ, Elliott JP Thromboangiitis obliterans:
clinico-pathologic correlations. Arch Surg 1964; 88:824.

44. Hagen B, Lohse S. Clinical and radiologic aspects of Buerger 's dis-
ease. Cardiovasc Intervent Radiol 1984; 7:283.

45. Suzuki A, Mine H, Yoshida T, Okada Y Buerger's disease (throm-
boangiitis obliterans): an analysis of the arteriograms of 119 cases.
Clin Radiol 1982; 33:235.

pa rt I Vascular pathology and physiology

46. Suzuki S, Yamada I, Himeno Y. Angiographic findings in Buerger
disease. Int J Cardiol 1996; 54:S189.

47. Barlas S, Elmaci T, Dayioglu et ah Has the clinical definition of
thromboangiitis obliterans changed indeed? Int } Angiol 1997;
6:49.

48. Farberow NL, Nehemkis AM. Indirect self-destructive behavior
in patients with Buerger's disease. / Personal Assess 1979; 43:86.

49. Olin JW. Current concepts: thromboangiitis obliterans (Buerger's
disease). N Engl} Med 2000; 343:864.

50. Fiessinger JN, Schafer M. Trial of iloprost versus aspirin treatment
for critical limb ischaemia of thromboangiitis obliterans. The TAO
Study. Lancet 1990; 335:555.

51. The European TAO Study Group. Oral iloprost in the treatment of
thromboangiitis obliterans (Buerger's disease): a double-blind,
randomised, placebo-controlled trial. Eur } Vase Endovasc Surg
1998; 15:300.

52. Shionoya S, Ban I, Nakata Y et ah Surgical treatment of Buerger's
disease. J Cardiovasc Surg 1980; 21:77.

53. Sasajima T, Kubo Y, Inaba M, Goh K, Azuma N. Role of infrain-
guinal bypass in Buerger's disease: an eighteen-year experience.
Eur J Vase Endovasc Surg 1997; 13:186.

54. Swigris JJ, Olin JJ, Mekhail NA. Implantable spinal cord stimula-
tor to treat the ischemic manifestations of thromboangiitis obliter-
ans (Buerger's disease). / Vase Surg 1999; 29:28.

55. Nakata Y, Suzuki S, Kawai S et ah Effects of lumbar sympathecto-
my on thromboangiitis obliterans. / Cardiovasc Surg 1975; 16:415.

56. Sayin A, Bozkurt AK, Tuzun H et ah Surgical treatment of
Buerger's disease: experience with 216 patients. Cardiovasc Surg
1993; 1:377.

57. Nakajima N. The change in concept and surgical treatment on
Buerger's disease— personal experience and review. Int } Cardiol
1998;66:S273.

58. Isner JM, Baumgartner I, Rauh G et ah Treatment of thromboangi-
itis obliterans (Buerger's disease) by intramuscular gene transfer
of vascular endothelial growth factor: preliminary clinical results.
} Vase Surg 1998; 28:964.

59. Pairolero PC, Joyce JW, Skinner CR et ah Lower limb ischemia in
young adults: prognostic implications. / Vase Surg 1984; 1:459.

60. McPherson JR, Juergens JL, Gifford RW. Thromboangiitis obliter-
ans and arteriosclerotic obliterans. Ann Int Med 1963; 59:288.

61. Ohta T, Shionoya S. Fate of the ischaemic limb in Buerger's
disease. Br } Surg 1988; 75:259.

100

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Ergotism

Roger F.J. Shepherd

Although ergotism occurs less frequently today because of de-
clining use of preparations such as Cafergot (Sandoz Pharma-
ceuticals, East Hanover, NJ) for the treatment of migraine
headaches, it remains an important cause of widespread arter-
ial disease that often is misdiagnosed. As a result of inade-
quate or inappropriate therapy, potential consequences of
limb and organ ischemia can be severe.

Since the early part of this century, ergotamine tartrate has
been a drug of choice for the treatment of acute attacks of mi-
graine headaches. 1 This is a relatively common disease affect-
ing nearly 10 million people in the United States and causing
significant disability, with over 3 million days per month spent
bedridden. 2 Ergotamine derivatives such as ergotamine tar-
trate owe their effectiveness in the treatment of migraine in
part to the fact that they contract cerebral vascular smooth
muscle.

Unfortunately, the action of ergot compounds is not
limited to the cranial vessels, and multiple case reports have
demonstrated the potential for widespread arterial vasocon-
striction with resulting extremity and organ ischemia. Iatro-
genic ergotism is caused most commonly by excessive dosage
of ergotamine tartrate administered as a rectal suppository
but also can occur in patients receiving therapeutic oral
dosages. 3

Ergotism is the great masquerader, simulating other disease
entities such as atherosclerotic occlusive disease, 4 throm-
boembolic disease, 5 arteritis, 6 including Takayasu and giant
cell arteritis, fibromuscular dysplasia, 7 and vasospastic dis-
ease such as the Raynaud phenomenon. 8 Most commonly,
ergotism involves the extremities, although multiple organ
systems may be involved. For example, gastrointestinal
ischemia may result from ergot ingestion leading to intestinal
angina and possible bowel infarction. 9 Renal artery constric-
tion may cause hypertension and renal failure. Coronary
artery vasospasm can cause myocardial ischemia with resul-
tant angina pectoris, myocardial infarction, cardiac arrest, and
sudden death. 10 Intense vasospasm of the leg and arm vessels
has resulted in critical ischemia with gangrene in some cases,
necessitating amputation. 11

This chapter presents the great variety of clinical manifesta-
tions of ergot toxicity, differential diagnosis, angiographic fea-
tures, pharmacology and mechanism of action of the ergot
derivatives, and, finally, management options for ergotism.

History

Ergot occurs naturally as a product of a fungus, Claviceps
purpura, which is indigenous to both North America and
Europe. The contamination of edible grain by this parasitic
fungus has been known for centuries, and, around 600 bc, an
Assyrian tablet depicted "the noxious pustule in the ear of
grain/' 12 Rye is the most susceptible grain, and becomes in-
fected by fungal spores that germinate into hyphal filaments,
causing a reaction with a dense tissue formation that becomes
a purple, curved body called the sclerotium 13 (Fig. 9.1). The
sclerotium continues to be a major pharmaceutical source of
ergot alkaloids. 12

Ergotism was unknown to the early Romans because rye
was not a staple grain. The ancient Greeks recognized the toxic
effects of spoiled rye, and also did not consume this grain. Rye
was consumed by the conquering Teutons, who are credited
with the spread of ergotism throughout Western Europe, re-
sulting in many large epidemics of ergotism that occurred
from the 9th to the 19th century in Scandinavia, Bohemia, and
Russia. 13 A description of gangrenous ergotism from the 9th
century ad comes from George Barger's 1931 book, Ergot and
Ergotism: "A great plague of swollen blisters consumed the
people by a loathsome rot so that their limbs were loosened
and fell off before death." 14/15 Further written accounts of
ergotism during the Middle Ages portrayed strange epi-
demics of agonizing, burning pain of the extremities in which
the skin turned black; in severe cases, mummification with
spontaneous amputation of the limbs without blood loss was
described.

The intense burning pain became known as "The Holy Fire"
or "St. Anthony's Fire" after a monastery where sufferers
were believed to have obtained relief. St. Anthony of Egypt,

101

pa rt I Vascular pathology and physiology

Figure 9.1 Early 1 9th century drawing of three ears of rye (left) containing
ergot sclerotia, and wheat grain (right), which rarely is infected by Claviceps
purpura. (Reprinted from Prescott Jr. A Dissertation on the Natural History
and Medicinal Effects of Secale Cornutam, or ergot. Boston, Cummingsand
Hilliard, 1813. In: Tanner JR. St. Anthony's fire, then and now: a case report
and historical review. Can J Surg 1987; 30:292, with permission from the
Canadian Medical Association.)

patriarch of monks and healer of both men and animals, lived
between approximately 251 and 356 ad, and adopted the
monastic life at 20 years of age, giving away all his worldly be-
longings. 13 He was reputed to be a miracle worker, and had
many followers. When he died, he was buried in Egypt in an
unknown place. Many centuries later, his relics were claimed
by the city of LaMotte in southeast France, where a shrine
was erected. About 1100 ad, the Order of Hospitallers of St. An-
thony was founded, devoted to the care of those with epidemic
gangrene. 13 The houses of the Order were said to have flame-
red walls. Relief from the intense burning pain of ergotism was
said to occur by visiting the shrine of St. Anthony. This relief
probably was real, because the long sojourn to the shrine at
LaMotte may have taken pilgrims out of the area of infected
rye, and thus their ischemic symptoms would have improved
on an ergot-free diet. It was not until 1676 that the cause of epi-
demic ergotism was linked to foods such as bread and cereal
made from rye contaminated with this poisonous fungus.

Outbreaks of ergot poisoning continued to occur in the 20th
century, and cases were reported in Russia in 1926 and Ireland
in 1929 that resulted from improper storage and processing of
grain. In 1953, an outbreak occurred in France when a baker
tried to circumvent a grain tax using contaminated, bootleg
flour. 11 Occasional outbreaks of ergotism continue to occur in
underdeveloped countries, as reported in Ethiopia in 1977. 16

There have been no recent large outbreaks in North America
because government inspection procedures reject rye if it con-
tains more than 0.3% infected grain. Annual rejection rates
vary from less than 1% to as high as 36% in wet seasons. 12

Obstetrics and ergot

In obstetrics, the adverse effects of ergot have been noted for
more than 2000 years. The sacred book of Parsees (about 350
bc) contains the following quote: " Among the evil things cre-
ated by Angro Maynes are noxious grasses that cause preg-
nant woman to drop the womb and die in child bed/' 12 The
earliest Western description of the medical use of ergot was in
obstetrics in 1582 by Lonicer, who noted the resultant painful
effect of ergot on the uterus. 11 ' 12 Its official introduction into
medicine is attributed to John Stearns in 1808, when he pub-
lished a letter entitled "Account of the Pulvis Parturiens, a
Remedy for Quickening Childbirth/' By 1822, it was recog-
nized that the use of this drug greatly increased the risk for
stillborn children, and an inquiry into its misuse was conduct-
ed by the Medical Society of New York. Hosack in 1824 coined
the term pulvis ad mortem, and recommended that the drug be
used only to control postpartum hemorrhage. 13 In 1820, ergot
was listed in the United States Pharmacopoeia for obstetric
use. 14 The ergot derivative ergonovine continues to be a useful
agent for controlling postpartum or postabortion bleeding by
its sustained contraction of uterine smooth muscle. These
same actions, however, preclude its use for induction or facili-
tation of labor.

Pharmacology of ergot alkaloids

The actions of ergot alkaloids are complex and not completely
understood. Initially, it was thought that these substances
acted directly on the vascular smooth muscle to cause its con-
traction, presumably by increasing the influx of calcium ions.
There is evidence, however, that they also act as partial ago-
nists on smooth muscle receptors. These receptors include the
oc-adrenoreceptors and serotonergic receptors, and their acti-
vation also causes vasoconstriction (Fig. 9.2). The binding to
these receptors may be irreversible, and hence serotonergic
and a-receptor antagonists may be without effect in inhibiting
constriction. This accords with clinical observations that the
directly acting relaxing agent nitroprusside is most effective in
opposing the vasoconstriction. In addition, small amounts of
the ergot alkaloids increase the vasoconstrictor action of nor-
epinephrine, serotonin, and angiotensin. 12 ' 17 ' 18

Ergot preparations

In 1906, Ergotoxine was first isolated from ergot by Barger and
colleagues. 19 In 1920, Stol isolated the first pure ergot alkaloid,

102

CHAPTER 9 Ergotism

SYMPATHETIC NERVE

Enhance the
vasoconstrictor action

Partial agonist

ARTERY

Direct action
increasing Ca 2+ entry

Endothelium

Smooth muscle

Partial agonist

NE = Norepinephrine
a1 = Alpha r adrenoceptor
a 2 = Alpharadrenoceptor
S — Serotoninergic receptors

Figure 9.2 Complex actions of ergot alkaloids. These act directly on the
vascular smooth muscle and cause contraction by increasing the entry of
calcium ions. In addition, they may act as partial agonists or a 2 - and a 2 -
adrenoreceptors and serotonergic receptors. In this way, the vasoconstrictive
action of norepinephrine released from the sympathetic nerves is enhanced.

ergotamine, and its salt, ergotamine tartrate. Since then there
have been numerous semisynthetic derivatives of ergot
alkaloids.

Available ergot preparations include the following:

• Ergotamine tartrate (Ergostat), 2-mg sublingual tablets, and
oral inhalation 0.36 mg per dose.

• Mixtures of ergotamine tartrate and caffeine (Cafergot);
tablets are lmg ergotamine tartrate and 100 mg caffeine,
suppositories are 2 mg/ 100 mg.

• Ergonovine maleate (ergotrate maleate) intramuscular or
intravenous injection, 0.2 mg/ml, 0.2-mg oral tablets.

• Dihydroergotamine mesylate (DHE 45), 1 mg/ml solution
for injection.

• Ergoloid mesylates (Hydergine), 0.5- and 1-mg tablets.

• Bromocriptine mesylate (Parlodel), 2.5- and 5-mg capsules.
Pharmacologically, these preparations are divided into

three categories: amino acid alkaloids (ergotamine); dehydro-
genated amino acid alkaloids (dihydroergotamine and dihy-
droergotoxine); and the amine alkaloids (ergonovine). 12 These
ergot alkaloids are used in a number of therapeutic modalities,
including obstetrics, prophylaxis of deep vein thrombosis,

and orthostatic hypotension. Although rare, there have been
case reports of vascular ischemia in all these applications.

Ergotamine tartrate

Pharmacokinetics

Ergotamine tartrate is a widely used medication for the treat-
ment of acute attacks of migraine headache. The most
common preparation of ergot is ergotamine tartrate caffeine,
marketed as Cafergot, and is available as 2-mg suppositories
and 1-mg oral tablets. It is reported to be about 90% effective in
the treatment of migraine when given parenterally, as op-
posed to 80% with rectal administration and only 50% when
given orally. 1 The poor therapeutic effect of oral and sub-
lingual administration is a result of very low blood levels
due to extensive first-pass hepatic metabolism. Concurrent
administration of caffeine (100 to 200 mg) improves the rate
of absorption and peak plasma concentration, but the oral
bioavailability remains low, at less than 1% 70 min after a 2-mg
tablet; with a rectal suppository, bioavailability is much
greater, with a 20-fold increase in plasma concentration to
400 pg/ml after a similar 2-mg dosage. Parenteral administra-
tion by intramuscular injection increases bioavailability by
50-fold. 12

After intravenous administration of ergotamine, plasma
levels decline rapidly, with an initial distribution half-life of 3
min and a mean terminal half-life of 1.9 h. 20 Plasma concentra-
tion and biologic effect do not correlate. 21 Despite the short
plasma half-life of 2 h, ergot frequently produces vasoconstric-
tion that persists for more than 24-48 h. The prolonged action
of ergotamine has been hypothesized to be due to sequestra-
tion in the tissues. More recent studies, however, using intra-
venous tritium-labeled ergotamine, have demonstrated that
there is a long elimination half-life of 21 h, despite the lack of
detectable ergotamine in the plasma. The slow elimination of
its active metabolites is the most likely explanation for the long
duration of biologic action of ergotamine. 20

Action in headaches

The earliest reference was by Dr. William Thompson of New
York City, who in 1884 recommended the fluid extract of ergot
for relief of periodic headaches. The recommended dosage
was 2 to 4 mg orally or rectally every hour for 3 h or until effec-
tive. 13 Graham and Wolff in 1938 proposed that ergotamine
tartrate acts to alleviate migraine headache by preventing
vasodilation of scalp arteries. 22

The source of pain in migraine headaches and consequently
the action of ergotamine in relieving these symptoms contin-
ues to be disputed. It has been long believed that the prodro-
mal symptoms of classic migraine with aura are initiated by
vasoconstriction with resultant cerebral hypoperfusion, fol-
lowed by painful dilation of extracranial and intracerebral

103

pa rt I Vascular pathology and physiology

vessels. The headache is aggravated by increased amplitude
of pulsations, especially those involving the meningeal
branches. Ergotamine is effective even though it is neither a
sedative nor an analgesic, and has been postulated to reduce
extracranial blood flow, resulting in a decrease in the ampli-
tude of the pulsations. This may alleviate the discomfort of mi-
graine, as has been shown clinically by applying direct
pressure over the carotid artery. Transcranial Doppler has
been used to study fluctuations in intracranial vessel flow, and
in some studies demonstrates a predominantly constrictor ac-
tion in migraine with aura. This vascular theory therefore indi-
cates that cerebral blood flow reduction is the primary cause
of neurologic deficits associated with migraine aura, and sub-
sequent vasodilation produces the headache. 12

With new advances in migraine research using methods to
measure cerebral blood flow and metabolism, a neurogenic
theory has become popular in which cerebral blood flow re-
duction is secondary to neuronal dysfunction. A transient
neuronal excitatory wave is believed to be the primary event
leading to active constriction of resistance vessels and long-
lasting reduction in cortical blood flow. The excitatory electri-
cal wave activates free pain fiber endings— for example, a
slowly conducting C-fiber in a venule initiating a local axon
reflex. The resultant release of a transmitter could cause en-
hanced leakage of plasma proteins across the endothelium,
with subsequent edema formation of the vessel wall and low-
ering of pain threshold locally 23

It therefore is difficult to fully explain the therapeutic action
of ergotamine solely by its vascular component. It is likely that
ergotamine has other actions, such as blocking extravasation
of plasma from the dura, as demonstrated by electrical stimu-
lation of the trigeminal nerve.

Other ergotamine derivatives

Ergonovine

Of the ergot alkaloids, ergonovine has the most potent action
on uterine smooth muscle. By increasing force and duration of
contraction, it continues to be a useful agent to control post-
partum and postabortion bleeding. Ergonovine also may be
helpful in the diagnosis of Prinzmetal angina because it may
precipitate coronary artery spasm in affected people (Fig. 9.3).
It is rapidly absorbed after oral administration.

Dihydroergotamine

Dihydroergotamine is used parenterally for the treatment of
migraine headaches. It has low bioavailability because of
rapid hepatic clearance, and is less completely absorbed and
eliminated more rapidly than ergotamine. Dihydroergota-
mine is more effective on venous capacitance than on arterial
resistance vessels. It is combined with heparin for prophylaxis
against deep venous thrombosis, and decreases venous stasis
by smooth muscle contraction. It also has been useful in the
treatment of orthostatic hypotension by decreasing venous
pooling. Stimulation of oc 2 -adrenoreceptors on the blood ves-
sels may produce a modest increase in systemic vascular resis-
tance; however, there is little sustained elevation in arterial
blood pressure with any of the ergotamine preparations. Even
the mild increase in blood pressure with intravenous ergota-
mine dissipates after a few hours.

Schulman and Rosenberg reported two patients in whom
vasospasm developed with dihydroergotamine. 24 One pa-
tient treated for intractable migraine with dihydroergotamine,

Figure 9.3 Ergonovine-induced coronary artery spasm after intracoronary administration for diagnosis of Prinzmetal angina (From Bove AA, Vlietstra RE.
Spasm in ectatic coronary arteries. Mayo Clin Proc 1 985; 60:822.)

104

CHAPTER 9 Ergotism

1 mg intravenously every 8 h, along with atenolol had right
calf pain associated with decreased pulses in the right lower
extremity; another had right arm pain with diminished pulses
in the right upper extremity 24

Although the incidence of vasospastic reactions to heparin
dihydroergotamine, 5000IU/0.5mg twice a day, is low, esti-
mated to be approximately 0.01-0.002%, there are several re-
ports of severe vasospasm, especially in trauma victims. In one
report, a 27-year-old woman with multiple fractures after a fall
was treated with heparin dihydroergotamine for thromboem-
bolic prophylaxis. After 7 days of therapy, absence of ankle and
arm pulses was noted, and, despite intravenous vasodilators,
necrosis of all toes developed and subsequently the foot had to
be amputated. A repeat arteriogram 48 h after discontinuation
of heparin dihydroergotamine showed resolution of lower
leg arterial vasospasm. 25 At least 12 other cases have been
reported, and in many the vasospastic reactions occurred in
the injured limb. Diffuse arterial spasm has been reported in
another patient with intestinal gangrene and extensive
gangrene of the abdominal wall while receiving heparin
dihydroergotamine after a total hip replacement. 26 It is recom-
mended that heparin dihydroergotamine be avoided in pa-
tients with impaired circulation, those who have recently
undergone arterial surgery, patients with sepsis or liver dys-
function, and in patients who are hemodynamically unstable.

Methysergide

Methysergide (Sansert), which causes smooth muscle contrac-
tion, also inhibits the action of 5-hydroxytryptamine (sero-
tonin), which also may be involved in the causation of vascular
headaches. It therefore has been used in the prophylaxis of
migraine headaches. Although fibrotic complications pre-
dominate, particularly retroperitoneal fibrosis, there are some
reports of arterial vasospasm. 27

Hydergine

Hydergine has been touted for improvement of symptomatic
decline in mental capacity, and as having beneficial effects on
mental alertness, memory, and orientation. It therefore has
been used in the treatment of senile dementias. Each 0.5-mg
tablet contains a combination of dihydroergocorine, dihy-
droergocristine, and dihydroergocryptine as a mixture of
ergoloid mesylates. No significant ischemic episodes have
been reported from the use of this agent.

Bromocriptine

Bromocriptine is effective in decreasing secretion of prolactin
in patients with galactorrhea. As a dopamine receptor antago-
nist, it is also effective in the treatment of Parkinson disease.
There have been rare reports of mild digital vasospasm associ-
ated with this drug. 8

Toxic effects of ergot

Cleveland and King reported that the injection of fowls with
ergotoxine resulted in cyanosis and eventual gangrene of the
combs. It was recognized that this resulted from powerful
constriction of the blood vessels occluding the vasa vasorum,
and resultant vascular damage with the formation of thrombi
leading to gangrene. 28

Von Storch in 1938 found 42 reports of ergot toxicity. 29 All pa-
tients had gangrene or impending gangrene, of which 23 were
obstetric cases and 11 were being treated for thyrotoxicosis,
and eight of the 42 patients died. It was noted that the smallest
dose was 0.5 mg of ergotamine tartrate, and the maximum
150 mg over 14 days. He recognized severe hepatic or renal dis-
ease, sepsis, and obliterative vascular disease to be contraindi-
cations to therapy. 28

Contraindications to ergotamine tartrate (Cafergot) listed in
the package insert (December 1, 1990; Sandoz) include some of
the following: peripheral vascular disease, coronary heart dis-
ease, hypertension, impaired hepatic or renal function, sepsis,
and pregnancy. Additionally stated under "Precautions" is
that ergotism is manifested by intense arterial vasoconstric-
tion producing signs and symptoms of vascular ischemia.
Ergotamine induces vasoconstriction by a direct action on
smooth muscle. In chronic intoxication with ergot derivatives,
headaches, intermittent claudication, muscle pains, and
numbness, coldness, and pallor of the digits may occur. If the
condition is allowed to progress untreated, gangrene can re-
sult. When ergotamine is prescribed in correct dosages in the
absence of contraindications, it is a safe and useful drug; few
serious complications have been reported from its use in the
migraine syndrome.

The incidence of ergot toxicity is estimated at less than 0.01 %
of patients taking ergot preparations and in most cases is due
to its excessive consumption. 14 For this reason, the manufac-
turer recommends that care should be taken to remain within
the limits of recommended dosage. Ergot, however, can be
toxic in patients with liver disease because the liver is a major
site of its detoxification.

The constrictive action of ergotamine may be accentuated
by cigarette smoking because nicotine stimulates the sympa-
thetic ganglion causing vasoconstriction. 30 Accentuation of
vasospasm has been reported with erythromycin. 31 Through
their inhibition of (3-mediated vasodilation, (3-blockers have
been reported in several clinical studies to have a synergistic
constrictive effect with the ergot alkaloids. 32

Clinical manifestations

Extremity ischemia

Ergotism usually causes signs and symptoms of vasospasm,

105

pa rt I Vascular pathology and physiology

especially in the extremities. Initial symptoms may be subtle,
with coolness and pallor, cyanosis, numbness, and tingling in
the feet, less often in the hands. Paresthesias and intermittent
pain may occur early with loss of arterial pulses, or may be-
come apparent only during activity. The involved extremities
are pale and cool, with diminished or absent arterial pulses.
Pallor on elevation and dependent rubor also may be ob-
served. As the symptoms progress, typical features of inter-
mittent claudication and ischemic rest pain occur; in severe
cases, ischemia may result in local tissue necrosis and digital
gangrene. Reports of resultant gangrene and amputation were
more common in the 1940s, when ergot was used for obstruc-
tive jaundice and its toxicity was increased because of liver
damage. 33 Cleveland and King reported on a 42-year-old
woman who developed bilateral gangrene after she received 2
ounces of fluid extract of ergot after an abortion, necessitating
amputation of both legs. 28 Since that time, there have been a
number of reports of gangrene of the extremities from taking
parenteral and rectal preparations of ergot. Lower extremity
gangrene also has been reported after ergotrate taken orally. A
more frequent presentation, however, is ischemic rest pain
without skin necrosis.

Greenberg and Hallett discussed a typical case of a 48-year-
old woman with sudden onset of bilateral lower extremity
pain and numbness and absent pulses distal to the femoral
level. Vasospasm was documented by angiography and
responded to drug withdrawal. 34 Vasospasm may be aggrav-
ated by (3-blockers (propranolol, 10 mg, four times a day), as
occurred in this patient.

Most cases consist of women between the ages of 20 and 45
years, usually taking Cafergot for treatment of their long-
standing migraine headaches. Sometimes symptoms of is-
chemic rest pain and impending gangrene develop acutely
and cannot be differentiated from an acute thrombosis or
thromboembolism. These features frequently are misdiag-
nosed by patient and physician alike as atherosclerosis obliter-
ans. Abrupt onset of symptoms with bilateral leg discomfort
has been misdiagnosed as possible saddle embolus or acute
aortic dissection. Angiographic findings have been confused
with vasculitis. Although presentations most commonly in-
volve lower and upper extremities, ergot may cause con-
striction of smooth muscle in all vascular beds. From cases
reported in the literature, initial diagnoses include aortic dis-
section in a patient with absent femoral pulses, atherosclerosis
obliterans with claudication and Leriche syndrome, thrombo-
sis in situ, and arteritis. Angiographic signs also have been
confused with fibromuscular dysplasia.

Typically, involvement of the extremities is symmetric,
although symptomatic presentation may be unilateral
ischemia. 35,36 Cases of unilateral brachial artery thrombosis
have been reported in patients receiving ergotamine tartrate
for migraine headaches. In one report, 37 an embolism to the
axillary artery was incorrectly diagnosed based on an arteri-
ogram. A repeat angiogram showing no residual obstruction

led to the correct diagnosis of Caf ergot-induced vasospasm. In
another example, a 61-year-old farmer presenting with pain
and weakness of the left forearm was thought to have a throm-
botic occlusion of the brachial artery, with clinical findings of a
cyanotic, cool forearm and absent pulses. Because of persistent
vasospasm for 3 days after withdrawal of oral ergotamine and
persistence of brachial artery obstruction despite sublingual
nitroglycerin, he underwent saphenous vein bypass. A repeat
angiogram 10 days later showed a patent native brachial
artery and bypass graft. Of note, he also had intermittent
angina despite no significant coronary artery disease by
angiography, and this probably was secondary to coronary
vasospasm. 38

After drug withdrawal, symptoms in most patients begin to
improve within 24 h, and pulses return within 48 h; however,
in some the vasospasm can last as long as 8 days 38 to 2 weeks. 39
Arterial narrowing produced by ergotamine may not always
be reversible. Persistent narrowing of a proximal celiac artery
has been reported in a patient presenting with diffuse arm, leg,
and mesenteric vasospasm who had been taking up to 30 mg
of ergotamine weekly for 14 years. 40 An iatrogenic renal artery
aneurysm has been documented at the site of a previous
ergot-induced stenosis. 41 Arterial damage resulting in late
aneurysm formation is rare and may be a complication of
vasculitis.

Systemic symptoms have been reported, including nausea,
vomiting, diarrhea, and headache. Confusion, depression,
drowsiness, and convulsion may occur rarely. Cases of
headaches and dysphoria improved by cessation of ergota-
mine have been reported. In some patients, small cerebral in-
farcts seen on computed tomography and magnetic resonance
imaging have been attributed to occlusion of superficial
cortical vessels. 42

Dosage

Most cases of ergotism result from a dosage in excess of recom-
mended amounts. There is, however, great variability in ergot
tolerance in that some patients may take huge doses (42 mg by
suppositories every week) 14 for many years with few symp-
toms, whereas ergotism has been reported in patients on ther-
apeutic doses taken for as little as 24 h and with as little as
2.5 mg ergotamine tartrate. 38

The cause of vascular insufficiency frequently is missed, as
noted in multiple case reports. Patients frequently fail to men-
tion ergotamine as a currently taken medication. Often, the di-
agnosis of ergotism is made long after hospital admission, and
the patient has started to improve because of abstinence from
ergotamine while in the hospital. Surreptitious use occurs,
however, and there are examples of patients who worsen after
several days of hospitalization because of continued use of
ergot suppositories. Psychiatric illness such as schizophrenia
may predispose to abuse, but frequently ordinary people are
afflicted.

106

CHAPTER 9 Ergotism

At the Mayo Clinic, we identified 38 cases of ergotism be-
tween 1945 and 1985. In most of these patients, the dosage was
excessive, although in two patients two suppositories over 4
days produced significant vasospastic effects. An unusual pre-
sentation involved a 40-year-old white woman who was ad-
mitted in January 1983 with severe Raynaud phenomenon of
the hands with cold, painful, cyanotic fingers and bilateral calf
claudication at 30 steps. She had been using Cafergot supposi-
tories for migraine headaches. She was treated with nitroprus-
side after an unsuccessful trial of nifedipine, and all pulses
returned with relief of symptoms. Four months later she was
readmitted with return of symptoms, and Cafergot again was
discontinued. In January 1984, she was admitted with severe
hand and foot ischemia while taking verapamil and prazosin
for Raynaud phenomenon diagnosed elsewhere. She was
again treated with nitroprusside, intravenous fluids, and he-
parin. Although the patient denied further ergot use, it was
found that she recently had filled three prescriptions for
Cafergot. Surreptitious use was confirmed by the finding of a
small box filled with tissue paper and Cafergot suppositories.
Another case is presented in Figure 9.4.

Often it is hard for the patient to distinguish a vascular
headache from a nonvascular headache, and ergot may be
taken more frequently. Patients usually realize they are taking
excessive amounts of ergot but may be reluctant to admit this

to the doctor because of the concern that he or she will reduce
the medication, which they believe is vital to them. In addition,
cessation of ergot frequently leads to a severe rebound
headache and thus initiates a vicious circle of excessive dosage,
as recognized by Peters and Horton in 1925. 43 Saper has des-
cribed the rebound headache as "a predictable protracted and
severely debilitating headache accompanied by autonomic
disturbances and other somatic and mental complaints" }

This syndrome can be prevented by eliminating the use of
ergotamine to no more than 2 dosage days per week. 1 Alterna-
tive medications may assist with drug withdrawal, including
(3-blockers, calcium antagonists, tricyclic antidepressant med-
ications, Midrin (Carnrick Laboratories, Cedar Knolls, NY),
analgesics, or nonsteroidal antiinflammatory agents. 1

Angiographic findings

The angiographic features of ergotism involving the leg ves-
sels were first reported in 1936 by Yater and Cahill. 33 They
discussed a 64-year-old fisherman who was admitted to the
hospital for malaise and fever and was treated with ergota-
mine tartrate for pruritus secondary to jaundice. He received
0.5 mg intramuscularly three times a day continuously for 13
days. During that time in the hospital, progressive gangrene
of the feet developed, ultimately necessitating bilateral

Figure 9.4 Operative angiogram in a 54-year-old woman who presented to
the emergency department with sudden pain, numbness, and cyanosisof
both lower extremities, who initially was suspected of having a saddle
embolus (pulses distal to the femoral arteries were absent). (A) Left

transfemoral angiogram: superficial femoral artery and distal vessels not
visualized because of severe vasospasm. (B and C) After Fogarty balloon
dilation and papaverine: superficial femoral artery popliteal arteries patent
with residual distal spasm.

107

pa rt I Vascular pathology and physiology

below-knee amputation. Findings on a preoperative arteri-
ogram were classic for ergotism and reported "the main leg ar-
teries of the leg to be smooth in outline and apparently normal
down to the lower third of the leg where they faded out into a
point. Small, long and somewhat tortuous collateral arteries
passed downward toward the feet from the arteries above the
point of occlusion". 33 During the past 50 years, multiple indi-
vidual case reports have documented the angiographic spec-
trum of ergotism. 38-40 Variable vasoconstriction of large and
medium-sized arteries generally is seen, with smooth narrow-
ing of vessels, which often appears as tapering and sometimes
is so severe that only thread-like lumen can be seen. Diffuse or
segmental vessel occlusion also may occur. In chronic cases,
extensive collateral formation sometimes is seen, with collat-
eral vessels that may be larger than the main arteries. In the
legs, vasospasm may begin in the superficial femoral arteries
and becomes more severe distally with narrowing and distal
occlusions. Common femoral and iliac arteries and abdominal
aorta are involved to a lesser extent. In the upper extremity,
brachial, radial, and ulnar arteries may be segmentally in-
volved. Coronary, carotid, vertebral, and cerebral artery in-
volvement has been reported. Superior mesenteric artery and
renal artery vasoconstriction due to ergotism have been angio-
graphically documented. 11/44/45

Angiography is unnecessary in cases in which the diagnosis
of peripheral ischemia due to ergotism has been clearly made
on clinical grounds, and there is no evidence of critical is-
chemia or gangrene. Angiography, however, remains impor-
tant when the cause of arterial insufficiency is uncertain and
the diagnosis of ergotism is unsuspected clinically, as occurs in
surreptitious use of ergot.

Unusual effects of ergotism

Ophthalmic complications

Ocular manifestations of ergot toxicity are rare. Bilateral is-
chemic papillitis and decreased central visual acuity have
been reported in patients receiving ergotamine. 46 Severe,
generalized retinal vasoconstriction has been reported in a
young woman who was treated with oral ergotamine tartrate,
6mg/day, for symptomatic orthostatic hypotension after a
suicide attempt by ingestion of a rodenticide. Blurred vision
developed, and ophthalmoscopy revealed marked constric-
tion of all retinal vessels, especially the arterioles. 47

Hypertension

Ergometrine administered postpartum in obstetric practice
has been reported to cause hypertension in three patients and
a cardiac arrest in another. In another case, severe hyper-
tension (190/120) developed in a 17-year-old, normotensive,
primigravida girl after intravenous administration of

ergometrine 0.2 mg, and she died of an intracerebral hem-
orrhage. 48 Although the effect of ergot alkaloids on blood
pressure varies, some patients may have a hypertensive
response to ergometrine, especially if they have preexisting
hypertension. In one study, in which 0.5 mg ergometrine was
given intravenously during general anesthesia for cesarean
section, of women whose diastolic blood pressure was
100-109 mmHg, approximately half had rises of 20mmHg or
more. If diastolic blood pressure was less than 90 mmHg, only
3.5% were found to have an increase of this magnitude. 49

Renal artery

Renovascular hypertension and renal failure have been asso-
ciated with ergot poisoning. In 1970, Fedotin and Hartman
reported the case of a 40-year-old woman taking Cafergot
suppositories for migraine headaches who presented with
hypertension. A renal arteriogram showed bilateral renal
artery stenosis due to suspected fibromuscular dysplasia;
however, 24 h later, a second angiogram was negative. 7 Renal
failure also has been associated with ergot poisoning. In an-
other case, a 35-year-old man was found to have a blood pres-
sure of 190/100 while taking eight suppositories a day for
several years for migraine headaches. A renal arteriogram
showed smooth segmental narrowing of the right renal artery,
and a repeat angiogram 10 months after stopping the drug
showed resolution of spasm and a small aneurysm at the site of
constriction. 40

Gastrointestinal vascular ischemia

Reports of vasospastic complications of ergot therapy affect-
ing the mesenteric vessels are rare, probably because of the
unrecognized association of abdominal pain with ergotism.
Holmes and associates described a series of five pregnant Fiji
women with fatal intestinal gangrene who were suspected of
taking large quantities of ergot to induce abortion. 50 Re-
versible vasospasm of the superior mesenteric artery has been
well documented by arteriography. Buenger and Hunter de-
scribed a patient taking methysergide who had abdominal
pain, and documented multiple areas of spasm of the superior
mesenteric artery. After stopping ergot, resolution of va-
sospasm was documented on repeat arteriography 11 weeks
later. 51 Green and colleagues described a 40-year-old woman
with cyanosis and pain of both feet who, on arteriography, had
severe constriction in lower limb arteries, superior mesenteric
artery, and intestinal branches. After vasodilator therapy, dis-
tal pulses returned, and on repeat arteriography, the superior
mesenteric artery had returned to normal caliber. 52 Stillman
and coworkers reported on a 50-year-old woman who pre-
sented with abdominal pain and rectal bleeding while taking
estrogen and oral ergotamine. Clinical findings included an
epigastric bruit; in this patient, however, there was no evi-
dence of vasospasm involving the extremities. 53

108

CHAPTER 9 Ergotism

Coronary artery spasm

Since it first was recommended by Stein in 1949, ergonovine
maleate has been widely used as a provocative test to induce
coronary artery spasm in patients with Prinz-metal angina 54
(see Fig. 9.3). Therapeutic doses of ergotamine have been
associated with myocardial ischemia, infarction, cardiac
arrest, and sudden death. 10 Although ergonovine may pre-
cipitate angina in patients with preexisting, asymptomatic
coronary artery disease, there a number of reports describing
recurrent episodes of angina after administration of ergota-
mine in younger patients with no evidence of coronary athero-
sclerosis. 3 ' 10 One patient took a massive overdose of Migral
(60 tablets, equivalent to 120 mg ergotamine) with resultant
ventricular fibrillation and acute inferolateral myocardial
infarction. He also had absent peripheral pulses with pregan-
grenous changes in his toes. He responded to intravenous
sodium nitroprusside continued for 5 days. 55 Patients with
ergotamine-induced myocardial ischemia may present with
typical symptoms of unstable angina, with left substernal
chest pain radiating to the left arm and ischemic T-wave
changes on electrocardiogram. Intravenous nitroglycerin also
is the drug of choice in these patients.

Treatment of ergotism

There is no specific antidote for ergot-induced peripheral va-
sospasm (Table 9.1). The initial treatment of ergot intoxication
is the immediate and permanent discontinuation of ergot ad-
ministration. This is the cornerstone of treatment, and in
milder cases may be all that is necessary 14 ' 56 General manage-
ment of lower extremity ischemia includes bed rest, a vascular
boot to maintain warmth and protection of the extremity, and a
dependent position for the leg to increase perfusion pressure.
Intravenous fluids along with heparin anticoagulation should
be considered in more severe cases.

Intravenously infused nitroprusside and nitroglycerin are
effective agents in the treatment of vasospasm. Nitroprusside
has emerged as the drug of choice in the management of severe
ergotism, as documented by multiple case reports citing both
clinical and angiographic improvement. During the past 40
years, numerous other pharmacologic agents have been tried
and given intravenously, intraarterially, and orally in an at-
tempt to overcome the intense vasospasm induced by ergota-
mine. Most of these agents have been used empirically, with
varying results reported but with no clear evidence of benefit
in many.

Sodium nitroprusside

This drug is used chiefly in the management of acute con-
gestive heart failure with pulmonary edema, and also in
hypertensive emergencies. Sodium nitroprusside is called a

Table 9.1 Treatment of ergotism

Effective therapies for vasoconstriction due to ergotism

Immediate and permanent cessation of ergot administration

Supportive measures
Intravenous fluids to maintain adequate hydration
Care and protection of ischemic extremities
Vascular boot

Mild dependency of foot to enhance perfusion pressure
Anticoagulation
Intravenous heparin to prevent thrombosis

Intravenous vasodilators
Sodium nitroprusside for severe ischemia or impending gangrene

Other measures of possible benefit

Intravenous nitroglycerin

Intravenous prostacyclin
Oral calcium blockers
Intraarterial balloon dilation

Measures of little or no benefit

Sympathectomy (surgical sympathectomy, chemical sympathectomy,
epidural block)

Drugs
Ethyl alcohol
Phentolamine
Amyl nitrate
Scopolamine
Theophylline
Papaverine

Procaine hydrochloride
Lidocaine
Tolazoline
Streptokinase

balanced vasodilator in that it relaxes vascular smooth muscle
of both arterioles and venules. This accounts for its value in the
treatment of congestive heart failure; it decreases preload sec-
ondary to venous dilation and reduces arterial compliance,
improving cardiac output. Arteriolar vasodilation induces
hypotension, so it also is a drug of first choice in hypertensive
emergencies. Because it is a nonselective vasodilator, renal
blood flow and glomerular filtration rate are maintained.
Sodium nitroprusside must be given by continuous intra-
venous infusion to be effective; it has an onset of action within
30 s and a peak hypotensive effect within 2min. A major ad-
vantage is the disappearance of its effects within 3min when
the infusion is stopped. Breakdown of nitroprusside occurs
rapidly as it reacts with membrane-bound sulfhydryl groups
of the vascular wall and erythrocytes, causing its disassocia-
tion into cyanide and nitric oxide. Its main action is through
production of nitric oxide (NO). 57 NO also is produced en-
dogenously by vascular endothelium (endothelium-derived

109

pa rt I Vascular pathology and physiology

relaxing factor), 58 and causes vasodilation by activating
guanylate cyclase in vascular smooth muscle. Cyanide is me-
tabolized in the liver to thiocyanate, which has a mean half-life
of 3 days and is eliminated by the kidney The risk of thio-
cyanate toxicity increases with infusions of sodium nitroprus-
side for more than 24-48 h and with infusion rates greater than
2jig/kg/min. With prolonged infusions, plasma concentra-
tion of thiocyanate can be monitored and should not exceed
O.lmg/ml. 57 Clinical features of thiocyanate toxicity include
anorexia, nausea, fatigue, disorientation, and psychosis. The
use of sodium nitroprusside requires a variable-rate infusion
pump and direct monitoring of radial artery pressure in an
intensive care unit. Most patients respond to doses of 0.5-
1.5jig/kg/min, although higher doses may be necessary in
some cases. The optimal infusion rate and duration of therapy,
however, have not been established, and administration of
nitroprusside needs to be individualized. In general, the infu-
sion rate should be titrated to obtain maximal clinical im-
provement as judged by disappearance of cyanosis and return
of warmth and pulses, while maintaining an adequate blood
pressure and urine output. Administration time varies from 24
to 72 h, because vasospasm may return after nitroprusside
infusion is stopped. 59 Even at 72 h, repeat angiography has
shown residual vasospasm in some patients despite reappear-
ance of peripheral pulses. 60

Nitroglycerin

Nitroglycerin administered intravenously also has been
reported to be effective for ergotamine-induced peripheral
ischemia. The onset of action is within 1 min, and it can be
rapidly titrated to effective doses. Nitroglycerin also produces
smooth muscle relaxation mediated by activation of soluble
guanylate cyclase and increased rate of guanosine monophos-
phate synthesis. 61 NO produced endogenously from endothe-
lial cells also is formed from organic nitrates with resultant
arterial and venous dilation. With lower doses, only venodila-
tion occurs. Larger doses are necessary to induce arteriolar va-
sodilation, and use of nitroglycerin therefore may be limited
by hypotension. Nitrates are potent vasodilators of epicardial
coronary arteries, and promote resolution of ergonovine-
induced spasm. 61 Nitroglycerin is metabolized by the liver
and other tissues. Side-effects include hypotension,
headaches, and methemoglobinemia.

In 1982, Tfelt-Hansen and colleagues studied the effect of
nitroglycerin on segments of human temporal arteries con-
tracted with ergotamine. 62 Addition of nitroglycerin effec-
tively and rapidly relieved the ergotamine-induced arterial
contraction. Vasoconstriction recurred when nitroglycerin
was removed, indicating that nitroglycerin is not an antidote
but acts by a direct vasodilatory action. Although sodium
nitroprusside remains the drug of first choice in treating
ergotamine-induced peripheral ischemia, nitroglycerin may
be useful in patients with concomitant angina pectoris. It may

be an alternative agent when large doses or prolonged infu-
sions of nitroprusside cause cyanide or thiocyanate toxicity,
especially in the presence of renal insufficiency.

Anticoagulation

Anticoagulation with full-dose intravenous heparin has
been advocated as routine therapy to counteract the devel-
opment of stasis thrombosis due to vasospasm. 63 Systemic
heparin as well as low-molecular-weight dextran also may
improve microvascular flow by inhibiting distal throm-
bosis. Streptokinase also has been used to dissolve thrombus
in patients with incipient gangrene. 64 Thrombolytic agents,
however, are unlikely to be useful in most cases of ergotism.

Prostanoids

Prostanoids (prostaglandin-1, PGI 2 , or the prostacyclin ana-
logue iloprost) may be helpful in patients with severe leg
ischemia. 65 All prostanoids have to be given by intravenous
infusion. Because they are metabolized rapidly by the liver
and lungs, they are more effective given intraarterially They
prevent platelet activation, aggregation, and adhesion, have a
stabilizing effect on leukocytes, and are potent vasodilatory
agents mediated by stimulation of adenyl cyclase, with result-
ing mild side-effects of flushing, headaches, and nausea. 66
Levy and associates reported the use of prostaglandin-1
infused into the distal abdominal aorta at 1 mg/min in a 38-
year-old woman with ergotamine-induced lower extremity is-
chemic rest pain. Within 10 min, leg pain subsided and limb
warmth increased. After a 12-h infusion, pedal pulses were
present by Doppler measurement, and there was angiogra-
phic evidence of increased caliber of the iliac and femoral
arteries, although significant vasospasm persisted. 67

Calcium channel blockers

Calcium channel blockers relax arterial smooth muscle with
little effect on venous beds by inhibiting voltage-dependent
calcium channels in vascular smooth muscle . Calcium channel
blockers such as nifedipine have been reported to be effective
in relieving peripheral ischemia associated with ergotamine
tartrate. 68 Kemerer and colleagues described a 53-year-old
woman with severe vasospasm involving the distal aorta and
iliofemoral, superficial femoral, and trifurcation vessels by an-
giography; treatment with oral nifedipine, 10 mg three times a
day, caused resolution of symptoms within 2 days and no evi-
dence of residual spasm on repeat angiography 5 days later. 69
It is unknown, however, whether this response is any more
rapid than that provided by conservative therapy with simple
drug withdrawal, fluids, and heparinization. Clearly, in cases
of critical leg ischemia, intravenous nitroprusside is necessary
for prompt vasodilation. Nifedipine in some patients is unable
to overcome the constrictive effects of ergotamine. Wells and

110

CHAPTER 9 Ergotism

associates reported on a patient who was taking nifedipine for
arterial hypertension but in whom ischemic symptoms from
Cafergot still developed. 11

The dihydropyridines, especially nifedipine and nicardip-
ine, are more potent vasodilators than is verapamil, which in
turn is more potent than diltiazem. In mild cases of ergotism,
use of a calcium antagonist as empiric therapy may be reason-
able; however, there is no evidence that this shortens the dura-
tion or intensity of symptoms due to ergotism.

Prazosin

Prazosin hydrochloride is a selective a 1 -adrenergic receptor
blocker in arterioles and veins; it causes a decrease in peri-
pheral vascular resistance similar to the effect of nitro-
prusside, and has the advantage of oral administration.
Prazosin also has been used successfully in the treatment of
ergotamine-induced peripheral ischemia, but its use is not
advocated for severe ischemia or impending gangrene. 11

Other vasodilators

A number of other vasodilators have been used with varying
results, including intraarterial tolazoline, phentolamine,
papaverine, procaine, and phenoxybenzamine. Tolazoline is
an a-adrenergic receptor-blocking agent similar to phento-
lamine, with affinity for both a t - and oc 2 -adrenoreceptors. In
theory, oc-blockers should be of benefit in reducing vasocon-
striction; however, in multiple case reports they are unable to
overcome the vasospasm induced by ergotamine, which acts
through other mechanisms in addition to a-receptor stimula-
tion. Reserpine depletes norepinephrine from sympathetic
nerves, attenuating adrenergic constriction; however, it also is
ineffective in ergotamine-induced arterial spasm.

During the past 30 years, procedures to interrupt sympa-
thetic tone, including surgical sympathectomy and epidural
and spinal anesthetics, have been tried. These are ineffective,
however, and do not reverse the vasoconstriction produced by
ergot with its direct action on vascular smooth muscle.

Mechanical dilation

Reversal of ergotamine-induced arterial spasm by mechanical
intraarterial dilation has been reported. Shifrin and associates
described the use of a Fogarty balloon-tip catheter for at-
tempted thrombectomy of a 49-year-old woman who pre-
sented with severe pain and numbness of both legs, absent
pulses, and severe arterial spasm distal to the abdominal
aorta, as demonstrated by translumbar arteriography 70
Passage of a Fogarty balloon catheter did not retrieve any
thrombi; however, when the catheter was removed, the
arteries had distended to their normal diameter and pulses
were again palpable in both feet. A second patient, a 38-year-
old woman with pain and paresthesias in both hands and feet,

was treated with tolazoline fluids, low-molecular-weight
dextran, 20% mannitol, continuous epidural blockade, full
heparinization, intravenous nitroprusside, and finally a stel-
late ganglion block, but all this did not restore pulses to the
left hand. With progressive cyanosis of all four extremities,
she underwent Fogarty catheter dilation of the femoral and
brachial arteries, which resulted in immediate resumption of
flow to both hands and legs. 70

Successful intraarterial balloon angioplasty also has been
reported in patients with severe and drug-refractory leg or
arm artery spasm. Wells and colleagues reported a popliteal
stenosis that was successfully dilated in a 37-year-old woman
with ischemic rest pain in her foot after intraarterial tolazoline
failed to relieve the vasospasm. 11 In another report, severe
popliteal and superficial femoral vasospasm developed in a
patient with right foot ischemia secondary to methysergide
and intramuscular ergotamine tartrate; catheter dilation treat-
ment restored peripheral pulses within 3 days. 67 ' 71

Intraarterial dilation may be warranted when extremities
are in imminent danger of gangrene. The mechanism by which
dilation produces immediate and permanent relief of va-
sospasm is not known, but it is hypothesized that mechanical
stretching in some manner interrupts the sustained contrac-
tion. With more forceful overdilation of the arteries, it is pos-
sible to produce damage to the smooth muscle, as occurs in
balloon angioplasty of fixed stenotic coronary and peripheral
artery stenosis, rendering the artery noncontractile in re-
sponse to ergonovine.

Surgical revascularization also has been performed in cases
of severe ischemia not responding to drug withdrawal or va-
sodilators. Demartini and coworkers reported such a case in a
61-year-old farmer who had left forearm cyanosis, hand
ischemia, as well as angina pectoris. An arteriogram de-
monstrated occlusion of the brachial artery at the level of the
mid-humerus. Sublingual nitroglycerin increased the diame-
ter of the brachial artery but did not change the obstruction.
Forty-eight hours later, the pulse remained absent and the pa-
tient underwent surgical bypass using saphenous vein graft
from the brachial artery to the radial artery. The surprise
finding on a postoperative arteriogram was a patent brachial
artery along with patent saphenous vein graft. 38

Magee reported a case of a 48-year-old woman treated with
methysergide for migraine headaches in whom angiographi-
cally confirmed claudication developed secondary to aor-
toiliac stenosis, which was treated with a right-to-left femoral
cross-over graft. A second patient underwent aortobilateral
common femoral artery bypass with bifurcated Dacron for
iliofemoral stenosis attributed to fibrous reaction from
methysergide, with additional vascular spasm resulting from
concomitant use of ergotamine tartrate orally 6

Prognosis

After cessation of drug use, complete resolution has been

111

pa rt I Vascular pathology and physiology

noted in as soon as 1 day, 6/35 or as long as 14 days. 39 In some
patients, complete resolution of symptoms does not occur
despite restoration of extremity pulses. This is probably due
to small vessel and peripheral nerve damage, resulting in
chronic or intermittent episodes of burning pain, especially in
the feet, similar to peripheral neuropathy. This may last for
many months or longer despite discontinuation of ergot
preparations. 72 Intermittent episodes of vasospasm involving
digital small vessels with skin color changes and reactive
hyperemia of the foot despite normal pedal pulses have been
reported 4 months after cessation of ergot. 6 Permanent
ischemic lateral popliteal nerve palsy due to ergot toxicity also
has occurred. 73 In most patients with ergotism who have full
recovery from vasospasm and no tissue damage, the long-
term prognosis is excellent if recurrent abuse of ergotamine
can be prevented.

Conclusion

Severe cases of ergotism today are rare; however, migraine is
common and continues to be treated effectively with ergot
medications. Thus, the potential for significant vasospastic
episodes will continue to exist. It is therefore important for
both the general internist and vascular surgeon to be aware of
possible ischemic complications that may involve any part of
the arterial system. Early recognition is important to avoid
gangrenous complications and to initiate appropriate treat-
ment. In unusual cases, angiography frequently is helpful in
the diagnosis. Ergotism can be treated successfully in most
situations by drug withdrawal alone, with restoration of nor-
mal circulation. In severe cases with threatened limb or organ
loss, sodium nitroprusside is the therapy of choice to reverse
vasospasm due to ergotism.

References

1. Saper JR. Ergotamine dependency: a review. Headache 1987;27:435.

2. Stang PE, Osterhaus JT. Impact of migraine in the United States:
data from the National Health Interview survey. Headache 1992;
33:29.

3. Goldfischer JD. Acute myocardial infarction secondary to ergot
therapy. NEngl J Med 1960; 262:860.

4. Abercrombie D, Oehlert WH. Ergotism as a cause of acute va-
sospastic disease with features mimicking severe atherosclerosis.
OK State Med Assoc 1984; 77:86.

5. McLoughlin MG, Sanders RJ. Ergotism causing peripheral vascu-
lar ischemia. Rocky Mount Med J 1972; 69:45.

6. Magee R. Saint Anthony's fire revisited: vascular problems associ-
ated with migraine medication. Med] Aust 1991; 154:145.

7. Fedotin M, Hartman C. Ergotamine poisoning producing renal
arterial spasm. N Engl J Med 1970; 283:518.

8. Wass JAH, Thorner MO, Besser GM. Digital vasospasm with
bromocriptine. Lancet 1976; 1:1135.

9. Rogers DA, Mansberger JA. Gastrointestinal vascular ischemia
caused by ergotamine. South Med J 1989; 82:1058.

10. Galer BS, Lipoton RB, Solomon S et ah Myocardial ischemia
related to ergot alkaloids: a case report and literature review.
Headache 1991; 31:446.

11. Wells KE, Steed DL, Zajko AB, Webster MW. Recognition and
treatment of arterial insufficiency from Cafergot. / Vase Surg 1986;
4:8.

12. Rail TW. Oxytocin, prostaglandins, ergot alkaloids, and other
drugs; tocolytic agents. In: Gilman A, Rail TW, Nies A, Taylor P,
eds. Goodman and Gilman's Pharmacological Basis of Therapeutics. 8th
edn. New York: Pergamon Press-Macmillan, 1990:933.

13. Tanner JR. St. Anthony's fire, then and now: a case report and
historical review. Can J Surg 1987; 30:291 .

14. Merhoff GD, Porter JM. Ergot intoxication: history review and
description of unusual clinical manifestations. Ann Surg 1974;
180:773.

15. Barger G. Ergot and Ergotism. London: Gurney & Jackson, 1931.

16. Demeke T, Kidane Y, Wuhib E. Ergotism: a report of an epidemic,
1977-1978. EthiopMedJ 1979; 17:107.

17. Berde B. Pharmacology of the ergot alkaloids in clinical use. Med J
Aust 1978;2(Suppl3):3.

18. Aellig WH. Influence of pizotifen and ergotamine on the venocon-
strictor effect of 5-hydroxytryptamine and noradrenaline in man.
Eur J Clin Pharmacol 1983; 25:759.

19. Dale HH. On some physiological actions of ergot. / Physiol 1906;
34:163.

20. Ibraheem JJ, Paalzow L, Tfelt-Hansen P. Kinetics of ergotamine
after intravenous and intramuscular administration to migraine
sufferers. Eur J Clin Pharmacol 1982; 23:235.

21. Graham AN, Johnson ES, Persaud NP et al. Ergotamine toxicity
and serum concentrations of ergotamine in migraine patients.
Hum Toxicol 1984; 3:193.

22. Graham JR, Wolff HG. Mechanism of migraine headache
and action of ergotamine tartrate. Arch Neurol Psychiatr 1938;
39:737.

23. Hardebo JE. Migraine: why and how a cortical wave may initiate
the aura and headache. Headache 1991; 31:213.

24. Schulman EA, Rosenberg SB. Claudication: an unusual side effect
ofDHE administration. Headache 1991; 31:237.

25. Cunningham M, A de Torrente JM, Ekoe JP et al. Vascular spasm
and gangrene during heparin-dihydroergotamine prophylaxis.
BrJSurg 1984; 71:829.

26. Van den Berg E, Rumf KD, Frohlich H. Vascular spasm during
thromboembolism prophylaxis with heparin-dihydroergota-
mine. Lancet 1982; 2:268.

27. Katz J, Vogel RM. Abdominal angina as a complication of methy-
sergide maleate therapy. JAMA 1967; 199:160.

28. Cleveland FE, King RL. Gangrene following ergotamine tartrate
therapy oi migraine. Bull Mason Clin 1948; 2:19.

29. Von Storch TJC. Complications following the use of ergotamine
tartrate: their relation to the treatment of migraine headache.
JAMA 1938; 111:293.

30. Fielding JWL, Donovan RM, Burrows FGO, Hurlow RA.
Reversible arteriopathy following an ergotamine overdose in a
heavy smoker. BrJSurg 1980; 67:247.

31. Francis H, Tyndall A, Webb J. Vascular spasm due to ery-
thromycin-ergotamine interaction. Clin Rheum 1984; 3:243.

32. Venter CP, Joubert PH. Severe peripheral ischemia during

112

CHAPTER 9 Ergotism

concomitant use of beta-blockers and ergot alkaloids. Br Med J
1984; 289:288.

33. Yater WM, Cahill JA. Bilateral gangrene of feet due to ergotamine
tartrate used for pruritus of jaundice. JAMA 1936; 106:1625.

34. Greenberg DJ, Hallett JW. Lower extremity ischemia due to com-
bined drug therapy for migraine. Postgrad Med 1982; 72:103.

35. Tator CH, Heimbecer RO. Unilateral arm ischemia due to ergota-
mine tartrate. Can Med Assoc J 1960; 95:1319.

36. Atwell D, Pois A, Moriedge J et ah Severe unilateral ischemia
secondary to ergot intoxication. Wis Med J 1976; 75:S33.

37. Herlache J, Hoskins P, Schmidt CM. Ergotism. Unilateral brachial
artery thrombosis secondary to ergotamine tartrate. Angiology
1973; 24:369.

38. Demartini DR, Pluncker MW, Johnson F et ah Ergot induced
unilateral brachial artery occlusion. Minn Med 1979; 62:719.

39. Imrie CW. Arterial spasm associated with oral ergotamine
therapy. Br] Clin Pract 1973; 27:457.

40. Corrocher R, Brugnara C, Maso Ret ah Multiple arterial stenoses in
chronic ergot toxicity. N Engl J Med 1981; 311:261.

41. Pajewski M, Modai D, Wisgarten J. Iatrogenic arterial aneurysm
associated with ergotamine therapy. Lancet 1981; 2:935.

42. Fincham RW, Perdue Z, Dunn VD. Bilateral focal cortical atrophy
and chronic ergotamine abuse. Neurology 1985; 35:720.

43. Peters GA, Horton BT. Headache: with special reference to the
excessive use of ergotamine preparations and withdrawal effects.
Proc Staff MeetMayo Clin 1925; 9:153.

44. Syme J, Whitworth JA. Ergotamine-induced peripheral arterial
spasm: clinical and angiographic diagnosis. Australas Radiol 1971;
15:45.

45. Richer AM, Banker VP Carotid ergotism: a complication of
migraine therapy. Radiology 1973; 106:339.

46. Mindel JS, Rubenstein AE, Franklin B. Ocular ergotamine tartrate
toxicity during treatment of vacor-induced orthostatic hypoten-
sion. Am J Ophthalmol 1961; 92:492.

47. Gupta DR, Strobos RJ. Bilateral papillitis associated with Cafergot
therapy. Neurology 1972; 22:793.

48. Browning DJ. Serious side effects of ergometrine and its use in
routine obstetric practice. Med] Aust 1974; 1:957.

49. Baillie TW. Vasopressor activity of ergometrine maleate in
anesthetized parturient women. Br Med] 1963; 1:585.

50. Holmes G, Martin E, Tabua S. Mesenteric vascular occlusion in
pregnancy: suspected ergot poisoning. Med] Aust 1969; 2:1009.

51. Buenger RE, Hunter JA. Reversible mesenteric vascular artery
stenosis due to methysergide maleate. JAMA 1966; 198:144.

52. Green FL, Ariyan S, Stansel HC. Mesenteric and peripheral vascu-
lar ischemia secondary to ergotism. Surgery 1977; 81:176.

53. Stillman AE, Weinberg M, Mast WC, Palpant S. Ischemic bowel
disease attributable to ergot. Gastroenterology 1977; 72:1336.

54. Heupler FA, Proudfit WL, Razavi M et ah Ergonovine maleate
provocative test for coronary artery spasm. Am J Cardiol 1978;
41:631.

55. Carr P. Self-induced myocardial infarction. Postgrad Med J 1981;
57:654.

56. Perry MO. Ergot induced vascular insufficiency. West J Med 1977;
127:246.

57. Gerber JG, Nies AS. Antihypertensive agents and the drug the-
rapy of hypertension. In: Gilman A, Rail TW, Nies A, Taylor P, eds.
Goodman and Gilman's Pharmacologic Basis of Therapeutics. 8th edn.
New York: Pergamon Press-Macmillan, 1990:784.

58. Moncada S, Radomski MW, Palmer RM. Endothelial-derived
relaxing factor: identification as nitric oxide and role in the control
of vascular tone and platelet function. Biochem Pharmacol 1988;
37:2495.

59. Carliner NH, Denune DP, Finch CS, Goldberg LI. Sodium nitro-
prusside treatment of ergotamine-induced peripheral ischemia.
JAMA 1974; 227:308.

60. O'Dell CW, Davis GB, Johnson AD et ah Sodium nitroprusside in
the treatment of ergotism. Radiology 1977; 124:73.

61. McGoon MD, Vlietstra RE, Shub C. Antianginal agents. In:
Giuliani ER, Fuster V, Gersh B, McGoon MD, McGoon DC, eds.
Cardiology Fundamentals and Practice. 2nd edn. St. Louis: Mosby,
1991:730.

62. Tfelt-Hansen P, Ostergaard JR, Gothgen I et ah Nitroglycerin
for ergotism: Experimental studies in vitro and in migraine
patients and treatment of an overt case. Eur J Clin Pharmacol 1982;
22:105.

63. Bhuta I. Acute lower extremity ischemia due to ergotism. / Med
Assoc Ala 1982; 33:28.

64. Brismar B, Somell A, Lockner D. Arterial insufficiency caused by
ergotism: report of a case treated with streptokinase. Acta Chir
Samd 1977; 143:319.

65. Dormandy JA. Clinical experience with iloprost in the treatment
of critical leg ischemia. In: Rubanyi GM, ed. Cardiovascular Signifi-
cance ofEndothelium-Derived Vasoactive Factors. New York: Futura,
1991.

66. Dormandy JA. Critical leg ischemia. In: Clement DL, Shepherd JT,
eds. Vascular Disease in the Limbs: Mechanisms and Principles of
Treatment. St. Louis: Mosby, 1993:91.

67. Levy JM, Ibrabim F, Nykamp PW et ah Prostaglandin El for alle-
viating symptoms of ergot intoxication: a case report. Cardiovasc
Intervent Radiol 1984; 7:28.

68. Dagher FJ, Pais SO, Richards W et ah Severe unilateral ischemia
caused by ergotamine: treatment with nifedipine. Surgery 1985;
97:369.

69. Kemerer VF, Dagher FJ, Osher P. Successful treatment of ergotism
with nifedipine. AJR 1984; 143:333.

70. Shifrin E, Olschwang D, Perel A et ah Reversal of ergotamine
induced arteriospasm by mechanical intra-arterial dilatation.
Lancet 1980; 2:1278.

71. Joyce OA. Arterial complications of migraine treatment
with methysergide and parenteral ergotamine. Br Med J 1982;
285:261.

72. Maples M, Mulherin JL, Harris J et ah Arterial complications of
ergotism. Am Surg 1981; 47:224.

73. Perkin GD. Ischemic lateral popliteal nerve palsy due to ergot
intoxication. J Neurol Neurosurg Psychiatry 1974; 34:1389.

113

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Arteritis

Francis J. Kazmier

Arteritis is a form of vasculitis. Vasculitis is defined as inflam-
mation and necrosis of blood vessels leading to destruction of
vessel walls, producing local bleeding or thrombosis, and a
variable degree of vascular occlusion. 1 Because vessels other
than arteries (including arterioles, veins, and venules) often
are involved in these syndromes, vasculitis is clearly the more
inclusive term.

Our knowledge of the precise cause of most of these syn-
dromes is incomplete, and the pathogenesis often also is in-
completely understood. Immune mechanisms are associated
with some vasculitis syndromes and extrapolated to involve
others. Immune complex-mediated or, less commonly, cell-
mediated immune mechanisms are invoked. The association
of vasculitis with connective tissue diseases, with serologic
abnormalities (circulating immune complexes, hypocomple-
mentemia, hepatitis B antigenemia, mixed cryoglobulins,
monoclonal immunoglobulins) and with drug reactions and
infections is stressed in the literature and supported by ex-
perimental evidence derived from animal models of immune
complex-induced vasculitis. 2

In 1990, the American College of Rheumatology (ACR) re-
classified vasculi tidies and eliminated from statistical analysis
those vasculitis syndromes associated with clearly distin-
guishable features and serology, as is the case with the
vasculitis associated with systemic lupus erythematosus and
rheumatoid arthritis. 3 The search for a specific diagnostic
laboratory test is a continuing effort. 4 Autoantibodies against
neutrophils may be relatively specific for Wegener's granulo-
matosis, but the discovery of antineutrophil cytoplasmic
antibodies in other vasculitis syndromes means clinical pre-
sentation and biopsy cannot yet be replaced. 5-7

The ACR classified seven major types of systemic vasculitis
that are sufficiently different to warrant individual considera-
tion. The actual comparisons among them are based on types
of vessels involved (elastic arteries in Takayasu arteritis), dis-
tribution and localization of disease (predominantly skin in
hypersensitivity vasculitis), type of cell infiltrate (necrotizing
in polyarteritis nodosa), highly specific features (aneurysmal
changes in Takayasu arteritis), and demographics (in tempo-

ral arteritis, all patients are older than 50 years of age, and in
Takayasu arteritis, virtually all are younger than 40 years of
age in the active phase of disease). 3 ' 8

J.T. Lie, a major contributor to the pathology and classifi-
cation of the vasculitidies, made the following additional
observations. 8

1 A negative biopsy does not rule out the presence of vasculi-
tis, because the disease often is segmental in distribution.

2 Mixed cell infiltrates are the rule in vasculitis, and the type of
cell infiltrate is independent of the size of the affected blood
vessels.

3 Despite all attempts to characterize the vasculitidies, there is
continued overlap in the size of blood vessels involved
in the major syndromes, and overlap within individual
syndromes as well.

The 1990 "Magnificent Seven" include polyarteritis nodosa,
Churg-Strauss syndrome, Wegener granulomatosis, hyper-
sensitivity vasculitis, Henoch-Schonlein purpura, giant cell
(temporal) arteritis, and Takayasu arteritis. It is the giant cell
arteritidies, both temporal arteritis and Takayasu arteritis, that
the vascular surgeon and specialist are apt to encounter with
some frequency.

Physicians need clinical, serologic, histopathologic
(biopsy), and often angiographic features to make a diagnosis
of arteritis. Angiograms may be less than specific, as in the
findings associated with drug abuse (ergot), or those associ-
ated with pheochromocytoma. 9 There are in addition several
vasculitis look-alikes, or, as several authors refer to them, mim-
ics. 10 ' 11 Among the most important are endocarditis, cardiac
myxoma, and the multiple cholesterol embolization syn-
drome associated with the so-called shaggy aorta. 10-13 Skin le-
sions may mimic vasculitis, but be unrelated, as, for example,
in the sweet syndrome (neutrophilic dermatosis) character-
ized by fever, neutrophilic polymorphonuclear leukocytosis
of the blood, raised plaques on the face and neck, and a dense
dermal infiltration with mature polymorphonuclear leuko-
cytes. 14 This syndrome responds to steroids in just a few days.
Degos syndrome also may mimic vasculitis and be seen both
with and without associated antiphospholipid antibodies. 15

114

CHAPTER 10 Arteritis

The association of antiphospholipid antibodies with the
presence of skin lesions, such as ulcers and livedo reticularis,
recurrent arterial and venous thrombosis in multiple vascular
beds, and cutaneous necrosis frequently mimics a vasculitis. 16
These antibodies include reagins produced in lues, the lupus
anticoagulant, and anticardiolipin antibodies.

Reagins are important because of the high incidence of false-
positive serologic test results for lues in patients with the lupus
anticoagulant or the antiphospholipid antibody syndrome. In
spite of the clinical mimic, the vascular changes seen with
these antibodies are thrombosis and occlusion of arteries and
veins, with little inflammatory response. 16 Nishino and col-
leagues described five patients who initially were diagnosed
as having giant cell arteritis of the elderly, but in whom the full
clinical expression of Wegener granulomatosis subsequently
developed. 17 All five patients were older than 60 years of age,
had jaw claudication, sudden loss of vision, severe headache
with or without diplopia, or polymyalgia rheumatica at the
time of initial examination. Nongiant cell arteritis was noted
on biopsy of the superficial temporal artery in four of the five.
All five subsequently demonstrated pulmonary and renal
lesions and positive-staining antineutrophil cytoplasmic
antibodies typical of Wegener granulomatosis. This reempha-
sized the overlap among the individual syndromes. Both
polyarteritis and Wegener granulomatosis may involve the
superficial temporal artery. 18 The implications for treatment
are significant, because not all the vasculi tidies respond well to
steroids alone, and the use of steroids alone in polyarteritis or
Wegener granulomatosis can result in bowel infarction due to
active vasculitis, a complication not usually seen in giant cell
arteritis.

Giant cell arteritis and Takayasu arteritis

In both giant cell arteritis (temporal arteritis) of the elderly and
Takayasu arteritis, during the active stage, the pathologic pat-
tern in large vessels is characterized by a granulomatous pan-
arteritis with lymphoplasmacytic infiltrate and giant cells. 18
Both are more common in women, but temporal arteritis is a
disease of the elderly, and Takayasu arteritis is seen in the rela-
tively young. Giant cell (temporal) arteritis is about 10 times
more common than Takayasu arteritis in the community-
based incidence studies in Olmsted County, Minnesota, and
the incidence may be increasing. 19 In contrast to Takayasu
arteritis, giant cell arteritis of the elderly is less common in
non whites. Patients with giant cell arteritis are invariably
older than 50 years. The most common symptoms include
headache, an abnormal superficial temporal artery, jaw
claudication, constitutional symptoms, and polymyalgia
rheumatica. Neurologic findings are not rare and include
amaurosis fugax, visual loss, scotomata, diplopia, peripheral
neuropathies, and cerebral ischemia. 20,21 Visual loss is likely to
be permanent when seen; however, visual loss is rare after

steroid treatment has been started in adequate doses. To avoid
a tragic loss of vision, treatment need not be delayed for
biopsy results.

Laboratory data are nonspecific, but anemia and elevation
in the sedimentation rate are common; in only about 1% of pa-
tients with giant cell arteritis of the elderly is the sedimenta-
tion rate completely normal. 18 Spiera has stressed that an
otherwise unaccountable presentation of the described symp-
toms in an elderly patient should suggest the possibility of
giant cell (temporal) arteritis. 22

Typically, medium-sized arteries are involved in temporal
arteritis, as reflected in symptoms related to the head and neck
arteries. A frequency of involvement of major head and neck
arteries at postmortem examination demonstrates the super-
ficial temporal and vertebral arteries to be diseased in almost
every case. 23 The ophthalmic and posterior ciliary arteries are
involved 75% of the time or more, with the external carotid, its
other branches, and part of the internal carotids and central
retinal arteries showing less frequent severe involvement. In-
volvement always ends near where the arteries cross the dura
mater or enter the substance of the optic nerve. There appears
to be a correlation between involvement and the amount of
elastic tissue in the media and adventitia of the individual
arteries affected. 23

Large arteries are involved in giant cell arteritis of the
elderly only in about 10% of patients. 24 Symptoms include
claudication of an extremity, paresthesias, and the Raynaud
phenomenon. Findings include decreased or absent upper ex-
tremity pulses and bruits over larger arteries. Rarely, the lower
extremity vessels are involved. Large brachiocephalic arterial
involvement can occur as steroids are tapered in classic
temporal arteritis, and on occasion we have seen large vessel
brachiocephalic arteritis as the sole initial presentation of
temporal arteritis.

When the brachiocephalic arteries are involved, angio-
graphic findings are noted from the distal subclavian arteries
extending to the proximal brachial artery 9 Stenosis is seen
most frequently with artery of normal caliber between tapered
stenoses. Occlusive and local aneurysmal changes also are
seen. Findings are bilateral but not necessarily symmetric. An-
giographic changes are not found in the common carotids, the
extracranial internal carotids, the innominate, or the sub-
clavian proximal to the origin of the vertebral artery. When the
lower extremities are involved, distribution favors the deep
and superficial femoral arteries, popliteals, and proximal
tibial artery segments. Long, smooth stenoses with or without
occlusion are typical. 9

Symptoms of brachiocephalic involvement improve
when steroid therapy is introduced, but it is unusual to see
return of absent pulses in the upper extremities in giant cell
arteritis of the elderly. Whether this relates to the frequently
prolonged delay in diagnosis is not entirely clear. Reconstruc-
tive surgery in the upper extremity due to giant cell arteritis
rarely is necessary, but it is needed in a dominant extremity on

115

pa rt I Vascular pathology and physiology

occasion. An intervention should be delayed until the disease
is no longer active or is sufficiently suppressed with steroid
therapy

The treatment for giant cell arteritis of the elderly is steroids,
usually with prednisone orally. Because the risks of prolonged
steroid treatment are higher in the elderly, a definite diagnosis
is appropriate and usually obtained by temporal artery biopsy.
Temporal artery biopsy is not associated with major complica-
tions, but adequate length of artery (often several centimeters)
or biopsy of both arteries may be needed because of the
segmental nature of the pathologic process. 19 Most physicians
use 40-60 mg of prednisone daily initially, with an attempt at
tapering after approximately 4-6 weeks. Should symptoms
recur or the sedimentation rate rise over 50 mm in 1 h, particu-
larly in the absence of another responsible disease process, a
suppressive dose is again instituted for a period of time and
then gradually tapered. In the Olmsted County, Minnesota,
experience, most patients were no longer taking steroids after
1 year, but there are exceptions to this rule. 19 Osteopenia and
steroid myopathy can be troublesome, along with hyper-
tension, glucose intolerance, fluid retention, skin changes, and
persistent gastric hyperacidity.

Death due directly to giant cell arteritis in the elderly is un-
common, but is noted on occasion after aortic dissection with
coronary arteritis, or secondary to stroke. 18

Polymyalgia rheumatica without giant cell arteritis needs
to be differentiated as a clinical syndrome in adults older
than 50 years presenting with aching and stiffness in the
neck, shoulder, or hip girdle for at least 1 month or more. The
sedimentation rate often is elevated, but no other specific
disease is noted. In spite of symptoms directed to muscles,
little is found objectively relating to any muscle disease.
Polymyalgia rheumatica accompanies giant cell arteritis of the
elderly 35-40% of the time. In contrast, the incidence of a posi-
tive superficial temporal artery biopsy is less than 15% in
patients with polymyalgia rheumatica where there are no
associated symptoms of arteritis. Most patients with arteritis
have symptoms in addition to the musculoskeletal ones
that characterize the disease. A response to smaller doses of
prednisone (i.e. 5-15 mg orally per day) is characteristics of
polymyalgia. 22

Takayasu arteritis is a chronic inflammatory disease of
arteries affecting the aorta and its major branches, including
the proximal coronary and renal arteries and the elastic
pulmonary arteries. 25 As mentioned, patients most often are
women, and worldwide more often Asian or North African.

Our own experience in North America with 32 patients with
Takayasu arteritis seen between 1971 and 1983, however,
included 23 North American whites, four Mexicans, three
Orientals, and one of Middle East origin. 25 The actual county
incidence for Olmsted County, Minnesota, is 2.6 per mil-
lion/year, and all of the country residents affected were white
and not related. Although claudication of an upper extremity,
decreased arterial pulses in the upper extremity, differential

brachial blood pressure, and bruits were noted in all 32 pa-
tients, only two of 32 patients had a provisional diagnosis of
Takayasu arteritis. Postural dizziness reflecting severe carotid
or vertebral involvement is common. 26 Visual disturbances
also are common, although classic Takayasu retinopathy usu-
ally is not seen in North American patients. Although pul-
monary artery involvement has been noted to be common in
postmortem series, clinical expression of pulmonary involve-
ment was uncommon in our North American series. The coro-
nary arteries are involved in about 15% of patients with aortic
disease, and osteal lesions are due to inflammation in the prox-
imal coronary artery segments. Hypertension is noted in at
least 50% of patients, and reflects renal artery involvement.
Aortic regurgitation occurs in anywhere from 5% to 20% of
patients, and valve replacement may prove necessary on
occasion.

Takayasu arteritis presents with variable patterns of arterial
involvement. This variable distribution has led to categoriza-
tion according to the anatomy involved. Type I involves the
arch and brachiocephalic vessels; type II involves the descend-
ing thoracic aorta and abdominal aorta (but spares the arch);
and type III is a combination of types I and II. Type IV includes
pulmonary artery involvement. 18

Many authors divide Takayasu arteritis into stages. The first
is an acute inflammatory (prepulseless) stage, characterized
by systemic symptoms; the second is a chronic stage character-
ized by vessel occlusion and vascular insufficiency. 25 It may
take several years for the disease to pass through the full spec-
trum of involvement, and hence the separation may be some-
what arbitrary. In our North American experience, there was a
distinct overlap of the inflammatory stage with vascular oc-
clusions. There may well be a difference in clinical expression
of the disease in different racial groups as well as in different
countries. 25

In the active stage of the inflammatory process, a biopsy is
diagnostic, but as stressed by Lie, Takayasu arteritis cannot be
diagnosed by biopsy alone in its chronic phase with complete
confidence. 18 Angiography is diagnostic, however. Findings
are strikingly similar to those seen with large artery involve-
ment in giant cell arteritis (temporal arteritis), and include
arterial occlusions, stenosis, lumen irregularity, and ectasia or
aneurysm formation. 9

The distribution of typical lesions is key to diagnosis.
Takayasu arteritis usually involves the proximal brachio-
cephalic arteries in addition to their distal segments, the com-
mon carotids, proximal subclavian, and innominate arteries.
The carotid arteries above the bifurcation are not affected. The
aorta is frequently abnormal and may be either ectatic or
frankly aneurysmal. Stenosis or occlusion of the visceral ar-
teries also is frequent, and on occasion the pelvic arteries are
involved. The entire aorta, including lateral films to assess the
visceral arteries, needs to be imaged when Takayasu arteritis is
suspected clinically. This extends to imaging of the head, arm,
and pelvic vessels. 9

116

CHAPTER 10 Arteritis

Treatment may involve medicine, surgery, or both, depend-
ing on disease activity level and the vascular bed affected.
Results with steroid therapy had been difficult to assess in the
past from data available in the literature. 27-29 Whether this re-
lates to an insufficient dose to suppress disease activity or to
stage of disease at treatment is unclear. In Japan it has been
noted that patients with elevated sedimentation rates respond
well to steroids. 30

In our North American experience, 29 patients were treated
with steroids with suppressive doses. 25 Systemic symptoms
improved in a relatively short interval in all of these patients.
Eight of 16 patients with active disease demonstrated return of
pulses. We have been dismayed occasionally, however, when
we elected operative intervention with vascular reconstruc-
tion in a patient with persistent minimal to moderate sedimen-
tation rate elevation, only to find active arteritis involving the
arterial wall at the time of surgery. Three patients from our se-
ries suffered graft occlusion or stenosis at an anastomotic site.
Within 3 months of treatment, one patient had both stenosis of
a renal artery graft and severe stenosis in the angioplasted con-
tralateral renal artery. Sufficient steroid dose suppress both to
symptoms and return the sedimentation rate to normal is im-
portant in planning surgery. Takayasu arteritis often demands
prolonged steroid treatment, but, fortunately, this is better
tolerated in younger than older subjects. Shelhamer and col-
leagues have added cyclophosphamide to the steroid treat-
ment of patients with clinical or angiographic progression of
arteritis, and obtained a good result in most patients. 31
Ishikawa has noted that in patients with severe or multiple
complications, a decreased 6-year survival is seen. 32 Conges-
tive heart failure and stroke were major causes of death. Five-
year survival in our North American experience from the time
of diagnosis was 94%. Early diagnosis and treatment decrease
both mortality and morbidity in Takayasu arteritis.

References

1. Fauci AS, Haynes BF, Katz P. The spectrum of vasculitis: clinical
pathologic, immunologic and therapeutic considerations. Ann
Intern Med 1978; 89:660.

2. Conn DL, McDuffie FC, Holley KE et al. Immunologic mecha-
nisms in systemic vasculitis. Mayo Clin Proc 1976; 51:511.

3. Hunder GG, Arend WP, Bloch DA et al. The American College of
Rheumatology 1990 criteria for the classification of vasculitis:
introduction. Arthritis Rheum 1990; 33:1065.

4. Lie JT. Classification and immunodiagnosis of vasculitis: a new
solution or promises unfulfilled. / Rheumatol 1988; 15:728.

5. Abbot F, Jones S, Lockwood CM, Rees AG. Autoantibodies to
glomerular antigens in patients with Wegener's granulomatosis.
Nephrol Dial Transplant 1989; 4:1.

6. Nolle B, Specks U, Ludenmann J et al. Anticytoplasmic autoanti-
bodies: their immunodiagnostic value in Wegener's granulo-
matosis. Ann Intern Med 1989; 111:28.

7. Lockwood CM, Bakes D, Jones S et al. Association of alkaline
phosphatase with an autoantigen recognized by circulatory

antineutrophil antibodies in systemic vasculitis. Lancet 1987;
1:716.

8. Lie JT Classification criteria and histopathologic specificity
of major vasculitis syndromes. In: Tanabe T, ed. Intractable
Vasculitis Syndrome. Sapporo, Japan, Hokkaido University Press,
1993:17.

9. Stanson AW. Roentgenographic findings in major vasculitic
syndromes. Rheum Dis Clin North Am 1990; 16:293.

10. Byrd W, Matthews O, Hunt R. Left atrial myxoma presenting as a
systemic vasculitis. Arthritis Rheum 1980; 23:240.

11. Lightfoot RW, Jr. Classification of polyarteritis nodosa. In: Tanabe
T, ed. Intractable Vasculitis Syndromes. Sapporo, Japan: Hokkaido
University Press, 1993:167.

12. Kazmier FJ, Hollier LH. The shaggy aorta. Heart Disease and Stroke
1993;2:131.

13. Cappiello R, Espinoza L, Adelman H et al. Cholesterol embolism: a
pseudovasculitic syndrome. Semin Arthritis Rheum 1989; 18:240.

14. Sweet RD. An acute febrile neutrophilic dermatosis. Br J Dermatol
1964; 76:349.

15. Englert H, Hawkes C, Boey M et al. Degos' disease: association
with anti-cardiolipin antibodies and the lupus anticoagulant. Br
Med J 1984; 289:576.

16. Bowles CA. Vasculopathy associated with the antiphospholipid
antibody syndrome. Rheum Dis Clin North Am 1990; 16:471.

17. Nishino H, Remee RA, Rubino FA, Parisi JE. Wegener's granulo-
matosis associated with vasculitis of the temporal artery: report of
5 cases. Mayo Clin Proc 1993; 68:115.

18. Lie JT Diagnostic histopathology of major systemic and pul-
monary vasculitis syndromes. Rheum Dis Clin North Am 1990;
16:259.

19. Hunder GG. Giant cell arteritis. Rheum Dis Clin North Am 1990;
16:399.

20. Caselli RJ, Hunder GG, Whisnant JP Neurologic disease in
biopsy-proven giant cell (temporal) arteritis. Neurology 1988;
38:352.

21. Caselli RJ, Daube JR, Hunder GG, Whisnant JP Peripheral neuro-
pathic syndromes in giant cell (temporal) arteritis. Neurology 1988;
38:685.

22. Spiera H. Polymyalgia rheumatica and cranial arteritis. In: Katz
W, ed. Diagnosis and Management of Rheumatic Diseases. 2nd edn.
Philadelphia: JB Lippincott, 1988:514.

23. Wilkinson IMS, Russell RWR. Arteries of the head and neck in
giant cell arteritis. Arch Neurol 1972; 27:378.

24. Klein RG, Hunder GG, Stanson AW, Sheps SG. Large artery in-
volvement in giant cell (temporal) arteritis. Ann Intern Med 1975;
83:806.

25. Hall S, Barr W, Lie JT, Stanson AW, Kazmier FJ, Hunder GG.
Takayasu arteritis: a study of 32 North American patients.
Medicine 1985; 64:89.

26. Hall S, Buchbinder R. Takayasu's arteritis. Rheum Dis Clin North
Am 1990; 16:411.

27. Ishikawa K. Natural history and classification of occlusive throm-
boaortopathy (Takayasu's disease). Circulation 1978; 57:27.

28. Lupi-Herrela E, Sanchez-Torres G, Marcushamer J, Mispireta J,
Horwitz S, Vela JE. Takayasu's arteritis: clinical study of 107 cases.
Am Heart} 1977; 93:94.

29. Fraga A, Mintz G, Valle L, Flores-Izquardo G. Takayasu's
arteritis: frequency of systemic manifestations (study of 22 pa-
tients) and favorable response to maintenance steroid therapy

117

pa rt I Vascular pathology and physiology

with adrenocorticosteroids (12 patients). Arthritis Rheum 1972; 31. Shelhammer JR, Volkman DJ, Parillo JR et ah Takayasu's arteritis

15:617. and its therapy. Ann Intern Med 1983; 103:121.

30. Ishikawa K, Yorekawa Y. Regression of carotid stenosis after corti- 32. Ishikawa K. Pattern of symptoms and prognosis in occlusive

costeroid therapy in occlusive thromboaortopathy (Takayasu dis- thromboaortopathy (Takayasu's disease). / Am Coll Cardiol 1986;

ease). Stroke 1987; 18:677. 8:1041.

118

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Adventitial cystic disease

Carlos E. Donayre

Adventitial cystic disease is a curious and uncommon disor-
der that causes a localized extraluminal arterial stenosis or ob-
struction. Pathologically, it is characterized by the presence in
the adventitia of multiple cystic spaces containing a viscous
gel accompanied by a surrounding fibrosis. Atkins and Key
described the first case in 1947, in a 40-year-old man with
intermittent claudication and a palpable mass above the
inguinal ligament. At surgery, a cyst was dissected from the
external iliac artery and was described as a " typical ganglion''
in appearance, but depicted as a myxomatous tumor histolog-
ically. 1 Since then, nearly 200 cases have been reported, with
most affecting the midpopliteal artery. It also has been de-
scribed in the common and internal iliac, femoral, radial, and
ulnar arteries. Involvement of adjacent veins is extremely
unusual.

Histology

The first case of adventitial cystic disease of the popliteal
artery was operated on by Hierton in 1953. 2 A 32-year-old man
with intermittent claudication was found to have a popliteal
stenosis on arteriographic evaluation. An enlarged and thick-
ened popliteal artery was encountered at surgery, and two
thimblefuls of a raspberry jelly-like substance were emptied
from an intramural, multilocular cavity. The involved
popliteal artery was resected and a vein graft bypass was per-
formed successfully. Hierton later collected three more cases;
the arteries were described as sausage shaped, and on gross
examination an intramural cyst filled with a clear, thickened
gel was encountered. The popliteal arteries were occluded
owing to compression by the cysts, which were full of gel and
under high pressure. 3

Careful histologic studies have shown that adventitial cysts
initially do not involve the media or intima. Hematoxylin and
eosin (H&E) stain shows the cysts as having a smooth, thin
wall composed of fibroconnective tissue, and lacking an ep-
ithelial lining. The main structural component of the cyst wall
is compact collagen interspersed with strands of elastin fibers.

Basophilic degenerative changes of the fibroconnective tissue
of the cyst wall, similar to that seen in ganglion cysts, also have
been observed. Masson's trichrome stain and elastic stain
show the cysts to be situated in the tunica adventitia, lacking a
communication with the arterial lumen. The rest of the arterial
wall appears normal without any evidence of degenerative,
atherosclerotic, or aneurysmal changes. Fibrosis, necrosis,
calcification, and ulceration of the media, intima, or both occur
only as a result of hemodynamic alterations brought about
by the luminal encroachment of the adventitial cysts as they
grow in size. Alterations typical for a primary dysplasia of the
media, such as cystic degeneration, focal necrosis of collage-
nous or muscular fibers, or destruction of elastic fibers, have
not been observed in the media adjacent to the adventitial
cysts when specimens have been studied by light microscopy.
Electron microscopy of the interior surface of the cyst wall re-
veals a lack of cellular lining or a partial lining several layers
thick of cuboidal cells with round nuclei. Ultrastructurally,
these mesothelium-like cells are in loose connection with
each other, and their luminal surface contains cytoplasmic
pseudopodia and villi. 4

Immunohistochemical examinations of the adventitial cell
lining show a lack of factor VIII. This finding, along with the
lack of basement membrane on electron microscopy, permits
the exclusion of an endothelial nature for the cellular lining of
these cysts. Examinations for keratin also have been negative,
excluding an epidermal origin. Because a specific reaction for
identifying synovial cells does not yet exist, a definitive corre-
lation between adventitial cysts and a synovial origin cannot
be made. 4

When retrieved, the cyst fluid has been subjected to exten-
sive examination. Hierton and associates were the first to per-
form a chemical analysis of the cyst fluid, which showed no
increase of calcium or cholesterol, but abundant fibrinogen
and mucoprotein. 3 The mucoid, gelatinous material found
inside the cysts remained unstained by H&E, but stained as
Prussian blue on Rinehart-Abul-Haj preparation, which is
specific for the presence of mucopolysaccharide. Initially, it
was suggested that the main cyst constituent was mucin, but

119

pa rt I Vascular pathology and physiology

incubation with hyaluronidase turned the thick, viscous gel
into a thin, cloudy fluid, proving the presence of hyaluronic
acid in a high content. 5 Histochemical characterization of the
acid mucopolysaccharide within the cyst has shown it to be
rich in hyaluronic acid radicals. 6 A comparison of the chemical
and physical characteristics of the cyst contents with syno-
vial fluid shows a variety of surprising differences. The
hyaluronidate concentration measured 16.4mg/ml, approxi-
mately five times the value of normal human or bovine syn-
ovial fluid. The adventitial cyst fluid had to be diluted 20-fold
to have the same viscosity, 4.87 cp, as that of undiluted bovine
synovial fluid. On digestion with Streptomyces hyaluronidase,
the viscosity of the adventitial cyst gel decreased 70%. The
high specificity of this enzymatic digestion clearly document-
ed that hyaluronidate, not mucin, is the foremost source of the
viscosity exhibited by the cyst fluid. Also, the coefficient of
friction for the 20-fold dilution of the cyst contents has been
found to be 0.220 jam, 10 times higher than the value for bovine
synovial fluid, indicating the absence of lubricating mucin or
the presence of molecules that disturb the small lubricating
ability of sodium chloride. Unlike human and bovine synovial
fluid, adventitial cyst fluid is unable to lubricate a latex glass
bearing in vitro. Only synovial mucin and its purified lubricat-
ing glycoprotein, lubricin, have this ability 7 Hyaluronic acid,
although viscous like synovial fluid in some respects, does not.
Furthermore, the total protein content of the cyst fluid was
200mg/ml, in contrast to 17.2mg/ml, which is considered
normal for human synovial fluid. 8

Based on histologic and histochemical findings, ganglia and
adventitial cysts appear to be similar. Both have a unilocular-
or multilocular-based structure. The wall of adventitial cysts
exhibits a basophilic degenerative change in the fibroconnec-
tive tissue similar to that seen in a ganglion cyst. Histochemi-
cal staining of the cyst wall and cyst gel demonstrated a rich
content of hyaluronic acid, and a strong Alcian blue positivity,
which indicates that adventitial cysts are more comparable to
ganglia of the wrist than to normal knee synovium when all
three are compared. 9

Etiology

The rare and uncommon nature of this disease has led to a
variety of theories regarding its etiology. The most common
theories are summarized and conclusions are drawn based on
the histologic and histochemical findings described in the
preceding section.

Traumatic origin

The popliteal artery is bound by a fibrous tunnel formed by the
fascia of the deep surface of the gastrocnemius muscle. The
popliteal artery finds itself relatively fixed owing to this

anatomic arrangement, and is submitted to constant bending
or stretching by a mobile knee joint. Hierton and associates
were the first to suggest that repetitive popliteal trauma result-
ed in the degeneration of arterial adventitia with subsequent
cyst formation. 3 Furthermore, in several patients definitive
traumatic events such as a fall, use of a pedal cycle, and repeat-
ed kneeling have preceded the discovery of the cyst. Although
simple and straightforward, the theory of trauma as an etiolo-
gy for this disease seems unlikely. The relatively low incidence
of adventitial cystic disease (1 in 1000 angiograms per-
formed), 10 rare bilateral involvement, and the uncommon
involvement of adjacent veins make this theory highly
questionable. Also, knee dislocations or subluxations tend to
cause intimal flaps, because it is the intima that is the most
fragile component of the arterial wall, not the adventitia. The
ulnar artery, when submitted to repetitive trauma such as in
the hypothenar hammer syndrome, develops mural degener-
ation. Intimal damage results in thrombosis, whereas injury to
the media leads to true arterial aneurysm formation. Adventi-
tial cysts never have been reported when this occupation-
related disorder is encountered. 11,12

Embryologic origin

The popliteal space is the site of several anatomic rearrange-
ments during embryonic life. Thus, the theory that mucin-
secreting cells from adjacent endothelium-derived joint tis-
sues can become incorporated into the adventitia of neighbor-
ing arteries during embryonic development may be easily
supported. With time, these cells would then continue to se-
crete mucin, enlarge, and coalesce, leading to cyst formation.
DeLaurentis and coworkers have also proposed a modified
view that these mucinous cysts are the result of an intrinsic de-
fect at the time of vessel wall formation. 6 Ultrastructural stud-
ies, however, have shown that the lining cells of adventitial
cysts do not have any of the characteristics of endothelial cells,
such as the presence of basement membrane or micropinocy-
totic vesicles. 4 As mentioned, immunohistologic examina-
tions have failed to find any factor VIII, which is specific for
endothelium. The chemical content of the cysts does not re-
semble epithelial secretions, but is rich in hyaluronic acid.

Also, if adventitial cystic disease is the direct result of a con-
genital abnormality, an increased incidence in children would
be expected; however, although cases have been reported in
children, the mean age at presentation is 42 years, with an age
range of 11-70 years. 13 Furthermore, the increased incidence in
men, with a male /female ration of 5 : 1, cannot be explained by
this theory.

Connective tissue disorder

This theory proposes that adventitial cystic disease is a muci-
nous condition associated with a generalized body disorder.

120

chapter 11 Adventitial cystic disease

Adventitial cystic disease has been reported in patients af-
flicted with cutaneous elastic tissue deficiencies, such as the
nail-patella syndrome or hereditary osteoonychodysplasia.
There is a miniscule probability that these two rare disorders
occurred simultaneously by chance alone. Thus, the argument
is made that these two disorders must be etiologically
related. 14 The nail-patella syndrome is a mixed ectodermal
and mesodermal autosomal dominant disorder, with an inci-
dence of 22 per million people. 15 Elastic tissue abnormalities,
as well as increased urinary levels of acid mucopolysaccha-
ride, 16 have been reported in skin biopsy specimens obtained
from a patient with the nail-patella syndrome. 17 If a similar en-
zymatic defect was the etiologic factor for both the nail-
patella syndrome and adventitial cystic disease of the
popliteal artery, symmetrical popliteal artery involvement
would be expected; however, adventitial cystic disease is uni-
focal, unilateral, and recurrences are rare. In addition, adventi-
tial cystic disease is sporadic in its occurrence, and usually is
not seen with any other fibrovascular disorders.

Connective tissue disorders associated with vascular struc-
tures and cyst formation primarily affect the medial muscula-
ture of the aorta and other elastic arteries. The lesions seen in
Erdheim medial necrosis and lathyrism are due to a vacuolar
degeneration of smooth muscle cells, can occur anywhere in
the arterial wall, but are not accompanied at any time by
adventitial cysts. In contrast, light microscopy examinations
have demonstrated repeatedly that adventitial cystic disease
is localized to the adventitial layer, is clearly separated from
the medial musculature, and that the media is built up of well
preserved, normal, smooth muscle cells. Furthermore, arterial
medial necrosis is associated with aneurysm formation, not
the mechanical luminal narrowing and occlusion seen with
adventitial cystic disease. 4

Ectopic ganglia

It also has been postulated that adventitial cysts are true
ganglia that originate from adjacent joint capsules or associat-
ed tendon sheaths. McEvedy, in his thesis on simple ganglia,
advocated that these lesions arose as capsular extensions of
the neighboring joint. 18 The occurrence of adventitial cysts
only in the vicinity of joints, such as the knee, hip, and wrist,
points toward an etiology contingent on articular pathologic
processes.

Cysts containing mucin but involving nerves, not arteries,
have been described only in the popliteal region. 19 ' 20 The find-
ing of intraneural ganglia in the peroneal nerve, arising from
the superior tibiofibular joint, is strong evidence that this also
is the etiology for the formation of adventitial cysts in the
popliteal artery. Histologic examination of these intraneural
cysts reveals a wall composed of fibrous tissue, a lining with
great nuclear proliferation, and a cyst content rich in mucin, all
of which are characteristics typical of a simple ganglion as

well. 18 Furthermore, Parkes also found that some of these
neural ganglia were found to be in direct communication, via a
small pedicle, to the superior tibiofibular joint. 20

Ganglia develop owing to a dysontogenic hyperplasia of
persistent rests of scleroblastema, which may or may not de-
velop a synovial lining. The term Baker cyst usually is reserved
for large, simple cysts that form in the popliteal space and orig-
inate from an adjacent bursa. Both ganglia and the so-called
Baker cysts, however, communicate with joints or tendon
sheaths, and their histologic character is identical. Histologi-
cally, adventitial cysts closely resemble ganglion cysts in the
following ways: (i) they both have a smooth, thin wall of fibro-
connective tissue composed of compact collagen interspersed
with strands of elastin fibers; (ii) both exhibit basophilic de-
generative changes of the cyst wall; and (iii) the cyst lining,
when present, does not resemble endothelium. Furthermore, a
communication between the adventitial cyst and the adjacent
joint, as seen occasionally in ganglion cysts, has been de-
scribed for the popliteal artery, 21,22 radial and ulnar arter-
ies, 20,23 ' 24 and the common femoral artery. 25 Hunt and
colleagues were able to demonstrate with arthrography and
subsequent exploration a direct communication between an
adventitial popliteal cyst and the knee joint. 26 It also has been
postulated that in adventitial cysts lacking a communication
with the adjacent joint, the communication may have become
obliterated, or formed a fibrous band that may be barely
noticeable and easily missed at the time of operative
exploration. 21 The adventitial cyst fluid is similar to the
fluid encountered in ganglia, with a gel-like viscosity and
hyaluronic-rich chemical content comparable in both. 6,8

The point also has been made that a ganglion is a benign le-
sion found more frequently in the second and third decades of
life, whereas adventitial cystic disease occurs in the 30-40-year
age range. 10 Ganglia, however, tend to occur in the wrist,
where they are readily seen, and compress adjacent structures,
causing pain. Adventitial cysts are found more commonly in
the popliteal fossa, where despite enlargement they are not
readily visible or palpable. They are diagnosed only when
claudication symptoms develop, and thus their increased inci-
dence in the fourth and fifth decade can be attributed to a delay
in presentation and diagnosis.

Despite extensive research, the etiology of adventitial cysts
remains a topic of controversy. The most logical explanation is
that adventitial cysts arise from ectopic ganglionic tissue,
which originates from scleroblastema, mesenchymal cells re-
sponsible for the formation of joint capsules, and bursae. Mes-
enchymal stem cells are widely distributed in adult connective
tissues, including granulation tissue, and form hyaluronic
acid when stimulated. The cyst thus grows either slowly, from
accumulation of muciform fluid produced by the mesenchy-
mal cells lining the cyst, or rapidly owing to trauma, cyst rup-
ture, or the presence of a direct communication with an
adjacent joint space.

121

pa rt I Vascular pathology and physiology

Clinical presentation and physical findings

Diagnosis

Because adventitial cystic disease predominantly seems to
affect the popliteal artery, patients usually present with an
abrupt history of calf claudication. Typically, the afflicted pa-
tients are men in 80% cases, with a mean age of 42 years, and a
range of 11-70 years. 13 Their presentation also is unusual in
that atherosclerosis, which might be expected, is either mini-
mal or nonexistent.

Clinically, this disease presents with unilateral exertional
cramping in the calf, which later develops into typical inter-
mittent claudication. The onset may be gradual, but more
often tends to be sudden and of rather short duration, but with
eventual recurrence of symptoms. The abrupt onset is attrib-
uted to cyst rupture or hemorrhage within it, 2 which is accom-
panied by an acute occlusion or severe narrowing of the
popliteal artery. In the usual atherosclerotic claudicant, the
recovery time after exercise is short, and the claudication
distance is constant or tends to improve with exercise. The
claudication seen with adventitial cystic disease has a longer
recovery time, and there is a reduction in claudication distance
with increasing exercise. It is known that the pressure in the
knee during exercise may rise to over 1000 mmHg if an effu-
sion is present. 27 This pressure can then easily force fluid into
an adventitial cyst from the knee joint or adjacent cysts. This el-
evated pressure is then maintained until the knee joint returns
to normal. Rest would then allow for fluid resorption from the
cyst, with a longer rest associated with greater fluid resorption
and decreased luminal narrowing of the popliteal artery. An
increased exercise tolerance would therefore follow a pro-
longed period of rest. 23 The high cyst content of mucopolysac-
charides, with their ability to swell due to passive water
diffusion, also may contribute to adventitial cyst enlargement
and subsequent arterial narrowing.

The physical examination is peculiar for the lack of stigmata
of generalized arterial disease. The femoral pulses almost al-
ways are normal, but the popliteal and pedal pulses are either
diminished or absent at rest, or may disappear after exercise. If
the popliteal vessel is only partially occluded, the pedal pulses
may be present with the knee extended but disappear when
the knee is sharply flexed —the Ishikawa sign. Only rarely is the
cyst palpable in the popliteal fossa; this finding probably de-
pends on the size the popliteal cyst is able to attain. The pres-
ence of a murmur or bruit due to flow disturbances secondary
to popliteal arterial narrowing also has been reported. 28

Although a poststenotic dilation may be present on occa-
sion, aneurysm formation is not encountered with adventitial
cystic disease. Ischemic skin changes, signs of distal emboliza-
tion, or gangrene rarely are seen with this disorder. Ischemic
neuropathy, which may cause paresthesias, burning pain, or a
cool-feeling foot, is seldom present.

A high index of suspicion in the young male claudicant, along
with the proper use of noninvasive tests, can be used to make
the correct diagnosis of adventitial cystic disease. The easy ac-
cess to the popliteal fossa, accompanied by blood flow alter-
ations due to a stenosis of the popliteal artery, makes the
diagnosis of adventitial cystic disease particularly suited to a
variety of noninvasive arterial examinations.

Doppler pressure readings may be normal at rest, but
drop to abnormal levels during continued exercise, such as
walking at a fast pace. Return to normal levels can then be seen
when the patients are allowed to rest. Abnormal pressure
readings reappear not only after repeated, but at lower levels
of exercise. 29

Ultrasonography of the popliteal fossa can demonstrate
the eccentric compression of the arterial lumen, and the pres-
ence of low-level echoes within the gelatinous cyst contents. 30
Ease of test performance, minimal discomfort to the patient,
and the short time required for the examination make ultra-
sonography an ideal screening test for adventitial cystic
disease.

Ultrasonography can be combined with computed tomog-
raphy (CT) to provide a highly specific preoperative diagnosis
of adventitial cystic disease. The CT findings associated with
popliteal adventitial cystic disease include an eccentric
compression of the arterial lumen due to the presence of a thin
wall mass with an enhancing rim, and a cyst whose contents
show no enhancement and exhibit attenuation values inter-
mediate between those of water and muscle. 31 ' 32 Occasionally,
CT is able to make the diagnosis of adventitial cyst disease de-
spite a normal arteriogram. 33 Magnetic resonance imaging,
which avoids the use of ionizing radiation and intravenous
contrast material, can provide definition similar to CT, but at a
higher cost. Its ability to image the popliteal fossa has yet to be
evaluated.

In the past, the diagnosis of adventitial cystic disease rested
on angiographic studies of the popliteal artery. Most of the pa-
tients are found to have a popliteal stenosis on diagnostic arte-
riography, but one-third have a total popliteal occlusion. The
early angiographic findings associated with this disorder are
(i) normal arterial vessels proximal and distal to the popliteal
artery; (ii) minimal stenosis of the popliteal artery without
poststenotic dilation; and (iii) lack of collateral circulation. An
hour-glass deformity is seen when the tapering or stenosis is
concentric in nature, and indicates extrinsic compression. The
classic scimitar sign occurs when there is a smooth tapering
above and below the adventitial cyst at the site of the stenosis
(Fig. 11. 1). 34 Depending on the size of the cyst, the popliteal
artery may be displaced medially, or more commonly laterally.
Furthermore, the stenosis may be missed if lateral views are
not obtained at the time of arteriography. A higher yield can be
obtained if stress views with the knee in flexion are obtained;

122

chapter 11 Adventitial cystic disease

Figure 11.1 Expected arteriographic findings
in adventitial cystic disease. (A) The scimitar sign
is observed when the cyst displaces the popliteal
artery medically or laterally. (B) An hour-glass
sign occurs when the cyst surrounds the artery in
a concentric fashion. (C) Total arterial occlusion
is seen when the cyst is large enough to occlude
the lumen, or narrows it so as to produce
hemodynamic changes conducive to thrombus
formation. (Adapted from Flanigan DP,
Burnham SJ, Goodreau JJ, Bergan JJ. Summary
of cases of adventitial cystic disease of the
popliteal artery. Ann Surg 1 979; 1 89: 1 67.)

this is particularly helpful in patients afflicted by minor cyst
encroachments .

Popliteal vascular entrapment is another cause of calf clau-
dication in the young patient without associated atheroscle-
rotic disease. It differs from adventitial cystic disease in that
the claudication is precipitated by walking, not running, is
more commonly bilateral, and presents clinically with a
greater degree of ischemia. Angiographically, the popliteal
artery is displaced medially, and the presence of poststenotic
dilation or aneurysm formation is frequent. Buerger's disease,
or thromboangiitis obliterans, is another clinical syndrome
that can cause claudication in young male patients. It differs
from adventitial cystic disease in that it is associated with a
strong history of tobacco addition, usually presents with signs
of severe ischemia such as digital gangrene, and is associated
with migratory superficial phlebitis. Angiographically, the
proximal femoral arteries are spared, but the infrageniculate
vessels are extensively diseased, and collateral vessels have a
corkscrew configuration.

Treatment

As in any other disease process, making a correct and early di-
agnosis is essential for a long-term therapeutic success in the
treatment of adventitial cystic disease. The high degree of suc-
cess that has been achieved in the treatment of adventitial dis-
ease of the popliteal artery is well summarized by Flanigan
and colleagues. 13 Of a total group of 98 patients (106 proce-
dures) operated on for adventitial disease of the popliteal
artery, only one patient required a late amputation due to graft
failure. Hierton and Hemingsson further demonstrated the
durable success of autogenous vein grafting for this disorder
by using angiography to document graft patency 27-30 years
after surgery 35 The vein grafts exhibited only widening and

mild iregularities, and there was no sign of recurrent disease in
the adjacent femoral or popliteal arteries.

The status of the popliteal artery at the time of operative
intervention determines the type of vascular reconstruction
performed. If the involved artery is not occluded, and
degeneration of the arterial wall is not present, nonresectional
therapies such as open cyst aspiration or evacuation can be
performed. In a review of the world literature, there were only
six failures in 47 patients treated in this fashion. 13 Because of
the success of simple surgical cyst aspiration, CT-guided per-
cutaneous aspiration has been attempted. Probably owing to
inadequate aspiration with this technique, early cyst recur-
rence has been observed. 36 In the cases in which the adventitial
cyst happens to have an open communication with the adja-
cent joint, early cyst recurrence also can be expected. If the cyst
is drained using an open technique, a possible communication
with the joint should be searched for and ligated if found. It
also has been shown that if the cyst material is too thick to aspi-
rate via the percutaneous route, the cyst can be drained if mul-
tiple punctures are performed. 37

The use of local angioplasties with either autologous vein or
synthetic material to bolster the involved artery is not recom-
mended. If the popliteal artery requires partial resection or
seems to need some kind of reinforcement, it probably has suf-
fered substantial injury and should be excised. A 20% recur-
rence rate, as reported by Flanigan and colleagues, can be
expected when patch angioplasties are used to treat adventi-
tial cystic disease. 13

Percutaneous transluminal angioplasty (PTA), although
successful in the treatment of short-segment stenosing lesions
in the arterial tree, has failed to treat adventitial cystic disease
with durable success. 38 Vessels with adventitial cystic disease
usually are void of atherosclerotic changes, and are young in
age. As a result, they are more compliant and muscular, and
are more likely to resume their original stenotic configuration

123

pa rt I Vascular pathology and physiology

after balloon dilation. Even if the cyst material is extruded dur-
ing PTA, recurrences are likely because the cyst will continue
to secrete or accumulate fluid if a communication with the
joint space is present. Balloon angioplasty dilates stenotic ar-
teries by cracking the intima and adjacent media. In adventi-
tial cystic disease, the lesion resides in the outer wall of the
vessel, not in the layers affected by balloon therapy.

When total occlusion of the popliteal artery is encountered,
replacement of the artery or bypass yields the best results.
Autologous vein grafts should be used because of their
proven long-term patency compared with synthetic grafts. A
95% success rate can be expected when vein graft interposition
or primary end-to-end arterial anastomosis is used to treat this
condition. 13 As mentioned, patency of 30 years can be seen
when vein graft interposition is used. 35

In the patient with adventitial cystic disease and total occlu-
sion of the popliteal artery, urokinase lytic therapy has been
used to treat the thrombosis and unmask the underlying
pathologic process. 39 If the intima is spared, and no evidence
of atherosclerotic change is seen after thrombolysis, nonresec-
tional adventitial cystotomy seems reasonable; otherwise, ar-
terial reconstruction should be performed.

A high success rate can be achieved in the treatment of this
fascinating disorder if the proper procedure is performed.
Drainage of the cyst and careful dissection to identify any con-
nection with the adjacent joint space should result in few re-
currences. Increased awareness of the existence of adventitial
cystic disease, with its peculiar clinical presentation, should
lead to a more frequent diagnosis and a clearer understanding
of its etiology.

References

1. Atkins HJB, Key J A. A case of myxomatous tumor arising in the
adventitia of the left external artery. Br J Surg 1947; 34:217.

2. Ejrup B, Hierton T. Intermittent claudication: three cases treated
by free vein graft. Acta Chir Scand 1954; 108:217.

3. Hierton T, Lindberg K, Rob C. Cystic degeneration of the popliteal
artery. Br J Surg 1957; 44:348.

4. Leu HJ, Largiader J, Odermatt B. Pathogenesis of the so-called ad-
ventitial degeneration of peripheral blood vessels. Virchows Arch
[A] 1984; 404:289.

5. Hierton T, Lindberg K. Cystic adventitial degeneration of the
popliteal artery. Acta Chir Scand 1957; 113:72.

6. DeLaurentis DA, Wolferth CC Jr, Wolf FM et at. Mucinous adventi-
tial cysts of the popliteal artery in an 11 year old girl. Surgery 1973;
74:456.

7. Swann DA, Silver FH, Slayter HS, Stafford W. The molecular struc-
ture and lubricating activity of lubricin isolated from bovine and
human synovial fluids. Biochem J 1985; 225:195.

8. Jay GD, Ross FL, Mason RA, Giron F. Clinical and chemical char-
acterization of an adventitial cyst. / Vase Surg 1989; 9:448.

9. diMarzo L, Peetz DJ, Bewtra C, Schultz RD, Feldhaus RJ, Anthone
G. Cystic adventitial degeneration of the femoral artery: is evacu-

ation and cyst excision worthwhile as a definitive therapy?
Surgery 1987; 101:587.

10. Lewis GJT, Douglas DM, Reid W, Watts JK. Cystic adventitial dis-
ease of the popliteal artery. Br Med J 1967; 3:411.

11 . Conn J Jr, Bergan JJ, Bell JL. Hypothenar hammer syndrome: post-
traumatic digital ischemia. Surgery 1970; 68:1122.

12. Vayssairat M, Debure C, Cormier JM. Hypothenar hammer syn-
drome: seventeen cases with long term follow-up. / Vase Surg 1987;
5:838.

13. Flanigan DP. Burnham SJ, Goodreau JJ, Bergan JJ. Summary of
cases of adventitial cystic disease of the popliteal artery. Ann Surg
1979; 189:165.

14. Mark TM, Rywlin AM, Unger H. Cystic adventitial degener-
ation of the popliteal artery: its occurrence in a patient with
the nail-patella syndrome. Arch Pathol Lab Med 1983; 107:
186.

15. RenwickJH. Nail-patella syndrome: evidence for modification by
alleles at the main locus. Ann Hum Genet 1956; 21:159.

16. Lorincz AE. Urinary acid mucopolysaccharide in hereditary
arihrodysplasia. South Med J 1960; 53:1588.

17. Gibbs RC, Berczeller PH, Hyman AB. Nail-patella-elbow syn-
drome. Arch Dermatol 1964; 89:194.

18. McEvedy BV. Simple ganglia. Br J Surg 1962; 49:585.

19. Clark K. Ganglion of the lateral popliteal nerve. / Bone Joint Surg
1961;43B:778.

20. ParkesA. Intraneural ganglion of the lateral popliteal nerve. J Bone
Joint Surg 1961; 43B:784.

21. Shute K, Rothne NG. The aetiology of cystic arterial disease. Br J
Surg 1973; 60:397.

22. Lassonde J, Laurendeau F. Cystic adventitial disease of the
popliteal artery: clinical aspects and etiology. Am Surg 1982;
48:341.

23. Durham JR, Mclntyre KE Jr. Adventitial cystic disease of the
popliteal artery. / Cardiovasc Surg 1989; 30:517.

24. Absoud E. Recurrent cystic adventitial disease of the radial artery.
Angiology 1984; 35:257.

25. Campbell WB, Millar AW. Cystic adventitial disease of the com-
mon femoral artery communicating with the hip joint. Br J Surg
1985; 72:537.

26. Hunt BP, Harrington MG, Goode JJ, Galloway JMD. Cystic adven-
titial disease of the popliteal artery. Br J Surg 1980; 67:811 .

27. Jayson MIV, Dixon A. Intra-articular pressure in rheumatoid
arthritis of the knee: pressure changes during joint use. Ann Rheum
Dis 1970; 29:401.

28. Eastcott HHG. Cystic myxomatous degeneration of the popliteal
artery. Br Med J 1963; 2:1270.

29. Schoolhorm J, Arnolds B, von Reuten GM, Schlosser V. Cystic
adventitial degeneration as a cause of dynamic stenosis of the
popliteal artery: a case report. Angiology 1985; 36:809.

30. Bunker SR, Laufen GJ, Hutton JE Jr. Cystic adventitial disease of
the popliteal artery. AJR 1981; 136:1209.

31. Wilbur AC, Woelfel GF, Meyer JP, Flanigan DP, Spigos DG.
Adventitial cystic disease of the popliteal artery. Radiology 1985;
155:63.

32. Fitzjohn TP, White FE, Loose HW, Proud G. Computed tomogra-
phy and sonography of cystic adventitial disease. Br J Radiol 1986;
59:933.

33. Rizzo RJ, Flinn WR, Yao JST, McCarthy WJ, Vogelzang RL, Pearce

124

chapter 11 Adventitial cystic disease

WH. Computed tomography for evaluation of arterial disease in
the popliteal fossa. / Vase Surg 1990; 11:112.

34. Velasquez G, Zollikofer C, Hrudaya PN et ah Cystic arterial adven-
titial degeneration. Radiology 1980; 134:19.

35. Hierton T, Hemingsson A. The autogenous vein graft as popliteal
artery substitute: long term follow-up of cystic adventitial degen-
eration. Acta Chir Scand 1984; 150:377.

36. Sieurine K, Lawrence-Brown MM, Kelsey P. Adventitial cystic dis-
ease of the popliteal artery: early recurrence after CT guided per-
cutaneous aspiration. / Cardiovasc Surg 1991; 32:702.

37. Wilbur AC, Spigos DG. Adventitial cyst of the popliteal artery:
CT-guided percutaneous aspiration. / Comput Assist Tomogr 1986;
10:161.

38. Fox RL, Kahn M, Adler J et ah Adventitial cystic disease of the
popliteal artery: failure of percutaneous transluminal angioplasty
as a therapeutic modality. / Vase Surg 1985; 2:464.

39. Samson RH, Willis PD. Popliteal artery occlusion caused by cystic
adventitial disease: successful treament by urokinase followed by
nonresectional cystotomy. / Vase Surg 1990; 12:591.

125

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Entrapment syndromes

Carlos E. Donayre

The abnormal muscular compression of an artery can lead to
the development of symptoms that mimic atherosclerotic vas-
cular disease. The presence of intermittent claudication in a
young patient with an otherwise normal vascular system is
usually associated with an entrapment of the popliteal artery as it
traverses the popliteal fossa. Impingement of the popliteal
artery by the medial head of the gastrocnemius muscle is
the most common form of arterial entrapment in the lower
extremity, but compression of the distal superficial femoral
artery as it exits the adductor canal also has been described. 1 A
similar form of entrapment also can be acquired after a below-
knee femoropopliteal bypass, when the vein graft is placed
medial to the medial head of the gastrocnemius muscle,
making the graft vulnerable to compression when the patient
dorsiflexes their foot. 2

Comprehension of the embryology and anatomy of the
popliteal fossa has been essential to the understanding of the
pathophysiologic process and treatment of this condition.
Furthermore, the association of a variety of anatomic anoma-
lies with this peculiar vascular syndrome has attracted global
attention and has led to numerous reports in the surgical liter-
ature. As a result, vascular entrapment of the lower extremity
is being recognized with greater frequency as one of the causes
of intermittent claudication and threatened limb loss in young
people.

Historical background

In 1879, T.P. Anderson Stuart, a medical student at the Univer-
sity of Edinburgh, described the popliteal artery coursing
around and deep to the medial head of the gastrocnemius
muscle after dissecting the amputated leg of a 64-year-old
man. 3 He also noted that there were aneurysmal changes in the
popliteal artery distal to the point of external compression.
This anomaly had not been recorded previously, and later
became associated with the popliteal artery entrapment
syndrome.

Chamberdel-Dubreuil in 1925 reported a case in the French
literature in which the popliteal artery followed a normal
anatomic course but was separated from the popliteal vein by
an accessory bundle of the gastrocnemius muscle. 4 The first
operative correction was not reported until 1959 by Hamming
at Ley den University in the Netherlands. He described the
same anomaly reported by Stuart 80 years earlier, but this time
it was in a 12-year-old boy with intermittent claudication. 5
Hamming transected the gastrocnemius and performed a
popliteal artery thromboendarterectomy. His case is signifi-
cant in that it illustrated the first successful surgical treatment
of complications arising from popliteal artery entrapment. He
continued to look for the presence of this syndrome, and in
1965 he and Vink reported on four more cases, giving him the
largest personal experience with this anomaly. 6

Servello, at the University of Padua, described in 1962 a clin-
ical case similar to that of Hamming's in a 28-year-old Italian
farmer. He also found a small aneurysm of the popliteal artery
distal to the area of compression, analogous to the one de-
scribed by Stuart in his original report. 7 In 1965, Love and
Whelan at the Walter Reed General Hospital in Washington,
DC, reported two more cases, and were the first to use the term
popliteal artery entrapment syndrome? Insua and colleagues, in
1970, collected all 17 cases of popliteal artery entrapment
syndrome present in the world literature, and added two of
their own. 9 They theorized that entrapment of the popliteal
artery was probably more common than the few reports in
the literature seemed to indicate, and during a spirited dis-
cussion of their paper, eight more cases were added by the
audience! In the United States this condition has been de-
scribed in greater frequency among military personnel. In
1989, Collins and associates identified 20 extremities with
popliteal vascular entrapment using noninvasive screen-
ing methods followed by diagnostic arteriography at the
Letterman Army Medical Center. 10 There have been over 200
reported cases of popliteal artery entrapment, and many
more are sure to follow.

126

chapter 12 Entrapment syndromes

Embryology

In the simplest terms, the popliteal artery entrapment syn-
drome consists of an anomalous course of the popliteal artery,
which loops around and passes medially to the inner head of
the gastrocnemius muscle. The popliteal artery is then subject
to entrapment and compression between the origin of the me-
dial head of the gastrocnemius muscle and the back of the
medial femoral epicondyle. This abnormal condition thus can
be classified as an embryologic disorder.

In the 5.5-week embryo, the muscular blastema that gives
rise to the gastrocnemius muscle originates from the calca-
neus, grows in a lateral and cephalad direction, and divides
into a lateral and medial head. The lateral head attaches itself
first to the lateral femoral epicondyle. The medial head lags be-
hind, crosses the midline posterior to the popliteus muscle,
and inserts much higher than the lateral head into the medial
femoral epicondyle. 11

At about the same time, the popliteal artery also develops.
The formation of the popliteal artery is a complex process that
occurs from the union of two embryonic vessels: the deep
popliteal artery, which is the terminal end of the sciatic artery,
and the later developing superficial popliteal artery. The part of
the deep popliteal artery located anteriorly to the popliteus
muscle atrophies. The superficial popliteal artery develops
posterior to the popliteus muscle and unites with the remain-
ing deep popliteal artery at the knee to form the anatomically
correct popliteal artery. 12

Because the cephalad migration of the medial head of the
gastrocnemius muscle occurs at the time the popliteal artery is
being formed, anomalous entrapment of the popliteal artery
may arise either because of early migration of the medial head
of the gastrocnemius muscle, or a late development of the
popliteal artery. The popliteal artery would then be swept me-
dially and impinge against the femur in either of these cases.
Only two cases of entrapment of the popliteal artery due to
compression from the lateral head of the gastrocnemius mus-
cle have been described. 13 ' 14 This is probably because the
lateral head attaches itself to the lateral femoral epicondyle at
the 20-mm stage of embryonic growth well before the
popliteal artery changes from the anterior to the posterior sur-
face of the popliteus muscle. 15

Variation in the development of the popliteal artery also
can lead to entrapment by the popliteus muscle, as has been
described by Love and Whelan. 8 If the superficial popliteal
artery fails to develop, the deep popliteal must persist to sup-
ply blood to the lower leg. The popliteal artery would then
develop anterior to the popliteal muscle, and could be sub-
jected to muscular entrapment.

The popliteal vein is the last vascular structure to develop,
and normally would not be available for entrapment at the
time the medial head of the gastrocnemius muscle attaches to
the femur. Reports of entrapment of the popliteal artery and

vein have been sparse. Rich and Hughes were the first to report
this unusual form of entrapment in 1967. 16 Exploration of the
popliteal fossa in a 47-year-old man with acute left leg is-
chemia demonstrated the typical abnormal lateral attachment
of the medial head of the gastrocnemius muscle, with the dif-
ference that both popliteal artery and vein were compressed.
Distally, the popliteal artery contained a thrombosed post-
stenotic aneurysm, and the vein also had an irregular saccular
dilation. 16 Disruption of the cephalad migration of the medial
head of the gastrocnemius muscle leads to the many variations
and degrees of entrapment described in the literature, due
mainly to the diverse ways in which the muscle attaches to the
femur. Abroad attachment forces the popliteal artery to pierce
the medial head of gastrocnemius. 8 ' 17 Another variation is a
third head arising from the normal medial head of the gastroc-
nemius (bifid head) and attaching laterally to the normally
positioned popliteal artery and vein. 18 The accessory head
probably results from a congenital growth of excess muscle. It
also must be associated with a late attachment to the femur be-
cause it always encircles the popliteal vein, which is the last
vascular structure to form in the popliteal fossa. Although ex-
tremely rare, isolated popliteal vein entrapment has also been
reported. 16-19

Fibrous bands arising from the medial or lateral head of the
gastrocnemius muscle also have been described. The popliteal
artery lies in its normal position but is subjected to compres-
sion by an extrinsic fibrous band. 20 ' 21 These bands are similar
to the compressive bands that have been described in the
upper extremity with thoracic outlet syndrome.

Anatomy

The diversity of anatomic anomalies associated with entrap-
ment of the popliteal vessels has led to a variety of classifica-
tion schemes. Insua and associates, in 1970, were the first to
propose an anatomic classification based on review of 19
cases. 9 Just 1 year later, Delaney and Gonzalez proposed a
modified classification comprising four types of anatomic
compressions. 22 Rich and associates added a fifth type in 1979,
and this classification is the one quoted most often in the
literature. 23

Anatomically, the femoral artery becomes the popliteal
artery at the tendinous opening of the adductor magnus.
The popliteal artery then courses distally between the
lateral and medial heads of the gastrocnemius muscle, and
terminates at the distal border of the popliteus muscle
by dividing into the anterior and tibioperoneal arteries
(Fig. 12.1).

There are five types of anomalies associated with entrap-
ment of the popliteal artery, and most are related to an abnor-
mal configuration of the gastrocnemius muscle. A complete
classification remains elusive owing to the many anatomic
variants that continue to be reported.

127

pa rt I Vascular pathology and physiology

Popliteal vein
Popliteal artery

Gastrocnemius muscle

Tibial nerve

Figure 12.1 Normal anatomy. Popliteal artery and vein course between the
lateral and medial heads of the gastrocnemius muscle.

Type I: The medial head of the gastrocnemius attaches its
normal insertion to the medial femoral epicondyle. The
popliteal artery, however, instead of passing between the
two heads of the gastrocnemius muscle, passes medial and
deep to the medial head of the gastrocnemius muscle. This is
the anomaly described by Stuart in 1879 (Fig. 12.2). 3
Type II: The medial head of the gastrocnemius attaches itself to
the femur in a more lateral position. The course of the
popliteal artery is straighter than in type I, but still passes
medial and deep to the medial head of the gastrocnemius
muscle (Fig. 12.3).
Type III: The popliteal artery descends in a relatively straight
course, but is compressed by an accessory bundle of muscle
from the medial head of the gastrocnemius. This third (bifid)
head inserts on the femur in a more lateral position (Fig.
12.4).
Type IV: The popliteal artery courses deep to, and is com-
pressed by the popliteus muscle or by a fibrous band in the
same location. The artery in this type may or may not pass
medially to the medial head of the gastrocnemius, as in type
I (Fig. 12.5).
Type V: Both popliteal artery and vein are entrapped by any of
the types of compression described (Fig. 12.6).
Other rare variations and degrees of popliteal entrapment
also exist, such as hypertrophy of the gastrocnemius, plan-
taris, or semimembranosus muscle in highly trained and
athletic people. 24 ' 25 Acute popliteal vascular entrapment sec-
ondary to blunt trauma resulting in massive swelling of the

Popliteal vein
Popliteal artery

Gastrocnemius muscle

Figure 12.2 Type I anomaly. The popliteal artery courses medially and
posteriorly to the normally attached medial head of the gastrocnemius
muscle.

gastrocnemius muscle also has been described. 26 In this report,
surgical exploration revealed that the popliteal artery and vein
had a normal anatomic relationship to the muscle insertions,
but that both were compressed by an edematous gastrocne-
mius muscle.

Another unusual type of acute vascular entrapment of the
lower extremity that also tends to occur in younger men is the
adductor canal syndrome. 1 The adductor canal (Hunter's
canal) is an aponeurotic tunnel in the middle third of the thigh,
bounded by the vastus medialis anteriorly and laterally, and
by the adductor longus and magnus posteriorly. The canal is
then surrounded by a strong aponeurosis that extends from
the vastus medialis across the femoral vessels to the adductors
longus and magnus. Hypertrophy of this aponeurosis, or
development of tendinous bands at the outlet of the adductor
canal, cause a scissors-like compression of the femoral vessels
by the vastus medialis and the adductor magnus. Repeated
extrinsic trauma to the femoral artery by bands at Hunter's
canal can give rise to an arterial intimal tear and proximal
thrombosis. 27-29

Clinical presentation

The young, athletic patient who presents with symptoms of in-

128

Popliteal vein
Popliteal artery

Gastrocnemius muscle

Popliteal vein
Popliteal artery

Accessory slip

Gastrocnemius muscle

Figure 12.3 Type II anomaly. The medial head of the gastrocnemius
attaches to the femur in a more lateral position. The popliteal artery courses
through the medial head of the gastrocnemius or medial to it.

Figure 12.4 Type III anomaly. The popliteal artery descends in a relatively
straight course but is compressed by an accessory bundle of muscle from the
medial head of the gastrocnemius muscle.

Popliteal vein
Popliteal artery

Popliteus muscle

Gastrocnemius muscle

Popliteal vein
Popliteal artery

Gastrocnemius muscle

Figure 12.5 Type IV anomaly. Popliteal artery courses deep to the popliteus
muscle or a fibrous band in the same location.

Figure 12.6 TypeVanomaly. Both popliteal artery and vein course medial
to and are compressed by the medial head of the gastrocnemius muscle.

pa rt I Vascular pathology and physiology

termittent claudication or an acute occlusion of the popliteal
artery should be suspected of having an associated vascular
entrapment. It tends to occur more frequently in men, in a ratio
of 5 : 1 . In the past this was attributed to increased muscular de-
velopment in men. 24 Also, interest in this disorder has always
been high among military personnel, who traditionally have
been predominantly male. Now that more women are in-
volved in strenuous physical activities and are seeking enroll-
ment in the armed forces, however, the incidence of lower
extremity vascular entrapment in women is bound to increase.

Popliteal artery entrapment usually becomes symptomatic
before 30 years of age. Hamming and Vink claim that the inci-
dence of this syndrome is 40% in patients younger than 30
years with foot and calf claudication. 6 Review of 150 cases
showed the mean age at presentation to be 28 years, with 68%
of the patients younger than 35 years. 30 The diagnosis of vas-
cular entrapment is uncommon in patients younger than 12
years, and older than 50 years. Only 10% of reported cases in-
volve patients older than 50 years, but it is possible that in this
age group symptoms of claudication or the presence of a
popliteal aneurysm are erroneously attributed to atheroscle-
rotic disease.

The clinical presentation of patients with vascular
entrapment will vary depending on the pathophysiologic
status of the popliteal vessels. Symptoms range from a
bothersome, mild claudication to acute ischemia with limb
threat.

Symptomatic patients present with an unusual history of
sudden, rather than gradual onset of intermittent claudica-
tion. Furthermore, the symptoms frequently are associated
with some type of strenuous physical activity. The claudica-
tion is peculiar in that it can be elicited by walking, and not by
running. 31 It has been postulated that the running gait allows
the knees to remain in a flexed position, whereas walking
forces knee extension and a sustained plantar flexion with sub-
sequent impingement on the popliteal artery by the contract-
ing gastrocnemius muscle. 32 The claudication is also unusual
in that it may begin with the first steps rather than after walk-
ing a finite distance, can be precipitated by climbing stairs, and
involves the foot as well as the calf. These uncommon claudi-
cation characteristics may be related to the fact that the ob-
struction of arterial flow is at first mechanical in nature and not
due to atherosclerotic luminal narrowing. Also, because the
level of vascular obstruction is at a different location than that
usually seen with atherosclerotic claudication, symptoms at
time of presentation are atypical. In the vascular entrapment
syndrome, the popliteal artery is compressed at its distal
segment, which allows greater collateral flow from the
suprageniculate to the infrageniculate popliteal branches.
Claudication symptoms secondary to atherosclerosis most
commonly result from stenosis or occlusion of the superficial
femoral artery at the level of the adductor canal, and thus flow
to the suprageniculate arterial arcade is diminished.

Patients may present with acute limb ischemia due to

popliteal artery thrombosis. Muscular entrapment of the
popliteal artery leads to repetitive mechanical trauma to the
vessel wall. The delicate intima tears, the media undergoes
premature and localized atherosclerosis, and as the vessel
lumen narrows, the incidence of thrombosis increases. The sta-
tus of the collateral circulation through the geniculate arteries
at the time of popliteal obstruction determines the severity of
limb ischemia at time of presentation. In a review of 13 cases of
popliteal entrapment by Brightmore and Smellie, only two pa-
tients presented with acute limb ischemia. 33 Angiography
showed a filling defect of the popliteal artery and poor distal
runoff in one, and a popliteal aneurysm with incomplete
thrombosis in the other. Both of these patients had patent
popliteal arteries, but of the remaining 11 patients, all of whom
presented with intermittent claudication, seven had an oc-
cluded popliteal artery at time of angiography. Both of the
patients treated surgically by Brightmore and Smellie gave a
history of acute limb ischemia. The first patient described a
history of his right lower leg suddenly becoming white, cold,
and numb but returning to normal a few hours later. There-
after, he suffered from intermittent claudication of his calf
when walking. The second patient also had sudden pain in his
right lower leg and foot that was followed by numbness and
coolness in the same region, but resolved over a few days.
Seven years later he presented again in similar fashion, but this
time the ischemic symptoms persisted. In these patients, the
history is consistent with an acute occlusion of the popliteal
artery followed by transient ischemia. Because the adjacent
arteries are normal, flow to the distal lower limb can be
augmented quickly after popliteal occlusion by relying on a
rich geniculate collateral system.

Increased turbulence in the popliteal artery distal to the site
of muscular compression can lead to poststenotic dilation and
even true aneurysm formation. 34 Histologic examinations of
abnormally dilated vessels reveal thinning of the arterial wall,
destruction of the internal elastic lamina, and thrombus for-
mation. 23 ' 25 These changes are more prominent on the side of
the vessel wall that is in contact with the rigid and unyielding
femoral surface. Accumulation of thrombus in the abnormally
dilated popliteal artery results in either total arterial occlu-
sion or distal embolization to the infrageniculate or pedal
arteries. 31,36 It is distal embolization to the outflow vessels
that probably is responsible for the development of ischemic
symptoms long after popliteal arterial occlusion occurs. Distal
embolization of thrombi results in the occlusion of outflow
vessels, which are critical to the maintenance of collateral flow.
Furthermore, loss of outflow vessels may compromise the
results of arterial reconstruction undertaken for the treatment
of popliteal vascular entrapment.

The incidence of bilateral disease is approximately 30%,
but often is asymptomatic on one side. Bouhoutsos and
Daskalakis 24 found 33 patients with vascular compression in
the popliteal fossa, 12 of whom had bilateral popliteal involve-
ment. The widespread knowledge of this syndrome, coupled

130

chapter 12 Entrapment syndromes

with a broader use of noninvasive diagnostic techniques, will
undoubtedly unmask asymptomatic cases with greater
frequency

Entrapment of both popliteal artery and vein is seen in
about 10% of cases, but, as mentioned, involvement of the
popliteal vein alone also can occur. 16 ' 18 ' 24 Young patients pre-
senting with leg swelling associated with strenuous exercise,
recurrent popliteal vein thrombosis, or varicosities of the
popliteal fossa should be suspected of having vascular entrap-
ment and should be properly evaluated.

The older patient who has undergone an infragenicular
femoropopliteal bypass but presents with recurrent symp-
toms shortly after operative intervention could have an ac-
quired or iatrogenic form of popliteal entrapment. 37-40 If the
vein used for the arterial bypass is placed medially and super-
ficially to the medial head of the gastrocnemius, and excessive
angulation is used to enter the popliteal fossa, the type I
anomaly of popliteal entrapment is recreated. Obliteration of
distal pulses with knee extension confirms the presence of
iatrogenic entrapment if thrombosis of the graft has not
occurred at the time of presentation.

From this discussion it is evident that the clinical features
and presentations seen in conjunction with the vascular
entrapment syndromes of the lower extremity are extremely
diverse, and are related to the anatomic anomaly encountered.
This is further modulated by the degree of disease within the
afflicted vessels, and the status of the collateral circulation.

Diagnosis

As with any other disorder, a careful and appropriate history is
essential to the early diagnosis of vascular entrapment of the
lower limb. The most remarkable finding on physical exa-
mination is that of reduced or absent pulses in the distal
extremity of a young, physically active patient, who has
an otherwise normal vascular examination. The absence of
any risk factors for early atherosclerotic disease should lead
the astute clinician to the diagnosis of popliteal vascular
entrapment.

The physical findings may be nonspecific in the patient who
presents with claudication alone, and normal resting popliteal
and pedal pulses may be present. 32 Maneuvers that tighten the
gastrocnemius muscle overriding the popliteal artery, such as
knee extension, dorsiflexion of ankle, or active plantar flexion
against resistance may reduce or abolish pedal pulses. The
asymptomatic contralateral extremity also should be exam-
ined carefully, using the same maneuvers, because bilaterality
in this disorder is common. Some normal people, however,
also lose their pedal pulses with these maneuvers, and a
false-negative test occurs in afflicted patients with total occlu-
sion of the popliteal artery if collateral pathways are well
established. 25 ' 41

Auscultation of the popliteal fossa can on occasion demon-

strate a systolic bruit if the artery is compressed but not
occluded. Finding a popliteal aneurysm on palpation of the
fossa in a young person also is highly suggestive for the pre-
sence of vascular entrapment. It has been suggested that in
some elderly patients with popliteal aneurysms, the cause of
the aneurysm may not be atherosclerotic but the end-result
of years of unrecognized popliteal entrapment. 10 Distal em-
bolization from the thrombus of a popliteal aneurysm due to
entrapment usually is to the outflow vessels, with resultant
calf claudication. Blue toe syndrome, characterized by small
scattered areas of painful, bluish discoloration of the toes
due to atheromatous microemboli, is rarely encountered on
physical examination in the patient with lower limb vascular
entrapment. Equally rare is the presence of digital gangrene
on initial evaluation.

Calf and foot paresthesias are common complaints of
patients with popliteal entrapment, but decreased cutaneous
sensation in the tibial nerve distribution seldom is reported in
these patients, presumably because of underdiagnosis. The
tibial nerve travels in close proximity to the popliteal artery,
and probably is subjected to varying degrees of compression
in this syndrome. 42

Noninvasive vascular diagnosis of popliteal entrapment
has varied with the tests available to various investigators. The
tests have ranged from changes in the magnitude of the oscil-
lometry pulse wave during forced plantar flexion, used by
Servello in 1962/ to magnetic resonance imaging (MRI). 43/44

Hand-held Doppler ultrasound, which is readily available,
should be used to assess the status of the posterior tibial and
dorsalis pedis arteries, and this can be supplemented easily
with an ankle-brachial index (ABI) determination. Patients
with an ABI below 1 .0 must be suspected of having an arterial
occlusive lesion. 45 If the ABI is normal at rest, the patient can be
challenged with a physically strenuous treadmill test. The
patient is place on a treadmill set at 6.8 km/h and a 10% grade
for 10 min or until claudication develops. Using this approach,
20 limbs in 12 patients were identified prospectively by
Collins and associates as having an abnormal extrinsic com-
pression or occlusion of the popliteal artery 10 Eight of the
patients had bilateral involvement of the popliteal artery, but
only three of them were symptomatic in both legs. This is an
excellent technique for screening groups of patients at risk.

Duplex ultrasonography also has been used to document
changes in popliteal blood flow with stress maneuvers, and to
confirm the presence of poststenotic dilation or aneurysmal
formation. 46 This technique, however, has been found at times
to provide false-positive results in normal individuals. 10

To minimize the false-positive rate seen with these tech-
niques, computed tomography (CT) and MRI have been
evaluated as diagnostic alternatives. CT can show the altered
position of the popliteal artery in relation to its accompanying
vein, its abnormal relationship to the surrounding muscles,
and the presence of thrombosed vessels not imaged on an-
giography. 35 MRI has proved to be a more specific technique

131

pa rt I Vascular pathology and physiology

because it can define the morphology external to the affected
vessels by relying on a diversity of planes, coronal and sagittal
reconstructions, and a higher contrast resolution in soft tis-
sues. 43 ' 44 Both of these diagnostic modalities are effective in
excluding alternative diagnoses. Widespread use of CT and
MRI is hampered by lack of patient access and a high cost.

When properly performed, angiography remains the gold
standard for the diagnosis of popliteal artery entrapment syn-
drome. When this condition is suspected, a careful assessment
of both the asymptomatic and contralateral limb must be
made. The vessels proximal and distal to the knee usually are
normal, helping to distinguish popliteal entrapment from
early-onset atherosclerosis in the young patient. The most
common finding on angiography is that of a localized occlu-
sion of the popliteal artery with an extensive collateral pattern,
found in 66% of angiograms reviewed. 30 Medial deviation or
an anomalous course of the popliteal artery were found in only
29%. Popliteal occlusion at the time of angiography may mask
the presence of this finding, however, as the complementary
use of CT and MRI in patients suspected of having this syn-
drome has demonstrated. 43 ' 44 Popliteal stenosis alone is seen
in 11%, with poststenotic dilation in another 8%. Angiography
may fail to document accurately the presence of a popliteal
aneurysm in this disorder.

At rest, with the foot in a neutral position, the angiographic
appearance of the popliteal artery may be entirely normal. If
popliteal vascular entrapment is suspected, angiographic
views must be obtained with the foot passively dorsiflexed,
and also actively plantarflexed with maximum active exten-
sion of the knee. Lateral and oblique views also should be
obtained. The use of dynamic and stressed angiography is
essential to the early diagnosis of vascular entrapment in the
afflicted person. 47

Treatment

All cases of vascular entrapment of the lower limb should be
treated surgically regardless of the status of the involved ves-
sels. Most patients are young, active, and at risk for limb loss if
the progression of disease is only temporized and not altered
by operative intervention. A variety of procedures have been
used to release the entrapped vessels, repair luminal stenosis,
bypass and exclude aneurysms, or restore distal arterial flow.
Simple division of the obstructing muscular structure or
fibrous band is curative only if the affected artery is com-
pressed, not occluded, and secondary fibrotic changes have
not taken place in the vascular wall. Because most patients
already have an occluded popliteal artery at the time of
symptom development, simple myotomy is reserved for the
contralateral patent and asymptomatic artery in the patient
with bilateral popliteal artery entrapment. The medial head
of the gastrocnemius can be transected readily or partially
resected without producing physical disability, even in the

young and athletic patient. 23 It is important that the transec-
tion of the offending structure be complete, and the artery be
fully mobilized to prevent overlooking associated obstructive
fibrous bands or a more distally located muscle bundle, which
can lead to recurrence of symptoms. The posterior approach
to the popliteal fossa using a S-shaped incision has been rec-
ommended because it allows greater access to the popliteal
neurovascular bundle, and ensures clear delineation of all
the varied anomalies that exist in this syndrome. 23

Thromboendarterectomy accompanied by patch closure
can be used successfully only if a short segment of the popliteal
artery is thrombosed, a good cleavage line is present, and
disease does not spread to the infrageniculate vessels. Using a
similar strategy, Hamming and Vink effectively treated the
first clinical case of popliteal artery entrapment in 1962. 5

When thromboendarterectomy cannot be performed safely,
or associated complications such as aneurysm formation or
midpopliteal thrombosis are present, autogenous vein graft
bypass is indicated. A standard medial approach to the
popliteal artery has been recommended because it affords a
better exposure of the infrapopliteal vessels, and provides
easier access to the greater saphenous vein. The posterior
approach also should be considered because it still provides
adequate exposure of the infrapopliteal vessels, and allows
easy access to the lesser saphenous vein. The use of the internal
iliac artery as an autogenous graft has been proposed in young
patients because vein grafts seem to degenerate and become
aneurysmal with time. 22 The use of artificial grafts is strongly
discouraged owing to their lower patency in the infragenicu-
late location compared with autogenous tissue.

In patients presenting with acute ischemia due to vascular
entrapment, intraarterial thrombolytic agents have been used
to dissolve fresh thrombus, and to improve distal outflow. 48
This approach is supported by reports of improved results
with thrombolytic therapy in patients with popliteal
aneurysms due to atherosclerotic disease who present with
thrombosis and acute ischemia.

Long-term follow-up of the treatment of lower extremity
entrapment syndromes comes mainly from individual case
reports in the literature. The only larger study is by di Marzo
and colleagues. 49 When simple myotomy was possible, the
patency rate in 11 patients was 94%, with a mean of 46
months. 49 When vein reconstruction was undertaken in 12
cases, the patency rate dropped to 58% over a mean period of
43 months. The durability of the vein graft has been
questioned, but because vein reconstruction had to be used in
the more complicated situations, distal outflow also probably
was compromised.

In summary, vascular entrapment of the lower limb must be
considered in the differential diagnosis of intermittent claudi-
cation or acute limb ischemia. It occurs in high frequency in
the young, physically active patient with calf or foot claudica-
tion and no risk factors for early-onset atherosclerosis.
The prevalence of this syndrome probably is greater than

132

chapter 12 Entrapment syndromes

presumed, and it probably contributes to the popliteal
occlusions and aneurysmal disease encountered in the older
patient with associated atherosclerotic disease. The variety of
noninvasive techniques and stress maneuvers available
should lead to an early diagnosis, allowing treatment with a
simple myotomy, and ensuring long-term patency without
physical disability

References

1. Lee BY, La Pointe DG, Madden JL. The adductor canal syndrome.
Am J Surg 1972; 123:617.

2. Baker WH, Stoney RJ. Acquired popliteal entrapment syndrome.
Arch Surg 1972; 105:780.

3. Stuart TPA. Note on a variation in the course of the popliteal
artery. JAnat 1879; 13:162.

4. Chamberdel-Dubreuil L. Variations des Arteres du Pelvis et du Mem-
bre Inferieur. Paris: Masson & Cie, 1925.

5. Hamming JJ. Intermittent claudication at an early age, due to
an anomalous course of the popliteal artery. Angiology 1959;
10:369.

6. Hamming JJ, Vink M. Obstruction of the popliteal artery at an
early age. / Cardiovasc Surg 1965; 6:516.

7. Servello M. Clinical syndrome of anomalous position of the
popliteal artery: differentiation from juvenile arteriopathy.
Circulation 1962; 26:885.

8. Love JW, Whelan TJ. Popliteal artery entrapment syndrome. Am J
Surg 1965; 109:620.

9. Insua JA, Young JR, Humphries AW. Popliteal artery entrapment
syndrome. Arch Surg 1970; 101:771.

10. Collins PS, McDonald PT, Lim RC. Popliteal artery entrapment
syndrome: an evolving syndrome. / Vase Surg 1989; 10:484.

11. Becquemin JP, Melliere D. The popliteal entrapment syndrome
AnatClin 1984; 6:203.

12. Senior HD. The development of the arteries of the human lower
extremity. Am JAnat 1919; 25:55.

13. di Marzo L, Cavallaro A, Sciacca V, Mingoli A, Tamburellia A.
Surgical treatment of popliteal entrapment syndrome: a ten year
experience. Eur J Vase Surg 1991; 5:59.

14. Fontanettta AP, Kirshblom I, Fisher MM, Katz M, Claus RH.
Popliteal artery entrapment: lateral deviation and compression of
artery. VflSfl 1974; 3:399.

15. Gibson MHL, Mills JG, Johnson GE, Downs AR. Popliteal entrap-
ment syndrome. Ann Surg 1977; 185:341.

16. Rich NM, Hughes CW. Popliteal artery and vein entrapment. Am
Surg 1967; 113:696.

17. Edmondson HT, Crow JA. Popliteal arterial and venous entrap-
ment. Am Surg 1972; 38:657.

18. Iwai T, Sato S, Yamada T et ah Popliteal vein entrapment caused
by the third head of the gastrocnemius muscle. Br J Surg 1987;
74:1006.

19. Connell J. Popliteal vein entrapment. Br J Surg 1978; 65:351.

20. Ezzet F, Yettra M. Bilateral popliteal artery entrapment: case
report and observations. / Cardiovasc Surg 1971; 12:71.

21. Haimovici H, Sprayregen S, Johnson F. Popliteal artery entrap-
ment by fibrous band. Surgery 1972; 72:789.

22. Delaney TA, Gonzalez LL. Occlusion of popliteal artery due to
muscular entrapment. Surgery 1971; 69:97.

23. Rich NM, Collins GJ, McDonald PT, Kozloff L, Clagett GP, Collins
JT. Popliteal vascular entrapment: its increasing interest. Arch
Surg 1979; 114:1377.

24. Bouhoutsos J, Daskalakis E. Muscular abnormalities affecting the
popliteal vessels. Br J Surg 1981; 68:501.

25. Rignault DP, Pailler JL, Lunely F. The "functional" popliteal artery
syndrome. IntAngiol 1985; 4:341.

26. Evans WE, Bernhard V. Acute popliteal artery entrapment. Am J
Surg 1971; 121:739.

27. Balaji MR, DeWeese JA. Adductor canal syndrome. JAMA 1981;
245:167.

28. Verta MJ, Vitello J, Fuller J. Adductor canal compression syn-
drome. Arch Surg 1984; 119:345.

29. Ezaki T, Nagasue N, Ogawa Y, Yamada T Popliteal artery entrap-
ment: an unusual case. / Cardiovasc Surg 1986; 27:51.

30. Murray A, Halliday M, Croft RJ. Popliteal artery entrapment
syndrome. Br J Surg 1991; 78:1414.

31. Carter AE, Eban R. A case of bilateral developmental abnormality
of the popliteal arteries and gastrocnemius muscles. Br J Surgl964;
51:518.

32. Darling RC, Buckley CJ, Abbott WM, Raines JK. Intermittent
claudication in young athletes: popliteal artery entrapment
syndrome. / Trauma 1974; 14:543.

33. Brightmore TG, Smellie WAB. Popliteal artery entrapment. Br J
Surg 1971; 58:481.

34. Gedge SW, Spittel JA Jr, Irvins JC. Aneurysm of the distal popliteal
artery and its relationship to the arcuate popliteal ligament.
Circulation 1961; 24:270.

35. Iwai T, Konno S, Soga K et ah Diagnostic and pathological consid-
erations in the popliteal artery entrapment syndrome. / Cardiovasc
Surg 1983; 24:243.

36. Fong H, Downs AR. Popliteal artery entrapment syndrome with
distal embolization: a report of two cases. / Cardiovasc Surg 1989;
30:85.

37. Baker WH, Stoney RJ. Acquired popliteal entrapment syndrome.
Arch Surg 1972; 105:780.

38. Downs AR. Discussion of Insua JA, Young JR, Humphries
AW. Popliteal artery entrapment syndrome. Arch Surg 1970;
101:775.

39. Guitierrez IZ, Barone DL, Currier C, Makula PA. Iatrogenic
entrapment of the femoropopliteal bypass. / Vase Surg 1985; 2:468.

40. Van Damme H, Ballaux, Dereume JP. Femoro-popliteal venous
graft entrapment. / Cardiovasc Surg 1988; 29:50.

41. McDonald PT, Easterbrook JA, Rich NM et ah Popliteal artery
entrapment syndrome: clinical, noninvasive and angiographic
diagnosis. Am J Surg 1980; 139:318.

42. Podore PC. Popliteal entrapment syndrome: a report of tibial
nerve entrapment. / Vase Surg 1985; 2:335.

43. Fukiwara H, Sugano T, Fujii N. Popliteal artery entrapment
syndrome: accurate morphological diagnosis utilizing MRI. /
Cardiovasc Surg 1992; 33:160.

44. Di Cesare E, Simonetti C, Morettini G, Spartera C. Popliteal artery
entrapment: MR findings. / Comput Assist Tomogr 1992; 16:295.

45. Yao ST, Hobbs JT, Irvine WT. Ankle systolic pressure measure-
ments in arterial disease affecting the lower extremities. Br J Surg
1969; 56:676.

133

pa rt I Vascular pathology and physiology

46. Miles S, Roediger W, Cooke P, Mieny CJ. Doppler ultrasound in 48. Greenwood LH, Yiezarry JM, Hallet JW. Popliteal artery entrap-
the diagnosis of popliteal artery entrapment syndrome. Br J Surg ment: importance of the stress run-off for diagnosis. Cardiovasc
1977; 64:883. Intervent Radiol 1986; 9:93.

47. Hallett JW Jr, Greenwood LH, Robinson JG. Lower extremity 49. di Marzo L, Cavallaro A, Sciacca V, Mingoli A, Tamburelia A.
arterial disease in young adults: a systematic approach to early Surgical treatment of popliteal artery entrapment syndrome: a
diagnosis. Ann Surg 1985; 202:647. ten year experience. Eur J Surg 1991; 5:59.

134

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Intimal hyperplasia

Ted R. Kohler

Intimal hyperplasia is the primary cause of restenosis follow-
ing revascularization. It develops after all forms of arterial
injury including endarterectomy, angioplasty, hydrostatic
stretch, and exposure to toxins. There are four stages in this
process: (i) elastic recoil; (ii) laying down of thrombus on the
injured surface; (iii) proliferation and migration of myofibrob-
lasts from the media and adventitia into the intima; (iv) re-
modeling. The cellular component arises from myofibroblasts
that proliferate and migrate from the media and adventitia
into the intima. These cells produce abundant matrix com-
ponents (collagen, elastin, and proteoglycans) that further
thicken the wall. The resulting lesion is smooth, fibrous, and
nonthrombogenic. Although it does not ulcerate or embolize,
it often narrows the lumen significantly, causing deceased
pressure and flow and, ultimately, thrombosis. Animal re-
search has provided significant insight into the pathophysiol-
ogy of this process. Our developing understanding of the
cellular and molecular events underlying intimal hyperplasia
may soon lead to effective means to control it. This chapter will
outline the nature of this process, its pathophysiology, and
some of these potential methods for control.

The clinical problem

Intimal hyperplasia as a complication of revascularization has
been recognized since Carrel and Guthrie 1 first described the
use of vein grafts as arterial substitutes in 1906. They noted
that "the vein was enormously thickened" due to an "increase
of the fibrous tissue, seemingly of the elastic variety." These
workers concluded that "the vein quickly undergoes anatomi-
cal changes and has a tendency to assume the characteristics of
an artery." During the subsequent decades the nature of this
thickening has been better characterized as smooth muscle cell
(SMC) proliferation and matrix deposition (with considerable
amounts of elastin), and it is now understood that this is a re-
sponse to vein injury that occurs at the time of transplantation,
as well as to increased wall stress caused by arterial pres-
sure. 2-9 The process is similar to that of arterial restenosis. Un-

fortunately, we still cannot control it. Approximately 30% of
vein bypass grafts fail within 5 years of operation, both in the
coronary and lower extremity circulation, 3 ' 10-15 and intimal
hyperplasia is responsible for about a third of the failures dur-
ing the first 12-18 months. 10/16/17 The lesions of restenosis occur
mainly in areas of injury including angioplasty sites, anasto-
moses, valves, and regions of clamp injury.

The magnitude of the problem is similar following all forms
of revascularization. About 30% of coronary arteries restenose
following percutaneous angioplasty, with most of these
failures occurring within the first 6 months. 4 ' 18-21 This rate
has not been reduced by antiplatelet agents, steroids, calcium
channel blockers, or heparin. 20-22 Local pharmacology and
brachytherapy hold promise and will be discussed later in this
chapter.

It was hoped that long-term results of revascularization
would be improved by removing the lesion with atherectomy
catheters; however, restenosis rates of up to 50% have been
reported with this technique. 23 Restenosis is also frequent
following carotid endarterectomy, occurring in 10-30% of pa-
tients during the first year. 24-27 Again, these early lesions are
composed of SMCs and matrix. They are generally asympto-
matic because they are nonthrombogenic and do not tend to
ulcerate. Furthermore, flow reduction is well tolerated in the
carotid system due to the abundant collateral circulation to the
brain. 26,27 Approximately 10% of carotid restenosis regresses
over time. 24,26 These lesions can be a site for atherosclerosis,
which may occur after a year or more, with the typical features
of calcium deposits, increased amounts of collagen, foam cells,
and eventual ulceration and thromboembolism. 28 ' 29

Animal models

Most of our understanding of the cellular biology of intimal
hyperplasia comes from work done with animal models. The
most common of these is balloon injury of the rat common
carotid artery. 30 Passage of an inflated balloon catheter
through the artery causes complete denudation of the en-

135

pa rt I Vascular pathology and physiology

dothelium and loss of approximately 20% of the medial SMCs.
Loss of endothelium results in a highly thrombogenic surface
that is covered immediately by a carpet of aggregated
platelets. These platelets spread and degranulate, releasing a
number of vasoactive substances and growth factors. Interest-
ingly, thrombosis is not a prominent feature of this process —
very little fibrin is formed. Within days the surface becomes
nonthrombogenic, and few platelets remain at the surface.
Viable endothelium at the border of the denuded regions re-
generates after the injury. In rats, this occurs at a rate of about
0.2 mm per day and stops after 8-12 weeks, even if denuded re-
gions remain. 31 This is not due to cell senescence, since reinjury
of the endothelium stimulates further replication. 32 There is
strong evidence that basic fibroblast growth factor (bFGF)
plays an important role in regulating this process. Regenerat-
ing endothelial cells produce bFGF, whereas quiescent cells do
not. 33 Addition of bFGF stimulates endothelial proliferation in
normal and injured arteries and can cause reendothelializa-
tion to go to completion within 10 weeks in the rat carotid in-
jury model. 34 It is not known to what extent endothelium can
regenerate in man since this is difficult to analyze either in vivo
or post mortem. The clinical finding that endothelial ingrowth
onto prosthetic grafts in the arterial circulation is limited to ap-
proximately 2 cm suggests that this process is limited. There
are species differences in capillary and endothelial growth,
and there may also be an age factor, with older individuals
having a diminished ability to regenerate endothelium.
Whereas porous polytetrafluoroethylene (PTFE) grafts in ju-
venile baboons completely endothelialize by ingrowth and
spreading of capillaries, this does not appear to occur in el-
derly men. 35 Increased age also has been demonstrated to have
an adverse effect on neovascularization in a murine model of
tumor angiogenesis. 36

Normally, adult medial SMCs are quiescent with less than
0.06% of cells dividing daily. Within 25 h of rat carotid balloon
injury approximately 40% of the viable cells remaining in the
media synchronously enter into the cell cycle. 37 Cell division

■*■ '

rf #

^L~

^•*|r*.

ff '

*■

1 4

— m "^^Uk

■

^■^■st**

■m

i£"^i

^^r ^X7 •

reaches its peak in 2 days and then returns to near-normal lev-
els by 1 week. Lesion thickening continues for about 12 weeks,
mainly due to the continued secretion of matrix components
by the SMCs (Fig. 13.1). Studies using continuous infusion of
tritiated thymidine to label all cells that enter the growth cycle
reveal that the number of nondividing cells does not change
over time. Thus, it appears that cells that do not enter into the
cell cycle during the initial wave of proliferation do not do so at
later times.

Both proliferating and nonproliferating cells contribute to
intimal thickening since both can migrate across the internal
elastic lamina into the new intima. This early, synchronous
and limited entry of SMCs into the cell cycle suggests that the
factor or factors that initiate cell proliferation are present only
transiently at the time of injury. Platelet-derived growth factor
(PDGF) was initially proposed to be one of these agents. This
factor is released from aggregating platelets present on the
surface of the injured vessel and is known to stimulate SMC
growth and migration. A similar mechanism was proposed
as the cause of SMC proliferation in early atherosclerotic
plaques. 38,39 Initial experiments suggested that intimal hyper-
plasia following injury was reduced in thrombocytopenic ani-
mals. 40 However, it has subsequently been shown that platelet
aggregation is not a feature of early atherosclerosis, and that
platelets are not essential for SMC proliferation following arte-
rial injury. More recent experiments with thrombocytopenic
animals demonstrate that arterial lesions are reduced due to
decreased SMC migration rather than proliferation, which is
unaltered. 41 Similarly, antibodies against PDGF cause reduced
intimal lesions following injury with no change in SMC pro-
liferation. 42 Finally, administration of PDGF following injury
stimulates SMC migration more than proliferation. 43

There is now good evidence that bFGF released from the in-
jured SMCs themselves is the main factor stimulating early
proliferation. This growth factor is normally present in SMCs
but is not secreted by them. Injury results in release of bFGF,
which then stimulates SMC proliferation. 44 This is supported

Figure 13.1 Histologic cross-sections of the
region lacking endothelium of injured left
carotid arteries. (A) Normal vessel. Note the
thin intima with a single layerof endothelium.
(B) Denuded vessel at 2 days. Note the loss of
endothelium. (C) Denuded vessel at 2 weeks.
The intima is markedly thickened due to smooth
muscle proliferation. (D) Denuded vessel at
1 2 weeks. Further intimal thickening has
occurred mainly due to continued matrix
deposition. The internal elastic lamina is
indicated by arrows. Lumen is at the top.
(Original magnification x 260.) (From Clowes
AW, ReidyMA, Clowes MM. Kinetics of cellular
proliferation after arterial injury. I. Smooth
muscle growth in the absence of endothelium.
Lab Invest 1 983;49:327, with permission from
Lippincott, Williams & Wilkins.)

r -

136

chapter 13 Intimal hyperplasia

by the finding that very little SMC proliferation occurs when
the endothelium is removed by using a fine loop that does not
damage the media, even though early platelet adhesion still
occurs. 45 ' 46 Conversely, medial SMC proliferation takes place
in the absence of significant endothelial injury if arteries are
hydrostatically distended without balloon catheter injury. 47 In
this model very little SMC migration occurs, perhaps due to
lack of PDGF release from platelets. Further evidence for the
importance of bFGF comes from studies demonstrating that
intimal lesions are significantly reduced if antibodies to bFGF
are given following injury, and infusion or local administra-
tion of bFGF following injury causes an increase in lesion
size. 44 ' 48 This growth factor probably acts in conjunction with
other substances that are present following injury since it is not
mitogenic for cells in uninjured vessels. Basic FGF does not ap-
pear to influence chronic SMC proliferation since the amount
of bFGF in the wall is decreased after balloon injury, and
delayed administration of neutralizing antibodies to bFGF,
4-5 days following injury, has no effect on SMC proliferation. 49
This is also consistent with the fact that this growth factor
cannot be secreted from normal, uninjured cells.

It is likely that all cell types present in the arterial wall (en-
dothelial cells, macrophages, and smooth muscle cells) inter-
act both by cell-cell contact and secretion of various growth
factors to produce intimal hyperplasia. There are cell-cell con-
tacts among and between endothelial and smooth muscle cells
that may be important for regulation of cell growth. Endothe-
lial cells in culture and in regenerating areas of denuded ar-
teries cease their growth when they achieve confluence.
Growth of SMCs in vitro is inhibited when these cells are co-
cultured with endothelial cells.

All of the cell types in the vessel wall are capable of produc-
ing a number of growth factors such as PDGF, bFGF, insulin-
like growth factor, transforming growth factor (TGF)-p, and
epidermal growth factor-like protein. They may release these
factors to regulate their own growth in an autocrine fashion or
the growth of other cells in a paracrine fashion. The normal
balance of growth vs. inhibition is changed following injury.
SMCs recovered from injured arteries and grown in culture
secrete up to five times the amount of PDGF as cells from
uninjured arteries. 50 They also express messenger RNA for in-
sulin-like growth factor and TGF-p. 51 Activated T lympho-
cytes, which are present in the injured wall in small numbers,
make interferon (IFN)-y, which inhibits SMC proliferation in
vitro and induces expression of class II major histocompatibil-
ity complex antigens (la). A negative correlation between la
expression and uptake of thymidine has been observed fol-
lowing balloon injury suggesting that IFN-y regulates SMC
proliferation after injury 52

Growth inhibitors are also produced in the vessel wall. For
example, endothelium makes a heparin-like molecule, which
normally may inhibit SMC growth. 53 Removal of endothelium
may promote SMC growth through removal of this growth in-
hibitor. It has been observed that intimal hyperplasia in in-

jured vessels is decreased in areas of reendothelialization. Loss
of endothelium may also promote proliferation by removal of
the barrier function of these cells. Unlike SMCs that line the
lumen of injured vessels, endothelial cells have tight junctions
that may prevent growth factors in the plasma from reaching
the underlying cells in the vessel wall.

The angiotensin system appears to be active in the vessel
wall and may contribute to regulation of wall structure. 54
Angiotensin converting enzyme (ACE) activity is present in
endothelial cells and in other vascular cells, and angiotensin II
is capable of stimulating SMC proliferation. Intimal hyper-
plasia following balloon injury is reduced by ACE inhibitors
and the specific angiotensin II receptor antagonist Dup
753 55,56 jY\e various factors that are thought to contribute to
intimal thickening in response to injury are summarized in
Figure 13.2.

While the initial growth of new intima is a response to
events related to the injury, hemodynamic factors play an im-
portant role in determining the ultimate extent of wall thicken-
ing (Fig. 13.3). It has long been known that in both developing
and mature arteries intraluminal pressure (wall tension) af-
fects wall thickness while flow (shear) affects diameter. When
pressure is increased the wall thickens to maintain a normal
level of wall tension. This is seen in arteries of hypertensive
adults and in veins grafted into the arterial circulation. Ar-
terial diameter during development is regulated by flow.
Mature arteries can decrease diameter when flow is reduced
by ligation of outflow vessels and can dilate in response
to increased flow (e.g. following creation of an arteriovenous
fistula). 42 ' 57-60 This process appears to be endothelium
dependent. 61 ' 62 Diseased arteries also respond to flow. Coro-
nary arteries dilate when their lumen is compromised by
atherosclerotic plaque. 63 Furthermore, atherosclerosis is
increased in areas of diminished shear. 64-66 It is not surprising,
therefore, to find that intimal hyperplasia is also influenced by
these hemodynamic factors.

The intimal thickening that occurs in response to changes in
flow or pressure has been distinguished from intimal hyper-
plasia and referred to as intimal fibromuscular hypertrophy
(IFH). 67 In this process the SMCs are arranged in an orderly
fashion with a layering that is similar to the lamellar organiza-
tion of the normal vessel wall. This contrasts with intimal
hyperplasia, which consists of a fairly uniform distribution of
SMCs without the layered organization. Glagov and Zarins 67
suggest that IFH is a normal, adaptive response that eventu-
ally returns the local conditions of the vessel wall to normal
levels of wall tension and shear at the luminal surface. In con-
trast, intimal hyperplasia results when the normal, self-limit-
ing process cannot reach a state of equilibrium because the
changes in diameter or thickness are inadequate to restore nor-
mal shear or wall stress. This may be due to compliance mis-
match at sites of anastomoses, obliteration of normal vascular
tissue by scar, or abnormally low flow due to extensive distal
disease.

137

pa rt I Vascular pathology and physiology

Initial Injury

Wall Thickening and
Chronic Adaptation

Blood flow

Platelets and
thrombus

Injury

Blood flow

Injured SMC ^
Macrophage

Proliferation

bFGF

PDGF

TGF

IGF

EGF

IL-1

Angiotensin

Others?

Endothelium

IEL

Figure 13.2 Diagram illustrating how arterial injury might result in intimal
thickening. Macrophages and injured endothelial cells (ECs) and smooth
muscle cells (SMCs) release intracellular mitogens such as basic fibroblast
growth factor (bFGF) that stimulate proliferation of SMCs in the media.
Platelet-derived factors, such as PDGF, stimulate movement of the smooth

Hyperplasia

Migration

• i

Number
^ of SMC

^x^Iedrf

^r^and structural
-=s^ modifications
of SMC

Heparan suffate

TEndothelin

SMC Growth
Inhibition

Decreased
lumen

Increased SMC
proliferation

muscle cells from the media into the intima. Angiotensin II is also involved in
the intimal thickening process. (From Liu MW, Roubin GS, King SB.
Restenosis after coronary angioplasty, potential biologic determinents and
role of intimal hyperplasia. Circulation 1989; 79: 1374 by permission of the
American Heart Association, Inc.)

fif^&*

Figure 13.3 Histologic cross-sections of balloon-injured
rat common carotid arteries 2 weeks after injury. High
flow was created by ligation of the opposite common
carotid artery (A), and low flow was created by ligation of
the ipsilateral internal carotid artery (B). The lumen is at
the top. (From KohlerTR, Jawien A. Flow affects
development of intimal hyperplasia following arterial
injury in rats. ArteriosderThromb 1 992; 1 2:963 with
permission.)

138

chapter 13 Intimal hyperplasia

Increased pressure appears to enhance intimal hyperplasia.
Spontaneously hypertensive rats develop greater intimal
thickening than do normal animals unless their blood pres-
sure is controlled medically. 68 Vein graft thickening also is re-
lated to wall tension. Vein grafts tend to thicken until wall
tension is reduced to normal levels. 69 Furthermore, grafts that
are wrapped with an external support to relieve wall tension
undergo less thickening. 70 These experiments, however, are
confounded by the fact that the narrowed lumen also causes
shear stress to be increased. Wall thickening is reduced in seg-
ments of vein grafts where flow is increased. 71-74 Increased
flow also dramatically reduces the thickness of the endothe-
lialized intima that forms in highly porous PTFE grafts placed
in baboons. 75 Similarly, intimal hyperplasia is reduced in
denuded segments of balloon-injured rat carotid arteries. 76

To date most pharmacology that reduces intimal hyperpla-
sia in animal models has failed in clinical studies. This has led
to questioning of the relevance of these models. However, we
can learn a great deal about the cellular and molecular events
following revascularization from laboratory studies, which
allow quantitative analysis over time. The rat carotid balloon
injury is similar to the human restenosis in many respects —
endothelial loss and medial damage seen after ballooning
occur with all forms of clinical revascularization, including
vein grafting, angioplasty, and endarterectomy — and the cell
types in the neointima are the same. Specimens from human
lesions of restenosis retrieved by atherectomy catheters reveal
very similar histology to the rat lesion. 9/23/77 Both are com-
posed primarily of smooth muscle cells. The clinical failure of
many agents that reduce SMC proliferation in animal models
has taught us that laboratory studies must be interpreted with
caution. Some of the differences in response may be due to
species and age differences. Most human lesions are in elderly
individuals, while laboratories usually study juveniles. As-
pects of vascular biology affected by species and age include
the ability of endothelium to regenerate, the proportion of
growth factor isoforms present in various cell types, and the
vigor of the angiogenic response.

Potential methods of controlling
intimal hyperplasia

Efforts to control intimal hyperplasia may be directed at any of
the fundamental processes involved: elastic recoil, injury to
the wall, platelet aggregation, SMC migration or proliferation,
matrix deposition, endothelial regeneration, and adaptation
to hemodynamic changes. The potential methods for control-
ing these processes include drugs, genetic manipulation,
brachy therapy, and mechanical devices.

Many different classes of drugs may influence intimal
hyperplasia including antiplatelet, antihypertensive, anti-
coagulant, and antiproliferative agents. The effective use of
antiplatelet agents for reduction of intimal hyperplasia is

hampered by the fact that total inhibition of platelet function
would cause significant hemorrhage. In clinical trials, the peri-
operative use of aspirin has reduced early failure of coronary
artery bypass grafts but has not been shown to influence inti-
mal hyperplasia. 78-80 This is consistent with the limited role of
platelets in animal models of intimal hyperplasia. Studies
using aspirin and dipyridamole for lower extremity bypass
grafts have had mixed results. 81-83 In one trial, ticlopidine im-
proved 2-year patency of vein grafts. 84 Low-molecular-weight
dextran has been shown to improve the early patency rates of
difficult lower extremity bypass grafts but does not appear to
affect long-term patency, suggesting that it, like aspirin, can
help prevent early thrombosis in marginal reconstructions but
has little or no effect on intimal hyperplasia. 85 A monoclonal
chimeric antibody, abcixamab, directed against the glycopro-
tein lib /Ilia integrin, inhibits platelet aggregation and reduces
cardiac events following coronary angioplasty 86 Patients
treated with the drug have an improved outcome as long as
3 years later, suggesting that this antiplatelet agent affects
long-term healing in addition to any immediate antithrom-
botic effects. 87

The anticoagulant heparin reduces intimal hyperplasia fol-
lowing experimental vein grafts and in the injured carotid
artery. 37,88,89 It primarily inhibits SMC proliferation although it
also reduces migration. 37 ' 90-94 Heparin affects the matrix com-
position resulting in a reduction in elastin and collagen and an
increase in proteoglycans. 95 It does not inhibit endothelial cell
regrowth and in fact may enhance this process. 92 ' 93 Heparin is
effective at early times after injury and appears to block SMC
proliferation in the late G or early G 1 phase of the cell cycle. To
be effective it must be present 24-72 h after injury. The mecha-
nism of inhibition is not dependent on the drug's effect on the
coagulation system; nonanticoagulant fractions of heparin
that do not bind anti thrombin III are effective in reducing inti-
mal hyperplasia. 96 Heparin may act by decreasing the expres-
sion of tissue plasminogen activator and displacement of
urokinase. 97 ' 98 The action of these proteases is necessary for the
breakdown of matrix components that allows SMCs to move
and proliferate. 99 Heparin may also bind to bFGF, thus inhibit-
ing early SMC proliferation. Clinical trials using heparin to
reduce restenosis following percutaneous transluminal coro-
nary angioplasty have been disappointing.

As mentioned previously, intimal hyperplasia following
balloon injury is reduced by ACE inhibitors and the specific
angiotensin II receptor antagonist Dup 753. 55/56 The mecha-
nism for reduction of proliferation by ACE inhibitors is not
fully understood, but seems to be different from that of he-
parin since the combination of these two agents is much more
effective in reducing wall thickening than either alone. 100

Another class of antihypertensive agent, the calcium chan-
nel blockers, can also reduce intimal hyperplasia. In animal
models these agents have been shown to inhibit vein graft hy-
perplasia, atherosclerosis secondary to fat feeding, and inti-
mal hyperplasia following balloon injury 56 ' 101-105 Nifedipine

139

pa rt I Vascular pathology and physiology

may be able to reduce formation of new atherosclerotic lesions
in coronary arteries. 106 The mechanism may involve alteration
of SMC migration and matrix synthesis as well as cellular
handling of lipoproteins and intracellular cholesterol ester
stores. 107

Other drugs that have been used to reduce intimal hyper-
plasia include cytotoxic agents, steroids, which can inhibit ex-
perimental intimal hyperplasia but to date have not worked in
clinical trials, and fish oil, which may reduce intimal hyper-
plasia by reducing platelet aggregation and production of
PDGF. 108-112 Genetically altered growth factors or growth-
factor receptors are another means to modulate growth. For
example, mutant PDGF receptors that bind PDGF but do not
promote SMC growth inhibit PDGF-stimulated proliferation
in animal models. 113 ' 114 Unfortunately, clinical trials with this
agent have not been successful. Along similar lines, mutated
acidic FGF (aFGF) molecules have been produced that do not
stimulate mesenchymal cell growth but do bind to the aFGF
receptor blocking the action of authentic aFGF. 115-117 It is also
possible to make specific antibodies that bind and inactivate
these growth factors. As mentioned earlier, in animal models
antibodies that neutralize PDGF can block SMC migration and
antibodies against bFGF block proliferation following arterial

• • 42 48

injury. '

Vein grafts have the advantage of local treatment after they
have been excised and before they are reimplanted in the
arterial circulation. In a pilot study, bathing the vein in a
decoy oligodeoxynucleotide that binds the E2F transcription
factor necessary for cell cycling reduced rates of proliferation
and restenosis in human lower extremity bypass grafts. 118 A
prospective, multicenter trial of this protocol is under way.

Mechanical methods for reducing intimal hyperplasia in-
clude techniques to reduce the initial injury at the time of
revascularization, stents to prevent elastic recoil, and atherec-
tomy devices to debulk lesions. Drug-eluting stents provide
local delivery of antiproliferative agents and have shown
promise in clinical trials. Among the agents being used are
actinomycin D and paclitaxel. Sirolimus, a natural macrolide
immunosuppressant, has been particularly effective in pre-
venting stent restenosis. It acts by binding with its cognate
immunophilin and increasing p27 concentrations, which
stops G 1 /S cell cycle progression. 119 Restenosis was virtually
absent at 1 year in 45 patients treated with drug-eluting stents
impregnated with sirolimus. 120 A 237-patient cooperative
study in Europe (the RAVEL trial) has had similar results. 119

Injury may be reduced by improvements in surgical en-
darterectomy, balloon catheter technology, lasers, and atherec-
tomy devices. Vein graft injury is reduced by gentle surgical
dissection, avoidance of forceful dilation of the graft during
preparation, and possibly by using an in-situ technique rather
than removing the vein from its native bed. Rough handling or
an overvigorous distention of veins causes significant loss of
endothelium and results in more wall thickening than does
gentle handling. 16 Careful balloon angioplasty avoiding ex-

cessive vessel distention can reduce damage to the media and
thereby the response to injury. Some studies suggest that in-
creasing the depth of atherectomy results in an increased rate
of restenosis, possibly due to the increased damage to the ves-
sel wall. 23 Compliance mismatch, which may cause a chronic
form of injury at the prosthesis-vessel interface, may be re-
duced by using more compliant graft materials or interposi-
tion of a short segment of vein graft between the prosthesis
and the vessel. 121

Stents are effective in maintaining adequate lumen
following angioplasty, particularly when the vessel tends to
collapse or "recoil" to its initial luminal diameter immediately
following angioplasty. There is no evidence that stents
can reduce intimal hyperplasia. The neointima can grow
through the interstices to cause restenosis in small vessels.
This may be reduced by impregnating the stents with
antiproliferative or radioactive materials. Brachytherapy has
shown promise in reducing restenosis in coronary arteries, but
a candy-wrapper effect often causes stenosis at either end of
the delivery site where a subtherapeutic radiation dose is
given. 122 This effect was not seen when gamma radiation was
delivered locally by catheter to reduce the incidence of in-
stent restenosis in coronary artery bypass grafts following
treatment by atherectomy, balloon angioplasty, or additional
stenting. 123

In experimental models, increased flow results in reduced
intimal hyperplasia, therefore any method that will increase
flow through an injured segment may improve outcome.
Methods to achieve this include vigorous treatment of distal
lesions to improve outflow and creation of a distal arteriove-
nous fistula. The latter should not be undertaken unless fur-
ther studies demonstrate a clear benefit to offset the inherent
disadvantages of this procedure. There are reports that the use
of arteriovenous fistulae can increase lower extremity bypass
graft patency rates. 124 This may be due to the reduced throm-
bogenicity of high flow grafts or possibly reduction of intimal
hyperplasia at the distal anastomosis.

Reendothelialization may be complete in short segments of
arteries injured by angioplasty, in stented segments, and in
short grafts (a few centimeters in length). Longer devices en-
dothelialize only for a limited region at either end. For decades
various laboratories have tried to produce a durable, viable,
and complete endothelial lining of prosthetic grafts in the
hopes of reducing thrombogenicity to allow grafting to small-
er vessels. In animal models, successful endothelial cell cover-
age of PTFE grafts has been achieved by ex-vivo cell seeding
and by increasing porosity to allow capillary ingrowth
throughout the graft. Some clinical studies have demon-
strated reduced thrombogenicity of cell-seeded PTFE graft
segments. 125 However, it remains to be proven that cell seed-
ing results in improved graft patency. Endothelial cells grown
on prosthetic grafts may not function normally and could
promote intimal hyperplasia. In the baboon model, these cells
have an increased rate of turnover suggesting chronic in-

140

chapter 13 Intimal hyperplasia

jury. 126 ' 127 Injured endothelial cells produce bFGF, which
is not made by normal, quiescent cells, and production of
endothelial-derived relaxing factor is reduced. 128 In animal
models, cell seeding of deendothelialized segments has been
successful using balloon catheter techniques. Bush and col-
leagues demonstrated decreased wall thickening of denuded
segments that were successfully seeded with endothelium. 129
This is consistent with the finding in balloon-injured vessels
that wall thickening is less in segments where endothelium
has regrown and that SMCs cease proliferating in reendothe-
lialized regions but continue to proliferate in chronically
denuded regions near the lumen. 53 ' 127

An alternative approach to cell seeding is to make the graft
sufficiently porous to allow ingrowth of capillaries. In baboon
models using highly porous PTFE and Dacron, capillaries
grow through the graft to the lumen. 75 ' 130-132 On reaching the
lumen they change from a tubular morphology and spread,
eventually coalescing into an intact endothelial lining. A
neointima is formed when SMCs grow in under the endothe-
lium and proliferate. To date, there is no evidence that this
occurs in humans. 133-136 This may be due to a difference in
the potential for angiogenesis in adult humans compared with
juvenile baboons.

Endothelial cells can be genetically altered to produce anti-
thrombotic agents or factors that will inhibit SMC growth or
promote endothelial regrowth following injury. Many work-
ers have demonstrated function of genes implanted into en-
dothelial cells or SMCs that have been seeded onto PTFE grafts
or onto denuded arterial segments. 137-141 Similarly, genetically
altered endothelial cells, producing large amounts of tissue
plasminogen activator, have been successfully grown on
stents placed in the arterial circulation in the hope that stent-
related thrombosis could be eliminated. 142 It is possible that
this type of genetic engineering may be useful not only for con-
troling vessel wall growth, but for treating some metabolic
derangements such as diabetes or adenosine deaminase
(ADA) deficiency. SMCs retrovirally infected with genes
expressing human ADA have been seeded into injured rat
arteries and have yielded potentially therapeutic levels of the
enzyme for up to 6 months. 141

At present, the two main factors that reduce the success rate
of revascularization for arterial occlusive disease are thrombo-
sis of small-caliber grafts and intimal hyperplasia causing
restenosis and long-term graft failure. We now know a great
deal about the vascular biology of intimal hyperplasia and are
beginning to understand the molecular biology of this process.
Effective new therapies directed at the molecular control of
intimal hyperplasia are on the horizon.

References

1. Carrel A, Guthrie CC. Uniterminal and biterminal venous trans-
plantations. Surg Gynecol Obstet 1906; 2:266.

2. Barboriak JJ, Pintar K, Van Horn DK, Batayias GE, Korns ME.
Pathologic findings in the aortocoronary vein grafts. A scanning
electron microscope study. Atherosclerosis 1978; 29:69.

3. Lawrie GM, Lie JT, Morris GC, Beazley HL. Vein graft patency
and intimal proliferation after aortocoronary bypass. Early and
long-term angiopathologic correlations. Am J Cardiol 1976;
38:856.

4. Liu MW, Roubin GS, King SB. Restenosis after coronary
angioplasty, potential biologic determinants and role of intimal
hyperplasia. Circulation 1989; 79:1374.

5. Marti MC, Bouchardy B, Cox JN. Aorto-coronary bypass with
autogenous saphenous vein grafts: histopathological aspects.
Virchows Arch [Pathol Anat] 1971; 352:255.

6. Unni KK, Kottke BA, Titus JL, Frye RL, Wallace RB, Brown AL.
Pathologic changes in aortocoronary saphenous vein grafts. Am J
Cardiol 1974; 34:526.

7. Vlodaver Z, Edwards JE. Pathologic changes in aortic-coronary
arterial saphenous vein grafts. Circulation 1971; 44:719.

8. Jones M, Conkle DM, Ferrans VJ, Roberts WC, Levine FH.
Lesions observed in arterial autogenous vein grafts. Light and
electron microscopic evaluation. Circulation 1973; 47:198.

9. Garratt KN, Edwards WD, Kaufmann UP, Vlietstra RE, Holmes
DR Jr. Differential histopathology of primary atherosclerotic and
restenotic lesions in coronary arteries and saphenous vein
bypass grafts: analysis of tissue obtained from 73 patients by
directional atherectomy. J Am Coll Cardiol 1991; 17:442.

10. Szilagyi DE, Elliott JP, Hageman JH, Smith RF, Dall'Olmo CA.
Biologic fate of autogenous vein implants as arterial substitutes.
Ann Surg 1973; 178:232.

11 . Whittemore AD, Clowes AW, Couch NP, Mannick JA. Secondary
femoropopliteal reconstruction. Ann Surg 1981; 193:35.

12. Lie JT, Lawrie GM, Morris GC. Aortocoronary bypass saphenous
vein graft atherosclerosis. Anatomic study of 99 vein grafts from
normal and hyperlipoproteinemic patients up to 75 months post-
operatively. Am } Cardiol 1977; 40:906.

13. Grondin CM, Campeau L, Lesperance J, Solymoss BC. Athero-
sclerotic changes in coronary vein grafts six years after operation:
angiographic aspect in 110 patients. / Thorac Cardiovasc Surg 1979;
77:24.

14. Campeau L, Enjalbert M, Lesperance J, Bourassa MG. The
relation of risk factors to the development of atherosclerosis in
saphenous-vein bypass grafts and the progression of disease in
the native circulation. N Engl] Med 1984; 311:1329.

15. Bourassa MG, Fisher LD, Campeau L, Gillespie MJ, McConney
M, Lesperance J. Long-term fate of bypass grafts: the coronary
artery surgery study (CASS) and Montreal Heart Institute
experiences. Circulation 1985; 72 (Suppl. V):V-71.

16. LoGerfo FW, Quist WC, Cantelmo NL, Haudenschild CC. In-
tegrity of vein grafts as a function of initial intimal and medial
preservation. Circulation 1983; 68 (Suppl. II):II-117.

17. Imparato AM, Bracco A, Kim GE, Zeff R. Intimal and neointimal
fibrous proliferation causing failure of arterial reconstructions.
Surgery 1972; 72:1007.

18. Macdonald RG, Henderson MA, Hirshfeld JW Jr et al. Patient-
related variables and restenosis after percutaneous transluminal
coronary angioplasty —a report from the M-HEART group. Am]
Cardioll990;66:926.

19. Holmes DR Jr, Vlietstra RE, Smith HC. Restenosis after percuta-
neous transluminal coronary angioplasty (PTC A): a report from

141

pa rt I Vascular pathology and physiology

the PTCA registry of the National Heart, Lung and Blood Insti-
tute. Am } Cardiol 1984; 53:77C.

20. Meier B. Restenosis after coronary angioplasty: review of the
literature. Eur Heart J 1988; 9 (Suppl.C):l.

21 . Pepine CJ, Hirshfeld JW, Macdonald RG et al. A controlled trial of
corticosteroids to prevent restenosis after coronary angioplasty.
Circulation 1990; 81:1753.

22. Meier B. Prevention of restenosis after coronary angioplasty: a
pharmacological approach. Eur Heart] 1989; 10:64.

23. Garratt KN, Holmes DR Jr, Bell MR et al Restenosis after
directional coronary atherectomy: differences between primary
atheromatous and restenosis lesions and influence of subintimal
tissue resection. J Am Coll Cardiol 1990; 16:1665.

24. Zierler RE, Bandyk DF, Thiele BL, Strandness DE Jr. Carotid
artery stenosis following endarterectomy Arch Surg 1982;
117:1408.

25. Thomas M, Otis SM, Rush M, Zyroff J, Dilley RB, Bernstein EF.
Recurrent carotid artery stenosis following endarterectomy. Ann
Surg 1984; 200:74.

26. Healy DA, Zierler RE, Nicholls SC et al. Long-term follow-up and
clinical outcome of carotid restenosis. / Vase Surg 1989; 10:662.

27. Bernstein EF, Torem S, Dilley RB. Does carotid restenosis predict
an increased risk of late symptoms, stroke, or death? Ann Surg
1990; 212:629.

28. Clagett CP, Robinowitz M, Youkey JR et al. Morphogenesis and
clinicopathologic characteristics of recurrent carotid disease. /
Vase Surg 1986; 3:10.

29. Sterpetti AV, Schultz RD, Feldhaus RJ et al. Natural history of re-
current carotid artery disease. Surg Gynecol Obstet 1989; 168:217.

30. Clowes AW, Reidy MA, Clowes MM. Mechanisms of stenosis
after arterial injury. Lab Invest 1983; 49:208.

31. Clowes AW, Clowes MM, Reidy MA. Kinetics of cellular pro-
liferation after arterial injury. Ill Endothelial and smooth muscle
growth in chronically denuded vessels. Lab Invest 1986; 54:295.

32. Reidy MA, Clowes AW, Schwartz SM. Endothelial regeneration.
V. Inhibition of endothelial regrowth in arteries of rat and rabbit.
Lab Invest 1983; 49:569.

33. Lindner V, Reidy MA, Fingerle J. Regrowth of arterial endo-
thelium: denudation with minimal trauma leads to complete
endothelial cell regrowth. Lab Invest 1989; 61 :556.

34. Lindner V, Majack RA, Reidy MA. Basic fibroblast factor stimu-
lates endothelial regrowth and proliferation in denuded arteries.
/ Clin Invest 1990; 85:2004.

35. Kohler TR, Stratton JR, Kirkman TR, Johansen KH, Zierler BK,
Clowes AW. Conventional versus high-porosity polytetra-
fluoroethylene grafts: clinical evaluation. Surgery 1992; 112:901.

36. Kreisle RA, Stebler BA, Ershler WB. Effect of host age on tumor-
associated angiogenesis in mice. J Natl Cancer Inst 1990; 82:44.

37. Majesky MW, Schwartz SM, Clowes MM, Clowes AW. Heparin
regulates smooth muscle S phase entry in the injured rat carotid
artery. Circ Res 1987; 61:296.

38. Ross R, Glomset JA. The pathogenesis of atherosclerosis (second
of two parts) . N Engl J Med 1976; 295:420.

39. Ross R, Glomset JA. The pathogenesis of atherosclerosis (first of
two parts). N Engl} Med 1976; 295:369.

40. Friedman RJ, Stemerman MB, Wenz B, Moore S, Gauldie J. The ef-
fect of thrombocytopenia on experimental arteriosclerotic lesion
formation in rabbits. Smooth muscle cell proliferation and re-
endothelialization./C/m Invest 1977; 60:1191.

41 . Fingerle J, Johnson R, Clowes AW, Majesky M W, Reidy MA. Role
of platelets in smooth muscle cell proliferation and migration
after vascular injury in rat carotid artery. Proc Natl Acad Sci USA
1989; 86:8412.

42. Ferns G AA, Raines EW, Sprugel KH, Motani AS, Reidy MA, Ross
R. Inhibition of neointimal smooth muscle accumulation after
angioplasty by an antibody to PDGF. Science 1991; 253:1129.

43. Jawien A, Bowen-Pope DF, Lindner V, Schwartz SM, Clowes AW.
Platelet-derived growth factor promotes smooth muscle migra-
tion and intimal thickening in a rat model of balloon angioplasty.
} Clin Invest 1992; 89:507.

44. Lindner V, Lappi DA, Baird A, Majack RA, Reidy MA. Role of
basic fibroblast growth factor in vascular lesion formation. Circ
Res 1991; 68:106.

45. Tada T, Reidy MA. Endothelial regeneration. IX. Arterial injury
followed by rapid endothelial repair induces smooth-muscle-
cell proliferation but not intimal thickening. Am } Pathol 1987;
129:429.

46. Fingerle J, Au YPT, Clowes AW, Reidy MA. Intimal lesion forma-
tion in rat carotid arteries after endothelial denudation in
absence of medial injury. Arteriosclerosis 1990; 10:1082.

47. Clowes AW, Clowes MM, Reidy MA. Role of acute distension in
the induction of smooth muscle proliferation after endothelial
denudation. FASEB 1990; Abstract.

48. Lindner V, Reidy MA. Proliferation of smooth muscle cells after
vascular injury is inhibited by an antibody against basic fibrob-
last growth factor. Proc Natl Acad Sci USA 1991; 88:3739.

49. Olson NE, Chao S, Lindner V, Reidy MA. Intimal smooth muscle
cell proliferation after balloon catheter injury. Am } Pathol 1992;
5:1017.

50. Walker LN, Bowen-Pope DF, Reidy MA. Production of platelet-
derived growth factor-like molecules by cultured arterial
smooth muscle cells accompanies proliferation after arterial
injury. Proc Natl Acad Sci USA 1986; 83:7311.

51. Cercek B, Fishbein MC, Forrester JS, Helfant RH, Fagin JA. In-
duction of insulin-like growth factor I messenger RNA in rat
aorta after balloon denudation. Circ Res 1990; 66:1755.

52. Hansson GK, Jonasson L, Holm J, Clowes MM, Clowes AW.
Gamma-interferon regulates vascular smooth muscle prolifera-
tion and la antigen expression in vivo and in vitro. Circ Res 1988;
63:712.

53. Castellot JJ Jr, Addonizio ML, Rosenberg R, Karnovsky MJ.
Cultured endothelial cells produce a heparin-like inhibitor of
smooth muscle cell growth. / Cell Biol 1981; 90:372.

54. Pipili E, Manolopoulos VG, Catravas JD, Maragoudakis ME.
Angiotensin converting enzyme activity is present in the
endothelium-denuded aorta. Br J Pharmacol 1989; 98:333.

55. Powell JS, Clozel JP, Muler RKM et al. Inhibitors of angiotensin-
converting enzyme prevent myointimal proliferation after vas-
cular injury. Science 1989; 245:186.

56. Powell JS, Miiller RKM, Rouge M, Kuhn H, Hefti F, Baumgartner
HR. The proliferative response to vascular injury is suppressed
by angiotensin-converting enzyme inhibition. / Cardiovasc Phar-
macol 1990; 16 (Suppl. 4):S42.

57. Holman E. The anatomic and physiologic effects of an arteriove-
nous fistula. Surgery 1940; 8:362.

58. Hull SS, Romig GD, Sparks HV, Jaffe MD. Flow induced vasodi-
lation of large arteries is dependant on endothelium. Fed Proc
1984; 43:900.

142

chapter 13 Intimal hyperplasia

59. Kamiya A, Togawa T. Adaptive regulation of wall shear stress to
flow change in the canine carotid artery Am J Physiol 1980; 239:14.

60. Zarins CK, Zatina MA, Giddens DP, Ku DN, Glagov S. Shear
stress regulation of artery lumen diameter in experimental
atherogenesis. / Vase Surg 1987; 5:413.

61. Langille BL, O'Donnell F. Reductions in arterial diameter pro-
duced by chronic decreases in blood flow are endothelium-
dependent. Science 1986; 231:405.

62. Tohda K, Masuda H, Kawamura K, Shozawa T. Difference in di-
latation between endothelium-preserved and -desquamated
segments in the flow-loaded rat common carotid artery. Arte-
rioscler Thromb 1992; 12:519.

63. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R. Compen-
satory enlargement of human atherosclerotic coronary arteres. N
Engl J Med 1987; 316:1371.

64. Ku DN, Giddens DP, Zarins C, Glagov S. Pulsatile flow and
atherosclerosis in the human carotid bifurcation. Arteriosclerosis
1985; 5:293.

65. Zarins CK, Giddens DP, Bharadvaj BK, Sottiurai VS, Mabon RF,
Glagov S. Carotid bifurcation atherosclerosis: quantitative corre-
lation of plaque localization with flow velocity profiles and wall
shear stress. Circ Res 1983; 53:502.

66. Friedman MH, Hutchins GM, Bargeron CB, Deters OJ, Mark FF.
Correlation between intimal thickness and fluid shear in human
arteries. Atherosclerosis 1981; 39:425.

67. Glagov S, Zarins CK. Is intimal hyperplasia an adaptive response
or a pathologic process? Observations on the nature of non-
atherosclerotic intimal thickening. / Vase Surg 1989; 10:571 .

68. Clowes AW, Clowes MM. Influence of chronic hypertension on
injured and uninjured arteries in spontaneously hypertensive
rats. Lab Invest 1980; 6:535.

69. Zwolak RM, Adams MC, Clowes AW. Kinetics of vein graft
hyperplasia. Association with tangential stress. / Vase Surg 1987;
5:126.

70. Kohler TR, Kirkman TR, Clowes AW. The effect of rigid external
support on vein graft adaptation to the arterial circulation. / Vase
Surgl989;9:277.

71 . Berguer R, Higgins RF, Reddy DJ. Intimal hyperplasia. Arch Surg
1980; 115:332.

72. Rittgers SE, Karayannacos PE, Guy JF. Velocity distribution and
intimal proliferation in autologous vein grafts in dogs. Circ Res
1978; 42:792.

73. Morinaga K, Okadome K, Kuroki M, Miyazaki T, Muto Y. Effect
of wall shear stress on intimal thickening of arterially trans-
planted autogenous veins in dogs. / Vase Surg 1985; 2:430.

74. Mii S, Okadome K, Onohara T, Yamamura S, Sugimachi K.
Intimal thickening and permeability of arterial autogenous vein
graft in a canine poor-runoff model: transmission electron
microscopic evidence. Surgery 1990; 108:81.

75. Kohler TR, Kirkman TR, Kraiss LW, Zierler BK, Clowes AW. In-
creased blood flow inhibits neointimal hyperplasia in endothe-
lialized vascular grafts. Circ Res 1991; 69:1557.

76. Kohler TR, Jawien A. Flow affects development of intimal hyper-
plasia following arterial injury in rats. Arterioscler Thromb 1992;
12:963.

77. Safian RD, Gelbfish JS, Erny RE, Schnitt SS, Schmidt DA,
Bairn DS. Coronary atherectomy. Clinical, angiographic, and
histological findings and observations. Circulation 1990; 82:
69.

78. Goldman S, Copeland J, Moritz T et al. Improvement in early
saphenous vein graft patency after coronary artery bypass
surgery with antiplatelet therapy: results of a Veterans Adminis-
tration Cooperative Study. Circulation 1988; 77:1324.

79. Chevigne M, David J-L, Rigo P, Limet R. Effect of ticlopidine on
saphenous vein bypass patency rates: a double-blind study. Ann
Thor Surg 1984; 37:371.

80. Goldman S, Copeland J, Moritz T et al. Saphenous vein graft
patency 1 year after coronary artery bypass surgery and effects
of antiplatelet therapy: results of a Veterans Administration
Cooperative Study. Circulation 1989; 80:1190.

81. Clagett GP, Genton E, Salzman EW. Antithrombotic therapy in
peripheral vascular disease. Chest 1989; 95:128S.

82. Clowes AW. The role of aspirin in enhancing arterial graft
patency. / Vase Surg 1986; 3:381 .

83. Clowes AW, Reidy MA. Prevention of stenosis after vascular re-
construction: pharmacologic control of intimal hyperplasia— a
review. / Vase Surg 1991; 13:885.

84. Becquemin JP Effect of ticlopidine on the long-term patency of
saphenous-vein bypass grafts in the legs. N Engl J Med 1997;
337:1726.

85. Rutherford RB, Jones DN, Bergentz SE et al. The efficacy of
dextran 40 in preventing early postoperative thrombosis follow-
ing difficult lower extremity bypass. / Vase Surg 1984; 1:765.

86. Randomised placebo-controlled and balloon-angioplasty-
controlled trial to assess safety of coronary stenting with use of
platelet glycoprotein-IIb/IIIa blockade. The EPISTENT Investi-
gators. Evaluation of Platelet lib /Ilia Inhibitor for Stenting.
Lancet 1998; 352(9122):87.

87. Topol EJ, Ferguson JJ, Weisman HF et al. Long-term protection
from myocardial ischemic events in a randomized trial of brief
integrin beta3 blockade with percutaneous coronary interven-
tion. EPIC Investigator Group. Evaluation of Platelet lib /Ilia In-
hibition for Prevention of Ischemic Complication. JAMA 1997;
278:479.

88. Hirsch GM, Karnovsky MJ. Inhibition of vein graft intimal pro-
liferative lesions in the rat by heparin. Am J Pathol 1991; 139:581.

89. Kohler TR, Kirkman TR, Clowes AW. Effect of heparin on adapta-
tion of vein grafts to arterial circulation. Arteriosclerosis 1989;
9:523.

90. Clowes AW, Clowes MM. Kinetics of cellular proliferation after
arterial injury. IV Heparin inhibits rat smooth muscle mitogene-
sis and migration. Circ Res 1986; 58:839.

91. Majack RA, Clowes AW. Inhibition of vascular smooth muscle
cell migration by heparin-like glycosaminoglycans. / Cell Physiol
1984; 118:253.

92. Hoover RL, Rosenberg R, Haering W, Karnovsky MJ. Inhibition
of rat arterial smooth muscle cell proliferation by heparin. II. In
vitro studies. Circ Res 1980; 47:578.

93. Clowes AW, Clowes MM. Kinetics of cellular proliferation after
arterial injury. II Inhibition of smooth muscle growth by heparin.
Lab Invest 1985; 52:611.

94. Clowes AW, Karnovsky MJ. Suppression by heparin of smooth
muscle cell proliferation in injured arteries. Nature 1977; 265:625.

95. Snow AD, Bolender RP, Wight TN, Clowes AW. Heparin modu-
lates the composition of the extracellular matrix domain sur-
rounding arterial smooth muscle cells. Am J Pathol 1990; 137:313.

96. Guyton JR, Rosenberg RD, Clowes AW, Karnovsky MJ. Inhibi-
tion of rat arterial smooth muscle cell proliferation by heparin. I.

143

pa rt I Vascular pathology and physiology

In vivo studies with anticoagulant and non-anticoagulant
heparin. Circ Res 1980; 46:625.

97. Au YPT, Kenagy RD, Clowes AW. Heparin selectively inhibits
the transcription of tissue-type plasminogen activator in primate
arterial smooth muscle cells during mitogenesis. / Biol Chem
1992; 70:3438.

98. Kenagy RD, Welgus HG, Clowes AW. Heparin inhibits the
expression of matrix metalloproteases by smooth muscle cells. /
Cell Biol 1991; 115 (138a): Abstract.

99. Clowes AW, Clowes MM, Au YPT, Reidy MA, Belin D. Smooth
muscle cells express urokinase during mitogenesis and tissue-
type plasminogen activator during migration in injured rat
carotid artery. Circ Res 1990; 67:61.

100. Clowes AW, Clowes MM, Vergel SC et ah Heparin and cilazapril
together inhibit injury-induced intimal hyperplaisa. Hyperten-
sion 1991; 18 (Suppl. II):II-65.

101. El-Sanadiki MN, Cross KS, Murray JJ et ah Reduction of intimal
hyperplasia and enhanced reactivity of experimental vein by-
pass grafts with verapamil treatment. Ann Surg 1990; 212:87.

102. Atkinson JB, Swift LL. Nifedipine reduces atherogenesis in
cholesterol-fed heterozygous WHHL rabbits. Atherosclerosis
1990; 84:195.

103. Paniszyn CC. Reduction of intimal hyperplasia and enhanced
reactivity of experimental vein bypass grafts with verapamil
treatment. Ann Surg 1991; 213:374.

104. Jackson CL, Bush RC, Bowyer DE. Mechanism of antiatherogenic
action of calcium antagonists. Atherosclerosis 1989; 80:17.

105. Jackson CL, Bush RC, Bowyer DE. Inhibitory effect of calcium
antagonists on balloon catheter-induced arterial smooth
muscle cell proliferation and lesion size. Atherosclerosis 1988;
69:115.

106. Lichtlen PR, Hugenholtz PG, Rafflenbeul W, Hecker H, Jost S,
Deckers JW. Nifedipine trial. Lancet 1990; 335:1109.

107. Sowers JR. Calcium channel blockers and atherosclerosis. Am J
Kidney Dis 1990; 16 (Suppl. 1):3.

108. Chervu A, Moore WS, Quinones-Baldrich WJ, Henderson T. Effi-
cacy of corticosteroids in suppression of intimal hyperplasia. /
Vase Surg 1989; 10:129.

109. Berk BC, Gordon JB, Alexander RW. Pharmacologic roles of
heparin and glucocorticoids to prevent restenosis after coronary
angioplasty. J Am Coll Cardiol 1991; 17 (Suppl. B):111B.

110. Fanelli C, Aronoff R. Restenosis following coronary angioplasty.
Am Heart} 1990; 119:357.

111. Goodnight SH, Fisher M, Fitzgerald GA, Levine PH. Assessment
of the therapeutic use of dietary fish oil in atherosclerotic vascu-
lar disease and thrombosis. Chest 1989; 95:19S.

112. Landymore RW, Manku MS, Tan M, MacAulay MA, Seridan B.
Effects of low-dose marine oils on intimal hyperplasia in auto-
logous vein grafts. / Thorac Cardiovasc Surg 1989; 98:788.

113. Duan D, Pazin MJ, Fretto LJ, Williams LT. A functional soluble
extracellular region of the platelet-derived growth factor (PDGF)
B-receptor antagonized PDGF-stimulated responses. J Biol Chem
1991; 266:413.

114. Ueno H, Colbert H, Escobedo JA, Williams LT. Inhibition of
PDGF (3 receptor signal transduction by coexpression of a trun-
cated receptor. Science 1991; 252:844.

115. Moncada S. The first Robert Furchgott lecture: From
endothelium-dependent relaxation to the L-arginine:NO path-
way. Blood Vessels 1990; 27:208.

116. Rubanyi GM, Freay AD, Kauser K, Johns A, Harder DR.
Mechanoreception by the endothelium: mediators and mecha-
nisms of pressure- and flow-induced vascular responses. Blood
Vessels 1990; 27:246.

117. Anggard EE, Botting RM, Vane JR. Endothelins. Blood Vessels
1990; 27:269.

118. Mann MJ, Whittemore AD, Donaldson MC et ah Ex-vivo gene
therapy of human vascular bypass grafts with E2F decoy: the
PREVENT single-centre, randomised, controlled trial. Lancet
1999;354(9189):1493.

119. Poon M, Badimon JJ, Fuster V Overcoming restenosis with
sirolimus: from alphabet soup to clinical reality. Lancet 2002;
359(9306):619.

120. Sousa JE, Costa MA, Abizaid AC et ah Sustained suppression of
neointimal proliferation by sirolimus-eluting stents: one-year
angiographic and intravascular ultrasound follow-up. Circula-
tion 2001; 104:2007.

121. Neville RF, Tempesta B, Sidway AN. Tibial bypass for limb sal-
vage using polytetrafluoroethylene and a distal vein patch. / Vase
Surg 2001; 33:266.

122. Kim HS, Waksman R, Cottin Y et ah Edge stenosis and geographi-
cal miss following intracoronary gamma radiation therapy for
in-stent restenosis. J Am Coll Cardiol 2001; 37:1026.

123. Waksman R, Ajani AE, White L et ah Intravascular gamma radia-
tion for in-stent restenosis in saphenous-vein bypass grafts. N
Engl J Med 2002; 346:1194.

124. Dardik H, Sussman B, Ibrahim IM, Kahn M, Svoboda JJ. Distal
arteriovenous fistula as an adjunct to maintaining arterial and
graft patency for limb salvage. Surgery 1983; 94:478.

125. Ortenwall P, Wadenvik H, Risberg B. Reduced platelet deposi-
tion on seeded versus unseeded segments of expanded polyte-
trafluoroethylene grafts: clinical observations after a 6-month
follow-up. / Vase Surg 1989; 10:374.

126. Clowes AW, Gown AM, Hanson SR, Reidy MA. Mechanisms of
arterial graft failure. I. Role of cellular proliferation in early heal-
ing of PTFE prostheses. Am J Pathol 1985; 118:43.

127. Clowes AW, Reidy MA, Clowes MM. Kinetics of cellular pro-
liferation after arterial injury. I. Smooth muscle growth in the
absence of endothelium. Lab Invest 1983; 49:327.

128. Chesebro JH, Fuster V, Webster MW. Endothelial injury and
coronary vasomotion./ Am Coll Cardiol 1989; 14:1191.

129. Bush HL Jr, Jakubowski JA, Sentissi JM, Curl GR, Hayes JA,
Deykin D. Neointimal hyperplasia occurring after carotid en-
darterectomy in a canine model: effect of endothelial cell seeding
vs. perioperative aspirin. / Vase Surg 1987; 5:118.

130. Sauvage LR, Berger KE, Wood SJ, Yates SG, Smith JC, Mansfield
PD. Interspecies healing of porous arterial prostheses. Arch Surg
1974; 109:698.

131. Clowes AW, Zacharias RK, Kirkman TR. Early endothelial cover-
age of synthetic arterial grafts— porosity revisited. Am } Surg
1987; 153:501.

132. Zacharias RK, Kirkman TR, Clowes AW. Mechanisms of healing
in synthetic grafts. / Vase Surg 1987; 6:429.

133. Sauvage LR, Berger K, Beilin LB, Smith JC, Wood SJ, Mansfield
PB. Presence of endothelium in an axillary-femoral graft of
knitted Dacron with an external velour surface. Ann Surg 1975;
182:749.

134. Szilagyi DE, Smith RF, Elliott JP, Allen HM. Long-term behavior
of a Dacron arterial substitute. Ann Surg 1965; 162:453.

144

chapter 13 Intimal hyperplasia

135. Wesolowski SA, Fries CC, Hennigan G, Fox LM, Sawyer PN,
Sauvage LR. Factors contributing to long-term failures in human
vascular prosthetic grafts. / Cardiovasc Surg 1964; 5:544.

136. Kohler TR, Stratton JR, Kirkman TR, Johansen KH, Zierler BK,
Clowes AW. Conventional versus high-porosity polytetrafluo-
roethylene grafts: clinical evaluation. Surgery 1992; 112:901.

137. Wilson JM, Birinyi LK, Salomon RN, Libby P, Callow AD,
Mulligan RC. Implantation of vascular grafts lined with geneti-
cally modified endothelial cells. Science 1989; 244:1344.

138. Plautz G, Nabel EG, Nabel GJ. Introduction of vascular smooth
muscle cells expressing recombinant genes in vivo. Circulation
1991; 83:578.

139. Nabel EG, Plautz G, Nabel GJ. Site-specific gene expression in

vivo by direct gene transfer into the arterial wall. Science 1990;
249:1285.

140. Nabel EG, Plautz B, Boyce FM, Stanley JC, Nabel J. Recombinant
gene expression in vivo within endothelial cells of the arterial
wall. Science 1989; 244:1342.

141 . Lynch CM, Clowes MM, Osborne WRA, Clowes AW, Miller AD.
Long-term expression of human adenosine deaminase in vascu-
lar smooth muscle cells of rats: a model for gene therapy. Proc Natl
Acad Sci USA 1992; 89:1138.

142. Flugelman MY, Virmani R, Leon MB, Bowman RL, Dichek DA.
Genetically engineered endothelial cells remain adherent and
viable after stent deployment and exposure to flow in vitro. Circ
Res 1992; 70:348.

145

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Thoracic outlet syndrome

Herbert I. Machleder

The term thoracic outlet compression syndromes has evolved to
describe a group of disorders arising from compression of the
neurovascular structures in an area described medially by the
sternum and laterally by the insertion of the pectoralis minor
muscle on the coracoid process. It encompasses the passage
of the brachial plexus, axillary-subclavian vein, and artery
through the outlet from chest and neck to the upper extremity.
It is bound anteriorly or ventrally by the clavicle and inferior ly
and dorsally by the first thoracic rib (Figs. 14.1 and 14.2).

The anatomic site is structurally unique in at least two re-
gards: (i) the neurovascular structures are confined between
two movable bony structures, the clavicular portion of the
shoulder girdle and the first rib of the thoracic cage, and (ii) the
neurovascular bundle is traversed (and consequently subdi-
vided) by a normal anatomic structure, the anterior scalene
muscle, which can function either as a postural muscle of the
neck or an accessory muscle of respiration. The discrete symp-
tom complexes derive from the focus of compressive forces on
either brachial plexus, artery, or vein or on a combination of
these structures.

Although the medical literature on the subject is rich with
clinical descriptive material (as well as increasing editorial
musings), there is a relative paucity of fundamental scientific
and experimental research. Nevertheless, three areas of basic
investigation have begun to clarify the origins of the clinical
syndromes.

Morphologic research

The earliest basic science studies have appropriately been
morphologic in nature since thoracic outlet compression is still
fundamentally considered to be an anatomic problem. Conse-
quently, throughout the history of its medical and surgical
treatment, attempts were made to understand the underlying
structural abnormalities.

Anatomic variation at the region of the thoracic outlet has
intrigued surgical anatomists from early descriptions of
supernumerary ribs to modern studies of ultrastructural

changes in the scalene muscles. 1-3 Additionally, recent studies
have expanded earlier observations that described the
clinical significance of discrete abnormalities recognized in
routine anatomic dissections, for example, cervical ribs, cervi-
cal fibrocartilagenous bands, and supernumerary scalene
muscles. 4 ' 5

Makhoul and Machleder recently redefined these abnor-
malities in the context of embryologic and developmental
variation. 6 These authors indicated that, in their most obvious
manifestations, the congenital and acquired abnormalities are
generally considered discrete anomalies. However, they are
more effectively classified in a continuum of developmental
variation. The particular spectrum of developmental charac-
teristics may be clinically significant as predisposing elements
when complicated by increased functional requirements or a
change in muscle fiber type (or isoforms of myosin) conse-
quent to trauma.

Coincident with the range and continuum of developmen-
tal abnormalities, there is a range and continuum of compres-
sion of the normal structures that traverse the thoracic outlet.
Symptoms associated with the extremes of the compressive
abnormalities are easy to distinguish. They represent the clas-
sic cases: (i) Paget-Schroetter axillosubclavian vein occlusion;
(ii) hand ischemia from thrombosis or embolization from the
compressed or aneurysmal subclavian artery; and (iii) the
"wasted hand" of cervical band compression of the brachial
plexus. Nevertheless, lesser degrees of symptoms arise,
which may be disabling in settings of specific physical or
occupational requirements.

Greater resilience of the arterial and venous system seems to
result in relatively innocuous symptoms until compression
reaches levels of hemodynamic significance or structural
damage. It is evident that the threshold for neurogenic symp-
toms from the brachial plexus is lower than that of the vascular
system.

In recent studies of the embryologic development of the sca-
lene muscles, Milliez emphasized that anomalies rarely exist
in isolation because there is interaction of development of the
different elements. 7 White and colleagues elaborated on this

146

CHAPTER 14 Thoracic outlet syndrome

Scalenus med. m.
\

Clavicle

Brachial plexus

Scalenus ant m.

Subclavian:
-artery

r* — vein

^- — 1st rib

^Pectoralis
minor m.

Figure 14.1 Drawing of an anatomic dissection showing the thoracic
outlet area. The clavicle has been disarticulated from the sternum and
reflected laterally with the subclavius muscle and tendon (not shown). The
"floor" is the first rib, curving posteriorly from its attachment at the sternum.
Note the position of the artery and brachial plexus between the anterior and
middle scalene muscles. (From MachlederHI, ed. Vascular Disorders of the
Upper Extremity. Mt Kisco, NY: Futura Publishing, 1 989: 1 56.)

fo C, Vertebra
-A 2

Scalenus

medius

Spina! nerve 5
6

Scalenus
posterior

Scalenus
anterior

Clavicle
(cut)

Axillary
artery— 1

Axillary
vein

Median

nerve-

Radial J
nerve

Ulnar
nerve

Pectoralis
minor

fc^i PL7V*_—^ . fl3 _j^rf

wpr

Figure 14.2 Anatomic drawing of the interscalene triangle portion of the
thoracic outlet. This area is a frequent site of scalene muscle abnormalities
and post-traumatic brachial plexus compression. (A segment of the clavicle
has been removed.) (From TravellJG, Simons DG. Myofascial Pain and
Dysfunction: The Trigger Point Manual. Baltimore: Williams & Wilkins,
1983:348.)

concept in describing cervical ribs and congenital malforma-
tions of first thoracic ribs as linked to errors of bodily segmen-
tation in early embryonic development, where variation in the
formation of the brachial plexus appears before bony skeletal
development. 8

Cervical rib development, for example, is determined by the
formation of the spinal nerve roots. The regression of the C5
through C7 ribs is occasioned by the rapid development of the
enlarging roots of the brachial plexus in the region of the limb
bud. In cases of a cervical C7 rib, there is generally a prefixed
plexus with only a small neural contribution from the Tl nerve
root. The inhibition to rib development at that level is lost or
reduced, and the size of the cervical rib is then related to the
extent of contribution of this Tl root to the brachial plexus. As
a corollary, in the post fixed plexus, where there is a contribution
of the T2 root to the brachial plexus, the first thoracic rib is often
rudimentary, having been inhibited in its development by the
unusual nerve growth. 8 This embryologically determined
morphologic interdependence is evident with other structural
relationships at the thoracic outlet.

Embryologic considerations

The abdominal, thoracic, and cervical musculature develops
from the hypomeric portion of the paraxial and epaxial meso-
derm. The scalene and prevertebral muscles in the neck corre-
spond to the intercostal and ventrolateral abdominal muscles
in the thorax and abdomen, respectively. 9 In the embryo,
plates of axially running muscle segments differentiate into
the discrete muscle groups seen in the adult.

The subclavian artery, which is the artery of the seventh cer-
vical segment, as well as the spinal nerves from C5 to Tl, pierce
the muscle plates in the cervical segment much the same as the
intercostal nerve and artery do in the thoracic segments. The
growth of the limb bud and development of the pectoral
girdle then lead to the particular structural changes seen in this
region.

The important causal relationship between supernumerary
cervical ribs and compression of the subclavian artery was
recognized in the earliest descriptions and remains the most
durable concept after more than a century of clinical observa-

147

pa rt 1 Vascular pathology and physiology

Figure 14.3 Occlusion of the left subclavian artery as it passes over a
rudimentary cervical rib. This is a typical picture of an advanced vascular
injury in the arterial manifestation of thoracic outlet compression syndrome.

Figure 14.4 A characteristic venogram in a patient with right-sided
Paget-Schroetter syndrome of long standing. The thrombus in the axillary
vein has resolved, leaving the high-grade compressive abnormality at the
axillary-subclavian vein junction (A) (arrow). This corresponds to the
anatomic abnormality seen in Figure 14.10. Note the "first rib collaterals"
reconstituting central flow via the jugular vein (B) into superior vena cava (C).
This patient, while asymptomatic at rest, will have disabling symptoms of
venous hypertension with right arm exercise. (From Machleder HI. Vaso-
occlusive disorders of the upper extremity. CurrProblSurg 1988; 25:1, with
permission from Elsevier.)

tion. 1 Anatomic investigations of the Paget-Schroetter syn-
drome have likewise documented the morphologic changes
at the site of thoracic outlet compression, which results in
the so-called effort thrombosis of the axillosubclavian vein
(Figs. 14.3 and 14.4). 10 - 12

Cervical rib

During development, the C7 rib forms, then regresses to the C7
transverse process. Various stages in this evolution range from
a complete C7 rib to rudimentary forms associated with a
fibrocartilagenous band. 13,14 The only radiologic indication
of this residual band may be an enlarged C7 transverse
process. 15

Makhoul and Machleder reported that anomalies of the first
or cervical rib were encountered in 8.5% of 200 patients under-
going surgery for correction of thoracic outlet compression
(Figs. 14.5 through 14.7). 6 The incidence of these autosomal
dominant abnormalities in the general population can be esti-
mated from the medical literature. Adson and Coffee reported
that Galen and Vesalius both described cervical ribs in their
anatomic dissections. 16 ' 17

In a study of 40 000 consecutive chest X-rays in American
army recruits, Etter encountered 68 complete articulated cer-
vical ribs (0.17%), 31 anomalous first ribs (0.08%), and 67 rudi-
mentary first ribs (0.25%). He recognized 77 synostoses of
first and second ribs and 16 bifid first ribs. 18 In 1947, Adson
reviewed his experience with cervical ribs, citing a Mayo
Clinic radiologic study prior to 1927, which identified an
incidence of 0.563% or 5.6 patients per thousand with cervical
rib. Twenty-eight percent were men, 72% were women, and
47% of the cervical ribs were bilateral. The right side was in-
volved in 23% and the left side in 30% of cervical rib cases.
Forty-five percent of the group was symptomatic. 19 Adson
quotes a review by Haven of 5000 routine roentgenograms of
the thorax in which he found 38 first rib abnormalities and 37
cervical ribs, an incidence of 0.74% or 7.4 persons per 1000.
Adson subscribed to the concept linking formation of cervical
ribs to a failure of rudimentary rib regression as the nerve roots
form. 20

Firsov, in the Soviet Union, reported fluorographic exami-
nation of 510 893 people, observing 1379 cervical ribs with an
incidence of 0.27%. Women accounted for 76.8%, men for
23.2%, and 33.3% were bilateral. 21 In 102 dissections, Lang 33
found 67% of cervical ribs were bilateral, an additional 13% of
dissections had an enlarged C7 transverse process. He recog-
nized if the cervical rib measured 5.6 cm or greater the subcla-
vian artery passed over the cervical rib. If the rib measured less
than 5.1 cm, the artery crossed over the first rib, an observation
that helps clarify the likelihood of arterial injury.

As previously described, a cervical rib or C7 rib is most often
associated with a prefixed type of brachial plexus where there
is a minor contribution to the brachial plexus from the Tl nerve
root and a major contribution from C4. When the C7 rib is

148

chapter 14 Thoracic outlet syndrome

Scalenus
medius m.

Subclavian:
artery vein

Figure 14.5 Transaxillary view of the thoracic
outlet, showing the normal anatomic
relationships. (From Machleder HI, ed. Vascular
Disorders of the Upper Extremity. 2nd edn. Mt
Kisco, NY: Futura Publishing, 1989.)

Skin incision

Intercosto-
brachial n.

\ Scalenus 1st nb
Brachial plexus \ anticus m,

2nd rib

Subclavius m.

Figure 14.6 Appearance of a cervical rib from the transaxillary surgical
approach. There is generally more distortion with a greater degree of anterior
displacement and compression of the brachial plexus and axillary-subclavian
artery. (From Makhoul RG, Machleder HI. Developmental anomalies at the
thoracic outlet. J Vase Surg 1 992; 1 6:534.)

incomplete, there is often a rudimentary band in place of the
regressed portion of the rib.

From these studies, it appears that first rib and cervical rib
abnormalities occurred in a significantly higher incidence in
Makhoul and Machleder 's series of patients with thoracic out-
let compression than it does in the general population. 6 This
variation represents one of the predisposing abnormalities for
development of the clinical syndromes. The significantly
higher incidence of recognized congenital abnormalities in
women compared with men may also be reflected in the high-
er incidence of nerve compression symptoms in women.

Figure 14.7 Appearance of residual fibrocartilaginous band, attaching to
the first rib, from a cervical rib that has undergone partial regression during
embryologic development. The radiologic picture will often show an
enlarged C7 transverse process. The symptom complex can be similar to that
seen in the presence of a complete cervical rib. (From Makhoul RG,
Machleder HI. Developmental anomalies at the thoracic outlet. J Vase Surg
1992; 16:534.)

Clinical correlation

The symptom characteristics associated with the presence of
cervical ribs have been addressed by several authors. In 205
patients treated for cervical rib or thoracic outlet syndrome
(TOS) at Montefiore Medical Center, 12 patients had arterial
lesions (5.9%). 22 Halsted indicated that in 716 cases of cervical
rib he had reviewed, there were 27 cases of subclavian artery
dilatation (3.7%). He indicated that of 360 symptomatic

149

pa rt 1 Vascular pathology and physiology

clinical cases, 235 (63.3%) had nerve symptoms alone; 106
(29.4%) had nerve and vascular symptoms; and 19 (53.%) had
vascular symptoms alone. 23

Charlesworth and Brown performed 23 cervical rib exci-
sions; they indicated 15 (65%) had neurologic symptoms and
eight (35%) had vascular symptoms. 24

Telford and Mottershead 29 encountered anomalous fibrous
bands from rudimentary cervical ribs in 12 and complete
cervical ribs in 70 of 105 surgical cases. Telford and Stopford
attributed the vascular complications of cervical rib to stimu-
lation of the sympathetic innervation through the lowest
two nerve roots, a view that prevailed for many years. In
their discussion they neglected the possibility of arterial em-
bolism. 25

Fibrocartilagenous bands extending from the end of incom-
pletely formed cervical ribs are best thought of as an anomaly
of cervical rib formation. These abnormalities have been
referred to as type 1 and type 2 bands by Roos. 5

Scalene muscle

The significance of the anterior scalene muscle in thoracic out-
let compression was first recorded by Adson, who noted in the
course of treating neurovascular compression associated with
a cervical rib, that simple section of the anterior scalene muscle
relieved the compression without additional intervention. 19
Adson applied scalene tenotomy to a subset of patients who
had symptoms of cervical rib syndrome without evidence of a
supernumerary cervical rib. In 1935, Ochsner and coworkers
described a group of patients who were successfully treated
with scalenectomy, crediting Naffziger with naming the
scalenus anticus syndrome. 26 They postulated that the neu-
rovascular compressive phenomenon was a result of entrap-
ment of the subclavian artery and brachial plexus by a taut,
chronically contracting anterior scalene muscle. The increased
tone in this muscle would elevate the first rib, thereby com-
pressing the brachial plexus against the undersurface of the
clavicle. They commented on possible structural changes and
dynamic mechanisms.

Despite these and similar intraoperative observations
of muscle abnormalities, histopathologic substantiation of
changes in scalene muscle had been poorly characterized by
the available cytochemical methods.

Milliez in his recent studies of scalene muscle in a 2.5-cm
embryo emphasized the influence of neurovascular structure
development on the ultimate configuration of the scalene
muscle mass. 7 He recognized a confluent scalene muscle, dis-
tinguished only by a groove at the site of anterior and middle
scalene differentiation. Actual separation into two muscles oc-
curred at the points where the muscle mass was traversed by
the roots of the brachial plexus. It was postulated that rather
than the scalenus minimus muscle forming as a separate
muscle entity, it represents one form of segmentation of the
scalenic mass (Fig. 14.8).

Brachial
plexus

Subclavian:
artery vein

Scalenus
medius m.

Subclavius
tendon

Scalenus
anticus m.

Scalenus
minimus m.

Figure 14.8 The most common supernumerary scalene muscle
abnormality: the scalenus minimus muscle. (From Makhoul RG, Machleder
HI. Developmental anomalies at the thoracic outlet. J Vase Surg 1992;
16:534.)

Milliez quotes Poitevin's studies of 1988 indicating that the
scalenic mass is only differentiated into specific muscle groups
by the traversing of the neurovascular bundle. The persistence
of certain muscle inclusions in the brachial plexus, as well as
muscle groups that traverse various elements of the brachial
plexus, is related to the original mass of the scalene variously
fragmented by the passage of these developing structures
as the limb bud develops. He acknowledges, however, an
alternate opinion that accessory scalene muscles maybe repre-
sentations of phylogenetic recapitulation wherein muscles
persist, which ordinarily would regress either completely or to
a small tendon.

Clinical correlation

The separation of muscle bundles interdigitating between the
neurovascular structures accounts for the muscular bridges
seen between the middle and anterior scalene, which often
penetrate the brachial plexus. Sanders and Roos, studying the
anatomy of the interscalene triangle, found interdigitating
fibers between the scalene muscles through the brachial
plexus in 75% of dissections in thoracic outlet syndrome (TOS)
patients and in 40% of 60 cadaver dissections. In 45% of cadav-
er dissections, the C5, C6 nerve roots emerged between fibers
of the anterior scalene muscle, rather than between the an-
terior and middle scalene muscles. 27 Although not pathologic,
these fibers may result in neurogenic symptoms as a conse-
quence of later abnormal growth, peculiarities of occupational
or recreational activities, or post-traumatic changes.

150

chapter 14 Thoracic outlet syndrome

Adson, Ochsner, Naffziger and coworkers made the initial
observations of scalene muscle involvement in neurovascular
compression at the thoracic outlet. 26 ' 28 Other investigators
added quantitative anatomic measurements. In a series of me-
thodical dissections, Lang 33 found the anterior scalene inser-
tion to be a mean of 16.15 mm with a range of 7.2-23 mm. The
width of insertion of the scalenus medius was a median of 23.2
cm (range 15-28 mm). The base of the interscalene triangle at
the first rib averaged 8.3 mm and ranged from to 19 mm in
102 dissections.

Telford and Mottershead found the ventral attachment of
the anterior scalene to vary from 2.4 to 6.0 cm from the chon-
drosternal junction, and the width of the anterior scalene in-
sertion to vary from 0.4 to 2.5 cm with the interscalene interval
varying from to 2.4 cm. They encountered crossing of the in-
sertions of the anterior and middle scalene muscles in eight of
105 operative cases and in 15 (15%) of 102 cadaver dissec-
tions. 29 Sunderland and Bedbrook encountered this crossing
of the scalene insertions in 15% of 35 cadaver dissections
(Fig. 14.9). 30

This abnormality, which in the extreme is a sling-like cross-
ing of the scalene muscles described by some as a V-shaped de-
formity that traps the subclavian artery and brachial plexus,
corresponds to the type 4 band of Roos. 5 It is described by
Makhoul and Machleder as intercostalizaion to reflect the em-
bryonic derivation of these muscles (first recognized by
Todd). 31 Roos subdivided these crossing bands into an eighth
and ninth abnormality in 1980; the type 8 represents a band
from the middle scalene to the costochondral junction, and the
type 9, a fascial band in the concave curve of the first rib. 32

Despite some common variations of scalene muscle inser-
tion, these unique configurations that tend to compress the
neurovascular structures in the interscalene triangle have
been identified, and accounted for 43% of the variations seen
in Makhoul and Machleder 's series. 6

The scalenus minimus muscle, present in 10% of this series,
can be represented by a residual ligament, which has been
called the costovertebral or pleurospinal ligament when the mus-
cle has regressed. Roos described these as type 5 or type 6
bands, depending on the site of insertion to rib or apical
pleura. Lang noted a scalenus minimus or residual ligament in
39% of dissections. 33

Subclavius anomalies

Telford and Mottershead noted in anatomic dissections that
during the movements of abduction or retraction of the shoul-
der the tendon of the subclavius muscle compressed the sub-
clavian vein against the first rib. 29 In the Paget-Schroetter type
deformity, Sampson emphasized the striking hypertrophy
of the subclavius tendon; these observations were later
corroborated by Aziz and colleagues and Kunkle and
colleagues. 12,34,35

This manifestation of compression at the thoracic outlet is
related to progressive enlargement of the subclavius muscle
system with repetitive compressive trauma to the subclavian
vein followed by fibrosis, stricture, and thrombosis. An ab-
normality in this system was found in 19.5% of Makhoul and
Machleder 's cases with 15.5% having an exostosis at the sub-
clavius tubercle (Fig. 14.10 and see Fig. 14.4). 6

Scalenus
medius m.

Scalenus
anticus m.

Figure 14.9 Incomplete separation of the scalene muscle mass or
intercostalization of the anterior and middle scalene. This leads to a sling-like
entrapment of the neurovascular structures. (From Makhoul RG, Machleder
HI. Developmental anomalies at the thoracic outlet. J Vase Surg 1992;
16:534.)

Subclavian v.

Exostosis

Subclavius m
tendon

Scalenus
anticus m.
tendon

Figure 14.10 The typical abnormality seen in the Paget-Schroetter
syndrome of axillary— subclavian vein thrombosis. Compare this with the
radiologic picture seen in Fig. 1 4.4. (From Kunkel JM, Machleder HI.
Treatment of Paget-Schroetter syndrome. Arch Surg 1989; 124:1 153.)

151

pa rt 1 Vascular pathology and physiology

Clinical correlation

The Paget-Schroetter syndrome of spontaneous, effort-
related axillosubclavian vein thrombosis is a consequence of
repetitive trauma of the vein in the thoracic outlet. The algo-
rithm for management is based on this renewed understand-
ing of the pathologic anatomy 12 ' 36 In a review of the anatomic
abnormalities in 200 consecutive clinical cases, correlations
can be drawn with the embryologic investigations. Forty cases
(20%) presented with problems related to venous obstruction,
or arterial insufficiency and embolization. Twenty-six cases
(65%) of this vascular subgroup had additional neurogenic
symptoms associated with abnormal sensory evoked respons-
es in the affected extremity.

In the group of 200 symptomatic extremities, 153 (76.5%)
exhibited signs of arterial compression in stress positions on
clinical examination or with digital photoplethysmography.

In 68 cases (34%), there was no abnormality discernible from
the transaxillary surgical approach. Seventeen cases (8.5%)
had a cervical rib articulating with the first rib directly or by fi-
brocartilagenous extension. This group included associated
anomalies of the first rib. Twenty cases (10%) had a scalenus
minimus abnormality inserting either on the first rib or
Sibson's fascia. Thirty-nine patients (19.5%) had an anomaly
of the subclavius tendon or its insertion tubercle. Eighty-six
cases (43%) had an anomaly of scalene muscle development or
insertion. In 15 cases (7.5%), there were other categories of
anomalies that could not be related to specific developmental
characteristics. These included ligamentous or fibrous struc-
tures that did not correspond to regression residua of recog-
nized embryologic structures. More than one abnormality was
recognized in 22.5% of the cases, and 32% of 25 patients
undergoing bilateral procedures had similar anomalies on
both sides.

In review, there was only one clinical setting that could be
correlated with characteristic anatomic abnormalities. Of
33 patients presenting with spontaneous axillosubclavian
vein thrombosis (Paget-Schroetter syndrome), 18 (55%)
had hypertrophy of the subclavius tendon associated with
enlargement of the insertion tubercle. Among male patients
with Paget-Schroetter syndrome, 14 (70%) had this
anomaly 6 ' 12

Despite the congenital nature of these deformities, the onset
of symptoms in early to mid adult life has been recorded by
virtually all physicians treating the clinical disorders. This
delay in onset is most likely related to postnatal development.
The chest widens and the clavicle continues its growth to ap-
proximately age 25, after which the pectoral girdle begins to
descend. With loss of strength or tone in the supporting mus-
culature of the thoracic girdle, there is further traction on the
neurovascular structures at the thoracic outlet. 37

Ultrastructural studies

In 1986, initial ultrastructural studies of the anterior scalene
muscle were reported from the UCLA Neuropathology Labo-
ratories. Distinctive fiber type changes were seen in patients
with post-traumatic compression of the brachial plexus. 2,38
These observations have subsequently been amplified by
other laboratories. 3 There is also an increasing body of evi-
dence that myosin isoforms and muscle fiber phenotype
changes occur in response to injury or altered physiologic
environments both in the experimental situation and in a
number of human disorders. 39-41 Similar distinctive changes
have been described in muscular systems as diverse as the
perineum, nasopharynx, and myocardium. 42-45

Overview

Historically, intraoperative observations and basic histologic
studies have suggested structural abnormalities in scalene
muscle in some patients with neurogenic thoracic outlet
compression. Recently developed histochemical and mor-
phometric techniques have facilitated reassessment of the
pathophysiology of possible muscular dysfunction in thoracic
outlet compression syndrome. Morphologic transformations
of muscle fibers reflecting metabolic and enzymatic changes,
characteristic of various adaptive and pathologic responses,
can be demonstrated by specific staining techniques.

Vertebrate skeletal muscle is composed of several distinc-
tive myofiber types; each has different morphologic, metabol-
ic, and contractile characteristics that are distinguishable by
specific histochemical staining methods. 46 Most mammalian
muscle is composed of a mosaic of two muscle fiber types
conventionally referred to as type I and type II, with the latter
predominating.

Type II has been called fast twitch to characterize its physio-
logic response to stimulation. It is white, reflecting its vascular
supply and paucity of blood pigments consistent with the
predominantly anaerobic, glycolytic nature of its metabolism.

Type I fibers are referred to as slow twitch to characterize the
propensity for fatigue-resistant sustained contraction. They
are red, reflecting increased blood pigments and vascular sup-
ply. They stain weakly for glycogen content and glycolytic
enzymes, and strongly for oxidative enzymes, consistent with
metabolism via predominantly aerobic pathways.

In a number of congenital and metabolic muscle disorders,
selective atrophy and hypertrophy occur preferentially in one
or the other fiber system, as well as alterations of normal ratios
of fiber types and subtypes. Changes of fiber subtype group-
ing may also occur in denervation and reinnervation of muscle
in many types of lower motor neuron disorders.

Each motor unit, composed of a single motor neuron and its
axon in a peripheral nerve, innervates a group of muscle fibers,

152

chapter 14 Thoracic outlet syndrome

all of the same histochemical and physiologic type. These
groups are arranged randomly through the muscle fascicles
with fibers of one motor unit overlapping and interdigitating
with those of adjacent motor units of opposite fiber type. This
results in a mosaic or checkerboard pattern of histochemical
fiber types when viewed in cross-section and stained histo-
chemically Most descriptions of normal human muscle indi-
cate a type II fiber predominance with several exceptions, such
as deltoid and soleus muscles, which may have a 60-80% type
I myofiber predominance. This type I predominance has been
demonstrated in normal anterior scalene muscle. 2 ' 47 It has
been reasoned that the wide distribution of motor units and
the intermingling with muscle fibers belonging to other motor
units serves to disperse muscle action potentials so that
self-excitation will not lead to continuous contraction.
This arrangement facilitates the feedback control of muscle
tension. 48

Despite a high degree of specialization in mammalian mus-
cle, a considerable capacity for accommodating changes in
demand and patterns of stimulation is retained by responding
with alterations in basic biochemical elements. Under
conditions of chronic stimulation, a predominantly type II
fast twitch muscle can undergo an orderly sequence of
changes, ultimately causing complete transformation to
a slow twitch, predominantly type I fiber muscle. 49-51
When subjected to prolonged periods of increased stimulation
or contraction, the activity of enzymes of aerobic metabolism
increases; the activity of enzymes of anaerobic metabolism
decreases; and predominant myosin isoforms change.
Such changes in myosin synthesis, characteristic of either
type I or II fibers, are reversible. The transformation
occurs in individual fibers adaptively responding to
developmental and pathologic departures from normal
activation patterns. 52

Human skeletal muscle usually comprises predominantly
type II, quick-reacting fibers that have low oxidative enzyme
capacity and increased reactivity with phosphorylase and
myosin adenosine triphosphatase. A smaller percentage of
slow tonic contracting type I fibers, characterized by greater
oxidative capacity, completes the complement. These enzyme
systems within muscle fibers are largely determined by the
pattern of contractile activity, with tonic stimulation increas-
ing oxidative activity and decreasing glycolytic activity. These
changes are manifest histochemically by reduced staining
reactivity with phosphorylase and myosin adenosine tri-
phosphatase, and greater reactivity with nicotinamide
adenine dinucleotide-tetrazolium reductase, a mitochondrial
oxidative enzyme. 53-55

Histochemical studies

Initial muscle studies in thoracic outlet compression patients
have followed a comprehensive investigative protocol.

The algorithm for analysis has been described by Dubowitz

and Brooke. 56

Hematoxylin and eosin: This basic staining procedure provides
an overview of pathologic changes, primarily evidence of
vasculitis and inflammatory cell infiltrates. Some fiber type
differentiation is possible.

Modified Gomori's trichrome: This technique augments the
hematoxylin and eosin stain in the evaluation of general
pathologic processes and highlights abnormal accumula-
tion of mitochondria, such as in the "ragged-red" fibers,
which may be present in various pathologic processes, such
as storage myopathies or polymyositis.

Oil red O: This technique is used to detect neutral lipid
droplets, which are more abundant in type I than type II
fibers. Mitochondrial myopathies and the myelin sheaths of
intramuscular nerve twigs are also demonstrated with this
stain.

Periodic acid-Schiff: This staining technique is used with
and without diastase digestion to demonstrate normal and
abnormal deposits of glycogen. Basement membrane and
plasmalemma about myofibers are stained, outlining the
shape of the myofibers and facilitating the recognition of
abnormally small or irregular fibers. Target fibers, which
reflect focal degenerative changes within denervated type I
muscle fibers, are identified and document denervation as
opposed to myopathic atrophy.

Succinic dehydrogenase: This highlights mitochondria and
identifies oxidative enzyme activity.

Nicotinamide adenine dinucleotide-tetrazolium reductase: Type
I myofibers are stained darker than type II with this stain
as it reflects oxidative enzyme activity. Atrophic and
denervated fibers are identified by their dark staining
characteristics with this histochemical reaction, as well as
by the appearance of target, targetoid, and "moth-eaten"
fibers.

Myofibrillar adenosine triphosphatase: This is used to demon-
strate the various fiber types after preincubation at a pH
of 9.4, 4.6, and 4.3. Types I and II and fiber subtypes can be
identified to assess relative numbers, size, and distribution
of the fiber types.

Myophosphorylase: This is used to detect the presence of glyco-
gen and of myophosphorylase enzyme. This enzyme is
absent in necrotic fibers and demonstrates the target phe-
nomenon in the presence of denervation.

Phosphatase: Muscle specimens are subject to both acid and
alkaline phosphatase reactions. The acid phosphatase
reaction accentuates increased lysosomal activity, which is
seen in necrotic and degenerating myofibers in many my-
opathies. Ordinarily, normal muscle shows little activity
with this stain. The alkaline phosphatase reaction demon-
strates abnormally reactive blood vessels in inflammatory
conditions. Regenerating myofibers are highlighted, as
well as fibers of denervated muscle. As with the acid phos-

153

pa rt 1 Vascular pathology and physiology

phatase reaction, little or no alkaline phosphatase activity is
seen in normal skeletal muscle sections.

Morphometric studies

Researchers studying fiber typing in anterior scalene muscle
have reported their analyses according to the nomenclature
of Brooke and Kaiser, based on myosin adenosine triphos-
phatase with acid or alkaline preincubation. 57 The lesser diam-
eters of at least 250 fibers are analyzed in three random
microscopic fields at final photographic magnification of xl50.

Hypertrophy and atrophy factors are determined from
muscle fiber histograms using the method of Brooke and
Engle. 53 This quantitative assessment represents the percent-
age and distribution of large (hypertrophic) and small
(atrophic) fibers in the biopsy.

In the initial studies of anterior scalene muscle reported in
1986, control muscle demonstrated a 70% type I fiber predomi-
nance (ratio 4.7 : 1). The mean type I fiber size was 5.08+2.1 jam,
with a normal hypertrophy factor and atrophy factor distribu-
tion. 2 Both hypertrophy factor and atrophy factor were less
than 0.10 for type I fibers. In this early study of biopsy speci-
mens mainly from women, limits of 30 and 70 jam have been
used to calculate the atrophy factor and hypertrophy factor,
respectively.

Scalene muscle from patients with post-traumatic neuro-
genic thoracic outlet compression was characterized by
marked type I fiber predominance of (5.7: 1) 85.1 + 5.1% and
type I fiber hypertrophy, with mean + SD fiber size of 55.6 ±
2.7 jim and a hypertrophy factor of 0.28 ± 0.08. The paucity of
small fibers is indicated by an atrophy index under 0.05. There
was no concomitant hypertrophy of type II fibers, and the
mean size was 32.2 ± 3.1 jam.

After scalene tenotomy for thoracic outlet compression,
biopsies of anterior scalene muscle at various intervals show
reduction in type I fibers, representing 77.0 ± 6.8% of the distri-
bution. There is marked atrophy of type I fibers (atrophy factor
0.66 ± 0.24) and in the type II fiber system (atrophy factor 0.74 ±
0.23). Hypertrophy indices were less than 0.10 in both fiber
systems (Fig. 14.11).

Fiber type transformation is even more striking in patients
who have relatively long-standing TOS with type I /type II
ratios up to 66 : 1, representing subtotal type I fiber transfor-
mation with hypertrophy factor of 0.34 and atrophy factor of
less than 0.01.

In a study of 66 scalene muscle specimens from patients
with thoracic outlet compression syndrome, all showed type I
fiber predominance varying from slightly over 50% to over
95%. Twenty-three specimens had over 80% type I, with 32
specimens showing subtotal or 90-95% type I predomi-

Figure 14.1 1 Representative microscopic fields of two anterior scalene
muscles, demonstrating the typical mosaic pattern with fiber-type staining.
(A) Type 1 fiber (pale staining) hypertrophy and predominance. This is the
characteristic pattern seen in patients with post-traumatic thoracic outlet
compression syndrome. (B)A muscle after tenotomy with type 1 fiber

atrophy and increased proportion of type 2 (dark staining) fibers (myosin
adenosine triphosphatase, pH 9.3 x 1 10). Both slides are at the same
magnification. (From Machleder HI, Moll F, Verity A. The anterior scalene
muscle in thoracic outlet compression syndrome. Arch Surg 1 986;
121:1141.)

154

chapter 14 Thoracic outlet syndrome

nance. 40 Type IA and IB myofiber subtypes were identified in
14% of specimens. Additional myopathic changes were signs
of focal denervation and reinnervation in 11 specimens (17%)
with seven cases of atrophy in the type II system (limited by
the small number of type II fibers available for analysis).
Areas of reduced myophosphorylase were seen in 17%
of specimens with no evidence of either necrosis, hyaliniza-
tion, or inflammation (except in one patient with prior
brachial plexus exploration). Three specimens showed minor
changes of endomesial sclerosis. Other scattered myopathic
changes were encountered including elevated vascular
alkaline phosphatase (15%), areas of dehydrogenase lucency
(35%); central nucleation (3%), and sarcolemmal nuclear
aggregates (18%).

Clinical correlation

Fiber type transformation in skeletal muscle has been de-
scribed in the clinical situation by Munsat and coworkers. 58
They reported myofiber transformation as a consequence of
prolonged stimulation with a relative increase in numbers of
type I fibers. Salmons and Sreter and Salmons and Vrbova
demonstrated that a continuous low-frequency discharge of
motor neurons can establish and maintain a slow time course
of contraction in postural muscle, producing fiber-type
changes as a result of altered levels of contractile or metabolic
activity. 49 ' 59 Even more pertinent to the studies of anterior
scalene muscle is the reported increase of relatively fatigue-
resistant type I fibers (up to 54% of fiber content) demonstrat-
ed in human respiratory muscle in response to increased
respiratory loads. 60 ' 61 These changes may be analogous to the
transformation occurring in scalene muscle, which can per-
form either as an accessory muscle of respiration or a postural
muscle, depending on the pattern of contraction.

The facility of fiber transformation when documented in
human fitness training is confined to younger individuals. 62 ' 63
The demonstrable loss of this capacity in middle-aged men
may explain the predominance of the observed scalene muscle
transformation and subsequent TOS predominantly in
women in the second and third decades of life. 64

After tenotomy, muscle fiber composition can reverse de-
spite continued stimulation, and selective atrophy can occur
in the type I fiber system with little change in the type II fiber
system. This previously described transformation in human
muscle is consistent with the observations in post-tenotomy
scalene muscle.

All control and affected scalene muscle studied in patients has
demonstrated type I fiber predominance, indicating that this
muscle is at the outset uniquely structured in fiber composi-
tion to sustain protracted periods of tonic contraction. The
striking increases in the type I fiber system observed in pa-
tients with TOS may be a consequence of changes in metabolic
demands or stimulation, or a response to denervation and
reinnervation. These initiating events may arise anywhere in

the lower motor neuron system or following trauma to the sca-
lene muscle, disrupting the continuity of the motor unit. After
tenotomy of the scalene muscle, selective atrophy of the type I
fiber system occurs, followed by reverse transformation of
type I to type II fibers, or selective loss of the hypertrophied
type I fibers.

The anterior scalene muscles in patients with TOS, there-
fore, demonstrate an extraordinary adaptive transformation
and recruitment response in the type I fiber system, possibly
reflecting chronic increased tone or motor neuron stimulation.
It seems particularly likely that in post-traumatic TOS, stretch
injury to the muscle initiates a response that, if uninterrupted,
serves to accentuate and perpetuate the neurovascular com-
pressive phenomenon. In a small percentage of individuals
involved in hyperextension neck injuries (such as a "rear-
ended" automobile accident), there will be gradual develop-
ment of signs and symptoms of brachial plexus compression
in the interscalene triangle (see Fig. 14.2). This often will occur
months after resolution of the initial symptoms of musculo-
ligamentous strain.

Electrophysiologic studies

When advanced, the neurogenic symptom complexes of
the TOS are associated with characteristic electrophysiologic
changes. At the outset, however, it is important to recognize
the limitations of the electrophysiologic evaluation. Pain and
dysesthesia are the predominant symptoms of patients pre-
senting with brachial plexus compression at the thoracic
outlet, particularly when these symptoms are positionally
enhanced. This pain is considered to be mediated by the
smaller myelinated or unmyelinated nerve fibers, the integrity
or function of which is not tested by the standard electrophys-
iologic techniques. In advanced cases of brachial plexus com-
pression where weakness and atrophy are evident on clinical
examination, there will generally be concomitant abnormali-
ties on electromyography and nerve conduction studies.

A more sensitive method for determining abnormalities
in these larger fibers would be the use of electrical tests in
the symptomatic positions. Although this is feasible for carpal
tunnel compression of the median nerve, the parameters have
not been established for the brachial plexus. It is likewise im-
portant to recognize that there are substantial variations in
normal nerve conduction that tend to limit sensitivity and
specificity. The electrophysiologic findings must frequently be
interpreted in light of the presenting signs and symptoms.

Electromyography

Reviewing the anatomic relationships at the thoracic outlet, it
should be recalled that the C8-T1 nerve roots, or inferior trunk
of the brachial plexus are most likely to be compressed against
the bony resistance of the first rib. In advanced cases of neuro-

155

pa rt 1 Vascular pathology and physiology

genie thoracic outlet compression with muscle atrophy,
characteristic electrophysiologic changes can be seen in the
abductor pollicis, opponens pollicis, first dorsal interosseous,
and abductor digiti minimi muscles. The physiologic
changes seen are those of chronic denervation with reinner-
vation, specifically, prolonged and polyphasic motor unit
potentials.

With the degeneration of some axons, their muscle fibers
will become "orphans". The orphaned muscle fibers begin
to atrophy, then become reinnervated by peripheral axon
sprouts growing from nearby intact motor units. The sprouts
may conduct a motor impulse slowly and irregularly so that
the newly reinnervated fibers are activated asynchronously
compared with the directly innervated fibers in the same
motor unit. This is manifest by prolongation of the electrical
potentials, as well as irregular configurations (polyphasic) in
these reinnervated motor units.

Nerve conduction

In compressive neuropathic processes, like the thoracic outlet
compression syndrome, abnormalities of nerve conduction
may underly the clinical manifestations. Damage to a myelin
sheath due to a mild nerve compression injury produces either
conduction slowing or total block of a nerve impulse traveling
distally through a single nerve fiber when stimulated proxi-
mally With a proximal stimulus, the distal recording may show
a low-amplitude electrical potential or a delay in the distal
response (increase in latency), dependent on the number of
damaged fibers in a given nerve. When a nerve is stimulated
distal to the site of injury the impulse characteristics will de-
pend on whether there is only myelin damage or axonal de-
generation. The technical difficulty of stimulating the C8-T1
never roots or lower trunk of the brachial plexus (that is proxi-
mal to the site of usual compression) and recording distally
has largely invalidated the use of peripheral nerve conduction
for the diagnosis of neurogenic thoracic outlet compression
syndrome.

In far advanced neurogenic cases, there may be sufficient
axonal degeneration that the sensory nerve action potential is
reduced in size all along the course of the ulnar nerve. This
type of abnormality can be detected with measurement of
ulnar sensory nerve action potentials generated and recorded
distally.

F-wave studies

When a peripheral nerve is stimulated percutaneously, nerve
action potentials are propagated in opposite directions, proxi-
mally and distally. When the centrally propagated impulse
(termed retrograde or antidromic) reaches the motor neuron in
the spinal cord, some of the impulses will be reflected back
down the axon in an orthodromic direction. This will produce

secondary action potentials recordable from muscles in the
hand. This reflected potential is called an F-wave and represents
successful antidromic and orthodromic impulse propagation
occurring in an individual axon. This implies that there has
been satisfactory propagation of the impulse across the root
and proximal portions of the brachial plexus. The test is
limited by the effectiveness of even a few functioning fibers
in eliciting a near normal F-wave response. In some settings,
quantitative measurements can enhance the sensitivity of this
electrophysiologic response.

Somatosensory evoked potentials

Evoked potentials are voltage changes that can be recorded
over many portions of the nervous system in response to a
variety of peripheral sensory stimuli. These recordings repre-
sent the functional integrity of the neural pathway. When
median or ulnar nerves are stimulated at the wrist, they give
rise to a typical high-amplitude peak at the brachial plexus, re-
ferred to as the Erb's point peak and designated by the charac-
ters N9. The conduction occurs principally along the lemniscal
system, a large fiber sensory pathway associated with vibra-
tion and joint-position sense. These pathways do not conduct
pain and temperature sense and impose another limitation in
the electrophysiologic characterization of neurogenic TOS
patients who, in the early stages, present predominantly with
pain.

Measurement of the voltage amplitude at the brachial
plexus as well as the time from stimulation to appearance of
the potential (latency) will reflect any pathologic process in
the referenced nerve. Peripheral neuropathies that disrupt
myelin can slow conduction velocity and thereby delay
the latency to the Erb's point peak. Neuropathies that impair
axonal function without disrupting myelin are associated
with decreased peripheral peak amplitude but relatively nor-
mal latency. Compression can result in a partial block of nerve
function along a brief stretch of the overall peripheral path-
way. This type of compression can result in a loss of amplitude
from the reduction in conduction and the loss of some path-
way fibers. This does not usually result in any substantial in-
crease in latency.

The evoked potential can be used to determine anatomi-
cally the level of lesion in the sensory pathway. The general
level of impairment can be deduced to be at peripheral, cervi-
cal, brainstem, or hemispheric level, depending on the charac-
ter of peaks recorded at Erb's point (at the brachial plexus); at
the cervical spinal cord (N13); or over the contralateral cortex.
Changes in latencies of the peaks generally indicate a demyeli-
nating process, whereas changes in amplitudes generally indi-
cate an axonal or compressive lesion. 65

For evaluation of brachial plexus compression in TOS, both
median and ulnar nerve stimulation at the wrist is used.
As in the electromyographic findings, abnormalities are

156

CHAPTER 14 Thoracic outlet syndrome

RIGHT ULNAR PREOPERATIVE

Figure 14.1 2 Preoperative sensory evoked
potential (SEP) recording in a patientwith
bilateral thoracic outlet compression syndrome.
On the vertical axis, nerve action potentials from
three separate anatomic sites are recorded:
central, over the contralateral cortex; cervical,
over the cervical spine; and Erb's point (star),
over the brachial plexus. Note the small
peripheral peakoverthe brachial plexusat8 ms.
The cervical peak is also small, indicating
interference with the nerve action potential
proceeding centrally from the peripheral ulnar
nerve stimulation. The central peak is normal
and reflects the phenomenon of central
augmentation of a weak peripheral signal.

Cenlra!

Periphera
peak

Erb's

Figure 14.13 Preoperative recording from left
ulnar nerve stimulation. Note the low peripheral
peak (arrow) at N9 (the brachial plexus
potential). This indicates impairment of
peripheral nerve impulse conduction through
the brachial plexus.

LEFT ULNAR PREOPERATIVE

Centra

found predominantly in the ulnar nerve pathway, which
traverses the lower trunk of the brachial plexus exclusively
The effect of this compression is to cause a loss of amplitude of
the Erb's point peak (N9) for the ulnar nerve (Figs. 14.12
through 14.14).

The clinical observation that these patients are considerably
more symptomatic with their arms in the overhead position
(abduction and external rotation) is likewise represented in
the evoked potential response. When the arm is placed in the
stress position, the abnormality is enhanced.

The normal population variations in evoked potential
amplitude make it difficult to use absolute amplitude as a
criterion for abnormality. This problem has been solved by
normalizing the measurements using ratios of amplitudes
within individual patients or subjects. All investigators have
taken this general approach, comparing the ratios between left
side and right side or between the median nerve and ulnar
nerve. Another method has been to compare peak amplitude
with the arm in a neutral position along the patent's side, with
the peak amplitude with the arm in a stressed position with the

157

PART 1 Vascular pathology and physiology

LEFT ULNAR POSTOPERATIVE

Central

Figure 14.14 After left transaxillary first rib
resection, repeat sensory evoked potential (SEP)
recordings show improvement in the brachial
plexus action potential peak at Erb's point
(arrow). This amplitude is now normal and
reflects improved nerve conduction through the
brachial plexus, coincidentwith relief of
radicular neuropathic symptoms. Compression
of the inferiortrunkof the brachial plexus was
evident intraoperatively. (From Machleder HI,
Moll F, Nuwer M, Jordan S. Somatosensory
evoked potentials in the assessment of thoracic
outlet compression syndrome. J Vase Surg
1987;6:177.)

hand positioned above the head. The latter position tends to
increase compression at the thoracic outlet in symptomatic in-
dividuals. This exacerbation can be measured as a reduction in
evoked potentials in that position compared with the neutral
position.

The effect of additional compression at the thoracic outlet
may cause changes because of direct mechanical effects, com-
pression of the vasa nervorum, or as a result of ischemia to the
limb from arterial compression. Normal parameters for ampli-
tude ratios have been established by a number of investigators
and lie in a narrow range. 66-68

Many patients with neurogenic thoracic outlet compression
gradually develop bilateral symptoms, which may render the
right /left amplitude comparisons insensitive for diagnosis. In
these situations, calculation of median to ulnar amplitudes,
as well as neutral to stressed position amplitudes, has been
found to increase both sensitivity and accuracy without affect-
ing specificity.

The median nerve serves a sensory function for the thumb
and the two to three adjacent fingers. That sensory region is
extensively represented on the cortical homunculus, corre-
sponding to the importance of the thumb in humans. The
peripheral and central portions of this pathway both produce
large, well defined evoked responses. The ulnar-generated
evoked response corresponding to the sensory function for
the fourth and fifth fingers is considerably smaller. The ulnar
nerve peripheral peak is often only about two-thirds the
height of the corresponding median nerve peak. The limit of
normal of 30% ulnar /median would reflect a drop of 50% from
the usual average, and is the threshold for diagnosing signifi-
cant brachial plexus compression.

With change in arm position to the abducted externally

rotated position, there is commonly a peripheral peak ampli-
tude change of 20-30%. The limit of normal variability appears
to be about 50%. A drop of more that 50% is considered indica-
tive of compression. In patients with neurogenic TOS, such a
change in position often completely abolishes the peripheral
peak despite a constant stimulus represented by its associated
abducting twitch of the fifth digit.

The accuracy of the sensory evoked potential (SEP) test in
TOS has been further validated by preoperative and postoper-
ative studies. Ninety- two percent of patients with abnormal
evoked potentials preoperatively had normalization of the
evoked potential postoperatively if they had relief of symp-
toms. Studies of patients with classical clinical findings of
TOS but normal SEP tests have additionally shown increase
in the SEP amplitudes, over and above the previous normal
amplitudes, after undergoing corrective surgery that resulted
in symptomatic improvement. The accuracy of the SEP
test is enhanced by near 100% specificity at the cost of lower
sensitivity. 69 ' 70

Clinical correlation

Patients with neurogenic TOS often incur symptoms in the
overhead position (abduction and external rotation), typically
developing numbness of the fourth and fifth fingers and ulnar
aspect of the forearm (as they do during the SEP test). They
rapidly note impairment in strength and endurance with
exercise in abduction and external rotation. This will affect
mechanics, painters, stone masons, electricians, and others
who work with arms in abduction.

In addition, fine finger movement and grip strength become
progressively impaired as patients develop difficulty

158

chapter 14 Thoracic outlet syndrome

performing tasks in both the abduction and external rotation
positions and with their arms in neutral positions. This
accounts for the disabilities in many patients in industrial and
white collar repetitive-motion occupations.

The loss of tactile ability and fine motor coordination is a
consequence of impairment in the C8-T1 nerve root contribu-
tion to both median and ulnar nerves.

Median nerve innervated muscles such as the opponens
pollicis, and lumbricalis-interossei on the radial side are sup-
plied completely by the C8-T1 roots. The flexor pollicis longus
and brevis, as well as the abductor pollicis brevis, derive sub-
stantial innervation from C8-T1 as does flexor digitorum pro-
fundus and sublimis.

Most of the ulnar innervated muscles derive their entire in-
nervation from C8-T1. That includes the flexor digitorum pro-
fundus (fingers 3 and 4), adductor pollicis, dorsal interossei,
palmar interossei, and abductor digiti quinti. The motor inner-
vation of the fifth finger, opponens digiti quinti and flexor
digiti quinti is predominantly from the inferior trunk of
brachial plexus (C8-T1).

These sensory and motor changes in neurogenic thoracic
outlet compression explain the complaints and disabilities de-
scribed by patients in various occupational and daily life situ-
ations. They also emphasize the necessity for differentiating
compression of the inferior trunk of brachial plexus from iso-
lated median nerve compression at the wrist and ulnar nerve
compression at the cubital tunnel or Guyon's tunnel. In this
regard, the clinical evaluation is enhanced by careful electro-
physiologic assessment.

Conclusion

It should be evident that many of the clinical conundrums that
still surround the thoracic outlet compression syndromes will
yield to continued basic experimental efforts, and represent a
fertile field for the inquisitive investigator. This research is of
particular consequence when it is recognized that these dis-
abling disorders primarily affect otherwise healthy working
men and women.

References

1. Coote H. Exostosis of the left transverse process of the seventh cer-
vical vertebra, surrounded by blood vessels and nerves: success-
ful removal. Lancet 1861; 1:360.

2. Machleder HI, Moll F, Verity A. The anterior scalene muscle in
thoracic outlet compression syndrome: histochemical and
morphometric studies. Arch Surg 1986; 121:11414.

3. Sanders RJ, Jackson CGR, Banchero N, Pearce WH. Scalene
muscle abnormalities in traumatic thoracic outlet syndrome. Am
J Surg 1990; 159:231.

4. Law AA. Adventitious ligaments simulating cervical ribs. Ann
Surg 1920; 72:497.

5. Roos DB. Congenital anomalies associated with thoracic outlet
syndrome. Am J Surg 1976; 132:771 .

6. Makhoul RG, Machleder HI. Developmental anomalies at the
thoracic outlet: an analysis of 200 consecutive cases. / Vase Surg
1992; 16:534.

7. Milliez PY. Contribution a V Etude de VOntogenese des Muscles
Scalenes (Reconstruction d'un Embryon de 2.5 cm). Universite Paris
TPantheon-Sorbonne Musee de l'Homme, Museum d'Histoire
Naturelle, Paris, June 28, 1991.

8. While JC, Poppel MH, Adams R. Congenital malformations
of the first thoracic rib: a cause of brachial neuralgia which
simulates the cervical rib syndrome. Surg Gynecol Obstet 1945;
81:643.

9. Hamilton WJ, Boyd JD, Massman WJ. Human Embryology, 3rd edn.
Cambridge, 1952:548.

10. Hughes ESR. Venous obstruction in the upper extremity (Paget-
Schroetter 's syndrome). Int Abstr Surg 1949; 88:89.

11. Machleder HI. Vaso-occlusive disorders of the upper extremity.
Curr Probl Surg 1988; 25:1.

12. Kunkel JM, Machleder HI. Treatment of Paget-Schroetter syn-
drome by a staged multidisciplinary approach. Arch Surg 1989;
124:11538.

13. Todd TW. The relations of the thoracic operculum considered in
reference to the anatomy of cervical ribs of surgical importance.
JAnat 1911; 45:293.

14. Gruber W. Uber die Halsrippen des Menschen mit vergle-ichend-
anatomischen Bemerkunge. Mem Acad Imper Sci 1869; 7
(serl3).

15. Gilliatt RW, Le Quesne PM, Logue V, Summer AJ. Wasting of the
hand associated with a cervical rib or band. / Neurol Neurosurg
Psychiatry 1970; 33:615.

16. Adson AW, Coffee JR. Cervical rib. Ann Surg 1927; 85:839.

17. Schapera J. Autosomal dominant inheritance of cervical ribs. Clin
Genet 1987; 31:386.

18. Etter LE. Osseous abnormalities of the thoracic cage seen in
forty thousand consecutive chest photoroentgenograms. Am }
Roentgenol 1944; 51:359.

19. Adson AW. Surgical treatment for symptoms produced by cervi-
cal ribs and the scalenus anticus muscle. Surg Gynecol Obstet 1947;
85:687.

20. Adson AW. Cervical ribs: symptoms, differential diagnosis, and
indication for section of the insertion of scalenous anticus muscle.
J Int Coll Surg 1951; 16:546.

21. Firsov GI. Cervical ribs and their distinction from under-devel-
oped first ribs. Arkh Anat Gistol Embriol 1974; 67:101.

22. Scher LA, Veith FJ, Haimovici H et ah Staging of arterial complica-
tions of cervical rib: guidelines for surgical management. Surgery
1984; 95:644.

23. Halsted WS. An experimental study of circumscribed dilation of
an artery immediately distal to a partially occluding band, and its
bearing on the dilation of the subclavian artery observed in certain
cases of cervical rib./ Exp Med 1916; 24:271.

24. Charlesworth D, Brown SCW. Results of excision of a cervical
rib in patients with the thoracic outlet syndrome. Br } Surg 1988;
75:431.

25. Telford EC. Stopford JSB. The vascular complications of cervical
rib. BrfSurg 1931; 18:557.

26. Ochsner A, Gage M, DeBakey M. Scalenus anticus (Naff ziger) syn-
drome. Am J Surg 1935; 28:669.

159

pa rt 1 Vascular pathology and physiology

27. Sanders RJ, Roos DB. The surgical anatomy of the scalene triangle.
Contemp Surg 1989; 35:11.

28. Naffziger HC, Grant WT. Neuritis of the brachial plexus mechani-
cal in origin: the scalenus syndrome. Surg Gynecol Obstet 1938;
67:722.

29. Telford ED, Mottershead S. Pressure at the cervico-brachial junc-
tion (an operative and anatomical study). / Bone Joint Surg 1948;
30B:249.

30. Sunderland S, Bedbrook GM. Narrowing of the second part of the
subclavian artery. Anat Rec 1949; 104:299.

31. Todd TW. Cervical rib: factors controlling its presence and its size,
its bearing on the morphology of the shoulder, with four cases.
J Anat 1911-1912; 45:293.

32. Roos DB. Pathophysiology of congenital anomalies in thoracic
outlet syndrome. Acta Chir Belg 1980; 79:353.

33. Lang J. Topographische Anatomie des Plexus brachialis und Thoracic-
Outlet Syndrom. Berlin: Walter de Gruyter, 1985.

34. Sampson JJ. Medico-Surgical Tribute to Harold Brunn. Berkeley:
University of California Press, 1942:453.

35. Aziz S, Straehley CJ, Whelan TJ. Effort-related axillo-
subclavian vein thrombosis: a new theory of pathogenesis
and a plea for direct surgical intervention. Am } Surg 1986;
152:57.

36. Machleder HI. Evaluation of a new treatment strategy for Paget-
Schroetter's syndrome: spontaneous thrombosis of the axillary-
subclavian vein. / Vase Surg 1993; 17:305.

37. Todd TW. The descent of the shoulder after birth. Anat Anz 1912;
41:385.

38. Machleder HI. Role du muscle scalene anterieur dans les syn-
dromes de la traversee thoraco-brachiale. In: Kieffer E, ed. Les
Syndromes de la Traversee Thoraco-Brachiale. Paris: Editions AERCV,
1989:69.

39. Eisenberg BR, Dix DJ, Kennedy JM. Physiological factors influ-
encing the growth of skeletal muscle. In: Ciba Foundation Sympo-
sium 138. Plasticity of the Neuromuscular System. Chichester: John
Wiley & Sons, 1988:3.

40. Pette D, Staron RS. Molecular basis of the phenotypic characteris-
tics of mammalian muscle fibers. In: Ciba Foundation Symposium
138. Plasticity of the Neuromuscular System. Chichester: John Wiley
& Sons, 1988:22.

41. Dubowitz V. Responses of diseased muscle to electrical and
mechanical intervention. In: Ciba Foundation Symposium 138.
Plasticity of the Neuromuscular System. Chichester, UK: John Wiley
& Sons, 1988:240.

42. Nag AC, Lee ML, Shepard D. Effect of amiodarone on the expres-
sion of myosin isoforms and cellular growth of cardiac muscle
cells in culture. Circ Res 1990; 67:51.

43. Bugaisky LB, Anderson PG, Hall RS, Bishop SP Differences in
myosin isoform expression in the subepicardial and subendocar-
dial myocardium during cardiac hypertrophy in the rat. Circ Res.
1990; 66:11272.

44. Caforio ALP, Rossi B, Risaliti R et ah Type I fiber abnormalities
in skeletal muscle of patients with hypertrophic and dilated car-
diomyopathy: evidence of subclinical myogenic myopathy. / Am
Coll Cardiol 1989; 14:14643.

45. Smirne S, Iannaccone S, Ferini SL, Comola M, Colombo E,
Nemni R. Muscle fiber type and habitual snoring. Lancet 1991;
337:597.

46. Riley DA, Allin EF. The effects of inactivity, programmed stimula-
tion and denervation on the histochemistry of skeletal muscle
fibers. Exp Neurol 1973; 40:391.

47. Johnson MA, Polgar J, Weightman D, Appleton D. Data on distrib-
ution of fiber types in thirty-six human muscles: an autopsy study.
J Neurol Sci 1973; 18:111.

48. Sarnat HB. Muscle Pathology and Histochemistry. Chicago:
American Society of Clinical Pathologists Press, 1983.

49. Salmons S, Sreter FA. Significance of impulse activity in
the transformation of skeletal muscle type. Nature 1976; 263:
30.

50. Salmon S, Hendriksson J. The adaptive response of skeletal mus-
cle to increased use. Muscle Nerve 1981; 4:94.

51. Sreter FA, Gergely J, Salmons S et al. Synthesis by fast muscle of
myosin light chains characteristic of slow muscle in response to
long-term stimulation. Nature N Biol 1973; 241:17.

52. Salmons S, Gale DR, Sreter FA. Ultrastructural aspect of the trans-
formation of muscle fiber type by long-term stimulation: changes
in Z discs and mitochondria. ] Anat 1978; 127:17.

53. Brooke MH, Engel WK. The histographic analysis of human
muscle biopsies with regard to fiber types. I. Adult male and
female. Neurology 1969; 19:221 .

54. Ianuzzo DC, Gollnick PD, Armstrong RB. Compensatory adapta-
tions of skeletal muscle fiber types to a long-term functional over-
load. Life Sci 1976; 19:15174.

55. Morris CJ, Salmons S. The innervation pattern of fast muscle
fibers subjected to long-term stimulation. / Anat 1975; 120:
142.

56. Dubowitz V, Brooke MH. Muscle Biopsy: A Modern Approach.
Philadelphia: WB Saunders, 1973.

57. Brooke MH, Kaiser KK. Muscle fiber types: how many and what
kind? Arch Neurol 1970; 23:369.

58. Munsat T, McNeal D, Waters R. Effects of nerve stimulation on
human muscle. Arch Neurol 1976; 33:608.

59. Salmons S, Vrbova G. The influence of activity on some contractile
characteristics of mammalian fast and slow muscles. / Physiol
1969; 201:535.

60. Keens TG, Bryan AC, Levison HH et al. Developmental pattern of
muscle fiber types in human ventilatory muscles. / Appl Physiol
1978; 44:909.

61 . Keens TG, Chan V, Patel Petal. Cellular adaptations of the ventila-
tory muscles to a chronic increased respiratory load. J Appl Physiol
1978; 44:905.

62. Nygaard JE. Adaptational changes in human skeletal muscle with
different levels of physical activity. Acta Physiol Scand Suppl 1976;
440:176.

63. Jansson E, Sjodin B, Tesch P. Changes in muscle fiber type dis-
tribution in man after physical training. Acta Physiol Scand 1978;
104:235.

64. Kiessling K, Pilstrom L, Bylund Aetal. Enzyme activities and mor-
phometry in skeletal muscle of middle aged men after training.
Scand J Clin Lab Invest 1974; 33:63.

65. Chiappa KH, ed. Evoked Potentials in Clinical Medicine. New York:
Raven Press, 1983.

66. Siivola J, Sulg I, Pokela R. Somatosensory evoked responses as
diagnostic aid in thoracic outlet syndrome. Acta Chir Scand 1982;
148:647.

67. Yiannikas C, Walsh JC. Somatosensory evoked responses in the

160

chapter 14 Thoracic outlet syndrome

diagnosis of thoracic outlet syndrome. J Neurol Neurosurg Psychia- potential in the assessment of thoracic outlet compression syn-

try 1983; 46:234. drome. / Vase Surg 1987; 6:177.

68. Yiannikas C. Plexopathies and radiculopathies. In: Chiappa KH, 70. Glover JL, Worth RM, Bendick PJ, Hall PV, Markand OM. Evoked
ed. Evoked Potentials in Clinical Medicine. New York: Raven Press, responses in the diagnosis of thoracic outlet syndrome. Surgery
1983:278. 1981; 89:86.

69. Machleder HI, Moll F, Nuwer M, Jordan S. Somatosensory evoked

161

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Aneurysmal disease

Juan Carlos Jimenez
Samuel Eric Wilson

An aneurysm is a localized or diffuse arterial dilation usually
considered to be twice the normal arterial diameter. True
aneurysms have a three-layer wall with thinning of the adven-
titia, media, and intima. False aneurysms possess only adven-
titia with a surrounding capsule of compressed fibrous tissue
and thrombus. Aneurysms may be saccular or fusiform, and
differ in etiology. Saccular aneurysms arise from a distinct,
limited portion of the arterial wall, both longitudinally and
circumferentially. The more common fusiform aneurysms
involve the total wall circumference and are often diffuse. A
small number of aneurysms are present at birth; however, the
majority are acquired. Susceptibility to aneurysm develop-
ment, however, may be genetically predetermined. Approx-
imately 1 .7% of deaths in men aged 65-74 are due to rupture of
abdominal aortic aneurysms (AAAs). 1 Most are preventable
with elective surgery. Venous autograft was first used in repair
of a popliteal aneurysm as early as 1906. Successful excision
and replacement of an AAA with arterial homograft was per-
formed by Dubost and coworkers in 1952, 2/3 and soon there-
after, textile prosthetics were introduced. Small aneurysms
less than 5.5 cm in diameter have a low rupture rate, but
enlargement should be anticipated. Larger aneurysms are
surgically curable but have potentially grave consequences
if left untreated.

Epidemiology

The prevalence of aortic aneurysm has been studied in diverse
populations. In an ultrasonographic screening program in
Great Britain, Scott and coworkers found AAAs 3 cm or more
in diameter were visualized in 4.3% of 65- to 80-year-old men
and women. 4 Autopsy records of more than 7000 patients from
the University of Kansas Medical Center over a 34-year period
of study revealed unsuspected AAAs in 1.9% of men and 0.9%
of women. 5 Taylor and Porter 6 summarized 10 studies and
found the highest rates of AAA among patients with popliteal
or femoral aneurysms (53%). Unselected adults screened by
ultrasound had a 3.2% incidence of AAAs. Patients with coro-

nary artery disease had a 5% incidence by ultrasound, while
9.6% of those with peripheral vascular disease had AAAs
(Table 15.1). The age-standardized mortality rate for AAA in
Western Australia as determined by health department mor-
tality data increased 36% for men and 27% for women from
1980 to 1988. 7 This increase appears to be caused by more
patients presenting with rupture, even though the number of
elective operations increased (Fig. 15.1). This suggests a true
increase in prevalence of AAAs. Fowkes and coworkers 8 iden-
tified similar trends in English and Welsh patient populations.

Ninety percent of aortic aneurysms occur in the abdominal
aorta. Over 90% of these are below the renal arteries. While
unique characteristics of each location exist, risk factors are
similar. The risk factors for aneurysms are similar to those for
atherosclerosis including cigarette smoking, hypertension,
and age. 9 These factors are thought to determine disease
progression, stabilization, regression, and possibly genesis.
Cigarette smoking is believed to increase blood collagenolytic
and elastolytic activity. 10 ' 11

Cohen and coworkers postulate that factors in cigarette
smoke may block the active site of a 1 -antitrypsin, a substance
thought to enhance the stability of the aortic matrix. 12
MacSweeney and colleagues 13 followed 43 patients with
small aneurysms over 3 years with serial ultrasound and
found that growth rates were higher in patients who contin-
ued to smoke vs. nonsmokers (0.16 cm/year vs. 0.09 cm/year,
respectively). Higher growth rates were also significantly
correlated with concentration of serum cotinine. Carotid,
coronary, and peripheral arterial disease are also indicators
of aneurysm disease.

Knowledge of the common locations of aneurysms is useful
when evaluating individuals and populations for disease pat-
terns (Fig. 15.2). The primary site is the abdominal aorta, and
fortunately for both patient and surgeon, less than 5% of ab-
dominal aneurysms extend above the renal arteries. 6 Found
almost exclusively in men, 70% of peripheral aneurysms occur
in the popliteal arteries. Two-thirds of these are bilateral: ap-
proximately 50% have a simultaneous AAA. 6 Femoral artery
aneurysms are noted for their presence in older, hypertensive

162

chapter 15 Aneurysmal disease

Table 15.1 Incidence of abdominal aortic aneurysms

Category

Incidence (%)

Autopsy series 1.5

Unselected adults screened by ultrasound 3.2

Patients with coronary artery disease (by ultrasound) 5.0

Patients with peripheral vascular disease (by ultrasound) 9.6

Patients with popliteal or femoral aneurysms 53.0

From Taylor LM, Porter JM. Basic data related to clinical decision making in
abdominal aortic aneurysms. Ann Vase Surg 1986; 1 :502.

men. A significant proportion of patients with femoral
aneurysms have other sites of aneurysmal change; 51-85%
have A A As while 17-44% have popliteal aneurysms. 14-16
Aneurysms are seen in the common, external, and internal
iliac arteries. Coronary artery aneurysms are rare, and
typically atherosclerotic in origin. 17 One should always search
for bilateral and aortic dilation when a peripheral aneurysm
is detected. 18

Pathogenesis

c
g

500 r-

400 -

300-

200-

100 -

■ < ■

1 980*82

1986-88

1 983-85

Triennia

Figure 1 5.1 Total operations for abdominal aortic aneurysm in Western
Australia for patients aged 55 years and over. The number of elective and
emergent operations for AAA has increased from 1980 to 1988, suggesting
an increase in prevalence. Solid = emergency; empty = elective. (From
Norman PE, Castleden WM, Hockey RL. Prevalance of abdominal aortic
aneurysm in Western Australia. Br J Surg 1 991 ; 78: 1 1 1 8.)

. J c

Figure 15.2 Common locations of arterial aneurysms. Ninety percent of
aneurysms arise from the infrarenal abdominal aorta. (From Sabiston DC Jr.
Aneurysms. In: Sabiston DC, ed. Textbook of Surgery, 14th edn. Philadelphia:
WB Saunders, 1991:1540.)

Anatomy and structure

Study of the arterial anatomy of aneurysms reveals mechani-
cal weaknesses in areas prone to dilation. These weaknesses
may increase vessel susceptibility to ischemia and consequent
dilation. The key structural element of the arterial wall is the
media, which consists of smooth muscle cells with elastic
layers in a collagen network. While the elastin gives the wall
distensibility on pulse propagation, the collagen contributes
tensile strength and prevents overdistention. The aortic wall
possesses at least three types of collagen. 19 Type IV collagen is
the basal lamina layer, arranged around each smooth muscle
cell group. The meshwork of fine fibrils interspersed with the
basal lamina consists of type I collagen. The thick type III colla-
gen fibers are the strongest and make up the structural portion
of the aortic wall. Chronic medial injury to the canine aorta
may be experimentally induced with acetrizoate. The result-
ing, progressive destruction in the middle region of the media
exhibits no evidence of repair, and aneurysms develop. 20

Twenty-nine lamellae of the media can be nourished by dif-
fusion from the vessel lumen. 21 Each lamellar unit consists of
an elastic lamella, its adjacent smooth muscle cells, and colla-
gen fibers arranged in sheets. 22 In most mammals, when there
are more than 29 lamellar layers in the media, luminal diffu-
sion is inadequate to supply the full thickness of the aortic
wall, and adventitial vasa vasorum supply the outer media.
The number of thoracic and abdominal aortic lamellar units
varies linearly with the diameter with little reserve to protect
against increased pressure and aneurysm formation.

The human abdominal aorta has a unique structure. Com-
pared with the thoracic aorta, the abdominal aorta has fewer
lamellar units relative to its diameter and wall thickness. 23 The
human abdominal aortic wall is approximately 0.7 mm thick.
If the human aorta had the same design as other mammals, it
would have 40 elastic lamellae; however, it has only 30 lamel-
lar units. 22 Each layer of the abdominal aorta is thicker, relative
to vessel diameter, compared with other arteries, which in-
creases the tension per lamellar unit. 24 When the critical
loading level of the elastic lamellae is exceeded, wall failure
and aneurysm formation may result.

The human abdominal aortic media is devoid of vasa vaso-
rum; the aorta possesses fewer adventitial vasa vasorum than

163

pa rt I Vascular pathology and physiology

other mammals. 21 ' 25 Most vessels with wall thickness more
than 0.5 mm possess vasa vasorum. Without this source of
blood supply in the abdominal aorta, luminal diffusion is the
sole source of nourishment and oxygenation to the vessel wall.
The vulnerability of medial smooth muscle to relative is-
chemia may lead to atrophy and contribute to aneurysmal
changes. Relative ischemia may occur secondary to demand
for oxygen from the increased load per lamellae ratio or from
decreased oxygen supply. Decreased oxygen supply can occur
when luminal diffusion is obstructed by an intimal atheroscle-
rotic plaque. The lack of medial vasa vasorum also predis-
poses to ischemia.

This structural and nutrient hypothesis of aneurysm
formation was developed and supported further by Palma. 26
He created canine aortic wall ischemic lesions in the adventitia
and outer media by placing costal cartilage fragments near
each lumbar artery. Over 8-10 months, the lumbar arteries
were obliterated and ischemia and aortic dilation occurred.

If circulation to the outer wall fails when atherosclerotic dis-
ease compromises luminal diffusion, outer medial injury may
result. Rhesus monkeys with atherosclerotic disease have
aortic vasa vasoral blood flow that is 17- to 30-fold higher
than controls. 27 This may be an important difference between
humans and monkeys. Deliberate occlusion of the thoracic
vasa vasorum in experimental animal models produces
medial necrosis but not acute aneurysm formation. 25,28 ' 29
Medial injury may, however, predispose to late vessel dilation.

AAA MODEL (5 slice)

Finite Element Discretization
i

Single Brick Element

Figure 15.3 Finite element discretization foran aneurysm. Simple
continuous segments are drawn from complex geometric structures to
determine the architectural characteristics with known equations. (From
StringfellowMM, Lawrence PF, StringfellowRG.The influence of aortic
aneurysm geometry upon stress in the aneurysm wall. J Surg Res 1 987;
42:425, with permission from Elsevier.)

Hemodynamic factors

Hemodynamic studies have shown that if a wall weakness
predisposes to a slight increase in vessel diameter, aneurysm
formation, enlargement, and rupture can follow. The law of
Laplace states that wall tension equals the product of the pres-
sure on the vessel and the vessel radius (T=Pxr). This explains
why as aneurysms increase in size, the rate of expansion
increases as well. 30

With increased radius, the tension will increase despite
constant or even normal pressures.

Aneurysm growth can be viewed as a passive yield to blood
pressure with reactive thickening of the vessel wall. Since the
abdominal aorta is one of the largest central arteries, it is
subject to high pulse pressure, producing high oscillating and
distending forces on the arterial wall.

Small aneurysms also rupture. Thus, mechanics that more
fully explain the phenomenon of aneurysm formation and
development must be found. Fine element analysis models
the characteristics of continuous structures by examining
finite elements of complex geometric structures and arranging
them into simple segments (Fig. 15.3). The behavior of these
continuous segments is evaluated mathematically and used to
determine characteristics of the more complex whole. Thus,
simple equations can be used to evaluate complex forces. Fine
element analysis may reveal mechanical parameters that

enable the clinician to make more informed decisions about
the clinical significance of aneurysms.

Stringfellow and coworkers 7 fine element analysis to deter-
mine the wall stress distribution in models of infrarenal
AAAs. 31 Their work shows that aorta to aneurysm geometry
can determine aneurysm wall stress. The aortic size affects the
wall stress via the ratio of the aortic diameter to aneurysm di-
ameter; the larger the aneurysm relative to the aorta from
which it arises, the greater the aneurysm wall stress. Arterial
wall thinning may also increase wall stress, allowing an
aneurysm to form and enlarge. Dilation and thinning of an ar-
terial wall increases stresses in all directions (Fig. 15.4).

Stringfellow and colleagues note that an analysis of the
stresses in cylindrical and spherical systems considers three
forces: radial, longitudinal, and circumferential (Fig. 15.5). 31
Studies indicate that given equal wall thickness and diameter,
cylindrical aneurysms are more likely to rupture. In all cases,
rupture was most common at the point of maximum
aneurysm diameter. Therefore, the ability of an aneurysm wall
to withstand stress in the longitudinal and circumferential
direction is an important factor in determining aneurysm
rupture. 31 The pulse pressure in the abdominal aorta is
higher than in the thoracic aorta. Also, the distal aorta tapers
and stiffens. The cross-sectional area ratio compares the cross-
sectional area of an arterial bed distal to a bifurcation with that
of the proximal vessel. The ratio of the cross-sectional area of

164

chapter 15 Aneurysmal disease

Figure 15.4 Dilation and thinning of the
arterial wall increase stresses in all directions. The
larger the aneurysm relative to the aorta from
which it arises, the greater the aneurysm wall
stress. o c = circumferential stress; o L =
longitudinal stress. (From Stringfellow MM,
Lawrence PF, Stringfellow RG. The influence of
aorta aneurysm geometry upon stress in the
aneurysm wall. J Surg Res 1 987; 42:425, with
permission from Elsevier.)

Minimum Wall Thickness
<r c (xlO 5 dynes/cm 2 )
<7 L (xlO 5 dynes/cm 2 )

0.1

22.4
16.0

0.33
60.6
57.7

the two common iliac arteries compared with the cross-
sectional area of the proximal aorta decreases from approxi-
mately 1.1 at infancy to 0.75 by age 50 (Fig. 15.6). 32 Occlusive
atherosclerotic disease in the iliac arteries can make the ratio
even smaller. As the cross-sectional area gradient increases,
the pressure wave reflection back to the proximal vessel also
increases. This reflection adds to systolic blood pressure and
increases the pulse pressure. The decrease in cross-sectional
area between the two common iliacs and the abdominal aorta
results in a significant increase in pulse pressure in the abdom-
inal aorta. The increased stress on the abdominal aorta further
increases the tensile strength required to prevent dilation.
Oscillating and deforming forces stimulate smooth muscle
cells to secrete collagen and elastin. 33 Aneurysm walls are stiff;
therefore, there are less oscillation and deformation forces
acting on the smooth muscle cells. Thus, there is a decrease in
the stimulus for tissue repair processes. As the aneurysm
dilates, wall thinning decreases the content of smooth muscle
cells. These cells are replaced by fibrous connective tissue,
which lacks the capacity for repair. The role of these events in
aneurysm pathogenesis is not yet determined.

The effects of (3-adrenergic blockade and slowing of
aneurysm growth rate has been measured in several recent
studies. Leach and colleagues 34 studied 136 patients retrospec-
tively and found that the growth rate for control subjects not
treated with (3-blockers was 0.44 cm/year compared with
0.17 cm/year for those treated with (3-blockade. Slaiby et al. 35
tested similar effects on rats with chemically induced AAA
using direct injection of elastase into the infrarenal aorta. Both

a r radia!

b - longitudinal

c = circumferential

Figure 1 5.5 Components of stress in cylindrical or spherical structures.
(From Stringfellow MM, Lawrence PF, Stringfellow RG. The influence of aorta
aneurysm geometry upon stress in the aneurysm wall. J Surg Res 1 987;
42:425, with permission from Elsevier.)

normotensive and genetically hypertensive rats were used.
Propranolol was found to reduce significantly the size of ex-
perimentally induced AAA in these animals. This theory was
further extended in relation to large aneurysms in a human
study by Gadowski etal. 36 They found that in 138 patients with
large (>5 cm) aneurysms monitored with serial aortic ultra-
sound, those receiving (3-blockers had a significantly reduced
mean expansion rate compared with untreated patients. Large
aneurysms expanded more rapidly than small aneurysms in
patients not treated with (3-blockers.

Conflicting results, however, were found in a multicenter
randomized controlled trial studying the effects of pro-
pranolol on patients with asymptomatic small aneurysms

1.2r

1.1-

O 1.0-

<
LU

cr
<

0*9-

0.8-

0.7-

0.6

(19)

(15)

(18) T

(20) _

(19)

0-10 11-20 21-30 31-40 41-50

AGE (years)

Figure 1 5.6 The variation of the cross-sectional area ratio of the aortic
junction in man with age. As the cross-sectional area ratio decreases with
age, the pressure wave reflection and pulse pressure increase. This increases
the stress in the abdominal aorta. (From Gosling RG, Newman DC, Bosden LR
etal. The area ratio of normal aortic junctions. Br J Radiol 1 971 ; 44:850.)

165

pa rt I Vascular pathology and physiology

(between 3.0 and 5.0 cm). 37 Over 500 patients receiving either
placebo or propranolol were followed for a mean of 2.5 years.
Growth rates were similar between the two groups and there
was no difference in death rates. Incidentally, poorer quality
of life scores were noted in patients receiving propranolol
compared with placebo, indicating poor overall tolerance to
the drug in patients with AAA. 37

Atherosclerosis

The traditional view of aneurysm formation is that arterial
dilation is a consequence of degenerative atherosclerotic
disease, which results in acquired wall weakness. The ex-
perienced vascular surgeon is well aware that peripheral arte-
riosclerosis and aneurysmal disease often coexist. Severe
atherosclerotic calcification in the aortoiliac vessels presents
a technical challenge in aneurysm surgery. Epidemiologic,
radiographic, and histologic data support the association
between aneurysm disease and atherosclerosis.

AAAs and atherosclerosis share many risk factors and
frequently occur simultaneously. The frequency of aortic
aneurysms closely parallels the prevalence of atherosclerosis;
for example, the low abdominal aneurysm rate in Asia corre-
lates with the decreased incidence of atherosclerosis.

Radiographic and histopathologic studies support the link
between atherosclerosis and aneurysms. Ultrasound screen-
ing of patients with peripheral vascular disease detects a 5.9%
rate of AAA, double that of the general population. 38 Studies
of patients suffering from coronary and carotid artery occlu-
sive disease detect an aortic aneurysmal rate of 11-13. 5%. 39
Histologic evaluations of sections from aortic aneurysms
show atherosclerotic changes and thinning of the media.

Pathophysiologic principles also support the concept that
atherosclerosis contributes to aneurysm formation. Athero-
sclerotic plaques may obstruct nutrient diffusion from the
lumen to the media. The needs of the media must then be
supplied exclusively by vasa vasorum from the adventitia.
However, this may be inadequate due to incomplete distribu-
tion of vasa vasorum throughout the human arterial system. 25
Aortic vasa vasorum usually arise from the renal arteries,
accounting for the relative sparing of the perirenal aorta from
aneurysm formation. 40

Structural changes induced by atherosclerosis may con-
tribute to aneurysm formation. As atherosclerosis progresses
in humans, friable type I collagen replaces native type III
collagen. 41 Thus, the architectural integrity of the vessel is
impaired, leading to a predilection to aneurysm formation.

An association between aortic aneurysms and atherosclero-
sis is not surprising since the geometry and hemodynamics of
arterial dilation predispose to atherosclerosis formation. 42
Aneurysms have increased in incidence, prevalence, and mor-
tality over the last 30 years, while coronary artery and cere-
brovascular diseases have not. 8 The divergence of these
diseases in prevalence and mortality indicates that while risk

factors are shared, the development of aneurysm disease is not
entirely explained by atherosclerosis. 9

Although the epidemiologic link between the two is strong,
Tilson and Stansel 43 propose that occlusive atherosclerotic
aortic disease and aortic aneurysmal disease are distinct enti-
ties. This is based on the different characteristics of these
groups including age of onset, male-female ratio, clinical
course, and prognosis. Evidence found to correlate with the
size and state of aneurysm indicates that aneurysms reflect a
heterogeneous disease with multiple forms and etiologic
factors. 9

Histolytic enzymes

New evidence indicates that biochemical events in the arterial
wall are pivotal factors in aneurysm development, growth,
and rupture. Following laparotomy for nonvascular diseases,
there is an increase in acute rupture of preexisting AAAs. 44
Busuttil and Cardenas 45 ' 46 attribute this increase to postopera-
tive collagen breakdown. In 1970, Sumner and colleagues 47
noted a decrease in collagen and elastin in aneurysmal vessel
walls. Tilson 48 reported histochemical studies of aneurysmal
aortas that showed a specific deficiency of elastin compared
with atherosclerotic controls.

Proteolytic degradation of the vessel wall is postulated as a
factor in the pathogenesis of aneurysm formation. Increased
enzymatic activity may be attributed to a direct increase in en-
zymes or a failure of normal antiproteolytic processes. Busuttil
and colleagues demonstrated an increase in collagenase
and elastase enzymatic activity in the walls of human aortic
aneurysms, compared with those of nonaneurysmal athero-
sclerotic aortas. 49 Increases in proteolytic activity have been
found to correlate with the size and state of aneurysm
development, the highest levels of enzymatic activity being
demonstrated in ruptured aneurysms. Menashi and associ-
ates 50 detected collagenase in tissue from ruptured AAAs.

Zarins and coworkers 51 studied proteolytic degradation of
collagen in a cynomolgus monkey model of poststenotic dila-
tion. The collagenase activity was shown to increase after the
aneurysmal dilation occurred. This work raises the question
of cause and effect; the increased enzyme activity may be
a byproduct of aneurysm formation.

Dobrin and coworkers, 52 in a series of enzymatic degrada-
tion experiments, evaluated the characteristics of canine ves-
sel walls. Arteries from different locations were treated with
elastase, collagenase, or elastase and collagenase. Aneurysmal
dilation of arterial walls was experimentally induced by prote-
olytic degradation of elastin; however, there was no rupture.
Conversely, degradation of collagen invariably precipitated
rupture with only slight dilation. These studies indicate
elastin is responsible for maintaining normal vessel dimen-
sions and providing wall compliance; collagen provides
tensile strength and stability against rupture (Fig. 15.7).

When carotid arteries were treated with collagenase alone,

166

chapter 15 Aneurysmal disease

»0r

Figure 1 5.7 Pressure-diameter data for 56
canine common carotid arteries under relaxed
pretreatment conditions and after treatment
with proteolytic enzymes. All vessels dilated but
did not rupture after treatment with elastase.
All vessels ruptured following collagenase
treatment. Collagen provides tensile strength.
Elastin provides wall compliance. (From Dobrin
PD. Pathophysiology and pathogenesis of aortic
aneurysms: current concepts. Surg Clin North
Am 1 989; 69:687, with permission from
Elsevier.)

eo-

S 60

q
in

9 so

30-
5?

ELASTASE

CONTROL

150

COLLAGINASE

COLLAGENASE

CONTROL

"sfc ' i<Jo" ' ISO

PRESSURE mm Hg

investigators noted a slight increase in compliance and diame-
ter; however, all walls ruptured. Studies with human external
iliac arteries yielded similar results. Treatment of the internal
iliac artery revealed a dramatic dilation and rupture after
exposure to collagenase and elastase. Collagen and elastin
breakdown in aneurysm formation probably act synergistical-
ly in vivo, although the exact interrelations are not clear.

Matrix metalloproteinases (MMP) are a group of enzymes
produced by several cell types including inflammatory cells,
fibroblasts, and smooth muscle cells and have been found to
degrade components of the extracellular matrix such as elastin
and collagen. 53 Several authors have linked increased levels of
circulating and tissue MMPs to pathogenesis of aneurysmal
disease. 53-57 McMillan and colleagues 55 found that MMP-9
mRNA expression was significantly higher in moderate-
diameter AAA (5.0-6.9 cm) compared with smaller aneurysms
(<4.0 cm). Further studies have localized increased levels of
MMPs to macrophages within the aneurysmal aortic wall. 54,56

Doxycycline has been shown to inhibit elastin degradation
and reduce MMP activity within the porcine aneurysmal ab-
dominal aorta. 58 Thus, the use of doxycycline has been studied
as a potential inhibitor of aneurysmal growth. Several
randomized, double-blind, placebo-controlled studies have
shown a reduced expansion rate of AAA with administration
of doxycyline, tetracycline, and roxithromycin. 59-61 Further
large studies are ongoing.

Inflammatory aortic aneurysms

Inflammatory aortic aneurysms are characterized by excess
fibrotic thickening of the aortic wall and perianeurysmal
adhesion to adjacent structures. There is commonly an infil-
trate made up of lymphocytes and plasma cells in the vessel

wall. About 4.5% of AAAs are inflammatory; however, this
specific diagnosis is rarely made preoperatively. 7 ' 62-64 Instead,
the thickened aneurysmal wall is readily evident on computed
tomography (CT). Histologic analysis of 60 cases of infrarenal
AAA revealed no fundamental differences warranting
formal diagnostic differentiation between inflammatory and
atherosclerotic aneurysms. 65 Despite a different originating
stimulus, inflammatory aneurysms may share a final common
pathway of formation with atherosclerotic aneurysms (i.e.
the inflammatory process may occlude the nutrient vessels).
Some authors postulate inflammatory AAAs are atheroscle-
rotic in origin with an exaggerated chronic inflammatory

process

Genetics

Epidemiologic review indicates an aneurysm gene expression
that is typically delayed until at least the sixth decade. There
is strong evidence for inherited predisposition, and possibly
an association with generalized arteriomegaly 67 ' 68 Johansen
and Koepsell 69 demonstrated an incidence of 20% aortic
aneurysms among first order relatives of aneurysm patients.
Tilson and Seashore 67 ' 70 showed genetic linkages, accounting
for abdominal aneurysm formation in 50 families, who had
clustering of the lesion in two or more first order relatives.
Possibly, they possessed a common metabolic disorder
affecting the arterial wall.

A retrospective study of hospital patients in Zimbabwe
demonstrated a higher incidence of aneurysms among whites
than Africans. 71 Webster et al.'s 72 ultrasound screening of first
degree relatives demonstrated aortic aneurysms in 20-30% of
male siblings over 55 years of age (Table 15.2). Case reports
of familial aneurysm disease in patients without connective

167

pa rt I Vascular pathology and physiology

tissue or vascular diseases add validity to the theory of genetic
linkage. 73

The occurrence of multiple aneurysms in individuals is con-
sistent with a genetic foundation. Many authors suggest
aneurysm disease is a systemic process. Frequently, patients
suffer from generalized arteriomegaly; often this is accom-
panied by multiple aneurysms. 74,75

Several cross-linking defects have been associated with
aneurysm formation. Tilson 70 studied the biochemistry
of a collagen component deficiency that predisposes to
aneurysms. They evaluated pyridine cross-linkages and
found fewer cross-linkages per collagen molecule in human
skin samples. This suggests a genetic basis for aneurysm dis-
ease. Experiments with sex-linked defects of collagen and
elastin demonstrate the blotchy BLO allele. 76 These models

Table 1 5.2 Prevalence of aneurysms in men as reported in screening
studies

Year

Country

Age (years)

Prevalence (%)

Selection basis

1984

>50

10.7

Hypertension

1985

Sweden

50-69

0.9

Hypertension

1986

USA

Mean 67

5.0

Cardiology
appointment

1987

65-79

2.8

Unselected

1988

39-90

14.0

PVD

1988

34-86

3.0

Bronchogenic CAD

1988

USA

60-75

9.0

Hypertension or CAD

1989

Mean 68

8.5

PVD

1989

Sweden

39-82

29.0

Siblings

1989

Sweden

34-74

16.4

Claudication

1990

65-74

6.3

Unselected

1990

USA

>55

25.0

Siblings

USA, United States; UK, United Kingdom; CAD, coronary artery disease; PVD,
peripheral vascular disease.

From Webster MW, Ferrell RE, St Jean PL, Majumder PP, Fogel SR, Steed DL.
Ultrasound screening of first-degree relatives of patients with abdominal
aortic aneurysm. J Vase Surg 1991; 12:9.

exhibit aortic aneurysms and diminished skin tensile strength.
The pattern of expression indicates the trait is related to the X
chromosome.

Kuivaniemi and coworkers 77 reviewed the literature and
found clear evidence for an independent genetic defect in
most A A As. Their work centered on a genetic analysis of
collagen genes. Genetic collagen defects causing architectural
defects are established in osteogenesis imperfecta (type I colla-
gen of bone) and chondrodysplasias (type II collagen of carti-
lage). New evidence implicates mutations in the type III
procollagen gene in the pathogenesis of aneurysmal disease.
Various mutations have been confirmed in studies of patients
with type IV Ehlers-Danlos syndrome (EDS). 78,79

Studies of patients with aneurysms clearly demonstrate
family linkage, and the data strongly suggest a genetic defect.
Statistical analysis supports a recessive inheritance pattern in
approximately 10% of men who have aneurysms. 80 Research
in this area is active and implicates an autosomal diallelic
major locus.

Thus, several factors are involved in AAA formation and en-
largement. These include environmental risk factors, genetics,
and directional forces. Multiple etiologies can involve: athero-
sclerosis, collagen, and proteolytic disorders, mechanical and
hemodynamic processes, and anatomic predisposition.

Natural history

Most aneurysms are asymptomatic but are readily detected by
noninvasive testing. Aneurysms undergo progressive wall
weakening and dilation with rupture unless the patient dies
first of intercurrent disease. Hemodynamic principles deter-
mine the growth of aneurysms. To maintain a stable diameter,
there must be a balance between distending forces and
retractile circumferential forces. Etiologic factors previously
discussed also contribute to growth.

Laplace's law (T = P x r) explains the propensity for
aneurysms in larger vessels to enlarge and rupture (Fig. 15.8).
This principle states circumferential tension, a retractive force,

LA PLACE EQUATION

Area of cross-sectional plane:

A = 2 r L

Distending force acting outward

F = P*A=P-2rL
Restraining force acting inward:

F» = T • 2 L
At equilibrium:

Fo= Fi

2 Pr L = 2 T L

Pr = T

Figure 1 5.8 Derivation of Laplace's law. (From
Webster MW, Ramadan F. Vascular physiology.
In: Simmons RL, Stead DL, eds. Basic Science
Review for Surgeons. Philadelphia: WB
Saunders, 1992:214, with permission from
Elsevier.)

168

chapter 15 Aneurysmal disease

is equal to the product of the transmural pressure, a distending
force, and the radius. This assumes an infinitely thin vessel
wall. The concept relates the increased tension of an aneurysm
to the increased diameter and predicts a positive feedback
loop once there is an initial diameter increase. High tension
can develop even with a normal arterial blood pressure
because circumferential distending forces increase directly
with diameter.

Certain physical forces prevent rapid increases in aneurysm
diameter. Forces resisting an increase in wall diameter
include wall thickness, tensile strength (provided by colla-
gen), and retractive ability (secondary to elastin). The fusiform
shape of most aneurysms can be compared to that of a sphere.
The wall stress for a sphere is approximately half that for a
cylinder. 52 This is represented as T = P x (r 17) (Fig. 15.9). The
second radius of a sphere provides another retractive force
and decreases the stress 50% of the usual tension required to
maintain equilibrium. Structural factors also limit the rate of
expansion. The dilated wall recruits previously unstretched
collagen fibers. Some authors believe adventitial fibers are
also recruited. 47,52

Despite the factors working toward stabilization, the
aneurysm wall inevitably thins and ruptures. The laminated
thrombus does not help prevent this process. It transmits arte-
rial pressure, and does not diminish the distending force. This
is an important consideration in follow-up of endovascular re-
pair of aortic aneurysms where in the case of a type I endoleak,
aneurysm sac thrombus may be pressurized by contact with
the pulsatile flow channel via the leak. The friable nature of the
thrombus renders it inconsequential in altering retractive
force. Thus, the thrombus does not affect the radius when

applying the law of Laplace. A laminated thrombus can make
arteriograms deceiving; serial follow-up imaging should be
performed with CT or ultrasonography. 81

Aneurysms have a tendency to elongate and become tortu-
ous. The longitudinal force is proportional to the product of
the pressure and the square of the radius. This force is borne by
the elastic lamellae which are disrupted in aneurysms; hence,
the longitudinal expanding force is greater than its resistance
vector. As aneurysm dilation and subsequent force increase
the aneurysm length also increases. The force to lengthen is
constrained by arterial branches. Segmental elongation re-
sults, leading to buckling. As aneurysmal dilation increases,
the escalation of forces causes the vessel to become tortuous
(Fig. 15.10). 82

The biologic fate of aneurysms to increase in size to eventu-
al rupture is predestined; however, the rate of growth and time
to rupture are not. Usually aneurysm enlargement is slow but
progressive. The mean expansion rate of an infrarenal AAA is
approximately 0.4 cm per year, 48 ' 83-86 but there is wide varia-
tion. In a study using ultrasound screening, the expansion rate
was greater for aneurysms of greater transverse diameter.
Other studies cite expansion rates as high as 0.52 cm per year. 81
The growth rate is augmented in patients with compromised

F D = Pn r :

\ C

Cylinder: T=Pxr

Sphere: T=Px(r/2)

Figure 1 5.9 Wall stress required to maintain equilibrium is reduced by half
as the vessel changes from cylindrical to a more spherical aneurysm. T =
tension; P = transmural pressure; r = vessel radius. (From Dobrin PB, Baker
WH, Gley WC. Elastolytic and collagenolytic studies of arteries: Implications
for the mechanical properties of aneurysms. Arch Surg 1984; 1 19:405.)

F p + F T = F Z

Figure 15.10 Forces acting longitudinally on an artery. The elastic lamellae
of aneurysms are disrupted, allowing the longitudinal force to overcome its
resistance vector. The force to lengthen is constrained by arterial branches,
resulting in sequential elongation and buckling. At a stable length, FZ = FR.
FT = traction force; FP = pressure force; FZ = net elongating force; FR =
retractive force (from artery wall). (From Dobrin PB, Baker WH, SchwarczTH.
Mechanisms of arterial and aneurysmal tortuosity. Surgery 1 988; 1 04:568.)

169

pa rt I Vascular pathology and physiology

Table 15.3 Causes of death in small (61 cases) and large (40 cases) nonsurgical abdominal aortic aneurysm

Myocardial
infarction

Rupture

Cerebrovascular thrombosis
or hemorrhage

Carcinoma

Other

Small
Large

22(36.1%)
15(37.5%)

19(31.1%)
17(42.5%)

9
2

9
1

From Szilagyi DE, Elliott J P, Smith RF. Clinical fate of the patient with asymptomatic abdominal aortic aneurysm and unfit for surgical treatment. Arch Surg 1972;
104:600.

100

<
>

Table 1 5.4 Comparative survival experience of nonsurgical patients with
small (92) and large (46) aneurysms

SERIES H

1
14

YEARS OF OBSERVATION

Figure 15.11 Survival in two series of nonsurgically managed abdominal
aortic aneurysms. As observation continues, there is a sharp decrease in
survival. (From Szilagyi DE, Elliott JP, Smith RF. Clinical fate of the patient with
asymptomatic abdominal aortic aneurysm and unfit for surgical treatment.
Arch Surg 1972; 104:600.)

immune systems. For example, in patients with cardiac trans-
plants and subsequent immunosuppressive drug mainte-
nance, the average expansion rate of AAAs was 0.71 cm per
year. 87

During formation and enlargement, most aneurysms are
clinically silent. When the transverse diameter is less than
5 cm, abdominal aneurysms are difficult to palpate and usual-
ly discovered incidentally. Slowly enlarging aneurysms cause
little or no symptoms. Rapidly enlarging aneurysms often
cause deep abdominal or back pain. This pain is a ramification
of pressure on the somatic sensory nerves of the retroperi-
toneal soft tissues. It is a severe, constant, dull pain unrelated
to the position or activity. This pain indicates rapid aneurysm
growth and impending rupture. 40

In 1972, Szilagyi and coworkers 88 reported the results of 156
patients with asymptomatic AAAs followed over 19 years. For
various reasons, usually medical risk, 127 patients did not un-
dergo operation. Of these patients, 90 died during the follow-
up period. Approximately 28% of deaths were traceable to
aneurysm rupture. There was a sharp decrease in survival as
observation continued (Fig. 15.11). While rupture rate is in-
creased in larger aneurysms, smaller dilations are not risk free

Length of survival (yea

irs)

7-9

Small (<6 cm)
Large (>6 cm)

19
22

17
11

6
4

9
4

2
1

3
1

From Szilagyi DE, Elliott JP, Smith RF. Clinical fate of the patientwith
asymptomatic abdominal aortic aneurysm and unfit for surgical treatment.
Arch Surg 1972; 104:600.

(Tables 15.3 and 15.4). 88 This finding was confirmed by a recent
autopsy series. 89

Follow-up of patients with asymptomatic aneurysms is
essential to ensure that operation is done if the aneurysm
enlarges to a dangerous degree. Ultrasound and CT are good
methods for screening and follow-up. Angiography is not a
good method because of the increased risks of an invasive pro-
cedure. Also, it may falsely show a normal-appearing lumen.
There are no universal protocols for frequency of reevaluation
since many individual characteristics and patterns must be
considered. Yet, criteria are needed to guide operation and
appropriate follow-up. Cohn and colleagues 86 suggest that
AAAs less than 4 cm in diameter need not be restudied by
ultrasound more often than every sixth month.

Size is the most important prognostic feature to determine
the probability of aneurysm rupture (Fig. 15. 12). 40 In one
report, aneurysms less than 4 cm in transverse diameter
ruptured in less than 15% of cases over 5 years. A 6-cm AAA
is associated with a 30% rate of rupture over 5 years,
whereas aneurysms > 8 cm in diameter ruptured in 75% of

cases. Kaufman and Bettmann's review showed a 5%
rupture rate at 9 years for aneurysms 3.5-4.9 cm at first
ultrasound and 25% rupture at 8 years when aneurysms
were greater than 5 cm in diameter at initial discovery. The
latter study is one of the few major data collections from a
nonreferral population.

It is important to distinguish results of operative and non-
operative management of aneurysms. Taylor and Porter 6
chose data from multiple centers for rupture risk and results of
emergent and elective repair (Tables 15.5, 15.6 and 15. 7). They

170

chapter 15 Aneurysmal disease

lOOr

90 -

Table 1 5.7 Results of elective surgical treatment of abdominal aortic
aneurysm

■^~

R0 -

70-

60 -

;

50 -

C£

40/

t/>

ac:

30 -

70 -

10 -

4 6 8

DIAMETER (cm)

Figure 1 5.1 2 Relation between rupture risk and abdominal aortic
aneurysm diameter. The most important predictor of rupture is aneurysm
size. (FromThieleBL, BandykDF. The peripheral vascular system. In: Miller TA,
ed. Physiologic Basis of Modern Surgical Care. St Louis: CV Mosby,
1988:848.)

Table 15.5 Natural history of untreated abdominal aortic aneurysm rupture
risk

AAA diameter (cm)

Yearly rupture incidence (%)

5.0
5.7
7.0

4.1

6.6

19.0

From Taylor LM, Porter JM. Basic data related to clinical decision making in
abdominal aortic aneurysms. Ann Vase Surg 1986; 1 :502.

Table 15.6 Resultsof treatment of abdominal aortic aneurysms ruptured
and emergent

Patients with rupture, who die before reaching hospital 50%

Patients with rupture reaching hospital alive but dying before 24%

operation can be performed

Mortality of patients operated on for rupture 42%

Overall mortality of rupture 78%

Operative mortality for emergent operation for suspected rupture 1 9%

when none isfound

From Taylor LM, Porter JM. Basic data related to clinical decision making in
abdominal aortic aneurysms. Ann Vase Surg 1986; 1 :502.

found 5-year survival after successful AAA repair to be 67%,
compared with the 80% survival of age-matched controls
without AAAs. In an 11-year study involving 225 patients
undergoing repair electively, the mortality rate was 4%. 91
Conversely, 253 patients had emergency operations after non-
operative management and the mortality was 31.2%. Thus,

Group

Operative mortality

Best modern individual series 3.5%

State and community-wide surveys 10.5%

Patients over 80 years of age 10.0%

High-risk patients undergoing conventional intra- 7.0%

abdominal graft repair

High-risk patients undergoing nonresective 1 6.0%

treatment (ligation plus extraanatomic bypass)

From Taylor LM, Porter JM. Basic data related to clinical decision making in
abdominal aortic aneurysms. Ann Vase Surg 1986; 1 :502.

there is a critical distinction to make between operative and
nonoperative management.

Classic indications for surgery include onset of symptoms
and absolute size. The rupture rate of untreated AAA is
directly proportional to the size. A transverse diameter of
greater than 5.5 cm on CT scan or ultrasound confers an
increased risk of rupture. 84 ' 88 Also, relative increases in
aneurysmal diameter over time can indicate the need for oper-
ation. One must consider the patient's medical condition in
evaluating risk. White and colleagues 92 found a correlation
between the Goldman cardiac risk index and long-term sur-
vival after AAA repair. Cronenwett and coworkers 83 studied
30 medical indicators that might predict aneurysm rupture.
They found only diastolic blood pressure, initial aneurysm an-
teroposterior diameter, and degree of obstructive pulmonary
disease were independently predictive of rupture.

Survival of patients with ruptured AAA who reach the hos-
pital alive is less than 20% if there is intraperitoneal bleeding
and shock. Johansen and coworkers 93 at Harborview Medical
Center in Seattle found survival in these circumstances to be as
low as 10%. With more stable ruptures, 50% may survive if
given prompt surgical treatment. Most investigators docu-
ment a greater than 90% survival overall for electively re-
paired AAAs. Ivers and Bourke 94 found mortality for repair of
a nonruptured aneurysm to be 0.9%, while that for a ruptured
one was up to 55%.

Popliteal aneurysms are usually symptomatic when discov-
ered. Approximately one-half of these patients present with
complications; the most common is thrombosis. 95 Other
significant complications include embolization, rupture, and
venous compression with resultant edema and pain. While
there is minimal limb loss with asymptomatic popliteal
aneurysms, symptomatic presentation portends a 34% rate of
limb loss. 95

Dawson and coworkers 96 report long-term follow-up of a
popliteal aneurysm study group. The probability of complica-
tions was 74% over 5 years if untreated, compared with a 64%
graft patency rate and 95% foot salvage at 10 years if recon-
struction was done. They found the greatest risk factor for

171

pa rt I Vascular pathology and physiology

popliteal aneurysm occurrence is the presence of multiple
aneurysms at initial evaluation, indicating this population
should be followed more closely after repair. The high risks of
complications with popliteal and femoral aneurysms arise
primarily from their late clinical presentation.

In 1998, the UK Small Aneurysm trial randomly assigned
1090 patients between the ages of 60 and 76 years with
aneurysms measuring between 4.0 cm and 5.5 cm to either re-
ceive elective early open repair or undergo continued ultra-
sonographic surveillance. Patients were followed for a mean
of 4-6 years and surgical repair was performed once the size of
the aneurysm measured 5.5 cm or greater. The 30-day opera-
tive mortality in the early-surgery group was 5-8% demon-
strating a survival disadvantage for those receiving surgery
for small, asymptomatic A A As that had only a 1% risk of
rupture per year. Thus ultrasound surveillance for small
aneurysms is relatively safe and early surgery does not confer
a long-term survival advantage. 97

It is important to detect patients at high risk for rupture and
complications of aneurysm, and to perform repair. Ultrasound
screening for AAAs is highly effective, especially when ap-
plied to patients with coronary artery disease or peripheral
vascular disease. 98 A 2-year prospective analysis at Oxford
University confirms the benefit of elective surgery. 99

Technological advances in medical imaging provide a sup-
erior array of diagnostic modalities for detecting developing
aneurysms. Recently, three-dimensional reconstruction has
been used to assess various factors contributing to patho-
genesis and rupture of AAAs. 100-103 Data obtained from these
studies include aneurysmal tortuosity, maximum transverse
diameter, aneurysm volume, length, and cross-sectional area.
Hatakeyama et al. 100 found that the most efficient predictors of
aneurysmal rupture following three-dimensional reconstruc-
tion were maximum transverse diameter, diastolic blood pres-
sure, and ratio of transverse aneurysmal diameter to length of
the aneurysm. Kato et al. 101 have shown recently that accurate
life-sized aortic replicas can be reproduced using 3D CT data.
These constructs may be useful for preoperative evaluation of
complicated aneurysms. Recently, spiral-CT angiography has
been shown to be effective in evaluating patients for endovas-
cular repair of ruptured aortoiliac aneurysms. 102

Connective tissue disorders

Marfan's syndrome

Marfan's syndrome is an autosomal dominant disorder mani-
fested through abnormalities in the cardiovascular, muscu-
loskeletal, and ocular systems. It is one of the most common
genetic disorders of connective tissue, occurring in about 1 in
20 000 persons. 104 Common physical findings in these patients
include arachnodactyly, pectus excavatum or carinatum,
scoliosis, joint laxity, ectopia lentis, and a diastolic murmur.

The prominent cardiovascular defects are mitral valve pro-
lapse, aortic valve incompetence, and ascending aortic
aneurysm and dissection. Cardiovascular abnormalities
account for approximately 90% of deaths and the 32-year
mean life span found in patients with Marfan's syn-
drome. 105 ' 106 Aneurysm formation of isolated segments of the
descending thoracic and abdominal aorta has been reported
but it is rare. 107,108 For years, defects in collagen or elastin genes
were suspected as causes of the disorder but genetic analysis
failed to make a connection. 109 ' 110 Recently, immunohisto-
chemical studies of the skin and the extracellular matrix of
cultured fibroblasts from patients with Marfan's syndrome
have shown abnormal deposition of the protein fibrillin. 110 ' 111
Genetic analyses have linked Marfan's syndrome to the fib-
rillin gene on the long arm of chromosome is. 10A ' 111,112 Fibrillin
is a 350-kDa glycoprotein that either alone or in conjunction
with other proteins (especially elastin) forms the microfibrillar
network of the extracellular matrix. 110 Microfibrils are 10-12-
nm fibers that constitute a major structural component of
many tissues. They are ubiquitously distributed in the extra-
cellular matrix and are particularly abundant within the aorta,
in ligaments, at sites of epiphyseal growth, and in the zonular
fibers that maintain the lens in its normal position. 110

Though strong data support linkage between the chromo-
some 15 fibrillin gene and the syndromes, only a few specific
mutations in the fibrillin gene have been identified in patients
with Marfan's. 112 ' 113 Kainulainen and colleagues 112 screened
20 unrelated patients with Marfan's syndrome and found two
mutations, each of which coded for a shortened fibrillin
polypeptide. One mutation was a glycine to arginine substitu-
tion that created a stop codon in place of a tryptophan codon,
resulting in premature termination of the protein. 112 The other
mutation resulted in a 366-base deletion of fibrillin mRNA,
causing a shortened polypeptide. 112 This research group
screened 60 other nonrelated Marfan's patients for these muta-
tions (and a previously reported arginine to proline mutation),
and found no other cases of these genetic defects. 112

In experiments on fibroblasts from patients with Marfan's
syndrome, Milewicz and colleagues 110 found defects in fib-
rillin production and in extracellular matrix formation in 22 of
26 patient cell lines. In addition, they found significant hetero-
geneity in fibrillin-related defects in the cells from different pa-
tients. One-third of the cell strains synthesized one-half the
normal amount of fibrillin; one-third of the strains produced
fibrillin that was slowly secreted from cells and poorly incor-
porated into the extracellular matrix. In eight of the cell strains,
the quantity and efficiency of fibrillin synthesis were normal
but assimilation of fibrillin into the extracellular matrix was
abnormal (Fig. 15.13). 110

These data from DNA and protein analyses demonstrate
marked heterogeneity in mutations identified in the fibrillin
gene product and in extracellular matrix formation. Given this
heterogeneity, it has been suggested that most Marfan's syn-
drome families carry their own distinct mutation. 94 Further,

172

chapter 15 Aneurysmal disease

Figure 1 5.1 3 Diagrammatic representation of
fibrillin synthesis, secretion, proteolytic
conversion, and microfibril formation. The
roman numerals indicate the apparent location
of defects in synthesis (I), secretion (II), and
matrixaggregation (III) of fibrillin in individuals
with Marfan's. (From Milewicz DM, Pyeritz RE,
Crawford ES, Byers PH. Marfan syndrome:
defective synthesis, secretion, and extracellular
matrix formation of fibrillin by culture dermal
fibroblasts. J Clin Invest 1 992; 89:79.)

Decreased
Synthesis

II Defective Secretion

profibrilitn
fibrillin ttj

Defective matrix
incorporation

Extracellular Matrix
(microfibrils)

the molecular heterogeneity may explain the great diversity
in the clinical findings observed in patients with the
syndrome. 110 ' 111

Genetic linkage analysis may be used in the future for pre-
natal and postnatal diagnosis of Marfan's syndrome. 111 Given
the severe morbidity and the difficult task of making the diag-
nosis (particularly in youths), such genetic testing would en-
hance the identification and management of these patients. 111
Potentially, correlations could be found between specific mu-
tations or defects in fibrillin synthesis and disposition and in
specific clinical presentations. 111 If so, it may be possible to
identify those with Marfan's syndrome who are at highest risk
for aneurysm formation or other potentially morbid manifes-
tations of the disease.

Despite the exciting discoveries at the molecular level, the
clinical management of patients with Marfan's syndrome re-
mains the same. Echocardiography should be performed
starting at youth, repeated yearly until the size of the ascend-
ing aorta exceeds 50% of normal for the body surface area, and
then performed every 6 months thereafter. 105 Some authors
suggest that (3-b lockers be given to retard aortic dilation. 106 ' 109
Marsalese and colleagues recommend elective repair of the
aortic valve and ascending aorta when the aneurysm reaches
a diameter of 6 cm. 106

Ehlers-Danlos syndrome (EDS)

EDS is a connective tissue disorder characterized by hyper-
mobility of the joints, hyperelasticity and fragility of the skin,
subcutaneous nodules, and fragile blood vessels. 114 The
predominant vascular complications, which are the most
life-threatening, include dissecting aneurysm of the aorta,
systemic and intracranial aneurysm, spontaneous rupture of
the arteries, and carotid cavernous sinus fistula. 114 Marked
biochemical, genetic, and clinical heterogeneity is observed
among patients with EDS. To accommodate this diversity, the
disease is divided into 10 subtypes (Table 15.8). However,
heterogeneity exists in each subtype so only 50% of patients
with EDS will be easily diagnosed into one of these distinct
categories. 115

EDS type IV (also called the ecchymotic or arterial form) is

the severest form of the disease and is frequently associated
with arterial aneurysms and rupture of hollow organs such as
the intestine. 77 ' 80 The cardinal features of type IV EDS are ex-
tensive ecchymoses, bony prominences covered with thick
and darkly pigmented scars, and skin so thin that subcuta-
neous vessels are visible. 116 Unlike other forms of EDS, the skin
is not hyperelastic and joint hypermobility is limited to the
digits. 116 Despite the biochemical and phenotypical diversity
of these patients, genetic analyses of some patients with EDS
type IV have revealed DNA base pair mutations responsible
for defective collagen production. 77 ' 116-120 Tensile strength of
large blood vessels primarily depends on fibrils of type III col-
lagen. In type IV EDS, type III procollagen synthesis is faulty,
resulting in either an altered rate of synthesis of type III procol-
lagen or a defective structure of the procollagen molecule. 116 In
a few cases, defects in the type III collagen DNA locus, gene
COL3A1, have been identified. 77 ' 104 ' 116-119 In two patients, sin-
gle base substitutions in COL3A1 were found where glycine
codons were replaced with bulkier amino acids. 116 A larger
amino acid results in relative instability of the tertiary confor-
mation of the final collagen product that decreases the temper-
ature at which the triple helix of the protein unfolds. (Similar
single base mutations causing disruption of the tertiary struc-
ture of type I collagen are seen in osteogenous imperfecta.) 116
Further, Superti and colleagues 117 reported a 3.3-kilobase pair
deletion in one of the alleles of COL3A1 in a patient with severe
type IV EDS. This patient's fibroblasts produced normal and
shortened procollagen chains and the mutant procollagen had
decreased thermal stability and was less efficiently secreted. 117
Kontusaan and associates 116 ' 119 extended this analysis to
families with familial aneurysms but with no overt signs of
connective tissue disease. One case involved postmortem
analysis of fibroblasts from a 34-year-old man who died of in-
trathoracic and intraabdominal hemorrhage. The only aspect
of this patient's medical history that could be related to a con-
nective tissue disease was a history of easy bruisability The
patient's father died at age 43 of a ruptured AAA; his brother
died at age 35 of a ruptured aneurysm of the proximal de-
scending thoracic aorta. 116 Genetic analysis showed a muta-
tion in one allele for type III procollagen that converted a
glycine to arginine. This mutation caused aberrant splicing of

173

pa rt I Vascular pathology and physiology

Table 1 5.8 Ehlers-Danlos syndrome: classification and characteristics

Type

Mode of inheritance Molecular abnormality

Clinical features

Complications, comments

I — Gravis

Autosomal dominant

Unknown

ll-Mitis

Autosomal dominant

— Benign hypermobile Autosomal dominant

IV — Ecchymoticor
arterial form

Vl-Ocular

VII— Arthrochasasis
multiplex congentia

Autosomal dominant
and autosomal recessive

X-linked recessive

Autosomal recessive

Autosomal dominant
and autosomal recessive

Autosomal dominant

X-linked recessive

Autosomal recessive

Unknown

Decreased production
and/or synthesis of defective
type III collagen

Unknown

Lysyl hydroxylase deficiency

Procollagen accumulation
in the tissues due to either
inactive procollagen
N-terminal peptidase or to
amino acid substitution
causing resistance to
N-terminal peptidase

Unknown

Low lysyl oxidase activity,
low serum copper levels

Possible fibronectin deficiency

Classic features: hypermobility
of joints, stretchability of skin,
bruisability, arterial aneurysms,
molluscoid pseudotumors

Mild to moderate forms of the
features seen in type I

Hypermobile joints

Extensive ecchymoses, thick,
darkly pigmented scars over
bony prominences, thin skin
and visible subcutaneous
vessels

Hyperextensible and fragile
skin

Scoliosis, ocular problems,
ecchymoses, hypermobile
joints

Hyperextensible skin,
bruisability

Periodontitis, fragile skin

Occipital horns, skeletal
dysplasia, bladder neck
obstruction

Major vessel rupture,
visceral rupture, hernias,
varicose veins

Hernias, varicosities,
ecchymoses

Generalized musculoskeletal
complaints recurrent joint
dislocation

Rupture of vessels and hollow
organs, uterine rupture may
complicate pregnancy

Bruisability; may have molluscoid
pseudotumors and
recurrent dislocations

Scoliosis, ocular complications,
bruisability

Joint dislocations, ecchymoses

Early loss of teeth, dystrophic
scarring

Obstructive uropathy, osteoporosis

Hypermobile joints, bruisability, Only one case reported
mitral valve prolapse

about 50% of all the mRNA for procollagen alpha-I (type III)
chains from the patient's fibroblasts. Further analysis showed
the mutation to be dominantly inherited to one of the patient's
three children and to his 41-year-old sister. 116 In another case, a
37-year-old woman underwent genetic analysis because
many of her relatives died of ruptured AAAs. 119 Likewise, she
had a heterozygous single base mutation that converted a
glycine residue of the alpha-1 chain of type III procollagen to
arginine. The same mutation was found in pathologic speci-
mens from her mother and aunt who both died of ruptured
AAAs. The mutation was also found in her two children and in
two other first degree relatives. 119

These analyses indicate a genetic and phenotypic overlap
with patients with EDS type IV, 77 ' 116 ' 119 raising questions
regarding these and other potential molecular bases for
aneurysm disease. Presently, genetic screening is limited by
the time and resources required to sequence a large gene, such
as the type III procollagen gene. However, it is likely that
techniques will improve to allow inspection of DNA from all
of those with AAAs to assess for potentially well defined
molecular defects, and to screen other family members to
assess who is at increased risk for aneurysm formation. It is
possible such analyses will reveal greater genetic links and
incidence of aneurysm disease than are found today. 77/116/119

174

chapter 15 Aneurysmal disease

All patients with EDS should be followed from an early age
with noninvasive techniques to assess the aorta for aneurysm.
Also, even in the absence of a definitive DNA screening test,
given the high incidence of aortic aneurysm among family
members, some suggest that direct blood relatives of those
with aortic aneurysm should be followed by ultrasound at
regular intervals. 77

Infected aneurysms

The term mycotic aneurysm was coined in 1885 by Osier to de-
scribe the saccular dilations of the arch of the aorta in a man
who died from endocarditis. He based this on the morphology
of the process, not on the microbiology, since he found the
aneurysms to have "the appearance of fresh fungus vegeta-
tion/' 121 Since then, considerable change and confusion have
occurred over the nomenclature describing arterial infections.
Many still refer to all types of infected aneurysms as mycotic.
From the pathogenesis standpoint, however, there are five
types of infected aneurysms: those that form as a result of in-
travascular microbial seeding from (i) septic emboli of cardiac
origin or (ii) bacteremia; (iii) preexisting true aneurysms that
become infected; (iv) pseudoaneurysms that become infected;
and (v) aneurysms caused by an extravascular, contiguous
infection. 122

True mycotic aneurysms are caused by septic emboli of car-
diac origin. In the preantibiotic era, 86% of infected aneurysms
were sequelae of endocarditis. 122 With the subsequent anti-
biotic treatment and earlier diagnosis of bacterial endocardi-
tis, mycotic aneurysm has become an infrequent arterial
infection. In a 15-year review of the English language litera-
ture, 17% of the 220 infected aneurysms reported were associ-
ated with bacterial endocarditis. 123 More indicative of the
rarity of mycotic aneurysms in the United States today is
that in 32 patients with infected aneurysms of the aorta report-
ed in three reviews, only 6% had associated bacterial endo-
carditis. 123-125

The pathogenesis of mycotic aneurysm is based on a septic
embolus lodging in an artery causing bacterial colonization,
invasion, and disruption of the arterial wall. In large arteries,
emboli lodge in the vasa vasorum. Septic emboli may also lodge
in areas of arterial injury or disease, especially atherosclerotic
plaques. Bifurcations, acquired or congenital narrowings,
and other sites of disrupted blood flow predispose to arterial
seeding and aneurysm formation. Once a septic embolus
has lodged, acute and chronic inflammation of the arterial wall
ensues and necrosis, hemorrhage, and abscess formation
may follow. 126 Most often, rupture occurs and is contained
locally in the periarterial connective tissue, resulting in the
formation of a saccular pseudoaneurysm. Less typically, incom-
plete degradation of the arterial wall occurs, resulting in loss of
structural integrity (but not perforation), so a saccular true
aneurysm forms. Fusiform mycotic aneurysms are rare. 122,126

Mycotic aneurysms have been reported in nearly every ves-
sel including approximately 30% in visceral arteries, 30% in
the aorta, and 3-4% intracranially 122 ' 126 As in the original case
described by Osier, multiple mycotic aneurysms may be
found in the same vessel. However, multiple mycotic
aneurysms in different vessels in the same patient are rare. 122
The bacteriology of these aneurysms is the same as that of
infective endocarditis. Approximately 80% are Gram-positive
cocci in origin; streptococci species comprise 60% and staphy-
lococci species, 20%. 126 Less than 13% are caused by Gram-
negative organisms. Ironically, less than 4% of mycotic
aneurysms are caused by fungi. 126

The essential pathogenic differences between microbial ar-
teritis leading to aneurysm and mycotic aneurysm are the
source and type of bacteria. Rather than an infected embolus,
arterial seeding occurs from a distant source of bacteremia.
The normal arterial wall is relatively resistant to infection so
arterial seeding occurs at sites of atherosclerotic disease, in-
jury, narrowing, or abnormal blood flow. Once colonization
occurs, arterial degradation and contained rupture follows;
the characteristic saccular pseudoaneurysm forms as a result.
These aneurysms most often occur in vessels with advanced
atherosclerotic changes, primarily the aorta, common iliac,
femoral, and popliteal arteries. 122 Though it is typically less
atherosclerotic, the superior mesenteric artery may also be
affected. 122 Of the 22 infected aortic aneurysms reported by
Chan and associates, 124 at least 16 resulted from microbial ar-
teritis. In that series, a diverse array of bacterial sources was re-
ported including urinary tract infection, salmonellosis,
pneumonia, cellulitis, dental extraction, and intravenous line
infection. 124

An infected aneurysm may result from infection of a pre-
existing aneurysm, usually of atherosclerotic origin. 122 As
such, most infected aneurysms occur in the abdominal aorta
and are fusiform. Analogous to microbial arteritis, bacteria
from distinct sites of infection colonize the intraluminal
thrombus and atherosclerotic plaques of the preexisting
aneurysm. The wall of the aneurysm thins as acute polymor-
phonuclear inflammation, necrosis, and abscess formation
occur. 126 On gross inspection an infected aortic aneurysm
appears similar to an atherosclerotic fusiform aneurysm. The
primary distinction is the particularly thin wall of an infected
aneurysm, and necrosis and abscess may only be identified by
the pathologist. 122 Since most aneurysms are not cultured at
operation, infected atherosclerotic aneurysms are probably
underreported. 122

Microbial arteritis and infected aneurysms have a markedly
different bacteriologic profile compared with mycotic
aneurysms. Gram-positive cocci are found in 60% of cases; the
causative organism is staphylococcus in 40% of cases over-
all. 122 ' 124,126 Gram-negative infection is present in approx-
imately 35% of cases, the majority from salmonella. 122 ' 125 ' 126
The presumed portal of entry of salmonella is the gastrointesti-
nal tract; however, the predilection for the involvement of this

175

pa rt I Vascular pathology and physiology

organism is not understood. 126 Syphilitic aneurysms are a type
of microbial arteritis. The spirochetes penetrate the aortic wall
through the vasa vasorum, destroy the elastic and muscular
elements of the arterial wall, cause ingrowth of fibrotic gra-
nulation tissue, and predispose to formation of saccular
aneurysms. 122 Now that tertiary syphilis is rare, so are these
aneurysms. There is a predilection for infected aneurysms in
immunosuppressed patients. In Johansen and Devin's review,
24% of patients with infected aneurysms had depressed im-
munocompetence. In Chan et al.'s review of infected aortic
aneurysms, over 50% of the patients had factors predisposing
to defects in immunocompetency including steroid or cytotox-
ic drug use, chronic renal failure, and severe alcoholism. 123 ' 124
In infected pseudoaneurysm, the pseudoaneurysm forms
following injury to the arterial wall and infection from the
original traumatic inoculum follows. Intravenous drug
abusers comprise the predominant group. Infected pseudo-
aneurysms are found in arteries of the upper extremity and
groin, which are easily accessible for drug injection. Johnson
and colleagues found 76% of infected pseudoaneurysms in
intravenous drug abusers to be infected with Staphylococcus
aureus, and 18% with Pseudomonas aeruginosa. 127 Infected
pseudoaneurysm may follow iatrogenic injury from percuta-
neous vessel cannulation or vessel cutdown and even after
percutaneous insertion of devices to close femoral artery
catheterization sites. In trauma patients, most infected
pseudoaneurysms follow penetrating trauma. Extravascular
infection contiguous to an artery may erode the adventitia and
cause pseudoaneurysm formation. Infections next to the tho-
racic and abdominal aorta and some peripheral vessels are the
most common sources. 122 In this setting, salmonella is the
organism most often found in the abdominal aorta; there is a
high rate of associated lumbar osteomyelitis. 122 Staphylococcus
is also common. Aneurysms caused by tuberculosis result
from contiguous infection and are rare. 126 However, with the
increasing incidence of tuberculosis, vascular surgeons in
certain regions have again encountered this infectious
etiology of aortic aneurysm.

References

1. Collins I. The epidemiology of abdominal aortic aneurysm. Br J
Hosp Med 1988; 40:64.

2. Goyanes DJ. Substitution plastica de las arterias por las venae,
o arterioplastia venosa, aplicada, como neuvo metodo, al
tratarniento de los aneunsmas. SigloMed 1906; 53:346.

3. Dubost C, Allary M, Oeconomas N. Resection of an aneurysm of
the abdominal aorta: re-establishment of the continuity by a pre-
served human arterial graft, with results after five months. Arch
Surg 1952; 64:405.

4. Scott RAP, Ashton HA, Kay DN. Abdominal aortic aneurysm in
4237 screened patients: prevalence, development and manage-
ment over six years. Br J Surg 1991; 78:1122.

5. McFarlane MJ. The epidemiologic necropsy for abdominal aortic
aneurysm. JAMA 1991; 265:2085.

6. Taylor LM, Porter JM. Basic data related to clinical decision mak-
ing in abdominal aortic aneurysms. Ann Vase Surg 1986; 1:502.

7. Norman PE, Castleden WM, Hockey RL. Prevalence of ab-
dominal aortic aneurysm in Western Australia. Br } Surg 1991;
78:1118.

8. Fowkes FG, Macintyre CC, Ruckley CV. Increasing incidence of
aortic aneurysms in England and Wales. Br Med J 4989; 298:33.

9. Reed D, Reed C, Stemmermann G, Hayashi T Are aortic
aneurysms caused by atherosclerosis? Circulation 1992; 85:203.

10. Reed RC. Abdominal aortic aneurysm. Leriche's syndrome,
inguinal herniation, and smoking. Arch Surg 1984; 119:387.

11. Benditt EP, Benditt JM. Evidence for a monoclonal origin of
human atherosclerotic plaques. Proc Natl Acad Sci USA 1973;
70:1753.

12. Cohen JR, Sarfati I, Ratner L, Tilson MD. Alpha-I antitrypsin
phenotypes in patients with aortic aneurysms. / Surg Res 1990;
49:319.

13. MacSweeney ST, Ellis M, Worrell PC et al. Smoking and growth
rate of small abdominal aortic aneurysms. Lancet 1994; 344
(8923):651.

14. Baird RI, Gurry JS, Kellan J, Plume SK, Arteriosclerotic femoral
artery aneurysms. Can Med Assoc J 1977; 117:1306.

15. Cutler BS, Carling RC. Surgical management of arteriosclerotic
femoral aneurysms. Surgery 1973; 74:764.

16. Graham L, Zelenock GB, Whitehouse WM et al. Clinical signifi-
cance of arteriosclerotic femoral artery aneurysms. Arch Surg
1980; 115:502.

17. Lenthan DJ, Zeman HS, Collins GJ. Left main coronary artery
aneurysm in association with severe atherosclerosis: a case re-
port and review of the literature. Cathet Cardiovasc Diagn 1991;
23:28.

18. Breslin DJ, Jewel ER. Peripheral aneurysms. Cardiovasc Clin 1991;
9:489.

19. Courtney DF, Bergan JJ. Vascular systems: the peripheral vascu-
lar system. In: Kyle J, Carey CC, eds. Scientific Foundations of
Surgery, 4th edn. Chicago: Year Book Medical Publishers,
1989:328.

20. Economou SG, Taylor CB, Beattie El Jr et al. Persistent experimen-
tal aortic aneurysms in dogs. Surgery 1960; 47:21.

21. Wolinsky H, Glagov S. Nature of species' differences in the medi-
al distribution of aortic vasa vasorum in mammals. Circ Res 1967;
20:409.

22. Wolinsky H, Glagov S. Lamellar units of aortic medial structure
and function in mammals. Circ Res 1967; 20:99.

23. Wolinsky H. Comparison of medial growth of human thoracic
and abdominal aortas. Circ Res 1970; 27:531.

24. Glagov S. Hemodynamic risk factors: mechanical stress, mural
architecture, medial nutrition and the vulnerability of arteries to
atherosclerosis. In: Wisller RW, Geer JC, Kaufman N, eds. The
Pathogenesis of Atherosclerosis. Baltimore: Williams & Wilkins,
1972:164.

25. Heistad DD, Marcus ML, Carsen GE et al. Role of vasa vasorum in
nourishment of the aortic wall. Am J Physiol 1981; 240:11781.

26. Palma EC. Etiopatogenia. Cir Urug 1976; 46:425.

27. Heistad DD, Armstrong ML, Marcus ML. Hyperemia of the
aortic wall in atherosclerotic monkeys. Circ Res 1981; 48:669.

28. Zatina MA, Zanns CK, Genertz BL et al. Role of medial lamellar
architecture in the pathogenesis of aortic aneurysms. / Vase Surg
1984; 1:442.

176

chapter 15 Aneurysmal disease

29. Wilens SL, Malcom JA, Vasquez JM. Experimental infarction
(medial necrosis) of dog aorta. Am J Pathol 1965; 47:645.

30. Guirguis EM, Barber GO. The natural history of AAAs. Am } Surg
1991; 162:481.

31. Stringfellow MM, Lawrence PF, Stringfellow KG. The influence
of aorta aneurysm geometry upon stress in the aneurysm wall.
J Surg Res 1987; 42:425.

32. Gosling RG, Newman DC, Bowden LR et ah The area ratio of
normal aortic junctions. Br J Radiol 1971; 44:850.

33. Leung DYM, Glagov S, Mathews MB. Cyclic stretching stimu-
lates synthesis of matrix components by arterial smooth muscle
cells in vitro. Science 1976; 191:475.

34. Leach SD, Toole AL, Stern Hetal. Effect of beta-adrenergic block-
ade on the growth rate of abdominal aortic aneurysms. Arch Surg
1988; 123:606.

35. Slaiby JM, Ricci MA, Gadowski GR et ah Expansion of aortic
aneurysms is reduced by propranolol in a hypertensive rat
model. / Vase Surg 1994; 20:178.

36. Gadowski GR, Pilcher DB, Ricci MA. Abdominal aortic
aneurysm expansion rate: effect of size and beta-adrenergic
blockade. / Vase Surg 1994; 19:727.

37. Propranolol Aneurysm Trial Investigators. Propranolol for small
abdominal aortic aneurysms: results of a randomized trial. / Vase
Surg 2002; 35:72.

38. Shapira OM, Pakis S, Wassermann JP, Barzlllai N, Mashlah A.
Ultrasound screening for abdominal aortic aneurysms in
patients with atherosclerotic peripheral vascular disease.
/ Cardiovasc Surg 1990; 31:170.

39. Bengtson H, Ekberg O, Aspdlin P, TakoLander R, Bergqvist D.
Aneurysmal disease: abdominal aortic dilatation in patients
operated on for carotid artery stenosis. Acta Chit Scand 1988;
143:441.

40. Thiele BL, Bandyk DF. The peripheral vascular system. In: Miller
TA, ed. Physiologic Basis of Modern Surgical Care. St Louis: CV
Mosby, 1988:848.

41. McCullagh KG, Duance VC, Bishop KA. The distribution of col-
lagen types I, II, and V(AB) in normal and atherosclerotic human
aorta. / Pathol 1980; 130:45.

42. Sherer PW. Flow in an axisymmetrical glass model aneurysm.
JBiomech 1973; 6:695.

43. Tilson MD, Stansel HC. Differences in results for aneurysm
versus occlusive disease after bifurcation grafts: results of 100
elective grafts. Arch Surg 1980; 115:1173.

44. Swanson RI, Littooy RN, Hunt TK. Laparotomy as a precipitat-
ing factor in the rupture of intra-abdominal aneurysms. Arch
Surg 1980; 115:299.

45. Busuttil RW, Cardenas A. Collagenase and elastase activity in the
pathogenesis of the abdominal aortic aneurysms. In: Bergan JJ,
Yao JST, eds. Aneurysms: Diagnosis and Treatment. Orlando: Grune
&Stratton, 1982:83.

46. Busuttil RW, Rinderbiaecht H, Flesher A et ah Elastase activity:
the role of elastase in aortic aneurysm formation. / Surg Res 1982;
32:214.

47. Sumner DS, Hokanson DE, Strandness DE Jr. Stress-strain char-
acteristics and collagen-elastin content of abdominal aortic
aneurysms. Surg Gynecol Obstet 1970; 130:459.

48. Tilson MD. Histochemistry of aortic elastin in patients with non-
specific abdominal aortic aneurysmal disease. Arch Surg 1988;
123:503.

49. Busuttil RW, Abon-Zamzam AM, Machledar HI. Collagenase ac-
tivity of the human aorta: a comparison of patients with and
without abdominal aortic aneurysms. Arch Surg 1980; 115:1373.

50. Menashi S, Campa J, Greenhalgh R et ah Collagen in abdominal
aortic aneurysms: typing, content, and degradation. / Vase Surg
1987; 6:578.

51. Zarins CK, Runyon-Hass A, Zatina MA, Lu CT. Increased colla-
genase activity in early aneurysmal dilatation. / Vase Surg 1986;
3:2 38.

52. Dobrin PB, Baker WH, Gley WC. Elastolytic and collagenolytic
studies of arteries: implications for the mechanical properties of
aneurysms. Arch Surg 1984; 119:405.

53. Carrell TWG, Burnand KG, Wells GMA et ah Stromelysin-1
(matrix metalloproteinase-3) and tissue inhibitor of
metalloproteinase-3 are overexpressed in the wall of abdominal
aortic aneurysms. Circulation 2002; 105:477.

54. Nollendorfs A, Greiner TC, Nagase H, Baxter BT. The expres-
sion and localization of membrane type-1 matrix metallopro-
teinase in human abdominal aortic aneurysms. / Vase Surg 2001;
34:316.

55. McMillan WD, Tamarina NA, Cipollone M et ah Size matters: the
relationship between MMP-9 expression and aortic diameter.
Circulation 1997; 96:2228.

56. Newman KM, Jean-Claude J, Li H et ah Cellular localization of
matrix metalloproteinases in the abdominal aortic aneurysm
wall. / Vase Surg 1994; 20:814.

57. Yamashita A, Noma T, Nakazawa A et ah Enhanced expression of
matrix metalloproteinase-9 in abdominal aortic aneurysms.
World J Surg 2001} 25:259.

58. Boyle JR, McDermott E, Crowther M et ah Doxycycline inhibits
elastin degradation and reduces metalloproteinase activity in a
model of aneurysmal disease. / Vase Surg 1998; 27:354.

59. Mosorin M, Juvonen J, Biancari F et ah Use of doxycycline to de-
crease the growth rate of abdominal aortic aneurysms: a random-
ized, double-blind, placebo-controlled pilot study. / Vase Surg
2001; 34:606.

60. Curci JA, Mao D, Bohner DG et ah Preoperative treatment with
doxycycline reduces aortic wall expression and activation of
matrix metalloproteinases in patients with abdominal aortic
aneurysms. / Vase Surg2000; 31:325.

61. Vammen S, Lindholt JS, Ostergaard L et ah Randomized
double-blind controlled trial of roxithromycin for prevention of
abdominal aortic aneurysm expansion. Br] Surg 2001; 88:1066.

62. Boontje All, van den Dugen JJ, Blanksma C. Inflammatory
abdominal aortic aneurysms. / Cardiovasc Surg 1990; 31:661.

63. Fennel RC, Holier LH, Lie JT et ah Inflammatory abdominal
aortic aneurysms: a 30-year review. / Vase Surg 1985; 2:859.

64. Crawford JL, Sorwe CL, Safi RI et ah Inflammatory aneurysms of
the aorta. / Vase Surg 1985; 2:113.

65. Leu RI. Inflammatory abdominal aortic aneurysms: a disease
entity? Histological analysis of 60 cases of inflammatory aortic
aneurysms of unknown etiology. Virchows Arch 1990; 417:427.

66. Sterpetti AV, Hunter WI, Feldhaus RI et ah Inflammatory
aneurysms of the abdominal aorta: incidence, pathologic and
etiologic considerations. / Vase Surg 1989; 9:643.

67. Tilson MD, Seashore MR. Fifty families with abdominal aortic
aneurysms in two or more first-order relatives. Am } Surg 1984;
147:551.

68. Tilson MD, Dang C. Generalized arteriomegaly: a possible

177

pa rt I Vascular pathology and physiology

predisposition to the formation of abdominal aortic aneurysms.
Arch Surg 1982; 117:1212.

69. Johansen K, Koepsell T. Familial tendency for abdominal aortic
aneurysms. JAMA 1986; 256:1934.

70. Tilson M. Further studies of a putative crosslinking amino acid
(3-deoxy-pyridinoline) in skin from patients with AAAs. Surgery
1985; 98:888.

71. Kitchen ND. Racial distribution of aneurysms in Zimbabwe. JR
Soc Med 1989; 82:136.

72 . Webster M W, Ferrel RE, St Jean FL, Majuroder PP, Fogel SR, Steed
DL. Ultrasound screening of first-degree relatives of patients
with abdominal aortic aneurysm. / Vase Surg 1991; 12:9.

73. Teien D, Finley JP, Murphy DA, Lacson A, Longhi J, Gills DA.
Idiopathic dilatation of the aorta with dissection in a family
without Marfan syndrome. Acta Paediatr Scand 1991; 80:1246.

74. Hoffler LII, Stanson AW, Gloriczki P et ah Arteriomegaly: classifi-
cation and morbid implications of diffuse aneurysmal disease.
Surgery 1983; 93:700.

75. Tilson MD. Generalized arteriomegaly: a possible predisposition
to the formation of abdominal aortic aneurysms. Arch Surg 1981;
116:1030.

76. Rowe DW, McGoodwin EB, Martin GR. A sex-linked defect in the
crosslinking of collagen and elastin associated with the mottled
Cocusin mice. / Rep Med 1974; 139:180.

77. Kuivarilemi H, Tromp G, Prockop DJ. Genetic causes of aortic
aneurysms: unlearning at least part of what the textbooks say.
} Clin Invest 1991; 88:1441.

78. Superti-Furga A, Steinmann B, Ramirez F, Byers PH. Molecular
defects of type III procollagen in Ehlers-Danlos syndrome type
IV. Hum Genet 1989; 82:104.

79. Tromp G, Kuivaniemi H, Shikata H, Prockop DJ. A single base-
mutation that substitutes serine for glycine 790 of the a 1 (III)
chain of type III procollagen exposes an arginine and causes
Ehlers-Danlos syndrome IV JBiolChem 1989; 264:1349.

80. Majumder PP, St Jean PL, Ferell RE, Webster MW, Steed DL. On
the inheritance of abdominal aortic aneurysms. Am J Hum Genet
1991;48:164.

81. Delin A, Ohlsen H, Swedenborg J. Growth rate of abdominal
aortic aneurysms as measured by computed tomography. Br J
Surg 1985; 72:530.

82. Dobrin PB. Pathophysiology and pathogenesis of aortic
aneurysms: current concepts. Surg Clin North Am 1989; 69:687.

83. Cronenwett IL, Murphy TF, Zelenock GB et ah Actuarial analysis
of variables associated with rupture of small abdominal aortic
aneurysms. Surgery 1985; 98:472.

84. Bernstein EF. The natural history of abdominal aortic aneurysms.
In: Najarian IS, Delaney JP, eds. Vascular Surgery. Miami:
Symposia Specialists, 1978:441.

85. Bernstein EF, Chan EL. Abdominal aortic aneurysm in high risk
patients. Ann Surg 1984; 200:255.

86. Cohn J, Heather B, Walter J. Growth rates of subclinical abdomi-
nal aortic aneurysms: implications for review and rescreening
programmes. Eur] Vase Surg 1991; 5:141.

87. Piotrowski JJ, Mcintyre XE, Hunter GC, Sethi GK, Bernhard VM,
Copeland JC. Abdominal aortic aneurysm in the patient under-
going cardiac transplantation. / Vase Surg 1991; 14:460.

88. Szilagyi DE, Smith RE. Clinical fate of the patient with
asymptomatic abdominal aortic aneurysm and unfit for surgical
treatment. Arch Surg 1972; 104:600.

89. Matsuchita S, Kuroo M, Takagi T, Hou B, Kuramoto K. Cardio-
vascular disease in the aged: overview of an autopsy series.
JpnCircJ 1988; 52:442.

90. Kaufman JA, Bettmann MA. Prognosis of abdominal aortic
aneurysms: a population-based study. Invest Radiol 1991; 26:612.

91. Castleden WM, Merceri L. Abdominal aortic aneurysms in west-
ern Australia: descriptive epidemiology and patterns of rupture.
Br J Surg 1985; 72:109.

92. White GH, Advani SM, Williams RA, Wilson SE. Cardiac risk
index as a predictor of long-term survival after repair of abdomi-
nal aortic aneurysm. Am } Surg 1988; 156:108.

93. Johansen K, Kohler TR, Nicholls SC, Zierler RE, Clowes AW,
Kaymers A. Ruptured AAA: the Harborview experience. / Vase
Surg 1991; 13:240.

94. lvers CR, Bourke BM. Elective aneurysm repair and the incidence
of aortic rupture in an aging population. Aust NZ J Surg 1990;
60:203.

95. Evans WE, Connolly SE, Bernhard V Popliteal aneurysms.
Surgery 1971; 70:762.

96. Dawson I, van Bockel JH, Brand R, Terpstra IL. Popliteal artery
aneurysms: long-term follow-up of aneurysmal disease and
results of surgical treatment. / Vase Surg 1991; 13:398.

97. UK Small Aneurysm Trial Participants. Mortality results for
randomised controlled trial of early elective surgery or ultraso-
nographic surveillance for small abdominal aortic aneurysms.
Lancet 1998; 352:1649.

98. Quill DS, Colgan MP, Sumner DS. Ultrasonic screening for the
detection of abdominal aortic aneurysms. Surg Clin North Am
1989; 69:713.

99. Cohn J, Murie J, Moms PJ. Two-year prospective analysis of the
Oxford experience with surgical treatment of abdominal aortic
aneurysm. Surg Gynecol Obstet 1989; 169:527.

100. Hatakeyama T, Shigematsu H, Muto T. Risk factors for rupture of
abdominal aortic aneurysm based on three-dimensional study.
} Vase Surg 2001; 33:453.

101 . Kato K, Ishiguchi T, Maruyama K et ah Accuracy of plastic replica
of aortic aneurysm using 3D-CT data for transluminal stent-
grafting: experimental and clinical evaluation. / Comput Assist
Tomogr 2001; 25:300.

102. Willmann JK, Lachat ML, Von Smekal A et ah Spiral-CT
angiography to assess feasibility of endovascular aneurysm
repair in patients with ruptured aortoiliac aneurysm. Vasa 2001;
30:271.

103. Di Martino ES, Guadagni G, Fumero A et ah Fluid-structure
interaction within realistic three-dimensional models of the
aneurysmatic aorta as a guidance to assess the risk of rupture
of the aneurysm. Med Eng Phys 2001; 23;647.

104. Lee B, D'Alessio M, Vissing H, Ramirer F, Steinmann B, Superti
FA. Characterization of a large deletion associated with a
polymorphic block of repeated dinucleotides in the type UI
pro-collagen gene (COL3A1) of a patient with Ehlers-Danlos
syndrome type IV Am J Hum Genet 1991; 48:511.

105. Rowe DW, Shapiro JR. Disorders of bone and structural proteins.
In: Kehey WN, Harris JED, Ruddy S, Sledge CB, eds. Textbook
of Rheumatology, 3rd edn, Vol 1. Philadelphia: WB Saunders,
1989:1691.

106. Marsalese DL, Moodie DS, Vacante M et ah Marfan's syndrome:
natural history and long-term follow-up of cardiovascular
involvement. J Am Coll Cardiol 1989; 14:422.

178

chapter 15 Aneurysmal disease

107. van Ooijen B. Marfan's syndrome and isolated aneurysm of the
abdominal aorta. Br Heart J 1988; 59:81 .

108. Pruzinsky MS, Katz NM, Green CE, Satler LF. Dilated descend-
ing thoracic aortic aneurysm in Marfan's syndrome. Am } Cardiol
1988; 61:1159.

109. Joyce W. Uncommon arteriopathies. In: Rutherford RB, ed.
Vascular Surgery, 3rd edn. Philadelphia: SYB Saunders, 1989:276.

110. Milewicz DM, Pyeritz RE, Crawford ES, Byers PH. Marfan syn-
drome: defective synthesis, secretion, and extracellular matrix
formation of fibrillin by cultured dermal fibroblasts. / Clin Invest
1992; 89:79.

111. Tsipouras P, Del MR, Sarfarazi M et ah Genetic linkage of the
Marfan syndrome, ectopia lentis, and congenital contractual
arachnodactyly to the fibrillin genes on chromosomes 15 and 5:
the international Marfan Syndrome Collaborative Study. N Engl J
Med 1992; 326:905.

112. Kainulainen K, Sakal LY, Child A et ah Two mutations in Marfan
syndrome resulting in truncated fibrillin polypeptides. Proc Natl
Acad Sci USA 1992; 89:5917.

113. Dietz HC, Cutting GR, Pyeiitz RE et ah Marfan syndrome caused
by a recurrent de novo missense mutation in the fibrillin gene.
Nature 1991; 353:337.

114. Serry C, Agomuoh OS,. Goldln MD. Review of Ehlers-Danlos
syndrome. / Cardiovasc Surg 1988; 29:530.

115. Uitto J, Murray LW, Blumberg B, Shamban A. Biochemistry of
collagen in diseases. Ann Intern Med 1986; 105:740.

116. Kontusaan S, Tromp G, Kuivaniemi H, Ladda RL, Prockop DJ.
Inheritance of an RNA splicing mutation (G(llivs20)) in the type
UI procollagen gene (COL3A 1) in a family having aortic
aneurysms and easy bruisability Phenotypic overlap between
familial arterial aneurysms and Ehlers-Danlos syndrome type
IV. Am} Hum Genet 1990; 47:112.

117. Superti FA, Gugler B, Girzelmann R, Steinmann B. Ehlers-
Danlos syndrome type IV: a multi-exon deletion in one of the
two COL3AI alleles affecting structure, stability, and processing
of type 111 procollagen. J Biol Chem 1988; 263:6226.

118. Superti FA, Steinmann B, Ramirez F, Byers PH. Molecular defects
of type UI procollagen in Ehlers-Danlos syndrome type IV Hum
Genet 1989; 82:104.

119. Kontusaan S, Tromp G, Kuivaniemi H, Romanic AM, Prockop
DI. A mutation in the gene for type UI procollagen (COL3 AI) in a
family with aortic aneurysms. / Clin Invest 1990; 86:1465.

120. Lee B, Godfrey M, Vitale B et ah Linkage of Marfan syndrome and
a phenotypically related disorder to two different fibrillin genes.
Nature 1991; 353:330.

121. Osier W. The Gulstonian lectures on malignant endocarditis.
Br Med J 1885; 1:467.

122. Wilson SE, Wagenen PY, Passaro B Jr. Arterial infection. In:
Ravitch MM, ed. Current Problems in Surgery. Chicago: Year
Book Medical Publishers, 1978:6.

123. Johansen K, Devin J. Mycotic aortic aneurysms. Arch Surg 1983;
118:583.

124. Chan FY, Crawford ES, Coseill JS, Safi HJ, Williams TJ. In situ
prosthetic graft replacement for mycotic aneurysm of the aorta.
Ann ThoracSurg 1989; 47:193.

125. Gomes MN, Choyke FL, Wallace RB. Infected aortic aneurysms:
a changing entity. Ann Surg 1992; 215:435.

126. Scheld WM, Sande MA. Endocarditis and intravascular infec-
tions. In: Mandeil GL, Douglas RG Jr, Bennett SE, eds. Principles
and Practice of Infectious Diseases, 3rd edn. New York: Churchill
Livingstone, 1990:670.

127. Johnson JR, Ledgerwood AM, Lucas CE. Mycotic aneurysm: new
concepts in therapy. Arch Surg 1983; 118:577.

179

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Pathophysiology of renovascular
hypertension

David L. Robaczewski
Richard H. Dean
Kimberley J. Hansen

The importance of the kidney in hypertension was first recog-
nized by Richard Bright when he described the association be-
tween left ventricular hypertrophy and contracted kidneys in
1827. By 1836, Bright had reported the association of albumin-
uria, cardiac hypertrophy, shrunken kidneys, and hardness of
the pulse. His impact on the understanding of the cause of
hypertension is unparalleled. For much of the 19th century, the
medical community blindly accepted the premise that all
causes of hypertension emanated from the kidney. This per-
ception was augmented when Tigerstedt and Bergman sug-
gested the presence of a pressor substance in the kidney in
1898. In their experiments, a crude saline extract derived from
rabbit kidneys was shown to increase the blood pressure when
injected into other rabbits. They termed the uncharacterized
chemical in the kidney extract renin. Their observations were
controversial and a matter of curiosity for over three decades.
Goldblatt provided more support for the concept that a renal
hormonal mechanism could induce hypertension. By the late
1920s and early 1930s, the association between hypertension
and arteriolar nephrosclerosis was widely recognized. What
was not agreed on was the order of occurrence. Did hyperten-
sion cause arteriolar nephrosclerosis or did primary arteriolar
nephrosclerosis produce the hypertension? Goldblatt rea-
soned arteriolar nephrosclerosis was the equivalent of
millions of tiny vascular clamps limiting inflow into the
glomerulus. Since placement of such microscopic clamps was
impossible, he concluded that constriction of the main renal
artery would serve the same purpose of reducing flow to the
glomeruli. His observation that hypertension developed after
the clamp placement seemed to confirm what he suspected:
arteriolar nephrosclerosis produced hypertension. But he
did not expect to find the dissipation of hypertension that
occurred when he subsequently removed the occluding
clamp. These events made up the beginning of our under-
standing of renovascular hypertension.

In 1939, cooperative studies in the laboratories of Page in
the United States and Braun-Menendez in Argentina led to the
characterization of renin as a proteolytic enzyme and the dis-
covery of a rapidly acting, potent pressor byproduct. This was

followed by the structural characterization of angiotensin I,
angiotensin-converting enzyme, and angiotensin II by Skeggs
and Lentz in the 1950s. Cook localized renin production to the
juxtaglomerular apparatus by the end of that decade while the
rapid conversion of angiotensin I to angiotensin II during
its passage through the lungs was demonstrated by Ng and
Vane in the late 1960s. 1 Subsequent research resulted in iden-
tification and manipulation of the renin gene, angiotensin-
converting enzyme inhibitors, angiotensin receptors, active
metabolic byproducts of angiotensin, and discovery of local
renin-angiotensin systems.

Sixty years of laboratory investigation allowed the charac-
terization of the mechanisms responsible for renovascular
hypertension and led to new methods of medical and surgical
intervention. Clinical application of these discoveries has
significantly improved diagnostic capabilities and overall
patient management. The renin-angiotensin system is respon-
sible for the hypertensive response seen with renal artery dis-
ease. The physiologic effect of the renin-angiotensin system
influences renal, cardiovascular, neural, adrenal, and micro-
circulatory function. Intuitively, restoration of blood flow to
the chronically ischemic kidney should correct the hormonal
overactivity and result in normalization of blood pressure.
Unfortunately, this is frequently not the case. 2-5 The patho-
physiologic mechanisms that limit the success of reperfusion
are incompletely characterized. After reviewing the basics of
angiotensin metabolism and renal physiology, this chapter
will present the physiologic mechanisms associated with
renovascular hypertension.

Renin-angiotensin system

The metabolic process responsible for renovascular hyper-
tension is complex and continues to be investigated. The major
components of this process are the renin-angiotensin system,
the sympathetic nervous system, and vasoactive hormones.
Though these components have global influences, they
especially affect renal, cardiovascular, central nervous, and

180

chapter 16 Pathophysiology of renovascular hypertension

adrenal function. In the classic two-kidney /one-clip Goldblatt
model, renal perfusion pressure and glomerular filtration
are reduced by critical renal artery stenosis. The renin-
angiotensin system becomes activated by tubuloglomerular
and renal baroreceptor mechanisms. These mechanisms in-
duce production of the proteolytic enzyme, renin. This rate-
limiting enzyme initiates the cascade of angiotensin peptides
that are responsible for the changes associated with the
renin-angiotensin system. This hormonal system can be
viewed in short- and long-term clinical settings. With regard to
short-term decreases in renal perfusion pressure, such as in
hypovolemia, this system reestablishes normal rates of renal
plasma flow and glomerular filtration by increasing blood
pressure, and sodium retention, and enhancing behavior ac-
tivities associated with thirst. In long-term disease processes,
such as renovascular occlusive disease, renal blood flow (RBF)
and glomerular filtration are chronically decreased, resulting
in sustained activation of the renin-angiotensin system. This
chronic activation of the renin-angiotensin system is respon-
sible for the dangerous elevations in blood pressure associ-
ated with renovascular hypertension.

In this enzyme cascade, renin enzymatically cleaves a plas-
ma oc 2 -globulin, angiotensinogen, to produce the decapeptide,
angiotensin I (Fig. 16.1). Hemodynamically inactive angioten-
sin I is then metabolized by angiotensin-converting enzyme
to produce the potent vasoconstrictor, angiotensin II. An-
giotensin II increases systemic arterial pressure by increasing
vascular resistance and enhancing salt- and water-conserving
mechanisms. These activities are accomplished through its
influence on the renal, cardiovascular, central nervous, and
adrenal systems. Our understanding of the renin-angiotensin
system has been markedly enhanced by new gene technology.
With these techniques, the capacity to produce the compo-
nents of the renin-angiotensin system has been found in the
renal, cardiovascular, central nervous, and adrenal systems. 6
The activity of these tissue renin-angiotensin systems is regu-
lated by various factors. This has led to new hypotheses
regarding the mechanisms that maintain renovascular hyper-
tension. Contemporary schemes incorporate the classic sys-
temic renin-angiotensin system as the developmental phase
of hypertension. This relates to the fact that immediately
following critical renal artery stenosis plasma renin activity,

ANGIOTENSINOGEN

renin

Angiotensin-(2-10) -*■

Ami nope ptidase A

Angiotensin I

[NH2-Asp-Aig Val-Tyr-Ile His Pro-Phe-His-Leu-COOH]

ACE

ACE I

Tonin
Cathepsin G

Angiotensin II,

Ang-(l-8)

Ami no peptidase A

t
Angiotensin-(2-8)

(Ang ID)

Prolyl-endopeptidase
Endopepiidase 24.11
M em brane- bound
endopepiidase

ProJyl-endopeptidase
Carboxipepiidases

Am i nopeptidascs
Endopeptidases
C arbox t peptidases

*Xngiotensin-(l-7)

Smaller Fragments

Figure 16.1 Schematic outline of alternate enzymatic cascades based on data obtained by this laboratory. (From FerrarioCM, Barnes KL,

participating in generation of biologically active angiotensin peptides in Block CH etal. Pathways of angiotensin formation and function in the brain,

tissues. Putative pathways contributing to formation of angiotensin-(1 -7) are Hypertension 1990; 15[Suppl. I]:1 13.)

181

pa rt 1 Vascular pathology and physiology

angiotensin II levels and systemic arterial pressure increase.
These changes are completely reversed by early revasculariza-
tion or administration of angiotensin-converting enzyme
inhibitors. 7 After this initiating phase, the salt- and water-
conserving actions of the renin-angiotensin system gradually
become more responsible for the maintenance of hyperten-
sion. Chronic hypertension and elevated angiotensin II levels
upregulate the production of angiotensin peptides by local
integrated renin-angiotensin systems of the renal, cardiovas-
cular, adrenal, and central nervous systems. Eventually, these
tissue renin-angiotensin systems gain a prominent position in
the maintenance of renovascular hypertension. 8

Renin

Renin is a proteolytic enzyme that is produced and released
from the juxtaglomerular apparatus. Specialized vascular
smooth muscle cells of the afferent and efferent arterioles,
juxtaglomerular cells, are major sites of production and stor-
age. Stimuli for renin production and secretion include
tubuloglomerular feedback, hyponatremia, renal barorecep-
tors, prostaglandin I 2 , angiotensin II, and postganglionic sym-
pathetic nerves. Once produced, renin is stored in granules or
is released into the plasma. Metabolism and renin clearance
occurs in the liver.

Tigerstedt and Bergman's finding that renal vein effluent
increases levels of renin is clinically applied in the diagnostic
approach to renovascular hypertension by direct measure-
ment of renal vein renin levels. Such functional studies are sug-
gested to determine the significance of stenotic lesions and,
thus, enhance patient selection for surgical repair. The mea-
sured levels are analyzed via the renal vein : renin ratio and the
renal : systemic renin index. In the renal vein : renin ratio,
renin activity from ischemic and nonischemic kidneys is com-
pared. This method is most applicable in unilateral renal artery
stenosis. A ratio of 1.5 is consistent with a significant, cor-
rectable renal artery lesion. 2 The renal : systemic renin index
test compares renin activity from each renal vein with respect
to systemic renin levels. The findings from these tests are used
to predict the success of blood pressure response to revascular-
ization or nephrectomy. High false-negative rates limit the use-
fulness of renal vein : renin ratio studies. The renal : systemic
renin index is better able to predict hypertension cure.

Renin production is not limited to the afferent arterioles.
In fact, renin is present in the plasma of nephrectomized
patients. The sources of this renin are many and include the
vascular endothelium. This discovery of specific tissue
renin-angiotensin systems has led to reevaluation of the fac-
tors involved in the development and maintenance of reno-
vascular hypertension.

Angiotensin peptides

Angiotensinogen is an oc 2 -globulin abundantly produced by

the liver and secreted in the plasma. Recently, the vascular en-
dothelium of the brain, heart, adrenal glands, and kidneys has
also been found to produce this substrate. For all of these organ
beds, gene expression and substrate production increase in the
presence of angiotensin II and chronic hypertension. 9-11 The
byproduct of this substrate, angiotensin I, has no significant
vasoactivity However, it may influence renal tubular func-
tion. 12 Angiotensin metabolism continues as angiotensin-
converting enzyme removes two amino acids from the car-
boxyl end of angiotensin I to form the potent vasoconstrictor,
angiotensin II. Though the traditional description of the
renin-angiotensin system limits angiotensin-converting en-
zyme activity to the pulmonary circulation, it is accepted that
angiotensin-converting enzyme is produced and enzymati-
cally active in many vascular endothelial cells. Angiotensin-
converting enzyme production also increases under the
influence of hypertension and angiotensin II. Pharmacologic
inhibitors of this enzyme block the activities associated with
angiotensin II. These agents have had a major impact on the
clinical management of hypertension and our understanding
of the renin-angiotensin system. Their development has led to
the identification of new angiotensin peptides, angiotensin
receptor subtypes, and their specific tissue activities. To ap-
preciate the physiologic importance of converting enzyme
inhibitors, it is necessary to understand the activities of
angiotensin I, angiotensin II, their metabolites, and different
receptor subtypes.

The major metabolites of angiotensinogen are shown in
Figure 16.1. Besides angiotensin-converting enzyme, sev-
eral other angiotensin I and II processing enzymes exist.
They produce the heptapeptides angiotensin^ 1-7) from an-
giotensin I and angiotensin III from angiotensin II. The activi-
ties of the major angiotensinogen metabolites are included in
Table 16.1. The octapeptide, angiotensin II, has numerous
effects in many organ systems. As with other endothelial
hormones, angiotensin II acts through specific cell surface
receptors to produce its effects. 13 The carboxy terminus,
phenylalanine, is thought to mediate binding to the receptor
associated with vasoconstriction. Thus far, two angiotensin
receptor subtypes have been identified. The angiotensin
type 1 receptor is the best characterized. It is given credit for
mediating the vasoconstricting properties of angiotensin
II. This receptor activates a G-protein/phospholipase
apparatus to increase intracellular inositol triphosphate and
diacylglycerol. The resulting increase in intracellular calcium
triggers the contractile mechanisms in vascular smooth
muscle and mesangial cells. Systemic blockade of this receptor
results in a pharmacologic pattern similar to angiotensin-
converting enzyme inhibition. 14 ' 15 Of note, the specific
angiotensin type 1 receptor blocker Losartan (DuP 753) is
undergoing clinical trials for treatment of hypertension in
1993. Though the angiotensin type 2 receptor is incompletely
characterized, it is known to mediate physiologic activities in
several organs. The activities associated with type 2 receptors

182

chapter 16 Pathophysiology of renovascular hypertension

include natriuresis and vasodilation. The specific blocking
agents for this receptor are known as CGP 42 112 A and
PD1231 77. The existence of other receptors is highly likely
from studies of specific angiotensin type 1 and angiotensin
type 2 receptor antagonists. 16

Renal function and angiotensin peptides

Angiotensin peptides have direct local influences on renal
function. In addition to the arteriolar and glomerular effects
of angiotensin peptides, they directly influence tubular
reabsorption (see Table 16.1). Mitchell demonstrated that the
influence of angiotensin II on proximal tubular sodium
reabsorption is dose-dependent in studies using micro-
puncture techniques. The specific ability of angiotensin II to
augment sodium reabsorption by the proximal tubule has
been localized to the luminal sodium /hydrogen exchange
pump. 17 However, the influence of angiotensin II on basolat-
eral sodium/bicarbonate cotransport mechanisms remains

Table 16.1 Renal actions of angiotensin peptides

Specific renal response

Angiotensinogen Receptor subtype
metabolite mediating effect

Proximal tubular ion transport Ang I*, Ang II,

andAng-(1-7)

Afferent /efferent arteriolar Ang II

constriction

AT1/AT2

AT1

Inhibition of renin release

Increased mesangial tone

Ang II
Ang II

AT1
AT1

*Most likely, angiotensin I is metabolized locally to form angiotensin II.
Locally produced angiotensin II then directly influencestubular function.

unresolved. Though angiotensin I has been associated with
increased sodium reabsorption in the proximal tubule, it
is more likely that angiotensin II is being generated by
local angiotensin-converting enzymes. The finding that
angiotensin-converting enzyme inhibitors diminish the
sodium reabsorption associated with these angiotensin
peptides lends support to this hypothesis. 13

As in other organ systems, intrarenal angiotensin activity is
dependent on receptor-mediated mechanisms that probably
vary between species. The locations of various angiotensin
peptides and receptor subtypes in normal and chronically
ischemic human kidneys have recently been described by Diz
and coworkers. 18 Using receptor autoradiography, the densi-
ties of the two known receptors have been characterized in the
renal circulation and the nephron. These findings point to a
significant intrarenal influence of angiotensin peptides. In the
normal kidney, angiotensin activity is highest in the large
preglomerular arteries. Activity sequentially decreases in the
glomeruli and tubulointerstitial areas of the cortex (Fig. 16.2).
Regarding specific angiotensin receptor subtypes, type 1
receptors predominate in the glomerulus and the efferent
arteriole. Mesangial cells also have mostly type 1 receptor
populations. Type 2 receptors, on the other hand, predominate
in the large preglomerular arteries. In the tubulointerstitial
areas of the renal cortex, the mixtures of type 1 and 2 receptors
were found to be 60% and 40%, respectively. The functional
impact of this configuration relates to the ability of angiotensin
peptides to regulate vascular tone, mesangial tone, the filtra-
tion fraction, and tubular function. While angiotensin II is able
to interact with both receptor subtypes, angiotensin^ 1-7) in-
teracts only with type 2 receptors. Activation of type 2 recep-
tors is thought to decrease the tone of the larger preglomerular
vessels, thus enhancing blood flow to the afferent arterioles.
On the other hand, stimulation of type 1 receptors in the
glomeruli and postglomerular vessels by angiotensin II is as-

Figure 16.2 Angiotensin receptor subtypes in
the nonischemic human kidney. This diagram
illustrates the relative predominance of different
angiotensin receptor populations in the renal
circulation and proximal tubule. While
angiotensin II competes at both receptor
subtypes, angiotensin-(1 -7) only competes at
type 2 receptors. Type 1 receptors are associated
with vasoconstriction and increased tubular
reabsorption of sodium. Type 2 receptors are
associated with vasodilation and natriuresis.
(From Diz D, Goldfarb DA, Jaiswal N etal.
Parallel changes in receptors sensitive to
angiotensin-(1 -7) and AT2 antagonists in human
kidneys with renal artery disease. Hypertension
1 993;2 1 :528; and Goldfarb DA, Diz Dl, Tubbs
RR, Ferrario CM, Novick AC. Characterization of
angiotensin subtypes in human kidney and renal
carcinoma [in press].)

Afferent
arteriole

Intralobular
artery

Large Preglomerular

arteries: site of

highest density of

angiotensin II binding

Proximal

convoluted

tubule

Angiotensin

type 2 receptors

predominate

Bowman's
capsule

Receptor subtypes

are mixed:

60% type 1

40% type 2

Efferent
arteriole

Angiotensin

type 1 receptors

predominate

Angiotensin

type 1 receptors

predominate

183

pa rt 1 Vascular pathology and physiology

sociated with vasoconstriction. Vasoconstriction is greatest in
the postglomerular arterioles. The combination of increased
flow through the larger preglomerular arteries and increased
resistance in the postglomerular arterioles produces higher
glomerular hydrostatic pressure and leads to an increase in the
filtration fraction. Regarding tubular function, type 1 recep-
tors are associated with increased sodium reabsorption while
type 2 receptors augment natriuresis. Angiotensin^ 1-7) may
reduce preglomerular resistance and tubular sodium reab-
sorption as a result of its ability to stimulate local release of
prostacyclin. 16

Interestingly, the densities of these receptor populations
change in patients with critical renovascular occlusive dis-
ease. 18 This may impact on angiotensin peptide influence on
renal hemodynamics and tubular function in this setting.
When comparing normal and ischemic human kidneys, our
laboratory demonstrated that the number of receptors in large
preglomerular and glomerular vessels decreased by 50% in is-
chemic kidneys, while the number of receptors in the tubules
increased by 65%. The percentage of type 1 and type 2 receptors
in these areas was also altered by renovascular hypertension.
In particular, the percentage of type 2 receptors decreased in
the smooth muscle of the main renal artery while it increased in
large preglomerular arteries and the glomeruli. The functional
significance of these changes remains to be proven.

Mechanisms of renal autoregulation

The kidneys have two major mechanisms of autoregulation:
tubuloglomerular feedback and a myogenic vascular re-
sponse. External regulatory control occurs through multiple
vasoactive and natriuretic substances and the sympathetic
nervous system. Data from several different animal models
suggest these regulatory mechanisms have various complex
interactions, 19-21 which occur at many levels including the
vascular smooth muscle, glomerular mesangium, tubular sys-
tem, and vascular endothelium. Importantly, these regulatory
mechanisms are of primary importance in the development
and maintenance of renovascular hypertension.

Myogenic mechanism and angiotensin peptides

The renal microcirculation has a unique design that controls
blood flow into the glomeruli, the degree of plasma ultrafiltra-
tion, and the degree of reabsorption from the tubules. 22-25
Glomerular blood flow is tightly regulated between 80 mmHg
and 180 mmHg. Critical components of the microcirculatory
design include the afferent arterioles, the glomeruli, the effer-
ent arterioles, the peritubular capillaries and vasa recta, and
the vascular endothelium. The pattern of resistor(aa) — » capil-
lary bed — > resistor(ea) — > capillary bed in close approximation
to the tubular apparatus helps account for the unique auto-
regulatory abilities of the kidneys and allows for multiple

regulation patterns. 26 The baseline tone of the resistors is al-
tered in response to variations in blood pressure, neural activ-
ity, and vasoactive hormones. In such a scheme, baroreceptors
in the renal arteries alter sympathetic output to the renal circu-
lation and tubules. This directly affects RBF, glomerular filtra-
tion rate (GFR), and tubular reabsorption. The two major
opposing vasoactive hormones thought to be responsible for
myogenic autoregulation are nitric oxide (NO) and an-
giotensin II. 27-33 Endothelial-derived relaxing factors, particu-
larly NO, have a tonic relaxing effect on the afferent and
efferent arterioles and are associated with increased natriure-
sis. According to Romero et ah, NO production and release
from endothelial cells throughout the renal circulation in-
crease with perfusion pressure between 80 mmHg and 180
mmHg. Prostacyclin production, on the other hand, increases
when perfusion pressure drops below these levels of renal au-
toregulation. 27 This configuration augments diuresis during
times of high renal perfusion pressure while maintaining
tubular oxygen delivery and excretion of systemic waste when
renal perfusion is low.

The precision of this control mechanism can be appreciated
by reviewing the effects of isolated resistor changes. 22 Multi-
ple combinations of resistor tones occur in various physio-
logic conditions, resulting in precision autoregulation. In
the face of critical renal artery stenosis, angiotensin II directly
increases the tone of the afferent and efferent arterioles. Inter-
estingly, the tone of the efferent arteriole increases more in this
setting. This augments glomerular hydrostatic pressures and,
thus, glomerular filtration. This increase in the filtration frac-
tion helps maintain delivery of systemic waste products to the
tubules in the face of low renal perfusion pressures. While the
filtration fraction increases in this setting, medullary blood
flow and interstitial pressure decrease. This configuration
enhances the countercurrent multiplication mechanism and
tubular reabsorption of sodium and free water. Since renal
artery stenosis resembles the hemodynamic characteristics of
hypovolemia distal to the stenosis, the kidney interprets
the condition of renal artery stenosis as hypovolemia and
augments those actions that increase intravascular volume.

Regulation of regional blood flow in the kidney also impacts
on reabsorption of the tubular fluid. Under normal circum-
stances, more than 90% of RBF is distributed to the renal cor-
tex. The remainder of blood flow is distributed to the outer and
inner medulla. The gradation of flow in these areas is such
that the outer medulla receives the bulk of medullary blood
flow and the inner tissues receive progressively less flow.
Medullary blood flow dynamics are influenced by the small-
caliber vasa recta, increased viscosity of medullary blood,
hormonal activity of the vasa recta endothelium, and renal
sympathetic nerve activity. Finally, the pattern of regional RBF
has a direct impact on the countercurrent multiplier mecha-
nism of the renal medulla. In renovascular hypertension,
angiotensin II diminishes blood flow to the medulla by
increasing efferent arteriolar tone.

184

chapter 16 Pathophysiology of renovascular hypertension

j. d* c.

Figure 16.3 The juxtaglomerular apparatus and tubuloglomerular
feedback. The open section of the distal thick ascending loop of Henle (a)
reveals the cells of the macula densa(b), while similar sections of the afferent
(d) and efferent (e) arterioles reveal the renin-producing juxtaglomerular cells
(c). The macula densa estimates the tubular flow rate from distal NaCI
concentrations. When NaCI is low [decreased renal blood flow (RBF) and
glomerular filtration rate (GFR)], afferent and efferent arteriole resistance

decreases while renin secretion increases. When NaCI is high (increased RBF
and GFR), arteriole resistance increases while renin secretion decreases. The
afferent arteriole is the primary site of autoregulation. Arteriole constriction is
probably facilitated by adenosine byproducts from macula densa
sodium/potassium/chloride pump metabolism, f, glomerular capillary; g,
mesangial cell; h, Bowman's space; i, proximal convoluted tubule; j, renal
sympathetic nerves.

Tubuloglomerular feedback and
angiotensin peptides

In this mechanism of autoregulation, the contents of tubular
fluid influence glomerular blood flow. The unique anatomy of
the glomerulus and its associated tubule allows for this feed-
back mechanism (Fig. 16.3). The distal end of the thick ascend-
ing loop of Henle lies adjacent to its glomerulus of origin. This
anatomic entity is known as the juxtaglomerular apparatus. It in-
cludes the macula densa of the thick ascending loop of Henle,
the afferent and efferent arterioles, and the mesangial cells.
The cells of the macula densa are able to detect the concentra-
tion of NaCI of the local tubular fluid. Na + /K + /2C1~ cotrans-
port pumps, long known to exist in the thick ascending loop of
Henle, have been identified in the macula densa and probably
play an important role in the assessment of NaCI levels. 19,20
The macula densa accurately translates the tubular flow rate
from the local tubular salt concentration and alters afferent
and efferent arteriole resistance as needed to maintain normal
distal tubular salt delivery. Though both arterioles are affected
by this mechanism, the afferent arteriole is the primary site of

autoregulation. The effect on afferent and efferent arteriole
resistance is accomplished via local intercellular mecha-
nisms and systemic release of renin. Locally active renin-
angiotensin systems may also influence this aspect of auto-
regulation. 16 ' 18 Adenosine byproducts from Na + /K + /2C1~
cotransport pump ATP metabolism are thought to drive the
vasoconstriction of the arterioles. 19 This mechanism may also
alter GFR through direct effects on the mesangial cell tone and,
thus, the filtration surface area.

Under normal conditions, solute reabsorption from the
more proximal nephron produces tubular fluid that has a
lower NaCI concentration relative to plasma by the time it
arrives at the distal thick ascending loop of Henle. During
periods of low GFR, as in hemorrhagic shock or critical renal
artery stenosis, the decreased tubular flow results in maximal
reabsorption of solute and, hence, fluid with low NaCI concen-
trations. The macula densa senses the low NaCI concentra-
tions, decreases afferent arteriolar resistance, and increases
renin production and secretion. The exact mechanism of
increased renin production is unresolved. Renin causes the
production of angiotensin II. Angiotensin II increases systemic

185

pa rt 1 Vascular pathology and physiology

blood pressure through various effects on several organ sys-
tems, including the cardiovascular, central and autonomic
nervous, adrenal, and renal systems. Under otherwise normal
circumstances, these systemic actions restore renal plasma
flow, GFR, and distal NaCl delivery to normal levels. During
increased tubular flow and distal NaCl delivery, the macula
densa increases afferent arteriole resistance to decrease renal
plasma flow and GFR. Usually, this mechanism allows for fine
control of renal plasma flow and GFR. However, in pathologic
situations such as critical renal artery stenosis, the change
in afferent arteriolar resistance does not adequately correct
tubular flow and distal NaCl delivery. This results in chronic
activation of the renin-angiotensin system and leads to reno-
vascular hypertension and may activate local angiotensin
peptide production. 6

Angiotensin II also has direct local renal effects. In addition
to its effect on sodium reabsorption by the proximal tubule,
angiotensin II has been found to decrease the sensitivity of the
afferent and efferent arterioles to tubuloglomerular feedback.
In other words, tone in these vessels remains high, despite the
influence of the macula densa. This acts in tandem with its
other effects to enhance tubular sodium reabsorption while
maintaining renin release by the juxtaglomerular cells. Thus,
a locally active renal renin-angiotensin system may exacer-
bate the increased volume status and hypertension associated
with critical renovascular occlusive disease. This becomes im-
portant clinically when one considers the deleterious impact
of angiotensin-converting enzyme inhibitors in patients with
renal artery stenosis. By blocking local angiotensin II effects on
efferent arteriolar tone, angiotensin-converting enzyme in-
hibitors lower glomerular hydrostatic pressures. This causes a
reduction in the filtration fraction and diminishes the ability of
the kidneys to excrete systemic waste. The clinical result is an
increase in plasma creatinine, urea nitrogen levels, and renal
insufficiency. This deleterious intrarenal effect on GFR is prob-
ably more significant than the decrease in systemic arterial
pressure, also induced by these inhibitors. 18

Vascular endothelial substances

Increasing attention has been given to the effects of vascular
endothelial substances on the renal microcirculation and
autoregulation. These substances interact with myogenic and
tubuloglomerular feedback mechanisms. Changes in vascular
shear stress induce production and release of different vasoac-
tive substances by the endothelial cells. In general, hormones
induce vasodilation or constriction. NO and endogenous
angiotensin II are considered the most likely determinants of
myogenic renal autoregulation. 27-30 Importantly, these are not
the only vasoactive hormones that influence normal and
pathologic renal function. In addition to their direct vascular
effects, these compounds have regulatory effects on endothe-
lial production and release of prostaglandins, thromboxanes,
and other vasoactive hormones. The other recognized vasoac-

tive hormones influencing renal hemodynamics and function
are endothelin, thromboxane A 2 , prostaglandins, platelet
activating factor, and atrial natriuretic factor. All of these have
profound influences on the renal microcirculation and
mesangium (Table 16.2). Their opposing actions influence the
tone of the renal circulation and, thus, renal plasma flow, GFR,
and natriuresis. 34 Furthermore, their production, secretion,
and activity are directly influenced by the renin-angiotensin
system. In the setting of renovascular occlusive disease, sev-
eral physiologic factors work together to enhance renal hemo-
dynamics and filtration despite low perfusion pressures.
Specifically, the local intrarenal levels of angiotensin II,
endothelin 1, thromboxane A 2 , prostaglandin I 2 , and
prostaglandin E 2 are elevated in ischemic kidneys. 16,18,35-38
Angiotensin II and endothelin 1 increase efferent arteriolar re-
sistance, mesangial cell tone, and reabsorption of sodium by
the tubules. 39-41 Thromboxane A 2 increases resistance of the
afferent arteriole and probably has direct effects on sodium
reabsorption as well. 36/42/43 Though prostacyclin does not exert
a tonic influence on the vasculature in normal circumstances, it
maintains RBF during states of low renal perfusion pres-
sure. 27,34,44 ' 45 This is in contradistinction to the vasodilating
activity of NO, which has basal effects on autoregulation
and increases activity during states of high renal perfusion
pressure. This eicosanoid, like NO, selectively dilates the
afferent arteriole. 28,29,32 Recently, our laboratory has shown
that angiotensin II and angiotensin^ 1-7) directly increase
prostaglandin release from human renal arteries. During low
flow states, such as main renal artery occlusion, it is likely that
angiotensin peptides increase the release of prostacyclin. In
the presence of selective efferent arteriolar constriction from
angiotensin II and endothelin I, the resulting resistor pattern
attempts to maintain GFR by increasing the filtration fraction.
Finally, angiotensin II stimulates production and release of
atrial natriuretic factor by the heart. The balance of these
compounds determines the overall effect on renal function.
Thus, their activities are an important consideration in the
pathophysiology of renovascular hypertension. 46

Renal nerves and function regulation

The autonomic nervous system plays an important role in
modulation of renal function in states of normotension and
hypertension. The sympathetic nervous system innervates
both the renal microvascular and tubular structures. Major
vascular structures directly innervated by adrenergic nerves
include the afferent and efferent arterioles. Tubular structures
so influenced, in descending magnitude of innervation, in-
clude the thick ascending loop of Henle, the juxtaglomerular
apparatus, the distal convoluted tubules, and the proximal
tubules. Renal sympathetic nerves regulate renal sodium
excretion, RBF, GFR, and renin secretion rate through a x -
adrenergic receptors in the vessels and tubules and a 1 -
receptors in the juxtaglomerular apparatus. 47 Renal sympa-

186

chapter 16 Pathophysiology of renovascular hypertension

Table 16.2 Vasoactive hormones that interact with the angiotensin peptides and influence renal hemodynamics and tubular function

Vasoactive hormone Mechanism of action

Effect on
arteriole tone

Effect on
mesangial tone

Medullary blood flow Natriuretic effect

Angiotensin II
(Ang II)

Endothelin 1 (ET1)

Thromboxane A-
(TXAJ

Nitric oxide (NO)

Prostacyclin/
prostaglandin E 2
(PGI 2 /PGE 2 )

Atrial natriuretic
factor (AN F)

AT1 receptor/phospholipase/
ITP and diacylglycerol -» ft
cellular Ca 2+

ET 1 receptor/phospholipase/
ITP and diacylglycerol -» ft
cellular Ca 2+

TXA 2 receptor/phospholipase/
ITP and diacylglycerol -» ft
cellular Ca 2+

Diffuses through cell membrane
then IT cyclic GMP

PGI 2 receptor/then ft cyclic

AMP

(PGI 2 )

ANF receptor/then IT cyclic
GMP

ttttAAtone

1t1l"fT EAtone

IT Filtration percent

ft AA tone

TTITTr EAtone

ft Filtration percent

fTftAAtone
IT EAtone

MAAtone
W EAtone

MAAtone
U- EAtone
(PGI 2 )

TiAAtone
li EAtone

tfbne/UGFSA

ftTone/^GFSA

fTTone/UGFSA

illbne/ftGFSA

UTone/flGFSA

Decreased

Increased

Increased
(PGI 2 )

Increased

Decreased

Increased

Increased
(PGE 2 )

Increased

AA, afferent arteriole; EA, efferent arteriole; GFSA, glomerular filtration surface area; ITP, inositol triphosphate; GMP, guanosine monophosphate; AMP,
adenosine monophosphate.

thetic nerve activity varies with afferent stimulation from
renal vascular baroreceptors, ureteral mechanoreceptors, and
pelvic chemoreceptors. Additionally, renal sympathetic tone
is influenced by the degree of sympathetic response to stress-
ful stimuli. 8 Interestingly, the input from one kidney modu-
lates the function of the other and is termed the renorenal reflex.
Afferent input from ureteral mechanoreceptors is increased
with obstruction and results in decreased ipsilateral and
increased contralateral natriuresis. Increased sodium concen-
tration detected by renal pelvic chemoreceptors results in a
similar situation. According to Dibona, graded frequency of
efferent nerve stimulation results in variable effects on renal
function. 47 With increasing stimulation, renal function para-
meters are affected in the following order: (i) renin secretion
rate is increased; (ii) sodium and water reabsorption increases;
and (iii) GFR and RBF are decreased. Baroreceptor stimuli in-
fluence the sensitivity of juxtaglomerular apparatus renin pro-
duction. With increased baroreceptor stimulation during
states of hypotension, renin secretion is augmented. The oppo-
site occurs during states of high renal artery pressure and low
baroreceptor stimulation.

It is likely that renal nerve activity contributes to renovascu-
lar hypertension. Studies by Kopp and Buckley-Bleiler have
shown that the renorenal reflex is directly affected in two-
kidney /one-clip renovascular hypertension. 48 In this setting,

efferent renal nerve activity is increased. Denervation of the
undipped kidney results in an ipsilateral increase in sodium
excretion and GFR while clipped kidney denervation results
in bilateral natriuresis and increased GFR. 8 Additionally,
selective afferent denervation of the clipped kidney results
in decreased hypothalamic norepinephrine stores, decreased
peripheral sympathetic nerve activity, and reduced arterial
pressure. With regard to renovascular hypertension, an-
giotensin II directly increases central and renal sympathetic
nerve activity.

Tissue renin-angiotensin system

In 1827, Bright observed the association between renal disease
and cardiac hypertrophy. Multidisciplinary investigations of
angiotensin peptides have led to a new appreciation of the
scope of their systemic effects, and the identification of locally
integrated tissue renin-angiotensin systems. 6 The activities of
these local systems are listed in Table 16.3.

Angiotensin peptides, especially angiotensin II, influence
the metabolic activities of endothelial and vascular smooth
muscle cells. 49-53 Acutely, stimulation of angiotensin type 1 re-
ceptors effects an increase in intracellular calcium and vaso-
constriction. Over increased periods of exposure, angiotensin

187

pa rt 1 Vascular pathology and physiology

Table 16.3 Actions of angiotensin peptides and receptor subtypes in major organ systems

Organ system involved

Specific tissue response

Angiotensinogen

Receptor subtype

metabolite

mediating effect

Ang Hand Ang III

AT1

Angll

AT1

Angll

AT1/AT2

Angll

AT1

Angll

AT2

Ang-(1-7)

Undetermined

Angll

AT1

Angll

AT1

Angll

Undetermined

Ang-(1-7)

Undetermined

Ang-(1-7),Angll, and

AT1

Anglll

AT1

Ang Hand Ang III

AT1/AT2

Ang-(1-7),Angll, and

Anglll

Ang Hand Ang III

AT1

Ang Hand Ang III

AT1

Angll

AT1

Blood vessels and vascular smooth
muscle cells

Myocardium

Central nervous system

Sympathetic actions
Adrenal cortex

Medulla

Vasoconstriction

Mitogenesis/hypertrophy

Angiogenesis

Intimal hyperplasia

Endothelial prostaglandin release

Vasodilation

Positive inotrope

Hypertrophy

Stimulation of atrial natriuretic release

Coronary vasoconstriction

Arginine vasopressor (ADH) release

Thirst/drinking
Baroreceptor

Potentiation of norepinephrine release from nerve terminal

Aldosterone release
Corticotropin release
Catecholamine release

II induces hypertrophy of the smooth muscle cells. 54 ' 55 Addi-
tionally, angiotensin II increases smooth muscle proliferation
by inducing production of transforming growth factor 2 and
platelet-derived growth factors. 56 ' 57 Recent investigations
have found that not all angiotensin peptides induce the above
changes in vascular smooth muscle cells or systemic vascular
resistance. In particular, angiotensin-(l-7) is associated with
vasodilation. 58 ' 59

Regarding the heart, it has been shown that angiotensin
peptides have positive inotropic activity, stimulate release of
atrial natriuretic factor, and induce myocyte hypertro-
phy. 49 ' 60-62 After the introduction of angiotensin-converting
enzyme inhibitors to clinical practice, physicians observed a
dramatic effectiveness in the management of heart failure.
Subsequent laboratory investigation found that the myocar-
dium is also influenced by a local, active renin-angiotensin
system. 49 Angiotensin peptides exert positive inotropic and
chronotropic influences on the myocardium, increase coro-
nary artery tone, and stimulate the release of atrial natriuretic
factor. When coupled with its influence on adrenomedullary
secretion of catecholamines and cardiac sympathetic and
vagal nerve activity, the potential of the renin-angiotensin sys-
tem to modulate inotropic and chronotropic activity in the
healthy heart can be appreciated. In the setting of congestive
heart failure and myocardial ischemia, angiotensin peptides
worsen coronary blood flow and diminish myocardial func-
tion. By increasing afterload and coronary artery tone, these

peptides increase myocardial demand and wall tension while
diminishing diastolic perfusion.

Angiotensin-converting enzyme successfully blocks
these effects in the diseased myocardium while decreasing
preload and afterload. 63 ' 64 In addition, angiotensin-
converting enzyme inhibition results in increased produc-
tion of angiotensin^ 1-7) and decreased metabolism of
bradykinin. Angiotensin-converting enzyme, also referred to
as kininase II, is also responsible for the inactivation of
bradykinin. When this enzyme is blocked, the vasodilator
activity of bradykinin is maintained while the activity of
angiotensin^ 1-7) is increased. These hormones stimulate
the production and release of the vasodilators, NO and
prostaglandins. The cumulative effect is considerable. Thus,
angiotensin-converting enzyme may affect systemic blood
pressure by altering the balance of opposing vasoactive
hormone systems. 65

Angiotensin influence on central nervous system activity
complements this gathering systemic effect by increasing
sympathetic output, thirst and drinking behavior, and antidi-
uretic hormone release, and altering baroreceptor reflex activ-
ity 66 ' 67 By directly increasing norepinephrine secretion from
nerve terminals and simultaneously inhibiting its reuptake
from the synapse, angiotensin II increases peripheral sympa-
thetic nerve activity 66 ' 68 ' 69 Centrally, angiotensin peptides
influence the activity of the baroreceptor reflex, vagal tone,
sympathetic tone, and hormone release from the pituitary

188

chapter 16 Pathophysiology of renovascular hypertension

gland. Regarding its effect on baroreceptor activity, an-
giotensin II prevents the reflex bradycardia usually associated
with systemic hypertension. 66 The specific site in the medulla
where this occurs is one of the major centers for vagal efferent
activity, the nucleus of the tractus solitarius. It appears
that plasma angiotensin peptides are able to infiltrate the
blood-brain barrier at a site directly dorsal to the nucleus of
the tractus solitarius called the area postrema. These peptides
reduce vagal efferent activity from the nucleus of the tractus
solitarius that would otherwise decrease the heart rate in the
face of systemic arterial hypertension. 72 Since systemic pres-
sure is a consequence of cardiac stroke volume, heart rate, and
systemic vascular resistance, the activity of angiotensin pep-
tides on the baroreceptor reflex supports its cardiovascular
effects to increase systemic arterial blood pressure. Finally,
angiotensin peptides act in the hypothalamus to increase
thirst-related behavior and in the pituitary to increase the re-
lease of antidiuretic hormone (vasopressin) and adrenocorti-
cotropic hormone. 68 All of these activities can be diminished or
blocked by angiotensin-converting enzyme inhibitors or spe-
cific angiotensin receptor blockers. Because of their ability to
modulate these central and peripheral neural activities, an-
giotensin peptides are also considered neurotransmitters. 68,69

In the adrenal glands, angiotensin peptides enhance hor-
mone production and secretion in both cortical and medullary
tissues. In the medulla, both epinephrine and norepinephrine
production and release are increased by angiotensin II. 73 An-
giotensin II accomplishes this action directly by stimulation of
the medullary cells and indirectly by increasing central ner-
vous system sympathetic output. The resulting increase in
plasma catecholamines enhances the vasoconstricting and
inotropic effects of angiotensin II. In the adrenal cortex,
angiotensin II and III are equally able to increase production of
aldosterone by cells in the zona glomerulosa. By increasing
adrenocorticotropic hormone secretion by the anterior pitu-
itary, angiotensin peptides indirectly augment aldosterone
and Cortisol production. Aldosterone enhances sodium reab-
sorption and potassium excretion by the distal convoluted
tubules. Thus, the increase in adrenal hormones acts in concert
with intrarenal angiotensin peptide activity to diminish RBF
and glomerular filtration while maximizing tubular reabsorp-
tion of sodium. Nishimura and colleagues have shown the
adrenals also have local integrated renin-angiotensin systems
that are able to produce angiotensin peptides. 6 Though local
production has not been found to alter systemic levels, this
local adrenal renin-angiotensin system does have the ability
to act in a paracrine fashion to influence adrenal cortical and
medullary activity.

Therefore, the renin-angiotensin system can induce severe
hypertension by amplifying multiple determinants of sys-
temic blood pressure. In addition to the effects of systemic an-
giotensin peptides on these organ beds, each of these tissues
has locally integrated renin-angiotensin systems. The obser-
vation that these individual tissue systems can function in

isolation underlies the hypothesis that they may be respon-
sible for the chronic or maintenance phase of renovascular
hypertension. 6 ' 8

Conclusion

Our understanding of the events responsible for the develop-
ment and maintenance of renovascular hypertension has in-
creased dramatically over the past 60 years. Despite this, the
renin-angiotensin system remains incompletely understood.
Though once seen as a purely systemic hormonal mechanism
responsible for hypertension, the presence and significant
activities of renal, cardiovascular, central nervous, and adrenal
renin-angiotensin systems are widely recognized. In addition
to being responsible for many of the pathophysiologic changes
observed in renovascular hypertension, these local systems
probably have baseline tonic influence on the functions of
these organ systems. Research in this area has yielded potent
antihypertensive agents, such as angiotensin-converting
enzyme inhibitors and angiotensin receptor blockers. These
agents have dramatically improved our ability to manage
disease states such as essential hypertension and congestive
heart failure. No similar medical therapy has been effec-
tive at treating renovascular hypertension. On the contrary,
angiotensin-converting enzyme inhibitors have actually been
shown to increase the development of renal insufficiency in pa-
tients with critical renal artery stenosis. Importantly, research
findings have improved our ability to determine which pa-
tients will benefit from revascularization or removal of chroni-
cally ischemic kidneys. Examples of these diagnostic advances
include assessment of renal vein renin levels and captopril
renography. Centers experienced in the management of reno-
vascular hypertension have reported significant improvement
or total cure of hypertension in 90-98% of patients. 2-5

The hypertension and renal function responses to interven-
tion appear dichotomous. Improved hypertension does not
ensure improved renal function. In fact, nearly one-third of pa-
tients with disease caused by atherosclerosis continue to have
some measure of renal dysfunction after intervention. 4 ' 5 ' 74-77
Though the reasons for this are unknown, the function of
vascular endothelial cells in the diseased kidney may be
important. Disturbances in such opposing local vasoactive
hormones as angiotensin II and NO may impact on renal resis-
tance and tubular function. 39 Additionally, the ability of the
kidney to produce and systemically release such recently
identified compounds as platelet activating factor and medul-
lipin may influence both renal function and the reversal of
systemic hypertension after reperfusion. 78-80 Moreover, a
reperfusion injury may impact on all of these variables. 81 ' 82 It is
hoped that ongoing research in these areas will improve
our ability to cure patients with renovascular hypertension.
In the future, discoveries relating to organ-specific renin-
angiotensin systems will hopefully improve our understand-

189

pa rt 1 Vascular pathology and physiology

ing of and ability treat other cardiovascular and neurologic
disease processes.

References

1. Peart WA. Evolution of renin. Hypertension l99l:l8(Supp\.III):lOO.

2. Dean RH, Hansen KJ. Renovascular disease. In: Moore WS, ed.
Vascular Surgery: A Comprehensive Review, 6th edn. Philadelphia:
WB Saunders Co., 2002:548.

3. Benjamin ME, Hansen KJ, Craven TE, Keith DR, Plonk GW, Geary
RL, Dean RH. Combined aortic and renal artery surgery: a con-
temporary experience. Ann Surg 1996; 233:555.

4. Hansen KJ. Renovascular hypertension: an overview. In:
Rutherford RB, ed. Vascular Surgery, 5th edn. Philadelphia: WB
Saunders Co., 2000:1593.

5. Dean RH, Benjamin ME, Hansen KJ. Surgical management of
renovascular hypertension. In: Wells SA Jr., ed. Current Problems in
Surgery. St Louis: Mosby Year Book, Inc., 1997; 34:209.

6. Nishimura M, Milsted A, Block CH, Brosnihan KB, Ferrario CM.
Tissue renin-angiotensin systems in renal hypertension. Hyper-
tension 1992; 20:158.

7. Masaki Z, Ferrario CM, Bumpus FM. Effects of SQ 20881 on the
intact kidney of dogs with two-kidney, one clip hypertension.
Hypertension 1980; 2:649.

8. Martinez-Maldonado M. Pathophysiology of renovascular hy-
pertension. Hypertension 1991; 17:708.

9. Masaki Z, Ferrario CM, Bumpus FM, Bravo EL, Khosla MC. The
course of arterial pressure and the effect of Sar 1 -Thr 8 -angiotensin
II in a new model of two-kidney hypertension in conscious dogs.
Clin SciMolMedl977;52:l.

10. Ganten D, Takahashi S, Lindpainter K, Mullins J. Genetic basis of
hypertension: the renin-angiotensin paradigm. Hypertension
1991;18(Suppl.III):109.

11. Jeunemaite X, Soubrier F, Kotelevstev YV et ah Molecular basis of
human hypertension: role of angiotensinogen. Cell 1992; 71:169.

12. Gomez RA, Chevalier RL, Carey RM, Peach MJ. Molecular biol-
ogy of the renal renin-angiotensin system. Kidney Int 1990;
38(Suppl.30):S18.

13. Modrall R, Hagaki J, Okuniashi H et ah Changes in gene expres-
sion of the renin-angiotensin system in two-kidney, one-clip
hypertensive rats. / Hypertens 1991; 9:187.

14. Timmermans PBMWM, Benfield P, Chiu AT, Herblin WF, Wong
PC, Smith RD. Angiotensin II receptors and functional correlates.
Am J Hypertens 1992;5:221S.

15. Timmermans PBMWM, Carini DJ, Chiu AT et ah Angiotensin II
receptor antagonists: from discovery to antihypertensive drugs.
Hypertension 1991; 18(Suppl. III):136.

16. Wang ZQ, Millat LJ, Heiderstadt NT, Siragy HM, Johns RA, Carey
RM. Differential regulation of renal angiotensin subtype AT1A
and AT2 receptor protein in rats with angiotensin-dependent
hypertension. Hypertension 1999; 33:96.

17. Saccomani G, Mitchell KD, Navar LG. Angiotensin II stimulation
of Na + H + exchange in proximal tubular cells. Am J Physiol 1990;
258:F1188.

18. Diz DI, Goldfard DA, Jaiswal N et ah Parallel changes in receptors
sensitive to angiotensin-(l-7) and AT2 antagonists in human kid-
neys with renal artery disease. Hypertension 1993; 21 :528.

19. Holstein-Rathlou NH. Dynamic aspects of the tubuloglomerular
feedback mechanism. Dan Med Bull 1992; 39:134.

20. Briggs JP, Schnermann J. The tubuloglomerular feedback mecha-
nism: functional and biochemical aspects. Annu Rev Physiol 1987;
49:251.

21. Aukland K, Oien AH. Renal autoregulation: models combining
tubuloglomerular feedback and myogenic response. Am J Physiol
1987;252:F768.

22. Valtin H. Renal Function: Mechanisms Preserving Fluid and Solute
Balance in Health, 2nd edn. Boston: Little, Brown, 1983.

23. Berne RM, Levy MN. Physiology, 2nd edn. St Louis: CV Mosby,
1988:745.

24. Guyton AC. Kidneys and fluids in pressure regulation: small vol-
ume but large pressure changes. Hypertension 1992; 19(Suppl. I):2.

25. Granger JP. Pressure natriuresis: role of interstitial hydrostatic
pressure. Hypertension 1992; 19(Suppl. I):9.

26. Beeuwkes R III, Brenner BM. The renal circulation. In: Brenner
BM, Rector FC Jr, eds. The Kidney, 2nd edn. Philadelphia: WB
Saunders, 1981.

27. Romero JC, Lahera V, Salom MG, Biondi ML. Role of the
endothelium-dependent relaxing factor nitric oxide on renal
function. J Am Soc Nephrol 1992; 2:13717.

28. Sigmon DH, Carretaro OA, Beierwaltes WH. Plasma renin activi-
ty and the renal response to nitric oxide synthesis inhibition. } Am
Soc Nephrol 1992; 3:12884.

29. Sigmon DH, Carretero OA, Beierwaltes WH. Angiotensin depen-
dence of endothelium-mediated renal hemodynamics. Hyperten-
sion 1992; 20:643.

30. Salazar FJ, Pinilla JM, Lopez F, Romero JC, Quesada T. Renal
effects of prolonged synthesis inhibition of endothelium-derived
nitric oxide. Hypertension 1992; 20:113.

31. Zatz R, De Nucci G. Effects of acute nitric oxide inhibition on rat
glomerular microcirculation. Am} Physiol 1991; 261:F360.

32. Ito S, Juncos LA, Nushiro N, Johnson CS, Carretero OA.
Endothelium-derived relaxing factor modulates endothelin
action in afferent arterioles. Hypertension 1991; 17:10526.

33. Salom MG, Lahera V, Miranda-Guardiola F, Romero JC. Blockade
of pressure natriuresis induced by inhibition of renal synthesis of
nitric oxide in dogs. Am J Physiol 1992; 262:F718.

34. Henrich WL. The endothelium: a key regulator of vascular tone.
Am J Med Sci 1991; 302:319.

35. Katusic ZS, Shepherd JT. Endothelium-derived vasoactive fac-
tors. II. Endothelium-dependent contraction. Hypertension 1991;
(Suppl.III):86.

36. Ballermann BJ, Marsden PA. Endothelium-derived vasoactive
mediators and renal glomerular function. Clin Invest Med 1991;
14:508.

37. Shepherd JT, Katusic ZS. Endothelium-derived vasoactive fac-
tors. I. Endothelium-dependent relaxation. Hypertension 1991;
(Suppl.III):76.

38. DohiY,LuscherTF. Endothelin in hypertensive resistance arteries:
intraluminal and extraluminal dysfunction. Hypertension 1991;
18:543.

39. Lopez-Farre A, Gomez-Garre D, Bernabeu F, Montane I, Millas I,
Lopez-Novoa JM. Renal effects and mesangial cell contraction
induced by endothelin are mediated by PAR Kidney Int 1991;
39:624.

40. Hirata Y, Matsuoka H, Kimura K, Sugimoto T, Hayakawa H,
Suzuki E, Sugimoto T. Role of endothelium-derived relaxing

190

chapter 16 Pathophysiology of renovascular hypertension

factor in endothelin-induced renal vasoconstriction. / Cardiovasc
Pharmacol 1991; 17(Suppl. 7):169.

41 . Kohno M, Yasunari K, Murakawa KI et ah Plasma immuno reactive
endothelin in essential hypertension. Am J Med 1990; 88:614.

42. Remuzzi G, Fitzgerald GA, Patrino C. Thromboxane synthesis
and action with the kidney. Kidney Int 1992; 41:14833.

43. Anggard EE. The regulatory functions of the endothelium. Jpn J
Pharmacol 1992; 58(Suppl. 2):200.

44. Smith WL. Prostanoid biosynthesis and mechanisms of action. Am
J Physiol 1992; 263:F181.

45. McGiff JC, Carroll MA, Escalante B. Arachidonate metabolites
and kinins in blood pressure regulation. Hypertension 1991;
18(Suppl.III):150.

46. Cooke JP Endothelium-derived factors and peripheral vascular
disease. Cardiovasc Clin 1992; 22:3.

47. Dibona GF. Sympathetic neural control of the kidney in hyperten-
sion. Hypertension 1992; 19(Suppl. I):28.

48. Kopp UC, Buckley-Bleiler RL. Impaired renorenal reflexes in two-
kidney one clip hypertensive rats. Hypertension 1989; 14:445.

49. Lindpainter K, Ganten D. The cardiac renin angiotensin system:
an appraisal of present experimental and clinical evidence. Circ
Res 1991; 68:905.

50. Bumpus FM. Angiotensin I and II: Some early observations made
at the Cleveland Clinic Foundation and recent discoveries relative
to angiotensin II formation in human heart. Hypertension 1991;
18(Suppl.III):122.

51. Lever AF, Lyall F, Morton JJ, Folkow B. Angiotensin II, vascular
structure and blood pressure. Kidney Int 1992; 41(Suppl. 37):51.

52. Gohlke P, Biinning P, Unger T. Distribution and metabolism of an-
giotensin I and II in the blood vessel wall. Hypertension 1992;
20:151.

53. Shiota N, Miyazaki M, Okunishi H. Increase of angiotensin con-
verting enzyme gene expression in the hypertensive aorta. Hyper-
tension 1992; 20:168.

54. Rakugi H, Jacob HJ, Krieger JE, Ingelfinger JR, Pratt RE. Vascular
injury induces angiotensin gene expression in the media and
neointima. Circulation 1993; 87:283.

55. Krug LM, Berk BC. Na + ,K + -adenosine triphosphate regulation in
hypertrophied vascular smooth muscle cells. Hypertension 1992;
20:144.

56. Dzau VJ, Gibbons GH. Endothelium and growth factors in vascu-
lar remodeling of hypertension. Hypertension 1991; 18(Suppl.
III):115.

57. Bohr DF, Dominiczak AF, Webb RC. Pathophysiology of the
vasculature in hypertension. Hypertension 1991; 18(Suppl. III):69.

58. Benter IF, Diz DI, Ferrario CM. Cardiovascular actions of
angiotensin-(l-7). Peptides 1993; 14:679.

59. Jaiswal N, Diz DI, Chappell MC, Khosla MC, Ferrario CM. Stimu-
lation of endothelial cell prostaglandin production by angiotensin
peptides: characterization of receptors. Hypertension 1992;
19(Suppl.II):49.

60. Focaccio A, Volpe M, Ambrosio G et al. Angiotensin II directly
stimulates release of atrial natriuretic factor in isolated rabbit
hearts. Circulation 1993; 87:192.

61. Sarzani R, Arnaldi G, Takasaki I, Brecher P, Chobanian AV. Effects
of hypertension and aging on platelet-derived growth factor and
platelet-derived growth factor receptor expression in rat aorta and
heart. Hypertension 1991; 18(Suppl. III):93.

62. Sunga PS, Rabkin SW. Angiotensin II-induced protein phosphory-
lation in the hypertrophic heart of the Dahl rat. Hypertension 1992;
20:633.

63. Ondetti MA. Angiotensin converting enzyme inhibitors: an
overview. Hypertension 1991; 18(Suppl. Ill): 134.

64. Hirooka Y, Imaizumi T, Masaki H et al. Captopril improves im-
paired endothelium-dependent vasodilation in hypertensive
patients. Hypertension 1992; 20:175.

65. Campbell DJ, Kladis A, Duncan AM. Bradykinin peptides in kid-
ney, blood, and other tissues of the rat. Hypertension 1993; 21:155.

66. Ferrario CM, Barnes KL, Block CH et al. Pathways of angio-
tensin formation and function in the brain. Hypertension 1990;
15(Suppl.I):13.

67. Barnes KL, Diz DI, Ferrario CM. Functional interactions between
angiotensin II and substance P in the dorsal medulla. Hypertension
1991; 17:216.

68. Ferrario CM, Brosnihan KB, Diz DI et al. Angiotensin-(l-7): a new
hormone of the angiotensin system. Hypertension 1991; 18(Suppl.
III):126.

69. Diz DI, Pirro NT. Differential actions of angiotensin II and
angiotensin-(l-7) on transmitter release. Hypertension 1992;
19(Suppl.II):41.

70. Reid IA. Interactions between angiotensin II, sympathetic ner-
vous system, and baroreceptor reflexes in regulation of blood
pressure. Am] Physiol 1992; 262:E763.

71. Kumagai H, Averill DB, Khosla MC, Ferrario CM. Role of nitric
oxide and angiotensin II in the regulation of sympathetic nerve
activity in spontaneously hypertensive rats. Hypertension 1993;
21:476.

72. Ferrario CM, Jaiswal N, Yamamoto K, Diz DI, Schiavone MT.
Hypertensive mechanisms and converting enzyme inhibitors.
Clin Cardiol 1991; 14(Suppl. IV):56.

73. Williams GH, Hollenberg NK. Functional derangements in the
regulation of aldosterone secretion in hypertension. Hypertension
1991;18(Suppl.III):143.

74. Derkx FHM, Schalekamp MA. Renal artery stenosis and hyper-
tension. Lancet 1994; 344:237.

75. Cherr GS, Hansen KJ, Craven TE et al. Surgical management of
atherosclerotic renovascular disease. / Vase Surg 2002; 35:236.

76. Hansen KJ, Cherr GS, Craven TE et al. Management of ischemic
nephropathy: dialysis-free survival after surgical repair. / Vase
Surg 2000; 32:472.

77. Hansen KJ, Starr SM, Sands RE, Burkart JM, Plonk GW Jr, Dean
RH. Contemporary surgical management of renovascular dis-
ease. / Vase Surg 1992; 16:319.

78. Susie D. The role of the renal medulla in blood pressure control.
Am J Med Sci 1988; 295:234.

79. Gothberg G, Karlstrom G. Physiological effects of the humoral
renomedullary antihypertensive system. Am } Hypertens 1991;
4:569.

80. Muirhead EE, Brooks B, Byers LW, Brown P, Pitcock JA. Medul-
lipin system: generation of medullipin II by isolated kidney-liver
perfusion. Hypertension 1991: 18(Suppl. III):158.

81. Perry MO. Ischemia/reperfusion syndromes: clinical relevance.
Ischemia Reperfus Organ Dysfunct Res Init Vase Dis 1993:3.

82. Granger DN, Villareal D. Mechanisms of reperfusion-induced mi-
crovascular dysfunction. Ischemia Reperfus Organ Dysfunct Res Init
Vase Dis 1993:7.

191

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Pathophysiology, hemodynamics, and
complications of venous disease

Harold J. Welch
Kevin B. Raftery
Thomas R O'Donnell, Jr.

Deep vein thrombosis (DVT) occurs when a thrombus, formed
from elements of the blood, develops in the deep veins of
the lower extremity. Since many patients with acute DVT are
asymptomatic, the true prevalence of DVT in the population is
unknown. Furthermore, many studies have relied on the clini-
cal diagnosis of DVT so the actual prevalence of DVT even in
hospitalized patients is underestimated. Some have suggested
at least 2-3% of the population have experienced a DVT at
some time in their life. This chapter will review the causes
of acute DVT and pulmonary embolus, and relate the acute
pathophysiology of thrombus formation within the venous
system to the clinical state. DVT and pulmonary embolus rep-
resent the early manifestations of thrombus formation in the
venous system, while venous stasis disease is a late sequelae of
acute DVT. The venous hemodynamics in normal limbs will be
contrasted with the circulatory changes in limbs with venous
stasis disease or chronic venous insufficiency (CVI). In addi-
tion, microcirculatory alterations in CVI, an area of new in-
tense focus, will be reviewed.

Typically, the red clot seen in venous thrombosis contrasts with
the white clot composed mainly of platelets and fibrin seen in
arterial thrombosis.

Etiology of venous thrombosis

The process of thrombosis is a complex interaction of many
factors involving the blood and the vessel wall (Table 17.1).
The coagulation proteins, in concert with their activators and
inhibitors, platelet activation, adherence and recruitment, and
endothelial cell modulation, play an important, intertwining
role in thrombus formation. In addition, the fibrinolytic sys-
tem restrains the growth of the thrombus. Despite major
advances in coagulation research, the basic etiologic factors in
venous thrombosis can still be categorized by Virchow's triad
originally described in 1856: (i) stasis of blood flow; (ii) injury
to the vessel wall; and (iii) hypercoagulable blood. 1

Formation of venous thrombosis usually begins in the valve
cusp sinuses where eddy currents under phasic flow produce
relative stasis. Lowered venous blood flow combined with
a hypercoagulable state or local injury initiates thrombus
formation composed mainly of fibrin and red blood cells.

Thrombus formation

Ultimate formation of a fibrin clot is accomplished by a series
of integrated reactions between the blood and the vessel wall.
The intrinsic pathway is activated when blood contacts a non-
endothelial surface. The foreign surface interacts with factor
XII, resulting in activated factor XII (Xlla). Factor Xlla then
activates factor XI, a reaction that is calcium dependent.
Factor XIa next activates factor IX in the presence of factor VIII,
and phospholipid activates factor X. This last reaction is
greatly magnified if factor VIII has been exposed to thrombin
or factor Xa.

The extrinsic pathway is initiated by tissue thromboplastins
released by injured cells. These phospholipoproteins combine
with and activate factor VII; this complex then activates factor
X. Thus, activated factor X is the reaction where the intrinsic
and extrinsic pathways meet; beyond this step is a common
pathway. Factor Xa complexes with factor Va in the presence
of calcium and phospholipid and converts prothrombin to
thrombin. The presence of factor Villa and factor Va, catalytic
complexes optimally located on the surface of platelets, allows
the rate of thrombin generation to be increased 300000 times.
Thus, thrombin and factor Xa act as a positive feedback mech-
anism for the conversion of prothrombin to thrombin on the
platelet surface.

Platelet involvement in thrombus formation is very com-
plex, with adhesion and aggregation resulting from a multi-
tude of reactions. Glycoproteins on the platelet surface bind
to exposed adhesive proteins in the vascular subendothelial
collagen layer, including von Willebrand factor (vWF),
thrombospondin, fibronectin, and vitronectin. 2 ' 3 Platelet
phospholipase activity is initiated with adhesion, leading to
thromboxane A 2 production from arachidonic acid, and the
platelets secrete active compounds. Adenosine diphosphate,

192

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

Table 17.1 Etiology of venous thrombosis

Stasis

Immobilization

Chronic venous insufficiency

Congestive heart failure

Intraluminal obstruction

Wall injury

Venodilation

Hip, knee surgery

Blunt, penetrating trauma

Thermal injury

Hypercoagulable sta te
Protein C deficiency
Protein S deficiency
Antithrombin III deficiency
Dysfibrinogenemia
Heparin cofactor II deficiency
Plasminogen deficiency
Plasminogen activator deficiency
Homocystinuria

Extrinsic compression
Increased blood viscosity
Venous dilation
Surgery/anesthesia

Chemical injury

Immune complex mechanisms

Clamps/balloon injury

? Tobacco smoke

Increased blood viscosity

Age

Obesity

Pregnancy

Oral contraceptives

Sepsis

Congestive heart failure

Previous venous thrombosis

adenosine triphosphate, calcium, and serotonin are released
by the dense granules, while the alpha granules release vWF,
fibronectin, thrombospondin, vitronectin, platelet-derived
growth factor, and (3-thromboglobulin. Thromboxane A 2 is a
potent vasoconstrictor and adenosine diphosphate is a potent
aggregating agent in a positive feedback role. Fibrinogen
and the other adhesive proteins interact with platelet sur-
face glycoproteins in calcium-dependent reactions to form
platelet-platelet bonds.

The vascular endothelium is a heterogeneous, actively func-
tioning unit, through which several homeostatic mechanisms
work to prevent thrombus formation. Endothelial cells have
high-affinity binding sites for thrombin, which can lead to
the inactivation of thrombin. 4 Heparan sulfate on the cell
surface catalyzes the thrombin-antithrombin III reaction.
Prostaglandin generated by the vessel wall causes vasodila-
tion and inhibits platelet aggregation. Normally, platelets may
adhere to endothelial cells but do not necessarily aggregate.
Low levels of prostacyclin may lead to platelet aggregation
and thrombus formation. 5 The enzyme responsible for prosta-
cyclin production, prostacyclin synthetase, has a high con-
centration in the intima and progressively decreases in the
external layers of the vessel wall, providing an antithrombo-
genic environment near the lumen and a more thrombogenic
environment deeper in the wall.

Stasis

Venous thrombosis often begins in areas of relative or
actual stasis: valve cusps and the venous sinuses of the calf

muscles. Stasis itself, however, does not cause blood to clot
when it is in contact with intact endothelium, 7 a fact known for
several hundred years. Stasis contributes to thrombosis by
allowing a localized hypercoagulable state. Static blood does
not clear activated coagulation factors, nor does it allow
dilution of these activated coagulation factors by nonactivated
blood. Additionally, inhibitors of the activated coagulation
factors cannot effectively mix with the activated factors in
static blood. Finally, increased blood viscosity can be present
in areas of decreased blood flow. In addition to patients
with polycythemia vera, erythrocytosis, and dysprotein-
emias, postoperative and inflammatory states can lead to
increased blood viscosity.

Venous stasis can result from immobility, or obstruction to
blood flow. Immobility is seen most frequently during surgery
and in the early postoperative course, as well as in advanced
age and obesity. Extremities immobilized by splints or casts,
traction, or paralysis are also associated with venous stasis.
The effect of immobility is reflected in higher rates of DVT in
patients who have undergone hysterectomy or prostatectomy
via the transabdominal route, as opposed to a transvaginal
or transurethral procedure. 8-10 Limbs that are paralyzed by
stroke have a four to nine times higher incidence of DVT vs.
nonaffected limbs, compared with an equal incidence of DVT
in the legs of paraplegic individuals. 11 ' 12 Prevention of stasis
due to immobility is the rationale behind the use of pneumatic
compression boots in patients confined to bed.

Obstruction to venous blood flow can take several forms. In-
traluminal obstruction can result from a previous thrombus,
intraluminal web, or a tumor, such as a renal clear cell cancer
invading the inferior vena cava. Extraluminal extrinsic com-
pression by tumors, a gravid uterus, or aortic aneurysm may
result in stasis. Elevated venous pressure resulting from right
heart failure will produce decreased venous flow in the pe-
riphery, leading to a high incidence of venous thrombosis in
patients with congestive heart failure. 13

Venous dilation can lead to stasis and perhaps endothelial
damage. Venodilation occurs in pregnancy, in oral contra-
ceptive use, and in varicose veins. Surgery causes release of
humeral mediators, which produce venoconstriction and ven-
odilation, the action of which can be blocked by dihydroergot-
amine. 14 Studies in dogs undergoing hip replacement have
shown endothelial injury far from the operative site, thought
to be related to operative venodilation. 15

Vessel wall injury

Injury to the venous endothelium can occur by twisting and
stretching, as seen with hip and knee surgery. Vascular clamps
and balloon catheters can cause denudation of endothelial
cells. 16 Thermal injury to the vein wall can result from electro-
coagulation and the acrylic glue used in total hip replacement.
Chemicals such as intravenous contrast agents and

193

pa rt I Vascular pathology and physiology

chemotherapeutic drugs injure endothelial cells. Circulating
immune complexes and endothelial cell antibodies may result
in endothelial cell injury 17 It is postulated that tobacco smoke
may cause damage via the mechanism of immune complexes.
Burns, blunt and penetrating trauma, varicose vein stripping,
and indwelling venous catheters are other etiologies of vessel
wall injury.

With injury to the endothelial cell layer, the subendo-
thelium, composed of collagen fibrils and the basement
membrane, is exposed to blood, which can initiate thrombus
formation via several mechanisms. Glycoproteins on the sur-
face of platelets mediate adhesion of the platelets to the colla-
gen fibrils. Other glycoproteins, including vWF, fibronectin,
vitronectin, and thrombospondin, which are found in plasma,
platelet alpha granules, and the subendothelium, also mediate
platelet adhesion and recruitment. 2 ' 3 ' 18

Injured tissue and aggregated activated leukocytes release
tissue factor, which can initiate coagulation via the extrinsic
system by activating factor VII. The intrinsic system con-
tributes to thrombosis via the activation of factor XII by the
exposed collagen and elastin, and platelet activation of factors
XII and XL

There has not been conclusive evidence that vessel wall
damage is a sole initiator of thrombosis, however. Studies
performed in rabbits where the venous endothelium was
denuded by mechanical crushing did not lead to thrombosis. 16
Autopsy studies have also failed to identify any significant
pathologic endothelial lesion.

Hypercoagulable states

Inherited or primary hypercoagulable conditions

Protein C deficiency

Protein C is a vitamin K-dependent protein synthesized
in the liver. Its action is initiated by the binding of thrombin
to an endothelial cell receptor, thrombomodulin. This com-
plex converts protein C to its activated form, which acts
as a potent anticoagulant. The activated protein C can then
inhibit activated factor V and factor VIII, and enhance
fibrinolysis.

Decreased circulating levels of protein C can be on a genetic
or an acquired basis. Protein C deficiency is an inherited con-
dition expressed in an autosomal dominant fashion, while
the acquired condition is seen in patients with liver failure,
postoperative states, disseminated intravascular coagulation
(DIC), and chronic renal failure. Homozygotes frequently die
in infancy from thrombotic complications, though treatment
with plasma and with purified protein C concentrate have
been successful. Although heterozygotes are often asympto-
matic, many will develop DVT or a pulmonary embolus before
age 50. Arterial thrombosis due to protein C deficiency is

uncommon, but patients who thrombose a bypass graft post-
operatively with no indentifiable technical reason should be
screened for a hypercoagulable state.

Protein C levels are measured by both immunologic and
functional assays. Activated protein C activity can also be
measured in a chromogenic assay. Since Coumadin decreases
levels of protein C, patients should not be tested while taking
this drug. Protein C has a short (11 h) half-life and is thus
rapidly depleted with the initiation of warfarin therapy. Be-
cause other vitamin K-dependent coagulation factors have
longer half-lives, this results in a relative protein C deficiency.
Therefore, despite the initiation of anticoagulation therapy, a
paradoxical hypercoagulable state is produced; this some-
times results in warfarin-induced skin necrosis. Thus, patients
should be heparinized while on warfarin therapy.

Protein S deficiency

Protein S is a vitamin K-dependent protein synthesized by en-
dothelial cells that in vitro acts as a cofactor for activated pro-
tein C as it inhibits factors Va and Villa. Protein S circulates in
both an active free form and an inactive form bound to C4b
binding protein.

Like protein C, protein S deficiency can be acquired or inher-
ited. Inherited protein S deficiency can present in two forms: (i)
markedly decreased total protein S concentration; or (ii) nor-
mal or near normal levels of total protein S but with signifi-
cantly decreased free protein S concentration.

Protein S deficiency can be seen in the same clinical condi-
tions described for protein C, and may manifest by superficial
or deep thrombophlebitis, as well as mesenteric vein thrombo-
sis. Levels should be measured with the patient off warfarin
for at least 2 weeks. Both total and free concentrations should
be measured.

Antithrombin III deficiency

Antithrombin III (AT III) is a circulating plasma protein
synthesized by the liver, which inhibits thrombin and the
activated factors IX, X, XI, and XII. Inhibition of thrombin by
AT III occurs by binding to form a complex, which occurs at a
relatively slow rate. The activity of AT III is greatly increased
by heparin, and the heparin-like substance bound to endothe-
lial cells, heparan sulfate. When AT III is complexed to the
endothelial surface heparan sulfate, it neutralizes thrombin
and guards against thrombosis. Heparin accelerates the AT
III-thrombin complex formation by 1000-fold. 19

Inherited deficiency of AT III is by an autosomal dominant
transmission, with affected patients having AT III levels
40-60% of normal. This deficiency accounts for approximately
2-4% of venous thrombosis in patients under the age of 50.
Affected patients will frequently have femoral or popliteal,
or iliac DVT, often under the age of 25 and usually connected
with an inciting event such as trauma or pregnancy.

194

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

Diminished levels of AT III are also seen with hepatic insuf-
ficiency, with shock, with nephrotic syndrome, in women tak-
ing oral contraceptives, and with heparin therapy. Clinical AT
III deficiency can be due to decreased levels (quantitative defi-
ciency) or a dysfunctional state caused by inherited structural
abnormalities of the molecule (qualitative deficiency). Tests
should include immune assays to measure the concentration
and the functional ability of the patient's plasma to inhibit
thrombin in the presence of heparin, and levels should be mea-
sured with the patient off heparin.

Heparin cof actor II deficiency

A circulating glycoprotein, heparin cofactor II, also inhibits
thrombin but not other coagulation factors. As with AT III, the
rate of heparin cofactor II inhibition of thrombin is markedly
potentiated by heparin. Inherited as an autosomal dominant
disorder, affected patients will have levels approximately
50% of normal due to diminished synthesis. Both arterial and
venous thrombosis can occur.

Hyperprothrombinemia

Poort and associates 20 have described a genetic mutation
which occurred on the prothrombin gene located at nucleotide
20210. This mutation is found in approximately 20% of pa-
tients who have familial episodes of DVT and is associated
with increased levels of plasma prothrombin. This abnormal-
ity is found concomitantly with other inherited hypercoagul-
able states.

Plasminogen and plasminogen activator deficiency

Opposing the normal reaction of clot formation is the fibrin-
olytic system. Circulating plasminogen and the enzyme
tissue plasminogen activator bind to fibrin, after which tis-
sue plasminogen activator converts plasminogen to plasmin.
Plasmin then degrades fibrin and hydrolyzes other plasma
coagulation proteins — fibrinogen, factor V, and factor VII. De-
creased clot lysis will be seen in patients with low circulating
plasminogen levels or a dysfunctional molecule, abnormali-
ties that are inherited in an autosomal recessive manner.
Venous endothelial cells produce plasminogen activator. Low
levels of this protein have been described in patients with
recurrent thromboembolism. 21 Low levels can result from
decreased plasminogen activator production, or conversely
from supranormal levels of plasminogen activator inhibitor,
which complexes with plasminogen activator and thereby de-
creases its activity.

Deficiencies of this portion of the fibrinolytic system can be
measured with functional assays, determining the levels of
both plasminogen activator and its inhibitor.

Dysfibrinogenemia

Numerous patients have been described with various inher-
ited functional abnormalities of fibrinogen. 22 These fibrinogen
abnormalities include a diminished capacity to bind plas-
minogen or plasminogen activator, a defective polymeriz-
ation of fibrin monomers, decreased ability of fibrin to bind
thrombin, and resistance of fibrin degradation by plasmin.

Homocystinuria

The genetic disorder homocystinuria, due to cystathionine
synthase enzyme deficiency, results in an accumulation of
homocystine in the tissues and plasma. Homocystinuria is
associated with increased platelet activation and local de-
nudation of venous and arterial endothelium, which may
result in thrombosis.

Acquired or secondary
hypercoagulable states

Surgery and trauma

All three parts of Virchow's triad figure in venous thrombosis
when the patient is traumatized or undergoes surgery. Vessel
wall injury occurs with blunt and penetrating trauma or
operative injury and clamping. Twisting and distortion of the
femoral vein as documented by intraoperative venography
occurs with total hip replacement, 23 probably leading to inti-
mal disruption. Ninety percent of femoral vein thrombosis oc-
curs ipsilateral to the side of hip surgery, while calf vein DVT is
more evenly distributed. 23 ' 24

Decreased blood flow is present in the immobilized patient
perioperatively with the induction of anesthesia and in the
limb immobilized by a cast or by traction.

Soft tissue trauma, surgery, and burns all stimulate the
release of tissue thromboplastin, 25 ' 26 which can lead to
thrombosis via activation of the extrinsic pathway. There is
also reduced fibrinolytic activity in the early postoperative
period. 27

Pregnancy and oral contraceptives

The use of oral contraceptives has been shown to increase
the risk of thromboembolic events by 4 to 11 times that
seen in women not taking oral contraceptives. 28,29 Pregnancy
and oral contraceptives induce a hypercoagulable state by
several mechanisms. Fibrinogen and factors VII, VIII, and IX
are increased while fibrinolytic activity is concomitantly
decreased. Reduced levels of plasminogen activator have
been identified in each trimester of pregnancy, along with low
levels of AT III. Tissue thromboplastin is also released into the
circulation with placental separation. 30 Oral contraceptives

195

pa rt I Vascular pathology and physiology

have also been reported to decrease AT III levels, changes that
take several months to approach normal levels after stopping
the medication.

Estrogens also cause increased distensibility of veins, lead-
ing to decreased velocity of flow and relative stasis. Increased
venous pressure due to the enlarged uterus and during deliv-
ery also contributes to stasis.

Malignancy

Malignant tumors increase the likelihood of developing
venous thrombosis by various mechanisms. The neoplasm
can extrinsically compress or invade veins leading to
thrombosis. Malignant tumor cells from the breast, colon, and
vagina can produce factor X activation. Multiple myeloma,
mucin-secreting adenocarcinoma, and promyelocytic leuk-
emia cells secrete tissue thromboplastin. Many cancer patients
will have decreased fibrinolytic activity with low levels of
AT III and increased concentrations of fibrinogen, and factors
V, VIII, IX, and X. It is probable that these humoral
substances increase the risk of venous thrombosis two to
three times that of patients who undergo similar surgical
procedures for nonmalignant disease. These patients will
frequently develop superficial vein thrombosis, venous
thrombosis in unusual locations, and thrombosis resistant to
anticoagulant therapy.

Sepsis

While septic patients are usually immobile in bed and likely to
have hepatic, renal, or cardiac insufficiency, Gram-negative
and Gram-positive bacteria can stimulate platelet aggrega-
tion. Endotoxin from Gram-negative bacteria can also lead to
tissue factor-like activation of the coagulation system.

Heart failure

Congestive heart failure leads to increased venous pressure.
Combined with the immobility of these ill patients, venous
stasis is increased.

Obesity

Obesity has been associated with an increased risk for
thromboembolic disease. The contributing factors are a de-
creased fibrinolytic activity seen in overweight patients, and a
tendency to be less mobile postoperatively.

Age

Elderly people are at higher risk for thromboembolism due to
several possible mechanisms, though definite causes have not
been identified. There is a decrease in fibrinolytic activity in

patients over age 65, and venous dilation is also seen in the
elderly. These factors, combined with decreased mobility and
associated disease states, are likely to contribute to the higher
incidence of venous thrombosis.

Previous venous thrombosis

Patients who have had an episode of venous thrombosis may
suffer vein valve damage leading to insufficiency and stasis, or
residual clot may initiate recurrent thrombosis. One episode of
DVT makes a patient two to three times more likely to have
a subsequent episode after undergoing abdominal surgery 31
Those patients with recurrent thrombosis have a threefold to
fourfold chance of developing another thrombosis after dis-
continuation of anticoagulant therapy.

Etiology of varicose veins

Varicose veins are common in the Western world, indicating
both a genetic and environmental influence on their devel-
opment. The prevalence of varicose veins in developed
countries ranges from 10% to 64%, depending on the age
and sex of the population. Conversely, in developing coun-
tries, the prevalence is much less, ranging from 1% to 10%.
It had been postulated that one important factor in the
geographic /environmental differences in prevalence is the
amount of dietary fiber. 32 The low fiber diet of developed
countries results in high intraabdominal pressures required
to evacuate firm stools. This pressure is transmitted to the
venous system.

Varicose veins are two to eight times more common in
women than in men. This is due to hormonal influence, partic-
ularly estrogen, causing relaxation of smooth muscle and col-
lagen fibers with subsequent venodilation.

CONTRIBUTING FACTORS IN THE DEVELOPMENT
OF VARICOSE VEINS

• Vein wall weakness

• Valvular incompetence (superficial and perforator)

• Arteriovenous shunts

• Hormonal (estrogen)

• Genetic

• Environmental /dietary

Valve incompetence is another important etiologic factor in
the development of varicose veins. Of particular importance is
the saphenofemoral valve, where incompetence can lead to
venous dilation and further varicosities. The question of con-
siderable debate is which came first: venous dilation or valve
incompetence. The likely answer is both. Weakness in the vein
wall can lead to dilation with subsequent failure of the valve
cusps to coapt and, thus, permit reflux. In other patients, how-
ever, there is likely to be primary valve incompetence. This
condition is well recognized as a leading cause of deep venous

196

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

insufficiency, and it is likely to be present in superficial veins as
well.

Arteriovenous shunts have been suggested to play a role in
the development of varicose veins. Anastomoses between
arterioles and varicose veins have been demonstrated angio-
graphically and by microsurgical dissection. To assess the
significance of the arterial inflow into the varicosities, several
studies have measured the oxygen tension in blood sampled
from varicose veins. These studies produced diametrically op-
posite results, however, showing both higher and lower oxy-
gen tension in varicose veins compared with other veins. 33 ' 34 It
has been suggested that arteriole inflow may be a hemody-
namic factor, affecting a weakened vein wall leading to dila-
tion, though this is speculative. 35

Incompetent perforating veins have long been implicated in
the development of varicose veins. Valve failure in the perfor-
ating vein allows the 150-200 mmHg pressure developed in
the deep muscle compartment during exercise to be trans-
mitted to the superficial veins. It has been documented, how-
ever, that even in normal limbs, there can be a reversal of
flow through the perforating veins. Thus, there are probably
varying degrees of incompetence. 36

The leading theory as to why veins become varicose is
that some veins have an inherent weakness in their wall.
Normal lower extremity vein walls are composed of three
muscle layers supported by a matrix of collagen and elastic
fibers. The elastic fibers are dispersed throughout the vein wall
and seem to provide elastic recoil. As with arteries, the mor-
phology of the vein wall and number of valves present vary ac-
cording to their location in the body. The further distally on the
leg, the greater the number of valves and the thicker the vein
wall.

Varicose veins show marked changes in their walls, with a
significant increase in fibrous tissue interspersing among the
muscular layers, resulting in separation and disruption of the
muscular bundles. 37 This fibrous infiltration is not uniform
within the vein wall, and areas of "blowouts" or "blebs" may
have only collagen, endothelium, and subendothelial tissue.
Transmission electron microscopy also shows significant
changes in varicose veins. There is a marked increase in colla-
gen fibers, most of which have lost their regular bundle forma-
tion, with marked irregular scattering of both collagen and
elastic fibers. As a result, there is a separation of the muscle
cells. Additionally, some collagen fibers seem to have been
phagocytized by the smooth muscle cells. The increased peri-
cellular fibrous tissue infiltration also appears to be caused by
secretions of the smooth muscle cell. 37

Proponents of the wall weakness etiology theorize there is
an imbalance in the collagen-elastic fiber-smooth muscle cell
makeup of the vein wall. This results in lack of contractility.
Though unidentified, there is probably a genetic factor respon-
sible for the imbalance.

Normal venous physiology

To understand the pathophysiology of venous stasis disease,
the normal circulatory responses in the venous system must be
outlined. It has been well accepted that an elevated ambula-
tory venous pressure is fundamental to the development of
the abnormal skin changes, as well as to producing the micro-
circulatory alterations. 38 Venous pressure is measured con-
ventionally in a superficial dorsal foot vein. In the standing
position, the pressure in a dorsal foot vein is related to two
factors: (i) the pressure gradient between the arteries and the
veins through the capillary bed; and (ii) the hydrostatic pres-
sure. 39 The hydrostatic pressure is defined as the gravitational
force produced by a column of blood between the right atrium
and the measuring point on the foot. As Browse and associ-
ates 39 calculated, in a typical upright 1.8-m individual, the
hydrostatic pressure would be approximately 100 mmHg,
which when added to the measured foot vein pressure of
15 mmHg would represent 115 mmHg. In the lower extremity,
venous pressure is not static and is under the influence of the
calf muscle pump, as well as changes in intraabdominal and
intrathoracic pressure. In addition, venous vasomotor re-
sponses and venous valves play a role. Venous tone regulates
the relative partitioning of blood flow in the limb, while ve-
nous valves control the direction of blood flow. 40 The sympa-
thetic nervous system regulates venous tone much as it does in
the arterial system. This response is an important thermoregu-
latory mechanism, particularly in the skin and fatty tissue
layers. For example, with an increased core temperature,
venodilation occurs, while during exercise or with pain, veno-
constriction occurs.

Because of the intense interest in in situ bypass and venous
reconstructive surgery of the deep system, venous valves have
been closely studied. Vein valves are bicuspid and usually
occur at venous tributaries. While valves are unusual in the
inferior vena cava and upper iliac veins, they are more numer-
ous below the level of the common femoral vein. For example,
there are several important and relatively constant valves in
the upper thigh. One is at the superficial femoral vein just be-
fore its junction with the common femoral vein; another is in
the greater saphenous vein at its junction with the common
femoral vein. 41 Finally, a valve is located at the origin or just
after the origin of the profunda femoris vein. The valves in the
lower superficial femoral vein and popliteal vein appear to
play a critical role in modulating the hemodynamic responses
of the calf muscle pump . 42 The function of vein valves is to pre-
vent retrograde flow. With Valsalva maneuver or, better, with
the sudden release of an occluding tourniquet, there is a mo-
mentary reversal of flow across the vein valve. Valve closure is
related to both the magnitude of the retrograde flow and to
the attendant pressure differential across the valve structure. 43
That is, the supravalvular pressure must rise to cause closure

197

pa rt I Vascular pathology and physiology

of the valve. The anatomic structure of the valve is important
for appropriate valve closure. The valve edge or the free bor-
der of the valve cusp must be tight and nonredundant. With
valve closure, there is distention of the valve sinus and tight-
ening of the valve edge. If this cannot occur, then valve incom-
petence usually takes place.

Anatomic units of the lower extremity

The superficial system includes the small veins in the skin and
subcutaneous tissue, as well as their major tributaries, the
greater and lesser saphenous veins. In general, the communi-
cating or perforating veins are grouped with the superficial
system. The greater saphenous vein runs at a layer deeper than
its tributaries (on the fascia) while the tributaries lie in the
subcutaneous tissue. The smaller branch veins in the subcuta-
neous tissue are situated in loose areolar tissue, which pro-
vides little support against distention. The deep system is
composed of the paired tibial and peroneal veins, which join
together to form the popliteal vein, the superficial femoral and
profunda femoris vein, and the common femoral vein. These
latter veins lie in between muscle bundles, in contrast to the
veins of the true muscle pump —the soleal and gastrocnemius
veins— which lie within muscle tissue. Activation of the calf
muscle pump occurs with contraction of the gastrocnemius
and soleal muscles when walking. This phase is termed the
systolic phase of the calf muscle pump cycle. 39 Blood is expelled
from the gastrocnemius and soleal veins, which act as reser-
voirs so that blood flows cephalad toward the heart. It has been
calculated that approximately 60-70 ml of blood constitutes
the calf blood volume and approximately 50% of the blood
(30 ml) is expelled with each contraction. Forward flow to-
ward the heart is aided by one-way valves in the deep venous
system. During the systolic phase of the calf muscle pump in
the normal limb, there is no flow of blood from the deep to the
superficial system. With increased levels of exercise, a greater

volume of arterial blood flow is provided to the muscles so the
calf muscle pump must increase its rate of contraction and vol-
ume flow. During the diastolic phase of the calf muscle pump,
the valves in the perforating veins open, allowing superficial
to deep blood flow. In addition to the contribution of blood
flow from the superficial venous system, the gastrocnemius
and soleal muscle veins are filled by arterial flow.

Intrathoracic and intraabdominal pressure modulate ve-
nous blood flow by their effect on outflow tract resistance. For-
ward flow of blood is encouraged during the expiratory phase
of the respiratory cycle, while during inspiration, lower ex-
tremity blood flow to the abdomen is reduced. The situation is
different with flow from the abdominal to thoracic segments,
which is encouraged by a drop in intrathoracic pressure dur-
ing inspiration. By contrast, during expiration, abdominal to
thoracic blood flow is reduced.

Venous stasis disease

While the pathophysiologic changes associated with chronic
venous insufficiency (CVI) are recognizable at the gross
level— recanalization changes, valvular damage— the ulti-
mate effect is on the microcirculation. An increased ambu-
latory venous pressure is the hallmark of CVI. Numerous
clinical studies where venous pressure measurements were
carried out have detailed the differences in venous pressure
between involvement of the superficial venous system alone,
with perforating vein incompetence and, finally, with deep
venous disease. 44-46 The following sections will describe the
pathophysiology of CVI, including (i) changes in the large
veins as observed by such physiologic measurements as inva-
sive venous pressure, duplex flow velocities, and electromag-
netic flow assessment, and (ii) microcirculatory alterations as
detailed by TCP0 2 , laser Doppler flow, and histologic exami-
nation. Table 17.2 compares the changes in foot vein pressure
during exercise in 38 normal subjects with 21 patients with

Table 17.2 Percentage of change in foot vein pressure during exercise of normal subjects and patients with venous disease'

Reduction of

pressu

re/standard error

mean (%)

Normal

subjects

Group 1

Group 2

Group 3

Group 4

Group 5

(n = 38)

(n = 21)

(n = 10)

(n = 11)

(n = 37)

(n = 40)

With tourniquet

68/12

45/23

37/25

2/23

31/27

17/18

With thigh tourniquet

38/15

56/16

30/27

23/22

47/18

16/21

With below-knee tourniquet

43/22

40/29

33/25

28/20

25/24

12/15

*Group 1 : Saphenofemoral incompetence alone; normal deep veins demonstrated on ascending phelography.

Group 2: Saphenopopliteal incompetence alone; normal deep veins demonstrated on ascending phelography.

Group 3: Calf-communicating vein incompetence alone: normal deep veins demonstrated on ascending phelography.

Group 4: Calf-communicating vein and saphenous incompetence: normal deep veins demonstrated on ascending phelography.

Group 5: Phelographic evidence of deep vein damage by thrombosis.
(From Burnard KG, O'Donnell TF, Lea-Thomas M, Browse NL. The relative importance of incompetent communicating veins in the production of varicose veins
and venous ulcers. Surgery 1 977 82: 9.)

198

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

clinical and phlebographic evidence of greater saphenous
incompetence (GSI), and 37 patients with both perforating
vein (ICPV) and saphenous incompetence (GSI/ICPV). 47 In
the normal limbs, the percent change in venous pressure with
exercise was 68%. Both the GSI and GSI/ICPV groups had less
of a reduction in pressure with exercise than the normal limbs.
A thigh tourniquet normalized the decrease in venous pres-
sure with exercise patients with GSI but failed to do so in the
GSI/ICPV group.

Air plethysmography permits measurements of leg
volume changes with exercise and defines the degree of super-
ficial and deep venous reflux, the status of the calf muscle
pump, and indirectly measures ambulatory venous pressure.
Christopolous and coworkers 48 have shown that the total
venous volume is increased in most patients with any form of
CVI, and there is considerable overlap among the various
anatomic groups: GSI, ICPV, and deep venous disease. The
venous filling index, a measure of venous reflux, is calculated
from the time that it takes to achieve 90% of the venous
volume. In Christopolous' original study, 48 the venous filling
index increased from the normal value of 2 to 5ml/s in pa-
tients with primary varicose veins. Skin changes were com-
mon in patients with a venous filling index greater than 7 ml/ s,
a range characteristic of patients with deep venous insuffi-
ciency. A tourniquet failed to normalize venous filling index
in patients with deep venous disease but reduced it to less than
5 ml/s in patients with superficial venous disease. Calf muscle
pump function can be determined by measuring the ejection
fraction or the amount of volume expelled from the calf with
one contraction of the calf muscles. The ejection fraction was
reduced from the normal value of 75% to approximately 50%
in patients with superficial venous disease, and decreased fur-
ther to 35% in those with deep venous involvement. Finally,
the residual venous volume fraction is determined by measur-
ing limb volume after 10 tiptoe movements and comparing it

with total venous volume. The residual venous volume frac-
tion is markedly elevated in both limbs with superficial
venous disease and those with deep venous involvement, and
is an indirect measurement of ambulatory venous pressure.

Pathologic involvement of the deep venous system occurs as
a result of obstruction, valvular incompetence, or a combina-
tion of both. The clinical state of the limb has been related to
the level of involvement. For example, patients with stage 1,11
disease by the ICVS/SVS clinical classification, 49 who show
evidence of mild varicosities and no cutaneous or subcuta-
neous findings, would be less likely to have deep venous in-
volvement than patients with stage V/VI disease who have a
healed or active venous ulcer. The proportion of limbs with
deep venous involvements in stage III CVI varies from 22% to
73% and can be influenced by several biases: (i) referral bias:
reviews from larger referral hospitals might contain patients
with disease less treatable by straightforward surgical ap-
proaches; and (ii) surgical bias: patients who comprise surgi-
cal series may have venous disease either resistant to, or
amenable to, reconstructive surgery 50 For example, in our ex-
perience with patients who had venous ulcers, only 15% of
limbs had superficial involvement alone, 51 while the majority
had deep venous involvement. By contrast, Darke and An-
dress 52 had a proportion of deep venous involvement less than
ours; they observed a correspondingly higher proportion with
superficial venous system involvement.

Pathologic causes of deep venous involvement

Table 17.3 summarizes the type and location of deep venous
involvement among eight series. 50 It is apparent that obstruc-
tion is an infrequent cause of venous disease in these large
series, averaging about 7%. By contrast, valvular incompe-
tence is the chief cause of deep venous involvement, compris-
ing nearly 90% of cases.

Table 17.3 Type and location of deep venous insufficiency

No. of

Criteria

Methods

fordiag

nosis

Proximal/distal (popliteal)

Study

limbs

for entry

Clin

Dop

AVP

Phleb

Obstruction

vascular incompetence

Duke and Andress

100

Ulcer

—

19/88

NEMCH

346

CVI

—

11/80

Moore eta/.

113

—

22/78

Rajuand Fredericks

100

Schanzerand Pierce

Surg

Peace et al.

47/14

Bruns-Slotefa/.

194

14/86

Gooleyand Somner

54/46

CVI, chronic venous insufficiency; AVP, venous pressure; Surg, candidate for surgery; Phleb, attending or descending phelogram; Clin, clinical; PT, post-
thrombotic syndrome; Dop, bidirectional Doppler.

(From O'DonnellTF. Chronic venous insufficiency and varicose veins. In: Young JR, GraorRA, Olin JW, BartholemewTR, eds. Peripheral Vascular Disease. St.
Louis: Mosby Year Book, 1 991 : 467.)

199

pa rt I Vascular pathology and physiology

Type of valvular incompetence

Many physicians assume that valvular incompetence is the se-
quelae of acute DVT. As early as the 1940s, Bauer 53 described
patients with deep venous valvular incompetence who did
not manifest the typical post- thrombotic changes. He recog-
nized a new cause of valvular incompetence where the valve
cusps were redundant and failed to coapt. After selecting
limbs for surgery by high quality ascending and descending
phlebograms that showed post-thrombotic change, Kistner 54
observed this entity directly at surgery, which he termed
primary valvular incompetence. He ascribed primary valvular
incompetence to degeneration of fibroelastic tissue. Forty
to fifty percent of limbs in Kistner 's series had findings
compatible with primary valvular incompetence. Darke and
Andress 52 observed a comparable incidence of patients
with primary valvular incompetence who had this diagnosis
established on phlebography. None of these patients exhibited
typical post-thrombotic changes on ascending and descend-
ing phlebograms. Finally, patients may have primary valvular
incompetence proximally in the superficial femoral vein with
typical post-thrombotic changes distally. It has been theorized
that the stasis produced by the proximal primary valvular
incompetence leads to thrombotic interaction with the vein
wall and typical recanalization changes. 55

Location of deep venous involvement

Several studies have shown that the popliteal vein valve is crit-
ical in the genesis of venous ulceration. Schull and colleagues 56
showed a high rate of popliteal vein valve dysfunction by bidi-
rectional Doppler in the limbs with venous ulcer and a previ-
ous episode of acute DVT, while Darke and Andress's 52
phlebographic evaluation of 100 limbs with venous ulcer
demonstrated popliteal vein valve incompetence in 80% of
these limbs. Our evaluation of 225 limbs showed a comparable
incidence to Darke and Andress's evaluation of popliteal
valve dysfunction. 51

Sequelae of an acute episode of DVT

Not all limbs that have had acute DVT will develop advanced
skin changes characteristic of the post-thrombotic limb. The
relationship between a high proportion of limbs developing
venous ulcer following an episode of acute DVT was
suggested by Bauer 's 53 early studies in the 1940s and our high-
ly selected group of patients with iliofemoral DVT, 80% of
whom developed venous ulcers by 10 years. 57 Markal and as-
sociates 58 conducted a prospective study of 268 patients who
had sustained an acute venous thrombosis, and followed 123
limbs over a 5-year period. Duplex assessment of valve func-
tion was performed initially after the episode of DVT, at 1
month and every 3 months for the first year. The patients were

then seen yearly and were compared with a cohort of patients
without a previous history of DVT or CVI. Approximately 15%
of limbs had valvular incompetence immediately following
the episode of deep venous thrombosis. The incidence of re-
flux had doubled by the first month after the initial episode of
acute DVT. By the first year, approximately 70% of the limbs
had valvular incompetence. The popliteal vein was the most
frequently involved site for valvular incompetence (58%)
while the superficial femoral vein (37%) was next. Valve in-
competence developed more frequently in venous segments
that had contained thrombus. During the short-term follow-
up period, the incidence of lipodermatosclerosis was low
despite the relatively high incidence of reflux.

In a parallel study, Killewich and associates 59 followed 21
patients with acute DVT sequentially with duplex scanning.
By 90 days after onset of DVT, over half of these patients had
recanalization changes in all segments. Valvular incompe-
tence was demonstrated in 13 of 21 patients during the follow-
up period. Patients who developed leg swelling within the
first month following DVT usually had residual obstruction
rather than valvular incompetence. By contrast, edema that
developed late after the onset of DVT was more likely to be due
to valvular incompetence.

During the systolic phase of the calf muscle pump, flow is
outward and pressure is increased within the superficial
venous system. The failure of the perforating vein valves to
impede flow from the deep to the superficial system has been
fundamental to the pathophysiology of CVI. 38 Bjordal 60 mea-
sured the direction and volume of blood flow with electro-
magnetic flow meters in patients with varicose veins. He
demonstrated bidirectional flow but the dominant flow direc-
tion was inward. Linton 61 in the United States and Cockett and
Jones 62 in the United Kingdom emphasized the important
role of ICPV in the genesis of lipodermatosclerosis and venous
ulcer. Through postmortem studies and clinical inference,
Cockett and Jones 62 suggested that incompetent perforating
veins are the key factor in the post-thrombotic syndrome. Evi-
dence for the important role of ICPV was inferred from clinical
series where interruption of the incompetent perforating
veins led to healing of the venous ulcer. A subsequent study
in which all patients undergoing subfascial vein ligation of
ICPVs underwent preoperative phlebography emphasized
that deep venous valvular incompetence played an important
role. 63 Patients with deep venous involvement detailed on
preoperative phlebography had a high rate of recurrence
with treatment of ICPV alone.

Other than Bjordal's 60 study, little documentation exists
on direction of blood flow in patients with ICPVs. Two
studies that used duplex scanning have cast doubt on the
traditional concepts of ICPVs. Hanrahan and coworkers 64
carried out duplex imaging in 95 extremities with venous
ulcer. Nearly 70% of these patients had both superficial
and deep venous system involvement, and most had ICPVs.

200

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

The same authors 65 subsequently examined 30 patients
with nonhealing venous ulcers and compared them with
20 normal volunteers, all of whom underwent duplex evalua-
tion of the perforating veins. Duplex scanning showed sig-
nificant difference in the mean diameter of the perforating
veins between normal limbs and patients with venous
ulcers; in the latter, the perforating veins were much wider.
Direction of flow was not addressed in that study. A
recent study by Sarin and associates 66 examined direction of
blood flow in perforating veins. They verified that the direc-
tion of flow could either be deep to superficial or superficial to
deep, but were unable to demonstrate a consistent pattern
of deep to superficial blood flow in limbs with incompetent
perforating veins and ulcer. They showed during the diastolic
phase of the calf muscle pump that patients with CVI had flow
through their perforators in contrast to normals who had
none.

Microcirculatory changes

While most investigators agree elevated ambulatory venous
pressure is characteristic of patients with advanced CVI, the
causes of microcirculatory changes are debated. Previous in-
vestigators suggested hypoxia at the skin and subcutaneous
level was the dominant mechanism. Originally, arteriovenous
shunts were incriminated as the cause of the cutaneous hy-
poxia. 67 Measurements with microspheres and macro-
aggregates, however, have not substantiated the presence of
significant arteriovenous shunting in patients with CVI. 68 A
second mechanism proposed for hypoxia was the fibrin cuff
theory of Burnand and colleagues. 69 In a series of clinical and
experimental studies, they outlined a mechanism for cellular
injury. On histologic examination of skin biopsies, these au-
thors noted the capillary plexus in the skin had responded to
increased ambulatory pressure by increasing their redun-
dance and capillary folds. The number of folds appeared relat-
ed to the severity of the ambulatory venous pressure. In
addition, the capillaries were leaky due to widened inter-
endothelial cell pores, which allowed the accumulation of
macromolecules in the interstitium. Fibrinogen was one of the
dominant molecules and underwent polymerization to fibrin.
Due to diminished fibrinolytic activity within the vein walls,
fibrin is not broken down, so it coats the capillary walls. Histo-
logic examination of skin from patients with advanced CVI
showed the presence of pericapillary fibrin cuffs, while assess-
ment of fibrinolytic activity in these patients showed reduced
activity. Fibrin cuffing was theorized to reduce the delivery of
oxygen and substrates to tissue.

Fundamental to this thesis is that oxygen is reduced in the
tissue surrounding the ulcer. There are conflicting studies as to
whether there is a true oxygen deficit. Mani and associates 70
carried out transcutaneous measurements in patients with leg

ulcers and demonstrated a significant decrease in TCP0 2 . Tra-
vers and associates 71 compared preoperative values of TCP0 2
in patients with uncomplicated varicose veins with those with
suspected deep venous involvement and lipodermatosclero-
sis. They demonstrated that only the patients with lipoder-
matosclerosis had a reduced TCP0 2 (mean 39.6 ± 8.2mmHg)
vs. control (80.8 + 8.33 mmHg) and patients with uncompli-
cated varicose veins, 72.8 mmHg. Postoperatively, the TCP0 2
rose to near normal levels. Contrary evidence was forwarded
by other investigators. 72 Inherent in the assessment of TCP0 2
is difficulty with the measuring device. Multiple factors affect
the measurement of TCP0 2 ; the major one is skin temperature.
Indeed, patients with CVI have been shown to have increased
tissue oxygen with heating. Finally, the use of positron
emission tomography, which measures both blood flow and
oxygen extraction, revealed an increase in cutaneous flow
but a reduced oxygen extraction. 73

Further evidence to support the fibrin cuff theory was sug-
gested by a therapeutic protocol which used the anabolic
steroid stanozolol, which enhances fibrinolysis. 74 Histologic
examination of skin sampled during and after treatment with
this drug showed reduction of the fibrin cuff. A subsequent
study by Layer and associates, 75 however, could not confirm
this original finding. McMullen and colleagues 76 also con-
ducted a clinical trial of longer duration. While they demon-
strated a decrease in the area of lipodermatosclerosis in the
treated group, they were unable to observe an improvement
in tissue oxygenation.

White cell activation of cytokines

Following up on the work of Moyses and associates, 77 Thomas
and colleagues 78 sampled blood from the long saphenous vein
at the ankle level in a group of normal volunteers and in pa-
tients with CVI. They showed a significant reduction in the
white blood cell count in patients with CVI and suggested that
trapping of the white cells within the microcirculation was oc-
curring in these patients. Coleridge-Smith and associates 79
used capillary microscopy to detail further changes in white
blood cells within the microcirculation. They observed an in-
crease in the number of capillary loops in patients with lipo-
dermatosclerosis, similar to Burnand and Browse's original
study 69 They noted a decrease in the number of capillary loops
available and related this to white cell trapping. Since white
blood cells are larger than red blood cells, they suggested the
white blood cells became trapped in the microcirculation.
They proposed the white blood cells became attached to capil-
lary endothelium and were then activated. Potent proteolytic
enzymes were released and superoxide radicals developed.
Combined with the physical effect of capillary occlusion by
white blood cells, both cytoxic injury and heterogeneous
perfusion occurred. The latter aspects of this theory are yet to
be proved.

201

pa rt I Vascular pathology and physiology

Pulmonary embolization

Physiologic results

Pulmonary embolism (PE) is the most common cause of mor-
tality associated with deep venous thrombosis. 80 The alter-
ations in cardiopulmonary hemodynamics are important
indicators of the presence and severity of PE. Understanding
the physiologic consequences of PE can aid in the recognition
and treatment of patients with this disease.

Hemodynamic consequences

Acute PE increases pulmonary vascular resistance by reduc-
ing the cross-sectional area available for pulmonary arterial
flow. In patients without preexisting lung disease, pulmonary
hypertension occurs when angiographically demonstrable
obstruction exceeds 30% of the pulmonary arterial tree. The
degree of pulmonary hypertension has been shown to be
directly proportional to the degree of pulmonary vascular
obstruction. 81 Mean pulmonary artery pressures of 30-
40 mmHg represent severe pulmonary hypertension in previ-
ously healthy patients since 40 mmHg is approximately the
maximum pressure that a normal right ventricle can
generate. 81 ' 82 By contrast, patients with chronic pulmonary
vascular disease and a superimposed acute PE can generate
higher pulmonary artery pressures because of preexisting
right ventricular hypertrophy. Sharma and coworkers 82 found
a mean pulmonary artery pressure of 40 mmHg was average
in patients with preexisting pulmonary vascular disease and
acute PE. Furthermore, no correlation existed between the de-
gree of pulmonary artery obstruction (scored by pulmonary
arteriogram) and the mean pulmonary artery pressure in these
patients. 82

Pulmonary hypertension and increased pulmonary vascu-
lar resistance can also occur due to pulmonary vasospasm
caused by the release of vasoactive substances. 83 Circulating
platelets are activated by thrombin contained in the embolus
and, in turn, release thromboxane A 2 and 5-hydroxytrypta-
mine (serotonin, 5-HT), which are both potent pulmonary
vasoconstrictors. A morphologic basis for the humoral role of
platelets in pulmonary thromboembolism was demonstrated
by Thomas and coworkers 84 using New Zealand white rabbits
in 1966. Experimentally produced pulmonary emboli harvest-
ed from rabbit lungs were found to be coated with tightly
aggregated degranulated platelets. Serotonin released from
aggregating platelets causes contraction of canine pulmonary
arteries in vitro and is blocked by serotonin antagonists. 85
Huval and associates 86 found a dramatic reduction in
pulmonary vascular resistance and mean pulmonary artery
pressure when ketanserin (a selective serotonin receptor
antagonist) was given to dogs subjected to experimental PE
with autologous clot in vivo. Ketanserin has been shown

to have a similar, albeit less dramatic, effect in humans with
acute PE. 87

Acute PE has been associated with a reduction in cardiac
output in humans. 88 The acute rise in right ventricular after-
load results in increased right ventricular wall stress and myo-
cardial oxygen consumption. 89 This insult is exacerbated in
patients with underlying coronary artery disease. The degree
of cardiac output reduction is related to both the size of the
pulmonary embolus and the extent of underlying coronary
artery disease.

Impairment of gas exchange

Increased alveolar dead space

When an embolus occludes a pulmonary artery, it obstructs
perfusion to a segment of the lung, making that zone of lung
unavailable for gas exchange. Ventilation of this zone of lung is
wasted and has been described as alveolar dead space. Increased
physiologic dead space, whether caused by complete or
partial obstruction of a pulmonary artery, impairs efficient
elimination of C0 2 by the lung.

Dead space ventilation can be measured using the Bohr
technique. 90 Expired gas is collected for the measurement of
mean expired pC0 2 and the simultaneous measurement of ar-
terial pC0 2 . The tidal volume, ventilatory rate, and minute
ventilation are also measured during the collection of expired
gas. The ratio of dead space to tidal volume (Vd/Vt) provides
a measure of how efficiently the lungs eliminate C0 2 . The nor-
mal Vd/Vt is considered to be less than 35%. In one group of
patients with documented PE by pulmonary arteriogram, all
16 patients had abnormally elevated Vd/Vt (greater than
41 %). 90 This technique has not been clinically useful, however,
since it is cumbersome to perform and because Vd/Vt can be
affected by other variables including cardiac output, posture,
and ventilatory pattern. 92 ' 93

Although pulmonary emboli make elimination of C0 2 by
the lungs less efficient, hypercapnia and respiratory acidosis
are rarely seen in patients with PE. 94 Almost all patients with
PE develop compensatory hyperventilation sufficient to
produce hypocapnia and respiratory alkalosis. 95

Hypoxemia

Systemic arterial hypoxemia is the earliest and most frequent
manifestation of acute pulmonary thromboembolism. 81 In
two series of patients with documented PE, 95% of the patients
had abnormal alveolar-arterial oxygen gradients, and of
patients entered into the Urokinase Pulmonary Embolism
Trial, nearly 90% had an arterial p0 2 (Pa0 2 ) less than
80 mmHg. 96 Many patients with PE have mild to moderate
hypoxemia so only one-third demonstrate a Pa0 2 less than
60 mmHg. 97 In acute PE, the most important physiologic
mechanisms of hypoxemia are intrapulmonary and intra-

202

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

cardiac shunts. Other contributing mechanisms include
ventilation-perfusion inequality, diffusion impairment, and
mixed venous hypoxemia.

Intrapulmonary right to left shunt has been documented as
the principal cause of hypoxemia in two patients with acute
massive PE. 98 Intrapulmonary shunting may occur due to
perfusion of lung that is unventilated because of atelectasis or
pulmonary edema. Acute hypoperfusion of a lung segment
causes regional hypocapnia, which induces bronchiolar con-
striction and atelectasis so that the area around the embolus
may be perfused but not ventilated. Serotonin release from
platelets trapped within the embolus has also been implicated
as a cause of atelectasis by causing airway constriction. 99 ' 100
Pulmonary edema has been shown to occur after pulmonary
embolization in several animal models but this has been more
difficult to demonstrate in humans. 101 ' 102 Shunting through a
newly opened intrapulmonary arteriovenous anastomosis
following acute pulmonary hypertension associated with PE
has been proposed as another possible mechanism of intra-
pulmonary right to left shunt, but could not be demonstrated
in experimental animal studies. 103

Elevated right heart pressures following massive PE may
produce intracardiac shunting through an atrial septal
defect in some patients. It is estimated that 15% of patients
have a patent foramen ovale. The presence of intracardiac
right to left shunting has been reported in two patients
with large alveolar-arterial gradients following massive
pulmonary embolus. 104

Control of ventilation

One of the most common clinical findings in patients with
acute pulmonary thromboembolism is hyperventilation. The
increased minute ventilation usually leads to hypocapnia
and respiratory alkalosis. Normal ventilation is controlled
through a complex interaction between central and peripheral
chemoreceptors, the cerebral cortex, hypothalamus, pons, and
proprioceptors located within the lung. Correction of hypox-
emia with supplemental oxygen generally does not reverse
the hyperventilation and respiratory alkalosis. Although the
actual mechanisms leading to hyperventilation following PE
are undefined, limited data suggest that juxtacapillary sensors
and irritant receptors contribute to the reflex stimulation of
ventilation by pulmonary emboli. 105 Irritant receptors have
been activated by experimental embolization and can initiate
tachypnea, hyperventilation, and reflex bronchoconstriction
in rabbits. However, human data are lacking. 106

Pulmonary mechanics

While the effects of acute PE on pulmonary mechanics have
not been studied adequately in humans, a number of clinical
and experimental observations suggest airway resistance in-
creases and lung compliance decreases following pulmonary

emboli. 83 ' 107-109 Clinicians have frequently described bron-
chospasm and wheezing in patients acutely following
pulmonary emboli. 110

Atelectasis or elevated hemidiaphragm on chest radio-
graph are frequently found in patents with pulmonary
emboli, suggesting decreased lung volume in these patients. 96

Pulmonary infarction

Pulmonary infarction is a relatively uncommon occurrence
following pulmonary thromboembolism. Unlike other tis-
sues, the pulmonary parenchyma derives oxygen from three
sources: (i) the alveoli where oxygen tension is approximately
HOmmHg; (ii) the bronchial arteries where oxygen tension is
at the systemic arterial level; and (iii) the pulmonary artery
where oxygen tension is approximately 40 mmHg. In PE, the
pulmonary artery source of oxygen is lost but the more impor-
tant sources remain and infarction usually does not occur.
Pulmonary infarction associated with PE most commonly
occurs when there is occlusion of distal pulmonary arteries in
association with left ventricular heart failure. 110-112 Dalen and
coworkers suggested in an embolized segment of lung,
bronchial arterial blood enters the pulmonary capillary via
anastomotic channels and extravasates into the alveolus. 110 In
patients with left ventricular heart failure, pulmonary venous
hypertension makes this extravasation worse, leading to
pulmonary hemorrhage and ultimately infarction. 110 ' 113

Conclusion

Pulmonary thromboembolism results in a number of
pathophysiologic disturbances involving cardiopulmonary
hemodynamics, gas exchange, pulmonary mechanics, and
ventilatory control. These effects are mediated by a complex
interaction between the direct occlusion of the pulmonary
vasculature, the release of vasoactive and bronchoactive
substances, and several cardiopulmonary reflexes.

References

1. Virchow R. Neuer Fall von todlicher Emboli der Lungenarterie.
Arch Pathol Anat 1856; 10:225.

2. Houdijk WPM, Sakariassen KS, Nievelstein PFEM, Sixma JJ.
Role of factor VHI-von Willebrand factor and fibronectin in the
interaction of platelets in flowing blood with monomeric and
fibrillar collagen types I and III. / Clin Invest 1985; 75:531 .

3. Ginsberg MH, Loftus JC, Plow EE Cytoadhesions, integrins and
platelets. Thromb Haemost 1988; 59:1.

4. Lollar P, Owen WG. Clearance of thrombin from circulation
in rabbits by high affinity-binding sites on endothelium. / Clin
Invest 1980; 66:1222.

5. Harker LA, Schwartz SM, Ross R. Endothelium and arterioscle-
rosis. Clin Haematol 1981; 10:283.

203

pa rt I Vascular pathology and physiology

6. Moncade S, Vane JR. Unstable metabolites of arachidonic acid
and their role in haemostasis and thrombosis. Br Med Bull 1978;
34:129.

7. Lister J. On the coagulation of blood. Proc R Soc 1863; 12:560.

8. Mayo M, Halil T, Browse NL. The incidence of deep vein throm-
bosis after prostatectomy. Br J Urol 1971; 43:738.

9. Nicholaides AN, Field ES, Kakkar VV et al. Prostatectomy and
deep vein thrombosis. Br] Surg 1972; 59:487.

10. Walsh JJ, Bonnar J, Wright FW. A study of pulmonary embolism
and deep vein leg vein thrombosis after major gynaecologic
surgery using labelled fibrinogen-phlebography and lung scan-
ning. / Obstet Gynaecol Br Commonw 1974; 81:311.

11. Warlow C, Ogston D, Douglas AS. Deep venous thrombosis of
the legs after strokes. Br Med] 1976; 2:1178.

12. Bors E, Conrad CA, Massell TB. Venous occlusion of lower ex-
tremities in paraplegic patients. Surg Gynecol Obstet 1954; 99:451.

13. Simmons AV, Sheppard MA, Cox AF. Deep venous thrombosis
after myocardial infarction: predisposing factor. Br Heart } 1973;
35:623.

14. Comerota AJ, Sterwart GJ, Alburger PD, Smalley K, White JV
Operative venodilation: a previously unsuspected factor in
the cause of postoperative deep vein thrombosis. Surgery 1989;
106:310.

15. Stewart GJ, Alburger PD, Stone EA, Suszka TW. Total hip
replacement induces injury to remote veins in a canine model.
/ Bone Joint Surg 1983; 65 A:97.

16. Thomas DP, Merton RE, Wood RD, Hockley DJ. The relationship
between vessel wall injury and venous thrombosis: an experi-
mental study. Br } Haematol 1985; 59:449.

17. Shingu M, Hurd ER. Sera from patients with systemic lupus ery-
thematosus reactive with human endothelial cells. / Rheumatol
1981; 8:851.

18. Asch AS, Barnwell J, Silverstien AL, Nachman RL. Isolation of
the thrombospondin membrane receptor. / Clin Invest 1987;
79:1054.

19. Rosenber RD, Lam LH. Correlation between structure and func-
tion of heparin. Proc Natl Acad Sci USA 1979; 76:1218.

20. Poort SR, Rosendaal FR, Reitsma PH et al. The common genetic
variation in the 3' untranslated region of the prothrombin gene is
associated with elevated prothrombin levels and an increase in
venous thrombosis. Blood 1996; 88:3698.

21. Jorgensen M, Bonnevie-Neilson V Increased concentration of
fast acting plasminogen activator inhibitor in plasma associated
with familial venous thrombosis. Br J Haematol 1987; 65:175.

22. Egeberg D. Inherited fibrinogen abnormality causing throm-
bophilia. Thromb Haemost 1967; 17:176.

23. Stamatakis JD, Kakkar VV, Sagar S et al. Femoral vein thrombosis
and total hip replacement. Br Med J 1977; 2:213.

24. Hull R, Hirsh J, Sackett DL et al. The value of adding impedance
plethysmography to 125 I-fibrinogen leg scanning for the detec-
tion of deep vein thrombosis in high risk surgical patients: a com-
parative study between patients undergoing general surgery
and hip surgery. Thromb Res 1979; 15:227.

25. Borgstrom S, Gelin LE, Zenderfeldx B. The formation of
vein thrombi following tissue injury. Acta Chir Scand 1959;
247(Suppl.):l.

26. Bjorklid E, Gierksy KE, Prydz H. An immunoradiometric
assay for factor III (tissue thromboplastin). Br } Haematol 1978;
39:445.

27. Gordon-Smith IC, Hickman JA, LeQuesne LP. Post-operative
fibrinolytic activity and deep vein thrombosis. Br J Surg 1974;
61:213.

28. Vessey MP, Mann JI. Female sex hormones and thrombosis.
Epidemiological aspects. Br Med Bull 1978; 35:157.

29. Bottiger LE, Boman G, Eklund G, Westerholm B. Oral contracep-
tives and thromboembolic disease: effects of lowering estrogen
content. Lancet 1980; 1:1097.

30. Bonnar J, Prentice CR, McNichol GP, Douglas AS. Haemostatic
mechanism in the uterine circulation during placental separa-
tion. Br Med J 1970; 1 :564.

31. Kakkar VV, Howe CT, Nicholaides AN et al. Deep vein throm-
bosis of the leg: is there a "high risk" group? Am } Surg 1970;
120:527.

32. Burkitt DP. Varicose veins: facts and fantasy. Arch Surg 1976;
111:1327.

33. Blumoff RL, Johnson G. Saphenous vein p0 2 in patients with
varicose veins. / Surg Res 1977; 23:35.

34. Reikaras O, Sorlie D. The significance of arteriovenous shunting
for the development of varicose veins. Acta Chir Scand 1983;
149:479.

35. Baron HC, Cassaro S. The role of arteriovenous shunts in the
pathogenesis of varicose veins. / Vase Surg 1986; 4:124.

36. Sarin S, Scurr JH, Coleridge-Smith PD. Medial calf perforators in
venous disease: the significance of outward flow. / Vase Surg
1992; 16:40.

37. Rose SS, Ahmed A. Some thoughts on the aetiology of varicose
veins. / Cardiovasc Surg 1986; 27:534.

38. O'Donnell TF. The surgical management of deep venous valvu-
lar incompetence. In: Rutherford RB, ed. Vascular Surgery, 3rd
edn. Philadelphia: WB Saunders, 1989:1612.

39. Browse NL, Burnand KG, Lea-Thomas M, eds. Diseases of
the Veins: Pathology, Diagnosis and Treatment. London: Edward
Arnold, 1988:411.

40. Sumner D, Strandness DE, eds. Hemodynamics for Surgeons. New
York: Grune & Stratton, 1975:120.

41. Basmajian JV. The distribution of valves in the femoral, external
iliac and common iliac veins and their relationship to varicose
veins. Surg Gynecol Obstet 1952; 95:537.

42. O'Donnell TF, Mackey WC, Shepard AD. Clinical, hemo-
dynamic and anatomic follow-up of direct venous reconstruc-
tion. Arch Surg 1987; 122:474.

43. van Bemmelen AN, van Bemmelen PS, Bedford G, Beach K,
Strandness DE. Quantification of venous valvular reflux with
duplex ultrasound scanning. / Vase Surg 1989; 10:425.

44. Arnoldi CC. Venous pressure in patients with valvular in-
competence of the veins of the lower limb. Acta Chir Scand 1966;
132:628.

45. Walker AJ, Longland CJ. Venous pressure measurements in the
foot in exercise as an aid to investigation of venous disease of the
leg. Clin Sci 1950; 9:101.

46. Warren R, White EA, Belcher CD. Venous pressures in the saphe-
nous system in normal and postphlebitic limbs: alternatives fol-
lowing femoral vein ligation. Surgery 1949; 26:435.

47. Burnand KG, O'Donnell TF, Lea-Thomas M, Browse NL. The
relative importance of incompetent communicating veins in the
production of varicose veins and venous ulcers. Surgery 1977;
82:9.

48. Christopolous DV, Nicholaides AN, Szendro G et al. Air-

204

chapter 17 Pathophysiology, hemodynamics, and complications of venous disease

plethysmography and the effect of elastic compression on leg
vein hemodynamics. / Vase Surg 1987; 5:148.

49. Porter JM, Rutherford RB, Claggett GP, Cranley JJ, O'Donnell TF.
Reporting standards in venous disease. / Vase Surg 1988; 8:172.

50. O'Donnell TF. Chronic venous insufficiency: an overview of
epidemiology classification and anatomic considerations. Semin
Vase Surg 1988; 1:60.

51 . McEnroe CS, O'Donnell TF, Mackey WC. Correlations of clinical
findings with venous hemodynamics in 386 patients with
chronic venous insufficiency. Am } Surg 1986; 156:148.

52. Darke SG, Andress MR. The value of venography in the manage-
ment of chronic venous disorders of the lower limb. In:
Greenhalgh RM, ed. Diagnostic Techniques and Assessment Proce-
dures in Vascular Surgery. London: Grune & Stratton, 1985.

53. Bauer G. The etiology of leg ulcers and treatment by resection of
the popliteal veins. JInt Chir 1948; 8:937.

54. Kistner RL. Surgical repair of the incompetent femoral vein
valve. Arch Surg 1975; 110:13369.

55. Kistner RL. Primary venous valve incompetence of the leg. Am J
Surg 1980; 140:218.

56. Shull KC, Nicolaides AN, Fernandes JF et al. Significance of
popliteal reflux in relation to ambulatory venous pressure and
ulceration. Arch Surg 1979; 114:13046.

57. O'Donnell TF, Browse NL, Burnand KG et al. The socioeconomic
effects of an iliofemoral venous thrombosis. / Surg Res 1977;
22:483.

58. Markal A, Manzo RA, Bergelin RO, Strandness DE. Valvular re-
flux after deep vein thrombosis: incidence and time of occur-
rence. / Vase Surg 1992; 15:377.

59. Killewich LA, Bedford GR, Beach KW, Strandness DE. Sponta-
neous lysis of deep venous thrombotic rate and outcome. / Vase
Surg 1989; 9:89.

60. Bjordal RL Circulation patterns in incompetent perforating veins
in the calf and in the saphenous system in prominent varicose
veins. Acta Chir Scand 1972; 138:251.

61. Linton RR. The communicating veins of the lower leg and the
operative technique for their ligation. Ann Surg 1938; 107:582.

62. Cockett FB, Jones BE. The ankle blow-out syndrome: a new
approach to the varicose ulcer problem. Lancet 1953; 1:17.

63. Burnard KG, O'Donnell TF, Lea Thomas M et al. The relationship
between post-phlebitic changes in the deep veins and the results
of surgical treatment of venous ulcers. Lancet 1976; 1:936.

64. Hanrahan LW, Araki CT, Rodriquez AA et al. Distribution of
patients with venous stasis ulceration. / Vase Surg 1991; 13:805.

65. Hanrahan LW, Aracki CT, Fisher JB et al. Evaluation of the perfo-
rating veins of the lower extremity using high resolution duplex
imaging. / Cardiovasc Surg 1991; 32:87.

66. Sarin S, Scurr JH, Coleridge-Smith PD. Medial calf perforators in
venous disease: the significance of outward flow. / Vase Surg
1992; 16:40.

67. Pratt GH Arterial varices. A syndrome. Am J Surg 1949; 77:456.

68. Lindmayr W, Lofferer O, Mostbeck A, Partsch H. Arteriove-
nous shunts in primary varicosis: a critical essay. Vase Surg 1972;
6:9.

69. Burnand KG, Whimster IW, Clemenson G et al. The relationship
between the number of capillaries in the skin of the venous ulcer
bearing area of the lower leg and the fall in foot vein pressure
during exercise. Br J Surg 1981; 68:297.

70. Mani R, Gorman FW, White JE. Transcutaneous measurements of

oxygen tension at levels at edges of leg ulcers: preliminary com-
munication. / R Soc Med 1982; 79:650.

71. Travers JP, Barridge DC, Makin GS. Surgical enhancement of
skin oxygenation in patients with venous lipodermatosclerosis.
Phlebology 1990; 5:129.

72. Michel TC. Oxygen diffusion in edematous tissue and through
pericapillary cuffs. Phlebology 1990; 5:223.

73. Hopkins NFG, Sphinx TJ, Rhodes CG et al. Positron emission to-
mography in venous ulcerations in liposclerosis. V R Med } 1982;
286:333.

74. Browse NL, Jarrett PEM, Morland M, Burnand KG. Treatment of
liposclerosis of the leg by fibrinolytic enhancement: a prelimi-
nary report. Br Med J 1977; 2:434.

75. Layer GT, Stacey MC, Burnand KG. Stanozolol in the treatment
of venous ulceration: an interim report. Phlebology 1986; 1:197.

76. McMullen GM, Watkin GT, Coleridge-Smith PD, Scurr JH. The
efficacy of fibrinolytic enhancement with stanozolol in the treat-
ment of venous insufficiency. Br] Surg 1990; 77:A700.

77. Moyses C, Cedarholm-Williams SA, Michel CC. Hemoconcen-
tration and the accumulation of white cells in the feet during
venous stasis. Int J Microcirc Clin Exp 1987; 5:311.

78. Thomas PRS, Nash GB, Dorma DGA. White cell accumulation
in the dependent legs of patients with venous hypertension: a
possible mechanism for trophic changes in the skin. Br Med }
1988; 296:16935.

79. Coleridge-Smith PD, Thomas P, Scurr JH, Dormandy JA. Causes
of venous ulceration: a new hypothesis. Br Med } 1988; 296:17262.

80. Dalen JE, Alpert JS. Natural history of pulmonary embolism.
Prog Cardiovasc Dis 1975; 17:259.

81. Mclntyre KM, Sasahara AA. The hemodynamic response to pul-
monary embolism in patients without prior cardiopulmonary
disease. Am J Cardiol 1971; 28:288.

82. Sharma G, Mclntyre KM, Sharma S et al. Clinical and hemo-
dynamic correlates in pulmonary embolism. Clin Chest Med 1984;
5:421.

83. Comroe JH Jr, VanLingen B, Stroud RC et al. Reflex and direct
cardiopulmonary effects of 5-OH-tryptamine(serotonin): their
possible role in pulmonary embolism and coronary thrombosis.
Am} Physiol 1953; 173:379.

84. Thomas DP, Gurewich V, Ashford TP Platelet adherence to
thromboemboli in relation to the pathogenesis and treatment of
pulmonary embolism. N Engl J Med 1966; 274:953.

85. McGoon MD, Vanhoutte PM, Stemp LI et al. Aggregating
platelets contract isolated canine pulmonary arteries by releas-
ing 5-hydroxytryptamine. / Clin Invest 1984; 74:828.

86. Huval WV, Mathieson MA, Stemp LI et al. Therapeutic benefits
of 5-hydroxytryptamine inhibition following pulmonary em-
bolism. Ann Surg 1983; 197:200.

87. Huet Y, Brun-Buisson C, Lemaire F et al. Cardiopulmonary
effects of ketanserin infusion in human pulmonary embolism.
Am Rev Respir Dis 1987; 135:114.

88. Manier G, Castaing Y, Guenard H. Determinants of hypoxemia
during the acute phase of pulmonary embolism in humans. Am
Rev Respir Dis 1985; 132:332.

89. Brooks H, Kirk ES, Pantel SV et al. Performance of the right
ventricle under stress: relation to right coronary flow. Circ Res
1954; 11:326.

90. Enghoff H. Volumen inefficax bemerkungen zur frage des
schadlichen Raumes Uppsala. LakarefFork 1938; 44:191.

205

pa rt I Vascular pathology and physiology

91. Burki NK. The dead space to tidal volume ratio in the diagnosis
of pulmonary embolism. Am Rev Respir Dis 1986; 133:679.

92. Riley RL, Permutt S, Said S et ah Effect of posture on pulmonary
dead space in man. / Appl Physiol 1959; 14:339.

93. Baker R, Burki NK. The effects of alternations in ventilatory pat-
tern on the ratio of dead space to tidal volume. Chest 1982; 82:243.

94. Bouchama A, Curley W, Al-Dossary S et ah Refractory hyper-
capnea complicating massive pulmonary embolism. Am Rev
Respir Dis 1988; 138:466.

95. Citranic D, Merino PL. Improved use of arterial blood gas
analysis in suspected pulmonary embolism. Chest 1989; 95:48.

96. The Urokinase Pulmonary Embolism Trial: a cooperative study.
Circulation 1973;47(Suppl. II):1.

97. Dantzker DR, Bower JS. Clinical significance of pulmonary
function tests: alterations in gas exchange following pulmonary
thromboembolism. Chest 1982; 81:495.

98. D'Alonzo GE, Bower JS, DeHart P et ah The mechanisms of
abnormal gas exchange in acute massive pulmonary embolism.
Am Rev Respir Dis 1983; 128:170.

99. Levy SE, Simmons DH. Mechanisms of arterial hypoxemia
following pulmonary thromboembolism in dogs. / Appl Physiol
1975; 39:41.

100. Levy SE. Simmons DH. Redistribution of alveolar ventilation
following pulmonary thromboembolism in the dog. / Appl Phy-
siol 1974; 36:60.

101. Weisberg H, Lopez JF, Luria MH et ah Persistence of lung edema
and arterial pressure rise in dogs after lung emboli. Am } Physiol
1964; 207:641.

102. Malik AB, VanderZee H. Mechanism of pulmonary edema
induced by microembolization in dogs. Circ Res 1978; 42:72.

103. Cheney FW, Pavlin J, Ferens BS et ah Effect of pulmonary
microembolism on arteriovenous shunt flow. / Thorac Cardiovasc
Surg 1978; 4:473.

104. Herre PH, Petitperez P, Simonneau G et ah The mechanisms of ab-
normal gas exchange in acute massive pulmonary embolism. Am
Rev Respir Dis 1983; 128:1101.

105. Kelley MA, Fishman AP Pulmonary thromboembolic disease.
In: Fishman AP, ed. Pulmonary Diseases and Disorders. New York:
McGraw-Hill, 1988:1062.

106. Mills JE, Sellick H, Widdicombe JG. Activity of lung irritant
receptors in pulmonary microembolism, anaphylaxis and drug-
induced bronchoconstrictions. ] Physiol 1969; 203:337.

107. Gurewich V, Thomas DP, Stein M et ah Bronchoconstriction in the
presence of pulmonary embolism. Circulation 1963; 27:339.

108. Thomas DP, Stein M, Tonabe G et ah Mechanism of bronchocon-
striction produced by thromboemboli in dogs. Am J Physiol 1964;
206:1207.

109. Cahill JM, Attinger EO, Byrne JJ. Ventilatory responses to
embolization of lung. J Appl Physiol 1961; 16:469.

110. Dalen JE, Haffajee CI, Alpert JS et ah Pulmonary embolism, pul-
monary hemorrhage and pulmonary infarction. N Engl } Med
1977; 296:1431.

111. Tsao MS, Schraufnagel DE, Wang NS. Pathogenesis of pul-
monary infarction. Am J Med 1982; 72:599.

112. Schraufnagel DE, Ming-Soung T, Yao YT et ah Factors associated
with pulmonary infarction. Am } Clin Pathol 1985; 84:15.

113. Ellis FH Jr, Grindlay JH, Edwards JE. The bronchial arteries. II.
Their role in pulmonary embolism and infarction. Surgery 1952;
31:167.

206

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Physiologic changes in lymphatic
dysfunction

Peter Gloviczki

Failure of the lymphatic system to transport lymph from the
interstitial space back to the bloodstream results in lymphatic
stasis. If the collateral lymphatic circulation is insufficient and
all compensatory mechanisms are exhausted, the protein-rich
interstitial fluid accumulates and lymphedema develops. In
lymphedema, caused by either congenital or acquired dys-
function of the lymphatic system, the microcirculation in the
affected area of the body is disrupted. The transport of the
excess tissue fluid containing lymphocytes, different plasma
proteins, immunoglobulins, and cytokines is impaired
and chronic inflammatory changes in the subcutaneous tis-
sue and skin develop. Progress in ultrastructural, cytochemi-
cal, and imaging studies and improvement in conservative
and surgical treatment of lymphedema have stimulated
substantial interest in lymphatic disease.

Historical background

Lymph vessels were mentioned more than 2000 years ago by
Aristotle, who described "nerves which contain colorless
liquids" and later by members of the Alexandrian School of
Medicine, who recognized "arteries" in the mesentery "full of
milk." This knowledge, however, was lost during the Middle
Ages, and it was only in the Renaissance that attention was fo-
cused again on the lymphatic system. The thoracic duct was
observed in 1563 by Eustachius, who called it "vena alba
thoracis." He failed, however, to recognize the function of the
thoracic duct and its relation to the lymphatic system. The
discovery of the lymphatics is attributed to the Italian
anatomist Gasparo Aselli, who in 1622 observed the mesen-
teric lymphatics in a well-fed dog. He also recognized the
function of the lacteals, although he suggested mistakenly that
the chyle absorbed from the intestine by the mesenteric lym-
phatics was transported to the liver. In 1651, Pecquet described
the thoracic duct and recognized the correct route of lym-
phatic transport from the mesenteric lymphatics through the
"receptaculum chyli" and the thoracic duct to the subclavian
vein. Further details on the anatomy of the lymphatic system

were published in the 17th century by Bartholin and Rudbeck,
and by the great anatomists of the 18th century, Mascagni and
Cruikshank. It was most likely William Hunter who recog-
nized the lymphatics as a separate system responsible for
absorption.

Although Hunter suggested that the lymphatics were
closed tubes, one of his students, Hewson, recognized that
they had physiologic orifices, which, like "capillary tubes"
sucked up tissue fluid. It was not until the turn of the 20th
century, however, that Starling confirmed the relationship
between the oncotic pressure of the plasma proteins and the
hydrostatic pressure in the capillaries. 1 ' 2 Starling suggested
that lymph formed by filtration of the blood through the capil-
lary walls. Drinker, 3 and later Rusznyak and colleagues, 4 de-
serve the credit for clarifying the details of protein absorption
from the intercellular space via the lymphatic system. Interest
in lymphatic diseases was greatly enhanced by Kinmonth,
who described a clinically usable technique of direct contrast
lymphangiography in 1952. 5 Improvement in other imaging
techniques, such as lymphoscintigraphy, 6 ' 7 indirect lymphan-
giography, 8 ' 9 and magnetic resonance imaging, 10-12 furthered
the understanding of the structure and function of the lym-
phatic system in different lymphatic disorders. Progress in
conservative management 13 ' 14 and development of micro-
surgical operations on the lymph vessels 15-17 also have stimu-
lated experimental and clinical research in lymphatic diseases.

Development of the lymphatic system

The lymphatic system is first apparent in the human fetus at 6
weeks of gestation, and it consists of paired jugular, iliac, and
retroperitoneal lymph sacs (Fig. 18. 1). 18 The origin of the lym-
phatic system is controversial, but it is most likely a derivative
from the venous system. Another possible theory is that it
develops independently of the veins from the mesenchymal
tissue. The lymph vessels grow from the paired primitive
lymphatic sacs and coalesce along the major veins to form the
afferent vessels, nodes, and efferent lymphatic ducts. The

207

pa rt I Vascular pathology and physiology

Anterior
7 weeks

Posterior
7 weeks

Right lymphatic duct

Internal jugular vein

Subclavian vein

Superior vena cava

Anastomosis

Thoracic duct

Left and right thoracic ducts

Cisterna chyJi

Retroperitoneal lymph sac

Iliac (ymph sac

B C

Figure 18.1 Development of the lymphatic system. (A) Seven-week embryo
with paired iliac, retroperitoneal, and jugular lymph sacs. (B) At 9 weeks of
gestation, paired thoracic ducts are present with numerous connections
across the midline. (C) Portions of both primitive thoracic ducts persist to
form the thoracic duct in the adult. The right lymphatic duct is formed from
the primitive right jugular lymphatic sac. (From Cambria RA, Gloviczki P.
Lymphedema: pathophysiology and management. In: Callow AD, ed.
Vascular Surgery. Norwalk, CT Appleton&Lange, 1995:1593.)

cisterna chyli develops from one of the large retroperitoneal
lymph sacs, whereas the other forms the mesenteric lymphat-
ic system. There are paired thoracic ducts in the embryo, and
the mature thoracic duct develops from fusion of the upper
portion of the left and the lower portion of the right thoracic
duct. The right cervical lymphatic duct is formed by the right
jugular lymphatic sac. This receives lymph from the right face,
neck, and the right upper extremity, and from the upper part of

the right thorax and mediastinum. Abnormalities in the devel-
opment of the lymphatic system include agenesis, hypoplasia,
or hyperplasia of the lymphatics with valvular incompetence.
They may result in lymphedema or in abnormalities in the
circulation of the chyle, such as chylous ascites, chylothorax,
reflux of chyle to the pelvis or lower extremities, or protein-
losing enteropathy. Persistence of some of the embryonic
sacs may cause the development of lymphatic cysts, which
may or may not communicate with the lymphatic system.

Anatomy of the lymphatic system

The adult lymphatic system consists of peripheral lymph ves-
sel, lymph nodes, and major lymphatic trunks. The peripheral
lymph vessels collect lymph from the lymphatic capillaries,
which absorb a portion of the interstitial fluid from the inter-
stitial space. Afferent lymph channels transport lymph to the
lymph-conducting elements of the lymph nodes, which filter
and further conduct the lymph fluid to efferent lymphatic
channels. Significant communications between the lymphatic
and venous system in lymph nodes normally do not exist.

Eighty percent of the lower extremity lymph is carried by
the superficial lymphatic system. Although there is a lateral
superficial bundle located around the lesser saphenous vein,
most of the lower extremity lymph is transported by lymph
channels of the superficial medial bundle (Fig. 18.2). There is a
deep lymphatic network that runs in close proximity to the
tibial and peroneal vessels and transports lymph through
the popliteal lymph nodes into the deep femoral lymphatics.
The superficial and the deep lymphatics join in the inguinal
lymph nodes and drain lymph toward the aortoiliac lym-
phatic system. The cisterna chyli is located between the aorta
and the inferior vena cava, usually at the level of LI to L2.
Mesenteric lymphatics join the lower extremity and pelvic
lymphatics at this level and drain through the thoracic duct to
the left subclavian vein (Fig. 18.3). A very small amount of
mesenteric lymph is drained toward the liver around the he-
patic vein and the diaphragm to the mediastinal lymphatics.

The upper extremity lymphatics run along the major veins
of the arm. Although the medial arm bundle is the most signif-
icant route of lymph drainage in normal patients, after axillary
node dissection lymph is drained primarily through the
lateral lymphatic bundle to the deltoideopectoral and supra-
clavicular nodes (Fig. 18.4).

A single layer of endothelial cells forms the inner layer of the
lymphatic capillaries. Basal membranes similar to blood capil-
laries are not present. The lymphatic capillaries contain
bicuspid lymphatic valves, which play a crucial role in the
initial lymphatic transport and are responsible for the unidi-
rectional lymphatic flow. The capillaries are anchored by small
microfibrils that expand the endothelial cells and increase the
lumen of the capillaries if the tissue pressure is elevated. 19,20
Although smaller molecules may traverse the lymphatic

208

CHAPTER 18 Physiologic changes in lymphatic dysfunction

Superficial

lateral

bundle

Inquinal

lymph

nodes

'Superficial
medial
bundle

■ Popliteal

lymph

nodes

Posterior
tibial and
peroneal
lymphatics

Figure 18.2 Anatomy of major lymph vessels and lymph nodes of the lower
extremity. (From Gloviczki P. Microsurgical treatment for chronic
lymphedema: an unfulfilled promise? In: Bergan JJ, Yao JST, eds.
Venous Disorders. Philadelphia: WB Saunders, 1990:344.)

endothelial cells with active phagocytosis, large molecules
enter through the gaps between the endothelial cells of the
lymphatic capillary

Lymphatic physiology

According to Starling's law, hydrostatic and osmotic pressures
in the capillaries and in the interstitial space determine the
amount of interstitial fluid that is ultrafiltered from the blood
plasma. Additional factors responsible for interstitial fluid
exchange include capillary permeability, the number of active
capillaries, the ratio of precapillary arteriolar to postcapillary
venular resistance, and the total extracellular fluid volume.
The amount of fluid that moves across the capillary wall is
tremendous, considering that the cardiac output is about
8000 1 during a 24-h period. It is likely that an amount equal to
the total plasma volume enters the interstitial place and leaves
through the venous end of the capillaries and the lymphatics
every minute. 21 The lymphatic system is responsible for the
transport of 2-4 1 of interstitial fluid daily. During the same
time, approximately 100 g of plasma protein is carried back to
the circulation by the lymphatics. 22 The protein content of the
lymph is somewhat less than that of the plasma, and lymph
vessels from various parts of the body contain different
amounts of protein (Table 18.1). The lymphatic capillaries are
able to transport large molecules, even those with a molecular
weight over 1 kDa. 23

Figure 18.3 Anatomy of the thoracic duct.
(From Gloviczki P, Noel AA. Lymphatic
reconstructions. In: Rutherford RB, ed. Vascular
Surgery, 5th edn. Philadelphia: WB Saunders,
2000:2 1 59, with permission from Elsevier.)

Carotid Av Va ? usN

Left subclavian A

Vertebral V

Thoracic duct
termination

Thoracic duct
Cisterna chyli

Left subclavian V

209

pa rt I Vascular pathology and physiology

Supraclavicular nodes

Deltoideopectoral node.

^~F~

-mx^Z^z^ Infraclavicular

1 -atoral Hiinril^, ' ** ***^

n^^ nodes
^Axillary nodes

LdlCldl UUilUltfi

SI f t f If

*"^ i 1 1 1 II
I / 1 1 II

I I ' f li\
I f ' f ft m

Medial upper J~JMJ f[ ft\

arm bundle / JYIm
} /i iff

I S i 1*1

i S I I * I

i X 1 i > 1

is til/

r tiff
I If t It ,

Medial / J 1 \f

bundle v / J // y j

JyCj if j— Deep lymphatic

if /// / system

/ / / / /

/ . Au

Inarl

DundJe

actin and are able to contract actively, contraction of terminal
lymphatics with the help of competent valves enables rapid
lymphatic transport. Intrinsic contractility of the smooth
muscle in larger collecting vessels allows further propulsion of
the lymph. Strength and frequency of the contractions are
greatly influenced by changes in intraluminal pressure. 24
Adrenergic stimulation 25 and endothelin 26 also have been
shown to result in contraction of the lymph vessels. Patent
blue dye injected into the subcutaneous tissue is transported
centrally in the lymph vessels at the rate of 4-5mm/s, even
without any muscular exercise. Intrinsic contractions of the
lymph vessel wall with competent valves are able to propel
lymph intermittently against a pressure as high as 50 mmHg.

The major difference that distinguishes the lymphatic sys-
tem from the venous system is that the veins are filled with a
continuous liquid column. The lymphatic system, however, is
not fully // primed ,, / and only if there is longstanding stasis
does the lymph column fill the lymphatic channels complete-
ly. 23 It is only in these conditions that muscular contraction or
external massage play an important role in forward propul-
sion of the lymph and facilitate lymphatic transport.

Figure 18.4 Anatomy of major lymph vessels and lymph nodes of the upper
extremity. (From Gloviczki P. Microsurgical treatment for chronic
lymphedema: an unfulfilled promise? In: Bergan JJ, Yao JST, eds.
Venous Disorders. Philadelphia: WB Saunders, 1990:344.)

Table 18.1 Approxiamte protein content of lymph in humans*

Lymph origin

Protein content (g/dl)

Ankle

0.5

Limbs

Intestine

Liver

Thoracic duct

*Data based on various studies in humans and animals.

(From Ganong WF. Review of Medical Physiology, 1 0th edn. Los

Altos, CA: Lange Medical Publications, 1 981 : 452.)

The single most important determinant of lymph flow
through the lymphatic capillaries and the collecting lymph
vessels is the intrinsic contractility of the lymph vessels. In
addition, lymph flow in influenced by increased interstitial
pressure, muscular activity, arterial pulsation, respiratory
pressure, and gravity. Increase in interstitial volume and inter-
stitial pressure results in opening of the gaps between the
endothelial cells of the terminal lymphatics and an increase in
lymphatic transport. Because the endothelial cells contain

Pathophysiology of lymphedema

Lymphedema develops when the lymphatic load exceeds the
transport capacity of the lymphatic system. In patients with
lymphatic obstruction, numerous compensatory mechanisms
develop. These include collateral lymphatic circulation, de-
velopment of spontaneous lymphovenous anastomoses, and
increased activity of tissue macrophages to split macro-
molecules in the interstitial space, enabling them to be reab-
sorbed through the venous end of the capillaries (Fig. 18.5).

If the lymphatic transport is impaired due to injury or ob-
struction to the lymph vessels and lymph nodes, the different
compensatory mechanisms can function effectively for a
period of time. This explains why chronic lymphedema of the
limbs may develop several months or even years after an
edema-free state after inguinal or axillary node dissection or
irradiation.

Lymphedema is a high-protein edema that, except very
early in the course of the disease, is nonpitting in nature (Fig.
18.6). Without treatment, the high-protein edema fluid in the
subcutaneous tissue will be replaced by fibrous material, in-
flammatory cells accumulate, and progressive fibrosis of the
subcutaneous tissue and skin develops. Fibrosis of the lymph
vessels leads to loss of permeability and loss of intrinsic con-
tractility. Dilation of the lymph vessels causes valvular incom-
petence, and the inflammatory and fibrotic changes destroy
the valve leaflets, further decreasing the transport capacity of
the lymphatic system. Microsurgical reconstruction in this
late stage of lymphedema, using fibrotic and incompetent
lymphatics, cannot restore normal lymphatic transport.

Progression of lymphedema results in fibrotic obstruction of

210

Figure 18.5 Stages in development of
postsurgical lymphedema. (Modified from
Gloviczki P, Schirger A. Lymphedema. In: Spittell
JA, ed. Clinical Medicine. Philadelphia: Harper &
Row, 1985:1.)

Excision of Lymph Vessels and Nodes

Rapid regeneration of lymphatics,
restored lymph drainage

No acute edema

Inadequate

Acute edema (4-6 weeks)

No chronic edema

Latent edema

(8 months to years)

Compensated [ymph circulation

Growing lymph stasis

Fibrosis of lymphatics with

loss of permeability and loss

of intrinsic contractility

Dilation of lymph vessels
with valve incompetency

Incompetency of endothelial
junctions in lymph capillaries

High lymph vessel compliance,
lax skin

Ineffective muscular pump

Extralymphatic mastering of proteins exhausted

Lymph drainage through collaterals or
lymphovenous shunts inadequate

Decompensated iymph circulation

No chronic edema

Chronic lymphedema

Figure 18.6 Chronic secondary lymphedema of the left lower extremity in a 47-year-old
man after iliac node dissection, followed by irradiation.

pa rt I Vascular pathology and physiology

lee iiin

1 i^H mi •

*£ t 1
J 1 •" ■ J -

" ■ . - : -

Figure 18.7 Lymphoscintigram in a 44-year-old woman with secondary
lymphedema of the right lower extremity. (A) Note absence of right iliac
nodes and the presence of right inguinal nodes and collaterals. (B) Note
deterioration of lymphatic drainage 1 months later. There is no filling of the
right inguinal nodes or collaterals. The patient had a recent episode of
lymphangitis.

the lymph nodes and the major lymph vessels. Even the larger
lymphatic collaterals, which functioned effectively in the ini-
tial period after lymphatic obstruction, may occlude with
time. In this stage, dilated dermal lymphatics provide the only
lymphatic drainage of the extremity. Using noninvasive func-
tional tests, such as radionuclide lymphoscintigraphy per-

Figure 18.8 Contrast lymphangiogram in an 1 8-year-old man with
lymphangiectasia, protein-losing enteropathy, and chylous ascites
demonstrates dilated and tortuous thoracic duct.

formed with technetium-labeled antimony sulfur colloid, it is
possible to repeat the studies in the same patient and docu-
ment progression of the disease (Fig. 18.7).

Lymphatic stasis also results in deficiency of important im-
munoglobulins, cytokines, and plasma proteins. Because of
chronic inflammatory changes in the subcutaneous tissue and
the skin, there frequently is increased vascularity in the lym-
phedematous limb, and inflammatory cells accumulate. The
affected limb has an increased sensitivity to fungal and bacter-
ial infections. Obstructive lymphangitis further destroys the
lymphatic system and results in progression of the lymphede-
ma. In long-standing, neglected lymphedema, irreversible
sclerosis of the subcutaneous tissue and skin develops. Lym-
phangiosarcoma, which is a severe late complication of
secondary lymphedema, fortunately is rare.

Pathophysiology of chylous disorders

Disorders in the circulation of chyle usually are caused by lym-
phangiectasia or megalymphatics, with or without obstruc-
tion of the thoracic duct (Fig. 18.8). 27/28 Because of valvular
incompetence, chyle in these patients may reflux to the pelvis
or lower extremities, causing chylorrhea from small vesicles in
the skin of the limb, scrotum, or labia (Fig. 18.9). Reflux to the
kidney may lead to chyluria, whereas transudation through or
rupture of abdominal lymphatics results in chylous ascites.
Rupture of the lymphatics into the lumen of the gut causes
protein-losing enteropathy, and chylothorax develops if the

212

CHAPTER 18 Physiologic changes in lymphatic dysfunction

Figure 18.9 (A) Chyle draining through small
vesicles of the skin at the left groin of a 16-year-
old girl with lymphangiectasia and severe reflux
of the chyle. (B) Intraoperative photograph of
dilated, incompetent iliac lymphatics
containing chyle.

<^H

^^B

-^r

_ i _^ ta __ | ^^^^^ |ba

thoracic duct or mediastinal, intercostal, or diaphragmatic
lymphatics rupture.

Secondary chyloperitoneum or chylothorax is caused most
frequently by malignant tumors, primarily lymphoma, or by
injury to the thoracic duct. The latter usually is iatrogenic, oc-
curring during operations on the thoracoabdominal aorta 29-31
or, rarely, after a high translumbar aortography. 32

Chyle is a sterile alkaline fluid, odorless, and milky in ap-
pearance. Its protein content is around 4 g/dl and the fat con-
tent ranges from 0.4 to 4 g/dl. The fat stains with Sudan stain

and this test confirms the diagnosis of chyle in the peritoneal or
thoracic aspirate. The specific gravity of chylous fluid is
greater than 1012 g/dl.

Loss of chyle into the body cavities or through chylocuta-
neous fistulas has important physiologic consequences. If not
treated, it leads to malnutrition, hypoproteinemia, hypo-
cholesterolemia, hypocalcemia, immunodeficiency, and se-
vere metabolic disturbances. 27 ' 28 Lymphopenia and anemia
contribute to the poor immune function in these patients.

Chylous effusion in a patient with malignancy usually

213

pa rt I Vascular pathology and physiology

carries an ominous prognosis. The outcome of patients with
primary chylous disorders and reflux of the chyle depends on
the effectiveness of medical treatment. To compensate for the
physiologic changes caused by the loss of chyle, treatment is
directed at decreasing production of the chyle with a medium-
chain triglyceride diet, or by parenteral nutrition. In addition
to adequate calorie and protein replacement, calcium, lost in
chyle, also should be replaced. Reflux can be controlled effec-
tively with radical excision and ligation of the retroperitoneal
lymphatics in most cases. In patients with chylous effusion,
the site of lymphatic rupture should be oversewn if medical
treatment, paracentesis, or thoracentesis are ineffective. In
some patients with protein-losing enteropathy, the most dis-
eased segment of the small bowel may have to be resected to
decrease loss of chyle into the gastrointestinal tract. 27 ' 28 Trans-
plantation of small bowel for severe mesenteric lymphangiec-
tasia remains a task of the future, and it requires, as do many
other aspects of lymphatic disorders, further clinical research.

References

1. Starling EH. The influence of mechanical factors on lymph
production. J Physiol (Lond) 1894; 16:224.

2. Starling EH. On the absorption of fluids from the connective tissue
spaces. J Physiol (Lond) 1986; 19:312.

3. Drinker CK. The Lymphatic System: Its Part in Regulating Composi-
tion and Volume of Tissue Fluid. Stanford, CA: Stanford University
Press, 1942.

4. Rusznyak I, Foldi M, Szabo G. Lymphatics and Lymph Circulation.
New York: Pergamon Press, 1960.

5. Kinmonth JB. Lymphangiography in man: a method of outlining
lymphatic trunks at operation. Clin Sci 1952; 11:13.

6. Stewart G, Gaunt JI, Croft DN, Browse NL. Isotope lymphogra-
phy: a new method of investigating the role of the lymphatics in
chronic limb oedema. Br J Surg 1985; 72:906.

7. Gloviczki P, Calcagno D, Schirger A et al. Noninvasive evaluation
of the swollen extremity: experiences with 190 lymphoscinti-
graphic examinations. / Vase Surg 1989; 9:683.

8. Partsch H, Urbanek A, Wenzel-Hora B. The dermal lymphatics
in lymphoedema visualized by indirect lymphography. Br }
Dermatol 1984; 110:431.

9. Weissleder R, Thrall JH. The lymphatic system: diagnostic imag-
ing studies. Radiology 1989; 172:315.

10. Case TC, Witte CL, Witte MH et al. Magnetic resonance imaging in
human lymphedema: comparison with lymphangioscintigraphy.
Magn Reson Imag 1992; 10:549.

11. Weissleder R, Elizondo G, Wittenburg J, Lee AS, Josephson L,
Brady TJ. Ultrasmall superparamagnetic iron oxide: an intra-
venous contrast agent for assessing lymph nodes with MR
imaging. Radiology 1990; 175:494.

12. Duewell S, Hagspiel KD, Zuber J, von Schulthess GK, Bollinger A,
Fuchs WA. Swollen lower extremity: role of MR imaging.
Radiology 1992; 184:227.

13. Foldi E, Foldi M, Clodius L. The lymphedema chaos: a lancet. Ann
Plast Surg 1989; 22:505.

14. Pappas CJ, O'Donnell TF Jr. Long-term results of compression
treatment for lymphedema. / Vase Surg 1992; 16:555.

15. Gloviczki P, Fisher J, Hollier LH, Pairolero PC, Schirger A, Wahner
HW. Microsurgical lymphovenous anastomosis for treatment of
lymphedema: a critical review. / Vase Surg 1988; 7:647.

16. O'Brien BM, Mellow CG, Khazanchi RK, Dvir E, Kumar V,
Pederson WC. Long-term results after microlymphatico-venous
anastomoses for the treatment of obstructive lymphedema.
Plast Reconstr Surg 1990; 85:562.

17. Baumeister RG, Siuda S. Treatment of lymphedemas by micro-
surgical lymphatic grafting: what is proved? Plast Reconstr Surg
1990; 85:64.

18. Moore KL. The circulatory system. In: Moore KL, ed. The Develop-
ing Human, 3rd edn. Philadelphia: WB Saunders, 1982:296.

19. Leak LV. Electron microscopic observations on lymphatic capillar-
ies and the structural components of the connective tissue-lymph
interface. Microvasc Res 1970; 2:361.

20. Leak LV, Burke JF. Electron microscopic study of lymphatic capil-
laries in the removal of connective tissue fluids and particulate
substances. Lymphology 1968; 1:39.

21. Ganong WE Review of Medical Physiology, 10th edn. Los Altos, CA:
Lange Medical Publications, 1981:452.

22. Adair TH, Guyton AC. Physiology: lymph formation, its control,
and lymph flow. In: Clouse ME, Wallace S, eds. Lymphatic Imaging
Lymphography, Computed Tomography and Scintigraphy, 2nd edn.
Baltimore: Williams & Wilkins, 1985;123.

23. Witte CL, Witte MH. Circulatory dynamics and pathophysiology
of the lymphatic system. In: Rutherford RB, ed. Vascular Surgery,
5th edn. Philadelphia: WB Saunders, 2000:2110.

24. McHale NG, Roddie IC. The effect of transmural pressure on
pumping activity in isolated bovine lymphatic vessels. / Physiol
1976; 261:255.

25. Dobbins DE. Catecholamine-mediated lymphatic constriction:
involvement of alpha 1 and alpha 2 adrenoreceptors. Am } Physiol
1992;263:H473.

26. Dobbins DE, Dabney JM. Endothelin-mediated constriction of
prenodal lymphatic vessels in the canine forelimb. Regul Pept
1991;35:81.

27. Kinmonth JB. Chylous diseases and syndromes, including refer-
ences to tropical elephantiasis. In: Kinmonth JB, ed. The Lymphat-
ics: Surgery, Lymphography and Diseases of the Chyle and Lymph
System, 2nd edn. London: Edward Arnold, 1982:221.

28. Servelle M. Congenital malformation of the lymphatics of the
small intestine. / Cardiovasc Surg 1991; 32:159.

29. Garrett HE Jr, Richardson JW, Howard HS et al. Retroperitoneal
lymphocele after abdominal aortic surgery. / Vase Surg 1989;
10:245.

30. Williams RA, Vetto J, Quinones-Baldrich W et al. Chylous ascites
following abdominal aortic surgery. Ann Vase Surg 1991; 5:247.

31. Gloviczki P, Bergman RT Lymphatic problems and revasculariza-
tion edema. In: Bernhard VM, Towne JB, eds. Complications in
Vascular Surgery, 2nd edn. St Louis: Quality Medical Publishing,
1991:366.

32. Negroni CC, Ortiz VN. Chylothorax following high translumbar
aortography: a case report and review of the literature. Bol Assoc
Med PR 1988; 80:201.

214

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Physiologic changes in visceral ischemia

Tina R. Desai
Joshua A. Tepper
Bruce L. Gewertz

Interruption of intestinal blood flow can result in syndromes
of acute or chronic ischemia. Specific symptoms depend on
the nature, degree, and duration of blood flow interruption
as well as individual comorbidities and differences in
mesenteric anatomy and collateral development. The mortal-
ity in patients affected by acute mesenteric ischemia remains
high, often exceeding 60%, despite recent advances in
operative management and critical care. This poor prognosis
probably reflects frequent delays in diagnosis and our limited
ability to identify patients at risk for acute ischemia prior
to its onset. Continued advancements in the understanding
of the pathophysiology of intestinal ischemia are essential
to the improvement of outcomes from these syndromes.
This chapter will review anatomic and physiologic factors
important in the intestinal circulation and their relevance to
ischemic states.

Clinical ischemic syndromes

Acute insufficiency of the blood supply to the small bowel
and /or right colon may result from mesenteric arterial occlu-
sion (embolus or thrombosis), mesenteric venous occlusion,
and nonocclusive processes. Embolization to the superior
mesenteric artery (SMA) accounts for approximately 50% of
all cases of acute mesenteric ischemia; 20% of cases are sec-
ondary to thrombosis of a preeexistent atherosclerotic lesion. 1
Nonocclusive mesenteric ischemia accounts for approximately
20% of all episodes of intestinal ischemia. The cause of nonoc-
clusive mesenteric ischemia is most commonly multifactorial
but usually involves moderate or severe arterial atheroscle-
rotic lesions in association with a low cardiac output state
and /or the administration of vasoactive agents. Mesenteric
venous thrombosis is an unusual cause of mesenteric ischemia
accounting for, at most, 10-15% of cases. 2 Other unusual
arteropathies such as Takayasu's arteritis, fibromuscular dys-
plasia, and polyarteritis nodosa may first present with intesti-
nal ischemia. Isolated dissections of the SMA also have been
reported, 3 although the more common mechanism is exten-

sion of dissections of the descending thoracic aorta into the
SMA and celiac axis. 4 ' 5

While symptomatic chronic mesenteric ischemia (CMI) re-
mains relatively rare, its prevalence appears to be increasing. 2
This syndrome occurs most frequently in the setting of ad-
vanced atherosclerosis, although the female preponderance of
CMI is a unique feature compared with other atherosclerotic
complications. As a result of the extensive collateralization
among the celiac artery, SMA, and the inferior mesenteric
artery (IMA), it is usually true that two of these three main vis-
ceral trunks must be compromised before symptoms develop
(Fig. 19.1). Nonatherosclerotic causes of CMI include
thrombosis associated with thoracoabdominal aneurysm,
aortic coarctation, aortic dissection, mesenteric arteritis,
fibromuscular dysplasia, neurofibromatosis, middle aortic
syndrome, Buerger's disease, and extrinsic celiac artery
compression by the median arcuate ligament.

Anatomy

The mesenteric circulation consists primarily of three
branches of the abdominal aorta (Fig. 19.2): the celiac axis
(C A), the SMA, and the IMA. Their multiple branch points and
interconnections form a rich anastomotic network, such that
compromise of two of the three major arteries is usually re-
quired for the development of chronic ischemic symptoms.

The CA supplies the stomach, liver, spleen, portions of
the pancreas, and the proximal duodenum. It originates from
the ventral portion of the abdominal aorta, near the level of the
T 12 -L 1 , between the diaphragmatic crura. Its origin is encased
in the median arcuate ligament, a dense fibrous portion of the
central posterior diaphragm draped across the aortic hiatus. In
most patients, the CA branches soon after its origin into the
common hepatic, splenic, and left gastric arteries. In 1% of
cases, the SMA arises from the CA as well, forming a common
celiacomesenteric trunk.

The hepatic artery is usually the first branch of the CA. It
may also arise from the SMA (the so-called "replaced right

215

pa rt I Vascular pathology and physiology

Figure 19.1 Aortogram demonstrating meandering mesenteric artery
collateral originating from the inferior mesenteric artery in a patient with
chronic celiac and superior mesenteric artery occlusion.

Left Gastric

Hepatic

Right Gastric
Gastroduodenal

Superior .
Mesenteric

Middle Colic

Right Colic

Celiac Trunk

Splenic

Marginal Artery
of Drummond

Inferior
Mesenteric

Left Colic

Ileo Colic

Figure 19.2 Anatomy of the mesenteric circulation.

hepatic artery") in about 12% of cases. Additional variants in-
clude the "replaced common hepatic artery" (about 2.5%), and
direct origin of the common hepatic artery from the aorta
(about 2%). The common hepatic artery gives rise to the right
gastric artery and gastroduodenal artery which further di-
vides into the right gastroepiploic and superior pancreatico-
duodenal arteries. The remaining proper hepatic artery gives
rise to the cystic, right hepatic, and left hepatic arteries which

respectively serve the gallbladder, right and caudate, and
middle and left hepatic lobes.

The second branch of the CA is the splenic artery. Its first
named branch is the dorsal pancreatic artery supplying the
posterior body and tail of the pancreas. Just before entering the
splenic hilum, the splenic artery gives rise to the left gastro-
epiploic artery and multiple short gastric arteries, providing
blood flow to the gastric fundus.

The final branch of the CA is the left gastric artery. It courses
cephalad and to the left to supply the gastric cardia and fundus
along the lesser curvature of the stomach, joining centrally
with the right gastric artery from the hepatic artery. In approxi-
mately 12% of the population, the left hepatic artery originates
from the left gastric artery.

The SMA arises from the aorta just distal to the CA at the
level of L 1 -L 2 . It passes behind the neck of the pancreas, in
front of the uncinate process, and over the third portion of
the duodenum. Its first branch, the inferior pancreaticoduode-
nal artery, courses superiorly to join the superior pancreatico-
duodenal artery (from the gastroduodenal), and forms
the proximal-most collateral pathway with the CA. The
central branches of the SMA supply the midgut from the
ligament of Treitz to the midtransverse colon. These include
the middle colic (serving the proximal two-thirds of the trans-
verse colon), right colic (mid and distal ascending colon), and
ileocolic (distal ileum, cecum, appendix, and proximal as-
cending colon).

The IMA arises from the left side of the aorta 8-10 cm distal
to the SMA at the level of the third lumbar vertebra. It travels
caudad and to the left before dividing into the left colic and
sigmoid arteries. The IMA supplies the distal third of the
transverse colon, the descending and sigmoid colon, and the
proximal rectum. It has anastomotic communications with
the left branch of the middle colic from the SMA, and portions
of the middle and inferior rectal arteries from the internal iliac.

The mesenteric circulation has a redundant collateral
network, which serves to maintain perfusion even with com-
promise of the proximal main channels. The CA and SMA
communicate primarily via the superior and inferior pan-
creaticoduodenal arteries (via the gastroduodenal artery). The
SMA and IMA communicate via the centrally located arc of
Riolan (often referred to as the meandering mesenteric artery) as
well as by the multiple communications at the periphery of the
colon called the marginal arteries of Drummond. In addition to
these collateral pathways, muscular branches of the aorta may
contribute to intestinal perfusion including the lumbar inter-
costal arteries, internal mammary arteries (via the deep epi-
gastric arteries), middle sacral artery, and internal iliac arteries
(via collaterals between the inferior and superior rectal ar-
teries). Because of this plentiful collateral network, it is under-
standable that in most instances of gradual occlusion, at least
two of the three major mesenteric orifices must be blocked to
produce the clinical syndromes of chronic intestinal ischemia.
In contrast, sudden occlusion of one widely patent vessel

216

chapter 19 Physiologic changes in visceral ischemia

can cause acute ischemia since collaterals may be
underdeveloped. 6

Regulation of the mesenteric circulation

The mesenteric vascular bed contains as much as 25% of the
total blood volume, and extracts 15-20% of the 2 delivered at
baseline levels. 7 This large capacity can prove critical in help-
ing the body to compensate for hypovolemic states. It has been
estimated that maximal constriction of the splanchnic vascu-
lar bed may redistribute up to 1 .5 1 of blood per minute into the
systemic circulation. 7 With the ingestion of food, blood flow to
this region may increase up to 25% to meet the increased cellu-
lar needs associated with absorption and transport. 8 Although
controversy exists as to the distribution of intestinal blood
flow (IBF) throughout the gut, there is agreement that blood
flow supplying the mucosal and submucosal layers exceeds
that of the muscularis and serosal layers by about 40% of the
cardiac output. Intramural arteries, located in the deep sub-
mucosal plexus, comprise the extensive intestinal microcircu-
lation. Eccentrically located vessels provide arterial blood
flow to the villus tip, while the base is supplied by arterioles
from adjacent crypts. Regulation of IBF occurs both intrinsi-
cally through local mechanisms as well as extrinsically
through systemic control of the circulation.

Intrinsic control

Local intestinal mechanisms can regulate blood flow
independent of neural input and systemically circulating
vasoactive substances. Both metabolic and myogenic mecha-
nisms have been supported in experimental models.

The "metabolic" theory is based on oxygen availability as
the controlling variable in regulating perfusion. Decreases in
tissue oxygen supply relative to oxygen demand, such as in a
postprandial state, lead to the release of reactive metabolites,
which diffuse into surrounding tissues. Their overall effect
upon arteriolar smooth muscle cells is a net reduction in
tone, which leads to increased blood flow and increased
delivery of oxygen. A number of substances have been
implicated in this process including oxygen-derived free
radicals, 9 endothelium-derived relaxing factor (EDRF), 10-12
leukotrienes, 13 eicosanoids, 11 and adenosine. 14 These metabo-
lites also act on precapillary sphincters, further regulating 2
extraction through changes in capillary surface area and diffu-
sion distance. 14 The primary characteristic of this regulatory
process is that oxygen is the controlled variable and not blood
flow or pressure.

The second, "myogenic" theory is based on the ability of
vascular smooth muscle to maintain constant vessel wall ten-
sion despite variations in transmural perfusion pressure. The
Bayliss principle states that changes in vascular smooth mus-
cle tone maintain a constant transmural tension at the arteriole

despite variations in transmural pressure. For example, in-
creased transmural pressure leads to arteriolar vasoconstric-
tion, while decreased perfusion pressure leads to vasodilation.
By these adjustments, the intestinal circulation maintains rela-
tively constant capillary pressures and transcapillary fluid ex-
change. It is currently believed that the myogenic response is
composed of two distinct phases. The initial phase requires
calcium influx through voltage-dependent calcium channels,
supporting the prevailing hypothesis that myogenic contrac-
tion is initiated by smooth muscle cell depolarization. The
sustained phase requires calcium influx through voltage-
dependent calcium channels in addition to a cytochrome P450
metabolite, possibly 20-HETE. 15

Depending upon local conditions, either one of these basic
mechanisms (metabolic or myogenic) may predominate. At
extremes, both processes are active, facilitating pressure-flow
autoregulation (the ability to maintain near normal blood flow
in the face of changing perfusion pressures). This phenome-
non has been consistently demonstrated in various animal
models although its power can be compromised by extrinsic
neural innervation and systemically circulating vasoactive
substances. For example, in a denervated rat perfusion pre-
paration utilized in our laboratory, 16 systemic arterial pressure
was maintained constant while the SMA perfusion pressure
was progressively reduced. Intestinal blood flow remained
within normal limits until a perfusion pressure of approxi-
mately 70 mmHg was reached (the pressure-flow autoregula-
tory limit; Fig. 19. 3). 17 Below this pressure, intestinal blood

69.3 mmHg

o
u

o
o

H
in

1.1
1.0
0.9

0.8 -

0.7 -

0.6 -

0.5 -

0.4 -

0.3 -

0.2 -

0.1 -

V|<*## • # •

y = 0.0146 (x) - 0.062
r = 0.811

100

120

PERFUSION PRESSURE (mmHg)

Figure 19.3 Pressure-flow autoregulation: below the autoregulatory limit
of approximately 70 mmHg pressure incremental decreases in pressure result
in a decrease in blood flow (reprinted with permission from Sisley etal. Basic
mechanisms in mesenteric ischemia. In: Sidawy AN, Sumpio BE, DePalma RG,
eds. The Basic Science of Vascular Disease. New York: Futura Publishing Co.,
1997:723.)

217

pa rt I Vascular pathology and physiology

flow progressively decreased. Oxygen extraction progres-
sively increased as perfusion pressure decreased below the
pressure-flow autoregulatory limit so that oxygen consump-
tion was preserved until the point of maximal arterial-venous
oxygen difference (approximately 30 mmHg) after which oxy-
gen consumption progressively decreased. Further studies
performed in human small intestinal segments perfused with
an ex-vivo circuit identified a critical flow rate of 30 ml/min per
100 g tissue (Fig. 19.4). 18 Below this flow rate, oxygen con-
sumption became flow dependent as oxygen extraction could
not be increased above the maximal arterial-venous oxygen
difference.

Unfortunately, the intrinsic regulatory mechanisms are not
perfect, and may even pose a threat in certain clinical situa-
tions. For example, a heightened myogenic response may
occur in patients with chronic mesenteric arterial occlusions
following revascularization and the reestablishment of nor-

(a)

^ 7

♦••* ^

- 6

& 4

a 3

• • •
• *

^^^

< 1

* # ^^^

Blood flow (ml/ mm • lOOg)

2.0-1

(b)

*
e
£

o
>

i.5-

1.0

,5"

T-

—r-
60

20 30 40

Blood flow (ml/ min • lOOg)

Figure 19.4 (A) Relationship between human intestinal blood flow and
oxygen extraction (arteriovenous oxygen difference) (reprinted from Desai
etal. Defining the critical limit of oxygen extraction in the human small
intestine. J Vase Surg 1996; 23:832, with permission from Elsevier). (B) Below
a critical flow rate of 30 ml/min • 1 00 g tissue human intestinal oxygen
consumption becomes flow dependent (Desai etal. Defining the critical limit
of oxygen extraction in the human small intestine. J Vase Surg 1996; 23:832,
with permission from Elsevier.)

mal perfusion pressures, resulting in edema of the gut and
symptoms of continued intestinal ischemia in the postopera-
tive period despite patent reconstructions. Vasospasm is the
apparent etiological agent in this perplexing syndrome.

Extrinsic control

Neural and hormonal mechanisms assist in the regulation of
mesenteric blood flow by altering its distribution in response
to a variety of influences including altered levels of physical
activity, the fasting or postprandial state of the intestine,
sepsis, and stress.

Sympathetic nervous system input is provided by pregan-
glionic cholinergic fibers of the greater splanchnic nerves
which synapse in celiac ganglia adjacent to the celiac axis. An
extensive submucosal nervous plexus and a series of nerves
that penetrate the gut wall innervate the arteries of the gut,
whereas an insignificant amount of nerves supplies lymphat-
ics and veins. Stimulation of postganglionic fibers leads to
mesenteric artery and arteriolar vasoconstriction. Mesenteric
venous capacitance is also regulated via this network. Direct
stimulation of the sympathetic nerves has been shown to de-
crease splanchnic blood volume by more than half, with major
reduction occurring in the first 30 seconds. 19 Continued stimu-
lation results in an "autoregulatory escape" allowing partial
recovery of blood flow. This reproducible phenomenon,
present predominantly in the mucosa more so than the
muscularis, 20 ' 21 is believed to occur secondarily to a lactic
acidosis-induced inhibition of the postjunctional oc 2 -receptor
response to norepinephrine. 22,23 The parasympathetic ner-
vous system is thought to play a smaller role in the control of
the mesenteric circulation, despite its rich supply of nerves to
the small intestine via roots of the vagi.

Numerous circulating compounds contribute to the hu-
moral regulation of intestinal perfusion. The most potent
vasoactive compounds present in blood include vasopressin
and angiotensin II which produce uniform decreases in both
IBF and V0 2 . These agents are responsible for the sympatheti-
cally induced vasoconstriction seen posthemorrhage. An-
giotensin II receptor antagonists administered during
hemorrhage attenuate the decrease in IBF without systemic ef-
fects, in dogs. 24 Epinephrine has been shown to produce a
dose-dependent response in flow and V0 2 , whereas nor-
epinephrine decreases both IBF and V0 2 uniformly. 25

Furthermore, substances that have similar effects upon the
circulation in general may exhibit differing effects at the cellu-
lar level. For example, the catecholamines, epinephrine and
norepinephrine, both potent vasoconstrictors, have opposite
effects on intestinal capillary exchange; epinephrine increases
capillary exchange capacity while norepinephrine decreases
it. Epinephrine has been shown to have a differential pattern of
effects upon the intestine depending on the dose; it results in
vasodilation at low doses and vasoconstriction at higher
doses. This disparity is due to the differing affinities of the

218

chapter 19 Physiologic changes in visceral ischemia

drug or hormone for multiple receptor types at increasing
doses.

Adenosine and histamine are potent intestinal vasodilators.
Exogenously administered compounds such as the oc-
antagonist, phentolamine, or the (3-antagonist, propranolol,
are also capable of producing profound vasodilation as a re-
sult of their effects at the specific target receptors. Gastroin-
testinal peptides such as glucagon, CCK, VIP, and serotonin
have also been shown to be capable of reducing capillary fil-
tration 26 and eliciting arteriolar vasodilation. 27 ' 28

Clinical implications

Acute venous hypertension is a well-studied experimental
perturbation which results in increased vascular resistance
and decreased IBF. This phenomenon is associated with the
clinical syndrome of mesenteric venous occlusion and is most
consistent with the myogenic theory of regulation. Variations
of the venous hypertensive response may reflect the metabolic
state of the intestine and may be of physiologic significance
with respect to syndromes of nonocclusive mesenteric is-
chemia. For example, dogs given a luminal food source
or intraarterial dinitrophenol (a drug which induces a
hypermetabolic state) showed reduction or reversal of the
vasoconstriction elicited by acute venous hypertension. 29 In
contrast, chronic digitalis administration appeared to exacer-
bate the myogenic response to acute venous hypertension re-
sulting in chronic venoconstriction and decreased splanchnic
venous capacitance. The clinical syndrome of nonocclusive
mesenteric ischemia associated with the administration of
digitalis and other vasoactive drugs may result from a similar
mechanism.

An accentuated myogenic response may be of clinical im-
portance when patients with chronic mesenteric arterial occlu-
sions are revascularized and normal perfusion pressure is
reestablished. Gewertz and Zarins 30 reported three patients
who demonstrated intramural edema of the gut and symp-
toms of continued intestinal ischemia in the postoperative pe-
riod despite patent reconstructions. These patients were noted
to have diffuse mesenteric vasospasm and responded to con-
servative treatment including bowel rest, vasodilators, and
calcium channel blockers. This syndrome, characterized by an
inability of the intestinal microcirculation to prevent the accu-
mulation of absorbed fluid within the interstitium of the gut,
may represent a failure of capillary "derecruitment" after rein-
stitution of blood flow to previously ischemic tissue. In sup-
port of this hypothesis, the density of perfused capillaries
increases in the intestine in response to both feeding (metabol-
ic hyperemia) and local hypoxia (reactive hyperemia). 31 ' 32 The
nonocclusive mesenteric ischemia resulting after reperfusion
of ischemic intestine may result from myogenic regulation of
IBF in a maximally dilated capillary bed distal to a previously
occluded superior mesenteric artery. The vasospasm of
medium-sized vessels may actually be an adaptive response

to "protect" the maximally dilated capillary bed. The vascular
smooth muscle in these distal vessels which have been chroni-
cally exposed to low perfusion pressures may be exquisitely
sensitive and respond disproportionately to even slight in-
creases in pressure.

Variations in hematocrit and p0 2 are involved in the regula-
tion of IBF, as well. A linear relationship has been shown to
exist between decreasing hematocrit and increasing IBF. In
isolated canine intestinal loops perfused at a constant pres-
sure, oxygen extraction exhibits a parabolic function to chang-
ing hematocrit. 33 Maximum V0 2 occurs when the hematocrit
is approximately 50%. The increase in IBF secondary to a de-
crease in hematocrit has also been demonstrated in a rat model
where a hematocrit reduction from 41% to 17% resulted in a
twofold increase in blood flow. 34 Increased IBF is also seen
when arterial oxygen content is decreased without alterations
in oxygen-carrying capacity (i.e. hematocrit). 17 ' 35

Pathophysiologic mechanisms in intestinal
ischemia and reperfusion

An acute reduction in blood flow to the intestine results in
tissue damage from both the hypoxia incurred during flow in-
terruption as well as the deleterious effects of reperfusion.
Ischemia induces a complicated series of cellular events and
has the potential to cause permanent loss of cell function. The
cellular damage depends on the degree and duration of inter-
ruption of blood flow as well as the metabolic activity of the
cells. Reperfusion injury is primarily mediated by inflamma-
tory mediators and reactive oxygen species and may occur
even after brief periods of ischemia.

Ischemia

Intestinal ischemia results in a spectrum of functional and
morphologic alterations, ranging from minor changes in mu-
cosal permeability to full-thickness tissue necrosis affecting
the entire intestinal wall. An increase in mucosal permeability
has been shown to be the earliest physiological change, occur-
ring after as little as 10-30 min of ischemia. 36 This ultrastruc-
tural damage increases the fluid layer between the cells and
the basement membrane and is followed by a loss of villus tips,
which are at highest risk for necrosis since the greatest oxygen
concentration is provided at the crypt base. Permanent cell
damage occurs by 3 hours, which is clearly demarcated by the
loss of villus crypts. The resulting tissue destruction occurs in
a centrifugal manner, proceeding sequentially from the mu-
cosa to the submucosa and then to the muscularis. Reepithe-
lialization can occur if flow is restored before the crypt stem
cells are irreparably damaged and is usually complete within
24 hours after the ischemic insult.

Arterial occlusion produces an immediate alteration in the
gross appearance of the intestines; initial pallor progresses to

219

pa rt I Vascular pathology and physiology

cyanosis with prolonged ischemia. Peristaltic activity may
increase initially but eventually abates and is followed by
edema, intramural hemorrhage, and ultimately gangrene as
the tissue becomes vulnerable to intraluminal hydrolases, bile,
and bacteria. An enormous shift of extracellular fluid occurs
into the lumen, bowel wall, mesentery, and peritoneal cavity
with the increase in interstitial and intraluminal pressure
promoting a further reduction in perfusion.

Direct studies of isolated cells have elucidated some of the
biochemical events associated with hypoxic cell death. Cells
undergo specific changes in enzyme activities, mitochondrial
function, cytoskeletal structure, membrane transport, and
antioxidant defenses in response to ischemia. Mitochondrial
dysfunction leading to the depletion of cellular adenosine
triphosphate (ATP) appears to be the critical factor, initiating a
cascade of events which ultimately results in hypoxic injury 37
The decrease in aerobic metabolism noted during hypoxia is
also associated with an alteration in cellular pH (lactic acido-
sis), increased production of hypoxia inducible factors (HIF-
l), 38 and the generation of reactive oxygen species (ROS). The
contribution of these factors to the development of ischemic
injury has not been clearly defined.

The depletion of cellular ATP has long been believed
to be the premier event in initiating a complex cascade of
intracellular biochemical events. 37 Hypoxia prohibits aerobic
metabolism, decreasing ATP and increasing adenosine
monophosphate (AMP), causing the downregulation of both
the Na-K-ATPase pump and the epithelial Na + channel
(ENaC). Subsequently, ionic homeostasis is no longer main-
tained and membrane damage ensues. Mitochondrial calcium
efflux ensues, resulting in an increased cytosolic calcium con-
centration which activates destructive proteases and lipases
and leads to cell surface bleb formation. These blebs grow and
coalesce until one to three large terminal blebs remain. Ulti-
mately, cytolysis occurs as one of the terminal blebs ruptures,
marking the transition from reversible to irreversible injury.

Despite abundant support for this hypothesis, recent tech-
nological advances, especially the use of multiparameter
digitized video microscopy (MDVM), have produced
contradictory evidence. In particular, several studies have
failed to demonstrate a rise in intracellular calcium levels
until long after irreversible injury was evident. 39 Aw and
colleagues 40 showed that ischemic mitochondria possess
adaptive mechanisms to maintain calcium homeostasis.
Ionic gradients are maintained during early hypoxia by a
nonenergy-dependent inhibition of ion movement across the
inner mitochondrial membrane. Lemasters et al. 41 observed no
increase in cytosolic free calcium with hypoxia, even though
they did observe terminal bleb formation. Such recent data
challenge the idea that the loss of calcium homeostasis is the
final common pathway to cell death.

An alternative hypothesis was proposed by Gores et al. A2
after they demonstrated that intracellular pH dropped by

more than one full point during hypoxia. Experiments also
showed that the maintenance of this intracellular acidosis pro-
longed cell survival whereas inhibition of pH reduction ac-
celerated cell death. They proposed that intracellular acidosis
depresses the activity of the critical degradative enzymes
(phospholipases and proteases) which are activated during
ATP depletion. These enzymes damage the cytoskeleton caus-
ing increased membrane permeability. The resulting hydro-
gen ion leakage out of the cell raises the intracellular pH and
prevents further inhibition of the degradative enzymes, al-
lowing continued membrane damage and further leakage of
hydrogen ions out of the cell. The protective effect of intracel-
lular acidosis carries implications for reperfusion injury as
well, since the abrupt rise in pH occurring upon reoxygenation
may accentuate cell damage. Cellular injury can be prevented
if pH is slowly increased after reoxygenation. 43

Another potential mechanism of hypoxic cellular injury is
the formation of ROS. Although the release of ROS into tissues
by inflammatory cells is well documented during reperfusion
of ischemic organs, small amounts of these compounds may
be equally important intracellular messengers. 44-46 Super-
oxide, hydrogen peroxide, and hydroxyl species are thought
to be released from mitochondria of hypoxic cells and initiate a
series of secondary cellular responses. Hastie et al. 47 applied
exogenous hydrogen peroxide to study the mechanisms of
ROS effects on endothelial monolayers. Application of this
compound resulted in a decrease in intracellular cAMP, a re-
distribution of cytoskeletal elements, and an increase in the
gap area. 13 This effect was inhibited by the adenylate cyclase
stimulator, forskolin. Other studies have implicated protein
kinase C (PKC), platelet activating factor, and cAMP in
permeability changes mediated by ROS. 48-50 Studies in our
laboratory have correlated the production of ROS by hypoxic
endothelial cells with impaired barrier function as measured
by trans-endothelial electrical resistance (TEER). 51 Further
data have demonstrated that addition of menadione, a stimu-
lus of endogenous ROS production, to normoxic cells repro-
duces the permeability increases seen with hypoxia. 52

Reperfusion

Reestablishment of blood flow is essential to prevent death of
ischemic tissues, but intestinal reperfusion results in an addi-
tional tissue injury. Parks and Granger demonstrated that the
mucosal injury observed after 3 h of ischemia followed by 1 h
of reperfusion was more severe than the injury observed after
4 h of ischemia alone. 53 Clark and Gewertz 54 showed that
intermittent episodes of ischemia and reperfusion resulted
in significantly worse histologic injury than a comparable time
of continuous ischemia.

Reperfusion of the ischemic intestine incurs further local
damage by a variety of mechanisms. Reactive oxygen metabo-
lites released from reperfused tissue are known to cause

220

chapter 19 Physiologic changes in visceral ischemia

significant microvascular and parenchymal injury This
phenomenon can be prevented in animal models by the use of
oxygen free radical scavengers such as superoxide dismutase,
mannitol, and allopurinol. 55 ' 56 The relatively high concentra-
tions of xanthine dehydrogenase in intestinal mucosal tissue
are thought to be an important source of reactive oxygen
metabolites in reperfused intestine. This enzyme is converted
to xanthine oxidase in an ischemic environment, and the xan-
thine oxidase enzyme system results in superoxide radical and
peroxide release. 55 ' 57-59

Recent studies have suggested that ischemia reperfusion-
associated microvascular dysfunction may result from an
imbalance between superoxide and nitric oxide resulting in
impaired arteriolar vasodilation and an acute inflammatory
response in venules. 60 Reperfusion of ischemic tissue is
thought to increase the production of superoxide by endothe-
lial cells and decrease the synthesis of nitric oxide. This results
in a significant decrease in the baseline ability of nitric oxide to
scavenge intracellular superoxide, maintain arteriolar vasodi-
lation, prevent platelet aggregation and intravascular throm-
bosis, and minimize adherence of leukocytes to endothelium.

Additional important sources of reactive oxygen metabo-
lites during reperfusion are circulating polymorphonuclear
leukocytes (PMN) and other inflammatory cells, especially
mast cells. 61 The interaction between PMN and endothelial
cells is now known to be critical in reperfusion-associated in-
jury in the intestine and other tissues, including myocardium
and skeletal muscle. In order for inflammatory cells to partici-
pate in reperfusion injury, they must be attracted to the site of
postischemic tissue, adhere to the microvascular endo-
thelium, and migrate through the vessel wall to infiltrate the
tissue. The adhesion, diapedesis, and activation of PMNs in
reperfused tissue is mediated by a complex series of interac-
tions between cytokines [especially tumor necrosis factor
(TNF)-oc, interleukin (IL)-l, platelet-derived growth factor
(PDGF)], the CD11/CD18 complex on the PMN cell mem-
brane, and endothelial cell adhesion molecules.

ROS derived from xanthine oxidase promote leukocyte ad-
herence as demonstrated by the fact that xanthine oxidase in-
hibitors significantly decrease the number of adherent PMN
in reperfused tissue. 56 Zimmerman and Granger 62 have
suggested that PMN are attracted to ischemic microvascular
endothelium by a two-step mechanism. First, ROS activate
phospholipase A 2 in endothelial cell membranes. Phospholi-
pase A 2 activation leads to the formation of leukotriene B 4 and
platelet activating factor which are both chemoattractants for
PMN. Supporting this theory, it has been demonstrated that
leukotriene B 4 and platelet activating factor levels increase
dramatically on reperfusion of ischemic intestine in canine
and feline models. 63 ' 64 Additionally, PMN infiltration of reper-
fused tissue was significantly decreased in animals treated
with either a leukotriene B 4 or platelet activating factor
receptor antagonist. 62

Neutrophil binding to the vascular endothelium is
mediated by a glycoprotein adhesion complex on the PMN
(CD11/CD18) and corresponding endothelial-based adhe-
sion complexes (E-selectin, intercellular adhesion molecules).
The use of monoclonal antibodies against the CD11/CD18
complex or against endothelial-based adhesion complexes in-
hibits PMN adherence to the microvascular endothelium and
results in a decrease in observed reperfusion injury. 56 ' 65-67
Other endothelial cell adhesion complexes (vascular cell adhe-
sion molecule, platelet-endothelial cell adhesion molecule)
are involved in adhesion of leukocytes and platelets to reper-
fused microvascular endothelium, further contributing to
tissue injury upon reperfusion.

A variety of cytokines contribute to the pathogenesis of
ischemic and reperfusion injury. Both pro- and antiinflamma-
tory cytokines are elaborated in the local circulation and their
balance is critical to normal homeostasis. Secretion of proin-
flammatory cytokines such as TNF-oc, IL-1, and IL-6 leads
directly or indirectly to chemoattraction of inflammatory
mediators, upregulation of cell adhesion molecules, and
alteration of vascular permeability.

An ex-vivo perfusion model of human small intestine devel-
oped in our laboratory has facilitated the study of the venous
effluent from reperfused human intestine. Application of per-
fusate from this ex-vivo model to isolated cultures of endothe-
lial cells resulted in increased expression of cellular adhesion
molecules (intercellular adhesion molecule-1 and E-selectin)
as measured by flow cytometry and Northern blot analysis.
Venous effluent was analyzed for cytokine contents and
revealed increased levels of IL-1, TNF, and IL-6 in a time-
dependent fashion with maximal increases in levels of IL-6. 68
Application of TNF and IL-1 to endothelial cell monolayers re-
sulted in increased adhesion molecule expression similar to
the increases seen with application of the venous effluent of
reperfused intestine. Application of IL-6 did not demonstrate
this effect on adhesion molecule expression but did modify en-
dothelial permeability, an effect which is mediated through ef-
fects of the PKC enzyme system on junctional proteins. 68-70
These effects of cytokines further contribute to the recruitment
of inflammatory cells in reperfused intestine.

The end-result is the activation of myeloperoxidase and
other destructive enzymes (collagenase, elastase) released
from inflammatory cells, which further augment the ischemic
damage that has already occurred. Importantly, such reperfu-
sion phenomena are a promising target for therapy. It has been
shown in clinically relevant experimental models that damage
can be reduced by the filtration of leukocytes in the reperfusate
as well as by pharmacologic antagonism of ROS, cellular ad-
hesion molecules, or cytokines. 65 ' 71-73

Gut mucosal barrier dysfunction is an important conse-
quence of intestinal reperfusion injury. Disruption of the
intestinal mucosal barrier may allow translocation of
bacteria and bacterial products (e.g. endotoxin) into the

221

pa rt I Vascular pathology and physiology

portal and systemic bloodstream. Roumen and colleagues 74
have demonstrated systemic endotoxemia after major vascu-
lar operations involving aortic cross clamping. Alternatively,
the bacteria may act locally to activate an inflammatory
cascade which may then have local and systemic effects.
Several studies have suggested that decontamination of the
gut prior to an ischemic insult ameliorates systemic effects of
reperfusion. 75 ' 76

A final and devastating effect of reperfusion of ischemic in-
testine is the adverse effect on distant tissues via systemic acti-
vation of an inflammatory response. Cardiac, pulmonary,
hepatic, and other organ system injury results from activation
and release of inflammatory cytokines (TNF-oc and IL-1),
arachidonic acid metabolites [prostacyclin, thromboxane
A 2 (TxA 2 ), leukotriene B 4 (LTB 4 )], endothelium-derived
relaxing factor, endothelin, platelet activating factor, and
complement. 77

Respiratory insufficiency is the most frequent systemic
complication of intestinal ischemia and reperfusion, occurring
in approximately 10% of patients after intestinal revascular-
ization. This syndrome of acute respiratory distress syndrome
is characterized by increased microvascular permeability in
the lung resulting in an accumulation of PMN-rich alveolar
fluid. It typically occurs 24-72 h after reperfusion and results
in significant oxygen requirements and prolonged ventilatory
support.

Failure of multiple organ systems (MOSF) contributes to the
significant morbidity and mortality of intestinal ischemic
syndromes. This leading cause of death in critically ill patients
has been documented to result from intestinal ischemia
and reperfusion. 78 Organ system failure resulting from
diverse shock states may also be the result of nonocclusive
mesenteric ischemia associated with splanchnic vasoconstric-
tion. Proposed mechanisms for the initiation of distal effects
include the loss of mucosal barrier integrity, bacterial trans-
location to mesenteric lymph nodes and portal venous blood,
and stimulation of inflammatory cells by these bacteria and
their products resulting in a systemic release of inflammatory
cytokines.

The mesenteric circulation has complex anatomy and op-
portunities for rich perfusion. It is regulated by both metabolic
and myogenic mechanisms which allow fairly good autoregu-
lation of oxygen delivery. The injury resulting from blood flow
interruption involves both ischemia and reperfusion pheno-
mena. Recent research has offered insight into these processes
and holds much promise for more targeted cellular therapies
in the future.

Acknowledgments

The authors wish to thank Ms Lydia Johns for her assistance
in the preparation of the figures and Ms Karen Hynes for her
assistance in the preparation of the manuscript.

References

1. Stoney RJ, Cunningham CG. Acute mesenteric ischemia. Surgery
1993; 114:489.

2. Kairaluoma MI, Karkola P, Heikkinen D et ah Mesenteric infarc-
tion. Am } Surg 1977; 133:188.

3. Vignati PV, Welch JP, Ellison L et ah Acute mesenteric ischemia
caused by isolated superior mesenteric artery dissection. / Vase
Surg 1992; 16:109.

4. Cambria RP, Brewster DC, Gertler J et ah Vascular complications
associated with spontaneous aortic dissection. / Vase Surg 1988;
7:199.

5. Chopra PS, Grassi CJ. Superior mesenteric artery angioplasty
with the TEGwire: usefulness and technical difficulties. / Vase
Interv Radiol 1992; 3:523.

6. Schwartz LB, Gewertz BL. Intestinal ischemic disorders. In: Yao
JST, Pearce WH, eds. Modern Trends in Vascular Surgery. New York:
Appleton and Lange, 1999:347.

7. Donald DE. Splanchnic circulation. In: Shepherd JT, Abboud FM,
eds. Handbook of Physiology— The Cardiovascular System. Baltimore,
MD: Williams & Wilkins, 1983:219.

8. Granger DN, Richardson PD, Kvietys PR et ah Intestinal blood
flow. Gastroenterology 1980; 78:837.

9. Myers SI, Hernandez R. Oxygen free radical regulation of rat
splanchnic blood flow. Surgery 1992; 112:347.

10. Fan WQ, Smolich JJ, Wild J et ah Nitric oxide modulates regional
blood flow differences in the fetal gastrointestinal tract. Am } Phys-
iol 1996; 271:G598.

11. Kodama T, Marmon LM, Vargas R et ah The interaction between
endothelium-derived relaxing factor (EDRF) and eicosanoids in
the regulation of the mesenteric microcirculation. / Surg Res 1995;
58:227.

12. Shen W, Lundborg M, Wang J et ah Role of EDRF in the regulation
of regional blood flow and vascular resistance at rest and during
exercise in conscious dogs. JAppl Physiol 1994; 77:165.

13. Chapnick BM. Divergent influences of leukotrienes C4, D4,
and E4 on mesenteric and renal blood flow. Am } Physiol 1984;
246:H518.

14. Shepherd AP Metabolic control of intestinal oxygenation and
blood flow. FedProc 1982; 41:2084.

15. Chlopicki S, Nilsson H, Mulvany MJ. Initial and sustained phases
of myogenic response of rat mesenteric small arteries. Am } Physiol
Heart Circ Physiol 2001; 281:H2176.

16. Mesh CL, Gewertz BL. The effect of hemodilution on blood flow
regulation in normal and postischemic intestine. / Surg Res 1990;
48:183.

17. Sisley AC, Tullis MJ, Clark ET et ah Basic mechanisms in mesen-
teric ischemia. In: Sidawy AN, Sumpio BE, DePalma RG, eds. The
Basic Science of Vascular Disease. Armonk, NY: Futura Publishing
Co., 1997:723.

18. Desai TR, Sisley AC, Brown S et ah Defining the critical limit of
oxygen extraction in the human small intestine. / Vase Surg 1996;
23:832; discussion 838.

19. Brooksby GA, Donald DE. Dynamic changes in splanchnic blood
flow and blood volume in dogs during activation of sympathetic
nerves. Circ Res 1971; 29:227.

20. Bohlen HG, Henrich H, Gore RW et ah Intestinal muscle and

222

chapter 19 Physiologic changes in visceral ischemia

mucosal blood flow during direct sympathetic stimulation. Am }
Physiol 1978; 235:H40.

21. Shepherd AP, Riedel GL. Intramural distribution of intestinal
blood flow during sympathetic stimulation. Am } Physiol 1988;
255:H1091.

22. Chen LQ, Riedel GL, Shepherd AP. Norepinephrine release dur-
ing autoregulatory escape: effects of alpha 2-receptor blockade.
Am J P/zys/o/ 1991; 260:H400.

23. Chen LQ, Shepherd AP. Role of H+ and alpha 2-receptors in escape
from sympathetic vasoconstriction. Am } Physiol 1991; 261 :H868.

24. Suvannapura A, Levens NR. Local control of mesenteric blood
flow by the renin-angiotensin system. Am J Physiol 1988; 255:G267.

25. Kvietys PR, Granger DN. Vasoactive agents and splanchnic
oxygen uptake. Am J Physiol 1982; 243:G1 .

26. Harper SL, Bohlen HG, Granger DN. Vasoactive agents and the
mesenteric microcirculation. Am} Physiol 1985; 249:G309.

27. Banks RO, Gallavan RH Jr, Zinner MH et al. Vasoactive agents in
control of the mesenteric circulation. Fed Proc 1985; 44:2743.

28. Lanciault G, Jacobson ED. The gastrointestinal circulation.
Gastroenterology 1976; 71:851.

29. Granger HJ, Norris CP Intrinsic regulation of intestinal oxygena-
tion in the anesthetized dog. Am J Physiol 1980; 238:H836.

30. Gewertz BL, Zarins CK. Postoperative vasospasm after antegrade
mesenteric revascularization: a report of three cases. / Vase Surg
1991; 14:382.

31. Pawlik WW, Fondacaro JD, Jacobson ED. Metabolic hyperemia in
canine gut. Am J Physiol 1980; 239:G12.

32. Shepherd AP. Intestinal capillary blood flow during metabolic
hyperemia. Am J Physiol 1979; 237:E548.

33. Shepherd AP, Riedel GL. Optimal hematocrit for oxygenation of
canine intestine. CircRes 1982; 51:233.

34. Kiel JW, Riedel GL, Shepherd AP. Effects of hemodilution on
gastric and intestinal oxygenation. Am J Physiol 1989; 256:H171.

35. Shepherd AP. Intestinal 02 consumption and 86Rb extraction dur-
ing arterial hypoxia. Am J Physiol 1978; 234:E248.

36. Patel A, Kaleya RN, Sammartano RJ. Pathophysiology of
mesenteric ischemia. Surg Clin North Am 1992; 72:31 .

37. Kehrer JP, Jones DP, Lemasters JJ et al. Mechanisms of hypoxic cell
injury. Summary of the symposium presented at the 1990 annual
meeting of the Society of Toxicology. Toxicol Appl Pharmacol 1990;
106:165.

38. Wang GL, Semenza GL. Purification and characterization of
hypoxia-inducible factor l.JBiolChem 1995; 270:1230.

39. Nieminen AL, Gores GJ, Wray BE et al. Calcium dependence of
bleb formation and cell death in hepatocytes. Cell Calcium 1988;
9:237.

40. Aw TY, Andersson BS, Jones DP. Suppression of mitochondrial
respiratory function after short-term anoxia. Am } Physiol 1987;
252:C362.

41. Lemasters JJ, DiGuiseppi J, Nieminen AL et al. Blebbing, free Ca2+
and mitochondrial membrane potential preceding cell death in
hepatocytes. Nature 1987; 325:78.

42. Gores GJ, Nieminen AL, Fleishman KE et al. Extracellular acidosis
delays onset of cell death in ATP-depleted hepatocytes. Am J
Physiol 1988; 255:C315.

43. Currin RT, Gores GJ, Thurman RG et al. Protection by acidotic pH
against anoxic cell killing in perfused rat liver: evidence for a pH
paradox. FASEB J 1991; 5:207.

44. Chandel NS, Maltepe E, Goldwasser E et al. Mitochondrial reac-

tive oxygen species trigger hypoxia-induced transcription. Proc
Natl Acad Sci USA 1998; 95:11715.

45. Zulueta JJ, Sawhney R, Yu FS et al. Intracellular generation of
reactive oxygen species in endothelial cells exposed to anoxia-
reoxygenation. Am } Physiol 1997; 272:L897.

46. Vanden Hoek TL, Becker LB, Shao Z et al. Reactive oxygen species
released from mitochondria during brief hypoxia induce pre-
conditioning in cardiomyocytes./B/o/ C/zem 1998; 273:18092.

47. Hastie LE, Patton WF, Hechtman HB et al. H202-induced filamin
redistribution in endothelial cells is modulated by the cyclic
AMP-dependent protein kinase pathway. / Cell Physiol 1997; 172:
373.

48. Ochoa L, Waypa G, Mahoney JR Jr et al. Contrasting effects of
hypochlorous acid and hydrogen peroxide on endothelial perme-
ability: prevention with cAMP drugs. Am J Respir Crit Care Med
1997; 156:1247.

49. Johnson A, Phillips P, Hocking D et al. Protein kinase inhibitor pre-
vents pulmonary edema in response to H202. Am } Physiol 1989;
256:H1012.

50. Siflinger-Birnboim A, Goligorsky MS, Del Vecchio PJ et al. Activa-
tion of protein kinase C pathway contributes to hydrogen
peroxide-induced increase in endothelial permeability. Lab Invest
1992; 67:24.

51. Ali MH, Schlidt SA, Chandel NS et al. Endothelial permeability
and IL-6 production during hypoxia: role of ROS in signal trans-
duction. Am } Physiol 1999; 277:L1057.

52. Schlidt SA, Ali MH, Chandel NS et al. Quenching of reactive oxy-
gen species prevents endothelial permeability changes during
hypoxia. Surg Forum 1998; 49:343.

53. Parks DA, Granger DN. Contributions of ischemia and reperfu-
sion to mucosal lesion formation. Am J Physiol 1986; 250:G749.

54. Clark ET, Gewertz BL. Intermittent ischemia potentiates intestinal
reperfusion injury. / Vase Surg 1991; 13:601.

55. Granger DN, McCord JM, Parks DA et al. Xanthine oxidase
inhibitors attenuate ischemia-induced vascular permeability
changes in the cat intestine. Gastroenterology 1986; 90:80.

56. Suzuki M, Inauen W, Kvietys PR et al. Superoxide mediates
reperfusion-induced leukocyte-endothelial cell interactions. Am}
Physiol 1989; 257:H1740.

57. Granger DN. Role of xanthine oxidase and granulocytes in
ischemia-reperfusion injury. Am ] Physiol 1988; 255:H1269.

58. Grisham MB, Hernandez LA, Granger DN. Xanthine oxidase and
neutrophil infiltration in intestinal ischemia. Am } Physiol 1986;
251:G567.

59. Parks DA, Granger DN. Xanthine oxidase: biochemistry, distribu-
tion and physiology. Acta Physiol Scand Suppl 1986; 548:87.

60. Grisham MB, Granger DN, Lefer DJ. Modulation of leukocyte-
endothelial interactions by reactive metabolites of oxygen and
nitrogen: relevance to ischemic heart disease. Free Radic Biol Med
1998; 25:404.

61. Kanwar S, Kubes P. Ischemia /reperfusion-induced granulocyte
influx is a multistep process mediated by mast cells. Microcircula-
tion 1994; 1:175.

62. Zimmerman BJ, Granger DN. Reperfusion injury. Surg Clin North
Am 1992; 72:65.

63. Kubes P, Suzuki M, Granger DN. Platelet-activating factor-
induced microvascular dysfunction: role of adherent leukocytes.
Am/P/zys/o/1990;258:G158.

64. Otamiri T, Lindahl M, Tagesson C. Phospholipase A2 inhibition

223

pa rt I Vascular pathology and physiology

prevents mucosal damage associated with small intestinal is-
chaemia in rats. Gut 1988; 29:489.

65. Hernandez LA, Grisham MB, Twohig B et ah Role of neutrophils in
ischemia-reperfusion-induced microvascular injury. Am } Physiol
1987;253:H699.

66. Friedman G, Jankowski S, Shahla M et ah Administration of an
antibody to E-selectin in patients with septic shock. Grit Care Med
1996; 24:229.

67. Kurose I, Anderson DC, Miyasaka Metal. Molecular determinants
of reperfusion-induced leukocyte adhesion and vascular protein
leakage. Circ Res 1994; 74:336.

68. Wyble CW, Desai TR, Clark ET et ah Physiologic concentrations
of TNFalpha and IL-lbeta released from reperfused human intes-
tine upregulate E-selectin and ICAM-1. / Surg Res 1996; 63:333.

69. Wyble CW, Hynes KL, Kuchibhotla J et ah TNF-alpha and IL-1 up-
regulate membrane-bound and soluble E-selectin through a com-
mon pathway. / Surg Res 1997; 73:107.

70. Desai TR, Leeper NJ, Hynes KL et ah Interleukin-6 causes endothe-
lial barrier dysfunction via the protein kinase C pathway. / Surg
Res 2002; 104:118.

71. Korthuis RJ, Granger DN. Reactive oxygen metabolites,

neutrophils, and the pathogenesis of ischemic-tissue/reperfu-
sion. Clin Cardiol 1993; 16:119.

72. Sisley AC, Desai T, Harig JM et ah Neutrophil depletion attenuates
human intestinal reperfusion injury. / Surg Res 1994; 57:192.

73. Vedder NB, Winn RK, Rice CL et ah A monoclonal antibody to
the adherence-promoting leukocyte glycoprotein, CD18, reduces
organ injury and improves survival from hemorrhagic shock and
resuscitation in rabbits. / Clin Invest 1988; 81:939.

74. Roumen RM, Frieling JT, van Tits HW et ah Endotoxemia after
major vascular operations. / Vase Surg 1993; 18:853.

75. Sorkine P, Szold O, Halpern P et ah Gut decontamination reduces
bowel ischemia-induced lung injury in rats. Chest 1997; 112:491.

76. Alverdy J, Piano G. Whole gut washout for severe sepsis: review of
technique and preliminary results. Surgery 1997; 121:89.

77. Tullis MJ, Brown S, Gewertz BL. Hepatic influence on pulmonary
neutrophil sequestration following intestinal ischemia-
reperfusion. / Surg Res 1996; 66:143.

78. Neary P, Redmond HP. Ischaemia-reperfusion injury and the
systemic inflammatory response syndrome. In: Grace PA, Mathie
RT, eds. Is chaemia-Reper fusion Injury. London: Blackwell Science,
1999:123.

224

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Natural history of atherosclerosis in
the lower extremity, carotid, and
coronary circulations

Daniel B. Walsh

Arterial occlusion with consequent ischemia is the leading
cause of death due to myocardial infarction and stroke in the
United States and other industrialized nations. Most com-
monly, arterial occlusion is the result of atherosclerosis, but it
also can be caused by vasculitis, vasospasm, embolism, dissec-
tion, fibromuscular dysplasia, congenital abnormalities, and
hypercoagulability. Because of their lower frequency, nonath-
erosclerotic causes of arterial occlusion result in significantly
less morbidity and mortality than atherosclerosis. Accord-
ingly, this chapter will examine the natural history of athero-
sclerotic lesions in the lower extremity, carotid, and coronary
circulations. Only with this knowledge can rational decisions
for therapeutic intervention be made.

Atherosclerosis is a difficult entity to study scientifically.
The complex environment of the arterial wall, including adja-
cent pulsatile blood flow, requires consideration of velocity,
shear stress, pulsatility, elasticity, and the microenvironment
of the blood-endothelial interface. This apparent polyfactorial
milieu is complicated further by a slow pathologic process and
risk factors peculiar to the human lifestyle that make animal
models of atherosclerosis both difficult and expensive to pro-
duce. Species differences in endothelial and smooth muscle
cell biology further confound the interpretation of animal
experiments.

Despite its complex etiology, the macropathology of athero-
sclerosis as demonstrated by histologic study can be described
as one of three types of lesions: the fatty streak, the fibrous
plaque, and the complex calcified lesion. Each lesion is
composed of differing proportions of three elements: intimal
smooth muscle cells, connective tissue, and accumulated in-
tracellular and extracellular lipid. Fatty streaks are minimally
elevated intimal lesions that are ubiquitous in all but the very
young. They do not impinge significantly on the arterial
lumen and probably never cause downstream, end-organ
damage. Fibrous plaques are heaped-up accumulations of de-
generated foam cells covered by a thick layer of proliferated
smooth muscle cells. In contrast to fatty streaks, fibrous
plaques may produce clinical ischemia by slowly enlarging
and restricting the lumen of an artery enough to reduce arte-

rial flow. Complex calcified lesions of atherosclerosis repre-
sent a progression of fibrous plaques with increases in size of
the lipid-rich core, precipitation and accumulation of calcium,
deterioration of endothelial integrity, platelet thrombus for-
mation at the flow surface, and hemorrhage into the core of
these lesions. These lesions can cause acute ischemia due to ar-
terioarterial embolism of plaque-associated platelet thrombi
and atheromatous debris, as well as terminal thrombosis of
arteries distal to nearly occlusive plaques.

As atherosclerotic lesions develop and progress, clinical
symptoms caused by flow limitation, embolization, or throm-
bosis occur. Initially, a stenosis impedes arterial flow during
periods of peak demand, such as during lower extremity exer-
cise when calf muscle ischemia causes symptoms of intermit-
tent claudication. As the plaque enlarges to a critical stenosis,
flow becomes sufficiently limited so that ischemia causes
symptoms at rest in the heart or extremity. Plaque contents
may be discharged into the arterial lumen, causing emboli that
produce transient cerebral ischemia or stroke. Ultimately, an
enlarging plaque may progress to cause luminal obliteration
and distal arterial thrombosis. Unfortunately, animal experi-
mentation has not clearly documented this progression, osten-
sibly because of the difficulty in producing a valid model that
can mimic advanced human fibrous plaques. 1 This chapter ex-
amines what is known about the correlation of specific athero-
sclerotic lesions and clinical symptoms, to shed light on the
natural history of atherosclerotic lesions in the critical circula-
tions of the limbs, the brain, and the heart.

Lower extremity arteries

Clinical progression of disease

Before the era of arterial reconstruction, atherosclerotic lesions
were seen only at autopsy or amputation. Treatment was re-
stricted to patients with severe symptoms such as gangrene.
At that time, intermittent claudication was thought to be a
relatively benign process, with less than 10% of patients ever

225

pa rt I Vascular pathology and physiology

requiring amputation, implying that lower extremity athero-
sclerosis was a stable or slowly progressive process. With the
advent of arterial reconstructive surgery, the frequency at
which patients with claudication progressed to critical is-
chemia requiring intervention has been studied in more detail.
In one such study, Cronenwett et al. 2 followed 116 men with
claudication that had been stable for at least 6 months after
initial presentation. After 2.5 years 7 mean follow-up, 60% of
patients reported worsening claudication. At the same time,
operations for progressive lower extremity atherosclerosis
were performed at the rate of 9% per year, despite the initial in-
tent to treat these patients without surgery. In 257 patients
with intermittent claudication, Jelnes and associates 3 reported
that 7.5% progressed to critical ischemia in 12 months, but only
2.2% per year progressed thereafter. Rosenbloom and col-
leagues 4 found that critical ischemia developed within 5 years
in 24% of patients with stable claudication. In a larger study,
Dormandy and Murray 5 followed 1969 patients with claudica-
tion and found that 7.3% progressed to critical lower extremity
ischemia in 1 year. In a 15-year study of 2777 claudicants,
Muluk et al. 6 found a 10-year cumulative amputation rate of
less than 10% while revascularization procedures occurred at
a 10-year cumulative rate of 18%.

Patients who present with intermittent claudication as their
major symptom of atherosclerosis are also at risk for systemic
progression of atherosclerotic disease. In Cronenwett et al.'s 2
study of men with claudication, the annual mortality rate was
5%, mostly due to complications of atherosclerosis. O'Rior-
dain and O'Donnell 7 reported a 5-year actuarial mortality of
23% in an analysis of 112 patients with stable claudication. In
this study, patients with an initial ankle-brachial index (ABI)
less than 0.5 suffered 50% mortality, whereas patients whose
ABI was greater than 0.5 had only 16% mortality. This empha-
sizes the association between more severe lower extremity
atherosclerosis and the risk of complications from generalized
atherosclerosis, because 72% of deaths were due to stroke or
myocardial infarction. 7 In Muluk et al.'s 6 study of 2777 claudi-
cants, mortalities occurred at a rate of 12% per year, 66% of
which were due to heart disease.

When it is recognized that a subgroup of patients presenting
with apparently stable lower extremity atherosclerosis will
progress to critical ischemia, identification of patient charac-
teristics associated with progression becomes useful. In the
Cronenwett et al. study, 2 patients with intermittent claudica-
tion who had smoked at least 40 pack years subsequently re-
quired operation 3.3 times more frequently than those who
smoked less. Jonason and Ringqvist 8 reported that smoking
best predicted development of rest pain in patients with ini-
tially stable intermittent claudication. In their review of clau-
dication, Dormandy et al. 9 state that the amputation rate is 11
times greater in smokers than in nonsmokers. Although the
Cronenwett et al. study 2 also found that significant daily exer-
cise was associated with more stable claudication, the study's
design precluded understanding whether this was a causal
relation.

Several reports have found that a lower initial ABI is associ-
ated with more frequent symptom progression and operation
requirement. 2 ' 4 Not surprisingly, a significant decrease in the
ABI during follow-up was an even more accurate predictor of
disease progression. 2 Jelnes and associates 3 emphasized the
importance of an initial ankle pressure of less than 70 mmHg, a
toe pressure less than 40 mmHg, and an ABI less than 0.5 in
predicting subsequent critical ischemia. In this experience, no
patients with ABI greater than 0.7 required amputation over a
mean follow-up of 6.5 years. Dormandy and Murray 5 also
found that an initial ABI less than 0.5 predicted a subsequent
operation with a rate 2.3 times greater than among patients
whose initial ABI was more than 0.5.

Studies of patients with intermittent claudication neces-
sarily differ with respect to risk factors, anatomic location,
initial symptoms, and sampling bias. McDaniel and
Cronenwett 10 summarized the status of the natural history of
intermittent claudication by reviewing 48 published series.
The prevalence of claudication increases progressively with
age, from 1% (younger than 40 years) to 5% (older than 70
years). Prevalence approximately doubles in both patients
with diabetes and smokers, each of which represents an inde-
pendent risk factor. This summary of the best available data at
that time predicts the following 5-year outcome for patients
who present with intermittent claudication: 29% mortality;
among survivors, 25% arterial reconstruction; 4% amputa-
tions; 16% worsening claudication; and 55% stable (or pos-
sibly improved) claudication. Smoking cessation will reduce
this mortality by half (to 12%), eliminate major amputation,
and reduce operation requirement to 8%. Unfortunately, dia-
betes increases the projected 5-year mortality to 49%, with 21%
of survivors requiring amputation and 35% requiring arterial
reconstruction. Regular exercise will result in a 67% improve-
ment of walking distance. Using structured treadmill exer-
cises, a fourfold increased walking distance can be expected,
especially in patients who stop smoking. 10

Anatomic progression of disease

Although the natural history of intermittent claudication is
well established, the rate of anatomic progression of specific
atherosclerotic plaques in the iliac and superficial femoral
arteries (SFA) is unknown. Previously, this has not been an
important issue, because bypass surgery can be performed
equally well whether the affected artery is stenotic or totally
occluded. The advent of endovascular techniques such as
balloon angioplasty and stent placement has focused more
attention on predicting progression of specific atherosclerotic
lesions. These techniques are more effective when applied to
stenotic arteries before total occlusion occurs, and are cost
effective only if their results are durable.

To examine these issues as they relate to atherosclerosis
plaques, we have examined the natural history of lesions in the
SFA and the impact of disease distribution on endovascular

226

chapter 20 Natural history of atherosclerosis in the lower extremity, carotid, and coronary circulations

therapy in the iliac arterial system. Because the SFAis the most
frequent site of atherosclerosis causing claudication, we fol-
lowed 46 patients whose SFAs were asymptomatic or associ-
ated with minimal claudication, initially evaluated at the time
of an incidental arteriography 11 After 3 years' mean follow-
up, these SFA stenoses had progressed from a mean stenosis of
33% reduction in diameter to a mean stenosis of 45%. Most SFA
stenoses (72%) did not progress. However, 28% did progress,
including 17% which progressed to total occlusion. We then
examined risk factors associated with stenosis progression
and found that patients with longer smoking history and
those with contralateral SFA occlusion were more likely to ex-
perience progression of their asymptomatic SFA stenoses. Pro-
gression of SFA stenoses also was highly correlated with
symptom progression. Overall, the mean rate of stenosis pro-
gression was 5% per year, with a maximum rate of 15% per
year. Based on these progression rates, we concluded that pa-
tients with SFA stenoses of less than 70% reduction in luminal
diameter can be followed safely at yearly intervals with du-
plex scanning, which should allow the detection of stenosis
progression before occlusion. However, we noted extensive
variability in stenosis progression that could not be explained
by these above-mentioned predictors. We postulated that
individual lesion characteristics might predict stenosis
progression.

To address this question, we identified 19 patients who re-
quired arteriography for treatment of critical lower extremity
ischemia who had previously undergone arteriography where
minimal or no symptoms were present in the leg under
study 12 These incidental arteriograms had been performed a
mean of 32 months previously. Distinct SFA lesions were
characterized by their location, length, stenosis severity, and
morphologic appearance. Comparison was then made with
the appearance at the second arteriogram. The contribution of
patient-specific risk factors to disease progression was also
assessed.

We found, not surprisingly, that stenosis progression
occurred independently among multiple lesions within the
same patient. Our study confirmed a fact long known among
vascular surgeons —lesions in the adductor canal region were
more likely to occlude than lesions elsewhere in the SFA. As is
obvious, severity of initial lesion stenosis was also predictive
of occlusion. However, most progressing lesions (93%) arose
in areas of initially mild disease despite more severe lesions
elsewhere. Increasing patient age and the need for contra-
lateral surgery also were associated with lesion progression.
Finally, we found the smooth, asymmetric lesions progressed
11% more slowly than smooth symmetric lesions or lesions
which were irregular, ulcerated or complex.

The other large patient group who present with claudica-
tion have atherosclerosis in their iliac systems. We chose to
study those who failed endovascular therapy in an attempt to
discover determinants of failure so that they might be assessed
by different therapies including bypass surgery. 13 ' 14 We found

that in patients with multisegment iliac artery disease, those
with external iliac artery lesions had lower primary patency
after endovascular therapy. Patients with bilateral external
iliac artery disease, particularly those patients with external
iliac lesions greater than 5 cm in length, suffered the worst
results.

In summary, lower extremity atherosclerosis severe enough
to cause symptoms appears to progress significantly in
approximately 5-10% of patients per year. The stenoses
themselves progress at a similar rate. In diabetic or smoking
patients, this progression is accelerated. Interestingly, lesion
locations and characteristics of shape and length are also
useful in predicting their natural history. Mechanical stress
from movement to portions of arteries anchored in place
by muscle attachments is one possible cause. 15 Others
have speculated that areas which are mechanically restricted
from compensatory dilation in response to a progressive
plaque are the areas of most common lesion progession. 16 The
external iliac and adductor canal portion of the SFAs would
both qualify by this criterion. Progression of disease in other
critical areas such as the coronary or carotid circulations
appears to parallel events in the lower extremities of these
patients.

Carotid arteries

Clinical progression of disease

Stroke remains the third leading cause of death in the United
States. 17 Within 30 days of stroke, as few as 15% to as many as
33% of patients will die. Within 1 year, 30-52% of patients who
suffer ischemic stroke will be dead. 18-20 Atherosclerosis is the
most common disorder causing stroke, due to thrombotic oc-
clusion or embolization of atherosclerotic or thrombotic ma-
terial. Nearly 75% of strokes occurring in a community appear
to be caused by atherosclerosis of the extracranial carotid
circulation. 21 Unfortunately, this lesion is quite common.
Ramsey and coworkers 22 found extracranial carotid athero-
sclerosis in 11% of 102 asymptomatic volunteers older than 50
years; 6% of this group had hemodynamically significant
stenoses. In unselected populations between the ages of 65 and
74 years, the annual risk of stroke is approximately 0.8%. 20/23_25
As atherosclerotic disease progresses within the carotid artery,
the risk of stroke increases. In asymptomatic patients with
nonstenotic carotid ulceration, the annual risk of stroke in-
creases to 4%. 26/27 As the carotid artery disease becomes hemo-
dynamically significant, the annual stroke rate rises to more
than 6%. 28-32 Once the patients with a significant carotid steno-
sis become symptomatic, the annual risk of stroke increases to
15-20%. 33-35 Stroke risk is even higher in patients whose
carotid atherosclerosis has already resulted in a previous
stroke. 36,37
Just as with intermittent claudication, patients who present

227

pa rt I Vascular pathology and physiology

with evidence of carotid artery atherosclerosis are also at
risk for complications caused by atherosclerosis in other
circulations, most notably myocardial infarction. In a classic
study by Hertzer and colleagues, 38 ' 39 506 patients with
extracranial carotid artery disease as their primary diagnosis
underwent coronary angiography at the Cleveland Clinic.
Only 7% of these patients had normal coronary arteries. Severe
coronary lesions representing significant risk for myocardial
infarction were documented in 35% of patients. The angio-
graphic findings were correctable in 28%, but were already
inoperable in the remaining 7%. Left ventriculography
demonstrated myocardial impairment consistent with previ-
ous myocardial infarction in 24% of patients with unremark-
able cardiac histories, and in fully 25% of patients whose
electrocardiogram showed no evidence of myocardial
infarction. 39

Risk factors for progression of carotid atherosclerosis and
stroke abound. Increasing age, male gender, and black race are
proven factors that increase the risk for stroke due to carotid
atherosclerosis but, unfortunately, they are not treatable. 40 ' 41
As in other circulations, smoking is a great risk to the well-
being and function of the carotid circulation. 42-45 Using data
from the Framingham Study, Wolfe and associates 46 described
the general cerebrovascular risk profile that identifies 10% of
the population who will have 33% of the strokes. This risk pro-
file includes systolic hypertension, elevated serum choles-
terol, glucose intolerance, smoking, and electrocardiographic
evidence of left ventricular hypertrophy.

Anatomic progression of disease

With the maturation of noninvasive duplex ultrasound tech-
nology, several facts concerning the natural history of carotid
artery atherosclerosis have become clear. Roederer and col-
leagues 30 demonstrated that 36% of patients with carotid
artery stenoses of less than 50% progressed to greater than 50%
in 3 years. There was an 8% average annual progression rate of
lesions with less than 50% to greater than 50% reduction in
diameter. When all lesions were considered, 60% progressed
within 3 years. The average progression rate, although diffi-
cult to measure exactly, appears to be approximately 10% per
year. In a more recent study Liapis et al. 47 followed 442 carotid
arteries over a 10-year period with duplex. They found that
significant stenosis progression occurred in 19% of patients. In
the Liapis et al. study, 12% of patients suffered neurologic
events. As these stenotic lesions progress, symptoms occur
more frequently 48-50 Intraplaque hemorrhage, one indication
of severe progression that can be documented as heteroge-
neous plaque by duplex ultrasound, appears to be particularly
associated with lesion instability and the development of
stroke. 51-59 Since Moore and coworkers' association of com-
plex carotid bulb ulceration with increased risk of stroke, the
threat that ulcerated lesions pose for development of symp-

toms of cerebrovascular ischemia has been debated. 60 ' 61 A cor-
relation exists, however, between the presence of ulceration
within the carotid bulb and the presence of cerebral infarction
by computed tomography scanning, and underscores the
pathologic potential of these ulcers. 62 With the widespread use
of duplex, another risk factor for symptom progression has
been described— plaque echolucency. Nicolaides and others
have demonstrated ultrasonographic characteristics can pre-
dict an increased risk of neurologic events. 47,63-65

In summary, as in the lower extremity, symptoms and
signs of cerebral ischemia appear associated with progression
of atherosclerosis at the bifurcation of the carotid artery.
Lesions appear to progress at a rate of approximately 5-10%
reduction in diameter per year; in individual patients, this
rate is variable and directly related to risk factors such as
smoking, age, and hypertension. Characteristics of individual
lesions such as ulceration at arteriogram or echolucency at
duplex ultrasound are also predictive of an increased risk of
neurologic events. There is also a striking association of
cerebrovascular disease with the presence and progression
of coronary artery disease.

Coronary arteries

Clinical progression of disease

Atherosclerosis is the most common cause of chronic obstruc-
tion of the coronary arteries, which results in chronic ischemic
heart disease. Patients with known coronary artery atheroscle-
rosis and stable symptoms suffer an annual mortality of 4%. 66
There is a direct relationship between the severity of symp-
toms and patient outcome. 67,68 If angina progresses, requiring
hospitalization, medication change, or more noninvasive
treatment, coronary artery obstruction usually has progressed
in extent beyond that seen in patients with chronic stable angi-
na. 69 These patients die at a rate of 8-18% per year and have
myocardial infarction at a rate of 14-22% per year. 70 ' 71 Risk fac-
tors for development of coronary artery atherosclerosis have
been among the most extensively studied areas in modern
medical science. Every major epidemiologic study performed
has shown a significant correlation between serum cholesterol
at the time of entry and risk for development or progression of
coronary atherosclerosis. 72 The potential for development of
coronary atherosclerosis in male cigarette smokers is approxi-
mately twice that of nonsmokers. 73 Hypertension also has
been well established as a major risk factor for development of
coronary atherosclerosis in both sexes and among various age
and racial groups. 74 Diabetes mellitus is a well-known risk fac-
tor for development of coronary artery disease. 75 Finally, the
presence of peripheral vascular occlusive disease adversely
affects survival in medically managed patients with chronic
coronary artery disease. 76

228

chapter 20 Natural history of atherosclerosis in the lower extremity, carotid, and coronary circulations

Anatomic progression of disease

When coronary atherosclerosis is examined in relation to the
anatomic lesions that appear, a striking similarity to the be-
havior of atherosclerosis in the other circulations is seen. The
severity of left ventricular dysfunction and the extent of coro-
nary artery disease directly affect the patient's prognosis. In
general, the severity of left ventricular dysfunction is a more
important prognostic sign than the severity of coronary artery
disease. 77 Follow-up of patients in the Coronary Artery
Surgery Study registry has accurately documented the
survival of medically treated patients with angiographically
assessed coronary artery disease. Both the number of major
coronary arteries with severe obstruction and the degree of de-
pression of left ventricular ejection fraction were independent,
adverse risk factors, with the latter again exerting a dominant
influence. 78 These two risk factors are synergistic in that
adverse effects in the prognosis of impaired ventricular func-
tion are more pronounced as the number of stenotic vessels
increases.

Studies of symptomatic patients have revealed that if
only one of the three major coronary arteries has more
than a 50% stenosis, the annual mortality rate is approxim-
ately 2%. 79/80 In symptomatic patients with severe stenosis in
two of the three major arteries, the annual mortality rate
increases to approximately 7%. If all three vessels are stenotic,
mortality rises to approximately 11%. 81-83 In addition to the
number of vessels involved, the severity of obstruction also
is important. The prognosis for patients with 50-70% narrow-
ing is better than in those with more than 75% reduction in
diameter. 84

Ambrose and colleagues examined coronary morphology
at cardiac catheterization in patients with either stable or un-
stable angina. 85 ' 86 All patients had coronary lesions that ob-
structed the luminal diameter by 50% or more. They found,
just as in the superficial femoral artery, that asymmetrical
coronary lesions with a narrow neck and irregular borders are
more likely to be found in patients with unstable angina. In
contrast, lesions with an asymmetrical narrowing but with
smooth borders and a broad neck, or lesions with concentric,
symmetric narrowing, were more common in patients with
stable angina. In patients with stable angina who were re-
studied after an acute episode of unstable angina, it appeared
that acute symptomatic progression had resulted in most in-
stances from a change in a previously insignificant coronary
lesion. In these cases, eccentric coronary lesions (i.e. an eccen-
trically placed convex stenosis with a narrow neck and over-
hanging edges or irregular, scalloped borders) were the most
common morphologic feature of disease progression and may
represent either a disrupted atherosclerotic plaque, a partially
lysed thrombus, or both. These lesions appear to be the major
cause of unstable angina. This finding has been confirmed in
patients with the abrupt onset of stable angina. 87 Intracoro-

nary filling defects that appear consistent with thrombus have
also been found more commonly in patients with recent rest or
unstable angina pectoris. Coronary angioscopy frequently re-
veals complex plaques with thrombi not detected by coronary
angiography in such patients. 88

Thus, as in the lower extremity and carotid arteries, worsen-
ing symptoms predict progression of anatomic coronary dis-
ease, with increased morbidity and mortality. Smoking,
hypertension, diabetes, and elevated serum cholesterol have a
deleterious effect on the coronary circulation. Finally, exami-
nation of unstable coronary artery lesions suggests that in-
traplaque hemorrhage is a critical element in abrupt
deterioration within the coronary circulation.

Conclusion

There appears to be a common mechanism for progression of
atherosclerosis in the major circulations of the body. Anatomic
progression of disease is associated with progression of symp-
toms and leads inexorably to more frequent morbidity and
mortality. The mechanism appears related to deterioration
within an atherosclerotic plaque, leading to fissuring, in-
traplaque hemorrhage, and rupture, with accompanying arte-
rioarterial embolism. These increases in luminal obstruction
caused by a rapidly enlarging plaque lead to distal thrombosis
and end-organ damage. Risk factors are shared in all circula-
tions and are particularly related to age, smoking, hyperten-
sion, diabetes mellitus, and higher levels of cholesterol.
Interventions to prevent progression of these lesions can be
better planned with knowledge of progression rates and
plaque morphology. Application of noninvasive vascular
techniques to these areas should improve our knowledge of
the natural history of atherosclerotic lesions.

References

1. McGill HC Jr. Persistent problems in the pathogenesis of athero-
sclerosis. Arteriosclerosis 1984; 4:443.

2. Cronenwett JL, Warner KG, Zelenock GB et ah Intermittent claudi-
cation: current results of nonoperative management. Arch Surg
1984; 119:430.

3. Jelnes R, Gaardsting O, Jensen KH, Baekgaard N, Tonnesen KH.
Fate in intermittent claudication: outcome and risk factors. Br Med
/ 1986; 293:1137.

4. Rosenbloom MS, Flanigan DP, Schuler JJ et ah Risk factors affecting
the natural history of intermittent claudication. Arch Surg 1988;
123:867.

5. Dormandy JA, Murray GD. The fate of the claudicant: a prospec-
tive study of 1969 claudicants. Eur] Vase Surg 1991; 5:131.

6. Muluk SC, Muluk VS, Kelley ME et ah Outcome events in patients
with claudication: a 15-year study in 2777 patients. / Vase Surg
2001; 33:251.

229

pa rt I Vascular pathology and physiology

7. O'Riordain DS, O'Donnell JA. Realistic expectations for the
patient with intermittent claudication. Br J Surg 1991; 78:861 .

8. Jonason T, Ringqvist I. Factors of prognostic importance for sub-
sequent rest pain in patients with intermittent claudication. Acta
MedScand 1985; 218:27.

9. Dormandy J, Heeck L, Vig S. The natural history of claudication:
risk to life and limb. Semin Vase Surg 1999; 12:123.

10. McDaniel, Cronenwett JL. Basic data related to the natural history
of intermittent claudication. Ann Vase Surg 1989; 3:273.

11. Walsh DB, Gilbertson JJ, Zwolak RM et al. The natural history of
superficial femoral artery stenoses. / Vase Surg 1991; 14:299.

12. Walsh DB, Powell RJ, Stukel TA et al. Superficial femoral artery
stenoses: characteristics of progressing lesions. / Vase Surg 1997;
25:512.

13. Powell RJ, Fillinger MF, Bettmann M. The durability of endovas-
cular treatment of multisegment iliac occlusive disease. / Vase Surg
2000; 31:1178.

14. Powell RJ, Fillinger MF, Walsh DB et al. Predicting outcome of
angioplasty and selective stenting of multisegment iliac artery
occlusive disease. / Vase Surg 2000; 32:564.

15. Mavor GE. The pattern of occlusion in atheroma of the lower limb
arteries. Br] Surg 1952; 40:352.

16. Blair JM, Glagov S, Zarins CK. Mechanism of superficial femoral
artery adductor canal stenosis. Surg Forum 1990; 41:359.

17. Culicchia F, Mohr JP Morbidity and mortality of stroke. In: Moore
WS, ed. Surgery for Cardiovascular Disease. New York: Churchill
Livingstone, 1987:35.

18. Sacco RL, Wolfe PA, Kannel WB, McNamara PM. Survival and
recurrence following stroke: the Framingham Study. Stroke 1982;
13:290.

19. Mohr JP, Caplan LR, Melski JW et al. The Harvard Cooperative
Stroke Registry: a prospective registry. Neurology 1978; 28:754.

20. Soltero L, Lui K, Cooper R, Stamler J, Garside D. Trends in mortal-
ity from cerebrovascular diseases in the United States, 1969 to
1975. Stroke 1978; 9:549.

21. Whisnant JP, Fitzgibbons JP, Kurland LT, Sayre GP Natural
history of stroke in Rochester, Minnesota, 1945 through 1954.
Stroke 1971; 2:11.

22. Ramsey DE, Miles RD, Lambeth A, Sumner DS. Prevalence of ex-
tracranial carotid artery disease: a survey of an asymptomatic
population with noninvasive techniques. / Vase Surg 1987; 5:584.

23. Taylor LM Jr, Porter JM. Basic data related to carotid endarterect-
omy Ann Vase Surg 1986; 1:264.

24. The Joint Committees for Stroke Facilities. Transient focal cerebral
ischemia: epidemiological and clinical aspects. Stroke 1974; 5:277.

25. Kannel WB, Gordon T. Evaluation of cardiovascular risk in the
elderly: the Framingham Study. Bull NY Acad Med 1978; 54:573.

26. Dixon S, Pais SO, Raviola C et al. Natural history of nonstenotic,
asymptomatic ulcerative lesions of the carotid artery. Arch Surg
1982; 117:1493.

27. Kroener JM, Dom PL, Shoor PM, Wickborn IG, Bernstein EF. Prog-
nosis of asymptomatic ulcerating carotid lesions. Arch Surg 1980;
115:1387.

28. Meissner L Wiebers DO, Whisnant JP, O'Fallon WM. The natural
history of asymptomatic carotid artery occlusive lesions. JAMA
1987; 258:2704.

29. Taylor LM, Loboa L, Porter J. The clinical course of carotid bifurca-
tion stenosis as determined by duplex scanning. / Vase Surg 1988;
8:255.

30. Roederer GO, Langlois YE, Jager KA et al. The natural history of
carotid arterial disease in asymptomatic patients with cervical
bruits. Stroke 1984; 15:605.

31 . Busutil RW, Baker JD, Davidson RK, Machleder HI. Carotid artery
stenosis: hemodynamic significance and clinical course. JAMA
1981; 245: 1438.

32. Moore DJ, Miles RD, Gooley N A, Sumner DS. Noninvasive assess-
ment of stroke risk in asymptomatic and nonhemispheric patients
with suspected carotid disease. Ann Surg 1985; 202:491.

33. North American Symptomatic Carotid Endarterectomy Trial
Collaborators. Beneficial effect of carotid endarterectomy in
symptomatic patients with high-grade carotid stenosis. N Engl J
Med 1991; 325:445.

34. European Carotid Surgery Trialists' Collaborative Group. MRC
European Carotid Surgery Trial: interim results for symptomatic
patients with severe (70%) or mild (0%) carotid stenosis. Lancet
1991; 337:1235.

35. Mayberg MR, Wilson SE, Yatsu F et al. Carotid endarterectomy and
prevention of cerebral ischemia in symptomatic carotid stenosis.
JAMA 1991; 266:3289.

36. McCoulloch JL, Mentzer RM Jr, Harmon PK, Kaiser DL, Kron IL,
Crosby IK. Carotid endarterectomy after a completed stroke: re-
duction in long-term neurologic deterioration. / Vase Surg 1985;
2:7.

37. Asplund K, Liliequist B, Fodstad H, Wester PO. Long-term out-
come in cerebrovascular disease in relation to findings at aortocer-
vical angiography: a 12 year follow-up. Stroke 1981; 12:307.

38. Hertzer NR. Coronary artery disease in patients with cerebrovas-
cular disease: evaluation, risk assessment, timing, and manage-
ment. In: Moore WS, ed. Surgery for Cerebrovascular Disease. New
York: Churchill Livingstone, 1987:679.

39. Hertzer NR, Young JR, Beven EG et al. Coronary angiography in
506 patients with extracranial cerebrovascular disease. Arch Intern
Med 1985; 145:849.

40. Kurtzke JF. Epidemiology of cerebrovascular disease. In:
Cerebrovascular Survey Report for the Joint Council Subcommittee on
Cerebrovascular Disease. National Institute of Neurological and
Communicative Disorders and Stroke and National Heart
and Lung Institute. Rev. ed. Rochester, MN: Whiting Press,
1980:135.

41. Caplan LR, Gorelick PB, Hier DB. Race, sex and occlusive cere-
brovascular disease: a review. Stroke 1986; 17:648.

42. Tell GS, Howard G, McKinney WM, Toole IF. Cigarette smoking
cessation and extracranial carotid atherosclerosis. JAMA 1989;
261:1178.

43. GilllS, Shipley MJ, Tsementzis SA et al. Cigarette smoking: a risk
factor for hemorrhagic and nonhemorrhagic stroke. Arch Intern
Med 1989; 149:2053.

44. Whisnant JP, Homer D, Ingall TJ, Baker HL Jr, O'Fallon WM,
Wiebers DO. Duration of cigarette smoking is the strongest pre-
dictor of severe extracranial carotid artery atherosclerosis. Stroke
1990; 21:707.

45. Haapanen A, Koskenvuo M, Kaprio 1, Kesaniemi YA, Heikkila K.
Carotid arteriosclerosis in identical twins discordant for cigarette
smoking. Circulation 1989; 80:10.

46. Wolfe PA, Kannel WB, Verter 1. Current status of risk factors for
stroke. Neurol Clin 1983; 1:317.

47. Liapis CD, Kakisis JD, Kostakis AG. Carotid stenosis: factors
affecting symptomatology. Stroke 2001; 32:2782.

230

chapter 20 Natural history of atherosclerosis in the lower extremity, carotid, and coronary circulations

48. Moneta GL, Taylor DC, Zierler RE, Kazmers A, Beach K,
Strandness DE Jr. Asymptomatic high-grade internal carotid
artery stenosis: is stratification according to risk factors or duplex
spectral analysis possible? / Vase Surg 1989; 10:475.

49. Norris JW, Zhu CZ. Stroke risk and critical carotid stenosis. } Neu-
rol Neurosurg Psychiatry 1990; 53:235.

50. Weinberger J, Ramos L, Ambrose JA, Fuster V. Morphologic and
dynamic changes of atherosclerotic plaque at the carotid artery
bifurcation: sequential imaging by real-time B-mode ultrasono-
graphy. J Am Coll Cardiol 1988; 12:15151.

51. Gray-Weale AC, Graham JC, Burnett JR, Byrne K, Lusby RJ.
Carotid artery atheroma: comparison of preoperative B-mode
ultrasound appearance with carotid endarterectomy specimen
pathology. / Cardiovasc Surg 1988; 29:676.

52. Persson AV, Robichaux WT, Silverman M. The natural history of
carotid plaque development. Arch Surg 1983; 118:1048.

53. Imparato AM, Riles TS, Mintzer R, Baumann FG. The importance
of hemorrhage in the relationship between gross morphologic
characteristics and cerebral symptoms in 376 carotid artery
plaques. Ann Surg 1983; 197:195.

54. Lusby RJ, Ferrell LD, Ehrenfeld WK, Stoney RJ, Wylie EL. Carotid
plaque hemorrhage: its role in production of cerebral ischemia.
Arch Surg 1982; 117:1479.

55. Langsfeld M, Gray-Weale AC, Lusby RJ. The role of plaque
morphology and diameter reduction in the development of new
symptoms in asymptomatic carotid arteries. / Vase Surg 1989;
9:548.

56. Ammar AD, Ernst RL, Lin JJ, Travers H. The influence of repeated
carotid plaque hemorrhages on the production of cerebrovascular
symptoms. / Vase Surg 1986; 3:857.

57. Sterpetti AV, Schultz RD, Feldhaus RJ et al. Ultrasonographic fea-
tures of carotid plaque and the risk of subsequent neurologic
deficits. Surgery 1988; 104:652.

58. Leahy AL, McCollum PT, Feeley TM et al. Duplex ultrasono-
graphy and selection of patients for carotid endarterectomy:
plaque morphology or luminal narrowing? / Vase Surg 1988; 8:
558.

59. Feeley TM, Leen EL, Colgan MP, Moore DL, Hourihane DO'B,
Shanik GD. Histologic characteristics of carotid artery plaque. /
Vase Surg 1991; 13:719.

60. Moore WS, Boren C, Malone LM et al. Natural history of non-
stenotic, asymptomatic ulcerative lesions of the carotid artery.
Arch Surg 1978; 113:1352.

61. Harward TRS, Kroener JM, Wickbom IG, Bernstein EF. Natural
history of asymptomatic ulcerative plaques of the carotid bifurca-
tion. Am } Surg 1983; 146:208.

62. Zukowski AJ, Nicolaides AN, Lewis RT et al. The correlation
between carotid plaque ulceration and cerebral infarction seen on
CT scan. / Vase Surg 1984; 1:782.

63. Biasi GM, Mingazzini PM, Baronio L et al. Carotid plaque charac-
terization using digital image processing and its potential in fu-
ture studies of carotid endarterectomy and angioplasty. / Endovasc
Surg 1998; 5:240.

64. Tegos TJ, Stavropoulos P, Sabetai MM et al. Determinants of
carotid plaque instability: echoicity versus heterogeneity. Eur J
Vase Endovasc Surg 2001; 22:22.

65. Sabetai MM, Tegos TJ, Nicolaides AN et al. Hemispheric
symptoms and carotid plaque echomorphology / Vase Surg 2000;
31:39.

66. Kannel WB, Feinleib M. Natural history of angina pectoris in
the Framingham Study: progress and survival. Am } Cardiol 1972;
29:154.

67. Detre K, Peduzzi P, Murphy M et al. Effect of bypass surgery on
survival in patients with low- and high-risk subgroups delineated
by the use of simple clinical variables. Circulation 1981; 63: 1329.

68. Kaiser GC, Davis KB, Fisher LD et al. Survival following coronary
artery bypass grafting in patients with severe angina pectoris
(CASS). J Thorac Cardiovasc Surg 1985; 89:513.

69. Moise A, Theroux P, Taeymans Y et al. Unstable angina and pro-
gression of coronary atherosclerosis. N Engl} Med 1983; 309:685.

70. Roberts KB, Califf RM, Harrell FE Jr, Lee KL, Pryor DB, Rosati RA.
The prognosis for patients with new-onset angina who have
undergone cardiac catheterization. Circulation 1983; 68:970.

71. Mulcahy R, Awadhi AHA, de Buitleor M, Tobin G, Johnson H,
Contoy R. Natural history and prognosis of unstable angina. Am
H^rf J 1985; 109:753.

72. Gotto AM Jr, Farmer JA. Risk factors for coronary artery disease.
In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular
Medicine, 3rd edn, Vol. 2. Philadelphia: WB Saunders, 1988:1153.

73. Shapiro S, Weinblatt E, Frank CW, Sager RV Incidence of coronary
heart disease in a population insured for medical care (HIP):
myocardial infarction, angina pectoris, and possible myocardial
infarction. Am J Public Health 1969; 59:1 .

74. Starnler J, Starnler R, Liu KJ. High bloodpressure. In: Connor WE,
Bristow JD, eds. Coronary Heart Disease: Prevention, Complications
and Treatment. Philadelphia: JB Lippincott, 1985:85.

75. Garcia MJ, McNamara PM, Gordon T, Kannell WB. Morbidity and
mortality in diabetics in the Framingham population. Diabetes
1974; 23:105.

76. European Coronary Surgery Study Group. Long-term results of
prospective randomized study of coronary artery bypass surgery
in stable angina pectoris, Lancet 1982; 2: 1173.

77. Sam G, Castaftero A, Betriu A et al. Determinants of prognosis in
survivors of myocardial infarction: a prospective clinical angio-
graphic study. N Engl} Med 1982; 306:1065.

78. Mock MB, Ringqvist I, Fisher LD et al. and participants in the
Coronary Artery Surgery Study. Survival of medically treated
patients in Coronary Artery Surgery Study (CASS) registry.
Circulation 1982; 66:562.

79. Reeves TJ, Oberman A, Jones WB, Sheffield LT. Natural history of
angina pectoris. Am J Cardiol 1974; 33:423.

80. Califf RM, Tomabechi Y, Lee KL et al. Outcome in one-vessel coro-
nary artery disease. Circulation 1983; 67:283.

81. Burggraf GW, Parker JO. Prognosis in coronary artery disease:
angiographic hemodynamic and clinical factors. Circulation 1975;
51:146.

82. Humphries JO, Kuller L, Ross RS, Friesinger GC, Page EE. Natural
history of ischemic heart disease in relation to arteriographic
findings. Circulation 1974; 49:489.

83. Oberman A, Jones WB, Riley CP, Reeves TJ, Sheffield LT. Natural
history of coronary artery disease. Bull NY Acad Med 1972;
48:11095.

84. Harris PJ, Bahar VS, Conley MJ et al. The prognostic significance of
50% coronary stenosis in medically treated patients with coronary
artery disease. Circulation 1980; 62:240.

85. Ambrose JA, Winters SL, Stem A et al. Angiographic morphology
and the pathogenesis of unstable angina pectoris. ] Am Coll Cardiol
1985; 5:609.

231

pa rt I Vascular pathology and physiology

86. Ambrose JA, Winters SL, Arora RR et ah Angiographic evolution morphology of coronary stenoses leading to myocardial infarc-
of coronary artery morphology in unstable angina. / Am Coll Car- tion or unstable angina. Circulation 1986; 73:286.

diol 1986; 7:472. 88. Sherman CT, Litvack F, Grundfest W et al. Coronary angioscopy

87. Wilson RF, Holida MD, White CW. Quantitative angiographic with unstable angina pectoris. N Engl} Med 1986; 315:913.

232

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Neurologic basis for sympathetically
maintained pain: causalgia and reflex
sympathetic dystrophy

Marco Scoccianti
Rodney A. White

It is commonly accepted that Weir Mitchell, G.R. Moorhouse,
and W.W. Keen in 1864 were the first to accurately describe
a syndrome characterized by persistent burning pain and
trophic and behavioral changes in Civil War soldiers who
had sustained gunshot wounds to their limbs. 1 It was R.
Dunglison, however, who coined the term causalgia (from the
Greek words kausos, ''burning/' and algos, "pain") that
Mitchell then used in the 1872 edition of his book, Injuries of
Nerves and Their Consequences. 2

Since then, a wide spectrum of similar clinical conditions
with clear signs of sympathetic overactivity has been de-
scribed in the literature, and different terms have been used to
illustrate them (Table 21.1). A widely established term is reflex
sympathetic dystrophy (RSD) first used by Evans 3 and later
reproposed by Bonica. 4 Because the manifestations of
both causalgia and RSD are eliminated by sympathectomy,
Roberts 5 in 1986 proposed the unifying name of sympathetically
maintained pain (SMP).

In 1986, the International Association for the Study of
Pain (IASP) defined causalgia as a "syndrome of sustained,
diffuse, burning pain, allodynia, and hyperpathia after
traumatic nerve lesions, often combined with vasomotor
and sudomotor disturbances and later trophic changes." IASP
also proposed to reserve the term reflex sympathetic dystrophy
for those conditions with similar symptoms but without major
nerve injury. 6

Pain pathways

Cutaneous, deep, and visceral sensations are carried to the
central nervous system (CNS) by three types of primary affer-
ent fibers: large myelinated A-(3 fibers that transmit impulses
from mechanoreceptors; small myelinated A-5 fibers that
transmit impulses from both primary afferent nociceptors
(PANs) and mechanothermal receptors; and small unmyeli-
nated C fibers originating from polymodal PANs (Table 21.2).
Therefore, almost all of the noxious stimuli are carried by A-5
and C fibers. Evidence suggests, however, that in particular

situations such as causalgia, pain can be transmitted by A-p
fibers or when the PANs are not activated. 5,7 ~ 9

The primary afferent fibers, having their cell bodies in the
dorsal root ganglia, enter the spinal cord and reach specific
areas in the dorsal horn where they make polysynaptic
reflex connections that relay impulses to the cerebral cortex.
Nociceptive fibers mainly project to laminae I, II, and V of
the dorsal horn (Fig. 21.1). These laminae also contain projec-
tions from wide dynamic range neurons that are local in-
terneurons receiving inputs from nonnociceptive primary
afferents as well as from PANs. They are called wide dynamic
range neurons because they are excited by innocuous
mechanical or thermal stimuli but can increase their discharge
frequency as the stimulus intensity increases and becomes
damaging to tissue. 7

From the spinal cord, second order neurons ascend to the
thalamus (mainly the ventrobasal nucleus) in the lateral
spinothalamic tract. From the thalamus, the third order
neurons project to the somatosensory areas of the cortex
(postcentral gyrus and lateral cerebral sulcus), where
the painful stimulus is recognized as painful (perception). In
addition to the spinothalamic tract, nociceptive fibers also
project to the brain stem via the spinoreticulothalamic tract,
which connects the reticular formation and periaqueductal
gray to the thalamus.

The nociceptive message is not relayed unchanged to the
supraspinal regions but undergoes significant transforma-
tions in the dorsal horn of the spinal cord, which acts as a gate
or impulse processor. In fact, the activity evoked in spinal neu-
rons by PANs is modified by activity in other PANs within
the spinal cord, activity in nonnociceptive primary afferents,
and activity in descending projections from the supraspinal
structures. 10

Furthermore, transmission of nociceptive stimuli is inhib-
ited by a specific CNS network that has endogenous opioid
peptides as neurotransmitters 11 ' 12 and also by the myelinated
peripheral afferents (afferent inhibition). Actually, when these
fibers are blocked, the response of dorsal horn neurons to
noxious stimuli is greatly exaggerated (hyperalgesia). 13 ' 14

233

pa rt I Vascular pathology and physiology

Mechanisms for sympathetically
maintained pain

In both causalgia and RSD, pain may occur spontaneously
or in response to the slightest sensory stimulus, becoming
progressively worse and spreading to larger or even unrelated
areas of the body 7 Common accompaniments of pain are signs
of vasomotor instability and excessive sweating and swelling
of the affected extremity These symptoms and the fact they are
immediately alleviated by sympathectomy have prompted a
search to explain the role of the sympathetic nervous system
in starting and sustaining these syndromes. Several theories

Table 21 .1 Some terms applied to reflex sympathetic dystrophy

Causalgia
Minor causalgia
Mimocausalgia
Acute atrophy of bone
Sudeck's atrophy
Traumatic angiospasm
Traumatic vasospasm
Post-traumatic osteoporosis
Post-traumatic pain syndrome
Post-traumatic spreading neuralgia
Post-traumatic vasomotor disorder
Post-traumatic edema
Chronictraumatic edema
Peripheral acute trophoneurosis
Reflex dystrophy of the extremities
Reflex neurovascular dystrophy
Algoneurodystrophy
Shoulder-hand syndrome
Sympathalgia
Causalgia-dystonia syndrome

have then been proposed, but it is likely that more than one
contribute to the variegated manifestations of this disease.

Peripheral mechanisms

In 1944, Granit and colleagues 15 proposed that after nerve
damage, regenerating sprouts of sympathetic efferents and of
PANs form an artificial electrical junction (synapse) called an
ephapse. This causes discharges in sympathetic efferents to
directly activate adjacent pain afferents. Evidence exists
that ephaptic connections occur in neuromas. 16-18 Devor 16
similarly proposed the presence of a chemical ephapse— a
noradrenergic synapse between the sympathetic postgan-
glionic axons and the PANs within a neuroma. The neuroma
model, however, does not explain the immediate onset of pain
typical of causalgia, the efficacy of nerve blocks performed
distal to the nerve injury in relieving the pain, and the pro-
longed period of action of sympathetic blocks. In addition, pa-
tients with RSD do not have any significant nerve injury.

To explain the nature of the burning pain, of hypoalgesia
and hyperalgesia, researchers 18-20 proposed the existence of a
cross-modality threshold sensitization and of a dishomogeneous
sensitization of polymodal nociceptors. Whereas burning
pain normally requires the activation of specific A-5 cold
receptors and C-fiber nociceptors, some patients with SMP
have C nociceptor sensitization to both thermal and mechani-
cal stimuli. At the same time, the intensity of the perceived
pain is related to the status (frequency of discharge) of these
receptors, with hypoalgesia reflecting receptor fatigue and
hyperalgesia receptor sensitization. 21

Another possibility is that nonadrenergic and non-
cholinergic sympathetic cotransmitters are involved in mediat-
ing SMP by activation of PANs. 22 ' 23 Neuropeptide Y and
adenosine triphosphate have been found in sympathetic post-
ganglionic neurons, and their release depends on the state of
activity of these neurons. The release of specific cotransmitters

Table 21.2 Characteristics of sensory fibers and receptors

A-p fibers

A-8 fibers

C fibers

Dorsal horn projections

Transmission

Mostly sensitive to

Fiber diameter (jim)

Coverage

Conduction velocity (m/s)

Receptors*

Laminae III, IV, V, VI

Touch, pressure, mechanoreceptors

Pressure

6-22

Heavily myelinated

33-75

Mechanoreceptors

Nonnociceptive primary afferents

Laminae I, V

Fast (first) pain; sharp sensation

Pressure

2-5

Poorly myelinated

5-30

PANs: High-threshold mechanoreceptor

High-threshold mechanothermal receptor

Laminae I, II

Slow (second) pain; dull sensation

Local anesthetics

0.3-3

Unmyelinated

0.5

PANs: Polymodal — pain, temperature

*PANs, primary afferent nociceptors. The peripheral terminals of PANs are sensitive to thermal, mechanical, or chemical stimuli. The polymodal PANs can be
activated by all of them, and their axons are C fibers. The other PANs respond only to intense mechanical or mechanothermal stimulation (high-threshold PANs),
and their fibers are A-8.

234

chapter 21 Neurologic basis for sympathetically maintained pain

DESCENDING
CONTROLS

Heavily myelinated
OOOOOOCXD Lightly myelinated
jd Unmyelinated
_ Ascending pathway
- Descending pathway

PrGS

Fig. 21 .1 Schematic representation of a peripheral nerve and its projections to the spinal cord. The shaded areas of the dorsal horn (laminae I, II, and V or
Rexed) are of special importance for nociception. DRG, dorsal root ganglion; PoGS, postganglionic sympathetic; SG, sympathetic ganglion; PrGS, preganglionic
sympathetic; IML, intermediolateral column; WDR, wide dynamic range neurons.

in the dorsal horn and in the peripheral terminals may then be
associated with different types of sensation. It has been shown
that nociceptive fibers mediating temperature liberate so-
matostatin, mechanociceptive fibers substance P, and other
pain afferents glutamate. 24-28

Central mechanisms

There are numerous reasons to believe the CNS plays a critical
role in these syndromes: 29 causalgia may occur spontaneously
or with diseases localized to the CNS; the distribution of pain
does not always follow the pattern of a spinal or cranial nerve
distribution; the sympathetic block may be effective even
when the lesion is proximal to the block; patients with causal-
gia may have disturbances of motor function with tremor,
spasms, dystonia, and so forth; and causalgia is worsened by
stress and emotional stimuli.

Livingston in 1943 proposed his vicious circle hypothesis. 30 He
suggested that a cutaneous stimulus activates a PAN which in
turn activates the preganglionic sympathetic neuron in the
intermediolateral column of the spinal cord. This activates
the noradrenergic postganglionic neuron in the sympathetic
ganglion, which sensitizes and activates other PANs that
feedback to the spinal cord, maintaining the pain (Fig. 21.2).

Bonica 31 and Melzack and Wall 32 formulated the hypothesis
of a central biasing mechanism. According to them, the reticular
formation acts as a central biasing mechanism by exerting a
tonic inhibitory control (bias) on the gate of the dorsal horn, as
well as on other segments of the neuraxis. This inhibitory
action depends on a normal sensory input. Lesions of large
myelinated afferent fibers alter this input and result in an
inhibition of the reticular center, thus causing persistent pain.
This hypothesis explains the therapeutic mechanism of
electrical stimulation, in which an increase in the sensory
input increases the inhibition and decreases the pain.

Roberts 5 pointed out the importance of sympathetic activa-
tion of low-threshold afferents. 8,9 ,33 He posited that in certain con-
ditions, spontaneous and touch-evoked pain are mediated by
low-threshold mechanoreceptors and not by nociceptors. In
these situations, trauma to peripheral tissues activates C
nociceptors that in turn activate wide dynamic range neurons
that send axons through traditional pain pathways to the cor-
tex. Once sensitized, wide dynamic range neurons respond to
A-(3 afferents (mechanoreceptors) activated by light touch
(Fig. 21.3). This state produces allodynia (pain to nonnoxious
stimulation). Wide dynamic range neurons also respond to A-
(3 afferents stimulated by local sympathetic afferents in the ab-
sence of any stimulation, and this explains the spontaneous

235

pa rt I Vascular pathology and physiology

Fig. 21.2 Vicious circle hypothesisof Livingston. A tissue-damaging
stimulus (S) activates one primary afferent nociceptor (PAN 1 ), which
activates the sympathetic preganglionic neuron in the intermediolateral

column (IML). The activated postganglionic neuron activates more PANs (PAN
2, PAN 3). The sympathetic system thus spreads and maintains the pain.

DESCENDING
CONTROLS

Fig. 21 .3 Sympathetically maintained pain hypothesis of Roberts. Wide
dynamic range neurons (WDR) receive afferent input from both nociceptive
and nonnociceptive receptors. Once sensitized, they perceive every stimulus

(S) as painful. Tonic sympathetic activation of low-threshold
mechanoreceptors maintains WDR neuron sensitization.

236

chapter 21 Neurologic basis for sympathetically maintained pain

occurrence of pain. This hypothesis may also explain the rela-
tion between CNS activity or disorders and RSD. Major evi-
dence that low-threshold mechanoreceptors are involved in
SMP is given by the fact that selective blockage of large myeli-
nated afferents abolishes spontaneous pain and allodynia 34
and that the threshold intensity for pain in RSD patients is low
and similar to that for touch perception.

In conclusion, evidence suggests that both peripheral sensi-
tization of receptors and central sensitization of dorsal horn
neurons contribute to the pain of causalgia and RSD and that
the sympathetic system is instrumental in maintaining or
exacerbating it. 7/35

Mechanisms of related sympathetic
dysfunction

A common component of RSD, edema is a result of the axon
reflex. Stimulation of a peripheral nerve releases several
neuropeptides from the peripheral terminals of PANs. 36/37
Substance P has been shown to increase capillary permeability
and to liberate histamine from mast cells. Calcitonin gene-
related peptide (CGRP) causes vascular smooth muscle
relaxation, 38 and neurokinine A and B produce vasodilatation
and increase vascular permeability 39 The end-result is called
reflex neurogenic inflammation. 40 Another aspect of RSD is the
generalized dysfunction of the cutaneous circulation. It has
been demonstrated 41 that stimulation of arterial baroreceptors
and chemoreceptors after nerve injury causes an altered reflex
discharge of postganglionic sympathetic neurons, which
is probably secondary to an altered reflex control of vasocon-
strictive neurons at the spinal level. 42

Patients with SMP have variation in the skin temperature of
the affected extremity. Typically, the skin is warm and red in
the early stages and later becomes cold and pale. A possible
mechanism for this phenomenon is that nerve damage causes
the rapid onset of spontaneous firing of A-5 fibers and this
would increase sympathetic efferent activity 43 Hypothermia
would then be maintained by active vasoconstriction. Instead,
patients with warm extremities may show C nociceptor
overactivity, with resultant neurogenic vasodilatation
that predominates over the vasoconstrictive sympathetic
activity. 19

The sympathetic nervous system has critical effects on
skeletal muscle contraction, neuromuscular transmission, and
spinal cord reflexes. Sympathetic projections have been found
in the bag and the intrafusal fibers of the muscle spindle, and in
the ventral efferent fibers. Sympathetic nerve stimulation
increases firing of muscle spindle afferents; therefore, the
increase in muscle tone, reflexes, and spasms seen in SMP may
be spindle mediated by an interaction with the sympathetic
system. 35

Clinical features of
sympathetically maintained pain

Etiology

Causalgia is rare and affects 1-5% of patients with peripheral
nerve injuries. Almost all develop it after high-velocity missile
injuries that cause stretching or shearing of the nerve without
complete transection. However, it can also be found after
avulsion of the roots of the brachial plexus such as after a
motorcycle accident. 44 Although the pain is usually referred to
the distal part of the affected extremity, most of the peripheral
nerve lesions associated with causalgia are above the elbow or
the knee.

Bonica 45 found that the upper extremity was involved in
56.8% of cases and the lower extremity in 41 .4%, with the most
commonly injured nerves being the sciatic nerve (27.3%), the
median nerve (18%), and the brachial plexus (10.5%). The high
incidence of sciatic and median nerve injuries is attributed to
the fact that these two nerves carry most of the sensory and
postganglionic sympathetic fibers of the hand and foot. 45

RSD is not associated (by definition) with major nerve injury
but usually develops after some sort of trauma— over 50% of
recognized cases are seen after peripheral bone fractures. 46 It is
well known that RSD can develop even after minor injuries
such as sprains, dislocations, and skin lacerations. Less
often, it is found after myocardial infarction (hand-shoulder
syndrome), cerebrovascular accidents, inflammatory dis-
eases, degenerative joint diseases, nerve infiltration by
metastatic cancer, burns, or drug use (phenobarbital). In 35%
of patients, no cause can be found. 47

Manifestations

Pain is the key feature of causalgia, starting soon after the
injury in 37% of patients, in the first week in 45%, and after 1
month only in 5%. 45 The pain is usually burning and superfi-
cial and referred to the fingers and toes, but 66% of patients
also complain of episodes of stabbing, tearing, or crushing
pain. The pain is so severe as to preclude the usual daily
activities and eventually causes profound behavioral and
personality changes. It is initially localized to the distribution
of the affected nerve; if left untreated, however, it quickly
spreads to the entire limb, becomes diffuse and poorly local-
ized, and may even extend to the contralateral limb. Typically,
the pain is aggravated by the slightest movement or touch,
any change in temperature, and by any visual, auditory, or
emotional stimulus.

The duration of pain is usually related to its intensity, with
mild pain lasting 1-3 months and severe pain even more than
1 year; Bonica 45 found that pain persisted more than 6 months
in 85% of patients and more than 1 year in 25%. The pain is
often associated with hyperpathia (delay, overreaction, and

237

pa rt I Vascular pathology and physiology

afterreaction to a stimulus) and allodynia (pain after
nontissue-damaging stimulation).

As causalgia progresses into the later stages, trophic
changes appear and involve the skin, subcutaneous tissues,
and bones and joints. The skin becomes red and glossy, the
fingers become tapering as a result of loss of subcutaneous fat,
and the nails become curved and brittle. Eventually, ankylosis
of the interphalangeal joints, muscle wasting and contrac-
tions, and osteoporosis supervene.

RSD presents some differences from causalgia: the pain usu-
ally starts several weeks or months after the triggering event,
and it does not match a peripheral nerve distribution. In this
syndrome, one symptom is frequently out of proportion to the
others: some patients have only pain, whereas edema is the
only sign in others. If left untreated, RSD evolves into three
classic stages 48 :

First acute stage: characterized by pain, hyperalgesia, allody-
nia, and signs suggestive of sympathetic denervation or un-
deractivity, including increased skin blood flow, increased
skin temperature, heat intolerance, and accelerated nail and
hair growth. At this stage, bone radiographs are normal and
bone scans show only increased uptake by the small joints.
Second dystrophic stage: characterized by signs of sym-
pathetic overactivity with decreased blood flow, decreased
skin temperature, cold intolerance, and decreased nail and
hair growth. If present, edema becomes brawny; muscle
wasting appears. Radiographs show spotty osteoporosis,
and bone scans show normalization of blood flow.
Third atrophic stage: characterized by irreversible trophic
changes. The subcutaneous tissue becomes thin, leading to
a glossy appearance of the skin, and the muscles become
atrophic and the joints ankylosed with flexor contractures.
Behavioral and emotional changes secondary to chronic
pain are common. Radiographs show diffuse osteoporosis,
and bone scans reveal bone hypofixation.

Diagnosis

The diagnosis of SMP is usually made on the basis of patient
history and clinical findings. However, several diagnostic
methods based on measuring abnormalities of sympathetic
tone 47 and on radiologic imaging can confirm the clinical
impression.

Skin blood flow can be measured by plethysmography, by
xenon-133 clearance methods, or by thermography (con-
sidered positive if there is a temperature increase or decrease
of at least 1°C in the painful area compared with the opposite
side). 49 The sudomotor function can be assessed by skin
galvanic resistance or by a sweat test.

Plain radiographs may show patchy demineralization of
the epiphyses and short bones of the diseased extremity
(Sudeck's atrophy), and three-phase radionuclide bone scans
may demonstrate increased blood flow in the angiogram

phase and increased uptake of the radionuclide in the blood
pool (early static) and delayed (late static) phases of the study.
Kozin and associates 50 suggested that three-phase radionu-
clide bone scanning is more specific than radiography and that
the increased uptake of radionuclide may reflect the vasomo-
tor instability of these patients. They also noted a correlation
between the positivity of the scan and the response to cortico-
steroids, suggesting scintigraphy could be useful to assess the
response of patients to treatment.

The most valuable diagnostic test remains the sympathetic
block. In fact, relief of pain or modification of signs after a
regional sympathetic block is considered typical of causalgia
and of RSD. It is helpful not only in confirming the diagnosis
of SMP but also in predicting the response to surgical sym-
pathectomy. However, Bonica 45 noted that to be a reliable
diagnostic— prognostic test, the sympathetic block should
produce complete denervation of the affected limb. This
means that it should extend in the upper extremity from the
middle cervical ganglion to the fourth thoracic ganglion, and
in the lower extremity from the lowest thoracic ganglion to the
fourth lumbar ganglion. After the block, a skin conductance
test or a sweat test should be done to confirm the completeness
of the procedure.

Patients in the third stage of RSD may not respond to a sym-
pathetic block because the pain produced by secondary mus-
culoskeletal changes has become more important than the
SMP.

Treatment

Interruption of the activity in the efferent sympathetic fibers
is the objective of the therapy in SMP. It can be achieved by
temporary or permanent anesthetic block of the sympathetic
ganglia, by oral or intravenous administration of adrenergic
blocking drugs, or by surgical sympathectomy.

A regional sympathetic block should always be tried in
these patients and, if successful, repeated with a long-lasting
agent such as 0.25% bupivacaine to maintain continuous pain
relief. If SMP is diagnosed at an early stage, such a block may
result in cure in as many as 70% of patients. 51 A successful
block produces almost instant remission of pain, an increase in
the skin temperature with dry skin, a decrease in edema, and
an increase in mobility of the affected limb. Typically, the block
lasts for several days or even weeks, well beyond the effect of
the pharmacologic agent used. 8 If the relief of pain after the
block becomes too short, infusion of the anesthetic through an
epidural catheter may be useful, but it is more likely that surgi-
cal sympathectomy is indicated at this stage.

An intravenous regional sympathetic block with guan-
ethidine or reserpine 52 is indicated in patients receiving
anticoagulants or when a local block may be difficult, as in
postoperative patients. Guanethidine acts by depleting
norepinephrine from synaptic vesicles in sympathetic efferent

238

chapter 21 Neurologic basis for sympathetically maintained pain

fibers and is effective in producing pain relief in 60-80% of
patients. 53

Surgical sympathectomy has a well-established role in the
treatment of both causalgia and RSD. 45/47/54_56 It is mainly indi-
cated in patients who have responded initially to a local or re-
gional sympathetic block but in whom the efficacy of the block
becomes shorter and shorter. It is especially beneficial in the
early stages of the disease, when it is successful in relieving the
symptoms in 97% of patients and in obtaining long-term cure
in 89-95% of cases. 54

Patients who usually do not respond well to sympa-
thectomy are those referred too late and who have already
developed significant trophic changes (third stage). Even in
these patients, however, sympathectomy can alleviate some of
the symptoms. The procedure has a low morbidity rate and
almost no mortality; the only disturbing complication is
postsympathectomy neuralgia, which can occur in 40% of
patients. However, it is usually easily treatable with oral anal-
gesics and almost invariably disappears in 2-3 months. 54 ' 55
Lumbar sympathectomy is usually performed through a
retroperitoneal approach via a flank incision. All ganglia from
LI to L4 are removed. Dorsal sympathectomy can be carried
out through a transpleural approach, entering the third inter-
costal space, or through an extrapleural approach, with resec-
tion of the first rib. The resected chain includes the lower part
of the stellate ganglion and the Tl, T2, and T3 ganglia.

Several pharmacologic agents have also been used in
patients with SMP, including oral sympatholytics, calcium
channel blockers, steroids, tricyclic antidepressants, and
anticonvulsants, 47 but they have been found useful only in
selected subgroups and are not considered for wide use. One
treatment that has been found useful in patients with pain is
transcutaneous nerve stimulation. To be effective, however, it
should be applied proximally to the lesion, and several combi-
nations of repetition rates, pulse frequencies, and pulse widths
should be tried to get a satisfactory result. 52

Of critical importance if long-term success is to be
achieved in these patients is early and continuous use of
physical therapy. It should be started soon after a local block
has produced pain relief and should include limb elevation,
pressure pumps, and range-of-motion and stretching

exercises.

52,57

Summary

In conclusion, considerable advances have been made in the
understanding of SMP and its treatment. The role of the sym-
pathetic system has been elucidated, and the utility of local
sympathetic blocks and of surgical sympathectomy are now
well established. Emphasis should be given to an early
diagnosis and treatment of this now-curable condition,
which if left undiagnosed, progresses to irreversible
functional changes and permanent disability.

References

1. Mitchell SW, Moorehouse GR, Keen WW. Gunshot Wounds and
Other Injuries of Nerves. Philadelphia: JB Lippincott, 1864.

2. Mitchell SW. Injuries of Nerves and Their Consequences. London:
Smith Elder, 1872.

3. Evans JA. Reflex sympathetic dystrophy. Surg Gynecol Obstet 1946;
82:36.

4. Bonica JJ. The Management of Pain, 1st edn. Philadelphia: Lea &
Febiger, 1953:913.

5. Roberts WJ. An hypothesis on the physiological basis for causalgia
and related pains. Pain 1986; 24:297.

6. Merskey H. Classification of chronic pain: descriptions of chronic
pain syndromes and definitions of pain terms. Pain 1986;
28(Suppl):l.

7. Roberts WJ, Kramis RC. Sympathetic nervous system influence
on acute and chronic pain. In: Fields HL, ed. Pain Syndromes in
Neurology. London: Butterworths, 1990:87.

8. Loh L, Nathan PW. Painful peripheral states and sympathetic
blocks. / Neurol Neurosurg Psychiatry 1978; 41:664.

9. Wiesenfeld-Hallin Z, Hallin RG. The influence of the sympathetic
system on mechanoreception and nociception: a review. Hum
Neurobioll984:}3-Al.

10. Fields HL. Introduction. In: Fields HL, ed. Pain Syndromes in
Neurology. London: Butterworths, 1990:6.

11. Basbaum AI, Fields HL. Endogenous pain control mechanisms:
review and hypothesis. Ann Neurol 1978; 4:451.

12. Fields HL, Basbaum AI. Brainstem control of spinal pain transmis-
sion neurons. Annu Rev Physiol 1978; 40:217.

13. Torebjork HE, Hallin RG. Identification of afferent C units in intact
human skin nerves. Brain Res 1973; 16:321.

14. Campbell JN, Long DM. Peripheral nerve stimulation in the
treatment of intractable pain. / Neurosurg 1976; 45:692.

15. Granit R, Leksell L, Skoglund CR. Fibre interaction in injured or
compressed region of nerve. Brain 1944; 67:125.

16. Devor M. Nerve pathophysiology and mechanisms of pain in
causalgia. JAutonom New Syst 1983; 7:371.

17. Seltzer Z, Devor M. Ectopic transmission in chronically damaged
peripheral nerves. Neurology 1979; 29:1061.

18. Sato J, Perl ER. Adrenergic excitation of cutaneous pain receptors
induced by peripheral nerve injury. Science 1991; 251:1608.

19. Ochoa JL. The newly recognized painful ABC syndrome: thermo-
graphic aspects. Thermology 1986; 2:65.

20. Ochoa JL, Torebjork HE. Paresthesia from ectopic impulse
generation in human sensory nerves. Brain 1980; 103:835.

21. Torebjork HE, Lamotte RH, Robinson CJ. Peripheral neural corre-
lates of magnitude of cutaneous pain and hyperalgesia: simulta-
neous recordings in humans of sensory judgments of pain and
evoked responses in nociceptors with C fibers. / Neurophysiol 1984;
2:325.

22. Burnstock G. Do some nerve cells release more than one transmit-
ter? Neuroscience 1976; 1:239.

23. Pernow J. Co-release and functional interactions of neuropeptide
Y and noradrenaline in peripheral sympathetic vascular control.
Acta Physiol Scand 1988; 133(Suppl. 568):1.

24. Curtis DR, Johnston GAR. Aminoacid transmitters in mammalian
central nervous system. Ergbn Physiol 1974; 69:97.

239

pa rt I Vascular pathology and physiology

25. Hokfelt T, Elde R, Johansson O. Immunohistochemical evidence
for separate populations of somatostatin-containing and
substance P-containing primary afferent neurons in the rat.
Neuroscience 1976; 1:131.

26. Hokfelt T, Johansson O, Ljungdahl A. Peptidergic neurons. Nature
1980; 284:515.

27. Kuraishi Y, Hirsta N, Sato Y Evidence that substance P and
somatostatin transmit separate information related to pain in
the spinal dorsal horn. Brain Res 1985; 325:294.

28. Wiesenfeld-Hallin Z. Substance P and somatostatin modulate
spinal excitability via physiologically different sensory pathways.
Brain Res 1986; 372:172.

29. Schott GD. Mechanisms of causalgia and related clinical condi-
tions. Brain 1986; 109:717.

30. Livingston WK. Pain Mechanisms: a Physiologic Interpretation of
Causalgia and Its Related States. New York: Macmillan, 1943:1.

31. Bonica JJ. General considerations of chronic pain. In: The Manage-
ment of Pain, Vol. 1, 2nd edn. Philadelphia: Lea & Febiger, 1990:187.

32. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;
150:971.

33. Maixner W, Dubner R, Bushnell MC, Kenshalo DR, Oliveras JL.
Wide dynamic range dorsal horn neurons participate in the
encoding process by which monkeys perceive the intensity of
noxious heat stimuli. Brain Res 1986; 374:385.

34. Campbell JN, Raja SN, Meyer RA, Mackinnon SE. Myelinated
afferents signal the hyperalgesia associated with nerve injury.
Pain 1988; 32:89.

35. Schwartzman RJ. Reflex sympathetic dystrophy and causalgia.
Neurol Clin 1992; 10:953.

36. Dalsgaard CJ, Jonsson CJ, Hokfelt T, Cuello AC. Localization of
substance P immunoreactive nerve fibers in the human digital
skin. Experimentia 1983; 39:1018.

37. Fuller RW, Conradson TB, Dixon CHS. Sensory neuropeptide
effects in human skin. Br J Pharmacol 1987; 92:781.

38. Brain SD, Williams TJ, Tippens JR. Calcitonin gene related peptide
is a potent vasodilator. Nature 1985; 313:54.

39. Foreman JC, Jordan CC, Oehme P, Renner H. Structure-activity
relationship for some substance P-related peptides that cause
wheal and flare reaction in human skin. / Physiol 1983; 335:449.

40. Levine JD, Dardick SJ, Basbaum AI, Scipio E. Reflex neurogenic
inflammation. I. Contribution of the peripheral nervous system
to spatially remote inflammatory responses that follow injury.
JNeurosci 1985; 3:1380.

41. Blumberg H, Griesser HJ, Hornyak HE. Mechanisms and role of
peripheral blood flow dysregulation in pain sensation and edema
in reflex sympathetic dystrophy. In: Stanton-Hicks M, Jang W,
Boas RA, eds. Reflex Sympathetic Dystrophy. Boston: Kluwer
Academic Publishers, 1990:8.

42. Blumberg H, Janig W. Reflex patterns in postganglionic vasocon-
strictor neurons following chronic nerve lesions. / Autonom New
Syst 1985; 14:157.

43. Beacham WS, Perl ER. Characteristic of a spinal sympathetic
reflex./ Physiol 1964; 173:431.

44. Horowitz SH. Iatrogenic causalgia: classification, clinical findings
legal ramifications. Arch Neurol 1984; 41 :82 1 .

45. Bonica JJ. Causalgia and other reflex sympathetic dystrophies. In:
The Management of Pain, Vol. 1, 2nd edn. Philadelphia: Lea &
Febiger, 1990:220.

46. Rowlingson JC. The sympathetic dystrophies. Int Anesthesiol Clin
1983; 21:117.

47. Payne R. Reflex sympathetic dystrophy syndrome: diagnosis and
treatment. In: Fields HL, ed. Pain Syndromes in Neurology. London:
Butterworths, 1990:107.

48. Demangeat JL, Constantinesco A, Brunot B, Foucher G, Farcot JM.
Three phase bone scan in reflex sympathetic dystrophy of the
hand. JNuclMed 1988; 29:26.

49. Ecker A. Contact thermography in diagnosis of reflex sympathetic
dystrophy: a new look at pathogenesis. Thermology 1985; 1:106.

50. Kozin F, Soin JS, Ryan LM, Carrera GF, Wortmann RL. Bone
scintigraphy in the reflex sympathetic dystrophy syndrome.
Radiology 1981; 138:437.

51. Wang JK, Johnson KA, Ilstrup DH. Sympathetic blocks for reflex
sympathetic dystrophy. Pain 1985; 23:13.

52. Girgis FL, Wynn-Parry CB. Management of causalgia after
peripheral nerve injury. Int Disabil Stud 1989; 11:15.

53. Hannington-Kiff JG. Relief of causalgia in limbs by regional
intravenous guanethidine. Br Med] 1979', 2:3.

54. Mockus MB, Rutherford RB, Rosales C, Pearce WH. Sym-
pathectomy for causalgia, patient selection and long-term results.
Arch Surg 1987; 122:668.

55. Rutherford RB. Reflex sympathetic dystrophy: a treatable cause of
post-traumatic and post-ischemic pain. Semin Vase Surg 1991; 4:31.

56. Shumacker HB Jr. A personal overview of causalgia and other
reflex dystrophies. Ann Surg 1985; 201:278.

57. Headley B. Historical perspective of causalgia: management of
sympathetically maintained pain. Phys Ther 1987; 67:1370.

240

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Compartment syndromes physiology

Malcolm O. Perry

Compartment syndrome has been described by Matsen
and colleagues "as a condition in which increased pressure
within a limited space compromises the circulation and func-
tion of the tissues within that space/' 1 The rise in intracom-
partmental pressure is usually the result of an increase in
interstitial fluid, but some studies have suggested that there
also may be cell swelling. 2-4 Direct muscle trauma from a
crushing injury, or bleeding within the compartment may
acutely raise the compartmental pressure, but more often
compartment syndromes are delayed and follow a period of
ischemia. 5 These clinical syndromes almost certainly are
caused by an injury to the capillaries that results in increased
permeability.

Etiology

Compartment syndromes usually are associated with trauma,
either as a result of direct injury to the skeletal muscle by blunt
trauma or fractures, or after a major arterial or venous injury
that produces significant periods of ischemia. In Patman's re-
port, 32% of the patients with an arterial injury underwent a
fasciotomy for clinical indications, yet he noted that only 2% of
his patients who had an arterial embolus required decompres-
sion of the compartments. 6 This is in sharp contrast to the
experience of Patman and associates when they performed
fasciotomy in only 0.45% of 2000 patients undergoing periph-
eral arterial reconstructions. As the following list shows, there
are many causes of compartment syndromes, but regardless of
the method of classification, most will be associated with
injuries of one type or another. 7

MECHANISMS OF ACUTE COMPARTMENT SYNDROMES

Decrease in compartment size

• Constriction by casts, dressing, pneumatic garments

• Surgical closure of fascial defects

• Eschar formation, thermal injuries, frostbite
Increase in compartment contents

• Edema

• Ischemia and reperfusion from arterial injuries, emboli,
or thrombosis, limb replantation, or tourniquet

• Limb compression and immobilization as a result of drug
overdose or general anesthesia for surgery

• Hemorrhage

• Trauma from fractures or vascular wounds

• Bleeding disorders

• Anticoagulant therapy

In patients where the increased compartmental pressures
are the result of direct muscular injury or the accumulation of
blood within the compartment, the pathophysiologic mecha-
nisms initially are different from those seen when ischemia is
the etiology. There may be a final common pathway, however.
As the blood accumulates within the compartment it can pro-
duce significant ischemia, which leads to injuries associated
with reperfusion syndromes related to increased capillary
permeability and an increase in interstitial fluid. 5

Hemodynamics

Blood flow through the capillaries depends on the gradient be-
tween arterial and venous blood pressure. These changes in
transmural pressure are central to the hypothesis of capillary
equilibrium, which accounts for the normal exchange of body
fluid between the capillaries and the interstitium. It is appar-
ent that situations that reduce the arteriovenous (AV) gradient
(decreasing the arterial pressure, increasing the venous pres-
sure, or increasing the interstitial hydrostatic pressure) will
ultimately result in a reduction in capillary blood flow.
The pressure at the arterial end of the capillary is usually
30-35 mmHg. It is apparent that when compartmental pres-
sures are in excess of this value, disturbances in the AV gradi-
ent will appear, and ischemia is likely to occur even though
systemic arterial pressure is normal. Peripheral pulses can
remain detectable beyond the area of increased compart-
mental pressure despite an ischemic process. 1/5/7 In patients
who are at risk for the development of compartment syn-
dromes, reductions in arterial pressure from any cause, or

241

pa rt I Vascular pathology and physiology

pathologic increases in venous pressure, are likely to aggra-
vate the ischemia. 1

There are studies showing that small, muscular arteries
require a minimal level of intraluminal pressure to remain
patent; a pressure lower than this has been called the "critical
closing pressure". 8 When the intraluminal arteriolar pressure
drops below 50 mmHg, small muscular arteries may close, and
to reopen these arteries requires a finite increase in pressure.

A similar phenomenon can result in closure of the thin post-
capillary venules; this probably occurs at a much lower inter-
stitial pressure. Because blood is a non-Newtonian fluid and
has an anomalous viscosity, there is a "yield pressure" that
must be achieved to reinitiate blood flow against inertia from a
state of rest. This may pose another impediment to reestablish-
ment of normal blood flow.

Pathophysiology

It is difficult to assess the tolerance of the extremities to is-
chemia because some cells are more susceptible, presumably
as a result of differences in oxygen requirements. Peripheral
nerves and muscles are relatively less resistant to ischemia.
The skin and subcutaneous tissues are capable of surviving
periods of hypoxia that are not tolerated by skeletal muscle or
peripheral nerves. There is a considerable body of evidence
suggesting that the outcome of tissues after a period of is-
chemia depends not only on the specific tissue tolerance to
hypoxia, and the duration of the ischemia, but also on local
changes that interfere with normal blood flow.

Ames and coworkers have called this "the impaired re-flow
phenomenon"; they assign the cause to a narrowing of the
vascular lumen secondary to compression by swollen cells. 9
These investigators also demonstrated capillary narrowing as
a result of intravascular blebs of capillary endothelium. Other
investigators have shown that red cells can be trapped in the
narrow capillaries, contributing further to impaired reflow.
Other studies confirmed that plugs made up of white blood
cells can be seen in skeletal muscle and the lungs during hem-
orrhagic shock 10 ' 11 ; this phenomenon also has been observed
in cardiac muscle capillaries after coronary artery occlusion.
These studies strongly suggest that cellular swelling does fol-
low some episodes of ischemia, and may play an important
role in the development of cell damage.

Early studies by Harman demonstrated that after complete
ischemia of the hind limb in rabbits there was a delay in the
penetration of injected bromophenol blue dye. 12 The vital dye
stained normal and mildly ischemic tissue immediately after
the tourniquet was released, but if the ischemia was extended
beyond 3 h there was considerable retention of the dye within
the muscle. If ischemia lasted for more than 6 h, the bromophe-
nol blue did not enter the muscle for at least 30 min. Local cir-
culatory stagnation was seen, and capillaries were filled with
red blood cells, but no clots were observed. Generalized

swelling of the tissue was documented by weight gain of the
ischemic limb.

Studies of the myocardium suggested that cellular edema
occurred with ischemia because of failure of the sodium
pump. Willerson and coworkers demonstrated that these
changes could be reduced by pretreatment with mannitol,
suggesting that this diuretic helps reduce cell swelling. 13
Other studies have shown that mannitol is a scavenger of the
toxic oxygen-derived hydroxyl radical. This free radical can
cause lipid peroxidation of the cell membrane; neutralization
of these radicals may prevent cellular damage that contributes
to increased permeability 2 ' 14

The maintenance of normal cell volume depends on tissue
respiration, and in the absence of oxygen tissues gain weight
because of an increase in water content. This suggests that
damage from these episodes may be a consequence not only of
swelling of individual cells, but also of the subsequent effect
on vascular perfusion. Because of the impaired reflow, the
final duration of the ischemic insult may be greater than is ap-
parent. These studies suggest that impaired reflow may be an
important phenomenon, although the exact mechanisms re-
sponsible have not been completely established. 3

Experimental evidence suggests that short periods of is-
chemia can cause cell damage that may not be apparent when
the overall function of the organ system is examined. Eklof and
colleagues demonstrated that temporary aortic occlusion in
patients undergoing vascular surgery resulted in metabolic
changes in the skeletal muscle of the legs that persisted for at
least 16 h. 15 Perry and associates in two experimental studies
in animals demonstrated that 3 h of partial ischemia at a mean
arterial pressure of 50 mmHg caused longer lasting cell mem-
brane dysfunction than 3h of tourniquet ischemia. 16 Subse-
quently, Perry and Fantini found that superoxide dismutase
prevented progressive cell membrane dysfunction in rats if
administered simultaneously with the restoration of aortic
blood flow. 3

Additional studies have shown that oxygen-derived free
radicals can be released by white blood cells. 14 Walker and col-
leagues found that skeletal muscle necrosis after 5 h of total
ischemia in dogs could be reduced by controlling oxygen de-
livery in the reperfusion period. 17 The beneficial effect of
hemodilution also was increased if free radical scavengers
were given.

All of these studies suggest that skeletal muscle is
susceptible to injury by oxygen-derived free radicals. This
phenomenon has been demonstrated in the intestine, kidney,
and heart. Furthermore, it appears that partial ischemia,
which often is the case in the clinical practice of surgery, causes
cell damage that in some instances is more severe than that
caused by total ischemia. 16 These observations suggest that
ischemic injury and reperfusion injury may occur simul-
taneously. Although the final common pathway of cell mem-
brane damage after episodes of ischemia has not been clearly
delineated, there is abundant evident that toxic mediators

242

chapter 22 Compartment syndromes physiology

cause direct cell membrane damage and increased cell
wall permeability

Compartmental pressures

Measurements of intracompartmental tissue pressures have
varied depending on the techniques used. One study using an
invasive technique found that the resting value of normal in-
tracompartmental pressure in humans was approximately 4 +
4mmHg. 1 ' 7,18 Although the value obtained by other methods
varies, it is clear that the normal resting compartment pressure
is less than the mean arterial capillary blood pressure. In most
situations the clinical features of a compartment syndrome
demonstrated by frequent examinations establish the diagno-
sis. In uncooperative patients and in those with multiple
injuries, measurements of compartmental pressures may be
required to determine critical pressure levels.

Initially measurements of intracompartmental pressure
were based on needle manometer techniques. Whitesides
et al.'s technique requires a needle, intravenous tubing, sa-
line, and a mercury manometer. 18 The needle is inserted into
the compartment, and after zero calibration of the apparatus,
pressure is measured. For longer term surveillance, Matsen
and associates used a Harvard pump and a slow infusion of
saline (0.7ml/day). He measured pressure continuously for
up to 3 days. 1 This technique has been shown to provide mea-
surements almost identical to those obtained with a wick
catheter, the most popular method of measuring compart-
mental pressures. In a wick catheter, fibrils of polyglycolic su-
ture protrude from the central lumen of the catheter and
provide increased surface area for fluid equilibration. The per-

meation of this wick by fluid permits direct contact with a large
volume of the interstitial fluid; the pressure is transmitted via
the catheter filled with sterile saline. This system is inherently
accurate and has been used widely for the measurement of
compartmental pressures. 7 A modification of the wick catheter
is the slit catheter, which, although a less expensive item, also
has been shown to provide accurate measurements. More re-
cently, a solid-state transducer instrument has been devel-
oped. The transducer is in the tip of the needle, and thus it is
automatically at the proper level when inserted into the com-
partment. The zero baseline is obtained electronically and
there is an immediate display of intracompartmental pres-
sures. The transducer-tipped instrument is easier to use and
can be used repeatedly; in many centers it has become the pre-
ferred method. 19

Continuing difficulties in establishing the diagnosis of
compartment syndromes led Perry and associates to investi-
gate noninvasive methods. 20 The normal resting pressure
in the recumbent tibial veins of humans is 7-16 mmHg.
This study was based on the concept that when intracompart-
mental pressures exceeded resting tibial venous blood pres-
sure, alterations in venous blood flow would be present and
these could be detected transcutaneously using a Doppler de-
vice. It was postulated that early rises in pressure would dis-
turb spontaneous, phasic, and augmented blood flow in the
veins. The hypothesis was validated in a prospective study
of 26 patients; the group was small, however. Nevertheless,
all patients who had abnormal Doppler venous flow had
increased compartmental pressures on subsequent ex-
amination. In patients who were treated by compartment
decompression, there was a return to normal pressure and
flow in the tibial veins.

Figure 22.1 Schematic diagram of the
compartments of the lower leg. The anterior
compartment is bounded by thick fascial layers
and is more vulnerable to the development of a
compartment syndrome. If an operation for
decompression is required, it is best to open all
four compartments.

Deep posterior comportment

Peroneal artery

Superficial posterior compartment

ibial nerve

Lateral compartment

Anterior compartment

Anterior tibial
artery

Posterior tibial
artery

Deep peroneal
nerve

243

pa rt I Vascular pathology and physiology

Controversy still surrounds what constitutes abnormal
pressures, although Matsen and associates have concluded
that any sustained pressure greater than 45 mmHg is capable
of inflicting injury 1 Although there is no agreement on the
level that produces injury, it is clear that sustained pressures
above capillary arterial pressures are capable of producing
damage.

Clinical features and syndromes

The clinical diagnosis of a compartment syndrome is based on
a careful neuromuscular evaluation. As described in the pre-
ceding sections, ischemic injury can occur with intracompart-
mental pressures of 40-50 mmHg, and therefore it is not
prudent to delay intervention until distal pulses disappear. On
examination, there may be tenseness of the compartment to
palpation, and circumferential enlargement in the extremity,
but in patients with multiple trauma or crush injuries these
may be less reliable indicators.

Evaluation of the function of the blood vessels and nerves
that traverse the compartment is important (Fig. 22.1). The
most vulnerable compartment is the anterior one in the lower
leg. The peroneal nerve passes through this muscular com-
partment, and disturbances in its function may occur early.
Hypoesthesia of the skin between the first and second
metatarsal on the dorsum of the foot is an early sign. Subse-
quently there may be a footdrop, pain on dorsiflexion of the
foot, and increasing tenderness of the compartment. If un-
treated, sustained increases in pressure are likely to result in
permanent nerve and muscle damage, and finally produce
muscle necrosis. Early recognition and proper decompression
of the compartments before permanent nerve and muscle
damage occurs are possible if serial clinical evaluations are
done, if there is surveillance of tibial venous blood flow, and if
appropriate measurement of intracompartmental pressures
are obtained. 1/5/7/20

References

1. Matsen FA, Winquist RA, Krugmire RB. Diagnosis and man-
agement of compartmental syndromes. / Bone Joint Surg 1980;
62A:286.

2. McCord JM. Oxygen-derived free radicals in post-ischemic tissue
injury. N Engl} Med 1985; 313:154.

3. Perry MO, Fantini G. Ischemia: profile of an enemy. / Vase Surg
1987; 6:231.

4. Perry MO, Shires GT III, Albert SA. Cellular changes with graded
limb ischemia in reperfusion. / Vase Surg 1984; 1:536.

5. Russell WL, Burns RP Acute upper and lower extremity compart-
ment syndromes. In: Bergen JJ, Yao JST, eds. Vascular Surgical
Emergencies. Orlando: Grune & Stratton, 1987:203.

6. Patman RD. Fasciotomy: indications and technique. In: Ruther-
ford RR, ed. Vascular Surgery. Philadelphia: WB Saunders;
1984:513.

7. Mubarak SJ, Hargens AR. Acute compartment syndromes. Surg
Clin North Am 1983; 63:539.

8. Burton AC. On the physical equilibrium of small blood vessels.
Am} Physiol 1951; 164:319.

9. Ames A, Wright RL, Kowada M et al. Cerebral ischemia: the no re-
flow phenomenon. Am J Pathol 1968; 52:437.

10. Bagge U, Amundson B, Lauritzen C. White blood cell deformabil-
ity and plugging of skeletal muscle capillaries in hemorrhagic
shock. Acta Physiol Scand 1980; 108:159.

11. Ernst E, Hammerschmidt DE, Bagge U et al. Leukocytes and the
risk of ischemic diseases. JAMA 1987; 257:2318.

12. Harman JW. The significance of local vascular phenomena in the
production of ischemic necrosis in skeletal muscle. Am } Pathol
1948; 24:625.

13. Willerson JT, Powell WJ, Guiny TE. Improvement in myocardial
function and coronary blood flow in ischemic myocardium after
mannitoL J Clin Invest 1972; 11:2981.

14. Bulkley GB. Pathophysiology of free radical-mediated reperfu-
sion injury. / Vase Surg 1987; 5:512.

15. Eklof B, Neglan P, Thompson D. Temporary incomplete ischemia
of the legs induced by aortic clamping in man. Ann Surg 1980;
93:89.

16. Roberts JP, Perry MO, Hariri RJ et al. Incomplete recovery of mus-
cle cell function following partial but not complete ischemia. Circ
S/zod: 1985; 17:253.

17. Walker PM, Lundsay TF, Lable R et al. Salvage of skeletal muscle
with free oxygen radical scavengers. / Vase Surg 1987; 5:68.

18. Whitesides TE, Haney TC, Morimoto K et al. Tissue pressure
measurements as a determinant of the need of fasciotomy. Clin
Orthop 1975; 113:43.

19. McDermott AGP, Marble AE, Yabsley RH. Monitoring acute com-
partment pressures with the S.T.I.C. catheter. Clin Orthop 1984;
190:192.

20. Jones WG, Perry MO, Bush HL. Changes in tibial venous blood
flow in the evolving compartment syndrome. Arch Surg 1989;
124:801.

244

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Physiology of reperf usion injury

Shervanthi Homer- Vanniasinkam
D. Neil Granger

Revascularization of ischemic tissue is clearly necessary for its
preservation, although it is becoming increasingly apparent
that this may be associated with a series of pathological events
that may culminate in irreversible injury to that organ and
systemic organ dysfunction. 1 ' 2 Tissues subjected to partial or
total ischemia are destined to undergo eventual cellular
dysfunction and death without timely reoxygenation. If
ischemic tissues are to be saved and their normal function
and metabolic activity preserved, there is little doubt
that blood flow and tissue oxygenation must be reestablished.
Reperfusion with oxygen-rich blood, however, results
in a multifactorial, paradoxic cascade of events (Fig. 23.1),
termed reperfusion injury, that exacerbates the damage
to already compromised tissues and may result in a
progressive loss of cell viability, tissue necrosis, and organ
dysfunction.

Many studies illustrate the concept of reperfusion injury
and have helped to elucidate some of the mechanisms in-
volved in this process. For example, small bowel ischemia of
4 h duration causes much less damage than ischemia of 3 h fol-
lowed by 1 h of reperfusion. 3 Similarly, limited myocardial is-
chemia produces minor alterations in function, whereas the
same ischemic insult followed by reperfusion induces a num-
ber of more severe physiologic changes. 4 Evidence strongly
suggests that with reperfusion, molecular oxygen is converted
to highly reactive radical species that inactivate endothelial
cell-derived nitric oxide, activate enzymes (phospholipase A 2 )
that generate lipid mediators of inflammation (e.g. platelet-
activating factor), and initiate the transcription-dependent
production of endothelial cell adhesion molecules. The net
effect of these oxidant-dependent changes is an acute and
often intense inflammatory response that can result in leuko-
cyte-mediated tissue injury (Fig. 23.1). This chapter sum-
marizes some of the evidence that supports a role for the
oxidant-initiated, inflammation-dependent mechanisms in
mediating the microvascular and parenchymal cell changes
that are associated with reperfusion of ischemic tissues. This is
followed by a discussion of the clinical manifestations of
ischemia-reperf usion (I/R) injury.

Parenchymal cell changes

It is well accepted that ischemia and reperfusion result in
different degenerative changes in the parenchyma and
microvasculature of the affected tissue. These changes include
alterations in microvascular function, intracellular and inter-
stitial edema formation, and suppression of normal metabolic
activity. 5-7 Accordingly, the phenomenon of reperfusion in-
jury has become the focus of many investigations that have
attempted to characterize the structural and biochemical
changes involved in this disease process. Unfortunately, the
pathophysiologic mechanisms involved in I/R injury are
quite diverse and differ from tissue to tissue, thereby creating
a difficult challenge for both the investigator and clinician.

Tissue susceptibility to I/R-induced injury is not uniform
throughout the body. Both the brain and intestine are exquis-
itely sensitive to the effects of ischema and reperfusion, with
ischemic insults of as little as 5-15 min in duration causing sig-
nificant alterations in normal cellular functions. 2 ' 8 Conversely,
skeletal muscle is quite resistant to ischemic injury, withstand-
ing ischemic insults of several hours with minimal irreversible
cellular damage. 7 Additionally, certain cell types or subpopu-
lations within an organ are more vulnerable to the effects of
I/R. For example, intestinal mucosal epithelial cells and cer-
tain neuronal subpopulations (hippocampal field CA1) ex-
posed to I/R suffer irreversible degenerative changes long
before like changes are observed in other cells of the same
tissue. 6/8/9

Partial and total ischemia

The effects of total as opposed to partial ischemia in relation to
reperfusion injury are not well understood. Many investiga-
tions using partial ischemia have demonstrated protective
effects provided by oxygen radical scavengers (dimethyl
sulfoxide) and substances that either inhibit the formation of
oxygen radicals (allopurinol) or enzymatically detoxify the

245

pa rt I Vascular pathology and physiology

Ischemia-Reperfusion

| Reactive Oxygen
1 Metabolites

1 Nitric Oxide

Upregulation of

Endothelial Cell

Adhesion Molecules

Platelet-endothelial
cell adhesion

Leukocyte-endothelial
cell adhesion

Proinflammatory

. Prothrombogenic .

Phenotype

Tissue Injury
Organ Dysfunction

Figure 23.1 Cascade of events that have been implicated in ischemia-
reperfusion-induced tissue injury and organ dysfunction. The mechanism
proposes that ischemia and reperfusion lead to an increased production of
reactive oxygen metabolites (ROM) and a decreased bioavailability of nitric
oxide (NO). The resulting imbalance between ROM and NO leadstoan
increased expression of endothelial cell adhesion molecules, which initiate
the recruitment of leukocytes and platelets. The proinflammatory and
prothrombogenic phenotypes that follow lead to tissue injury and organ
dysfunction.

reactive oxygen species (superoxide dismutase). 6 ' 7 When the
same compounds have been used in some models of reperfu-
sion injury involving total ischemia, however, little or no pro-
tection is sometimes observed, particularly when the tissue is
exposed to a lengthy period of reperfusion. This has raised the
question as to whether reperfusion injury exists after total is-
chemia. The issue has been addressed by a study that showed
no histologic evidence of reperfusion injury in the small intes-
tine when near-total ischemia was accompanied by venous
congestion. In the absence of venous congestion, however,
there was histologic evidence of reperfusion injury, provided
the ischemic insult was not too severe. 10

Role of reactive oxygen metabolites

The concept that reoxygenation is required to produce reper-
fusion injury is well established and has led to investigations
into the role of reactive oxygen metabolites (ROM) in this in-
jury process. The molecular oxygen introduced into ischemic
tissues on reperfusion is primarily reduced to water, but a sig-
nificant fraction of the oxygen is reduced to produce highly
reactive oxygen intermediates, namely superoxide (0 2 -), hy-
drogen peroxide (H 2 2 ), and the hydroxyl radical (*OH), all of
which have been implicated in reperfusion injury. A conse-
quence of the production of these thermodynamically unsta-
ble metabolites is reaction with (and injury to) the entire
spectrum of cellular contents, including nucleic acids, mem-
brane, lipids, enzymes, and receptors. The danger of random
"nicking" and destruction of DNA is readily apparent. De-
struction of membrane lipids leads to altered membrane fluid-
ity and allows for the leakage of various molecules (lactate
dehydrogenase, creatine phospokinase, Na + , Ca 2+ , and the
like) into and out of the cytosol. It is possible that the decreased
contractile function observed in skeletal muscle, myocardium,
and vascular smooth muscle after reperfusion may result from
ROM-mediated damage to contractile proteins. Although sev-
eral approaches have been used to assess the role of ROM in
reperfusion injury, most of the supportive evidence is based on
observations that agents that either scavenge or inhibit the
production of ROM significantly attenuate reperfusion in-
jury 2,11 The importance of ROM in reperfusion injury has also
been demonstrated using mutant mice that are genetically en-
gineered to overexpress antioxidant enzymes such as super-
oxide dismutase, catalase, or glutatione peroxidase. 12

Xanthine oxidase

Xanthine oxidase (XO), in addition to its function as the rate-
limiting enzyme in nucleic acid degradation, also is able to
generate H 2 2 and 2 - during the oxidation of hypoxanthine
or xanthine. Because this pathway is a significant source of
ROM in certain tissues (e.g. intestine and liver), it is predicted
that tissues with a high activity for this enzyme are more likely
to be susceptible to I/R injury 13 Indeed, XO-specific antibod-
ies applied to frozen sections of liver, heart, skeletal muscle,
lung, intestine, and kidney reveal that capillary endothelial
cells contain about 100 times more XO activity than other cells
in the same tissue. 13 Epithelial cells lining the tips of villi in the
small intestine also are a rich source of XO. 14 Since vascular en-
dothelial cells are exposed to high concentrations of molecular
oxygen upon reperfusion of ischemic tissues, the ability of the
vessel wall to generate ROM is significant. Endothelial cell-de-
rived ROM appear to promote the adhesion, activation, and
emigration of leukocytes by eliciting the production of lipid
mediators of inflammation (e.g. platelet-activating factor),

246

CHAPTER 23 Physiology of reperfusion injury

and initiating the transcription-dependent production of
endothelial cell adhesion molecules. 11

Leukocyte-endothelial cell adhesion in
reperfusion injury

It has long been known that neutrophils adhere to vascular en-
dothelium and that this adhesive interaction is responsible for
the physiologic trafficking of leukocytes between intravascu-
lar and extravascular compartments. 15 Adherence and
emigration of neutrophils are critical for their immunologic
function in the interstitium, as is demonstrated clearly in
patients afflicted with leukocyte adhesion deficiency
syndromes. The hypothesis that neutrophils may be a key
component of I/R injury has been proposed. It has been sug-
gested that a major source of ROM is the neutrophil itself since
it is ideally equipped to produce large amounts of these highly
reactive molecules, and reperfusion generally results in a sig-
nificant influx of neutrophils in affected tissues. 11 In addition,
neutrophils secrete myeloperoxidase, which catalyzes the
production of hypochlorous acid from H 2 2 and chloride ions.
Hypochlorous acid is approximately 100 times more reactive
than H 2 2 .

Two experimental approaches have been used to determine
whether reperfusion-induced neutrophil accumulation is a
cause or a consequence of this I/R injury: (i) neutrophil deple-
tion with antineutrophil serum; and (ii) prevention of
neutrophil adherence with monoclonal antibodies directed
against leukocyte or endothelial cell adhesion molecules.
Using antineutrophil serum, it has been demonstrated that re-
duction of circulating neutrophils to less than 5% of control
significantly attenuates the increased microvascular perme-
ability seen after I/R of the small bowel. 16 Likewise, infarct
size is reduced in postischemic canine myocardium in animals
rendered neutropenic with antineutrophil serum. 17

If adhesion is required for neutrophil-mediated vascular in-
jury, then it can be hypothesized that inhibition of leukocyte
adherence with monoclonal antibodies would attenuate
reperfusion injury in a manner similar to neutropenia. Intra-
vital microscopic techniques have been used to demonstrate
enhanced leukocyte adhesion and emigration in postcapillary
venules exposed to I/R. 18/19 When monoclonal antibodies
against the neutrophil adhesion glycoprotein CD18 are ad-
ministered before the induction of ischemia, both the in-
creased leukocyte adhesion and microvascular permeability
normally observed after reperfusion are largely prevented. 11
Similarly, postischemic canine myocardium experiences a re-
duction in infarct size in the presence of adhesion molecular
directed antibodies. 17

In addition to the directly toxic effects of adherent neu-
trophils, it has been hypothesized that capillary plugging by
neutrophils decreases blood flow, thereby exacerbating the
hypoxic insult. Indeed, neutropenic rats exhibit higher blood

flow in the gastrointestinal tract than controls subjected to the
same degree of hypotension shock. 20 Furthermore, skeletal
muscle exhibits a no-reflow phenomenon in which some cap-
illaries fail to perfuse when flow is reestablished. Several
potential mechanisms have been proposed to explain this
phenomenon: arteriolar spasm, microemboli, interstitial
edema, endothelial bleb formation, and platelet thrombi. Evi-
dence suggests, however, that this no-reflow phenomenon is
the result of leukocyte plugging. It has been shown that the in-
crease in vascular resistance on reperfusion of ischemic skele-
tal muscle is obliterated in animals that are neutropenic. 21
Inhibition of leukocyte adhesion by monoclonal antibodies
produces similar beneficial effects in reperfused skeletal
muscle. 22

Platelet-endothelial cell adhesion

While the results of numerous studies suggest that the recruit-
ment of activated and adherent leukocytes (primarily neu-
trophils) is a rate-determining step in the development of
microvascular dysfunction and tissue injury following is-
chemia and reperfusion, there is a growing body of evidence
that also implicates other circulating blood cells, including
platelets, as potential modulators of I/R-induced microvascu-
lar alterations and tissue injury. A role for platelets in the
pathogenesis of I/R injury is supported by reports describing
a beneficial effect of platelet depletion. 23 Further support is
provided by recent studies which demonstrate that intestinal
I/R is associated with the recruitment of rolling and adherent
platelets in postcapillary venules and that the density (cells
per unit vessel area) of recruited platelets can exceed the
density of adherent leukocytes by an order of magnitude. 24
Several adhesion molecules (P-selectin and GPIIb/IIIa) and
procoagulant factors (e.g. fibrinogen) have been implicated in
the platelet-endothelial cell adhesion that is elicited by I/R.
Furthermore, the possibility exists that I/R-induced recruit-
ment of leukocytes may be influenced by the initial adhesion
of platelets to venular endothelium. Three lines of evidence
support the possibility that I/R-induced leukocyte recruit-
ment is dependent on the expression of P-selectin by platelets
that are adherent to venular endothelium: (i) P-selectin ex-
pressed on the surface of adherent, activated platelets can sus-
tain leukocyte rolling and adherence in vitro; (ii) P-selectin is a
major determinant of I/R-induced leukocyte recruitment in
postcapillary venules; and (iii) thrombocytopenic animals ex-
hibit a profound attenuation of both P-selectin expression and
neutrophil accumulation after intestinal I/R. 23/24

Nitric oxide

Nitric oxide (NO), an endothelium-derived smooth muscle
relaxant, is synthesized by nitric oxide synthase (NOS), a

247

pa rt I Vascular pathology and physiology

calcium-dependent enzyme, from the amino acid L-arginine.
NO plays an important role in modulating vascular tone and
consequently blood flow in various tissues. 25 Although this
substance has been implicated as a modulator of several cellu-
lar and physiological processes, it is also believed to be in-
volved in a number of acute and chronic pathological states,
including I/R, sepsis, inflammatory bowel diseases, and
arthritis. The prevailing opinion is that NO confers a protec-
tive action on the vasculature in acute inflammatory condi-
tions, but NO can exert a toxic effect (via its conversion
to peroxynitrite) on tissues during chronic inflammatory
states. While studies of I/R injury have suggested that NO
may be either protective or injurious, the overwhelming
majority of reports support a protective effect of NO in
reperfusion injury.

Inhibition of endogenous NO production by addition of
NG-nitro-L-arginine methyl ester (l-NAME), an L-arginine
analogue, dramatically enhances the injurious effects of I/R in
the feline small intestine. 26 Augmentation of endogenous NO
production by addition of the NO precursor L-arginine signif-
icantly attenuates reperfusion injury in the same model. 26
Endogenous NO production by the feline mesenteric mi-
crovasculature has been shown to prevent leukocyte adhesion
to endothelial cells, whereas inhibition of NO synthesis with
l-NAME elicits increased leukocyte adhesion and enhanced
microvascular protein leakage. 27 Because leukocytes are a
well-established contributor to the tissue damage associated
with reperfusion injury, it is likely that the protective effects of
NO in intestinal reperfusion injury are due in part to preven-
tion of leukocyte-endothelial cell adhesion.

The superoxide radical (0 2 -) is known to react avidly with
NO to form a variety of highly reactive and potentially damag-
ing radical species, including peroxynitrite. In the brain, so-
called delayed neuronal death, which may occur for 24-72 h
after reperfusion is initiated, has been linked to NO and the ex-
citatory neurotransmitter, glutamate. 28 With neuronal depo-
larization, glutamate is released into the synapse in small
amounts. Under normal conditions, the free glutamate is ac-
tively and rapidly reabsorbed; however, under ischemic con-
ditions this reabsorption does not occur, leading to a dramatic
increase in tissue glutamate levels after brief episodes of is-
chemia. 8 Free glutamate binds to N-methyl-D-aspartate
(NMDA) receptors on postsynaptic neurons and subsequent-
ly opens receptor-associated calcium channels, thus allowing
an influx of calcium into the cytosol. 5 ' 25 The calcium influx
probably activates the calcium-dependent enzyme NOS, re-
sulting in NO production, which in turn reacts with super-
oxide radicals produced during reperfusion to form the
peroxynitrite anion (ONOO-). 25 Impressive neuroprotective
effects have been seen with administration of either NMDA re-
ceptor antagonists (MK-801) or a NOS inhibitor (l-NAME) in
cerebral I/R models. 5,8 ' 25 The damaging effects of peroxyni-
trite may be related to: (i) reaction with transition metals to
form a potent nitrating substance; (ii) initiation of lipid peroxi-

dation and direct interaction with sulfhydryl groups; or (iii)
hydroxyl radical and nitrogen dioxide formation from hydro-
gen ion catalyzed hemolytic cleavage of peroxynitrite. 25 The
injury from any of these pathways may lead to a cascade of
events that exacerbate reperfusion injury.

Clinical manifestations of I/R injury

The clinical sequelae of I/R injury are seen in a number of or-
gans and tissues. These may be manifested as (i) local and/or
(ii) systemic effects. In vascular surgical practice, reperfusion
of an acutely or critically ischemic limb (i.e. limb revascular-
ization by surgery or radiological intervention) may be associ-
ated with the development of skeletal muscle reperfusion
injury characterized by muscle edema, impaired contractile
function and, in extreme cases, muscle necrosis necessitating a
major limb amputation. 29 In addition to this local tissue dam-
age, limb revascularization may result in deleterious systemic
effects. In 1960, Haimovici 30 described a complex of cardiac,
renal, and pulmonary complications, the "metabolic syn-
drome/' which may account for up to 25% of deaths in patients
presenting with acute limb ischemia.

Aortic surgery has been studied in some detail with respect
to the I/R-induced inflammatory response and the ensuing
organ failure. 31 Patients are especially at risk of developing
I/R-induced multiorgan failure following repair of thoracoab-
dominal aortic aneurysms (TAAA) due to visceral I/R. 32 In
this study of 28 patients undergoing TAAA repair, 36% devel-
oped multiorgan failure, 43% developed pulmonary dysfunc-
tion, and 36% developed renal failure. These patients had
elevated levels of cytokines [tumor necrosis factor (TNF)-oc,in-
terleukin (IL)-6, IL-8, and IL-10] and shed TNF receptors p55
and p75. TAAA repair patients thus provide an ideal "human
model" of I/R-induced, and cytokine-mediated, multiorgan
failure. 32 While less common when compared with TAAA re-
pair, I/R-related organ dysfunction is also seen following
infrarenal AAA repair. 33 Ruptured AAA patients present a
model of whole-body I/R and Lindsay et al. have proposed a
"two-hit" I/R scenario wherein the initial I/R of hemorrhagic
shock and resuscitation is followed by the lower torso I/R
during surgical repair of the ruptured aneurysm. 34 When
compared with elective AAA repair, these patients had signifi-
cantly elevated levels of lipid peroxidation products and neu-
trophil oxidant production.

Although a number of preclinical studies have addressed
I/R injury in the gut, clinical manifestations of this are gener-
ally seen following surgery for bowel obstruction, resuscita-
tion from hemorrhagic shock, in necrotizing enterocolitis, and
during surgery for acute and chronic mesenteric ischemia.
Harward et al. 35 reported significant morbidity (hepatic, renal
and pulmonary organ dysfunction, and coagulopathy) and
mortality in patients undergoing revascularization for symp-
tomatic chronic mesenteric arterial occlusive disease. These

248

CHAPTER 23 Physiology of reperfusion injury

deleterious effects are thought to be due to inflammatory me-
diator release from the reperfused gut. I/R-induced organ and
tissue damage is also seen in patients following coronary
revascularization and solid organ transplantation.

Conclusions

Reperfusion injury is a complex disease process that involves a
paradoxic and multifactorial cascade of events. The injury in-
curred is above and beyond that caused by ischemia alone and
leads to alterations in the function of both parenchymal cells
and the microvasculature. The role of ROM in reperfusion in-
jury and its two most prevalent sources, XO and neutrophils, is
well established. There also is substantial evidence that altered
NO bioavailability and interactions between leukocytes and
vascular endothelium play an important role in modulating
reperfusion injury. The reperfusion-induced alterations in
ROM and NO appear to activate numerous proteins (e.g.
enzymes and adhesion molecules) through transcription-
dependent and -independent pathways and consequently
amplify the inflammatory response. Clinical studies generally
support the existence of I/R-induced tissue injury in a variety
of organ systems. An improved understanding of the basic
pathophysiologic mechanisms that underlie reperfusion in-
jury may lead to therapeutic interventions that improve tissue
viability and patient survival.

Acknowledgment

Supported by grants (HL26441 and DK43785) from the
National Institutes of Health, Bethesda, Maryland.

References

1. Grace PA. Ischaemia-reperfusion injury. Br J Surg 1994; 81:637.

2. Carden DL, Granger DN. Pathophysiology of ischaemia-
reperfusion injury. J Pathol 2000; 190:255.

3. Parks DA, Granger DN. Contributions of ischemia and reperfu-
sion to mucosal lesion formation. Am } Physiol 1986; 250:G749.

4. Hearse DJ. Reperfusion of the ischemic myocardium. / Mol Cell
Cardioll977;9:605.

5. Nowicki JP, Duval D, Poignet H et ah Nitric oxide mediates
neuronal death after focal cerebral ischemia in the mouse. Eur }
Pharmacol 1991; 204:339.

6. Granger DN, Hollwarth ME, Parks DA. Ischemia-reperfusion
injury: role of oxygen-derived free radicals. Acta Physiol Scand
(Suppl) 1986; 548:47.

7. Lindsay T, Romaschin A, Walker PM. Free radical mediated
damage in skeletal muscle. Microcirc Endothelium Lymphatics 1989;
5:157.

8. Choi DW. The role of glutamate neurotoxicity in hypoxic-ischemic
neuronal death. Annu Rev N euros ci 1990; 13:171.

9. Carati CJ, Rarnbaldo S, Gannon BJ. Changes in macromolecular

permeability of microvessels in rat small intestine after total
occlusion ischemia /reperfusion. Microcirc Endothelium Lymphat-
ics 1988; 4:69

10. Park PO, Haglund U, Bulkley GB et ah The sequence of develop-
ment of intestinal tissue injury after strangulation ischemia and
reperfusion. Surgery 1990; 107:574.

11. Granger, DN. Role of xanthine oxidase and granulocyte in is-
chemia-reperfusion injury. Am] Physiol 1988; 255:H1269.

12. Lefer DJ, Granger DN. Oxidative stress and cardiac disease. Am }
Med 2000; 109:315.

13. Jarasch ED, Bruder G, Heid HW. Significance of xanthine oxidase
in capillary endothelial cells. Acta Physiol Scand 1986; 548:39.

14. Auscher C, Amory N, Emp P et ah Xanthine oxidase activity in
human intestines: histochemical and radiochemical study. Adv
Exp Med Biol 1979; 122:197.

15. Harlan JM.Leukocyte-endothelial interactions. Blood 1985; 65:513.

16. Hernandez LA, Grisham MB, Twohig B, Arfors KE, Harlan JM,
Granger DN. Role of neutrophils in ischemia-reperfusion-in-
duced microvascular injury. Am } Physiol 1987; 253:H699.

17. Romson JL, Hook BG, Kunkel SL, Abrams GD, Schork MA, Lucch-
esi BR. Reduction of the extent of ischemic myocardial injury by
neutrophil depletion in the dog. Circulation 1983; 67:1016.

18. Granger DN, Benoit JN, Suzuki M et ah Leukocyte adherence
to venular endothelium during ischemia-reperfusion. Am }
Physiol 1989; 257:G683.

19. Oliver MG, Specian RD, Perry MA, Granger DN. Morphologic as-
sessment of leukocyte-endothelial cell interactions in mesenteric
venules subjected to ischemia and reperfusion. Inflammation 1991;
1:331.

20. Smith SM, Grisham MB, Manci EA et ah Gastric mucosal injury in
the rat: role of iron and xanthine oxidase. Gastroenterology 1987;
92:950.

21. Korthius RJ, Grisham MB, Granger DN. Leukocyte depletion
attenuates vascular injury in postischemic skeletal muscle. Am }
Physiol 1988; 254:H823.

22. Carden DL, Smith JK, Korthius RJ. Neutrophil mediated mi-
crovascular injury. Am EmergMed 1989; 18:476.

23. Salter JW, Krieglstein CF, Issekutz AC, Granger DN. Platelets
modulate ischemia /reperfusion-induced leukocyte recruitment
in the mesenteric circulation. Am J Physiol Gastrointest Liver Physiol
2001;28:G1432.

24. Massberg S, Enders G, Matos FC et ah Fibrinogen deposition at
the postischemic vessel wall promotes platelet adhesion during
ischemia-reperfusion in vivo. Blood 1999; 94:3829.

25. Beckman JS. The double-edged role of nitric oxide in brain
function and superoxide mediated injury. / Dev Physiol 1991;
15:53.

26. Kubes P. Ischemia /reperfusion in the feline small intestine: a role
for nitric oxide. Am J Physiol 1993; 264:G143.

27. Kubes P, Granger DN. Nitric oxide modulates microvascular per-
meability. Am Physiol 1992; 262:H611 .

28. Hallenbeck JM, Dutka AJ. Background review and current con-
cepts of reperfusion injury. Arch Neurol 1990; 47:1245.

29. Crinnion JN, Homer- Vanniasinkam S, Gough MJ. Skeletal muscle
reperfusion injury: pathophysiology and clinical considerations.
Cardiovasc Surg 1993; 1: 317.

30. Haimovici H. Arterial embolism with acute massive ischemic
myopathy and myoglobinuria. Surgery 1960; 47:739.

249

pa rt I Vascular pathology and physiology

31. Groeneveld AB, Raijmakers PG, Rauwerda JA, Hack CE. The in-
flammatory response to vascular surgery-associated ischaemia
and reperfusion in man: effect on postoperative pulmonary func-
tion. Eur J Vase Endovasc Surg 1997; 14:351.

32. Welborn MB, Oldenburg HS, Hess PJ et al. The relationship be-
tween visceral ischemia, proinflammatory cytokines, and organ
injury in patients undergoing thoracoabdominal aortic aneurysm
repair. CritCare Med 2000; 28: 3191.

33. Huber TS, Harward TRS, Flynn TC, Albright JL, Seeger JM.
Operative mortality rates after elective infrarenal aortic
reconstructions. / Vase Surg 1995; 22:287.

34. Lindsay TF, Luo XP, Lehotay DC et al. Ruptured abdominal aortic
aneurysm, a "two-hit" ischemia /reperfusion injury: evidence
from an analysis of oxidative products. / Vase Surg 1999; 30:219.

35. Harward TR, Brooks DL, Flynn TC, Seeger JM. Multiple organ
dysfunction after mesenteric revascularization. / Vase Surg 1993;
18:459.

250

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Cerebral ischemia

Hao Bui

Christian de Virgilio

Anatomy

The cerebral circulation is composed of the anterior circula-
tion, which includes the carotid arteries, anterior and middle
cerebral arteries and their branches, and the posterior circula-
tion, which includes the vertebral, basilar, and posterior cere-
bral arteries. The circle of Willis consists of the proximal
portion of the anterior and posterior cerebral arteries as well
as the anterior and posterior communicating arteries. The
anterior communicating arteries provide collateral circu-
lation between the two hemispheres, whereas the posterior
communicating arteries form collaterals between the anterior
and posterior circulations. Thus, the circle of Willis connects
the two carotid arteries with each other and with the basilar
artery. A "normal" circle of Willis is present in about half of
cases 1 (Fig. 24.1).

Cerebral metabolism

The human brain relies on a constant supply of oxygen and
glucose to maintain homeostasis. At rest, this requires
50-55 ml of blood per 100 g cerebral tissue per minute to be
delivered through the cerebral circulation. 2 Energy in the
form of adenosine triphosphate (ATP) is necessary to maintain
neuronal integrity. The source for the vast majority of the
ATP is oxidative metabolism of glucose. Lactate is also
consumed in very small quantities by the brain under normal
circumstances. There is little storage capacity for energy
substrate in the brain, as demonstrated by the fall in ATP
levels to zero within 7 min after termination of the oxygen
supply. 3 About 40% of the energy is used for basal needs,
whereas functional activity consumes about 60%. When
the oxygen supply is decreased, energy production con-
verts to anaerobic glycolysis. This conversion eventually
leads to acidosis and failure of the Na + /K + pump. Because
of the intracellular release of K + , a large rise in extracellular
K + occurs.

Cerebral autoregulation

The brain is able to maintain a constant blood flow indepen-
dent of moderate changes in mean arterial perfusion pressure,
cardiac output, and body activity. This mechanism of control
is termed cerebral autoregulation. Autoregulation is not an
all-or-none phenomenon, but rather represents a continuous
spectrum of adaptive response in cerebrovascular resistance
to a change in perfusion pressure. Without autoregulation,
systemic hypertension may lead to cerebral hemorrhage and
edema formation. Conversely, a decrease in systemic blood
pressure may cause ischemia and infarction. 3

Normal cerebral blood flow (CBF) is approximately
50 ml/ 100 g/min. This represents the average blood flow for
the whole brain; blood flow to the gray matter is higher at
80 ml/ 100 g/min, whereas flow to the white matter averages
20 ml/ 100 g/min. The average brain receives about 14% of
the cardiac output. Under normal physiologic conditions,
changes in mean arterial pressure (MAP) between 60 and
160 mmHg in the average individual produces little or no
change in CBF. Cerebral autoregulation thus protects the brain
from fluctuations in MAP. When the MAP falls outside these
limits (60-160 mmHg), autoregulation fails and CBF becomes
directly proportional to the MAP. In these circumstances, CBF
becomes pressure dependent. Several conditions such as
trauma, hypoxemia, hypercapnia, and high-dose volatile
anesthetics can impair or abolish autoregulation.

The exact mechanism by which the brain regulates blood
flow is not known. There is some evidence that the autoregula-
tion control may be a combination of metabolic, myogenic,
and neurogenic mechanisms. 2 Current evidence suggests that
local metabolic factors are of primary importance in the local
tissue regulation of CBF. Under normal conditions, regional
cortical blood flow is reflective of localized brain activity.
There are several major metabolic factors that play a role in
autoregulation of CBF under normal conditions, including
carbon dioxide, potassium, adenosine, prostaglandins, and
nitric oxide (NO). The myogenic factor that may affect CBF is

251

pa rt I Vascular pathology and physiology

Anterior communicating artery
Anterior cerebral artery -

Middle cerebral artery

Posterior communicating artery

Posterior cerebral artery

Basilar artery

Vertebral arteries

Figure 24.1 Classically described circle of Willis. (From Baker WH,
Cerebrovascular occlusive disease. In: Greenfield U, Mulholland MW,
Oldham KT, Zelnock GZ, eds. Surgery: Scientific Principles and Practice.
Philadelphia: JB Lippincott, 1 993:1 600.)

the basal tone of the vascular smooth muscle. This tone may be
affected by changes in perfusion or transmural pressure,
whereby the smooth muscle contracts with increased MAP
and relaxes with decreased MAP. Studies suggest that there
may be two myogenic mechanisms involved in cerebral auto-
regulation: a rapid fast reaction to pressure pulsations, and a
slower reaction to changes in MAP.

Perivascular innervations of cerebral resistance vessels and
the specific neurotransmitters contained within the perivascu-
lar nerve fibers may modulate vascular response to changes in
blood pressure. The specific mechanisms by which the central
nervous system exerts control on the cerebral vasculature are
poorly understood. The current notion is that sympathetic and
trigeminal neuronal activity can modify the pressure-flow
relationship, but the role of these systems is minimal under
normal blood pressure conditions.

The cerebral circulation is exquisitely sensitive to changes in
carbon dioxide tension (P a co 2 ), which is the most potent phys-
iologic cerebral vasodilator. In normal subjects, CBF changes
linearly by 2-4% for every 1 mmHg change in P a co 2 (within the
range of 25-75 mmHg). 4 As an example, inhalation of a mix-
ture of 5% carbon dioxide increases CBF by 50%. Conversely,
CBF falls by approximately 35% if the P a co 2 decreases from 45
to 26 mmHg. 1 Changes occur within seconds. Complete equi-
libration occurs within 2 min. Carbon dioxide rapidly diffuses
across the blood-brain barrier and into the perivascular fluid
and cerebral vascular smooth muscle cell. Carbon dioxide de-
creases perivascular pH. The pathway by which perivascular
pH influences cerebral vascular tone has not been clearly
defined, though some studies suggest a role for NO and
prostaglandin E 2 . Certain conditions, such as severe carotid

stenosis, head injury, cardiac failure, and severe hypotension
can attenuate the C0 2 cerebral response.

Potassium also acts as a potent vasodilator, but its mechanism
of action is not well understood. It is known that during periods
of hypoxia, electrical stimulation, and seizures, increases in
perivascular K + coincide with increases in CBF. Furthermore,
when K is applied topically, vessel dilation occurs. Evidence
suggests that K + acts as an early signal in vasodilation.

Adenosine can be found in increased concentration in cerebral
tissue as systemic blood pressure falls toward the lower limit of
autoregulation. Brain adenosine concentration doubles within
5 s of decreasing blood pressure. Unlike pH and K + , adenosine
levels remain elevated through the entire period of hypoxia and
act for longer periods to mediate vascular dilation. 5

Prostaglandins are another group of important endogenous
compounds that are increased in the extracellular fluid during
hypotension. Prostaglandin E and prostacyclin are two
prostaglandins that possess vasodilator properties. In animal
models, cerebrospinal fluid levels of these compounds have
been demonstrated to be increased after arterial hypotension.
Use of prostaglandin inhibitors such as indomethacin can
block the ability of the brain to maintain constant cerebral
perfusion during arterial hypotension. These findings sup-
port the role of prostaglandins as modulators of CBF, but the
mechanisms need to be further clarified. Recent evidence sug-
gests that prostaglandins may mediate at least some of their
function through interactions with NO. 6

NO is an intercellular messenger in the peripheral circula-
tion and in the central nervous system, and causes vascular
smooth muscle relaxation and inhibition of platelet aggrega-
tion. NO has a very short half-life of approximately 6 s and is
synthesized from L-arginine by a group of enzymes known as
NO synthases. In animal models of global ischemia, NO levels
were found to be increased 4-6 h after insult.

The effect of P a o 2 on the cerebral circulation is mild and of
much less clinical significance. In one study, reduction of the
Po 2 from 89 to 35 mmHg resulted in an increase in the CBF
from 45 to 77 ml/ 100 g/min. 7 Thus, the brain is able to main-
tain constant oxygen consumption despite a drop in Po 2 to
40 mmHg by increasing the CBF. The exact mechanism by
which changes in Po 2 alter CBF is unclear, but it may be medi-
ated by alterations in tissue pH caused by an increase in anaer-
obic metabolism, 1 or it may be a direct effect of oxygen on the
cerebral vessel smooth muscle. 2

Complete cessation of cerebral circulation results in irre-
versible cell damage within minutes. The mechanism involves
depletion of high-energy phosphates; membrane ion pump
failure; efflux of cellular potassium; influx of sodium, chloride,
and water; and membrane depolaration. 8 More commonly,
however, the cerebral circulation is only partially interrupted,
with focal ischemia resulting from occlusion of a cerebral ves-
sel. In this setting, the degree of cell damage depends on the
duration of ischemia, the efficiency of the collateral circula-
tion, and local perfusion pressure.

252

chapter 24 Cerebral ischemia

Thresholds in cerebral ischemia

Considerable knowledge about the pathophysiology of focal
cerebral ischemia has been gained through animal experi-
ments involving middle cerebral artery occlusion in baboons.
These studies have led to the concept of thresholds of cerebral
blood flow for reversible dysfunction and irreversible infarc-
tion. Branston and colleagues 9 demonstrated that evoked so-
matosensory potential in baboon cortex was abolished when
CBF fell below 15 ml/ 100 g/min. This level was termed the
threshold for electrical failure in the cerebral cortex. This threshold
correlates in humans with electroencephalographic (EEG)
flattening when CBF falls below 16 ml/ 100 g/min. 10

In baboon studies, Na + /K + pump failure did not occur at
this level of CBF reduction, as evidenced by a normal extra-
cellular K + level. 11 A further reduction in CBF to below
10 ml/ 100 g/min was necessary before massive K + release
from the cell into the extracellular fluid occurred. This level of
CBF reduction has been termed the flow threshold for energy and
ion pump failure, and is considered irreversible.

The term ischemic penumbra was coined to describe the area
of ischemic brain in which CBF is between these two thresh-
olds—electrical failure without pump failure. 8 In this penum-
bral range, synaptic activity is abolished, but the structural
integrity is preserved. Cerebral ischemia within this penum-
bra may be reversible if CBF is increased. Morawetz et al. 12
found that recovery without histological signs of structural in-
farction, following a 2- to 3-h period of focal ischemia in the
monkey, could only be found at sites where local blood flow
was sustained above 12 ml/ 100 g/min. An example of the
penumbra phenomenon with complete reversibility can be
seen in patients with embolic transient ischemic attacks
(TIAs). Another example is in the acute stroke setting, where
areas of the brain adjacent to the infarction may be in the
penumbra, and thus potentially reversible. Avoidance of sys-
temic hypotension in this setting may prevent a further drop of
the CBF to below the irreversible level of ischemia. Converse-
ly, elevation of blood pressure can improve perfusion in the
ischemic zone. In addition, patients undergoing carotid end-
arterectomy after recovery from a stroke are usually shunted
since an area of ischemic but viable brain tissue exists around
the infarcted area that is at risk if CBF drops further to a critical
level.

Occlusion of the middle cerebral artery does not lead to im-
mediate infarction, because CBF usually does not drop to zero.
Whether infarction occurs depends on the severity and dura-
tion of the ischemia as well as the effectiveness of the collateral
circulation, as measured by the residual flow. Collaterals in-
clude extracranial to intracranial circuits, the circle of Willis,
and the leptomeningeal end-to-end collaterals. In an animal
study, 13 middle cerebral artery occlusion with residual flows
below 12 ml/ 100 g/min leads to infarction after 2-3 h, whereas
flows of 6-8 ml/ 100 g/min lead to infarction after 1 h. In

humans, marked EEG changes during carotid clamping can
be tolerated for 30 min with a CBF reduction to 12-15 ml/
100 g/min. 14 Clinical application of this threshold is used dur-
ing carotid endarterectomy, as some surgeons selectively insert
a carotid shunt based on the development of EEG changes. 15

These studies have prompted interest in attempts at early
reperfusion of impaired but not infarcted areas of ischemic
brain (in the penumbra) in the hopes of restoring neurologic
function. Experimental work has shown evidence of improve-
ment in blood flow and tissue function with early reperfusion. 16
One major concern is the effect of reperfusion on edema forma-
tion. Bell and coworkers 16 showed that cerebral blood flow re-
duction to 19 ml/ 1000 g/min for more than 30 min results in
edema formation in baboon cortex. Once edema has formed,
reperfusion to the edematous brain resulted in exacerbation of
the edema. Furthermore, as demonstrated by Tamura and col-
leagues, 17 reperfusion may result in extreme hyperemia result-
ing from dysautoregulation of cerebral vessels with resultant
petechial hemorrhages and vasculogenic edema. These studies
may explain why some patients subjected to urgent carotid
endarterectomy in the setting of acute stroke have deteriorated
clinically. Likewise, these studies have sparked an interest in
early reperfusion of cerebral infarction using thrombolytic
agents. However, these clinical studies have shown that a delay
in reperfusion beyond 3 h after onset of ischemic stroke dra-
matically increases the risk of intracranial hemorrhage.

Etiology of cerebral ischemia

The etiology of extracranial cerebral ischemia can be loosely
categorized into flow-restrictive and embolic lesions. Some
causes of cerebral ischemia are embolization from cardiac
sources, fibromuscular dysplasia, arteritides, aneurysm,
radiation damage, trauma, hematologic disorders, hyperten-
sive hemorrhage, and severe systemic hypotension.

A proposed classification system of ischemic infarction was
modified by Caplan and coworkers 18 into four groups based
on the site of arterial involvement.

Group 1: Patients have atherosclerosis of the extracranial
large arteries, most commonly the origins of the internal
carotids and vertebral arteries. They often have other periph-
eral sites of atherosclerosis.

Group 2: Patients have lipohyalinosis of smaller arteries re-
sulting in lacunar infarcts in the territories of the basal ganglia,
pons, internal capsule, and thalamus. They have a strong his-
tory of hypertension.

Group 3: Patients have intraarterial thrombi, primarily in
the middle, anterior, and posterior cerebral arteries. In most
instances, the sources of the occlusion are emboli from a
more proximal source such as the heart.

Group 4: Patients are the least recognized and have primary
atherosclerosis of the intracranial vessels, with a low incidence
of concomitant peripheral vascular disease.

253

pa rt I Vascular pathology and physiology

Carotid bifurcation

The carotid bifurcation is particularly predisposed to athero-
matous plaque formation. Zarins 19 conducted plaque loca-
lization studies in the carotid bifurcation. These studies
demonstrated that plaque formed preferentially in areas of
low shear stress, low velocity, flow separation, and stasis,
which corresponded to the outer wall of the carotid sinus.
Conversely, along the inner wall, flow was laminar and rapid,
with a high shear stress, and plaques did not form. Low shear
stress is thought to prolong the contact of atheromatous
substances with the arterial wall, and thus promote plaque
deposition.

Carotid siphon

After the carotid bifurcation and the sinus portion, the siphon
region is the most common site of atherosclerotic plaque for-
mation in the carotid artery 20 The siphon is the segment of in-
ternal carotid artery between the exit from the petrous bone
and its division into the anterior and middle cerebral arteries.
The predilection for atheroma in this region may be a result of
disturbances in laminar flow resulting from the curvilinear
configuration of the siphon. 20

The carotid siphon plaque differs from carotid bifurcation
plaque in its propensity toward early and marked calcifica-
tion. 21 The calcification preferentially involves the media
and the elastic lamina, which seems to render stability to the
plaque. Deep or irregular ulcerations are uncommon, and sig-
nificant stenoses are less common than at the bifurcation. 22
This explains why patients with TIA and tandem lesions in the
carotid bifurcation and siphon still benefit from standard
endarterectomy in most cases, since the carotid bifurcation is
more likely to be the source of atheromatous debris or platelet
aggregate emboli. 23

Vertebrobasilar system

The most common site of atherosclerosis in the vertebral artery
is the origin from the subclavian artery. 18 Castaigne and asso-
ciates 24 studied 44 patients with arterial occlusion in the verte-
brobasilar system and found atherosclerosis as the cause in
79% and cardiac embolus in 9%. The occlusions were most
often the result of thrombosis in an area of tight atherosclerotic
stenosis, and usually involved the vertebral and basilar arter-
ies, with only two thrombotic occlusions from atherosclerosis
in the posterior cerebral arteries.

Emboli from the vertebrobasilar system likewise most often
originate from atherosclerotic lesions in the vertebral and basi-
lar arteries, and frequently deposit in the posterior cerebral
artery. In the series by Castaigne and coworkers, 24 evidence of

infarction was present in only half of patients with vertebral
artery occlusion. The authors suggested that occlusion of one
vertebral artery can lead to infarction in the basilar artery terri-
tory if the contralateral vertebral artery is tightly stenosed or
atretic, even with a normal circle of Willis.

A distinct clinical entity of flow reversal in the vertebral
artery in the presence of a proximal subclavian /innominate
stenosis or occlusion is known as subclavian steal syndrome.
Blood is siphoned away from the basilar artery in a retrograde
fashion down the vertebral artery. This reversal of flow may
produce symptoms of vertebrobasilar insufficiency; depend-
ing on the adequacy of collateral circulation, flow in the con-
tralateral vertebral artery is increased, as is flow in both carotid
arteries. Exercise in the ipsilateral arm increases the amount of
vertebral flow reversal. In a review of 168 patients with subcla-
vian steal syndrome, Fields and Lemak 25 noted that the left
subclavian was involved in 70% of cases. The most common
neurologic symptoms were vertigo, limb paresis, and pares-
thesias. Intermittent arm claudication was less common. Radi-
ologic demonstration of flow reversal in the vertebral artery
was not necessarily accompanied by symptoms.

Cerebral emboli

Cerebral emboli most commonly arise from atheromatous
plaques at the carotid bifurcation, followed by cardiac sources.
Fibrous plaques can undergo degeneration with calcification
and ulceration. Several researchers have pointed out the sig-
nificance of ulcerative lesions in the carotid bifurcation as
sources of cerebral emboli. 26-28 Turbulence at the site of an ir-
regular plaque leads to platelet aggregation. These aggregates
may break off and lodge in the cerebral circulation. The seve-
rity and duration of subsequent neurologic symptoms depend
on where the embolus lodges and whether the embolus dissi-
pates quickly, causing a TIA, or persists, causing an infarction.
Emboli from carotid atheromatous plaques may also be the
result of intraplaque hemorrhage. With time, some atheroscle-
rotic plaques themselves become vascularized with many
thin-walled vessels. If these vessels tear, hemorrhage can
occur within the plaque, which may lead to acute expansion
with occlusion of the vessel. More commonly, the plaque may
rupture, releasing its contents (cholesterol crystals, calcific and
thrombotic material) into the cerebral circulation. If these em-
boli lodge in the central retinal artery or its branches, they may
be visualized on fundoscopic examination as a shower of
bright orange crystals know as Hollenhorst plaques. Acute or
recent hemorrhage was present in 92% of carotid endarterecto-
my specimens from symptomatic patients vs. 27% of speci-
mens from asymptomatic patients in one study. 28

254

chapter 24 Cerebral ischemia

Cerebral hypoperfusion

Cerebral hypoperfusion is another mechanism of cerebral
ischemia. Transient hypotension in the face of a hemodynami-
cally significant carotid stenosis was at one time thought to be
the most common cause of TIA. Using a tilt table to induce hy-
potension, however, numerous investigators were unable to
reproduce neurologic symptoms. 29 Likewise, pharmacologic
lowering of blood pressure in hypertensive patients with TIAs
does not typically reproduce neurologic symptoms. More-
over, intraoperative clamping of the internal carotid artery
during endarterectomy is tolerated by 90% of patients.

Thus, although cerebral hypoperfusion does play a role in
the pathogenesis of focal cerebral ischemia, it may play a more
important role in the setting of a highly stenotic internal
carotid artery with a contralateral occlusion, with vertebral
occlusions, or with an incomplete circle of Willis. 30

Transient ischemic attack

A TIA is defined as a focal neurologic deficit of ischemic origin
that lasts less than 24 h; most resolve within 30 min. TIAs are
most commonly embolic, but they can also be of hemodynam-
ic origin. Carotid system TIAs typically involve the cerebral
hemisphere, in the distal distribution of the middle cerebral
artery, producing symptoms of numbness or weakness in the
contralateral arm or leg, although a wide variety of symptoms
can occur. Ocular attacks, known as amaurosis fugax, present
with transient ipsilateral monocular blindness and are associ-
ated with carotid bifurcation disease. Vertebrobasilar
TIAs have a varied presentation and can include diplopia,
dysarthria, dizziness, facial numbness, and weakness or
numbness of one or both sides of the body.

The proposed mechanism of TIA is embolization of fib-
rin-platelet material from atherosclerotic sites. As the platelet
material fragments and dissolves, the neurologic symptoms
resolve. Multiple TIAs with the same neurologic pattern in the
distribution of the middle cerebral artery suggest a carotid bi-
furcation source, as does amaurosis fugax, whereas multiple
TIAs with differing neurologic patterns are more likely to be of
cardiac origin.

The fact that carotid system TIAs produce the same symp-
toms repeatedly can be explained by the laminar flow of blood.
A particle that enters the bloodstream from the same fixed pos-
ition will deposit in the same distal site each time.

Role of carotid endarterectomy

To define appropriately the role of carotid endarterectomy in
the management of carotid artery stenosis one must know the
natural history of symptomatic and asymptomatic lesions, as

well as the risk of surgical intervention. After a TIA, the prob-
ability of stroke at 1 year was 13% in a Rochester, Minnesota,
population-based study 31 Chambers and Norris 32 found that
in asymptomatic patients with greater than 75% carotid steno-
sis, the risk at 1 year of developing a neurologic event (stroke or
TIA) was 18% and that of completed stroke was 5%. The risk of
stroke and death following carotid endarterectomy (CEA) de-
pends on the patient's medical risk factors and neurologic
status, as well as the expertise of the surgeon. In a good-risk pa-
tient with unilateral carotid stenosis and no history of stroke,
the combined risk of stroke and death should be less than 3%.

The North American Symptomatic Carotid Endarterectomy
Trial (NASCET) I convincingly showed that carotid end-
arterectomy benefited patients with recent hemispheric and
retinal TIAs or nondisabling strokes and ipsilateral high-
grade stenosis (70-99%). Two-year ipsilateral stroke rates
were 9% for the endarterectomy group compared with 26% in
the medical (aspirin) group. 33 The European Carotid Surgery
Trial showed a 3-year stroke rate of 14.9% for the surgery
group compared with 26.5% for the control group, an absolute
benefit of 11.6% for symptomatic patients with stenosis more
than 80% of diameter. 34 In NASCET II, patients with sympto-
matic carotid stenosis of 70% or more continued to derive a
substantial benefit from endarterectomy that persisted up to
8 years of follow-up. In this study, investigators also looked at
the benefit of endarterectomy in symptomatic patients with
moderate carotid stenosis of 50-69%. At 5 years, the risk of
ipsilateral stroke was 15.7% in patients treated with carotid
endarterectomy vs. 22.2% in patients treated medically. The
subgroups that benefited the most were men, patients with re-
cent stroke as the qualifying event, and patients with hemi-
spheric symptoms. Symptomatic patients with less than 50%
stenosis showed no benefit from carotid endarterectomy. 35

Patients with asymptomatic carotid artery stenosis are also
potential candidates for endarterectomy. The Asymptomatic
Carotid Artery Stenosis (ACAS) study showed that patients
with carotid artery stenosis of 60% or greater reduction in di-
ameter and whose general health makes them good candidates
for elective surgery have a 5-year stroke and death rate of 5.1%,
compared with 11.0% for patients treated medically. In order
to obtain this reduction in stroke, the authors noted a requisite
perioperative morbidity and mortality of less than 3%. 36/37

Recently with the explosion of endovascular technology,
carotid angioplasty stenting (CAS) has been widely popular-
ized. In a group of high-risk patients, the SAPPHIRE Trial
reported a rate of death/stroke of 4.5% in the CAS group
compared with 6.6% in the CEA group, which was not signifi-
cant. However, when the perioperative myocardial infarction
rate was added to the analysis, CAS had a 5.8% rate of
stroke /death /MI, which was significantly lower than the
12.6% in the CEA group. 38 An ongoing National Institutes
of Health-sponsored trial, the Carotid Revascularization vs.
Stent Trial (CREST) is under way comparing carotid stenting
with surgery in lower risk patients.

255

pa rt I Vascular pathology and physiology

References

1. McPherson RW, Traystman RJ. Physiology of the cerebral circula-
tion. In: Giordano JM, Trout HH III, DePalma RG, eds. The Basic
Science of Vascular Surgery. Mount Kisco, NY: Futura, 1988:569.

2. Sundt TM Jr. Metabolic-cerebral blood flow couple. In: Occlusive
Cerebrovascular Disease: Diagnosis and Surgical Management.
Philadelphia: WB Saunders, 1987:3.

3. Vaililala M, Lee L, Lam A. Cerebral blood flow and vascular
physiology. Anesthes Clin North Am 2002; 20:247.

4. Severinghaus JW, Lassen N. Step hypocapnia to separate arterial
from tissue PC02 in the regulation of cerebral blood flow. Circ Res
1967; 20:272.

5. Laudignon N, Beharry K, Farri E, Aranda JV. The role of adenosine
in the vascular adaptation of neonatal cerebral blood flow during
hypotension. / Cereb Blood Flow Metab 1991; 1:24.

6. Najarian T, Marrache AM, Dumont I et ah Prolonged hypercapnia-
evoked cerebral hyperemia via K channel- and prostaglandin
E2-dependent endothelial nitric oxide synthase induction. Circ
Res 2000; 87:1149.

7. Cohen PJ, Alexander SC, Smith TC et ah Effects of hypoxia and nor-
mocarbia on cerebral blood flow and metabolism in conscious
man. JAppl Physiol 1967; 23:183.

8. Astrup J, Siesjo BK, Symon L. Thresholds in cerebral ischemia: the
ischemic penumbra. Stroke 1981; 12:723.

9. Branston NM, Symon I, Crockard HA et ah Relationship between
the cortical evoked potential and local cortical blood flow follow-
ing acute middle cerebral artery occlusion in the baboon. Exp
Neurol 1974; 45:195.

10. Sundt TM Jr, Sharbrough PW, Anderson RE et ah Cerebral blood
flow measurement and electroencephalograms during carotid en-
darterectomy JNeurosurg 1974; 41:310.

11 . Branston NM, Strong AJ, Symon L. Extracellular potassium activ-
ity, evoked potentials and tissue blood flow: relationship during
progressive ischaemia in baboon cerebral cortex. J Neurol Sci 1977;
32:305.

12. Morawetz RB, Crowell RH, DeGirolani U et ah Regional cerebral
blood flow thresholds during cerebral ischemia. Fed Proc 1979;
38:2493.

13. Branston NM, Hope T, Symon L. Barbiturates in focal ischemia of
primate cortex: effects on blood flow distribution, evoked poten-
tial and extracellular potassium. Stroke 1979; 10:647.

14. Heiss WD. Flow thresholds of functional and morphological
damage of brain tissue. Stroke 1981; 14:329.

15. Jafar JJ, Crowell RM. Focal ischemic thresholds. In: Wood JH, ed.
Cerebral Blood Flow. New York: McGraw-Hill, 1987;452.

16. Bell BA, Symon TD, Branston NM. Cerebral blood flow and time
thresholds for the formation of ischemic cerebral edema, and
effect of reperfusion in baboons. / Neurosurg 1985; 62:41 .

17. Tamura A, Asano T, Sano K. Correlation between cerebral blood
flow and histological changes following temporary middle cere-
bral artery occlusion. Stroke 1980; 11:487.

19. Caplan LR, Gorelick PB, Hier DB. Race, sex and occlusive cere-
brovascular disease: a review. Stroke 1986; 17:648.

19. Zarins C. Hemodynamic factors in atherosclerosis. In: Moore W,
ed. Vascular Surgery. New York: Grune and Stratton, 1986:99.

20. Solberg LA, Eggen DA. Localization and sequence of develop-

ment of atherosclerotic lesions in the carotid and vertebral arter-
ies. Circulation 1971:711.

21. Mackey WC. O'Donnell TF, Callow AD. Carotid endarterectomy
in patients with intracranial vascular disease: short-term risk and
long-term outcome. / Vase Surg 1989; 10:432.

22. Roederer GO, Langlois YE, Chan ARW, Chikos PM, Thiele BL,
Strandness E. Is siphon disease important in predicting outcome
of carotid endarterectomy? Arch Surg 1983; 118:1177.

23. Moore WS. Does tandem lesion mean tandem risk in patients with
carotid artery disease? / Vase Surg 1988; 7:454.

24. Castaigne P, Lhemmitte F, Gautier JC et ah Arterial occlusions in
the vertebrobasilar system: a study of 44 patients with post-
mortem data. Brain 1973; 96:133.

25. Fields WS, Lemak NA. Joint study of extracranial arterial occlu-
sion. VII. Subclavian steal: a review of 168 cases. JAMA 1972;
111:1139.

26. Moore WS, Hall AD. Ulcerated atheroma of the carotid artery. Am
J Surg 1968; 116.

27. Moore WS, Hall AD. Importance of emboli from carotid bifurca-
tion in pathogenesis of cerebral ischemic attacks. Arch Surg 1970;
101:708.

28. Lusby RJ, Ferrell LD, Wylie EJ. The significance of intraplaque
hemorrhage in the pathogenesis of carotid atherosclerosis. In:
Bergan JJ, Yao JST, eds. Cerebrovascular Insufficiency. New York:
Grune & Stratton, 1983;41.

29. Millikan CH. The pathogenesis of transient focal cerebral is-
chemia. Circulation 1965; 32:438.

30. Moore WS. Fundamental considerations in cerebrovascular dis-
ease. In: Rutherford RB, ed. Vascular Surgery. Philadelphia: WB
Saunders, 1989.

31. Whisnant JP, Wiebers DO. Clinical epidemiology of transient
ischemic attack (TIA) in the anterior and posterior circulation.
In: Sundt TM Jr, ed. Occlusive Cerebrovascular Disease: Diagnosis
and Surgical Management. Philadelphia: WB Saunders, 1987:
63.

32. Chambers BR, Norris JW. Outcome in patients with asymptomatic
neck bruits. N Engl} Med 1986; 315:860.

33. North American Symptomatic Carotid Endarterectomy Trial col-
laborators. Beneficial effect of carotid endarterectomy in sympto-
matic patients with high-grade carotid stenosis. N Engl] Med 1991;
325:445.

34. European Carotid Surgery Trialists' Collaborative Group. Ran-
domised trial of endarterectomy for recently symptomatic carotid
stenosis: final results of the MRC European Carotid Surgery Trial
(ECST). Eancet 1998; 351:1370.

35. Barnett HJM, Tayor DW, Eliasziw M et ah Benefit of carotid en-
darterectomy in patients with symptomatic moderate or severe
stenosis. N Engl} Med 1998; 339:1415.

36. Executive Committee for the Asymptomatic Carotid Atheroscle-
rosis Study. Endarterectomy for asymptomatic carotid artery
stenosis. JAMA 1995; 27 '3:1 421.

37. Hobson RW II, Weiss DG, Fields WS et ah Efficacy of carotid en-
darterectomy for asymptomatic carotid stenosis. N Engl J Med
1993; 328:221.

38. Yadav J. SAPPHIRE: Stenting and Angioplasty with protection in
patients at high risk for endarterectomy. Presented at 75th Scien-
tific Sessions of the American Heart Association. Nov 17-20, 2002.
Chicago, Illinois.

256

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Pathophysiology of spinal cord ischemia

Larry H. Hollier

Spinal cord ischemia is one of the most dreaded complications
of aortic surgery. Its occurrence has largely been unpredictable
and unpreventable. Moreover, the etiology of spinal cord
injury during aortic surgery has been poorly understood
(Fig. 25.1).

The highest risk of ischemic spinal cord injury in aortic
surgery is associated with repair of type I and II thoraco-
abdominal aortic aneurysms 1-3 (Fig. 25.2). This is especially
true if the repair is performed for aortic dissection in this area,
where the incidence of paraplegia may be as high as 40% of
cases. 3 For isolated aneurysms of the thoracic aorta, or even of
the lower thoracoabdominal area (e.g. type III and IV thoraco-
abdominal aneurysms), the risk of paraplegia is less— in the
neighborhood of 1-5% in many published reports. 1-14 If the
aortic replacement is limited to the infrarenal aorta, the risk of
ischemic spinal cord injury is less than 1%. 15/16

Historical approaches to prevention

The precise cause of paraplegia or paraparesis occurring after
aortic surgery is poorly understood. Simplistic views held that
failure to implant intercostal arteries was the primary cause of
spinal cord injury and paraplegia. However, paraplegia some-
times occurred even though intercostal arteries were reim-
planted. 1-4 ' 17 In these cases, it was assumed that perioperative
hypotension, thrombosis of the reimplanted intercostal arter-
ies, or embolization into the anterior spinal artery were sec-
ondary mechanisms for the development of paraplegia in
those patients; however, this often seemed unlikely. 1 ' 4 ' 18 Early
attempts at the prevention of paraplegia centered primarily on
trying to maintain collateral flow to the cord during the opera-
tive procedure. 8-12 ' 19

One protective technique that proved to be effective during
repair of isolated thoracic aneurysms is use of a heparin-
bonded shunt or arteriofemoral or femorofemoral bypass;
several series have demonstrated excellent results with this
technique for repair of thoracic aneurysms. 10-14 ' 20-22 However,
the use of shunts or bypass in the management of extensive

type I or II thoracoabdominal aneurysms does not appear to
protect against spinal cord injury. 3 ' 17 ' 23

Other authors have suggested somatosensory or motor-
evoked potential monitoring as a means of determining which
patients might benefit from intercostal reimplantation. 9 ' 10 ' 24-26
These techniques are safe and relatively simple, but they can
provide erroneous information when peripheral nerve is-
chemia occurs. Moreover, they have shown poor correlation
with postoperative results. Of greater significance is the fact
that spinal cord monitoring does not in itself prevent para-
plegia. It simply indicates when the cord is ischemic; some
specific measure (e.g. intercostal artery reimplantation) is still
necessary in these patients. Moreover, although the monitor-
ing may suggest maintenance of cord function during the op-
erative procedure, delayed onset of the neurologic deficit has
been clearly recognized to occur in up to half of patients who
develop paraplegia or paraparesis after aortic surgery. 1 ' 3 ' 17

Some authors have suggested routine spinal artery angio-
graphy to identify the location of critical intercostal arteries
that supply the anterior spinal artery. 2 ' 27 With extensive thoraco-
abdominal aneurysms, about half of patients have one or
more critical intercostal arteries arising from the aneurysmal
aorta. 2 Thus, two things seem evident: first, some patients
clearly do need to have one or more intercostal arteries reim-
planted; second, intercostal artery reimplantation, even when
the critical intercostal arteries are identified preoperatively,
does not entirely prevent paraplegia.

Since 1987, we have advocated the routine use of cere-
brospinal fluid (CSF) drainage and have now done this
in more than 100 patients undergoing thoracoabdominal
aneurysm repair. 4 ' 28 In our early experience with the first 50
patients undergoing CSF drainage without other adjunctive
methods, no patient awoke with a neurologic deficit. How-
ever, two patients developed delayed-onset paraparesis, one
after 5 days and another after 24 h. The first patient has not had
intercostal artery reimplantation because of technical diffi-
culties. The second patient, however, after developing a
rapidly progressive neurologic deficit the day after operation,
underwent withdrawal of an additional 40 ml of CSF. Within

257

pa rt I Vascular pathology and physiology

Good collaterals

No ischemia

No deficit

Aortic cross-clamping

Marginal
collaterals

No clinical deficit

incomplete spinal
cord ischemia

Delayed paraplegia
or paraparesis

Absent collaterals

Severe ischemia

Immediate paraplegia

Immediate paraparesis

Figure 25.1 Clinical implications of adequacy
of collateral event in relation to adequacy of
collateral blood flow and spinal cord perfusion.

6h his neurologic deficit had entirely reversed, and he made
an uneventful recovery without neurologic sequelae. CSF
drainage appeared beneficial, but it was clear that additional
etiologic factors were also important.

Some researchers advocate the diagnostic use of hydrogen
ion injection in the intercostal arteries at the time of aneurysm
repair. 29 Early results with this technique indicated that one
could indeed identify intercostal arteries that supplied the
spinal cord. However, one was still left with the need for intra-
costal reimplantation. Although somewhat easier than inter-
costal angiography, it does not appear that hydrogen ion
injection is any more efficacious than is preoperative angiogra-
phy. Also, intercostal artery reimplantation alone does not
prevent paraplegia.

Investigators have added papaverine to the CSF as a means
of dilating the anterior spinal artery and presumably improv-
ing collateral blood flow. 30 While this theoretically would be
of added advantage when combined with intercostal CSF
drainage, there is no evidence that papaverine has changed
the incidence of spinal cord injury.

Pathophysiology of paraplegia

Each of the approaches just outlined focused on one specific
aspect thought to be associated with spinal cord injury.
However, the pathophysiology of spinal cord injury during
thoracoabdominal aneurysm repair has not been fully de-
fined. Because of this, we undertook a series of experiments to
try to identify the variables that play a role in the causation of
paraplegia and paraparesis, both acute and delayed onset. The
findings from these studies have led us to postulate the mech-
anisms that can cause paraplegia during thoracoabdominal
aneurysm repair.

We postulate that ischemic spinal cord injury is due to the
following interrelated variables:

• Severity of the initial ischemic event.

• Rate of metabolism of neurons during the ischemic period.

• Reperfusion injury that occurs after restoration of blood
flow.

Severity of ischemia

If there is little or no ischemia of the cord, despite the extent of
resection of an aortic resection, one would not expect to get
ischemic spinal cord injury. Thus, if a patient has adequate
collateral blood flow to the anterior spinal artery such that flow
is not significantly diminished during aortic clamping, the pa-
tient should remain free of cord injury (Fig. 25.3). Conversely,
if the patient's critical blood supply to the spinal cord arises
from the aorta that is replaced and no attempts are made to
revascularize the cord, that patient can be expected to have
severe permanent paralysis. If the patient has marginal arterial
collaterals to the spinal cord or if intercostal artery reimplanta-
tion is performed expeditiously, the patient may develop an
ischemic cord but have complete reversal of that injury after
flow is restored. However, if the ischemic time before restora-
tion of flow was prolonged and collaterals were insufficient to
adequately oxygenate the neurons, paraplegia could occur
despite intercostal reimplantation. Kieffer and colleagues
clearly documented this in patients who underwent spinal
cord angiography and subsequent aneurysm repair. 2

In a previous study published by Bower and colleagues,
blood flow to the spinal cord was studied using injection of
radiolabeled microspheres. 31 Clamping of the thoracic aorta
caused a significant reduction in blood flow to the spinal cord,
particularly in the lower thoracic and lumbar areas of the cord.
Drainage of surplus CSF resulted in improved blood flow to
the spinal cord.

This latter phenomenon— namely, the improvement in
blood flow to the cord by reduction of CSF pressure— has
been extensively studied by Miyamoto and others. 29,32-34 The

258

CHAPTER25 Pathophysiology of spinal cord ischemia

Figure 25.2 Classification of thoracoabdominal aortic aneurysms based on extent of aneurysmal changes.

259

Figure 25.3 (A) Critical intercostal arteries supplying the anterior spinal artery arise
above the level of aortic aneurysmal involvement. Repairof this aneurysm would pose
minimal risk of paraplegia. (B) Critical intercostal arteries supplying the anterior artery
arise below the aneurysm. Use of shunts or bypass would be protective in minimizing
spinal cord ischemia. Similarly, rapid aneurysm repair would minimize cord ischemia,
and the risk of paraplegia would be low. (C) Critical intercostal arteries arise from the
midportions of the aneurysm. If these vessels are the sole blood supply to the thoracic
spinal cord, reimplantation is mandatory. Additionally, if the ischemia interval before
restoration of flow to the cord is too long, paraplegia may occur despite intercostal
artery reimplantation.

CHAPTER25 Pathophysiology of spinal cord ischemia

Spinal cord

Vertebral
foramen

Arterial
inflow

Figure 25.4 Relationship of cerebrospinal
fluid to arterial perfusion of the spinal cord.

Cerebrospinal
fluid

Venous
outflow

perfusion dynamics can best be understood by considering
the elements of the spinal canal to function as a Starling resis-
tor. Arterial blood flow goes to the cord through the anterior
and posterior spinal arteries and drains through the spinal
veins. The spinal cord is surrounded by CSF, all of which is en-
cased within the bony unyielding spinal canal. When a proxi-
mal aortic clamp is placed and the distal aortic pressure falls,
CSF pressure rises. 28 ' 32-35 The elevation in CSF pressure is
probably due to a combination of factors, including increased
production of CSF and increased brain volume. The net effect
of increase in CSF pressure is reduction of the arterial perfu-
sion pressure to the spinal cord. 32,33,36 (Spinal cord perfusion
pressure equals spinal artery pressure minus CSF pressure.)
When the thoracic aorta is clamped, distal aortic pressure de-
creases in the intercostal artery and the CSF pressure may rise
significantly, worsening the spinal cord perfusion. This mech-
anism explains the beneficial effects demonstrable with CSF
drainage, which may reduce the severity of the ischemic
event 37 (Fig. 25.4).

Neuronal metabolism

Extensive experimental and clinical work has clearly
documented that the ischemic tolerance of the brain can be
extended by reducing neuronal metabolism. 18 ' 38-53 High-dose
intravenous pentothal and systemic cooling have both been
shown to prolong safe ischemia time of the brain. Indeed, deep
hypothermic cardiac arrest can be tolerated safely for 30 min or
more.

Experimentally, we have demonstrated that both systemic
cooling and regional (CSF) cooling are highly protective
against ischemic spinal cord injury 38 ' 52 Maughan and
colleagues also documented the protective effect of CSF
cooling. 46

Thus, it would appear that the rate of metabolism of neurons
in the spinal cord may play a role in the relative susceptibility

of the neurons to ischemic injury. This has clinical implica-
tions, since some researchers have strongly advocated the
maintenance of normothermia during thoracoabdominal
aneurysm repair. 3 ' 17 Although normothermia may be helpful
in minimizing cardiac irregularities and coagulopathy, it may
increase the risk of spinal cord injury during thoracoabdomi-
nal aortic repair. Conversely, cooling of the spinal cord can
minimize cord injury during ischemia.

Reperfusion injury

Reperfusion injury is a broad term that encompasses changes
that occur following the ischemic event. The ischemic injury
cascade includes both direct and complement-mediated
leukocyte activation, the endothelial cell production of adhe-
sion molecules, vasoconstriction secondary to arachidonic
acid metabolism, neuronal membrane injury, and spinal cord
hyperperfusion and edema. 1 ' 7 ' 39 ' 54-64

The full-blown injurious effects of reperfusion do not neces-
sarily happen immediately but may occur over hours and
days. Obviously, a spectrum of injury can occur. This is the
most common etiology of delayed-onset paraplegia. 18,65
Severe prolonged ischemia can result in neuronal infarction
and neuronal death with resultant immediate paraparesis or
paraplegia. Mild ischemia of short duration may cause mo-
mentary neuronal dysfunction with complete return of spinal
cord function and no late sequelae. However, intermediate
levels of ischemia— not so severe as to cause infarction yet se-
vere enough to initiate reperfusion phenomena— may cause
no immediate neuronal dysfunction but instead cause
delayed-onset paraplegia or paraparesis 1-5 days later as the
injury cascade progresses. 18

By using different radiolabeled microspheres, we demon-
strated that marked hyperperfusion occurs following sig-
nificant ischemia of the spinal cord. 37 This pronounced
hyperperfusion after ischemia can lead to neuronal dysfunc-

261

pa rt I Vascular pathology and physiology

tion on either a transient or permanent basis. Additionally,
spinal cord edema can result in secondary elevation of CSF
pressure, which in turn can further compromise blood flow in
the spinal cord.

We previously demonstrated that intermediate levels of
spinal cord ischemia could reliably produce delayed-onset
paraplegia in rabbits. Brief periods of spinal cord ischemia,
produced in awake and active New Zealand rabbits, resulted
in transient paralysis that was totally reversible after blood
flow was restored. Prolonged periods of ischemia resulted in
permanent paralysis despite complete restoration of blood
flow. Most significantly, however, intermediate levels of is-
chemia (21-22 min in rabbits) resulted in initial paralysis with
complete return of spinal cord function after restoration of
blood flow, but with complete permanent paralysis occurring
24 h later. This clearly showed that the initial period of inter-
mediate ischemia initiated the mechanisms that resulted in
subsequent delayed-onset paraplegia. 18 The clinical implica-
tions of this are illustrated in Figure 25.1.

Clark and others confirmed the importance of leukocyte ac-
tivation and production of adhesion molecules in the causa-
tion of paraplegia when they documented a reduction of
paraplegia by administration of monoclonal antibodies to the
CD18 adhesion molecule produced by leukocytes. 66 ' 67
Numerous other studies have similarly demonstrated that
avoidance of hyperglycemia, the use of free radical scav-
engers, calcium-channel blockers, prostaglandins, steroids,
and multiple other agents that ameliorate various aspects of
reperfusion injury may reduce neuronal injury. 39,68-87

Recommendations for clinical practice

All of these experimental and clinical data provide compelling
evidence that paraplegia, both acute and delayed onset, fol-
lowing thoracoabdominal aneurysm repair is due to the inter-
related variables of the severity of the ischemic event, the rate
of neuronal metabolism during the time of ischemia, and the
secondary effects of the reperfusion phenomena. If one wants
to reduce the risk of paraplegia, it is evident that each of these
injurious mechanisms should be addressed. The following is
the approach that we have used to minimize the risk of neu-
ronal injury from each of these components. 65

Severity of ischemia

To maximize collateral blood flow during the time of aortic
cross-clamping, we start volume loading the patient as soon as
the operation is begun. Anitroprusside drip is instituted when
the skin incision is made and crystalloid solution is infused at
an accelerated rate while blood pressure, filling pressure, and
cardiac indices are carefully monitored. At the time of aortic
cross-clamping, an attempt is made to maintain the arterial
systolic blood pressure proximal to the clamp at a slightly ele-

vated level, about 170-180 mmHg proximal to the aortic clamp
to improve collateral flow. Cardiac function is carefully
monitored at this stage by the use of transesophageal two-
dimensional echocardiography. To control the elevation of
CSF pressure that frequently occurs following aortic cross-
clamping, an intrathecal catheter is routinely inserted before
preparation and draping of the patient and CSF pressure is
monitored continually. CSF is removed as necessary to keep
the CSF pressure below 10 mmHg.

Reducing neuronal metabolism

Cooling is the most effective way of reducing neuronal metab-
olism and thus of protecting the spinal cord from ischemic in-
jury. Several techniques have been tried, including induced
systemic hypothermia with cardiopulmonary bypass, cooling
of the CSF intrathecally, and mild passive systemic hypother-
mia. We prefer the last approach, in which passive cooling of
the patient is allowed during the early stages of the operation.
No attempt is made to warm the inspired gases or the intra-
venous fluids administered; the warming blanket is not
turned on and the ambient temperature of the room is signifi-
cantly reduced. This generally allows the patient's tempera-
ture to drift down to about 32-34°C. After completion of
intercostal artery reimplantation and restoration of blood flow
to the spinal cord, rewarming of the patient proceeds vigor-
ously. The ambient temperature of the room is increased, in-
spired air and intravenous fluid are warmed, and the warming
blanket is turned on. Although it generally has not been pos-
sible to return the temperature to normal, we find that the
temperature does rise sufficiently to avoid cardiac and
coagulopathic complications.

A large bolus of intravenous pentothal, 10-20 mg/kg, is
administered intravenously 5 min before application of the
proximal aortic clamp. No attempt is made to monitor burst
suppression on the electroencephalogram.

Minimizing reperfusion injury

Cooling in itself is effective in reducing reperfusion injury
since it decreases neuronal metabolism and thus reduces pro-
duction of injurious metabolic byproducts. Additionally, we
administer high-dose intravenous steroids at the beginning of
treatment in hopes of providing some degree of membrane
stabilization. Mannitol, a mild free radical scavenger, is ad-
ministered intravenously at 25 g just before aortic clamping;
12.5 g is also administered just before removing the aortic
clamp. Prostaglandin, superoxide dismutase, calcium-
channel blockers, naloxone, and various opiate antagonists
have all been used by some investigators in an effort to reduce
reperfusion injury. We have insufficient experience with these
modalities to comment about their efficacy, and we do not use
any of these adjuncts at this time.

As mentioned, one of the detrimental aspects of the reperfu-

262

Table 25.1 Thoracoabdominal aneurysm repair: mortality

CHAPTER25 Pathophysiology of spinal cord ischemia

Type

III

Total

Patients

Deaths in operating room

Deaths at 30 days

Total

34
3
2
5(14.7%)

53
2
2
4(7.4%)

4
4(7.4%)

1
1 (1.6%)

203

5(2.5%)

9(4.4%)

14(6.9%)

Table 25.2 Thoracoabdominal aneurysm repair: neurologic deficit

Type

III

Total

Patients

Patients with neurologic deficit

34
3(8.8%)

53
3(5.7%)

54
2(3.7%)

62
1 (1.6%)

203

9(4.4%)

sion phenomena is hyperperfusion of the spinal cord with re-
sultant spinal cord edema, elevation of CSF pressure because
of cord edema and expansion within the closed space of the
bony spinal canal, and thus decrease in spinal cord perfusion.
This edema appears to become progressively worse 1-3 days
after the ischemic insult. Because of this, in addition to moni-
toring CSF pressure and draining the fluid as necessary during
operation to keep the CSF pressure below lOmmHg, we con-
tinue to monitor and drain CSF for 1-3 days postoperatively.
We have strong anecdotal evidence that suggests that CSF
pressure elevation in the postoperative period is one of the
major contributors to delayed-onset paraplegia. Indeed, since
instituting prolonged CSF drainage, we have seen no further
incidence of delayed-onset paraplegia when the drainage is
performed in conjunction with the adjunctive modalities pre-
viously described. 4 The amount of CSF drained in the post-
operative period varies considerably from patient to patient.
Some patients require drainage of more than 1 1 to maintain a
CSF pressure below 10 mmHg.

Results

As in most large series spanning several years, techniques
to provide spinal cord protection during thoracoabdominal
aneurysm repair have evolved gradually over the years, as has
the technique of spinal cord protection just described. Thus, an
overall analysis of neurologic injury in these patients does not
fully demonstrate the protective effect that is provided by ad-
junctive treatment. Nonetheless, since our experience in tho-
racoabdominal aneurysm repair now numbers more than 200

cases, one can derive some evidence of relative success with
these techniques when compared with other reports in the lit-
erature (Table 25.1 and 25.2). In an analysis of 203 thoraco-
abdominal aneurysms between 1980 and 1993, the immediate
perioperative mortality was 2.5%. The importance of this is
that 97% of the patients survived at least long enough for an
adequate assessment of their neurologic status postopera-
tively. The overall incidence of paraplegia or paraparesis is
4.4%. With one exception, every patient who developed a
neurologic deficit had incomplete or no intercostal artery
reimplantation performed. The patient who developed
paraparesis despite intercostal reimplantation was operated
on early in the series, before the use of CSF drainage. The spe-
cific incidence of neurologic deficit by aneurysm type in this
series is presented in Table 25.2.

Of more importance than an analysis of neurologic deficits
in a large series of heterogeneous thoracoabdominal
aneurysms is the analysis of those patients undergoing repair
of type I and II thoracoabdominal aneurysms. In a prospective
randomized trial of CSF drainage vs. no CSF drainage in pa-
tients undergoing thoracoabdominal aneurysm repair for
type I or II thoracoabdominal aneurysms, Crawford and col-
leagues showed no significant difference in neurologic injury
whether or not CSF drainage was used. 17 They reported an
overall incidence of neurologic deficit in both groups in excess
of 30%. Because of institutional regulations, however, they
were not allowed to withdraw more than 50 ml of CSF at the
time of the operation and were not allowed to perform any
postoperative CSF drainage. Most significantly, they docu-
mented that intercostal reimplantation was not performed
routinely and no adjunctive pharmacologic treatment was
given to specifically minimize reperfusion injury. A subse-

263

pa rt I Vascular pathology and physiology

Decreased spinal
artery pressure

Aortic cross-clamping

Shunt t increase
blood pressure, bypass

Drain CSF

Increased
CSF pressure

Decreased spinal
cord perfusion pressure

Reduce cord metabolism:
cooling, barbiturates, anesthetics

Decrease glucose load
No decrease in fluids

Ischemia

P read minister steroids

t Shorten clamp time

Free radical scavengers

Hyperperfusion

Spinal cord edema

Free radical injury
Leukocyte injury

Drain CSF

■

Membrane
stabilizers

Decreased spinal cord
perfusion pressure

Cellular damage

___...,. .

Neuronal death

Figure 25.5 Ischemic neurologic injury cascades and possible methods of
intervention. (From Hollier LH, Marino RJ. Thoracoabdominal aneurysms. In:

Moore WS, ed. Vascular Surgery: A Comprehensive Review, 3rd edn.
Philadelphia: WB Saunders, 1 990:301 .)

quent study in the same institution by Safi and colleagues doc-
umented a reduction of neurologic deficit to 9% by the use of
CSF drainage and distal aortic perfusion 88 in a comparable
group of patients with type I and II aneurysms.

Comparison of these data strongly suggests that a com-
bined approach including routine CSF drainage preopera-
tively and postoperatively, routine intercostal artery
reimplantation, moderate hypothermia, and adjunctive phar-
macologic therapy is able to significantly reduce the risk of
paraplegia and paraparesis following thoracoabdominal
aneurysm repair.

Summary

The pathophysiology of spinal cord ischemia associated with
thoracoabdominal aneurysm repair is best described as the
consequence of the interrelated variables of the severity of the
ischemic insult, the rate of metabolism of the neurons during
the ischemic interval, and the secondary reperfusion pheno-
menon including hyperperfusion, cord edema, oxygen-de-
rived free radical injury, endothelial cell activation, and both
direct and complement-mediated leukocyte activation, as
well as other factors. If one is to hope to reduce the risk of para-
plegia during thoracoabdominal aneurysm repair, attention
should be paid to each of these variables and attempts made to
minimize the effect of each (Fig. 25.5).

References

1. Hollier LH, Moore WM Jr. Avoidance of renal and neurological
complications following thoracoabdominal aortic aneurysm
repair. Acta Chir Scand 1990; 555(Suppl.):129.

2. Kieffer E, Richard R, CHivas J et ah Preoperative spinal cord arteri-
ography in aneurysmal disease of the descending thoracic and
thoracoabdominal aorta: preliminary results in 45 patients. Ann
Vase Surg 1989; 3:34.

3. Crawford ES, Crawford JL, Safi HJ et ah Thoracoabdominal aortic
aneurysms: preoperative and intraoperative factors determining
intermediate and long-term results in 605 patients. / Vase Surg
1986; 3:389.

4. Hollier LH, Money SR, Naslund TC et ah Risk of spinal cord dys-
function in patients undergoing thoracoabdominal aortic replace-
ment. Am J Surg 1992; 164:120.

5. Cox GS, O'Hara PJ, Hertzer NR etah Thoracoabdominal aneurysm
repair: a representative experience. / Vase Surg 1992; 15:780.

6. Golden MA, Donaldson MC, Whittemore AD, Mannick JA. Evolv-
ing experience with thoracoabdominal aortic aneurysm repair at a
single institution. / Vase Surg 1991; 14:792.

7. Naslund TC, Hollier LH. Etiology, prevention and treatment of
delayed-onset paraplegia after aortic surgery. In: Veith FJ, ed. Cur-
rent Critical Problems in Vascular Surgery, Vol. 4. St Louis: Quality
Medical Publishing.

8. Livesay JJ, Cooley DA, Ventemiglia RA et ah Surgical experience in
descending thoracic aneurysmectomy with and without adjuncts
to avoid ischemia. Ann Thorac Surg 1985; 39:37.

264

CHAPTER25 Pathophysiology of spinal cord ischemia

9. Laschinger JC, Cuningham JN Jr, Catinella FP et al. Detection and
prevention of intraoperative spinal cord ischemia after cross-
clamping of the thoracic aorta: use of somatosensory evoked
potentials. Surgery 1982; 92:1109.

10. Cuning JN Jr, Laschinger JC, Merkin HA et al. Measurement of
spinal cord ischemia during operations upon the thoracic aorta:
initial clinical experience. Ann Surg 1982; 34:299.

11. Molina JE, Cogordam J, Einzig S et al. Adequacy of ascending
aorta-descending aorta shunt during cross-clamping of the tho-
racic aorta for prevention of spinal cord injury. / Thorac Cardiovasc
Surg 1985; 90:126.

12. Verdant A, Page A, Cossette R et al. Surgery on the descending tho-
racic aorta: spinal cord protection with the Gott shunt. Ann Thorac
Surg 1988; 46:147.

13. Carlson DE, Karp RB, Kochoukos NT. Surgical treatment of
aneurysms of the descending thoracic aorta: an analysis of 85
patients. Ann Thorac Surg 1983; 35:58.

14. Donahoo JS, Brawley RK, Gott VL. The heparin-coated vascular
shunt for thoracic aortic and great vessel procedures: a ten-year
experience. Ann Thorac Surg 1977; 23:507.

15. Brown OW, Hollier LH, Pariolera PA et al. Abdominal aortic
aneurysm and coronary disease: a reassessment. Arch Surg 1981;
116:1484.

16. Elliott JP, Szilagzi DE, Hageman JH et al. Spinal cord ischemia sec-
ondary to surgery of the abdominal aorta. In: Towne J, Bernhard V,
eds. Complications in Vascular Surgery. Orlando: Grune & Stratton,
1985:291.

17. Crawford ES, Svensson LG, Hess KR et al. A prospective random-
ized study of cerebrospinal fluid drainage to prevent paraplegia
after high-risk surgery on the thoracoabdominal aorta. / Vase Surg
1990; 13:36.

18. Moore WM Jr, Hollier LH. The influence of severity of spinal cord
ischemia in the etiology of delayed-onset paraplegia. Ann Surg
1991; 213:427.

19. Wadouh F, Lindemann EM, Arndt CF et al. The arteria radicularis
magna anterior as a decisive factor influencing spinal cord dam-
age during aortic occlusion. / Thorac Cardiovasc Surg 1984; 88:1.

20. Walls JT, Boley T, Curtis J et al. Centrifugal pump support for repair
of thoracic aortic injury. MO Med 1991 :811 .

21. Kochoukos NT, Rokkas CK. Descending thoracic and thoracoab-
dominal aortic surgery for aneurysms or dissection: how do we
minimize the rise of spinal cord injury. Semin Thorac Cardiovasc
Surg 1993:47.

22. de Mol B, Hamerlijnck R, Boezeman E, Vermeulen FE. Prevention
of spinal cord ischemia in surgery of thoracoabdominal
aorta aneurysms: the Bio Medicus pump, the recording of
somatosensory evoked potentials and the impact on surgical
strategy. Eur J Cardiothorac Surg 1990; 12:658.

23. Lowell RC, Gloviczki P, Bergman RT et al. Failure of selective
shunting to intercostal arteries to prevent spinal cord ischemia
during experimental thoracoabdominal aortic occlusion. IntAngi-
o/1992;4:281.

24. Coles JG, Wilson GJ, Sima AF, Klement P, Tait GA. Intraoperative
detection of spinal cord ischemia using somatosensory cortical
evoked potentials during thoracic aorta occlusion. Ann Thorac
Surg 1982; 14:159.

25. Reuter DG, Tacker WA Jr, Badylak SF et al. Correlation of motor-
evoked potential response of ischemic spinal cord damage. / Tho-
rac Cardiovasc Surg 1992; 2:262.

26. Marini CP, Cunningham JN Jr. Evoked potentials: ten year experi-
ence with a value research and clinical tool. Semin Thorac Cardio-
vasc Surg 1992; 2:262.

27. Savander SJ, Williams GM, Trerotola SO et al. Preoperative spinal
artery localization and its relationship to postoperative neurolog-
ic complications. Radiology 1993; 1:167.

28. McCullough JL, Hollier LH, Nugent M. Paraplegia after thoracic
aortic occlusion: influence of cerebrospinal fluid drainage. / Vase
Surg 1988; 7:153.

29. Svensson LG, Patel V, Robinson MF et al. Influence of preservation
of perfusion of intraoperatively identified spinal cord blood sup-
ply on spinal motor-evoked potentials and paraplegia after aortic
surgery. / Vase Surg 1991; 13:355.

30. Svensson LG, Von Ritter CM, Groeneveld HT et al. Cross-
clamping of the thoracic aorta: influence of aortic shunts, laminec-
tomy, papaverine, calcium channel blocker, allopurinol, and
superoxide dismutase on spinal cord blood flow and paraplegia in
baboons. Ann Surg 1986; 204:38.

31. Bower TC, Murray MJ, Gloviczki P et al. Paraplegia after thoracic
aortic occlusion: influence of cerebrospinal fluid drainage. / Vase
Surgl988;7:153.

32. Oka Y, Miyamoto T. Prevention of spinal cord injury after cross-
clamping of the thoracic aorta. Jpn } Surg 1984; 14:159.

33. Miyamoto K, Ueno A, Wada T, Kimoto S. A new and simple
method of preventing spinal cord damage following temporary
occlusion of the thoracic aorta by draining the cerebrospinal fluid.
/ Cardiovasc Surg 1960; 1:188.

34. Blaisdell FW, Cooley DA. The mechanism of paraplegia after tem-
porary thoracic aortic occlusion and its relationship to spinal fluid
pressure. Surgery 1963; 51:351.

35. Gelman S, Reves JG, Fowler K et al. Regional blood flow during
cross-clamping of the thoracic aorta and infusion of sodium nitro-
prusside. / Thorac Cardiovasc Surg 1983; 85:287.

36. Griffiths IR, Pitts LH, Crawford RA, Trench JG. Spinal cord com-
pression and blood flow: the effect of raised cerebrospinal fluid
pressure on spinal cord blood flow. Neurology 1978; 28:1145.

37. Woloszyn TT, Marini CP, Coons MS et al. Cerebrospinal fluid
drainage and steroids provide better spinal cord protection dur-
ing aortic cross-clamping than does either treatment alone. Ann
Thorac Surg 1990; 49:78.

38. Naslund TC, Hollier LH, Money SR et al. Protecting the ischemic
spinal cord during aortic clamping: the influence of anesthetics
and hypothermia. Ann Surg 1992; 215:409.

39. Hollier LH. Protecting the brain and spinal cord. / Vase Surg 1987;
5:524.

40. VacantiFX, Ames A III. Mild hypothermia and magnesium protect
against irreversible damage during CNS ischemia. Stroke 1984;
15:695.

41 . Busto R, Dietrich WD, Glubus MYT et al. Small differences in intra-
ischemic brain temperature critically determine the extent of
ischemic neuronal injury. / Cerebral Blood Flow Metab 1987; 7:729.

42. Marin J, Labafo RD, Rico ML et al. Effect of pentobarbital on the
reactivity of isolated human cerebral arteries. / Neurosurg 1981;
54:521.

43. Drummond JC, Shapiro HM. Cerebral physiology. In: Miller RD,
ed. Anesthesia. New York: Churchill Livingstone, 1990:621 .

44. Berendes JN, Bredee JJ, Schipperheyn JJ, Mashhour YA. Mecha-
nisms of spinal cord injury after cross-clamping of the descending
thoracic aorta. Circulation 1982;66(Suppl. 1):112.

265

pa rt I Vascular pathology and physiology

45. Coles JH, Wilson GJ, Sima AF et ah Intraoperative management
of thoracic aortic aneurysm: experimental evaluation of perfu-
sion cooling of the spinal cord. / Thorac Cardiovasc Surg 1983;
85:292.

46. Maughan RE, Mohan C, Nathan IM et ah Intrathecal perfusion of
an oxygenated perfluorocarbon prevents paraplegia after aortic
occlusion. Ann Thorac Surg 1992; 5:818.

47. Rosomoff HL, Holaday DA. Cerebral blood flow and cerebral
oxygen consumption during hypothermia. Am } Physiol 1954;
179:85.

48. Davson H, Spaziani E. Effect of hypothermia on certain aspects of
the cerebrospinal fluid. Exp Neurol 1962; 6:118.

49. Stein PA, MitchenfelderJD. Cerebral protection with barbiturates:
relation to anesthetic effect. Stroke 1978; 9:140.

50. Arnfred I, Secher O. Anoxia and barbiturates: tolerance to anoxia
in mice influenced by barbiturates. Arch Intern Pharmacodyn Ther
1962; 139:67.

51 . Stein PA, Mitchenf elder JD. Cerebral protection with barbiturates:
relation to anesthetic effect. Stroke 1978; 9:1402.

52. Wisselink W, Becker M, Nguyen J et ah Protection of ischemic
spinal cord during aortic clamping: the influence of selective
hypothermia and spinal cord perfusion pressure. / Vase Surg 1994;
19:788.

53. Westaby S. Hypothermic thoracic and thoracoabdminal
aneurysm operative: a central cannulation technique. Ann Thorac
Surg 1992; 2:253.

54. Halliwell B, Gutteridge JMC. Oxygen-free radicals and the ner-
vous system. Trend Neurosci 1985; 8:22.

55. Chen ST, Hsu CY, Hogan EL et ah Thromboxane, prostacyclin, and
leukotrienes in cerebral ischemia. Neurology 1986; 36:466.

56. Barone GW, Joob AW, Flanagan TL, Dunn CE, Kron IL. The effect
of hyperemia on spinal cord function after temporary thoracic
aortic occlusion. / Vase Surg 1988; 8:535.

57. Jacobs TP, Shohami E, Baze W et ah Deteriorating stroke model:
histopathology, edema, and eicosanoid changes following spinal
cord ischemia in rabbits. Stroke 1987; 18:741.

58. Schmidley JW. Free radicals in central nervous system ischemia.
Stroke 1990; 27:1086.

59. Cao W, Carney JM, Duchon A et ah Oxygen free radical involve-
ment in ischemia and reperfusion injury to the brain. Neurosci Lett
1988; 88:233.

60. Oehmichen M. Inflammatory cells in the central nervous system.
ProgNeuropathol 1983; 5:277.

61. North RJ. Concept of activated macrophage. / Immunol 1979;
121:806.

62. Giulian D. Ameboid microglia as effectors of inflammation in the
central nervous system. J Neurosci Res 1987; 18:155.

63. Hickey R, Albin MS, Bunegin L, Gelineau J. Autoregulation of
spinal cord blood flow: is the cord a microcosm of the brain? Stroke
1986; 17:1183.

64. Hayashi N. Self propagation injured tissue by localized changes
of free radicals in the central nervous system. Hum Cell 1992;
4:354.

65. Hollier LH, Marino RJ, Kazmier FJ. Thoracoabdominal aortic
aneurysms: In: Moore WS, ed. Vascular Surgery: A Comprehensive
Review, 4th edn. Philadelphia: WB Saunders, 1993.

66. Clark WM, Madden KP, Rothlein R et ah Reduction of central ner-
vous system ischemic injury in rabbits using leukocyte adhesion
antibody treatment. Stroke 1991; 22:877.

67. Clark WM, Madden KP, Rothlein R, Zivin JA. Reduction of central
nervous system ischemic injury by monoclonal antibody to inter-
cellular adhesion molecule. JNeurosurg 1991; 75:354.

68. Pickard JD, Murray GD, Illingworth R. Effect of oral nimodipine
on cerebral infarction and outcome after subarachnoid hemor-
rhage: British aneurysm nimodipine trial. Br } Med 1989; 298:638.

69. Laschinger JC, Cunningham JN, Cooper MM et ah Prevention of
ischemic spinal cord injury following aortic cross-clamping: use
of corticosteroids. Ann Thorac Surg 1984; 38:500.

70. Granke K, Hollier LH, Zdrahal P et ah Longitudinal study of cere-
bral spinal fluid drainage in polyethylene glycol-conjugated
superoxide dismutase in paraplegia associated with thoracic
and aortic cross-clamping. / Vase Surg 1991; 13:615.

71 . Coles JC, Ahmed SN, Mehta HU, Raufman JCE. Role of free radical
scavenger in protection of spinal cord during ischemia. Ann Tho-
rac Surg 1986; 41:551.

72. Robertson CS, Foltz R, Grossman RG et ah Protection against
experimental ischemic spinal cord injury. / Neurosurg 1986; 64:
633.

73. Rhinehart JJ. Effects of corticosteroids upon human monocyte
function. N Engl} Med 1975; 292:236.

74. Norris JW, Hachinski VC. Megadose steroid therapy in ischemic
stroke. Stroke 1985; 16:150.

75. Norris DA, Weston WL, Sams WM. The effects of immunosup-
pression and anti-inflammatory drugs upon monocyte function in
vitro. / Lab Clin Med 1977; 90:569.

76. Steen PA, Newberg LA, Milde JH et ah Nimodipine improves cere-
bral blood flow and neurologic recovery after complete cerebral
ischemia in the dog. / Cereb Blood Plow Metab 1983; 3:38.

77. Lundy EF, Ball TD, Mandell MA et ah Dextrose administration
increases sensory/motor impairment and paraplegia after
infrarenal aortic occlusion in the rabbit. Surgery 1987; 102:737.

78. Johnson SH, Kraimer JM, Graeber GM. Effects of flunarizine on
neurological recovery and spinal cord ischemia in rabbits. Stroke
1993; 10:1547.

79. Katircioglu SF, Kucukaksu DS, Kuplulu S et ah Effects of prostacy-
clin on spinal cord ischemia: an experimental study. Surgery 1993;
1:36.

80. Tymianski M, Wallace MC, Spiegelman I et ah Cell-permeant Ca 2+
chelators reduce early excitotoxic and ischemic neuronal injury in
vitro and in vivo. Neuron 1993; 2:221.

81. Seibel PS, Theodore P, Kron IL, Tribble CG. Regional adenosine
attenuates postischemic spinal cord injury. / Vase Surg 1993; 2:
153.

82. Qayumi AK, Janusz MT, Jamieson WR, Lyster DM. Pharmacol-
ogic interventions for prevention of spinal cord injury caused by
aortic crossclamping. / Thorac Cardiovasc Surg 1992; 2:256.

83. Mohan RE, Ascer E, Marini CP et ah Does iliprost mediate throm-
boxane activity and polymorphonuclear leukocyte sequestration
in ischemic skeletal muscle? / Cardiovasc Surg 1992; 33:613.

84. Breckwoldt WL, Genco CM, Connolly RJ et ah Spinal cord protec-
tion during aortic occlusion: efficacy of intrathecal tetracaine. Ann
Thorac Surg 1991; 5:959.

85. Agee JM, Flanagan T, Blackbourne LH et ah Reducing post-
ischemic paraplegia using conjugated superoxidase dismutase.
Ann Thorac Surg 1991; 6:911.

86. Hemmila MR, Zelenock GB, D'Alecy LG. Postischemic hyper-
glycemia worsens neurologic outcome after spinal cord ischemia.
J Vase Surgl993;4::661.

266

CHAPTER25 Pathophysiology of spinal cord ischemia

87. Svensson LG, Grum DF, Bednarski M et ah Appraisal of cere- 88. Safi HJ, Bartoli S, Hess KP et ah Neurologic deficit in high-risk

brospinal fluid alterations during aortic surgery with intrathecal patients with thoracoabdominal aortic aneurysms: the role of

papaverine administration and cerebrospinal fluid drainage. / cerebral-spinal fluid drainage and distal aortic perfusion. / Vase

Vase Surg 1990; 3:423. Surg 1994; 20:434.

267

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Vascular erectile dysfunction:
mechanisms and current approach

Ralph G. DePalma

Impotence has been of interest to vascular surgeons since
Leriche's 1923 observation 1 that erectile dysfunction was
often the first signal of aortoiliac occlusive disease, particu-
larly in young men. In these cases erectile failure was due
to compromised arterial inflow to the corpora cavernosa.
Using techniques to minimize damage to the pelvic nerves and
restoring flow into the internal iliac arteries, potency can be
restored after aortoiliac reconstruction. 2 ' 3 Aortoiliac recon-
struction itself can also cause erectile dysfunction not present
preoperatively due to failure to perfuse the internal iliac arter-
ies or to damage autonomic genital nerves. 4 These cases are of
particular interest to vascular surgeons, while the general
problem of erectile dysfunction as a chief complaint has
proven to be much more complex.

Early in the 1970s it was postulated that erectile dysfunction
was due to arterial insufficiency and that the effect was pro-
gressive with aging. In the 1980s, with observations that
intracavernous injection with vasoactive agents such as pa-
paverine 5 and phentolamine 6 caused erection by relaxing cor-
poral smooth muscle, a new era began in diagnosis and
treatment of this disorder. Subsequently abnormalities in cor-
poral smooth muscle were defined and erectile dysfunction
due to venous leakage was recognized and treated. 7 In addi-
tion to vascular factors, neurogenic, endocrine, and medica-
tions such as antihypertensive agents, tranquilizers, alcohol,
and other drugs contribute to erectile dysfunction. Finally, the
introduction of oral medication, sildnafil, 8 for treatment of
erectile dysfunction in the late 1990s further improved treat-
ment for most men with this complaint. From the standpoint
of vascular surgical practice, in the author's experience, 9
approximately 6-7% of men become candidates for vascular
surgical interventions after failing to respond to risk factor
modifications, treatment with drugs, or other modalities. 10 In
this select cohort responses to vascular interventions appear
most effective.

Hemodynamics of normal erection

Erection begins with relaxation of the smooth muscle of the
cavernosal bodies and dilation of their arterial blood supply.
The increase in arterial flow is neurally mediated, and as arte-
rial flow increases, the veins draining the cavernosal bodies
are compressed against the tunica albuginea, causing almost
complete venous outflow in the erect state (Fig. 26.1). When
the penis is flaccid, corporal smooth muscle is contracted and
arterial inflow and venous outflow are balanced. With onset of
erection intracavernous pressure increases from resting pres-
sures approximating 5-15 mmHg to levels of 80-90 mmHg. 11
With full rigid erection cavernosal flow virtually ceases and
suprasystolic pressures contributing enhanced penile rigidity
are generated by contraction of the perineal muscles.
Table 26.1 summarizes vascular factors contributing to erectile
failure. Most frequent among these in many men with this
complaint is failure of smooth muscle relaxation, and among
these most frequently this is caused by diabetes, which im-
pedes smooth muscle relaxation. 12 In the author's experience
failure to respond to injection of intracavernous agents often
relates to high blood glucose in poorly controlled diabetics.

Neural factors in erectile function

Eckart in 1863 13 stimulated pelvic nerves in dogs, which pro-
duced erection and which he named nervi erigenti. Classically,
erection had been considered a parasympathetic function;
however, sacral elements responsible for enervation of the uri-
nary tract, bladder, colon, rectum, and penis contain both sym-
pathetic and parasympathetic fibers. 14 When these neural
elements are diseased or injured, erectile function is frequently
disturbed. In man the exact courses of preganglionic and
postganglionic fibers responsible for the integrated responses
of tumescence, emission, ejaculation, and detumescence are
incompletely delineated. The human penis contains abundant
adrenergic nerves with catechol neurotransmitters within the

268

chapter 26 Vascular erectile dysfunction: mechanisms and current approaches

Cavernous smooth

muscle contracted

Vein open

Figure 26.1 Current concept of penile
erection. Three stages are characterized by
pressure measurements obtained during
dynamic infusion cavernosometry after
intracavernous injection of a vasoactive agent.
(DePalma RG. New developments in the
diagnosis and treatment of impotence. West J
Med 1996; 164:54.)

Cavernous artery contracted

Cavernous smooth
muscle relaxed
Vein occluded

Cavernous artery open-

Cavernous smooth
muscle relaxed

PENIS

FLACCID

PENDANT

Pressure: 10-15 mmHg

PENIS

ERECT

HORIZONTAL

Pressure: 80-90 mmHg

Cavernous flow ceases-

PENIS

ERECT

UPWARD

Pressure >120 mmHg

with perineal/muscle contraction

Table 26.1 Factors in vasculogenic impotence

Factor

Probable etiology

Cavernosal
Refractory smooth muscle

Arteriolar
Fibrosis
Peyronie's disease

Arterial
Aortoiliac occlusion or aneurysm
Atheroembolism
Pudendal artery occlusion
Penile artery occlusion

Venous leakage
Acquired
Congenital

Hormonal; metabolic: diabetes
Blood pressure medication
Functional or anatomic; medications
Postpriapic; drug injection
Invasion of corpora; venous leakage

Atherosclerosis
Atherosclerosis

Atherosclerosis; trauma; idiopathic
Idiopathic proliferative
Atherosclerosis; trauma

Trauma; tunica lesions
Developmental leakage from corpora
into spongiosum

corpora and blood vessels; these are probably responsible for
maintenance of a flaccid state due to smooth muscle contrac-
tion. Smooth muscle relaxation, as noted, is associated with
the erect state. Little adrenergic activity exists in the glans and

spongiosum. Cholinergic nerves are also present within the
corporal bodies. 15 In addition to adrenergic and cholinergic
control of smooth muscle function, nonadrenergic, noncholin-
ergic (NANC) systems with neuronal nitric oxide (NO) and
vipergic systems also exist. In addition endothelial factors in-
cluding NO, endothelin, and prostglandins also contribute to
smooth muscle relaxation. 16 The process is complex and local
mediators are multiple.

Similarly the central and peripheral neural pathways are
complementary. Centrally mediated psychogenic erections
occur in most men. Such erections are integrated with im-
pulses in an erection center in the thorocolumbar spinal cord.
Central excitatory or inhibitory centers have been localized in
the cortex, limbic system, medial preoptic area, paraventricu-
lar nucleus of the hypothalamus and the hippocampus, which
acts in concert with the former two areas. 15 Reflexogenic erec-
tions are mediated by a sacral spinal center responding to di-
rect stimulation or enteroceptive stimuli from the bladder and
rectum. During normal erection these two pathways interact,
but in certain cases of spinal cord injury, where reflex stimuli
are not perceived, erection can occur by purely psychogenic
stimulation. For example, experimentally, after excision of the
feline male lumbosacral spinal cord, 17 exposure to a female cat
in estrous caused erection, while direct genital stimuli failed to
elicit this response. Both a second transection in the lower

269

pa rt I Vascular pathology and physiology

Greater

sciatic
notch

Gluteal (superior) artery

Internal iliac artery

Ischial
spine

Ischial (inferior-
gluteal) artery

Exit of

Alcock's

canal

Dorsal penile artery

Accessory deep branch
(variable)

Deep cavernosa! artery

Penile buJb and artery

Branch of superficial perineal artery
and scrotal arteries

Figure 26.2 Right oblique schematic of the
internal iliacartery, pudendal arteryand penile
branches. This view is preferred for highly
selective pudendal arteriography. Note
landmarks. (DePalma RG. New developments in
the diagnosis and treatment of impotence.
West J Med 1996; 164:54.)

spinal cord and resection of the inferior mesenteric ganglion
and hypogastric nerves in this species then abolished erection.

Psychogenically mediated erections can occur in men with
lesions up to T12; these may be mediated through sympa-
thetic pathways. 16 Bilateral sympathectomy involving LI
produces erectile failure in about 30% of men. 18 Variable
responses to denervation accompanying vascular surgery are
caused by the rich interconnections of the vegetative nervous
system around the aorta and major pelvic arteries. The spinal
nuclei for control of erection in man, according to Lue and col-
leagues, 19 are located in the anteromedial gray matter at S2-S4
and the higher centers at the T10-L2 cord levels. In humans the
inferior mesenteric ganglia and nerve outflow are located
mainly to the left accompanying the mesenteric artery and left
iliac artery. Axons that issue vertically from the sacral neurons
merge with axons of the nuclei for the bladder and rectum
forming the main sacral visceral efferent fibers. These join
sympathetic fibers to constitute the pelvic plexus and ulti-
mately branch out to supply the bladder, rectum, and penis.
The nerve fibers innervating the penis, i.e. the paired caver-
nosal nerves, travel along the posterior aspect of the seminal
vesicles and prostate accompanying the membranous urethra
as it pierces the genitourinary diaphragm. These nerve fibers
are applied closely to the prostate gland. Advances in prostate
surgery have enabled nerve-sparing dissections to allow post-
operative potency following prostatectomy, 20 as can be accom-
plished with vascular interventions for aortoiliac disease 21 by
minimizing dissection of the pelvic and periaortic plexi, par-
ticularly on the left.

The pelvic nerve plexus can be visualized as a rectangular
plate that spreads over the lateral aspect of the rectum to inner-
vate that structure, the bladder, seminal vesicles, and prostate.
Most of these nerves travel with blood vessels and the distal
fibers of the plexus, as mentioned, closely encompass the pos-
terolateral aspect of the prostate and then pass into the penis
innervating the arteries and the erectile tissue of the corpora.
That these are nerves responsible for erection has been demon-

strated by electrostimulation of the caudal bundles of the
plexus. 19

Vascular anatomy

Arterial supply

The anterior divisions of both internal iliac arteries give rise
to the internal pudendal arteries, which exhibit frequent
anatomic variation. The most common arrangement is shown
in Figure 26.2. The pudendal artery leaves the pelvis between
the pyriformis and coccygeus muscles, crosses the ischeal
spine externally, reentering the ischiorectal space through the
lesser sciatic foramen. It courses to the base of the penis pass-
ing along the lateral wall of the ischiorectal fossa, emerging
from Alcock's canal into the perineum about 4 cm anterior to
the lower margin of the ischial tuberosity. In the perineum
these arteries are exposed and vulnerable to injury.

Figure 26.3 illustrates the usual arterial arrangement in the
penis. The deep cavernosal arteries supply blood to the corpo-
ral erectile tissue, while the dorsal penile and urethral arteries
supply mainly the glans and the spongiosum, respectively.
These arteries also vary and the radiologic literature depicts
important variations in detail. 22 Issues for microvascular re-
construction include the existence of branches of the dorsal
artery that pierce the tunica albuginea, enter the cavernosal
spaces, to permit penile revascularization using the dorsal
arteries. Microvascular anastomoses to the deep cavernosal
arteries are impractical; these vessels are usually minute and
difficult to expose.

Venous drainage

Figure 26.4 illustrates the penile veins, which consist of super-
ficial, intermediate, and deep vessels and associated emissary,
circumflex, and communicating veins. Superficial veins course

270

chapter 26 Vascular erectile dysfunction: mechanisms and current approaches

Internal
pudendal

artery

Dorsal

artery

with

circumflex

branches

Accessory

branch

from

dorsal

artery

Bulbourethral
artery

Corpus
spongiosum

Perforating
arteries from
glans

Figure 26.3 Penile arterial supply: note deep
cavernosal and helicine arteries within corpora.

Deep cavernosal artery
within cavernous body
giving off helicene arteries

To saphenous
system and anterior
abdominal wall

Deep Superficial
dorsal dorsal
vein vein

To pudendal
and pelvic
plexi

Circumflex
and lateral

emissary
veins

Superficial

dorsal _

Deep

dorsal

vein

Lateral

emissary

vein

Inferior

emissary

vein

Veins inferior.
to corpus
cavernosum

Figure 26.4 Penile venous drainage: note absence of discrete cavernous vein and interconnections on cross-sectional view.

Skin

Dartos fascia
A — Buck's fascia

Corpus cavernosum

Tunica albuginea

Emissary
vein from
spongiosum

just under the skin and form a superficial vein normally visible
on inspection. Prominence of these veins is usually not relevant
to venogenic impotence. The intermediate veins lie deep to
Buck's fascia and just above the tunica. The most important
from a surgical standpoint is the deep dorsal vein formed by the
confluence of 15 to 16 short straight vessels from the glans and
retrocoronal sulcus. The deep dorsal vein is found in the sulcus
between the two corpora in close relationship to the paired
dorsal arteries and nerves. This vein has a thick muscular coat;
proximally emissary and circumflex veins from the corpora
and the spongiosum join it. At the penile base multiple veins
merge with several trunks entering the pudendal vein or

plexus. Valves exist at this confluence and three or more valves
are located along the penile course of the dorsal vein. All of
these valves are arranged in a conventional fashion to facilitate
blood flow to the heart and prevent distal reflux. The dorsal
vein is an important surgical structure as it may be resected for
venous leakage, be a target for a distal inflow procedure, or a
source of tissue for patching the tunica albuginea.

There are no deep discrete veins in the cavernosal spaces
and one to four large veins exit the proximal extremity of each
crus to provide deep venous drainage. These empty into the
pudendal vein or plexus, while the corpora distal to the arcu-
ate ligament empty into the deep dorsal vein, circumflex ves-

271

pa rt I Vascular pathology and physiology

sels, and veins posterior to the corpora. The functional closure
of the venous system during erection ensues when the smooth
muscle of the corpora relaxes, pressurization occurs, and ve-
nous drainage is occluded by subalbugineal smooth muscle
and the tunica itself. Venous leakage from the corporal bodies
can be due to abnormal congenital connections, following
trauma, or acquired defects in the tunica itself. In well
characterized cases division of offending leakage can correct
venogenic impotence; however, a complicating issue is that
arterial insufficiency itself is associated with venous leakage.

Approaches to diagnosis and treatment

In the past a variety of invasive diagnostic methods and diag-
nosis and treatment including intracavernous injection of
vasoactive agents have been used. Currently, for patients
presenting with the chief complaint of erectile dysfunction,
treatment is begun with medical measures including risk fac-
tor reduction and a comprehensive history and physical exam-
ination searching for the likely etiology of erectile dysfunction.
With availability of specific pharmocotherapy, paradigm
shifts in diagnosis and treatment have occurred so that, absent
contraindications, sildenaphil, a specific inhibitor of CGMP-
specific phosphodiestrase type 5, is employed as a first step.
Currently other drugs acting in a similar fashion have been ap-
proved including vardenaphil and tadalafil. With titration this
drug may be effective in up to 60% of men with erectile dys-
function. 23 At the same time treatment of erectile dysfunction
has become a sophisticated subspecialty. Our colleagues in
urology have contributed to enormous scientific progress in-
cluding strategies of gene therapy 24 with constructs of cDNA
that induce nitric oxide synthetase activity in cavernosal
smooth muscle and possibly in the vascular system generally.
Many men have used cavernosal self-injection with pro-
staglandin E 1 (PGE 1 ) and other mixtures, and vacuum devices
have also been used. These are important advances for the
many men, generally younger than the atherosclerotic popu-
lation treated by vascular surgeons. In atherosclerotics, impo-
tence may accompany lower extremity symptoms or, less
commonly, is a presenting complaint for large vessel disease
including aneurysms. 25 The vascular surgeon will more likely
encounter patients who do not initially complain of erectile
dysfunction. However, the results of large vessel reconstruc-
tion (as opposed to microvascular procedures) can be reward-
ing and durable in this group of men and even in some women.
The sections that follow outline diagnostic and vascular inter-
ventions of interest to vascular surgeons.

Diagnostic methods

Penile plethysmography and measurements of penile brachial
indices are a useful noninvasive method that can be employed
in most vascular laboratories. 26,27 A penile brachial index less

than 0.6 and flat pulse waves may indicate proximal large ves-
sel obstruction. Waveforms are recorded on a polygraph at a
chart speed of 25 mm/s and a sensitivity setting of 1. A cuff
with a self-contained transducer is inflated to diastolic plus
one-third of pulse pressure for recording waves, which when
normal resemble those found in the normal lower extremity.
The sensitivity and specificity of this method in predicting an
abnormal arteriogram were 85% and 70%, respectively. 27

Color duplex Doppler ultrasound provides a detailed view
of penile vascular and corporal anatomy when performed at
intervals after intracavernous injection of a vasoactive agent
such as PGE 1 . 28 This study offers objective flow data and
anatomic details as well as information about the rigidity of
erectile response to injection, information about Peyronie's
disease, and provides a basis for further vascular investigation
of patients who fail to respond to therapy. A review of penile
vascular evaluation illustrating normal and abnormal find-
ings has been provided by Sanchez-Ortiz and Broderick. 29
Other research-oriented methods include magnetic resonance
imaging and radionuclide phallography.

Dynamic infusion cavernosometry and cavernosography
are used to detect and document sites of venous leakage after
injection of a vasoactive agent to dilate arterial inflow maxi-
mally. These can provide information about cavernosal artery
occlusion pressure (CAOP), which reflects inflow. Methods
used by my associates 27 were modified after those described
by Goldstein. 30 The procedure is carried out using two fine
needles inserted into the corpora, one for infusion of warm he-
parinized saline, the other for pressure monitoring. With a
steady state of pressure intracavernous at 100 mmHg, flows to
maintain and to induce erection are measured. A flow rate of
greater than 45 ml/min exceeds the ability of a normal arterial
system to compensate venous leakage, and a rate of fall in
pressure of greater than 1 mmHg/s suggests venous leakage.
CAOP is recorded as the pressure at which the Doppler signal
in the corporal arteries disappears as full erection is induced. A
CAOP gradient of 30 mmHg less than brachial artery pressure
indicates compromised arterial inflow and is an indication for
arteriography in selected candidates.

Arteriography is needed in aortoiliac occlusive disease to
plan reconstruction, and for planning endovascular proce-
dures to treat aneurysmal disease. In cases requiring large ves-
sel reconstruction, special attention to the status of the internal
iliac arteries is important; however, in the presence of athero-
matous debris highly selective internal iliac arteriography
should not be attempted due to the risk of atheroembolism. Se-
lective pudendal arteriography employs nonionic contrast
material after intracorporal injection of a vasoactive agent in-
tended to induce partial tumescence. Optimal visualization of
the distal penile branches is needed for distal reconstruction,
particularly for younger patients after traumatic perineal or
pelvic injuries.

Neurologic testing is now much less commonly employed.
The author and associates 31 and others 32 used pudendal-

272

chapter 26 Vascular erectile dysfunction: mechanisms and current approaches

evoked potentials and measurements of bulbocavernous re-
flex times to assess neurologic contributions to erectile dys-
function. About 30% of patients with this complaint were
found to exhibit abnormalities, and when detected, these were
felt to contraindicate microvascular vascular reconstructions
for impotence. 9 However, since many patients with neurolo-
gic abnormalities respond to pharmocotherapy, this testing
is now less frequently used.

Vascular interventions

Surgical interventions in the aortoiliac segment to prevent
erectile failure or restore normal function are relatively
straightforward. These are: (i) preservation of nerve fibers in
the periaortic and pelvic fields, and (ii) perfusion of the inter-
nal iliac arteries. For open aneurysm repair the author prefers
an inlay technique with entry into the sack well to the right
avoiding the mesenteric leaf, the inferior mesenteric artery,
and the pelvic neural plate. The internal iliac artery can be
closed by suture from within the sack. These maneuvers use
minimal dissection and probably help to avoid colon and
spinal ischemia. I cannot recall ever having to reimplant an in-
ferior mesenteric artery. In a personal series followed up to
3 years until 1990, four of 125 men operated on suffered erectile
dysfunction, two after ruptured aneurysms and two with in-
ternal iliac aneurysms requiring ligation. Fifty-three men with
average age of 65 years were impotent preoperatively and
postoperatively, 30 men average age 57 years were potent pre-
operatively and remained so postoperatively and 39 men av-
erage age 58 years, with erectile dysfunction preoperatively,
regained function postoperatively. Overall it appeared possi-
ble to restore or maintain erectile function in about half of the
men with open surgery, though age appears to be a factor in
postoperative potency. For occlusive disease a right-sided
nerve-sparing approach is also useful in preserving erectile
and ejaculatory function; 2 van Schaik et al. 33 from Leiden have
provided a useful recent anatomic review of the periaortic
nerve plexi.

Femorof emoral bypass is also a useful procedure and can be
combined with transluminal dilation of a donor iliac artery. 34
Transluminal dilation of the external iliac can improve penile
perfusion by relieving steal through the internal iliac and
gluteal arteries. In the author's experience the primary focus
of endovascular interventions of erectile dysfunction has been
in the common or external iliac arteries, 35 and others have de-
scribed selective dilation on the internal iliac arteries. 36 Proce-
dures attempting endovascular interventions below this level,
i.e. in the pudendal arteries, have usually failed. 37

Other open techniques of aortic surgery, mainly the ex-
traperitoneal exposure aorta, have not been investigated by
the author, though the ability to dissect the nerve bundle from
the aortic bifurcation and left iliac artery using this approach
seems clear. Division of the inferior mesenteric artery is re-

quired. The extraperitoneal approach to the internal iliac
artery is quite useful for endarterectomy. The use of endovas-
cular grafts for aneurysms may compromise the orifice inter-
nal iliac artery to achieve a landing site or actually require
occlusion of this vessel. The proposed Veterans Administ
ration prospective randomized trial of open vs. endovascular
repair (OVER) offers to provide insight into endovascular
techniques as related to erectile function which will be as-
sessed pre- and postoperatively.

Microvascular surgery has been employed much less fre-
quently since effective oral agents and other therapies became
readily available. Experience with microvascular bypass into
the dorsal artery and dorsal vein arterialization have been
practiced with results that vary widely according to different
authors with varying follow-up intervals. The results of large
vessel reconstruction for erectile dysfunction are better than
those using microvascular interventions, even though the lat-
ter are required by younger individuals. These procedures
have been performed only in individuals who completely fail
medical and intracavernous therapy and who choose this pro-
cedure over a prosthetic device. 9 ' 10 The results are that about
27-30% functioning after 3 years of follow-up can erect spon-
taneously, and if intracavernous injection is offered up to 70%
can achieve intromission. Jarrow 38 has summarized varying
results and limitations of these operations, as well as for ve-
nous interruption, which, with careful patient selection, has
been effective in my experience with an overall supplemental
function rate of 70% using injection and /or medical treatment.

Summary

The vascular surgeon in practice will be familiar with tech-
niques to minimize or prevent erectile dysfunction associated
with aortoiliac disease. It is important to obtain a careful his-
tory and to ascertain whether or not this might be a problem
for the patient postoperatively. It should be noted that, in spite
of best efforts, erectile dysfunction can become a postopera-
tive problem and the surgeon will to deal sensitively with this
distressing complaint. The treatment of the primary com-
plaint of erectile dysfunction has become, in the main, the
province of capable colleagues in urology who have made re-
markable strides in its treatment. Vascular surgeons, in turn,
will need to be aware of the challenges of new approaches to
aortoiliac disease as these influence erectile function.

References

1. Leriche R. Des obliterations arterielle hautes (obliteration de la
termination de l'aorte) comme cause de insufficances circulatoire
des members inferiors (Abstr). Bull Mem Soc Chir 1923; 49:1401.

2. DePalma RG, Levine SB, Feldman S. Preservation of erectile func-
tion after aortoiliac reconstruction. Arch Surg 1978; 113:958.

273

pa rt I Vascular pathology and physiology

3. DePalma RG, Kedia K, Persky L. Surgical options in the correction
of vasculogenic impotence. Vase Surg 1980; 14: 92.

4. May AG, DeWeese JA, Rob CG. Changes in sexual function fol-
lowing operation on the abdominal aorta. Surgery 1969; 65; 41.

5. Virag R. Intracavernous injection of papaverine for erectile failure
(Letter). Lancet 1982; 2:938.

6. Brindley GS. Pilot experiments on the actions of drugs injected
into the human corpus cavernosum penis. Br J Pharmacol 1986;
87:495.

7. DePalma RG, Schwab F, Druy EM, Miller HC, Emsellem HA. Ex-
perience in diagnosis and treatment of impotence caused by cav-
ernosal leak syndrome. / Vase Surg 1989; 10:117.

8. Goldstein I, Lue TF, Padma-Nathan H et ah Oral sildenaphil in the
treatment of erectile dysfunction. / Urol 1998; 338:1397.

9. DePalma RG, Olding M, Yu GW et ah Vascular interventions for
impotence: lessons learned. / Vase Surg 1995; 21 :576.

10. DePalma RG. Vascular surgery for impotence: a review. IntJImpot
i^s 1997; 9:61.

11. DePalma RG. New developments in the diagnosis and treatment
of impotence. West J Med 1996; 164:54.

12. Saenz de Tejada I, Goldstein I, Azadoi K, Krane RJ, Cohen RA. Im-
paired neurogenic and endothelium-mediated relaxation of pe-
nile smooth muscle in diabetic men with impotence. N Engl} Med
1989; 320:1025.

13. Eckart C. Untersuchinger iiber die Erection des Penis beim Hund.
BeitrAnat Physiol 1863; 3:123.

14. Pick J. The Autonomic Nervous System. Philadelphia: Lippincott
1970:439.

15. Rehman J, Melman A. Normal anatomy and physiology. In:
Mulcahy JJ, ed. Male Sexual Function: A Guide to Clinical Manage-
ment. Towata, NJ: Humana Press 2001:1.

16. Anderson KE, Wagner G. Physiology of penile erection. Physiol
Rev 1995; 75:191.

1 7. Root WS, Bard D. The mediation of feline erection through sympa-
thetic pathways. Am} Physiol 1947; 151:180.

18. Whitelaw GP, Smithwick RE Some secondary effects of sympa-
thectomy with particular reference to sexual dysfunctions. N Engl
J Med 1951; 245:121.

19. Lue TF, Zeneh SJ, Schmidt RA, Tanagho E A. Neuroanatomy of pe-
nile erection: its relevance to iatrogenic impotence. / Urol 1984;
131:273.

20. Walsh PC, Donker PJ. Impotence following radical prostatectomy:
insight into etiology and prevention. / Urol 1982; 128:492.

21. DePalma RG. Impotence in vascular disease: relationship to vas-
cular surgery. Br J Surg 1982; 69:514.

22. Ginestie JF, Romieu A. A Radiologic Exploration of Impotence. The
Hague: Martinus Nijoff, 1978.

23. Boolell M, Allen MJ, Ballard SA et ah Sildenaphil: an orally
active type 5 cyclic GMP-specific phosphodiestrase inhibitor for

the treatment of penile erectile dysfunction. Int } Impot Res 1996;
8:47.

24. Gonzolez-Cavidad NF, Ignarro LJ, Rajfer J. Gene therapy for erec-
tile dysfunction. In: Mulcahy JJ, ed. Male Sexual Function: A Guide
to Clinical Management. Towata, NJ: Humana Press 2001:371.

25. DePalma RG, Massarin E. Occult aortoiliac disease in men with
primary complaint of erectile dysfunction. Int J Impot Res 1994;
6(Suppl.I):A10.

26. DePalma RG, Emsellem HA, Edwards CM et ah A screening se-
quence for vasculogenic impotence. / Vase Surg 1987; 5:228.

27. DePalma RG, Schwab FJ, Emsellem HA, Massarin E, Bergsrud D.
Noninvasive assessment of impotence. In: Pearce WH, Yao JST,
eds. The Surgical Clinics of North America. Philadelphia: WB
Saunders Co., 1990; 70:119.

28. Lue TF, Hricak H, Marich KW, Tanagho EA. Vasculogenic impo-
tence evaluated by high-resolution ultrasonography and pulsed
Doppler spectrum analysis. Radiology 1985; 155:777.

29. Sanchez-Ortiz RF, Broderick GA. Vascular evaluation of erectile
dysfunction. In: Mulcahy JJ, ed. Male Sexual Function: A Guide to
Clinical Management. Towata, NJ: Humana Press 2001:167.

30. Goldstein I. Vasculogenic impotence, its diagnosis and treatment.
In: White dV, ed. Problems in Urology— Sexual Dysfunction.
Philadelphia: Lippincott, 1987:547.

31. Emsellem H, Bergsrud DW, DePalma RG, Edwards CM. Puden-
dal evoked potentials in the evaluation of impotence. / Clin Neuro-
physiol 1988; 5:359.

32. Fabra M, Porst H. Bulbocavernous-reflex latencies and pudendal
nerve SSEP compared to penile vascular testing in 609 patients
with erectile failure and other sexual dysfunctions. IntJImpot Res
1999; 11:167.

33. van Schaik J, van Baalan JM, Visser MJT, De Ruiter MC. Nerve-
preserving aortoiliac surgery: anatomical study and surgical
approach. / Vase Surg 2001; 33:983.

34. Merchant RF Jr, DePalma RG. The effects of femoro-femoral grafts
on postoperative sexual function: correlation with penile pulse
volume recordings. Surgery 1981; 90:962.

35. DePalma RG. Iliac artery occlusive disease: impotence and colon
ischemia. In: Moore WS, Ahn SS, eds. Endovascular Surgery, 3rd
edn. Philadelphia: WB Saunders Co., 2001:355.

36. Urigo F, Pischedda A, Maiore M et ah The role of arteriography and
percutaneous transluminal angioplasty in the treatment of arteri-
ogenic impotence. (Italian) Radiol Med (Torino) 1994; 88:80.

37. Valji K, Bookstein JJ. Transluminal angioplasty in the treatment
of arteriogenic impotence. Cardiovasc Intervent Radiol 1988; 11:
245.

38. Jarrow JP Vascular surgery for erectile dysfunction. In: Mulcahy
JJ, ed. Male Sexual Function: A Guide to Clinical Management.
Towata, NJ: Humana Press 2001:293.

274

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Portal hypertension: pathophysiology and
clinical correlates

David Rigberg
Hugh A. Gelabert

As care for patients suffering end-stage liver disease is pro-
gressively directed to specialized liver and transplantation
units, the knowledge base and approach has evolved to a high-
ly refined set of protocols. The fundamental aspects of portal
hypertension remain unchanged: the diseases are chronic and
progressive, the presentations may be daunting, and the clini-
cal decision tree is unforgiving: errors made in the haste of the
moment may rapidly escalate into an uncontrollable situation.
For these reasons a well-rehearsed familiarity with the current
management is essential.

Our objective is to present the information required for
competent management of these patients in a coherent and
organized manner. The basic physiologic abnormalities of
portal hypertension are reviewed along with the common
complications. Based on these, a rationale for intervention is
outlined along with algorithms for managing the various
presentations.

Etiology and classification
of portal hypertension

Extrahepatic presinusoidal obstruction

Presinusoidal extrahepatic obstruction is most commonly due
to thrombosis of the portal vein. While less common than other
forms of obstructive portal hypertension, portal vein throm-
bosis in an adult is progressive decrease in portal flow from
progression of intrahepatic sinusoidal obstruction. Other
causes in adults include pancreatitis, hypercoagulable states,
or tumor thrombus and mechanical obstruction of portal
venous flow.

The most common presentation of portal vein thrombosis is
hemorrhage from gastric varices and ascites. In children this is
often seen in association with minor infections such as upper
respiratory infections (URI). The diagnosis of portal vein
thrombosis may be established by Doppler ultrasound, or
the lack of visualization of the portal vein on the venous phase
of a celiac or mesenteric angiogram. Wedged hepatic venous
pressure is usually normal, as is the hepatic synthetic function.
The preservation of hepatic function is important since it
allows patients to tolerate hemorrhagic events with less
decompensation than other portal hypertensive patients.

Definition

Portal hypertension is an abnormal elevation of the portal
blood pressure which results most commonly from an
obstruction to the flow of blood in the portal system. Less
commonly, it may result from an increase in the portal
blood flow. The obstructions to the flow of blood result from a
variety of pathological entities, and have been classified
according to the location of the obstructive elements relative
to the portal sinusoid (Table 27.1). Thus a given disease maybe
said to cause sinusoidal or presinusoidal portal hypertension.
The presinusoidal and postsinusoidal varieties of portal
hypertension have been further subclassified as intra- or
extrahepatic.

Intrahepatic presinusoidal obstruction

Most causes of intrahepatic presinusoidal obstructive portal
hypertension relate to fibrosis and compression of the portal
venuoles with subsequent restriction of portal flow. Included
amongst these diseases are congenital hepatic fibrosis,
sarcoidosis, chronic arsenic exposure, Wilson's disease,
hepatoportal sclerosis, primary biliary cirrhosis, schistosomi-
asis, and myeloproliferative disorders.

Schistosomiasis is the most common cause of portal hyper-
tension in the world. It is particularly common in Third- World
countries. Deposition of ova in the portal venule walls results
in a granulomatous inflammatory reaction which in turn
results in fibrosis and portal flow restriction. Hepatic function
is preserved in early stages, but later stages of this disease are
characterized by advanced cirrhosis and loss of hepatic func-
tion. 1 Myelosclerosis and myeloid leukemia will occasionally

275

pa rt I Vascular pathology and physiology

Table 27.1 Etiology and classification of portal hypertension

Class

Characteristics

Etiology

Hemodynamics

Angiographic

Increased resistance
Presinusoidal extrahepatic

Presinusoidal intrahepatic

Sinusoidal/postsinusoidal
(intrahepatic) obstruction

Postsinusoidal extrahepatic
obstruction

Increased blood flow
Arteriovenous fistula

Portal vein thrombosis

Schistosomiasis
Sarcoidosis

Cirrhosis
Hemochromatosis

Budd-Chiari

Malignancies

Trauma

Pregnancy

CHF

Constrictive pericarditis

Hepatic or splenic trauma

t whvp, t ppv
Iwhpv,Tppv

T whvp, T ppv

WHVP unobtainable,? PPV

T WHVP, T PPV

PVopen
PVopen

PVopen,
Collateralization

Hepatic vein occlusion

PVopen,

Collateralization

PVopen

CHF, congestive heart failure.

lead to presinusoidal hypertension by virtue of the deposition
of primitive cellular material infiltrating the portal zones. 2
Sarcoidosis causes portal hypertension by two mechanisms:
first, granulomas within the portal vein lead to obstruction,
and second, there is increase in portal blood flow. Hemody-
namic characteristics are similar to those of extrahepatic portal
vein obstruction: low hepatic wedge pressure and elevated
portal venous pressure.

Intrahepatic sinusoidal and
postsinusoidal obstruction

Sinusoidal portal hypertension may be the sequela of alcoholic
hepatitis or hepatitis. Pure sinusoidal obstruction is a rare
cause of portal hypertension and is caused by the same
hepatic insults.

Sinusoidal obstruction frequently is present as part of a
combined sinusoidal and postsinusoidal obstructive picture.
Postsinusoidal obstruction is seen most commonly in cases of
alcoholic liver disease, postnecrotic cirrhosis, or hemochro-
matosis. As would be expected in these diseases, the hepatic
function is usually significantly impaired.

In the United States, combined sinusoidal and postsinu-
soidal obstructive disease is the most common type of portal
hypertension, and is estimated to be the tenth leading cause of
death. Two mechanisms account for the portal hypertension in
these patients. First is the mechanical obstruction of the portal
blood flow by regenerating hepatic nodules and cirrhous
bands. These lesions may extend beyond the confines of the
hepatic sinusoids, and account for the presinusoidal, sinu-
soidal, and postsinusoidal distortion of the hepatic architec-
ture. The second element is increase in splanchnic perfusion,

partly due to the multiple arteriovenous shunts and collateral
channels which develop within the liver. One-third of portal
blood flow may bypass functional hepatocytes through these
channels. 3 The clinical correlate of this increased blood flow
is the hyperdynamic state which typifies cirrhotics: elevated
cardiac output and a diminished systemic resistance. 4

The portal hemodynamic characteristics of these diseases
are characterized by elevated hepatic wedge pressure and ele-
vated portal vein pressure. Since most of these diseases affect
hepatocytes, hepatic function is frequently impaired even in
the early stages of these diseases. These patients will fre-
quently have poor hepatic reserve, and will decompensate
with each bleeding episode.

Extrahepatic postsinusoidal obstruction

Postsinusoidal hepatic vein obstruction is usually caused by
thrombosis in the hepatic veins. While the underlying cause of
most cases is unknown, a series of associated diseases have
been identified. Membranous webs of the hepatic veins,
malignancies (hepatomas, renal carcinomas, adrenal carcino-
mas), trauma, pregnancy, contraceptives, acute alcoholic he-
patitis, and senecio toxicity may all result in hepatic vein
thrombosis. 5 Constrictive pericarditis and chronic congestive
heart failure may also cause functional postsinusoidal
obstruction.

Budd-Chiari syndrome is the result of hepatic venous
occlusion and is characterized by massive ascites, esophageal
varices, variceal hemorrhage, hepatic failure, and death.
Chiari's disease is due to primary hepatic vein ostial occlusion.
The clinical progression following hepatic vein occlusion may
be fulminant or gradual. Hepatic failure is the result of chronic

276

chapter 27 Portal hypertension: pathophysiology and clinical correlates

congestion and ischemia from impaired hepatic blood flow.
The diagnosis is established by angiographic characteristics. 6 ' 7

The fulminant course is marked by rapid development
of ascites, fatigue, and jaundice. Additionally elevations of
liver enzymes and prothrombin times indicate hepatocellular
damage. Patients who do not improve with anticoagulation
should be considered for either shunting or liver transplanta-
tion. The more gradual presentation may have many similar
features such as ascites and chronic fatigue. Hepatic function
is fairly well preserved in the early stages.

Anticoagulation may allow resolution of the venous throm-
bosis by endogenous lytic systems. Patients whose course is
gradually progressive and who have intact hepatocellular
function should be considered for portal decompression by
surgical shunt. The selection of shunt is dependent on the
patient's anatomy as defined by angiography. When the
Budd-Chiari syndrome leads to deterioration of hepatic func-
tion, liver transplantation should be performed. 8

Arteriovenous fistulas

Arteriovenous (AV) fistulas are a relatively rare cause of portal
hypertension. Most fistulas are the result of hepatic or splenic
trauma. Traumatic AV fistulas may occur as a consequence of
transhepatic biliary manipulations, or as a result of penetrat-
ing trauma. Splenic fistulas may be associated with splenic
artery aneurysms, sarcoidosis, Gaucher 's disease, myeloid
metaplasia, or tropical splenomegaly. The portal hypertension
results from increased flow in the portal circulation. At later
stages fibrosis of the presinusoidal spaces exacerbates the
portal hypertension.

Clinical manifestations
of portal hypertension

Esophageal varices

Esophageal varices develop in about 30% of cirrhotics. Of cir-
rhotics who present with upper gastrointestinal (GI) bleeding,
about 30% will bleed from their varices. The remainder bleed
from erosive gastritis, hypertensive gastritis, ulceration,
mucosal tears, and neoplastic tumors. Between 5 and 15%
of cirrhotic variceal bleeders will experience a massive
hemorrhage which is difficult to control. The mortality of these
patients is in the range of 30-50%.

The pathogenesis of esophageal varices centers around the
development of collateral circulatory pathways for blood exit-
ing the portal circulation. The development of these collaterals
is driven by the difference in pressure between the portal
system and the systemic venous circulation. Several major
collateral networks have been described in cirrhotics: the
coronary-esophageal veins, the umbilical vein, the hemor-
rhoidal veins, and the retroperitoneal veins (veins of Retzius).
Each of these venous systems may develop into significant

Table 27.2 Clinical manifestations of portal hypertension

Pathogenesis

Clinical significance

Varices

Encephalopathy

Ascites

Collateral vessel
development

Increased cerebral uptake
of neutral amino acids

Altered Starling forces

Gastric or esophageal
hemorrhage

Neurologic alterations
from irritability to coma

Respiratory compromise
Bacterial peritonitis

collateral networks. Hemorrhoidal vessels, intestinal varices,
and stomal varices have all been documented as bleeding sites
in cirrhotic patients. The mechanical and chemical irritants
which bathe the gastroesophageal region result in esophagi tis,
attenuation of the mucosal layers, and predispose the varices
to bleed. Increases in the blood pressure within the varices lead
to an elevated chance of hemorrhage (Table 27.2).

Characterization of the severity of portal hypertension on
the basis of corrected sinusoidal pressure has not correlated
with risk of hemorrhage. Factors which do predict the risk
of bleeding include the size of the varices, the Child's class
of the patient, and the presence of erosions on the varices (red
dot signs). 9-13

Encephalopathy

Encephalopathy may have a profoundly disabling effect on
patients. The clinical manifestations of encephalopathy cover
a spectrum from mild inattention to coma. The most com-
monly used system of staging encephalopathy classifies pa-
tients on a scale from stage I through stage IV. The progression
begins with mild personality alterations, occasionally with
asterixis or clonus in stage I. Stage II may be characterized
by drowsiness, sometimes with mild confusion. Stage III is
typified by stupor and obtundation. Coma is the hallmark of
stage IV. Electroencephalography is not specifically diagnos-
tic, characteristically showing only slow-wave activity pri-
marily in the frontal regions. 14

How liver failure leads to coma is not clearly understood.
Several agents have been postulated as encephalopathic: am-
monia, nitrogenous amines, increased false neurotransmitters,
decreased true neurotransmitters, and an increased ratio of
aromatic to branched chain amino acids have been implicated.

Elevated ammonia levels have several significant effects.
First, they cause elevated glucagon levels, which stimulate
gluconeogenesis and produce more ammonia. Additionally,
the gluconeogenesis leads to elevated insulin levels. The ele-
vated insulin promotes catabolism of branched chain amino
acids and leads to increased levels of straight chain amino
acids such as phenylalanine, tyrosine, and methionine. An
elevated ratio of straight chain to branched chain amino acids
drives neutral amino acids past the blood-brain barrier. The

277

pa rt I Vascular pathology and physiology

cerebral uptake of these neutral amino acids is possible be-
cause ammonia stimulates brain glutamine synthesis, allow-
ing rapid equilibration of brain glutamine for straight chain
neutral amino acids. These same neutral amino acids may
act as false neurotransmitters, and are thought to produce
encephalopathy. 15

The treatment of encephalopathy is based on reduction of
the ammonia levels, and supplementation of branched chain
amino acids. Lactulose and neomycin will reduce ammonia
uptake from the gut by altering the intestinal pH, reducing the
number of intestinal bacteria, and reducing intestinal transit of
protein. Other agents such as L-Dopa have been used with
mixed results in improving encephalopathy 16

Ascites

Ascites is a common symptom of portal hypertensive patients.
As many as 80% of these patients may have some degree of as-
cites. The mechanism by which ascites develops is a combina-
tion of hemodynamic, physiologic, and metabolic factors. The
hemodynamics of the portal circulatory system in the face of
cirrhosis is primarily driven by the increased portal venous
pressure. In such a state, the Starling forces will force serum
into the interstitial space. Compounding this problem is the
low oncotic pressure which characterizes many cirrhotics
by virtue of their hypoalbuminemia. Finally, many of these
patients chronically register relatively low effective intravas-
cular volume, which in turn triggers the renal aldosterone,
renin-angiotensin system, and perhaps an additional natri-
uretic hormone. The net effect is a state in which the patients
retain free water and salt, both of which aggravate the ascites.
The cornerstones to the management of ascites are the re-
striction of salt intake and diuretic therapy. In only 5% of cases
can ascites be considered to be intractable, and other means of
addressing the ascites are required.

Diagnosis

The diagnosis of portal hypertension is made by demonstrat-
ing the increased portal venous pressure. In practical terms,
the diagnosis of portal hypertension is accomplished by
identifying signs of the elevation of portal venous pressure
and a history which would support these findings (Table 27.3) .

The physical signs of portal hypertension include
esophageal varices, splenomegaly, ascites, or abdominal wall
collaterals. Of these, ascites, splenomegaly, and abdominal
wall collateralization may be apparent on physical examina-
tion. Endoscopy is the most reliable means of identifying
gastroesophageal varices.

Signs of underlying hepatic disease include spider
angiomata, palmar erythema, gynecomastia, muscle wasting,
loss of pubertal hair growth, and testicular atrophy.
Encephalopathy, asterixis, fetor hepatis, and fatigue may also

Table 27.3 Diagnosis of portal hypertension

Procedure

Role

Liver function tests
Upper Gl endoscopy

Liver biopsy

Duplex scanning

Angiography

Assess severity of liver dysfunction

Identify source of bleeding
Identify and control varices

Identify active hepatitis
Define stage of liver disease

Determine patency of portal vein and direction of
portal venous blood flow

Outline vascular anatomy in preparation for bypass
ofTIPSS

TIPSS, transjugular intrahepatic portosystemic shunt.

be noted in chronic hepatic insufficiency. The presence of liver
disease does not conclusively signify that the patient has
significant portal hypertension, but these are frequently
associated.

History supporting the diagnosis of portal hypertension
is based on identifying diseases which are known to lead to
portal hypertension (alcoholisim, hepatitis, hepatotoxins,
etc.). The duration of these problems is important in substanti-
ating the diagnosis of portal hypertension, since the time
between the onset of these insults and the development of
the hypertension may be as long as 10 or more years.

Angiography and hemodynamic measurements are impor-
tant adjuncts in diagnosing portal hypertension. Angiogra-
phy may reveal both splenomegaly and collateral vessels in
the portal region as well as the gastroesophageal axis. Angio-
graphy will also reveal the direction of portal blood flow
(hepatopetal or hepatofugal).

Hemodynamic definition of the portal circulation consists
of measurement of the wedged hepatic vein pressure
(WHVP). 17 This technique records the pressure in the hepatic
sinusoids in a manner analogous to the Swann-Gantz catheter
measurement of left atrial pressure. Elevations of the hepatic
wedge pressure reflect elevations in the portal venous
pressure. False-negative results occur in patients with pre-
sinusoidal obstruction. Normal hepatic venous pressure is
essentially the same as the right atrial pressure (0-5 mmHg),
portal venous pressure is about 2-6 mmHg higher than the
hepatic venous pressure. A gradient greater than 10 mmHg is
considered abnormal.

Other methods of measuring portal venous pressure have
been developed, but have largely been abandoned because of
increased risk of bleeding. This group includes surgical cann-
ulation of the portal vein, percutaneous transhepatic portal
venous catheterization, transjugular portal vein catheteriza-
tion, and percutaneous splenic pulp pressure measurement. 18

Duplex scanning has been used to measure the direction
and velocity of portal blood flow. This technique may detect

278

chapter 27 Portal hypertension: pathophysiology and clinical correlates

portal vein thrombosis, hepatopetal and hepatofugal flow,
and support the diagnosis of portal hypertension. 17-21

Laboratory testing

The initial point of departure in evaluating most patients with
portal hypertension is to asses the liver function tests. Specific
attention should be placed on the liver enzymes (SGOT, SGPT,
LHD, alkaline phosphatase), and tests of hepatic synthetic
function (prothrombin time and serum albumin). Informa-
tion from these two sets of tests will identify patients with
acute hepatocellular damage, and those with reduced syn-
thetic function. The presence and degree of abnormalities in
liver function tests will correlate with the outcome of patients:
the more abnormal the tests, the worse the prognosis. 22,23

The Child's classification, or more recently the Child-Pugh
classification, serve as prognosticators of survival in cirrhotic
patients who undergo both emergent and elective surgery.
The criteria expressed in this classification represent the
collation of historical information, physical examination, and
laboratory testing. 24

Additional laboratory investigation should include a deter-
mination of serum ammonia and complete blood count (CBC:
WBC, RBC, and platelets) . Serum ammonia may be elevated in
instances of severe hepatic dysfunction and coma. It correlates
loosely with mentation, but is more important as an indicator
of a treatable cause of encephalopathy: hyperamonemia. 25 The
CBC may identify both anemia and hypersplenism.

While splenomegaly is present in virtually all portal hyper-
tensive patients, hypersplenism may not develop until later in
the course of their disease. The size of the spleen does not cor-
relate directly with either the degree of portal hypertension or
the severity of hypersplenism. 26 Hypersplenism is character-
ized by splenic sequestration and destruction of platelets and
WBCs leading to significant depressions in the circulating
level of both. A platelet count below 50 000 and WBC count
below 2000 support this diagnosis.

Upper Gl endoscopy

Endoscopy plays a pivotal role in management of portal
hypertensive patients. Both for diagnostic and therapeutic
reasons, endoscopy should be one of the first tests performed.
Endoscopy will not only define the presence of varices, but
also identify the source of bleeding in patients who present
with hemorrhage.

The diagnosis of portal hypertension may be established by
noting the presence of varices. The size, appearance, and loca-
tion of the varices are important points to be noted. Endoscopy
may identify other frequent sources of bleeding in portal hy-
pertensive patients such as hypertensive gastropathy, gastri-
tis, gastric ulceration, duodenal ulceration, gastric mucosal
lacerations (Mallory Weiss tears), or esophageal ulcerations.
Because of the variety of possible bleeding lesions and

significant difference in the management of these lesions, pa-
tients admitted for hemorrhage must undergo upper en-
doscopy on each admission. As many as 40-60% of patients
with documented varices have associated gastritis or peptic
ulcer disease causing an acute hemorrhage. 27 In patients with
both esophageal and gastric varices, the gastric varices may be
the cause of bleeding in as many as 18%. 28

Liver biopsy

The role of liver biopsy in the preoperative evaluation of portal
hypertensive patients has been debated. The goal of liver bio-
psy in this setting is to identify those patients who have active
hepatitis. In alcoholic patients this is established by the pres-
ence of Mallory bodies which signify acute hyaline necrosis.
Mallory bodies may also be seen in patients with Wilson's dis-
ease, cholestasis, and primary biliary cirrhosis. Patients with
acute hepatic necrosis are at increased risk of dying in the
course of shunt surgery. Mikkelsen and associates noted oper-
ative mortality of 69% in elective shunt cases and 83% in emer-
gent cases in the presence of acute hyaline necrosis. 28,29 Other
authors have contested the point of whether acute alcoholic
hepatitis alters survival. 22 ' 30 Since Mallory bodies will disap-
pear as patients abstain from alcohol the liver recovers from its
insult; these may also signify active alcohol ingestion. 31

Patients suspected of having acute hepatitis (based on
elevated liver enzymes) and who are candidates for elective
shunt surgery should undergo percutaneous liver biopsy. If
Mallory bodies are identified, then consideration should be
given to postponing the elective operation for a period of time
to allow the liver to recover.

Duplex scanning

Duplex scanning is finding increased application in the evalu-
ation of portal hypertensive patients. Duplex scanning will
determine both the patency of the portal vein and the direction
of portal venous blood flow. This is the minimal anatomical
information required to proceed with portacaval shunting or
liver transplantation. Color flow imaging, a recent advance
in duplex scanning, has improved accuracy and diagnostic
abilities of the duplex scanners. 19

Angiography

Angiography should be performed on all patients who are
to undergo elective shunting procedures. The importance of
angiography is that is allows an accurate anatomical record
which may be used in planning surgery. Techniques primarily
of historical interest include splenoportography 32 (percuta-
neous needle into the spleen), umbilical vein catheterization,
and transhepatic percutaneous portal venography 29-38

Current portal angiography is performed by selective
cannulation of the celiac and superior mesenteric arteries, and

279

pa rt I Vascular pathology and physiology

observation of the venous phase of these angiograms. Addi-
tional studies which should be obtained include an injection of
the renal veins, and a hepatic wedge angiogram, and pressure
recording. The combination of these studies is commonly
referred to as a "liver package".

The goal of these studies is to delineate the major portal trib-
utaries: the splenic vein, the superior mesenteric vein, the por-
tal vein itself, and their relation to the renal vein. An additional
goal of the "liver package" study is to measure the hepatic
wedge pressure and visualize the hepatic sinusoidal circula-
tion. Low hepatic wedge pressure (less than 10-12 mmHg) in
a patient with variceal hemorrhage should prompt a careful
search for evidence of portal vein thrombosis. 39 The wedged
hepatic vein catheter can be used to produce an image of
sinusoid architecture. Wedge hepatic venogram in cirrhotics
demonstrates irregular sinusoids with multiple scattered
filling defects. Retrograde portal vein filling indicates
hepatofugal flow. 32

Angiographic findings correlate with the degree of cirrho-
sis. In early cirrhosis, no definite angiographic abnormalities
are present. As cirrhosis becomes more severe one sees the
development of collateral pathways, dilatation of the hepatic
artery, and pruning of intrahepatic portal vein branches. In
advanced cirrhosis, reversal of flow in the portal vein may be
detected.

Table 27.4 Treatment of portal hypertension

Treatment of portal hypertension

Non-operative management of portal
hypertension: variceal sclerosis

Variceal sclerosis or sclerotherapy is the technique by which
esophageal varices are obliterated as the consequence of injec-
tion with inflammatory and thrombogenic solutions. The
objective of this treatment is to ablate the esophageal varices
and by doing so prevent their bleeding (Table 27.4).

While esophageal variceal sclerosis was first introduced by
Crafoord and Frenckner in 1939, their technique using rigid
esophagoscopy was cumbersome and unappealing. 40 In addi-
tion to its technical difficulty, early interest in sclerotherapy
was eclipsed by the emergence of surgical procedures
designed to reduce portal pressure. Thus it was not until the
emergence of flexible fiberoptic esophagoscopy and the iden-
tification of the significant limitations of portal shunt proce-
dures that sclerotherapy resurged in its appeal. In the course of
the past two decades, sclerotherapy has gained a more promi-
nent role in the care of portal hypertensive patients.

While initial evaluations asked the questions of efficacy in
controlling hemorrhage, more recent studies have addressed
the issue of the optimal role of therapy in improving survival.
Additional interest has focused on the role of sclerotherapy in
controlling the various presentations of portal hypertensive
patients: emergency sclerotherapy for control of acute hemor-

Type

Indication

Limitation

Variceal sclerosis

Bleeding varices

Portacaval shunt

Mesocaval shunt

Distal splenorenal
shunt

TIPSS

Livertransplant

Bleeding varices
Emergency ascites
Budd-Chiari syndrome

Bleeding varices

Ascites

Hepatorenal syndrome

Bridge to transplantation

End-stage liver failure

Ineffective in treating
gastric varices and
gastritis

Potentially high
mortality

Mesenteric vein
thrombosis

Intractable ascites
Anatomical restrictions

Must be transplant
candidate

TIPSS, transjugular intrahepatic portosystemic shunt.

rhage, chronic sclerotherapy for suppression of varices follow-
ing a hemorrhagic episode, and prophylactic sclerotherapy for
suppression of varices prior to episodes of bleeding.

The efficacy of sclerotherapy was addressed in a review by
Johnston and Rodgers in 1973. 41 In 117 patients with 194 ad-
missions for acute variceal hemorrhage, bleeding was initially
controlled in 92% and the hospital mortality per admission
was 18%. The average time to recurrence of variceal hemor-
rhage was 10 months. Similar mortality and variceal hemor-
rhage control rates were reported by Terblanche and
colleagues from South Africa. 42-45

Sclerotherapy for control of acute hemorrhage has been
advocated as the treatment of choice following failure of initial
treatment with vasopressin or balloon tamponade. 42 ' 43 Several
controlled trials have compared sclerotherapy with medical
management with the Sengstaken-Blakemore tube in the
acute setting. 46-49 Three studies demonstrated a significantly
lower early rebleed rate. 46-48 Long-term results were not as
good. Only Paquet and Feussner demonstrated significantly
improved overall survival with sclerotherapy 46

The use of sclerotherapy to prevent recurrent variceal hem-
orrhage has been examined in several controlled trials. 49-53
When sclerotherapy was performed repeatedly to eradicate
all varices and compared with conservative medical manage-
ment, sclerotherapy patients had fewer recurrent bleeds and
improved long-term survival. 50 ' 51 ' 53 Terblanche et al. were able
to eradicate varices effectively in 95% of sclerotherapy
patients; however, they could not demonstrate a significant
difference in survival. They also noted that varices recurred
in greater than 60% of the sclerosed patients. 51

280

chapter 27 Portal hypertension: pathophysiology and clinical correlates

Sclerotherapy has been compared with portosystemic
shunts in the management of variceal hemorrhage. Cello et al.
studied these treatments on Child's class C patients and
showed an increased incidence of rebleeding in the sclero-
therapy group, with 40% of the sclerotherapy patients
ultimately requiring surgical therapy They were unable to
demonstrate any significant difference in survival and con-
cluded that in high-risk patients, sclerotherapy and portacav-
al shunting are equal in the acute setting. Surgery should be
considered if variceal obliteration is not successful. 54,55

The distal splenorenal shunt (DSRS) has been compared
with sclerotherapy for the long-term management of variceal
bleeds in three controlled trials. 46 Warren and coworkers
showed that there is a higher rebleeding rate with sclero-
therapy than with shunting; furthermore, one-third of scle-
rotherapy patients required surgery 56 Treatment with
sclerotherapy allowed significant improvement in liver func-
tion when successful, with less encephalopathy and improved
survival when backed up by surgical therapy for patients with
uncontrolled bleeding. Neither Teres et al. nor Rikkers et al.
showed a difference in early and long-term mortality between
DSRS and sclerotherapy 57 ' 58 The rebleeding rate was greater
in those patients who had sclerotherapy and encephalopathy
rates were higher in shunted patients in Teres et al.'s study 57

Staple transection has been compared with sclerotherapy for
emergency control of variceal bleeding. Burroughs and col-
leagues found no difference in overall mortality and improved
control of bleeding with esophageal transection when compared
with a single injection. Bleeding rates were similar, however,
following three injection treatments. 59 Teres and coworkers
randomized cirrhotic patients with uncontrolled bleeding to
staple transection or sclerotherapy in high-risk patients. 60
Sclerotherapy and staple transection had similar rebleed rates
and survival, but fewer complications were observed in the
sclerotherapy group. Although a consensus opinion on scle-
rotherapy has not yet been reached, sclerotherapy may well
represent an appropriate alternative ablative procedure in se-
lected patients with hepatic dysfunction. 46 ' 60-67 ' 109

Several complications have been documented as related to
sclerotherapy. These range from minor problems such as ret-
rosternal pain and mild fever through esophageal ulceration
and perforation. Pleural effusion may occur but does not nec-
essarily indicate esophageal perforation. 68 Variceal ulceration
is not uncommon and resolves spontaneously in most cases. 69
Serious complications such as esophageal perforation, spinal
cord paralysis, and bradyarrhythmias are rare but recognized
complications of this procedure. 33 ' 35 ' 36 ' 38 ' 45 ' 64 ' 70

Endoscopic variceal sclerosis has several distinct advan-
tages. It affords good control of variceal hemorrhage and easy
accessibility for recurrent sclerotherapy. These are accom-
plished rapidly and with minimal morbidity. Additionally, it
allows the opportunity to convert an emergent surgical proce-
dure to an elective one —with attendant improvement in out-
comes. In patients presenting with an acute variceal bleed,

who have had a prior splenectomy or portosystemic shunt,
variceal sclerosis is clearly beneficial. Variceal sclerosis in con-
junction with pitressin is the initial therapy of choice in the
acute management of variceal hemorrhage, especially in
Child's class C patients.

A significant limitation of sclerotherapy is its inability to
control gastric varices effectively. Also, it does not prevent
severe gastritis frequently associated with hemorrhage in
these patients. Finally, recurrent variceal hemorrhage is likely
if follow-up routine sclerotherapy is not performed. Thus
sclerotherapy is an adjunctive therapeutic maneuver, rather
than the definitive treatment of portal hypertension.

Comment should also be made regarding drug infusion for
the treatment of bleeding varices. Recently, somatostatin and
related agents (octreotide, terlipressin) have been evaluated
with regard to their utility in this clinical scenario. Data have
been mixed, but several studies have demonstrated decreased
blood loss when these drugs are used. For example, a meta-
analysis comprising 12 trials (1452 patients) showed an overall
decrease in transfusion requirements of one unit of blood per
patient. 71 Mortality, rebleeding, and need for balloon tampon-
ade were not decreased. As the authors concluded, the trans-
fusion requirements were significant, but it is not evident that
there is that much of a clinical benefit to a single unit of blood
saved.

Terlipressin is a synthetic vasopressin analog which can be
given intermittently as an injection, unlike vasopressin which
must be administered as an intravenous drip. In comparison
with vasopressin, no significant differences in outcome were
shown in a meta-analysis of 20 studies comprising 1609
patients. 72 It is not clear whether this drug will become a front-
line treatment agent. In summary, because these drugs are
easy to administer, have relatively safe profiles, and may be ef-
fective, many centers use them routinely for bleeding varices.

Surgical management of portal hypertension

Shunt nomenclature

Several systems have been devised for the naming of shunts.
Two basic sets of nomenclature prevail: the anatomic-based
descriptive names, and the taxonomic names. The anatomic
naming of shunts is based on the participant elements of the
shunt. Thus the principal shunts are portacaval, mesocaval, or
splenorenal. Because of some ambiguity associated with these
names, modifiers are applied. A portacaval shunt may be
either a side-to-side portacaval shunt or an end-to-side porta-
caval shunt. Similarly a splenorenal shunt may be either a
proximal or a distal splenorenal shunt.

The second set of shunt names is derived from physiologic
and anatomic considerations. These names, which we refer to
as a taxonomic nomenclature, are older, and not as frequently
employed. The two principal sets of names are central/
remote and selective/nonselective. A central shunt is one

281

pa rt I Vascular pathology and physiology

which is constructed in the region of the porta hepatis, or at the
center of the portal confluence. Principal amongst these are the
various portacaval shunts. The term is used to distinguish
those shunts which involve the portal vein itself from those
shunts which are remote from the portal vein such as the
splenorenal and the mesocaval shunts.

Selective and nonselective shunts are the second set of taxo-
nomic names. Selectivity of a shunt refers to the effect of the
shunt on the portal venous blood flow. Selective shunts pre-
serve the flow of mesenteric blood through the portal vein to
the liver. Nonselective shunts drain the portal blood flow in to
the vena cava. Selective shunts such as the distal splenorenal
shunt are thought to preserve hepatic function to a greater ex-
tent than the nonselective shunts. The selective shunts include
the distal splenorenal (Warren) shunt and the coronary-caval
(Inokuchi) shunt. Nonselective shunts include end-to-side
portacaval shunt, side-to-side portacaval shunt, mesocaval
shunt, and proximal splenorenal shunt. Currently, the term
selective is used almost as a synonym for a distal splenorenal
shunt.

The last set of shunt names is the eponyms. Many shunts
have been associated with the name of an avid advocate. In-
cluded amongst these are the Warren shunt (distal splenore-
nal), the Linton shunt (proximal splenorenal), the Clatworthy
shunt (mesocaval shunt using the inferior vena cava), the
Drapanas shunt (mesocaval shunt using a Dacron inter-
position graft), the Inokuchi shunt (coronary-caval), and the
Sarfeh shunt [portacaval polytetrafluoroethylene (PTFE)
interposition graft].

Portacaval shunts

The portacaval shunts divert the portal blood into the inferior
vena cava by anastomosing the portal vein to the infrahepatic
vena cava either in an end-to-side manner or as a side-to-side
anastomosis. Both end-to-side and side-to-side portacaval
shunts are nonselective shunts used for controlling variceal
hemorrhage. Intractable ascites was previously treated
with a side-to-side portacaval shunt; however, this tech-
nique has been superseded in favor of peritoneovenous
shunting. 73 ' 74

The question of whether an end-to-side or side-to-side por-
tacaval shunt is a more effective procedure is still debatable.
Each has advantages and limitations. Budd-Chiari syndrome
is a clear reason for using an end-to-side portacaval shunt, be-
cause the portal vein serves as a decompressive outflow tract
for intrahepatic portal blood. Uncontrolled studies comparing
end-to-side with side-to-side shunts indicate that the side-to-
side shunt is associated with a lower surgical mortality in
patients with poor hepatic functions. 75 On the other hand, the
side-to-side portacaval shunt is more technically demanding.
Encephalopathy has been shown to be somewhat more com-
mon in side-to-side than in end-to-side portacaval shunts. 76
This may be in part due to a siphon effect which permits egress

of hepatic arterial blood through the portal vein rather than the
hepatic vein.

There are certain technical considerations that preclude a
portacaval shunt: portal vein thrombosis, a small diameter
portal vein, and prior surgery in the portal area should be con-
sidered contraindications.

Mesocaval shunts

The Clatwothy shunt consists of a division of the inferior vena
cava at the level of the iliac bifurcation, and the anastomosis of
the proximal segment to the mesenteric vein. It has become the
operation of choice in children with extrahepatic portal vein
thrombosis. Its usefulness in adults is limited because of mas-
sive lower extremity edema. Alternative graft materials have
been used in performing a similar operation: the mesocaval
shunt. This operation places a vascular graft between the vena
cava and the portal vein. Several graft materials have been
used including cadaveric inferior vena cava, iliac vein, PTFE,
and Dacron. 77-82

Two principal concerns have emerged with regards to the
mesocaval shunt: the question of selectivity and the durability
of the shunt. This shunt allows blood to drain from the entire
portal system and thus is not selective. The duration of
patency of this shunt is thought to be less than that of the
portacaval or DSRS.

The primary use of the mesocaval shunt has been for the ur-
gent control of massive variceal hemorrhage. 83 If a selective
shunt is contraindicated, a mesocaval shunt may be a good
alternative, especially in patients with significant ascites.
Other factors which favor mesocaval shunting include prior
surgery in the right upper quadrant, extensive periportal fi-
brosis, a large overriding caudate lobe, an obliterated portal
vein, extreme obesity, and Budd-Chiari syndrome. 84

Distal splenorenal shunts

The total diversion of portal blood away from the liver may
hasten the patient's hepatic failure and postshunt en-
cephalopathy. It is thought that this is mediated by the loss of
hepatotrophic substances found within portal blood. In an
attempt to avoid the complications of total diversion of portal
flow, the distal splenorenal (DSRS) or Warren shunt was
developed in the late 1960s. 85,86

This operation is based on the principle of compartmental-
ization as discussed by Malt: the portal azygous system and
the portal splanchnic system may be surgically separated into
parallel and independent hemodynamic units. Decompres-
sion of the portal azygous system may be accomplished
without affecting the portal splanchnic system perfusion pres-
sure. 1 ' 84 This requires two steps: first the coronary vein and
right gastroepiploic vein are ligated, reducing blood flow into
the esophageal variceal system. Second, the distal splenic vein
is anastomosed to the left renal vein, without performing a

282

chapter 27 Portal hypertension: pathophysiology and clinical correlates

splenectomy This allows blood to drain from the esophageal
varices through the short gastrics into the lower pressure
systemic circulation.

DSRS vs. other shunts

Three prospective randomized clinical trials comparing the
efficacy of the selective shunt (DSRS) to the end-to-side
portacaval shunt have been published: in Toronto, in the
Boston-New Haven trial and at USC. 87-89 Similarly, the distal
splenorenal shunt has been compared with mesenteric-
systemic shunts in two randomized trials: in Atlanta and in
Philadelphia. 90 ' 91

The distal splenorenal shunt has been compared with
the end-to-side portacaval shunt in three randomized
studies: by Langer et ah in Toronto and Resnick et ah in the
Boston-New Haven trial, and Harley et ah at USC. 87-89
Langer 7 s group in Toronto reported that the DRSR was effec-
tive in reducing encephalopathy 87 The Boston and USC trials
were not conclusive. 88

Both the Atlanta and Philadelphia studies demonstrated
persistent hepatopetal flow in the distal splenorenal shunt pa-
tients, but not in the nonselective shunt patients. Correspond-
ing to this, quantitative measurements of hepatic function
were similar to the preoperative values in the distal splenore-
nal group but greatly decreased in the nonselective group. The
incidence of encephalopathy correlated with preservation of
hepatopetal blood flow. 87 ' 90

Splenopancreatic disconnection

In 1986 Warren modified the operation, proposing that com-
plete dissection of the splenic vein and division of the spleno-
colic ligament (splenopancreatic disconnection) should be
included as part of the operation. 56 This is intended to reduce
the pancreatic siphon effect: the tendency of pancreatic
branches of the splenic vein to progressively enlarge and serve
to shunt portal mesenteric blood into the systemic circulatory
system. This modification is intended to prolong and preserve
the selective quality of the DSRS. Clinically, Warren's group
found that this modification considerably extended the mag-
nitude of the operation. 92

Results: DSRS— patency and portal perfusion

The maintenance of portal perfusion in the early postoperative
period has been documented in greater than 90% of pa-
tients. 83,90,93 ' 94 A 10-year follow-up by Warren of his distal
splenorenal shunt group revealed that 75% have persistent
portal perfusion at 10 years. Patients who were demonstrated
to have portal perfusion at 3 years maintained this until the
conclusion of the study at 7 years. 85,91,95

The distal splenorenal shunt has several advantages
over other nonselective shunts: it preserves portal flow to the
liver; it maintains hepatotrophic perfusion; it permits the me-
tabolism of toxic metabolites; and it maintains a high portal
perfusion pressure in the intestinal venous bed, decreasing the

absorption of toxic substances. 85 It is the best procedure to
perform under elective conditions. It should be noted that
the DSRS has not been clearly demonstrated to improve
survival by itself, but when used in conjunction with judicious
sclerotherapy, it appears to provide the best survival in these
patients.

Surgical management of portal hypertension:
nonshunt procedures

Splenectomy

Based on Banti's theory that the diseased spleen caused portal
hypertension and ascites, splenectomy was one of the first
operations proposed for the treatment of Banti's syndrome
(splenomegaly, hypersplenism, ascites, often accompanied
by esophageal varices). The use of splenectomy in this setting
was in great part due to its advocacy by Osier, a great admirer
of Banti's work. It was not until 1936, when Rouselot reviewed
the experience at Columbia University in New York, that
the failings of this operation were noted: a significant inci-
dence of recurrent hemorrhage following splenectomy and
the consequent loss of the splenic and portal veins which
would preclude possible shunt surgery 96 In 1940 Thompson
was able to demonstrate statistically that splenectomy was
of value only to those patients with isolated splenic vein
thrombosis. 97 Additionally, Pemberton and Kierman reported
a 54% incidence of recurrent variceal hemorrhage with
splenectomy alone. 98

Collateralization

Development of collateral pathways between the portal circu-
lation and the systemic circulation was the goal of several pro-
cedures. Omentopexy, introduced by Talma in 1898, produces
collateral pathways by suturing the omentum to the peri-
toneum. 99 It was thought to be particularly beneficial in the
resolution of ascites. It was sometimes used in conjunction
with splenectomy for the relief of ascites associated with
splenomegaly and decreased WBC counts.

Another collateral promoting operation was the transposi-
tion of the spleen into the thorax. 100 ' 101 Like omentopexy, its
goal was to allow the development of large venous collateral
pathways between the portal venous system and the systemic
venous circulation. Unfortunately, these collateral path-
ways were never able adequately to decompress esophageal
varices. Thus the patients were doomed to repeated hemor-
rhage and these operations have been abandoned.

Ablation

Ablative procedures to remove the source of bleeding were
first advocated by Phemister and Humphreys in 1944, who
recommended total gastrectomy 28 Peters and Womack later

283

pa rt I Vascular pathology and physiology

encouraged splenectomy with obliteration of both the intra-
and extraluminal vasculature of the distal variceal bearing
esophagus and proximal stomach. 102 Keagy et al. reviewed the
long-term results of the "Womack" procedure and found the
risk to be prohibitively high, with a 54% incidence of rebleed
and a 35% operative mortality. They concluded that this proce-
dure should be used only in highly selected patients who do
not have suitable anatomy for a shunt. 103

Despite these results, the procedure was not abandoned.
The EEA stapler allowed transection and reanastomosis
of the distal esophagus with greater facility. The innovation
resulted in reduced operative mortality and improved reduc-
tion of postoperative hemorrhage. Wexler and Cooperman
et al. reported two groups of patients who underwent trans-
abdominal EEA variceal stapling, without recurrent
bleeding. 104 ' 105

Perhaps the most successful devascularization procedure
was developed by Suguira and colleagues in Japan. 99 ' 106 ' 107
The procedure is performed via separate thoracic and ab-
dominal incisions; in poor-risk patients, a two-stage proce-
dure is indicated. The esophagus is devascularized from the
gastroesophageal junction to the left inferior pulmonary vein.
The vagus nerve is carefully preserved. At the level of the di-
aphragm, the esophagus is partially transected, leaving only
the posterior muscular layer intact. Esophageal varices are
occluded, not ligated, by oversewing each with interrupted
sutures. The esophageal muscle is closed, but the mucosa is
not sutured. The abdominal operation is performed through a
separate midline incision and includes splenectomy, devascu-
larization of the abdominal esophagus and proximal stomach,
and a pyloroplasty and fundoplication.

Suguira and colleagues have reported an overall operative
mortality of 4.6%. Their emergency operative mortality is 20%.
Varices were eradicated in 97% of patients, and recurrent
bleeding occurred in only 2.5%. Their long-term survival
was 84%. 99 A follow-up report by this group on 276 patients
indicated equally good survival rates with excellent control
of variceal bleeding and no encephalopathy 106 In a later
report they analyzed their results according to the patients'
Child's classification. They found that in class C patients both
operative mortality and long-term survival were discourag-
ing. For the class A and B patients, the results were very good
with combined (A, B, and C) 15-year survivals as high as
72%. 108

Reports by Suguira's group have been confirmed by others
in Japan as well as by selected investigators in this coun-
try 109 ' 110 The EEA stapler has made this a relatively simple
procedure, but the possibility of esophageal perforation and
leakage still makes the procedure one of considerable risk. 111
Overall these studies demonstrate that esophageal transection
should be considered a reasonable option in the management
of acute hemorrhage in the debilitated patient with both
gastric and esophageal varices. 112-114

New developments in the treatment
of portal hypertension

Nonsurgical shunting: TIPSS

The transjugular intrahepatic portosystemic shunt (TIPSS)
has become a commonly utilized technique for controlling
variceal hemorrhage. This has had a dramatic effect on the
performance of operative shunting procedures, despite the
fact that TIPSS has been used clinically only since 1988. Over
the last few years, the indications for TIPSS have expanded,
although these applications have not always been backed
by randomized trials. Recent developments in TIPSS research,
most notably the use of covered stents, may improve the
patency of TIPSS, and thus lead to even more widespread use
of this procedure.

Rosch et al. demonstrated 30 years ago that intrahepatic
portacaval shunts could be created in a minimally invasive
fashion. 115 Palmaz et al. applied their stent to this type of
procedure, leading to prolonged patency of these shunts in
animal models. 116 After its initial application in human sub-
jects, there was a rapid expansion in its application, so that a
large number of interventionalists became skilled in this tech-
nique. Most major medical centers now perform TIPSS.

TIPSS can be performed with conscious sedation, and is
frequently carried out with this level of sedation in interven-
tional suites. Ultrasonography should be performed prior to
the procedure to document patency of the portal vein and to
assess the need for peritoneal tap. Most interventionalists
minimize the amount of ascites before performing TIPSS so
that the liver is not "floating" in the abdomen. Access is ideally
attained via the right internal jugular vein, which provides the
most direct route for cannulation of right hepatic vein branch-
es. Wedged venogram and portal pressures are obtained, and a
portal vein branch is then punctured. There are a number of
commercially available devices for advancing through the
liver parenchyma to the portal vein, with the Rosch-Uchida
transjugular liver access set (Cook Surgical) in use at our insti-
tution. These devices combine a cutting needle with an aspira-
tion port and are usually designed to fit a 10-French sheath.
Portal access is confirmed by aspirating blood during needle
advancement. A guidewire is advanced so that the hepatic and
portal circuits are in continuity. Finally, the track is balloon-
dilated and stented. At the end of the procedure, pressure
measurements can be repeated to confirm that the portal-
systemic gradient is less than 12mmHg. Completion veno-
gram must also be performed to assess stent placement and to
ensure that varices are no longer filling.

TIPSS for acute variceal bleeding stops the hemorrhage in
over 90% of patients. 117 This compares favorably with endo-
scopic sclerotherapy in many series, including at least eight
prospective randomized trials. 118 ' 119 However, most studies
have shown a decreased rate of rebleeding with TIPSS vs. scle-

284

chapter 27 Portal hypertension: pathophysiology and clinical correlates

ro therapy Rosch and Keller reported a 5.6% rebleeding rate at 3
months' follow-up, compared with the 20-30% usually seen
following endoscopic therapy 120 As with rebleeding following
other therapies, a significant number (25-30%) of these patients
will have a different, nonvariceal lesion as their hemorrhage
source. Confirmation of source is thus extremely important.

When the bleed is from varices, it is almost invariably asso-
ciated with the shunt not functioning properly. As with other
vascular conduits, short-term problems are usually technical
in nature, while those developing later are related to neoin-
tima formation. Although TIPSS can be revised with control of
the bleeding, one of the chief criticisms of these shunts is their
need for repeated interventions. 121 In fact, primary patency
of TIPSS is roughly 40% at 1 year, but secondary patency
approaches 90%. 122

Despite the benefit in preventing rebleeding, TIPSS is
considered second-line therapy to sclerotherpy due to several
factors, most notably the increased incidence of hepatic
encephalopathy (roughly 50% vs. 20%). In addition, for
patients whose bleeding is controlled by sclerotherapy, there
is no survival benefit for TIPSS.

In comparison with surgical shunts, it is clear that TIPSS has
replaced open procedures for most Child's B and C patients.
This is underscored by the fact that it is not uncommon for a
surgical resident to finish his or her training without perform-
ing a single shunting procedure. Interestingly, in the largest
prospective trial comparing TIPSS and surgical shunts (8 mm
prosthetic H-grafts), TIPSS was found to have higher rates of
death, rebleeding, and treatment failures. 123 This trial, as well
as an additional 4 years of follow-up, was published by Rose-
murgy et ah and evaluated 35 patients in each arm, with pair-
ing of patients by their Child's class. In addition, comparison
of rates of post-treatment encephalopathy generally favors
surgery as well. Further data may compel practitioners to con-
sider surgical shunts in patients able to tolerate them.

TIPSS has been applied in several other clinical situations.
One broad category is as a bridge to hepatic transplantation.
Theoretically, the procedure leaves the portal hepatis unper-
turbed and allows for easier subsequent operation. Some have
argued that TIPSS actually accelerates liver failure in patients
awaiting transplant. 124 However, several studies support the
use of TIPSS in this setting, particularly as a means of improv-
ing patients' general condition, sometimes with decreased
blood loss at time of surgery 125-129 One technical note of im-
portance is placement of the distal end of the stent well above
the inferior vena cava so that it does not interfere with the
transplant itself or cause damage to the portal vein.

Another application of TIPSS has been for refractory ascites
management. There are numerous retrospective trials docu-
menting the effectiveness of TIPSS in this setting, but only a
few randomized trials. In a randomized study by Rossle and
coauthors, TIPSS was shown to be more effective than para-
centesis in controlling refractory ascites. 130 In a study by Lebrec

et ah, TIPSS was shown to benefit Child's B patients with as-
cites, but was of no help in Child's C patients, actually having
little effect on ascites while increasing mortality 131 In a related
process, Siegerstetter and colleagues reported retrospectively
on the benefits of TIPSS in treating refractory hydrothorax. 132

TIPSS has also been utilized with some success in the treat-
ment of Budd-Chiari syndrome. There are several small stud-
ies investigating its use in this setting. Perello and coauthors
managed 13 such patients with TIPSS and found that 11 pa-
tients were doing well following placement, with follow-up of
4 years. 133 Interestingly, many of the patients' TIPSS were no
longer patent but did not require treatment due to absence of
signs of portal hypertension. Blum and coauthors reported on
12 Budd-Chiari patients who underwent TIPSS. Two of the pa-
tients had fulminant hepatic failure and died, but the remain-
ing 10 had resolution of their ascites. 134 Clearly, more data
must be collected before the role of TIPSS for Budd-Chiari
syndrome is fully appreciated.

One additional application of TIPSS has been for hepatore-
nal syndrome. Brensing and colleagues reported an improve-
ment in renal function in these patients compared with
nonshunted patients. 135 Studies by Ochs and Alam also
support the use of TIPSS in this setting, but no prospective data
are available at this time. 136 As mentioned previously, the main
drawback of TIPSS is the poor primary patency and need for
repeated interventions to maintain the shunt. It has been
shown in animal studies that the use of covered stents im-
proves patency for TIPSS. There are now human data to sup-
port this contention. Otal and coauthors used PTFE-covered
nitinol stents in a recent series of 20 patients, and reported pri-
mary and secondary patency rates of 80% and 100% at 387 days
(as opposed to the noncovered stent literature rate of 58% pri-
mary patency at 1 year). 137 Several other small series have been
published, but long-term data are not yet available. If these de-
vices fulfill their promise of impoved patency, perhaps TIPSS
will be applied with more success in a wider range of clinical
situations.

Liver transplantation

With the advent of liver transplantation as an established
modality for the care of patients with end-stage liver failure,
the role of nontransplant procedures (shunt surgery in par-
ticular) has been the subject of considerable debate. Current
best transplant survival rates are generally more favorable
than most of the reported survival of Child's class C patients
following the best care with combination of sclerotherapy and
shunting. Iwatsuki and colleagues have presented their data
in two reports. 138,139 They have described the survival of 302
liver transplant patients who presented with bleeding
esophageal varices. All patients were Child's class C. The sur-
vival of these patients is 71% at 5 years. The conclusion is that
in Child's class C patients, liver transplantation should be con-

285

pa rt I Vascular pathology and physiology

sidered as the treatment of choice —assuming that the patients
are reasonable transplant candidates.

At UCLA liver transplantation has resulted in improved
survival of Child's class C patients. In a 6-year period, of 761
patients 77 underwent portosystemic shunting as their initial
procedure, and 684 underwent hepatic transplantation. Of
those transplanted, 86% were Child's class C patients, where-
as only 16% of the shunt patients were Child's' class C patients.
Despite this, 15% of shunt patients eventually required liver
transplantation for progressive hepatic deterioration.
Furthermore, the 5-year survival of the shunt group was 64%
in contrast to a 73% 5-year survival of the transplanted pa-
tients. Portosystemic shunting appears to be an appropriate
form of therapy for Child's A and B patients, but Child's C pa-
tients who are transplant candidates are best managed by liver
transplantation.

Postoperative care

Control of ascities

The most common problem in the postoperative period is as-
cites. Fluid management is the key to minimizing this prob-
lem. Postoperative fluids should be restricted to free water and
salt-poor albumin or fresh frozen plasma. These should be
given to maintain adequate intravascular volume, yet avoid-
ing overexpansion of the patient's intravascular space.

Ascites predisposes these patients to potentially lethal
complications, including renal failure, peritonitis, variceal
hemorrhage, pleural effusions with respiratory insufficiency,
abdominal wall hernias, anorexia, and sepsis. Most patients
with ascites may be controlled with a restricted salt diet and
diuretic regimen. Only 5% of ascitic patients can be considered
to have truly intractable ascites, and it is these patients who
may require surgical intervention. 40

A fluid restriction of no more than 1 1/day should be ordered
in addition to a 20 mEq sodium diet per day. With this regimen
one would expect a diuresis of 500ml to 1 1/day. If such a di-
uresis does not occur, then progressive diuretic therapy is indi-
cated. This usually involves gradual increase in the dosages
and varieties of diuretics. Frequently the first diuretic used is
spironolactone, a potassium-sparing diuretic. It should be
started at a dosage of lOOmg/day and doubling it at 2-day
intervals until a maximum dosage is obtained. If further
diuresis is required, the other agents such as metalazone,
hydrochlorothiazide, and lasix may be used.

At the first sign of encephalopathy or elevation of blood
urea nitrogen by lOmg/dl or creatinine by 0.5mg/dl, all
diuretics must be discontinued so as to avoid development of
the hepatorenal syndrome. If, after such a trial of intensive
nonsurgical therapy, no significant response occurs, surgery
is indicated. 41

Peritoneovenous shunting

In 1974, LeVeen introduced the peritoneovenous shunt. This
device consists of a silastic tube which runs from the peri-
toneum to the superior vena cava. It is controlled by a one-way
valve so that a pressure gradient of 5 cm H 2 will suffice to
transfer the ascitic fluid from the abdomen to the intravascular
space. 42 The Denver shunt is a variation of the LeVeen shunt
which incorporates a pump in line with the shunt, which is im-
planted in the subcutaneous tissues on the chest wall. Its pro-
posed benefit is its ability to clear the shunt of debris by using
the pump mechanism. 140

The use of peritoneovenous shunts is associated with a
number of complications. The most significant early com-
plications include congestive failure and disseminated in-
travascular coagulopathy (DIC). Almost all patients may
demonstrate some serological characteristics of DIC. Clini-
cally apparent DIC is much less common. 43 ' 141 The serological
indices will normalize as the shunt flow reduces towards a
baseline volume, usually about the end of the first week. 44 If
clinically apparent DIC develops the shunt must be ligated.
Other attempts at treating DIC such as infusions of blood
products may be ineffective and result in hemorrhagic
death. 43 Recurrent variceal hemorrhage in the early postshunt
period has generated the concern that the rapid expansion
of intravascular volume from the shunt may result in in-
creased distention of varices and bleeding. In all, early
postoperative complications have resulted in a mortality
approaching 20% in some institutions. 45 Late complications
include shunt infection, shunt occlusion, and death. Infection
may occur in as many as 26% of patients. Shunt occlusion is
thought to result from the precipitation of the ascitic protein in
the shunt tubing. Death related to inherent liver dysfunction is
not uncommon.

The infusion of ascitic fluid which follows use of a perito-
neovenous shunt will frequently result in a brisk diuresis. The
exact mechanism is not clear. Volume expansion, increased
renin levels from the ascitic fluid, and the relief of intraab-
dominal pressure have been proposed as explanations for
the diuresis. 6 Because of its impact on kidney function,
peritoneovenous shunting is considered a potential treatment
of hepatorenal syndrome. Successful resolution of the syn-
drome by placement of such a shunt has been reported. 46

Contraindications to the placement of peritoneovenous
shunts include infection of ascitic fluid, recurrent sepsis, or
encephalopathy. Other absolute contraindications include a
bilirubin greater than 6 or prolongation of prothrombin time
more than 4 s. These are associated with postoperative coagu-
lopathy and fatal hemorrhage. DIC resulting from an intra-
venous test infusion of ascitic fluid is another relative
contradiction to peritoneovenous shunting. Finally, a large
pleural effusion associated with an elevated intrathoracic
pressure may preclude use of these shunts. 40

In general, peritoneovenous shunting is an effective method
to treat ascites. It should be kept in mind as an adjunct to the

286

chapter 27 Portal hypertension: pathophysiology and clinical correlates

Resuscitate
Endoscopy for diagnosis
Emergent sclerotherapy

Bleeding controlled

Long-term
sclerotherapy

Breakthrough
Bleeding

Observation

Figure 27.1 Acute variceal hemorrhage.

Continued bleeding

Minimal to moderate bleed

/ \

Poor hepatic function Good hepatic function

Liver transplant Surgical shunt

Massive bleed

Emergency shunt
TIPSS

care of these patients, and may not be as simple in practice as it
is in concept. It should not be expected to result in prolonga-
tion of survival or alteration of the natural course of their
hepatic disease.

Recurrent bleeding

If variceal hemorrhage occurs within the immediate post-
operative period, angiography should immediately be per-
formed to define accurately the anatomy and patency of
the shunt. If a patent collateral such as the coronary vein is
identified, one should consider percutaneous transhepatic
embolization. If other collateral pathways are present, reex-
ploration may be indicated in order to ligate them. If the
shunt is occluded and the patient is bleeding, then reexplo-
ration should be undertaken to repair the shunt or to perform
another shunt.

Treatment plant for variceal hemorrhage

Because of the acute nature of a variceal hemorrhage, physi-
cians must have a clearly defined approach outlined to care for
these patients (Fig. 27.1). As with trauma patients whose initial
care is critical, portal hypertensives require large-bore intra-
venous lines, judicious fluid resuscitation, assessment of their
hepatic status, and diagnosis of the source of hemorrhage. A
nasogastric tube must be placed to assess the presence of blood
in the stomach and to prepare for endoscopy. Endoscopy is es-
sential for the diagnosis of the site of bleeding. Vasopressin
should be started on all patients diagnosed with esophageal

variceal bleeding or hypertensive gastropathy. If pitressin
does not stop the hemorrhage, then balloon tamponade
should be instituted promptly. Emergency sclerotherapy
should be considered if balloon tamponade is shunt surgery is
indicated if all of the above procedures fail.

If bleeding is controlled, then a course of periodic
sclerotherapy is indicated to ablate the esophageal varices.
Elective shunt surgery should be considered in those patients
who can not be controlled by periodic sclerotherapy. The DSRS
should be attempted whenever anatomy allows. If patients
have poor hepatic function, then they should be considered for
liver transplantation.

Summary

The management of portal hypertensive patients requires
an insightful, judicious approach based on a clear understand-
ing of the patient's hepatic disease, the stage of hepatic dys-
function, and the anatomic aberrations which underlie the
presentation. The physicians must be familiar with the full
spectrum of alternatives as well as the patient's tolerance for
these (and their potential complications). Therapy must be in-
dividualized since the patients vary considerably with regard
to their particular set of problems. Finally, as newer interven-
tions are refined, physicians must thoughtfully incorporate
these into the care which they offer. In deciding how best to
care for these patients it should be recalled that the optimal
outcome is not merely survival but also considers the quality
of life.

287

pa rt I Vascular pathology and physiology

References

1. Malt R. Portasystemic venous shunts. N Engl J Med 1976; 295:24.

2. Shaldon S, Sherlock S. Portal hypertension in the myeloprolifera-
tive syndrome and the reticuloses. Am } Surg 1962; 32:758.

3. Shaldon S, Chiandussi L, Guevara L. The measurement of hepa-
tic blood flow and intrahepatic shunted blood flow by colloid,
heat denatured serum albumin labeled with 1131. / Clin Invest
1969; 40:1038.

4. Gordon M, DelGuerco L. Late effects of portal systemic shunting
procedures on cardiorespiratory dynamics in man. Ann Surg
1972; 176:672.

5. Langer B, Stone R, Colapinto R. Clinical spectrum of the Budd-
Chiari syndrome and its surgical management. Am } Surg 1975;
129:137.

6. Ludwick J, Markel S, Child L. Chiari's disease. Arch Surg 1965;
91:697.

7. Sherlock S. Classification and functional aspects of portal hyper-
tension. Am J Surg 1974; 127:121.

8. Ahn S, Yellin A, Sheng F et ah Selective surgical therapy of the
Budd-Chiari syndrome provides superior survival rates than
conservative medical management. / Vase Surg 1987; 5:28.

9. Lebrec D, DeFleury P, Rueff B et ah Portal hypertension, size of
esophageal varices and risk of gastrointestinal bleeding in alco-
holic cirrhosis. Gastroenterology 1980; 79:1139.

10. Paquet K. Prophylactic endoscopic sclerosing treatment of the
esophageal wall varices. A prospective controlled randomized
trial. Endoscopy 1982; 14:4.

11. Wizel L, Wolbergs E, Merki H. Prophylactic endoscopic scle-
rotherapy of esophageal varices. A prospective controlled study.
Lancet 1985; 1:773.

12. The North Italian Endeoscopic Club for the Study and Treatment
of Esophageal Varices. Prediction of the first variceal hemor-
rhage in patients with cirrhosis of the liver and esophageal
varices: A prospective multicenter trial. N Engl } Med 1988;
319:983.

13. Dagradi ZA. The natural history of esophageal varices in
patients with alcoholic cirrhosis. Am J Gastroenterol 1972; 57:
520.

14. Schenker S, Breen K, Hoyumpa A. Hepatic encephalopathy:
current status. Gastroenterology 1974; 66:121.

15. Fischer J, James J, Jeppsson B. Hyperammonaemia, plasma
amino acid imbalance and blood-brain amino acid transport.
Lancet 1979; 2:772.

16. Fischer, J, Furovics J, Folcao H. L-Dopa in hepatic coma. Ann Surg
1976; 183:386.

17. Bosch J, Navasa M, Garcia-Pagan J et ah Portal hypertension. Med
Clin North Am 1989; 73:931.

18. Bosch J, Mastai R, Kravetz D et ah Hemodynamic evaluation
of the patients with portal hypertension. Semin Liver Dis 1986; 6:
309.

19. Koslin D, Berland L. Duplex Doppler examination of the liver
and portal system. Clin Ultrasound 1987; 15:675.

20. Ohnishi K, Saito M, Koen H et ah Pulsed Doppler flow as a criteri-
on of portal venous velocity: comparison with cineangiographic
measurements. Radiology 1985; 154:495.

21. Bolondi L, Gandolfi L, Arienti V et ah Ultrasonography in the

diagnosis of portal hypertension: diminished response of portal
vessels to respiration. Radiology 1982; 142:167.

22. Cello J, Deveney K, Trunkey D. Factors influencing survival after
therapeutic shunts. Am J Surg 1981; 141:257.

23. Orloff M, Duguay L, Kosta L. Criteria for selection of patients for
emergency portacaval shunt. Am } Surg 1977; 134:146.

24. Schwartz S. Liver. In: Schwartz S, ed. Principles of Surgery. New
York: McGrawHill, 1984:1257.

25. Rueff B, Benhamou J. Management of gastrointestinal bleeding
in cirrhotic patients. Clin Gastroenterol 1975; 4:426.

26. Rikkers L. Operations for management of esophageal variceal
hemorrhage. West J Med 1982; 136:107.

27. Resnick R. Portal hypertension. Med Clin North Am 1975; 59:945.

28. Phemister D, Humphreys E. Gastroesophageal resection and
total gastrectomy in the treatment of bleeding varicose veins in
Banti's syndrome. Ann Surg 1947; 125:397.

29. Mikkelsen W. Therapeutic portacaval shunt. Arch Surg 1974;
108:302.

30. Bell R, Miyai K, Orloff M. Outcome in cirrhotic patients with
acute alcoholic hepatitis after emergency portacaval shunt for
bleeding esophageal varices. Am J Surg 1984; 147:78.

31. Eckhauser F, Appelman H, O'Leary T. Hepatic pathology as a
determinant of prognosis after portal decompression. Am }
Surg 1980; 139:105.

32. Viamonte M, Warren W, Famon J. Angiographic investigations in
portal hypertension. Surg Gynecol Obstet 1970; 130:37.

33. Huizinga W, Keenan J, Marszaley A. Sclerotherapy for bleeding
esophageal varices. A case report. SAfrMed J 1984; 65:436.

34. Mikkelsen V, Turrill F, Kern W. Acute hyaline necrosis of the liver.
A surgical trap. Am J Surg 1968; 116:266.

35. Seidman E, Neber A, Morin C. Spinal cord paralysis following
sclerotherapy for esophageal varices. Hepatology 1981; 4:950.

36. Perakos P, Cirbus J, Camara D. Persistent bradyarrhythmia after
sclerotherapy for esophageal varices. South Med J 1984; 77:531.

37. Lunderquist A, Vang J. Transhepatic catheterization and obliter-
ation of the coronary vein in patients with portal hypertension
and esophageal varices. N Engl J Med 1974; 291 :646.

38. Abecossis M, Makowka L, Lanser B. Sclerotherapy for
esophageal varices. Can J Surg 1984; 27:561.

39. Reynolds T. The role of hemodynamic measurements in porta-
systemic shunt surgery. Arch Surg 1974; 108:276.

40. Crafoord C, Frenckner P. New surgical treatment of varicose
veins of the esophagus. Acta Otolaryngol 1939; 27:422.

41. Johnston G, Rodgers H. A review of 15 years experience in the
use of sclerotherapy in the control of acute hemorrhage for
esophageal varices. Br J Surg 1973; 60:797.

42. Paquet K, Kalk J, Koussouris P. Immediate endoscopic sclerosis
of bleeding esophageal varices: a prospective evaluation over 5
years. SurgEndosc 1988; 2:18.

43. Schubert T, Smith O, Kirkpatrick S et ah Improved survival
in variceal hemorrhage with emergent sclerotherapy. Am }
Gastroenterol 1987; 82:1134.

44. Terblanche J, Northover J, Bornman P et ah A prospective evalua-
tion of injection sclerotherapy in the treatment of acute bleeding
esophageal varices. Surgery 1979; 85:239.

45. Terblanche J, Yakoob H, Bornman P et ah Acute bleeding varices:
a five-year prospective evaluation of tamponade and sclerother-
apy. Ann Surg 1981; 194:521.

46. Paquet K, Feussner H. Endoscopic sclerosis and esophageal bal-

288

chapter 27 Portal hypertension: pathophysiology and clinical correlates

loon tamponade in acute hemorrhage from esophagogastric
varices: a prospective controlled randomized trial. Hepatology
1985; 5:580.

47. Larson A, Cohen H, Zweiban B et ah Acute esophageal variceal
sclerotherapy: results of a prospective randomized controlled
trial. JAMA 1986; 255:497.

48. Barsoum M, Bolous F, El-Rooby A et ah Tamponade and injection
sclerotherapy in the management of bleeding oesophageal
varices. Br] Surg 1982; 69:76.

49. Project CEVS. Sclerotherapy after first variceal hemorrhage in
cirrhosis. A randomized multicenter trial. N Eng } Med 1984;
311:1594.

50. Westaby D, MacDougall B, Williams R. Improved survival fol-
lowing injection sclerotherapy for esophageal varices: final
analysis of a controlled trial. Hepatology 1985; 5:827.

51. Terblanche J, Bornman P, Kahn D et ah Failure of repeated
injection sclerotherapy to improve long term survival after
esophageal variceal bleeding. A five year prospective controlled
clinical trial. Lancet 1983; 2:1328.

52. Soderlund C, Ihre T. Endoscopic sclerotherapy v. conservative
management of bleeding oesophageal varices. Acta Chir Scand
1985; 151:449.

53. Korula J, Balart L, Radvan G et ah A prospective randomized
controlled trial of chronic esophageal variceal sclerotherapy.
Hepatology 1985; 5:584.

54. Cello J, Grendell J, Crass R et ah Endoscopic sclerotherapy versus
portacaval shunt in patients with severe cirrhosis and variceal
hemorrhage. N Engl J Med 1984; 311:1589.

55. Cello J, Grendell J, Crass R et ah Endoscopic sclerotherapy versus
portacaval shunt in patients with severe cirrhosis and acute
variceal hemorrhage. Long-term follow-up. N Engl } Med 1987;
316:11.

56. Warren W, Henderson J, Millikan W et ah Distal splenorenal
shunt versus endoscopic sclerotherapy for long-term manage-
ment of variceal bleeding. Preliminary report of a prospective,
randomized trial. Ann Surg 1986; 203:454.

57. Teres J, Bordas J, Bravo D et ah Sclerotherapy vs. distal splenore-
nal shunt in the elective treatment of variceal hemorrhage: a
randomized controlled trial. Hepatology 1987; 7:430.

58. Rikkers L, Burnett D, Volentine G et ah Shunt surgery versus
endoscopic sclerotherapy for long-term treatment of variceal
bleeding. Early results of a randomized trial. Ann Surg 1987;
206:261.

59. Burroughs A, Hamilton G, Phillips A et ah A comparison of
sclerotherapy with staple transection of the esophagus for the
emergency control of bleeding from esophageal varices. N Engl J
Med 1989; 321:857.

60. Teres J, Baroni R, Bordas M et ah Randomized trial of portacaval
shunt, stapling transection and endoscopic sclerotherapy in
uncontrolled variceal bleeding. J Hepatol 1987; 4:159.

61. Terblanche J, Bornman P, Kirsch R. Sclerotherapy for bleeding
esophageal varices. Annu Rev Med 1984; 35:83.

62. Yassim Y, Sherif S. Randomized controlled trial of injection
sclerotherapy for bleeding esophageal varices. Br J Surg
1983; 70:20.

63. Reilly J, Schade R, Roh M et ah Esophageal variceal sclerosis. Surg
Gynecol Obstet 1982; 155:497.

64. Palani L, Abvabara S, Kraft A et ah Endoscopic sclerotherapy in
acute variceal hemorrhage. Am } Surg 1981; 141:164.

65. Lewis J, Chung R, Allison J. Sclerotherapy of esophageal varices.
Arch Surg 1980; 115:476.

66. Johnston G. Bleeding esophageal varices: the management of
shunt rejects. Ann R Coll Surg Engl 1981; 63:3.

67. Goodale R, Silvis S, O'Leary J et ah Early survival for bleeding
esophageal varices. Surg Gynecol Obstet 1982; 155:523.

68. Bacon A, Bauley-Newton R, Connors A. Pleural effusions
after endoscopic variceal sclerotherapy. Gastroenterology 1985;
88:1910.

69. Tripodis S, Buenskin A, Wenser J. Gastric ulcers after endoscopic
sclerosis of esphogeal varices. / Clin Gastroenterol 1985; 7:77.

70. Lewis J, Chung R, Allison J. Injection sclerotherapy for control of
acute variceal hemorrhage. Am } Surg 1981; 142:592.

71. Gotzsche PC. Somatostatin analogues for acute bleeding
oesophageal varices. Cochrane Database Syst Rev 2002;
1:(1)CD000193.

72. Ioannou G, Doust J, Rockey DC. Terlipressin for acute
esophageal variceal hemorrhage. Cochrane Database Syst Rev
2001;1:CDC002147.

73. LeVeen H, Wapnick S, Grosberg S et ah Further experience with
peritoneovenous shunt for ascites. Ann Surg 1976; 184:574.

74. Burchell A, Rousselot L, Panke W. A seven-year experience with
side-to-side portacaval shunts for cirrhotic ascites. Ann Surg
1968; 168:655.

75. Turcotte J, Wallin V, Child C. End-to-side versus side-to-side
portacaval shunts in patients with hepatic cirrhosis. Am } Surg
1979; 117:108.

76. Iwatsuki S, Mikkelsen W, Redeker A et ah Clinical comparison of
the end-to-side and side-to-side portacaval shunt. Ann Surgl973;
178:65.

77. Clatworthy H, Wall T, Watman R. A new trial of portal-to-
systemic venous shunt for portal hypertension. Arch Surg
1955; 71:588.

78. Drapanas T, LoCicero J, Dowling J. Interposition meso-caval
shunt for treatment of portal hypertension. Ann Surg 1972;
176:435.

79. Lord J, Rossi G, Daliana M et ah Mesocaval shunt modified
by the use of a Teflon prosthesis. Surg Gynecol Obstet 1970;
130:525.

80. Nay H, Fitzpatrick H. Mesocaval "H" graft using autogenous
vein graft. Am Surg 1976; 183:114.

81. Read R, Thompson B, Wise W et ah Mesocaval H venous homo-
grafts. Arch Surg 1970; 101:785.

82. Thompson B, Reed B, Casall R. Interposition grafting for portal
hypertension. Am J Surg 1975; 130:733.

83. Cargenas A, Busuttil R. A comparative analysis of the mesocaval
H graft versus the distal splenorenal shunt. Curr Surg 1982;
39:151.

84. Malt R. Portasystemic venous shunts. N Engl J Med 1976; 295:
80.

85. Warren W. Control of variceal bleeding. Reassessment of
rationale. Am J Surg 1983; 145:8.

86. Warren W, Zeppa R, Fomon J. Selective transsplenic decompres-
sion of gastroesophageal varices by distal splenorenal shunt.
Ann Surg 1967; 166:431.

87. Langer B, Rotstein L, Stone R. A prospective randomized trial of
the selective distal splenorenal shunt. Surg Gynecol Obstet 1980;
150:45.

88. Harley H, Morgan T, Redeker A et ah Results of a randomized

289

pa rt I Vascular pathology and physiology

trial of end-to-side portacaval shunt and distal splenorenal shunt
in alcoholic liver disease and variceal bleeding. Gastroenterology
1986; 91:802.

89. Resnick R, Atterbury L, Grace N et ah Distal splenorenal shunt
versus portal systemic shunt: current status of a controlled trial.
Gastroenterology 1979; 77:433.

90. Rikkers L, Rudman D, Galambos J. A randomized controlled trial
of distal splenorenal shunts. Ann Surg 1978; 188:271.

91. Millikan W, Warren W, Henderson J et ah The Emory prospective
randomized trial: selective versus nonselective shunt to control
variceal bleeding. Ten year follow-up. Ann Surg 1985; 201:712.

92. Maksoud J, Mies S. Distal splenorenal shunt in children. Ann
Surg 1982; 195:401.

93. Reichle F, Fahmy W, Golsorkhi M. Prospective comparative
clinical trial with distal splenorenal and mesocaval shunts. Am }
Surg 1979; 137:13.

94. Warren W, Millikan W, Henderson J. The years of portal hyper-
tension surgery at Emory. Ann Surg 1982; 195:530.

95. Fulenwider J, Smith R, Millikan W et ah Variceal hemorrhage in
the veteran population. To shunt or not to shunt? Am Surg 1984;
50:264.

96. Rousselot L. The role of congestion (portal hypertension) in so
called Banti's syndrome: a clinical and pathological study of
thirty one cases with late results following splenectomy. JAMA
1936; 107:1788.

97. Thompson W. The pathogenesis of Banti's disease. Ann Intern
Med 1940; 14:255.

98. Pemberton J, Kierman P. Surgery of the spleen. Surg Clin North
Am 1945; 25:880.

99. Suguira M, Futagawa S. Anew technique for treating esophageal
varices. / Thorac Cardiovasc Surg 1973; 66:677.

100. Strauch G. Supradiaphragmatic splenic transposition. Am J Surg
1970; 119:379.

101. McClelland R, Bashour F. Supradiaphragmatic transposition of
the spleen in portal hypertension. Arch Surg 1969; 98:175.

102. Peters R, Womack N. Surgery of vascular distortions in cirrhosis
of the liver. Ann Surg 1961; 154:432.

103. Keagy B, Schwartz J, Johnson G. Should ablative operations be
used for bleeding esophageal varices? Ann Surg 1986; 203:463.

104. Wexler M. Treatment of bleeding esophageal varices by transab-
dominal esophageal transection with the EEA stapling instru-
ment. Surgery 1980; 88:406.

105. Cooperman M, Fabri P, Martin E et ah EEA esophageal stapling
for control of bleeding varices. Am J Surg 1980; 140:821 .

106. Suguira M, Futagawa S. Further evaluation of the Suguira
procedure in the treatment of esophageal varices. Arch Surg
1977; 112:1317.

107. Koyarna K, Takagi Y, Ouchi K et ah Results of esophageal transec-
tion for esophageal varices. Am J Surg 1980; 139:204.

108. Suguira M, Futagawa S. Results of six hundred and thirty-six
esophageal transections with paraesophagogastric devascular-
ization in the treatment of esophageal varices. / Vase Surg 1984;
1:254.

109. Superina R, Weber J, Shandling B. A modified Suguira operation
for bleeding varices in children. / Pediatr Surg 1983; 18:794.

110. Weese J, Starling J, Yale C. Control of bleeding esophageal varices
by transabdominal esophageal transection, gastric devascular-
ization and splenectomy. Surg Gastroenterol 1984; 3:31.

111. Wanamaker S, Cooperman M, Carey L. Use of the EEA stapling

instrument for control of bleeding esophageal varices. Surgery
1983; 94:620.

112. Wexler M. Esphageal procedures to control bleeding from
varices. Surg Clin North Am 1983; 63:905.

113. Spence R, Anderson J, Johnston G. Twenty-five years of injection
sclerotherapy for bleeding varices. Br J Surg 1985; 72:195.

114. Huizinga W, Angorn I, Baker L. Esophageal transection versus
injection sclerotherapy in the management of bleeding
esophageal varices in patients of high risk. Surg Gynecol Obstet
1985; 160:539.

115. Rosch J, Hanafee W, Snow H, Barenfus M, Gray R. Transjugular
intrahepatic portacaval shunt. An experimental work. Am } Surg
1971; 121:588.

116. Palmaz JC, Garcia F, Sibbitt RR et ah Expandable intrahepatic
portacaval shunt stents in dogs with chronic portal hyperten-
sion. Am JRoentgenol 1986; 147:1251.

117. Richter G, Noeledge G, Palmaz J et ah The transjugular intra-
hepatic portosystemic stent-shunt (TIPSSS); results of a pilot
study. Cardiovasc Intervent Radiol 1990; 13:200.

118. Luca A, D'Amico G, La Galla R, Midiri M, Morabito A, Pagliaro L.
TIPS for prevention of recurrent bleeding in patients with cirrho-
sis: meta-analysis of randomized clinical trials. Radiology 1999;
212:411.

119. Narahara Y, Kanazawa H, Kawamata H et ah A randomized
clinical trial comparing transjugular intrahepatic portosystemic
shunt with endoscopic sclerotherapy in the long-term manage-
ment of patients with cirrhosis after recent variceal hemorrhage.
Hepatol Res 2001; 21:189.

120. Rosch J, Keller FS. Transjugular intrahepatic portosystemic
shunt: present status, comparison with endoscopic therapy and
shunt surgery, and future prospectives. World} Surg 2001; 25:337.

121. Latimer J, Bawa SM, Rees CJ, Hudson M, Rose JD. Patency
and reintervention rates during routine TIPSS surveillance.
Cardiovasc Intervent Radiol 1998; 21:234.

122. Saxon RS, Ross PL, Mendel-Hartvig J et ah Transjugualr intra-
hepatic portosystemic shunt patency and the importance
of stenosis location in the development of recurrent symptoms.
Radiology 1998; 207:683.

123. Rosemurgy AS, Serafini FM, Zweibel BR et ah Tranjugular intra-
hepatic portosystemic shunt vs small-diameter prosthetic H-
graft portacaval shunt: extended follow-up of an expanded
randomized prospective trial. / Gastrointest Surg 2000; 4:589.

124. Freeman RB Jr, FitzMaurice SE, Greenfield AE, Halin N, Haug
CE, Rohrer RJ. Is the transjugular intrahepatic portacaval shunt
procedure beneficial for liver transplant recipients? Transplanta-
tion 1994; 58:297.

125. Woodle ES, Darcy M, White HM et ah Intrahepatic portosystemic
vascular stents: a bridge to hepatic transplantation. Surgery 1993;
113:344.

126. Millis JM, Martin P, Gomes A et ah Transjugular intrahepatic
portosystemic shunts: impact on liver transplantation. Liver
TransplSurg 1995; 1:229.

127. Woodle ES, Darcy M, White HM et ah Intrahepatic portosystemic
vascular stents: a bridge to hepatic transplantation. Surgery 1993;
113:344.

128. Lerut JP, Laterre PE, Goffette P et ah Transjugular intrahepatic
portosystemic shunt and liver transplantation. Transpl Int 1996;
9:370.

129. John TG, Jalan R, Stanley AJ et ah Transjugular intrahepatic

290

chapter 27 Portal hypertension: pathophysiology and clinical correlates

portosystemic stent-shunt (TIPSS) insertion as a prelude to
orthotopic liver transplantation in patients with severe portal
hypertension. Eur J Gastroenterol Hepatol 1996; 8:1145.

130. Rossle M, Ochs A, Gulberg V et ah A comparison of paracentesis
and transjugular intrahepatic portosystemic shunting in patients
with ascites. NEngl J Med 2000; 8:1701.

131. Lebrec D, Giuily N, Hadengue A et ah Transjugular intrahepatic
portosystemic shunts: comparison with paracentesis in patients
with cirrhosis and refractory ascites: a randomized trial. French
Group of Clinicians and a Group of Biologists. / Hepatol 1996;
25:135.

132. Siegerstetter V, Deibert P, Ochs A, Olschewski M, Blum HE,
Rossle M. Treatment of refractory hepatic hydrothorax with
transjugular intrahepatic portosystemic shunt: long-term results
in 40 patients. Eur J Gastroenterol Hepatol 2001; 13:529.

133. Perello A, Garcia-Pagan JC, Gilabert R et ah TIPS is a useful
long-term derivative therapy for patients with Budd-Chiari
syndrome uncontrolled by medical therapy. Hepatology 2002;
35:132.

134. Blum U, Rossle M, Haag K et ah Budd-Chiari syndrome: techni-
cal, hemodynamic, and clinical results of treatment with

transjugular intrahepatic portosystemic shunt. Radiology 1995;
197:805.

135. Brensing KA, Textor J, Perz J et ah Long term outcome after
transjugular intrahepatic portosystemic stent-shunt in non-
transplant cirrhotics with hepatorenal syndrome: a phase II
study. Gut 2000; 47:288.

136. Allgaier HP, Haag K, Ochs Aet ah Hepato-pulmonary syndrome:
successful treatment by transjugular intrahepatic portosystemic
stent-shunt (TIPS). } Hepatol 1995; 23:102.

137. Otal P, Smayra T, Bureau C et ah Preliminary results of a new
expanded polytetrafluoroethylene-covered stent-graft for
transjugular intrahepatic portosystemic shunt procedures.
Am J Roentgenol 2002; 178:141.

138. Iwatsuki S, Starzl T, Todo S et ah Liver transplantation in the
treatment of bleeding esophageal varices. Surgery 1988; 104:697.

139. Reyes J, Iwatsuki S. Current management of portal hypertension
with liver transplantation. Adv Surg 1992; 25:189.

140. Lund R, Newkirk J. Peritoneo-venous shunting system for
surgical management of ascites. Contemp Surg 1979; 14:31 .

141. Greig P, Langer B, Blendis L et ah Complications of the
peritoneovenous shunting for ascites. Am } Surg 1980; 139:125.

291

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Noninvasive vascular

diagnostics

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Physiologic basis of hemodynamic
measurement

R. Eugene Zierler

Although the clinical presentation of peripheral arterial occlu-
sive disease typically includes signs and symptoms such as
claudication or ischemic rest pain, the physiologic abnormali-
ties produced by arterial lesions are more appropriately
described in terms of changes in flow velocity, pressure
gradients, and energy loss. The use of objective hemodynamic
measurements to assess the location and severity of arterial
disease is an essential step in the management of patients with
vascular problems. Thus, an understanding of the basic princi-
ples of normal and abnormal arterial flow is a prerequisite for
the correct performance and interpretation of hemodynamic
tests. Doppler ultrasound is an ideal noninvasive method for
characterizing both normal blood flow and the changes pro-
duced by arterial disease. Doppler instruments can be used to
measure blood flow velocity, estimate pressure gradients, and
determine the location of arterial lesions. Duplex scanning,
which combines Doppler flow detection with B-mode imag-
ing, gives additional information on arterial wall anatomy. To
provide a theoretical basis for the clinical use of hemodynamic
measurements, this chapter reviews the principles that govern
normal arterial flow and discusses the characteristic altera-
tions produced by arterial occlusive lesions.

Normal arterial flow

Blood flow patterns in arteries can be described using terms
such as velocity, resistance, pressure, and energy. These flow
patterns are influenced by a variety of factors, including car-
diac events, arterial wall compliance, vascular smooth muscle
tone, and vessel geometry.

Pressure and flow relationships

Pressure is defined as force per unit area and is given in units of
dynes/cm 2 or mmHg (ImmHg = 1333 dynes/cm 2 ). The in-
travascular arterial pressure (P) is determined by the dynamic
pressure of cardiac contraction, hydrostatic pressure, and the
static filling pressure. Hydrostatic pressure represents the

weight of a column of blood and has the expression -pgh,
where p is the specific gravity of blood (1.056 g/cm 3 ), g is the
acceleration due to gravity (980 cm/ s 2 ), and h is the distance
above or below the reference level of the right atrium. Static
filling pressure is the pressure that would exist in the arterial
system without cardiac contraction and is related to the vol-
ume of blood and the elastic properties of the vessel wall. The
static filling pressure is typically in the range of 5-10 mmHg.

Although it may appear that pressure gradients drive the
circulation, blood flows through the vascular system accord-
ing to differences in total fluid energy. This total fluid energy
(E) can be divided into potential (Ep) and kinetic (Ek) compo-
nents. Potential energy consists of intravascular pressure (P)
and the gravitational potential energy (+pgh) that represents
the ability of blood to do work because of its height above a
specific reference level. Thus, potential energy can be ex-
pressed as:

E P = P + (pgh)

(1)

Because the gravitational potential energy and hydrostatic
pressure cancel each other out when measurements are made
in the supine position, and the static filling pressure is rela-
tively low, the dynamic pressure of cardiac contraction is the
predominant component of Ep. Kinetic energy represents the
ability of blood to do work on the basis of its motion and is pro-
portional to the specific gravity of blood (p) and the square of
blood velocity v:

Ek = y 2 pv :

(2)

An expression for total fluid energy is obtained by combining
equations 1 and 2:

E = P + ( P gh) + (V 2 pv 2 )

(3)

Bernoulli's principle states that energy is conserved when
fluid flows, and its total energy remains constant, provided
that flow is steady and there are no frictional energy losses.
Ideal flow conditions, however, are not present in the arterial
system, and a portion of the total fluid energy is lost, primarily
as heat. In this situation, the Bernoulli equation is:

295

pa rt 1 1 Noninvasive vascular diagnostics

Pi + Pgh + y 2 pVi 2 = P 2 + pgh 2 + V 2 pv 2 2 + heat

(4)

Viscosity describes the resistance to flow resulting from the
intermolecular attractions between layers of fluid. The coeffi-
cient of viscosity (r|) is defined as the ratio of shear stress (x) to
shear rate (D):

r| =

(5)

Shear stress results from friction between adjacent fluid layers,
and shear rate refers to the relative velocity of these layers. The
concentration of red cells, or hematocrit, is the most important
determinant of blood viscosity, with viscosity increasing expo-
nentially as hematocrit increases. Plasma viscosity depends
primarily on the concentration of proteins. Because blood is a
suspension of cells and large protein molecules, viscosity
varies with the shear rate or flow velocity. This characteristic is
referred to as non-Newtonian. The viscosity of blood increases
at low shear rates but approaches a constant value at higher
shear rates.

Poiseuille's law describes the relationship between pressure
and flow in an idealized fluid system. It states that when fluid
of viscosity (r|) flows through a tube with length L and radius r,
the pressure difference P 1 - P 2 is proportional to volume flow
Q, as follows:

P.-P^Q 8 -^

(6)

The strict application of Poiseuille's law requires a straight,
rigid cylindrical tube and steady flow of a Newtonian fluid.
Because these conditions are not present in the arterial system,
equation 6 will predict only the minimum pressure gradients
that may exist.

Arterial flow patterns

A laminar flow pattern develops under the steady-state condi-
tions specified by Poiseuille's law. In laminar flow all motion is
parallel to the walls of the tube and the fluid is arranged in con-
centric layers or laminae, each of which has a constant flow
velocity. The velocity is lowest adjacent to the tube wall and in-
creases toward the center of the tube, creating a parabolic flow
profile.

Turbulent flow is characterized by chaotic or random move-
ments of the blood elements and a rectangular or blunt flow
profile. These random velocity changes can result in signifi-
cant losses of fluid energy. The transition from laminar to tur-
bulent flow depends on the velocity of flow (v), along with the
viscosity (r|) and specific gravity (p) of the blood and the vessel
diameter (d). These factors can be expressed as a dimension-
less quantity called the Reynolds number (Re):

Re =

dvp

(7)

Turbulence tends to occur at Reynolds numbers greater than
2000. Reynolds numbers are usually less than 2000 in normal
arterial flow, although higher Reynolds numbers are known to
occur in the ascending aorta. With arterial disease, turbulence
often is present during systole immediately distal to a focal
stenosis where the critical value of the Reynolds number is
exceeded.

Arterial geometry plays an important role in determining
the velocity profile and other features of the flow pattern. A
converging tube tends to stabilize laminar flow while flatten-
ing the velocity profile. Flow in a diverging tube becomes less
stable, with an elongated velocity profile and an increased ten-
dency toward turbulence. As blood flows along a curved ves-
sel such as the aortic arch, the rapidly moving blood at the
center of the lumen is accelerated toward the outer edge of the
curve. This can result in complex helical flow patterns that in-
clude a skewing of the velocity profile.

The region of relatively low-velocity flow adjacent to the
vessel wall is referred to as the boundary layer. This layer is af-
fected by both frictional interactions with the wall and viscous
forces from the higher flow velocities located toward the cen-
ter of the vessel. The specific size and shape of the boundary
layer vary according to local vessel geometry and flow condi-
tions. At points where arteries curve, branch, or change diam-
eter, pressure gradients can occur that cause the boundary
layer to stop or reverse direction. This results in a localized
flow pattern called an area of flow separation or boundary
layer separation. Flow separation has been observed in
models of arterial anastomoses and bifurcations. 1-4 In models
of the carotid artery bifurcation, a complex area of flow separa-
tion is found along the outer wall of the bulb. A more laminar
flow pattern is present in the distal internal carotid artery.
These areas of flow separation also have been detected in
human subjects by pulsed Doppler and color-flow imaging
studies (Fig. 28. 1). 5-7 Because the vascular endothelium ad-
jacent to an area of flow separation would be subject to
relatively low or oscillating shear stress, the flow pattern in the
boundary layer may contribute to the initiation and develop-
ment of atherosclerotic plaques. 8 In the carotid bifurcation, in-
timal thickening and atherosclerotic plaque formation tend to
occur along the outer wall of the bulb, whereas the inner wall
along the flow divider is relatively spared. 9 ' 10

Pulsatile pressure and flow changes

A portion of the energy from each cardiac systole distends the
large proximal arteries that store both blood volume and ener-
gy. These can then be returned to the circulation in diastole. Al-
though the large arteries of the thorax and abdomen are highly
elastic, the arteries become progressively stiffer as the blood
moves through the arterial system. This results from a relative
decrease in the amount of elastic tissue in the arterial wall and
a corresponding increase in the quantity of collagen. Because
the speed of propagation for the arterial pressure wave in-

296

CHAPTER28 Physiologic basis of hemodynamic measurement

Figure 28.1 Region of flow separation along
the outer wall of the normal carotid bulb shown
by pulsed Doppler spectral analysis. The flow
pattern near the apical divider (A) is forward
throughout the cardiac cycle, but adjacent to
the outer wall (B) the spectral waveform
contains both forward (positive) and reverse
(negative) components. This low-velocity,
oscillating flow pattern indicates an area of flow
separation. (Courtesy of David J. Phillips, PhD,
University of Washington, Seattle.)

F d
kHz 1

3-1
2-

-w

JSmJ^jSm

3"'
2"'
1 -'

o -■

* i» t

Apical divider

Outside wall

opposite apical

divider

creases with the stiffness of the arterial wall, the pressure wave
travels faster as it moves along the arterial tree. Thus, the pres-
sure and flow waves generated by cardiac contraction are
altered as they traverse the arterial system by the elasticity of
the proximal arteries and the stiffness of the more peripheral
vessels.

The large and medium-sized arteries offer relatively little re-
sistance to normal flow, so the mean pressure drop between
the aorta and the small arteries of the limbs is minimal. 11 The
systolic pressure and pulse pressure (difference between sys-
tolic and diastolic pressures), however, increase as the pres-
sure wave moves distally owing to the effects of reflected
waves and the increasing stiffness of the arterial walls. Re-
flected waves originate at points of branching or changes in
vessel diameter, and are augmented by increases in peripheral
resistance. 11 Consequently, in the arteries of the limbs, pul-
satile pressure changes are enhanced by vasoconstriction and
diminished by vasodilation.

The pulsatile variations of the pressure wave are associated
with corresponding changes in the flow velocity. Although
storage of energy in the elastic arterial walls tends to promote
continuous forward flow throughout the cardiac cycle, cessa-
tion of forward flow or reversal of flow in diastole normally oc-
curs in certain segments of the arterial system. Diastolic
reverse flow is most prominent in arteries with a high periph-
eral resistance, so it tends to be reduced or absent in low-
resistance vessels and those with a large mean forward flow
component. Because of the low resistance of the normal cere-

bral circulation, the internal carotid artery shows continuous
forward flow through systole and diastole. In contrast, the ex-
ternal carotid circulation has a high peripheral resistance, and
diastolic reverse flow usually is present.

Flow patterns in arterial disease

Energy losses in arterial stenoses

The energy lost as blood flows through the arterial system is
the result of viscous losses from the friction between adjacent
layers of blood and inertial losses related to changes in the
velocity and direction of flow. Poiseuille's law indicates
that viscous energy losses in an arterial stenosis are directly
proportional to its length and inversely proportional to the
fourth power of the radius (equation 6). Thus, the radius of a
stenosis is the predominant factor determining viscous energy
losses. Inertial energy losses are related to changes in kinetic
energy and are proportional to the square of blood velocity
(equation 2). The high-velocity jets and turbulence produced
by arterial stenoses contribute substantially to inertial losses.
Thus, in arterial lesions, inertial energy losses usually exceed
viscous energy losses. The viscous losses originate within the
stenotic segment, as specified by Poiseuille's law. Inertial loss-
es predominate at the entrance (contraction effects) and exit
(expansion effects) of a stenosis. Poststenotic turbulence is an
expansion effect that can result in considerable energy loss.

297

pa rt 1 1 Noninvasive vascular diagnostics

Hemodynamic resistance (R) can be expressed as the ratio of
the energy drop between two points in the arterial system (E 1 -
E 2 ) and volume flow (Q):

R =

E 1 -E 1
Q

(8)

If the pressure gradient (P 1 - P 2 ) is substituted in equation 8 as
an approximation for energy drop, then the resistance term
can be taken from Poiseuille's law:

R =

8Lrj

(9)

The actual hemodynamic resistance increases as blood velo-
city increases, even when the lumen size remains constant,
and these additional energy losses are related to inertial effects.
In the normal arterial circulation, over 90% of the total vascu-
lar resistance is related to flow through the arteries, arterioles,
and capillaries, whereas the remaining portion is related to ve-
nous flow. The large and medium-sized arteries, however, ac-
count for only about 15% of the total. 12 Therefore, the vessels
that are most commonly affected by arterial occlusive disease
normally have a very low hemodynamic resistance.

When arterial obstruction is present, a network of collateral
vessels is recruited to bypass the diseased segment. A major
stimulus for collateral development is the abnormal pressure
gradient that exists across the collateral system. 13 Because col-
lateral vessels are smaller, longer, and more tortuous than the
major arteries they replace, the resistance of the collateral net-
work is always higher than that of the corresponding normal
artery. Furthermore, changes in the resistance of the collateral
circulation in response to exercise are minimal, so the high col-
lateral resistance is considered to be fixed.

The hemodynamic resistance of the arteries in the lower
limb can be divided into segmental and peripheral compo-
nents. Segmental resistance includes the fixed parallel resis-
tances of the major arteries and any collateral vessels. The
peripheral resistance consists mainly of the variable resistance
offered by the arterioles in the muscular and cutaneous circu-
lations. In the normal resting state, segmental resistance is
very low, peripheral resistance is relatively high, and the pres-
sure drop across the major segmental arteries is minimal. With
leg exercise, peripheral resistance falls as the muscular arteri-
oles dilate, and blood flow through the segmental arteries in-
creases with little or no pressure drop.

When the major segmental arteries of the lower limb are dis-
eased, segmental resistance increases, and an abnormal pres-
sure gradient appears. Because of a compensatory decrease in
peripheral resistance, however, the total resistance of the limb
and resting limb blood flow usually remain in the normal
range. 14 Thus, during leg exercise, the segmental resistance as-
sociated with collateral flow remains high and fixed, although
the peripheral resistance may decrease further. In this situa-
tion, the capacity of the peripheral circulation to compensate
for the high segmental resistance is limited, so blood flow dur-

ing exercise is lower than normal, and the pressure gradient in-
creases. These alterations in the distribution of hemodynamic
resistance explain the changes in blood pressure and flow ob-
served in the lower limbs of patients with arterial occlusive
disease.

Critical arterial stenosis

The term critical stenosis refers to the degree of arterial narrow-
ing required to produce a significant reduction in distal blood
pressure or flow. 15 Because the viscous energy losses asso-
ciated with a stenosis are inversely proportional to the fourth
power of the radius at the stenotic site, there is an exponential
relationship between pressure drop and lumen size. There-
fore, once the critical stenosis value is reached, distal pressure
and flow will diminish rapidly with any further narrowing of
the lumen (Fig. 28.2). Because blood flow velocity is the major
determinant of inertial energy loss, the critical stenosis value
also depends on the flow rate (Fig. 28.3). An arterial segment
with a high flow velocity (low resistance) shows a pressure
drop with less narrowing than a segment with a lower flow ve-
locity (high resistance). For the large and medium-sized arter-
ies typically affected by atherosclerosis, the critical stenosis
value is approximately a 50% diameter reduction or 75% area
reduction. Because the critical stenosis value depends on the
flow rate, however, a lower limb arterial stenosis that is not sig-
nificant at rest may become critical when flow rates are in-
creased by exercise. These observations provide a physiologic
basis for assessing the severity of arterial lesions by blood
pressure measurements before and after exercise. 16-18

When stenoses that are not critical or hemodynamically sig-
nificant individually are arranged in series, large pressure and
flow reductions can occur. 19 Because the energy losses from a
stenosis are related primarily to inertial effects at the entrance
and exit, multiple short stenoses tend to be more significant
than a single longer stenosis of similar severity. Thus, several
moderate or subcritical stenoses can have the same effect on
distal pressure and flow as a single critical stenosis. When two
stenoses in series have nearly the same diameter, removal of
one will result in a modest increase in blood flow. If stenoses in
series have different diameters, removal of the least severe is
unlikely to improve flow significantly, whereas removal of the
most severe will increase flow substantially.

Clinical applications

Blood flow detection by Doppler ultrasound

The change in frequency of reflected ultrasound that results
from relative motion between the ultrasound source and a re-
flector is referred to as the Doppler effect. In the case of blood
flow detection by ultrasound, the main reflectors are the mov-
ing red blood cells, and the ultrasound source is a stationary

298

CHAPTER28 Physiologic basis of hemodynamic measurement

Figure 28.2 Effect of increasing stenosis on
blood flow and pressure drop (AP) across a
stenotic arterial segment. Peripheral resistance
is considered to be fixed, so autoregulation does
not occur. (From Strandness DE, Sumner DS.
Hemodynamics for Surgeons. New York:
Grune & Stratton, 1 975, with permission from
Elsevier.)

1 .0 cm Long stenosis in 1 cm Long

artery with radius of 0.5 cm

(flow = 0.5cm 3 /s)

Percent stenosis

Flow

cm 3 /s

mmHg

102030 40 50 60 70

80 90 95 100

■ /~»o /"*

100 "

28.6

90 -

x *

\ t

\ r

■ 25.0

Q.
O

80 "

\ /

^ 4.0

MOW

2 3

70 -

\ l
\ *
\ i

- 20.0

■*- en

\ l

mum
pres

60 "

\ t

• 3.0

CO 3

\ i
Vi

* 15.0

E E

50 -

° CC

1 \

t \

E E

40 -

t \
t \

- 2.0

O O

t \
t \

- 10.0

CD £

3- ©

30 -

t \

CD
Q_

20 -
10 -

t \
t \

t \

* \

-1.0

- 5.0

/-\

. n n

L o.o

H
0.

5 0.4 0.3

0.2 0.1 0.0

Inside radius of stenosis, cm

% stenosis

transducer placed on the skin. The Doppler frequency shift (f d )
is a function of the ultrasound transmitting frequency (/" t ),
blood cell velocity (v), the speed of sound in soft tissue (C), and
the cosine of the angle (9) between the ultrasound beam and
the direction of blood flow:

fd -

2f t v cos 9
C

(10)

Because the speed of sound in soft tissue is relatively constant
(approximately 1540 m/s), the Doppler frequency shift is de-
termined by blood cell velocity if the angle is held constant. It
is necessary to use the cosine of 9 to account for the component
of the velocity vector in the direction of the ultrasound beam.
For the special case in which the direction of the ultrasound
beam and the direction of blood flow are the same, 9 = and
cos 6 = 1; however, in most clinical applications the transducer
is positioned to create an angle of 30-60° between the ultra-
sound beam and the presumed direction of blood flow along
the longitudinal axis of the artery. When 9 = 90°, cos 9 = and
there is, theoretically, no Doppler shift. Therefore, beam angles
approaching 90° should be avoided when using Doppler tech-
niques for blood flow detection.

The attenuation of ultrasound as it travels through tissue is
directly proportional to the transmitting frequency. Thus,
higher frequency ultrasound does not penetrate as deeply
into tissue as lower frequency ultrasound. Doppler transmit-

100

102030 40 50 60 70

0.4 0.3 0.2 0.1

Inside radius of stenosis, cm

Figure 28.3 Relationship of pressure drop (AP) across a stenosis to the
severity of the stenosis and the flow velocity. Higher velocities produce
pressure drops with less arterial narrowing than lower velocities, and each
velocity curve has its own "critical stenosis." (From Strandness DE, Sumner
DS. Hemodynamics for Surgeons. New York: Grune & Stratton, 1 975, with
permission from Elsevier.)

299

pa rt 1 1 Noninvasive vascular diagnostics

ting frequencies in the range of 2-10 MHz are suitable for most
peripheral vascular examinations. In general, lower frequen-
cies (5 MHz or less) must be used to evaluate deeply located
vessels such as those in the abdomen. It is a fortunate coinci-
dence that the Doppler frequency shift in most vascular diag-
nostic applications is in the audible range. This permits both
subjective interpretation of the Doppler signal by the ex-
aminer and more objective or quantitative methods such as
analogue waveforms and spectral analysis.

The simplest and least expensive Doppler instruments op-
erate in the continuous-wave (CW) mode, in which ultra-
sound is continuously transmitted from one transducer and
continuously received by another transducer. An important
limitation of C W instruments is that they provide no informa-
tion on the distance between the transducer and the ultra-
sound reflectors, so a complex Doppler signal often is obtained
that represents velocities from all the vessels traversed by the
ultrasound beam. Thus, interpretation of CW Doppler signals
can be difficult when vessels are superimposed within the ul-
trasound beam or there are complex flow disturbances within
a single vessel.

With pulsed Doppler ultrasound, flow is detected at a dis-
crete site called the sample volume that can be positioned any-
where along the axis of the ultrasound beam. Pulsed Doppler
instruments use the principle of range-gating, in which a short
burst of ultrasound is emitted from the transducer and, after
waiting a specific period of time, the same transducer is used
as a receiver to sample the reflected ultrasound from a selected
depth in tissue. The speed of ultrasound in tissue (C) deter-
mines the time needed for a round trip from the transducer to a
particular depth (d) in tissue and back again. This requirement
for a round trip of each ultrasound pulse limits the pulse repe-
tition frequency (PRF), or rate at which pulses may be
transmitted:

PRF '(maximum) =

(11)

A pulsed Doppler measures the frequency shift by sampling
flow from a specific depth at the PRF. With a higher PRF, more
pulses are available to sample flow, and a better representation
of the Doppler frequency-shifted signal is obtained. If the fre-
quency of the Doppler signal is greater than twice the PRF,
however, the output frequency from the instrument will be
anomalously low. The generation of these artifactual, low-
frequency Doppler signals is called aliasing. The Nyquist
frequency is the highest Doppler-shifted frequency that can be
accurately detected by a pulsed Doppler instrument, and is
equal to one-half the PRF.

Both CW and pulsed Doppler instruments can be designed
to distinguish between flow toward and away from the trans-
ducer. Directional capabilities are extremely important for
clinical applications, because both forward and reverse flow
may be present within normal or diseased arteries at various
times during the cardiac cycle.

Doppler signal analysis

Although audible interpretation of the Doppler frequency-
shifted signal may be of diagnostic value to an experienced ex-
aminer, additional information can be obtained by more
sophisticated methods of Doppler signal analysis. A simple
and commonly used method is the zero-crossing detector. This
device is based on a frequency-to-voltage converter that
measures the number of times the signal crosses the zero-volts
line and produces a voltage output that drives a strip-chart
recorder. An analogue waveform is generated that represents
the change in Doppler frequency shift over time, but the out-
put of the zero-crossing detector is not equivalent to the actual
frequency of the Doppler signal. 20 In addition, the zero-
crossing detector may not be able to characterize rapid
changes in blood velocity such as those that occur during pul-
satile flow. Other disadvantages of this method include sus-
ceptibility to electrical noise and amplitude dependency. 21

Spectral analysis

The technique of spectral analysis separates a complex
Doppler signal into its individual frequency components, thus
overcoming many limitations of the analogue zero-crossing
method. A Doppler-shifted signal can be considered as the
sum of a series of single-frequency signals, each having a par-
ticular amplitude and phase. The amplitude of a specific
frequency component is proportional to the number of blood
cells moving through the ultrasound beam that causes that
particular frequency shift. Spectral information generally is
presented graphically, with frequency on the vertical axis,
time on the horizontal axis, and amplitude represented by a
gray scale. In contrast to the analogue waveform, which
depicts the Doppler signal as a single line, spectral analysis
displays the entire frequency and amplitude content of the
Doppler signal.

Techniques for real-time spectral analysis include parallel
filter systems and the fast Fourier transform (FFT). 22 Most
available instruments use a digital FFT that provides a spectral
analysis on sequential signal segments of approximately 5 ms
duration. This approach produces a wide dynamic range and
gives a continuous display of frequency and amplitude. The
appearance of the Doppler spectral waveform is determined
in part by the portion of the flow stream that is sampled by the
ultrasound beam. Because the flow profile within an artery
may contain a wide range of blood cell velocities, CW and
pulsed Doppler systems produce spectra with different char-
acteristics. Because the sample site extends over the entire
ultrasound beam, C W Doppler spectra tend to be complex and
difficult to interpret. The relatively small sample volume of a
pulsed Doppler system permits evaluation of flow patterns at
specific sites in the arterial lumen, and is more suitable for
spectral analysis. In general, samples should be obtained from
the center of the flow stream, because the laminar flow pattern

300

CHAPTER28 Physiologic basis of hemodynamic measurement

Table 28.1 Criteria for classification of internal carotid artery disease by
spectral analysis of pulsed Doppler signals

Table 28.2 Criteria for classification of lower extremity arterial lesions by
spectral analysis of pulsed Doppler signals

Arteriographic lesion

Spectral criteria

Arteriographic lesion Spectral criteria

A. Normal

Peak systolic frequency less than 4 kHz;
no spectral broadening

B. 1-15% diameter reduction Peak systolic frequency less than

4 kHz; spectral broadening in
deceleration phase of systole only

C. 16-49% diameter
reduction

D. 50-79% diameter
reduction

D+. 80-99% diameter
reduction

Peak systolic frequency less than 4 kHz;
spectral broadening throughout systole

Peak systolic frequency greater than or
equal to 4 kHz; end-diastolic frequency
less than 4.5 kHz

End-diastolic frequency greater than or
equalto4.5kHz

E. Occlusion (100% diameter No internal carotid flow signal; flow to
reduction) zero in common carotid artery

Criteria are based on a pulsed Doppler with a 5-MHz transmitting frequency,
a sample volume that is small relative to the internal carotid artery, and a 60°
beam-to-vessel angle of insonation. Approximate angle-adjusted velocity
equivalents are: 4 kHz = 125 cm/sand 4.5 kHz = 140cm/s.

normally present at that site is readily distinguished from the
disturbed or turbulent flow patterns associated with arterial
lesions.

The application of Doppler spectral analysis to the diagno-
sis of arterial disease is based on the observation that specific
lesions give rise to characteristic flow patterns. Because
normal center-stream arterial flow is relatively laminar, the
normal spectral waveform shows a narrow band of frequen-
cies, particularly during the high-velocity systolic phase of the
cardiac cycle. Stenoses and wall irregularities caused by ather-
osclerotic plaques disrupt this laminar pattern and produce
more random movements of the blood cells. The resulting
flow disturbances create spectra with a wider range of fre-
quencies and amplitudes, and this widening of the frequency
band is referred to as spectral broadening. With minor lesions
that do not affect distal pressure or flow, spectral broadening
occurs in late systole and early diastole; severe lesions with
marked turbulence result in spectral broadening throughout
the cardiac cycle. Hemodynamically significant or critical
stenoses narrow the arterial diameter by 50% or more and pro-
duce localized high-velocity jets that appear in the spectral
waveform as increased peak systolic frequencies. These flow
disturbances are present only in close proximity to the respon-
sible lesion, so accurate detection of arterial disease by spectral
waveform analysis requires that flow be sampled at or very
near the site of the lesion. For example, the spectral features as-
sociated with high-velocity jets and poststenotic turbulence

Normal

1-19% diameter
reduction

20-49% diameter
reduction

50-99% diameter
reduction

Occlusion

Triphasic waveform, no spectral broadening

Triphasic waveform with minimal spectral
broadening only; peak systolic velocities
increased less than 30% relative to the adjacent
proximal segment; proximal and distal waveforms
remain normal

Triphasic waveform usually maintained, although
reverse flow component may be diminished;
spectral broadening is prominent with filling-in of
the clear area underthe systolic peak; peak
systolic velocity is increased from 30% to 100%
relative to the adjacent proximal segment;
proximal and distal waveforms remain normal

Monophasic waveform with loss of the reverse
flow component and forward flow throughout
the cardiac cycle; extensive spectral broadening,
peak systolic velocity is increased over 1 00%
relative to the adjacent proximal segment; distal
waveform is monophasic with reduced systolic
velocity

No flow detected within the imaged arterial
segment; preocclusive "thump" may be heard
just proximal to the site of occlusion; distal
waveforms are monophasic with reduced systolic
velocities

are most prominent within and immediately distal to a
stenosis. 23

Clinical studies correlating various spectral characteristics
with the results of contrast arteriography have been used to
develop classifications for disease involving specific segments
of the arterial system. Spectral analysis criteria for classifica-
tion of carotid and lower extremity arterial disease by duplex
scanning are summarized in Tables 28.1 and 28.2. The criteria
for carotid stenoses can distinguish between normal and dis-
eased internal carotid arteries with a specificity of 84% and a
sensitivity of 99%; the accuracy for detecting 50-99% diameter
stenosis or occlusion is 93%. 24-26 For identifying lower extrem-
ity arterial lesions that produce a significant pressure gradient
or greater than 50% diameter reduction at the time of arterio-
graphy, spectral analysis has a sensitivity of 82%, a specificity
of 92%, a positive predictive value of 80%, and a negative pre-
dictive value of 93%. 27 Duplex scanning with spectral analysis
is particularly useful for assessing the aortoiliac segment, a
portion of the arterial system difficult to evaluate by any other
noninvasive method. Significant iliac artery stenoses are
detected with a sensitivity of 89% and a specificity of 90%.

301

pa rt 1 1 Noninvasive vascular diagnostics

Color-flow imaging

The real-time, color-flow image is an alternative to spectral
analysis for displaying the pulsed Doppler information
obtained by duplex scanning. Whereas spectral analysis
evaluates the entire frequency and amplitude content of the
pulsed Doppler signal at a selected arterial site, color-flow
imaging provides a single estimate of the Doppler shift fre-
quency or flow velocity for each site within the B-mode image.
Thus, spectral waveforms actually contain more information
on the flow pattern at each individual site than the color-flow
image. The principal advantage of the color-flow display is
that it presents flow information on the entire image in real
time, although the absolute amount of data for each site is
reduced. Because of these differences, it is difficult to compare
the Doppler information from spectral waveforms and color-
flow imaging. Spectral waveforms contain a range of frequen-
cies and amplitudes, allowing determination of flow direction
and frequency parameters such as mean, mode, and peak. In
contrast, color assignments are based on flow direction and a
single mean or average frequency estimate. Consequently, the
peak or maximum Doppler frequency shifts seen with spectral
waveforms generally are higher than the frequencies in-
dicated by color-flow imaging.

The addition of color-flow imaging can facilitate certain as-
pects of the arterial duplex examination. Color-flow imaging
is extremely helpful for identifying vascular structures, espe-
cially when they are deeply located, such as the abdominal
vessels, or small like the arteries below the knee. 28-30 It is diffi-
cult, however, to assess the severity of arterial lesions based on
the color-flow image alone, and spectral waveforms remain
necessary for the most accurate classification of arterial
disease. 29

20 1

Normal

<? 20 n

Stenotic iliac

20 1

Occluded iliac

Figure 28.4 Velocity waveforms obtained from the common femoral
artery of a normal subject, a patient with external iliac artery stenosis, and a
patient with common iliac artery occlusion. The normal waveform is
triphasic; the abnormal waveforms are monophasic and damped, with
the most extreme changes noted distal to the occluded arterial segment.
(From Strandness DE, Sumner DS. Hemodynamics for Surgeons. New York:
Grune & Stratton, 1 975, with permission from Elsevier.)

Velocity waveforms

The shape or frequency envelope of the Doppler waveform,
obtained with either a zero-crossing detector or spectrum ana-
lyzer, can be used as a guide to the severity of arterial occlusive
disease. A triphasic flow pattern is normally present in the
major arteries of the limbs. This consists of an initial, high-
velocity, forward flow phase during cardiac systole, a brief
phase of reverse flow in early diastole, and a low-velocity, for-
ward flow phase in late diastole. The factors that modify this
normal pattern include the presence of arterial occlusive dis-
ease and changes in peripheral vascular resistance. For exam-
ple, body warming diminishes the second phase of flow
reversal as a result of vasodilation and decreased peripheral
resistance. Exposure to cold causes vasoconstriction with in-
creased resistance, and the reverse flow phase becomes more
prominent. An arterial stenosis acts like a filter that removes
rapidly changing components of the flow velocity waveform.
This, together with the compensatory decrease in peripheral
resistance that occurs distal to a stenosis, results in the disap-

pearance of the reverse flow phase. Thus, waveforms obtained
distal to stenotic lesions are described as monophasic with a
single forward velocity phase and a low, rounded peak (Fig.
28.4). Waveforms taken from within hemodynamically signif-
icant stenoses have an abnormally high peak systolic velocity
that represents the high-velocity jet generated by the lesion.
Loss of the reverse flow phase and increased diastolic forward
flow have been related to the severity of a stenosis. 31 The fea-
tures of waveforms taken proximal to stenotic lesions are vari-
able and depend on the capacity of the intervening collateral
circulation. If the collateral system is well developed, the
proximal waveform may appear normal.

Velocity waveforms from various segments of the arterial
system have certain recognizable characteristics. Arteries that
supply low-resistance organs, such as the internal carotid,
renal, and celiac, normally show forward flow throughout the
cardiac cycle and relatively high diastolic flow velocities.
Some waveforms may change in response to physiologic

302

CHAPTER28 Physiologic basis of hemodynamic measurement

stimuli. In the fasting state, superior mesenteric artery wave-
forms are triphasic with early diastolic reverse flow, but
mesenteric vasodilation in the postprandial state results in
loss of diastolic reverse flow. 32

Indirect blood pressure measurements

When blood flows through an arterial stenosis, the distal pulse
pressure is reduced to a greater extent than the mean pres-
sure. 33 This indicates that systolic pressure is the component of
the pressure pulse most sensitive to the presence of hemo-
dynamically significant arterial lesions. Because the peak sys-
tolic pressure is amplified as the pulse wave proceeds away
from the heart, the systolic pressure measured at the ankle is
normally higher than that taken from the upper arm. These
pressures are measured easily with a pneumatic cuff and
Doppler flow detector. 34 Although the reduction in ankle sys-
tolic pressure is proportional to the overall severity of arterial
occlusive disease in the lower limb, ankle blood pressure also
varies with the central aortic pressure. Because subclavian and
axillary artery occlusive disease is uncommon, however, the
brachial artery systolic pressure closely approximates the
central aortic pressure. The ratio of ankle systolic pressure to
brachial systolic pressure, referred to as the ankle pressure index,
compensates for variations in central aortic pressure and facil-
itates comparisons between measurements taken at different
times. 35 The normal ankle pressure index has a mean value of
1.11 + 0.10. In lower limbs with intermittent claudication, the
mean value of the index is 0.59 + 0.15. 36 The ankle pressure
index does not indicate the exact location or relative severity of
lesions at multiple levels in the lower extremity. Such infor-
mation can be obtained with segmental pressure cuffs or by
duplex scanning. 34

Another approach to the indirect assessment of pressure
gradients is to use parameters derived from Doppler flow ve-
locity waveforms. An example of this is the pulsatility index
(PI), which is calculated by dividing the peak-to-peak frequen-
cy difference by the mean frequency. These measurements can
be based on either analogue or spectral waveforms. A close
correlation has been observed between reduction in PI and the
severity of arterial occlusive disease as documented by arteri-
ography and measurement of ankle pressure index. 37 The nor-
mal PI of the common femoral artery has a mean value of 6.7.
When the common femoral artery PI was compared with
intraarterial pressure measurement, a PI of 4.0 or greater was
highly predictive of a hemodynamically normal aortoiliac
segment. 38 If the superficial femoral artery was patent, a PI less
than 4.0 indicated a hemodynamically significant aortoiliac le-
sion; however, a PI less than 4.0 with an occluded superficial
femoral artery was not diagnostic.

A more direct method for the assessment of arterial pressure
gradients by Doppler ultrasound is the modified Bernoulli
equation. 39 This relationship between the pressure gradient
across a lesion (P 1 - P 2 ) and the maximal velocity of the

stenotic jet (V max ) has been used to estimate the gradients
across cardiac valves:

Pi-P 2 =W I

max

(12)

Because stenotic cardiac valves represent very short lesions,
viscous forces within the stenosis are negligible. Assuming
that the kinetic energy of the stenotic jet is completely dissipat-
ed beyond the stenosis in a region of turbulence, equation 12
gives the maximum pressure gradient across the lesion. Al-
though this approach generally works well for cardiac valves,
it may overestimate the pressure gradients across subcritical
peripheral arterial stenoses that do not produce turbulence
and therefore do not result in complete loss of the kinetic
energy in the stenotic jet. In this situation, some of this kinetic
energy is presumably converted back to pressure energy be-
yond the stenosis. Equation 12 also may underestimate the
pressure gradients associated with long, high-grade stenoses
where viscous energy losses are significant. The modified
Bernoulli equation does not account for lesion length, taper-
ing, surface irregularity, branching, or tortuosity. All these fac-
tors are likely to be important with peripheral atherosclerotic
lesions, and could limit the application of this method. Al-
though the modified Bernoulli equation may be of some value
in assessing peripheral artery lesions, experimental and clini-
cal studies generally have confirmed the disadvantages dis-
cussed earlier. 39 ' 40 The absence of diastolic reverse flow in the
waveform taken at the site of stenosis has shown a strong
correlation with a pressure gradient greater than 15 mmHg. 39

Direct measurement of arterial blood pressure

As discussed, the relationship between pressure drop, flow,
and resistance is expressed by Poiseuille's law, and the degree
of narrowing at which pressure and flow begin to decline is
called the critical stenosis. In the intact arterial circulation,
however, autoregulation can maintain normal flow rates dis-
tal to a critical stenosis, even when a significant pressure drop
is present. Therefore, pressure measurements are more likely
to reflect accurately the presence of arterial disease than flow
measurements. Furthermore, measurements of flow rates or
peripheral resistance are extremely difficult to perform in the
clinical setting.

The direct measurement of arterial pressure avoids the
potential errors associated with noninvasive pressure mea-
surements. Direct pressure measurements have been applied
primarily to the assessment of lower extremity arterial dis-
ease. Specific approaches include pull-through aortoiliac
artery pressures during arteriography and percutaneous mea-
surement of common femoral artery pressures. As with the
indirect noninvasive methods, direct pressure measurements
can be made both in the resting state and after some form of
hemodynamic stress. A pedal ergometer exercise test has been
described for use with percutaneous common femoral artery
pressure measurements; however, a large proportion of

303

pa rt 1 1 Noninvasive vascular diagnostics

patients are unable to perform this test, and it has not been
widely used. 41 A simpler technique that does not require
strict patient cooperation or any specialized equipment is
intraarterial injection of papaverine to produce peripheral
vasodilation.

Direct pressure measurement can be used to assess the
physiologic severity of aortoiliac disease found either on arte-
riography or noninvasive testing. Although arteriography
usually is adequate for evaluating the significance of infrain-
guinal arterial disease, the same is not true for more proximal
arterial lesions. 17 Even biplane arteriography may not allow
an accurate assessment of the aortoiliac system. 42 Because ar-
teriographic procedures most commonly are performed using
a femoral puncture site, direct measurements of arterial pres-
sure during arteriography generally include the aortic, iliac,
and femoral segments. Consequently, pull-through pressures
taken with the arteriogram catheter indicate the hemodynam-
ic significance of any lesions present in the aortoiliac system.
Intraarterial injection of papaverine can be used as a pharma-
cologic stress test to assess the pressure gradients during high-
flow conditions. Studies of hemodynamically normal patients
suggest that a hemodynamically significant lesion in the aor-
toiliac segment is present when the systolic pressure gradient
is more than lOmmHg at rest or 20mmHg after injection of
papaverine hydrochloride (30 mg) into the arteriogram
catheter. 38

Direct measurement of femoral artery pressure is per-
formed by percutaneous puncture of the common femoral
artery with a 19-G needle attached by rigid, fluid-filled tubing
to a calibrated pressure transducer. The femoral artery systolic
pressure is compared with the brachial artery systolic pres-
sure, and the femoral brachial index (FBI) is calculated. As for the
ankle pressure index, the brachial artery pressure, as mea-
sured by Doppler ultrasound, is presumed to represent the
central aortic pressure. A resting FBI of greater than or equal to
0.9 is considered normal. 42 Values less than 0.9 indicate the
presence of a hemodynamically significant lesion proximal to
the common femoral artery. If the resting FBI is normal, the in-
jection of papaverine can be used to look for less severe lesions
that are apparent only at increased flow rates. This is ac-
complished by injecting 30 mg of papaverine hydrochloride
directly through the needle in the common femoral artery and
monitoring both the common femoral and brachial artery
pressures. It is particularly important to measure the brachial
artery pressure during this test, because papaverine often
causes a slight decrease in systemic arterial pressure. The
mean decrease in FBI after papaverine injection is 6% for nor-
mal subjects, and a decrease of 15% or more is indicative of a
hemodynamically significant lesion. 42 Apeak flow increase of
50% or greater is sufficient for a valid test; reasons for an in-
valid test include fixed outflow resistance and extravascular
injection of papaverine.

Conclusions

The hemodynamic measurements used in the assessment of
arterial disease are based on the physiologic relationships
between vessel anatomy, blood flow patterns, and changes in
blood pressure. Duplex scanning with spectral analysis has
emerged as the preferred method for most noninvasive vascu-
lar evaluations. Although duplex scanning initially was devel-
oped as a means for examining the carotid artery bifurcation,
advances in technology and clinical experience have broad-
ened the applications to include the lower limb, renal, and
mesenteric vessels. 7 ' 24,27,32,43,44 The clinical role of direct arteri-
al blood pressure measurement has diminished as the
diagnostic capabilities of noninvasive testing have expanded.
Direct pressure measurements, however, are still regarded as
the reference standard for the physiologic evaluation of
peripheral arterial disease and will continue to be a valuable
adjunct, particularly when invasive diagnostic or therapeutic
procedures are required.

References

1. Bharadvaj BK, Mabon RF, Giddens DP. Steady flow in a model of
the human carotid bifurcation. I. Flow visualization. / Biomech
1982; 15:349.

2. Ku DN, Giddens DP. Pulsatile flow in a model carotid bifurcation.
Arteriosclerosis 1983; 3:31.

3. LoGerf o FW, Nowak MD, Quist WC. Structural details of bound-
ary layer separation in a model human carotid bifurcation under
steady and pulsatile flow conditions. / Vase Surg 1985; 2:263.

4. LoGerfo FW, Soncrant T, Teel T et ah Boundary layer separation in
models of side-to-end anastomoses. Arch Surg 1979; 114:1369.

5. Phillips DJ, Greene FM, Langlois Y et ah Flow velocity patterns
in the carotid bifurcations of young, presumed normal subjects.
Ultrasound Med Biol 1983; 9:39.

6. Ku DN, Giddens DP, Phillips DJ et ah Hemodynamics of the
normal human carotid bifurcation: in vitro and in vivo studies.
Ultrasound Med Biol 1985; 11:13.

7. Zierler RE, Phillips DJ, Beach KW et ah Noninvasive assessment
of normal carotid bifurcation hemodynamics with color-flow
ultrasound imaging. Ultrasound Med Biol 1987; 13:471.

8. Fox JA, Hugh AE. Localization of atheroma: a theory based on
boundary layer separation. Br Heart J 1966; 28:388.

9. Zarins CK, Giddens DP, Bharadvaj BK et ah Carotid bifurcation
atherosclerosis: quantitative correlation of plaque localization
with flow velocity profiles and wall shear stress. Circ Res 1983;
53:502.

10. McMillan DE. Blood flow and the localization of atherosclerotic
plaques. Stroke 1985; 16:582.

11. Carter SA. Effect of age, cardiovascular disease, and vasomotor
changes on transmission of arterial pressure waves through the
lower extremities. Angiology 1978; 29:601.

12. Burton AC. Physiology and Biophysics of the Circulation, 2nd edn.
Chicago: Year Book Medical Publishers, 1972.

304

CHAPTER28 Physiologic basis of hemodynamic measurement

13. John HT, Warren R. The stimulus to collateral circulation. Surgery
1961;49:14.

14. Sumner DS, Strandness DE Jr. The effect of exercise on resistance
to blood flow in limbs with an occluded superficial femoral artery.
Vase Surg 1970; 4:229.

15. Berguer R, Hwang NHC. Critical arterial stenosis: a theoretical
and experimental solution. Ann Surg 1974; 180:39.

16. Carter SA. Response of ankle systolic pressure to leg exercise
in mild or questionable arterial disease. N Engl J Med 1972; 287:
578.

17. Moore WS, Hall AD. Unrecognized aortoiliac stenosis: a physio-
logic approach to the diagnosis. Arch Surg 1971; 103:633.

18. Sumner DS, Strandness DE Jr. The relationship between calf
blood flow and ankle blood pressure in patients with intermittent
claudication. Surgery 1969; 65:763.

19. Flanigan DP, Tullis JP, Streeter VL et ah Multiple subcritical arterial
stenosis: effect on poststenotic pressure and flow. Ann Surg 1977;
186:663.

20. Lunt MJ. Accuracy and limitations of the ultrasonic Doppler blood
velocimeter and zero crossing detector. Ultrasound Med Biol 1975;
2:1.

21. Johnston KW, Maruzzo BC, Cobbold RSC. Errors and artifacts
of Doppler flowmeters and their solution. Arch Surg 1977;
112:1335.

22. Zierler RE, Roederer GO, Strandness DE Jr. The use of frequency
spectral analysis in carotid artery surgery. In: Bergan JJ, Yao JST,
eds. Cerebrovascular Insufficiency. New York: Grune & Stratton,
1983:137.

23. Thiele BL, Hutchison KJ, Greene FM et ah Pulsed Doppler wave-
form patterns produced by smooth stenosis in the dog thoracic
aorta. In: Taylor DEM, Stevens AL, eds. Blood Flow Theory and
Practice. New York: Academic Press, 1983:85.

24. Fell G, Phillips DJ, Chikos PM et ah Ultrasonic duplex scanning for
disease of the carotid artery. Circulation 1981; 64:1191.

25. Langlois YE, Roederer GO, Chan AW et ah Evaluating carotid
artery disease: the concordance between pulsed Doppler/
spectrum analysis and angiography. Ultrasound Med Biol 1983;
9:51.

26. Roederer GO, Langlois YE, Chan AW et ah Ultrasonic duplex
scanning of extracranial carotid arteries: improved accuracy
using new features from the common carotid artery. / Cardiovasc
Ultrasonogr 1982; 1:373.

27. Kohler TR, Nance DR, Cramer MM et ah Duplex scanning for
diagnosis of aortoiliac and femoropopliteal disease: a prospective
study. Circulation 1987; 76:1074.

28. Hatsukami TS, Primozich J, Zierler RE et ah Color-Doppler charac-

teristics in normal lower extremity arteries. Ultrasound Med Biol
1992; 18:167.

29. Hatsukami TS, Primozich J, Zierler RE et ah Color-Doppler imag-
ing of infrainguinal arterial occlusive disease. / Vase Surg 1992;
16:527.

30. Moneta GL, Yeager RA, Antonovic R et ah Accuracy of lower
extremity arterial duplex mapping. / Vase Surg 1992; 15:275.

31. Nicholls SC, Kohler TR, Martin RL et ah Diastolic flow as a predic-
tor of arterial stenosis. / Vase Surg 1986; 3:498.

32. Jager K, Bollinger A, Valli C et ah Measurement of mesenteric
blood flow by duplex scanning. / Vase Surg 1986; 3:462.

33. Keitzer WF, Fry WT, Kraft RO et ah Hemodynamic mechanism for
pulse changes seen in occlusive vascular disease. Surgery 1965;
57:163.

34. Zierler RE, Strandness DE Jr. Doppler techniques for lower
extremity arterial diagnosis. In: Zwiebel WJ, ed. Introduction to
Vascular Ultrasonography, 2nd edn. Orlando: Grune & Stratton,
1986:305.

35. Yao JST, Hobbs JT, Irvine WT. Ankle systolic pressure measure-
ments in arterial diseases affecting the lower extremities. Br J Surg
1969; 56:676.

36. Yao JST. Hemodynamic studies in peripheral arterial disease. Br]
Surgl970;57:761.

37. Johnston KW, Taraschuk I. Validation of the role of pulsatility
index in quantitation of the severity of peripheral arterial occlu-
sive disease. Am J Surg 1976; 131:295.

38. Thiele BL, Bandyk DF, Zierler RE et ah A systematic approach to
the assessment of aortoiliac disease. Arch Surg 1983; 118:477.

39. Kohler TR, Nicholls SC, Zierler RE et ah Assessment of pressure
gradient by Doppler ultrasound: experimental and clinical obser-
vations. / Vase Surg 1987; 6:460.

40. Langsfeld M, Nepute J, Hershey FB et ah The use of deep duplex
scanning to predict hemodynamically significant aortoiliac
stenoses. / Vase Surg 1988; 7:395.

41. Sobinsky KR, Williams LR, Gray G et ah Supine exercise testing in
the selection of suprainguinal versus infrainguinal bypass in
patients with multisegmental arterial occlusive disease. Am } Surg
1986; 152:185.

42. Flanigan DP, Williams LR, Schwartz JAet ah Hemodynamic evalu-
ation of the aortoiliac system based on pharmacologic vasodilata-
tion. Surgery 1983; 93:709.

43. Taylor DC, Strandness DE Jr. Carotid artery duplex scanning.
J Clin Ultrasound 1987; 15:635.

44. Taylor DC, Kettler MD, Moneta GL et ah Duplex ultrasound
scanning in the diagnosis of renal artery stenosis: a prospective
evaluation. / Vase Surg 1988; 7:363.

305

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Spectral analysis

Christopher R.B. Merritt

The development of safe, accurate, and relatively inexpensive
methods to assist in the screening of patients suspected of
having vascular disease has been a major accomplishment of
the past 20 years. Although angiographic techniques remain
the gold standard for confirming diagnoses affecting the peri-
pheral arterial and venous systems, indirect noninvasive
tests, particularly Doppler spectral analysis, now play a major
role in establishing the presence or absence of disease and
determining the degree of stenosis, particularly in the arterial
circulation. Leading these noninvasive methods is duplex
Doppler ultrasonography, combining high-resolution imag-
ing of vessel wall and lumen with Doppler spectrum analysis
to characterize the nature of the flow, identify hemodynamic
disturbances, and detect alterations in organ perfusion.

As noninvasive tests have improved in accuracy, they have
become more competitive with angiography as the ultimate
standard for clinical decision making. A high degree of corre-
lation of duplex ultrasonography with angiography in the pre-
diction of stenosis has been established in numerous studies. 1
Spectral analysis is the foundation for the use of ultrasound in
the evaluation of blood flow. Despite the introduction of new
and graphic methods of showing flow information, such as
Doppler color imaging (DCI), spectral analysis remains an es-
sential component in the ultrasonographic assessment of flow
in vessels throughout the body. To achieve high levels of sensi-
tivity and specificity using Doppler spectral analysis, the basic
principles of Doppler ultrasound must be understood thor-
oughly, along with examination techniques, instrumentation,
and diagnostic criteria. This chapter reviews the basics of
Doppler spectral analysis and its uses.

Basic Doppler principles

Diagnostic ultrasound is based on the detection of echoes scat-
tered by reflecting interfaces within the body. Ultrasound
imaging uses pulse-echo transmission, detection, and display
techniques. Brief pulses of ultrasound energy emitted by a
transducer are directed into tissue, and echoes arise from

acoustic interfaces within the body. Precise timing allows de-
termination of the depth from which each echo originates. The
returning echo signal contains amplitude, phase, and fre-
quency information related to the position, nature, and motion
of reflecting structures. In conventional B-mode ultrasound
imaging, amplitude information in the backscattered signal is
used to generate a gray-scale image. Small, rapidly moving
targets, such as red blood cells within a vessel, produce echoes
of such low amplitude that they are not commonly displayed.
Gray-scale imaging uses only the amplitude of the backscat-
tered ultrasound signal and ignores changes in the frequency
of the reflected sound that arise if the target is moving relative
to the transducer. Doppler ultrasound uses this additional
frequency information to evaluate moving targets. When
high-frequency sound impinges on a stationary interface, the
reflected ultrasound has essentially the same frequency or
wavelength as the transmitted sound (Fig. 29.1 A). 2 If, how-
ever, the reflecting interface is moving with respect to the
sound beam emitted from the transducer, there is a change in
the frequency of the sound scattered by the moving object (see
Fig. 29. IB and C). This change in frequency is directly propor-
tional to the velocity of the reflecting interface relative to the
transducer and is a result of the Doppler effect. The relation-
ship of the returning ultrasound frequency to the velocity of
the reflector is described by the Doppler equation:

AF = (F r -F t ) = 2F t x-

Where:
AF = Doppler frequency shift;

F r = frequency of sound reflected from the moving target;
F t = frequency of sound emitted from the transducer;
v = velocity of the target toward the transducer;
c = velocity of sound in tissue.

The Doppler frequency shift AF, as described, applies only if
the target is moving directly toward or away from the trans-
ducer, as is shown in Figure 29. IB and C.

In the clinical setting, the direction of the ultrasound beam is
seldom directly toward or away from the direction of flow, and

306

chapter 29 Spectral analysis

F t = F}

\ < F r

3fc*C

F f > F,

c -t - -r

Figure 29.1 Doppler effect. When ultrasound impinges on a stationary
target (A), the reflected frequency (F r ) is identical to the transmitted
frequency (F t ), and there is no Doppler frequency shift. If the target is moving
toward the transducer (B), the reflected frequency (F r ) is greater than the
transmitted frequency (F t ), and there is a positive Doppler frequency shift. If
the target is moving away from the transducer (C), the reflected frequency
(F r ) is less than the transmitted frequency (F t ), and there is a negative Doppler
frequency shift.

the ultrasound beam usually approaches the moving target at
an angle designated as 9, the Doppler angle (Fig. 29.2). In this
case, the frequency shift AF is reduced in proportion to the
cosine of this angle, and

AF = (F r -F t ) = 2F t x-xcos9

Where:

9 = angle between axis of flow and incident ultrasound
beam.

If the Doppler angle can be measured, estimation of flow
velocity is possible. Accurate estimation of target velocity re-
quires precise measurement of both the Doppler frequency
shift and the angle of insonation to the direction of target
movement. As the Doppler angle 9 approaches 90°, the cosine
of 9 approaches 0. At an angle of 90°, there is no relative move-
ment of the target toward or away from the transducer, so no
Doppler frequency shift is detected. Accurate angle correction
requires that Doppler measurements be made at angles of less
than 60° because above 60°, relatively small changes in the

Doppler angle are associated with large changes in cos 9, and a
small error in estimation of the Doppler angle may result in a
large error in estimation of velocity. In general, the Doppler
angle should be kept as near 60° as possible, and the same
angle should be used for serial studies.

Spectral display

Several options exist for the processing of AF, the Doppler
frequency shift, to provide useful information on the direction
and velocity of blood flow. Doppler frequency shifts found
in clinical examinations performed with typical Doppler
instruments operating at 3-5 MHz fall in the range of
several hundred to several thousand hertz and are audible.
The audible Doppler frequency shift may be analyzed by
ear, and with training many flow characteristics may be
identified. When ultrasonic Doppler methods were first
introduced into medical practice in the late 1950s, high blood
flow velocities and stenosis were identified by the audible
sounds produced.

Most Doppler shift data are displayed in graphic form as a
time-varying plot of the frequency spectrum of the returning
signal. 3 In the 1960s, real-time Doppler wave forms were gen-
erated using a zero-crossing technique, but spectral waveform
analysis was limited to off-line methods. With the later intro-
duction of real-time fast Fourier transform (FFT) spectral ana-
lyzers, a practical approach for clinical spectral waveform
analysis became a reality. When using the FFT to perform fre-
quency analysis, the analogue Doppler signal containing
the frequency shift data from the target is digitized by an
analogue-to-digital converter into numerous segments, each
a few milliseconds in duration. The frequency data from each
short time interval are then transformed from the frequency
domain to the time domain by the FFT. The data are displayed
as a graphic plot of frequency (or velocity, if the Doppler angle
is taken into account) plotted against time. The Doppler
frequency spectrum displays the variation with time of the
Doppler frequencies present in the volume sampled, with the
envelope of the spectrum representing the maximum and
minimum frequencies present at any given point in time, and
the width of the spectrum at any point indicating the range of
frequencies present. In many instruments, the amplitude of
each frequency component is displayed in gray scale (Fig.
29.3). The presence of a large number of different frequencies
at a given point in the cardiac cycle results in so-called spectral
broadening.

In DCI systems, frequency shifts determined from Doppler
measurements are displayed as a feature of the image itself
(Fig. 29.4; also in colour, see Plate 1, facing p. 370). 4 A single
color pixel can display only a portion of the information pro-
vided in the spectral display, and, in most instruments, the
color assigned to each pixel indicates a weighted mean of the
Doppler frequency shifts detected.

307

pa rt 1 1 Noninvasive vascular diagnostics

AF=(F t -F r ) = 2»vF t /c

AF= (F t -F r ) = 2 • vF t • cos 0/ c

Figure 29.2 Dopplerangle. The Doppler
equation describes the relationship of the
Doppler frequency shift AFto the velocity of the
target, v, and the velocity of sound in tissue, c. If
the direction of motion of the target is parallel
to the ultrasound beam (A), the Doppler angle is
0° and is not considered in the Doppler
equation. If target motion is not parallel to the
ultrasound beam (B), the angle between the
direction of insonation and the direction of
flow, the Doppler angle (9), must be taken into
account in the Doppler equation. In this case,
the velocity represented by a given frequency
shift is reduced in proportion to the cosine of 0.

,7 3 „
-53-

£FT*:i£Ffc

6,2 MH2

UF 53Hr
P'.:_E ^E
Sv l'.5rr
SV£ ZZrr
FS-_533E

C\t2 5^/S

2-/ s as

1-13

lin.! ii iIiiiiIiiii In ii lllllllllllllill ii illllllkllll

Figure 29.3 Spectral display. The
conventional Doppler spectrum displays
variation with time of the Doppler frequency
shifts present in the sample volume. At any
instant in time, the maximum and minimum
frequency shifts present are displayed. Gray
scale is used to indicate the distribution of the
frequencies. Here, arrow A indicates the
maximum frequency shift detected in
middiastole, and arrow B indicates the
minimum frequency at this point.

308

chapter 29 Spectral analysis

Figure 29.4 With Doppler color imaging,
static interfaces are displayed as conventional
gray-scale images. The mean Doppler frequency
shift of moving targets is displayed in color.
Color is used to indicate the direction of flow
relative to the transducer, as well as the
magnitude of the frequency shift. In this scan of
the carotid bifurcation, flow away from the
transducer is shown in shades of red, with less
saturated colors indicating frequency shifts
associated with minimal stenosis of the internal
carotid artery. See also Plate 1 , facing p. 370.

Doppler instrumentation

The simplest Doppler devices use continuous (continuous-
wave Doppler) rather than pulsed ultrasound. Two trans-
ducers transmit and receive ultrasound continuously, and the
"transmit" and "receive" beams overlap in a sensitive volume
at some distance from the transducer face (Fig. 29. 5 A). Al-
though direction of flow can be determined with continuous-
wave Doppler, these devices do not allow discrimination of
motion coming from various depths, and the source of the sig-
nal being detected is difficult if not impossible to ascertain
with certainty. Continuous-wave Doppler instruments are
used primarily at the bedside or during surgery to confirm the
presence of flow in superficial vessels.

Flow analysis is performed most often using range-gated
pulsed Doppler systems. Pulsed Doppler devices emit brief
pulses of ultrasound energy. Because the velocity of sound in
tissue is relatively constant, measurement of the time interval
between transmission of a pulse and the return of the echo
provides a means of determining the depth from which the
Doppler shift arises (see Fig. 29. 5B). An electronic gate allows
selection of the location and size of the flow volume from
which the Doppler data are displayed. When this is combined
with a two-dimensional, real-time, B-mode imager in the form
of a duplex scanner, the position of the Doppler sample can be
precisely controlled and monitored.

In color-flow imaging systems, flow information deter-
mined from Doppler measurements is displayed as a feature
of the image itself. Stationary or slowly moving targets pro-

vide the basis for the B-mode image. Signal phase provides in-
formation about the presence and direction of motion, and
changes in frequency relate to the velocity of the target.
Backscattered signals from red blood cells are displayed in
color as a function of their motion toward or away from the
transducer, and the degree of the saturation of the color is used
to indicate the relative velocity of the moving red cells. The use
of color saturation to display variations in Doppler shift fre-
quency allows a semiquantitative estimate of flow from the
image alone, provided that variations in the Doppler angle are
noted. The display of flow throughout the image field allows
the position and orientation of the vessel of interest to be ob-
served at all times. The display is ideal for identification of
small, localized areas of turbulence within a vessel, which pro-
vide clues to stenosis or irregularity of the vessel wall caused
by atheroma, trauma, or other disease. Flow within the vessel
is observed at all points, and stenotic jets or focal areas of tur-
bulence are shown that might be overlooked with duplex in-
strumentation. Flow imaging permits small vessels to be seen
that are invisible in conventional real-time images. DCI aids in
precise determination of the direction of flow and measure-
ment of the Doppler angle. Limitations of DCI include inabil-
ity to display the entire Doppler spectrum in the image and
limited sensitivity of some instruments.

Interpretation of the Doppler spectrum

Components of the Doppler data that must be evaluated in
spectral display include the Doppler shift frequency and am-

309

pa rt 1 1 Noninvasive vascular diagnostics

Sample volume

Figure 29.5 Continuous-wave and pulsed
Doppler. Simple continuous-wave Doppler
devices use two transducers that transmit and
receive ultrasound continuously (A). The beams
overlap in a sensitive volume at some distance
from the transducer face, but do not allow
discrimination of motion coming from various
depths. Pulsed Doppler devices (B) emit brief
bursts of ultrasound. Through precise timing
of pulses and detection of echoes, flow
information from small sample areas can be
obtained.

plitude, the Doppler angle, the spatial distribution of frequen-
cies across the vessel, and the temporal variation of the signal.
Because the Doppler signal itself has no anatomic significance,
the examiner must interpret the Doppler signal and then de-
termine its relevance in the context of the image.

Detection of a Doppler frequency shift indicates movement
of the target, which in most applications is related to the
presence of flow. The sign of the frequency shift (positive or
negative) indicates the direction of flow relative to the trans-
ducer. Analysis of the Doppler shift frequency with time can be
used to infer both proximal stenosis and changes in distal vas-
cular impedance. Most work using pulsed Doppler has em-
phasized the detection of stenosis, thrombosis, and flow
disturbances in major peripheral arteries and veins. The abili-
ty of spectral waveform analysis to identify stenosis is based
primarily on detection of increased blood velocity through the
stenotic region of reduced cross-sectional area; in peripheral
vessels, analysis of the Doppler spectrum allows accurate
prediction of the degree of vessel narrowing. Vessel stenosis
typically is associated with large Doppler frequency shifts in
both systole and diastole at the site of greatest narrowing, with
turbulent flow in poststenotic regions. The maximum peak
Doppler frequencies associated with 50-70% diameter reduc-
tion are typically two to four times normal peak Doppler
frequencies.

In addition to characterization of stenosis, Doppler spectral

display can provide information related to resistance to flow in
the distal vascular tree (Fig. 29.6). Doppler indices such as the
systolic-diastolic ratio, resistive index, and pulsatility index,
which compare the flow in systole and diastole, indicate resis-
tance in the peripheral vascular bed (Fig. 29.7). Changes of
these indices from normal may be important in early identifi-
cation of rejection of transplanted organs, parenchymal dys-
function, and malignancy. Although these indices are useful, it
is important to keep in mind that these measurements are in-
fluenced not only by resistance to flow in peripheral vessels
but by many other factors, including heart rate, blood pres-
sure, vessel wall length and elasticity, and extrinsic organ com-
pression. Interpretation must therefore always take into
account all of these variables.

Although the more graphic presentation of DCI suggests
that interpretation is made easier, the complexity of the color
Doppler image actually makes this a more demanding exami-
nation to evaluate than the Doppler spectrum. Nevertheless,
DCI has several advantages over duplex Doppler, in which
flow data are obtained only from a small portion of the area
being imaged. For a conventional Doppler study to achieve
reasonable sensitivity and specificity in detection of flow dis-
turbances, a methodical search and sampling of multiple sites
within the field of interest must be performed. DCI devices
permit simultaneous sampling of multiple sites and are less
susceptible to this error.

310

chapter 29 Spectral analysis

Figure 29.6 Peripheral resistance. Changes in the Doppler spectral
waveform resulting from physiologic changes in the resistance of the vascular
bed supplied by the brachial artery of a normal person are illustrated. (A) The
waveform shows a typical high-resistance pattern with little diastolic flow. (B)
A blood pressure cuff has been inflated to above systolic pressure to occlude
the distal branches supplied by the brachial artery. This causes a drop in
systolic amplitude and reversal of diastolic flow, resulting in a quite different
waveform than found in the normal resting state (A). (C) The effect of
removing the peripheral resistance on the waveform after release of 3 min of
occluding pressure. Within a single heartbeat, the waveform changes from
the damped pattern (A) of high resistance to a low-impedance pattern (B)
with high systolic amplitude and greatly increased diastolic flow as a result of
vasodilation induced by the brief period of ischemia.

Other considerations

S/D ratio = A / B
Resistive index = (A - B) / A
Pulsatility index = (A - B) / Mean

Figure 29.7 Doppler indices. The most common indices used to
characterize peripheral impedance are derived from measurements of the
peakand minimal amplitudes of the spectral envelope. These include the
systolic— diastolic (S/D) ratio, the resistive index, and the pulsatility index.

Detection and display of frequency information related to
moving targets adds a group of special technical considera-
tions not encountered with other forms of ultrasonography It
is important to understand the source of these artifacts and
their influence on the interpretation of the flow measurements
obtained in clinical practice. Major sources of Doppler inaccu-
racy or artifact are related to the following.

Doppler frequency

The moving red blood cells that serve as the primary source of
the Doppler signal act as point scatterers of ultrasound, and
the intensity of the scattered sound varies in proportion to the
fourth power of the frequency. As the transducer frequency
increases, Doppler sensitivity improves, but this is offset by
increased tissue attenuation and diminished penetration.
The operator must ensure that a proper balance is achieved

311

pa rt 1 1 Noninvasive vascular diagnostics

Figure 29.8 Spectral broadening. Spectral broadening may arise undera
number of conditions. (A) Spectral broadening in late systole and diastole.
This isa normalfinding in most vessels and isa result of a change from plug
flow to laminar flow. (B) Spectral broadening may also reflect turbulent flow.
Here, the sample volume is located just distal to an atheromatous plaque that

disturbs flow in both systole and diastole. Artifactual spectral broadening
may be produced if the Doppler sample volume is too large (C), or placed
near the vessel wall (D). Excessive Dopplergain may also suggest spectral
broadening.

between the conflicting requirements for Doppler sensitivity
and penetration during a Doppler examination.

Wall filters

Doppler instruments may detect low-frequency motion from
vessel walls and adjacent structures as well as from blood flow.
To eliminate these low-frequency signals from the display,
most instruments use high-pass filters or "wall" filters, which
remove signals that fall below a given frequency limit. Al-
though these filters are effective in eliminating low-frequency
noise, they also may remove signal from low-velocity blood
flow. In certain clinical situations, measurement of these
slower flow velocities is of clinical importance, and improper
selection of the wall filter may result in serious errors of

interpretation. For example, low-velocity venous flow may
not be detected if an improper filter is used. Similarly, low-
velocity diastolic flow in certain arteries also may be eliminat-
ed from the display, resulting in errors in the calculation of
Doppler indices. In general, the filter should be kept at the
lowest practical level, usually in the range of 50-100 Hz.

Spectral broadening

The range of flow velocities at a given point in the pulse cycle
defines the spectral breadth, and spectral broadening is an im-
portant criterion of high-grade vessel narrowing (Fig. 29. 8 A
and B). Excessive system gain or changes in the dynamic range
of the gray-scale display of the Doppler spectrum may suggest
spectral broadening, whereas opposite settings may mask

312

chapter 29 Spectral analysis

broadening of the Doppler spectrum, causing diagnostic inac-
curacy. Spectral broadening also may be produced by the se-
lection of an excessively large sample volume (see Fig. 29. 8C)
or by placement of the sample volume too near the vessel wall,
where slower velocities are present (see Fig. 29.8D). Excessive
Doppler gain is another common cause of apparent spectral
broadening (see Fig. 29. 8E).

Aliasing

Aliasing is an artifact arising from ambiguity in the measure-
ment of high Doppler frequency shifts. To ensure that samples
originate only from a selected depth when using a pulsed
Doppler system, it is necessary to wait for the echo from the
area of interest before transmitting the next pulse. This limits
the rate at which pulses can be generated, since a lower pulse
repetition frequency (PRF) is required for greater depth. The
PRF also determines the maximum depth from which ambigu-
ous data can be obtained. If the PRF is less than twice the max-
imum frequency shift produced by movement of the target
(the Nyquist limit), an artifact called aliasing results (Fig.
29. 9 A). If the PRF is less than twice the frequency shift being
detected, lower frequency shifts than are actually present are
displayed (see Fig. 29 .9B). Because of the need for lower PRFs
to reach deep vessels, signals from deep abdominal arteries are
prone to aliasing if high velocities are present. In practice,
aliasing usually is recognized readily. Aliasing can be reduced
by increasing the pulse repetition frequency, increasing the
Doppler angle (thereby decreasing the frequency shift), or by
using a lower frequency Doppler transducer.

Doppler angle

When making Doppler measurements, it is desirable to correct
for the Doppler angle and display the measurements in terms
of velocity. These measurements are independent of the
Doppler frequency. The Doppler angle is an important factor
in the detection of high-grade stenosis, and the accuracy of a
velocity estimate obtained with Doppler is only as great as the
accuracy of the measurement of the Doppler angle. Compli-
cating measurement of flow velocity is the fact that the veloc-
ity vector in a vessel usually traces a helical pattern along the
cylindrical wall of the artery, and changes in Doppler fre-
quency result from changes in flow direction as well as in
velocity. Despite these problems related to the Doppler angle,
the clinical standard for evaluation of stenosis of peripheral
vessels has been single-gate pulsed Doppler at a standard
angle. In general, the Doppler angle is best kept at 60° or less,
because small changes in the Doppler angle above 60° result in
significant changes in the calculated velocity; therefore, mea-
surement inaccuracies result in much greater errors in velocity
estimates than similar errors at lower Doppler angles.

Sample volume size

With pulsed Doppler systems, the length of the Doppler

www

A A A A A A
V V V V V V

A High PRF

ttttttt

Low PRF

Figure 29.9 Aliasing. To measure accurately a Doppler frequency shift, the
sampling rate or pulse repetition frequency must be at least twice the
frequency being sampled (A). If the sampling rate istoo low, the sampled
frequency will be represented by a lower frequency than is actually present
(B). In the spectral display, these undersampled components are presented
beneath the baseline (C).

sample volume is controlled by the operator and the width is
determined by the beam profile. Analysis of Doppler signals
requires that the sample volume be adjusted to exclude as
much unwanted clutter from near the vessel walls as possible.
In high-grade stenosis in which there is more than a doubling
of the peak Doppler frequency, accurate measurements are
possible with either large or small Doppler sample volumes.
This is in contrast to the requirements for detection of minor
flow disturbances, in which the smallest possible sample vol-
umes are required.

Conclusion

Doppler spectral analysis is the mainstay of noninvasive eval-
uation of the peripheral arterial system, and serves as the
primary noninvasive screening test for hemodynamically sig-
nificant narrowing in the peripheral arterial system. Accurate
and reliable results are possible only with a detailed under-

313

pa rt 1 1 Noninvasive vascular diagnostics

standing of principles, instrumentation, and artifacts related
to Doppler spectral analysis, along with knowledge of normal
and abnormal hemodynamics.

References

1. Strandness D Jr. The gold standard in the diagnosis of vascular
disease. In: Labs KH, Jaeger KA, Fitzgerald DE, Woodcock JP,

Neuerburg-Heusler D, eds. Diagnostic Vascular Ultrasound.
London: Hodder & Stoughton, 1992:3.

2. Merritt CRB. Doppler US: the basics. Radiographics 1991; 11:109.

3. Beach KW, Phillips DJ. Sensitivity and precision of fast Fourier
transform spectral analysis in mild carotid atherosclerotic disease.
In: Labs KH, Jager KA, Fitzgerald DE, Woodcock JP, Neuerburg-
Heusler D, eds. Diagnostic Vascular Ultrasound. London: Hodder &
Stoughton, 1992:57.

4. Merritt CRB. Doppler color flow imaging. / Clin Ultrasound 1987;
15:591.

314

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Ultrasound imaging

Christopher R.B. Merritt

Modern vascular diagnosis depends on a variety of tech-
niques including clinical assessment, noninvasive testing, and
angiography. Magnetic resonance imaging, computed tomog-
raphy, and radioisotope imaging play useful complementary
roles. Ultrasound is the primary method for noninvasive
vascular evaluation. Doppler ultrasound provides a safe and
relatively inexpensive method to determine the presence,
direction, velocity, and character of flow in peripheral, abdom-
inal, and pelvic vessels. Doppler, however, represents only a
part of the contribution of ultrasound to vascular diagnosis.
Ultrasound imaging, using high-resolution scanners, permits
high-resolution imaging of vessel walls and vascular
abnormalities such as thrombus and plaque that often are as
important as and, in some cases, even more important
than flow characteristics. Ultrasound imaging thus plays an
essential role in vascular evaluation and the identification of
abnormalities.

To obtain maximum benefit from ultrasound imaging,
knowledge of basic principles of ultrasound imaging and ex-
amination techniques is required, including an understanding
of the interactions of acoustic energy with tissue and the meth-
ods and instruments used to produce the ultrasound display.
The use of expensive, state-of-the-art ultrasound instrumenta-
tion does not guarantee that high-quality images of diagnostic
value will be produced. As with angiography, special skills
both in performance and interpretation of the examination are
required, and pitfalls await the careless or poorly trained user.
Artifacts and the limitations of ultrasound must be considered
along with the advantages and disadvantages of alternative
approaches, and the user must have knowledge of sectional
anatomy and normal and abnormal sonographic patterns
necessary to establish a diagnosis. Finally, mastery of technolo-
gy must be matched by clinical skills, so that the sonographic
findings are related to the clinical problem under evaluation.

Basic principles of ultrasound

All forms of diagnostic ultrasound are based on back-

scattered information (echoes) from interfaces within the
body. Accurate interpretation of ultrasound images requires
an understanding of the way ultrasound interacts with tissues
and vessels to display normal anatomy as well as pathologic
features related to atheromatous disease, degeneration, in-
flammation, and trauma. Sound energy is transmitted thr-
ough matter as alternating waves of pressure and rarefaction.
Sound frequencies used for diagnostic applications typically
range from 2 to 10 MHz (2000000 to 10000000 cycles per sec-
ond) for clinical imaging, although frequencies as high as
50-60 MHz are under investigation for certain specialized in-
travascular imaging applications. In general, the frequencies
used for ultrasound imaging are somewhat higher than those
used for Doppler. The pressure waves produced by an ultra-
sound transducer travel at a velocity determined by the nature
of the propagating medium. In soft tissues, the average veloci-
ty of sound propagation is approximately 1540 m/s, although
certain tissues have propagation velocities significantly less
than (e.g. aerated lung and fat) or greater than (e.g. bone) those
of soft tissues.

In contrast to some ultrasound applications, such as
continuous-wave Doppler, imaging requires pulsed ultra-
sound. With pulsed ultrasound, the transducer introduces a se-
ries of brief bursts of sound into the body. Each ultrasound
pulse typically consists of about three cycles. The pulse length
is determined by the product of the wavelength and the num-
ber of cycles in the pulse. Axial resolution, the maximum reso-
lution along the beam axis, is determined by the pulse length
(Fig. 30.1). Ultrasound frequency and wavelength are inverse-
ly related, so the pulse length decreases as the imaging frequen-
cy increases. Since the pulse length determines the maximum
resolution along the axis of the ultrasound beam, higher trans-
ducer frequencies provide higher image resolution. For exam-
ple, a transducer operating at 5.0 MHz produces sound with a
wavelength of 0.308 mm. If eachpulse consists of three cycles of
sound, the pulse length is slightly less than 1 .0 mm, and this be-
comes the maximum resolution along the beam axis. If the
transducer frequency is increased to 10 MHz, the pulse length
is less than 0.5 mm, permitting resolution of smaller details.

315

pa rt 1 1 Noninvasive vascular diagnostics

C D

~ -%- ~

Image

Figure 30.1 Axial resolution. The pulse length (the
wavelength times the number of cycles in the pulse)
determines the maximum resolution of objects lying
along the beam axis. The vessel walls A and B are
separated by a distance greaterthan the pulse length
and will be resolved in the image as two separate
structures. The walls C and D are separated by a
distance less than the pulse length and will therefore
appear in the image as a single object. Axial resolution
is important for imaging because it determines the
smallest structure that may be imaged with a given
instrument and transducer.

Ultrasound is a tomographic method of imaging, producing
thin slices of information from the body, and the width and
thickness of the ultrasound beam are important determinants
of image quality. Excessive beam width and thickness limit the
ability to delineate small features such as the tiny cystic areas
in atheromatous plaque associated with intraplaque hemor-
rhage. The width and thickness of the ultrasound beam deter-
mine lateral resolution (Fig. 30.2) and elevation resolution
(Fig. 30.3), respectively. Lateral and elevation resolution are
significantly poorer than the axial resolution of the beam.
Lateral resolution is controlled by focusing the beam, usually
be electronic phasing to alter the beam width at a selected
depth of interest. Elevation resolution is determined by the
construction of the transducer and generally cannot be con-
trolled by the user.

Practical considerations in the selection of the optimal trans-
ducer for a given application include not only the require-
ments for spatial resolution but the distance of the target object
from the transducer, because penetration of ultrasound
diminishes as frequency increases. In general, the highest
ultrasound frequency permitting penetration to the depth of
interest should be selected. For superficial vessels within
1-3 cm of the surface, such as the carotid arteries or extremity
arteries and veins, imaging frequencies of from 7.5 to 10 MHz
usually are used. These high frequencies also are ideal for in-
traoperative applications. For evaluation of deep abdominal
vessels more than 10-15 cm from the surface, frequencies as
low as 2.25-3.5 MHz may be required. When maximal resolu-
tion is needed (as in the characterization of plaque), a high-
frequency transducer with excellent lateral and elevation
resolution at the depth of interest is required.

The quality of an ultrasound image is determined not only
by axial, lateral, and elevation resolution, but also by the char-
acteristics of the structures being examined. Reflection of
sound occurs at boundaries of tissues or propagating media
that have different acoustic properties. The most important
properties affecting the amount of sound reflected or transmit-
ted at a tissue boundary are the propagation velocities of
sound in the adjacent media, the physical densities of the

media, the smoothness of the interface, and the angle of in-
sonation with respect to the interface. Propagation velocity
and tissue density determine the acoustic impedance of the tis-
sues. The greater the difference in the adjacent acoustic imped-
ances, the greater the reflecting property of the boundary. If the
target is large compared with the wavelength of the ultra-
sound used, it will behave like a mirror, and is called a specular
reflector. A specular interface reflects sound back to the trans-
ducer only if the reflector lies at near 90° to the path of the ul-
trasound beam (Fig. 30.4A and B). In vascular imaging, most
vessel walls act as specular reflectors and are optimally im-
aged when insonated at angles near 90° (see Fig. 30.4C). With
smaller angles of incidence, the sound is reflected, but little of
the reflected sound returns to the transducer (see Fig. 30. 4D).
Because most vessel walls behave as specular reflectors,
conflicting conditions exist for optimal vessel imaging
and Doppler evaluation since Doppler requires a small angle
between the sound beam and the vessel.

In contrast to vessel walls, tissues, thrombus, and plaque
consist of small, irregular tissue interfaces near the size of the
wavelength of diagnostic ultrasound. These structures pro-
duce diffuse scattering of the incident sound. The image of
these tissues is derived from the portion of the scattered sound
that returns to the transducer (Fig. 30.5).

Imaging transducers

Like a motion picture, real-time ultrasound produces the
impression of motion by generating a series of individual two-
dimensional images at rates of from 15-60 frames per second.
Transducers used for real-time imaging may be classified by
the method used to steer the beam to rapidly generate each in-
dividual image. Beam steering may be by mechanical rotation
or oscillation of the transducer, or the beam may be steered
electronically. Electronic beam steering is used in linear-array
and phased-array transducers and permits a variety of image
display formats. Most electronically steered transducers in
use also provide electronic focusing adjustable for depth.

316

chapter 30 Ultrasound imaging

Lateral resolution

Image

Figure 30.2 Lateral resolution. (A) The width of the ultrasound beam
determines the maximum resolution of objects side by side in the plane
perpendicular to the axis of the beam. (B) The width of the beam in the focal
zone is less than the distance between the targets, permitting both objects to
be imaged as separate structures. (C) The distance between the targets is less

than the width of the beam, and only a single object will be displayed in the
image. With most transducers, focusing adjustment is possible, permitting
the user to maximize lateral resolution at a selected depth from the
transducer.

317

Elevation resolution

Figure 30.3 Elevation resolution. The width
of the ultrasound beam in the direction
perpendicular to the scan plane determines the
slice thickness or elevation plane. Unlike the
lateral resolution, this beam dimension is
determined by the construction of the
transducer and cannot be adjusted by the user.
Excessive slice thickness may hinder
identification of small imaging features such as
intraplaque hemorrhage and may render some
transducers inappropriate for this application.

Figure 30.4 Specular reflectors. A specular
interface reflects sound back to the
transducer like a mirror. If the reflector lies at
near 90° to the path of the ultrasound beam
(A), most of the reflected sound will return to
the transducer. With smalleranglesof
incidence (B), the sound is reflected but little
of the reflected sound returns to the
transducer. In vascular imaging, most vessel
walls act as specular reflectors and are
optimally imaged when insonated at angles
near 90°. The walls of the common carotid
artery are specular reflectors, and insonation
at 90° (C) produces a strong returning echo,
whereas images at a smaller angle (D) provide
poor definition of the vessel wall.

chapter 30 Ultrasound imaging

Figure 30.4 Continued

Mechanically steered transducers may use single-element
transducers with a fixed focus, or may use annular arrays of el-
ements with electronically controlled focusing.

For real-time imaging, transducers using mechanical or
electronic beam steering generate display in a rectangular or
pie-shaped format. For peripheral vascular examinations, lin-
ear-array transducers with a rectangular image format often

are used. The rectangular image display has the advantage of a
larger field of view near the surface, but requires a larger sur-
face area for transducer contact. These arrays permit imaging
of a relatively long segment of the vessel. Because most pe-
ripheral vessels run parallel to the skin surface, such transduc-
ers provide an angle of near 90° between the ultrasound beam
and the vessel, and thus are well suited for imaging the vessel

319

pa rt 1 1 Noninvasive vascular diagnostics

Figure 30.5 Diffuse reflectors. If the
reflecting surface is irregular, there is
scattering of the incident sound in many
directions. Even with low incident angles,
someof the sound returns to the transducer.

wall. Sector scanners with either mechanical or electronic
steering require only a small surface area for contact and are
better suited for examinations in which access is limited.

Image display

To provide clinically useful information, ultrasound signals
may be processed and displayed in several ways. 1 Over the
years, imaging has evolved from simple A-mode and bistable
display to high-resolution, real-time, gray-scale imaging. The
earliest A-mode devices displayed the backscattered echo as a
vertical deflection on the face of an oscilloscope. The horizon-
tal sweep of the oscilloscope was calibrated to indicate the dis-
tance from the transducer to the reflecting surface. In this form
of display, the strength or amplitude of the reflected sound is
indicated by the height of the vertical deflection displayed on
the oscilloscope. With A-mode ultrasound, only the position
and strength of a reflecting structure are recorded.

Another simple form of imaging, M-mode ultrasound, dis-
plays echo amplitude and shows the position of moving re-
flectors. M-mode uses the brightness of the display to indicate
the intensity of the reflected signal. The time base of the dis-
play can be adjusted to allow for varying degrees of temporal
resolution, as dictated by clinical application. M-mode ultra-
sound is interpreted by assessing motion patterns of specific
structures and determining anatomic relationships from
characteristic patterns of motion. The major application of M-
mode display is in the evaluation of the rapid motion of car-
diac valves and of cardiac chamber and vessel walls. M-mode
may play a future role in measurement of subtle changes in
vessel wall elasticity accompanying atherogenesis.

The mainstay of vascular imaging with ultrasound is pro-
vided by real-time, gray-scale B-mode display. Here, varia-
tions in display intensity or brightness are used to indicate
reflected signals of differing strength. When an ultrasound
image is displayed on a black background, signals of greatest
intensity appear as white, absence of signals is shown as black,

and signals of intermediate intensity appear as shades of gray.
If the ultrasound beam is moved with respect to the object
being examined and the position of the reflected signal is
stored, a two-dimensional image results, with the brightest
portions of the display indicating structures reflecting more of
the transmitted sound energy back to the transducer. Since B-
mode display relates the strength of a backscattered signal to a
brightness level on the display device (usually a video display
monitor), it is important that the operator understands how
the amplitude information in the ultrasound signal is translat-
ed into a brightness scale in the image display. Each ultra-
sound manufacturer offers several options for the way the
dynamic range of the target can be compressed for display, as
well as the transfer function that assigns a given signal ampli-
tude to a shade of gray. Although these technical details vary
from one machine to another, the way they are used (or
abused) by the operator of the scanner may have a profound
impact on the clinical value of the final image.

Real-time, two-dimensional B-mode ultrasound is now the
major method for ultrasound imaging throughout the body
and is the most common form of B-mode display. Real-time
ultrasound permits assessment of both anatomy and motion.
To produce the impression of motion, a rapid series of individ-
ual two-dimensional B-mode images is generated. When im-
ages are acquired and displayed at rates of several times per
second (typically 15-60), the effect is dynamic, and because
the image reflects the state and motion of the organ at the time
it is examined, the information is regarded as being shown in
real time. In cardiac applications, the terms 2-D echocardio-
graphy and 2-D echo are used to describe real-time B-mode
imaging; in most other applications, the term real-time
ultrasound is used.

Imaging pitfalls

In ultrasound, perhaps more than in any other imaging
method, the quality of the information obtained is determined

320

chapter 30 Ultrasound imaging

by the ability of the operator to recognize and avoid artifacts
and pitfalls. 2 Many imaging artifacts are induced by errors in
scanning technique or improper use of the instrument, and are
preventable. Artifacts may suggest the presence of structures
that are not present, causing misdiagnosis, or they may cause
important findings to be obscured. Because an understanding
of artifacts is essential for correct interpretation of ultrasound
examinations, several of the most important artifacts deserve
discussion.

Many artifacts suggest the presence of structures not
actually present. These include reverberation, refraction, side
lobes, and speckle. Reverberation artifacts arise when the ul-
trasound signal reflects repeatedly between highly reflective
interfaces that usually but not always are near the transducer
(Fig. 30.6A). This type of problem may occur in large vessels,
causing diagnostic problems by obscuring significant find-
ings. Reverberations also may give the false impression of
solid structures in areas where only fluid is present, suggest-
ing the presence of thrombus within a vessel, when in fact the
lumen is patent. Certain types of reverberation may be helpful
because they allow the identification of a specific type of
reflector, such as a surgical clip. Reverberation artifacts can
usually be reduced or eliminated by changing the scanning
angle or transducer placement to avoid the parallel interfaces
that contribute to the artifact.

Refraction causes bending of the sound beam so that targets
not along the axis of the transducer are insonated. Their reflec-
tions are then detected and displayed in the image. This may
cause structures to appear in the image that actually lie outside
the volume the investigator assumes is being examined.
Similarly, side lobes may produce confusing echoes that arise
from sound beams that lie outside of the main ultrasound
beam. These artifacts are of clinical importance because they
may create the impression of particulate debris in fluid-filled
structures such as large vessels, aneurysms, and pseudo-
aneurysms. Side lobes also may result in errors of measure-
ment by reducing lateral resolution. As with most other
artifacts, repositioning the transducer and its focal zone or
using a different transducer will usually allow the differentia-
tion of artif actual from true echoes. Artifacts also may remove
real echoes from the display or obscure information, and im-
portant pathologic features may be missed. Shadowing results
when there is a marked reduction in the intensity of ultra-
sound deep to a strong reflector or attenuator (see Fig. 30.6B).
Shadowing causes partial or complete loss of information due
to attenuation of the sound by superficial structures. Calcified
plaque is a common source of shadowing that may obscure
segments of vessel lumen and plaque from view, and interfere
with Doppler measurements. Another common cause of loss
of image information is improper adjustment of system gain
and time gain compensation settings. Many low-level echoes
are near the noise levels of the equipment, and considerable
skill and experience are needed to adjust instrument settings
to display the maximum information with the minimum

noise. Poor scanning angles, inadequate penetration, and poor
resolution also may result in loss of significant information.
Careless selection of transducer frequency and lack of atten-
tion to the focal characteristics of the beam will cause loss of
clinically important information from deep, low-amplitude
reflectors and small targets. Finally, ultrasound artifacts may
alter the size, shape, and position of structures. For example, a
multipath artifact is created when the path of the returning
echo is not the one expected, resulting in display of the echo at
an improper location in the image. 3

Clinical applications

Most use of ultrasound in vascular assessment involves the
combination of imaging with Doppler flow measurement;
however, there are several applications in which imaging
is the primary diagnostic method. Although a detailed
discussion of these applications is beyond the scope of
this chapter, a brief summary of the major clinical uses of
ultrasound in vascular imaging is provided in the following
list:

MAJOR APPLICATIONS OF ULTRASOUND IMAGING IN
VASCULAR DIAGNOSIS

• Determinations of vessel size

• Diagnosis and follow-up of aneurysms

• Diagnosis of deep vein thrombosis (DVT)

• Localization and characterization of atheromatous plaque

• Identification of adjacent pathology (e.g. adenopathy,
tumor, cyst)

• Identification of sites for Doppler scanning

• Accurate measurement of Doppler angle

Duplex evaluation

Duplex Doppler and Doppler color imaging are standard non-
invasive techniques for peripheral vascular assessment that
depend on a combination of high-resolution imaging of the
vessel wall and lumen with pulsed Doppler sampling of flow.
A careful imaging examination of the vessel is essential to se-
lect sites for Doppler sampling and ensure correct positioning
of the Doppler sample volume. Because accurate spectral
analysis requires precise estimation of the angle of insonation
to the direction of blood flow, imaging is critical in providing
this information by delineating the course of the vessel. Imag-
ing also may demonstrate changes such as intraluminal
thrombus, intraplaque hemorrhage, or dissection that explain
symptoms in patients who do not have Doppler evidence of
hemodynamically significant lesions.

Venous evaluation

Ultrasound, as the primary method for diagnosis of DVT, re-
lies heavily on imaging criteria. 4 The earliest manifestation of

321

pa rt 1 1 Noninvasive vascular diagnostics

Figure 30.6 Imaging artifacts. (A)
Reverberation artifacts arise when the
ultrasound signal is reflected repeatedly
between strong interfaces near the transducer.
The resulting echoes (arrows) may obscure
structures of interest. (B) Shadowing is a
common artifact due to a marked reduction in
the intensity of ultrasound deep to a strong
reflector or attenuator. Shadowing causes
partial or complete loss of information. Calcified
plaque is a common source of shadowing that
may obscure segments of vessel lumen and
plaque from view, and interfere with Doppler
measurements.

DVT detectable with ultrasound is the finding of small areas of
thrombus in the valve recess. This is a subtle finding demon-
strable only with high-resolution imaging equipment. As the
thrombus enlarges and propagates within the vein, it fre-
quently is imaged, although some thrombi are not sufficiently
echogenic to be readily identified by imaging alone. The lack

of vein compressibility is the most important imaging criteri-
on for the presence of intraluminal thrombus, and ability to
compress the vein completely is regarded as a reliable indica-
tor of the absence of thrombus. Other imaging findings helpful
in the diagnosis of DVT include venous enlargement and loss
of normal respiratory dynamics.

322

chapter 30 Ultrasound imaging

Figure 30.7 Measurements of vessel diameter or
aneurysm size with ultrasound require careful
selection of the scan plane to minimize distortion due
to transducer angulation. Scan plane A overestimates
the anteroposterior diameter of the vessel, whereas
scan plane B provides a correct measurement of
anteroposterior and transverse diameters.

Aneurysm evaluation

Ultrasound imaging is the primary screening method for pa-
tients suspected of abdominal aortic and peripheral arterial
aneurysms. The value of ultrasound imaging lies in its ability
to delineate vessel walls and lumen clearly and to provide
accurate measurement of vessel size. Properly obtained, mea-
surements of aneurysms with ultrasound should be accurate
to within 5% of the true outer wall diameter of the vessel.
Variations of measurement are related primarily to differences
in selection of measurement sites and the plane of measure-
ment. Unlike measurements obtained from plain abdominal
radiographs, ultrasound measurements are not magnified.
As a result, the diameter of an aneurysm measured on a
lateral abdominal radiograph is 20-25% greater than the true
diameter obtained with ultrasound. Because improperly
selected scan planes may produce artifactual distortion of
ultrasound measurements, meticulous scanning technique is
necessary to ensure reproducible and accurate measurements
(Fig. 30.7).

In early stages of aneurysm development, ultrasound
shows loss of the normal tapering of the vessel followed by lo-
calized ectasia. Enlargement of aneurysms occurs at varying
rates, and serial ultrasound examinations are helpful in char-
acterizing the rate of enlargement of newly diagnosed small
aneurysms. A follow-up interval of 6 months is recommended
for monitoring aneurysm enlargement, although for small
aneurysms that show only minimal change on serial evalua-
tion, annual examinations are appropriate. Ultrasound imag-
ing, in addition to providing a noninvasive and generally
accurate means of measuring the size of abdominal aortic
aneurysms, also permits the identification of intraluminal
thrombus and permits assessment of the residual lumen of the
vessel.

Ultrasound imaging aids in differentiation of true aneu-
rysms from pseudoaneurysms and dissecting aneurysms,
although complicated aneurysms with leak or rupture
generally are evaluated with more precision using angio-

graphic or computed tomographic techniques. Finally,
ultrasound imaging aids in identification of fluid collections at
or around the graft in postoperative assessment of patients
after aneurysm repair.

Plaque evaluation

Unlike angiography, which images only the vessel lumen,
ultrasound permits simultaneous imaging of lumen, wall,
and atheromatous plaque. With modern, high-resolution in-
struments, changes in vessel wall thickness and intimal thick-
ening may be demonstrated. Surface irregularities of the
vessel wall at early stages of atheromatous plaque develop-
ment are shown along with more advanced stages of plaque
development (Fig. 30.8). 5 Ultrasound imaging permits deter-
mination of the presence and location of plaque in most
peripheral vessels. Ultrasound imaging is increasingly
appreciated for its ability to identify intraplaque hemorrhage
and predict unstable plaque likely to lead to embolism or rapid
progression. 6

Evaluation of adjacent structures

By transmitting arterial pulsations, masses adjacent to the
aorta or peripheral arteries may mimic clinical findings of
aneurysms. Because ultrasound, unlike angiography, shows
structures adjacent to the vessel, aneurysms are easily differ-
entiated from adjacent nonvascular pathologic features.
Lymphadenopathy, solid tumors, cysts, hematomas,
pseudoaneurysms, and abscesses may mimic aneurysms in
the abdomen or lower extremities, and are readily identified
using ultrasound.

Limitations of ultrasound imaging in primary vascular as-
sessment include generally poor correlation of imaging esti-
mates of vessel narrowing with hemodynamic measurements
derived from Doppler spectral analysis. Although ultrasound
provides a more complete view of the effects of plaque and
thrombus on the vessel lumen than angiography, calcified

323

pa rt 1 1 Noninvasive vascular diagnostics

Figure 30.8 High-resolution ultrasound permits detailed imaging of
plaque and thrombus. A homogeneous plaque (A) is contrasted to a
heterogeneous plaque (B) containing intraplaque hemorrhage (arrows).

(From MerrittCRB, Bluth El. Ultrasonographic characterization of carotid
plaque. In: Labs KH, Jaeger KA, Fitzgerald DE, eds. Diagnostic Vascular
Ultrasound. London: Hodder&Stoughton, 1992:213.)

plaque and tortuosity of some vessels restrict the value of
ultrasound imaging as the sole means of assessing the pres-
ence of vessel occlusion or narrowing.

Conclusions

In vascular diagnosis, imaging plays an important role and
complements Doppler investigations, serving as an essential
adjunct in the selection of sampling sites and angle correction.
Imaging is particularly important in the diagnosis of DVT,
and plays a key role in the identification and follow-up of
aneurysms. A growing role for ultrasound is in the identifica-
tion and characterization of atheromatous plaque. To use
ultrasound successfully in these applications, a well trained
operator, thoroughly familiar with the basic principles of
ultrasound imaging, is essential.

References

1 . Merritt C, Hykes D, Hedrick W et ah Medical diagnostic ultrasound
instrumentation and clinical interpretation: report of the Ultra-
sonography Task Force. JAMA 1991; 265:1155.

2. Kremkau FW. Diagnostic Ultrasound: Principles, Instruments, and
Exercises, 3rd edn. Philadelphia: WB Saunders, 1989.

3. Kremkau FW. Principles and instrumentation. In: Merritt CRB, ed.
Doppler Color Imaging. New York: Churchill Livingstone, 1992:7.

4. Merritt C. Evaluation of peripheral venous disease. In: Merritt
CRB, ed. Doppler Color Imaging. New York: Churchill Livingstone,
1992:113.

5. Merritt CRB, Bluth EI. Ultrasonographic characterization of carotid
plaque. In: Labs KH, Jaeger KA, Fitzgerald DE, eds. Diagnostic
Vascular Ultrasound. London: Hodder & Stoughton, 1992:213.

6. Merritt C, Bluth E. The future of carotid sonography. Am }
Roentgenol 1992; 158:37.

324

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Radionuclide scanning

Robert E. Sonnemaker

Radionuclide imaging procedures span the breadth and depth
of vascular surgery. They are reviewed in this chapter accord-
ing to clinical application.

Peripheral vascular disease

A plethora of radionuclide imaging procedures exists that ad-
dresses the full spectrum of peripheral vascular management
decisions (Table 31.1). Those with clinical utility have not
been widely adopted, however, whereas others remain
investigational.

nous shunting in extremities may be quantified by comparing
lung activity after intraarterial and intravenous 99m Tc MAA
administration. 3 This procedure maybe used to determine the
extent of ligated or embolized arteriovenous malformation
after treatment. Selective chemotherapy by arterial catheteri-
zation is evaluated for true chemotherapeutic agent distribu-
tion and presence of arteriovenous shunting by infusion of
99m Tc MAA at the therapeutic flow rate using the drug delivery
system. 4 Cross-circulation during isolation chemotherapy is
detected and quantified using an intravascular agent such as
99m Tc autologous erythrocytes. Treatment is discontinued at a
predetermined systemic blood level to prevent toxicity. 5

Assessment of vessel patency

Radionuclide angiography may be performed with the bolus
administration of various technetium-99m ( 99m Tc) radiopharm-
aceutical agents to assess major vascular channel patency and
perfusion (Table 31.2). This technique is proposed as a primary
screening procedure to assess injuries, stenoses, obstructions,
true and false aneurysms, grafts, repairs, and altered sympa-
thetic vasomotor tone. Moss and Rudavsky report the most
extensive clinical experience, as well as high diagnostic utility,
in a series of 555 examinations (contrast arteriographic and
surgically confirmed sensitivity, 100%; specificity, 99%). 1 The
test is simple, minimally invasive, safe, quick, inexpensive,
universally available, but not frequently performed.

Shunt evaluation and quantification

Trapping of macroaggregated albumin particles by the first
distal capillary vascular bed is the mechanism used to quant-
ify arteriovenous shunts in extremities or to assess patency of
peritoneovenous shunts. Routine confirmation of LaVeen
and Denver peritoneovenous shunt obstruction involves in-
traperitoneal injection of 99m Tc macroaggregated albumin
(MAA) and sequential imaging over the lungs. Visualization
of lung activity is the criterion for patency, with sensitivity of
100%, specificity of 97.2%, and accuracy of 98.5%. 2 Arteriove-

Assessment of arterial insufficiency

Gradual decrease of muscle blood flow reserve has been docu-
mented by radionuclide techniques to be the first pathophysi-
ologic sequela of occlusive arterial disease resulting in
intermittent claudication. 6 Early investigations of peripheral
perfusion scanning of pressure stocking-induced reactive
hyperemia using arterial injection of radionuclide-labeled
albumin microspheres revealed 83% agreement with patient
symptomatology and 40% correlation with arteriography. 7
The authors stressed the importance of assessing perfusion at
the microcirculation level, permitting evaluation of arterial
lesion significance as well as the effect of both visualized
and non visualized collateral channels.

Analogous to decreased coronary flow reserve associated
with ischemic heart disease, the assessment of peripheral per-
fusion by stress-redistribution thallium-201 ( 201 T1) imaging is
a logical extension of this clinical procedure. Preliminary stud-
ies have reported the diagnostic accuracy of stress peripheral
perfusion scanning; however, extensive clinical series have
not been forthcoming. 8-12 Measurement of limb blood flow
during reactive hyperemia may be performed using 99m Tc au-
tologous erythrocytes or human serum albumin. 13 ' 14 Clinical
applications include: screening both symptomatic and
high-risk asymptomatic patients; confirming physiologic
significance of arteriographic and color-flow duplex

325

pa rt 1 1 Noninvasive vascular diagnostics

Table 31.1 Radionuclide imaging evaluation of peripheral vascular disease

Technique

Clinical assessment

Radionuclide angiography

Stress perfusion imaging

Radioxenon clearance

Tissue perfusion-blood pool
Tissue hyperemia

Vessel patency

Perfusion dynamics and distribution
Shunt evaluation and quantification
Limb blood flow

Screening arterial insufficiency
Physiologic significance of arterial
lesions and collateral channels

Amputation level selection

Tissue viability

Ulcer healing potential

Simultaneous blood inflow-outflow Vasculogenic impotence

Lymphoscintigraphy

Autologous leukocyte imaging
Autologous platelets

Peripheral edema diagnosis
Microvascular lymphovenous
anastomosis

Graft infection

Vascular integrity

Table 31 .2 Selection of radiopharmaceutical agents for radionuclide
angiography

Agent

Characteristic

99m Tcsulfurcolloid

99m TcDTPA

99 m

Tcpertechnetate

99 m Tc erythrocytes/ 99 m Tc human
serum albumin

99 m

Tc macroaggregated albumin

Rapid blood clearance by
reticuloendothelial system allows
sequential assessment of flow in two or
more regions

Renal clearance allows dynamic and
static imaging in two or more regions

Dynamic and static imaging in one
region

Evaluation of regional perfusion and
blood flow

Dynamic and static imaging in one
region with best-quality static imaging

Rapid clearance by first capillary bed
allows arteriovenous shunt
quantification

DTPA, diethylenetriaminepentaacetic acid.

ultrasonographic lesions; and evaluating improvement in
limb blood flow after revascularization.

Tissue viability: prediction of spontaneous healing

As peripheral vascular obstruction progresses, resting blood
flow decreases, threatening nutrient perfusion and tissue via-
bility. Nuclear assessment of tissue blood flow, hyperemic
response, and necrosis aid in management of end-stage
complications.

Amputation level selection

Estimation of cutaneous blood flow by measuring the clear-
ance after intradermal injection of xenon-133 ( 133 Xe) gas
dissolved in saline has been used successfully to select
amputation levels based on potential for spontaneous skin
healing. Diagnostic parameters for predicting healing are
good (sensitivity, 96%; specificity, 80%; accuracy, 95%); 15/16
however, multiple factors preclude routine clinical applica-
tion, including lack of a commercial supplier for the tracer and
the requirement for a demanding, meticulous technique. 17 Co-
existing conditions of muscle necrosis, infection, and os-
teomyelitis are important contributors to nonhealing that also
must be taken into account.

Frostbite injuries predominantly involve distal extremity
distributions. Although soft tissue injury is universal, bone
demineralization occurs as a late sequela up to 6 months after
the insult. 18 The use of 99m Tc phosphate bone scanning during

the first 3 months after deep frostbite injury has been sug-
gested to identify nonviable bone by absence of uptake. 19 A
more aggressive approach has been suggested using 99m Tc
pertechnetate flow and blood pool scintigraphy. 20 Persistent
perfusion defects (at 2 days and 2 weeks) identify nonviable
tissue requiring surgical resection. Use of these criteria or any
others to support amputation may be inappropriate, however,
in light of slow (6-12 months) spontaneous healing of most
frostbite cases. 21

Healing potential of skin ulcers

Although measurement of skin blood flow by intradermal
xenon clearance accurately predicts healing, 16 a more practical
clinical approach has been the assessment of local perfusion
status. Contrary to its name, an ischemic ulcer heals by an in-
flammatory response and hyperemia of the microcirculation.
When assessed by arterial injection of 99m Tc albumin micros-
pheres, a relative increase in ulcer perfusion of 3.5 times back-
ground predicts healing with 87% sensitivity, 90% specificity,
and 88% accuracy. 22 Similar results have been reported using
intravenous 201 T1 (thallous) chloride and a hyperemic index of
1.5. 23 The simplest and least expensive nuclear examination
reported to predict healing is 99m Tc phosphate radionuclide
angiography and blood-pool imaging. 24 Local increase in per-
fusion and blood-pool pattern predicted healing with 96%
sensitivity, 87% specificity, and 93% accuracy. A bone imaging
agent was selected to permit subsequent three- and four-phase
bone imaging to diagnose osteomyelitis.

326

CHAPTER31 Radionuclide scanning

Vascular assessment of impotence

Our increasing knowledge of the pathophysiology of vasculo-
genic impotence implies a role for vascular surgery in manag-
ing this disorder. 25 Abnormalities of both arterial insufficiency
and venous incompetence have been demonstrated, and diag-
nostic techniques are evolving to evaluate these vascular func-
tions. 26 Simultaneous measurement of corpora cavernosal
inflow with 99m Tc erythrocyte wash-in, and outflow with 133 Xe
wash-out has been reported in methodologic studies of small
patient groups. 27 ' 28

Evaluation of peripheral edema

Peripheral lymphoscintigraphy has been used successfully to
differentiate lymphatic and venous edemas, and has been rec-
ommended to replace contrast lymphangiography because of
its relative ease of performance and its accuracy (sensitivity,
92-97%; specificity, 100%). 29-31 The testis neither approved for
routine use nor standardized, although quantitative assess-
ment of lymphatic flow by limb clearance-ilioinguinal lymph
node uptake and lymphatic channel imaging patterns are re-
quired components of the examination. 99m Tc microcolloids
(antimony sulfide, rhenium sulfide, sulfur microcolloid, sul-
fur minicolloid) injected subcutaneously and imaged over 3 h,
and 99m Tc human serum albumin or dextran injected intrader-
mally and imaged over 1 .5 h have been used in peripheral lym-
phoscintigraphy. 32 Specific vascular surgery applications
include assessment of edema after arterial reconstruction and
selection of patients for microvascular lymphovenous anasto-
mosis. Suga and colleagues, investigating postarterial
reconstructive edema, demonstrated decreased lymphatic
clearance, suggesting disruption of lymphatic vessels as the
cause. 33 Lymphoscintigraphy appears to be useful in selection
and postoperative follow-up of lymphovenous anastomosis.
Patency of the anastomosis may be suggested by visualization
of lymph channels to the anastomotic site, increased clearance,
and increased liver uptake of colloid. 34 Lymphoscintigraphy
is also reported to be diagnostic of lymphangiectasia and use-
ful in documenting surgical treatment of chyle reflux. 29

Graft infection

It is recognized that both diagnosis and treatment of pros-
thetic arterial graft infection present a challenge because mor-
bidity and mortality remain high. 35 Early surgical treatment
with removal of infected material and repeat revasculariza-
tion requires accurate diagnostic confirmation. 36 Labeling au-
tologous neutrophils with indium-Ill ( m In) oxine is an
established technique for routine scintigraphic imaging of in-
fection, and has been reported by multiple investigators to be
efficacious in diagnosing graft infection (Fig. 31.1). Combining
results of 10 studies evaluating 234 grafts yields diagnostic
sensitivity of 92%, specificity of 86%, and accuracy of 88%. 37-46

Figure 31.1 1 1 1 1n WBC scan of anterior pelvis at 24 h showing infection of
aortobifemoral graft extending from proximal to distal anastomotic sites
(arrows). (From Williamson MR, Boyd CM, Thompson BW etal. 1 1 1 -In-
labeled leukocytes in the detection of prosthetic vascular graft infections. Am
J Roentgenol 1986; 147:173.)

An exception to these results is a single study of patients
screened before discharge and found to have a low specificity
of 50%. 47 These false-positive results all were associated with
femoral grafts and groin uptake that gradually resolved on
subsequent m In neutrophil scans. Although consistent with
noninfectious inflammatory wound healing and graft incor-
poration, surgical technique and graft material may have been
contributing factors. Nevertheless, these findings demon-
strate the requirement for cautious interpretation of positive
scans in the immediate postoperative period. Recognized
sources for false-positive scan interpretation include the fol-
lowing: pseudoaneurysm; hematoma; phlegmon; cellulitis;
wound healing; ischemic, infarcted, or inflamed bowel; and
accessory spleen.

Two investigational infectious imaging procedures promise
advantages over m m leukocyte imaging: 99m Tc leukocytes and
m In immunoglobulin G . Labeling autologous leukocytes with
99m Tc hexamethylpropyleneamine (HMPAO) permits earlier
imaging (2 to 3 vs. 24 h), better image quality, and more rapid
image acquisition. Two studies, including one of 20 patients
imaged immediately and at 1 week after aortic graft surgery,
and the other assessing potentially infected peripheral arterial
grafts, reported improved diagnostic parameters (sensitivity,
100%; specificity, 96%; accuracy, 97%). 48 ' 49 Fast (2-min, 30-min,

327

pa rt 1 1 Noninvasive vascular diagnostics

Figure 31.2 1 1 1 1n IgG scan of anterior pelvis at 48 h showing infection of
aortobifemoral graft at distal anastomotic sites: left greater than right
(arrows). (From Oyen WJG, ClaessensAMJ, vanderMeerJWMefa/. Indium-
1 1 1 -labeled human nonspecific immunoglobulin G: a new
radiopharmaceutical for imaging infectious and inflammatory foci. Clin
Infect Dis 1 992; 14:1110. Copyright 1 992, the University of Chicago. All
rights reserved.)

2-h, 4-h) 99m Tc HMPAO leukocyte imaging has been per-
formed in a prospective study of 80 patients evaluated for ab-
dominopelvic inflammation. 50 Overall test accuracy was 77%,
86%, and 92% through 2h, with no additional incremental
increase at 4 h.

m In-labeled human nonspecific immunoglobulin G (IgG)
obviates the need for complex, time-consuming, and poten-
tially hazardous harvesting, separating, and labeling of autol-
ogous leukocytes. Accumulated at sites of inflammation by
vascular leakage into the intravascular space and imaged at
4-48 h after administration, m In IgG has demonstrated sensi-
tivities of 91-100% and specificities of 100% in preliminary
studies (Fig. 31.2). 51

Other infectious imaging agents are being developed that
label leukocytes in vivo. These agents include radiolabeled
monoclonal antibodies to leukocyte antigens, radiolabeled
chemotactic polypeptides, and unlabeled xenopeptides (sub-
sequently targeted by high-affinity radiolabeled small mole-
cules). 52 The basis for targeting leukocytes for continued
development is their 20-fold increase in ratios of abscess to
blood and tissue compared with noncellular inflammatory
imaging agents. 53

Assessment of vascular integrity

The literature is rich in investigations of prosthetic and natural

vessel integrity with regard to endothelial maintenance, gene-
sis of atherosclerosis, thromboembolic complications, effects
of thrombolytic and platelet-inhibitory drugs, and effects of
revascularization by surgical and percutaneous techniques
using radionuclide-labeled platelets, endothelium, fibrino-
gen, lipoproteins, and immunoglobulins. 54-56

Cerebrovascular disease

The benefit of carotid endarterectomy to reduce the risk of sub-
sequent stroke is being tested in six prospective, randomized,
controlled, multicenter clinical trials of both symptomatic and
asymptomatic patients. 57 In one of these trials, the North
American Symptomatic Carotid Endarterectomy Trial, surgi-
cal benefit already has been established for symptomatic pa-
tients with high-grade (70-99%) stenosis. 58 This significant
benefit is most likely derived from both an improvement in the
hemodynamic status of the ipsilateral cerebral circulation, as
well as reduced atheroembolic potential. A second objective of
this trial is to determine whether the degree of angiographi-
cally defined carotid stenosis can discriminate patients who
will benefit most from surgery. 59 Because intracerebral collat-
eral circulation (which is variable) determines compromise to
ipsilateral cerebral circulation, no degree of carotid stenosis
consistently alters intracerebral hemodynamics. 60 Thus it is
important that investigations of prognosis and therapy selec-
tion for ischemic cerebrovascular disease include assessment
of intracerebral perfusion hemodynamics. Radionuclide
positron emission tomography (PET) and single photon emis-
sion computed tomography (SPECT) studies of cerebrovascu-
lar reserve and cerebral risk can provide the database for this
assessment.

Cerebral hemodynamics

The study of cerebral blood flow and oxygen metabolism by
PET imaging provides an understanding of the pathophysio-
logic mechanisms of ischemic cerebrovascular disease. 60-64
Regional measurement of cerebral blood flow (rCBF), cerebral
blood volume (rCBV), oxygen extraction fraction (rOEF), and
cerebral oxygen metabolism (rCMR0 2 ) permits characteriza-
tion of the cerebral vulnerability and viability associated with
large artery atherosclerosis. Cerebral blood flow remains con-
stant under varying cerebral perfusion pressures by an auto-
regulatory mechanism (Fig. 31.3). When large artery stenosis
reduces cerebral perfusion pressure (rCPP), arteriolar vasodi-
lation (increased rCBV) maintains rCBF. As rCPP continues to
fall, the ability of compensatory vasodilation to maintain rCBF
is exceeded. At that point, rOEF increases to maintain
rCMR0 2 . When oxygen extraction is maximal, further reduc-
tion in perfusion pressure results in depressed oxygen metab-
olism, brain dysfunction, and tissue damage.

328

CHAPTER31 Radionuclide scanning

CBF

? CBV

(D
O

CD
Q_

-100

+100 i
+50 -

, ro;:q,., _^<
•:-:-;o:o:->:-:o;o;-> '— ^'^

■^•;';-;o:->:-:<-!o^^.J:^r^:l^r^:o:o:ox"x;
*.->>'«'■ ■■•.■■•:•■■:■•:' ■.•:•.-.•••.■•.■..■..■.■.■.

I
I

OEF

+100 T^m

+OU " 5?.. >» ? ?> < ' •< )

o(o;o:':<I :o!o!o:':o!o!' d I:'j

CMRO,

-50
-100

,<«

'o;c;xx-;
<Ss>::>i:s:"<::o:J-- *,

•?■■••: a- o.^lf

oSS:o- ^i":

o> < • iv i;

?;° ; Jf .■:'f:!x':c-;

!i!-H.!i!4»HW»W.

ill;

P»I»BRBBPW1W^

IWWWWWPBWB^SIip

30 60 90 120

Perfusion pressure (mm Hg)

Figure 31 .3 Compensatory responses to reduced cerebral perfusion
pressure (CPP). As CPP falls, cerebral blood flow (CBF) is initially maintained
by dilation of precapillary resistance vessels, resulting in increased cerebral
blood volume (CBV). At maximal compensation by vasodilation, CBF begins
to fall (line A). Oxygen extraction fraction (OEF) progressively increases,
maintaining cerebral oxygen metabolism (CMR0 2 ). At maximal OEF
compensation (line B), further decline in CPPand CBF disrupts normal cellular
metabolism and function. Dashed lines indicate conditions for which data are
inadequate to draw firm conclusions. (From Powers WJ. Cerebral
hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1 991 ;
29:231.)

Cerebral viability

Cerebral tissue viability has been characterized by measure-
ments of rCBF and rCMR0 2 in normal subjects and in patients
with transient and reversible ischemic neurologic deficit and
with infarct. 62 Values observed in viable brain demonstrate
substantial overlap with nonviable tissue. Infarcts fall within
the viable range in 60% of rCBF values and 20% of rCMR0 2
values. Therefore, the most efficacious regional cerebral he-
modynamic measurement to assess viability is rCMR0 2 . PET
imaging will provide the physiologic data necessary to docu-
ment the investigation of stroke therapy for salvaging cerebral
tissue and function. 65

Cerebral vulnerability

Decreased compensatory reserve is the hallmark of cerebral
vulnerability, and is used to assess stroke risk.

Physiologic assessment of hemodynamic vulnerability has
followed two strategies to evaluate the overall effect of carotid
stenosis and the contribution of collateral channels to regional
cerebral function. Resting cerebral hemodynamics provide a
panel of indices by which position on the autoregulatory curve

determines cerebral perfusion reserve. Alternatively, by com-
paring baseline and stress hemodynamic indices, perfusion
reserve may be evaluated in a fashion analogous to coronary
flow reserve.

Resting cerebral perfusion reserve

The autoregulatory curve has been divided into stages based
on predictable changes in measured parameters (see Fig.
31.3). 60 Stage reflects normal perfusion pressure with rCBV/
rCBF ratio and rOEF normal. Stage 1 reflects autoregulatory
vasodilation in response to decreased perfusion pressure:
rCBF and rOEF are normal with rCBV and the rCBV/rCBF
ratio increased. Stage 2 occurs when vasodilation can no
longer compensate and rCBF falls: rOEF increases to maintain
rCMR0 2 ; and rCBV and the rCBV/rCBF ratio continue to in-
crease.

All parameters may be measured by PET imaging. Measure-
ments of rCBF, rCBV, and the rCBV/rCBF ratio may be ob-
tained using SPECT and two radiopharmaceutical agents. 66-68
Only PET imaging provides rOEF measurement.

PET assessment of cerebral perfusion reserve in carotid
artery occlusion and stenosis has clearly demonstrated hemo-
dynamic abnormalities that improve with revasculariza-
tion. 69 ' 70 In addition, the stage of hemodynamic abnormality
does not correlate with angiographically determined stenosis
or obstruction. 71 Therefore, potential exists for selection of pa-
tients most likely to benefit from revascularization by using
PET-derived parameters.

Two preliminary prognostic studies of ipsilateral stroke oc-
currence within 1 year have failed to demonstrate any differ-
ence between patients with normal and abnormal stage 1
(rCBV/rCBF) cerebral hemodynamics. 72 ' 73 Both medically
treated and extracranial-intracranial bypass-treated patients
were studied. More restrictive (stage 2) abnormalities have
identified a greater stroke risk at 2 years for both patient
groups, however. 60 Assessment of resting cerebral perfusion
reserve for selection of patients for carotid endarterectomy has
not been accomplished owing to the limited availability of PET
imaging.

Measurements of rCBF, rCBV, and the rCBV/rCBF ratio by
SPECT imaging have not been applied to assess stroke risk and
prognosis; however, greater access to this technology should
allow investigators to assess the role of intracerebral hemody-
namics in the diagnosis and management of ischemic cere-
brovascular disease.

Stress cerebral perfusion reserve

Assessment of cerebral perfusion reserve by the reactive
vasodilation response to physiologic (C0 2 ) or pharmacologic
(acetazolamide) stress permits hemodynamic characteriza-
tion in the early portion of stage 1 autoregulation. Marked va-
sodilation and rCBF increase occur with stress at normal rCPP;

329

pa rt 1 1 Noninvasive vascular diagnostics

Figure 31 .4 Stress (acetazolamide) cerebral
perfusion SPECT assessment of cerebrovascular
reserve in a 48-year-old woman with bilateral
internal carotid artery stenosis due to
fibromuscular dysplasia: simultaneous
acquisition of baseline 99m Tc HMPAO and stress
123 l IMP perfusion distribution. A representative
transaxial slice demonstrates mild baseline
perfusion decrease in the left parietal region
consistent with a previous small left parietal
stroke. Stress perfusion (post diamox) is
diminished in the anterior cerebral artery
distributions bilaterally, with marked decrease in
the left middle cerebral artery distribution. Such
relative decreases are consistent with areas of
poorvasoreactivity. (Courtesy of D. Mathews,
MD, PhD, and B. Walker, MD, University of Texas
Southwestern Medical School, Dallas.)

however, this response is blunted or lost as rCPP is reduced. 74
More consistent increase in cerebral blood flow is achieved by
acetazolamide administration (1.0 g intravenously), which
produces a normal rCBF increase of 50% in young and 30% in
elderly subjects. 75 Either PET or SPECT may be used; however,
the universal availability of SPECT favors its application and
use for this examination (Fig. 31.4). Initial studies in patients
undergoing carotid endarterectomy have demonstrated
improvement (45-84%) of cerebral perfusion reserve after
surgery 76 ' 77 Large-scale studies must address prognosis
(stroke risk) in addition to patient selection (hemodynamic
risk).

Opportunities for radionuclide investigations

As management strategies evolve for treatment of ischemic
cerebrovascular disease, radionuclide imaging provides a
powerful tool to resolve controversial issues. The ultimate
goal is to improve patient selection for revascularization and
optimize medical management. The basis for future research is
discussed in the following section.

Prognosis: recovery and risk

Determination of an individual patient's prognosis includes
assessment of recovery (when presenting with ischemic se-
quelae) and of risk for future ischemic events. Investigations of
acute changes in cerebral hemodynamics and oxygen
metabolism in ischemic stroke have provided insight into
pathophysiologic changes and significant relationships to
long-term prognosis, but have not yet yielded a model by
which recovery may be predicted. 63,78 Evaluation of water-
shed vulnerability, ischemic penumbra, "stunned and hiber-
nating" cerebrum, and neurotransmitter dysfunction all are
relevant issues for study. 79,80

Studies of regional cerebral glucose metabolism near
ictus using PET and fluorine-18 fluorodeoxyglucose have
demonstrated consistent relationships to clinical outcome
with important prognostic implications. 81 SPECT studies
have predicted recovery based on retention (delayed wash-
out, redistribution, filling-out phenomenon) of both iodine-
123 iodoamphetamine ( 123 I IMP) and 99m Tc HMPAO. 82,83 A
strongly positive correlation has been reported between
clinical outcome and the difference between defects as as-
sessed with computed tomography and HMPAO SPECT: the
greater the difference, the greater the recovery 84 Assessment
of brain activation, drug intervention, and brain receptors
may lead to important applications for prognosis
determination. 85,86

Risk assessment for stroke should be based on epidemi-
ologic, hemodynamic, and atheroembolic factors. Patients
with transient ischemic attacks have been shown by SPECT
brain perfusion imaging to be at increased risk (75% vs. 0%) for
early stroke if regional ipsilateral perfusion is reduced by more
than 30% within 1-2 days after the event. 87 This finding was
present in 33% of patients with less than 75% internal carotid
artery stenosis. Prolonged hypoperfusion may be a risk factor
for early stroke. IMP redistribution— increased tracer activity
between early (0.3 h) and delayed (2h) SPECT imaging— has
been interpreted as reflecting ischemic but viable tissue and
proposed as a method to assess results of carotid endarterecto-
my 88 On the other hand, HMPAO reverse redistribution —
decreased tracer activity between early (0.25-0.5 h) and de-
layed (4h) SPECT imaging— is interpreted as reflecting mild
ischemia (early stage l). 89 It is postulated that IMP retention is
related to delayed wash-out of metabolic products, whereas
HMPAO loss reflects blood clearance in regions of increased
blood volume secondary to autoregulatory changes. HMPAO
SPECT easily identifies the extent of hypoperfusion secondary
to vasospasm in patients with subarachnoid hemorrhage. 90,91

330

CHAPTER31 Radionuclide scanning

Figure 31.5 Carotid occlusion cerebral
perfusion SPECT assessment (Matas test) of
collateral circulation reserve in a 59-year-old
woman with left cavernous carotid aneurysm:
separate acquisition of baseline and balloon
occlusion 99m Tc HMPAO perfusion distribution.
A representative transaxial slice demonstrates
extensive perfusion decrease in the left anterior
and middle cerebral arterial distributions and in
the right cerebellum (arrow), representing
crossed cerebellar diaschisis. Based on the poor
collateral reserve demonstrated in this study,
platinum coil therapy was selected ratherthan
carotid artery ligation. (Courtesy of D.
Mathews, MD, PhD, and B. Walker, MD,
University of Texas Southwestern Medical
School, Dallas.)

SPECT also is helpful in evaluating adequacy of cerebral col-
lateral circulation (Matas test) in patients in whom carotid
artery ligation is anticipated. 92 Precompression and postcom-
pression obliteration examinations are obtained using two in-
jections of tracer (Fig. 31.5). Although no procedure assesses
thromboembolic potential and risk, radionuclide studies of
vascular integrity may find applications in this important area
of stroke risk determination.

Table 31 .3 Diagnostic parameters for tests proposed to screen for
renovascular hypertension

Renovascular hypertension

The management of renovascular hypertension (RVH) is con-
founded by its difficult diagnosis (low prevalence among hy-
pertensive patients, with difficult separation from essential
hypertension) and its unrealized potential for cure by revascu-
larization (success rates are 26% by percutaneous translumi-
nal angioplasty and 37% by surgical reconstruction). 93 ' 94
Given the estimated prevalence of RVH at 0.5% of the hyper-
tensive population, a diagnostic test must have near-perfect
specificity to be an effective screening modality. The diagnos-
tic specificity (ability correctly to identify those without the
disorder) of most tests is well below 100%, which is to be ex-
pected when dealing with biologic systems. Tests that have
been suggested to screen for RVH all fall below the ideal speci-
ficity (Table 31.3). The highest reported specificity of 95%
(captopril-enhanced plasma renin activity 95 and captopril
renography 96 ) falls short when applied to a nonselected hy-
pertensive population. Only 9% of all positive tests actually
are in patients with RVH; for each patient correctly identified
with RVH, 10 are misdiagnosed as false-positives. To rectify
this problem, and in the hopes of establishing preselection cri-
teria, the Cooperative Study of Renovascular Hypertension
identified clinical characteristics more frequently associated

Sensitivity (%)

Specificity (%)

PRAt

Rapid-sequence urogramt

131 l hippuran renographyt

99 m Tc DTPA renographyt

Captopril renography

Intravenous digital subtraction

angiographyt

Doppler ultrasound

Captopril PRA*

100

Furosemide-captopril-hippuran

renography

PRA, plasma reinin activity; DTPA, diethylenetriaminepentaaceticacid.

* Unreliable in renal insufficiency.

t Data reprinted by permission of Elsevier Science Publishing from Vidt DG.

The diagnosis of renovascular hypertension: a clinician's viewpoint. Am J

Hypertens 1 991 ; 4:663. Copyright 1 991 . American Journal of Hypertension,

Inc.

with RVH than with essential hypertension (Table 31.4). 97
When adjusted for prevalence of disease, however, these char-
acteristics are still 35-238 times more likely to reflect essential
hypertension. To improve the ability to detect pretest risk for
renal artery stenosis (RAS) in hypertensive patients submitted
to radionuclide imaging, various authors have successfully
applied combinations of clinical characteristics or clues that
increase RAS pretest risk to an optimal 46-64%. 98-100 Specific
combinations of clinical clues, however, have neither been
characterized as to exact risk nor submitted to a prospective
study to confirm increased pretest risk. Prevalence data of spe-

331

pa rt 1 1 Noninvasive vascular diagnostics

Table 31 .4 Frequency of clinical characteristics in hypertensive patients

Prevalence-

Frequency of

adjusted

association (%)

frequency (%)

Characteristic

Essential

RVH

Essential

RVH

Acceleration of

hypertension*

Age at onset*

>50 years

<20 years

100

Bruit*

49*

Fundi (grades 3 and 4)*

Negative family history*

Recent onset (<1 year)*

22*

Sex (female)*

Cigarette smoking§

Resistant hypertension§

RVH, renovascular hypertension.

* Significant difference established for atherosclerosis and fibromuscular
dysplasia (P< 0.05).

* Significant difference established for atherosclerosis only (P< 0.01 ).

* Frequency data from Simon N, Franklin SS, Bleifer KH etal. Clinical
characteristics of renovascular hypertension. JAMA 1 972; 220:1 209-1 218.
Copyright 1 972, American Medical Association.

§ Frequency data from Setaro JF, Saddler MC, Chen CC etal. Simplified
captopril renography in diagnosis and treatment of renal artery stenosis.
Hypertension 1991; 18:289-298. Copyright 1991, American Heart
Association.

cific clinical findings (see Table 31.4) suggest that an optimal
pretest risk of 50% may be difficult to achieve solely by apply-
ing highly selective clinical criteria. 101 One study has de-
scribed the typical history characteristics of its high-risk RAS
population, listing the risk factors of cigarette smoking,
abdominal bruit, and unexplained increased or medically
uncontrolled hypertension. 98 According to reported symptom
frequency, all these symptoms should be associated with es-
sential hypertension 99% of the time (see Table 31.4). Thus,
other preselection factors must be responsible for the high in-
cidence of RAS reported. Prospective studies are needed to es-
tablish clinical predictors and pretest risk. Until pretest risk
can be reliably estimated, RAS-RVH incidence must be estab-
lished for individual populations referred for diagnostic
imaging to avoid high false-positive rates during screening.

Angiotensin-converting enzyme-inhibited
renal scintigraphy

The captopril renogram identifies kidneys (or segments of kid-
neys) that have compensated for the reduction of renal perfu-
sion pressure due to renal artery stenosis. The preferential

efferent arteriolar vasoconstriction by angiotensin II main-
tains glomerular filtration by increasing the glomerular pres-
sure gradient when renal perfusion pressure is decreased.
Angiotensin-converting enzyme (ACE) inhibitors such as cap-
topril and enalapril block formation of angiotensin II and its
protective maintenance of perfusion pressure. In kidneys with
physiologically important RAS, glomerular filtration rate de-
creases and reabsorption of water and sodium increases. 102 ' 103
These findings form the basis for the use of filtered solutes such

as Wm Tc diethylenetriaminepentaacetic acid (DTPA) to charac-
terize renal response to ACE inhibitors. The basis for secreted
agents used in captopril renography such as 131 I orthoiodohip-
purate and 99m Tc mercaptoacetyltriglycine is parenchymal re-
tention secondary to increased water reabsorption. 104 Thus,
both filtered and secreted radiopharmaceuticals are being in-
vestigated for their ability to detect RAS and RVH. Diagnostic
sensitivities and specificities have been reported from multi-
ple centers using both filtered and secreted tracers to range
from 90% to 96%. 96 ' 99 ' 100 ' 105

The test is reliable in patients with azotemia, bilateral RAS,
and intrinsic renal disease. Chronic ACE inhibitor therapy
has been shown to decrease the test's sensitivity. Although test
sensitivity and specificity are high in groups preselected
for high incidence of RAS-RVH, findings may be abnormal
in other conditions in which the rennin-angiotensin-
aldosterone axis is activated to maintain renal function,
including volume depletion, hypotension, and renal vas-
cultis. 106 It has been reported that curable RVH in renal trans-
plant recipients can be reliably differentiated from native
kidney hypertension, chronic rejection, and recurrent renal
disease. 107 Uniform study protocols and interpretive criteria
have not yet been established for ACE-inhibited renal scin-
tigraphy and await the establishment of clinical efficacy.
Until this is achieved, a consensus report of the American
Society of Hypertension suggests interpretive criteria based
on a renogram grading system. 108 Deterioration of grade after
captopril administration suggests hemodynamically signifi-
cant RAS with high probability. Improvement over baseline
after ACE inhibition is low probability, and an abnormal grade
unchanged by ACE inhibition is indeterminant. A diagnostic
algorithm based on parenchymal retention of secreted agents
(determination of residual cortical activity at 20min) yields
sensitivity of 95% and specificity of 96% in preselected patients
(Fig.31.6). 96

A typical study of a patient with RVH is presented in Figure
31.7. Although potential exists for ACE inhibition inducing
hypotension or acute renal failure in patients with high-grade
stenosis or renal insufficiency, the risk is not as great as
that with contrast agents during angiography 109 ' 110 Cerebral
ischemia has been reported in patients with carotid occlusive
disease. 111

332

CHAPTER31 Radionuclide scanning

Figure 31.6 Diagnostic algorithm for work-
up of patients with suspected renovascular
hypertension. Residual cortical activity (RCA),
expressed in percentage as the ratio of 131 l
hippuran activity in the cortex at 20 min vs.
peak, determines subsequent management.
ACE, angiotensin-converting enzyme; RVH,
renovascular hypertension. (From Erbsloeh-
MoellerB, Dumas A, Roth Detal. Furosemide-
131 l hippuran renography after angiotensin-
converting enzyme inhibition for the diagnosis
of renovascular hypertension. Am J Med 1 991 ;
90:23.)

RCA i 30%
RVH unlikely

Angiography

ACE Inhibited

Hippuran

Renogram

RCA 100%

Peak at 20 min.
RVH likely

RCA 31-100%
Peak before 20 min.

Serum creatinine mg/dl

z 1.5
RVH likely

Angiography

RCA decreased
RVH likely

Angiography

t 1.8

Baseline
Hippuran

Renogram

RCA unchanged
or increased

Renal disease

Figure 31.7 Angiotensin-converting enzyme
(ACE)-inhibited 131 l hippuran renography in a
58-year-old hypertensive woman, showing
peak (P) and 20-min (T) scans and renograms
(R). (A) Normal pretreatment baseline (right and
left residual cortical activity [RCA] of 21 % and
20%, respectively). (B) Abnormal pretreatment
ACE inhibition (right and left RCA of over 100%
and 26%). (C) Normal postreatment baseline
(rightandleftRCAof29%and27%). (D)
Normal postreatment ACE inhibition (right and
left RCAof 23% and 23%). Critical right renal
artery stenosis was treated by percutaneous
transluminal renal angioplasty.

Prediction of revascularization outcome

The physiologic basis of ACE-inhibited renal scintigraphy
allows it to assess renal functional reserve and identify pa-
tients with functionally significant RAS. It is well documented
that RAS does not establish renovascular hypertension. 112 ' 113
Initial studies of ACE-inhibited renography have reported
good results in predicting both success (positive predic-
tive value, 83-97%) and failure (negative predictive value,
72-81%) of revascularization. 100 ' 114 Potential clinical applica-
tions have been identified by the consensus conference: (i) de-
termination of physiologic significance of angiographic
abnormalities; (ii) determination of culprit lesion in bilateral
RAS; (iii) prediction of clinical outcome (curability) from

revascularization; (iv) long-term follow-up of RAS-RVH; and
(v) safer diagnostic evaluation in high-grade stenosis, bilateral
disease, and renal insufficiency. 115

Venous thromboembolism

The natural history of venous thrombosis and pulmonary
embolism provides the rationale for current diagnostic and
management strategies, based on the following observations:
(i) pulmonary embolism (PE) is a complication of venous
thrombosis; (ii) deep venous thrombosis (DVT) of the lower
extremities is the predominant source of PE; (iii) residual DVT
clot burden is the major determinant of reembolization

333

pa rt 1 1 Noninvasive vascular diagnostics

morbidity and mortality; and (iv) most deaths from PE occur
within several hours of the event. 116 These concepts support
the need to document clot burden in both the lungs and lower
extremities with therapy directed at preventing subsequent
complications.

Diagnosis of pulmonary embolism

Strategies have stressed the need to estimate the probability of
PE by noninvasive screening with ventilation-perfusion lung
scanning. Scans are characterized by specific criteria to clas-
sify them as high (over 90%), low (less than 10%), and inter-
mediate in probability for detecting acute PE, as confirmed by
pulmonary angiography. 117 ' 118 A large prospective study, the
Prospective Investigation of Pulmonary Embolism Diagnosis
(PIOPED), determined the diagnostic power of this classifica-
tion, and is summarized in Table 31. 5. 119 In this scheme, pa-
tients with high-probability scans would be anticoagulated,
those with low-probability scans would be observed, and
those with intermediate-probability scans would require in-
vasive confirmation with pulmonary angiography. In theory,
this scheme requires a large number of patients to undergo an-
giography —39% of all patients scanned in the PIOPED study.
This percentage might be reduced to 29% if clinical pretest
likelihood is used together with lung scan results to determine
post- test likelihood. In practice, few of these patients undergo
angiography (the minority, as reported in an academic
medical center). 120 Thus, physicians are managing this inter-
mediate (indeterminant) group on clinical grounds only, and
treating empirically. According to the PIOPED study, one-
third of indeterminant scans have angiographically proved
emboli.

A controversy exists regarding the frequency of PE associ-
ated with low-probability scans. In a long-term investigation
of venous thromboembolism, the McMaster University group
has maintained that patients with low-probability scans have
frequencies of PE of 25-40% of total cases. 121 ' 122 High positive

Table 31.5 PIOPED study determination of pulmonary embolism frequency
association with ventilation-perfusion lung scan interpretation category

Post- test probability (%)

Scan category

(grou

ped by clinical pretest likelihood)

(% of total scans)

High

Intermediate Low

Total group

High (13)

88 56

Intermediate (39)

26 16

Low (34)

16 4

Normal (14)

6 2

(Modified from the PIOPED Investigators. Value of the ventilation/perfusion
scan in acute pulmonary embolism: results of the prospective investigation of
pulmonary embolism diagnosis [PIOPED]. JAMA 1 990; 263:2753-2759.)

predictive values and frequency of PE associated with low-
probability classification most likely reflect differences in scan
interpretation and category criteria. Reported values for low-
probability scans have varied from 4% to 24% for positive
predictive value and from 1% to 53% for percentage of total
cases. 119 ' 122-125 Perhaps reader variability and suboptimal fre-
quency distribution can be improved with simplification and
refinement of scan interpretation criteria. 126 A modification of
the Biello criteria has reported high-, intermediate-, and low-
probability scan positive predictive values of 94%, 52%, and
3%, and percentages of total PE cases of 48%, 51%, and 1%,
respectively 124

Although the McMaster experience is at variance with
most other reports, and is probably due to scan interpretation
criteria differences, it has stimulated a new approach to
evaluating the patient with suspected PE. Investigations
involving over 300 patients with low-probability lung scan
findings and not placed on anticoagulation therapy have
demonstrated the clinical course to be uniformly benign
with no evidence of subsequent PE. 127-129 In light of this
good prognosis, the absence of diagnostic distinction between
their low- and intermediate-probability groups, and the
reticence of physicians to request pulmonary angiography,
the McMaster group tested the hypothesis "that clinically
evident recurrent venous thromboembolism is unlikely
in the absence of proximal-vein thrombosis and, in such
patients, treatment is not required, even though pulmonary
embolism may be present." 130 In a prospective study of
874 consecutive patients with suspected PE, 370 subjects with
abnormal but nonhigh-probability lung scans and no evi-
dence of proximal DVT on serial (2 weeks) impedance plethys-
mography (IPG) were not anticoagulated. This group was
compared with a control group of 315 subjects with normal
scans and IPG. At 3-month follow-up, venous thromboem-
bolic events (PE, DVT) occurred in 2% of the abnormal scan-
untreated group and 1% of the normal control group. Eleven
percent of the initial nonhigh-probability abnormal lung scan
group was identified to have proximal venous thrombosis
by IPG (8% at entry and 3% on serial follow-up). The high-
probability lung scan group of 69 subjects was treated with
anticoagulation without further confirmation by pulmonary
angiography.

Diagnostic-management algorithm

Kelley and associates have proposed a diagnostic algorithm
(Fig. 31.8) based on a summary of knowledge regarding
the diagnosis of PE (Table 31. 6). 126 This post-test probability
scheme (supported by the PIOPED data in Table 31.5)
depends on a standardized protocol for ventilation-perfu-
sion lung scan performance and interpretation. Use of the
following imaging protocol and interpretation criteria
established for the PIOPED investigation should yield similar
results.

334

CHAPTER31 Radionuclide scanning

Normal

Lung-scan
result

(<1*)

Low probability
(15-30%*)

Intermediate

(>30%*)

High probability

(>85%*)

^ Cardiopulmonary

disease? /*-

Previous PE or low suspicion

(<75%**)
No previous PE, high suspicion

Serial IPG

W9£

(>90%**)

^ Angiogram or serial
IPG***

■► Treat

(<3%*)

Treat

'Strongly supported by clinical studies.

"Suggested by clinical studies, needs confirmation

**A serially negative IPG result may not be sufficient to rule out thromboembolism.

Figure 31 .8 Management algorithm for pulmonary embolism (PE), with
post-test likelihood values given in parentheses. Normal scans and the
combination of low-probability scan and low clinical suspicion effectively rule
outPE. High-probability scans support treatment except when combined
with low clinical suspicion or previous PE. All other patients require

pulmonary angiography or noninvasive studies of deep venous thrombosis
for assessment of reembolization potential. IPG, impedance
plethysmography. (From Kelley MA, Carson JL, Palevsky HI etal. Diagnosing
pulmonary embolism: new facts and strategies. Ann Intern Med 1 991 ;
114:300.)

Table 31.6 Current knowledge of pulmonary embolism diagnosis

1 . Normal lung scans and pulmonary angiograms exclude clinically
important pulmonary embolism.

2. High-probability lung scans indicate an approximate 85% probability
of pulmonary embolism.

3. Using the clinical assessment of pretest likelihood may improve the
accuracy of scan results.

4. Noninvasive investigation of proximal deep venous thrombosis may
improve management in patients with nondiagnostic lung scans.

(From Kelley MA, Carson, JL. Palevsky HI etal. Diagnosing pulmonary
embolism: new facts and strategies. Ann Intern Med 1991; 1 14:300-306.)

PIOPED VENTILATION-PERFUSION LUNG SCAN PROTOCOL

• Ventilation (before perfusion)

• 133 Xe

• Posterior single breath and equilibrium

• Posterior and both posterior obliques wash-out

• Perfusion

• 99m TcMAA

• Eight-projection imaging (anterior and posterior, both
anterior obliques, both posterior obliques, both laterals)

Techniques using radioaerosol, 127 Xe, krypton-81m assess-
ment of ventilation (before or after perfusion imaging), or
fewer than the eight-projection perfusion imaging sequence
may vary in diagnostic power. Although the PIOPED clinical

estimation of pretest likelihood when combined with scan
findings substantially improved the diagnostic ability of the
scan alone, the method of assigning a pretest likelihood has
not been reported. Determining pretest likelihood needs fur-
ther investigation and refinement. Assessing the presence of
proximal DVT with IPG is well established; however, due to
low sensitivity (20%) in evaluating the calves, serial studies
(seven examinations over 2 weeks) have been proposed to
identify patients who subsequently progress to more proximal
involvement. 131-133 Although demonstrated to be equivalent
to 24- and 72-h 125 I fibrinogen uptake, serial IPG has not been
vigorously compared in clinical utility and cost effectiveness
with the more accessible procedures of ultrasound and radio-
nuclide venography. Further investigation is thus required for
the noninvasive assessment of DVT.

Caveats

Matched defects

Reflex bronchoconstriction occurs early (initial hours) in the
course of PE in response to low alveolar carbon dioxide. 134
Although uncommon, it may occasionally result in a matched
segmental ventilation-perfusion abnormality of a slow air-
space. 135,136 Matched segmental defects thus may reflect PE.

Restoration of perfusion

Complete pulmonary arterial occlusion does not always

335

pa rt 1 1 Noninvasive vascular diagnostics

occur. Even when it does, endogenous fibrinolysis can quickly
promote partial reperfusion, observed as decreased but not
absent segmental perfusion. 137 Furthermore, resolution may
not be complete with slower reperfusion, reported in patients
with decreased cardiac output. 138 ' 139 The Urokinase Trial
documented 36% resolution in 5 days, 52% in 2 weeks, and
73% at 3 months, with 24% residual occlusion at 1 year. 140
Thus, obtaining baseline predischarge perfusion lung scans
with subsequent examinations as indicated in both high-risk
and chronic venous thrombosis patients may permit the future
diagnosis of reembolization to be made by comparison of lung
scans. Although patients with preexisting or concurrent car-
diac or pulmonary disorders may be at greater risk for compli-
cations of PE and have slower restoration of perfusion, the
diagnostic utility of ventilation-per fusion lung scanning is
not impaired in this group. 141

Baseline lung scans

Asymptomatic PE may occur in association with DVT. An inci-
dence of 43% abnormal lung scans (high and intermediate
probability) was reported in a prospective study of 116 con-
secutive patients with iliofemoral DVT without symptoms of
PE. 142 Six patients subsequently became acutely symptomatic
with either no change or improvement in their lung scans.
These patients might have received irreversible caval surgery
based on acute scan findings alone. These authors warn of this
possibility, and recommend baseline lung scans in asympto-
matic patients with DVT.

False-positive lung scans

Selective pulmonary vascular obstruction may occur as a
result of many pathologic processes in addition to acute and
chronic thromboembolism. Probably the most frequent cause
of a false-positive, high-probability scan reading is residual
pulmonary arterial occlusion from previous PE. 143 For this rea-
son, patients with previous PE and high-probability scans
are placed in a lower post-test risk category (see Fig. 31.8).
External vascular compression (tumor, hematoma, sarcoid),
primary vascular disorders (vasculitis, agenesis, stenosis,
arteriovenous malformation, venoocclusive disease), tumor
embolus, and wedged Swan-Ganz catheter are other fre-
quently cited causes. 144 ' 145

Venous thrombosis

The natural history of venous thrombosis supports the man-
agement algorithm of prophylaxis for asymptomatic, high-
risk patients and anticoagulation for patients diagnosed with
acute venous thrombosis. Although high-risk patients may
be identified by specific population (e.g. hip replacement
surgery), diagnosis of active thrombogenesis poses more of
a problem. 116 Because only half of clinically suspect patients
have venous thrombosis, 146 and half of patients with venous

thrombosis are asymptomatic, 147 confirmatory and screening
diagnostic testing must be used. Unfortunately, no noninva-
sive test has been shown to approach the accuracy of contrast
venography, especially in evaluation of calf thrombosis. The
clinical practice of serially assessing symptomatic patients
until disease is identified in the popliteal veins is thus dictated
more by limitations of noninvasive testing than by the natural
history of venous thrombosis. In fact, autopsy-based studies
have documented that 13-15% of fatal PE originate from DVT
in the calves. 148 ' 149 In a prospective study of patients with sus-
pected DVT, 62% were confirmed to have DVT; 56% of DVT
were limited to the calves; and 46% of patients with DVT lim-
ited to the calves had high-probability ventilation-perfusion
lung scans. 150 Therefore, methods that accurately identify calf
venous thrombosis will find widespread clinical application
and are being sought.

Three diagnostic techniques (contrast venography, IPG, and
125 I fibrinogen uptake) are considered to have confirmed diag-
nostic power and clinical utility 116 Because of its invasive
nature and complications, contrast venography serves more
as the test by which noninvasive procedures are compared,
and is not used in screening. IPG has demonstrated excellent
sensitivity and specificity for identifying proximal (femoral)
deep venous occlusion in patients suspected of having DVT,
but poorly detects calf involvement (20% sensitivity). 131 IPG
has been used in combination with the 125 I fibrinogen uptake
test (FUT) to improve clinical utility in patients with suspected
DVT. 151 When applied to an asymptomatic, high-risk popula-
tion (total hip replacement), however, the diagnostic parame-
ters for the combined IPG-FUT examination are low (IPG
thigh thrombosis sensitivity of 12%; FUT calf thrombosis
sensitivity of 59%; and combined total sensitivity of 47%). 152
Furthermore, IPG usually is not available outside of large
medical centers, and 125 I fibrinogen is no longer commercially
available because of the theoretical risk of hepatitis and
human immunodeficiency virus transmission in a pooled
blood product. 153 ' 154

Clinical evaluation of DVT relies on confirmation tests that
are more universally available: radionuclide venography and
compression B-mode ultrasound. These anatomic tests fail to
distinguish between new and old clots and are relatively in-
sensitive in the calves. Investigation continues for a test with
acceptable clinical utility in the calves and for a test that direct-
ly images venous thrombosis.

Radionuclide venography

Two nuclear imaging tests that indirectly identify the presence
of DVT are used clinically. Flow radionuclide venography
(FRV) evaluates antegrade venous flow after bolus dorsal
pedal vein injection of 99m Tc MAA. Perfusion lung imaging
may be performed subsequently using the same tracer injec-
tion to identify PE. Equilibrium radionuclide venography
(ERV) images deep venous vessel distribution and volume

336

CHAPTER31 Radionuclide scanning

Table 31 .7 Radionuclide venography diagnostic parameters for lower
extremity: comparison with contrast venography

Technique

Patients
studied

Location

Sensitivity
(%)

Specificity
(%)

Flow radionuclide

328

Total

venography

150

Proximal*

100

Calf

Equilibrium

232

Total

radionuclide

Proximal*

venography

94
97
95

84
94

Proximal to and including popliteal vein.

after a peripheral injection of 99m Tc pertechnetate to label
stannous-prepared erythrocytes in vivo. ERV is less technically
demanding than FRV, but cannot be followed by pulmonary
perfusion imaging. Both tests have excellent diagnostic para-
meters for proximal (proximal to and including the popliteal
vein) DVT (Table 31.7). 155-165 FRV may be used with higher
confidence to diagnose calf DVT. Combined FRV and ERV
have been reported to improve DVT detection further. 166 ' 167
Because of its high diagnostic accuracy, radionuclide venogra-
phy has been proposed for screening DVT, with negative tests
suggesting absence of clinically important DVT. Two studies
have shown the complete absence of morbidity and mortality
in 181 patients with negative radionuclide venograms. 168 ' 169

Interpretation criteria include major findings of absence or
cut-off of flow or vessel, and minor findings of collateraliza-
tion, irregular or decreased filling, tracer retention (FRV), or
blood-pool increase below the DVT level (ERV). One major or
three minor findings are required for a high-probability DVT
interpretation. Advantages of radionuclide venography in-
clude high patient acceptance, low morbidity, ability to assess
pelvic veins and inferior vena cava, a permanent record of
deep venous distribution for subsequent comparison to assess
resolution or progression of DVT, and the ability to evaluate
simultaneously for the presence of PE (FRV). Because radio-
nuclide venography indirectly identifies DVT, it cannot
distinguish intrinsic (DVT) from extrinsic (i.e. popliteal cyst)
processes. As with all other diagnostic imaging approaches,
it neither directly images thromboses nor distinguishes new
from old DVT.

Thrombus imaging

Radionuclide techniques to image thrombi are directed at the
two major constituents, platelets and fibrin (Table 31.8). By
labeling these components or their precursors, the resulting
radiopharmaceuticals may be used to image forming thrombi.
Applied to high-risk patient groups, these tracers screen for
silent DVT. A second group of radiopharmaceuticals identifies

Table 31.8 Carrying agents for thrombus-imaging radiopharmaceuticals

Agent

Forming thrombi Preformed thrombi

Fibrin-directed Fibrinogen

Platelet-directed

Autologous platelets
Antiplatelet

monoclonal

antibodies

to activated

platelets

Soluble fibrin

Antifibrin monoclonal antibodies

Fragment E 1

Plasmin

Plasminogen activators

Heparin

Fibronectin

Antiplatelet monoclonal
antibodies to circulating
platelets

preformed thrombi. These tracers, which include monoclonal
antibodies to platelets and fibrin, are being developed to iden-
tify thrombi in patients suspected of having DVT. Images are
taken within hours of radiopharmaceutical administration.

Detecting forming thrombus

The prototype radiopharmaceutical agent used to identify
clinically silent forming thrombi, 125 I fibrinogen, is detected
with a scintillation probe and measured as a percent of precor-
dial counts in the FUT. The diagnosis is established in 3-4 days,
when thrombus concentration exceeds that of background. 170
The advantages of the FUT, including cost, portability, and
high accuracy for calf DVT evaluation, have established its
utility as a screening test for forming DVT. Attempts to extend
the diagnostic utility to the remainder of the body using imag-
ing radionuclides ( 131 I, 123 I, 99m Tc, and gallium-67) have met
with limited success. Further development of fibrinogen
products probably will cease because human fibrinogen is a
pooled blood product with a theoretical risk for transmitting
both hepatitis and immunodeficiency viruses, and is no longer
commercially available.

Autologous and donor platelets labeled with m In oxine
provide the only approved method for imaging forming
thrombi. The clinical utility of this radiopharmaceutical as a
surveillance test in nonheparinized, high-risk patients has
been demonstrated in two studies of postabdominal, pelvic,
and major lower limb orthopedic surgery patients. 171,172 In 274
patients studied, 112 underwent other confirmatory proce-
dures to demonstrate diagnostic sensitivity of 96% and speci-
ficity of 98% for identifying peripheral DVT and PE (Fig. 31.9).
In a smaller subset undergoing heparin therapy, these para-
meters fell to 42% and 67%, respectively. Major drawbacks, in-
cluding the time, complexity, and expense of cell separation,

337

pa rt 1 1 Noninvasive vascular diagnostics

Figure 31.9 1 1 1 1n platelet scan (A) of anterior
calves at 24 h, showing focal accumulation in
the left calf corresponding to clot identified by
contrast venography (B) as a filling defect.
(Courtesy of M.D. Ezekowitz, MD, PhD, Yale
University School of Medicine, New Haven.)

handling, and labeling, and the required 24-72-h delay in im-
aging will most likely relegate this procedure to use primarily
for investigations, although it has the demonstrated advan-
tage of providing total-body postoperative thromboembolism
surveillance by a single method. Radionuclide-labeled anti-
platelet antibodies have been developed that can selectively
bind platelets when administered intravenously Obviating
the need for costly and time-consuming ex vivo labeling, circu-
lating platelets have been tagged after the intravenous admin-
istration of 99m Tc-labeled monoclonal antibody fragments that
recognize platelet surface antigens. 173 Using this in vivo label-
ing technique in dogs, fresh thrombi were imaged within
hours, with high thrombus-to-blood ratios.

Detecting preformed thrombus

The clinical need to obtain diagnostic information and initiate
therapy within a few hours for patients suspected of having
DVT precludes the use of labeled fibrinogen and platelets,
which have both decreased affinity for thrombi older than 1
day and require 24-72 h to permit optimal deposition and
blood clearance for successful imaging. 174 Radiopharmaceuti-
cal agents being developed to image thrombi rapidly recog-
nize molecular sites on thrombus (fibrin or activated platelets)
not present on circulating precursors (fibrinogen and non-
activated platelets), and are fragments with fast clearance
from blood and tissues and low immunogenicity and
risk. 175 ' 176

Assessment of cardiac risk and prognosis

The indirect assessment of coronary flow reserve with stress
201 T1 myocardial perfusion scintigraphy is the recognized
standard for diagnostic and prognostic evaluation of patients
with coronary artery disease. 177 The ability of this test to reflect
myocardium at risk has been documented in a large series of
patients. 178 In 1689 patients with clinical evidence of coronary
artery disease but without previous myocardial infarction,
revascularization, or imaging confirmation, extent of rever-
sible perfusion defects by 201 T1 scintigraphy was found to be
significantly (P < 0.001) and exponentially correlated with the
rate of major coronary events (death, nonfatal infarction, and
bypass surgery) within 1 year of the prognostic study (Fig.
31.10). Normal stress 201 T1 results in 3573 patients with known
or suspected coronary artery disease predicts a major cardiac
event (death, infarction) rate of 0.9% per year, approaching
that of the general population. 179 Evaluation of patients with
asymptomatic coronary artery disease with dipyridamole
stress 201 T1 scintigraphy has been shown to stratify risk for
major cardiac events (death, infarction) significantly (P <
0.001). 180 Patients with reversible thallium defects experi-
enced an event rate of 12.5% per year, vs. 0.85% per year for
patients without reversible defects. Brown has summarized
findings in more than 6000 stress 201 T1 examinations reported
in patients with known or suspected coronary artery dis-
ease. 179 The presence and extent of reversible defects were

338

CHAPTER31 Radionuclide scanning

60 r

Event
rate

(%)

.-'■

.-eX>"

2 4 6

Reversible defects
(number)

N =1078 294 157 93 46 17 4

Figure 31 .10 Risk of future cardiac events (death, nonfatal infarction, and
bypass surgery) as a function of extent of reversible myocardial stress 201 TI
defects. The number of reversible defects is exponentially related to the event
rate (P< 0.001 ). (From Ladenheim ML, Pollock BH, Rozanski Aef al. Extent
and severity of myocardial hypoperfusion as predictors of prognosis in
patients with suspected coronary artery disease. J Am Coll Cardiol 1 986;
7:464.)

Table 31 .9 Preoperative assessment of cardiac risk in elective peripheral
vascular reconstruction using dipyridamole 201 TI myocardial perfusion
imaging

Study

Boucher eta/. 182
Eagle et a/. 183

Leppoefa/. 184
Sachs eta/. 185
Eagle eta/. 186
Letteefa/. 187
Total

Complications/scan category

Scan findings*

Reversible

Nonreversible

(2/1 )/1 6

(0/0)/32

(3/2)/1 8

(0/0)/43

(1/3)/24

(1/0)/26

(13/1)/42

(1/0)/47

(0/2)/1 4

(0/0)/32

(7/6)/82

(2/0)/118

(2/7)/2 1

(0/0)/39

(28/22)7217

(4/0)/337

23%

1.2%

^Complications: (myocardial infarction/death)/total patients studied.

found to be the most consistent predictors of major events,
including cardiac death, myocardial infarction, and coronary
bypass graft surgery Furthermore, 201 T1 imaging has been
shown to be superior to clinical and stress electrocardiograph-
ic assessment and equivalent to coronary arteriography Thus,
stress myocardial perfusion imaging provides an objective
means of determining which patients may benefit the most
from revascularization.

Preoperative assessment of cardiac risk

Because of the high prevalence of associated coronary artery
disease (31%), cardiac death and myocardial infarction are the
leading causes of early and late mortality after vascular recon-
structive surgery 181 Initial investigations of the clinical utility
of 201 T1 imaging to assess cardiac risk in these patients have
focused on early or perioperative cardiac events. Six patient
groups have been studied with dipyridamole stress 201 T1
imaging before major elective peripheral vascular surgery.
Reversible perfusion defects were accompanied by a 23%
incidence of major perioperative cardiac complications
(death, infarction), compared with 1 .2% in patients with no re-
versible defects (Table 31.9). Although these findings clearly
demonstrate the prognostic importance of a patient having
myocardium at risk, the cost of universal screening and subse-
quent selective cardiac catheterization is substantial. The goal
is to develop the most cost-effective and clinically efficacious
algorithm to identify these patients. Problems result from the
low incidence of major cardiac complications occurring both
in the total population (9.7%, of which 92.6% are identified by
dipyridamole stress thallium scintigraphy) and in the re-
versible defect population (only 23% of positive scans). Be-
cause it is not economically prudent to screen all surgical
candidates or to revascularize all patients with positive scans,
investigators have proposed selective screening and revascu-
larization based on specific clinical and scan criteria, respec-
tively (Table 31.10). Eagle and colleagues identified five
clinical predictors of postoperative ischemic events by
multivariate analysis (P < 0.01): advanced age (older than 70
years); Q waves on electrocardiogram; diabetes requiring

Table 31 .10 Optimization of preoperative cardiac risk stratification in elective peripheral vascular reconstruction using dipyridamole 201 TI myocardial perfusion
imaging

Study

Method

Patients screened (%)

Scan

Angiogram

100

From Table 3 1.1
Eagle eta/. 186
Levinsonefa/. 188
Cambria eta/. 189
Taylor eta/. 190

All patients scanned/positive scan -» cardiac catheterization

Clinical criteria determine scan/positive scan -> cardiac catheterization

Clinical criteria determine scan/>1 reversible territory -> cardiac catheterization

Clinical judgment determines scan/>1 reversible territory -» cardiac catheterization

Screening limited to severely symptomatic patients

339

pa rt 1 1 Noninvasive vascular diagnostics

Figure 31.11 Dipyridamole 201 TI myocardial
perfusion SPECT study in a 51 -year-old woman
with bilateral renal artery stenosis, status after
angioplasty and stents, with restenosis for
presurgical evaluation. Midventricular short axis
(SA), vertical long axis (VLA), and horizontal
long axis (HLA) slices at pharmacologic stress
and redistribution show reversible defects in
two major coronary arterial territories (arrows:
1 , left anterior descending; 2, left circumflex).
This patient is at increased risk for a major
perioperative or postoperative cardiac event.

pharmacologic therapy; history of angina; and history of ven-
tricular ectopic activity requiring therapy 186 Patients with
none of these clinical factors had a very low risk for events
(3.1%), and thus would not benefit from scanning. Patients
with three or more clinical predictors had a high risk for events
(50%), suggesting a need for revascularization on clinical as-
sessment only. Only patients with one or two clinical factors
were shown to benefit from 201 T1 imaging stratification (posi-
tive scan, 29.6% risk; negative scan, 3.2% risk). This strategy re-
duces patients screened to 58% and patients catheterized to
37%. To improve scan specificity, these investigators demon-
strated that all major cardiac events could be identified by pos-
itive scans with two or three reversible coronary territories (P =
0.007). 188 This scan criterion further reduces patients catheter-
ized to 23% (Fig. 31.11).

Preoperative evaluation of cardiac risk remains controver-
sial, however. More conservative use of screening dipyri-
damole 201 T1 scans (29% vs. 58%) and cardiac catheterization
(11% vs. 23%) has been achieved using clinical judgment in
place of clinical criteria. 189 In this study, all clinical markers
failed to show statistical correlation with major cardiac events.
Failure of other investigators to demonstrate statistical corre-
lation of clinical variables with major postoperative cardiac
events, 184 ' 191 combined with the low positive predictive value
of dipyridamole 201 T1 scanning (38%), 192 the surgical mortality
of patients with peripheral vascular disease undergoing coro-
nary artery bypass graft (5.3%), 181 the significant statistical
correlation of operative variables with major cardiac compli-
cations 189 and improvement in perioperative management,
has led certain authors to limit screening for potential revascu-

larization to patients with severely symptomatic coronary
artery disease (unstable angina, uncontrolled arrhythmias,
severe congestive heart failure, and recent myocardial infarc-
tion). 190 Cardiac symptoms should be assessed relative to ac-
tivity level to identify asymptomatic or mildly symptomatic
patients who are sedentary or deconditioned. These patients
might benefit the most from screening dipyridamole 201 T1
scanning. 193

Risk stratification algorithm

A conservative management algorithm modeled after that of
Brown is offered (Fig. 31 .12). 179 Supported by preliminary data
on the significant impact of surgical factors on perioperative
events, which include procedure (aortic reconstruction),
operation time (greater than 5h), intraoperative ischemia,
and hemodynamic instability, 189 as well as by the important
incidence (22% per year) of major cardiac events experienced
by asymptomatic patients with coronary artery disease with
reversible thallium defects, 180 all patients with aortic aneu-
rysms and aortoiliac occlusive disease might benefit from
dipyridamole 201 T1 assessment. Patients with femoropopliteal
and carotid arterial disease should be screened by clinical
criteria (e.g. those enumerated by Eagle and associates). 186
Patients with extensive myocardium at risk (multiple vascular
territories with reversible defects) identified by thallium scan-
ning should be considered for revascularization. Until con-
trolled, prospective studies are performed that assess not only
cardiac risk but therapeutic risk, routine management should
use selective application of cardiac screening procedures and

340

chapter 31 Radionuclide scanning

AAA/AIOD

DIPYRIDAMOLE-THALLIUM SCAN

Clinical Criteria:
Q-Wave

Angina
Diabetes
Vea
Age <>70)

YES

REVERSIBLE DEFECTS

>1 Territory

F-P/CAROTID

Figure 31 .1 2 Management algorithm for preoperative assessment of
cardiac risk. Assessment of pretest risk by clinical criteria determines the need
for scanning. The major stress of aortic reconstructive surgery may justify
screening of all candidates. The significantly increased post-test risk for
cardiac events in patients with more than one coronary territory with
reversible thallium defects justifies consideration of subsequent cardiac

CATHETER IZE/

REVASCULARIZE

0-1 Territory

SURGERY

catheterization and revascularization. AAA, abdominal aortic aneurysm;
AIOD, aortoiliac occlusive disease; F-P, femoropopliteal occlusive disease;
VEA, ventricular ectopic activity. (Modified from Brown KA. Prognostic value
of thallium-201 myocardial perfusion imaging; a diagnostic tool comesof
age. Circulation 1 991 ; 83:363.)

prophylactic myocardial revascularization based on clinical
assessment of cardiac risk.

Assessment of late survival

Late survival after vascular reconstructive surgery is clearly
influenced by the presence of coronary artery disease, with
cardiac death the leading cause of mortality 181/194/195 Actuarial
comparison of 5-year survival in patients with peripheral vas-
cular reconstruction and suspected coronary artery disease
with age-matched patients with myocardial revascularization
demonstrates 23-29% better survival in the coronary artery
bypass group (Fig. 31.13). By performing coronary angiog-
raphy in most patients for elective peripheral vascular recon-
struction, the Cleveland Clinic group has documented severe
coronary artery disease in 31% of the entire population and in
44% of those suspected of having coronary artery disease on
clinical grounds. 181 As a result, 216 patients underwent myo-
cardial revascularization (130 before peripheral vascular
reconstruction). Long-term survival will be evaluated in this
group to determine whether high-frequency prophylactic
myocardial revascularization benefits candidates for elective
peripheral vascular reconstruction. If so, myocardial perfu-
sion imaging should find a role in the assessment and manage-
ment of long-term survival.

100

5 s 60

g 50

CO 40

30
20
10

.972

.619

O Coronary artery bypass, ages 60-64 (N = 757)
# Carotid endarterectomy, mean age 61 (N = 142)
Suspected coronary disease

12 3 4 5

Years

Figure 31.13 Comparison of 5-year survival rates for patients undergoing
peripheral vascular reconstruction with survival rates of patients undergoing
myocardial revascularization. Significant improvement in survival is
demonstrated by the coronary bypass group. (From Hertzer NR, Beven EG,
Young JR et al. Coronary artery disease in peripheral vascular patients: a
classification of 1 000 coronary angiograms and results of surgical
management. Ann Surg 1984; 199:223.)

Conclusion

Stress 201 T1 myocardial perfusion imaging has become a pow-
erful prognostic examination that objectively performs risk

stratification. This information may be incorporated into
rational management algorithms in patients with peripheral
vascular disease to optimize cost-effective and clinically effi-
cacious care.

341

pa rt 1 1 Noninvasive vascular diagnostics

References

1. Rudavsky AZ. Radionuclide angiography in the evaluation
of arterial and venous grafts. Semin NuclMed 1988; 18:261.

2. Stewart CA, Sakimura IT, Applebaum DM et ah Evaluation of
peritoneovenous shunt patency by intraperitoneal Tc-99m
macroaggregated albumin: clinical experience. Am } Roentgenol
1986; 147:177.

3. Rhodes BA, Rutherford RB, Lopez-Majand V et ah Arteriovenous
shunt measurements in extremities. ] NuclMed 1973; 13:357.

4. Ziessman HA, Thrall JH, Yang PJ et ah Hepatic arterial perfusion
scintigraphy with Tc-99m-MAA. Radiology 1984; 152:167.

5. Wile A, Smolin M. Hyperthermic pelvic isolation-perfusion
in the treatment of refractory pelvic cancer. Arch Surg 1987;
122:1321.

6. Lassen NA, Kampp M. Calf muscle blood flow during walking
studied by the Xe-133 method in normals and in patients with
intermittent claudication. ScandJClin Lab Invest 1965; 17:447.

7. Siegel ME, Giargiana FA Jr, White RL Jr et ah Peripheral vascular
perfusion scanning: correlation with the arteriogram and clinical
assessment in the patient with peripheral vascular disease. Am }
Roentgenol 1975; 125:628.

8. Christenson J, Larsson I, Svensson S-E et ah Distribution of intra-
venously injected 201-thallium in the legs during walking: a new
test for assessing arterial insufficiency in the legs. Eur } NuclMed
1977; 2:85.

9. Siegel ME, Stewart CA. Thallium-201 peripheral perfusion
scans: feasibility of single-dose, single-day, rest and stress study.
Am} Roentgenol 1981; 136:1179.

10. Seder JS, Botvinick EH, Rahimtoola EH et ah Detecting and
localizing peripheral arterial disease: assessment of 201-T1 scin-
tigraphy. Am J Roentgenol 1981; 137:373.

11. Oshima M, Akanabe H, Sakuma S et ah Quantification of leg
muscle perfusion using thallium-201 single photon emission
computed tomography. JNucl Med 1989; 30:458.

12. Sayman HB, Urgancioglu I. Muscle perfusion with technetium-
MIBI in lower extremity peripheral arterial diseases. / Nucl Med
1991; 32:1700.

13. Parkin A, Robinson PJ, Martinez D et ah Radionuclide limb blood
flow in peripheral vascular disease: a review of 1100 measure-
ments. NuclMed Commun 1991; 12:835.

14. Gehani A A, Thorley P, Sheard K et ah Value of a radionuclide limb
blood flow technique in the assessment of percutaneous balloon
and dynamic angioplasty. Eur J Nucl Med 1992; 19:6.

15. Malone JM, Leal JM, Moore WS et ah The "gold standard" for
amputation level selection: xenon-133 clearance. / Surg Res 1981;
30:449.

16. Silberstein EB, Thomas S, Cline J et ah Predictive value of intracu-
taneous xenon clearance for healing of amputation and cuta-
neous ulcer sites. Radiology 1983; 147:227.

17. Malone JM, Anderson GG, Lalka SG et ah Prospective compari-
son of non-invasive techniques for amputation level selection.
Am} Surg 1987; 154:179.

18. Tishler JM. The soft-tissue and bone changes in frostbite injuries.
Radiology 1972; 102:511.

19. Lisbona R, Rosenthall L. Assessment of bone viability by scinti-
scanning in frostbite injuries. / Trauma 1976; 16:989.

20. Salimi Z, Vas W, Tang-Barton P et ah Assessment of tissue viabil-
ity in frostbite by 99m-Tc pertechnetate scintigraphy. Am }
Roentgenol 1984; 142:415.

21. Ward M. Frostbite. Br Med J 1974; 1:67.

22. Siegel ME, Williams GM, Giargiana FA et ah A useful, objective
criterion for determining the healing potential of an ischemic
ulcer. J Nucl Med 1975; 16:993.

23. Siegel ME, Stewart CA, Kwong P et ah 201-T1 perfusion study of
"ischemic" ulcers of the leg: prognostic ability compared with
Doppler ultrasound. Radiology 1982; 143:233.

24. Alazraki N, Dries D, Lawrence P et ah Assessment of skin ulcer
healing capability by technetium-99m phosphate angiogram
and blood pool images. J Nucl Med 1985; 26:586.

25. Sharlip ID. The role of vascular surgery in arteriogenic and com-
bined arteriogenic and venogenic impotence. Semin Urol 1990;
8:129.

26. Seftel AD, Goldstein I. Vascular testing for impotence. JNucl Med
1992; 33:46.

27. Schwartz AN, Graham MM. Combined technetium radioisotope
penile plethysmography and xenon washout: a technique for
evaluating corpora cavernosal inflow and outflow during early
tumescence. /Nwd Med 1991; 32:404.

28. Miraldi F, Nelson AD, Jones WT et ah A dual-radioisotope tech-
nique for the evaluation of penile blood flow during tumescence.
] Nucl Med 1992; 33:41.

29. Gloviczki P, Calcagno D, Schirger Aet ah Non invasive evaluation
of the swollen extremity: experiences with 190 lymphoscinti-
graphic examinations. / Vase Surg 1989; 9:683.

30. Stewart G, Gaunt JI, Croft DN et ah Isotope lymphography: a new
method of investigating the role of the lymphatics in chronic limb
oedema. Br } Surg 1985; 72:906.

31. McNeill GC, Witte MH, Witte CL et ah Whole-body lymphan-
gioscintigraphy: preferred method for initial assessment of the
peripheral lymphatic system. Radiology 1989; 172:495.

32. Kramer EL, Sanger JJ. Lymphoscintigraphy in 1987: selected
aspects. In: Freeman LM, Weissmann HS, eds. Nuclear Medicine
Annual. New York: Raven Press, 1987:233.

33. Suga K, Uchisako H, Nakanishi T et ah Lymphoscintigraphic
assessment of leg oedema following arterial reconstruction
using a load produced by standing. Nucl Med Commun 1991;
12:907.

34. Vaqueiro M, Gloviczki P, Fisher J et ah Lymphoscintigraphy in
lymphedema: an aid to microsurgery. / Nucl Med 1986; 27:1125.

35. Bandyk DF. Graft infection: a dreadful challenge. Semin Vase Surg
1990; 3:77.

36. Bunt TJ. Synthetic vascular graft infections. I. graft infections.
Surgery 1983; 93:733.

37. McKeown PP, Miller DC, Jamieson SW et ah Diagnosis of arterial
prosthetic graft infection by indium-Ill oxine white blood cell
scans. Circulation 1982; 66 (Suppl. I): 1-130.

38. Lawrence PF, Dries DJ, Alazraki N et ah Indium Ill-labeled
leukocyte scanning for detection of prosthetic vascular graft
infection. / Vase Surg 1985; 2:165.

39. Mark AS, McCarthy SM, Moss AA et ah Detection of abdominal
aortic graft infection: comparison of CT and In-labeled white
blood cell scans. Am J Roentgenol 1985; 144:315.

40. Brunner MC, Mitchell RS, Baldwin JC et ah Prosthetic graft infec-
tion: limitations of indium white blood cell scanning. / Vase Surg
1986; 3:42.

342

CHAPTER31 Radionuclide scanning

41. Williamson MR, Boyd CM, Read RC et al. Ill In-labeled leuko-
cytes in the detection of prosthetic vascular graft infections. Am }
Roentgenoll986;U7:173.

42. Becker W, Duesel W, Berger P et al. The 111 In-granulocyte scan
in prosthetic vascular graft infections: imaging technique and
results. EurJNuclMed 1987; 13:225.

43. Forstrom LA, Dewanjee MK, Chowdhury S et al. Indium-Ill
labeled purified granulocytes in the diagnosis of synthetic vascu-
lar graft infection. Clin NuclMed 1988; 13:859.

44. Berridge DC, Earnshaw JJ, Frier M et al. 111-In-labelled leucocyte
imaging in vascular graft infection. Br } Surg 1989; 76:41.

45. Reilly DT, Grigg MJ, Cunningham DA et al. Vascular graft
infection: the role of indium scanning. Eur } Vase Surg 1989;
3:393.

46. Chung CJ, Hicklin OA, Payan JM et al. Indium-lll-labeled leuko-
cyte scan in detection of synthetic vascular graft infection: the
effect of antibiotic treatment. J NuclMed 1991; 32:13.

47. Sedwitz MM, Davies RJ, Pretorius HT et al. Indium Ill-labeled
white blood cell scans after vascular prosthetic reconstruction.
J Vase Surg 1987; 6:476.

48. Insall RL, Jones NAG, Chamberlain J et al. New isotopic tech-
nique for detecting prosthetic arterial graft infection: 99m
Tc-hexametazime-labelled leucocyte imaging. Br } Surg 1990;
77:1295.

49. Insall RL, Keavey PM, Hawkins T et al. The specificity of
technetium-labelled leukocyte imaging of aortic grafts in the
early postoperative period. Eur J Vase Surg 1991; 5:571.

50. Lantto EH, Lantto TJ, Vorne M. Fast diagnosis of abdominal
infections and inflammations with technetium-99m-HMPAO
labeled leukocytes. } NuclMed 1991; 32:2029.

51. Oyen WJG, Claessens RAMJ, van der Meer JWM et al. Indium-
lll-labeled human nonspecific immunoglobulin G: a new radio-
pharmaceutical for imaging infectious and inflammatory foci.
Clin Infect Dis 1992; 14:1110.

52. Thakur ML. Immunoscintigraphic imaging of inflammatory le-
sions: preliminary findings and future possibilities. Semin Nucl
Med 1990; 20:92.

53. McAfee JG, Gagne G, Subramanian G et al. The localization of
indium-lll-leukocytes, gallium-67-polyclonal IgG and other
radioactive agents in acute focal inflammatory lesions. / Nucl
Med 1991; 32:2126.

54. Dewanjee MK. Cardiac and vascular imaging with labeled
platelets and leukocytes. Semin NuclMed 1984; 20:154.

55. Kesler K, Herring MB, Arnold MP et al. Enhanced strength of en-
dothelial attachment on polyester elastomer and polytetrafluo-
roethylene graft surfaces with fibronectin substrate. / Vase Surg
1986; 3:59.

56. Vallabhajosula S, Goldsmith SJ. 99m-Tc-low density lipoprotein:
intracellularly trapped radiotracer for noninvasive imaging of
low density lipoprotein metabolism in vivo. Semin Nucl Med
1990; 20:68.

57. Baunett HJM. Symptomatic carotid endarterectomy trials. Stroke
1990;21(Suppl.lll):lll.

58. NASCET Investigators. Clinical alert: benefit of carotid
endarterectomy for patients with high-grade stenosis of the
internal carotid artery. Stroke 1991; 22:816.

59. NASCET Steering Committee. North American symptomatic
carotid endarterectomy trial: methods, patient characteristics,
and progress. Stroke 1991; 22:711.

60. Powers WJ. Cerebral hemodynamics in ischemic cerebrovascu-
lar disease. Ann Neurol 1991; 29:231.

61 . Wise RJS, Bernardi S, Frackowiak RSJ et al. Serial observations on
the pathophysiology of acute stroke: the transition from is-
chaemia to infarction as reflected in regional oxygen extraction.
Brain 1983; 106:197.

62. Powers WJ, Grubb RL Jr, Darriet D et al. Cerebral blood flow
and cerebral metabolic rate of oxygen requirements for cerebral
function and viability in humans. / Cereb Blood Flow Metab 1985;
5:600.

63. Powers WJ, Raichle ME. Positron emission tomography and its
application to the study of cerebrovascular disease in man. Stroke
1985; 16:361.

64. Sette G, Baron JC, Mazoyer B et al. Local brain haemodynamics
and oxygen metabolism in cerebrovascular disease. Brain 1989;
112:931.

65. Zivin JA, Choi DW. Stroke therapy. Sci Am 1991; 265:56.

66. Knapp WH, von Kummer R, Kuebler W. Imaging of cerebral
blood flow-to-volume distribution using SPECT. / Nucl Med
1986; 27:465.

67. Buell U, Braun H, Ferbert A et al. Combined SPECT imaging of re-
gional cerebral blood flow (99mTc-hexamethylpropyleneamine
oxime, HMPAO) and blood volume (99mTc-RBC) to assess re-
gional cerebral perfusion reserve in patients with cerebrovascu-
lar disease. NuclMed 1988; 27:51.

68. Toyama H, Takeshita G, Takeuchi A et al. Cerebral hemody-
namics in patients with chronic obstructive carotid disease by
rCBF, rCBV, and rCBV/rCBF ratio using SPECT. J Nucl Med 1990;
31:55.

69. Gibbs JM, Wise RJS, Leenders KL et al. Evaluation of cerebral per-
fusion reserve in patients with carotid-artery occlusion. Lancet
1984; 1:300.

70. Powers WJ, Martin WRW, Herscovitch P et al. Extracranial-
intracranial bypass surgery: hemodynamic and metabolic
effects. Neurology 1984; 34:1168.

71. Powers WJ, Press GA, Grubb RL Jr et al. The effect of hemo-
dynamically significant carotid artery disease on the hemo-
dynamic status of the cerebral circulation. Ann Intern Med 1987;
106:27.

72. Powers WJ, Tempel LW, Grubb RL Jr. Influence of cerebral hemo-
dynamics on stroke risk: one-year follow-up of 30 medically
treated patients. Ann Neurol 1989; 25:325.

73. Powers WJ, Grubb RL Jr, Raichle ME. Clinical results of
extracranial-intracranial bypass surgery in patients with hemo-
dynamic cerebrovascular disease. } Neurosurg 1989; 70:61.

74. Harper AM, Glass HI. Effect of alterations in the arterial carbon
dioxide tension on the blood flow through the cerebral cortex at
nomal and low arterial blood pressures. } Neurol Neurosurg Psy-
chiatry 65; 28:449.

75. Chadhuri TK, Fink S, Weinberg S et al. Pathophysiologic consid-
erations in carotid artery imaging: current status and physiologic
background. Am J Physiolog Imag 1992; 7:77.

76. Ramsay SC, Yeates MG, Lord RS et al. Use of technetium-
HMPAO to demonstrate changes in cerebral blood flow reserve
following carotid endarterectomy. J Nucl Med 1991; 32:1382.

77. Ciskrit DF, Burt RW, Dalsing MC et al. Acetazolamide enhanced
single photon emission computed tomography (SPECT) evalua-
tion of cerebral perfusion before and after carotid end-
arterectomy. / Vase Surg 1992; 15:747.

343

pa rt 1 1 Noninvasive vascular diagnostics

78. Heiss W-D, Zeiler K, Havelec L et ah Long-term prognosis in
stroke related to cerebral blood flow. Arch Neurol 1977; 34:671.

79. Heiss W-D. Flow thresholds of functional and morphological
damage of brain tissue. Stroke 1983; 14:329.

80. Yamauchi H, Fukuyama H, Kimura J et ah Hemodynamics in
internal carotid artery occlusion examined by positron emission
tomography. Stroke 1990; 21:1400.

81 . Kushner M, Reivich M, Fieschi Cetah Metabolic and clinical cor-
relates of acute ischemic infarction. Neurology 1987; 32:1103.

82. Defer G, Moretti J-L, Cesaro P et ah Early and delayed SPECT
using N-isopropyl P-iodoamphetamine iodine 123 in cerebral is-
chemia: a prognostic index for clinical recovery. Arch Neurol 1987;
44:715.

83. Costa DC, Ell PJ. 99mTc-HMPAO washout in prognosis of stroke.
Lancet 1989)1:213.

84. Mountz JM, Modell JG, Foster NL et ah Prognostication of
recovery following stroke using the comparison of CT and
technetium-99m HM-PAO SPECT. / Nucl Med 1990; 31:61.

85. Tikofsky RS, Hellman RS. Brain single photon emission com-
puted tomography: newer activation and intervention studies.
Semin Nucl Med 1991; 21:40.

86. Lequin MH, Blok D, Pauwels BKJ. Radiopharmaceuticals for
functional brain imaging with SPECT. In: Freeman LM, ed.
Nuclear Medicine Annual. New York: Raven Press, 1991, 37.

87. Bogousslavsky J, Delaloye-Bischof A, Regli F et ah Prolonged
hypoperfusion and early stroke after transient ischemic attack.
Stroke 1990; 21:40.

88. Maurer AH, Siegel JA, Comerota AJ et ah SPECT quantification of
cerebral ischemia before and after carotid endarterectomy J Nucl
Med 1990; 31:1412.

89. Hayashida K, Nishimura T, Imakita S et ah Filling out phenome-
non with technetium-99m HM-PAO brain SPECT at the site of
mild cerebral ischemia. / Nucl Med 1989; 30:591 .

90. Davis S, Andrews J, Lichtenstein M et ah A single-photon emis-
sion computed tomography study of hypoperfusion after sub-
arachnoid hemorrhage. Stroke 1990; 21:252.

91. Soucy JP, McNamara D, Mohr G et ah Evaluation of vasospasm
secondary to subarachnoid hemorrhage with technetium-99m-
hexamethyl-propyleneamine oxime (HM-PAO) tomoscintigra-
phy. J Nucl Med 1990; 31:972.

92. Matsuda H, Higashi S, Neshandar I et ah Evaluation of cere-
bral collateral circulation by technetium-99m HM-PAO brain
SPECT during Matas test: report of three cases. / Nucl Med 1988;
29:1724.

93. Tegtmeyer CJ, Kellum CD, Ayers C. Percutaneous transluminal
angioplasty of the renal artery: results and long-term follow-up.
Radiology 1984; 153:77.

94. Dean RH, Krueger TC, Whiteneck JM et ah Operative manage-
ment of renovascular hypertension: results after a follow-up of
fifteen to twenty-three years. / Vase Surg 1984; 1:234.

95. Muller FB, Sealey JE, Case DB et ah The captopril test for identify-
ing renovascular disease in hypertensive patients. Am } Med
1986; 80:633.

96. Erbsloeh-Moeller B, Dumas A, Roth D et ah Furosemide-1311-
hippuran renography after angiotensin-converting enzyme
inhibition for the diagnosis of renovascular hypertension. Am }
Med 1991; 90:23.

97. Simon N, Franklin SS, Bleifer KH et ah Clinical characteristics of
renovascular hypertension. JAMA 1972; 220:1209.

98. Chen CC, Hoffer PB, Vahjen G et ah Patients at high risk for renal
artery stenosis: a simple method of renal scintigraphic analysis
with Tc-99m DTPA and captopril. Radiology 1990; 176:365.

99. Mann SL, Pickering TG, Sos TA et ah Captopril renography in the
diagnosis of renal artery stenosis: accuracy and limitations. Am }
Med 1991; 90:30.

100. Setaro JF, Saddier MC, Chen CC et ah Simplified captopril renog-
raphy in diagnosis and treatment of renal artery stenosis. Hyper-
tension 1991; 18:289.

101. Davidson R, Wilcox C. Diagnostic usefulness of renal scanning
after angiotensin converting enzyme inhibitors. Hypertension
1991; 18:299.

102. Mueller CB, Surtshin A, Carlin MR et ah Glomerular and tubular
influences on sodium and water excretion. Am J Physiol 1951;
165:411.

103. Ploth DW. Angiotensin-dependent renal mechanisms in two-
kidney, one-clip renal vascular hypertension. Am } Physiol 1983;
245:131.

104. de Zeeuw D, Jonker GJ, Hovinga TKK et ah The mechanism and
diagnostic value of angiotensin I converting enzyme inhibition
renography. Am J Hypertens 1991; 4:741S.

105. Dondi M. Captopril renal scintigraphy with 99mTc-mercap-
toacetyltriglycine (99mTc-MAG 3) for detecting renal artery
stenosis. Am ] Hypertens 1991; 4:737S.

106. Sfakianakis G, Bourgoignie JJ. Renographic diagnosis of
renovascular hypertension with angiotensin converting enzyme
inhibition and furosemide. Am J Hypertens 1991;4:706S.

107. Dubovsky EV, Russell CD. Diagnosis of renovascular hyperten-
sion after renal transplantation. Am J Hypertens 1991; 4:724S.

108. Nally JV Jr, Chen C, Fine E et ah Diagnostic criteria of renovascu-
lar hypertension with captopril renography: a consensus state-
ment. Am J Hypertens 1991;4:749S.

109. Hricik DE, Browning PJ, Kopelman R et ah Captopril-induced
functional renal insufficiency in patients with bilateral renal-
artery stenoses or renal-artery stenosis in a solitary kidney. N
Engl J Med 1983; 308:373.

110. Jackson B, McGrath BP, Matthews PG et ah Differential renal
function during angiotensin converting enzyme inhibition in
renovascular hypertension. Hypertension 1986; 8:650.

111 . Hansen PB, Gardal P, Fruergaard P. The captopril test for identifi-
cation of renovascular hypertension: value and immediate ad-
verse effects. J Intern Med 1990; 228:159.

112. Holley KE, Hunt JC, Brown AL Jr et ah Renal artery stenosis: a
clinical-pathologic study in normotensive and hypertensive
patients. Am J Med 1964; 37:14.

113. Eyler WR, Clark MD, Garman JE et ah Angiography of the
renal areas including a comparative study of renal arterial
stenoses in patients with and without hypertension. Radiology
1962; 78:879.

114. Dondi M, Fanti S, DeFabritiis Aet ah Prognostic value of captopril
renal scintigraphy in renovascular hypertension. / Nucl Med
1992; 33:2040.

115. Vidt DG. The diagnosis of renovascular hypertension: a clini-
cian's viewpoint. Am J Hypertens 1991; 4:633S.

116. Moser KM. Venous thromboembolism. Am Rev Respir Dis 1990;
141:235.

117. Biello DR, Mattar AG, McKnight RC et ah Ventilation-perfusion
studies in suspected pulmonary embolism. Am J Roentgenol 1979;
133:1033.

344

CHAPTER31 Radionuclide scanning

118. Cheely R, McCartney WH, Perry JR et ah The role of noninvasive
tests versus pulmonary angiography in the diagnosis of pul-
monary embolism. Am J Med 1981; 70:17.

119. The PIOPED Investigators. Value of the ventilation/perfusion
scan in acute pulmonary embolism: results of the prospective in-
vestigation of pulmonary embolism diagnosis (PIOPED). JAMA
1990; 263:2753.

120. Sostman HD, Ravin CE, Sullivan DC et ah Use of pulmonary an-
giography for suspected pulmonary embolism: influence of
scintigraphic diagnosis. Am J Roentgenol 1982; 139:673.

121. Hull RD, Hirsch J, Carter CJ et ah Diagnostic value of
ventilation-perfusion lung scanning in patients with suspected
pulmonary embolism. Chest 1985; 88:819.

122. Hull RD, Raskob GE. Low probability lung scan findings: a need
for change. Ann Intern Med 1991; 114:142.

123. Sullivan DC, Coleman RE, Mills SR et ah Lung scan interpreta-
tion: effect of different observers and different criteria. Radiology
1983; 149:803.

124. Spies WG, Burstein SP, Dillehay GL et ah Ventilation-perfusion
scintigraphy in suspected pulmonary embolism: correlation
with pulmonary angiography and refinement of criteria for
interpretation. Radiology 1986; 159:383.

125. Webber MM, Gomes AS, Roe D et ah Comparison of Biello,
McNeil, and PIOPED criteria for the diagnosis of pulmonary
emboli of lung scans. Am J Roentgenol 1990; 154:975.

126. Kelley MA, Carson JL, Palevsky HI et ah Diagnosing pulmonary
embolism: new facts and strategies. Ann Intern Med 1991;
114:300.

127. Lee ME, Biello DR, Kumar B et ah "Low probability" ventilation-
perfusion scintigrams: clinical outcomes in 99 patients. Radiology
1985; 156:497.

128. Smith R, Maher JM, Miller RI et ah Clinical outcomes of patients
with suspected pulmonary embolism and low-probability
aerosol-perfusion scintigrams. Radiology 1987; 164:731.

129. Kahn D, Bushnell DL, Dean R et ah Clinical outcome of
patients with a "low probability" of pulmonary embolism
on ventilation-perfusion lung scan. Arch Intern Med 1989;
149:377.

130. Hull RD, Raskob GE, Coates G et ah A new noninvasive manage-
ment strategy for patients with suspected pulmonary embolism.
Arch Intern Med 1989; 149:2549.

131. Wilson JE. Diagnostic methods for deep venous thrombosis. Arch
Intern Med 1980; 140:893.

132. Hull RD, Hirsch J, Carter CJ et ah Diagnostic efficacy of imped-
ance plethysmography for clinically suspected deep-vein throm-
bosis. Ann Intern Med 1985; 102:21.

133. Huisman MV, Bueller HR, ten Cate JW et ah Serial impedance
plethysmography for suspected deep venous thrombosis in
out-patients: the Amsterdam general practitioner study. N Engl J
Med 1986; 314:823.

134. Severingham JW, Swenson EW, Finley JN et ah Unilateral hy-
poventilation produced in dogs by occluding one pulmonary
artery. J Appl Physiol 1961; 16:53.

135. Sandler MS, Velchik MG, Alavi A. Ventilation abnormalities
associated with pulmonary embolism. Clin Nucl Med 1988;
13:450.

136. Bushnell DL, Sood KB, Shirazi P et ah Evaluation of pulmonary
perfusion in lung regions showing isolated xenon-133 ventila-
tion washout defects. Clin Nucl Med 1990; 15:562.

137. James WS III, Menn SJ, Moser KM. Rapid resolution of a pul-
monary embolus in man. West J Med 1978; 128:60.

138. Tow DE, Wagner HN Jr. Recovery of pulmonary arterial blood
flow in patients with pulmonary embolism. N Engl } Med 1967;
276:1053.

139. Murphy ML, Bulloch RT. Factors influencing the restoration of
blood flow following pulmonary embolization as determined by
angiography and scanning. Circulation 1968; 38:1116.

140. Urokinase Pulmonary Embolism Trial Study Group. Phase 1
results. JAMA 1970; 214:2163.

141. Stein PD, Coleman RE, Gottschalk A et ah Diagnostic utility of
ventilation/perfusion lung scans in acute pulmonary embolism
is not diminished by pre-existing cardiac or pulmonary disease.
Chest 1991; 100:604.

142. Monreal M, Barroso CR-J, Manzano JR et ah Asymptomatic pul-
monary embolism in patients with deep vein thrombosis: is it
useful to take a lung scan to rule out this condition? / Cardiovasc
Surg 1989; 30:104.

143. Palevsky HI, Cone L. A case of "false-positive" high probability
ventilation-perfusion lung scan due to tuberculous mediastinal
adenopathy with a discussion of other causes of "false-positive"
high probability ventilation-perfusion lung scans. / Nucl Med
1991; 32:512.

144. Velchik MG, Tobin M, McCarthy K. Nonthromboembolic causes
of high-probability lung scans. Am J Physiol Imag 1989; 4:32.

145. Chung CJ, Grossnickle M, Rosenthal P et ah Postatelectatic
ventilation-perfusion mismatch simulating a pulmonary
embolus. J Nucl Med 1990; 31:1397.

146. Moser KM, LeMoine JR. Is embolic risk conditioned by location
of deep venous thrombosis? Ann Intern Med 1981; 91:439.

147. Kakkar VV, Howe CT, Flanc C et ah Natural history of post-
operative deep-vein thrombosis. Lancet 1969; 1:230.

148. Sevitt S, Gallagher N. Venous thrombosis and pulmonary
embolism: a clinico-pathological study in injured and burned
patients. Br J Surg 1961; 48:475.

149. Giachino A. Relationship between deep-vein thrombosis in the
calf and fatal pulmonary embolism. Can J Surg 1988; 31 :129.

150. Koehn H, Koenig B, Mostbeck A. Incidence and clinical feature of
pulmonary embolism in patients with deep vein thrombosis: a
prospective study. Eur J Nucl Med 1987; 13:S11.

151. Hull R, Hirsh J, Sackett DL et ah Combined use of leg scanning
and impedance plethysmography in suspected venous throm-
bosis: an alternative to venography. N Engl J Med 1977; 296:
1497.

152. Paiement G, Wessinger SJ, Waltman AC et ah Surveillance of
deep vein thrombosis in asymptomatic total hip replacement
patients: impedance phlebography and fibrinogen scanning
versus roentgenographic phlebography. Am J Surg 1988;
155:400.

153. Knight LC. Radiopharmaceuticals for thrombus detection. Semin
Nucl Med 1990; 20:52.

154. Schaible TF, Alavi A. Antifibrin scintigraphy in the diagnostic
evaluation of acute deep venous thrombosis. Semin Nucl Med
1991;21:313.

155. Henkin RE, Yao JST, Quinn JL III et ah Radionuclide venography
(RNV) in lower extremity venous disease. / Nucl Med 1974;
15:171.

156. Webber MM, Pollak EW, Victery W et ah Thrombosis detection
by radionuclide particle (MAA) entrapment: correlation

345

pa rt 1 1 Noninvasive vascular diagnostics

with fibrinogen uptake and venography. Radiology 1974; 111:
645.

157. Vlahos L, MacDonald AF, Causer DA. Combination of isotope
venography and lung scanning. Br J Radiol 1976; 49:840.

158. Ryo UY, Colombetti LG, Polin SG et ah Radionuclide venogra-
phy: significance of delayed washout; visualization of the saphe-
nous system. JNuclMed 1976; 17:590.

159. Van Kirk OC, Burry MT, Jansen AA et ah A simplified approach to
radionuclide venography: concise communication. / Nucl Med
1976; 17:969.

160. Beswick W, Chmiel R, Booth R et ah Detection of deep venous
thrombosis by scanning of 99m technetium-labelled red-cell
venous pool. Br Med] 1979; 1:82.

161. Kempi V, van der Linden W. Diagnosis of deep vein thrombosis
with in vivo 99mTc-labeled red blood cells. Eur JNuclMed 1981;
6:5.

162. Fogh J, Nielsen SL, Vitting K et ah The diagnostic value of
angioscintigraphy with 99mTc-labelled red blood cells for
detection of deep vein thrombosis. Nucl Med Commun 1982;
3:172.

163. Lisbona R, Stern J, Derbekyan V. 99mTc red blood cell venogra-
phy in deep vein thrombosis of the leg: a correlation with contrast
venography. Radiology 1982; 143:771.

164. Singer I, Royal HD, Uren RF et ah Radionuclide plethysmogra-
phy and Tc-99m red blood cell venography in venous thrombo-
sis: comparison with contrast venography. Radiology 1984;
150:213.

165. Littlejohn GO, Brand CA, Ada A et ah Popliteal cysts and deep
venous thrombosis: Tc-99m red blood cell venography. Radiology
1985; 155:237.

166. Snarski AM. Radionuclide venography: two-stage flow and
equilibrium technique using 99mTc pertechnetate, and 99mTc
RBC labelled "in vivo." Eur J Nucl Med 1989; 15:137.

167. Caner B, Ozmen M, Dincer A et ah Detection of deep vein throm-
bosis: combined flow and blood pool radionuclide venography
vs contrast venography. Angiology 1991; 42:796.

168. Uphold RE, Knopp R, dos Santos PAL. Radionuclide venogra-
phy as an outpatient screening test for deep venous thrombosis.
Ann EmergMed 1980; 9:613.

169. Williams VL, Higgins WL, Epstein DH. The negative radionu-
clide venogram: an indication for conservative therapy? Clin
NucZMed 1988; 13:26.

170. Kakkar VV. Fibrinogen uptake test for detection of deep vein
thrombosis: a review of current practice. Semin Nucl Med 1977;
7:229.

171. Clarke-Pearson DL, Coleman RE, Siegel R et ah Indium 111
platelet imaging for the detection of deep venous thrombosis
and pulmonary embolism in patients without symptoms after
surgery. Surgery 1985; 98:98.

172. Ezekowitz MD, Pope CF, Sostman HD et ah Indium-Ill platelet
scintigraphy for the diagnosis of acute venous thrombosis. Circu-
lation 1986; 73:668.

173. Som P, Oster ZH, Zamora PO et ah Radioimmunoimaging of
experimental thrombi in dogs using technetium-99m-labeled
monoclonal antibody fragments reactive with human platelets. /
Nucl Med 1986; 27:1315.

174. Knight LC, Primeau JL, Siegel BA et ah Comparison of In-111-
labeled platelets and iodinated fibrinogen for the detection of
deep vein thrombosis. / Nucl Med 1978; 19:891 .

175. Chadhuri TK, Fink S, Farpour A. Physiological considerations in
imaging of lower extremity venous thrombosis. Am} Physiol Imag
1991; 6:90.

176. Knight LC. Do we finally have a radiopharmaceutical for rapid,
specific imaging of venous thrombosis? JNuclMed 1991; 32:791.

177. Kotler TS, Diamond GA. Exercise 201-thallium scintigraphy in
the diagnosis and prognosis of coronary artery disease. Ann
Intern Med 1990; 113:684.

178. Ladenheim ML, Pollock BH, Rozanski Aet ah Extent and severity
of myocardial hypoperfusion as predictors of prognosis in pa-
tients with suspected coronary artery disease. / Am Coll Cardiol
1986; 7:464.

179. Brown KA. Prognostic value of 201-thallium myocardial perfu-
sion imaging: a diagnostic tool comes of age. Circulation 1991;
83:363.

180. Younis LT, Byers S, Shaw L et ah Prognostic importance of silent
myocardial ischemia detected by intravenous dipyridamole
thallium myocardial imaging in asymptomatic patients with
coronary artery disease. } Am Coll Cardiol 1989; 14:1635.

181 . Hertzer NR, Beven EG, Young JR et ah Coronary artery disease in
peripheral vascular patients: a classification of 1000 coronary
angiograms and results of surgical management. Ann Surg 1984;
199:223.

182. Boucher CA, Brewster DC, Darling RC et ah Determination of
cardiac risk by dipyridamole-thallium imaging before peripher-
al vascular surgery. N Engl J Med 1985; 312:389.

183. Eagle KA, Singer DE, Brewster DC et ah Dipyridamole-thallium
scanning in patients undergoing vascular surgery: optimizing
preoperative evaluation of cardiac risk. JAMA 1987; 257:2185.

184. Leppo J, Plaja J, Gionet M et ah Noninvasive evaluation of cardiac
risk before elective vascular surgery. / Am Coll Cardiol 1987;
9:269.

185. Sachs RN, Tellier P, Larmignat P et ah Assessment by
dipyridamole-thallium-201 myocardial scintigraphy of coro-
nary risk before peripheral vascular surgery. Surgery 1988;
103:584.

186. Eagle KA, Coley CM, Newell JB et ah Combining clinical and
thallium data optimizes preoperative assessment of cardiac
risk before major vascular surgery. Ann Intern Med 1989; 110:859.

187. Lette J, Waters D, Lapointe J et ah Usefulness of the severity
and extent of reversible perfusion defects during thallium-
dipyridamole imaging for cardiac risk assessment before
noncardiac surgery. Am J Cardiol 1989; 64:276.

188. Levinson JR, Boucher CA, Coley CM et ah Usefulness of semi-
quantitative analysis of dipyridamole-thallium-201 redistribu-
tion for improving risk stratification before vascular surgery. Am
J Cardiol 1990; 66:406.

189. Cambria RP, Brewster DC, Abbott WM et ah The impact of selec-
tive use of dipyridamole-thallium scans and surgical factors on
the current morbidity of aortic surgery. / Vase Surg 1992; 15:43.

190. Taylor LM, Yeager RA, Moneta GL et ah The incidence of periop-
erative myocardial infarction in general vascular surgery. / Vase
Surg 1991; 15:52.

191. Lette J, Waters D, Lassonde J et ah Postoperative myocardial
infarction and cardiac death: predictive value of
dipyridamole-thallium imaging and five clinical scoring sys-
tems based on multifactorial analysis. Ann Surg 1990; 211:84.

192. Yeager RA. Basic data related to cardiac testing and cardiac risk
associated with vascular surgery. Ann Vase Surg 1990; 4:193.

346

CHAPTER31 Radionuclide scanning

193. Hollier LH. Cardiac evaluation in patients with vascular reconstructive operation over a twenty-five-year period. Surgery
disease — overview: a practical approach. / Vase Surg 1992; 1981; 90:1055.

15:726. 195. Burnham SJ, Johnson G Jr, Gurri JA. Mortality risks for sur-

194. Crawford ES, Bomberger RA, Glaeser DH et ah Aortoiliac occlu- vivors of vascular reconstructive procedures. Surgery 1982;
sive disease: factors influencing survival and function following 92:1072.

347

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Computed tomography

Anton Mlikotic
Irwin Walot

Remarkable advances in computed tomography (CT) and
computer technology over the past decade have resulted in
exciting new applications that benefit both the physician and
patient. During the 1980s, single-detector CT scanners
emerged as an important modality in the evaluation of
vascular disease. It was now possible to study vessels with
cross-sectional images and semiautomated reformatted re-
constructions in various planes without relying exclusively on
invasive angiography. The further development and wide-
spread availability of spiral (helical) CT scanners have al-
lowed greater areas of coverage in shorter periods of time
while significantly reducing patient radiation exposure. With
the recent addition of multiple row detectors, larger datasets
with greater resolution could be acquired, leading to im-
proved image quality and decreased patient contrast load re-
quirements. At the same time, the development of powerfully
fast computers equipped with sophisticated software has fa-
cilitated data processing and paved the way for advanced
three-dimensional (3-D) image reconstructions. A simple axial
CT acquisition followed by 3-D reconstruction of the dataset
can now obviate the need for additional invasive catheter-
based angiography or other noninvasive imaging studies in
selected cases. In many instances, CT angiography (CTA) has
replaced catheter-based imaging in the investigation of vascu-
lar disease processes and is quickly becoming a standard of
practice.

Basic principles

With the introduction of single-slice CT imaging, it became
possible to generate transaxial images of reasonable quality
to study certain parts of the vasculature. It was suitable for
examining large structures oriented along the long axis of the
body, such as the aorta and vena cavae, as well as the brachio-
cephalic arteries and iliofemoral vessels. CT served a comple-
mentary role to catheter-based angiography— it could easily
detect the presence of hematomas, contrast extravasation, cal-
cifications, masses, fluid collections, inflammatory tissue or is-

chemic changes in end-organ structures. Moreover, it can
readily identify the synthetic fabric material of endoluminal
stent grafts. 1 This contributed important diagnostic informa-
tion for the assessment of aortic aneurysms, for example, and
steered CT imaging toward replacing traditional catheter-
based angiography in preparation for, and postprocedural
evaluation of, endovascular stent grafts. Compared with mod-
ern scanners, however, these nonspiral CT studies required
relatively longer scanning intervals (seconds as opposed to
subsecond acquisitions), allowing the introduction of artifacts
that result in suboptimal image quality. Maximal arterial con-
trast opacification is essential for diagnosing and excluding
certain conditions, requiring data capture during the narrow
temporal peak window of vascular opacification of the first
pass of intravascular contrast. 2 ' 3 This ability was limited by the
inherent slower acquisition times of older generation scan-
ners. Patient factors further contributed to reduced image
quality. An abnormal cardiac output or renal function can alter
the proper timing of arterial phase acquisitions, and mediasti-
nal motion and a limited patient breathhold capacity may
also introduce significant artifacts that may lead to diagnostic
inaccuracy. 4

The advent of spiral CT imaging established a more defini-
tive role for CTA in studying vascular disease. 5 The introduc-
tion of slip ring technology and improved CT tube cooling
methods allows for rapid acquisitions that capture a larger
patient volume during peak contrast enhancement and mini-
mize artifacts related to motion. 6 ' 7 Multiphasic imaging with
selective visualization of the arterial and venous phases of the
study has become easier to perform with test injection tech-
niques or utilization of contrast bolus tracking software. 8-10
Spiral scanning offers superior spatial resolution in the
transaxial plane (at 1 mm along the x and y axes), allowing con-
fident image assessment when viewing transaxial images. 11
This became important for identifying abnormalities such as
arterial dissections, for example, in which intimal flaps could
be more accurately determined. A relative disadvantage of
spiral CT, however, is poor resolution along the longitudinal
or z axis. This accounts for partial volume averaging effects

348

chapter 32 Computed tomography

and misregistration artifacts when 3-D or multiplanar recon-
structions are performed along the length of the patient. 12
Although optimization of scanning parameters (including
overlapping of images), narrow beam collimation, and vari-
ous postprocessing filtering and iterative deconvolution tech-
niques were used in an attempt to adjust for decreased
longitudinal resolution, image reconstruction remained less
than ideal. 13 ' 14

The emergence of multidetector row scanners has even fur-
ther decreased scanning times, yet significantly increased the
amount of imaging data per scan. Four or eight detector scan-
ners improve the quality of data by acquiring thinner (0.5 mm)
scan slices, improving contrast enhancement, and broadening
the range of patient coverage with greater beam widths. More
recently, highly advanced scanners with 16 or more detectors
further enrich the dataset by eliciting diagnostic information
in isotropic voxels that optimizes z axis resolution and appre-
ciably improves the quality of 3-D reconstructions. Of great
importance, diagnostic results for peripheral studies can be
achieved using a minimal contrast load. Therefore, the subset
of patients with renal disease and high serum creatinine levels
such as diabetic patients previously restricted from CTA
owing to exacerbation of renal function from contrast-
induced nephrotoxicity may now be safely included. 15

Multislice imaging presents a unique challenge to the physi-
cian evaluating the study. On average, single-slice scanners
would produce about 35 images per study. In contrast, multi-
detector scanners may generate up to 1000 images per exami-
nation, which is cumbersome to evaluate and may overload
one's sensory circuits! This has resulted in a shift in image
assessment practice from routine evaluation of sequential
two-dimensional (2-D) transaxial images to interactive,
subselective appraisals of 3-D volumetric renditions of the
imaging data and has created an increasingly important role
for the diagnostic computer workstation. Volumetric recon-
structions of the imaging data can be viewed in multiple
planes and the axial 2-D source images can be referenced
alongside the reconstructed image. This advanced data view-
ing capability in effect has encouraged a more interactive
environment for the surgeon and radiologist for patient
radiographic assessment and preprocedural planning. 16

Once the imaging data are acquired, various postprocessing
techniques can be applied to create an image display based
upon the specific information that is desired to evaluate the
patient. Although with single-slice or four-detector systems
most pertinent diagnostic information may be gleaned from
the axial source images alone, 3-D reconstruction algorithms
allow the added advantage of the creation of vascular images
that mimic traditional catheter-based angiograms that are
most familiar to the surgeon. Multiplanar reformation (MPR)
of axial images into the orthogonal (sagittal and coronal) and
oblique or curved planar reformatted views were the first to be
employed as important adjuncts to visualizing 2-D images.
The technique provides the observer with a conceptual 3-D

picture of the vasculature by allowing scrolling through a
series of 2-D images in various planes on a computer worksta-
tion. Both vascular and nonvascular structures are included,
which may be helpful in some instances but may interfere
with evaluation in others. Curved planar reformatting (CPR)
remains an important tool in the assessment of vascular
stenoses. 17-20

Shaded surface display (SSD) is a technique that relies on the
preselection of a threshold attenuation or "brightness" level.
All voxels in the dataset with Hounsfield values above the
threshold level are used to display a single structure in a 3-D
format. Depth cues are provided by shading from a computer-
generated light source and the applied algorithm displays the
surface features of blood vessels. Although the image detail is
appealing to most clinicians, this type of display may require
long time periods for editing at the computer workstation to
ensure that nonvascular structures are excluded. Moreover,
with this technique, the presence of mural calcifications tends
to underestimate the degree of vascular stenosis and the inter-
nal architecture of the vasculature cannot be evaluated. This
type of image display is valuable in demonstrating the rela-
tionship of the aorta to branching vessels, although review of
the axial images alongside the reconstructed image is neces-
sary. Although initially considered useful, this technique has
been largely replaced by more advanced projection and ren-
dering methods. 21

Maximum intensity projection (MIP) imaging differs from
SSD in that the resultant image is derived from maximum
Hounsfield value voxels. It is a type of 3-D rendering technique
that evaluates each voxel along a line from the viewer's eye
through a volume of data and then displays the brightest pixel.
In contrast to SSD, depth perception is lacking and therefore it
may be difficult to separate overlapping vessels, such as the
pulmonary arteries and thoracic aorta. Small intravascular le-
sions, such as intimal flaps or thrombosis, may be missed
entirely, and high-density material, such as calcification, may
obscure information obtained from intravascular contrast
material. This technique, however, provides a more global per-
spective of vascular disease processes by best demonstrating
collateralization and is most reliable for determining vascular
stenosis and evaluating smaller vessels. However, as opposed
to more sophisticated postprocessing applications such as
volume rendering, this projection technique tends to generate
vessels that appear "flattened" and may distort the spatial rela-
tionship of the vasculature to surrounding structures. 21

Volume rendering is the newest reconstruction technique
that takes advantage of current sophisticated computer tech-
nology. Specialized computer graphics hardware coupled
with commercially available parallel processing computers
provides "real time" reconstructions that have greatly facili-
tated interactivity with the dataset, such as editing with clip
planes. This capability is the result of the increased speed of
data management in terms of acquisition, resampling, editing,
and flow. Unlike SSD or MIP reconstructions that utilize

349

pa rt 1 1 Noninvasive vascular diagnostics

approximately 10% of the usable dataset, volume rendering
takes the entire volume of data, sums the contribution from
each voxel along a line from the viewer's eye, and displays the
array of the resulting composite voxels. The resultant 3-D
image resembles an SSD image but contains much more infor-
mation. The image can be manipulated with many degrees of
freedom to analyze a lesion from multiple perspectives. This is
well suited for the study of vascular geometry and can provide
identical diagnostic information compared with axial images.
Arguably, one of the greatest advantages of spiral CT with 3-D
volume rendering is that all the necessary information is pro-
vided in a single radiologic study, where two or sometimes
three independent studies were required in the past. Although
this technology is extremely attractive it is quite expensive.
Continued advances in computer image processing will con-
tinue to increase the speed although the effects on cost are
unclear. 22-24

Compared with other vascular imaging modalities, CTA
offers several advantages. Requiring only venous access
for a high-rate delivery of contrast, it is a minimally invasive
technique where the primary concern for patient safety is re-
lated to adverse effects due to contrast allergy or systemic re-
actions, extravasation of contrast into the soft tissues, and
potential nephrotoxicity. Unlike catheter-based angiography,
the risks of potentially serious complications of invasive
angiography that are induced by the catheter or related to the
vascular access site are eliminated. Moreover, postacquisition
3-D reconstruction capabilities allow lesions such as
aneurysms to be studied in many planes and axes following a
single dataset acquisition that would otherwise require addi-
tional traditional angiographic runs with increased loads of
intravascular contrast. CTA also permits visualization of
structures surrounding the vasculature not appreciated with
more invasive techniques. In combination with 3-D volume
rendering, spiral CT provides information at a lower cost than
conventional angiography. Unlike magnetic resonance imag-
ing (MRI), ferromagnetic surgical clips and pacemakers do not
preclude performing a CTA study. Also, both MRI and ultra-
sound may overestimate the degree of vascular stenosis due to
artifacts generated by turbulent flow. Unless contrast opacifi-
cation is suboptimal, CTA may reliably determine the extent of
compromise of vascular flow. The caveat of CTA, however, lies
not only in the burden to the patient in terms of relative
contrast load and radiation exposure but also in terms of the
time that is necessary to perform editing functions in image
reconstructions.

Clinical applications

The risks and expense associated with conventional angio-
graphy have contributed to a shift in practice towards non-
invasive imaging modalities. The accuracy of CTA in many
clinical applications makes it well suited for the study of

various vascular disease processes, both acute and chronic.
In many instances, CTA now has the important function of
preprocedural planning in preparation for a surgical or endo-
vascular intervention.

Perhaps the best example of the utility of CTA in replacing
invasive angiography is portrayed in the evaluation of pul-
monary thromboembolism. CTA can now detect pulmonary
emboli to the fourth order (segmental) vessels. According to
Remy-Jardin and coworkers, 25 a study of 360 patients sus-
pected of having pulmonary embolism using thin collima-
tion CT imaging revealed a false-negative rate of only 5% and
concluded that CTA was a technically acceptable examination
in nearly 97% of patients in the study group. The reported sen-
sitivity and specificity for detection of central emboli ranges
from 53% to 100% and from 78% to 100%, respectively. The
presence of higher order emboli with smaller subsegmental
vessels, however, appears to have little clinical significance.
At many institutions, pulmonary CTA has already become
the diagnostic standard, whereas at others its position in the
hierarchy of investigations has yet to be determined. 25-29

CTA is well suited for the evaluation of trauma. A screening
CT examination of the chest, abdomen, and pelvis is routine in
trauma centers today and may uncover clues suggesting acute
injury. With modern spiral scanners, most motion artifacts
are minimized, allowing for accurate determination of the
presence of a mediastinal (periaortic) hematoma or contrast
extravasation that should trigger a more focused examination
of the thoracic aorta (Fig. 32.1). CTA can identify aortic transac-
tion, and 3-D reconstruction can assist in preoperative plan-
ning (Fig. 32.2). 30-33

Although coarctation of the thoracic aorta may be
better evaluated by MRI than by cross-sectional CT alone,
a multiplanar, curved planar, or volumetric reconstruc-
tion provides a longitudinal view of the aorta allowing
better appreciation of the point of stenosis and poststenotic
dilation. Collateralization secondary to obstruction can be
demonstrated on volume-rendered reconstructions of the
chest (Fig. 32.3). Midaortic coarctations are suggested by a
sudden reduction in caliber of the midabdominal segment of
the aorta. 34

CT is an excellent diagnostic modality for evaluating tho-
racic aortic pathology. Dissection planes can be easily identi-
fied on dynamic contrast-enhanced cross-sectional images
obtained from helical CT scanners (Fig. 32.4). CT provides
additional information about the presence and size of
hematomas and hemothorax. Rupture into the pericardium
from an ascending aortic aneurysm is easily detected on axial
images. Multiplanar reconstructions will often identify the
dissection entry and reentry sites (Fig. 32.5A-C) as well as the
relationship of intimal flaps to the origins of the great vessels,
visceral vessels, and renal arteries (Fig. 32.5D,E). If endovascu-
lar repair is contemplated, a computer workstation analysis
can identify device landing zones and provide the necessary
diameters and lengths for device selection. Interrogation of

350

chapter 32 Computed tomography

Figure 32.1 Computed tomography angiography (CTA) of the thoracic
aorta. This study was acquired in a patient following a high-speed motor
vehicle collision. (A)There is a small periaortic hematoma (short arrow) and
minimal contrast extravasation into the aortic wall (long arrow). (B) Even a

low-resolution 3-D volumetric reconstruction of data acquired from a single-
detector CT scanner can suggest an abnormality (arrow). (C) The subsequent
catheter-based thoracic aortogram shows an increase in the size of the
pseudoaneurysm (arrow).

351

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.2 Thoracic computed tomography angiography (CTA) study
performed after a patientwasstruckbyacar. (A) At the level of the
pulmonary arteries there is a large periaortic hematoma and
pseudoaneurysm (arrow). (B) A low-resolution volumetric reconstruction of
the imaging data shows an aortic transection and large pseudoaneurysm
(arrow). (C) Following endovascular stent graft placement, the
pseudoaneurysm is obliterated.

352

Figure 32.3 Thoracic computed tomography angiography (CTA) study in a
young patient with hypertension of unknown etiology and a severe blood
pressure gradient between the arms and legs. (A) A 3-D volume-rendered
image displays many enlarged, superficial arteries (arrows). (B) A curved
planar reformatted image demonstrates a focal stenosis (arrow) at the level
of the ligamentumarteriosum with poststenotic dilation of the immediate
descending thoracic aorta. This is consistent with aortic coarctation. (C)The
low-resolution volumetric reconstruction of the thoracic aorta alone is
sufficient to make the diagnosis.

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.4 Thoracic computed tomography angiography (CTA) study.
There is an intimal flap that extends from the ascending aorta (long arrow)
into the descending thoracic aorta (short arrow) in a Stanford type A aortic
dissection.

access sites can provide assurance that access site diameters
are adequate or can determine whether a surgical conduit will
be necessary for device placement. 35-42

The ability of CTA to detect accessory vessels accurately
(particularly of the renal artery) is extremely valuable in the
preprocedural consideration of stent-graft placements in
cases of aortic aneurysm (Fig. 32.6). Moreover, its capability of
identifying early branching of the renal artery and aberrant
venous anatomy makes it a primary consideration in screen-
ing potential kidney donors. A prospective study of 52 poten-
tial donor patients who had undergone digital subtraction
angiography (DSA) followed by CTA showed that CTA identi-
fied 24 of 26 accessory renal arteries visualized on DSA, 10 of
11 early branching arteries seen on DSA, and even detected 15
accessory vessels not identified by DSA. 43

Before multidetector technology was incorporated into CT,
adequate morphological depiction of the renal artery caliber in
patients with hypertension was lacking. This is a consequence
of the perpendicular orientation of the renal vessels with re-
spect to the native aorta. The degree of luminal stenosis was
much more accurately defined with conventional catheter
angiography. CTA initially served as a screening modality in
a selected patient population with hypertensive disease.
Today, however, multislice capabilities have increasingly
redefined the role for CTA as a potential primary modality for
both qualitatively and quantitatively assessing renal artery
stenosis (Figs 32.7 and 32.8). 44 - 54

CT imaging has become an increasingly necessary tool for
accurate preoperative planning of endovascular-treated aortic
aneurysms. CTA demonstrates the extent of the aneurysm,
identifies accessory renal arteries, determines the patency of
the inferior mesenteric artery, and shows the position of the
left renal vein(s) (Fig. 32.9). Preassessment no longer simply
entails identifying an aneurysm, monitoring maximum diam-

eter, and noting the relationship of visceral vessels to the
aneurysm. Endovascular treatment planning now also re-
quires accurate measurements of the diameters and lengths
of the proximal placement and distal landing zones, lesion
lengths, neck angulations, and access vessel diameters. The
combination of multidetector CT scanners, computer-aided
workstations, and sophisticated software has simplified
analysis while maintaining necessary accuracy. CTA essen-
tially replicates calibrated catheters used in length measure-
ments and intravascular ultrasound in determining diameter
measurements. Thin section cross-sectional images and
reconstructions can also be used to identify the number
and positions of renal arteries. 55-60

Specialized software simplifies analysis, especially in pa-
tients with tortuous anatomy. Centerline tracking using seed
points and edge detection effectively remodels a sinuous aorta
in three dimensions to provide true length measurements and
diameter measurements that are perpendicular to the vessel
proper. The proximal landing zone measurement therefore
becomes straightforward (Fig. 32.10A,B), as well as the distal
landing zone, neck angulation measurement, and access
vessel diameters and tortuosity. Displaying data with inter-
mediate CT window and level settings allows the user to
evaluate calcification and plaque (Fig. 32. IOC). Alternatively,
vessel analysis and characterization could be performed
with workstations containing preexisting 3-D analysis tools
(Fig.32.10D,E). 61-71

CTA plays an important role in the routine postplacement
evaluation of aortic stent grafts. Volumetric renditions of
the imaging data may more accurately predict the type(s) of
endoleaks that are present compared with cross-sectional
imaging alone (Figs 32.11 and 32.12). In addition, aneurysmal
expansions related to endoleaks are better appreciated. CTA
also reliably detects other complications, including branch
vessel occlusion, stent kinking, stent migration, and graft
thrombosis. 72-80

Occasionally, CTA may detect abnormalities that suggest an
underlying connective tissue disorder. Multiple aneurysmal
dilations and chronic hematomas may suggest disorders such
as Ehlers-Danlos disease, subtype IV, in which abnormal col-
lagen production leads to vessel fragility. A volumetric recon-
struction provides a vascular map which can characterize both
the nature of the aneurysms and their locations (Fig. 32.13).
Affected individuals may also be at risk for arterial rupture
and dissection. 81 ' 82

Modern CTA is highly accurate in detecting carotid artery
stenosis and can replace DSA in many instances. Unlike more
invasive imaging, CTA permits direct visualization of the arte-
rial wall and atheromatous plaque, facilitating the degree of
stenosis. Difficulties in assessment associated with the pres-
ence of calcification can be overcome with curved planar
reformatting and volume-rendered techniques. CTA can
also identify ulcerated plaques that are believed to be strongly
associated with embolic events.

354

Figure 32.5 Thoracoabdominal aortic computed tomography
angiography (CTA) study performed with a single-detector spiral scanner.
(A f B)The 3-D volume-rendered images show a Stanford type B aortic
dissection. The proximal entry site is clearly visualized (long arrows). Note the
"stair step" misregistration artifact (short arrows) due to cardiac motion. (C)
An axial CT image at the level of the aortic arch again shows the proximal
entry site (arrow). (D) From the volumetric reconstruction, the origins of the
visceral arteries and left renal artery (arrows) clearly arise from the true lumen
of the dissection. Axial CT images show that the origin of the right renal
artery (E) comes from the false lumen of the dissection whereas the celiac
trunk (F) arises from the true lumen.

355

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.5 Continued

356

chapter 32 Computed tomography

Figure 32.7 3-D volume-rendered image from a computed tomography
angiography (CTA) study of the abdominal aorta using a Vitrea 2
workstation. This patient with hypertension has a severe right renal artery
stenosis (arrow). (Image courtesy of Vital Images, Inc.)

Figure 32.6 Volumetric reconstruction of an abdominal aortic computed
tomography angiography (CTA) study utilizing specialized software from
Medical Media Systems (MMS). The image reveals four renal arteries (arrows)
in a patient being evaluated for endovascular repair.

Figure 32.8 Automated vessel measurements can be performed to
quantify accurately the degree of vessel stenosis. The interrogated duplicated
right renal artery is highlighted (arrow). Serial transverse images of the two
right renal arteries can be viewed side by side and the amount of stenosis in
any vessel can be instantly calculated. (Image courtesy of Vital Images, Inc.)

357

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.9 Computed tomography angiography (CTA) study of the
abdominal aorta. (A) A sagittal multiplanar reformatted (MPR) image
provides a "one-shot" view of the extent of the aneurysm and shows the
origin of the inferior mesenteric artery (IMA) arising from the aneurysm
proper. (B)CT image at the level of the pelvic inlet. In addition to
demonstrating the take-off of the IMA from the aneurysm (arrow), the
cross-sectional image also reveals concentric intraluminal thrombus.

The assessment of atherosclerotic carotid arterial disease
has largely been governed by the North American Sympto-
matic Carotid Endarterectomy Trial (NASCET) criteria using
conventional catheter-based angiographic orthogonal mea-
surements of carotid artery stenosis as a standard. With subop-
timal imaging techniques of early-generation scanners, CTA
was not a suitable method for assessing stenosis. Suboptimal
contrast opacification may lead to overestimation of the de-
gree of stenosis and may inappropriately triage patients for an
intervention.However, using multidetector scanners with op-
timal contrast effect, CTA with volume rendering is remark-
ably equivalent to MIP imaging in characterizing stenosis
(Fig. 32.14). However, significant data related to the ability of
CTA to serve as a frontline screening modality to reliably dis-
tinguish moderate (50-69%) from severe stenosis (70-99%)
affecting treatment decisions are still lacking. 83-89

The benefits of multidetector CT scanning in assessing the
brain and neurovascular disease were first realized when a
retrospective analysis showed that in 89% of cases, there
was significantly reduced artifact in the posterior fossa in
multidetector row CT studies compared with single-detector
studies. 90 The potential for improved vascular imaging
quickly followed, and CTA has become a vital component in
the assessment of stroke patients. Patients with a nonhemor-
rhagic acute stroke may be appropriately triaged for throm-
bolytic therapy based on the results of a CT brain perfusion
and CTA study. Utilizing a minimal bolus of contrast, a dy-
namic perfusion study determines areas of nonperfused and
ischemic brain tissue. Perfusion maps are then generated by
specialized computer software generating values for mean
transit time, regional cerebral blood flow, and regional cere-
bral blood volume. The extent of perfusion disturbance and
"tissue at risk" can be determined and appropriate therapy
planned. In conjunction, a CTA study from the level of the left
atrium to the circle of Willis can help detect stenosis or occlu-
sion of extra- and intracranial arteries as well as thrombus that
may have precipitated the event. 91 ' 92

Once the gold standard for evaluating and triaging the neu-
rovascular patient, catheter-based angiography has largely
been replaced by CTA for the initial assessment of intracranial
aneurysms. With a single contrast bolus acquisition, the num-
ber, size, geometry, and orientations of aneurysms could be
accurately determined and a follow-up invasive diagnostic
procedure could be reserved for equivocal cases only. Volume
rendering permits accurate visualization of the aneurysmal
neck in most cases and is helpful in directing the patient to-
ward surgical or endovascular therapy (Fig. 32.15).

CTA is a viable alternative to DSA in detecting and charac-
terizing cerebral aneurysms. In a study designed to evaluate
the imaging assessment of middle cerebral artery (MCA)
aneurysms, 251 patients with suspected intracranial
aneurysms underwent evaluation with both CTA and DSA.
The sensitivity and specificity of both CTA and DSA for MCA
aneurysms were 97 and 100%, respectively. In addition, CTA

358

chapter 32 Computed tomography

Figure 32.10 Preprocedural evaluation in preparation forendovascular
stent placement. (A) Aortoiliac computed tomography angiography (CTA)
study. The volume-rendered image demonstrates an infrarenal aortic
aneurysm and allows for evaluation of vascular morphology. The native
vessel wall contains calcifications and is mildly tortuous. (B) Computer-
assisted evaluation of the proximal landing zone using Preview software
provided by Medical Media Systems (MMS). The dataset is sent to the
company and returned on a CD ROM that can be reviewed on any computer
with a Microsoft Windows operating system. The cross-sectional image
window (left side) shows how the diameter at the distal neck is evaluated
using a diameter tool. The 3-D image (right side) indicates the position of
diameter measurements^, see interpretive figure below) just below the
lowest renal artery and in the distal infrarenal neck. Blood flow (■), plaque
and thrombus (□), and wall calcification (□) are also shown. The
reformatted cross-sectional image allows for accurate diameter

measurements as the image reconstruction orthogonal to the centerline
minimizes errors associated with vessel obliquity. The software allows for a
simple calculation of centerline measurements (arrow). (Also in colour; see
Plate 2, facing p. 370.) (C) Another calculation using the MMS software
demonstrates the aneurysm diameter (left), and the centerline length
measurement of the distance from the lowest renal artery to the aortic
bifurcation (right, between arrows). Analysis of the same dataset on a Vitrea
2 Vital Images workstation with vessel analysis software has similar utility to
the Pre view software; however, the analysis can be done in-house. (D)The
automated curved reformatting of the dataset with centerline length
measurement of the infrarenal neck is shown. The diameter measurement
taken perpendicular to the centerline (E), and neck angulation measurement
(F) can be performed. Note that selecting appropriate window and level
settings allows one to distinguish between blood flow, plaque/thrombus,
and calcification.

359

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.10 Continued

360

Figure 32.1 1 Bifurcated stent graft evaluation. (A) A cross-sectional image
from a contrast-enhanced CT study shows contrast within the aneurysm sac
(arrows) and outside of the graft. The location initially suggests a type II
lumbar endoleak. (B) A multiplanar sagittal reformatted image again
demonstrates the endoleak (arrow). This is a suprarenal graft where the
fabric beginsjust below the renal arteries. (C) A volumetric reconstruction
(Preview software, Media Medical Systems) shows a contrast leak (arrows)
around the top of the suprarenal graft as well, making this a type I endoleak.
(D) A translumbar catheter angiogram with contrast injected directly into the
aneurysmal sac (long arrow) shows streaming of contrast upward around the
top of the graft (short arrow). This confirms a type I endoleak.

361

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.1 2 Bifurcated stent graft evaluation. (A) A cross-sectional image
from a contrast-enhanced CT study reveals a collection of contrast within the
aneurysm sac. The arrow marks the origin of the inferior mesenteric artery.
(B) Just below, there is another contrast collection near the origin of the
lumbar artery. (C) At the most inferior level of the aneurysm sac there is yet
another contrast collection (arrow). (D) Volumetric rendering of the dataset
using Medical Media Systems Pre view software both confirms a type II
endoleak (long arrow) and demonstrates a type I endoleak. The type I
endoleak is secondary to a short distal landing zone in the right limb of the
graft (short arrow).

362

chapter 32 Computed tomography

Figure 32.13 Aorto-bifemoral computed tomography angiography (CTA)
study ina patientwith Ehlers-Danlos disease. (A) ACT image at the level of
the upper femurs shows multiple low-density collections of various sizes
within the thigh musculature and scrotum (arrows). These represent chronic
hematomas. (B) A volume-rendered image demonstrates multiple saccular
and fusiform aneurysms arising from the iliac, superficial femoral, and deep
profunda arteries (arrows).

Figure 32.14 Carotid computed tomography angiography (CTA) study.
The curved reformatted planar maximum intensity projection (MIP) images
(left and middle) allow precise determination of the degree of stenosis in a
given vessel. There is a severe stenosis of the proximal internal carotid artery.
In this case, the volume-rendered image (right) provides identical
information. (Image courtesy of Toshiba Medical Systems.)

Figure 32.1 5 Computed tomography angiography (CTA) study of the
circle of Willis. The 3-D volume-rendered image demonstrates multiple
aneurysms (arrows) involving the middle cerebral arteries (MCA) and internal
carotid arteries. On a computerworkstation, the image can be rotated in
various planes to determine the geometry of the lesion to properly triage the
patient for treatment. For example, the aneurysm located at the proximal
right MCA (middle arrow) has an unfavorable conformity for unassisted
endovascular coiling. Such a lesion could be treated surgically orwith
endovascular therapy in which a stent is deployed in the native vessel prior to
coiling.

363

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.16 (A) A 3-D volume-rendered image shows bilateral occlusions
of the superficial femoral arteries with distal reconstitution. (B) A 3-D
maximum intensity projection (MIP) image better demonstrates collateral

vessels arising from the deep profunda vessels. (Image courtesy of Dr Scott
Lipson, Long Beach Memorial Medical Center, Long Beach, CA, USA.)

provided additional information for complex lesions not
detected by DSA, changing the initial treatment planning in
67% of this subset of patients. Moreover, the volume-rendered
depiction of the aneurysm was identical or nearly identical to
the operative findings in 17 of 19 lesions. 93 CTA has demon-
strated increased sensitivity for very small aneurysms com-
pared with DSA. A study of 51 patients harboring 41
aneurysms smaller than 5 mm in size revealed a sensitivity of
98% for detection (compared with 95% for DSA), 100% speci-
ficity, and 99% accuracy. Forty-eight percent of lesions were
detected in the presence of subarachnoid hemorrhage. 94 How-
ever, meta-analyses of studies comparing DSA, CTA, and
MRA have been performed as well as comparative assess-
ments of DSA and CTA in in-vitro models that do not suggest a
statistically significant difference of CTA over DSA and MRA
in the detection of aneurysms. 95-99

Catheter-based angiography once served as the primary
tool for providing detailed information in patients with clini-
cally suspected atherosclerotic peripheral vascular disease.
The location and nature of hemodynamically significant oc-
clusive lesions could be precisely characterized and surgical
or endovascular therapy planned accordingly. Although
modalities such as magnetic resonance angiography and
Doppler ultrasound serve as attractive noninvasive alterna-
tives, technical limitations related to flow dynamics cause
overestimations of the degrees of stenosis. The introduction of
16 detector row CT technology has revolutionized the ap-
proach to peripheral assessment and is quickly shifting the
role of invasive imaging from a definitive diagnostic modality
to a tool for therapeutic interventional preparation. With a sin-
gle bolus of contrast, a survey of the entire vasculature span-
ning the level of the clavicles to the ankles could be performed

364

chapter 32 Computed tomography

Figure 32.1 7 A 3-D volume-rendered image of the calf shows diffuse
irregularity of the vasculature related to atheromatous disease. On the
patient's right (left side of the figure) there is a two-vessel runoff and a severe
stenosis of the proximal right anterior tibial artery. On the left, only the
anterior tibial artery is patent to the ankle. (Image courtesy of Dr Scott Lipson,
Long Beach Memorial Medical Center, Long Beach, CA, USA.)

and accurately uncover segmental stenoses and focal occlu-
sions (Fig. 32.16). A single 3-D rendered volumetric image
from the dataset provides a global overview of the vasculature
and allows for assessment of the extent of atheromatous dis-
ease, for example (Fig. 32.17). These displays of arterial anato-
my reveal not only current vascular disease but also evidence
of previous surgical or endovascular intervention (Fig. 32.18).
Moreover, volumetric data acquisitions now provide precise
anatomical evaluation of abnormal arteriovenous com-
munications and aneurysms as well as ischemic disease in
the distal extremities (Figs 32.19 and 32.20). Entities such as
thoracic outlet syndrome are also easily demonstrated by

Figure 32.18 Arch to ankle computed tomography angiography (CTA)
study. At a glance, the 3-D volume-rendered image shows a bypass graft
extending from the right subclavian artery to the right femoral artery with a
second limb connecting to the left common femoral artery. There are bilateral
occlusions of the superficial femoral arteries. A two-vessel runoff is present to
each foot. This "ghost image" of the bony landmarks allows precise
localization of the diseased portionsof the vesselswhich may assist in
surgical or interventional planning. (Image courtesy of Dr Scott Lipson, Long
Beach Memorial Medical Center, Long Beach, CA, USA.)

365

pa rt 1 1 Noninvasive vascular diagnostics

Figure 32.19 A 3-D volume-rendered image of the hand clearly shows the
characteristics of an ulnar artery aneurysm. (Image courtesy of Toshiba
Medical Systems.)

Figure 32.20 A 3-D maximum intensity projection (MIP) image reveals a
paucity of distal flow in a patient with digital ischemia. (Image courtesy of
Toshiba Medical Systems.)

performing successive CTA studies, placing the patient's
extremity first in neutral position and then repositioning using
the Adson maneuver. 100

CTA has proven efficacy in detecting hemodynamically sig-
nificant vascular stenoses in the lower extremities. Rubin et al.
reported 100% concordance for the absence or extent of dis-
ease in 351 arterial segments when directly compared with
catheter-based angiography with optimal opacification of
the arterial system and minimal interference from venous
enhancement using a multidetector scanner. 101 A study of 48
arteries comparing CTA with catheter arteriography showed
an overall accuracy of 95% for identifying segmental occlu-
sions of greater than 50%. 102 A recent study comparing the effi-
cacy of multislice CTA with intraarterial DSA for detecting
peripheral arterial occlusive disease showed promise for the
less invasive method of screening. Evaluation of 1136 vascular
segments of the leg with MIP reconstructions and traditional
catheter-based angiography showed an overall concordance
of 86% using qualitative assessment categories for stenosis.
Match rates between MIP and DSA proximal and distal to the

trifurcation were 87% and 80% respectively. 103 In comparing
the utility of CTA with DSA when considering thresholds for
treatment, a prospective study found 92% agreement between
the two modalities, with a sensitivity and specificity of 91%
and 92%, respectively. 104

Image processing of the volumetric data using various ren-
dering techniques is crucial for maximizing the yield of avail-
able diagnostic information in terms of extent and severity of
disease as well as considerations for intervention. Volume-
rendered images remarkably simulate conventional angio-
graphic displays. Once a 3-D overview image is studied, the
datasets may be segmented to focus on a particular portion of
the vasculature and the thickness of the data slab can be altered
to optimize analysis. Using bone segmentation and changing
the opacity removes bony structures and extraneous high-
density structures allowing optimal visualization of the vas-
culature. When intervention is planned, ghost images of the
bony architecture may be applied that accentuate the display,
drawing together the anatomic relationships between the
bony elements and vascular disease. Careful manipulation of

366

chapter 32 Computed tomography

window and level settings allows soft tissue structures or
vessels to be emphasized or subtracted as desired.

Certain assessments of the vasculature may require more
than one type of 3-D rendering technique. Although both vol-
ume rendering and MIP can identify vascular calcifications,
determining the degree of stenosis of densely calcified and
tortuous vessels may be difficult. Curved planar reformatting
is an important adjunct in this circumstance, allowing accurate
evaluation of arterial branch ostia and the lumen of an arterial
stent. Overlay of venous structures in the setting of cellulitis
and arteriovenous fistula may obscure proper visualization
of arterial structures, necessitating the use of more than one
reconstruction technique.

The future of CTA will envision expansion of clinical appli-
cations to include improved assessment of the coronary circu-
lation. Scanners with 64 row detectors or more are on the
horizon and will continue to improve the speed of acquisition
and image resolution. The addition of automated stenosis siz-
ing software will more accurately define vascular stenosis.
With these continuing innovations, CTA will supplant
catheter angiography as the primary vehicle in diagnosing
vascular disease.

References

1. Rankin SC. CT angiography. Eur Radiol 1999; 9:297.

2. Bae KT. Peak contrast enhancement in CT and MR angiography:
when does it occur and why? Pharmacokinetic study in a porcine
model. Radiology 2003; 227:809.

3. Brink JA. Contrast optimization and scan timing for single
and multidetector-row computed tomography. / Comput Assist
Tomogr 2003;27 (Suppl. 1):S3.

4. Kuszyk BS, Fishman, EK. Technical aspects of CT angiography.
Semin US CTMR 1998; 19:383.

5. Rankin SC. Spiral CT: vascular applications. Eur } Radiol 1998;
28:18.

6. Brink JA, Heiken JP, Wang G, McEnery KW, Schlueter FJ, Vannier
MW. Helical CT: principles and technical considerations.
Radiographics 1994; 14:887.

7. Brink JA. Technical aspects of helical (spiral) CT Radiol Clin North
Am 1995; 33:825.

8. van Hoe L, Marchal G, Baert AL, Gryspeerdt S, Mertens L. Deter-
mination of scan delay time in spiral CT-angiography: utility of a
test bolus injection. / Comput Assist Tomogr 1995; 19:216.

9. Kirchner J, Kickuth R, Laufer U, Noack M, Liermann D. Opti-
mized enhancement in helical CT: experiences with a real-time
bolus tracking system in 628 patients. Clin Radiol 2000; 55:368.

10. Hittmair K, Fleischmann D. Accuracy of predicting and con-
trolling time-dependent aortic enhancement from a test bolus
injection. / Comput Assist Tomogr 2001; 25:287.

11. Bae KT, Tran HQ, Heiken JP Multiphasic injection method for
uniform prolonged vascular enhancement at CT angiography:
pharmacokinetic analysis and experimental porcine model.
Radiology 2000; 216:872.

12. Luboldt W, Weber R, Seemann M, Desantis M, Reiser M.

Influence of helical CT parameters on spatial resolution in CT
angiography performed with a sub-second scanner. Invest Radiol
1999; 34:421.

13. Fleischmann D, Rubin GD, Paik DS et ah Stair-step artifacts with
single versus multiple detector-row helical CT Radiology 2000;
216:185.

14. Schlueter FJ, Wang G, Hsieh PS, Brink JA, Balf e DM, Vannier M W.
Longitudinal image deblurring in spiral CT Radiology 1994;
193:413.

15. Berland LL, Smith JK. Multidetector-array CT: once again, tech-
nology creates new opportunities. Radiology 1998; 209:327.

16. Rubin GD. Techniques for performing multidetector-row com-
puted tomographic angiography. Tech Vase Interv Radiol 2001; 4:2.

17. Rubin GD, Shiau MC, Schmidt AJ et ah Computed tomographic
angiography: historical perspective and new state-of-the-art
using multi detector-row helical computed tomography. / Com-
put Assist Tomogr 1999 ;23 (Suppl. 1):S83.

18. Fishman EK, Ney DR, Kawashima A, Scott Jr WW, Robertson
DD. Effect of image display on the quality of multiplanar recon-
struction of computed tomography data. Invest Radiol 1993;
28:146.

19. Raman R, Napel S, Beaulieu CF, Bain ES, Jeffrey RB Jr, Rubin GD.
Automated generation of curved planar reformations from vol-
ume data: method and evaluation. Radiology 2002; 223:275.

20. Prokesch RW, Coulam CH, Chow LC, Bammer R, Rubin GD. CT
angiography of the subclavian artery: utility of curved planar
reformations. Comput Assist Tomogr J 2002; 26:199.

21. Herman GT, Liu HK. Display of three-dimensional information
in computed tomography. /ComjPWfAss/sf Tomogr 1977; 1:155.

22. Calhoun PS, Kuszyk BS, Heath DG, Carley CC, Fishman EK.
Three dimensional volume rendering of spiral CT data: theory
and method. Radiographics 1999; 19:745.

23. Johnson PT, Heath DG, Kuszyk BS, Fishman EK. CT angiogra-
phy with volume rendering: advantages and applications in
splanchnic vascular imaging. Radiology 1996; 200:564.

24. Johnson PT, Heath DG, Bliss DF, Cabral B, Fishman EK. Three-
dimensional CT: real-time interactive volume rendering. Am
} Roentgenol 1996; 167:581.

25. Remy-Jardin M, Remy J, Baghaie F, Fribourg M, Artaud D,
Duhamel A. Clinical value of thin collimation in the diagnostic
workup of pulmonary embolism. Am } Roentgenol 2000; 175:407.

26. Schlueter FJ, Zuckerman DA, Horesh L, Gutierrez FR, Hicks ME,
Brink JA. Digital subtraction versus film-screen angiography
for detecting acute pulmonary emboli: evaluation in a porcine
model. / Vase Interv Radiol 1997; 8:1015.

27. Goodman LR, Lipchik RJ, Kuzo RS, Liu Y, McAuliffe TL, O'Brien
DJ. Subsequent pulmonary embolism: risk after a negative
helical CT pulmonary angiogram — prospective comparison
with scintigraphy. Radiology 2000; 215:535.

28. Remy-Jardin M, Remy J. Spiral CT angiography of the
pulmonary circulation. Radiology 1999; 212:615.

29. Schoepf UJ, Costello P. CT angiography for diagnosis of pul-
monary embolism: state of the art. Radiology 2004; 230:329.

30. Munshi IA. Aortic dissection after blunt trauma. / Trauma 2003;
55:1181.

31. Bhalla S, Menias CO, Heiken JP. CT of acute abdominal aortic
disorders. Radiol Clin North Am 2003; 41:1153.

32. Ofer A, Nitecki SS, Braun J et ah CT angiography of the carotid ar-
teries in trauma to the neck. Eur J Vase Endovasc Surg2001; 21:401.

367

pa rt 1 1 Noninvasive vascular diagnostics

33. Soto JA, Munera F, Cardoso N, Guarin O, Medina S. Diagnostic
performance of helical CT angiography in trauma to large
arteries of the extremities. / Comput Assist Tomogr 1999; 23:188.

34. Becker C, Soppa C, Fink U et ah Spiral CT angiography and 3D
reconstruction in patients with aortic coarctation. Eur Radiol
1997; 7:1473.

35. Batra P, Bigoni B, Manning J et ah Pitfalls in the diagnosis of
thoracic aortic dissection at CT angiography. Radiographics
2000; 20:309.

36. Lawler LP, Fishman EK. Multi-detector row CT of thoracic
disease with emphasis on 3D volume rendering and CT
angiography. Radiographics 2001; 21:1257.

37. Ravenel JG, McAdams HP, Remy-Jardin M, Remy J. Multidimen-
sional imaging of the thorax: practical applications. / Thorac
Imaging 2001; 16:269.

38. Johnson PT, Heath DG, Kuszyk BS, Fishman EK. CT angiogra-
phy: thoracic vascular imaging with interactive volume render-
ing technique. J Comput Assist Tomogr 1997; 21:110.

39. Castaner E, Andreu M, Gallardo X, Mata JM, Cabezuelo MA,
Pallardo Y. CT in nontraumatic acute thoracic aortic disease:
typical and atypical features and complications. Radiographics
2003; 23 Spec No:S93.

40. Adachi H, Ino T, Mizuhara A, Yamaguchi A, Kobayashi Y, Nagai
J. Assessment of aortic disease using three-dimensional CT
angiography. / Card Surg 1994; 9:673.

41. Quint LE, Piatt JF, Sonnad SS, Deeb GM, Williams DM. Aortic
intimal tears: detection with spiral computed tomography. /
Endovasc Ther 2003; 10:505.

42. Halpern EJ, Nazarian LN, Wechsler RJ et ah US, CT, and MR
evaluation of accessory renal arteries and proximal renal
arterial branches. Acad Radiol 1999; 6:299.

43. Beregi JP, Mauroy B, Willoteaux S, Mounier-Vehier C, Remy-
Jardin M, Francke J. Anatomic variation in the origin of the
main renal arteries: spiral CTA evaluation. Eur Radiol 1999;
9:1330.

44. Rubin GD, Dake MD, Napel S et ah Spiral CT of renal artery steno-
sis: comparison of three-dimensional rendering techniques. Ra-
diology 1994; 190:181-9.

45. Galanski M, Prokop M, Chavan A, Schaefer CM, Jandeleit K,
Nischelsky JE. Renal arterial stenoses: spiral CT angiography.
Radiology 1993; 189:185.

46. Johnson PT, Halpern EJ, Kuszyk BS et ah Renal artery stenosis:
CT angiography — comparison of real-time volume-rendering
and maximum intensity projection algorithms. Radiology 1999;
211:337.

47. Fleischmann D. Multiple detector-row CT angiography of
the renal and mesenteric vessels. Eur } Radiol 2003; 45 (Suppl. 1):
S79.

48. Lufft V, Hoogestraat-Lufft L, Fels LM et ah Contrast media
nephropathy: intravenous CT angiography versus intraarterial
digital subtraction angiography in renal artery stenosis: a
prospective randomized trial. Am J Kidney Dis 2002; 40:236.

49. Kuszyk BS, Heath DG, Johnson PT, Eng J, Fishman EK. CT
angiography with volume rendering for quantifying vascular
stenoses: in vitro validation of accuracy. Am } Roentgenol 1999;
173:449.

50. Kim TS, Chung JW, Park JH, Kim SH, Yeon KM, Han MC. Renal
artery evaluation: comparison of spiral CT angiography to
intra-arterial DSA. / Vase Interv Radiol 1998; 9:553.

51. Kaatee R, Beek FJ, de Lange EE et ah Renal artery stenosis:
detection and quantification with spiral CT angiography versus
optimized digital subtraction angiography. Radiology 1997;
205:121.

52. Johnson PT, Halpern EJ, Kuszyk BS et ah CT angiography of renal
artery stenosis: comparison of a real-time volume rendering
algorithm with a maximum intensity projection algorithm.
Radiology 1999; 211:337.

53. Kuszyk BS, Heath DG, Johnson PT, Eng J, Fishman EK. CT
angiography with volume rendering for quantifying vascular
stenoses: in vitro validation of accuracy. Am } Roentgenol 1999;
173:449.

54. Smith PA, Fishman EK. Three-dimensional CT angiography:
renal applications. Semin US CTMR 1998; 19:413.

55. Rubin GD, Shiau MC, Leung AN, Kee ST, Logan LJ, Sofilos MC.
Aorta and iliac arteries: single versus multiple detector-row
helical CT angiography. Radiology 2000; 215:670.

56. Prokop M. CT angiography of the abdominal arteries. Abdom
Imaging 1998; 23:462.

57. Rubin GD, Walker PJ, Dake MD et ah Three-dimensional spiral
computed tomographic angiography: an alternative imaging
modality for the abdominal aorta and its branches. / Vase Surg
1993; 18:656.

58. Rubin GD. MDCT imaging of the aorta and peripheral vessels.
Eur } Radiol 2003; 45 (Suppl. 1):S42.

59. Chow LC, Rubin GD. CT angiography of the arterial system.
Radiol Clin North Am 2002; 40:729.

60. Costello P, Gaa J. Spiral CT angiography of abdominal aortic
aneurysms. Radiographics 1995; 15:397.

61. Schwartz LB, Baldwin ZK, Curi MA. The changing face of
abdominal aortic aneurysm management. Ann Surg 2003; 238
(6Suppl.):S56.

62. Tillich M, Hill BB, Paik DS et ah Prediction of aortoiliac
stent-graft length: comparison of measurement methods. Radiol-
ogy 2001; 220:475.

63. Sun Z. Helical CT angiography of abdominal aortic aneurysms
treated with suprarenal stent grafting. Cardiovasc Intervent Radiol
2003; 26:290.

64. Coenegrachts K, Rigauts H, De Letter J. Prediction of aortoiliac
stent graft length: comparison of a semi-automated computed
tomography angiography method and calibrated aortography. /
Comput Assist Tomogr 2003; 27:284.

65. Lutz AM, Willmann JK, Pfammatter T et ah Evaluation of
aortoiliac aneurysm before endovascular repair: comparison
of contrast-enhanced magnetic resonance angiography with
multidetector row computed tomographic angiography with an
automated analysis software tool. / Vase Surg 2003; 37:619.

66. Armon MP, Whitaker SC, Gregson RH, Wenham PW,
Hopkinson BR. Spiral CT angiography versus aortography in
the assessment of aortoiliac length in patients undergoing
endovascular abdominal aortic aneurysm repair. / Endovasc Surg
1998; 5:222.

67. Broeders IA, Blankensteijn JD, Olree M, Mali W, Eikelboom BC.
Preoperative sizing of grafts for transfemoral endovascular
aneurysm management: a prospective comparative study of
spiral CT angiography, arteriography, and conventional CT
imaging. / Endovasc Surg 1997; 4:252.

68. Wolf YG, Tillich M, Lee WA, Rubin GD, Fogarty TJ, Zarins CK.
Impact of aortoiliac tortuosity on endovascular repair of abdomi-

368

chapter 32 Computed tomography

nal aortic aneurysms: evaluation of 3D computer-based assess-
ment. / Vase Surg 2001; 34:594.

69. Baum RA, Carpenter JP, Golden MA et ah Treatment of type 2
endoleaks after endovascular repair of abdominal aortic
aneurysms: comparison of transarterial and translumbar tech-
niques. / Vase Surg 2002; 35:23.

70. Baum RA, Carpenter JP, Tuite CM et ah Diagnosis and treatment
of inferior mesenteric arterial endoleaks after endovascular
repair of abdominal aortic aneurysms. Radiology 2000; 215:
409.

71. Cejna M, Loewe C, Schoder M et ah MR angiography vs CT
angiography in the follow-up of nitinol stent grafts in endolumi-
nally treated aortic aneurysms. Eur Radiol 2002; 12:2443.

72. Pollock JG, Travis SJ, Whitaker SC et ah Endovascular AAA
repair: classification of aneurysm sac volumetric change using
spiral computed tomographic angiography. / Endovasc Ther 2002;
9:185.

73. Rial R, Serrano Fj F, Vega M et ah Treatment of type II endoleaks
after endovascular repair of abdominal aortic aneurysms:
translumbar puncture and injection of thrombin into the
aneurysm sac. Eur J Vase Endovasc Surg 2004; 27:333.

74. Lookstein RA, Goldman J, Pukin L, Marin ML. Time-resolved
magnetic resonance angiography as a noninvasive method to
characterize endoleaks: initial results compared with conven-
tional angiography. / Vase Surg 2004; 39:27.

75. Rozenblit AM, Patlas M, Rosenbaum AT et ah Detection of en-
doleaks after endovascular repair of abdominal aortic aneurysm:
value of unenhanced and delayed helical CT acquisitions.
Radiology 2003; 227:426.

76. Verhagen HJ, Prinssen M, Milner R, Blankensteijn JD. Endoleak
after endovascular repair of ruptured abdominal aortic
aneurysm: is it a problem? / Endovasc Ther 2003; 10:766.

77. Wicky S, Fan CM, Geller SC et ah MR angiography of endoleak
with inconclusive concomitant CT angiography. Am } Roentgenol
2003; 181:736.

78. Kasirajan K, Matteson B, Marek JM, Langsfeld M. Technique and
results of transf emoral superselective coil embolization of type II
lumbar endoleak. / Vase Surg 2003; 38:61.

79. Krueger K, Zaehringer M, Gawenda M, Brunkwall J,
Lackner K. Successful treatment of a type-II endoleak with
percutaneous CT-guided thrombin injection in a patient after
endovascular abdominal aortic aneurysm repair. Eur Radiol 2003;
13:1748.

80. Golzarian J, Murgo S, Dussaussois L et ah Evaluation of abdomi-
nal aortic aneurysm after endoluminal treatment: comparison of
color Doppler sonography with biphasic helical CT. Am }
Roentgenol 2002; 178:623.

81. de Paiva Magalhaes E, Fernandes SR, Zanardi VA et ah
Ehlers-Danlos syndrome type IV and multiple aortic
aneurysms — a case report. Angiology 2001; 52:223.

82. Abdul Wahab A, Janahi IA, Eltohami A, Zeid A, Ul Haque NF,
Teebi AS. A new type of Ehlers-Danlos syndrome associated
with tortuous systemic arteries in a large kindred from Qatar.
Acta Paediatr 2003; 92:456.

83. Randoux B, Marro B, Koskas F et ah Carotid artery stenosis:
prospective comparison of CT, three-dimensional gadolinium-
enhanced MR, and conventional angiography. Radiology 2001;
220:179.

84. Anderson GB, Ashforth R, Steinke DE. CT angiography for the

detection and characterization of carotid artery bifurcation dis-
ease. Stroke 2000; 31:2168.

85. Lev MH, Romero JM, Goodman D et ah Total occlusion versus
hairline residual lumen of the internal carotid arteries: accuracy
of single section helical CT angiography. Am } Neuroradiol 2003;
24:1123.

86. Walker LJ, Ismail A, McMeekin W et ah Computed tomography
angiography for the evaluation of carotid atherosclerotic plaque:
correlation with histopathology of endarterectomy specimens.
Stroke 2002; 33:977.

87. Hirai T, Korogi Y, Ono K et ah Maximum stenosis of extracranial
internal carotid artery: effect of luminal morphology on stenosis
measurement by using CT angiography and conventional DSA.
Radiology 2001; 221:802.

88. Ebert DS, Heath DG, Kusyzk BS et ah Evaluating the potential
and problems of three-dimensional computed tomography
measurements of arterial stenosis. / Digit Imag 1998; 11:151 .

89. Cumming MJ, Morrow IM. Carotid artery stenosis: a prospective
comparison of CT angiography and conventional angiography.
Am} Roentgenol 1994; 163:517.

90. Jones TR, Kaplan RT, Lane B, Atlas SW, Rubin GD. Single- versus
multi-detector row CT of the brain: quality assessment. Radiology
2001; 219:750.

91. Tomandl BF, Klotz E, Handschu R et ah Comprehensive imaging
of ischemic stroke with multisection CT. Radiographics 2003;
23:565.

92. Chuang YM, Chao AC, Teng MM et ah Use of CT angiography in
patient selection for thrombolytic therapy. Am J Emerg Med 2003;
21:167.

93. Villablanca JP, Hooshi P, Martin N et ah Three-dimensional heli-
cal computerized tomography angiography in the diagnosis,
characterization, and management of middle cerebral artery
aneurysms: comparison with conventional angiography and
intraoperative findings. JNeurosurg 2002; 97:1322.

94. Villablanca JP, Jahan R, Hooshi P et ah Detection and characteri-
zation of very small cerebral aneurysms by using 2D and 3D
helical CT angiography. Am J Neuroradiol 2002; 23:1187.

95. Hochmuth A, Spetzger U, Schumacher M. Comparison of three-
dimensional rotational angiography with digital subtraction
angiography in the assessment of ruptured cerebral aneurysms.
Am J Neuroradiol 2002; 23:1199.

96. Karamessini MT, Kagadis GC, Petsas T et ah CT angiography
with three-dimensional techniques for the early diagnosis of
intracranial aneurysms: comparison with intra-arterial DSA and
the surgical findings. Eur J Radiol 2004; 49:212.

97. Piotin M, Gailloud P, Bidaut L et ah CT angiography, MR angio-
graphy and rotational digital subtraction angiography for volu-
metric assessment of intracranial aneurysms. An experimental
study. Neuroradiology 2003; 45:404.

98. Chappell ET, Moure FC, Good MC. Comparison of computed
tomographic angiography with digital subtraction angiography
in the diagnosis of cerebral aneurysms: a meta-analysis.
Neurosurgery 2003; 52:624.

99. Kuszyk BS, Beauchamp NJ, Fishman EK. Neurovascular appli-
cations of CT angiography. Semin US CTMR 1998; 19:394.

100. Katz DS, Hon M. CT angiography of the lower extremities
and aortoiliac system with a multi-detector row helical CT
scanner: promise of new opportunities fulfilled. Radiology 2001;
221:7.

369

pa rt 1 1 Noninvasive vascular diagnostics

101. Rubin GD, Schmidt AJ, Logan LJ,SofilosMC. Multi-detector row peripheral arterial occlusive disease: comparison of multislice-
CT angiography of lower extremity arterial inflow and runoff: CT angiography (MS-CTA) and intraarterial digital subtraction
initial experience. Radiology 2001; 221:146. angiography (IA-DSA). Eur J Med Res 2003; 8:389.

102. Lawrence JA, Kim D, Kent KC, Stehling MK, Rosen MP, 104. Ofer A, NiteckiSS, Linn Seffl/.MultidetectorCT angiography of
Raptopoulos V. Lower extremity spiral CT angiography peripheral vascular disease: a prospective comparison with
versus catheter angiography. Radiology 1995; 194:903. intraarterial digital subtraction angiography. Am J Roentgenol

103. Heuschmid M, Krieger A, Beierlein W et ah Assessment of 2003; 180:719.

370

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Magnetic resonance imaging

David Saloner
Rem van Tyen
Charles M. Anderson
Gary R. Caputo

Vascular disease presents throughout the body with a variety
of associated disturbances in normal flow conditions that have
a profound impact on the vessel wall. 1-3 Magnetic resonance
imaging (MRI), like ultrasound, is a modality with which vas-
cular disorders may be studied noninvasively. MRI can be
used to determine the morphology of blood vessels, assess
blood flow velocities, evaluate the lumen for the presence of
thrombus, and visualize the surrounding tissue to evaluate for
the presence of hemorrhage or infection, and the status of the
end-organ.

Unlike ultrasound, MRI is not compromised by overlying
bone, bowel gas, or calcification. MRI vascular examinations
do not demand the same level of expertise of the operator that
is required by Doppler ultrasound. On the other hand, MRI is
relatively expensive and is limited in situations where metallic
instrumentation might be required.

MRI does not possess the high temporal and spatial resolu-
tion over a large field of view that can be obtained with catheter
angiography; however, MRI is cheaper and poses less risk.
Patients with compromised renal function or severe contrast
allergies can be safely scanned using MRI. There also is no risk
of dislodging emboli as there is with catheter angiography.

Exclusion criteria

There are several conditions that might exclude patients from
undergoing an MRI examination. Patients with pacemakers,
intracranial aneurysm clips, or metal fragments in the eyes
would be at risk and are excluded from MRI studies. Subjects
who are claustrophobic may feel uncomfortable and refuse the
examination. In such cases, sedatives can be administered that
will enable the person to complete the study. MRI studies are
difficult to perform on patients who require close monitoring
or mechanical respiration. Equipment containing metallic
components is strongly attracted by the magnet and should
not be brought into the scanning room. On the other hand,
most implanted devices such as stainless-steel joint prostheses
or heart valves are safe for MRI.

Although not posing any significant health risks, the pres-
ence of surgical clips next to a vessel can produce local distur-
bances of the magnetic field and could reduce the diagnostic
value of a study. Similarly, if the patient is unable to keep still
for the time over which an image series is acquired, image
quality will be compromised. Images can be acquired in times
as short as 10 s, but subjects also might be required to lie still for
images that take longer than lOmin.

Physical basis of magnetic
resonance imaging

MRI is a flexible modality in which parameters can be mani-
pulated to alter the signal characteristics of the tissues being
imaged. Images are classified as proton density-weighted, Tl-
weighted or T2-weighted, and acquisitions can be adjusted to
make blood appear either with high or with low signal. To un-
derstand how this is achieved, it is necessary to understand
some elementary principles of MRI signal production.

MRI methods create images from the signal produced by
protons, the nuclei of hydrogen atoms attached to water mole-
cules and triglycerides. Protons, like many other atomic nu-
clei, have a magnetic moment. When placed in a magnetic
field, the protons can be in a high- or a low-energy state. A pro-
ton in a low-energy state will move to the high-energy state if
the precise amount of energy, in the form of radio frequency
(RF) excitation, is deposited into the proton. Once a proton is in
a high-energy state, it can return to the low-energy state by im-
parting energy to the motion of surrounding molecules. The
rate at which this process takes place is termed Tl relaxation,
and is a property of the tissue surrounding the excited proton.
This is one key parameter in determining tissue contrast in an
MRI image.

When considering magnetic resonance phenomena, it often
is useful to think of the magnetization in a resolvable volume
element. MRI studies typically provide image resolution of the
order of 1 x 1 mm in the plane of the image. Slice thickness can
range from between 2 and 8 mm for two-dimensional (2-D)

371

pa rt 1 1 Noninvasive vascular diagnostics

studies to less than 1mm for three-dimensional (3-D) studies.
MRI images reflect the average magnetization strength of all
protons that are in these image volume elements or voxels.

To determine correctly the magnetization strength in each
voxel in space, the signal must be sampled and labeled using
magnetic fields, referred to as gradients because of their linear
spatial variation. This process must be repeated a large num-
ber of times, each time altering the strength of the gradients.
(For each spatial dimension this requires 128, 256 or even 512
different samples, depending on the required resolution.) The
practical implication of this is that 10 s or more are required to
acquire an image.

Magnetic resonance methods

MRI parameters can be manipulated to be sensitive to differ-
ent anatomic features. Conventional MRI methods make
heavy use of the spin-echo (SE) method, which is a robust
technique for generating images with desirable soft tissue
contrast. For vascular studies, vessel morphology is well
appreciated when the stationary material signal is suppressed
and the flow signal is retained. That is best accomplished using
gradient recalled echo (GRE) sequences. GRE sequences also
can be used with cardiac triggering to provide images at spe-
cific phases of the cardiac cycle. Data acquired over a number
of cardiac cycles (128-256, depending on desired resolution)
can be partitioned into images at short temporal intervals
(about 50 ms) throughout the cardiac cycle by acquiring a por-
tion of each image during each of several cardiac cycles.

Figure 33.1 T1 -weighted spin-echo study (TR = 600 ms, TE = 22 ms) of a

patientwith aortic dissection. Axial 8-mm slice with superior and inferior
presaturation bands. The intimal flap (arrows) is clearly seen as a bright line,
whereas flowing blood appears dark.

T2-weighted images are fluid-weighted and so are useful for
evaluating perigraft hematoma and studying graft infection.

Spin-echo images

Magnetic resonance angiography

Spin-echo images can be produced that are sensitive to the Tl
relaxation process described earlier. In that case, voxels con-
taining material with a rapid relaxation, such as hemorrhage
that is older than a few days or fat, will appear bright, whereas
voxels containing material with a slow relaxation, such as
cerebrospinal fluid, will appear dark. Additional detailed fea-
tures of SE sequences cause magnetization loss from moving
blood and moving protons produce little or no signal on
Tl-weighted images. Blood appears darker than would be
expected even given the relatively long Tl value of blood. 4
This feature is used often to assess patency of a vessel. Since
both the intravascular space and the surrounding tissue are
visualized, MRI permits the evaluation of the true size of a
partially thrombosed aneurysm. Aortic dissection also is well
illustrated using SE images (Fig. 33.1).

SE images also can be designed to be sensitive to a second re-
laxation process, termed T2 relaxation. In this process, one pro-
ton exchanges energy with a neighboring proton. Again, the
probability that this exchange will take place depends on the
specific molecular environment of the given voxel of material.

Magnetic resonance angiography (MRA) represents a class of
MRI sequences in which signal from flowing blood is bright
and signal from surrounding stationary material is dark. 5-10
There are a number of strategies to achieve this, and their dif-
ferences will be discussed later. Using the high contrast-to-
noise ratio between flowing material and stationary material,
images can be built up, either from multiple thin slices through
the vessels of interest or from a projection of those vessels
through an extended volume, to provide representations of
vascular morphology that are similar in appearance to con-
ventional X-ray angiograms. 11

It is important to appreciate the differences between con-
ventional X-ray angiography and MRA. MRA images do not
require the injection of any contrast agent. Flowing blood
emits a different signal from stationary material solely because
of its motion. As such, all flowing blood in the imaged volume
will appear bright, as opposed to X-ray angiography, in which
the only vessels visualized are those that are downstream from
the site-specific contrast injection. Because of the dependence
of signal strength on motion, the relationship of intraluminal

372

chapter 33 Magnetic resonance imaging

signal strength to lumen diameter may be subject to artefacts.
In MRA, flow dynamics contribute significantly to vessel sig-
nal, particularly in regions of flow disturbance where there can
be pronounced signal dropout. A benefit of MRA is that stud-
ies are 3-D. As such, the data can be viewed in projection on
any prescribed plane, providing a full 360° view of the vessels
of interest.

MRA contrast mechanisms

There are three major classes of MRA methods in clinical use —
time-of-flight (TOF), phase contrast (PC), and contrast en-
hanced (CE). 12 To appreciate their differences, certain
properties of the source of the MRA signal must be under-
stood. The signal-producing component of the magnetization
can be described by two components, a magnitude and a di-
rection. TOF and PC MRA methods differ in that the first class,
TOF, relies on generating substantial differences in the magni-
tude of magnetization of flowing spins and stationary spins,
whereas the second, PC, produces differences in the orienta-
tion of the magnetization. CE-MRA derives the high contrast
between vessels and surrounding stationary tissue from the
application of an intravenous injection of a contrast agent
which provides strong Tl-shortening in the intravascular
space, and hence high signal from the vascular lumen.

Time-of-flight contrast

TOF studies typically produce images in which flowing blood
appears bright and stationary material appears dark. 7 The
image contrast arises from properties of the GRE sequences,
invariably used for MRA. Two properties dominate the con-
trast characteristics: the sequences are run with a very short
repetition time, TR, and with a selectable flip angle, a. As noted
earlier, a large number of RF pulses is applied in the course of
producing an image. With each excitation pulse, the magneti-
zation modulus is reduced by a small amount in the selected
volume of excitation. Only limited Tl relaxation can occur be-
cause of the short repetition times used, and a steady-state
value is reached, which, for stationary spins, will be substan-
tially less than the equilibrium magnetization strength. On the
other hand, the signal strength of flowing spins is governed by
the rate at which the spins replenish the excitation volume. In
the limit of rapid blood flow, the entire intraluminal space in
the volume of excitation will be replenished between one RF
pulse and the next. The magnetization strength of that space
will then be registered with maximum strength, reflecting the
equilibrium magnetization value of the newly arrived flowing
spins. For cases of reduced blood flow velocity, the intralumi-
nal space may be only partially replenished between consecu-
tive RF pulses. The vascular signal will then be attenuated, a
process referred to as saturation in MRA. The specific contrast
between flowing blood and stationary material in TOF MRA

will depend on the choice of imaging parameters and on the
flow patterns in the vessels of interest.

TOF images can be acquired using either 2-D or 3-D meth-
ods. In the former, one thin slice at a time is collected, and this
is repeated, moving the slice until the entire volume is covered.
In 3-D methods, the volume is divided into multiple partitions
that are collected simultaneously. Two-dimensional studies
are acquired with slices as small as 2 mm, reducing saturation
effects even in slow flow situations, particularly when the flow
is through the imaging plane. To improve the contrast-to-noise
ratio, however, 2-D studies typically are run with relatively
large flip angles, resulting in significant saturation if the ves-
sels run in plane. Three-dimensional studies provide the high-
est-resolution studies, with voxels that are less than 1 mm in
each dimension; however, because 3-D studies cover a large
region, several centimeters thick, protons will have a lengthy
dwell time in the RF excitation volume, and vessels with slow
flow, such as arteriovenous malformations, veins, or arteries
distal to severe stenosis, may disappear because of saturation.

A different approach is to use advantages of both the 2-D
and 3-D methods. This is achieved by using a series of rela-
tively thin 3-D slabs acquired sequentially. 13 This retains some
of the sensitivity to slow flow that the 2-D studies have, while
keeping the high resolution of the 3-D studies.

An additional powerful tool for TOF MRA should be men-
tioned here. The effect of signal saturation can be deliberately
exploited to good advantage. In many cases, it is simpler to
evaluate the arterial supply or the venous supply without sig-
nal from the other being present. In that case, an additional RF
pulse can be used to saturate completely the signal from blood
entering on one side of the imaging volume. 14,15 If arteries and
veins are flowing in opposite directions, then the signal from
one or the other may be eliminated. These presaturation
pulses also can be used on both sides of the slice (e.g. in
conjunction with the SE technique described earlier) to guar-
antee a dark intraluminal space. Presaturation slabs used on
one side of the imaging volume are useful in determining flow
directionality as well.

The time needed to acquire images varies with the vascular
territory under investigation and the technique used. Single-
slice, 2-D images can be acquired in less than 10 s. It may, how-
ever, be necessary to collect a large number of such slices to
provide adequate coverage with reasonable resolution. A
series of that kind requires 5 min of scan time. Similarly,
3-D studies might require several minutes of acquisition
time, with large-volume, high-resolution studies taking
10 min or longer.

Phase contrast magnetic resonance angiography

As noted, the MRI measurement process measures both the
magnitude of the magnetization in each voxel, and the orienta-
tion of that magnetization in space . The orientation of the mag-
netization in each voxel with respect to the orientation of

373

pa rt 1 1 Noninvasive vascular diagnostics

stationary spins is referred to as the magnetization phase.
In practice, to determine the phase, it is necessary to collect
two datasets. Data subtraction eliminates the effects of
magnetic field imperfections and leaves a residual phase
that can be shown to be directly proportional to the flow
velocity An image that displays the phase of the subtracted
datasets is referred to as a PC study 5 PC images therefore have
midscale gray value for stationary material, whereas flowing
material will either appear bright or dark, depending on the
flow direction. This is used to display flow velocities. Alterna-
tively, the bright signal may be assigned to high velocities re-
gardless of direction to yield a more angiographic-appearing
image.

PC MRA is significantly less sensitive to saturation effects
than TOF MRA because it does not rely on the inflow of fully
magnetized spins into the volume of interest. As such, it is
more sensitive to vessels with slow flow that might not be
visualized in TOF MRA because of saturation. An additional
advantage is that stationary material with extremely short
Tl relaxation times, such as regions of hemorrhage or contrast-
enhanced lesions, which might show up as a vessel-
mimicking high-signal area on TOF studies, is completely
eliminated by the subtraction procedure. In PC, the acquired
data also can be displayed to show flow directionality. PC
methods do, however, require longer acquisition times and
place more stringent requirements on scanner performance
than do TOF methods.

Contrast-enhanced magnetic
resonance angiography

MR contrast agents can be applied by means of intravenous in-
jection and are very well tolerated. They provide a dramatic re-
duction in Tl properties while they remain in the intravascular
space and, following successful performance in research
studies, are now used widely in clinical practice. Because of
the short temporal window during which the benefits of Tl-
shortening are available in the arterial structures of interest,
CE-MRA, like spiral computed tomography (CT) angiogra-
phy, needs careful consideration of the initiation and duration
of the injection of the contrast bolus relative to the interval of
data acquisition for successful studies.

T1 shortening

At the dosages applied in MRA studies, contrast agents act to
reduce the Tl relaxation time of intraluminal blood. The Tl re-
laxation time of blood is of the order of 1 .2 s at field strengths of
1.5 T, the field strength of a high field magnet. When gado-
linium contrast agents are used at sufficiently high concentra-
tions, the Tl relaxation time of blood is reduced to less than
150 ms, well below the Tl of all tissue material. This means that
contrast-enhanced blood will rapidly recover magnetization

and will have high signal strength, even for short values of the
repetition time.

The acquisition of an MR angiogram while using contrast
agents requires a different approach than standard MRA. With
a bolus injection of the contrast agent, there is a short interval
during which the agent will be in the arterial phase. 16 ' 17 It is im-
portant to time the MR data acquisition so that it coincides
with the period during which there is peak arterial signal.
After reaching a peak, the arterial signal strength drops and
the venous signal starts to increase. In conventional MRA
studies of arteries, presaturation pulses are applied superior
or inferior to the volume of interest to eliminate signal from the
veins. In CE-MRA, the application of presaturation pulses is
not a viable strategy for eliminating venous signal. Most con-
trast-enhanced studies rely on using parameters providing the
shortest possible data acquisition time and the addition of a
presaturation pulse substantially increases that time. In any
event, presaturation is of limited utility because the reduced
Tl values of the blood rapidly restore saturated magnetization
strength.

In certain applications, it is important to be able to acquire
a 3-D study in a short time. This includes the extracranial
carotids where there is a short interval when the first pass of
the bolus provides maximal intraarterial signal and when the
venous enhancement, which occurs shortly after the arterial
phase because of the blood-brain barrier, has not yet occurred.
Similarly, short acquisition times are desirable for the vessels
of the abdomen, so that studies can be obtained within a
breathhold. The use of current high-performance gradient
systems permits the acquisition of 3-D studies in times in the
range of 10 s and 20 s.

Acquisition timing

As in other angiographic techniques that use a contrast agent,
timing of image acquisition relative to the passage of the con-
trast agent is of key importance for CE-MRA. In some applica-
tions, multiple injections of contrast material can be used.
However, in order to reduce effects from venous enhance-
ment, and to exploit fully the high magnetization strength that
prevails immediately following injection of the contrast agent,
timing of data acquisition remains important. Appropriate
timing of data acquisition will depend on the specifics of the
acquisition, but can be achieved with several different
approaches.

Acquisition timing: test bolus

A straightforward approach to sequence timing is to follow the
injection of a test bolus in an arm vein 16 ' 18 with image acquisi-
tion at the vessel of interest at 1-s intervals for about 50 s. The
transit time for the contrast to travel from the injector to the

374

chapter 33 Magnetic resonance imaging

vessel can be determined from the image series, and the full
study is then acquired with data acquisition centered on the
calculated arrival of the contrast material. An alternative ap-
proach to the test bolus technique is to use a fluoroscopic
method. 17/19/20 A pulse sequence is initiated that samples the
magnetization strength in the vessel of interest, or in a parent
vessel. The sampling sequence is chosen to be a low-resolution
2-D study with rapid image acquisition time and with imme-
diate reconstruction and display of images, a method referred
to as MR fluoroscopy. After contrast injection, the sampling
study is terminated as soon as contrast arrival is visualized,
and the CE-MRA study is begun.

Advantages of CE-MRA

There are three main advantages to the use of contrast agents
in MRA. First, the total study time required to collect the data
for a 3-D study is quite short, of the order of 10-20 s. This means
that gross patient motion can be substantially reduced.
Studies of the visceral arteries can be performed in a single
breathhold. 21-23 With TOF methods studies of the extracranial
carotid arteries take up to 10 min to acquire and patient motion
such as swallowing, snoring, and neck movement can sub-
stantially degrade image quality. Short-duration CE-MRA
largely avoids these problems. The second major advantage is
the increased coverage that is available with CE-MRA.
Because TOF methods rely on inflow enhancement, signal
strength in distal vessels is diminished, and the only way to en-
sure uniformly high vascular signal through a large volume is
to use multiple overlapping subvolumes to cover the region of
interest. This results in long acquisition times, data inefficien-
cies because of overlap requirements, and the increased possi-
bility of patient motion. Provided contrast material fills the
vessels of interest, CE-MRA can be used to cover a very large
volume with excellent contrast-to-noise properties (Fig. 33.2).
The third major benefit of CE-MRA is that because of the signal
strength, these sequences can be applied using a high band-
width to give very short echo times while still retaining an ad-
equate signal-to-noise ratio. All MRA sequences benefit from
reduced echo times because they restrict the extent of signal
loss that is associated with disordered flow.

Limitations of CE-MRA

While CE-MRA has become the sequence of choice for many
applications, there are some limitations. A principal disadvan-
tage is the need to inject a contrast agent, which means despite
the low-risk profile of side-effects for the agents that are
used, 24 that the study can be done only once. While the actual
data acquisition time is reduced, there is increased prepara-
tion time because of the need to place an intravenous line prior
to placing the patient in the scanner. The administration of an

Figure 33.2 Maximum-intensity projection of coronal three-dimensional
contrast-enhanced MRA acquisition (TR/TE/flip angle = 5 ms/1 .5 ms/35°) of
the aortic arch and great vessels of the neck showing an innominate artery
lesion. Good visual delineation is depicted over a large field of view.

injection requires the presence of additional personnel which,
together with the cost of the contrast agent, adds to the cost of
the study. As noted above, a major concern is the presence of
venous signal that increases with increasing time following in-
jection. This has proved to be a major obstacle in studies of the
intracranial circulation, particularly for the circle of Willis
where the venous signal in the cavernous sinus obscures a de-
lineation of the arterial lumen, and in locations, such as the
lower extremities, where the veins and arteries abut each
other. An additional limitation of CE-MRA is the time con-
straints imposed by the need to capture the high-intensity sig-
nal in the short interval that it is in the arterial phase . Even with
the extremely short repetition times used, 3-D studies can only
be acquired by compromising in terms of coverage and /or
resolution.

The very strong suppression of signal from stationary mate-
rial is advantageous for the visualization of vascular contours.
Conventional MRA sequences also strongly suppress station-
ary material signal but still retain considerably more of that
signal than do CE-MRA sequences. Stationary material signal
can add valuable information to a study of vascular pathology.
At regions of stenosis, conventional MRA sequences often
show features of the atheromatous plaque that cannot be seen
in CE-MRA. The extent of atheroma can be assessed and the
presence of features such as high-signal hematomas is easily
noted on TOF studies.

375

pa rt 1 1 Noninvasive vascular diagnostics

Image display

A three-dimensional dataset can be obtained using either a 3-D
acquisition or a series of contiguous or overlapping 2-D slices
to cover the territory of interest. Although the individual slices
are important in evaluating the vessels, a number of post-
processing tools can be used to aid in interpreting the cross-
sectional data.

The most common postprocessing method in use is referred
to as the maximum-intensity projection (MIP) algorithm. With
this algorithm, the data are projected onto the desired viewing
plane. As opposed to X-ray angiography, which projects a
summation of attenuation effects along the ray, the MIP image
selects out the maximum-intensity voxel lying along the pro-
jection ray and projects that onto the imaging plane. Using the
MIP algorithm, blood vessels will appear in a way similar to
that in X-ray angiograms. The operator can define a restricted
subset of the acquired volume for postprocessing to eliminate
overlapping vessels that might otherwise obscure the vessels
of interest. Restricting the postprocessing volume also
improves the contrast-to-noise properties of the displayed
image. The algorithm is quick and easy to apply. Arbitrary
viewing angles can be prescribed and multiple angles can be
calculated and played back in cine mode to provide the viewer
with a better appreciation of the morphology. The MIP algo-
rithm does, however, have some serious flaws, and it is impor-
tant always to return to the base images after identifying a
possible pathologic feature to confirm that it is not an artifact
of the algorithm.

Clinical applications

MRA techniques have varying success in depicting different
regions of vascular anatomy because of the interplay of flow
behavior, instrument sensitivity, pulse sequences, and signal
intensity at the different locations. Methods that work well at
one site may fail at another. Although imaging of a number of
sites has proved to be highly effective with MRA, consensus as
to the protocol to follow for all areas has not yet been achieved,
and is the subject of active research.

Intracranial vessels

The vessels of the head lend themselves well to high-resolu-
tion TOF studies. It is relatively easy to immobilize the head for
the extended periods needed for such studies. The course of
the carotids to the circle of Willis is well depicted, as are the
distal vertebral arteries and the basilar artery (Fig. 33.3). 25,26
Small branches off those vessels, such as the posteroinferior,
arteroinferior, and superior cerebellar arteries, are seen with
variable success. There typically is good visualization of the

Figure 33.3 Multiple, overlapping three-dimensional axial slabs. Study of
circle of Willis in a patient with stenosis of the basilar artery. Maximum-
intensity projection of three overlapping three-dimensional slabs acquired
with 3-D time-of-flight methods (TR/TE/flip angle = 35 ms/7 ms/25°).

vessels of the circle of Willis, although the extent to which the
more distal small vessels can be delineated varies from subject
to subject, depending on their flow rates. Imaging of the
carotid siphon also is challenging because of the tortuous
geometry there. Siphons often are visualized with fluctuating
signal intensity, reflecting the complex flow patterns.

Aneurysms of the circle of Willis are readily detected with
MRA. 27 The flow patterns in larger aneurysms may be fairly
complex, and it often is difficult to determine whether a region
is slowly rotating or if it is thrombosed. It can be difficult to
identify the neck of an aneurysm if it is small, although the use
of multiple viewing angles can be helpful in that case. Presatu-
ration pulses can also be used to selectively eliminate signal
originating from suspected feeding arteries to determine the
true origin of the aneurysm. Smaller aneurysms have less
complex flow patterns, and, provided they are not in the most
distal branches of the circle of Willis, can be well displayed
with MRA.

Venous flow is significantly slower than arterial flow.
Three-dimensional TOF sequences are poorly suited to those
studies. Vascular malformations that typically are completely
saturated in 3-D TOF studies are well appreciated with 2-D
TOF studies or PC methods. 28 Small arteriovenous malforma-
tions (AVMs) and venous angiomas often are visualized. The
sagittal and transverse sinuses are clearly depicted with 2-D
TOF methods or PC. 29 When using 2-D TOF, it is important to
avoid placing the image plane in the plane of those sinuses.
Coronal slices or oblique sagittal slices are effective ways to
cover those vessels (Fig. 33.4). The presence of thrombus can
be identified, as can invasion of the dural sinus by adjacent
tumor.

Carotid and vertebral vessels

The carotid and vertebral arteries, like the intracranial vessels,
are amenable to studies with extended acquisition times.

376

chapter 33 Magnetic resonance imaging

Figure 33.4 Sagittal maximum-intensity projection of multiple para-
sagittal 2D time-of-flight slices showing the intracranial venous anatomy.

Blood flow typically is rapid and unidirectional, without the
extreme pulsatility that characterizes blood flow in the lower
extremities, for example. The TOF technique has been success-
fully applied to these vessels from their origins at the aortic
arch to their distal branches at the circle of Willis.

The primary region for imaging is the bifurcation of the
carotid artery, site of the greatest prevalence of atherosclerotic
disease. There are proponents of both 2-D and 3-D meth-
ods. 30,31 One of the most challenging questions in evaluation
of carotid artery stenosis is the correct determination of its
grade. This difficulty arises because of the flow disturbances
distal to severe or critical stenoses. When there is disturbed
flow, there is mixing of a broad range of magnetization phases
and a resultant loss in signal. The loss in signal mimics the
effects of a tighter-appearing lesion than is truly present.
In particularly severe cases, the vessel will appear to be
completely interrupted. Three-dimensional sequences are
less susceptible to this signal loss (which generally decreases
as the imaging parameter, TE, the echo time, decreases). Flow
disturbance is less problematic when the stenosis is less than
about 85%. Many of the limitations of 3-D TOF MRA are over-
come by the use of CE-MRA methods. CE-MRA methods pro-
vide extended fields of view because they are less prone to
saturation effects. They also have much reduced sensitivity
to signal loss from flow disturbances. Finally, the short total
acquisition times (~ 20 s) reduce image quality degradation
that occurs in TOF MRA with the extended acquisition time
(lOmin).

If the stenosis is hemodynamically restrictive, reduced flow
can lead to saturation effects in 3-D TOF imaging, and the pres-
ence of a patent internal carotid artery is better depicted by the
2-D TOF sequence. Because the distinction of critical stenosis

Figure 33.5 Multiple, overlapping three-dimensional axial slabs. This study
shows three thin (40 mm) slabs covering the carotid bifurcation and the
vertebral artery. A superior presaturation slab removes the signal from the
jugular vein. The patient has a moderate stenosis of the origin of the internal
carotid artery and a severe internal carotid artery stenosis several centimeters
distal to the bifurcation.

from complete occlusion will determine whether the patient
will benefit from an endarterectomy, it is important in situa-
tions of suspected occlusion to confirm the absence of flow
with a slow-flow-sensitive, 2-D image at the level of the carotid
canal in the base of the skull. PC and CE-MRA studies are use-
ful in making this distinction as well.

Figure 33.5 is a study of a patient with moderate stenosis of
the internal carotid artery depicted using three overlapping
axial 3-D slabs. In all cases, the jugular vein has been elimi-
nated using an axial presaturation slab superior to the
imaging slab.

The origins of the great vessels of the neck from the aortic
arch require a relatively large field of view to accommodate the
relevant anatomy. Visualization of small structures such as the
origins of the vertebral arteries are therefore challenging. Also,
flow disturbances commonly are found where vessels change
direction or bifurcate, and evaluation of the origins of the
carotids is complicated by the signal loss seen with disturbed
flow. For these reasons, and particularly because of the
requirements of the extended field of view, a coronal 3-D

377

pa rt 1 1 Noninvasive vascular diagnostics

Figure 33.6 Maximum-intensity projection of a large coverage field of view
from the aortic arch to the circle of Willis. A coronal 3-D contrast-enhanced
MRA study is shown of a patient with a postendarterectomy patch of the left
carotid bifurcation (with an occluded external carotid artery) showing a
severe stenosis slightly distal to the origin of the right common carotid artery
and a severe stenosis of the common carotid artery proximal to the left
carotid bifurcation.

CE-MRA study is the method of choice for this territory

(Fig.33.6). 32 ' 33

The heart

There is high contrast between the myocardium and the blood
pool in both SE and GRE sequences. The signal intensity with-
in the blood pool is low on SE because of washout effects and
high on GRE because of flow-related enhancement. Slow flow
can be detected on SE images owing to reduction in washout,
whereas high-velocity flow abnormalities can be detected as
signal voids attributable to mixing of spin phase within a
voxel.

MRI has been effective in demonstrating complications of
myocardial infarction, such as intraventricular thrombus, as

well as true and false aneurysms of the left ventricle. MRI also
shows potential for demonstrating regional deficits of myo-
cardial perfusion when used in conjunction with contrast
agents. MRI can be used to provide direct visualization of the
pericardium, enabling the diagnosis of constrictive pericardi-
tis. It also permits identification of pericardial hematoma by its
characteristic signal intensity. Intracardiac thrombi, primary
and metastatic tumors to the heart, as well as mediastinal tu-
mors invading cardiovascular structures all are effectively
demonstrated with MRI. MRI is also extremely useful in defin-
ing cardiac function and can be used to show wall motion de-
fects, and cardiac ejection fraction.

The aorta

Conventional SE sequences applied with cardiac triggering
provide excellent anatomic studies of the aorta. The applica-
tion of presaturation bands above and below the acquired
slices ensures that flowing blood appears black. On the other
hand, high-quality bright blood images can be acquired using
the injection of contrast material. With appropriate timing, ex-
tensive coverage of the aorta can be achieved.

In the study of dissection, MRI clearly demonstrates the inti-
mal flap and the extent of dissection within the aorta. 4 It is,
however, often difficult to differentiate slow blood flow signal
from thrombus on SE sequences, and it is therefore necessary
to acquire images with additional sequences to determine if
the false lumen is patent or thrombosed. One suitable method
is to inject gadolinium contrast medium and acquire images
with very rapid GRE. If a region of the vessel fills with gadolin-
ium, it will be bright, indicating that it is not thrombosed.

CE-MRA studies provide excellent depiction of the aorta.
They can be used to delineate the aortic arch, the thoracic aorta,
and the abdominal aorta. MRI provides cross-sectional views
of aortic aneurysms that allow both the intravascular volume
as well as the mural plaque and the vessel wall to be visual-
ized. In addition, other viewing planes can be acquired to de-
termine the relationship of the aneurysm to other vessels (e.g.
coronal images clearly demonstrate the location of the origins
of the renal arteries with respect to the aneurysm).

Perigraft fluid can be expected to persist for approximately
4 weeks after aortic graft surgery. This fluid is bright on T2-
weighted SE sequences. Fluid increase after the postsurgical
period usually indicates infection. Hemorrhage surrounding
a graft will appear bright on a Tl-weighted sequence if the
hemorrhage is more than 1 week old. Hemorrhage that is
several months old usually will be surrounded by a very dark
rim of hemosiderin.

Abdominal vessels

Respiratory motion is a major consideration in imaging of the
abdomen. Gross motion results in severely degraded studies
if 3-D TOF methods are used. Three-dimensional CE-MRA

378

chapter 33 Magnetic resonance imaging

Figure 33.7 Thrombus in the iliac vein. A single-slice axial image acquired
with breathhold through the iliac vein shows intraluminal clot occupying a
significant area of the lumen. Arterial presaturation was used in this
venogram.

Figure 33.8 Maximum-intensity projection imagesof a small seriesof two-
dimensional coronal breathhold slices each 7 mm thick, acquired with
superior saturation of arterial flow. The portal vein (PV) is well depicted.

studies can be acquired in a breathhold and are therefore
generally more suitable for evaluating abdominal vessels than
are 3-D TOF methods. 17 ' 18 ' 23 ' 34

Conventional angiography of the abdominal veins is prob-
lematic because of the high-contrast loads involved. MRA
studies provide high-contrast images and often can answer
the clinical question with the acquisition of a small number of
2-D slices. The depiction of the venous structures is again sim-
plified by using arterial presaturation. Care must be taken
when placing these presaturation bands not to saturate territ-
ories supplying blood to the vessels of interest.

In cross-sectional images, venous thrombosis is seen as a
dark filling defect within the vascular space (Fig. 33.7). Screen-
ing of the iliac or inferior vena cava can be accomplished with
sequential 2-D slices. In this application, high resolution is not
critical, and the slices accordingly can be spaced apart to pro-
vide the needed coverage. Collateral veins also can be visual-
ized well with 2-D studies.

Portal venous anatomy can be shown reliably with MRA,
even when ultrasound access is difficult or the liver is highly
echogenic. Slice orientation is flexible, and coronal or trans-
verse images are convenient views for imaging the portal and
splenic veins. 35 The coronal acquisition provides a presenta-
tion similar to that obtained in conventional X-ray angio-
graphy studies (Fig. 33.8). The determination of vessel patency
is made easily with MRI, and methods exist (described later)
for determining velocity distributions and volume flow.

Renal arteries have been imaged using 2-D or 3-D TOF
methods and with PC methods. 10 However, imaging of the ab-
dominal vessels is greatly improved by the use of CE-MRA
methods since they can be acquired in a time of the order
of 20 s, permitting breathhold acquisition (Fig. 33.9). 17 ' 22
Accessory renal arteries can be displayed and information on
differential perfusion of the kidneys can be extracted.

Lower extremities

Magnetic resonance studies of the lower extremities have been
greatly improved in recent years by the development of high-
sensitivity coils that collect signal over the full length of the leg.
The availability of CE-MRA methods has revolutionized the
conduct of MRI studies of lower extremity anatomy. It is now
possible to conduct a three- or four-station runoff examination
covering from the level of the renal arteries to the feet follow-
ing a single injection (Fig. 33.10). 20 ' 36-38 Subtraction of a prein-
jection mask provides good depiction of the vessels of interest.
In this acquisition mode the patient table moves from one
imaging station to the next with each imaging station covering
a field of view of about 36 cm.

Magnetic resonance velocimetry methods

Both TOF and PC methods can be adapted to provide velocity
information. 39 TOF methods typically acquire images of the

379

pa rt 1 1 Noninvasive vascular diagnostics

Figure 33.9 Maximum-intensity projection of a breathhold contrast-
enhanced MRA study of the aorta and renal arteries. Total acquisition time
was 25 s providing good delineation of the aorta and the iliac arteries, and of
the renal arteries.

vessels with a long segment of the vessel in the plane of the
image. Saturation bands placed transverse to the vessel of in-
terest appear as dark bands across the vessel. The dark band in
the flowing blood is displaced relative to that in the stationary
material by blood transport effects. It is simple to measure that
displacement and infer the flow velocity, knowing the timing
parameters of the pulse sequence. This can be applied to areas
such as the portal vein, where the data can be collected in a
time short enough to permit a breathhold study 40

Arterial velocities also can be collected with cardiac trigger-
ing. Velocities can then be determined at multiple phases
through the cardiac cycle. These bolus tagging measurements
are complicated by the difficulty in determining the edge of
the tagged material and evaluating the net displacement of the
tag. Methods using inversion tagging have been shown to pro-
vide multiple tags, facilitating the evaluation of velocities. 41

PC methods are useful in determining the distribution of ve-
locities across the lumen of the vessel of interest. They can be
applied with cardiac gating as well. These techniques have
been successful in measuring velocities in complicated flow
regimes, such as in the aortic arch where there can be regions of
simultaneous antegrade and retrograde flow in different parts
of the aorta. 42 They can also be used to measure flow through
several arteries simultaneously provided a slice can be pre-
scribed that is transverse to all vessels at the same time
(Fig. 33.11).

Figure 33.10 Four-station runoff from a patient with extensive vascular
disease. The anatomy covers the thoracic and abdominal aorta, and the
runoff vessels down to the feet. The study method was a contrast-enhanced
MRA method with four coronal three-dimensional volumes. A precontrast
mask is subtracted to improve vessel conspicuity and acquisition time for
both the precontrast and the postcontrast study is of the order of 2 min.
There is slight overlap in coverage of adjacent slabs.

Conclusion

Magnetic resonance imaging methods provide a convenient,
noninvasive modality for imaging vessels in many different
locations. MRI is particularly valuable for patients who have
poor tolerance of contrast studies. Imaging of the vasculature
and the end-organ can be performed in the same session,
providing an integrated study of vascular disease. MRI has

380

chapter 33 Magnetic resonance imaging

Figure 33.1 1 Phase contrast velocity
measurement in the intracranial circulation in a
patient with a fusiform aneurysm of the basilar
artery. Flow velocities are measured in the
cavernous portion of the internal carotid
arteries and in the basilar artery proximal to the
aneurysm (white circled region). Peak velocities
in the basilar artery are plotted as a function of
time in the cardiac cycle.

: — r

1 r

- 1 —

■ ! p

— F 1 —

1 , .

1 1 f—

1 ■

— w—

1 '

•

■

- 1

■

■ i

M _,

250 500

Time (im)

750

greatly reduced the number of catheter angiograms per-
formed, resulting in lower cost and fewer complications.

References

1. Milnor WR. Hemodynamics. Baltimore, MD: Williams & Wilkins,
1989.

2. Strandness DE, Sumner DS. Hemodynamics for Surgeons. New
York: Grune & Stratton, 1975.

3. Stroud JS, Berger SA, Saloner D. Influence of stenosis morphology
on flow through severely stenotic vessels: implications for plaque
rupture. JBiomech 2000; 33:443.

4. Higgins CB. The vascular system. In: Helms CA, ed. Magnetic Res-
onance Imaging of the Body, 2nd edn. New York: Raven Press,
1992:629.

5. Dumoulin CL, Souza SP, Walker MF et al. Three-dimensional
phase contrast angiography. Magn Reson Med 1989; 9:139.

6. Haacke EM, Masaryk TJ. The salient features of MR angiography.
Radiology 1989; 173:611.

7. Laub G A, Kaiser WA. MR angiography with gradient motion refo-
cusing. JComput Assist Tomogr 1988; 12:377.

8. Lenz GW, Haacke EM, Masaryk TJ et al. In-plane vascular imag-
ing: pulse sequence design and strategy. Radiology 1988; 166:875.

9. Masaryk TJ, Tkach J, Glicklich M. Flow, radiofrequency pulse se-
quences, and gradient magnetic fields: basic interactions and
adaptations to angiographic imaging. Top Magn Res Imaging 1991;
3:1.

10. Saloner D, Anderson CM, Lee RE. Magnetic resonance angiogra-
phy. In: Helms CA, ed. Magnetic Resonance Imaging of the Body, 2nd
edn. New York: Raven Press, 1992:679.

11. Keller PJ, Drayer BP, Fram EK et al. MR angiography with two-
dimensional acquisition and three-dimensional display. Work in
progress. Radiology 1989; 173:527.

12. Saloner D. Determinants of image appearance in contrast-
enhanced magnetic resonance angiography. A review. Invest
Radiol 1998; 33:488.

13. Parker DL, Yuan C, Blatter DD. MR angiography by multiple thin
slab 3D acquisition. Magn Res Med 1991; 17:434.

14. Felmlee JP, Ehman RL. Spatial presaturation: a method for
suppressing flow artifacts and improving depiction of vascular
anatomy in MR imaging. Radiology 1987; 164:559.

15. Jara H, Barish MA. Black-blood MR angiography. Techniques, and
clinical applications. Magn Res Imag Clin N Am 1999; 7:303.

16. Kim JK, Farb RI, Wright G A. Test bolus examination in the carotid
artery at dynamic gadolinium-enhanced MR angiography. Radiol-
ogy 1998; 206:283.

17. Wilman AH, Riederer SJ, King BF et al. Fluoroscopically triggered
contrast-enhanced three-dimensional MR angiography with el-
liptical centric view order: application to the renal arteries. Radiol-
ogy 1997; 205:137.

18. Earls JP, Rofsky NM, DeCorato DR et al. Hepatic arterial-phase
dynamic gadolinium-enhanced MR imaging: optimization with
a test examination and a power injector. Radiology 1997; 202:
268.

19. Foo TK, Saranathan M, Prince MR et al. Automated detection of

381

pa rt 1 1 Noninvasive vascular diagnostics

bolus arrival and initiation of data acquisition in fast, three-
dimensional, gadolinium-enhanced MR angiography. Radiology
1997; 203:275.

20. Ho VB, Choyke PL, Foo TK et al. Automated bolus chase
peripheral MR angiography: initial practical experiences and
future directions of this work-in-progress. } Magn Res Imag 1999;
10:376.

21. Leung DA, Hany TF, Debatin JF. Three-dimensional contrast-
enhanced magnetic resonance angiography of the abdominal
arterial system. Cardiovasv Intervent Radiol 1998; 21:1.

22. Leung DA, McKinnon GC, Davis CP et al. Breath-hold, contrast-
enhanced, three-dimensional MR angiography. Radiology 1996;
200:569.

23. Prince MR, Narasimham DL, Stanley JC et al. Breath-hold
gadolinium-enhanced MR angiography of the abdominal aorta
and its major branches. Radiology 1995; 197:785.

24. Niendorf HP, Dinger JC, Haustein J et al. Tolerance data of Gd-
DTPA: a review. Eur J Radiol 1991; 13:15.

25. Lin W, Tkach JA, Haacke EM et al. Intracranial MR angiography:
application of magnetization transfer contrast and fat saturation
to short gradient-echo, velocity-compensated sequences. Radi-
ology 1993; 186:753.

26. Mathews VP, Ulmer JL, White ML et al. Depiction of intracranial
vessels with MRA: utility of magnetization transfer saturation
and gadolinium. / Comp Assist Tomogr 1999; 23:597.

27. Huston J 3rd, Rufenacht DA, Ehman RL et al. Intracranial
aneurysms and vascular malformations: comparison of time-of-
flight and phase-contrast MR angiography. Radiology 1991;
181:721.

28. Edelman RR, Wentz KU, Mattle HP et al. Intracerebral arteriove-
nous malformations: evaluation with selective MR angiography
and venography. Radiology 1989; 173:831.

29. Tsuruda JS, Shimakawa A, Pelc NJ et al. Dural sinus occlusion:
evaluation with phase-sensitive gradient-echo MR imaging. Am J
Neuroradioll991;12-A81.

30. Anderson CM, Saloner D, Lee RE et al. Assessment of carotid
artery stenosis by MR angiography: comparison with x-ray angio-
graphy and color-coded Doppler ultrasound. Am J Neuroradiol
1992; 13:989; discussion 1005.

31. Litt AW, Eidelman EM, Pinto RS et al. Diagnosis of carotid artery
stenosis: comparison of 2DFT time-of-flight MR angiography
with contrast angiography in 50 patients. Am J Neuroradiol 1991;
12:149.

32. Fain SB, Riederer SJ, Bernstein MA et al. Theoretical limits of spa-
tial resolution in elliptical-centric contrast-enhanced 3D-MRA.
Magn Res Med 1999; 42:1106.

33. Fellner FA, Fellner C, Wutke R et al. Fluoroscopically triggered
contrast-enhanced 3D MR DSA and 3D time-of-flight turbo MRA
of the carotid arteries: first clinical experiences in correlation with
ultrasound, x-ray angiography, and endarterectomy findings.
Magn Res Imag 2000; 18:575.

34. Prince MR. Contrast-enhanced MR angiography: theory and opti-
mization. Mag Res Imag Clin N Am 1998; 6:257.

35. Finn JP, Edelman RR, Jenkins RL et al. Liver transplantation: MR
angiography with surgical validation. Radiology 1991; 179:265.

36. Janka R, Fellner F, Requardt M et al. Contrast enhanced MRA of
peripheral arteries with the automatic "floating table". Rontgen-
praxis 1999; 52:15.

37. Westenberg JJ, Wasser MN, van der Geest RJ et al. Scan optimiza-
tion of gadolinium contrast-enhanced three-dimensional MRA of
peripheral arteries with multiple bolus injections and in vitro val-
idation of stenosis quantification. Magn Res Imag 1999; 17:47.

38. Westenberg JJ, van der Geest RJ, Wasser MN et al. Vessel diameter
measurements in gadolinium contrast-enhanced three-
dimensional MRA of peripheral arteries. Magn Res Imag 2000;
18:13.

39. Saloner D. Flow and motion. Magn Res Imag Clin N Am 1999; 7:699.

40. Edelman RR, Zhao B, Liu C et al. MR angiography and dynamic
flow evaluation of the portal venous system. Am J Roentgenol 1989;
153:755.

41. Saloner D, Anderson CM. Flow velocity quantitation using inver-
sion tagging. Magn Res Med 1990; 16:269.

42. Bogren HG, Klipstein RH, Firmin DN et al. Quantitation of ante-
grade and retrograde blood flow in the human aorta by magnetic
resonance velocity mapping. Am Heart J 1989; 117:1214.

382

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Invasive vascular
diagnostics

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Angiography

Anton Mlikotic
C. Mark Mehringer

Since its inception into clinical practice, invasive angiography
served as the standard for the radiographic evaluation of vas-
cular pathology. Vascular disease could be well characterized
and precisely localized. With the advent of computed tomo-
graphy (CT) and MRI (magnetic resonance imaging) and,
more recently, multidetector helical CT angiography (CTA)
with sophisticated image reconstruction capabilities, these
newer and less invasive options are quickly supplanting
catheter-based imaging in assessing many disease processes.
In particular, CTA provides a safer alternative for patients and
superior quality images that can be reconstructed to closely re-
semble catheter angiograms. As a result, the role of invasive
imaging as a primary diagnostic modality has significantly di-
minished. Nevertheless, catheter-based angiography remains
an important tool for the diagnosis of certain disorders.

Indications

Since the 1970s, catheter-based angiography has been the pri-
mary method for evaluation of vascular pathology. During
this time, the accepted indications for diagnostic angiography
remain largely unchanged (Table 34.1). On the other hand,
refinements in catheter technology and the development of
microcatheters and various therapeutic delivery devices have
greatly expanded the possibilities for endovascular treatment.
Moreover, in the past few years there has been a revolution in
CT scanner development and computer technology for med-
ical applications. 1 Multidetector helical CT scanners coupled
with powerful computer workstations have shifted practices
toward less invasive imaging for evaluating many vascular
diseases. The nearly instantaneous acquisition of imaging
data, single bolus contrast requirement, and superior resolu-
tion of transaxial and reformatted images argue against the
use of traditional invasive techniques.

Most structural vascular abnormalities are now reliably de-
tected by CTA and magnetic resonance angiography (MRA),
decreasing the need for catheter angiography. Cross-sectional
imaging has the added advantage of allowing evaluation of

the soft tissue structures adjacent to the vasculature. Invasive
angiography continues to serve an adjunctive role where re-
sults with less invasive imaging are equivocal. Normal variant
structures and anomalies may be uncovered that simulate
pathology on cross-sectional imaging (Fig. 34. 1). 2 Lesions,
such as arteriovenous malformations and arteriovenous
fistulae, where dynamic imaging reveals the temporal
relationships of arterial and venous components, benefit from
catheter angiography, especially when intervention is
planned (Fig. 34.2). A unique feature of catheter angiography
is the ability to obtain segmental intravascular pressure mea-
surements. Pressure gradients can accurately be determined
across areas of suspected hemodynamically significant steno-
sis. This is invaluable, for example, in determining the success
of a transjugular intrahepatic portosystemic shunt (TIPS)
placement.

Contraindications

Although there are no absolute contraindications to perform-
ing peripheral or neurovascular diagnostic arteriography, rel-
ative contraindications are considered on a per case basis.
These include severe hypertension, hypotension, uncor-
rectable coagulopathy, clinically significant iodina ted contrast
material sensitivity, renal insufficiency, congestive heart fail-
ure, and certain connective tissue disorders. 3 ' 4 For diagnostic
venography, evidence of active cellulitis of the extremity to be
imaged, contrast allergy, and renal insufficiency, particularly
in patients with diabetes or congestive heart failure, may pre-
clude performance of the study. 5

Technique

Arterial access is generally achieved via the common femoral
artery. Alternatively, in patients with occlusive aortofemoral
disease, 6 immature or infected femoral grafts, cardiac cathe-
terizations, 7 difficult approaches for neurovascular study, 8 or

385

Table 34.1 Accepted indications for diagnostic angiography

Pulmonary arteriography

Suspected acute pulmonary embolus, in particular, when other diagnostic tests are inconclusive or discordant with clinical findings

For example:
High-probability ventilation-perfusion scan when there is a contraindication to anticoagulation
Indeterminate ventilation-perfusion scan in a patient suspected of having pulmonary embolus
Low-probability ventilation-perfusion scan in a patient with a high clinical suspicion of pulmonary embolus
Ventilation-perfusion scan cannot be performed
Spiral computed tomography is inconclusive or not able to be performed
Suspected chronic pulmonary embolus

Other suspected pulmonary abnormalities, such as vasculitis, congenital and acquired anomalies, tumor encasement, and vascular malformations
Spinal arteriography

Evaluation for spine and spinal cord tumors, vascular malformations, and spinal trauma

Preoperative evaluation before aortic or spinal surgery
Bronchial arteriography

Evaluation for hemoptysis and suspected congenital cardiopulmonary anomalies

Assessment of distal pulmonary artery circulation (through collaterals) in patients who are potential candidatesfor pulmonarythromboendarterectomy
Aortography

Evaluation for intrinsic abnormalities, including transection, dissection, aneurysm, occlusive disease, aortitis, and congenital anomaly

Evaluation of aorta and its branches before selective studies
Abdominal visceral arteriography

Assessment of acute or chronic gastrointestinal hemorrhage; blunt or penetrating abdominal trauma; acute or chronic intestinal ischemia

Intraabdominal tumors; portal hypertension and varices

Evaluation for primary vascular abnormalities, including aneurysms, vascular malformations, occlusive disease, or vasculitis

Pre- and postoperative evaluation of portosystemic shunts

Pre- and postoperative evaluation of organ transplantation

Preliminary procedure for computed tomographic portography
Renal arteriography

Evaluation for renovascular occlusive disease (e.g. for hypertension or progressive renal insufficiency); renal vascular trauma

Evaluation for primary vascular abnormalities, including aneurysms, vascular malformations, and vasculitis

Assessment of renal tumors and hematuria of unknown cause

Pre- and postoperative evaluation for renal transplantation
Pelvic arteriography

Evaluation for atherosclerotic aortoiliac disease, gastrointestinal or genitourinary bleeding, trauma, and pelvic tumors

Evaluation for primary vascular abnormalities, including aneurysms, vascular malformations, and vasculitis

Assessment of male impotence caused by arterial occlusive disease
Extremity arteriography

Evaluation of atherosclerotic vascular disease, including aneurysms, emboli, occlusive disease, and thrombosis

Assessment of vascular trauma

Preoperative planning and postoperative evaluation for reconstructive surgery

Evaluation of surgical bypass grafts and dialysis grafts and fistulae

Assessment of vascular malformations, vasculitis, entrapment syndrome, and thoracic outlet syndrome

Evaluation of tumors
Cerebral arteriography

Define the presence and extent of vascular occlusive disease and thromboembolic phenomena

Define the etiology of hemorrhage

Define the presence, location, and anatomy of intracranial aneurysms and vascular malformations

Evaluate vasospasm related to subarachnoid hemorrhage

Evaluate trauma, vascular supply to tumors, vasculitis, congenital or anatomic anomalies, and venous occlusive disease

Outline vascular anatomy for planning and determining the effect of therapeutic measures

Perform physiologic testing of brain function (e.g. WADA)
Diagnostic venography

Diagnosis of DVT in a patient with a limited duplex exam, infrapopliteal disease, symptomatic extremity after joint replacement, or suspected DVT with
negative duplex exam results

Preprocedural venous mapping, evaluation of valvular insufficiency

Evaluation for venous hypertension, venous stenosis, and venous malformations

Evaluation of tumor involvement or encasement

Targeting for central venous catheter placement
Preprocedural evaluation

Adapted from Quality Improvement Guidelines for Diagnostic Arteriography (2003) and Quality Improvement Guidelines for Diagnostic Infusion Venography
(2003), Society of Interventional Radiology Standards of Practice Committee; and Quality Improvement Guidelines for Adult Diagnostic Neuroangiography:
Cooperative Study between ASITN, ASNR and SIR (2003).

CHAPTER 34 Angiography

Figure 34.1 A patient with a suspected iliac artery aneurysm istriagedfor
potential endovascular repair. During evaluation, the angiogram instead
uncovers an enlarged artery arising from the internal iliac artery. This is a
persistent sciatic artery, a normal variant that is well demonstrated by
angiography. The catheter courses through the common femoral and
external iliac arteries (arrow).

where direct assessment of the distal upper extremity or arm
dialysis fistulae is desired, an axillobrachial approach is used.
Direct carotid puncture arteriography is now performed only
in exceptional cases. A "translumbar puncture" is mainly re-
served for assessing the presence of an endoleak following bi-
furcated aortic stent graft placement and in rare situations
where multifocal occlusive disease leaves no other suitable
option for arterial access. 9-11 Catheterization through synthet-
ic aortofemoral bypass grafts (e.g. PTFE grafts) or superficial
arteriovenous dialysis grafts can be easily accomplished with
some precautions. 12

Located within the femoral sheath and at the medial aspect
of the femoral head, the common femoral artery lies anterolat-
eral to the accompanying vein. Of important note is the fact
that in up to 8% of vessel pairs more than one-quarter of the
vein will overlap the artery posteriorly and approximately 5%
of common femoral arteries bifurcate at the level of the mid-
femoral head. One to two centimeters caudal to the femoral
head, the superficial femoral artery (SFA) bears a constant an-
terior orientation to the corresponding vein. This directs the
location of the skin entry site for retrograde puncture at the in-
ferior edge of the femoral head and may be confirmed fluoro-
scopically An 18- or 19-G needle is then inserted into the artery
angulated upward 30-45° and medially 20° to the sagittal
plane. Using the Seldinger technique over a wire, the needle is
then removed and an appropriate vascular sheath is placed.
The sideport of the sheath is then maintained with a continu-
ous infusion of heparinized saline solution to help prevent
thromboembolic complications. Various catheters of appro-

priate size and shape may then be placed into the sheath over a
wire and introduced into the arterial system to perform a diag-
nostic study. The leg that is less symptomatic is chosen to opti-
mize cannulation and prevent exacerbation of symptoms on
the compromised side. 13-15

When an axillobrachial approach is employed, a right-sided
entry is used to study the thoracic aortic arch or carotid vessels
whereas the left is used to evaluate the descending thoracic
and abdominal aorta. Access is best accomplished at the level
of the proximal brachial artery, approximately 5 cm distal to
the pectoralis musculature. By elevating and abducting the
arm, this ensures avoiding the humeral circumflex vessels
during puncture. 16 Although the distal brachial artery may
serve as an alternative site of entry, its smaller caliber requires
the use of smaller catheters and systemic heparinization. The
use of systemic heparinization during femoral catheterization
remains controversial and is not commonly practiced. 17

Occasionally, direct arterial puncture using anatomic land-
marks becomes challenging due to smaller vessel calibers or
patient body habitus. Smaller access systems using a micro-
puncture needle, wire, and sheath may be used to minimize
puncture site injury. Alternatively, ultrasound guidance may
easily facilitate proper arterial needle placement. Arterial
structures are readily distinguished from their venous coun-
terparts by their thicker walls, relative lack of compressibility,
and Doppler waveform signatures. Certain devices provide
graded attachments and a needle guide to facilitate place-
ment. 18 ' 19 Depending upon the interrogated vessel, a variety of
catheter shapes and sizes is carefully selected to perform the
study. The success of visceral and spinal arteriography relies
heavily upon the shape of the catheter head. In neurovascular
studies, the size of the catheter is especially important where
vessels prone to vasospasm, such as the vertebral artery, are
investigated. 20

Dynamic radiographic vascular imaging has evolved from
cut-film techniques to digital subtraction angiography (DSA)
as image intensifiers improved and computers assumed a
more important role in image acquisition and processing. DSA
offers faster imaging with reduced radiation exposure and
intravascular contrast requirements. However, it lacks the
degree of spatial resolution and therefore imaging quality of
cut-film systems and provides a smaller field of view.

Intravascular contrast media

Since experimental intravascular contrast media were first
injected into cadavers to demonstrate vascular anatomy over
80 years ago, 21 there has been a steady evolution toward con-
trast agents that are safer and more efficacious with decreased
complication rates and improved patient tolerance. The cur-
rent Food and Drug Administration-approved angiographic
contrast agents fall into one of three categories: ionic, iodi-
nated; nonionic iodinated; and noniodinated. Media that have

387

pa rt 1 1 1 Invasive vascular diagnostics

Figure 34.2 Images from a catheter-based diagnostic pelvic and right
femoral arteriogram. The patient sustained a gun shot injury to the right
knee. (A) At the start of the study there is normal opacification of both iliac
and femoral arteries. (B) Within 1 s, however, abnormal early opacification
of the femoral vein (thin arrow) and inferior vena cava (thick arrow) occurs.
This suggests an arteriovenous fistula located below the field of view. (C) A
more selective evaluation of the superficial femoral artery at the level of the
injury site again shows early visualization of the femoral vein (thin arrow) and

a large arteriovenous fistulous communication (thick arrow). The popliteal
artery is transected and therefore not visualized. Below the level of
transection, the popliteal vein is partially seen. (D) An even more selective
injection near the fistula shows reconstitution of the arterial trifurcation via
collateralization from the geniculate arteries (thin arrows). This provides an
"arterial map" if surgical grafting is contemplated. The popliteal vein is even
better visualized (thick arrow).

388

CHAPTER 34 Angiography

high water solubility, high radio-opacity, low viscosity, low
toxicity, and suitable excretion pathways are generally desir-
able for patient evaluation. 22

Ionic contrast media represent the first generation of radio-
opaque compounds put into general clinical use. They consist
of monomeric salts of tri-iodinated benzoic acid with substi-
tuted side-chains containing iodine atoms and a sodium
cation. The iodine-containing anion imparts radio-opacity
whereas the cation makes the compound hypertonic to
plasma. In solution, the osmolality ranges from 1200 to
2400 mOsm/kg H 2 0. Side-effects result from the hypertonici-
ty, ionic charge, and inherent chemical toxicity of the contrast
agent. Ionic agents have five to eight times the tonicity of
plasma and may cause direct injury to blood vessel endo-
thelium, promoting increased capillary permeability and
increased risk of thrombus formation. There is a direct
relationship between the degree of hypertonicity and local
and generalized vasodilation and acute renal failure. The ionic
charge may have effects on electrolyte balance, nerve conduc-
tion, and cardiac activity. Toxicity is inversely proportional to
the degree of hydrophilia of the compound and may be harm-
ful especially to the heart, brain, and kidney. Moreover, pro-
tein (enzyme) binding by the agent may interfere with normal
metabolic activities. 23 ' 24

Nonionic contrast media have different intrinsic properties.
These are also tri-iodinated substituted ring compounds but
do not dissociate in solution and are not hypertonic. Alteration
of side-chain groups confers increased hydrophilicity, and the
osmolality is approximately one-third of their ionic counter-
parts. The lower osmotic dilution of these agents also con-
tributes to improved image quality. The caveat of using these
nonionic agents is their cost (three to four times that of ionic
media) and higher viscosity, which can make rapid injections
through small catheters difficult. 25-28

Nonionic contrast media have significant advantages over
ionic agents. Since most systemic side-effects are related to
hypertonicity, nonionic compounds markedly reduce the
amount of vasodilation that occurs and the resultant sensa-
tions of heat and flushing that cause patient discomfort, as
well as decreasing the degree of transient pain experienced in
some patients during injection. Nonionic compounds have in-
herently lower chemical toxicity due to the presence of longer
side-chains that increase their hydrophilic properties. As a re-
sult, they have a lower tendency to cross cell membranes and
the blood-brain barrier. Their decreased osmolality signifi-
cantly reduces the incidence of hypersensitivity reactions and
cardiotoxicity. In two large retrospective series, the mortality
rates from contrast reactions ranged from 1 : 15 000 to 1 : 75 000.
Frequencies of severe reactions to nonionic media were 0.02-
0.04% compared with 0.09-0.22% for ionic compounds. For
these reasons, ionic agents have nearly fallen out of clinical
practice. 29-31

In patients with hypersensitivity to iodinated contrast, risk
factors that may precipitate a serious adverse reaction, or in

whom a significant contrast load is anticipated, a noniodi-
nated agent such as carbon dioxide may serve as a suitable
alternative for angiography of the peripheral upper extremi-
ties and in cases performed below the diaphragm. This radio-
lucent medium produces excellent image quality with digital
subtraction technique with the advantages of low cost, very
low viscosity, and virtually no toxicity. Patients generally tol-
erate the agent without complaint although some report mild
transient sensory symptoms (i.e. the sensation of pins and
needles) during peripheral angiography. Carbon dioxide has
proven efficacy for use during interventions as well. 32 Success-
ful deployment of inferior vena cava filters using only carbon
dioxide has been demonstrated in 78% of cases in a short se-
ries. 33 In a prospective study, Kessel et al. M reported a prepon-
derance of successful angioplasty, stenting, grafting, and
embolization procedures with carbon dioxide alone, with only
12% of cases requiring a combination of the gas agent with
iodinated media.

Angiographic complications

Arteriographic complications

The Society of Interventional Radiology Standards of Practice
Committee regularly updates quality improvement guide-
lines for diagnostic angiography. Based on information in re-
cently published literature, a range of complication rates for
various indicators (e.g. procedure-related pseudoaneurysm
formation or contrast-induced nephrotoxicity) is reported and
recommended thresholds are established (Table 34.2). Docu-
mentation of physician and institutional complication rates is
strongly advised with the expectation that threshold levels for
each complication indicator are not exceeded. The overall pro-
cedure threshold for major peripheral angiographic complica-
tions is approximately 1%. 3 With strict adherence to proper
technique, complications associated with diagnostic angio-
graphy are quite uncommon. Modern digital subtraction
angiography permits faster acquisitions with decreased
required amounts of potentially harmful intravascular con-
trast media that may reduce procedure-related complications.
Adverse events may be related to the puncture site, induced by
the catheter, or caused by a systemic reaction.

Reported incidences of procedural complications related to
the entry site are lowest for femoral approaches (1.7%). Com-
plication rates at other access sites are slightly higher: 3.3% ax-
illary, 7% brachial, and 3% translumbar. The most common
complication is a minor hematoma with an incidence as high
as 10%. These are usually self-limiting and the patient does not
require additional supportive treatment. Major hematomas
that may require transfusion or surgical evacuation reportedly
occur in 0.5% of femoral punctures and 1.7% of axillary punc-
tures. 3 The morbidity associated with hematoma formation
may reflect its location and not necessarily its size. A small

389

pa rt 1 1 1 Invasive vascular diagnostics

Table 34.2 Indicators and thresholds for complications in diagnostic
arteriography

Complication indicator/major

adverse event threshold

Reported rates (%)

Puncture site complications

Major hematoma

0-0.68

0.5

Occlusion

0-0.76

0.2

Pseudoaneurysm/AV fistula

0.04-0.2

0.2

Catheter-induced complications

Distal emboli

0-0.10

0.5

Arterial dissection

0.43

0.5

Subintimal injection of contrast

0-0.44

0.5

Systemic adverse effects

Major contrast reactions

0-3.58

0.5

Contrast-associated nephrotoxicity

0.2-1.4

0.2

From Quality Improvement Guidelines for Diagnostic Arteriography (2003),
Society of Interventional Radiology Standards of Practice Committee.

hematoma in the axilla may cause significant neural injury to
the brachial plexus whereas a similarly sized collection in the
groin may be inconsequential. 35-37 Rare complications include
arterial dissection, thrombosis, pseudoaneurysm, or arterio-
venous fistula formation, occurring in less than 1% of femoral
punctures. 3

Puncture of the superficial femoral artery instead of the
common femoral artery may lead to pseudoaneurysm forma-
tion. 38 Without the support of the femoral head, a "low stick"
results in relatively less effective manual compression. A
pseudoaneurysm is suspected when a pulsatile hematoma is
present near the puncture site. Ultrasound-guided compres-
sion of the neck of the pseudoaneurysm is effective treatment
in 50-75% of cases. When unsuccessful, direct ultrasound-
guided introduction of thrombin has a demonstrated success
rate of 93% of complete obliteration without recurrence. 39
Pseudoaneurysms that are unamenable to more conservative
therapies may require surgical repair. If left untreated, the
pseudoaneurysm may expand and rupture.

Arteriovenous fistula also commonly results from a low ar-
terial entry below the common femoral artery and is related to
the anatomic relationship of the artery to the vein. Although
the common femoral artery lies lateral to the femoral vein,
more caudally the artery is anterior to the vein. Simultaneous

puncture of the artery and vein may occur with lower punc-
tures and inadequate manual compression at this location
may prevent proper arterial closure. Small arteriovenous fis-
tulae may be asymptomatic; however, larger ones may cause
patients to experience symptoms similar to arterial insuffi-
ciency due to a "steal" phenomenon through the abnormal
arteriovenous connection. 40

Puncture sites that are too high in location and above the in-
guinal ligament may result in a serious complication. Inade-
quate manual compression may lead to the development of a
retroperitoneal hemorrhage, a potentially life-threatening ad-
verse effect that may not be apparent to the physician until the
degree of extravasation becomes significant enough to result
in changes such as tachycardia and hypotension. A large
hematoma about the external iliac artery detected on a CT
study may prompt immediate surgical repair. 41 ' 42

Complications related to catheter manipulation reportedly
occur in 0.15-2.0% of cases. Arterial dissections may be caused
by subintimal passage of the guidewire or catheter and
thromboembolism may occur with catheter manipulation or
contrast injection. With advances in guidewire and catheter
technology, however, the rates of these types of complications
have significantly decreased to about 0.5%. Cholesterol em-
boli occur when cholesterol crystals detach from diseased por-
tions of the aortic wall and occlude small vessels distally.
Cholesterol emboli originating from the thoracic aorta may re-
sult in stroke, whereas those emanating from the abdominal
aorta may cause renal failure or bowel infarction. The mortal-
ity of this complication ranges from 50% to 80%, as these em-
boli are not amenable to anticoagulation therapy. 43 Arterial
spasm also may occur from endothelial cell and smooth mus-
cle irritation of the vessel wall by the presence of a catheter or
guidewire, and can be treated with intraarterial or sublingual
vasodilator and /or balloon angioplasty. 3

Systemic adverse effects overall occur in less than 5% of
cases. Most reactions are anaphylactoid and mild in degree,
manifesting as a sensation of warmth, flushing, pruritis, rhin-
orrhea, scattered urticaria, and diaphoresis. Moderate symp-
toms include repeated vomiting, headache, facial edema,
palpitations, dyspnea, and abdominal cramps. 44-46 The exact
mechanisms underlying these reactions are unknown but are
believed to be secondary to direct cellular effects, enzyme in-
duction, or activation of complement or other systems. 47 ' 48
Vasovagal syncope may occur and is usually characterized by
bradycardia, diaphoresis, lightheadedness, and hypotension.
True arteriographic or venographic contrast allergy is experi-
enced by less than 3% of patients who may present with
marked symptoms that include urticaria, periorbital edema,
and wheezing. Fortunately, most reactions are mild and re-
quire no therapy and less than 1% require hospitalization. 3
Fewer reactions are reported with lower osmolality agents and
prophylaxis with corticosteroids and antihistamines may ben-
efit patients with a history of a previous contrast reaction. 49 ' 50
Delayed reactions occur in 2.1-31% of cases 1 h to several days

390

CHAPTER 34 Angiography

following intravenous contrast administration although the
symptoms are usually mild in degree. 5156

Nephropathy may occur following intravascular contrast
administration. 57-60 There is variability in the quantitative de-
finition of contrast-induced nephropathy and therefore re-
ported rates are difficult to glean from the literature. Porter
recommended defining contrast-induced nephropathy as an
increase in the serum creatinine level by at least 25% if the base-
line level is less that 1.5 mg/dl or an increase of 1.0 mg/dl
above the preangiographic level if the baseline is greater than
1.5 mg/dl occurring within 72 h of contrast administration. 61
Using a definition of an increase in serum creatinine level of
0.3 mg/dl and a 20% increase above baseline, Lautin et al. 62
found the incidence of nephropathy to be 2% in nonazotemic
nondiabetic patients, 10% in nonazotemic diabetic patients
and 38% in diabetic azotemic patients.

Clinically and perhaps more importantly, the Standards of
Practice Committee defines it as "an elevation of serum creati-
nine requiring care that unexpectedly delays discharge or
results in unexpected admission, readmission, or permanent
impairment of renal function/' 3 Preexisting renal insufficiency
or cardiac disease are known risk factors for its development.
Other possible predisposing risk factors include insulin-
dependent diabetes, dehydration, advanced age, multiple
myeloma, high volumes of contrast, and recent exposure to io-
dinated contrast. Low osmolar contrast medium has a small
but definite benefit over high osmolar contrast media for pa-
tients with preexisting azotemia. Preprocedural hydration
may have a protective effect in high-risk patients and some
newer drugs (e.g. Mucomyst) may also have a role in protec-
tion from contrast-media-associated nephrotoxicity 63-67

Diabetic patients require special attention. In addition to mi-
crovascular disease that contributes to renal dysfunction, pa-
tients treated with metformin may develop potentially fatal
lactic acidosis when the drug is used concomitantly with iodi-
nated contrast media. The drug is discontinued at the time of
angiography and then resumed once the creatinine level re-
turns to baseline (usually after 48 h). 23,68

Neurovascular-specif ic angiographic complications

Neurovascular angiography is considered a safe and effective
technique for evaluating various intracranial and extracranial
disorders. Minimalization of angiographic complications
(i.e. thromboembolism or arterial dissection) is best achieved
by careful patient selection, preparation, and education; qual-
ity procedural performance; and patient monitoring. 69 A neu-
rologic event that occurs within 24 h of the angiogram is
generally attributed to the procedure and defined by its dura-
tion and severity. Major neurologic complications are divided
into reversible neurologic deficits, including transient is-
chemic attacks (resolving within 24 h) and reversible cerebral
ischemia (resolving within 7 days), and permanent deficits (ir-
reversible strokes lasting longer than 7 days). Reported rates

and suggested complication-specific thresholds are 0-2.3%,
2.5% for reversible deficits and 0-5%, 1% for permanent
deficits. The overall procedure threshold for all major compli-
cations resulting from adult diagnostic neuroangiography is
2%. 4 ' 70-74

The risks of neuroangiography are higher among patients of
advanced age with severe atherosclerosis, acute subarachnoid
hemorrhage, certain connective tissue disorders, and preexist-
ing symptomatic cerebrovascular disease. Risks are also
related to the length of the procedure, number of catheter
exchanges, tortuous anatomy, catheter size, extent of catheter
manipulation, and amount of contrast media used. 4/75/76

Selective vertebral arterial injections have inherently higher
associated risk. A vigorous contrast injection or the mere pres-
ence of a catheter within its lumen may trigger vessel wall
smooth muscle contraction and vasospasm. Transient cortical
blindness may occur from temporary catheter occlusion. 77 An-
terior spinal artery syndrome with cervical myelopathy is a
documented complication following vessel injection. Compli-
cations occur more often with nondominant right-sided injec-
tions related to smaller vessel calibers. 78

Spinal angiography carries a potential risk of injury to the
spinal cord. Vigorous overinjection of the intercostals or lum-
bar arteries that supply the radiculomedullary feeders may
cause spinal cord infarction. This may be related to vascular
occlusion or contrast toxicity. 79

Venographic complications

In addition to contrast sensitivity, adverse effects related to
contrast venography are uncommon and include effects from
extravasation and thromboembolism. Extravasation into the
adjacent soft tissues may cause significant pain and lead to
skin necrosis. Fortunately, skin necrosis occurs with a reported
rate of only 0.5%. Venous thrombosis at the entry site and with-
in the interrogated vessel that may potentially result in life-
threatening pulmonary embolism is a greater concern and
reported rates vary with ionic (2.6-10%, threshold 3%) and
nonionic (0-9%, threshold 3%) media. 5/80/81 Rare complica-
tions include death, cardiovascular collapse, and bron-
chospasm with both reported rates and suggested thresholds
of less than 1%. 5

Although less commonly performed today, pulmonary
angiography remains a safe procedure, especially if nonionic
contrast agents are used. 82 A retrospective review of 1434 pa-
tients where iopamidol, a nonionic agent, was used resulted in
only four major complications (0.3%), including two cases of
respiratory arrest and two patients with fatal ventricular ar-
rhythmias. There were 11 minor complications (0.8%) consist-
ing of six reversible catheter-induced arrhythmias, two
vasovagal episodes, two minor contrast reactions, and one
case of unexplained chest pain. 83 The routine assessment for
pulmonary embolism is now relegated to contrast-enhanced
CT imaging.

391

pa rt 1 1 1 Invasive vascular diagnostics

Arterial applications

With the widespread availability of advanced CT and MRI
scanners, catheter-based arteriography now has a limited role
in the evaluation of the acute trauma patient. CT is now the
appropriate screening modality for suspected cases of aortic
injury or acute dissection. MRI can further distinguish a dis-
section from an intramural hematoma or penetrating athero-
sclerotic ulcer. In occasional cases where a normal variant
ductus diverticulum cannot be distinguished from an aortic
tear, catheter angiography may be warranted, although false
positives are higher with DSA than conventional cut-film tech-
nique. It may also serve a supplementary role in identifying
the true lumen of a dissection if the CT or MRI imaging quality
is substandard. Invasive angiography may serve an adjunc-
tive role in delineating the nature of vascular disorders that are
not apparent on cross-sectional imaging. Patients with sus-
pected occult arteriovenous fistula or vascular malformation
can be diagnosed and potential intervention can be planned.
In the case of a fistula, a temporary occlusion balloon can be
strategically positioned to assist in control of flow during
surgery (Fig. 34.3). 84_91

Cross-sectional imaging is well suited to reveal anatomic
variants and anomalies (e.g. aortic coarctation) and provides
exquisite detail of the aortic wall and lumen in acquired dis-
ease. CTA has almost entirely replaced catheter-based angio-
graphy in the evaluation of aneurysmal disease. CT and MRI
can accurately depict aortic morphology, caliber, thrombus,
and periaortic structures; in addition, cine MRI has the added
advantage of assessing the functional status of the aorta. 92 ' 93 In
a study comparing the qualitative degree of ostial stenosis of
the great vessels originating from the aortic arch, 3-D MRA
consistently identified stenoses greater than 50% (sensitivity
100%, specificity 98%) in all vessels except the vertebral artery
(sensitivity 100%, specificity 85%) where stenoses were
overestimated. 94

Inflammatory infiltration of the aortic wall may lead to
aortitis. The angiographic findings in Takayasu's disease —
irregular areas of narrowing or occlusion involving medium-
or large-sized arteries emanating from the aorta, such as
the carotid, subclavian, and renal arteries— are usually well
demonstrated with catheter evaluation. As scanner tech-
nology and postprocessing reconstruction algorithms contin-
ue to improve, disorders such as vasculitis and fibrodysplasia
will undoubtedly be screened for with cross-sectional imaging
alone. In fact, MRI may detect aortitis before conventional
DSA. 95-99 Invasive imaging of vascular inflammatory disease
has a variety of radiographic presentations. Large-, medium-
and small-caliber vessels may specifically be involved or a
general pattern may be elicited from imaging. Vasculitis may
result in stenosis, occlusion, "beaded" areas of vascular
narrowing, thrombosis, vessel rupture, and aneurysm or
pseudoaneurysm formation (Fig. 34.4). Interrogation of the

Figure 34.3 A patient with a remote history of a stab wound injury to the
rightthigh presents with abdominal pain. He is initially evaluated with a
contrast-enhanced computed tomography study of the abdomen and pelvis.
(A) Although unremarkable for the patient's chief complaint, the study
incidentally did reveal marked dilation of the right iliac and common femoral
veins (arrow). (B) With the image intensifier positioned overthe rightthigh,
there is a pseudoaneurysm of the superficial femoral artery (thick arrow)
adjacent to an arteriovenous fistula. Abnormal, early visualization of the
femoral vein (long, thin arrow) and the normal distal superficial femoral
artery (short, thin arrow) is noted.

visceral vasculature (including the renal arteries) is usually
sufficient to reach a diagnosis. With continuing progress in
cross-sectional technology, CTA and MRA may predominate
in the diagnostic workup of these disorders. 100

Vasospastic diseases of the distal peripheral arteries of the
hand (Raynaud's disease or phenomenon) manifest angio-
graphically as small vessel spasm of the digits. Thromboangi-
itis obliterans (Buerger's disease) may also be associated with
Raynaud's phenomenon presenting as stenoses or occlusions
of medium-sized arteries with a characteristic "corkscrew"
appearance. 101 ' 102 In the past few years, CT visualization of the

392

CHAPTER 34 Angiography

Figure 34.4 Venographic evaluation of a patient with suspected
May-Thurner syndrome. (A) An initial assessment of the left femoral vein
revealsan unusual appearance of the left common iliac vein. (B)Avenogram
delineates the abnormality, a lucent diagonal band that crosses the iliac vein
near the venous confluence. This represents the impression made by the right
iliac artery. The vein is compressed between the artery and the underlying
spine, confirming the diagnosis. (C)The lesion istreated by placing an
endovascular stent (arrow).

distal vasculature has markedly improved and can suggest
these disorders and associated structural abnormalities. How-
ever, contrast arteriography can effectively exclude other
causes of ischemia when improved flow is noted following the
intraarterial administration of a vasodilating agent, such as
tolazoline or reserpine.

Catheter-based angiography has been a mainstay for evalu-
ating noninflammatory vascular diseases of the extremities.
Angiographic runs with supplemental views can readily un-
cover the nature of many disorders, including acute or athero-
sclerotic occlusive disease, malformations, aneurysms,
traumatic injury, and diabetic peripheral vascular dis-
ease. 103 ' 104 Popliteal cystic adventitial disease and popliteal
artery entrapment syndrome have characteristic angio-
graphic findings. 105-113 Rare entities such as Klippel-
Trenaunay-Weber syndrome are easily diagnosed by the
presence of an arteriovenous malformation and absence of a
deep femoral vein. Upper extremity arteriography is useful in
the evaluation of trauma and the subclavian steal phenome-

non. In suspected cases of thoracic outlet syndrome, imaging
of the arm in the neutral and abducted positions leads to a
diagnosis. 114-118

Cross-sectional imaging has proven utility in diagnosis of
many vascular disorders of the extremities. Contrast-
enhanced 3-D MRA can precisely localize occlusions in
Leriche syndrome. 119 For evaluating the patency of foot vascu-
lature in patients with diabetic vascular disease, a prospective
study indicated that 3-D MRA appears to be better than
conventional DSA. 120 MRI studies can now both diagnose and
determine the extent of peripheral arteriovenous malforma-
tions, a task previously performed by DSA alone. 121 ' 122 CTA
and dynamic MRI studies are becoming better appreciated for
their role in diagnosing positional disorders, including tho-
racic outlet syndrome and popliteal entrapment syndrome.
The diagnosis of subclavian steal syndrome can be made by
sonography alone, although CTA and MRI/MRA can both
diagnose a "steal" and demonstrate the subclavian artery
occlusion. 123-126

393

pa rt 1 1 1 Invasive vascular diagnostics

Figure 34.5 Visceral arterial angiogram. The patient is a middle-aged
woman with a history of systemic lupus erythematosus who presents with
abdominal pain. (A) A selective injection of a common origin celiac axis-
superior mesenteric artery shows an abnormal pattern of alternating
segmental stenoses (arrows) and dilation suggestive of arteritis. This departs
from the normal angiographic appearance of gradually tapering vessel
caliber. (B) Further interrogation of the inferior mesenteric artery (IMA) also
demonstrates a markedly abnormal vessel. There are again alternating
stenotic and dilated segments. Note the abrupt transition distally with hair-
thin opacification reflecting diffuse vessel wall inflammation (last, thin
arrow). The preceding segment (thick arrow) is abnormally of same or larger
caliber to the vessel at its origin at the catheter tip.

Venous applications

The widespread availability of sonography, CT, and MRI has
made systemic catheter-based venography a nearly obsolete
practice. 127 Superior reconstruction capabilities with CT and
MRA have further reduced the indications for an invasive ve-
nous procedure. With a sonographic accuracy of 97%, contrast
is rarely used today to evaluate deep venous thrombosis of the
lower extremities except in the calf, where indeterminate re-
sults may occur in approximately 30% of cases. 128 Certain MR
techniques, such as flow-independent venography, however,
are beginning to show promise in evaluation of the calf vascu-
lature. 129 Catheter venography does play a role in the evalua-
tion of dialysis access fistulae, pinpointing the nature and
location of obstructions in patients with clinically suspected
thoracic outlet or superior vena cava occlusions, and prepar-
ing for endovascular interventions. 130

Prosthetic dialysis arteriovenous grafts are prone to occlu-
sion at the site of venous anastomosis. Contrast venography
may readily identify the presence of thrombosis and delineate
the nature of the occlusion. However, 3-D CT or MRI maxi-
mum intensity projection (MIP) imaging can provide identical
diagnostic information as well as a vascular map of the entire
forearm when fistula placement or revision is considered.
When endovascular intervention is planned, contrast veno-
graphy becomes necessary.

Female patients near the age of 40 with left lower extremity
swelling, edema, and pain sometimes are diagnosed with
"May-Thurner syndrome" or "iliac compression syndrome,"
where contrast venography plays an important role in man-
agement. Venographic findings are characteristic and the ex-
cellent spatial resolution of DSA permits visualization of the
intraluminal spurs or webs that help define the anomaly 131,132
Recently, however, MRI has proved a better alternative in di-
agnosing this condition since it not only reproduces the veno-
graphic hallmarks in multiple imaging planes but also
excludes other causes of iliac venous thrombosis, such as
pelvic masses. 133 Catheter venography is necessary for proper
sizing of balloons and stents. There is excellent long-term suc-
cess of treatment with endovascular thrombolysis, balloon
angioplasty, and stent placement 134 (Fig. 34.5).

Upper extremity venography is indicated to evaluate the
cause of arm swelling and to provide a "venous map" for an-
ticipated surgical fistula creation. Catheter venography is use-
ful to determine the nature and location of a central occlusion
or stenosis that may result in arm or head swelling due to im-
peded flow. At the same time, thrombolysis and angioplasty
can be performed to relieve the obstruction and restore normal
flow 135,136 (Fig. 34.6).

Although once a primary diagnostic tool for the evaluation
of pulmonary thromboembolic disease, pulmonary angio-
graphy has fallen out of favor to advanced CT and MRI
imaging techniques. Pulmonary embolism and pulmonary

394

CHAPTER 34 Angiography

Figure 34.6 Thoracic venography. The patient presents with progressive
bilateral upper extremity swelling and has an underlying connective tissue
abnormality. (A) A left-sided subclavian vein contrast study shows gradual
tapering of the left innominate vein to a point of obstruction at the junction
of the innominate veins (thick arrow). There is retrograde flow into the
enlarged left internal jugularvein (thin, small arrows)aswell as visualization
of multiple thoracic collateral veins (thin, larger arrows). (B) Note the "waist"
in the inflated balloon during attempted angioplasty.

arteriovenous malformations are now routinely diagnosed
with less invasive modalities, although catheter-based
angiography may play a role in equivocal studies.

Neurovascular applications

Traditional four-vessel cerebral angiography plays a crucial
role in appropriately triaging the neurovascular patient for
either surgical or endovascular intervention. Its ability to
localize precisely and identify the flow characteristics and
geometry of aneurysms and vascular malformations makes it
an invaluable source of diagnostic information. State of the art
fluoroscopy and modern digital subtraction angiographic
techniques have made the examination considerably safer for
patients in terms of contrast load and radiation exposure.
Complicated lesions, however, may require multiple angio-
graphic runs that carry a significantly increased risk for the pa-
tient. With a single acquisition, CTA images of aneurysms can
now be viewed as a 2-D dataset or a volume-rendered or MIP
image on computer workstations equipped with specialized
software. The size, shape, and orientation of the lesion can be
readily determined. This becomes especially important where
the aneurysmal configuration may obscure visualization of
the neck. In cases of subarachnoid hemorrhage, however,
catheter-based angiography easily diagnoses intracranial
vasospasm and plays a role in management. 137,138

Although MRI can now diagnose many intracranial vascu-
lar lesions, invasive imaging is a necessary step in evaluating
carotid-cavernous fistulae. 139 Using various maneuvers to
alter the arterial flow into the lesion during the examination,
the nature of the rent of a direct-type fistula can be determined
and endovascular therapy with detachable balloons or coils
can be planned. The arterial pedicles supplying an indirect-
type fistula can be identified by selective injections of the
internal and external carotid arteries in preparation for
embolization. 140

Spinal arteriography with selective catheterization of the
radiculo-pial and radiculo-medullary arteries that supply the
anterior spinal artery is useful to characterize vascular malfor-
mations that potentially may be treated by catheter-delivered
embolization. A thorough search is required to prevent poten-
tial spinal cord injury. Typically, this arises from a left inter-
costal or lumbar artery from the T8 to L4 vertebral level,
although it may originate from the bronchial, costocervical,
thyrocervical, or thyroidal arteries. 141-143 Preangiographic
imaging studies with CT and /or MRI serve to tailor the exam-
ination to specifically targeted regions (Fig. 34.7). Patients
with severe aortic atherosclerotic disease may be difficult to
evaluate secondary to narrowing of the ostia of the arteries
that supply the lesion, thus generating false negatives or in-
complete assessment. 144 On occasion, invasive angiography
may be indicated in the workup of suspected vascular spinal
cord tumors, such as hemangioblastomas. 145

395

Figure 34.7 A 32-year-old man without a history of trauma presents with
sudden neck pain and meningismus. A lumbar puncture showed frank
subarachnoid hemorrhage. (A) A sagittal image from a magnetic resonance
imaging study of the spine shows a low intensity lesion at the T1 level
(arrow) posterior to the spinal cord and a similar lesion at the T6 level. This
may represent evidence of past hemorrhage or a flow void related to a
vascular lesion. (B) A spinal angiogram reveals an arteriovenous fistula during
injection of the left T-1 1 intercostal artery. An image captured in the early
arterial phase shows normal arterial opacification (white arrows). Note the
edge artifact (black arrow) produced by marked density differences at the

interface of the aerated lung and diaphragm. During the arterial phase (C)
there is normal contrast opacification of the intercostal artery (thin arrow) but
also early visualization of a paravertebral vein (thick arrow). Similar
arteriovenous fistulae were identified at the T1 0, L3 and L4 levels. Without a
history of trauma, a connective tissue disorder becomes a consideration. (D)
An injection of the right common carotid artery reveals a markedly narrowed
proximal internal carotid artery and then multiple saccularoutpouchingsor
ulcerations (thin arrows) separated by focal tight stenoses (thick arrows). (E)
Evaluation of the external iliac artery demonstrates a "beaded" appearance
supporting connective tissue disease.

CHAPTER 34 Angiography

With continued advances in noninvasive imaging, the role
of catheter-based diagnostic angiography has become rede-
fined. Once the primary method for unraveling vascular dis-
ease processes, invasive imaging now serves as an important
adjunct to cross-sectional imaging, an invaluable tool in
preparation for endovascular therapy, and the standard by
which other investigations are judged.

Acknowledgment

The authors thank Michael Douglas for assistance in prepar-
ing some of the images.

References

1. Bertland LL, Smith JK. Multidetector-array CT: once again
technology creates new opportunities. Radiology 1998; 209:
327.

2. Brancaccio G, Falco E, Pera M, Celoria G, Stefanini T, Puccianti F.
Symptomatic persistent sciatic artery. / Am Coll Surg 2004;
198:158.

3. Singh H, Cardella JF, Cole PE et ah Quality improvement guide-
lines for diagnostic arteriography. / Vase Interv Radiol 2003;
14:S283.

4. Citron SJ, Wallace RC, Lewis CA et ah Quality improvement
guidelines for adult diagnostic neuroangiography / Vase Interv
Radiol 2003; 14:S257.

5. Brown DB, Singh H, Cardella JF et ah Quality improvement
guidelines for diagnostic infusion venography. / Vase Interv Radi-
oZ 2003; 14:S289.

6. Gaines PA, Reidy JF. Percutaneous high brachial aortography: a
safe alternative to the translumbar approach. Clin Radiol 1986;
37:595.

7. Miller HC, Miller GA. Experience with systemic heparanization
during cardiac catheterization by brachial arteriotomy Br Heart}
1974; 36:1122.

8. Westcott, JL, Taylor PT. Transaxillary selective four-vessel
arteriography. Radiology 1972; 104:277.

9. Sonesson B, Dias N, Malina M et ah Intra-aneurysm pressure
measurements in successfully excluded abdominal aortic
aneurysm after endovascular repair. / Vase Surg. 2003; 37:733.

10. Lipski DA, Ernst CB. Natural history of the residual infrarenal
aorta after infrarenal abdominal aortic aneurysm repair. / Vase
Surg 1998; 27:805.

11. Henry GA, Williams B, Pollak J, Pfau S. Placement of an intra-
coronary stent via translumbar puncture. Catheter Cardiovasc
Interv 1999; 46:340.

12. Glanz S, Bashist J, Gordon DH et ah Angiography of upper
extremity access fistulas for dialysis. Radiology 1982; 143:45.

13. Lechner G, Jastch H, Waneck R, Kreschmer G. The relationship
between the common femoral artery, the inguinal crease, and the
inguinal ligament: a guide to accurate angiographic puncture.
Cardiovasc Intervent Radiol 1988; 11:165.

14. Baum PA, Matsumoto H, Teitelbaum GP et ah Anatomic relation-
ship between the common femoral artery and vein: CT evalua-
tion and clinical significance. Radiology 1989; 173:775.

15. Dotter CT, Rosch J, Robinson M. Fluoroscopic guidance in
femoral artery puncture. Radiology 1978; 127:2661.

16. Lipchik EO, Sugimoto H. Percutaneous brachial artery catheteri-
zation. Radiology 1986; 160:842.

17. Miller DL. Heparin in angiography: current patterns of use. Radi-
ology 1989; 172 (3 Pt 2):1007.

18. Yeow K-M, Toh C-H, Wu C-H et ah Sonographically guided ante-
grade common femoral artery access. / Ultrasound Med 2002;
21:1413.

19. Kwon TH, Kim YL, Cho DK. Ultrasound-guided cannulation of
the femoral vein for acute haemodialysis access. Nephrol Dial
Transplant 1997; 12:1009.

20. Soustiel JF, Shik V, Shreiber R. Basilar vasospasm diagnosis:
investigation of a modified "Lindegaard Index" based on imag-
ing studies and blood velocity measurements of the basilar
artery. Stroke 2002; 33:72.

21. Hoppe JO. Angiographic contrast media. In: Schobinger RA,
Ruzicka FF, eds. Vasular Roentgenology: Arteriography, Phlebo-
graphy, Lymphography. New York: Macmillan, 1964:8.

22. Keizur JJ, Das S. Current perspectives on intravascular contrast
agents for radiological imaging. / Urol 1994; 151 :1470.

23. Al Dieri R, Beguin S, Hemker HC. The ionic contrast medium
ioxaglate interferes with thrombin-mediated feedback activa-
tion of factor V, factor VIII and platelets. / Thromb Haemost 2003;
1:269.

24. American College of Radiology. Manual on Iodinated Contrast
Media, Edition 4.1, Reston, VA: American College of Radiology,
2001.

25. McClennan BL. Low osmolality contrast media: premises and
promises. Radiology 1987; 162 (1 Pt 1):1.

26. Ellis JH, Cohan RH, Sonnad SS, Cohan NS. Selective use of radio-
graphic low-osmolality contrast media in the 1990s. Radiology
1996; 200:297.

27. Bettmann MA. Guidelines for the use of low-osmolality contrast
agents. Radiology 1989; 172:901.

28. Spring DB, Quesenberry CP Jr. Costs of low-osmolar contrast
media./AMA1991;266:1081.

29. Lawrence V, Matthai W, Hartmaier S. Comparative safety of
high-osmolality and low-osmolality radiographic contrast
agents. Report of a multidisciplinary working group. Invest
Radiol 1992; 27:2.

30. Bettmann MA. Ionic versus nonionic contrast agents for intra-
venous use: are all the answers in? Radiology 1990; 175:616.

31 . Cohan RH, Ellis JH, Dunnick NR. Use of low-osmolar agents and
premedication to reduce the frequency of adverse reactions to
radiographic contrast media: a survey of the Society of Uro-
radiology Radiology 1995; 194:357.

32. Holtzman RB, Lottenberg L, Bass T, Saridakis A, Bennett VJ,
Carrillo EH. Comparison of carbon dioxide and iodinated con-
trast for cavography prior to inferior vena cava filter placement.
Am J Surg 2003; 185:364.

33. Sullivan KL, Bonn J, Shapiro MJ, Gardiner G A. Venography with
carbon dioxide as a contrast agent. Cardiovasc Intervent Radiol
1995; 18:150.

34. Kessel DO, Robertson I, Patel J 5th et ah Carbon dioxide digital
subtraction arteriography. Am J Roentgenol 1982; 139:19.

35. Tsao BE, Wilbourn AJ. The medial brachial fascial compartment
syndrome following axillary arteriography. Neurology 2003;
61:1037.

397

pa rt 1 1 1 Invasive vascular diagnostics

36. Smith DC, Mitchell DA, Peterson GW, Will AD, Mera SS, Smith
LL. Medial brachial fascial compartment syndrome: anatomic
basis of neuropathy after transaxillary arteriography. Radiology
1989; 173:149.

37. O'Keefe DM. Brachial plexus injury following axillary arterio-
graphy. Case report and review of the literature. JNeurosurg 1980;
53:853.

38. Rapoport S, Snidermas KW, Morse SS et ah Pseudoaneurysm:
a complication of faulty technique in femoral artery puncture.
Radiology 1985; 154:529.

39. Edgerton JR, Moore DO, Nichols D et ah Obliteration of femoral
artery pseudoaneurysm by thrombin injection. Ann Thorac Surg
2002;74:S1413.

40. Altin RS, Flicker S, Naidech HJ. Pseudoaneurysm and arteriove-
nous fistula after femoral artery catheterization: association with
low femoraipunctures. Am J Roentgenol 1989; 152:629.

41. Kent KC, Moscucci M, Mansour KA et ah Retroperitoneal
hematoma after cardiac catheterization: prevalence, risk factors,
and optimal management. / Vase Surg 1994; 20:905.

42. Lodge JP, Hall R. Retroperitoneal haemorrhage: a dangerous
complication of common femoral arterial puncture. Eur J Vase
Surg 1993; 7:355.

43. Hauben M, Norwich J, Shapiro E et ah Multiple cholesterol em-
boli syndrome — six cases identified through the spontaneous re-
porting system. Angiology 1995; 46:779.

44. Bettmann MA, Heeren T, Greenfield A et ah Adverse events with
radiographic contrast agents: results of the SC VIR contrast agent
registry. Radiology 1997; 203:611.

45. Ansell G. Adverse reactions to contrast agents. Scope of problem.
Invest Radiol 1990; 25:381.

46. Katayama H, Yamaguchi K, Kozuka T et ah Adverse reactions to
ionic and nonionic contrast media. A report from the Japanese
Committee on the Safety of Contrast Media. Radiology 1990;
175:621.

47. Fanning NF, Manning BJ, Buckley J, Redmond HP. Iodinated
contrast media induce neutrophil apoptosis through a mitochon-
drial and caspase mediated pathway. Br J Radiol 2002; 75:861.

48. Schick CS, Bangert R, Kubler W, Haller C. Ionic radiocontrast
media disrupt intercellular contacts via an extracellular calcium-
independent mechanism. Exp Nephrol 2002; 10:209.

49. Lasser EC, Berry CC, Tainer LB et ah Pretreatment with cortico-
steroids to alleviate reactions to intravenous contrast material. N
Engl J Med 1987; 317:845.

50. McClennan BL. Adverse reactions to iodinated contrast
media. Recognition and response. Invest Radiol 1994; 29 (Suppl
1):S46.

51. Yoshikawa H. Late adverse reactions to nonionic contrast media.
Radiology 1992; 183:737.

52. McCullough M, Davies P, Richardson R. A large trial of intra-
venous Conray 325 and Niopam 300 to assess immediate and de-
layed reactions. Br J Radiol 1989; 62:260.

53. Panto PN, Davies P. Delayed reactions to urographic contrast
media. Br J Radiol 1984; 59:41.

54. Yasuda R, Munechika H. Delayed adverse reactions to
nonionic monomeric contrast-enhanced media. Invest Radiol
1998; 33:1.

55. Webb JA, Stacul F, Thomsen HS, Morcos SK. Late adverse reac-
tions to intravascular iodinated contrast media. Eur Radiol 2003;
13:181.

56. Pedersen SH, Svaland MG, Reiss AL, Andrew E. Late allergy-like
reactions following vascular administration of radiography con-
trast media. Acta Radiol 1998; 39:344.

57. Berkseth RO, Kjellstrand CM. Radiologic contrast induced
nephropathy. Med Clin North Am 1984; 68:351.

58. Martin-Paredero V, Dixon SM, Baker JD et ah Risk of renal failure
after major angiography. Arch Surg. 1983; 118:1417.

59. Dawson P, Trewhella M. Intravascular contrast agents and renal
failure. Clin Radiol 1990; 41:373.

60. Schwab SJ, Hlatky MA, Pieper KS et ah Contrast nephrotoxicity: a
randomized control trial of a nonionic and an ionic radiographic
contrast agent. N Engl J Med 1989; 320:149.

61. Porter GA. Contrast-associated nephropathy: presentation,
pathophysiology, and management. Miner Electrolyte Metab
1994; 20:232.

62. Lautin EM, Freeman NJ, Schoenfeld AH et ah Radiocontrast-
associated renal dysfunction: a comparison of lower-osmolality
and conventional high-osmolality contrast media. Am J
Roentgenol 1991; 157:59.

63. Barrett BJ, Carlisle EJ. Meta-analysis of the relative nephro-
toxicity of high- and low-osmolality iodinated contrast media.
Radiology 1993; 188:171.

64. Harris KG, Smith TP, Cragg AH, Lemke JH. Nephrotoxicity from
contrast material in renal insufficiency: ionic versus nonionic
agents. Radiology 1991; 179:849.

65. Barrett BJ, Parfrey PS, McDonald JR et ah Nonionic low-
osmolality versus ionic high-osmolality contrast material for
intravenous use in patients perceived to be at high risk:
randomized trial. Radiology 1992; 183:5.

66. Eisenberg RL, Bank WO, Hedgock MW. Renal failure after major
angiography can be avoided with hydration. Am J Roentgenol
1981; 136:859.

67. Bettmann MA. The evaluation of contrast-related renal failure.
Am J Roentgenol 1991; 157:66.

68. Morcos SK, Thomsen HS. Radiology guidelines on administer-
ing contrast media. Abdom Imaging 2003; 28:187.

69. Cloft HJ, Jensen ME, Kallmes DF, Dion JE. Arterial dissections
complicating cerebral angiography and cerebrovascular inter-
ventions. AJNR 2000; 21:541.

70. Olivecrona H. Complications of cerebral angiography. Neuro-
radiology 1977; 14:175.

71. Heiserman JE, Dean BL, Hodak JA et ah Neurologic complica-
tions of cerebral angiography. AJNR 1994; 15:1401.

72. Hessel HJ, Adams DF, Abrams HL. Complications of angio-
graphy. Radiology 1981; 138:273.

73. Feild JR, Robertson JT, Desaussure RL Jr. Complications of
cerebral angiography in 2,000 consecutive cases. / Neurosurg
1962; 19:775.

74. Latchaw RE. Guidelines for diagnostic neuroangiography: a
model to emulate from a neuroradiologist's perspective. AJNR
2000; 21:44.

75. Hellmann DB, Roubenoff R, Healy RA, Wang H. Central nervous
system angiography: safety and predictors of a positive result in
125 consecutive patients evaluated for possible vasculitis. /
Rheumatol 1992; 19:568.

76. Leow K, Murie JA. Cerebral angiography for cerebrovascular
disease: the risks. Br J Surg 1988; 75:428.

77. Jackson A, Stewart G, Wood A, Gillespie JE. Transient global am-
nesia and cortical blindness after vertebral angiography: further

398

CHAPTER 34 Angiography

evidence for the role of arterial spasm. AJNR 1995; 16 (4
Suppl.):955.

78. Wishart DL. Complications in vertebral angiography as com-
pared to non-vertebral cerebral angiography in 447 studies. Am }
Roentgenol 1971; 113:527.

79. Di Chiro G. Unintentional spinal cord arteriography: a warning.
Radiology 1974; 112:231.

80. Blumgart RL, Immelman EJ, Jeffery PC, Lipinski JK. Thrombotic
side-effects of lower limb venography. The use of heparin-saline
flush S Afr Med J 1991; 79:88.

81. Albrechtsson U, Olsson CG. Thrombosis following phlebo-
graphy with ionic and non-ionic contrast media. Acta Radiol
Diagn 1979; 20:46.

82. Dorfman GS, Froehlich JA. The appropriate use of contrast
venography and pulmonary angiography in the orthopedic
population. Semin Arthroplasty 1992; 3:72.

83. Hudson ER, Smith TP, McDermott VG et ah Pulmonary angio-
graphy performed with iopamidol: complications in 1,434
patients. Radiology 1996; 198:61.

84. Fisher RG, Chasen MH, Lamki N. Diagnosis of injuries of the
aorta and brachiocephalic arteries caused by blunt chest trauma:
CT vs. aortography. Am] Roentgenol 1994; 162:1047.

85. Gavant ML, Flick P, Menke P, Gold RE. CT aortography of
thoracic aortic rupture. Am] Roentgenol 1996; 166:955.

86. Nunez D Jr, Rivas L, McKenney K, LeBlang S, Zuluaga A. Helical
CTof traumatic arterial injuries. Am] Roentgenol 1998; 170:1621.

87. Semba CP, Kato N, Kee ST et ah Acute rupture of the descending
thoracic aorta: repair with use of endovascular stent-grafts. /
Vase Interv Radiol 1997; 8:337.

88. Morgan PW, Goodman LR, Aprahamian C, Foley WD, Lipchik
EO. Evaluation of traumatic aortic injury: does dynamic
contrast-enhanced CT play a role? Radiology 1992; 182:661.

89. Dake MD, Kato N, Mitchell RS et ah Endovascular stent-graft
placement for the treatment of acute aortic dissection. N Engl J
Med 1999; 340:1546.

90. Macura KJ, Szarf G, Fishman EK, Bluemke DA. Role of computed
tomography and magnetic resonance imaging in assessment of
acute aortic syndromes. Semin Ultrasound CTMR 2003; 24:232.

91. Khan IA, Nair CK. Clinical, diagnostic, and management per-
spectives of aortic dissection. Chest 2002; 122:311.

92. Link KM, Lesko NM. The role of MR imaging in the evaluation of
acquired diseases of the thoracic aorta. Am J Roentgenol 1992;
158:1115.

93. Matsunaga N, Hayashi K, Okada M, Sakamoto I. Magnetic reso-
nance imaging features of aortic diseases. Top Magn Reson Imag-
ing 2003; 14:253.

94. Randoux B, Marro B, Koskas F, Chiras J, Dormont D, Marsault C.
Proximal great vessels of aortic arch: comparison of three-
dimensional gadolinium-enhanced MR angiography and digital
subtraction angiography. Radiology 2003; 229:697.

95. Kunzli A, von Segesser LK, Vogt PR et ah Inflammatory
aneurysm of the ascending aorta. Ann Thorac Surg 1998; 65:1132.

96. Wallis F, Roditi GH, Redpath TW, Weir J, Cross KS, Smith FW.
Inflammatory abdominal aortic aneurysms: diagnosis with
gadolinium enhanced Tl-weighted imaging. Clin Radiol 2000;
55:136.

97. Young TB, Paty J Jr, Panda M, Enzenauer RJ. Takayasu's arteritis:
isolated aortitis. / Rheumatol 2003; 30:2508.

98. Hata D. Fibromuscular dysplasia. Intern Med 2001; 40:978.

99. Surowiec SM, Sivamurthy N, Rhodes JM et ah Percutaneous
therapy for renal artery fibromuscular dysplasia. Ann Vase Surg
2003; 17:650.

100. Rooke TW, Joyce JW. Uncommon arteriopathies. In: Rutherford
RB, ed. Vascular Surgery, 5th edn. Philadelphia: WB Saunders,
2000:418.

101. Kransdorf MJ, Turner-Stepahin S, Merritt WH. Magnetic reso-
nance angiography of the hand and wrist: evaluation of patients
with severe ischemic disease. J Reconstr Microsurg 1998; 14:77.

102. Mills JL Sr. Buerger's disease in the 21st century: diagnosis, clini-
cal features, and therapy. Semin Vase Surg 2003; 16:179.

103. Lawler LP, Fishman EK. Multidetector row computed tomo-
graphy of the aorta and peripheral arteries. Cardiol Clin 2003;
21:607.

104. Akers DL Jr, Fowl RJ, Kempczinski RE Mycotic aneurysm of the
tibioperoneal trunk: case report and review of the literature. /
Vase Surg 1992; 16:71.

105. Macedo TA, Johnson CM, Hallett JW Jr, Breen JF. Popliteal artery
entrapment syndrome: role of imaging in the diagnosis. Am J
Roentgenol 2003; 181:1259.

106. Elias DA, White LM, Rubenstein JD, Christakis M, Merchant N.
Clinical evaluation and MR imaging features of popliteal artery
entrapment and cystic adventitial disease. Am } Roentgenol 2003;
180:627.

107. Turnipseed WD. Popliteal entrapment syndrome. / Vase Surg
2002; 35:910.

108. Elias DA, White LM, Rubenstein JD, Christakis M, Merchant N.
Clinical evaluation and MR imaging features of popliteal artery
entrapment and cystic adventitial disease. Am } Roentgenol 2003;
180:627.

109. Mellado JM, Salvado E. Adventitial cystic disease of the popliteal
artery: role of MRA. Eur Radiol 2002; 12:948.

110. Ruckert RI, Taupitz M. Cystic adventitial disease of the popliteal
artery. Am J Surg 2000; 180:53.

111. Gerasimidis T, Sfyroeras G, Papazoglou K, Trellopoulos G,
Ntinas A, Karamanos D. Endovascular treatment of popliteal
artery aneurysms. Eur J Vase Endovasc Surg 2003; 26:506.

112. Kopecky KK, Stine SB, Dalsing MC, Gottlieb K. Median arcuate
ligament syndrome with multivessel involvement: diagnosis
with spiral CT angiography. Abdom Imaging 1997; 22:318.

113. Shih A, Golden M, Mohler ER 3rd. Splenic artery aneurysm. Vase
Med 2002; 7:155.

114. Matsumura JS, Yao JS, Nemcek AA Jr. Helical CT angiography of
thoracic outlet syndrome. Am J Roentgenol 2001; 177:714.

115. Demondion X, Bacqueville E, Paul C, Duquesnoy B, Hachulla E,
Cotten A. Thoracic outlet: assessment with MR imaging in
asymptomatic and symptomatic populations. Radiology 2003;
227:461.

116. Hagspiel KD, Spinosa DJ, Angle JF, Matsumoto AH. Diagnosis
of vascular compression at the thoracic outlet using
gadolinium-enhanced high-resolution ultrafast MR angio-
graphy in abduction and adduction. Cardiovasc Intervent Radiol
2000; 23:152.

117. Dymarkowski S, Bosmans H, Marchal G, Bogaert J. Three-
dimensional MR angiography in the evaluation of thoracic
outlet syndrome. Am] Roentgenol 1999; 173:1005.

118. Westerband A, Rodriguez JA, Ramaiah VG, Diethrich EB. En-
dovascular therapy in prevention and management of coronary-
subclavian steal. / Vase Surg 2003; 38:699.

399

pa rt 1 1 1 Invasive vascular diagnostics

119. Ruehm SG, Weishaupt D, Debatin JR Contrast-enhanced MR
angiography in patients with aortic occlusion (Leriche syn-
drome) . / Magn Reson Imaging 2000; 11 :401 .

120. Kreitner KF, Kalden P, Neufang A et ah Diabetes and peri-
pheral arterial occlusive disease: prospective comparison of
contrast-enhanced three-dimensional MR angiography with
conventional digital subtraction angiography. Am J Roentgenol
2000; 174:171.

121. Herborn CU, Goyen M, Lauenstein TC, Debatin JF, Ruehm SG,
Kroger K. Comprehensive time-resolved MRI of peripheral vas-
cular malformations. Am } Roentgenol 2003; 181:729.

122. Simons ME. Peripheral vascular malformations: diagnosis
and percutaneous management. Can Assoc Radiol J 2001; 52:
242.

123. Baxter BT, Blackburn D, Payne K, Pearce WH, Yao JS. Noninva-
sive evaluation of the upper extremity. Surg Clin North Am 1990;
70:87.

124. Van Grimberge F, Dymarkowski S, Budts W, Bogaert J. Role of
magnetic resonance in the diagnosis of subclavian steal syn-
drome. J Magn Reson Imaging 2000; 12:339.

125. Ratanakorn D, Laothamatas J, Pongpech S, Tirapanich W,
Yamwong S. Pitfall of electron beam computed tomography
angiography in diagnosis of subclavian steal syndrome. /
Neuroimaging 2002; 12:80.

126. Bunce NH, Davies S, Mohiaddin RH. Magnetic resonance of ver-
tebral steal syndrome. Heart 2001; 85:638.

127. Athanasoulis CA. Vascular radiology: looking into the past to
learn about the future. Radiology 2001; 218:317.

128. Theodorou SJ, Theodorou DJ, Kakitsubata Y Sonography
and venography of the lower extremities for diagnosing
deep vein thrombosis in symptomatic patients. Clin Imaging
2003; 27:180.

129. Gallix BP, Achard-Lichere C, Dauzat M, Bruel JM. Flow-
independent magnetic resonance venography of the calf. J Magn
Reson Imaging2003; 17:421.

130. Schwab SJ, Oliver MJ, Suhocki P, McCann R. Hemodialysis arte-
riovenous access: detection of stenosis and response to treatment
by vascular access blood flow. Kidney Int 2001; 59:358.

131. Ahmed HK, Hagspiel KD. Intravascular ultrasonographic find-
ings in May-Thurner syndrome (iliac vein compression syn-
drome)./ Ultrasound Med 2001; 20:251.

132. Baron HC, Shams J, Wayne M. Iliac vein compression syndrome:
a new method of treatment. Am Surg 2000; 66:653.

133. Wolpert LM, Rahmani O, Stein B et ah Magnetic resonance veno-
graphy in the diagnosis and management of May-Thurner
syndrome. Vase Endovascular Surg 2002; 36:51.

134. Lamont JP, Pearl GJ, Patetsios P et ah Prospective evaluation of
endoluminal venous stents in the treatment of the May-Thurner
syndrome. Ann Vase Surg 2002; 16:61.

135. Lee AY, Ginsberg JS. Venous thrombosis of the upper extremities.
Curr Treat Options Cardiovasc Med. 2001; 3:207.

136. Dinkel HP, Mettke B, Schmid F, Baumgartner I, Triller J, Do DD.
Endovascular treatment of malignant superior vena cava syn-
drome: is bilateral wallstent placement superior to unilateral
placement? JEndovasc Ther 2003; 10:788.

137. van Gelder JM. Computed tomographic angiography for detect-
ing cerebral aneurysms: implications of aneurysm size dis-
tribution for the sensitivity, specificity, and likelihood ratios.
Neurosurgery 2003; 53:597.

138. Dehdashti AR, Rufenacht DA, Delavelle J, Reverdin A, de
Tribolet N. Therapeutic decision and management of aneurys-
mal subarachnoid haemorrhage based on computed tomo-
graphic angiography. Br J Neurosurg 2003; 17:46.

139. Ikawa F, Uozumi T, Kiya K et ah Diagnosis of carotid-cavernous
fistulas with magnetic resonance angiography: demonstrating
the draining veins utilizing 3-D time of flight and 3-D phase con-
trast techniques. Neurosurg Rev 1996; 19:7.

140. Cornelius RS. CCF: imaging evaluation. In: Tomsick TA, ed.
Carotid-Cavernous Fistula. Cincinnati, OH: Digital Educational
Publishing, 1997:23.

141 . Rodesch G, Lasjaunias P. Spinal cord arteriovenous shunts: from
imaging to management. Eur J Radiol 2003; 46:221.

142. Farb RI, Kim JK, Willinsky RA et ah Spinal dural arteriovenous
fistula localization with a technique of first-pass gadolinium-
enhanced MR angiography: initial experience. Radiology 2002;
222:843.

143. Oldfield EH, Bennett A 3rd, Chen MY, Doppman JL. Successful
management of spinal dural arteriovenous fistulas undetected
by arteriography. Report of three cases. / Neurosurg. 2002; 96b (2
Suppl.):220.

144. Andersson T, van Dijk JM, Willinsky RA. Venous manifestations
of spinal arteriovenous fistulas. Neuroimaging Clin N Am 2003;
13:73.

145. Lee DK, Choe WJ, Chung CK, Kim HJ. Spinal cord heman-
gioblastoma: surgical strategy and clinical outcome. JNeurooncol
2003; 61:27.

400

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Intravascular ultrasound

James T. Lee
George Kopchok
Rodney A. White

In the last half century, ultrasonographic imaging has flour-
ished and become an essential component and in most cases
the initial diagnostic approach preferred worldwide by most
clinicians. 1 The utility of this diagnostic principle continues to
expand and impact visceral imaging especially in the exami-
nation of the peripheral vasculature. Although external beam
imaging (B-mode, duplex, color flow) provides an excellent
anatomic depiction of the vascular system, frequent limita-
tions include decreased resolution of arterial wall elements
in deep vascular structures and overlying scatter produced
by surrounding soft tissues and bone causing a disruption
in representation continuity. The ambition of our con-
temporary society to surmount these impediments has influ-
enced an alternative approach to transcutaneous imaging.
The evolution of endovascular interventions and instrumen-
tation perpetuates this ethos, and through a wealth of
technological enhancements, now permits intravascular
sonographic interrogation of the peripheral and coronary ves-
sels. By combining a miniaturized transducer with an intralu-
minal catheter, precise anatomic information can be obtained
before, during, and after therapeutic intervention from within
the vessel lumen.

The diagnostic advantages of intravascular ultrasound
(IVUS) for examining arterial wall architecture and lesion
morphology are evident. Rapidly accumulating clinical expe-
rience has shown IVUS to accurately determine lesion shape,
length, and configuration, as well as to identify and examine
the origins of branch vessels and tributaries in preparation for
endoluminal intervention, either as a primary source or as a
complementary adjunct to current imaging modalities. 2-8 This
chapter describes the principles of IVUS imaging technology
in modern available designs. The techniques for use including
catheter delivery, manipulation, and methods to enhance
image interpretation are also reviewed. A discussion of clinical
applications emphasizing the role of IVUS in peripheral inter-
ventions is then presented in detail.

Principles of IVUS

Imaging systems

Extrapolating from the principles of conventional ultrasound,
IVUS instrumentation consists of mechanical transducers
with fixed or rotating elements, or electronically switched ar-
rays (phased). By scanning the ultrasound beam in a full circle
and synchronizing the beam direction and deflection with the
display, a 360° cross-sectional image is obtained (Fig. 35.1).

Phased array transducers

Electronically switched arrays consist of elements arranged
circumferentially along a catheter tip combined with a minia-
ture integrated circuit. This design provides sequenced trans-
mission and reception without requiring numerous electrical
circuits traveling the full length of the shaft thereby improving
catheter flexibility. Current multielement devices have fre-
quencies in the range of 15-25 MHz. Similar to mechanical
transducers, a 360° circumferential real-time intraluminal
image is produced perpendicular to the long axis of the
catheter.

Although excellent flexibility and resolution are advanta-
geous, a common limitation, as with all high-frequency
catheters (to some extent), is the inability to image structures
immediately adjacent to the transducer (termed "in the near
field"). Within this "near field," a bright circumferential arti-
fact known as the "ring down or halo" surrounds the catheter.
This occurs primarily due to the short distance between the
structure imaged and the imaging crystals in a phased array
configuration. Although this effect is also seen with transcuta-
neous ultrasound, it is of minimal significance since the struc-
tures of interest are often at some distance from the transducer
assembly. However, in devices with higher frequencies such as
IVUS, the "ring-down" artifact can adversely affect accurate
data acquisition due to the close proximity of the imaging
component to the target structure. This occurs when the

401

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.1 (A) Mechanical transducer with rotating (1) or fixed elements
(2). Either the crystal or the mirror may be fixed with the other in a rotating
position. (B) Electronically switched or phased array with transducer
elements arranged circumferentially.

Figure 35.2 (A) Rotating element device (left) where the ultrasound beam
is directed forward or perpendicular with respect to the catheter long axis,
producing a cone-shaped imaging plane (right). (B) Rotating mirror
configuration (left) where the ultrasound beam is directed at a 90° angle to
the catheter axis, creating a true cross-sectional imaging plane (right).

catheter tip approaches or comes into direct contact with the
vessel wall. The "halo or ring-down" effect can result in ob-
scured translation of transmural echoes. However, if the
catheter is maintained within the central lumen of the blood
vessel, only reflected signals from flowing blood elements im-
mediately surrounding the imaging assembly will be lost.
Current software can electronically remove the ring-down ar-
tifact; however, this results in suppression of all anatomic
structures within the masked region.

Mechanical transducers

Mechanical transducers consist of two basic configurations.
Either the transducer itself, or an acoustic mirror located at the
catheter tip, is rotated using a flexible, high-torque cable that
extends the length of the device. Catheters with a rotating
transducer that is angled slightly forward to the perpendicular
plane project a cone-shaped beam, creating an image of the
vessel slightly in front of the transducer assembly (Fig. 35. 2 A).
In devices manufactured with a rotating acoustic reflector, the
mirror is placed a short distance proximally and set at a 45°
angle to the rotating shaft. Beam emission then creates a cir-
cumferential image exactly perpendicular to the axis of the
catheter (Fig. 35.2B). Ultrasound frequencies vary between 10
and 30 MHz, although catheters can produce excellent images
with frequencies up to 45 MHz. 9

Strengths and limitations of differing
catheter configurations

Catheters manufactured with a rotating reflector have the ad-
vantage of a fixed transducer placed at a set distance from the

mirror within the imaging chamber. Emitted ultrasonic pulses
then travel the length of this chamber through echolucent
saline media. The scan converter in the central processing
unit will then compensate for this echoless area in the near
field of the beam and generate images precisely at the surface
of the catheter. This configuration partially eliminates the
ring-down artifact, and can attenuate the poor near field reso-
lution of the scan. The main disadvantage of this system is the
frequent need for low-pressure irrigation of the imaging
chamber. In order to prevent image distortion caused by
air/ fluid levels, intermittent manual flushing with saline is
required.

In catheters with rotating transducers or phased arrays, a
portion of the halo artifact in the near field zone of the energy
emission occurs outside the catheter resulting in poor image
capture in this area. Despite these limitations, mechanical
transducers overall have less image distortion in comparison
with phased arrays. 3

Other instrument designs feature a distally placed trans-
ducer with a proximal rotating mirror. This necessitates the
use of a communicating electrical wire that results in an arti-
fact occupying 15° of the image cross-section. Although a por-
tion of the circumference is lost, the wire artifact may be used
to facilitate rotational alignment of the catheter within the
vessel lumen. 10 This artifact which may be inconsequential
in two-dimensional (2-D) IVUS can significantly curtail the
precision of three-dimensional (3-D) reconstruction. In addi-
tion, it is also more noticeable in smaller, higher frequency
catheters. A recent modification to the mechanical transducer
design involves rotation of both the transducer and mirror,
which removes the wire artifact. However, as there is no
central lumen, catheter delivery and image acquisition

402

chapter 35 Intravascular ultrasound

must occur through an introducer sheath limiting catheter
flexibility.

Technological miniaturization of mechanical systems is an-
other major limitation that may ultimately differentiate these
devices from phase array catheters in regard to their useful-
ness in smaller vessels. On the other hand, the problems of
ring-down artifact and near field imaging also become more
apparent in progressively diminutive vessels. The smallest
mechanical transducers available approximate the luminal
diameters of the coronary arteries (2.9 Fr or 9 mm), and
when used with higher frequency ultrasound, prove to be
superior in most applications.

Catheter techniques

Access and preparation

IVUS catheters can be introduced either percutaneously or
through an open surgical technique. Regardless of the expo-
sure, access is obtained through a standard introducer sheath
(5-1 OFr). In most situations a retrograde femoral puncture al-
lows IVUS interrogation of the aortoiliac system with excellent
control when passed over a guidewire (0.009-0.038 in in dia-
meter). Phased array catheters are constructed with a central
guidewire channel while most mechanical transducers are
modified with a monorail or side-saddle coaxial lumen for
over-the-wire applications. Since catheter lengths vary from
90 cm to 140 cm and are flexible, it is possible to image the con-
tralateral iliofemoral system by crossing the bifurcation. In
this setting, both rotational orientation and catheter tracking
may be difficult (to a greater extent in coaxial systems). In
larger arteries and veins, the location of the guidewire is of
little import, but in the more distal vasculature or after inter-
vention (which may result in irregularities in the vessel wall), a
centrally placed guidewire may be advantageous in tracking
the catheter through the diseased segment.

Antegrade femoral artery puncture provides access to the
infrainguinal vessels and facilitates planned endovascular in-
tervention. In other instances, a retrograde popliteal approach
may be necessary to survey disease of a patent distal superfi-
cial femoral artery (SFA) when there is an occlusion at the ori-
gin. Upper extremity sites such as the brachial or axillary
artery are not often utilized but may provide added informa-
tion when imaging the thoracic aorta.

Catheter delivery and imaging techniques

Catheter fragility is a major concern when navigating through
tortuous and stenotic vessels. In monorail systems, simultan-
eous advancement of both catheter and guidewire maybe nec-
essary to expedite passage. The central lumen of electronic
arrays circumvents this issue and allows for better tracking
and is more kink resistant.

Once adequate vessel introduction is accomplished, accu-
rate anteroposterior orientation is examined. The most suc-
cessful methods of maintaining rotational alignment are use
of the image interference artifact produced by the connecting
transducer wires, and establishing correct initial positioning
during catheter insertion. 2 ' 6 ' 11 ' 12 When imaging the aortoiliac
segments, rotational accuracy can be confirmed by the relative
position of constant anatomic landmarks. Once the catheter
has reached the aortic bifurcation, both common iliac arteries
are visualized directly adjacent to one another in the horizon-
tal plane. Occasionally, this anatomic arrangement is skewed,
especially with tortuous dilated vessels. Visualizing other
adjunctive parameters then attains the proper orientation.
Often a combination of the known position of the catheter
at insertion, and the posteromedial location of the ipsilateral
hypogastric artery orifice, provides the best possible points
of reference (Fig. 35.3). 11-15 The origins of the splanchnic
vessels (i.e. superior mesenteric, celiac, and renal arteries
and left renal vein) are also important fixation points
(Fig. 35.4). Because the catheters are easily torqued, there is
very little loss of orientation with rotation and manipulation
during imaging.

Careful positioning of the imaging tip within the lumen and
appropriate size-matching of the device to the artery caliber
are essential to optimize visualization. Image quality is best
when the catheter is parallel to the vessel axis, with the ultra-
sound beam directed perpendicular to the intimal surface
of the vessel. This is best accomplished by obtaining cross-
sectional images and measurements during graded catheter
withdrawal rather than during advancement. The slight ten-
sion during manual pullback adds columnar support and
straightens the catheter while centering the tip. Manipulation
through difficult and meandering peripheral vessels, how-
ever, can obscure image interpretation, as the catheter tends to
engage in a more direct course through the lumen (Fig. 35.5).
This results in an eccentric intraluminal position that gener-
ates an artifactual difference in wall thickness between the ad-
jacent and contralateral walls. The endothelium closest to the
imaging chamber will appear as a hyperechoic image and
be misconstrued to have an increased density. Although this
rarely affects 2-D imaging, the results of 3-D reconstruction
may be severely altered. In a peripheral vessel such as the
femoral artery, this off-center position can be compensated by
external manual compression of the artery, which produces
excellent centering of the transducer.

The acoustic interface between the lumen and the intima can
be obscured by several factors, which include the echogenic
properties of flowing blood in small-caliber vessels, the close
proximity of the IVUS to the vessel wall, and the diminished
depth penetration by a higher frequency catheter. 8 Adjusting
the overall gain and use of imaging processing techniques
such as time gain compression and suppression, may only par-
tially compensate for the "ring-down or halo" effect. Flushing
the vessel lumen with saline solution or sonographic contrast

403

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.3 IVUS orientation is achieved
through visualization of relatively fixed
anatomic landmarks. As the catheter crosses
the aortic bifurcation, the iliac arteries are often
positioned side by side (top right) as seen on the
manual pullback. Catheteralignment may also
be maintained by identifying the posteromedial
origin of the ipsilateral hypogastric vessel
(bottom right).

Figure 35.4 The relative position of constant anatomic landmarks confirms
accurate rotational orientation. The left renal vein (solid arrow) crosses
anterior to the aorta and the origin of the right renal artery (open arrow).

agents may result in interface enhancement by displacing
erythrocyte elements with an echolucent area that greatly
facilitates identification of mural pathology (i.e. thrombus,
dissections, ulcerated plaques, etc.). 13 ' 16 This is an especially
useful technique when interrogating small low-flow caliber
vessels. Infusion with radiographic contrast, however, ren-
ders the vessel lumen hyperechoic and can serve to enhance
the acoustic /intimal interface from a different perspective.

Angular and radial position uncertainties create image arti-
facts observed in both rotating mirror and transducer designs.
A distorted image is displayed when the angulation of the
emitted beam fails to correspond accurately with the resultant
angular position on the screen. This is due to both the friction
of the drive shaft within the external catheter sheath and
limited torsional rigidity of the drive shaft. The process that
mistakenly reduces the radial extent of disease on the image
circumference is known as compression, and the opposite ef-
fect is known as expansion. 17 Radial position uncertainty is
produced by repetitive lateral movement of the catheter tip.
This is often undetectable unless there is a faulty rotating drive
shaft within the catheter or an unsatisfactory connection with
the motor drive exists. The effects are minor image distortion
and a reduction in radial resolution during real-time imaging.
In severely angulated anatomy, diameter measurements are
calculated by the shortest distance between two opposing par-
allel walls (Fig. 35.6).

404

chapter 35 Intravascular ultrasound

Figure 35.5 Longitudinal view of IVUS
catheter within a tortuous vessel. The tip of the
catheter lies within the lumen in the same plane
as the shaft, resulting in images of varying
eccentricity. During graded withdrawal of the
catheter, at certain positions the transducer
assemblywill lie against the vessel wall (solid
arrows); at other times, it will be centered within
the lumen (open arrow).

Figure 35.6 Diameter measurements within
tortuous vessels are obtained by calculating the
shortest distance between opposing walls.

Image intepretation

The images produced by IVUS catheters not only outline the
luminal and adventitial surfaces of normal arterial segments
but also have the potential to discriminate between normal
and diseased vessel wall. In healthy medium-sized muscular
arteries, three distinct sonographic layers are visible. The
media appears as an echolucent thin smooth muscular layer in
between the more hyperechoic intima and adventia (Fig. 35.7).
This is directly attributable to the denser elastin component in
the intima, and the increased collagen content in the adventi-
tia. Precise correlation between the ultrasound image and the
histologic anatomy of the muscular artery wall is still uncer-
tain, although the internal and external elastic laminae and
adventitia are considered to be the backscatter substrates for
the inner and outer echodense zones. 18,19 In smaller vessels the
adventitia may be less well defined unless the vessel is sur-
rounded by tissues of differing echogenicity (i.e. subcuta-
neous adipose tissue which is echolucent) . Also, in larger more
central vessels, the three-layer image seen in medium-sized

arteries may be lost due to the increased hyperechoic elastin
content within the media (Fig. 35.8).

The absorption coefficient for reflected ultrasound energy
varies dependent on structural thickness. In dense tissue (i.e.
bone), the absorption coefficient is proportional to the square
of the frequency, whereas for softer tissue it is proportional to
the frequency. 20 IVUS catheters are therefore sensitive in dif-
ferentiating between calcified and noncalcified vascular le-
sions. Because predominantly calcific plaque strongly reflects
ultrasound energy, it appears as a bright image with dense
acoustic shadowing behind it (Fig. 35.9). For this reason, the
exact location of the media and adventitia cannot be seen in
segments of vessels containing heavily calcified disease, and
dimensions must be estimated by interpolation of adjacent
size data. The accuracy of IVUS in determining luminal di-
mensions and wall thickness for normal to minimally diseased
arteries was observed to be within 0.05 mm in several in-vitro
and in-vivo studies. 4 ' 18-27 When measuring the outer wall
diameters, however, the margin of error is increased up to
0.5 mm.

405

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.7 IVUS cross-sectional viewwith
2-D manual pullbackof thefemoral artery.
Distinct sonographic layers are visible in
muscular arteries, with the media appearing as
an echolucent area (open arrows) in between
the more hyperechoic intima and adventitia.

Figure 35.8 In more centrally located larger caliber vessels (proximal
common iliac artery), the increased elastin content results in a loss of the
three-layer image seen in more muscular arteries.

In addition to providing accurate dimensional morphology,
IVUS can differentiate plaque from thrombus and determine
the consistency of lesions and degree of calcification present.
Gussenhoven and associates have described four basic plaque
components that can be distinguished using 40-MHz probes in
vitro: echolucent, soft echoes, bright echoes, and bright echoes
with acoustic shadowing. 18 ' 19 Echolucent images are pri-
marily from significant lipid deposits or 'lipid lakes/' while

soft echoes suggest the presence of fibromuscular tissue or in-
timal proliferation with varying amounts of dispersed lipid.
Bright echoes represent collagen-rich fibrous tissue, and
bright echoes with acoustic shadowing beyond the lesion sig-
nify the presence of calcifications.

Three-dimensional reconstruction

The development of 3-D IVUS image reconstruction has of-
fered a unique method for analyzing normal and diseased vas-
cular segments. 28 Computerized 3-D vessel representation
involves either surface- or volume-rendering algorithms.

In surface rendering, object surfaces are formed before the
creation of the image on a 2-D screen using methods such as
hidden-part removal, shading, dynamic rotation, and stereo-
projection. 29 ' 30 Based on the object's exterior, two types of sur-
face-rendering techniques can be identified: mosaic (triangular
shaped patches), or discrete boundary surface rendering. 31

Volume rendering is based on the generation of interfaces
and pseudosurfaces through digital voxel projection resulting
in image perspective, contour, and shape. 32 This technique
requires manipulation of large data volumes and differs
from surface rendering in that object surfaces are not explicitly
computed. 30

The initial step in image processing produces a real-time,
gray-scale, longitudinal section of the vessel that can be
rotated 360° to facilitate interrogation of specific sites. The
aligned set (up to 450 images per set) of consecutive 2-D IVUS
images is assembled in sequence along an axis plane to pro-
duce the 3-D image. 33 The final reconstruction is dependent on
the quality of these original images. The technique of data

406

chapter 35 Intravascular ultrasound

Figure 35.9 IVUS images of atherosclerotic
common (left) and external iliac (right) arteries.
A heavily calcified intimal surface on a large
complex plaque (arrow) produces a bright
luminal line with dense acoustic shadowing
behind it. U, ultrasound cathetervoid.

Figure 35.10 Complex IVUS image reconstruction. Two-dimensional cross-
sectional images are acquired along the length of the vessel (center) and are
"stacked" by an intricate computerized algorithm to create a longitudinal
gray-scale (left) and three-dimensional (right) image. (Sites A, B, and C on the

center images correspond to lines A, B, and C in the peripheral images.)
(From Cavaye DM, Tabbara MR, Kopchok GE etal. Three dimensional
vascular ultrasound imaging. Am Surg 1991; 57:751 .)

acquisition is therefore of paramount importance. The images
should be acquired by slowly withdrawing the IVUS catheter
through the vessel and sampling the cross-sectional images at
a defined rate. Carefully withdrawing the catheter, either
manually or using a mechanical device at a uniform rate
of 1 cm every 4 s, has been found to acquire the optimal raw
images. 34 Thus, a 20-s pullback is needed to collect data from
a 5-cm vessel segment. The total number of frames per
reconstruction is further determined by a combination of

withdrawal speed, total pullback time, and digital sampling
rate. A slower catheter removal rate will result in greater
anatomic detail. The pullback can then be reconstructed with
frame sampling rates up to 30 frames per second.

Three-dimensional reconstructions can be displayed as
either a complete vessel cylinder or as a luminal cast by
removing vessel wall signals, and using the lumen/intima in-
terface as the only connected surface (Fig. 35.10). The luminal
profile can then be examined on-screen and manipulated in

407

pa rt 1 1 1 Invasive vascular diagnostics

multiple orientations to allow inspection of the vessel segment
in an almost limitless number of projections, both from within
the lumen and from the adventitial surface. Other parameters
such as image sharpness, contrast, and ambient light can also
be altered to improve the resolution of particular features
being examined.

At present, the longitudinal view provides more clinically
useful information than volume representation. 34 The major
impediments to 3-D IVUS include imprecise spatial orienta-
tion within tortuous vessels, especially if the catheter does not
remain centrally positioned, and the inherent loss of gray-
scale with volume reconstruction. True gray-scale is lost when
projection of a 3-D image on a 2-D monitor results in graded
shading relative to perceptual distance. Distant parts of the
image appear with increased contrast as opposed to more ad-
jacent elements, which are a lighter gray. Thus, near field blood
components may be portrayed with the same echogenicity as
distant arterial wall. This will prevent visualization of the ves-
sel wall as a separate structure. Adjusting the time gain com-
pensation and energy output of the 2-D IVUS catheter may
partially reduce the blood artifact although further improve-
ments in imaging processing are needed. 34 As new catheter
modifications are implemented, gray-scale 3-D images, trans-
parent volume-rendered images, and the addition of color-
coded blood flow data may become available.

Clinical applications

The diagnostic applications, the enhancement of therapeutic
capabilities, and the direction of appropriate therapy utilizing
IVUS continue to evolve. By providing detailed information
on luminal morphology and characterization of vascular le-
sions before, during, and after intervention, IVUS provides a
method for both guidance of endoluminal devices and imme-
diate assessment of the results of the procedure (Table 35.1).
Vessel caliber and anatomic location often influence catheter
selection. Smaller caliber devices use higher ultrasound fre-
quencies that afford greater surface resolution with less depth
penetration than larger diameter probes. The most common
catheters used to image the coronary arteries are available in
2.9- and 3.2-Fr diameters with a preset frequency of 30 MHz.
Larger peripheral vessels require a wider range of devices
with diameters of 6 and 8.2 Fr with frequencies of 20 and
12.5 MHz, respectively.

Intimal flaps and dissections

Conventional diagnostic modalities for evaluating acute aor-
tic dissections include angiography, computed tomography,
magnetic resonance imaging, and transthoracic and trans-
esophageal echocardiography. Although technically accept-
able, these imaging tools are limited in image production and
resolution in certain clinical situations. IVUS promises a

Table 35.1 Applications of intravascular ultrasound

Diagnostic

Characterize luminal pathology and disease location

Eccentric vs. concentric plaque

Display luminal shape (circularor elliptical)

Defining the distribution of disease within the arterial lumen
Determine vessel cross-sectional dimensions

Accurately measure percent luminal stenosis

Quantitate medial and intimal thickening
Tissue characterization

Identify plaque composition (lipid, SMC, fibrous, calcium)

Differentiate plaque from thrombus
Arterial dissection

Identify entry site and intimal flap location

Determine false lumen relationship to origin of major branch vessels

Identify and localize intravasculartumors

Therapeutic

Match interventional method with lesion characteristic

Elucidate mechanisms of angioplasty
Quantitate wall stretching, dissection, plaque rupture, or ulceration
Assess recoil and spasm, need for further intervention

Guide angioplasty devices
Aid in selection of appropriate device (balloon)

Assess effects of therapy
Real-time intraluminal imaging of balloon angioplasty
Measure plaque and lumen areas before and after intervention
Provide accurate control data

Aid intravascular stent deployment
Identify anatomic landmarks

Confirm appropriate access (i.e. contralateral aortic stent-graft limb)
Guide stent selection based on intraluminal observations
Ensure accurate stent-graft sizing and positioning
Confirm adequacy of stent apposition and sealzone length

Adapted and modified from Cavaye DM. Intravascular ultrasound. In: White
RA, Hollier LH, eds. Vascular Surgery: Basic Science and Clinical Correlations.
Philadelphia, PA: JB Lippincott Co., 1 994:503.

unique approach in the identification and treatment of intimal
flaps and arterial wall dissections. It is especially useful when
the diagnosis is unclear or if the extent is uncertain. 35-37 Be-
cause IVUS is a dynamic, real-time imaging modality, the
movement of arterial flaps with pulse pressure variation can
be seen.

Although the indications for IVUS in clinical practice have
not been standardized, its utility in addressing thoracoab-
dominal dissections has been confirmed by several investiga-
tions. 35-41 Preliminary observations acknowledged several
parameters essential to the successful identification of the ex-
tent, and the possible treatment of acute aortic dissection by
means of endoluminal interventions. These include the fol-
lowing: accurate identification of the true and false lumen
(Fig. 35.11), isolating the site of proximal entry, evaluating the
extent of intimal injury and whether any fenestrations or re-
entry sites exist, determining the relationship between the

408

chapter 35 Intravascular ultrasound

Figure 35.11 Aortic dissection: IVUSand pullback demonstrating cannulationofthe false lumen (left)with the contralateral guidew ire (arrow). Repositioning
of the guidewire within the true lumen (right).

Figure 35.12 IVUS interrogation of an aortic
dissection: both the false and true lumen are
shown to supply the origin of the superior
mesenteric artery.

true /false lumen and the origin of major visceral branches
(Fig. 35.12), obtaining accurate luminal dimensions to facili-
tate accurate stent selection, and confirming the precision of
device implantation (Fig. 35.13). The ability of IVUS to recog-
nize luminal and vessel wall abnormalities that are not readily
apparent on conventional angiographic studies provides a
new dimension to future diagnostic evaluations and thera-
peutic interventions.

Interventions for occlusive disease

Transluminal balloon angioplasty

Although contrast angiography remains the gold standard for
determining patency and vascular continuity, IVUS offers a
unique perspective in the opportunity to inspect the distri-

bution of disease, and analyze the results of interventions
through intraluminal and transmural imaging. IVUS can
specifically differentiate between calcified and fibrous lesions,
and affords information on the morphologic eccentricity or
concentricity of a mural thrombus or plaque (Fig. 35.14). IVUS
is also particularly helpful in assessing the relationship of the
ostia of branch vessels to an atherosclerotic lesion, as well as in
determining the length and diameters of diseased segments.

Conclusions between contrast angiography and IVUS have
been discussed in several investigations concerning both the
coronary and peripheral vasculature. 42-45 In normal or mini-
mally atherosclerotic vessels, cross-sectional areas calculated
from biplanar arteriograms and those measured from IVUS
were statistically similar. Good correlation was also seen when
imaging mildly elliptical lumina. However, when applied to
assess severely diseased vessels, angiography often under-

409

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.13 Preprocedural interrogation of
a traumatic iliac artery dissection with IVUS
accurately identifies the site of proximal entry
(top left) and the extent of intimal disruption
(bottom left). The appropriate size stent-graft
is then selected based on intraluminal
sonographic data. Postendovascular exclusion,
IVUS confirms an adequate entry site seal (top
right) and reaffirms deployment of the
prosthesis in the true lumen with obliteration of
the false lumen (bottom right).

Figure 35.14 Concentric mural thrombus is observed during pre-
intervention IVUS examination using an 8.2-Fr, 1 2.5-MHz probe.

estimated the degree of stenosis. 42/43/45 In contrast, when calcu-
lating the cross-sectional area of significantly atherosclerotic
lumina with elliptical lesions, measurements extrapolated
from arteriography were greater than values obtained on
IVUS and overestimated the true cross-sectional area. 46

Long-term success in peripheral and coronary interven-

tions is governed by the ability to restore a healthy hemody-
namic equilibrium through adequate arterial dilation. In this
regard, data from conventional angiography has been less sen-
sitive in depicting the effects of endovascular therapies. 6,47-49
For an arteriogram to display successful continuity after treat-
ment of hard lesions, the resultant dissection must extend into
the vessel media. Failure occurs in nondisplaceable plaques or
if an intimal disruption or circumferential dissection ensues.
Success in soft lesions is associated with a superficial fissure or
fracture of the endothelial surface, whereas a poor outcome is
seen with vessel recoil, luminal disruption, or thrombosis at
sites of plaque rupture. Uniplanar or biplanar angiography, by
providing only a luminal "silhouette," correlates poorly with
information obtained from IVUS concerning these postinter-
ventional changes. 45 When comparing these two modalities
for occlusive disease, overall patency was related to the free
luminal area and lesion heterogeneity as evaluated by IVUS.
This is in contrast to arteriographic measurements, which
were found to have no predictive value. 14 IVUS can reveal the
presence and volume of atheromatous plaque with better pre-
cision, is more sensitive in detecting lesion eccentricity and
calcifications, and enhances clinical judgment in determining
the most suitable approach to revascularization using en-
dovascular methods. 14/48/50/51 Adjunctive use of IVUS has
also been invaluable in defining the factors associated with
immediate or late complications (e.g. perforation, thrombosis,
restenosis, or dissection) following percutaneous trans-
luminal balloon angioplasty (PTA) (Table 35.2). 15/51-53 In this
regard, during preprocedural interrogation, IVUS can

410

chapter 35 Intravascular ultrasound

Table 35.2 Risk of restenosis after PTA

Early

Late

Luminal thrombus
Raised intimal flap

Extensive medial dissection
Oversized balloon expansion

Residual stenosis > 30%
Concentric fibrous plaque
Absence of dissection or calcification
Undersized balloon selection

Adapted from: Landau C, Lange RA, Hillis LD. Percutaneoustransluminal
angioplasty. N Engl J Med 1 994; 330:981 ; Back MR, Kopchok GE, White RA.
Intravascular ultrasound imaging. In: White RA, FogartyTJ,eds. Peripheral
Endovascular Interventions, 2nd edn. New York: Springer-Verlag, 1 999:1 95;
Gussenhoven EJ, vanderLugtA, PasterkampG etal. Intravascular ultrasound
predictors of outcome after peripheral balloon angioplasty. Eur J Vase
Endovasc Surg 1995; 10:279.

differentiate plaque from thrombus and reveal the degree
of calcification, which is a predictor of more severe medial
dissection after balloon angioplasty than in vessels with mini-
mally calcified atheroma. 54

Other issues associated with restenosis have been defined
through the use of IVUS by several investigators. The et al. im-
aged 16 patients pre- and post-PTA of the superficial femoral
arteries. 48 IVUS was demonstrated to detect accurately the
presence of dissections, plaque fractures, and internal elastic
lamina ruptures with thinning of the media. Observations
from this study revealed that arterial remodeling post-PTA
resulted in luminal enlargement despite a constant lesion
volume. During PTA, embolization of fractured plaque and
thrombotic material was also seen. In a group of 40 patients,
Losordo et al. found that plaque fracture and displacement
contributed to 72% of the final luminal cross-sectional area
after PTA, with wall dilation or stretching accounting for an
additional 18% increase. 55

IVUS applications to the coronary vasculature found a
correlation between the morphometric characteristics of the
plaque, the mechanism of coronary angioplasty (PTCA), and
the risk of restenosis. 49 In this study, densely calcific lesions
were detected in 83% of cases by IVUS but in only 14% by an-
giography. These more calcified plaques were predisposed to
dissection and were associated with larger postprocedural
residual lumina than fibrous plaques. Thus, lesions at high
risk for restenosis were identified as fibrous plaques with con-
centric distribution that remained intact after PTCA. This re-
sults in an elastic recoil thought to be the primary mechanism
for early restenosis. Tobis et al. 47 confirmed these findings and
suggested further endoluminal intervention (i.e. stent deploy-
ment or directional atherectomy) be considered to augment
long-term patency.

Studies have also indicated that balloon size for PTA or
PTCA is often underestimated when selection is made using
quantitative angiography alone and that optimal balloon size
is more accurately determined by IVUS. 56/57 In addition to spe-
cific plaque characteristics, several cardinal variables includ-

ing oversized and undersized balloons have been associated
with an increased risk of early or late restenosis (Table 35.2).
These factors are identifiable by IVUS and influence peripro-
cedural decisions regarding correction of residual stenosis, ex-
tensive dissections, or luminal thrombus, which may ensure
the long-term success of peripheral interventions.

Intravascular stent placement

Since the 1980s, intravascular stents have been deployed for
various indications including postangioplasty. Careful stent
placement post-PTA has been found to reduce complications
attributed to vessel wall elastic recoil and spasm, residual
stenosis, intimal flaps, deep medial dissection, and plaque ul-
ceration with local thrombus accumulation. IVUS is particu-
larly suited to assess the transmural effects of angioplasty and
the change in vessel morphology as a result of stenting. 6 ' 50 ' 58,59

Essential requirements for successful stent placement in-
clude accurate initial positioning and full deployment at the
time of balloon expansion. 6 ' 7 ' 58-59 Incomplete stent apposition,
not detected by angiography but seen on IVUS, has been re-
ported to occur in up to 20-40% of cases. 60 Careful imaging
with intraluminal ultrasonography before stent insertion de-
termines the exact location and identifies the shape and di-
mensions of the arterial disorder to be corrected. Incomplete
expansion, as evidenced by unapposed stent struts to the ves-
sel wall (Fig. 35.15), may be associated with an increased inci-
dence of vessel wall thrombosis and stent migration whereas
overdilation can result in excessive intimal hyperplasia or
perforation. 61 This is especially crucial in PTCA where early
stent thrombosis has required the use of anticoagulation in up
to 25% of patients. By ensuring accurate deployment and stent
expansion, IVUS can decrease the incidence of thrombosis and
eliminate the need for long-term anticoagulation. 62

A recent vascular registry review demonstrated an in-
creased 3- and 6-year patency in patients with iliac occlusive
disease who underwent balloon angioplasty (PTA) and pri-
mary stenting. Buckley et al. 63 retrospectively reviewed 52
patients who underwent PTA and primary stenting for symp-
tomatic aortoiliac disease. The results are summarized in
Table 35.3. IVUS and arteriography were used to evaluate the
lesion in 36 patients (49 limbs), and arteriography alone in the
remaining 16 patients (22 limbs). Kaplan-Meier 3- and 6-year
primary patency estimates were 100% and 100% in the IVUS
group and 82% and 69% in those treated without IVUS
(P < 0.001). 63 In addition, there were no secondary procedures
in limbs treated with the benefit of IVUS and a 23% secondary
intervention rate in the non-IVUS group (P < 0.05). Poor stent
apposition was noted and corrected with larger balloon angio-
plasty in 40% of stented lesions by IVUS evaluation despite
the appearance of adequate expansion on arteriography.
The authors concluded that IVUS significantly improves the
long-term patency of iliac arterial lesions by defining the
appropriate angioplasty diameter endpoint and adequacy of

411

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.1 5 IVUS postiliac percutaneous
transluminal balloon angioplasty (PTA) and stent
placement clearly demonstrate that the stent is
underdeployed although the completion
arteriogram revealed adequate vessel patency
(left). Full stent expansion was accomplished with
repeat PTA and a larger balloon size.
Postprocedural IVUS results are shown with
complete apposition of the stent to the vessel
wall (right).

Table 35.3 Intravascular ultrasound and peripheral angioplasty

Study demographics

IVUS

Without IVUS

P- value

Number of patients

Numberof limbs

3-year patency

49(100)

18/22(82)

<0.001

6-year patency

49(100)

15/22(69)

<0.001

Secondary procedures

5/22 (23)

<0.05

Adapted from Buckley CJ, Arko FR, Lee S etal. Intravascular ultrasound
scanning improves long-term patency of iliac lesions treated with balloon
angioplastyand primary stenting. J Vase Surg 2002; 35:316.

stent placement. IVUS improves anatomical perception by
combining information about plaque and vessel wall consis-
tency with lesion location data, by quantitating residual steno-
sis and dissections, thereby influencing the adequacy of
arterial stent deployment. 64

Deployment of endovascular prosthesis

Since the first successful exclusion of an abdominal aortic
aneurysm (AAA) in 1991, endovascular interventions have
revolutionized the landscape in the treatment of vascular
disease. 65 One of the most fundamental and influential dif-
ferences between conventional surgery and endoluminal
grafting is the central role of imaging in every aspect of man-
agement. Current technologies offer new ways to evaluate pa-
tients for endograft procedures and to enhance the accuracy of
interventions. 66-68 Some data are complementary with other
methods, and others unique to a particular modality
(Fig. 35.16).

Contrast angiography is useful for defining the continuity
and morphology of vascular anatomy, and determining the
presence of associated abnormalities. Axial and 3-D computed
tomographic scans (CT) can noninvasively determine both
lumen and wall characteristics and provide anatomic infor-
mation on the location of surrounding structures. Exclusively,
arteriography is limited by magnification, thrombus effect,

foreshortening due to tortuosity, loss of luminal detail,
parallax error, and projection errors. 69-73 Difficulty with
CT involves methodologic differences in elliptical diameter
measurement, erroneous length determinations dependent
on slice thickness, and interobserver variability 74-76 On the
other hand, spiral or helical CT with 3-D processing has been
found to be particularly useful and universally well re-
ceived. 77-81 The greatest benefit of this modality lies in the area
of preoperative planning and in postprocedural assessment
after endovascular AAA repair. Efficient utilization, however,
often requires experience with interpretation algorithms and
use of a complex computerized workstation. The nature of
data acquisition is also limited by variations between software
vendors leading to labor-intensive reconstructions with a
margin for error accompanying interobserver variability. 82,83
Additional limitations include infusion of a larger contrast
volume than conventional arteriography, unavailability of in-
formation intraoperatively, and patient compliance with pro-
longed breathholding to enhance image resolution. 80 ' 84 ' 85
IVUS offers additional qualitative and quantitative data ad-
dressing these concerns and provides essential real-time
anatomic information from the aortic neck to the bifurcation,
and the distal attachment sites. 86 IVUS also significantly re-
duces fluoroscopy time and contrast use minimizing exposure
and allowing treatment of patients with renal compromise.
When supplemented with preoperative surface-rendered spi-
ral CT reconstruction and intraoperative cinefluoroscopy,
IVUS enables patient selection and precise deployment of en-
doluminal devices with better understanding.

Prior to endograft deployment, an important aspect
provided by IVUS is identification of vascular lesions not
visualized by CT or cinefluoroscopy. Critical factors in deter-
mining secure proximal device fixation involve the distribu-
tion of luminal thrombus, the eccentricity of aortic plaque, and
the presence of intimal disruption (Fig. 35.17). 87-89 Angiogra-
phy may be misleading regarding these pathologic findings
which may alter anatomic perceptions of infrarenal neck
length and configuration. These changes may also be obscured
during the arterial phase of spiral CT imaging and may not be
captured during inconsistent timing of the venous phase. The

412

chapter 35 Intravascular ultrasound

Figure 35.16 Vascular imaging modalities for endovascular aortic
interventions. (A) Spiral computed tomography (CT) and workstation
reconstruction, (B) contrast angiography, (C) intravascular ultrasound with

pullback, and (D) conventional axial CT slices are often utilized for
preoperative sizing, stent-graft selection, and during intraoperative
deployment of endoluminal devices.

Figure 35.17 IVUS assessment of the
proximal infrarenal aortic neck in preparation
for endovascular AAA repair. Intravascular
pathologicfindingsthat may influence or
preclude endoluminal repair include mixed
eccentric plaque (left), circumferential
thrombus (top right), or aortic dissection with
thrombus in the false lumen (bottom right). Left
renal vein (open arrow); plaque (arrowhead).

information provided by IVUS can be used to determine the
morphology of these anomalies and delineate precise anatom-
ic detail while performing real-time observations of device ex-
pansion to ensure firm endograft fixation. Longitudinal
display of IVUS images with automated analysis aids in eluci-
dation of neck anatomy into one of five categories guiding ap-

propriate stent-graft selection (Fig. 35.18). 79/90 As an example,
an aortic prosthesis with greater radial strength is preferen-
tially selected in aortic segments with a greater degree of
calcium or thrombus to achieve a more efficient seal and max-
imize the degree of stent-graft apposition within the entire
neck length. In other scenarios, devices with proximal attach-

413

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.18 Diagram of neck configuration determination guidelines.
Neck shape was described as conical, inverted conical, barrel, hourglass, or
straight based on infrarenal neck dimensions at the proximal, mid, and distal
infrarenal neck. (Adapted from Balm R, Stokking R, Kaatee R etal. Computed
tomographic angiographic imaging of abdominal aortic aneurysms;
implications for transfemoral endovascular aneurysm management. J Vase
Surg 1997; 26:231.)

ment capability or suprarenal fixation may be chosen for treat-
ment of a reverse-tapered infrarenal neck configuration to
minimize subsequent migration.

During endograft deployment, intravascular isonation is
particulary valuable in defining the relationship of the renal
artery ostia to the aneurysm as well as in evaluating the total
length and diameters along the lesion. IVUS avoids issues
with image magnification, parallax, and uniplanar views and
is useful for locating the appropriate proximal landing zone
and selecting the properly sized graft (Fig. 35.19A,B). 7/86/91/92

Intraluminal findings discovered by IVUS may also alter de-
ployment strategy. This is often seen in the setting of a saccular
AAA with a distal ulcerated plaque or partial dissection. En-
trapment of the contralateral limb of a modular stent-graft
within the false lumen can occur should the device be deliv-
ered in the anteroposterior plane. This creates a challenging
exercise in the cannulation of the contralateral limb. Careful
analysis during IVUS interrogation and manual withdrawal
("pullback") allows for a reversed limb deployment technique
to facilitate successful implantation (Fig. 35.20). Intraopera-
tively, IVUS is also an accurate way of determining full stent
expansion and for obtaining information regarding the conti-
nuity and the alignment of the graft material to the arterial
lumen (Fig. 35. 21 ). 66 ' 67 ' 93

Figure 35.19 (A) Not clear on arteriography due to
aortic neck angulation and parallax, aortic endograft
coverage of both renal artery ostia is confirmed on
intradeployment IVUS (white arrows). (B) After
careful positioning of the main body, completion
IVUS and arteriogram confirm accurate infrarenal
placement.

414

chapter 35 Intravascular ultrasound

Figure 35.20 Saccular abdominal aortic aneurysm with ulcerated plaque
proximal to the bifurcation. IVUS with pullback revealed an ulcerated luminal
plaque within the aneurysm (Aand center). Implantation strategy employed

a reversed-limb deploymenttechnique that ensured proper cannulation of
the contralateral limb seen on postoperative surveillance computed
tomographyangiogram (B).

Figure 35.21 (A) Difficult to visualize on
fluoroscopy, intraoperative IVUS accurately
identified the site of iliac limb dislocation
(arrow). (B) After determining the uncovered
length, the vessel wall was relined with the
properly sized stent.

415

pa rt 1 1 1 Invasive vascular diagnostics

Although much significance has been attributed to the mor-
phologic landscape of the infrarenal neck, distal fixation is of
equal import. Tortuous iliac vessels not only provide a chal-
lenge to access but can also obscure arteriographic findings.
Angiography demonstrates vascular continuity, but provides
ambiguous information concerning luminal involvement es-
pecially in identifying the origin of the hypogastric vessels.
IVUS can be complementary in this regard, and direct staged
intervention to ensure ample fixation and an appropriate seal-
zone length (Fig. 35.22).

The technology surrounding endoluminal procedures
continues to evolve. By fusing data obtained from IVUS with
information analyzed from angiography and helical CT,
improved ease and accuracy for conducting endovascular
procedures is realized.

Venous indications

The implementation of IVUS for diagnostic and therapeutic
objectives is not relegated exclusively to intraarterial applica-
tions. IVUS has been utilized successfully to assess the degree
of tumor extension into the inferior vena cava from renal cell or
hepatocellular carcinoma and establish resectability 94-96
Intraoperatively, IVUS has been used to determine portal
vein involvement from pancreatic adenocarcinoma through
the superior mesenteric vein route. 97

The clinical utility of IVUS-guided placement of inferior
vena caval filters for venous thromboembolic disease has re-
cently been explored. Deployment under real-time ultrasonic

imaging has been successful without the use of contrast
venography in a recent clinical trial. 98 The advantage of
bedside insertion obviates the need for cinefluoroscopy,
and allows for treatment of morbidly obese individuals with
renal impairment while precluding issues associated with
critical care transportation.

Color flow IVUS

One of the more remarkable developments with external
beam sonography is the introduction of color flow ultrasound
imaging. 99-101 This method superimposes a blood flow image
on a standard gray-scale ultrasound display, permitting in-
stantaneous, visual assessment of blood flow. Duplex or color
flow Doppler ultrasound is based on differences in the pulse
frequency transmitted by the transducer, and the signal re-
flected from moving blood known as the "Doppler frequency
shift/' 102,103 This value, which is proportional to blood flow
velocity, is calculated from the multiple emitted pulses that
are directed forward to the probe. 104 ' 105 Because IVUS trans-
ducers emit only one or two pulses per transmission, and
are not forward looking but send pulses perpendicular to
the catheter tip, the Doppler shift principle cannot be
applied. To circumvent these limitations without altering
sonographic technology, a new computerized software pro-
gram has been recently designed. 106 This program enables
the capture of up to 30 conventional IVUS frames per second
to produce a "real-time" image. Sequential axial images are

Wi .j

Figure 35.22 Endovascular AAA repair with hypogastric coil embolization.
Predeployment IVUS demonstrates extension of a right common iliac artery
aneurysm involving the origin of the ipsilateral hypogastric artery not seen on

initial arteriogram (left and center). Prestent coil embolization of the internal
iliac origin was then performed to create a more effective sealzone and
eliminate a potential source for endoleak (right).

416

chapter 35 Intravascular ultrasound

1 1 1 #1 1 l :

■-_

%
-

■■

j^^l^H

IpZ

■'*-... A \

r* _

»j i— J ^

*-,

r-fSV^ "

H. ■ :r. "^

■

^ ^""^

*■ •

ITT -.

_,— - .. ii_^ ■ _»

- „ — *=■

. ' -Hi m T '

™ ■

■

— . « - _->.

Figure 35.23 (A) Intraoperative color IVUS
images of multiple superficial femoral artery
pseudoaneurysms from penetrating trauma.
Acquired real-time axial images provide a 2-D
section of the vessel, while manual "pullback"
creates a 3-D longitudinal color image that
can be rotated around the catheter axis.
(B) These injuries were treated with a
polytetrafluoroethylene-covered self-expanding
nitinol stent. (The arrows above identify the
center of the IVUS probe.) See also Plate 3,
facing p. 370.

then acquired and differences in the position of echogenic
blood particles between images are compared. Larger dif-
ferences are categorized as "high flow" while minimal dis-
parity is considered "low flow." Axial and 3-D renderings
are then created while adding sufficient color dependent
on the degree of flow. Despite this level of sophistication,
since only a few pulses are transmitted, velocities cannot be
quantitated.

The advantage of colorized flow is the ability to recognize

more readily the true lumen by verifying the existence of circu-
lating blood flow. This new modification serves to enhance the
role of IVUS in both diagnostic and therapeutic applications.
In a traumatic setting, color IVUS was able to locate precisely
the areas of continuity disruption and grade the extent of in-
jury within the superficial femoral artery (Fig. 35.23A,B; also
in colour, see Plate 3, facing p. 370). 66 Accurate covered stent
placement was expeditiously performed and confirmed with-
out adverse sequelae.

417

pa rt 1 1 1 Invasive vascular diagnostics

Figure 35.24 Color IVUS images postdeployment of an aortic endograft.
IVUS interrogation with Chromaflow demonstrates adequate proximal seal
(left). However, inadequate distal stent-graft apposition evidenced by

independent arterial wall pulsation (arrow) at the stent interface resulted in a
retrograde type I endoleak (right). See also Plate 4, facing p. 370.

Figure 35.25 Three-dimensional segmentation and spectral analysis (left)
convert raw radio-frequency IVUS data (top right) into color-coded
parametric images (bottom right) emphasizing plaque boundary features.

Plaque composition is defined as fibrous (green), fibro-lipidic (yellow),
calcium (white), or lipid core (red). (Courtesy of Scott Huennekens, Volcano
Therapeutics Inc., Laguna Hills, CA, USA.) See also Plate 5, facing p. 370.

418

chapter 35 Intravascular ultrasound

Another promising feature of this emerging design is the
detection of blood flow within aortic stent-grafts, which may
prove invaluable in the diagnosis of endoleaks (Fig. 35.24; also
in colour, see Plate 4, facing p. 370). Catheters necessary to vi-
sualize the entire aortic lumen, however, have a lower fre-
quency that significantly reduces the overall functional color
effect. Furthermore, air trapped within the fabric of the pros-
thetic tends to reflect ultrasonic pulses. Using an angled glide
catheter, this limitation is counteracted by manipulating the
probe circumferentially within the perimeter of the device. 106
Color flow or ChromaFlo is only available in 3.5-Ff, 20-MHz
solid-state phased array catheters.

Future developments

Endovascular interventions have experienced exponential
growth and have dramatically altered the approach to vascu-
lar pathology. Potential new efforts have focused on combin-
ing an interventional component (i.e. balloon, stent, or
pressure wire) with the IVUS transducer to simplify overall as-
sessment, facilitate catheter exchanges, and to visualize imme-
diate device deployment or angioplasty results. 106-108 More
recently, plaque composition rather than lesion stenosis espe-
cially within the coronary circulation has been of major con-
cern. Unstable plaques may rupture or initiate an acute
thrombotic reaction resulting in chronic angina and acute
coronary syndromes, such as myocardial infarction or
sudden death. 109-113 Multiple investigations using IVUS
have attempted to define lesion morphology with finite
accuracy. 110 ' 114-117 However, current gray-scale imaging has
proven inconsistent with characterization of detailed plaque
composition. 118 Utilizing advanced signal processing
technology, unique boundary features within the plaque and
vessel wall have been identified by analyzing eight spectral
parameters of the reflected ultrasound signal. The newly
reconstructed images more closely approximate true an-
atomic findings and further delineate the composition of
athersclerotic plaques (Fig. 35.25; also in colour, see Plate 5,
facing p. 370). 119-123

By combining these new innovations with current IVUS
technology, a myriad of exciting possibilities can be realized
in peripheral vascular research. Blood vessel compliance,
dynamic changes in the vessel wall caused by disease or phar-
macologic interventions, and assessment of transmural
pathologic features of atherosclerosis are among the multi-
tude of topics whose investigation can only be enhanced with
intravascular ultrasound.

References

1. Donald I. How and why medical sonar developed. Ann R Coll
Surg Engl 1974; 54:132.

2. Bom N, ten Hoff H, Lancee CT et al. Early and recent intraluminal
ultrasound devices. Int J Card Imaging 1989; 4:79.

3. Yock PG, Linker DT, Angelsen BAJ. Two-dimensional intra-
vascular ultrasound: technical development and initial clinical
experience. J Am SocEchocardiogr 1989; 2:296.

4. Kopchok GE, White RA, Guthrie C et al. Intraluminal vascular ul-
trasound: preliminary report of dimensional and morphologic
accuracy. Ann Vase Surg 1990; 4:291.

5. van Urk H, Gussenhoven WJ, Gerritsen GP et al. Assessment of
arterial disease and arterial reconstructions by intravascular ul-
trasound. Int] Cardiac Imaging 1991; 6:157.

6. Scoccianti M, Verbin CS, Kopchok GE et al. Intravascular ultra-
sound guidance for peripheral vascular interventions. / Endovasc
Ther 1994; 1:71.

7. White RA, Verbin C, Kopshok GE et al. The role of cinefluo-
roscopy and intravascular ultrasonography in evaluating the de-
ployment of experimental endovascular prostheses. / Vase Surg
1995; 21:365.

8. Nishanian G, Kopchok GE, Donaryre et al. The impact of in-
travascular ultrasound (IVUS) on endovascular interventions.
Semin Vase Surg 1999; 12:285.

9. Lockwood GR, Ryan LK, Foster FS. High frequency intravascu-
lar ultrasound imaging. In: Cavaye DM, White RA, eds. A Text
and Atlas of Arterial Imaging: Modern and Developing Technologies.
London: Chapman and Hall, 1993:125.

10. Cavaye DM. Intravascular ultrasound. In: White RA, Hollier LH,
eds. Vascular Surgery: Basic Science and Clinical Correlations.
Philadelphia, PA: JB Lippincott Co., 1994:503.

11. Kobayashi Y, Yock P, Fitzgerald P. Vascular IVUS landmarks.
Intravasc Imaging 1998; 2:35.

12. White RA, Donayre CE, Kopchok GE et al. Utility of intravascular
ultrasound in peripheral interventions. Tex Heart Inst J 1997;
24:28.

13. van Urk H, Gussenhoven WJ, Gerritsen GP et al. Assessment of
arterial disease and arterial reconstructions by intravascular
ultrasound. Int J Card Imaging 1991; 6:157.

14. Vogt KJ, Rasmussen JG, Just S et al. Effect and outcome of
balloon angioplasty and stenting of the iliac arteries evaluated
by intravascular ultrasound. Eur J Vase Endovasc Surg 1999; 17:
47.

15. Back MR, Kopchok GE, White RA. Intravascular ultrasound
imaging. In: White RA, Fogarty TJ, eds. Peripheral Endovascular
Interventions, 2nd edn. New York: Sp ringer- Verlag, 1999:195.

16. Burns PN, Goldberg BB. Ultrasound contrast agents for vascular
imaging. In: Cavaye DM, White RA, eds. Arterial Imaging: Modern
and Developing Technology. London: Chapman and Hall, 1993:
61.

17. ten Hoff H, Korbijn A, Smith TH et al. Imaging artefacts in
mechanically driven ultrasound catheters. Int J Card Imaging
1989; 4:195.

18. Gussenhoven WJ, Essed CE, Frietman P et al. Intravascular echo-
graphic assessment of vessel wall characteristics: a correlation
with histology. Int J Card Imaging 1989; 4:105.

19. Gussenhoven EJ, Essed CE, Lancee CT. Arterial wall characteris-
tics determined by intravascular ultrasound imaging: an in-vitro
study. J Am Coll Cardiol 1989; 14:947.

20. West AL Endovascular ultrasound. In: Moore WS, Ahn SS,
eds. Endovascular Surgery. Philadelphia, PA: WB Saunders Co.,
1989:518.

419

pa rt 1 1 1 Invasive vascular diagnostics

21. Kopchok GE, White RA, White G. Intravascular ultrasound: a
new potential modality for angioplasty guidance. Angiology
1990; 41:785.

22. Tabarra M, Kopchok GE, White RA. In vitro and in vivo evalua-
tion of intraluminal ultrasound in normal and atherosclerotic ar-
teries. Am J Surg 1990; 160:556.

23. Meyer Cr, Chiang EH, Fechner KP et al. Feasibility of high resolu-
tion intravascular ultrasonic imaging catheters. Radiology 1988;
168:113.

24. Yock PG, Johnson EL, Linker DT. Intravascular ultrasound de-
velopment and clinical potential. Am J Card Imaging 1988; 2:185.

25. Nissen SE, Grines CL, Gurely JC et al. Application of new phased-
array ultrasound imaging catheter in the assessment of vascular
dimensions. Circulation 1990; 81:660.

26. Neville RF, Bartorelli AL, Leon MB et al. Validation and feasibili-
ty of in vivo intravascular ultrasound imaging with a new flexi-
ble catheter. SurgForumACS 1989; 75:314.

27. Mallery JA, Tobis JM, Griffith J et al. Assessment of normal and
atherosclerotic arterial wall thickness with intravascular ultra-
sound imaging catheter. Am Heart J 1990; 119:1392.

28. Liu JB, Bonn J, Needleman L et al. Feasibility of three-
dimensional intravascular ultrasonography: preliminary
clinical studies. / Ultrasound Med 1999; 18:489.

29. Cook LT, Dwyer SJ, Batnisitzky S et al. A three-dimensional dis-
play system for diagnostic imaging applications. IEEE Comput
Graph Appl 1983; 3:13.

30. Raya SP, Udupa JK, Barrett WA. A PC-based 3D imaging system:
algorithms, software and hardware considerations. Comput Med
Imag Graphics 1990; 14:353.

31. Herman GT, Liu HK. Three-dimensional display of human or-
gans from computed tomograms. Comput Graph Image Proc 1979;
9:1.

32. Cavaye DM, Tabbara MR, Kopchok GE et al. Three dimensional
vascular ultrasound imaging. Am Surg 1991; 57:751.

33. Reid DB, Douglas M, Diethrich EB. The clinical value of three-
dimensional intravascular ultrasound imaging. / Endovasc
Ther 1995; 2:356.

34. White RA, Kopchok GE, Donayre CE. Intravascular ultrasono-
graphy. In: Moore WS, Ahn SS, eds. Endovascular Surgery, 3rd
edn. Philadelphia, PA: WB Saunders Co., 2001:159.

35. Cavaye DM, French WJ, White RA et al. Intravascular ultrasound
imaging of an acute dissection aortic aneurysm: a case report. /
Vase Surg 1991; 13:510.

36. Weintraub AR, Erbel R, Gorge G et al. Intravascular ultrasound
imaging in acute aortic dissection. } Am Coll Cardiol 1994; 24:495.

37. Pande A, Meier B, Fleisch M et al. Intravascular ultrasound for
diagnosis of aortic dissection. Am J Cardiol 1991; 67:662.

38. Cavaye DM, White RA, Lerman RD et al. Usefulness of intravas-
cular ultrasound imaging for detecting experimentally induced
aortic dissection in dogs and for determining the effectiveness of
endoluminal stenting. Am J Cardiol 1992; 69:705.

39. White RA, Donayre CE, Walot I et al. Endovascular exclusion of
descending thoracic aortic aneurysms and chronic dissections:
initial clinical results with the AneuRx device. / Vase Surg 2001;
33:927.

40. Giudice R, Frezzotti A, Scoccianti M. Intravascular ultrasound-
guided stenting for chronic abdominal aortic dissection. /
Endovasc Ther 2002; 9:926.

41 . Koschyk DH, Meinertz T, Hofmann T et al. Value of intravascular

ultrasound for endovascular stent-graft placement in aortic dis-
section and aneurysm. J Card Surg 2003; 18:471.

42. Tabbara MR, White RA, Cavaye DM et al. In-vivo human com-
parison of intravascular ultrasound and angiography. / Vase Surg
1991; 14:496.

43. Tobis JM, Mahon D, Lehmann K et al. The sensitivity of ultra-
sound imaging compared to angiography for diagnosisng
coronary atherosclerosis. Circulation 1990; 82 (Suppl. 3):439
(Abstract).

44. Nissen SE, Gurley JC, Grines CL et al. Intravascular ultrasound
assessment of lumen size and wall morphology in normal sub-
jects and patients with coronary artery disease. Circulation 1991;
84:1087.

45. Van Lankeren W, Gussenhoven EJ, Pieterman H et al. Compari-
son of angiography and intravascular ultrasound before and
after balloon angioplasty of the femoropopliteal artery. Cardio-
vasc Intervent Radiol 1998; 21:367.

46. Nissen SE, Gurley JC, Grines CL et al. Comparison of intravascu-
lar ultrasound and angiography in quantitation of coronary di-
mensions and stenoses in man: impact of lumen eccentricity.
Circulation 1990; 82 (Suppl. 3):440 (Abstract).

47. Tobis JM, Mahon DJ, Goldberg SL et al. Lessons from intravascu-
lar ultrasonography: observations during interventional angio-
plasty procedures. / Clin Ultrasound 1993; 21 :589.

48. The SHK, Gussenhoven EJ, Zhong Y et al. Effect of balloon angio-
plasty on femoral artery evaluated with intravascular ultra-
sound imaging. Circulation 1992; 86:483.

49. Honye J, Mahon DJ, Jain Aetal. Morphological effects of coronary
balloon angioplasty in vivo assessed by intravascular ultra-
sound imaging. Circulation 1992; 85:1012.

50. Isner JM, Rosenfield K, Losordo DW et al. Combination balloon-
ultrasound imaging catheter for percutaneous transluminal
angioplasty. Circulation 1991; 84:739.

51. Tabbara MR, Mehringer CM, Cavaye DM et al. Sequential intra-
luminal ultrasound evaluation of balloon angioplasty of an iliac
artery lesion. Ann Vase Surg 1992; 6:179.

52 . Gussenhoven E J, van der Lugt A, Pasterkamp Getal. Intravascu-
lar ultrasound predictors of outcome after peripheral balloon
angioplasty. Eur J Vase Endovasc Surg 1995; 10:279.

53. Landau C, Lange RA, Hillis LD. Percutaneous transluminal
angioplasty. N Engl} Med 1994; 330:981.

54. Fitzgerald PJ, Ports TA, Yock PG. Contribution of localized cal-
cium deposits to dissection after angioplasty: an observational
study using intravascular ultrasound. Circulation 1992; 86:64.

55. Losordo DW, Rosenfield K, Pieczek Aet al. How does angioplasty
work? Serial analysis of human iliac arteries using intravascular
ultrasound. Circulation 1992; 85:1012.

56. Roubin GS, Douglas JS Jr, King SB III et al. Influence of balloon
size on initial success, acute complications, and restenosis after
percutaneous transluminal coronary angioplasty: a prospective
randomized study. Circulation 1988; 78:557.

57. Nichols AB, Smith R, Berke AD et al. Importance of balloon size in
coronary angioplasty. J Am Coll Cardiol 1989; 13:1094.

58. Diethrich EB. Endovascular treatment of abdominal aortic occlu-
sive disease: the impact of stents and intravascular ultrasound
imaging. Eur J Vase Endovasc Surg 1997; 3:228.

59. Cavaye DM, Diethrich EB, Santiago OJ et al. Intravascular ultra-
sound imaging: an essential component of angioplasty assess-
ment and vascular stent deployment. Int Angiol 1993; 12:212.

420

chapter 35 Intravascular ultrasound

60. Katzen BT, Benenati JF, Becker GJ et ah Role of intravascular ul-
trasound in peripheral atherectomy and stent deployment. Cir-
culation 1991; 84 (Suppl. II):2152 (Abstract).

61. Busquet J. The current role of vascular stents. Int Angiol 1993;
12:206.

62. Colombo A, Hall P, Nakamura S et ah Intracoronary stenting
without anticoagulation accomplished with intravascular ultra-
sound guidance. Circulation 1995; 91:1676.

63. Buckley CJ, Arko FR, Lee S et ah Intravascular ultrasound scan-
ning improves long-term patency of iliac lesions treated with bal-
loon angioplasty and primary stenting. / Vase Surg 2002; 35:316.

64. Arko F, Mettauer M, McCollough R et ah Use of intravascular ul-
trasound improves long-term clinical outcome in the endovascu-
lar management of atherosclerotic aortoiliac occlusive disease. /
Vase Surg 1998; 27:614.

65. Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal
graft implantation for abdominal aortic aneurysms. Ann Vase
Surg 1991; 5:491.

66. White RA, Donayre CE, Walot I et ah Preliminary clinical out-
come and imaging criterion for endovascular prosthesis devel-
opment in high-risk patients who have aortoiliac and traumatic
arterial lesions. / Vase Surg 1996; 24:556.

67. White RA, Donayre CE, Kopchok GE et ah Vascular imaging
before, during, and after endovascular repair. World] Surg 1996;
20:622.

68. Beebe HG. Imaging modalities for aortic endografting. / En-
dovascTher 1997; 4:111.

69. Elisevich K, Cunningham I A, Assis L. Size estimation and mag-
nification error in radiographic imaging: implications for classi-
fication of arteriovenous malformations. Am J Neuroradiol 1995;
16:531.

70. Wolf YG, Thomas WS, Brennan FJ et ah Computed tomography
scanning findings associated with rapid expansion of abdominal
aortic aneurysm. / Vase Surg 1994; 20:529.

71 . Jean-Claude J, Newman KM, Li H et ah Possible key role for plas-
min in the pathogenesis of abdominal aortic aneurysms. Surgery
1994; 116:472.

72. Quinones-Baldrich WJ, Deaton DH, Mitchell RS et ah Prelimi-
nary experience with the Endovascular Technologies bifurcated
endovascular aortic prosthesis in a calf model. / Vase Surg 1995;
22:370.

73. Gottwik MG, Siebes M, Bahawar H et ah Quantitative angio-
graphic assessment of coronary stenoses: problems and pitfalls.
ZKardiol 1983; 72 (Suppl. 3):111.

74. Ouriel K, Green RM, Donayre CE et ah An evaluation of new
methods of expressing aortic aneurysm size: Relationship to rup-
ture. / Vase Surg 1992; 15:12.

75. Beebe HG, Jackson T, Pigott JP Aortic aneurysm morphology for
planning endovascular aortic grafts: limitations of conventional
imaging methods. JEndovasc Ther 1995; 2:139.

76. Jaakkola P, Hippelainen M, Farin P et ah Inter-observer vari-
ability in measuring the dimensions of the abdominal aorta:
comparison of ultrasound and computed tomography. Eur }
Vase Endovasc Surg 1996; 12:230.

77. Rubin GD, Walker PJ, Dake MD et ah Three-dimensional spiral
computed tomographic angiography: an alternative imaging
modality for the abdominal aorta and its branches. / Vase Surg
1993; 18:656.

78. Bayle O, Branhereau A, Roisset E et ah Morphologic assessment

of abdominal aortic aneurysm by spiral computed tomographic
scanning. / Vase Surg 1997; 26:238.

79. Balm R, Stokking R, Kaatee R et ah Computed tomographic
angiographic imaging of abdominal aortic aneurysms; implica-
tions for transfemoral endovascular aneurysm management. /
Vase Surg 1997; 26:231.

80. Gomes MN, Davros WJ, Zeman RK. Preoperative assessment
of abdominal aortic aneurysm: the value of helical and three-
dimensional computed tomography. / Vase Surg 1994; 20:367.

81. White RA, Donayre CE, Walot I. Computed tomography assess-
ment of abdominal aortic aneurysm morphology after endograft
exclusion. / Vase Surg 2001; 33:S1.

82. Broeders I, Blankensteijn JD, Olree M et ah Preoperative sizing of
grafts for transfemoral endovascular aneurysm management: a
prospective comparative study of spiral CT angiography, arteri-
ography, and conventional CT imaging. / Endovasc Ther 1997;
4:252.

83. Aarts NJ, Schurink GW, Schultz LJ et ah Abdominal aortic
aneurysm measurements for endovascular repair: intra- and
interobserver variability of CT measurements. Eur J Vase
Endovasc Surg 1999; 18:475.

84. Rubin GD, Beaulieu CF, Argiro V et ah Perspective volume ren-
dering of CT and MR images: applications of endoscopic imag-
ing. Radiology 1996; 199:321.

85. Balm R, Kaatee R, Blankensteijn JD et ah CT-angiography of
abdominal aortic aneurysms after transfemoral endovascular
aneurysm management. Eur J Vase Endovasc Surg 1996; 12:182.

86. Garrett HE Jr, Abdullah AH, Hodgkiss TD et ah Intravascular
ultrasound aids in the performance of endovascular repair of
abdominal aortic aneurysm. / Vase Surg 2003; 37:615.

87. Gitlitz DB, Ramaswami G, Kaplan D et ah Endovascular stent
grafting in the presence of aortic neck filling defects: early clinical
experience. / Vase Surg 2001; 33:340.

88. Parra JR, Ayerdi J, McLafferty R et ah Conformational changes
associated with proximal seal zone failure in abdominal aortic
endografts. / Vase Surg 2003; 37:106.

89 . Dillavou ED, Muluk SC, Rhee RY et ah Does hostile neck anatomy
preclude successful endovascular aortic aneurysm repair? / Vase
Surg 2003; 38:657.

90. van Essen JA, Gussenhoven EJ, Blankensteijn JD et ah Three-
dimensional intravascular ultrasound assessment of abdominal
aortic aneurysm necks. / Endovasc Ther 2000; 7:380.

91. White RA, Donayre CE, Kopchok GE et ah Intravascular
ultrasound: the ultimate tool for abdominal aortic aneurysm as-
sessment and endovascular graft delivery. / Endovasc Ther 1997;
4:45.

92. Tultein Nolthenius RP, van den Berg JC, Moll FL. The value of
intraoperative intravascular ultrasound for determining stent
graft size (excluding abdominal aortic aneurysm) with a modu-
lar system. Ann Vase Surg 2000; 14:311.

93. van Sambeek MR, Gussenhoven EJ, van Overhagen H et ah
Intravascular ultrasound in endovascular stent-grafts for peri-
pheral aneurysm: A clinical study. JEndovasc Ther 1998; 5:106.

94. Barone GW, Kahn MB, Cook JM et ah Recurrent intracaval renal
cell carcinoma: the role of intravascular ultrasonography. / Vase
Surg 1991; 13:506.

95. Kaneko T, Nakao A, Endo T et ah Intracaval endovascular ultra-
sonography for malignant hepatic tumor: new diagnostic tech-
nique for vascular invasion. Semin Surg Oncol 1996; 12:170.

421

pa rt 1 1 1 Invasive vascular diagnostics

96. Kaneko T, Nakao A, Nomoto S et ah Intracaval endovascular
ultrasonography for preoperative assessment of retrohepatic
inferior vena cava infiltration by malignant hepatic tumors.
Hepatology 1996; 24:1121.

97. Hannesson PH, Stridbeck H, Lundstedt C et ah Intravascular
ultrasound for evaluation of portal venous involvement in
pancreatic cancer. Eur Radiol 1997; 7:21.

98. Wellons ED, Matsuura JH, Shuler FW et ah Bedside intravascular
ultrasound-guided vena cava filter placement. / Vase Surg 2003;
38:455.

99. Switzer DF, Nanda NC. Doppler color flow mapping. Ultrasound
Med Biol 1985; 403.

100. Merritt CRB. Doppler blood flow imaging: integrating flow with
tissue data. Diagn Imaging 1986; 11:146.

101. Powis RL. Color flow imaging: understanding its science and
technology. / Diagn Med Sonograph 1988; 4:236.

102. Currie GR, White DN. Color-coded ultrasonic differential
velocity arterial scanner (echoflow). Ultrasound Med Biol 1978;
14:27.

103. Rosenfield K, Kelly SM, Fields CD et ah Non-invasive assessment
of peripheral vascular disease by color flow Doppler/two-
dimensional ultrasound. Am J Cardiol 1989; 64:247.

104. Nelson TR, Pretorius DH. The Doppler signal: where does it
come from and what does it mean? Am J Roentgenol 1988; 151:439.

105. Scoutt LM, Zawin ML, Taylor KJ. Doppler ultrasound. Part II:
clinical applications. Radiology 1990; 174:309.

106. Irshad K, Reid DB, Miller PH et ah Early clinical experience with
color three-dimensional intravascular ultrasound in peripheral
interventions. / Endovasc Ther 2001; 8:329.

107. White RA, Kopchok GE, Tabbara MR et ah Intravascular ultra-
sound guided holmium: YAG laser recanalization of occluded
arteries. Lasers Surg Med 1992; 12:239.

108. Alfonso F, Flores A, Escaned J et ah Pressure wire kinking, entan-
glement, and entrapment during intravascular ultrasound
studies: a potentially dangerous complication. Catheter
Cardiovasc Interv 2000; 50:221.

109. Maehara A, Mintz GS, Bui AB et ah Morphologic and angio-
graphic features of coronary plaque rupture detected by
intravascular ultrasound. J Am Coll Cardiol 2002; 40:904.

110. Kotani J, Mintz GS, Castagna MT et ah Intravascular ultrasound
analysis of infarct-related and non-infarct-related arteries in pa-
tients who presented with an acute myocardial infarction. Circu-
lation 2003; 107:2889.

111 . Medina R, Wahle A, Olszewski ME et ah Three methods for accu-
rate quantification of plaque volume in coronary arteries. Int }
Cardiovasc Imaging 2003; 19:301.

112. Kotani J, Mintz GS, Pregowski J et ah Volumetric intravascular
ultrasound evidence that distal embolization during acute
infarct intervention contributes to inadequate myocardial perfu-
sion grade. Am J Cardiol 2003; 92:728.

113. Schaar JA, de Korte CL, Mastik F et ah Intravascular palpography
for high-risk vulnerable plaque assessment. Herz 2003; 28:488.

114. Fujii K, Kobayashi Y, Mintz GS et ah Intravascular ultrasound as-
sessment of ulcerated ruptured plaques. A comparison of culprit
and nonculprit lesions of patients with acute coronary syn-
dromes and lesions in patients without acute coronary syn-
dromes. Circulation 2003; 108:2473.

115. Schaar JA, De Korte CL, Mastik F et ah Characterizing vulnerable
plaque features with intravascular elastography. Circulation
2003; 108:2636.

116. White JA, Pflugfelder PW, Boughner DR et ah Validation of a
three-dimensional intravascular ultrasound imaging technique
to assess atherosclerotic burden: potential for improved assess-
ment of cardiac allograft coronary artery disease. Can J Cardiol
2003; 19:1147.

117. Schoenhagen P, Tuzcu EM, Stillman AE et ah Non-invasive as-
sessment of plaque morphology and remodeling in mildly
stenotic coronary segments: comparison of 16-slice computed
tomography and intravascular ultrasound. Coron Artery Dis
2003; 14:459.

118. Schoenhagen P, Stone GW, Nissen SE et ah Coronary plaque mor-
phology and frequency of ulceration distant from culprit lesions
in patients with unstable and stable presentation. Arterioscler
Thromb Vase Biol 2003; 23:1895.

119. Klingensmith JD, Shekhar R, Vince DG. Evaluation of three-
dimensional segmentation algorithms for the identification
of luminal and medial-adventitial borders in intravascular
ultrasound images. IEEE Transactions Med Imag 2000; 19:996.

120. Nair A, Kuban BD, Obuchowski NA, Vince DG. Assessing spec-
tral algorithms to predict atherosclerotic plaque composition
with normalized and raw intravascular ultrasound data. Ultra-
sound Med Biol 2001; 27:1319.

121. Klingensmith JD, Nair A, Kuban BD, Vince DG. Volumetric coro-
nary plaque composition using intravascular ultrasound: three-
dimensional segmentation and spectral analysis. Proc Comp
Cardiol2002;29:113.

122. Nair A, Kuban BD, Tuzcu EM, Schoenhagen P, Nissen SE, Vince
DG. Coronary plaque classification with intravascular ultra-
sound radiofrequency data analysis. Circulation 2002; 106:2200.

123. Nair A, Calvetti D, Kuban BD et ah Intravascular ultrasound
plaque characterization: spectral analysis and tissue maps. ] Am
Coll Cardiol 2003; 41 (Suppl. A):59A.

422

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Edited by Rodney A. White, Larry H. Hollier
Copyright © 2005 Blackwell Publishing

Angioscopy in peripheral vascular surgery

Arnold Miller
Thomas J. Holzenbein

The introduction of endoscopy into clinical medical practice
has exploded with advances in modern technology. Unlike
many of the subspecialties where endoscopy is routinely and
enthusiastically practiced such as gastroenterology, otolaryn-
gology, ophthalmology, and gynecology, the introduction of
routine vascular endoscopy during vascular surgery has met
with only halting and reluctant acceptance. Methods used to
visualize the interior of blood vessels include angiography,
B-mode ultrasound with or without Doppler analysis, intra-
luminal ultrasound, and magnetic resonance angiography.
Angioscopy, however, is the only method that allows direct in
vivo visualization of the interior of the blood vessels in real-life
colors, showing the subtle variations between the different
endoluminal states, both normal and pathologic. Only recent-
ly has the lack of endoluminal detail obtained by all the other
methods to detect and quantify endoluminal disease begun
to be appreciated. Angiography, still the mainstay and gold
standard for planning surgical procedures, provides only in-
verse shadows of the vessel wall. The radio-opaque contrast
media mixing within the blood often obscures intraluminal
abnormalities such as thrombus, dissections, false channels
characteristic of recanalized thrombus, or residual competent
valve leaflets in vein conduits in the in situ orientation.

The main reason limiting the widespread application of
angioscopy is the necessity to remove all the blood from the
field of vision for clear and consistent intraluminal visualiza-
tion, since blood is opaque to all forms of light. Achieving this
is not always easy and requires skill as well as an understand-
ing of the available instrumentation and techniques. Perhaps
just as important is the difficulty of introducing a relatively ex-
pensive and new technology into clinical practice in a time
of cost consciousness. Evaluation and acceptance of this tech-
nology based solely on the ability to demonstrate clear super-
iority ignores the obvious benefits of direct intraluminal
observation in normal and diseased states, such as the recogni-
tion of new or previously unappreciated intraluminal lesions
and their correlation with clinical outcome, as well as the use
of the angioscope for evaluating and teaching surgical tech-
nique. Superiority of one monitoring modality over another

defined only as determination of graft patency is a goal that
may be deceptively difficult to achieve. Determinants of graft
patency are multiple and variable. They include surgical skill,
extent of the disease, and the biology of the particular patient.
Just as was the case with the introduction of enhanced lighting
and loupe magnification, which initially met with resistance
by some vascular surgeons, benefits from direct endoluminal
observation, with its minimalization of the judgment factor,
experience, and "mystique" in the practice of vascular
surgery, maybe underestimated. Finally, endoscopic vascular
surgery is in its infancy, both in instrumentation evolution and
techniques, and remains one of the exciting and challenging
areas of exploration in vascular surgery.

This chapter briefly reviews the history of vascular endo-
scopy, the principles of irrigation, and the techniques of angio-
scopy, and presents an overview of the clinical experience of
angioscopy in the practice of modern vascular surgery, em-
phasizing its role in infrainguinal bypass grafting.

History

The major impetus to the investigation of in vivo endoscopy of
the cardiovascular system at the turn of the century was the
high incidence of fatal rheumatic heart disease with mitral
stenosis afflicting otherwise healthy young people. The first
"cardioscope" was built in 1913 by Rhea and Walker, 1 of the
Peter Bent Brigham Hospital (Boston). Their instrument, de-
signed to be inserted directly into the beating heart, included a
hooked cutting device to cut the chorda tendineae under direct
vision and so render the stenotic mitral valve incompetent.
Not appreciating that blood is opaque to all visible wave-
lengths of light, they recessed the distal viewing lens into the
rigid shaft of the cardioscope. This filled with blood and pre-
vented intracardiac visualization. In 1922, Allen and Graham
of St. Louis 2 modified the same cardioscope by adding a con-
vex Perspex lens to the distal end of the scope. Placing this
directly against the structures of the heart displaced the
blood from the adjacent cardiac structures and allowed clear

423

pa rt 1 1 1 Invasive vascular diagnostics

visualization of the chorda tendineae of the mitral valve. In-
serting the cardioscope through the auricular appendage of
the heart, they successfully rendered the mitral valve incom-
petent in 22 dogs with the hooked cutting device. Unfortu-
nately, most of the dogs died of empyema from infection of the
thoracotomy wound, and the technique was never systemati-
cally applied in humans. In 1943, Harken and Glidden 3 modi-
fied the device further by placing larger inflatable, transparent
balloons at the distal end of the cardioscope, and thus were
able to advance the instrument through the femoral vessels
and visualize the intracardiac structures in the anesthetized
experimental animal. With the advent of open heart surgery in
the 1950s, interest in cardioscopy soon waned, although spo-
radic reports of experimentation with cardioscopy continued
to appear in the literature. 4-11

Fiberoptic endoscopes were introduced into clinical
gastroenterologic practice by Hirschowitz and colleagues 12 in
1957, with a report of their initial experience with "a long fiber-
scope for the examination of the stomach and duodenum/' In
the early 1960s, Greenstone and associates, 13 using a flexible
choledochoscope, showed the feasibility of direct endoscopy
of the peripheral blood vessels in the examination of the aorta
and its major branches in the human cadaver, and in vivo in the
dog. In the next decade, Vollmar and Storz 14 in Europe, and
Towne and Bernard 15-17 in the United States, using both rigid
and flexible endoscopes in large series of vascular operations,
showed that the procedure indeed was more than a curiosity,
and had significant potential value in the clinical practice of
modern vascular surgery.

Progress in fiberoptic and electronic technology during the
1980s was rapid. An array of angioscopes with outer diameters
ranging from 0.5 to 3mm, with high resolution and excellent
visual quality, as well as miniature, gas-sterilizable CCD chip
video cameras are available for clinical practice. Linkage of
the angioscope's eyepiece to these video cameras, perhaps the
most important advance in the clinical application of angio-
scopy, allows the procedure to be visualized as an enlarged
image on a high-resolution video monitor, avoiding all prob-
lems of maintaining a sterile field in the operating room or
angiography suite.

There were many early reports on the usefulness and role of
angioscopy in the various vascular systems. 14 ' 15,17-20 Spears
and coworkers 21 attempted in vivo percutaneous coronary
angioscopy during cardiac catheterization. Grundfest and col-
leagues 19 ' 22 and Seeger and Abela 23 showed that potential
technical defects during bypass surgery could be detected and
corrected. Mehigan and Olcott 24 showed that intraoperative
angioscopy was a good alternative to the intraoperative
angiogram, and White and colleagues 25 reported on the
usefulness of angioscopy in thromboembolectomy

Since 1987, we have attempted to explore and optimize
angioscopic techniques and instrumentation and to assess crit-
ically the role of angioscopy in the practice of modern vascular
surgery, particularly during inf rainguinal bypass grafting.

Angioscopic equipment

Flexible angioscopes in clinical use range in external diameter
from 0.5 to 3.0 mm. Fiberoptic angioscopes of these sizes are
wonders of modern technology. They consist of bundles of
flexible glass fibers (3000 to as many as 30 000 or more) of vari-
ous types and refractive indices (clear glass or quartz) coher-
ently arranged and covered by an outer coating or cladding,
which ensures undistorted light and image transmission (Fig.
36.1). The number of fiber bundles (pixels) and the lensing sys-
tems are the main factors responsible for the resolution of the
angioscopic image; the more fibers, the more pixels and the
higher the resolution. The fiber bundles are organized into
those for imaging and those for conducting light. At the distal
end of the angioscope a convex lens is fitted to capture the light
emitted from the viewed intraluminal object and refocus the
image onto the mosaic of fibers of the optical bundle. Because
the fiber bundles are coherently arranged, this image is faith-
fully reproduced at the opposite end of the optical bundle,
where it may be magnified by an eyepiece and viewed directly,
or attached to a CCD chip video camera and viewed as
an image on a monitor, or attached to any other camera lens
system. To inject sufficient light for transmission through the
small volume of fiber bundles available in the modern angio-
scope for satisfactory intraluminal viewing, a very intense and
focused light source, usually derived from quartz-halogen or
xenon arc lamps, is used.

The definitive clinical angioscope may consist of only the
flexible light fibers, or include hollow channels that allow irri-
gation at the distal tip of the angioscope or for use as a working

Eyepiece and lens

CCD camera

Coupler

®-P> — Light fibers

Angioscope with irrigation channel

t t3t:

Ught
fibers

Irrigation
channei

Figure 36.1 The anatomy of an angioscope. (From Miller A, Jepsen S.
Angioscopy in arterial surgery. In: Bergan J, Yao J, eds. Techniques in Arterial
Surgery. Philadelphia: WB Saunders, 1 990:409.)

424

CHAPTER 36 Angioscopy in peripheral vascular surgery

channel for special intraluminal instrumentation. The distal
tip of the angioscope can be steered; this is usually done me-
chanically, and is facilitated by thin cables that extend along
the surface of the angioscope sheath. Such specialized features
as hollow channels or steering mechanisms increase the exter-
nal diameter of the angioscope as well as the overall rigidity
of the instrument. Inclusion of these special features into a
particular angioscope always entails a compromise between
the resolution and light intensity (total number of fiberoptic
bundles) and the external diameter of the angioscope.

Standard video and audio equipment is used for the intra-
operative recording of the angioscopic procedure. This allows
documentation of the procedure for review.

Basic techniques of
intraoperative angioscopy

Principles of saline irrigation

The fundamental problem with vascular endoscopy remains
the necessity to clear every last drop of blood from the visual
field. Different approaches to achieve this depend on the
anatomic region being examined, as shown in the following
list:

METHODS OF BLOOD DISPLACEMENT

• Saline irrigation (balanced salt solutions)

• Peripheral vessels

• Branches of the aorta

• Transparent balloon inflation (C0 2 , saline)

• Pulmonary arteries

• Heart

• Great veins

• Intraarterial injection of C0 2

• Peripheral vessels

• Branches of the aorta

In the intraoperative setting, complete isolation of the vas-
cular segment to be visualized may be obtained by isolating
the segment between arterial clamps and removing the blood
by flushing with a clear saline solution. This is standard prac-
tice during surgery for blood vessels in the suprainguinal and
abdominal vasculature and during venous thrombectomy. In
the infrainguinal region, only proximal control by occluding
the antegrade blood flow is necessary. The retrograde blood
flow from collaterals is cleared by flushing these vessels or
grafts with clear saline solution.

A novel but still experimental technique of blood displace-
ment, the intraarterial injection of C0 2 , 26 has been described.
Unlike saline, the gas is compressible, and thus the delivery of
C0 2 requires a special injector that delivers a precise volume
over a prolonged injection time. A standard angiographic con-
trast injector may compress the gas to the point of an explosive
delivery.

The most widely used method to clear the blood from a

restricted field in a particular vessel is local irrigation with a
balanced salt solution. Lack of appreciation of the factors gov-
erning successful irrigation and the difficulty in achieving
flow rates necessary for irrigation during surgery have de-
layed the incorporation of angioscopy as a routine procedure
in the practice of vascular surgery. Unlike angiography, where
a sufficient concentration of contrast media mixing with the
blood allows high-quality angiograms, in angioscopy the in-
traluminal blood must be totally replaced by a clear column of
fluid. A small volume of red cells causes blurring of the visual
field and the image appears to be out of focus. Addition of any
more blood makes meaningful visualization impossible.

Certain requirements are necessary to achieve a clear col-
umn of fluid in the vessels or grafts being angioscoped; these
are summarized in Table 36.1. All antegrade blood flow, both
from the main inflow vessel and collaterals, needs to be pre-
vented; otherwise, blood flowing in the same direction as the
irrigation fluid will join the irrigation fluid and a clear fluid
column will never be established. At surgery this usually en-
tails proximal clamp occlusion of the native arteries or graft.

To clear all blood and establish the clear fluid column, a
bolus of fluid injected at a high flow rate and large volume is
necessary. The more rapidly the column of fluid can be estab-
lished, the less total fluid will be needed for the angioscopic
study 27 Once the column of fluid is established, it can be main-
tained by irrigating at a much lower flow rate and smaller
volume. This prolongs the visualization time and minimizes
the total volume of irrigation fluid used.

The practical problem in achieving these flow rates is the
small size of the irrigation catheters necessary for intraopera-
tive use. Much less important is the use of the long lengths of
tubing necessary to maintain sterile fields for intraopera-
tive angioscopy. Flow in cylindrical pipes is described by
Poiseuille's law (Q = KAPr 4 /L, where Q = volume flow, K =
fluid viscosity constant, AP = P 1 - P 2 [pressure drop along
tube], r = inside tube radius, and L = tube length). This law
states that the pressure head necessary to generate flow is
directly proportional to the tube length, rate of flow, and vis-
cosity of the fluid, and inversely proportional to the fourth
power of the internal radius of the conduit. To achieve the high

Table 36.1 Principles of irrigation for intraoperative angioscopy

Aim

• To establish and maintain a column of clear fluid 'within the vessel
Requirements

• No antegrade flow in the main vessel or collateral vessels

• Initial fluid bolus of large volume and high flow rate to establish column of
clear fluid

• Subsequent small volume and low flow rate, with pressure in excess of
backflow pressure, to maintain clear fluid column

(From Miller A, Lipson W, Isaacsohn J, Schoen F, Lees R. Intraoperative
angioscopy: principles of irrigation and description of a new dedicated
irrigation pump. Am Heart J 1 989; 1 1 8:391 .)

425

pa rt 1 1 1 Invasive vascular diagnostics

Table 36.2 Consecutive measured flow rates over 1 min with pressure cuff
device inflated and pressures between 400 and 450 mmHg around a liter of
saline in a plastic container

Flow rate

Pressure = 400-450 mmHg

(ml/min)

IV set only

142

IV set + irrigation catheter

128

(2.5 mm ODx 300 mm)

106
90

IV set + 2.8-mm angioscope

with irrigation channel (1 mm

ID x 1200 mm)

IV, intravenous; OD, outer diameter; ID, Inner diameter.

*N= number of measurements

(From Miller A, Lipson W, Isaacsohn J, Schoen F, Lees R. Intraoperative

angioscopy: principles of irrigation and description of a new dedicated

irrigation pump. Am Heart J 1 989; 1 1 8:391 .)

flow rates through the small irrigation catheters used for
angioscopy during lower extremity revascularization, very
high pressures need to be generated at the fluid source.

Inflating a standard venous transfusion pressure cuff device
to 400-450 mmHg around a single liter of saline in a plastic
container allows a maximum flow rate of 142 ml/min (Table
36.2). As the saline container empties and alters its shape, de-
spite maintaining the pressure in the pressure cuff inflated to
400-450 mmHg, the flow rate decreases. With the addition of
various irrigation catheters, there is a further decrease in the
flow rate. From our experimental 27 as well as clinical experi-
ence, 28-30 the flow rates achieved with the pressure cuff device
are inadequate for routine intraoperative angioscopy except
in very limited circumstances. Furthermore, there is no control
over the flow rate. The flow rate cannot be varied and the exact
volume of fluid being injected into the patient is difficult to
monitor until termination of the procedure, when the pressure
cuff is removed and the saline bag examined.

Together with the Olympus Corporation (Lake Success,
NY), we developed a dedicated irrigation pump (Angio-
pump) for angioscopy. This peristaltic pump is designed to
provide flow rates between 10 and 400 ml/min and to generate
a maximum pressure of 2000 mmHg at the pump head. The
pump provides for the selection of two independent flow
rates, a high flow rate, or bolus, and a low flow rate, or mainte-
nance. These flow rate settings are variable and independent
of each other, and may be adjusted either before or during
the procedure. The flow rate is controlled remotely with a
foot pedal, allowing switching back and forth from bolus to
maintenance so that after the column of clear fluid is estab-
lished it may be maintained at all times in the vessel under
examination. 27

A serious concern during intraarterial infusion of fluid into

a relatively restricted outflow tract at high flow rates is that ex-
cessively high intraarterial pressures may be generated that
could damage either the intimal lining or inner layer of the
arterial wall, even to the extent of complete rupture. In our
experimental (Fig. 36.2A) and clinical studies (see Fig. 36.2B),
we have shown that this is not a problem provided the vessel
is not totally occluded or that irrigation is ceased as soon as
clearing of the visual field occurs. 27-29 ' 31

The most significant limitation of angioscopy is the volume
of irrigation fluid that can be infused safely into a particular
patient. In our experience of intraoperative angioscopy, the
volumes of fluid routinely required for irrigation with the
dedicated irrigation pump, less than 0.5 1, have not been exces-
sive, 28/29/31 and provided the patient is carefully monitored,
such volumes are safe even in the elderly and ill population
typically undergoing bypass surgery. To investigate this criti-
cal issue systematically, we compared the hemodynamic
response, anesthetic interventions during surgery, and the
postoperative 30-day morbidity and mortality in a prospec-
tive, randomized, controlled study of 110 patients under-
going infrainguinal bypass surgery with either completion
angioscopy (n = 60) or completion angiogram (n = 50). 31 Our re-
sults demonstrated that the hemodynamic changes that occur
with the volume of irrigation fluid required in completion
angioscopy for successful monitoring of infrainguinal bypass
procedures are not clinically significant (Fig. 36.3). They show
that with careful hemodynamic and anesthetic monitor-
ing, these intraarterial fluid volumes are safe and do not
require any increased anesthetic interventions (Table 36.3).

Table 36.3 Comparisons of interventions during anesthesia and
cardiovascular morbidity and mortality within the first 30 postoperative days
in 1 1 prospectively randomized infrainguinal bypass grafts

Angioscopy

Angiography

(n = 60)

(n = 50)

Anesthesia interventions

Nitroglycerine

43(72%)

34(68%)

Before

41 (68%)

32(64%)

During

36(60%)

28(56%)

After

37(62%)

23(46%)

Neosynephrine

8(13%)

11 (22%)

Furosemide

5(8%)

2 (4%)

Cardiovascular complications

19(32%)

15(30%)

4(7%)

6(12%)

CHF

4(7%)

1 (2%)

Other

11 (18%)

9(18%)

Mortality (<30 days)

1 (1.7%)

2(4.0%)

Ml, myocardial infarction; CHF, congestive heart failure.
(Modified from Kwolek C, Miller A, Stonebridge Petal. Safety of saline
irrigation for angioscopy: results of a prospective randomized trial. Ann Vase
Surg 1992; 6:62-68.)

426

CHAPTER 36 Angioscopy in peripheral vascular surgery

Figure 36.2 (A) Continuous blood pressure
measurements recorded in experimental animal
(pig). (Top) Baseline and postiliac artery ligation
pressure recordings in the brachial and femoral
arteries. (Bottom) Rise in baseline femoral artery
pressure after ligation of the iliac arteries at flow
rate settings of 1 00, 200, and 400 ml/min, with
no change in the brachial artery pressure. (B) A
typical continuous intraoperative pressure
tracing recorded during irrigation for
completion angioscopy. The intermittent high
peak pressures correspond to the high-volume,
high-flow-rate or "bolus" fluid, and the low-
volume, low-flow-rate or "maintenance" fluid
to the lower pressures. Infusion rate with a
dedicated pump is controlled by a foot pedal.
(A from Miller A, Lipson W, Isaacsohn J, Schoen
F, Lees R. Intraoperative angioscopy: principles
of irrigation and description of a new dedicated
irrigation pump. Am Heart J 1 989; 1 1 8:391 .
Bfrom KwolekC, Miller A, Stonebridge Petal.
Safety of saline irrigation for angioscopy: results
of a prospective randomized trail. Ann Vase
Surg 1992; 6:62.)

CO
CO
CD

"D
O

100

cd 60

I 40

20
01

LIGATION

i ; ■

i : V

Brachial

Femoral

INFUSION

Brachial

Femoral

100

200

Flow rate (ml/min)

400

Continuous pressure tracing
during irrigation for angioscopy

300-

200

100-

E
E

| 300 H

°- 200

100

Jfl

i — I
1 cm

10 cm = 1 minute

1 1 r

Time (min)

Furthermore, no increased patient morbidity or mortality in
the intraoperative, perioperative, or early postoperative (less
than 30 days) periods could be demonstrated, even in patients
with the highest preoperative cardiovascular risk status (see
Table 36.3).

The irrigation fluid volumes necessary for successful com-
pletion angioscopy are safe provided the anesthetist is aware
that angioscopy is to be performed, runs the patient "dry"
until the angioscopy is completed, and includes the irrigation

fluid in his or her calculations of the patient's total fluid
requirements.

Technique

We have described in detail the basic techniques for intraop-
erative angioscopy. The standard equipment and technique
for setting up in the operating room is shown in Figure 36.4. 32
For each angioscopic application, we use a standard method of

427

pa rt 1 1 1 Invasive vascular diagnostics

TIME POINTS llx-l

TIME POINTS Hx-IV

20-
15-

< 5

*■

-5

-10

P=00006 P

Adj. r'=0.21

* • *

'• *

n-|P=0.0002
- Adj. r*=0.22

7 A

< 3H

-1 -4

• **

-5

8-
6-
4-

B * .*

P=0.0001
Adj. ^=0.24 *

1 I I I I I

100 200 300 400 500 600 700

P=071
Adj. r 2 =0.00

i ,m r: ■ i . ; — 7

P=0.10
Adj. r'=0.07

P=0.05

Adj. r'=0.03 •

#• • #

--10
--20
--30

4
-0

--4

I " T "" | [■■■■!■■■■, |

100 200 300 400 500 600 700

Irrigation volume (ml)

Figure 36.3 Graphs of multivariate linear
regression analysis demonstrating no
statistically significant changes in pulmonary
artery systolic pressure (PAS), pulmonary artery
diastolic pressure (PAD), and central venous
pressure (CVP) in response to the volumes of
irrigation fluid used for angioscopy and their
return to baseline levels within 30min.Time
point llx, measurements at baseline and just
before angioscopy or arteriography; time point

III, measurements immediately after completion
of angioscopy but not arteriography; time point

IV, measurements 30 min after the completion
of angioscopy and arteriography. (From Kwolek
C, Miller A, Stonebridge Petal. Safety of saline
irrigation for angioscopy: results of a
prospective randomized trial. Ann Vase Surg
1992; 6:62.)

Camera
Character generator

Angioscopist

Instruments

Microphone

Light source
Video recorder

Pump

Surgeon

Pump and video
controls

Anesthesiologist

Figure 36.4 Angioscopy equipment and
setup in the operating room. (From Miller A,
Jepsen S. Angioscopy in arterial surgery. In:
Bergan J, Yao J, eds. Techniques in Arterial
Surgery. Philadelphia: WB Saunders,
1990:409.)

angioscopic examination. The following general principles

are important to achieve consistently good studies, maintain

safety, and avoid complications.

1 To avoid inducing spasm in the native artery or vein, choose
an angioscope for the procedure smaller than the lumen of
the smallest vessel to be intubated and pass it through these
vessels only in the presence of flowing blood or irrigation
fluid. Passage of an angioscope occupying almost the entire
lumen of the vessel or passage in a vessel emptied of all
blood or irrigation fluid may result in intense, irreversible
vasospasm.

2 To minimize the irrigation fluid volume and optimize the dura-
tion of angioscopic imaging: (i) perform angioscopy on
withdrawal whenever possible; (ii) during completion
angioscopy for bypass grafts, occlude the artery just proximal
to the distal anastomosis whenever possible —this reduces the
size of outflow tract as well as the likelihood of any blood
mingling with the clear column of irrigation fluid; and (iii)
whenever possible, reduce the retrograde flow or prevent it
from flowing into the clear fluid column. In bypass grafts,
on withdrawing the angioscope from the distal artery and
anastomosis and into the graft, occlude the distal end of the

428

CHAPTER 36 Angioscopy in peripheral vascular surgery

graft between the fingers or use a fine bulldog clamp to trap
the column of clear fluid within the graft. During thrombec-
tomy, external compression of the inflow or outflow vessels
may reduce the blood flow.

3 We perform irrigation most commonly through a separate
irrigation catheter or needle inserted collateral to the angioscope,
or through an irrigation sheath with a proximal hemostatic
valve, coaxial with the angioscope. Even with the dedicated
angioscopy pump, flow rates achieved through angio-
scopes with built-in irrigation channels usually are less
than 175ml/min, and are inadequate for most success-
ful angioscopic studies during infrainguinal bypass or
thrombectomy. In certain situations with limited blood flow,
such as vein conduit preparation, these flow rates may be
sufficient. 33 In our experience, angioscopes with an irriga-
tion channel are almost always too large and rigid to be
inserted through the distal anastomosis and into the distal
artery, particularly in bypass grafts distal to the popliteal
arteries.

4 Do not pass the angioscope through the native vessels or graft
unless flowing blood or normal vessel architecture is observed on
the video monitor. A "whiteout" of the image means that the
angioscope is abutting an obstruction, and insertion should
be halted and the angioscope withdrawn a few centimeters.
If the obstruction remains and freely flowing blood is not
seen, further insertion of the angioscope must be performed
under direct vision. This avoids injury to the vessels or the
anastomotic structures, even if it means using more irriga-
tion fluid, and prevents buckling of the optical fibers with
irreparable damage to the angioscope.

5 Visualize the entire lumen of the vessels being studied to
ensure completeness of the angioscopy study. This may be
achieved by using a steerable angioscope or by manipulat-
ing a nonsteerable angioscope. Steerable angioscopes
not only enhance the quality of the angioscopic study but
are much easier to manipulate intraluminally than the
standard, nonsteerable angioscope. Although useful in the
femoropopliteal and larger tibial vessels, the large size of
steerable angioscopes precludes their use in many of the
more distal bypass grafts. For the smaller nonsteerable
angioscopes, rolling or torquing the angioscope between
the plantar surfaces of the thumb and index finger allows
rotation of the angioscope and visualization of the entire
vessel circumference. Direct manipulation on the distal end
of the angioscope through the vessel wall is another method
of ensuring full visualization of the entire lumen and, in par-
ticular, the anastomosis. Coordinating these manipulations
is achieved best by watching the images produced on the
monitor and making adjustments of position accordingly,
and not by attempting to position the angioscope tip
directly within the anastomosis or lumen of the vessel or
graft. These techniques (Fig. 36.5) significantly enhance the
value of the studies and avoid missing relevant pathologic
findings.

Steerable angioscope

Torquing nonsteerable angioscope

External manipulation

Figure 36.5 Methods to enhance angioscopic visualization. (From Miller A,
Jepsen S. Angioscopy in arterial surgery. In: Bergan J, YaoJ, eds. Techniques in
Arterial Surgery. Philadelphia: WB Saunders, 1990:409.)

Interpretation

The value and reliability of the angioscopic examination is
enhanced by the interpretative skills and experience of the
endoscopist. These skills may be acquired from the review of
previous studies and findings, but are refined with the con-
tinued critical review and evaluation of the angioscopic find-
ings after each study. Many of the findings are new, subtle, and
of uncertain clinical significance. Careful follow-up and corre-
lation with the clinical course will eventually establish their
significance.

It is especially important to appreciate that the angioscope is
a qualitative instrument. The accurate assessment of size of an
angioscopic image on the video monitor remains problematic.
Magnification of the image changes with the distance of the
angioscope lens to an object; the closer the lens to the object, the
larger the image. 21 This makes much of the interpretation
of the angioscopic images subjective and the significance of
many of the more subtle endoluminal findings difficult to
assess, even with a large experience. Methods to quantify the
angioscopic image would substantially enhance the value of
angioscopy.

A fixed protocol for each angioscopic application is recom-
mended so that the findings may be recorded in a systematic
fashion, ultimately allowing clinicopathologic correlation.

429

pa rt 1 1 1 Invasive vascular diagnostics

Hood

Suture

Blood

Figure 36.6 Anatomy of end-to-side
anastomosis for systematic angioscopic
evaluation and interpretation. (From Miller A,
Stonebridge P, Kwolek C. The role of routine
angioscopy for infrainguinal bypass procedures.
In: Ahn S, Moore W, eds. Endovascular Surgery,
2nd edn. Philadelphia: WB Saunders, 1 992.)

For routine diagnostic completion angioscopy at the end
of an infrainguinal bypass procedure, we record the visual
quality and completeness of each study Incomplete studies
or failure to clear the blood completely from the visual field may
result in misinterpretation and missed significant findings.

The distal artery is graded as to the severity of disease —
minimal, moderate, or severe —taking into account the size of
the lumen, the regularity of the intimal surface, and the pres-
ence of protruding plaque. The anastomosis is observed sys-
tematically with particular attention to the size, shape, and
patency of the apex or outflow tract of the anastomosis; the
regularity of the suture lines, the hood, and arterial floor; the
presence of thrombus, its volume and site, whether fresh or or-
ganized, whether red or white (the latter composed mostly of
platelets); as well as the presence of intimal flaps, their size,
and degree of tethering and luminal obstruction (Fig. 36.6). For
vein grafts, the luminal size, the vein quality with regard to
surface characteristics such as sclerosis and presence of fresh
or organized thrombus, the number of valves and their con-
figuration, and any other abnormalities are noted. For in situ
and nonreversed bypasses, the number of valves and whether
they have been rendered incompetent, the number of unligat-
ed tributaries, and the presence of localized valvulotome
injury are recorded as well (Fig. 36.7; also in colour, see Plate 6,
facing p. 370).

Clinical applications

Indications

Our current indications for angioscopy are summarized in the
following list:

INDICATIONS FOR ANGIOSCOPY

Diagnostic

• Monitoring of surgical — interventional procedures

• Bypass, endarterectomy, thrombectomy, or embolectomy

• Angioplasty or atherectomy

• Clinicopathologic correlation

• Lesions responsible for anginal syndromes

• Endoscopic findings and graft failure

• Chronic pulmonary embolism
Therapeutic

• Surgical

• Endoluminal vein graft preparation (valvulotomy and
tributary occlusion)

• Catheter-directed thrombectomy or embolectomy

• Percutaneous

• Thrombolysis

• Assisted interventions (angioplasty, atherectomy,
stenting)

The usefulness and varied applications of angioscopy, how-
ever, depend in the main on the ingenuity and creativity of the
individual surgeon. We conceptualize and use the angioscope
as an adjunctive tool to see inside vessels, so that rational and
informed clinical and surgical decisions can be made based on
objective findings. We do not rely simply on experience.

Initially, we used angioscopy as a simple alternative or ad-
junct to the operative angiogram, as a means to avoid or cor-
rect technical errors. We soon appreciated, however, that the
rich and detailed endoluminal information not only provides
a sensitive and accurate method for the detection of technical
errors but allows continual assessment of technical profi-
ciency. It has become an excellent teaching tool, improving
and refining surgical technique of the surgeon and resident
staff. Angioscopy also has identified new or previously unap-
preciated endovascular pathologic conditions that have en-
hanced our understanding of the pathogenesis of graft
failure, 34 and fostered the development and design of new in-
strumentation for intraluminal manipulations such as valve
cutters, tributary occluders, and various grabbing and cutting
intraluminal instruments , 30 ' 33 ' 35 ' 36

430

■J

Lunan

™ thrombus

.^P Vwiyuft

^ CSmcfitha}

Figure 36.7 (A) Angioscopically directed valvulotomy using a modified
reversed Mills-type valvulotome allows accurate cutting of the valve leaflets.
(B) Valvulotome injury with furrowing and an intimal flap in a segment of
in situ saphenous vein following blind valvulotomy. (C) Dense webs in a
segment of recanalized saphenous vein. (D) Backbleeding into the clear
saline fluid column identifies an unligated tributary in an in situ saphenous
vein bypass graft. (E) Normal saphenodorsalis pedis anastomosis with no

technical deficits and clear visualization of the entire anastomosis. (F)
Abnormal saphenoanteriortibial artery anastomosis with an intimal flap
caught in the contralateral suture line and obstructing the lumen. (G) Patent
normal superficial femoral artery. The localized mural thrombus was present
before any endoluminal manipulations. (H) Residual mural thrombus
following a successful balloon thrombectomy of a failed 32-month-old
saphenous vein bypass graft. See also Plate 6, facing p. 370.

pa rt 1 1 1 Invasive vascular diagnostics

Clinical experience

Our clinical studies have been directed to the systematic eval-
uation of this new technology and its place in the practice of
vascular surgery Introducing a new technology into surgery
raises several questions. Is it feasible in a busy clinical setting
or is it really only a "research" tool? Is it safe? Does it facilitate
the particular surgical procedure? Does it provide new infor-
mation unavailable by standard or other technology? If so, is
this information useful? Does it influence and alter clinical and
operative surgical decisions, and, finally, how do these new
findings affect the results of the surgery as measured by clini-
cal outcome or graft patency, both in the short and long term?

Since performing our first clinical angioscopy on 1 May
1987, We have performed intraoperative angioscopy during
1207 revascularization procedures to February 1993 (Table
36.4), including infrainguinal bypass grafting, thrombectomy
or embolectomy, vascular access surgery, carotid endarterec-
tomy, and coronary artery bypass grafting.

Our largest experience is with angioscopy during infrain-
guinal bypass g ra fting, 28/29/34/37/38 and much of the discussion

Table 36.4 Details of 1 2 1 7 intraoperative angioscopies performed
between 1987 and 1993

Operation

Number

Infrainguinal bypass

1011

Femoropopliteal

240

Infrageniculate-pedal

771

Thrombectomy

Vascular access surgery

Carotid endarterectomy

Coronary artery bypass surgery

Miscellaneous

Total

1217

will focus on these studies. As mentioned, others have studied
the role of angioscopy during carotid endarterectomy, 15,39
thrombectomy and embolectomy, 25 ' 40,41 coronary artery by-
pass grafting, 19 ' 42 and venous valvular repair. 43 We are investi-
gating its application to vascular access surgery, 44 and our
preliminary results show findings and applications similar to
those in general vascular surgery.

As a clinical tool, angioscopy has proved most valuable and
clinically useful to us in graft monitoring, vein conduit prepa-
ration, and in reoperative surgery for the failing or failed by-
pass graft.

Monitoring infrainguinal bypass grafts

Our early studies showed that routine angioscopy to monitor
infrainguinal bypass grafting is feasible, safe, and a clearcut al-
ternative to routine intraoperative completion angiography,
except in cases where the runoff situation is not delineated on
the preoperative angiogram. In this situation, we often per-
form an adjunctive intraoperative angiogram confined to the
distal graft and runoff vasculature. Analysis of the first 355 an-
gioscopies during infrainguinal bypass grafting consolidated
our technique of intraoperative angioscopy, 28,29,32 and delin-
eated the most useful size of angioscope for a given procedure
and the volume of irrigation fluid required for consistently
high-quality studies (Table 36.5). Most important, these stud-
ies also showed that the angioscopic findings significantly
modified the process of intraoperative surgical decision mak-
ing (Table 36.6).

New or previously unappreciated findings on causes of
graft failure also came to light. These angioscopic findings in-
cluded segmental areas of previous thrombosis and recanal-
ization recognized angioscopically as webs, bands, or strands
(Fig. 36.8), vein stenosis or stricture, organizing nonoccluding
thrombus, or vein wall sclerosis. There also were findings re-
lated to surgical technique, with regard to residual competent
valves in the in situ and nonreversed vein graft configurations,

Table 36.5 Size of angioscope and mean volume of irrigation fluid used in 355 infrainguinal bypass operations

0.8-mm

1.4-mm

2.2-mm

2.8-mm

3.0-mm

Mean fluid volume

No.

(range in ml)

Femoral-AKP

—

458(64-1412)

Femoral-BKP

410(51-904)

Femoral-tibial

163

130

419(50-1098)

Femoral-pedal

—

433(77-1160)

Popliteal-distal

—

218(37-625)

Total

355

261

397(37-1412)

OD, outer diameter; AKP, above-knee popliteal; BKP, below-knee popliteal.

(From Miller A, Stonebridge P, KwolekC. The role of routine angioscopy for infrainguinal bypass procedures. In: Ahn S, Moore W, eds. E ndovascu la r Surgery, 2nd

edn. Philadelphia: WB Saunders, 1 992:66.)

432

CHAPTER 36 Angioscopy in peripheral vascular surgery

40(17.3%)*
65(32.9%)

2
23

Table 36.6 One hundred fifty-five clinical or surgical decisions in 355
infrainguinal bypass grafts

GRAFT

(in situ 1 97, NRV 93, RV 44, PTFE 8,

composite PTFE-vein 13)

Residual competent valve leaflets

Unligated tributaries

Graft torsion-inadequate graft tunneling

Recanalized vein

Vein discarded 3

Recanalized segment excised 6

Webs/bands/strands cut 14

Vein stenosis 3

Segment excised 2

Venoplasty 1

ANASTOMOSIS/DISTAL ARTERY

Revision of anastomosis 9

Technically inadequate 2

Postanastomotic stenosis 4

Initimal flap 3

Thrombectomy of distal artery 2

Primary siting of anastomosis 7

MISCELLANEOUS

Postoperative LMD/heparin for intraluminal

platelet thrombus 4

TOTAL 155

NRV, nonreversed; RV, reversed; PTFE, polytetrafluoroethylene; LMD,
low-molecular-weight dextran.

*Angioscopic valve lysis in 57 of 290 in situ and nonreversed vein grafts.
(From Miller A, Stonebridge P, KwolekC. The role of routine angioscopy for
infrainguinal bypass procedures. In: Ahn S, Moore W, eds. Endovascular
Surgery, 2nd edn. Philadelphia: WB Saunders, 1 992.)

or technical deficits in the anastomosis or distal runoff artery
(seeFig.36.7). 34

A marked difference in the prevalence of abnormal endolu-
minal findings was documented in the greater saphenous vein
(±12-20%) and in the veins of the arm (±60-70%) used for
bypass grafts. This difference suggests a distinct etiology for
abnormalities in the extremities, superficial thrombophlebitis
in the saphenous vein, and repeated trauma from phlebotomy
for blood analysis or intravenous infusions in the more easily
accessible veins of the upper extremity This hypothesis is sup-
ported from the distribution of intraluminal lesions found
in 113 arm veins harvested for infrainguinal bypass grafts
(Fig. 36.9), where the inaccessible basilic vein has an incidence
of intraluminal pathologic conditions similar to that of the
greater saphenous vein. 45

Vein conduit lesions in general, and especially the multiple
channels that result from this healing or recanalization
process, usually are undetectable at surgery by external in-
spection and palpation alone, even with the vein fully ex-
posed. Flushing with saline or blood is always possible unless
the vein is completely obstructed. Areas of recanalization may
cause early graft failure by acting as a filter and causing throm-
bosis, or cause late graft failure with the development of local-
ized vein stenosis and stricture.

Angioscopy is more sensitive and accurate in detecting and
delineating the various endoluminal pathologic conditions
than either the intraoperative angiogram, continuous-wave
Doppler examination, or duplex ultrasound. 46-48

In a series of arm vein conduits for infrainguinal bypass, we
showed that the angioscope accurately localizes the various
intraluminal lesions so that precise and directed interventions
to correct or eliminate them could be performed. 45,46 Such
interventions included excision of abnormal segments with
reanastomosis of the vein conduit when sufficient vein was
available, cutting fine bands with the valvulotome, or vein
patch angioplasty in the presence of stenosis. These "up-
graded" veins showed a patency at 1 year no different from
that of "normal" -quality arm vein grafts, but significantly dif-
ferent from "inferior" -quality vein grafts, where the intralu-
minal lesion could not be completely eliminated or corrected
(Fig. 36.10).

To determine whether use of the angioscope rather than the
gold-standard intraoperative angiogram for routine monitor-
ing of infrainguinal bypass grafting could improve early
(30-day) graft patency, we designed and performed such a
comparison in a prospective and randomized fashion. 38 Our
study was limited to primary bypass grafts using only autoge-
nous saphenous vein. All secondary bypasses and use of other
veins, such as arm vein, in which we had already shown a vast
improvement in early graft patency using the angioscope
rather than the angiogram, were excluded from this study. 45,46

The prospective randomization of the 293 patients for the
study resulted in well balanced groups for each of the moni-
toring modalities, but two significant, unforeseen biases
evolved during the study. The first was the preference of all the
participating surgeons for preparing the vein conduit with the
angioscope whenever they thought the quality of the vein was
in question. Of the 43 exclusions from the study after initial
randomization, 12 clinically assessed poor-quality veins were
excluded from the completion angiography group. In the an-
gioscopy group, such veins were prepared with the aid of the
angioscope and included in the study. The second bias to
evolve was the inclusion of a small group of 11 bypass grafts to
the plantar arteries of the foot. The patency for these 11 by-
passes to the plantar arteries was only 65%, compared with
95.4% at 30 days for the remaining 239 bypass grafts, reflecting
a poor runoff situation rather than a failure of either of the
monitoring techniques to detect technical errors. By chance,
more failures of this group occurred in the angioscopy group.

433

pa rt 1 1 1 Invasive vascular diagnostics

Figure 36.8 Saphenousvein partially occluded due to recanalized
organized thrombus. (A) Low-power photomicrograph of vein cross-section
demonstrating residual webs from recanalization of vessel. (B) High-power
photomicrograph demonstrating neovascularization with both large and
small blood channels, chronic inflammation, and hemosiderin pigment

indicative of organized thrombus. Both hematoxylin and eosin; Ax 20, Bx
150. (From Miller A, Jepsen S, Stonebridge Petal. New angioscopic findings
in graft failure after infra-inguinal bypass grafting. Arch Surg 1990;
125:749.)

Basilic Cephalic

A/ = 7/50(11.7%)

N = 0/2 (0%)

Arm

N = 28/78 (35.1%)

Median cubital

N = 16/48 (33.3%)

Forearm

W= 29/59(49.2%)

Figure 36.9 The incidence and distribution of the segmental lesion
detected with angioscopic preparation and monitoring of 1 1 3 arm veins
harvested for infrainguinal bypass conduits. (From Marcaccio E, Miller A,
Tannenbaum G etal. Angioscopically directed intervention upgrades arm
vein quality and i m proves ea rly graft patency. J Vase Surg 1993; 17:994.)

Table 36.7 Relevantfindings and clinical decisions resulting in 39
interventions in 36 bypass grafts during the completion monitoring of 250
infrainguinal bypass grafts

Interventions (n = 36 bypass grafts)

Vein conduit

Residual competent valve

Vein preparation-selection

Tributary ligation
Anastomosis
Distal artery
Total

*Two of these five findings were false-positive (see text).
(Modified from Miller A, Marcaccio E, Tannenbaum G etal. Comparison of
angioscopy and angiography for monitoring infrainguinal bypass grafts:
results of a prospective randomized trial. J Vase Surg 1993; 17:382.)

Anc

jioscopy

Angiography

Thus, although our study did not demonstrate a statistically
significant difference between the completion monitoring
techniques in early patency in this selected group of patients
with optimal-quality vein conduit, it did show a clear trend
favoring the angioscope as the preferable intraoperative
monitoring method (Fig. 36.11).

Perhaps the most important and least expected result of this
study was the paucity of findings in the completion angiogra-
phy group leading to subsequent surgical interventions —

only seven of 122 bypasses randomized to this group. Of the
seven findings, two were false positives resulting in unneces-
sary explorations of the distal anastomoses. In contrast, in the
angioscopy group, there were 32 findings in the 128 bypass
grafts that led to interventions, with no false-positive inter-
ventions. This difference was statistically significant (P <
0.0001; Table 36.7).

Review of the literature 29 ' 49-54 reveals a progressive im-
provement in the patency rates for infrainguinal bypass grafts,

434

CHAPTER 36 Angioscopy in peripheral vascular surgery

Figure 36.10 Comparison of primary graft
patency forthe 1 09 infrainguinal arm vein
bypass grafts as determined by life tables with
comparisons of "normal" vs. "upgraded" vs.
"inferior" quality arm vein grafts. (From
Marcaccio E, Miller A, Tannenbaum G etal.
Angioscopically directed intervention upgrades
arm vein quality and improves early graft
patency. J Vase Surg 1993; 17:994.)

100

-J?

80-

60-

20-

Normal
months

6
12

vs. Inferior
P-value
04
0002
00001
0001

Upgraded vs Inferior
months P-vaiue

1
3
6

0.02
0.0002
0.0003
01

Months

Normal (n=42)

Upgraded (n=47)

Inferior (n = 20)
1

Figure 36.11 Life table analysis to 1 month
comparing the proportions of primary graft
failure in three groups. (A) Angioscopy and
angiogram groups. The difference at 1 month is
not statistically significant (P= 0.2). (B) Eleven
bypasses to the plantar arteries and the
remaining 239 bypasses in the study. At 1
month the difference is statistically significant (P
= 0.04). (C) Angioscopy and angiogram groups
with the 1 1 plantar arteries excluded. The
difference at 1 month is statistically significant
(P = 0.03). (From Miller A, Marcaccio E,
Tannenbaum G etal. Comparison of angioscopy
and angiography for monitoring infrainguinal
bypass grafts: results of a prospective
randomized trial. J Vase Surg 1993; 17:382.)

CO
CL

100 H

90-

80 -J

100
90

■— ' 80
£ 70

a> eo

Angioscopy (n=128)
----Angiography (n=122)

50-
40-
30-

100 H

90-

Femoropopiiteal (n-65) --- - Femorodistal (n=99)
Pophteodistal (n=75) ——-Plantar (n=n)

Angioscopy (n=l22)
Angiography (n=117)

One month

104

P*0 20

■»- H*

■i

104

despite operations in the most severely threatened limbs and
elderly, ill patients, and extension of the grafts more distally in
the leg and foot. It appears that provided the conduit is of good
quality and the surgeon proficient in the surgical techniques,
good results are possible. In those bypasses where the vein
conduit is of good quality, the runoff vasculature adequate,
and the technique proficient, monitoring the surgery for tech-
nical or correctable errors by any means does not appear to
alter significantly the early graft patency. With a conduit of

less-than-optimal quality or a borderline runoff, however,
the role of monitoring in bypass surgery assumes a new
significance.

These studies clearly show that intraoperative angioscopy
is the most efficient way to monitor autogenous vein grafts, the
anastomoses, and adjacent distal artery for correctable intrin-
sic abnormalities or technical defects, additionally providing
a most effective way to ensure an optimally prepared graft
conduit.

435

pa rt 1 1 1 Invasive vascular diagnostics

70-
60 -

2 50

° 40

£ 30

5 20

ill

^^s

WW?

,-'%w/A

wmtm

In situ graft

non- reversed graft
residual competent valves

May '87-July *88 Aug '88-April '89 May '89-May '90

Figure 36.12 The incidence of retained
competent valves after "blind" retrograde
vavulotomy in vein graft preparation of the /n
s/ft/ and nonreversed bypass grafts during three
arbitrary time periods of data analysis: between
May 1 987 and July 1 988, 1 2/52 (23 %) /h s/fcv
and 2/31 (6.5%) nonreversed; between August

1988 and April 1989,9/61 (1 4.8%) in situ and
1/8 (12.5%) nonreversed; and between May

1 989 and May 1 990, 1 3/55 (23.6%) in situ and
3/26 (1 1 .5%) nonreversed. (From Miller A,
Stonebridge P, Kwolek C. The role of routine
angioscopy for infrainguinal bypass procedures.
In: Ahn S, Moore W, eds. Endovascular Surgery.
2nd ed. Philadelphia: WB Saunders, 1 992:68.)

Endoluminal vein conduit preparation of the in situ and
nonreversed vein

Despite a large and continuous experience with the technique
of blind retrograde valvulotomy using a Mills valvulotome,
the incidence of residual competent valve leaflets detected on
completion angioscopy in in situ and nonreversed veins
remained at approximately 20% and 6%, respectively (Fig.
36.12). Although complete disruption of the vein conduit wall
by engaging the tributary instead of the valve is a well
described complication of blind valvulotomy, 55 the high
incidence of intimal injury —almost 80% of vein conduits pre-
pared with the blind valvulotomy technique 30 — is not gener-
ally appreciated. The reasons for this high incidence of intimal
injury are related to the vein's normal endoluminal anatomy.
These injuries occur most commonly in the region of the valve
leaflets or related tributaries. Less commonly, injury is seen in
the intervalvular vein segment, extending as an intimal dissec-
tion from the cut that divided the leaflet, or as a furrow in the
intimal surface from blindly raking the vein with repeated up-
and-down movements of the valvulotome in an attempt to
engage a valve leaflet or to be sure that a valve leaflet was
not missed.

The bulbous dilation of the normal valve sinus forms a
"shoulder" at the base of the valve leaflet where the valvulo-
tome frequently engages, giving the characteristic tug that
confirms the cutting of a leaflet when the valvulotomy is done
blindly, by feel. With a sharp valvulotome, cutting the thin
valve leaflet can scarcely be appreciated. The presence of
empty sinuses, sinuses without valve leaflets, and those
valves with only one leaflet can be misleading, encouraging
the surgeon to continue to scrape the valvulotome along the
vein wall in an attempt to cut the nonexistent valves, mistak-
enly engaging the sinus shoulder without a valve leaflet and
increasing the likelihood of inducing intimal injury. With the
routine use of angioscopy to perform valvulotomy under
direct vision for the in situ and nonreversed vein, the problems

of residual competent valves and vein wall injury 30 ' 33 ' 56-58
have been almost completely eliminated.

Angioscopically directed valvulotomy is clearly a first step
in the complete endoluminal preparation of the in situ vein, in
which not only the valves are accurately rendered incompe-
tent but the tributaries occluded from within the vein.
Mehigan 33 has described the technique of limited incisions for
vein exposure and of ligating the identified tributaries
through multiple separate stab incisions after identifying the
location of the tributary with the angioscope's light. Others 59
have described similar techniques for closed or semiclosed
vein conduit preparation, with preoperative localization of
the main saphenous vein tributaries. We are conducting clini-
cal trials in the endoluminal occlusion of venous tributaries.

Preparation of the in situ vein by endoluminal methods is
very attractive. Not only would it provide the ideal conduit
and optimize the quality of the vein graft, but it also would
minimize the wound incisions and surgical trauma normally
required for vein graft preparation in the infrainguinal in situ
bypass graft, and thus significantly reduce patient morbidity
and medical costs.

Reoperative surgery for failed or failing grafts

In the reoperation of failing or failed infrainguinal bypass
grafts, angioscopy provides useful clinical information on the
endoluminal state of these grafts that otherwise would be un-
available to the surgeon, even with complete exposure of the
graft at surgery and a good preoperative angiogram. In many
instances, angioscopy not only accurately identifies the cause
of the graft failure, but localizes the problem and its extent,
thereby minimizing the surgical exposure necessary for the re-
vision surgery. In a retrospective study of 76 reoperations for
failing or failed infrainguinal bypass grafts, 34 of 76 (44%) of
the bypass operations required additional interventions based
on angioscopic examination, including removal of unsus-
pected thrombus, detection, localization, and correction of

436

CHAPTER 36 Angioscopy in peripheral vascular surgery

unsuspected pathologic conditions, and demonstration of us-
able graft despite nonvisualization on the preoperative an-
giogram. 60 In these complicated and challenging procedures,
angioscopy not only provides useful information in addition
to that of the preoperative angiogram, but may provide in-
sights into the pathogenesis of graft failure.

Conclusion

Endoscopy of the native vasculature and bypass grafts is in its
infancy. Technology is progressing with extraordinary speed
in the fields of imaging and manufacture of clinically useful
endoluminal microinstrumentation. Already, chip video
cameras as small as a few millimeters in diameter are being
developed and sited at the distal end of catheters, allowing all
information transfer to be performed electronically and in a
digital format, with vast improvement in the quality and the
resolution of the endoscopic image. Such advances make the
field of vascular endoscopy exciting, and portend a future of
vascular surgery and interventional techniques different from
those practiced today, in keeping with the trend toward mini-
mally invasive surgery.

References

1. Cutler E, Levine S, Beck C. The surgical treatment of mitral
stenosis: experimental and clinical studies. Arch Surg 1924;
9:689.

2. Allen D, Graham E. Intracardiac surgery: a new method. JAMA
1922; 79:10280.

3. Harken D, Glidden E. Experiments in intracardiac surgery. / Tho-
me Surg 1942; 11:566.

4. Murray G.Cardioscope.Angiology 1950; 1:334.

5. Butterworth R. A new operating cardioscope. / Thome Surg 1951;
22:319.

6. Bolton H, Bailey C, Costas-Durieux J, Gemeinhardt W. Car-
dioscopy: simple and practical. / Thorac Surg 1954; 27:323.

7. Bloomberg A, Hurwitt E. Endoscopy of the heart. Surg Clin North
Am 1957; 37:13374.

8. Carlens E, Silander T Method for direct inspection of the right
atrium: experimental investigation in the dog. Surgery 1961;
49:622.

9. Silander T Cardioscopy without thoracotomy. Acta Chir Scand
1964; 127:67.

10. Gamble W, Innis R. Experimental intracardiac visualization.
NEngl J Med 1967; 276:13972.

11. Crispin H, Van Baarle A. Intravascular observation and surgery
using the flexible fibrescope. Lancet 1973; 1:750.

12. Hirschowitz B, Peters C, Curtiss L. Preliminary report on a long
fiberscope for examination of stomach and duodenum. Univ Mich
Med Bull 1957; 23:178.

13. Greenstone S, Shore J, Heringman E, Massell T Arterial
endoscopy (arterioscopy). Arch Surg 1966; 93:811.

14. Vollmar J, Storz L. Vascular endoscopy. Surg Clin North Am 1974;
54:111.

15. Towne J, Bernhard V Vascular endoscopy: an adjunct to carotid
surgery. Stroke 1977; 8:569.

16. Towne J, Bernhard V. Vascular endoscopy: useful tool or interest-
ing toy. Surgery 1977; 82:415.

17. Towne J, Bernhard V Technique of intraoperative endoscopic
evaluation of occluded aortofemoral grafts following thrombec-
tomy. Surg Gynecol Obstet 1979; 148:87.

18. Shure D, Gregoratos G, Moser K. Fiberoptic angioscopy: role in the
diagnosis of chronic pulmonary artery obstruction. Ann Intern
Med 1985; 103:844.

19. Grundfest W, Litvack F, Sherman T et ah Delineation of peripheral
and coronary detail by intraoperative angioscopy. Ann Surg 1985;
202:394.

20. Sherman T, Litvack F, Grundfest W et ah Coronary angioscopy in
patients with unstable angina pectoris. NEngl] Med 1986; 315:913.

21. Spears R, Marais H, Serur J et ah In vivo coronary angioscopy. } Am
CollCardioll983;5:131U.

22. Grundfest W, Litvack F, Glick D et ah Intraoperative decisions
based on angioscopy in peripheral vascular surgery. Circulation
1985;78(Suppl.I):13.

23. Seeger J, Abela G. Angioscopy as an adjunct to arterial reconstruc-
tive surgery: a preliminary report. / Vase Surg 1986; 4:315.

24. Mehigan J, Olcott C. Videoangioscopy as an alternative to intra-
operative arteriography. Am J Surg 1986; 152:139.

25. White G, White R, Kopchok B, Wilson S. Angioscopic thrombe-
ctomy: preliminary observations on a recent technique. / Vase Surg
1988; 7:318.

26. Silverman S, Mladinich C, Hawkins I, Abela G, Seeger J. The use of
carbon dioxide gas to displace flowing blood during angioscopy.
} Vase Surg 1989; 10:313.

27. Miller A, Lipson W. Isaacsohn J, Schoen F, Lees R. Intraoperative
angioscopy: principles of irrigation and description of a new
dedicated irrigation pump. Am Heart J 1989; 118:391.

28. Miller A, Campbell D, Gibbons G et ah Routine intraoperative
angioscopy in lower extremity revascularization. Arch Surg 1989;
124:604.

29. Miller A, Stonebridge P, Jepsen S et ah Continued experience with
intraoperative angioscopy for monitoring infrainguinal bypass
grafting. Surgery 1991; 109:286.

30. Miller A, Stonebridge P, Tsoukas A et ah Angioscopically directed
valvulotomy: a new valvulotome and technique. / Vase Surg 1991;
13:813.

31. Kwolek C, Miller A, Stonebridge P et ah Safety of saline irrigation
for angioscopy: results of a prospective randomized trial. Ann
Vase Surg 1992; 6:62.

32. Miller A, Jepsen S. Angioscopy in arterial surgery. In: Bergan J, Yao
J, eds. Techniques in Arterial Surgery. Philadelphia: WB Saunders,
1990:409.

33. Mehigan JT. Angioscopic preparation of the in situ saphenous
vein for arterial bypass: technical considerations. In: White G,
White R, eds. Angioscopy: Vascular and Coronary Applications.
Chicago: Year Book Medical Publishers, 1989:72.

34. Miller A, Jepsen S, Stonebridge Fetal. New angioscopic findings in
graft failure after infra-inguinal bypass grafting. Arch Surg 1990;
125:749.

35. Stierli P, Aeberhard P. Angioscopy-guided semiclosed technique
for in situ bypass with a novel flushing valvulotome: early results.
} Vase Surg 1992; 15:546.

36. White G, White R, Kopock G, Colman P, Wilson S. Endoscopic

437

pa rt 1 1 1 Invasive vascular diagnostics

intravascular surgery removes intraluminal flaps, dissections,
and thrombus. / Vase Surg 1990; 11:280.

37. Miller A, Stonebridge P, Kwolek C. The role of routine angioscopy
for infrainguinal bypass procedures. In: Ahn S, Moore W, eds.
Endovascular Surgery, 2nd edn. Philadelphia: WB Saunders,
1992:58.

38. Miller A, Marcaccio E, Tannenbaum G et al. Comparison of
angioscopy and angiography for monitoring infrainguinal
bypass grafts: results of a prospective randomized trial. / Vase Surg
1993; 17:382.

39. Mehigan JT, DeCampli WM. Angioscopic control of carotid
endarterectomy In: Ahn S, Moore W, eds. Endovascular Surgery,
2nd edn. Philadelphia: WB Saunders, 1992:102.

40. White G, Kopchok G, White R. Current role of intraoperative
angioscopy for monitoring peripheral vascular surgery. Semin
Vase Surg 1989; 2:60.

41. Segalowitz J, Grundfest W, Treiman R et al. Angioscopy for intra-
operative management of thromboembolectomy. Arch Surg 1990;
125:13572.

42. Forrester J, Litvack F, Grundfest W, Hickey A. A perspective of
coronary disease seen through the arteries of living man. Circula-
tion 1987; 75:505.

43. Welch HJ, McLaughlin RL, O'Donnell TF. Femoral vein valvulo-
plasty: intraoperative angioscopic evaluation and hemodynamic
improvement. / Vase Surg 1992; 16:694.

44. Miller A, Marcaccio E, Goodman W, Gottlieb M. The role of
angioscopy in vascular access surgery. In: Ferguson R, Henry M,
eds. Proceedings of Symposium: Dialysis Access III. WL Gore & Asso-
ciates and Precept Press, 1993:210.

45. Marcaccio E, Miller A, Tannenbaum G et al. Angioscopically
directed intervention upgrades arm vein quality and improves
early graft patency. / Vase Surg 1993; 77:994.

46. Stonebridge P, Miller A, Tsoukas A et al. Angioscopy of arm vein
infrainguinal bypass grafts. Ann Vase Surg 1991; 5:170.

47. Baxter B, Rizzo R, Flinn W et al. A comparative study of intra-
operative angioscopy and completion arteriography following
femorodistal bypass. Arch Surg 1990; 125:9972.

48. Gilbertson J, Walsh D, Zwolak R et al. A blinded comparison of
arteriography, angioscopy, and duplex scanning in the intraop-
erative evaluation of in situ saphenous vein bypass grafts. / Vase
Surg 1992; 15:121.

49. Veith F, Moss C, Daly V et al. New approaches in limb salvage by
extended extraanatomic bypasses and prosthetic reconstructions
to foot arteries. Surgery 1978; 84:764.

50. Logerfo W, Gibbons G, Pomposelli F et al. Evolving trends in the
management of the diabetic foot. Arch Surg 1992; 127:617.

51. Taylor L, Edwards J, Phinney E, Porter J. Reversed vein bypass to
infrapopliteal arteries. Ann Surg 1987; 205:90.

52. Donaldson M, Mannick J, Whittemore A. Causes of primary graft
failure after in situ saphenous vein bypass grafting. / Vase Surg
1992; 15:113.

53. Leather R, Shah D, Chang B, Kaufman J. Resurrection of the in situ
vein bypass: 1000 cases later. Ann Surg 1988; 208:435.

54. Taylor L, Edwards J, Porter J. Present status of reversed vein
bypass: five year results of a modern series. / Vase Surg 1990;
11:207.

55. Corson J, Karmody A, Shah D, Leather R. Retrograde valve inci-
sion for in situ vein-arterial bypass utilising a valvulotome. Ann R
Coll Surg Engl 1984; 66:173.

56. Fleisher HI, Thompson B, McCowan T et al. Angioscopically
monitored saphenous vein valvulotomy. / Vase Surg 1986; 4:360.

57. Matsumoto T, Yang Y, Hashizume M. Direct vision valvu-
lotomy for nonreversed vein graft. Surg Gynecol Obstet 1987;
165:181.

58. Chin A, Fogarty T. Specialized techniques of angioscopic valvulo-
tomy for in situ vein bypass. In: White G, White R, eds. Angioscopy:
Vascular and Coronary Applications. Chicago: Year Book Medical
Publishers, 1989:76.

59. La Muraglia G, Cambria R, Brewster D, Abbott W. Angioscopy
facilitates a closed technique for in-situ vein bypass. / Vase Surg
1990; 12:601.

60. Holzenbein T, Miller A, Tannenbaum G etal. (Abstract). Presented
at the Peripheral Vascular Surgery Society Meeting, Washington,
DC, June 6, 1993.

438

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Medical management

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Atherosclerosis: risk factors and
medical management

Ralph G. DePalma
Virginia W. Hayes

At the beginning of the 21st century a variety of scientifically
based hypotheses and technical advances promise effective
prophylactic and therapeutic treatments for atherosclerosis.
Foremost among these are understanding of the pivotal role of
lipid abnormalities in pathogenesis, effects of treatment on
plaques and clinical outcomes, comprehension of the roles of
inflammatory and immune responses in this process, identifi-
cation of novel risk factors for late progression independent of
blood lipid levels, understanding of growth factors, particu-
larly in diabetes, and advances in magnetic resonance and ul-
trasound imaging that more clearly identify and characterize
plaques vulnerable to rupture, thrombosis, or downstream
embolization. Finally, many prospective randomized trials
using drugs, micronutrients, and other interventions for pri-
mary and secondary prevention continue to appear. This
chapter considers risk factors for atherosclerosis and its
prevention as well as new concepts of treatment particularly
for patients with peripheral arterial disease (PAD).

Risk factors

A risk factor is an individual characteristic that increases the
likelihood (risk) for development of the manifestations of dis-
ease. 1 Risk factors are derived from epidemiologic and case
studies, and the relationship between a risk factor and disease
is present no matter what population or study technique is
used. A strong association exists between the risk factor and
disease development; substantially more disease will occur in
patients who exhibit the risk factor than in those in whom it is
absent. Note, however, that a risk factor might not be an etio-
logic factor: for example, although lifestyle might predispose
to viral or bacterial infections, the etiologic agent is a virus or
bacterium.

A risk factor must exhibit biologic plausibility; an increase in
the exposure to the risk factor should cause an increase in the
incidence or intensity of disease. The presence of the risk factor
precedes the disease and reduction of risk factor exposure
causes a reduction in the risk for development of the disease.

In virtually all studies of coronary disease risk factor reduc-
tion, 2 lowering serum levels of low-density lipoprotein cho-
lesterol (LDLC) has been shown for more than a decade to
effect arrest or regression of established disease.

The intensity, number, and duration of specific risk factors
predict clinical complications of atherosclerosis: myocardial
infarction, stroke, and peripheral ischemia. Both coronary
artery disease (CAD) and PAD are associated with each other
and with risk factors for atherosclerosis. Risk factors may be
irreversible or potentially reversible abnormalities. Some are
controversial, such as personality traits and certain forms of
obesity. Because events associated with CAD, such as myocar-
dial infarction, are dramatic, considerable emphasis has been
placed on the relationship between risk factor reduction and
reduction of coronary events, with angiographic evidence of
regression of coronary artery plaques. 3-6 One study 7 indicated
that aggressive risk factor modification retarded progression
of symptomatic PAD consistent with observations of angio-
graphic plaque regression in response to lowered cholesterol
and cessation of smoking. 8 Regression is not regarded as a
common phenomenon as remodeling of the arterial wall
might accompany atherogenesis leading to less lesion intru-
sion on angiographic imaging. Stabilization of lipid-laden
plaques with lipid egress is now postulated to be more likely. 9

Irreversible risk factors associated with atherosclerosis
are genetic traits; aging and gender (male > 45 years; females
> 55 years) might relate to iron accumulation. 10 Potentially
reversible risk factors include cigarette smoking, diabetes
mellitus, hypertension, a variety of lipid abnormalities,
obesity, sedentary lifestyle, fibrinogen and hemostatic fac-
tors, and inflammation. Patterns of atherosclerosis vary with
risk factors, 11 e.g. infracrural involvement in diabetics and in
African- Americans with hypertension; smoking and hyper-
lipidemia are associated with aortic involvement. With coex-
isting multiple risk factors these anatomic patterns overlap.
Increased blood total cholesterol (TC) and increased LDLC
are associated with an increased risk for CAD. 2 Lowering of
TC and LDLC reduces this risk. High levels of high-density
lipoprotein cholesterol (HDLC) are also associated with low-

441

part iv Medical management

ered coronary heart disease (CHD) risk. 12 With PAD in partic-
ular, low levels of HDLC and high levels of lipoprotein a
[Lp(a)] are associated with claudication and PAD 12 ' 13 as well as
other recently described novel risk factors. 14 Lp(a) is com-
posed of a low-density lipoprotein (LDL) and a glycoprotein
called apolipoprotein(a) [apo(a)]. 15 The LDL and the apo(a)
are connected by a disulfide bridge. Lp(a) probably blocks
plasminogen's action in stimulating clot lysis. Modification of
Lp(a) in PAD has proved difficult in some conditions, par-
ticularly in renal failure. Low HDLC is a hallmark of PAD, and
systematic means of increasing HDLC remain challenging.

Irreversible risk factors

Aging is an irreversible risk factor, reflecting the culmination
of etiologic factors interacting over time. Atherosclerosis,
however, is not an inevitable consequence of aging. Complica-
tions of atherosclerosis may appear in the elderly in western
societies long after exposure to risk factors such as hyperlipi-
demia or smoking. 16 Complications relate to local processes
including endothelial dysfunction, thrombosis, deterioration
of elastin and collagen, and inflammatory responses, 14 which
contribute to plaque instability. Age and gender interact in
unique ways with environmental risk factors. For example,
men and women, in addition to differences in hormonal mi-
lieu and iron metabolism, differ in their responses to dietary
fat and cholesterol. 17 Women tend to respond to increased fat
intake with a greater rise of HDL, whereas men, especially
older men, tend to transport excess cholesterol preferentially
as LDL. In men, body fatness tends to modify the relationship
between dietary cholesterol and LDLC; leaner men are more
responsive than fatter men to increased dietary cholesterol
by increasing their concentration of LDLC. 18 In general, the
younger the patient, the more likely one is to encounter risk
factors that coincide with overt clinical events. Patients pre-
senting with PAD are usually a decade older than those pre-
senting solely with CAD. In patients presenting with CAD,
PAD is not commonly a clinical complaint, while virtually all
patients with PAD have CAD.

Genetic disorders predispose to lipid abnormalities. One of
these, familial hypercholesterolemia type 2, results from a lack
of LDL receptors on hepatocytes, causing an inability to inter-
nalize and metabolize LDL within the liver. 19-21 Cholesterol
elevation ranges above 600 mg/dl in homozygotes. These
individuals rarely survive beyond their third decade. The
cause of death is coronary atherosclerosis related directly to
hyperlipidemia. These rare conditions have been treated by
portacaval shunting or liver transplantation. Almost every
aspect of lipid transport and lipoprotein metabolism exhibits
genetic effects; 22 for example, familial hypercholesterolemia is
characterized by autosomal dominant disorders produced
by at least 12 different molecular defects of the LDL receptor.
Genetic factors also determine Lp(a) concentrations, which
are continuous and left-skewed in the white population, but

bell-shaped in black populations, where Lp(a) isoforms are
much higher. 15

Reversible risk factors

Hyperlipidemia

Based on extensive experimental and population data and ob-
servations of regression and arrested atherosclerosis in ani-
mals and humans, reduction of blood levels of TC, LDLC, and
triglycerides (TG) by diet and drugs is recommended. 23 Re-
duction of total and LDLC has had substantial effects in de-
creasing coronary events. Epidemiological and interventional
studies initially supported a view that serum TC values in the
general population should be below 200 mg/dl, with LDL
below 130 mg/dl, fasting triglycerides below 250 mg/dl, and
HDL levels above 40 mg/dl. These values applied to older age
groups as well. 24 Diet and exercise can be effective in favorably
influencing serum lipid levels, but these results are usually
modest. As will be discussed later, a case has been made for re-
duction of TC to 150 mg/dl and LDLC to below lOOmg/ml;
these newer recommendations imply that many more people
will be treated with drugs.

One simple approach involves screening of asymp-
tomatic individuals for total cholesterol on at least two
occasions in an ambulatory nonfasting state. If TC is greater
than 200 mg/dl, then fasting blood samples are obtained.
This allows calculation of LDLC as follows: LDLC = TC -
HDLC + TG/5. When LDLC is above 130 mg/dl, the first
step is dietary treatment. A total intake of fat of less than
30% of the total caloric intake, with less than 10% of total
calories derived from saturated fat and less than 300 mg of
cholesterol a day, has been recommended. If dietary therapy
for 3-6 months failed to achieve the goals for lipid lower-
ing (i.e. an LDL level below 130 mg/dl), drug therapy was
recommended.

Drugs available for lipid reduction in the United States
include 3-hydroxy-3-methylglutaryl coenzyme A reductase
inhibitors or statins, including lovastatin, pravastatin, sim-
vastatin, atorvastatin, and fluvastatin. Cerivastatin or
Baychol R has been removed from the market after being
linked to fatal outcomes due to rhabdomyolysis. Other agents
include cholestyramine and colestipol, bile acid sequestrants,
gemfibrozil, a fibric acid derivative, nicotinic acid and probu-
col. Previous intervention studies with drugs and ileal by-
pass 3-7 had proven effects in decreasing CAD risk, with serial
angiography showing that with lipid reduction, plaque pro-
gression is slowed or halted. Regression was seen in some
cases, and has also been described in patients after intra-
venous total parenteral nutrition. 25 An interesting observation
relating dietary total fat intake to progression of coronary
disease was obtained in a placebo trial group scrutinized by
serial angiography. Among nonsmoking men aged 40-
49 years, increased dietary fat consumption was associated

442

chapter 37 Atherosclerosis: risk factors and medical management

with a significant increase in the risk of developing new
coronary lesions. 26

When serum lipids were abnormally elevated, serum lipid
reduction provided a dimension of prevention of CAD in pa-
tients with PAD. Now, as will be discussed, most patients with
PAD have become candidates for drug treatment. Screening
PAD patients for lipoprotein abnormalities can be useful in
individualizing treatment. Blood should be obtained from
nonhospitalized, stable ambulatory subjects before disease
events, angiography, or surgery. Several reports 13 ' 27 demon-
strated increased levels of triglycerides or very low-density
lipoprotein (VLDL), increased LDL and apoprotein B (apo B),
and decreased levels of HDL in PAD. In one study of plasma
lipids, lipoproteins, and apoproteins in patients with intermit-
tent claudication and healthy, controlled subjects matched for
age, sex, and smoking habits, significant increases in total cho-
lesterol, triglycerides, VLDL cholesterol, VLDL triglycerides,
LDLC and decreased HDLC characterized PAD. 28 A prominent
feature, as mentioned, was elevation of Lp(a) in patients
with intermittent claudication compared with controls.
Thus Lp(a) concentration emerges as an important risk factor
for PAD independent of smoking, hypertension, diabetes
mellitus, elevated lipids, increased apo B, and low HDL
concentrations. The importance of VLDL, chylomicrons, and
postprandial lipemia in contributing to generalized athero-
sclerosis has been emphasized. 29

Clearly, the links between lipid abnormalities and athero-
sclerosis have been established so that patients with sympto-
matic atherosclerosis exhibiting coexisting lipid abnormalities
thus become universal candidates for aggressive lipid-
lowering therapy with modification of other concurrent risk
factors. Roberts 30 suggested that only one true risk (or etiolog-
ic) factor exists for CAD, that is the lifetime presence of serum
TC over 150 mg/dl. It is not clear, however, that such low
levels of cholesterol can be sustained in western society
even when consuming a prudent diet. This level of TC cor-
responds to epidemiologic observations as well as to observed
thresholds for plaque regression or progression in animal ex-
periments 31,32 using direct sequential observations. Recently,
in a study of patients with atherosclerotic disease, 33 the
efficacy of statin therapy has been demonstrated even when
lipids were not elevated. This prospective randomized
placebo-controlled trial with simvastatin demonstrated,
for the first time, in over 20 000 high-risk individuals, not only
a reduction in coronary events and deaths, but a reduction in
all-cause mortality.

Cigarette smoking

Smoking is one of the most powerful risk factors promoting
complications of atherosclerosis. Cigarette smoking relates di-
rectly to the progression of peripheral atherosclerosis to am-
putation, 34 high mortality rates for ischemic heart disease, 35
and failure of aortic 36 ' 37 and femoropopliteal grafts. 38 Cessa-

tion of smoking yields powerful and immediate benefits, par-
ticularly after any type of direct vascular intervention. The
means by which cigarette smoking promotes atherosclerosis,
graft thrombosis, and disease progression are incompletely
understood. Smoking is associated with low HDL levels. 39
Carbon monoxidemia predisposes to arterial wall injury and
also by producing increased plasma flux, with entry of LDL
and other proteins. Smoking causes increased platelet reactiv-
ity and peripheral vasoconstriction. From the standpoint of
pathogenesis, lipid abnormalities are emphasized because of
their known etiologic role and interactions with the arterial
wall. But from the standpoint of immediately effective clinical
interventions, smoking cessation is critical and more immedi-
ate in its effects than lipid interventions.

Diabetes

Among those patients with PAD who are nonsmokers, many
will be diabetic. Type 2 diabetes in an early stage, e.g. in the
morbidly obese, is potentially reversible by bariatric surgery 40
Diabetes causes increased death rates from CAD, as well as in-
creased PAD. Bierman 41 summarized the factors potentiating
atherosclerosis in diabetic patients. Epidemiologically, the
contribution of all of the commonly measured risk factors in
diabetic patients could account for no more than about 25% of
the excess CAD incidence. For every cholesterol level, diabetic
patients have a threefold to fivefold higher CHD mortality
rate. 42 Control of diabetes and progression of macrovascular
disease have not been well correlated, while diabetic mi-
crovascular manifestations are closely linked to tight glucose
control. Recently complications due to both types of vascular
involvement have been linked to control in terms of HbA lc . 43
Bierman noted, "The unique effects of hyperglycemia,
mediated through the mechanism of protein glycation and
glycooxidation . . . are in need of further focus/' 41 Controversy
persists about tight diabetic control and macrovascular dis-
ease progression. Data on overall complications seen from the
recent study of type 2 diabetes for endpoints diverged 43 from
the 1998 study 44 of noninsulin-dependent diabetes mellitus
showing that macrovascular complications were not
associated with glycemic control.

Diabetics exhibit several particular lipid abnormalities.
These include chylomicronemia, increased VLDL, increased
VLDL and chylomicron remnants, and triglyceride-rich LDL
and HDL concentrations. Mamo and Proctor 29 have em-
phasized the pathogenicity of these remnants. In insulin-
dependent diabetes, unique changes in lipoproteins might
increase the atherogenicity of most of these moieties even
when the serum lipid concentrations are similar to those of
nondiabetic subjects. 45

Diabetes also increases blood coagulation through several
mechanisms. Increased glucose levels are associated with ac-
celerated platelet aggregation in vitro. Hypertriglyceridemia
is associated with an increase in clotting activities of thrombo-

443

part iv Medical management

genie factors such as factors VII and X, 46 and in many diabetic
patients with proteinuria Lp(a) is increased. The concentration
of tissue plasminogen activator inhibitor is also reduced. 41 In-
sulin resistance and hyperinsulinemia are important issues for
increased coagulability. Insulin resistance and hyperinsuline-
mia play central roles in atherogenesis, particularly where in-
creased visceral abdominal adiposity correlates with insulin
resistance and compensatory hyperinsulinemia. 47 Upper ab-
dominal adiposity is associated with diabetes and cardiovas-
cular risk factors. Direct actions of insulin on the arterial wall
might include the promotion of arterial smooth muscle cell
proliferation and cholesterol ester accumulation by increasing
LDL delivery and biosynthesis. It has been shown that insulin
and insulin growth factor 1 downregulate HDL receptors
in human fibroblasts, possibly decreasing HDL receptor-
mediated cholesterol efflux. 48

Glycosylation of lipoproteins and collagen relates directly
to hyperglycemia and disease progression through several
possible mechanisms, including increased binding of LDL by
collagen. Another mechanism is lipoprotein oxidation, which
is promoted by a number of factors in diabetes; LDL does not
accumulate in its native form, but oxidized LDL is taken up
avidly by macrophages present in the vascular wall to trans-
form them into foam cells. 49 These interactions demonstrate
the importance of diabetes as an etiologic factor affecting the
arterial wall. The importance of insulin and glucose as growth
or mitogenic factors and factors promoting endothelial dys-
function has been noted. 50 These concepts as related to treat-
ment will be considered further.

Hypertension

Autopsy data show that atherosclerosis in the aorta, coronary
arteries, cerebral arteries, and other major vessels is more ex-
tensive and more severe in hypertensive than innormotensive
subjects. Prospective studies demonstrate, at least for affluent
populations, that hypertension is related to the risk for prema-
ture atherosclerotic disease independently of other major risk
factors such as hypercholesterolemia and cigarette smoking, 51
while in the elderly hypertension is associated with risk factor
clustering such as glucose intolerance, hyperinsulinemia, and
dyslipidemia promoted by abdominal obesity. 52

Hypertension predisposes to continued hemodynamic in-
jury, thus accelerating plaque complications and growth. Con-
trol of hypertension has been demonstrated to prolong life and
reduce coronary mortality. Treatment of hypertension with
thiazide diuretics, however, was thought to be disadvanta-
geous in a subgroup of men in the Mr. Fitt Trial. 53 Oriental and
Caribbean populations, however, in the presence of low lipids
exhibit hypertension with low prevalence of atherosclerotic
disease.

Obesity and sedentary lifestyle

Increased dietary fat intake and body fat are associated with an
increased risk for CHD. Results from the Chicago Western
Electric Study showed, many years ago, that people with
lower intakes of dietary cholesterol and lower body fat had the
lowest 25-year risk of coronary death. 18 These results were
unique in showing that fatter men apparently did not benefit
from a diet lower in cholesterol, whereas men who consumed
a diet very high in cholesterol apparently did not benefit from
leanness. Overall, maintenance of an ideal body weight and
avoiding excess caloric intake appear to reduce the risk of
coronary disease. The abdominal distribution of fat previous-
ly mentioned is of interest here. Obesity, hyperinsulinemia, el-
evated triglyceride levels, and low HDLC levels predispose
older obese men to CAD and atherosclerosis. 54 The fat distrib-
ution is characteristic, with male upper abdominal obesity
prominent compared with distribution of fat about the hips
and thighs seen more frequently in women. Upper abdominal
obesity is associated with abnormal postheparin plasma
lipoprotein lipase and hepatic lipase activities. Elevated fib-
rinogen level, also an independent risk factor for atherosclero-
sis, correlates with upper abdominal obesity. 55

Obesity is a major consequence of sedentary lifestyle cou-
pled with inappropriate dietary habits. Exercise contributes to
maintenance of proper body weight and probably exerts ef-
fects on atherosclerosis by influencing serum lipids favorably.
Effective regular exercise decreases serum TC, decreases fast-
ing triglycerides, and increases HDLC. No study of exercise,
however, demonstrates direct regressive effects on atheroscle-
rotic plaques, as do the extensive experimental and clinical
data that show plaque arrest or regression with threshold
serum lipid reduction.

Beneficial effects of exercise with increased walking dis-
tance relate mainly to improved skeletal muscle oxidative
phosphorylation, 56 and not collateral development insofar as
can be determined. For the patient with PAD, exercise is an im-
portant therapeutic alternative and an appropriate interven-
tion. This single step can be more effective than surgical or
endovascular interventions. 57 In patients with CAD, exercise
prescriptions must be carefully structured. Sudden coronary
events can cause death in apparently fit people. Patients with
known CAD require cardiac monitoring during periods of
high-stress aerobic exercises. Exercise is insufficient to offset
the adverse effects of elevated TC and LDL levels, nor should
exercise be considered a replacement for treatment of hyper-
tension. Weight loss and drugs are more effective. 58 Before
performing strenuous exercise, potentially atherosclerotic in-
dividuals should have stress testing, recognizing that this is
not absolutely predictive of sudden coronary events due to
plaque disruption. Recently, exercise has been shown to re-
duce C-reactive protein (CRP) levels suggesting a favorable ef-
fect on this novel risk factor, a surrogate for inflammation. 59
The importance of the inflammatory responses in atheroscle-

444

chapter 37 Atherosclerosis: risk factors and medical management

rosis will be considered subsequently. Inflammation is an im-
portant and recently delineated link to the complications of
atherosclerosis over and above dyslipidemia.

Fibrinogen and hemostatic factors

Although risk factor interventions have focused on choles-
terol and lipoprotein metabolism, the endpoint of epidemio-
logic studies of atherosclerosis usually involves a thrombotic
episode. 60 In a prospective study 61 of hemostatic function and
cardiovascular death, in addition to elevated plasma choles-
terol levels, elevated levels of factors VIIc, VIIIc, and particu-
larly fibrinogen were noted. Elevated fibrinogen 62 is a major
CAD risk factor and also highly significantly associated with
PAD in men. 15 Leukocyte counts may be elevated in PAD and
in CAD. 63 Hemostatic risk factors have been underempha-
sized as predictive risk factors for atherosclerotic events and
probably undertreated. Platelet levels are not elevated, but in
some studies of PAD, platelet survival seems to be decreased.
Interventions using chronic low-dose aspirin administration
might be more clearly related to decreased thrombotic
episodes than to prevention or arrest of plaque growth.

Novel risk factors: C-reactive protein, cytokines,
and inflammation

Elevated levels of the inflammatory cytokine interleukin
(IL)-6 have been reported to be associated with an increased
risk of future myocardial infarctions in apparently healthy
men, 64 while elevated levels of tumor necrosis factor (TNF)-oc
after myocardial infarction predicted subsequent myocardial
events. 65 Furthermore, a high CRP level, using the high
sensitivity method, has been determined to predict cardio-
vascular disease events. 66 Systemic and local effects of
infection, notably Chlamydia, cytomegalovirus, Helicobacter
pylori, and herpes virus have been postulated to play a role in
the pathogeneses of atherosclerosis. The deleterious effects of
infection presumably relate to inflammation and bacterial
heat shock proteins that produce inflammatory and auto-
immune reactions in the vascular system. 67 ' 68 A causal link
between infection, inflammatory biomarkers, and outcomes
remains to be established by prospective human antibiotic
trials. 69

Iron accumulation

Iron accumulation might predispose men and post-
menopausal women to atherosclerosis. Recent negative infor-
mation about hormone replacement for postmenopausal
women 70 tends to reinforce this provocative idea. Female hor-
mones, after all, might not be protective as had been thought
previously; menstrual bleeding might be protective. Since
Sullivan 71 first proposed in 1981 that iron accumulation might
be a risk factor for heart disease, this hypothesis has been de-

bated vigorously on epidemiological grounds. 72-75 Conflict-
ing opinions exist on possible beneficial effects of blood dona-
tion in preventing coronary events. 76-78 The recognition of
increasing ferritin levels with age in men and after menopause
in women has resulted in removal of iron supplementation of
flour in Denmark. 79 A Veterans Affairs cooperative study, The
Iron (Fe) and Atherosclerosis Study (FeAST), has accrued sub-
jects with stable atherosclerosis to test the iron accumulation
hypothesis. Primary endpoints include death, cardiovascular
events, and need for surgical interventions.

The relationship between dyslipidemia and atherosclerosis
might be regarded as more descriptive of disease severity long
term, while inflammation appears more predictive of short-
term clinical events. Early correction of dyslipidemia reduces
disease severity and improves long-term outcomes, but it re-
mains to be shown whether or not alteration of newly de-
scribed or novel risk factors by reduction of inflammation
might suppress immediate adverse clinical outcomes. Surgi-
cal interventions offer proven benefits when unstable plaques,
i.e. in the carotids, cause symptoms. Tables 37.1 and 37.2,
adapted from Pearson et al., 23 summarize guidelines and goals
for primary prevention applicable to ongoing modifiable risk
factors. Note that the typical patient with PAD, in addition to
established disease, generally exhibits more than two risks.
The subsequent section on secondary prevention outlines es-
tablished and novel risk factors along with standard, experi-
mental, and evolving approaches to treatment.

Medical management: secondary prevention

For patients with PAD presenting as stable claudicants, major
threats to survival are myocardial infarction and stroke.
Amputation occurs in only a minority of patients receiving
effective medical treatment, particularly when smoking is
relinquished. All patients with PAD should be considered can-
didates for secondary prevention. For those with PAD relative
risk of dying from cardiovascular disease was shown to be 5.9
[95% confidence interval (CI) 3.0, 11.4] and for death from
coronary disease was 6.6 (95% CI 2.9, 14.9). 80 As mentioned
previously, 33 a recent randomized, placebo-controlled trial of
simvastatin showed reduction of adverse cardiovascular out-
comes when used for primary and secondary prevention in
patients without abnormally elevated lipid levels. The antiin-
flammatory effect of statins has been documented by de-
creased CRP levels, 81 independent of LDL reduction. 82 The
results obtained with cerivastatin, now off the market, have
also been achieved with pravastatin. 83 Overall, many more in-
dividuals with additional risk factors are now candidates for
treatment to achieve recently revised National Cholesterol
Education Program III goals 84 of total cholesterol < 150 mg/dl
and LDL < 100 mg/dl. It appears unlikely that such stringent
target levels will be achieved in most individuals with dietary
measures alone.

445

Table 37.1 Treatment of modifiable risk factors

Modifiable risks and goals

Interventions

Risk: hyperlipidemia
Primary goals:

1. LDL-C < 160 mg/dl for < 1 RF present

2. LDL-C < 1 30 mg/dl for > 2 RF and 1 0-year CHD risk < 20%; or

3. LVDL-C<100 mg/dl if >2 RF+ 10-yearCHD is>20% orif
patient has diabetes.

Secondary goals (if LDL-C at goal):

For triglycerides > 200 mg/dl, usenon-HDL-C as secondary goal:

1. Non-HDL-C < 190 mg/dl for < 1 RF

2. Non-HDL-C < 1 60 mg/dl for > 2 RFs and 1 0-year CHD risk
is< 20%

3. Non-HDL-C < 130 mg/dl for diabetics or for > 2 RFand
10-year CHD risk > 20%

Other target goals:

Triglycerides > 1 50 mg/dl
HDL-C <40 mg/dl in men

< 50 mg/dl in women

Risk: cigarette smoking

Goals: Cessation and avoidance of second-hand smoke

Risk: diabetes mellitus
Goals:

1 . Normal fasting plasma glucose < 1 1 mg/dl and near
normal hemoglobin A 1c < 7%

2. BP< 130/80 mmHg

3. LDL-C to 100 mg/dl

4. For dysproteinuria: albuminuria < 1 mg/24 h

Risk: hypertension
Goal:

1. BP< 140/90 mmHg

2. BP < 1 30/85 mmHg in chronic renal insufficiency (CRI) or
chronic heart failure (CHF)

3. BP< 130/80 mmHg in diabetes mellitus (DM)

Risk: abdominal obesity

Goal: body mass index (BMI) 1 8.8-24.9 kg/m 2

If BMI > 25 kg/m 2 , use waist circumference at iliac crest level:

Men<102cm(<40in)

Women < 88 cm (< 35 in)

Risk: sedentary

Goal: At least 30 min of moderate-intensity physical activity

on most (preferably all) days of the week

Risk: fibrinogen/hemostatic factors
Goal:

1. No thrombolic or ischemic event

2. Normal sinus rhythm or if chronic atrial fibrillation,
anticoagulation with INR 2.0-3.0 (target 2.5)

Determine lipoprotein profile after a 9-1 2-h fast:

1 . If LDL-C is > goal range, recommend additional therapeutic lifestyle changes (TLC)
(Table 37.2)

• Rule out secondary causes (refer to primary care provider for work-up to include liver
function test; thyroid-stimulating hormone level, urinalysis)

2. After 12 weeks of TLC, consider LDL-lowering therapy if > 2 RFs are present, 10-year risk
is < 1 0% and LDL-C is > 1 90 mg/dl

• Start drugs and advance dose to bring LDL-C to goal range, usually a statin but a bile
acid-binding resin or niacin may also be considered

• If LDL-C goal not achieved, consider combination therapy (statin + resin or statin +
niacin)

3. After reaching LDL-C goal, assess triglyceride level:

• If 150-1 99 mg/dl, treat with TLC

• If 200-499 mg/dl, treat elevated non-HDL-C with TLC and if necessary, consider higher
doses of statins or adding niacin or fibrate

• If > 500 mg/dl, treat with fibrate or niacin to reduce risk of pancreatitis

4. If HDL-C is < 40 mg/dl in men or < 50 mg/dl in women, intensify lifestyle changes

5. For higher risk patients, consider drugsthat raise HDL-C (i.e. niacin, fibrates, statins)*

* Absolute contraindications for use of medications

1 . Instruct patient on effect of continued smoking on arteries

2. Provide referral to formal program which includes counseling and pharmacotherapy

3. Encourage avoidance of environmental tobacco smoke

1 st step: diet and exercise

2nd step: oral hypoglycemic drugs (sulfonylureas and/or metformin with ancillary use of

acarbose and thiazolidinediones)
3rd step: insulin

Consider use of statin drug and/or ACE inhibitor

1 . BP > 1 30 mmHg systolic or > 80 mmHg diastolic, initiate therapeutic lifestyle changes
(TLC). See Table 37.2

• ForBP> 140/90 mmHg prescribe drug if 6-12 months of TLC ineffective (dependent
on number of risk factors)

• Add BP medications individualized to individual requirements, ie. age, race, need for
drugs with specific benefits (ACE inhibitor, (3-blocker)

2. Drug therapy in CRI or CHF if systolic BP> 1 30 or diastolic > 85 mmHg

3. Drug therapy in DM if systolic BP> 130 mmHg or diastolics 80 mmHg
See Table 37.2: TLC program for diet, exercise, and weight management

See Table 37.2: TLC program

Before initiating vigorous exercise program, consult physician if cardiovascular, respiratory,
metabolic, orthopedic, or neurological disorders are suspected, or if patient is middle-aged
or older and is sedentary

1 . Prescribe antiplatelet medication:

• Aspirin 75-1 60 mg/day is as effective as higher doses and/or

• Clopidogrel 75 mg/day

2. For chronic or intermittent atrial fibrillation, use warfarin (aspirin can be used
alternatively for persons with certain contraindications)

BP, Blood pressure; CHD, coronary heart disease; LDL-C, low-density lipoprotein cholesterol, HDL-C, high-density lipoprotein cholesterol; RF, risk factor; TLC,

therapeutic lifestyle changes; INR, international normalized ratio.

^Absolute contraindications for use:

Statins: active or chronic liver disease (relative contraindications: concomitant use of certain drugs).

Bile acid sequestrants: dysbetalipoproteinemia; TG > 400 mg/dl (relative: TG > 200 mg/dl).

Nicotinic acid: chronic liver disease; severe gout (relative: diabetes, hyperuricemia, peptic ulcer disease).

Fibric acids: severe renal disease, severe hepatic disease.

Aspirin: allergy, gastrointestinal bleed, hemorrhagic stroke.

Adapted from Pearson etal 23

446

chapter 37 Atherosclerosis: risk factors and medical management

Table 37.2 Therapeutic lifestyle changes (TLC)

Diet:

• Advocate consumption of a variety of fruits, vegetables, grains, low-fat or nonfat dairy products, fish, legumes, poultry, and lean meats

• Modify food choices to reduce saturated fats (< 1 0% of calories), cholesterol (< 300 mg/dl), and trans-fatty acids by substituting grains and unsaturated
fatty acids from fish, vegetables, legumes, and nuts

• Limitsaltto<6g/day

• Limit alcohol consumption for those who drink to < 2 oz/day for men and < 1 oz/day for women

• For weight reduction: low-sodium and low-fat diet with consumption of fruit, vegetables

Additional TLC diet as therapeutic option to enhance LDL reduction:

• Saturated fat < 7% of calories

• Cholesterol < 200 mg/day

• Consider increased soluble fiber (1 0-25 g/day) and/or plant stanols/sterols (max, 2 g/day)

Exercise:

• Before initiating vigorous exercise program, consult physician if cardiovascular, respiratory, metabolic, orthopedic, or neurological disorders are suspected,
or if patient is middle-aged or older and is sedentary

• Moderate-intensity activities [40-60% of maximum capacity are equivalent to a brisk walk (1 5-20 min per mile)]

• Additional benefits are gained from vigorous-intensity activity (> 60% of maximum capacity for 20-40 min walk on 3-5 days/week)

• Recommend resistance training with 8-1 different exercises, 1 -2 sets per exercise, and 1 0-1 5 repetitions at moderate intensity > 3 days/week

• Flexibility training and an increase in lifestyle activities should complement this regimen. Exercise with systolic BP> 130 mmHgordiastolic>80 mmHg
(consult physician if comorbidities present or middle-aged or older and sedentary)

Weight management:

• Initiate weight-management program through caloric restriction and increased caloric expenditure as appropriate

• For overweight/obese persons, reduce body weight by 10% in the first year of therapy

Adapted from Pearson etal

The role of inflammation in the early 85 and later 86 stages of
atherosclerosis has received increased attention. Inflamma-
tion probably predisposes to progression and complications
independently of blood lipid levels. Recently recognized
genetic variants of toll-like receptors confer differences in
the inflammatory responses elicited by bacterial lipopolysac-
charides 87 and individuals capable of mounting brisk
proinflammatory responses maybe more susceptible to ather-
osclerosis. The inflammatory cascade includes the interaction
of proinflammatory and antiinflammatory cytokines within
the arterial wall. 88-90 Lipid accumulation appears to attract in-
flammatory cells that produce cytokines locally, which can be
detected systemically. Furthermore, an elevated level of a par-
ticular cytokine, such as TNF-oc, also affects the arterial wall.
The atherosclerotic plaque contains leukocytes of which
approximately 8% are monocytes or monocyte-derived
macrophages, while lymphocytes, predominantly memory T
cells, 91 comprise 5-20% of this cell population. Inflammation
may predispose to plaque vulnerability promoting sudden
expansion, rupture, and release of distal emboli prompting
vascular occlusion.

Trials using aspirin, which is also antiinflammatory, have re-
duced long-term adverse outcomes when used as primary
prevention. 92 In the case of aspirin, uncertainty remains about
recommended dosages (see Table 37.1). Aspirin in doses of
81 mg daily have been found to reduce serum levels of lipopro-

tein^), and this dose is probably most practical 93 for most
uses. Large doses of aspirin and dipyridimol were found to ac-
celerate atherosclerosis in a rhesus model of atherosclerosis; 94
and in extrapolating these results, smaller rather than larger
doses of aspirin and avoidance of dipyridimol appear desir-
able. Though clopidogrel is more effective than aspirin in
long-term secondary prevention, 95 cost may be an issue.
Hiatt 96 suggested that aspirin and /or clopidogrel are the anti-
platelet agents of choice for secondary prevention in patients
with PAD.

Angiotensin converter enzyme (ACE) inhibitors have im-
proved the primary endpoints of myocardial infarction,
stroke, or death from cardiovascular causes; a salient example
is the recent HOPE study 97 PAD patients have now become
potential candidates for treatment using three classes of drugs:
aspirin or clopidogrel; statins, which require monitoring of
liver, muscle, and peripheral nerve function; and ACE in-
hibitors requiring monitoring of blood urea nitrogen and
creatinine. Individuals with cardiac arrhythmias or those
undergoing cardiac or noncardiac surgery receiving (3 block-
ade 98 ' 99 exhibited small but consistent reductions in periopera-
tive mortality. Those with an ejection fraction of less than 30%
might not benefit and further studies are needed.

Vitamins, mainly C, E, and to a lesser extent vitamin A, have
been tested in prospective trials and the results in primary and
secondary prevention of myocardial infarction have been

447

part iv Medical management

found to be equivocal. More frequently than not, trials fail to
demonstrate positive impact of these agents on disease
processes, as in the most recent MRC/BHF trial of supplemen-
tation in over 20 000 high-risk individuals with antioxidant vi-
tamins. 100 Uncertainty about these results 101-103 has led to calls
for more trials stratified to include measurement of pro-
oxidant substrates in the subjects under test. Clearly, vitamin E
appeared effective in animal models, 104 supporting the oxida-
tive stress theory 49 Yet the oxidative stress theory itself has
been called into question by a recent study that shows foam
cell formation by activated macrophages in vitro when incu-
bated with native LDL. 105 Niacin in extended release form 106
should be considered a drug rather than a vitamin. It reduces
LDL and triglycerides and increases HDL. Niacin is poorly tol-
erated by about 10% of patients due to flushing. Titration with
gradually increasing doses and the use of aspirin 0.5 h prior to
dosing can mitigate the unpleasant flushing effects of niacin.
Whether used alone or with statins, monitoring of hepatic
function is required.

Homocysteinemia probably comprises an issue in superim-
posed blood clotting on an atherosclerotic substrate. Treat-
ment with large doses of folic acid (mg rather than jig
amounts), vitamin B6 and vitamin B12 has been studied in
patients with PAD 107 and after coronary angioplasty. 108 Al-
though primary outcome measures are not dramatically im-
proved, treatment of homocysteinemia, at the least, will do no
harm. Further studies are required.

Medical management 109 of claudication using cilostazol or
pentoxyphylline can improve walking distance, but does not
relate, so far as is known, to progression of atherosclerosis.
The enhanced activity may be of benefit. In prescribing
exercise for PAD, the clinician must differentiate between
preventive and therapeutic prescriptions. A recent review 110
summarizes an inverse linear dose-response between amount
of physical activity and all-cause mortality, cardiovascular
disease, CAD incidence, and death. The minimal effective
dose of exercise is unclear, but moderate activity, walking as
briskly as possible for 30-60 min on most days is sensible for
most individuals.

Smoking cessation requires reemphasis and more energetic
intervention strategies than have been employed to date. 111 All
clinicians treating atherosclerosis recognize that atherosclero-
sis and its complications progress inexorably with continued
cigarette smoking. Thrombosis occurs through the effects of
smoking on fibrinogen and increased blood coagulation. 112
Smoking also promotes abnormal immune and inflammatory
responses. 113 Much more than casual advice is needed to help
patients conquer this addiction. Casual advice by physicians
yields quit rates at 1 year in the low single digits. A Formal Pro-
gram combining counseling with drug treatment to achieve
adequate nicotine levels during withdrawal along with
buproprion 114 to treat depression is recommended. Bupropri-
on promotes prosexual side-effects, which can be important in
counteracting depression 115 and in substituting one pleasure

for another. Curiously, smoking cessation seems more difficult
for women, 116 who may need additional counseling. Vascular
surgeons can achieve success when advice to stop smoking,
given in a sympathetic and supportive way, accompanies op-
erative intervention. Casual advice to stop or "cut down" no
longer suffices and a coordinated proactive smoking cessation
program should be a part of every clinic treating vascular
disease.

Dietary advice remains surprisingly controversial in its
basic recommendations for proportions and types of fats, car-
bohydrates, and protein. Ongoing debates about proportions
and types of fat, carbohydrate, and protein have attracted sen-
sational publicity. As mentioned in risk factors, a diet low in
saturated fats, sugar, and refined carbohydrates, and calorical-
ly balanced to activity is desirable. Ideally, diets could be tai-
lored to specific metabolic individual characteristics and will
vary from individual to individual with an ultimate goal of
normalizing lipoprotein values and preventing obesity.
Unique lipoprotein patterns also need consideration. 117 Di-
etary trials using nonpharmacologic approaches have re-
duced adverse cardiovascular outcomes but the results have
not been as dramatic as those in which drugs were used. 118 On
the other hand, if diet is effective, potentially serious side-ef-
fects of drugs, although not common in data so far analyzed,
might be avoided.

The issue of high-fat diets and wine consumption, the so-
called "French Paradox" of lower incidence of CAD, has been
attributed to Gallic wine consumption; 119 either red or white
wine is supposed to be effective. 120 In offering dietary advice
in the past, 121 remarkable lipid reductions were seen in a few
highly motivated individuals without drug therapy. It is much
more difficult to obtain dietary compliance than to have
patients accept recommendations for major operative inter-
ventions; 122 consultation from an informed dietary service is
useful. That diet influences atherogenesis can be inferred from
epidemiology; however, evidenced-based prescription for di-
etary treatment of established atherosclerosis is no easy task.

Dietary supplements are constantly advertised. These
include garlic in various forms, native, pills or powders,
psyllium and fiber plans, cereals, phytosterols and cholestatin,
Benecol used in margarine, tocotrienols such as found in rice
and barley, Cholestatin, a dietary supplement, chromium pi-
colinate carnitine, and coenzyme Q. Others too numerous to
mention exist and certainly others will appear. All require not
only inferences of biologic plausibility, but also tests of con-
trolled trials.

New concepts about diabetes may affect future treatment,
particularly for patients with macrovascular disease. 50 High
levels of insulin and glucose promote smooth growth of cells
harvested from the infragenicular arteries; thiamine may
inhibit such growth. 123 Insulin-enhancing drugs such as
metformin (contraindicated in heart or renal disorder) or
glitazones, which improve impaired vasoreactivity 124 and fa-
vorably alter insulin resistance and inflammatory markers, 125

448

chapter 37 Atherosclerosis: risk factors and medical management

will probably be more often used for type 2 diabetes. The
efficacy of islet transplantation in type 1 diabetes has been
verified experimentally, but treatment requires chronic
immune suppression.

New and experimental treatment strategies

Phlebotomy is being tested based upon the iron accumulation
hypothesis. 10 ' 71 High levels of stored iron, in synergy with
smoking and dyslipidemia, possibly facilitate lipid peroxida-
tion and inflammatory responses associated with disease pro-
gression. Ferrous iron is a potent oxidizing agent capable of
promoting release of free radicals. The Veterans Administra-
tion Cooperative "Iron (Fe) and Atherosclerosis Study
(FeAST)," a single-blind randomized prospective trial, is ex-
amining the hypothesis that reduction in total body iron stores
by phlebotomy to a theoretically optimal serum ferritin of
25 ng/ml (approximating levels found in healthy menstruat-
ing females) will ameliorate the course of atherosclerosis in
subjects with stable PAD. Measured phlebotomy in subjects
with ferritin levels exceeding 25 ng/ml has been shown to be
safe using the calculation (ferritin -25) x 10 = ml blood donat-
ed, with an upper ferritin limit of 400 ng/ml. 126 Hematocrit
and hemoglobin do not fall as stored iron mobilizes to main-
tain their levels. Preliminary results from a FeAST substudy
showed an inflammatory signature in PAD 127 consisting of
elevated levels of TNF-oc, IL-6, and CRP, and reduced levels
of antiinflammatory IL-10. Measured phlebotomy reduced
inflammatory cytokine levels in high outliers and in nonsmok-
ers. Clinical outcomes of iron depletion using phlebotomy will
be determined by primary endpoints of death and cardiovas-
cular events when the trial is completed. It would be remark-
able to find that bleeding, a maligned relic of medicine's dark
ages, might under certain circumstances be shown to be of
benefit.

Another novel intervention includes treating possible infec-
tion, in the main, Chlamydia, since its DNA or organism is
found in 20-70% of plaques; seropositivity relates to disease
severity also including cytomegalovirus; and plaque T lym-
phocytes respond briskly to CLY antigen. 67-69 Trials using an-
tibiotics are in progress. A report of 325 patients with acute
myocardial infarction or unstable angina randomized to re-
ceive two antibiotic regimens (amoxicillin/metronidazole/
omeprazol and azitromycin/ metronidazole /amoxicillin)
showed a 36% reduction in the endpoints of death and read-
mission at 12 weeks and 1 year irrespective of antibiotics
used. 128 Note that certain of these drugs are considered antiin-
flammatory but in the STAMINA trial 128 the reduction in end-
points did not correlate with reduction in the inflammatory
marker CRP. On the other hand, doxycycline derivatives
under test for possible arrest of aneurysm growth 129 have
shown reduction in the biomarker MMP, but as yet no arrest of
aneurysm growth. Overall, in summarizing results currently

available, further large-scale trials of the infection-
atherosclerosis link are needed to define efficacy and mecha-
nisms, particularly with respect to Chlamydia. l3Q As with
bleeding, a role for infection in atherosclerosis, similar to
that found for peptic ulcer, would radically revise future
management strategies.

Gene therapy for PAD has received attention, though most
of these treatments are not designed to affect the atherosclerot-
ic process itself. This concept proposes that new blood vessel
collateral growth might be stimulated by the use of direct ap-
plication of growth factors. 131 These include fibroblast 132 and
vascular endothelial growth factors (VEGF). The delivery of
these growth factors includes direct gene application of DNA
by a viral vector using intravascular injection, or by balloons
or stents, and by direct application. Transfer of DNA coding
for angiogenesis remains an experimental approach requiring
clinical trials of safety and efficacy; 133 but little evidence exists
to support the theory that administration of angiogenic
growth factors might stimulate neoplasm growth or retinal
neoangiogenesis in diabetics. A variation on this theme is re-
cently reported therapeutic angiography in patients with se-
vere limb ischemia injecting autologous bone marrow cells. 134
In this study both ankle brachial indices and angiography
were interpreted as showing striking improvement.

Experimental concepts on the horizon include creation of a
vaccine against cholesterol esterase transfer protein (CETi-1
vaccine). 135 The vaccine, by lowering CETP, raises HDL and
lowers LDL and reduces atherosclerosis in rabbits. Another
experimental agent 136 now in human trial 137 also aims at low-
ering CETP. Genetic transfer of LDL receptors using a recom-
binant adenoviral vector to Watanabe rabbits reduced LDL
and atherosclerosis in this animal equivalent of human famil-
ial hypercholesterolemia. 138 However, trials with recombinant
adenoviral vectors have been performed for over 10 years, 139
and to date results applicable to human atherosclerosis have
not surfaced. Adeno-associated viruses from Rhesus mon-
keys have also been suggested as vectors for human gene
therapy, 140 and safety remains an issue with the use of such
vectors. The intramuscular injection of naked plasmid-DNA
encoding fibroblast growth factor type 1, an Escherichia coli de-
rivative, for end-stage unreconstructable ischemia has been
described to increase skin circulation without the need to use a
viral vector. 141

Summary

This chapter reviews accepted and new thinking about risk
factors and treatment of atherosclerosis. Secondary treatment
aims to prevent progression and complications. Stable claudi-
cants with inf rainguinal disease and those receiving carotid or
aortic reconstructions for occlusive and aneurysmal disease
will benefit from secondary prevention. Currently accepted
interventions in PAD include: (i) cessation of smoking; (ii) as-

449

part iv Medical management

pirin, preferably 81mg/day, or clopidogrel; (iii) reduction of
TC below 150 mg/dl, LDLC to below 100 mg/dl, and TG
below 150 mg/dl, goals possibly achieved with diet alone, but
more readily accomplished with statins, which also offer anti-
inflammatory benefits; (iv) walking briskly and as far as possi-
ble for 30-60 min daily A recent randomized trial supports
the addition of ACE inhibitors, particularly in diabetics
even when blood pressure is normal. Further advances in
this rapidly evolving field will require continued attention of
vascular practitioners.

Note

The opinions expressed in this chapter are those of the authors.

References

1. Susser M. Causal Thinking in the Health Sciences. New York:
Oxford University Press, 1973.

2. Rossouw JE, Rifkin BM. Does lowering serum cholesterol levels
lower coronary heart disease risk? Endocrinol Metab Clin North
Am 1990; 19:279.

3. Arntzenius AC, Kromhout D, Barth JD et ah Diet, lipoproteins
and the progression of coronary atherosclerosis: the Leiden
Intervention Trial. N Engl J Med 1985; 312:805.

4. Nikkila, E A, Viikinkowski P, Valle M et ah Prevention of coronary
atherosclerosis by treatment of hyperlipidemia: a seven-year
prospective angiographic study. Br Med J 1984; 289:220.

5. Blankenhorn DH, Nessim SA, Johnson RL et ah Beneficial
effects of combined colestipol-niacin therapy on coronary
atherosclerosis and coronary venous bypass grafts. JAMA 1987;
257:3233.

6. Buchwald H, Varco RL, Matts JP et ah Effect of partial ileal bypass
surgery on mortality and morbidity from coronary heart disease
in patients with hypercholesterolemia. N Engl J Med 1990;
323:946.

7. Duffield RGM, Miller NE, Brunt JNH et ah Treatment of
hyperlipidemia retards progression of symptomatic femoral
atherosclerosis: a randomized controlled trial. Lancet 1983;
2:639.

8. DePalma RG. Control and regression of atherosclerotic plaques:
a commentary. In: Dale WA, ed. Management of Arterial Occlusive
Disease. Chicago: Year Book Medical Publishers, 1971:63.

9. Fuster V. Plaque stabilization: present and future trends. In:
Fuster V, ed. The Vulnerable Atherosclerotic Plaque: Understanding,
Identification and Modification. Armonk, NY: Futura, 1999:393.

10. Howes PS, Zacharski LR, Sullivan J, Chow B. Role of stored iron
in atherosclerosis. / Vase Nurs 2000; 18:109.

11. DePalma RG. Patterns of peripheral atherosclerosis: implica-
tions for treatment. In: Shepherd J, ed. Atherosclerosis — Develop-
ments, Complications and Treatment. Amsterdam: Elsevier Science
Publishers BV, 1987:161.

12. Gordon DJ, Rifkind BM. High-density lipoprotein — the clinical
implications of recent studies. N Engl J Med 1989; 321:1311.

13. Seeger FM, Silverman SH, Flynn TC. Lipid risk factors in patients
requiring arterial reconstruction. / Vase Surg 1989; 10:418.

14. Ridker PM, Stampfer MJ, Rifai N. Novel risk factors for systemic
atherosclerosis. JAMA 2001 : 285:2481 .

15. Molgaard J, Klausen IC, Lassvik C et ah Significant association
between low-molecular-weight apolipoprotein (a) isoforms and
intermittent claudication. Arterioscler Thromb 1992; 12:895.

16. Rosen AB, DePalma RG, Victor Y Risk factors in peripheral
atherosclerosis. Arch Surg 1973; 107:303.

17. Clifton PM, Nestel PF. Influence of gender, body mass index, and
age on response of plasma lipids to dietary fat plus cholesterol.
Arterioscler Thromb 1992; 12:955.

18. Goff DC, Shelvelle, RB, Katan MB et ah Does body fatness modify
the association between dietary cholesterol and risk of coronary
death? Results from the Chicago Western Electric Study. Arte-
rioscler Thromb 1992; 12:755.

19. Goldstein JL, Brown MS. Familial hypercholesterolemia: identi-
fication of a defect in the regulation of 3 hydroxyl-3 methyl
glutaryl coenzyme A reductase activity associated with overpro-
duction of cholesterol. Proc Natl Acad Sci USA 1973; 70:2804.

20. Goldstein JL, Brown MS. Lipoprotein receptors, cholesterol me-
tabolism, and atherosclerosis. Arch Pathol 1975; 99:181.

21. Brown MS, Goldstein JL. Lipoprotein receptors in the liver: con-
trol signals for plasma cholesterol traffic. / Clin Invest 1983;
72:743.

22. Schonfeld G. Inherited disorders of lipid transport. Endocrinol
Metab Clin North Am 1990; 19:229.

23. Pearson TA, Blair SN, Daniels et ah AHA Guidelines for Primary
Prevention of Cardiovascular Disease and Stroke: 2002 Update.
Circulation 2002; 106:388.

24. Luepker RV. Dyslipoproteinemia in the elderly. Special consider-
ations. Endocrinol Metab Clin North Am 1990; 19:451.

25. Dudrick SJ. Regression of atherosclerosis by intravenous infu-
sion of specific biochemical nutrient substrates in animals and
humans. Am Surg 1987; 206:296.

26. Blankenhorn DH, Johnson RD, Mack WJ et ah The influence
of diet on the appearance of new lesions in human coronary
arteries. JAMA 1990; 263:1646.

27. Kaliman P, Widhalm K, Strobl W. Serum lipoproteins in patients
with clinical signs of atherosclerosis: peripheral vascular and
coronary heart disease. Artery 1980; 8:547.

28. Marrarino E, Siepi D, Pasqualini PL. The apoprotein pattern in
normolipemic peripheral vascular disease. Angiology 1988; 39:
555.

29. Mamo CL, Proctor SD. Chylomicron remnants and atherosclero-
sis. In: Barter PJ, Rye KA, eds. Plasma Lipids and their Role in Dis-
ease. Amsterdam: Harwood Academic Publishers, 1999:109.

30. Roberts WC. Atherosclerotic risk factors: are there ten or is there
only one? Am J Cardiol 1989; 64:552.

31. DePalma RG, Klein L, Bellon EM, Koletsky S. Regression of
atherosclerotic plaques in rhesus monkeys. Angiographic,
morphologic, and angiochemical changes. Arch Surg 1980;
115:1268.

32. DePalma RG, Koletsky S, Bellon EM, Insull W Jr. Failure of
regression of atherosclerosis in dogs with moderate
cholesterolemia. Atherosclerosis 1977; 27:297.

33. Heart Protection Study Group. MRC/BHF Heart Protection
Study of cholesterol lowering with simvastatin in 20,536 high-
risk individuals: a randomized placebo controlled trial. Lancet
2002; 360:7.

34. Juergens JL, Barker NW, Hines EA. Arteriosclerosis obliterans: a

450

chapter 37 Atherosclerosis: risk factors and medical management

review of 520 cases with special reference to pathogenic and
prognostic factors. Circulation 1960; 21:188.

35. Gordon T, Castelli WP, Hjortand MC, Kannel WB, Dawber TR.
Predicting coronary heart disease in middle-aged and older
persons: the Framingham Study. JAMA 1977; 238:497.

36. Wray R, DePalma RG, Hubay C A. Late occlusion of aortofemoral
bypass grafts: influence of cigarette smoking. Surgery 1971;
70:969.

37. Robicsek F, Daugherty HK, Mullen DC. The effect of con-
tinued cigarette smoking on the patency of synthetic vascular
grafts in Leriche syndrome. / Thome Cardiovasc Surg 1975; 70:
107.

38. Ameli FM, Stein M, Prosser RJ, Provan JL, Aro L. Effects of ciga-
rette smoking on outcome of femoropopliteal bypass for limb
salvage. ] Cardiovasc Surg 1989; 30:591.

39. Garrison RJ, Kannel WB, Feinleib M et ah Cigarette smoking and
HDL cholesterol: the Framingham offspring study. Atherosclero-
sis 1978; 30:17.

40. Sugarman HJ. Bariatric surgery for severe obesity. / Assoc Acad
Minor Phys 2001; 12:129.

41. Bierman EL. George Lyman Duff Memorial Lecture. Atherogen-
esis in diabetes. Arterioscler Thromb 1992; 12:647.

42. Pyorala K, Laakso M, Uusitupa M. Diabetes and atherosclerosis:
an epidemiological view. Diabetes Metab Rev 1987; 3:463.

43. Stratton IM, Adler AI, Neil HA et al. Association of glycaemia
with macrovascular and microvascular complications of type 2
diabetes (UKPDS 35): Br Med J 2000} 321:405.

44. Metabolic control and prevalent cardiovascular disease in
non-insulin dependent diabetes mellitus: The NIDDM Patient
Outcome Research Team. Am J Med 1997; 102:38.

45. Winocour PH, Durrington PV, Bhatnagar D et al. Abnormalities
of VLDL, IDL and LDL characterize insulin-dependent diabetes
mellitus. Arterioscler Thromb 1992; 12:920.

46. Skartlien AHJ, Lyberg-Beckmann S, Holme I et al. Effect of alter-
ation in triglyceride levels on factor VH-phospholipid complex-
es in plasma. Arteriosler Thromb 1989; 9:798.

47. St-Pierre J, Lemieux I, Vohl MC et al. Contribution of abdominal
obesity and hypertriglyceridemia to impaired fasting glucose
and coronary artery disease. Am J Cardiol 2002; 90:15.

48. Oppenheimer MF, Sundquist K, Bierman EL. Down regulation of
high density lipoprotein receptor in human fibroblasts by insulin
and IGF-I. Diabetes 1989; 38:117.

49. Steinberg D, Parthasarathy C, Carew TE et al. Beyond cholesterol:
modifications of low-density lipoprotein that increase its athero-
genicity. N Engl J Med 1989; 320:915.

50. Jones BA, Sidawy AN, LoGerfo FW. Diabetic vascular disease. In:
Sidawy AN, Sumpio BN, DePalma RG, eds. The Basic Science of
Vascular Disease. Armonk, NY: Futura, 1997:441.

51. Kannel WB. Hypertension and other risk factors in coronary
heart disease. Am Heart J 1987; 114:918.

52. OiDonnell CJ, Kannel WB. Epidemiologic appraisal of hyperten-
sion as a risk factor in the elderly. Am J Geriatr Cardiol 2002; 11 :86.

53. Multiple Risk Factor Intervention Trial Research Group. Multi-
ple risk factor intervention trial: risk factor changes and mortali-
ty results. JAMA 1982; 248:1465.

54. Katzel, LI, Coon PJ, Busby MJ. Reduced HDL2 cholesterol sub-
species and elevated postheparin hepatic lipase activity in older
men with abdominal obesity and asymptomatic myocardial
ischemia. Arterioscler Thromb 1992; 12:814.

55. Krobot K, Hense HW, Cremer P et al. Determinants of plasma fib-
rinogen: relation to body weight, waist to hip ratio, smoking,
alcohol, age and sex. Arterioscler Thromb 1992; 12:780.

56. Hiatt WR. Medical treatment of peripheral arterial disease and
claudication. N Engl J Med 2001; 344:1608.

57. Jonason T, Jonzon B, Ringquist I. Effect of physical training on
different categories of patients with intermittent claudication.
Acta Med Scand 1979; 206:253.

58. Blumenthal JA, Sherwood A, Gullette EC et al. Exercise and
weight loss reduce blood pressure in men and women with mild
hypertension: effects on cardiovascular, metabolic, and hemody-
namic functioning. Arch Int Med 2000; 160: 1947.

59. Ford ES. Does exercise reduce inflammation? Physical activity
and C reactive protein among US adults. Epidemiology 2002;
13:561.

60. Meade TW. Cardiovascular disease, linking pathology and epi-
demiology. Int J Epidemiol 2001; 30:1170.

61. Stone MC, Thorp JM. Plasma fibrinogen: major coronary risk fac-
tor. JR Coll Gen Pract 1985; 35:565.

62. Kannel WB, Wolf PA, Castelli WP, Dagostino RB. Fibrinogen and
the risk of cardiovascular disease. The Framingham Study.
JAMA 1987; 258;1183.

63. Ernst E, Hammerschmidt DE, Bagge U et al. Leukocytes and the
risk of ischemic diseases. JAMA 1987; 257:2318.

64. Ridker PM, Rifai N, Stampfer MJ, Hennekens CH. Plasma con-
centration of interleukin-6 and the risk of future myocardial
infarction among apparently healthy men. Circulation 2000;
101:1767.

65. Ridker PM, Rifai, N, Pfeffer M, Sacks F, Lepage S, Braunwald E.
Elevation of tumor necrosis factor alpha and increased risk of re-
current cardiac events after myocardial infarction. Circulation
2000:101:2149.

66. Ridker PM, Heinekens CH, Buring JE, Rifai N. C reactive protein
and other markers of inflammation in the prediction of cardio-
vascular disease in women. N Engl J Med 2000; 342:836.

67. Leinonen M, Saikku P. Evidence for infectious agents in cardio-
vascular disease and atherosclerosis. Lancet Infect Dis 2002; 2:11.

68. Lowe GD. The relationship between infection, inflammation and
cardiovascular disease: an overview. Ann Peridontol 2001; 6:1.

69. Ngeh J, Anand V, Gupta S. Chlamydia pneumonia and athero-
sclerosis, what we know and what we don't. Clin Microbiol Infect
2002; 8:2.

70. Rossouw JE, Anderson GL, Prentice RL et al. Risks and benefits of
estrogen plus progestin in healthy postmenopausal women.
JAMA 2002; 288:321.

71. Sullivan JL. Iron and the sex difference in heart disease risk.
Lancet 1981; 1:1293.

72. Sempos CT, Looker AC, Gillum RF. Iron and heart disease: the
epidemiologic data. Nutr Reviews 1996; 54:73.

73. Corti MC, Gaziano M, Hennekens CH. Iron status and risk of car-
diovascular disease. Ann Epidemiol 1997; 7:62.

74. Sempos CT, Looker AC, Gillum RE, McGee Dl, Vuong CV, John-
son C. Serum ferritin and death from all causes and cardiovascu-
lar disease: the NHANES II Mortality Study. Ann Epidemiol 2000;
10:441.

75. Haidari M, Javadi E, Sanati A, Hajilooi M, Ghanbili J. Association
of increased ferritin with premature coronary stenosis in men.
ClinChem 2001; 47:1666.

76. Tuomainen TP, Salonen R, Nyyssonen K, Salonen JT. Cohort

451

part iv Medical management

study of relation between donating blood and risk of myocardial
infarction in 2682 men in eastern Finland. Br Med J 1997; 314:
793.

77. Salonen JT, Tuomainen TP, Salonen R, Lakka TA, Nyyssonen K.
Donation of blood is associated with reduced risk of myocardial
infarction. The Kuopio Ischaemic Heart Disease Risk Factor
Study. Am J Epidemiol 1998; 148:445.

78. Ascherio A, Rimm EB, Giovannucci E, Willet WC, Stampfer MJ.
Blood donations and risk of coronary heart disease in men. Circu-
lation 2001; 103:52.

79. Osier M, Milman N, Heitmann BL. Consequences of removing
iron fortification of flour on iron status among Danish adults:
some longitudinal observations between 1987 and 1994. Prevent
Med 1999; 29:32.

80. Criqui MH, Langer RD, Fronek A et ah Mortality over a period of
10 years in patients with peripheral arterial disease. N Engl} Med
1992; 326:381.

81 . Ridker PM, Rif ai N, Lowenthal SP Rapid reduction in C-reactive
protein with cerivastatin among 785 patients with primary hy-
percholesterolemia. Circulation 2001; 6:1191.

82. Burmudez EA, Ridker PM. C-reactive protein, statins, and the
primary prevention of atherosclerotic cardiovascular disease.
Prev Cardiol 2002; 5:42.

83. Albert MA, Danielson E, Rifai N, Ridker PM. The pravastatin
inflammation/CRP evaluation (PRINCE): a randomized trial
and cohort study. JAMA 2001; 286:64.

84. Executive Summary of the Third Report of the National Choles-
terol Education Program (NCEP) Expert Panel on Detection,
Evaluation, and Treatment of High Blood Cholesterol in Adults
(Adult Treatment Panel III). JAMA 2001; 285:2486.

85. Ross R. Atherosclerosis is an inflammatory disease. Am Heart J
1999;138:S419.

86. Frostegard J, Ulfgren A-K, Nyberg P et ah Cytokine expression in
advanced human atherosclerotic plaques: dominance of proin-
flammatory (Thl) and macrophage-stimulating cytokines. Ath-
erosclerosis 1999; 145:33.

87. Kiechl S, Lorenz E, Reindl M et ah Toll-like receptor 4 polymor-
phisms and atherogenesis. N Engl J Med 2002; 347:185.

88. Desfaits AC, Serri O, Renier G. Normalization of lipid peroxides,
monocytes adhesion, and tumor necrosis factor-alpha produc-
tion in NIDDM patients after gliclazide treatment. Diabetes Care
1998; 21:487.

89. Winkler G, Lakatos P, Nagy Z et ah Elevated serum TNF-alpha
level as a link between endothelial dysfunction and insulin resis-
tance in normotensive obese patients. DiabetMed 1999; 16:207.

90. Fazio S, Linton ME The inflamed plaque: cytokine production
and cholesterol balance in the vessel wall. Am J Cardiol 2001;
88:122E.

91. Gerszten RE, Mach F, Sauty A, Rosensweig A, Luster AD.
Chemokines, leukocytes, and atherosclerosis. / Lab Clin Med
2000; 136:87.

92. Bredie SJ, Wollersheim, HC, Verheught FW, Thein T. Acetylsali-
cylic acid in primary prevention of cardiovascular events; litera-
ture study. Ned Tijdschr Geneeskd 2002; 146:687.

93. Akaike M, Azuma H, Kagawa A et ah Effect of aspirin treatment
on serum concentration of lipoprotein (a) in patients with athero-
sclerotic diseases. Clin Chem 2002; 48:1454.

94. DePalma RG, Bellon EM, Manalo P, Bomberger RA. Failure of an-
tiplatelet treatment in dietary atherosclerosis: a serial interven-

tion study. In: Gallo LL, Vahouny GV, eds. Cardiovascular Disease:
Molecular and Cellular Mechanisms, Prevention, Treatment. New
York: Plenum Press, 1987:407.

95. CAPRIE Steering Committee. A randomized, blinded, trial of
clopidogrel versus aspirin in patients at risk of ischemic events
(CAPRIE). Lancet 1996; 348:1329.

96. Hiatt WR. Preventing atherothrombotic events in peripheral ar-
terial disease: the use of antiplatelet therapy. / Intern Med 2002;
251:193.

97. Yusef S, Sleight P, Pogue J et ah Effects of an angiotensin-convert-
ing enzyme inhibitor, ramipril, on cardiovascular events in high-
risk patients. HOPE Study Investigators. N Engl J Med 2000;
342:145.

98. Auerbach AD, Goldman L. Beta-blockers and reduction of car-
diac events in noncardiac surgery: scientific review. / Earn Pract
2002; 287:1435.

99. Ferguson TB Jr, Coombs LP, Peterson ED. Preoperative beta-
blocker use and mortality and morbidity following CABAG
surgery in North America. JAMA 2002; 287:2221.

100. Heart Study Collaborative Group. MRC/BHF Heart protection
study of antioxidant vitamin supplementation in 20,536 high-
risk individuals: a randomized placebo controlled trial. Lancet
2002; 6:12.

101 . Pruthi S, Allison TG, Hensrud DD. Vitamin E supplementation in
the prevention of coronary heart disease. Mayo Clin Proc 2001;
76:1131.

102. Harris A, Devaraj S, Jialal I. Oxidative stress, alpha-tocopherol
therapy, and atherosclerosis. Curr Atheroscler Rep 2002; 4:373.

103. Blumberg JB. An update: Vitamin E supplementation and heart
disease. Nutr Clin Care 2002; 5:50.

104. Meagher E, Rader DJ. Antioxidant therapy and atherosclerosis:
animal and human studies. Trends Cardiovasc Med 2001; 11:162.

105. Kruth HS, Huang W, Ishii I, Zwang WY. Macrophage foam cell
formation with native LDL. J Biol Chem 2002; 277:34573.

106. Capuzzi DM, Guyton JR. Morgan JM et ah Efficacy and safety of
an extended-release niacin (Niaspan): a long-term study. Am J
Cardiol 1998; 82:74U.

107. Taylor LM Jr, Moneta GL, Sexon GL et ah Prospective blinded
study of the relationship between plasma homocysteine and pro-
gression of symptomatic peripheral arterial disease. / Vase Surg
1999; 8:8.

108. Schnyder G, Roffi M, Flammer Y et ah Effect of homosycteine low-
ering therapy with folic acid, vitamin B(12) and vitamin B(6) on
clinical outcomes after percutaneous coronary intervention: The
Swiss Heart study: a randomized controlled trial. JAMA 2002;
288:973.

109. Creager MA. Medical management of peripheral arterial dis-
ease. Cardiol Rev 2001; 9:238.

110. Haennel RG, Lemire F. Physical activity to prevent cardiovas-
cular disease. How much is enough? Can Earn Physician 2002; 48:
65.

111 . Wood-Baker R. Outcome of a smoking cessation programme run
in a routine hospital setting. Intern Med J 2002; 32:24.

112. Meade TW, Mellows S, Brozovic M et ah Haemostatic function
and ischaemic heart disease: principal results of the Northwick
Park Study. Lancet 1986; 2:533.

113. Zeidel A, Beilin B, Yardeni I. Immune response in asymptomatic
smokers. Acta Anaesthsiol Scand 2002; 46:959.

114. Jamerson BD, Nides M, Jorenby DE et ah Late term smoking

452

chapter 37 Atherosclerosis: risk factors and medical management

cessation despite initial failure: an evaluation of buproprion
sustained release, nicotine patch, combination therapy and
placebo. Clin Ther 2001; 23 ;7 44.

115. Modell JG, Katholi CR, Modell JD, DePalma RL. Comparative
sexual side effects of buproprion, fluoxetine, paroxitine, and ser-
traline. Clin Pharmocol Ther 1997; 61:476.

116. Perkins KA. Smoking cessation in women. Special considera-
tions. CNS Drugs 2001; 15:391.

117. Superko HR. Small dense, low density lipoprotein and athero-
sclerosis. Curr Atheroscler Rep 2000; 2:226.

118. Superko HR, Krauss RM. Coronary artery disease regression.
Convincing evidence for the benefit of aggressive lipoprotein
management. Circulation 1994; 90:1056.

119. Renaud S. Wine, alcohol, platelets and the French paradox for
coronary heart disease. Lancet 1992; 339:1523.

120. Jung K. Moderate red and white wine consumption and risk of
cardiovascular disease. Herz/Kreisl 1999: 31:25.

121. DePalma RG, Hubay CA, Botti RE, Peterka JL. Treatment of sur-
gical patients with atherosclerosis and hyperlipidemia. Surg
Gynecol Obstet 1970; 131:633.

122. DePalma RG, Clowes AW. Intervention in atherosclerosis:
a review for surgeons. Surgery 1978; 84:175.

123. Avena R, Arora S, Carmody BJ et ah Thiamine (Vitamin Bl) pro-
tects against glucose and insulin-mediated proliferation of
human infragenicular arterial smooth muscle cells. Ann Vase
Surg 2000; 14:37.

124. Avena R, Mitchell ME, Nylen ES et ah Insulin action enhance-
ment normalizes brachial artery reactivity in patients with
peripheral vascular disease and occult diabetes. / Vase Surg 1998;
28:1024.

125. Haffner SM, Greenberg AS, Weston WM et ah Effect of rosiglita-
zone treatment on nontraditional markers of cardiovascular dis-
ease in patients with type 2 diabetes mellitus. Circulation 2002;
106:669 pub. on line.

126. Zacharski LR, Chow B, Lavori PW et ah The Iron (Fe) and Athero-
sclerosis Study (FeAST): a pilot study of body iron stores in pe-
ripheral vascular disease. Am Heart J 2000; 139:337.

127. DePalma RG, Hayes VW, Cafferata HT et ah Cytokine signatures
in atherosclerotic claudicants. / Surg Res 2003; 111 :215.

128. Stone DM, Mendall MA, Kaski JC. Effect of treatment for Chlamy-
dia pneumoniae and Helicobacter pylori on markers of inflamma-
tion and cardiac events in patients with acute coronary
syndromes. South Thames Trial of Antibiotics in Myocardial
Infarction and Unstable Angina. STAMINA. Circulation 2002;
106:1219.

129. Baxter BT, Pearce WH, Waltke EAet ah Prolonged administration
of doxycycline in patients with small abdominal aneurysms: re-
port of a prospective phase II multicenter study. / Vase Surg 2002;
36:1.

130. Dugan JP, Feuge RR, Burgess DS. Review of the evidence for a
connection between Chlamydia pneumonia and atherosclerotic
disease. Clin Ther 2002; 24:719.

131. Lewis BS, Flugelman MY, Weisz A et ah Angiogenesis by gene
therapy: a new horizon for myocardial revascularization. Cardio-
vasc Res 1997; 35:490.

132. Tabata H, Silver M, Isner JM. Arterial gene transfer of acidic fi-
broblast growth factor for therapeutic angiogenesis in vivo: criti-
cal role of secretion signal in the use of naked DNA. Cardiovasc
Res 1997; 35:470.

133. Isner JM, Vale Pr, Symes JF, Losordo DW. Assessment of risks as-
sociated with cardiovascular gene therapy in human subjects.
Circ Res 2001; 89:389.

134. Tateishi-Yuyama E, Matsubara H, Murohara T et ah Therapeutic
angiogenesis for patients with limb ischaemia by autologous
transplantation of bone-marrow cells: a pilot study and a
randomized controlled trial. Lancet 2002; 310:427.

135. Maeder T. Down with the bad, up with the good. Scientific Ameri-
can Feb 2002; 32.

136. Kobayashi J, Okamoto H, Otabe H et ah Effect of HDL, from
Japanese white rabbit administered a new cholesterol ester pro-
tein inhibitor JTT-705, on cholesterol ester accumulation induced
by acetylated low-density lipoprotein in J 774 macrophage.
Atherosclerosis 2002; 162:131.

137. deGrooth GJ, Kuivenhoven JA, Stalenhoef AF et ah Efficacy and
safety of a novel cholesteryl ester transfer protein inhibitor, JYY-
705, in humans: a randomized phase II dose response study.
Circulation 2002; 105:2159.

138. Li J, Fang B, Eisensmith RC et ah In vivo gene therapy for
hyperlipidemia: phenotypic correction of Watanabe rabbits by
hepatic delivery of the rabbit LDL receptor gene. / Clin Invest
1995; 95:768.

139. Breyer B, Jiang W, Cheng H et ah Adenoviral vector-mediated
gene transfer for human gene therapy. Curr Gene Ther 2001; 1:149.

140. Gao GP, Alvira MR, Wang L et ah Novel adeno-asssociated virus-
es from rhesus monkeys as vectors for human gene therapy. Proc
Natl Acad Sci USA 2002; 99:11854.

141. Comerota AJ, Throm RC, Miller KA et ah Naked plasmid DNA
encoding fibroblast growth factor type 1 for the treatment of end-
stage unreconstructible lower extremity ischemia: preliminary
results of a phase 1 trial. / Vase Surg 2002; 35:930.

453

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Pharmacologic intervention
thrombolytic therapy

Anthony J. Comerota
A. Koneti Rao
Mohammad H. Eslami

In 1958, a dynamic equilibrium for the coagulation and fibri-
nolytic system was proposed by Astrup. 1 Under physiologic
conditions, the body maintains an equilibrium between the
coagulation and the fibrinolytic systems. Complex interrela-
tionships exist, such that fibrin formation actually stimulates
physiologic fibrinolysis. Since 1950, physicians have recog-
nized that pharmacologic stimulation of the patient's fibri-
nolytic system could be therapeutically effective. 2

Because of their close involvement in the management
of patients with thromboembolic disorders of the arterial and
venous systems, it is particularly important for vascular
surgeons to understand the interactions of the coagulation
and fibrinolytic systems and the pharmacologic applications
of plasminogen activation.

The fibrinolytic system

The primary purpose of the fibrinolytic system in humans is
the physiologic dissolution of thrombi. Fibrinolysis is initiated
by plasminogen activators, which activate the zymogen
plasminogen to plasmin, the key enzyme in this system. At
least two distinct physiologic plasminogen activators have
been identified: tissue-type plasminogen activator (t-PA) and
urokinase-type plasminogen activator (u-PA).

There are at least two physiologic pathways that activate
plasminogen to plasmin. The intrinsic activators consist of
components normally found in the blood, and include pro-
teins of the contact phase of blood coagulation such as factor
XII, and kallikrein, which can interact and generate plasmin, at
least in vitro. The physiologic relevance of these mechanisms is
uncertain. The extrinsic activators arise from cells and tissues,
including vascular endothelial cells and neoplastic cells, and
are the main physiologic activators. These include t-PA and u-
PA. The plasmin generated by these pathways is the principal
mechanism the body calls on to dissolve intravascular throm-
bi; however, the rate at which this occurs may be too slow to re-
solve pathologic thrombus. The overall goal of pharmacologic
manipulation of the fibrinolytic system is to supply sufficient

quantities of exogenous plasminogen activators in a well
controlled manner to induce rapid lysis of intravascular
thrombi and restore blood flow in order to minimize or avoid
the consequences of compromised perfusion. This has been
the topic of several excellent reviews. 3,4

Plasminogen

Plasminogen is synthesized in the liver and found in human
plasma and serum in an average concentration of 21 mg/dl. It
is a single-chain polypeptide with a molecular weight of
92kDa and contains 790 amino acids with 24 disulfide
bonds. 5 ' 6 Additionally, there are five homologous triple-loop
structures known as kringles. Activators convert plasminogen
to the two-chain plasmin molecule by cleavage of a single pep-
tide bond (arginine 560-valine 561 bond), which splits the
molecule into the heavy and light chains. The amino-terminal
76 residue of native plasminogen (Glu-plasminogen) consti-
tutes the activation peptide that is released by plasmin, pro-
ducing a smaller molecule containing an amino-terminal
lysine called LYS-plasminogen. 7 LYS-plasminogen has a much
higher affinity for binding to fibrin both in purified systems
and in plasma, and also has greater reactivity with plasmino-
gen activators. 8 Plasminogen binding sites are present on
fibrin molecules, are exposed by proteolysis, and have a
particular affinity for LYS-plasminogen. 9 Therefore, the
formation of LYS-plasminogen accelerates and improves
the efficiency of plasmin formation with subsequent fibrin
dissolution. The heavy-chain portion of the molecule contains
the kringles, which contribute to fibrin binding and interac-
tion with plasminogen activators. 6 ' 10

Plasmin

Plasmin is a serine protease composed of two polypeptide
chains linked by disulfide bonds. The light chain contains
the enzyme's catalytic site. 11 Because plasminogen (LYS-

454

chapter 38 Pharmacologic intervention: thrombolytic therapy

plasminogen) usually is bound to fibrin, it can be converted by
plasminogen activators to plasmin at the localized site of fibrin
deposition, which is the primary focus of fibrinolytic activity 12
Any plasminogen activation that occurs in the surrounding
fluid phase is promptly neutralized by a 2 -antiplasmin. Thus,
physiologic thrombolysis is a well controlled and localized
process. Plasmin cleaves protein and peptide molecules
at arginyl-lysyl bonds. In addition to fibrin and fibrinogen,
plasmin hydrolyzes the coagulation factors V and VIII,
components of serum complement, corticotropin (adrenocor-
ticotropic hormone), growth hormone, and glucagon. Plasmin
also cleaves the activation peptide from plasminogen,
which serves to accelerate further plasmin formation from
LYS-plasminogen.

Inhibitors

Plasmin's wide-ranging activity can have a profound effect on
a large number of plasma proteins. Human plasma contains
inhibitors designed to regulate the activity of proteolytic en-
zymes. 13-15 oc 2 -Plasmin inhibitor is the principal physiologic
plasmin inhibitor. It is fast acting and has the strongest affinity
for plasmin, creating inactive plasmin-plasmin inhibitor com-
plexes. 16 It is present in plasma in concentrations of 1 |imol/l.
A much slower acting inhibitor, oc 2 -macroglobulin, exists in a
concentration of approximately 3|imol/l. 16 The primary
function of a 2 -macroglobulin is to bind plasmin after
a 2 -plasmin inhibitor is depleted. Although the plasmin-
oc 2 -macroglobulin complexes are active, they are rapidly re-
moved from the circulation. Other plasmin inhibitors include
o^-antitrypsin, anti thrombin III, and C-l esterase inhibitor, but
they have a minimal physiologic effect in the blood. Plasmino-
gen activator inhibitors are also important in the control of fib-
rinolysis. Inhibitors of t-PA and u-PA have been identified in
human plasma 17 and derived from human platelets. 18 Other
inhibitors have been obtained from cultured endothelial cells,
from human umbilical vein, hepatoma cells, placenta, mono-
cytes, and human fibroblasts.

Breakdown products of fibrinolysis

Under physiologic conditions the action of plasmin is limited
to the site of fibrin deposition. 19 Circulating inhibitors bind to
plasmin and form inactive complexes, thus preventing break-
down of fibrinogen, clotting factors, and other circulating pro-
teins. With the exogenous administration of plasminogen
activators or under certain pathologic conditions, plasmin
levels exceed the inhibitor's capacity, resulting in a systemic
plasminemia with breakdown of plasma proteins, especially
fibrinogen. The action of plasmin on fibrinogen results in the
segmental formation of several peptides, including fragment
X (250 kDa), which is degraded to yield fragments Y (150 kDa)

and D (100 kDa). Fragment Y is degraded to yield fragments D
and E (50 kDa). 19 ' 20

The action of plasmin on noncross-linked fibrin is identical
to that on fibrinogen in the rate of breakdown and the end-
products except that BB 15-42 peptide rather that BB 1-42 is
produced on cleavage of the BB chain of fibrin. These peptides
have been used to assess specific breakdown of fibrinogen vs.
fibrin by plasmin. Mature fibrin contains factor XHIa-induced
intramolecular bonds, causing a slower degradation by plas-
min as well as different end-products. D-Dimer is a unique
derivative of the proteolysis of cross-linked moieties from ad-
jacent fibrin monomers, which have been covalently bound by
factor XlHa. 21 ' 22

Plasminogen activators

Since 1950, physicians have used plasminogen activators to
dissolve fibrin in an attempt to improve patients' clinical out-
come. 2 These activators have evolved from relatively impure
and highly antigenic substances to pure, less antigenic, and in-
creasingly fibrin-specific agents. The ongoing improvement of
these fibrinolytic agents, in addition to our increasing knowl-
edge of the coagulation and fibrinolytic systems, allows safer
and more effective use of these agents.

The plasminogen activators approved for clinical use in-
clude streptokinase (SK), recombinant t-PA (rt-PA), prouroki-
nase (proUK), and acylated plasminogen-streptokinase
activator complex (APSAC). Although urokinase (UK) was re-
moved from the marketplace in 1999 owing to concerns raised
by the Food and Drug Administration (FDA) over potential
viral contamination of the agent, these concerns were ad-
dressed by the manufacturer. New production and purifica-
tion facilities were constructed leading to the re-release of
urokinase in 2002. Pharmaceutical companies continue to
pursue newer plasminogen activators using recombinant
technology.

Streptokinase

Streptokinase (SK) was first discovered in 1933 by Tillet and
Garner, 23 and was the first thrombolytic agent approved for
clinical use. SK is a nonenzyme protein containing 415 amino
acids with a molecular weight of 437 kDa, produced by group
C B-hemolytic streptococci. 24 SK alone is incapable of directly
converting plasminogen to plasmin, and therefore is not an
enzyme. It indirectly activates plasminogen by forming a 1 : 1
complex with human plasminogen. This complex undergoes a
conformational change to expose an active site on the plas-
minogen molecule. The plasminogen-SK complex is then
capable of catalyzing plasminogen to plasmin. The various
cleavage products are SK fragments ranging from 10 to 40 kDa,
all of which are able to complex with plasminogen. These frag-

455

part iv Medical management

ments have from 50% to 60% of the antigenic potential of the
parent molecule. SK is highly antigenic, and has the potential
for causing allergic reactions. Patients with recent exposure to
streptococci and those recently treated with SK have a high
level of circulating antistreptococcal antibodies capable of
neutralizing SK. Therefore, patients who are resistant to SK
therapy at standard doses may respond to higher doses after
exceeding the saturation point of existing antibodies. In vivo,
the activator complex formed by SK has a half-life of approxi-
mately 23 min. 25 In-vitro testing is available to determine the
SK dose needed to achieve systemic fibrinogenolysis, al-
though 95% of patients will be treated effectively by the
standard recommended doses.

SK has various systemic effects on the plasma coagulation
and fibrinolytic systems, and on platelets. Circulating levels of
plasminogen and fibrinogen are markedly decreased during
SK therapy. Systemic fibrinogenolysis results in increased
hemorrhagic complications. Concomitantly, there is a de-
crease in plasma plasminogen and plasma oc 2 -plasmin in-
hibitor. In addition, there is evidence that platelets are
altered. 26 These effects are noted without exception with all of
the thrombolytic agents, although SK appears to have a more
pronounced effect than rt-PA. The major disadvantage of SK is
its bacterial origin and antigenicity, which result in antibody
formation in all patients, precluding reuse of SK within at
least 6 months. In 1999, the FDA warned of increasing
life-threatening events associated with the use of SK and
anistreplase. Current trends in thrombolytic therapy involve
the use of more promising agents such as recombinant-type
agents such as alteplase, reteplase, and recombinant prouroki-
nase. Because of its antigenicity, clinical unpredictability, and
complication rate, SK has fallen into disfavor, and has been re-
placed by other lytic agents in most medical centers in the
United States.

Urokinase

Urokinase (UK) was first isolated by MacFarlane and Pilling in
1946 27 and subsequently by Williams in 1951. 28 UK is a double-
chain, tryp sin-like protease with a molecular weight of
54-57 kDa. It was observed early by several investigators that
UK could exist in several molecular weights of approximately
22, 33, and 54 kDa. The lower molecular weight molecules are
fragments of the larger ones. The complete primary amino
acid sequence of UK has been characterized. 29 Plasmin and
kallikrein cleave proUK at position 156, producing UK in its
two-chain form. The two chains are held together by a disul-
fide bond that is important for the fibrinolytic activity of UK.

The preparations of UK used therapeutically are either ex-
tracted from human urine— although this method is largely
obsolete— or are isolated from cultures of human fetal kidney
cells. UK also can be produced by recombinant genetic engi-
neering in Escherichia coli. The problem in the production of

UK from urine is the need for large quantities to produce ade-
quate amounts of enzyme; 1500 1 of urine is required for the
production of enough enzyme to treat one patient. The tissue
culture techniques demonstrated by Bernik and Kwaan indi-
cated that improved production of UK could be achieved, 30
and that the best fibrinolytic activity was seen in cultures from
cells taken from fetuses at 26-32 weeks' gestational age.

UK converts the inactive forms of plasminogen to plasmin,
with greater affinity for the fibrin-bound LYS-plasminogen.
The conversion is due to the cleavage of a single arginine
560-valine bond. 11 Activation of the fibrin-bound plasmino-
gen allows fibrinolysis to occur in a relatively inhibitor-free
environment, because there are no competing substrates for
fibrin-bound plasmin. UK is rapidly cleared by the liver, with
about 3-5% cleared by the kidneys. It has a short half-life of
about 16 min, which might be prolonged in patients with
hepatic dysfunction.

Although UK induces systemic fibrinogenolysis, its sys-
temic effect is not as intense as that of SK. Due to the produc-
tion costs of UK its price is five to eight times that of SK per
patient treatment. In 1999, the FDA withdrew UK from the
marketplace because of concerns over potential viral contami-
nation. These concerns have been addressed, and with new
manufacturing and purification processes in place, UK was
approved in October 2002.

Tissue plasminogen activator

The development of relatively fibrin-specific plasminogen
activators was made possible once physiologic t-PA was
extracted from human vascular endothelium, 31 and it was
shown that the endothelial extract induced highly specific clot
lysis compared with the activity of SK or UK. 32 t-PA is a single-
chain polypeptide serine protease with a molecular weight of
approximately 68 kDa. The principal site of in vivo synthesis is
the endothelial cell. It was initially purified from culture fluid
of Bowes melanoma cells and other mammalian tissue, but is
now synthesized by recombinant DNA techniques. 33

t-PA exists as a single chain that is converted rapidly to a
double-chain form by enzymatic cleavage of the peptide
bond arginine 275-isoleucine 276. In the absence of fibrin, t-PA
is relatively inactive, but in the presence of fibrin, there is a
500-1000-fold increase of plasminogen activation. 34 This is
due to an increased affinity of fibrin-bound t-PA for plas-
minogen. The high affinity of t-PA for plasminogen in the pres-
ence of fibrin allows activation of plasminogen on the clot,
thus sparing plasma plasminogen. The half-life of t-PA is ap-
proximately 2-6 min owing to binding by the rapid t-PA in-
hibitors and clearance by the liver. t-PA exists in plasma in
both a free state and complexed with plasma serine protease
inhibitors.

Although preliminary studies suggested that t-PA is fibrin
specific and capable of inducing thrombolysis without caus-

456

chapter 38 Pharmacologic intervention: thrombolytic therapy

ing systemic lytic effects, 35-37 overwhelming evidence from
large trials indicates that therapeutic doses of rt-PA also in-
duce systemic fibrinogenolysis, albeit less intense than that
induced by SK. 38/39 The first human application of t-PA from
melanoma cells was in a renal transplant recipient who had
an iliofemoral venous thrombosis. 40 Extensive experience in
acute myocardial infarction and pulmonary embolism indi-
cate that rt-PA is a highly effective thrombolytic agent. 41-44

Although it was originally anticipated that the in-vitro fibrin
selectivity of rt-PA would lead to fewer bleeding complica-
tions, this has not been borne out in clinical trials. The main
reason for this is the inability of all of the potent thrombolytic
agents to discriminate hemostatic thrombi at sites of vascular
breach from the pathologic thrombi they are being adminis-
tered to dissolve.

Prourokinase

A single-chain precursor to high-molecular-weight UK,
proUK was isolated from urine in 1979, 45 and is also termed
single-chain UK plasminogen activator. ProUK has been
identified in human plasma, cultures of endothelial cells,
explants of fetal organs, and various malignant cell lines. 46
This proenzyme is derived from human urine or genetically
manipulated E. coli, and has a molecular weight of approxi-
mately 54 kDa.

ProUK is converted to high-molecular-weight UK by
hydrolysis of the lysine 158-isoleucine 159 peptide bond
following its binding to fibrin. ProUK is not very effective as a
plasminogen activator, but this cleavage converts proUK into
its two-chain structure and increases its activity 500-1000-
fold. 41 ProUK differs from UK in several characteristics,
mainly in its higher fibrin affinity, lower specific activity, and
stability in plasma. Fibrin specificity of proUK does not
depend on actual fibrin binding as with t-PA, and thus the
mechanism of clot lysis is different between the two. 46 The
half-life of prourokinase is approximately 7 min.

Studies with proUK in rabbits, dogs, and baboons demon-
strated fibrin-selective clot lysis without fibrinogenolytic ef-
fects or hemorrhagic complications. 47 ProUK testing in dogs
also revealed that although proUK had superior fibrin speci-
ficity, it was equal to UK in efficacy. 48

The initial clinical application of proUK in humans was for
actue myocardial infarction. A small pilot study followed by a
multicenter study of acute myocardial infarction demon-
strated a 60% reperfusion rate in patients with proven coro-
nary artery thrombosis. Fifty milligrams of proUK was used,
with a mean time of lysis of approximately 55 min. Increasing
the dose to 70-80 mg increased reperfusion to almost 70%;
however, time to lysis was prolonged and systemic fib-
rinogenolysis developed in several patients. 46 No hemorrhag-
ic effects were noted. Additional studies combining UK and
proUK as well as t-PA and proUK demonstrated a synergistic

response between these agents because of complementary
mechanisms of action. 49 Recombinant proUK is currently
being evaluated in clinical trials with acceptable results.

Acylated plasminogen-strepokinase
activator complex

SK has significant fibrinogenolytic effects reflecting the wide-
ranging proteolysis resulting from its use. In an effort to create
a more efficient SK molecule with greater fibrin specificity, its
molecular structure was modified. The addition of an acyl
group accomplished the goal of improving fibrin specificity.
Furthermore, the acyl group blocked the binding site of anti-
streptococcal antibodies, thereby eliminating its inactivation
in plasma by circulating inhibitors. 50,51

The acylated SK (APSAC) molecule therefore is essentially
inert with regard to the activation of plasminogen to plasmin in
vitro. On the other hand, the acylation has little, if any, effect on
the binding of the plasminogen-SK complex to fibrin. Because
the fibrin binding site and the functionally active catalytic site
are spatially separate, this compound retains its ability to bind
to fibrin despite its lack of enzymatic activity when freely
circulating. Activation then occurs via hydrolysis, in which the
compound deacylates to give free activator complex and anisic
acid in equal concentrations. Because most of the deacylation
occurs after fibrin binding, the result is improved concentra-
tion of the activator complex bound to fibrin and a relatively
little amount of free activator complex circulating systemic-
ally The deacylation half-life in plasma is approximately 90-
105 min. 25 Because deacylation governs plasma clearance, the
compound has a true half-life of approximately 90-105 min,
with fibrinolytic activity persisting for 4-6 h.

In trials of acylated SK in rabbits and guinea pigs, it was
shown that this compound had higher thrombolytic activity
compared with an equal amount of nonacylated SK complex,
and did not cause systemic fibrinogenolysis. A comparison of
SK with acylated SK in a dog model of jugular venous throm-
bosis and pulmonary embolism demonstrated a minimal lytic
response with SK; however, thrombolysis was radiographic-
ally complete with acylated SK. There was, however, signifi-
cant fibrinogenolysis, with a decrease in the circulating
fibrinogen levels to approximately 50%.

Studies in human volunteers demonstrated the relative po-
tency of acylated SK and confirmed its potential for breaking
down fibrinogen and depleting a 2 -antiplasmin. 52/53 A 5-mg
dose of acylated SK is equivalent to 178 000 IU of SK.

Human trials in acute myocardial infarction showed good
recanalization rates (approximately 70%), but these studies
demonstrated a higher systemic fibrinogenolytic effect than
was anticipated from the original animal data. Half of the
treated patients had fibrinogen levels of 30% or less. Decreases
in plasminogen and oc 2 -plasmin inhibitor levels likewise were
observed. 25

457

part iv Medical management

The primary advantages of acylated SK are the prolonged
half-life and the convenience of bolus administration.

Recombinant tissue plasminogen activator

Two recombinant agents, derivatives of t-PA, have been ap-
proved for use in thrombolytic therapy in coronary interven-
tion. Although these agents have similar molecular structures,
their difference leads to varied pharmacological properties.

Reteplase (r-PA) is produced by recombinant DNA technol-
ogy from E. coli and has a molecular weight of 39kDa. It is a
nonglycosylated deletion mutein of the wild-type human t-
PA. Reteplase contains 355 of the 527 amino acids in t-PA. The
epidermal growth factor, fibronectin-like finger, and kringle-1
domains of t-PA have been deleted. Because of the deletion of
these domains hepatic uptake of t-PA occurs and resulted in a
longer half-life of reteplase (r-PA) (13-16 min) than alteplase
(rt-PA)(5min).

Alteplase is a glycosylated single-chain serine protease. The
gene sequence that expresses t-PA is isolated from a human
melanoma cell line and inserted into the Chinese hamster
ovarian cell. Alteplase is isolated, purified, and packaged as a
lyophilized powder. The half-life of alteplase is (4-5 min)
shorter than r-PA as it contains the epidermal growth factor,
fibronectin-like finger, and kringle-1 domains on the mole-
cule. The molecular weight of alteplase is 65 kDa.

Reteplase and alteplase are fibrinolytic agents that attach to
fibrin and catalyze the cleavage of endogenous plasminogen
to plasmin, which subsequently cleaves fibrin into fibrin
degradation products and results in clot breakdown. Attach-
ment of these agents to fibrin requires the kringle domain of
the molecule. Reteplase binds fibrin via the kringle-2 domain
since the other domain has been removed. As a result the fibrin
affinity and specificity of reteplase is less than that of alte-
plase. 54 In-vitro experiments have shown that reteplase pen-
etrates the clot more efficiently than alteplase 55 and this was
explained by the lesser affinity of reteplase for the fibrin mole-
cule. 55 Both of these agents have high thrombolytic efficiency
in in-vitro experiments. 54 In 1997, the Gusto III trial compared
these two agents in the management of myocardial infarc-
tion. 56 In this clinical setting, these agents were shown to have
similar clinical efficacy and comparable adverse effects. 56

Indications for clinical use

Thrombolytic therapy has been used to treat a multitude of
thromboembolic disorders involving the pulmonary, venous,
peripheral arterial, and coronary arterial systems. The use of
thrombolytic agents in patients with acute ischemic stroke has
been established. As thrombolytic therapy is further refined,
and as our understanding of the pathophysiology of throm-
botic and embolic diseases increases, the role of thrombolytic

therapy will continue to expand. Available evidence clearly
establishes thrombolytic therapy as a major modality in our
armamentarium for thromboembolic disorders. Clinical trials
using thrombolytic therapy have shown beneficial results in
multiple clinical settings.

The majority of the applications for use of lytic agents
in patients with peripheral arterial and venous thrombotic
complications apply the principles of catheter delivery, with
intrathrombus infusion of the plasminogen activator. This
stimulates intrathrombus production of plasmin through
activation of fibrin-bound plasminogen. The common clinical
uses of lytic therapy by vascular surgeons are reviewed
below.

Therapy for acute limb ischemia

Prior to the 1990s, the therapy for acute limb ischemia (ALI)
was limited to surgical intervention. The results of surgical
interventions in this setting often were disappointing. During
the past decade, thrombolysis has become an important treat-
ment option for the management of patients with ALI. Be-
tween 1994 and 1996, three large prospective randomized
trials evaluated the use of catheter-directed thrombolysis
(CDT) and compared the results of CDT vs. surgery 57-59 The
theoretical advantages of thrombolytic therapy include re-
duced endothelial damage, reduction in the extent of surgical
intervention subsequently required, and improved outcome
of surgical intervention by restoring patency to thrombosed
run-off vessels. Although the endpoints of these studies var-
ied, results showed that in the appropriate setting CDT could
improve outcomes. 57-61

Despite the results from randomized trials, a definitive role
for CDT for acute arterial and graft occlusion has not been
clarified, yet clinicians familiar with the alternatives available
to reperfuse the acutely ischemic extremity recognize the
benefits of this treatment option in these high-risk patients
(Figs 38.1 and 38.2).

The Rochester Study was a single-center study of 114 pa-
tients suffering from < 7 days of acute ischemia related to the
occlusion of either native artery or bypass grafts. Patients were
randomized to either CDT or the routine surgical interven-
tion. 57 The endpoint was defined as "event-free survival/' It
was noted that limb salvage was identical between the two
groups. In-hospital cardiopulmonary complications were
higher among the surgical group (49% vs. 16%). Similarly, 30-
day amputation-free survival and 12-month survival were
significantly higher among the CDT group with no clear ex-
planation for these findings. It is possible that a systemic fibri-
nolytic effect (which invariably occurs) is beneficial in patients
at high risk of cardiovascular morbidity. The authors did not
report any secondary procedures beyond the relatively short
postprocedure follow-up.

The STILE trial 58 concluded that surgical revascularization
was more effective than CDT using UK or rt-PA in patients

458

chapter 38 Pharmacologic intervention: thrombolytic therapy

Figure 38.1 This case demonstrates the
benefits of catheter-directed thrombolysis in a
patient presenting with an acutely thrombosed
femoral popliteal bypass graft. The initial
arteriogram (A,B) demonstrates the occluded
origin of the bypass graft with a patent distal
popliteal and run-off arteries. (C) A catheter was
placed into the occluded bypass graft and
infused with urokinase. (D) Patency was
restored to the bypass, and a subsequent
arteriogram demonstrated the culprit lesion at
the distal anastomosis. These lesions can be
corrected by a direct operative approach or
percutaneous endovascular techniques.

with symptomatic lower extremity arterial or graft occlusion
occuring within 6 months of randomization. The endpoint
was one of a composite of endpoints including death, limb
loss, ongoing ischemia (treatment failure), bleeding, or major
complications. Unfortunately, the results of the STILE trial
were driven by the high failure rate of CDT in the subgroup of

patients with chronic native arterial occlusion. It was intuitive
to vascular surgeons that operative revascularization for
chronic arterial occlusive disease would be more successful
than CDT; however, interventionists thought (at the time the
protocol was written) that catheter-based techniques, begin-
ning with thrombolysis, held promise as a definitive therapeu-

459

part iv Medical management

Figure 38.2 This case demonstrates the benefit of catheter-directed
thrombolysis (CDT) for acute limb ischemia following lower extremity
embolic occlusion to the profunda femoris and common femoral arteries.
This patient had known lower extremity arterial occlusive disease. He
presented with acute limb ischemia thought to be due to an arterial embolus

resulting from chronic atrial fibrillation. Arteriography confirmed the
occluded left common femoral and profunda femoris arteries (left).
Subsequent CDT with urokinase improved perfusion within several hours
(middle) and following overnight infusion, the patient's perfusion was
restored to baseline and his ischemia resolved (right).

tic option. As it turned out, failure of CDT did not alter the ulti-
mate treatment outcome, since surgical revascularization sal-
vaged lytic failures. Interestingly, in the subgroups of diabetic
patients with native artery occlusion, those randomized to
CDT had a significantly better survival at 1 year compared
with the surgical group.

When results were analyzed according to acuity of symp-
toms, it became evident that patients with acute limb ischemia
(< 14 days) were best treated by surgical revascularization.
Results also demonstrated that CDT significantly simplified
anticipated surgical procedures.

The TOPAS investigators performed a randomized multi-
center study comparing CDT with surgery in patients with
< 14 days of lower extremity ischemia. 59 The endpoint of the
study was arterial recanalization and extent of lysis. This
study showed no significant difference in amputation-free
survival rates among the CDT group and the surgery group. It
was noted that the actual operations performed were less ex-
tensive than originally planned in 50% of the cases treated by
CDT compared with 14% of surgical patients. The authors did
not specify the secondary procedures which were required by
the CDT group. To address this criticism, the authors noted
that "The TOPAS trial was not designed to compare CDT vs.
surgery as definitive therapy but rather CDT as an adjunct to
surgical therapy of acute limb ischemia/' 60 A subsequent
analysis showed that thrombus size and the duration of the

ischemic events were the two critical factors that influenced
the outcome of the initial therapy. 61

The role of CDT in the treatment of thrombosed arterial by-
pass grafts is controversial and somewhat confusing. In the ex-
perimental setting, it was shown that use of urokinase for CDT
causes less severe endothelial injury than surgical thrombec-
tomy. 62 Despite the theoretical advantages of the use of CDT,
the long-term results of CDT do not approach the patency rate
of a new vein graft which replaces the thrombosed infrain-
guinal graft. 63,64 This is probably due to endothelial changes
associated with graft thrombosis, an inadequate graft, com-
promised inflow or outflow, undiagnosed /untreated hyper-
coagulable states or a combination of these. It has become clear
that the efficacy of an arterial graft surveillance program to re-
pair a stenotic graft, thereby avoiding thrombosis, exceeds
that of any subsequent interventions for a thrombosed graft. 65

In a subsequent report of the STILE trial, the outcome of CDT
vs. surgery was evaluated in the subgroup of patients with
occluded bypass grafts. 66 CDT was superior to a primary surgi-
cal approach in patients presenting with acutely occluded
bypass grafts, demonstrated by a higher limb salvage rate.

A retrospective review of CDT for thrombosed grafts
showed that the outcome of CDT was dependent on the qual-
ity of the run-off and that CDT alone was not effective for the
tibio-peroneal occlusion. 67 It was suggested that CDT of
thrombosed arterial grafts should be considered an adjunct to

460

chapter 38 Pharmacologic intervention: thrombolytic therapy

surgical revascularization. A similar, more recent retrospec-
tive single-center study echoes the same findings, showing
that in a selected group of nondiabetic patients with graft
thrombosis, primary-assisted patency may be improved by
thrombolysis. 68 The common conclusion is that CDT should
be used as an adjunct to surgical intervention.

Common to all these studies is the conclusion that in cate-
gory I and Ha ischemia, the use of CDT as an adjunct may im-
prove the outcome of surgical intervention but that CDT does
not substitute for definitive surgical revascularization. The
choice of therapy for ALI depends on four factors: (i) the cause
of ALI; (ii) co-morbidities of the patient with ALI; (iii) the dis-
tribution of disease; and (iv) the clinical degree of ischemia in
the affected limb.

Currently, the only FDA-approved agent for thrombolysis
of peripheral arterial occlusive disease is streptokinase. Its
side-effects, potential for bleeding complications, and unpre-
dictable lytic effect have rendered this agent obsolete. The ma-
jority of the contemporary experience has been achieved with
urokinase, and more recently rt-PA and reteplase. Davidian
et al. 68 reported the feasibility and efficacy of reteplase CDT for
lower extremity occlusive disease. These investigators re-
ported a 6% complication rate in 15 patients treated with
reteplase and the time to lysis was comparable to urokinase.
They concluded that monotherapy with reteplase is safe and
effective for treatment of patients with peripheral arterial
occlusive disease, although the number of patients was
exceptionally small. The PURPOSE trial investigators have
similarly reported the safety and efficacy of r-proUK in the
setting of peripheral arterial occlusive. 69

Intraoperative intraarterial thrombolytic therapy

Residual intraarterial thrombus is the rule rather than the ex-
ception following balloon catheter thromboembolectomy for
acute arterial occlusion. 70-72 This has been demonstrated both
clinically and experimentally. Residual thrombus forms the
nidus for recurrent thrombosis and additional distal emboli.
The existence of residual thrombus provides a strong rationale
to administer thrombolytic agents in the operating room. Tak-
ing advantage of their efficient action by direct intraarterial
delivery and their short half-lives, the benefits of thrombus
dissolution can be achieved with minimal additional risk of
hemorrhagic complications.

A number of early clinical reports, both experimental and
clinical, demonstrated that intraoperative intraarterial infu-
sion of thrombolytic agents could reduce residual thrombus
burden and improve reperfusion without exposing patients to
undue bleeding complications. 72-78

A prospective, randomized, blinded, and placebo-
controlled clinical trial evaluating multidose intraoperative,
intraarterial infusion of urokinase vs. saline control specifi-
cally addressed the safety of bolus intraoperative urokinase
infusion: 125 000 U, 250 000 U, and 500 000 U were infused into

the distal arterial bed at the time of lower extremity bypass
and compared with a saline control. Significant elevation in
D-dimer was observed indicating breakdown of fibrin in the
distal circulation. Although there was a step-wise decrease
in plasminogen, there was no significant reduction in plas-
minogen even at the highest dose compared with saline. There
was no significant breakdown in fibrinogen and there was no
increased risk of bleeding complications in any of the UK
groups compared with saline. 79 These observations led to our
routine use of intraoperative intraarterial UK in doses up to
500 000 U and intraoperative rt-PA in doses up to 6 mg when-
ever thrombus is extracted from an artery or bypass graft. An
intriguing observation was that patients receiving UK had a
significantly lower 30-day mortality compared with control
patients.

An interesting and potentially limb-saving use of intra-
operative intraarterial lytic therapy is the "high-dose isolated
limb perfusion" technique. 80 This procedure is indicated in pa-
tients with multivessel extremity occlusion in whom a single
or double bolus is unlikely to be adequate and in patients in
whom any degree of systemic fibrinolysis would pose signifi-
cant risk. This technique includes full anticoagulation, exsan-
guination of venous blood from the limb with a rubber
bandage, application of a proximal tourniquet to achieve com-
plete arterial and venous occlusion, direct arterial infusion
into the affected artery with a high dose of thrombolytic agent
(i.e. urokinase 1 000 000-2 000 000 U or more, or rt-PA 50 mg or
more), and drainage of the venous effluent. Infusion of a lytic
agent for 45-60 min has yielded impressive results in a small
number of patients suffering from acute multivessel distal
thrombi or acute emboli. Ten patients have been treated in our
institution for severe ischemia with the isolated limb perfu-
sion method. Five patients had a good response resulting in
limb salvage. If distal vessel occlusion is due to atheromatous
emboli or organized thrombus, treatment has not been suc-
cessful. This novel approach deserves further attention.

Thrombolytic therapy for thrombosed
hemodialysis access

End-stage renal disease (ESRD) affects about 350 000 patients
in the United States and the disease is increasing at a rate of
10% per year. 81 About 200 000 patients are estimated to be on
hemodialysis (HD) and the most common cause of hospital-
ization in these patients is vascular access occlusion. The ma-
jority of patients in the United States have prosthetic PTFE HD
grafts which have a 1-year patency rate of 60-70%. 82 A number
of surgical procedures are performed to restore and maintain
patency of failed dialysis access grafts, with modest mid-term
results. From 1960 to 1995, different techniques involving
thrombolysis were employed with some success but were
associated with significant complications. 83,84 Current
techniques have improved early success rates and reduced
complications. The two most commonly used lytic techniques

461

part iv Medical management

are the "pulse-spray" technique and the "lysis and wait"
protocol.

In 1995, Valji et al. 85 reported the use of thrombolysis em-
ploying pulse-spray pharmacomechanical thrombolysis. This
involves the placement of a catheter inside the clotted graft.
The lytic agent is administered locally in small aliquots using a
pulse technique via the multihole catheter inserted into the
thrombosed graft. The original protocol used UK, but since
1999 r-PA and rt-PA have replaced UK. After clot lysis abnor-
malities of arterial inflow or venous outflow are treated by
angioplasty or stenting. The goal here is complete or near
complete lysis, correction of any underlying stenosis with
restoration of a palpable thrill in the graft.

Cynamon et al. described a variation of the pulse-spray tech-
nique, called the "lyse and wait" method. 86 It involves the
direct injection of a thrombolytic agent into a thrombosed
dialysis graft. A slow injection is performed after the arterial
and venous anastomses are occluded manually. The patient is
observed for 30min to 2h (dwell time) and then evaluated
with arteriography. Using the same needle to infuse the
thrombolytic agent, contrast is injected and any arterial or
venous abnormality is managed endovascularly Similar to
the "pulse-spray" technique, the endpoint here is a palpable
thrill. The benefits of the "lyse and wait" technique include a
shorter time to lysis, cost savings (as no catheter is required),
less radiation, and more efficient use of arteriography suites. A
potential disadvantage of this technique is perigraft extrava-
sation of the thrombolytic agent leading to seroma formation.

The National Kidney Foundation Dialysis Outcome Quality
Initiative (DOQI) definition of acceptable results 87 is an
80-90% initial thrombolysis success rate, < 1% serious compli-
cation rate, and a 40% 90-day primary patency rate. Results of
therapy using either technique show promising outcomes
which should improve the functional life of a dialysis graft.
These procedures can be performed safely under local anes-
thesia as an outpatient and patients can be dialyzed the same
day. To achieve a durable result anatomical reasons for the
graft failure must be corrected. Flick and colleagues reported a
success rate of 94% with graft thrombolysis using "lysis and
wait" with reteplase. 88 Gibbens et al. 89 reported a 98% initial
success rate and a 53% 3-month patency rate using the pulse-
spray technique with reteplase. Alteplase has also been used
successfully with an initial success rate of 88% and 30- and 90-
day patency rates of 57% and 50%, respectively, using the
"lysis and wait" technique. 90 In summary, thrombolysis —
using a variety of fibrinolytic agents— is an effective and safe
alternative to surgical thrombectomy for acutely occluded
dialysis accesses. Excellent early success rates can be anticipat-
ed, therefore this approach will remain a mainstay in the treat-
ment of occluded hemodialysis grafts.

Thrombolysis for acute deep venous thrombosis

Venous occlusion can cause severe post-thrombotic conse-

quences, particularly in active patients. The post-thrombotic
syndrome is a morbid consequence of acute deep venous
thrombosis (DVT). In the majority of patients, treatment of
acute venous thrombosis is anticoagulation alone. Unfortu-
nately, many patients are denied the benefits of thrombus
removal/ resolution and needlessly suffer post-thrombotic se-
quelae. Thrombolysis has been effectively used in the initial
management of primary axillosubcalvain vein thrombosis
(Paget-Schroetter syndrome), iliofemoral DVT as a result of
iliac vein compression (May-Thurner syndrome, Fig. 38.3),
phlegmasia cerulea dolens, selected pulmonary embolism,
and superior vena cava syndrome. Thrombolysis in these clin-
ical settings can yield impressive results; however, due to its
limited use further investigation is necessary to achieve wide-
spread acceptance.

Thrombolytic therapy for iliofemoral
venous thrombosis

The benefits of CDT for acute DVT of the upper and lower
extremity are to relieve the acute symptoms of major central
venous obstruction and to avoid post-thrombotic sequelae. It
has been demonstrated that venous obstruction is an impor-
tant component of the severe post-thrombotic syndrome. 91 ' 92
When acute thrombi are lysed venous patency is restored, and
valvular function can be preserved if thrombus resolution
occurs in a timely fashion. 93

Iliofemoral DVT represents the most severe form of lower
extremity acute DVT, and is associated with the worst post-
thrombotic sequelae. 94 ' 95 The goals of treatment with CDT for
patients with iliofemoral DVT are to: (i) prevent pulmonary
embolism; (ii) reduce /eliminate the acute symptoms of il-
iofemoral DVT; and (iii) reduce /avoid the post-thrombotic
syndrome. Efficient elimination of thrombus is likely to pre-
serve valvular function, and will restore venous patency. Ad-
ditionally, many patients who are treated, especially those
with left-sided iliofemoral DVT, are likely to have an under-
lying iliac vein stenosis which will be uncovered following
successful CDT. 96 ' 97 If an underlying stenosis is identified, it
should be corrected in order to preserve long-term patency
and avoid recurrent thrombosis.

A large clinical experience has been obtained with CDT for
iliofemoral DVT, with consistent results observed between
centers. Bjarnason and colleagues 98 reported their 5-year ex-
perience treating 87 lower extremities with iliofemoral DVT in
77 patients. Mewissen et al." reported the largest series of CDT
for lower extremity DVT to date in a report of the national
multicenter venous registry. Two hundred and twenty-one
patients with iliofemoral DVT and 79 patients with femoral
popliteal DVT were treated with urokinase infusions.

The experience at Temple University Hospital encompasses
58 patients treated with CDT for iliofemoral DVT. 100 Surpris-
ingly consistent results were observed among the three
studies. Patients who were treated within 2 weeks of onset of

462

chapter 38 Pharmacologic intervention: thrombolytic therapy

Figure 38.3 This case demonstrates the benefits of catheter-directed
thrombolyis in a patient with iliofemoral deep venous thrombosis. A 21 -year-
old woman developed a painful and swollen left lower extremity 3 months
after beginning birth control pills. Phlebography confirmed the ultrasound
diagnosis of iliofemoral venous thrombosis. The phlebogram demonstrated
thrombus in the iliofemoral (A) and femoral-popliteal (B) venous segments.
Under ultrasound guidance, a catheter was placed into the popliteal vein and
advanced through the femoral vein into the iliofemoral system (B). Following

20 h of catheter-directed urokinase infusion, the thrombus was dissolved and
a tight stenosis of the left common iliac vein was demonstrated (C). After
balloon dilation of the stenotic iliac vein, partial luminal continuity was
restored (D); however, a residual stenosis persisted. A self-expanding stent
was placed which eliminated the stenosis (E) and restored unobstructed
venous drainage into the vena cava. The patient's symptoms were resolved
and she has been free of post-thrombotic sequelae since her treatment
5 years ago.

463

part iv Medical management

their acute iliofemoral DVT had approximately an 85% suc-
cess rate. Bleeding complications occurred in 7-12%, with the
majority being puncture site oozing. Intracranial bleeding oc-
curred in two patients and a fatal pulmonary embolism during
treatment in one patient, yielding a treatment mortality of less
thanl%.

A quality-of-life study was conducted in patients having
CDT for acute iliofemoral DVT as part of the National Venous
Registry and compared with a similar contemporary cohort of
patients treated with anticoagulation alone. 101 Patients were
queried with a validated health-related quality-of-life ques-
tionnaire with items specific to DVT and the post-thrombotic
syndrome. Patients who were treated with CDT reported bet-
ter overall physical functioning, less stigma of chronic venous
disease, less health distress, and fewer post-thrombotic symp-
toms compared with patients treated with anticoagulation
alone. Within the lytic group, phlebographically successful
thrombolysis correlated with an improved health-related
quality-of-life. Interestingly, patients who failed thrombolytic
therapy had similar treatment outcomes to patients managed
with anticoagulation alone. These observations corroborate
the clinically observed improvement and minimal post-
thrombotic sequelae in patients successfully treated with CDT
for iliofemoral DVT.

Axillosubclavian vein thrombosis

Improved understanding of the etiology of primary axillosub-
clavian vein thrombosis, as well as the evolution and high
success rate of CDT, has led to an important change in our ap-
proach to these patients. 102 The long-term consequences of
axillosubclavian vein thrombosis have demonstrated dis-
ability rates of 25-74%. 103 - 106 CDT followed by elimination of
anatomic subclavian vein compression reduces long-term
morbidity to 12% or less. 107 Our approach is to treat patients
presenting with symptomatic axillosubclavian vein thrombo-
sis with CDT. 107 Once the clot is resolved, a // temporizing /,
venoplasty can be performed followed by 4-6 weeks of anti-
coagulation, and subsequent first rib resection. Following first
rib resection phlebography is performed. If a residual subcla-
vian vein stenosis is present, a repeat venoplasty and stenting
(if necessary) is performed. Many clinicians are moving to first
rib resection immediately following successful CDT rather
than waiting a 4-6-week time interval with patients on
anticoagulation. 108

Summary

Thrombolytic agents are important for the management of the
spectrum of patients with acute arterial and venous thrombo-
sis or embolism. While the use of these agents has traditionally
encompassed many specialties, particulary those with special
interests in hematology, the benefits of acute thrombus disso-

lution have been recognized and have become part of the
therapeutic armamentarium of all clinicians caring for pa-
tients with acute vascular disorders.

Delivery of plasminogen activators by catheter-directed
techniques has improved efficiency and probably reduced
bleeding complications. Identifying and correcting anatomic
lesions contributing to the thrombotic event is crucial to long-
term success.

Modifying delivery systems by reducing their profile and
increasing mechanical penetration and extraction of thrombus
should improve early results. The rapid and ongoing ad-
vances in anticoagulant and antithrombotic therapy will un-
doubtedly improve long-term results by reducing the risks of
rethrombosis.

References

1. Astrup T. The haemostatic balance. Thromb Diath Haemorrh 1958;
2:347.

2. Sherry S. The origin of thrombolytic therapy. / Am Coll Cardiol
1989; 14:1085.

3. Marder VJ, Sherry S. Thrombolytic therapy. N Engl J Med 1988;
318:1512.

4. Marder VJ, Butler FO, Barlow GH. Antifibrinolytic therapy. In:
Colman RW, Hirsh J, Marder VJ, Salzman EW, eds. Hemostasis
and Thrombosis: Basic Principles and Clinical Practice, 2nd edn.
Philadelphia: JB Lippincott, 1987:380.

5. Sottrup-Jensen L, Claeys H, Zajdel M et al. The primary structure
of human plasminogen: isolation of two ly sine-binding frag-
ments and one // mini"-plasminogen (MW, 38,000) by elastase-
catalyzed specific limited proteolysis. In: Davidson JF, Rowan
RM, Samama MM, Desnoyers PC, eds. Progress in Chemical Fibri-
nolysis and Thrombolysis , Vol. 3. New York: Raven Press, 1978:191.

6. Dayhoff MO. Atlas of Protein Sequence and Structure, Vol. 5,
Suppl. 3. Silver Spring, MD: National Biomedical Research
Foundation, 1979.

7. Walther PJ, Steinmann HM, Hill RI, McKee PA. Activation of
human plasminogen by urokinase: partial characterization of
apreactivation peptide. JBiolChem 1974; 249:1173.

8. Wallen P, Wiman B. Characterization of human plasminogen. II.
Separation and partial characterization of different forms of
human plasminogen. Biochim Biophys Acta 1972; 257:122.

9. Varadi A, Patthy L. Location of plasminogen-binding sites in
human fibrin(ogen). Biochemistry 1983; 22:2440.

10. Markus G, DePasquale JL, Wissler FC. Quantitative determina-
tion of the binding of epsilon-aminocaproic acid to native
plasminogen. J Biol Chem 1978; 253:727.

11. Robbins KC, Summaria L, Hsieh B, Shah RF. The peptide chains
of human plasmin: mechanism of activation of human plasmino-
gen to plasmin. J Biol Chem 1967; 242;2332.

12. Alkjaesig N, Fletcher AP, Sherry S. The mechanism of clot
dissolution by plasmin. / Clin Invest 1959; 38:1086.

13. Mullertz S. Natural inhibitors of fibrinolysis. In: Davidson JF,
Rowan RM, Samama MM, Desnoyers PC, eds. Progress in Chimi-
cal Fibrinolysis and Thrombolysis, Vol 3. New York: Raven Press,
1978:213.

464

chapter 38 Pharmacologic intervention: thrombolytic therapy

14. Aoki N. Natural inhibitors of fibrinolysis. Prog Cardiovasc Res
1979; 21:276.

15. Wiman B. Human alpha-2-antiplasmin. Methods Enzymol 1981;
80:395.

16. Mullertz S, Clemmensen I. The primary inhibitor of plasmin in
human plasma. BiochemJ 1976; 159:545.

17. Thorsen S, Philips M. Isolation of tissue-type plasminogen
activator-inhibitor complexes from human plasma: evidence for
a rapid plasminogen activator inhibitor. Biochim Biophys Acta
1984; 802:111.

18. Erickson LA, Ginsbert MH, Loskutoff DF. Detection and partial
characterization of an inhibitor of plasminogen activator in
human platelets. / Clin Invest 1984; 74:1465.

19. Francis CW, Marder VJ. Concepts of clot lysis. Annu Rev Med
1986; 37:187.

20. Budzynski AZ, Marder VJ, Shainoff JR. Structure of plasmic
degradation products of human fibrinogen: fibrinopeptide and
polypeptide chain analysis. J Biol Chem 1974: 249:2294.

21. Francis CW, Marder VJ, Martin SE. Plasmic degradation of cross-
linked fibrin. I. Structural analysis of the particulate clot and
identification of new macromolecular soluble complexes. Blood
1980:56:456.

22. Kopec M, Tesseyre E, Dudek-Wojciechowska G et ah Studies on
"double D" fragment from stabilized bovine fibrin. Thromb Res
1973; 2:283.

23. Tillet WS, Garner RL. The fibrinolytic activity of hemolytic strep-
tococci. / Exp Med 1933; 58:485.

24. Kwaan HC. Hematologyic aspects of thrombolytic therapy.
In: Comerota AJ, ed. Thrombolytic Therapy. Orlando: Grune &
Stratton, 1988:17.

25. Sherry S. Pharmacology of anistreplase. Cliff Cardiol 1990; 5:3.

26. Coller B. Platelets and thrombolytic therapy. N Engl J Med 1990;
322:33.

27. MacFarlane RG, Pilling J. Fibrinolytic activity of normal urine.
Nature 1947; 159:779.

28. Williams JRB. The fibrinolytic activity of urine. Br J Exp Pathol
1951; 32: 520.

29. Grunzler WA, Steffens GJ, Otting F, Kim SM, Frankus E, Flohe L.
The primary structure of high molecular mass urokinase from
human urine: the complete amino acid sequence of a A chain.
Hoppe Seylers Z Physiol Chem 1982; 363:1155.

30. Bernik MB, Kwaan HC. Origin of fibrinolytic activity in cultures
of human kidney. J Lab Clin Med 1967; 70P:650.

31. Aoki N, von Kaulla K. The extraction of vascular plasminogen
activator from human cadavers: some of its properties. Am } Clin
Pathol 1971; 5:171.

32. Gurewich V, Hyde E, Lipinski B. The resistance of fibrinogen and
soluble fibrin monomer in blood to degradation by the potent
plasminogen activator derived from cadaver limbs. Blood 1975;
46:555.

33. Penneca D, Holmes WE, Kohn WJ et ah Cloning and expression of
human tissue plasminogen activator DNA in E. coli. Nature 1983;
301:214.

34. Holylaets M, Rijken DC, Lihnen HR, Collen D. Kinetics of the ac-
tivation of plasminogen by human tissue plasminogen activator:
role of fibrin. J Biol Chem 1982; 257:2912.

35. Collen D, Stass AJ, Verstraete M. Thrombolysis with human
extrinsic (tissue-type) plasminogen activator in rabbits with
experimental jugular vein thrombosis. / Clin Invest 1983; 71:368.

36. Van de Werf F, Ludbroo PA, Bergmann SR et ah Coronary throm-
bolysis with tissue-type plasminogen activator in patients with
evolving myocardial infarction. N Engl} Med 1984; 310: 609.

37. Korninger C, Matsuo O, Suy R, Stassen JM, Collen D. Throm-
bolysis with human extrinsic (tissue-type) plasminogen activa-
tor in dogs with femoral vein thrombosis. / Clin Invest 1982;
69:573.

38. Rao AK,PrattC,BerkeAe£ ah Thrombolysis in myocardial infarc-
tion (TIMI) trial, phase I: hemorrhagic manifestations and
changes in plasma fibrinogen and the fibrinolytic system in pa-
tients treated with recombinant tissue plasminogen activator
and streptokinase. J Am Coll Cardiol 1988; 11:1.

39. Mueller HS, Rao AK, Forman SA. Thrombolysis in myocardial
infarction (TIMI): comparative studies of coronary reperfusion
and systemic fibrinogenolysis with two forms of recombinant
tissue-type plasminogen activator. } Am Coll Cardiol 1987; 10:479.

40. Weimer W, Stibbe J, Van Seyen AJ, Billau A, DeSomer P, Collen D.
Specific lysis of an iliofemoral thrombus by administration of
extrinsic (tissue-type) plasminogen activator. Lancet 1981; 2:1018.

41. The TIMI Study Group. The thrombolysis in myocardial infarc-
tion (TIMI) trial: phase I findings. N Engl} Med 1985; 312:932.

42. Verstraete M, Bory M, Collen D et ah Randomized trial of intra-
venous recombinant tissue-type plasminogen activator versus
intravenous streptokinase in acute myocardial infarction: report
from the European Cooperative Study Group for Recombinant
Tissue-Type Plasminogen Activator. Lancet 1985; 325:842.

43. Goldhabor SZ. Thrombolytic therapy for pulmonary embolism.
Semin Vase Surg 1992; 5:89.

44. Come PC, Kim D, Parker JA et ah Early reversal of right ventricu-
lar dysfunction in patients with acute pulmonary embolism after
treatment with intravenous tissue plasminogen activator. / Am
Coll Cardiol 1987; 10:971.

45. Huasin SS, Lipinski B, Gurewich V Isolation of plasminogen
activators useful as therapeutic and diagnostic agents (single-
chain, high-fibrin affinity urokinase). US patent no. 4, 381, 346
(filed September 2, 1980, issued April 26, 1983).

46. Gurewich V Tissue plasminogen activator and pro-urokinase.
In: Comerota AJ, ed. Thrombolytic Therapy. Orlando: Grune &
Stratton, 1988:209.

47. Gurewich V, Pannel R, Louie S, Kelley P, Suddith RL, Greelee R.
Effective and fibrin-specific clot lysis by a zymogen precursor of
urokinase (pro-urokinase): a study in vitro and in two animal
species. J Clin Invest 1984; 73:1731.

48. Collen D, Stump D, Van de Werf F et ah Coronary thrombolysis in
dogs with intravenously adminstered human pro-urokinase.
Circulation 1985; 72:384.

49. Gurewich V, Pannell R. A comparative study of the efficacy and
specificity of tissue plasminogen activator and pro-urokinase:
demonstration of synergism and of different thresholds of non-
selectivity Thromb Res 1986; 44:217.

50. Smith R, Dupe R, English P, Green J. Fibrinolysis with acylen-
zymes: a new approach to thrombolytic therapy. Nature 1982;
290:505.

51. Matsuo O, Collen D, Verstaete M. On the fibrinolytic and throm-
bolytic properties of active-site p-anisoylated streptokinase-
plasminogen complex (BRL, 26921). Thromb Res 1981; 24:347.

52. Marder VJ, Francis CW, Norry EC. Dose-ranging study of acy-
lated streptokinase-plasminogen complex (BRL 26921). Thromb
Haemost 1983; 50:321.

465

part iv Medical management

53. Marder VJ, Rothbard RI, Fitzpatrick PG et al. Rapid lysis of coro-
nary artery thrombi with anisoylated plasminogen; streptoki-
nase activator complex. Ann Intern Med 1986; 104:304.

54. Martin U, Bader R, Bohm E et al. BM 06.022: a novel recombinant
plasminogen activator. Cardiovasc Drug Rev 1993; 11:299.

55. Fischer S, Kohnert U. Major mechanistic differences explain the
higher clot lysis potency of reteplase over alteplase: lack of fibrin
binding is an advantage for bolus application of fibrin-specific
thrombolytics. Fibrinolysis Proteolysis 1997; 11:129.

56. Global Use of Strategies to Open Occluded Coronary Arteries
(GUSTO III) Investigators. A comparison of reteplase with
alteplase for acute myocardial infarction. N Engl J Med 1997;
337:1118.

57. Ouriel K, Shortell CK, De Weese JA et al. A comparison of throm-
bolytic therapy with operative vascularization in the initial
treatment of acute peripheral arterial ischemia. / Vase Surg 1994;
19:1021.

58. The STILE Investigators. Results of a prospective randomized
trial evaluating surgery versus thrombolysis for ischemia. Ann
Surg 199 4; 220:251.

59. Ouriel K, Veith FL, Sasahara AA. For TOPAS Investigators.
Thrombolysis or peripheral arterial surgery: Phase I results. /
Vase Surg 1996; 23:64.

60. Ouriel K. A comparison of recombinant urokinase with vascular
surgery for acute arterial occlusion of the legs. (The authors
reply).NEnglJMed 1995; 339:564.

61. Ouriel K, Veith FJ. Acute lower limb ischemia: determinants of
outcome. Surgery 1998; 124:336.

62. Whitley D, Gloviczki P, Rhee R et al. Urokinase treatment pre-
serves endothelial and smooth muscle function in experimental
acute arterial thrombosis. / Vase Surg 1996; 23:851.

63. Faggioli GL, Peer RM, Pedrini L et al. Failure of thrombolytic
therapy to improve long-term vascular patency. / Vase Surg 1994;
19:289.

64. Whittemore AJ, Clowes AW, Couch NP et al. Secondary fem-pop
reconstruction. Ann Surg 1981; 193:35.

65. DaMariorbus CA, Mills JI, Fugitani RM et al. A reevaluation of
intraarterial thrombolytic therapy for acute lower extremity
ischemia. / Vase Surg 1993; 17:888.

66. Comerota AJ, Weaver FA, Hosking JD et al. Results of a prospec-
tive, randomized trial of surgery versus thrombolysis for oc-
cluded lower extremity bypass grafts. Am } Surg 1996; 172:105.

67. Nackman GB, Walsh DB, Filinger MF et al. Thrombolysis of
occluded infrainguinal vein grafts: predictors of outcome. / Vase
Surg 1997; 25:1023.

68. Davidian MM, Powel A, Benenati A et al. Initial results of
reteplase in treatment of acute lower extremity arterial occlusion.
/ Vase Interv Radiol 2000; 11:289.

69. Ouriel K, Kandarpa K, Schuerr DM et al. Prourokinase versus
urokinase for recanalization of peripheral occlusions, safety and
efficacy: the PURPOSE trial. / Vase Interv Radiol 1999; 10:1083.

70. Greep JM, Allman PJ, Janet F et al. A combined technique for
peripheral arterial embolectomy Arch Surg 1972; 105:869.

71. Plecha FR, Pories WJ. Intraoperative angiography in the
immediate assessment of arterial reconstruction. Arch Surg 1972;
105:902.

72. Quinones-Baldrich WJ, Ziomek S, Henderson TC et al. Intra-
operative fibrinolytic therapy: experimental evaluation. / Vase
Surg 1986; 4:229.

73. Comerota AJ, White JV, Grosh JD. Intraoperative, intraarterial
thrombolytic therapy for salvage of limbs in patients with distal
arterial thrombosis. Surg Gynecol Obstet 1989; 169:283.

74. Belkin M, Valeri R, Hobson RW. Intraarterial urokinase increases
skeletal muscle viability after acute ischemia. / Vase Surg 1989;
9:161.

75. Quinones-Baldrich WJ, Zierler RE, Hiatt JC. Intraoperative fibri-
nolytic therapy: an adjunct to catheter thromboembolectomy. /
Vase Surg 1985; 2:319.

76. Norem RF, Short DH, Kerstein MD. Role of intraoperative fibri-
nolytic therapy in acute arterial occlusion. Surg Gynecol Obstet
1988; 167:87.

77. Garcia R, Saroyan RM, Senkowski J et al. Intraoperative, intra-
arterial urokinase infusion as an adjunct to Fogarty catheter em-
bolectomy in acute arterial. Surg Gynecol Obstet 1990; 171:201.

78. Parent NE, Bernhard VM, Pabst TS et al. Fibrinolytic treatment of
residual thrombus after catheter embolectomy for severe lower
limb ischemia. / Vase Surg 1989; 9:153.

79. Comerota AJ, Rao AK, Thromb RC et al. A prospective, random-
ized, blinded, and placebo-controlled trial of intraoperative
intraarterial urokinase infusion during lower extremity
revascularization: regional and systemic effects. Ann Surg 1993;
218:534.

80. Comerota AJ. Intraoperative intraarterial thrombolytic therapy.
In: Yao JST, Pearce WH, eds. Progress in Vascular Surgery. Stam-
ford, CT: Appleton and Lange, 1997:341.

81. United States Renal Data System. Annual data report III: Treat-
ment modalities for ESRD patients. Am } Kidney Dis 1997; 30
(Suppl.I):54.

82. Hirth RA, Turenne MN, Woods JD et al. Predictors of type of
vascular access in hemodialysis patients. JAMA 1996; 276:1303.

83. Mangiarotti G, Canavese C, Thea A et al. Urokinase treatment for
arteriovenous fistulae declotting in dialyzed patients. Nephron
1984; 36:60.

84. Young AT, Hunter DW, Castaneda-Zuniga WR et al. Thrombosed
synthetic hemodialysis access fistula: failure to fibrinolytic
therapy. Radiology 1985; 154:639.

85. Valji K, Bookstein JJ, Roberts AC et al. Pulse-spray pharmacome-
chanical thrombolysis of thrombosed hemodialysis access
grafts; long-term experience and comparison of original and cur-
rent techniques. Am J Roentgenol 1995; 164:1495; discussion 1501.

86. Cynamon J, Lakritz PS, Wahl SI et al. Hemodialysis graft decod-
ing: description of the "lyse and wait" technique. / Vase Interv
Radiol 1997; 8:825.

87. Schwab S, Besaeab A, Beathard G et al. NKF-DOQI Clinical prac-
tice guidelines for vascular access. Am J Kidney Dis 1997; 30
(Suppl.3):150.

88. Flick P, Des M, Horton K et al. Initial experience with reteplase
using the "lyse and wait" technique in thrombosed dialysis
grafts (abstract). / Vase Interv Radiol 2000; 2 (Suppl.):252.

89. Gibbens D, Depalma J, Albanese J et al. Percutaneous thromboly-
sis of hemodialysis grafts using reteplase (abstract). / Vase Interv
Radiol 2000; 2 (Suppl.):250.

90. Falk A, Mitt H, Guller J et al. Thrombolysis of clotted hemodialy-
sis grafts with tissue-type plasminogen activator. / Vase Interv
Rad/o/ 2001; 12:305.

91. Shull KC, Nicolaides AN, Fernandes E, Fernandes J. Significance
of popliteal reflux in relation to ambulatory venous pressure and
ulceration. Arch Surg 1970; 114:1304.

466

chapter 38 Pharmacologic intervention: thrombolytic therapy

92. Johnson BF, Manzo RA, Bergelin RO, Strandness DE. Relation-
ship between changes in the deep venous system and the devel-
opment of the post-thrombotic syndrome after an acute episode
of lower limb deep vein thrombosis: a one-to-six year followup. /
Vase Surg 1995; 21:307.

93. Meissner MH, Manzo RA, Bergelin RO, Strandness DE. Deep ve-
nous insufficiency: the relationship between lysis and subse-
quent reflux. / Vase Surg 1993; 18:596.

94. O'Donnell TF, Browse NL, Burnand KG, Lea Thomas M. The
socioeconomic effects of an iliofemoral venous thrombosis. /
Surg Res 1977; 22:483.

95. Hill SL, Martin D, Evans P. Massive vein thrombosis of the
extremities. Am J Surg 1989; 158:131 .

96. O'Sullivan G J, Semba CP, Bittner C A et ah Endovascular manage-
ment of iliac vein compression (May-Thurner) syndrome. / Vase
Interv Radiol 2000; 11:823.

97. Seidensticker D, Wilcox J, Gagne P. Treatment of May-Thurner
syndrome with catheter directed thrombolysis and stent place-
ment, complicated by heparin-induced thrombocytopenia.
Cardiovasc Surg 1998; 6:607.

98. Bjarnason H, Kruse JR, Asinger DA et al. Iliofemoral deep venous
thrombosis: safety and efficacy outcome during 5 years of
catheter directed thrombolytic therapy. / Vase Interv Radiol 1997;
8:405.

99. Mewissen MW, Seabrook GR, Meissner MH, Cynamon J,
Labroupoulos N, Houghton SH. Catheter directed thrombolysis
for lower extremity deep venous thrombosis: report of a national
multicenter registry. Radiology 1999; 211:39.

100. Comerota AJ, Kagan SA. Catheter directed thrombolysis for the

treatment of acute iliofemoral deep venous thrombosis.
Phlebology 2001; 15:149.

101. Comerota AJ, Throm RC, Mathias S, Haughton SH, Mewissen
MW. Catheter directed thrombolysis for iliofemoral DVT
improves health-related quality-of-life. / Vase Surg 2000; 32:130.

102. Machleder HI. Thrombolytic therapy for acute primary axillo-
subclavian vein thrombosis. In: Comerota AJ, ed. Thrombolytic
Therapy for Peripheral Vascular Disease. Philadelphia: JB
Lippincott Co., 1995:197.

103. Donayre CE, White GH, Mehringer SM, Wilson SE. Pathogenesis
determines late morbidity of axillosubclavian vein thrombosis.
Am} Surg 1986; 152:179.

104. Gloviczki P, Kazmier FJ, Hollier LH. Axillary-subclavian venous
occlusion: the morbidity of a nonlethal disease. / Vase Surg 1986;
4:333.

105. Tilney NL, Grittiths HJG, Edwards EA. Natural history of major
venous thrombosis of the upper extremity. Arch Surg 1979;
101:792.

106. Linblad B, Mornmyer S, Kullendorff B, Bergqvist D. Venous
haemodynamics of the upper extremity after subclavian vein
thrombosis. Vasa 1990; 19:218.

107. Comerota AJ. Catheter directed thrombolysis for the treatment of
axillary-subclavian venous thrombosis. Persp Vase Surg Endovasc
Ther 2001; 14:51.

108. Kreienberg PB, Chang BB, Darling RC 3rd et al. Long-term results
in patients treated with thrombolysis, thoracic inlet decompres-
sion, and subclavian vein stenting for Paget-Schroetter syn-
drome. / Vase Surg 2001; 33 (2 Suppl.):S100.

467

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Pharmacologic intervention: vasodilation
therapy and rheologic agents

George Johnson, Jr.

Pharmacologic therapy for patients with peripheral vascular
disease has been evaluated repeatedly. Medications to prevent
or lyse clots, prevent or dissolve atherosclerotic plaques, dilate
vessels, or change the viscosity of the blood have had varying
degrees of success, depending on the pathologic process of the
disease. Several well controlled studies have demonstrated
that medication, combined with diet, can lower cholesterol
and triglyceride levels and prevent the development of
atherosclerotic plaques in certain people. On the other hand,
attempts to dissolve already-developed atherosclerotic
plaques have had only anecdotal results. No well controlled
studies showing benefits have been reported. This chapter will
concentrate on drugs that dilate vessels and drugs that affect
hemorrheology. Some of these drugs have complex effects,
however, and the beneficial results may not be a result of
vasodilation or improved hemorrheology. The pathologic
processes discussed will be ischemic symptoms due to fixed
occlusions from atherosclerosis of the lower extremity vessels
and vasospastic syndromes.

The European Working Group on Critical Leg Ischemia,
representing experts from Europe and supported by eight spe-
cialist societies, has published the Second European Consensus
Document on Chronic Critical Leg Ischemia. 1 Critical leg ischemia
(CLI) was defined by the Working Group as "that kind [of is-
chemia] that endangers the leg or part of leg." The Consensus
Document section on pharmacologic treatment is an excellent
summary of current knowledge in this area.

Vasodilators

Vasodilators work by lowering peripheral resistance, thus in-
creasing blood flow. Since ischemia itself is a strong stimulus
to vasodilation, one would not think that vasodilator drugs
would be of further benefit to patients with symptoms caused
by a fixed arterial occlusion. As recently stated by Shub,
"Vasodilators have various modes of action, but ultimately
all— either directly or indirectly — dilate blood vessels in one
or more vascular beds. Many vasodilators have been tried

over the years in patients with claudication, but with only lim-
ited, if any, success." 2

Lowe observed in 1990, "Some evidence also exists that
infusions of vasodilators, such as inositol nicotinate, nafti-
drofuryl, and prostanoids (prostaglandin E 1 , epoprostenol
[prostacyclin, prostaglandin I 2 ], or stable prostacyclin ana-
logues) may relieve subacute rest pain or promote ulcer heal-
ing, or both; however, the results of large controlled studies are
awaited before treatment with vasodilators can be recom-
mended with confidence . . . several mechanisms exist by
which vasodilatation may act, which can be helpful in 'buying
time' while patients are evaluated for surgery, angioplasty, or
thrombolytic treatment . . . the main problem ... is [that] . . .
rest pain and ulceration respond to placebo treatment and the
passage of time." 3

No vasodilator has been classified as "effective" for the
treatment of claudication by the US Food and Drug
Administration (FDA).

Although a classification of vasodilator drugs may be use-
ful, many of them have dual actions at different doses and may
work by multiple mechanisms. Vasospastic diseases, as op-
posed to fixed obstruction, should respond to vasodilator
drugs. Some of those in use will be reported.

Prostanoids

Prostanoids, including the stable prostacyclin analogue ilo-
prost, have received widespread attention in Europe for the
pharmacologic treatment of CLI. They induce vasodilation, as
well as inhibiting platelet aggregation. A number of large,
controlled trials of prostanoids (prostaglandin I 2 [PGI 2 ],
prostaglandin E 1 [PGEJ, and iloprost) as medical treatment
for severe arterial disease have been reported. 1

Prostanoids must be given intravenously or intraarterially.
Data suggest that prostanoids need to be given for at least 72 h;
short-term therapy has not been effective. The dosage of the
prostanoid infusion varies between the reported studies. In 16
studies summarized in the Consensus Document on leg is-
chemia, results do not seem to be dose related. 1

468

chapter 39 Pharmacologic intervention: vasodilation therapy and rheologic agents

Alprostadil (PGE^

The Consensus Document summarized three short-term (3-4
days) randomized trials comparing alprostadil with a
placebo, which failed to demonstrate a reduction in ischemic
pain. Longer-term trials (7-28 days) significantly reduced rest
pain more than a placebo, pentoxifylline, or adenosine
triphosphate (ATP). 1 Although it is approved in the United
States by the FDA for temporarily maintaining the patency of a
ductus arteriosus, it has not been approved for use in periph-
eral vascular disease.

Epoprostenal (PGI 2 )

Epoprostenal has been evaluated for severe rest pain in
several short-term (3-4 days) randomized trials comparing it
with a placebo, with equivocal results. One of two studies with
intravenous infusion showed slight relief of pain. Two studies
using arterial infusion therapy showed no relief of pain. The
FDA has approved this as an orphan drug (a drug that can be
used only in treatment of a rare disease with the approval of
the FDA).

Iloprost

Iloprost, the stable prostacyclin analogue, has been the
subject of several articles. Like the other prostanoids, it
has multiple actions, including prevention of platelet aggrega-
tion and some fibrinolytic activity that leads to a decrease in
neutrophil adhesion and chemotaxis, in addition to its va-
sodilatory action. It can be taken orally but is rapidly degraded
in the gut wall and liver. The FDA has approved it as an orphan
drug.

In 1991, Dormandy reviewed five prospective, randomized
trials of iloprost therapy for severe leg ischemia. 4 ' 5 There was a
response in 51.5% of iloprost-treated patients, based on pain
relief or decrease in ulcer size. A review article on iloprost 6 re-
ported by the United Kingdom Severe Limb Ischaemia Study
found an amputation rate of 32% for iloprost recipients and
47% for a placebo group at 6 months. After a survey of the
literature on claudication, the review article concluded,
"iloprost does appear to provide some benefit, but therapeutic
gains must be weighed against the difficulties associated with
lengthy intravenous therapy/' 6

Calcium channel blockers

Calcium channel blockers such as cinnarizine, flunarizine,
and nifedipine preferentially dilate skeletal muscle arteries,
increasing blood flow to exercising muscles. Although this
may in theory ameliorate claudication, no study has yet been
done to demonstrate conclusively that these drugs improve
patients with claudication. Their main use has been for
vasospastic disorders, such as Raynaud's phenomenon.

Nifedipine has been thoroughly evaluated for treatment of
this disease and is very effective in most patients.

Ketanserin

Ketanserin has a high affinity for 5-HT 2 receptors, which in-
hibit serotonin-induced vasoconstriction. Its main use may be
in the treatment of Raynaud's phenomenon. Several studies of
ketanserin in patients with intermittent claudication reveal
different results, but generally suggest that ketanserin is un-
likely to be associated with important improvement. 7-11 It is
not approved for use by the FDA.

Other vasodilating drugs

Blombery observes that tolazoline, nicotinyl alcohol, cyclan-
delate, all vasodilator drugs, have been used for claudication,
but there is no evidence that they are beneficial. 12 Shub added
reserpine, guanethidine, prazosin, terazosin, phenoxybenza-
mine, and papaverine to this list. 2

The Consensus Document concluded, "pharmacological
treatment of CLI should be considered when catheter proce-
dures or reconstructive surgery are not technically possible,
are contraindicated, have failed, or carry an unacceptable
risk/benefit ratio/' 1

Raynaud's phenomenon

Raynaud's phenomenon merits special attention. General
therapeutic measures include keeping warm, wearing gloves,
stopping smoking, and excluding causes such as drugs, vi-
brating tools, and diseases of the cervical rib or connective tis-
sue. Drug treatment should be used for severe cases.

Nifedipine is the drug of choice in the United States for the
treatment of this disease. The starting dosage is usually 10 mg
once or twice daily, increasing the dosage to balance relief of
symptoms with side-effects. Dosages may be as high as 20 mg,
three to four times a day if the side-effects are not too severe.
Several randomized trials have demonstrated the efficacy of
nifedipine in terms of a reduction in the number, duration, and
severity of attacks. 13 ' 14 Side-effects, however, were common
and often intolerable.

A 1992 report by Grant and Goa reviewed seven studies of
iloprost used for Raynaud's phenomenon. 6 It was noted these
were not well controlled; however, there was a consistent
tendency for improvement in the frequency, duration, and
intensity of ischemic episodes for up to at least 6 weeks. A
well controlled comparison of a short (3-6 days) course of
intravenous iloprost and oral nifedipine in 23 patients with
Raynaud's phenomenon associated with systemic sclerosis
confirmed the clinical efficacy of iloprost compared with
nifedipine. Although both drugs were beneficial in decreasing
the number, duration, and severity of attacks, decreasing the
number of digital lesions, and increasing digital blood flow,

469

part iv Medical management

side-effects with nifedipine were common, whereas the side-
effects of iloprost occurred only during the infusions and were
dose dependent. Short-term infusions of iloprost provided
long-lasting relief of symptoms. 15

A placebo-controlled study demonstrated complete healing
of digital ulcers in six of seven patients treated with iloprost
(0.5-2 ng/kg/min for 5 days), and no healing in controls. 16

Lukac and colleagues have suggested that serotonin may
not only participate in the pathophysiology of Raynaud's phe-
nomenon, but may play an important role in the pathogenesis
of scleroderma. 17 Thus, ketanserin may be beneficial in pre-
vention as well as treatment of this entity. Codella and associ-
ates found ketanserin to be superior to nifedipine in 28
patients, both clinically and as observed with computed digi-
tal thermometry. 18

Although reserpine had been widely used in the treatment
of Raynaud's phenomenon, the parenteral preparation is no
longer available, a- Adrenergic blocking agents, such as
methyldopa, tolazoline, guanethidine, phenoxybenzamine,
and prazosin have had little usefulness. Isoxsuprine, nylidrin,
isoproterenol, papaverine, niacin, griseofulvin, and
nitroglycerine all have been tried with varying results. 19

Hemorrheology

Blood flow through a tube, according to Poiseuille's law,
APnr 4

Q =

8|iL

(1)

is inversely related to the viscosity of the blood (P = pressure
drop across the tube, r = tube radius, L = tube length, and ji = the
viscosity coefficient). Several reports have documented the re-
lationship between viscosity and flow. 20-22 It would appear
that the quantity and quality of the red cells and fibrinogen are
the main constituents of the blood that determine viscosity.
Several drugs have been evaluated that alter the viscosity of
blood in anticipation of an improvement in blood flow to an is-
chemic extremity.

Pentoxifylline

Pentoxifylline has its primary effect by increasing the
deformability of the red cell. It also decreases fibrinogen,
however, and both of these effects will cause a decrease in
viscosity and presumably an increase in flow. The dosage is
400 mg three times a day with meals. It may take several weeks
before the effect is noticed. The only side-effect identified is
nausea, which is alleviated by stopping the medication.
Pentoxifylline is approved by the FDA as a treatment for
claudication.

There has been tremendous worldwide experience with this
drug for a great variety of illnesses. Some of these studies have
been well controlled. To add to the confusion, a 1987 review of

pentoxifylline concluded, "It would appear to be a useful
adjunct to conservative therapy in patients with mild to
moderate peripheral vascular disease and is almost certainly
useful in patients with more severe disease unable to undergo
surgery/' 23 On the other hand, Cameron and colleagues, after
noting that there was more evidence to support pentoxifylline
for intermittent claudication than any other drug, concluded
"we have important reservations about the trials." 24 The re-
sults for claudication probably can be summarized by the fact
that 30% of patients cannot tolerate the drug, 30% experience
no effect from the drug, and 30% have an improvement in their
claudication. Those patients showing an improvement in
claudication can walk about 30% further than those who do
not take the drug. The problem with determining its effect is
that all patients with recent development of claudication re-
ceive benefit from walking.

Ancrod

Ancrod is a defibrinogenating enzyme from the venom of the
Malayan pit viper that reduces plasma viscosity. A recent
study by Wiles and coworkers, however, in which blood flow
was measured with a laser Doppler velocimeter after giving
ancrod, showed a decrease in plasma fibrinogen and viscosity
after 48 h of an intermittent intravenous infusion, but no in-
crease in blood flow. 25 No report of its use for intermittent clau-
dication or for ischemic extremities has been found. It is an
orphan drug.

Glycosaminoglycan sulodexide

Glycosaminoglycan sulodexide, a mixture of dermatan sul-
fate and fast-moving heparin, was given orally in a double-
blind, crossover, placebo-controlled study of its effect on
blood hemorrheology. 26 It decreased plasma fibrinogen con-
centration and had a marked effect on plasma viscosity, but
had no effect on whole-blood viscosity. It had no anticoagulant
effect. No significant side-effects were noted. Its potential use
is in patients with claudication or vascular disease. It is not
approved by the FDA.

Dipyridamole

Dipyridamole has a broad spectrum of pharmacologic reac-
tions, including an inhibition of adenosine reuptake into blood
and vascular cells, inhibition of cyclic adenosine monophos-
phate phosphodiesterase, and inhibition of red cell-induced
platelet activation. It has been shown by Saniabadi and associ-
ates to increase human red cell deformability. 27 The mecha-
nism was not identified, and its effect on blood viscosity is not
known, nor is its benefit in patients with peripheral vascular
disease.

470

chapter 39 Pharmacologic intervention: vasodilation therapy and rheologic agents

Metabolic enhancers

Carnitine

Ischemia causes a disturbance in lipid and carnitine metabo-
lism. Because carnitine has a role in the oxidation of long-chain
fatty acids in skeletal muscle, it has received some attention in
the treatment of patients with peripheral vascular disease, l-
Carnitine has been reported to increase the walking distance in
patients with claudication, 28 but this needs further confirma-
tion. It is approved by the FDA for use in patients with sys-
temic carnitine deficiency.

Vitamin E (tocopherol)

Vitamin E protects mitochondria from the consequences of ex-
perimentally induced ischemia. In addition, the deformability
of the red cell may be enhanced by vitamin E. Kleijnen and col-
leagues reviewed studies of the effects of vitamin E by a
MEDLINE computer search (1963-1988). 29 Although the
dosage and duration of the studies varied considerably, there
seemed to be some positive effects of vitamin E in the treat-
ment of intermittent claudication. The authors noted that larg-
er, well designed, double-blind trials are necessary to confirm
or reject these suggestions.

Naftidrofuryl

Naftidrofuryl in animals has a diuretic effect on tissue oxida-
tive metabolism, activating succinic dehydroxygenase and
thereby promoting cellular glucose consumption and increas-
ing the supply of ATP in skeletal muscle. 30 Although clinical
trials for patients with claudication have shown improve-
ment, this has been small and probably not of clinical signifi-
cance. 30 It is an investigational drug in this country.

Conclusion

According to the Consensus Document on Leg Ischemia,
"there presently is inadequate evidence from published
studies to support the routine use of primary pharmacological
treatment in patients with CLI." 1 Likewise, after an analysis of
clinical trials of drug treatment of intermittent claudication,
Cameron and associates concluded, "Despite 75 trials of 33
drugs, it is still unclear whether any of these pharmacological
agents has a clinically relevant effect on intermittent claudica-
tion." 24 This could have been expected in the use of vasodila-
tors in patients with fixed arterial occlusion; however, it was
hoped that decreasing blood viscosity would improve flow
and perfusion. The results have been disappointing. Although
new drugs are being evaluated, the preliminary data do not
look promising.

On the other hand, certain vasodilator drugs have made a
dramatic improvement in Raynaud's phenomenon.
Nifedipine has been the standard therapy, and in recalcitrant
cases, intravenous iloprost can be useful.

References

1. European Working Group on Chronic Leg Ischemia. Second
European consensus document on chronic critical leg ischemia.
Circulation 1991; 84:1.

2. Shub C. Medical treatment of intermittent claudication. Hosp
Formul 1991; 26:575.

3. Lowe GDO. Drugs in cerebral and peripheral arterial disease. Br
Med J 1990; 300:524.

4. Dormandy JA. Use of the prostacyclin analogue iloprost in the
treatment of patients with critical limb ischemia. Therapie 1991;
46:319.

5. Dormandy JA. Clinical experience with iloprost in the treatment
of critical leg ischemia: cardiovascular significance of
endothelium-derived vasoactive factors. Therapie 1991; 46:335.

6. Grant SM, Goa KL. Iloprost: a review of its pharmacodynamic
and pharmacokinetic properties, and therapeutic potential in
peripheral vascular disease, myocardial ischaemia and extracor-
poreal circulation procedures. Drugs 1992; 43:889.

7. DeCree J, Leempoels J, Geukens H, Verhaegen H. Placebo-
controlled double-blind trial of ketanserin in treatment of inter-
mittent claudication. Lancet 1984; 2:775.

8. Bounameaux H, Holditch T, Hellemans H, Berent A, Verhaeghe R.
Placebo-controlled, double-blind, two-centre trial of ketanserin in
intermittent claudication. Lancet 1985; 2:1268.

9. Cameron HA, Waller PC, Ramsay LE. Placebo-controlled trial of
ketanserin in the treatment of intermittent claudication. Angiology
1987; 38:549.

10. Clement DL, Duprez D. Effect of ketanserin in the treatment of
patients with intermittent claudication: results from 13 placebo-
controlled parallel group studies. / Cardiovasc Pharmacol 1987;
10:589.

11. Prevention of Atherosclerotic Complications With Ketanserin
Trial Group: Controlled trial of ketanserin. Br Med J 1989; 298:
424.

12. Blombery PA. Intermittent claudication: an update on manage-
ment. Drugs 1987; 34:404.

13. Gjorup T, Kelback H, Hartling OJ, Nielsen SL. Controlled double-
blind trial of the clinical effect of nifedipine in the treatment of
idiopathic Raynaud's phenomenon. Am Heart J 1986; 111:742.

14. Corbin DO, Wood DA, Macintyre CC, Housley E. A randomized
double blind cross-over trial of nifedipine in the treatment of pri-
mary Raynaud's phenomenon. Eur Heart} 1986; 7:165.

15. Rademaker M, Cooke ED, Aimond NE et ah Comparison of intra-
venous infusions of iloprost and oral nifedipine in treatment of
Raynaud's phenomenon in patients with systemic sclerosis: a
double blind randomised study. Br Med J 1989; 298:561.

16. Wigley FM, Seibold JR, Wise RA, McCloskey DA, Dole WP Intra-
venous iloprost treatment of Raynaud's phenomenon and is-
chemic ulcers secondary to systemic sclerosis. / Rheumatol 1992;
19:1407.

17. Lukac J, Rovensky J, Tauchmannova H, Zitnan D. Long-term

471

part iv Medical management

ketanserin treatment in patients with systemic sclerosis and
Raynaud's phenomenon. Curr Ther Res 1991; 50:869.

18. Codella O, Caramaschi P, Olivieri O et ah Controlled comparison
of ketanserin and nifedipine in Raynaud's phenomenon.
Angiology 1989; 40:114.

19. Young JR. Treatment of upper extremity vasospastic disorders. In:
Ernst CB, Stanley JC, eds. Current Therapy in Vascular Surgery, 2nd
edn. Philadelphia: BC Decker, 1991:191.

20. Putnam TC, Kevy SV, Replogle RL. Factors influencing the
viscosity of the blood. Surg Gynecol Obstet 1967; 124:547.

21. Replogle RL, Kundler H, Gross RE. Studies on the hemodynamic
importance of blood viscosity. / Thorac Cardiovasc Surg 1965;
50:658.

22. Johnson GJ, Keagy BA, Ross DW, Gabriel DA, Lucas CL, Hardison
VC. Viscous factors in peripheral tissue perfusion. / Vase Surg
1985; 2:530.

23. Ward A, Clissold SP Pentoxifylline: a review of its pharmacody-
namic and pharmacokinetic properties, and its therapeutic
efficacy. Drugs 1987; 34:50.

24. Cameron HA, Waller PC, Ramsay LE. Drug treatment of intermit-
tent claudication: a critical analysis of the methods and findings of

published clinical trials, 1965-1985. Br J Clin Pharmacol 1988;
26:569.

25. Wiles PG, Nelson SR, Hampton KK, Casali B, Boothby M, Prentice
CRM. Therapeutic defibrinogenation by ancrod: effect on limb
blood flow in peripheral vascular disease. Blood Coagul Fibrin
1990; 1:385.

26. Lunetta M, Salanitri T. Lowering of plasma viscosity by the oral
administration of the glycosaminoglycan sulodexide in patients
with peripheral vascular disease. JInt Med Res 1992; 20:45.

27. Saniabadi AR, Fisher TC, Lau CS et ah Dipyridamole increases
human red blood cell deformability Eur J Clin Pharmacol 1992;
42:651.

28. Brevetti G, Chiariello M, Ferulano G et ah Increases in walking dis-
tance in patients with peripheral vascular disease treated with
L-carnitine: a double-blind, cross-over study. Circulation 1988;
77:767.

29. Kleijnen J, Knipschild P, ter Riet G. Vitamin E and cardiovascular
disease. Eur J Clin Pharmacol 1989; 37:541.

30. Bevan EG, Waller PC, Ramsay LE. Pharmacological approaches to
the treatment of intermittent claudication. Drugs Aging 1992;
2:125.

472

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Pharmacologic intervention
lipid-lowering agents

Ralph G. DePalma

This chapter reviews the uses of lipid-lowering agents when
diet and exercise fail to reduce serum lipids to ranges consid-
ered desirable. After 3-6 months of dietary intervention, if
total cholesterol is found to be above 240 mg/ dl (high), or if it is
above 200 mg/dl (borderline high) and coronary heart disease
(CHD) or two other risk factors for atherosclerosis are present,
drug therapy is indicated. The management of these people is
determined largely by the level of low-density lipoprotein
cholesterol (LDLC), which should be below 100 mg/dl, or
lower after coronary events . A variety of lipid-lowering agents
is available; this chapter outlines their indications, actions,
risks, and benefits.

It is estimated that about 25% of American adults have high
total cholesterol levels; another 30% are estimated to be in the
borderline range. 1 Lipid-lowering regimens slow the progres-
sion of coronary atherosclerosis, reduce the risk of coronary
events, and, in some cases, lead to angiographic evidences of
regression. 2 Because of this consensus, it is highly probable
that patients failing to respond to diet and exercise will be
placed on drugs by their physician.

Indications

A number of primary and secondary dietary intervention trials
have shown that the total cholesterol reduction over a 3- to 8-
year period ranged from a low of 9% to a high of 15%? In only
one of these studies 3 was there a significant difference at 5 years
between fatal and nonfatal CHD events. Because of the modest
effects of diet in usual circumstances, the use of drugs has
become more common. In addition, although homozygous
familial hypercholesterolemia is rare, the genes 3 for familial
hypercholesterolemia are not uncommon — the heterozygous
state exists in about 1 in 500 live births. 4 In a randomized,
controlled trial of 72 patients with heterozygous familial
hypercholesterolemia, unequivocal evidence of regression
of coronary lesions was observed during treatment with
colestipol and niacin, or colestipol and lovastatin regimens. 5
In all of the 13 trials involving angiographic studies, modest

improvements in angiographic appearance occurred in a
relatively short period of intensive lipid lowering. Improved
appearance relates to about a 60% reduction in coronary event
endpoints. 6 Similar results in angiographic appearance and
coronary endpoints were reported for ileal bypass. 7

The cholesterol-lowering atherosclerosis studies, CLAS
I and II, 8 are of particular interest. They showed at 4 years
that significantly more drug-treated subjects demonstrated
nonprogression and regression in native coronary lesions
compared with placebo-treated patients. These results also
indicated a need for long-term lipid-lowering therapy
after coronary bypass. The operant therapy in these studies
was the combination of colestipol and nicotinic acid.

Standard drugs in the United States

The available drugs in the United States are cholestyramine
and colestipol (bile acid sequestrants), nicotinic acid (a B-
complex vitamin that inhibits the hepatic synthesis of very-
low-density lipoprotein [VLDL]), lovastatin, pravastatin,
simvastatin and atorvastatin (HMG CoA reductase inhibitors
that inhibit cholesterol biosynthesis), gemfibrozil (a fibric acid
derivative whose mechanism of action is not established), and
probucol (an antioxidant that also reduces low-density
lipoprotein [LDL] levels). The most experience in trials has
been with the combination of the bile acid sequestrants and
nicotinic acid. Table 40.1 updates drug classes and actions as
described by Hunninghake. 9 In addition, dietary supplemen-
tation with certain vitamins may act by virtue of their antioxi-
dant properties. There has been interest in vitamin E as a
primary antioxidant of lipoprotein membranes. 10 ' 11 Vitamin E
is the principal tissue antioxidant in humans; it acts synergisti-
cally with vitamin C to provide a concentration-dependent
protection against oxidation of the lipid-soluble antioxidants.
Vitamin C also acts by regenerating the reduced form of
tocopherol. 12

Early studies reporting the CHD risk related inversely with
vitamin A, vitamin E, and (3-carotene consumption. 13 ' 14 In ad-

473

part iv Medical management

Table 40.1 Lipid-lowering agents

Class

Actions

DRUGS

Cholestyramine
Colestipol

Nicotinic acid

Lovastatin
Pravastatin
Simvastatin
Atorvastatin

Gemfibrozil

Probucol

NATURAL VITAMINS
Vitamin C

Vitamin E

Bile acid sequestrants

B-complex vitamin

HMG CoA reductase inhibitors

Fibric acid derivative

Antioxidant

Aqueous phase chain-breaking antioxidant

Lipid-phase (lipoprotein and membrane) chain-breaking
antioxidant

Increase fecal bile acid excretion

Inhibit hepatic synthesis of VLDL

Inhibit hepatic cholesterol biosynthesis and partially inhibit hepatic
lipoprotein synthesis
Anti-inflammatory activity

Mechanism not established

Block macrophage
LDL uptake

Block macrophage LDL uptake (?); increase HDL (?)
Block macrophage LDL uptake (?); increase HDL (?)

VLDL, very-low-density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein.

(From Hunninghake DB. Drug treatment of dyslipoproteinemia. Endocrinol Metab Clin North Am 1990; 19:345-360.)

dition, direct relationships between ascorbic acid levels and
high-density lipoprotein (HDL) levels have been described in
the presence of vitamin C deficiency. 15 In some studies, vita-
min E has also had beneficial effects in increasing HDL choles-
terol (HDLC) levels. 16 However in recent trials, vitamin C and
E supplementation had no effect in the prevention of CHD 17 or
in improving outcomes of established CHD. 18

Each of the drugs exhibits somewhat different effects on
blood lipid levels and lipoproteins; these are summarized rec-
ognizing recent anti-inflammatory effects of statins in Table
40.2. Cholestyramine and colestipol, which act by increasing
fecal excretion of bile acids, have the side-effect of constipation
and lack of palatability The sequestrants require administra-
tion as a powder. Based on personal experience with these
agents, patients have not accepted bile acid sequestrants with
enthusiasm. In the case of nicotinic acid, flushing and other
cardiovascular effects occur. Fulminant hepatic failure has
been reported after the ingestion of sustained-release nicotinic
acid; liver functions must be monitored. Niacin also may ag-
gravate hypercholesterolemia in certain cases, particularly in
diabetic patients. Lovastatin, pravastatin, and simvastatin all
interfere with HMG CoA reductase, the rate-limiting enzyme
in cholesterol synthesis. They are easy to administer, but re-
quire monitoring of liver functions at intervals of about 6
weeks. Lovastatin's long-term effects on CHD risk or on coro-
nary angiographic appearances have yet to be documented. It
is stated that pravastatin carries less of a risk for myopathy and
disturbed liver functions than does lovastatin. Gemfibrozil

Table 40.2 Effects of available drugs on blood lipids and lipoproteins (%
change)

Drug

LDLC

Triglyceride

HDLC

Cholestyramine

i 1 5-30

Variable

T3-5

Colestipol

Nicotinic acid

115-25

I 20-50

T 1 5-30

Lovastatin

120-40

110-25

T5-10

Gemfibrozil

Variable

i 20-50

T 1 0-1 5

Probucol

110-15

No change

T 20-25

LDLC, low-density lipoprotein cholesterol; HDLC, high-density lipoprotein

cholesterol.

(From Hunninghake DB. Drug treatment of dyslipoproteinemia. Endocrinol

Metab Clin North Am 1 990; 1 9:345-360.)

and clofibrate are fibric acid derivatives; their mechanisms of
action are not well established. These drugs have triglyceride-
lowering properties, and potential mechanisms include inhi-
bition of hepatic triglycerides, VLDL or apolipoprotein B
synthesis, and increased VLDL clearance. 9 This class of drugs
is useful to reduce the risk of pancreatitis associated with high
triglyceride levels, and is thought to benefit that risk class of
patients with elevated LDLC and triglycerides and lowered
HDLC. The antioxidant, probucol, is a lipid-phase lipoprotein
and membrane chain-breaking oxygen radical scavenger; it
also decreases LDLC levels. As described in Chapter 37, LDLC
that is oxidized or otherwise modified enters macrophages in

474

chapter 40 Pharmacologic intervention: lipid-lowering agents

the arterial wall. 19 Probucol has the disadvantage of decreas-
ing HDLC levels. Vitamin C and vitamin E appear safe at high
dosages.

Clinical correlates

Almost all of the trials related to aggressive antilipid therapy
for atherosclerosis, save ileal bypass, have used drugs and
focused on CHD. In at least two trials, 20 ' 21 including ileal
bypass, 22 there was a significantly decreased incidence of
peripheral arterial disease and slowing of progression during
aggressive therapy. The clinical benefits of drug therapy
almost always have accrued using combined treatment
regimens. The most dramatic has been the colestipol and
niacin combination; these unfortunately also are those with
the lowest compliance. 23 The HMG Co A reductase inhibitors
are considered to be a major advance in the treatment of
elevated LDLC. Inhibiting cholesterol synthesis stimulates
an increase in LDL receptors, reducing plasma concentrations
of LDL, IDL, and VLDL. Data on the efficacy of these drugs
for coronary artery disease reduction, or for slowing the
progression of atherosclerosis, are unavailable. HMG Co A
reductase inhibitors, however, are the most popular agents
prescribed. About 1.3% of patients exhibiting greater than
threefold elevations of hepatic transaminase will require dis-
continuance of the drug. Elevations of creatine phosphokinase
(CPK) may develop in some patients; in 0.1-0.2% myalgia may
develop; when associated with greater CPK elevations,
the drug must be discontinued. Severe myopathies with
rhabdomyolysis can occur in heart transplant patients when
lovastatin is taken with cyclosporin. Concomitant administra-
tion of niacin or gemfibrozil also has been limited because
of myopathy. 24 ' 25

Nicotinic acid, in spite of its unpleasant side-effects of flush-
ing as well as liver toxicity and elevation of serum glucose and
uric acid levels, provides the most favorable response in in-
creasing HDL, which is low in peripheral vascular disease.
Gemfibrozil is well tolerated by many patients; however, there
is an increased risk for cholesterol gallstones, as will occur on
any regimen in which body weight or lipids are suddenly
altered.

Oxidation of LDL and (3-VLDL within the arterial wall is
thought to be a critical step in rendering these lipoproteins
atherogenic. 19 Probucol is an antioxidant carried within the
lipoprotein particle that prevents its oxidation. Animal experi-
ments show probucol prevented progression of atherosclero-
sis in an inherited atherosclerosis model much more than
would have been expected from LDL reduction alone. There
has been interest in the role of natural vitamins, particularly E
and C, where experimental reports suggest that the combina-
tion of these two vitamins is effective in preventing LDL oxi-
dation, 10-12 but this is not supported by trials.

Vascular surgeons will have noted the heightened aware-

ness of the importance of elevated cholesterol and continued
stimuli for management. The public attention to lipids and ex-
ploration of the many links between lipid abnormalities and
atherosclerosis will continue. More patients with peripheral
artery disease being treated with lipid-lowering drugs will be
seen. Between 1983 and 1989, visits to physicians for elevated
cholesterol alone increased ninefold. 26 Between 1983 and 1988,
there was a fivefold increase in the dispensing of cholesterol-
lowering drugs by retail pharmacies, the most common of
which were gemfibrozil and lovastatin. 26

Attention has been drawn to the substantial overall expen-
ditures for these drugs when used for primary prevention. It
has been estimated that the annual cost of therapy with lovas-
tatin at 80 mg/day is $1881 P It is not known whether an ener-
getic population approach to cholesterol lowering by diet
ultimately will prevail over the need for drug therapy. There is
evidence to suggest that fat intakes have been falling steadily
since 1960 to approximately 30% of the total energy intake in
1984. This is still a relatively high fat intake measured against
the dietary guidelines that have been made for the step I diet,
which recommend a total fat intake of less than 30% of total
caloric intake, with further reduction in the step 2 diet, in
which fat is reduced to 7% of the total caloric intake, and
cholesterol intake to less than 200 mg/day. 27

Ultimately, trials comparing lipid interventions, particul-
arly those that increase HDL and decrease lipoprotein (a), will
be needed to determine efficacy for peripheral atherosclerosis.
The comparative importance of the antioxidants, platelet-
altering drugs, and anticoagulants will need to be integrated
into a rational medical approach to therapy. Such an integrat-
ed approach will complement and not compete with estab-
lished surgical treatment of advanced life- or limb-threatening
atheromatous disease.

A provocative issue is the applicability of diet or drugs in
lipid lowering when aneurysmal disease or a tendency to
aneurysmal disease coexists with obstructive atheromas.
There are experimental suggestions 28 ' 29 that lipid lowering
and atheroma regression predispose to aneurysm develop-
ment. This might relate to structural impairment due to reab-
sorption of the atheroma or possibly to altered structural
protein dynamics. Clinically, weight loss and drastic lipid re-
duction have been observed coincident with the appearance
and progression of aneurysmal disease. 30 Although this phe-
nomenon may be coincidental, further scrutiny is needed.
Overall, drugs for lipid lowering are an essential part of the
medical treatment for vascular patients. Further refinements
of indications and contraindications, along with data on long-
term effects, continue to emerge.

Acknowledgment

The author acknowledges the helpful suggestions of John C.
LaRosa, Dean for Research, George Washington University
School of Medicine, Washington, DC.

475

part iv Medical management

References

1. The Expert Panel. Report of the National Cholesterol Education
Program Expert Panel on detection, evaluation, and treatment of
high blood cholesterol in adults. Arch Intern Med 1988; 148:36.

2. Rossouw JE, Rifkin BM. Does lowering serum cholesterol levels
lower coronary heart disease risk? Endocrinol Metab Clin North Am
1990; 19:279.

3. Leren P. The Oslo diet — heart study: eleven year report. Circulation
1987; 76:515.

4. Schonfeld G. Inherited disorders of lipid transport. Endocrinol
Metab Clin North Am 1990; 19:211.

5. Kane JP, Malloy MJ, Ports TA et ah Regression of coronary athero-
sclerosis during treatment of familial hypercholesterolemia with
combined drug regimens. JAMA 1990; 264:3007.

6. Rossouw JE. Angiographic Studies in NIH Consensus Development
Conference. Bethesda: National Heart Lung and Blood Institute,
1992:71.

7. Buchwald H, Matts JP, Fitch LL et ah Changes in sequential
coronary arteriograms and subsequent coronary events. JAMA
1992; 268:1429.

8. Cashin-Hemphill L, Mack WJ, Pogoda JM et ah Beneficial effects
of cholestipol niacin on coronary atherosclerosis. JAMA 1990;
264:3013.

9. Hunninghake DB. Drug treatment of dyslipoproteinemia.
Endocrinol Metab Clin North Am 1990; 19:345.

10. Szczeklik A, Gryglewski RJ, Domalga B et ah Dietary supplemen-
tation with vitamin E in hyperlipidemia: effects on plasma lipid
peroxides, antioxidant activity, prostacyclin generation and
platelet aggregability. Thromb Haemost 1985; 54:425.

11. Princen HMG, vanPoppel G, Vogelezang C. Supplementation
with vitamin E but not B carotene in vivo protects low density
lipoprotein from lipid peroxidation in vitro. Arterioscler Thromb
1992; 12:554.

12. Jialal I, Vega GL, Grundy SM. Physiologic levels of ascorbate
inhibit the oxidative modification of low density lipoprotein.
Atherosclerosis 1990; 82:185.

13. Gey KF, Puska P, Jordan P et ah Inverse correlationbetween plasma
vitamin E and mortality from ischemic heart disease in cross
arterial epidemiology. Am J Clin Nutr 1991; 53:3265.

14. Gey FK, Bruhacher GB, Stabelin HB. Plasma levels of antioxidant
vitamins in relation to ischemic heart disease and cancer. Am J Clin
Nutr 1987; 45:1368.

15. Trout DL. Vitamin C and cardiovascular risk factors. Am J Clin
Nutr 1991; 55:3235.

16. Muckle TJ, Nazir DJ. Variation in human blood high density
lipoprotein response to oral vitamin E megadosage. Am J Clin
Pathol 1989; 91:165.

17. Muntwyler J, Hennekens CH, Manson JE et ah Vitamin supple-
ment use in a low-risk population of US male physicians and
subsequent cardiovascular mortality. Arch Intern Med 2002; 162:
1472.

18. MRC /BHF Heart Protection Study of antioxidant vitamin supple-
mentation in 20, 536 high risk individuals: a randomized placebo-
controlled trial. Lancet 2002; 360:22.

19. Steinberg D, Witzum JL. Lipoproteins and atherogenesis: current
concepts. JAMA 1990; 264:3047.

20. Blankenhom DH, Brooks SH, Seltzer RH, Barndt RJ. The rate of
atherosclerosis change during treatment of hyperlipoproteine-
mia. Circulation 1978; 57:355.

21 . Duffield RGM, Miller NE, Brunt HRT et ah Treatment of hyperlipi-
demia retards progression of symptomatic femoral atherosclero-
sis: a randomized controlled trial. Lancet 1983; 2:639.

22. Buchwald H, Varco RL, Matts JP et ah Effects of partial ileal bypass
surgery on mortality and morbidity from coronary heart disease
in patients with hypercholesterolemia: report of the program on
the surgical control of the hyperlipidemias (POSCH) . N Engl J Med
1990; 323:946.

23. Schulman KA, Kinosian B, Jacobson MD et ah Reducing high
blood cholesterol level with drugs: cost effectiveness of pharma-
cologic management. JAMA 1990; 264:3025.

24. Tobert JA. Efficacy and long-term adverse pattern of lovastatin.
Am J Cardiol 1988; 62:23].

25. Catalano PM, Masonson HN, Newman TJ et ah Clinical safety
of pravastatin. In: LaRosa JC, ed. New Advances in the Control
of Lipid Metabolism: Focus on Pravastatin. London: Royal Society
of Medicine Services, 1989:35.

26. Rifkin BM, Grouse LO. Cholesterol redux. JAMA 1990; 264:3061.

27. Stone WJ. Diets, lipids and coronary heart disease. Endocrinol
Metab Clin North Am 1990; 19:321.

28. DePalma RG, Kaletsky S, Bellon EM et ah Failure of regressions
of atherosclerosis in dogs with moderate cholesterolemia.
Atherosclerosis 1977; 27:297.

29. Zarins CK, Glagov S, Vesselinovitch et ah Aneurysm formation in
experimental atherosclerosis: relationship to plaque evolution /
Vase Surg 1990; 12:246.

30. DePalma RG, Sidawy AN, Giordano JM. Associated etiological
and atherosclerotic risk factors in abdominal aneurysms. In: EO
Greenhalgh RM, Mannick JA, eds. The Cause and Management of
Aneurysms. Philadelphia: WB Saunders, 1990:37.

476

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Infections and antibiotics in
vascular surgery

Martin R. Back

Bacterial infection complicates many of the wounds or lesions
resulting from pathological vascular conditions including is-
chemic tissues from arterial occlusive disease, diabetic foot
wounds, venous stasis ulcers due to chronic venous insuffi-
ciency, and recurrent cellulitis in patients with lymphedema.
Palliative procedures for these conditions are also associated
with an elevated risk of wound-related problems owing pri-
marily to the underlying vascular pathology and compro-
mised host defenses. Available surgical and endovascular
interventions for peripheral arterial occlusive disease and
aneurysms have been greatly expanded by the use of pros-
thetic biomaterials (polymer grafts, plastic catheters, metallic
stents) for vessel reconstruction and luminal recanalization.
Although the incidence of infection involving arterial prosthe-
ses is relatively low, the consequences of delayed or improper
treatment are potentially catastrophic. Surgical intervention is
required via one of several treatment options to avoid life-
threatening sepsis, rupture of mycotic pseudoaneurysms, or
limb-threatening graft thrombosis. Eradication of infection
primarily by removal of the involved prosthesis and mainte-
nance of perfusion to viable limbs and end-organs remain the
goals of surgical management. Treatment options are influ-
enced by clinical presentation, anatomic location, extent and
invasiveness of infection, virulence of the infecting organism,
type of graft material, patient comorbidities and overall clini-
cal status. Antibiotics are an important adjunct in the treat-
ment of infections involving wounds and lesions associated
with pathological vascular conditions, postoperative wound
infections, and established prosthetic infections. Nearly all in-
fections encountered in vascular surgery involve bacteria and
continued advances in pharmacotherapeutics have been
necessary to counteract evolving bacterial resistance mecha-
nisms. With the exception of late systemic Candidal infections
occurring in immunocompromised patients or in critically ill
patients after prolonged courses of antibacterial agents, vas-
cular infections caused by fungi and other microbes (e.g.
mycobacteria) are exceedingly rare. This chapter reviews the
etiology and pathophysiology of vascular prosthetic infect-
ions, current techniques and future directions for treating

infected arterial grafts, and recent antibiotic developments
pertinent to vascular surgery.

Pathophysiology of vascular prosthetic
infection

Classification and incidence

Two existing classification schemes have been used to catego-
rize infections involving prosthetic grafts used for vascular
reconstruction (Table 41.1). The Szilagyi et al. classification 1
describes the spectrum of early postoperative wounds compli-
cated by infection and culminating in direct involvement of the
prosthesis (grade III). Bunt 2 differentiated early (<4 months)
and late-appearing perigraft infections each of which can
be characterized by specific bacterial species involved and clini-
cal presentation. Graft-enteric erosions or fistulas (GEE/GEF)
occur late (typically >5 years) after implantation and account
for approximately 20% of aortic graft infections. Residual infec-
tion involving the infrarenal aortic "stump" after excision of
an infected aortic graft and extraanatomic lower limb revascu-
larization is also defined in the Bunt scheme.

Prosthetic grafts used for arterial reconstruction are at an
increased risk for infection compared with repairs done with
autologous conduits (superficial or deep limb veins). From
collected series, the overall incidence of vascular prosthetic in-
fection complicating arterial operations ranges from 0.5 to 5 %
(Table 41.2). Graft infection is more common when groin expo-
sure and anastomosis to the femoral artery is necessary or
when the graft is placed in a subcutaneous tunnel (e.g. ax-
illofemoral or cross-femoral bypass). Infrainguinal bypasses
performed with autologous vein have a low incidence of
infection (< 1%) relative to prosthetic graft reconstructions.

Endovascular interventions may lessen the overall phy-
siological stress to the patient compared with open surgical
reconstructions and appear to have a lower risk of
procedure-related infection. However, numerous case reports
exist of morbid outcomes associated with predominantly

477

part iv Medical management

Table41.1 Classification schemes for vascular graft infections

Szilagyi etal. classification 1 —applicable for early postoperative/wound
infections

Grade I— cellulitis involving implantation wound

Grade II — infection involving subcutaneous tissue

Grade III — infection involving vascular prosthesis

Bunt classification 2 —terminology for graft infections
Perigraft infection

Early postoperative (<4 months from implantation)

Late-appearing (>4 months)
Graft-enteric erosion or fistula (GEE/GEF)
Infrarenal aortic stump sepsis

Table 41 .2 Incidence of prosthetic vascular graft infections relative to
implant site

Graft site

Incidence (%)

Descending thoracic aorta

0.7-3

Aortoiliac or aortic tube graft

0.4-1.3

Aortofemoral

0.5-3

Extraanatomic (cross-fern, axfem)

1.3-6

Femoropopliteal/tibial

0.9-4.6

Carotid patch

0.2

Innominate, carotid, subclavian bypass

0.5-1.2

Arterial stent

<0.5

Endovascular stent-graft

<1?

early (<1 month) infections occurring in percutaneously
placed arterial or venous stents. 3 These cases far outnumber
reports of infection-related complications after balloon angio-
plasty, thus demonstrating the finite potential for bacterial
colonization of all implanted biomaterials. Iliac artery stents
in a swine model show a decreased susceptibility to infection
induced by bacteremic (Staphylococcus aureus) challenges
at 3 months after deployment relative to earlier exposures
(<1 month), probably due to protective arterial wall healing. 4
Prophylactic antibiotics (i.e. before bacterial challenge) sig-
nificantly reduced iliac stent infection at the susceptible early
time points. Similarly, infection complicating clinical use of
endovascular stent-graft devices for arterial aneurysms, oc-
clusive disease, or traumatic lesions appears to be uncommon
but only limited reporting and short-term follow-up are
available. Preliminary animal data suggest that endoluminal
stent-grafts may have a greater susceptibility for infection
than prosthetic interposition grafts when challenged by local
application of S. aureus. 5 Endoluminal grafts have neverthe-
less been used for exclusion of mycotic descending thoracic
aortic aneurysms 6 and in the presumed infected field of aor-
toenteric fistulas 7-11 with mixed results. Clinically significant
residual or recurrent infection described in several of these re-
ports suggests that endovascular treatment of these lesions
may only serve as a "bridge" to open surgical repair in hemo-

dynamically unstable or severely debilitated or ill patients. 9-11
Further understanding of endoluminal device susceptibility
to infection is clearly needed as endovascular techniques are
increasingly utilized and seek wider application.

Host defenses and prosthetic healing

Normal host defenses against infection include physical bar-
riers (e.g. skin), and tissue defenses including nonspecific
cellular and humoral components and specific immunologic
elements. Bacterial colonization on skin is minimized by dry-
ness and continual desquamation. Intertriginous regions such
as the groin, axilla, and perineum accumulate moisture and
harbor higher bacterial counts. Common flora in these
locations and present in sebaceous and sweat glands are
Staphylococcus epidermidis, diphtheroids, Streptococcus viridans
and faecalis, Candida species, and some Gram-negative col-
iform bacteria. Staphylococcus aureus is a resident of the anteri-
or nares and intertriginous skin regions in 10-40% of healthy
individuals and up to 70% of diabetics and patients with
chronic renal failure on hemodialysis. 12,13 Infection requires
disruption of the skin or mucosal surface and microbial inocu-
lation into the wound. Entry of large numbers of virulent
microbes, depressed host immune function, invasion of sites
more remote from host defenses, and the presence of foreign
bodies (e.g. biomaterials) are all factors favoring development
of an infection.

Neutrophils and macrophages provide an essential tissue
defense against infection. Phagocytic response is rapid, non-
specific and occurs well before antibody and cell-mediated
(i.e. lymphocytic) immune mechanisms. Despite having a
brief life span of 1-2 days, neutrophils are the initial phagocyte
in infected tissue and are recruited from large numbers nor-
mally circulating in the blood and from an even larger pool
marginating in perivascular spaces of the microcirculation.
Neutrophil function after bacterial invasion is dependent
on chemotaxis, phagocytosis, and intracellular killing. Neu-
trophil migration involves pseudopod formation along con-
centration gradients of chemotactic mediators and can occur
in anaerobic and acidic environments. Key chemotactic factors
include components of the clotting and complement cascades,
certain cytokines and bacterial peptides. Bacteria coated with
opsonins (IgG and C3b of complement cascade) are recog-
nized and bound to phagocytic cell membrane receptors,
engulfed by cell pseudopods and internalized in vacuoles.
Vacuoles fuse with intracellular neutrophil granules exposing
bacteria to lysozyme and cationic proteases, which degrade
microbial cell walls, membranes, and lactoferrin, which binds
iron required for bacterial growth. The respiratory burst
within neutrophils generates molecular oxygen byproducts
greatly accelerating intracellular killing of ingested bacteria.
Microbial destruction by mechanisms beyond neutrophil de-
granulation requires an aerobic environment, nicotinamide-
adenine dinucleotide phosphate (NADPH) oxidase present

478

CHAPTER41 Infections and antibiotics in vascular surgery

on the neutrophil membrane, and glucose as an oxidative
energy source. After initial conversion of molecular oxygen
to hydrogen peroxide by NADPH oxidase and superoxide
dismutase, further toxic intermediates are formed including
singlet oxygen, hydroxyl radicals, and hypochlorous acid
(formed in the presence of chloride anions and myeloperoxi-
dase). These intermediates are short-lived and destroy bacter-
ial membranes within phagocytic vacuoles. Ischemia and
associated local tissue hypoxia can hinder phagocytic killing
and may facilitate bacterial growth. Although macrophage
migration into infected tissue lags behind initial neutrophil re-
cruitment, their phagocytic contributions remain essential for
healing /remodeling, and both cell types perpetuate local in-
flammatory responses by secreting cytokines such as tumor
necrosis factor (TNF) and various interleukins (e.g. IL-lp).

Prosthetic biomaterials do not play a passive role in the de-
velopment of graft infection. Immediately after implantation,
graft material stimulates a chronic inflammatory response
aimed at isolating the foreign body from adjacent tissue by
production of a fibrous adherent collagen capsule. This "incor-
poration" of the prosthesis involves mononuclear cell in-
filtration (lymphocytes and macrophages), angiogenesis and
capillary ingrowth, fibroblast proliferation, and collagen de-
position. Along the luminal surface, endothelial cells, smooth
muscle cells, and fibroblasts migrate from adjacent artery
over deposited fibrin, platelets, and thrombus or potentially
through the porous biomaterial via transinterstice capillary
ingrowth from perigraft tissue. 14 Extracellular matrix produc-
tion thickens the pseudointimal coverage of the biomaterial,
and luminal endothelial cell coverage is typically present only
at the ends of the graft within 1-2 cm of arterial anastomoses
and is limited in the central portions of bypasses done in hu-
mans. 15 Complete connective tissue incorporation is not es-
sential for pseudointimal formation, but some perigraft tissue
support is critical to its long-term existence. This is based on
the observations that pseudointima is not formed in areas
where initial tissue incorporation is absent (e.g. early perigraft
seroma, lymphocele, or abscess) and an increased risk of graft
thrombosis exists, although pseudointima can be present in re-
gions where adjacent connective tissue support is minimal. 16

The degree of inflammatory response to the prosthetic
graft is dependent on the biomaterial used and its surface
characteristics (e.g. texture, porosity, hydrophobicity).
Relatively porous materials such as knitted or woven Dacron
polyester are relatively more thrombogenic, develop a thicker
pseudointima, and allow extensive ingrowth of fibrous tissue.
Expanded polytetrafluoroethylene (PTFE) grafts have lower
porosity, allow less tissue ingrowth, and develop a thinner
well defined fibrous capsule along their outer surface. PTFE
grafts are more thromboresistant than equivalent diameter
Dacron conduits owing to a smoother surface and inherent
biochemical and physical properties of the biomaterial. These
factors contribute to development of a thinner, confluent
pseudointimal layer within PTFE grafts.

The inflammatory response to an implanted prosthetic graft
creates an unfavorable environment characterized by local is-
chemia and an acidic pH that is potentially conducive to bacte-
rial colonization. Preliminary animal studies suggest that the
perigraft environment may possess similar biochemical de-
rangements to those seen in nonhealing, chronic skin wounds
such as venous stasis ulcers. Disruption of the fine balance be-
tween pro- and anti-inflammatory mediators locally may lead
to excess production of matrix metalloproteases (MMPs) by
TNF-stimulated macrophages. 17 Excessive degradation of se-
creted extracellular matrix and angiogenic growth factors by
MMPs may hinder optimal graft healing by restricting capil-
lary ingrowth, tissue incorporation, and potential luminal en-
dothelialization. Lack of perigraft ingrowth and vascularity
also favors greater exposure of the implanted biomaterial to
bacteria and sequestration within graft pores /interstices
away from activated phagocytic cells. Neutrophil function can
also be directly impaired in the presence of biomaterials.
Decreased neutrophil opsonic, phagocytic, and bactericidal
activities against S. aureus have been observed in PTFE tissue
cages implanted subcutaneously in guinea pigs. 18

Pathogenesis and predisposing factors

Potential routes of vascular graft exposure to bacteria include
perioperative contamination, hematogenous seeding, and
mechanical erosion to skin, gastrointestinal, or genitourinary
tract structures. Direct contact with bacteria present on skin of
the implantation site, disrupted lymphatic channels and
nodes draining sites of remote infection or lower limb ulcera-
tion, bacteria harbored within diseased arterial wall, cross-
contamination with endogenous flora during concomitant
surgical procedures (e.g. cholecystectomy, bowel resection, or
genitourinary interventions 2 ), or postoperative wound infec-
tions can each predispose graft material to early colonization.
Even in the absence of a "break" in sterile technique, skin bac-
terial counts are reduced for several hours but not eliminated
by iodine-based surgical scrub solutions. Staphylococcal
species within deeper dermal layers and sweat glands are
relatively protected from standard topical bactericidal agents.
Recent percutaneous arterial access in the femoral region for
peripheral or coronary arteriography may increase the risk of
subsequent infection for implanted grafts requiring groin ex-
posure presumably due to introduction of skin flora into sub-
cutaneous tissues and local hematoma formation. 19 Chronic
wounds in the lower extremity provide another important in-
fectious source for prosthetic bypasses placed in the groin after
routine transection of abundant lymphatics during implanta-
tion. In a canine model, lymphatics are readily able to trans-
port Escherichia coli and S. aureus to proximally placed vascular
grafts. 20 Positive bacterial cultures have been obtained from
atheromatous arterial wall, aneurysm thrombus and periarte-
rial tissues in 10-43% of patients undergoing clean, elective
vascular reconstructions in several studies. 21-23 The most com-

479

part iv Medical management

mon organism cultured is S. epidermidis, which has also been
associated with occult infections affecting 60% of anastomotic
aneurysms developing after prosthetic aortofemoral bypass.
Although the incidence of graft infection is small compared
with the frequency of positive operative cultures, it appears
that the presence of bacteria in the arterial wall increases the
risk of prosthetic infection. All graft infections in one study
occurred in patients with positive arterial wall cultures at
implantation. 21 Postoperative wound complications includ-
ing infection, skin or fat necrosis, and seroma or lymphocele
formation can adversely affect prosthetic healing and pre-
dispose to graft infection. A septic focus is more likely to dev-
elop in devascularized tissue occurring in large, previously
radiated or reoperative wounds and extension to deep, peri-
graft locations (Szilagyi grade II or III) may occur.

Transient bacteremia from remote sites of infection with
hematogenous seeding of the luminal surface of a vascular
conduit is an uncommon but potentially important mecha-
nism of graft infection. Elderly vascular patients frequently
possess indwelling vascular catheters, infected urine from ob-
structive or retentive processes, or other remote tissue infec-
tions (e.g. pneumonia) that can serve as bacteremic sources.
Experimentally, a single intravenous infusion of 10 7 colony-
forming units of S. aureus will produce a clinical graft infection
in nearly all animals if given up to 1 month after prosthetic im-
plantation. 24 In canines, the prosthesis becomes less suscepti-
ble to colonization as a neointimal lining develops and more
complete endothelial cell coverage of the luminal surface oc-
curs although vulnerability to bacteremic seeding exists for up
to 1 year. 25 Prophylactic parenteral antibiotics (i.e. prior to bac-
teremic challenge) significantly diminished infection risk in
this model. Since pseudointimal formation and luminal en-
dothelialization are limited in humans, some graft suscep-
tibility to hematogenous seeding may exist indefinitely after
implantation.

Mechanical erosion of prosthetic aortoiliofemoral bypasses
into adherent small or less commonly large bowel results in
formation of GEF/GEE. These morbid lesions should theoreti-
cally be avoided by assurance of adequate retroperitoneal
tissue interposition between graft and adjacent bowel during
implantation using soft tissues (fat, lymphatic structures,
peritoneal lining, mesentery, omentum) and aortic wall (after
aneurysmectomy). However, primary infection of the pros-
thetic graft with secondary communication with the gastro-
intestinal tract may account for a significant fraction of
GEF/GEE observed clinically. Gram-positive bacteria rather
than intestinal flora are cultured from adjacent graft seg-
ments not in direct contact with the enteric fistula in many
cases suggesting an underlying initial prosthetic infection.
Furthermore, nearly half of GEF/GEEs are associated with
pseudoaneurysm of the proximal aortic anastomosis and are
also consistent with a primary graft infection.

Multiple perioperative and patient-related factors can ele-
vate the risk of developing vascular graft infection (Table 41.3).

Table 41 .3 Risk factors predisposing to bacterial contamination and
prosthetic graft infection

Perioperative factors

1 . Prolonged preoperative hospitalization

2. Infection in remote site

3. Recent percutaneous arterial access at implant site (peripheral or
cardiac angiography)

4. "Break" in aseptictechniques

5. Emergent/urgent operation

6. Reoperative vascular procedure

7. Extended operating time

8. Concomitant gastrointestinal or genitourinary procedure

9. Postoperative wound complication (superficial infection, wound-edge
necrosis, lymphocele)

Patient-related factors/altered host defenses

1. Malignancy

2. Lymphproliferative disorder

3. Autoimmune disease

4. Corticosteroid administration

5. Chemotherapy

6. Malnutrition

7. Diabetes mellitus

8. Renal failure

In addition to those contributing perioperative factors dis-
cussed above, risk of infection is increased with emergent, ex-
tended length, and reoperative 26 reconstructions. A prolonged
preoperative hospital stay allows conversion of normal skin
flora to hospital-acquired strains that can possess resistance to
routinely administered antibiotics. Early graft infections are
usually the result of wound complications, unplanned re-
operation for hematoma or lymphocele, concomitant remote
infection, and impaired immunocompetence. Patients with
late-appearing graft infections often have a history of multiple
operations for thrombosis or anastomotic aneurysm of pre-
viously placed prosthetic bypasses.

Impaired host defenses from underlying systemic condi-
tions also can predispose patients to prosthetic graft infection.
Altered immune function associated with malnutrition, mal-
ignancy, lymphproliferative disorders, autoimmune diseases,
and drug administration (e.g. corticosteroids, antineoplastic
agents, and antirejection regimens) may potentiate graft infec-
tion with lower numbers of contaminating bacteria. Although
controversial, there appears to be an increased susceptibility
to infection in diabetic patients. Mean plasma glucose levels in
diabetics prior to the development of infection and the preva-
lence of subsequent infections have been tightly correlated in
clinical practice. 12 At the cellular level, neutrophil chemotactic
and intracellular bactericidal mechanisms are diminished in
diabetics. Opsonization of S. aureus and E. coli is impaired
in diabetics. Hyperglycemia in diabetics also decreases neu-
trophil adherence to foreign bodies. In addition to the frequent
hematogenous introduction of bacteria associated with re-
peated access of hemodialysis catheters, grafts, and fistulas,

480

CHAPTER41 Infections and antibiotics in vascular surgery

Table 41 .4 Bacteriology of prosthetic vascular graft infections from collected series

Incidence (%)

Microorganism

Staphylococcus aureus

Staphylococcus epidermidis

Streptococcus sp.

Pseudomonas sp.

Coliforms and Gram-negative sp.

Other species

No growth/no culture

Thoracic aorta

GEE/GEF

Aortofem

Fem-pop-tib

* Escherichia coli, Enterococcus, Bacteroides, Klebsiella, Enterobacter, Serratia, Proteus species.
GEE/GEF, graft-enteric erosions or fistula.

patients with chronic renal insufficiency also have a higher
susceptibility to infection owing to immune suppression
caused by uremia. 13 Depressed neutrophil function in uremia
appears multifactorial. Neutrophil chemotaxis is inhibited by
an unknown agent in uremic serum that specifically blocks
chemotactic factor synthesis. However, neutrophils from ure-
mic patients placed in normal plasma also exhibit impaired
chemotaxis. Adherence, phagocytosis, and bacterial killing by
neutrophils and lymphocyte numbers and function are also
diminished in patients with acute or chronic renal failure.

Bacteriology

Gram-positive cocci, especially S. aureus and S. epidermidis, are
the most prevalent pathogens causing vascular prosthetic in-
fection (Table 41.4). 1/27_39 Since the early 1970s, graft infections
due to S. epidermidis and Gram-negative bacteria have in-
creased in frequency. More recently, two reports from Britain
have documented methicillin-resistant S. aureus (MRSA) as
the most common pathogen involved in vascular wound and
graft infections and have shown high associated morbidity
and mortality rates. 40 ' 41 Late-appearing graft infections
caused by S. epidermidis and other coagulase-negative staphy-
lococci are less virulent by comparison and typically reveal
negative cultures of perigraft fluid or tissue and Gram stain
showing only white blood cells. Recognition of these
"biofilm" infections in patients with obvious clinical and
anatomic manifestations of perigraft infection and acknowl-
edgment of the potential microbiological sampling error
when low bacterial numbers are present have improved detec-
tion and allowed appropriate management. 42 Techniques to
optimize recovery of coagulase-negative staphylococci in-
clude mechanical disruption of the bacteria from graft
material by tissue grinding or ultrasonic disruption, and
incubation in trypticase soy broth medium for up to 14 days to
identify the slow-growing bacteria. 43 Gram-negative enteric
bacteria and anaerobic species (e.g. Bacteroides) are most
commonly present with graft-enteric erosions and fistulas.

Specific virulence factors enable bacteria to colonize tissue,
multiply, establish infection, invade adjacent tissues, and re-
sist host defenses. Staphylococcus aureus secretes coagulase
which coats the organism, facilitates bacterial aggregation,
and inhibits phagocytosis. Catalase production neutralizes
toxic hydrogen peroxide generated in the respiratory burst
of phagocytic cells. Release of hyaluronidase degrades tissue
extracellular matrix and potentiates spread of infection. The
cell wall components of Staphylococcal species also possess
virulent properties. The peptidoglycan layer can inhibit
leukocyte migration, behave like endotoxin, and is fairly
rigid against osmotic forces. Protein A on the bacterial wall
binds and inactivates IgG, thereby inhibiting opsonization
and phagocytosis. Several Gram-negative bacteria including
Pseudomonas species can be especially aggressive.
Pseudomonas aeruginosa produces elastase and alkaline pro-
tease that degrade elastin, collagen, fibrin, and fibrinogen and
facilitate invasion of perigraft tissue and degradation of vessel
walls.

Bacterial-graft interactions

After introduction to the perigraft region or along the luminal
surface, bacterial adherence to the prosthetic material is
the fundamental step necessary for eventual development
of graft infection. Bacterial adherence is influenced by the
bacterial species, physical (roughness, surface area) and
chemical (hydrophobicity, charge) surface properties of the
biomaterial, and duration of exposure. Cell wall glycoproteins
of both Gram-positive and -negative organisms assist in
adherence. Differential adherence of bacteria to vascular
prostheses have been observed in vitro. u Bacterial adherence
is greatest to velour knitted Dacron, less to woven Dacron,
and least to PTFE. In addition to differing chemical surface
properties, the more porous Dacron material has a larger
potential surface area for bacterial adhesion than PTFE
grafts. Staphylococcus aureus adheres more readily to PTFE
than does P. coli.

481

part iv Medical management

Graft infections occurring within 4 months of implantation
(i.e. early infections) are typically associated with the more
virulent bacterial strains including S. aureus, MRSA, E. coli,
Klebsiella, Enterobacter, Proteus, and Pseudomonas species.
Wound healing complications are most commonly responsi-
ble for early prosthetic bacterial inoculation. Established early
perigraft infections of prosthetic and autologous reconstruc-
tions involve high [10 5-7 colony-forming units (CFU)] concen-
trations of bacteria and invasive properties of these bacterial
strains contribute to an increased incidence of anastomotic
dehiscence, arterial wall or vein graft rupture, and resulting
catastrophic hemorrhage.

Late graft infections (i.e. >4 months postimplantation)
are commonly associated with the less virulent coagulase-
negative staphylococci including S. epidermidis. Despite likely
contamination of the graft during initial reconstruction, these
indolent perigraft infections possess relatively low numbers
(10 2-3 CFU) of bacteria and manifest an average of 40 months
after implantation. 45 ' 46 Staphylococcus epidermidis is nonmotile
and nonspore forming but produces a mucinous, extracellular
glycocalyx that increases its adherence to PTFE and Dacron
grafts relative to other bacterial strains. 47 The developing
"biofilm" on the outer graft surface provides not only bact-
erial nutrients promoting colonization and segmental spread
of infection, but also a protective layer from host defenses
and antibiotic penetration. The slow, progressive inflammat-
ory response in the perigraft tissue and adjacent artery leads
to eventual clinical recognition of local graft infection with an
absence of systemic signs of sepsis.

Current management of vascular
graft infections

can be absent. Vascular imaging studies are essential for
diagnosis and planning treatment. Perigraft fluid, gas, or in-
flammation, anastomotic pseudoaneurysm, and graft-enteric
communications can be accurately (sensitivity 90%) identified
by combinations of ultrasonography, contrast-enhanced
computed tomography (CT), magnetic resonance imaging
(MRI), 48 contrast arteriography, and endoscopy. Radionuclide
imaging using indium-Ill or technetium-99m hexametazime-
labeled leukocytes or gallium-67 citrate can confirm the pre-
sence of graft infection when equivocal CT or MRI studies
exist. 49 Localization and determination of the extent of graft
involvement with infection by available imaging studies is
critical to appropriate selection of therapeutic options. Abnor-
mal findings in adjacent structures may suggest more diffuse
graft involvement or invasive infections such as with hydro-
nephrosis and advanced retroperitoneal inflammation ac-
companying aortic graft infection. Diagnosis of the extent of
early graft infection can be particularly difficult since perigraft
fluid and air are common findings on CT or ultrasound studies
up to 2 or 3 months postoperatively, 50 and false-positive re-
sults from early radionuclide imaging can occur due to leuko-
cyte accumulation during normal inflammatory perigraft
healing. Contrast arteriography is generally recommended
for planning revascularization of affected organs or limbs if
excision of the infected graft is likely to result in significant is-
chemia. We maintain an aggressive policy for direct operative
exploration to make the diagnosis of graft infection in equivo-
cal cases and to routinely obtain initial perigraft cultures and
Gram stain in patients with more apparent infection to aid
proper selection of definitive treatment. Operative explo-
ration is mandatory in patients with prior aortic reconstruc-
tions and gastrointestinal bleeding in whom all other sources
of bleeding have been excluded.

Diagnostics

Prompt diagnosis and treatment of prosthetic graft infections
are essential to avoid morbidity and death. Early graft infec-
tions become apparent as they typically involve sepsis of the
implantation wound. Clinical presentations of later infections
of grafts confined to the abdomen or thorax may be subtle and
delay recognition. Although systemic sepsis is a relatively un-
common (10-20%) finding in most series of graft infections, its
presence in patients with abdominal or back pain, ileus or
anorexia, and an intracavitary bypass should raise suspicion.
The classic triad associated with aortoduodenal fistula (i.e.
gastrointestinal bleeding, sepsis, abdominal pain) occurs in
less than one-third of patients with GEE/GEF. Late prosthetic
graft infections with anastomoses to a femoral artery or subcu-
taneous locations are more likely to have specific symptoma-
tology including inflammatory perigraft mass, draining sinus
tract, anastomotic pseudoaneurysm, or limb ischemia due to
graft thrombosis. Fever, leukocytosis, or an elevated erythro-
cyte sedimentation rate are common with graft infection but

Treatment options

Specific therapies for prosthetic graft infection can be derived
from selection criteria based on clinical presentation, extent of
graft involvement, and microbiology (Table 41.5). Manage-
ment by either local therapy with graft preservation or graft
excision without revascularization is typically possible in only
a fraction of cases. Many experienced vascular centers utilize a
single preferred approach for definitive treatment of prosthe-
tic arterial bypass infection, especially aortoiliofemoral graft
infections. Our group has instead advocated a patient-specific
treatment algorithm involving use of conventional (total graft
excision and extraanatomic/ remote bypass) or in-situ replace-
ment modalities. 46 Important adjuncts to these treatment op-
tions include use of multiple, staged debridement/ "wash-out"
procedures to minimize residual bacterial counts with more
virulent infections, aggressive debridement of involved arte-
rial wall and perigraft tissues, intraoperative mechanical and
passive wound irrigations with dilute chlorapactin (bleach),
peroxide and /or betadine solutions, soft tissue coverage of

482

CHAPTER41 Infections and antibiotics in vascular surgery

Table 41 .5 Selection criteria for appropriate operative management of prosthetic vascular graft infection

Treatment option

Presentation

Extent of infection

Microbiology

Graft preservation/local therapy

Graft excision only

Excision and ex-situ bypass
Simultaneous

Staged

In-situ replacement
Prosthetic
Autogenous vein

Early infection, no sepsis

Graft thrombosis, viable
limb/adequate collaterals

Unstable patient, GEE/GEF

sepsis
Stable patient, GEE/GEF

sepsis

No sepsis, no GEE/GEF
No sepsis, no GEE/GEF, severe
occlusive disease

Not Dacron graft, graft body only, no

anastomosis
Diffuse or local

Invasive infection

Invasive infection, diffuse

Biofilm infection, local/segmental
Invasive or biofilm, diffuse or local

All except Pseudomonas

Any organism

Any organism
Any organism

S. epidermidis and negative Gram stain
All except Pseudomonas

*lnvasive infection due to more virulent bacteria or low-grade biofilm infection, diffuse graft involvement by infection, or localized/segmental process.
GEE/GEF, graft-enteric erosion or fistula.

arterial repairs with noninfected, well vascularized rotational
muscle or fasciocutaneous flaps, or omental pedicles, place-
ment of closed suction drainage in grossly infected tissue beds,
and perioperative, culture-specific parenteral antibiotics.

Graft preservation

Attempts at graft preservation using only local treatment
measures is possible under limited circumstances. Patent
grafts not constructed of Dacron with short length infection in-
volvement sparing anastomoses can be considered for preser-
vation. Infections should be limited to the immediate perigraft
region, be caused by bacteria with limited virulence (not
Pseudomonas), and not be associated with systemic sepsis.
Serial, aggressive wound debridements in the operating room
are necessary to minimize residual bacterial counts and subse-
quent rotational muscle flap coverage of the exposed graft seg-
ment should be employed. Approximately 70% of patients
with either early aortofemoral graft limb infection 51 or infrain-
guinal prosthetic bypass infection 32 have achieved complete
graft preservation and wound healing. However, initial treat-
ment failures due to fatal graft disruption, persistent graft
infection, and nonhealing wounds limit application of this
approach and emphasize the need for aggressive manage-
ment by graft excision if local sepsis persists.

Graft excision without revascularization

In patients where the initial indication for intervention was
claudication or when thrombosis of an infected graft does not
result in critical limb ischemia, graft excision alone may be
considered. Arterial collaterals around a thrombosed or
ligated bypass may take several weeks to develop and opti-

mization of inflow may be accomplished by endovascular in-
tervention if occlusive lesions exist (e.g. residual aortoiliac
stenosis with patent hypogastric artery and occluded, infected
aortofemoral graft limb). However, intraoperative decisions
regarding the need for immediate revascularization with
patent, infected grafts can be made with temporary bypass oc-
clusion, a sterile blood pressure cuff, and continuous wave
Doppler assessment of pedal outflow. Persistence of a pul-
satile pedal arterial signal and ankle pressure greater than
40 mmHg with bypass occlusion may allow initial graft exci-
sion and consideration for delayed revascularization after ir-
radication of local infection. Absence of pulsatile pedal flow
with bypass occlusion will probably result in critical ischemic
symptoms and potential limb loss and mandates concomitant
limb revascularization.

Graft excision with ex-situ revascularization

Conventional management of infected prosthetic arterial by-
passes involves total graft excision and revascularization via
extraanatomic or remote routes through uninfected tissue
planes. This approach is generally required in patients with
GEE/GEF and for more invasive infections associated with
septic presentations and extensive perigraft inflammation
(e.g. hydronephrosis with aortic graft infection). Timing of
limb revascularization in patients with aortic graft infection is
dependent on patient presentation as well. Simultaneous
ex-situ bypass and graft excision is necessary in hemodynami-
cally unstable patients with systemic sepsis or hemorrhage
from a graft-enteric communication. However, staged man-
agement with initial axillofemoral PTFE bypass followed in
1-2 days by aortic graft excision has been associated with a
lower perioperative mortality rate than a simultaneous ap-

483

part iv Medical management

Table 41 .6 Results of treatment for prosthetic graft infections involving the infrarenal aorta or aortoiliofemoral bypasses

Total graft excision and

ex-situ bypass

No. of

Operative

Early limb

Stump

Survival

Infection of

Author

cases

mortality (%)

loss(%)

blowout (%)

>1year(%)

EAB (%)

O'Haraefa/. 1986 35

Reillyefa/. 1987 37

Yeagerefa/. 1990 38

Seegerefa/. 2000 39

Bandykefa/. 2001 46

In-situ replacement

No. of

Operative

Early limb

1 -year graft

Survival

Recurrent

Author

cases

mortality (%)

loss(%)

patency (%)

>1 year (%)

infection (%)

NAIS/SFPV:

Nevelsteenefa/. 1995 55

Clagettefa/. 1997 56

Prosthetic:

Bandykefa/. 2001 46

100

Allograft:

Kieffer eta/. 1993 57

EAB, extraanatomic bypass (cross-femoral, axillofemoral); NAIS/SFPV, neo-aortoiliac system constructed of superficial femoral-popliteal vein segments.

proach in stable patients. 37 Several decades of experience with
conventional management of aortic graft infections have led to
minimal changes in operative mortality but significant reduc-
tions in early limb loss, residual infections of the infrarenal
aortic stump resulting in catastrophic hemorrhage, and
recurrent infections involving ex-situ prosthetic bypasses
(Table 41.6). The highest mortality rates continue to occur in
patients with GEE/GEF or systemic sepsis, and conventional
management is preferred over in-situ replacement techniques
for these presentations. Diffuse aortobifemoral graft infec-
tions involving groin regions and with significant associated
femoropopliteal occlusive disease can make ex-situ appr-
oaches to revascularization challenging. Limb loss is more
frequent after aortofemoral graft infection than aortoiliac
reconstructions due to the higher risk of ex-situ bypass throm-
bosis or recurrent infection. Unilateral axillofemoral bypasses
to the profunda femoris or superficial femoral artery through
an uninfected tissue plane have acceptable patency rates (94%
at 6 months), but distal anastomoses to the popliteal artery are
prone to early failure (42% at 6 months). 52 Preservation of ret-
rograde flow into the common femoral artery by vein patching
after excision of the infected aortofemoral graft is also im-
portant for maintaining pelvic, colonic, and potentially
lumbosacral neural viability. Compromised patency of ex-situ
bypasses has led to increased use of in-situ replacement with
lower limb deep vein conduit to better facilitate a durable "in-
line" reconstruction when significant femoral occlusive le-
sions exist. A further option for aortofemoral graft infection
with bilateral groin involvement is a combined unilateral ax-

illofemoral PTFE bypass, autogenous deep vein cross-femoral
bypass, and total graft excision.

In-situ replacement

Total graft excision and in-situ replacement with autogenous
venous conduit is appropriate for arterial prosthetic infections
in the absence of systemic sepsis, graft-enteric communica-
tions, and Pseudomonas involvement. Greater saphenous or
superficial upper extremity veins can be used for reconstruc-
tion of infrainguinal, visceral, cerebrovascular, and upper ex-
tremity arteries. Limited patency of saphenous vein grafts
used for cross-femoral or iliofemoral bypass, and supra-
aortic trunk reconstruction is primarily due to mismatch in
conduit and vessel diameters, progressive intimal prolifera-
tion, and stenosis. 53 Use of the superficial femoral-popliteal
vein segment (SFPV) has allowed larger diameter, autoge-
nous, arterial reconstruction without causing significant mor-
bidity in the donor limb (e.g. acute deep vein thrombosis, limb
edema, compartment syndrome). Adequacy of the lower ex-
tremity deep veins (absence of acute /chronic thrombus, di-
ameter >5-6mm) can be confirmed prior to harvest by duplex
imaging. The treatment-related morbidity and mortality of
conventional extraanatomic bypass and total graft excision for
aortic graft infection have also prompted use of in-situ replace-
ment techniques with lower limb deep vein, antibiotic-bond-
ed prosthetic conduit, and cryopreserved arterial allografts
(Table 41.6). Construction of an in-situ, neo-aortoiliac system
(NATS) from deep leg vein after removal of an infected aor-

484

CHAPTER41 Infections and antibiotics in vascular surgery

toiliofemoral graft was first described by Clagett et al. 54 and
Nevelsteen et al. 55 Compared with conventional management,
in-situ deep vein replacement for diffuse graft infection is
potentially associated with lower mortality and amputation
rates, improved graft patency rates, and a lower incidence of
recurrent infection. 56 Invasive infection confined to a single
limb of an aortofemoral graft (i.e. localized infection) is also
best treated by in-situ deep vein replacement. However, deep
vein replacement of aortic graft infections complicated by
GEE/GEF has been performed sporadically but is generally
not recommended.

Use of prosthetic conduit for in-situ replacement of infected
grafts has been proposed as a treatment option in selected
circumstances. The principles of graft excision and ex-
traanatomic bypass are not applicable to most cases of as-
cending, transverse arch, or descending thoracic aortic graft
infections. Since in-situ replacement with large-diameter
Dacron conduits is generally the only option, adjuncts includ-
ing wide debridement of infected tissues, soft tissue coverage
using rotated muscle flaps (pectoralis major, rectus abdo-
minis, latissimus dorsi) or pedicled omentum, and indefinite
antibiotics are necessary. 29-31 For aortoiliofemoral and
extracavitary (infrainguinal, axillofemoral, cross-femoral)
graft infections, in-situ prosthetic replacement with PTFE
conduit has been associated with recurrent infection in
10-20% of cases primarily due to Gram-negative bacteria and
MRSA involvement. These results have increased interest in
antibiotic-bonded prosthetic grafts. We have advocated graft
excision and in-situ prosthetic replacement for localized /seg-
mental, biofilm infections caused by coagulase-negative
staphylococci (S. epidermidis) (Table 41.5). Limited morbidity
has been shown with conduit replacement using untreated
PTFE and rifampin-soaked (60 mg/ml), gelatin-sealed Dacron
for low-grade aortoiliofemoral graft infections 46 (Table 41.6).
Recurrent infections were due to rif ampin-resistant S. epider-
midis, MRSA, or more extensive graft involvement (i.e. bilater-
al graft limbs or aortic graft body) and emphasize that optimal
results appear confined to limited biofilm infections. Despite
the risk of recurrent/ residual infection, in-situ prosthetic re-
placement of a distal aortofemoral graft limb through a groin
incision may be preferred over total graft excision in frail, de-
bilitated elderly patients with low-grade biofilm infection,
groin region symptomatology but more diffuse graft involve-
ment. However, in-situ replacement with autogenous deep
vein is preferred in patients with reasonable medical co-
morbidities and biofilm infections with diffuse aortoilio-
femoral graft involvement.

A third option for in-situ replacement is aortic and ilio-
femoral arterial segments harvested from transplant donors
and rendered nonantigenic by cryopreservation. Preliminary
experience with allografts has been in Europe as these con-
duits are not readily available in the United States at present.
Allograft replacement may have a role in primary mycotic or
prosthetic graft infections of the thoracic or visceral aorta

where in-situ reconstruction is mandatory. Kieffer et al. 57
reported a series that included 36 patients with allograft re-
placement of infrarenal aortic graft infections demonstrating
acceptable mortality and early limb loss rates after in-situ allo-
graft replacement (Table 41.6). However, significant numbers
of early (5%) and late (21%) allograft complications occurred
including thrombosis, graft dilation, and reinfection.

Antibiotic use in vascular surgery

Antibiotics serve as an important adjunct in management of
preexisting vascular conditions (soft tissue infections compli-
cating ischemic tissue loss, diabetic foot lesions, venous stasis
wounds, and lymphedema), surgical infection prophylaxis,
care of postoperative wound complications, and established
vascular prosthetic infections. Antibiotics inhibit bacterial
function at specific cellular sites necessary for microbial
growth and survival. The efficacy of an antimicrobial agent de-
pends on numerous factors including the location and extent
of infection, adequacy of host defenses and local vascularity,
presence of biomaterials, bacterial species and antibiotic sus-
ceptibility, local tissue concentration of antibiotic, and the exis-
tence of bacterial resistance mechanisms. Tissue concentration
of an antibiotic is the most important determinant of effective-
ness of infection control or prevention. Pharmacokinetic vari-
ables characterize absorption, metabolism, and excretion of
administered antibiotics and influence serum concentration.
However, tissue concentration may not reflect serum concen-
tration as local levels are dependent not only on tissue vascu-
larity but also on lipid content of the tissue and on the
lipophilic capacity and degree of protein binding of the drug.
Rapid delivery to infection sites occurs in well-vascularized
tissue with antibiotics having minimal plasma protein bind-
ing. Local concentrations will be sustained in relatively fatty
tissues if the antibiotic is lipid soluble. The choice of an appro-
priate antibiotic and its route of administration must be
weighed against its cost and potential side-effects. Adverse re-
actions including hypersensitivity and nephrotoxicity are spe-
cial concerns in elderly patients with multiple comorbidities
including diabetes and chronic renal insufficiency.

Available antibiotics

Antibiotics inhibiting cell wall synthesis

The presence of a bacterial cell wall outside its cytoplasmic
membrane helps shield it from osmotic forces and mechanical
trauma. The cell wall is thickest in Gram-positive organisms
and is predominantly composed of peptidoglycan. Gram-
negative bacteria have a thinner cell wall with a complex,
outer lipopolysaccharide layer and a thin, inner peptidogly-
can layer (Fig. 41.1). (3-Lactam antibiotics (penicillins,
cephalosporins, monobactams, and carbapenems) and van-

485

part iv Medical management

Gram +

Gram -

Teichoic
acid

Peptidoglycan
layer

. Porin
channel

oY*»lT** **Y*

Outer
wall

p-lactamases

Periplasmatic
space

Peptidoglycan
layer

Cell
membrane

Figure 41 .1 Schematic of cell wall stucture of
Gram-positive and -negative bacteria.
Antibiotics must penetrate the outer cell wall
and peptidoglycan layers to effect receptors in
the cytoplasmic membrane.

corny cin inhibit cell wall synthesis. (3-Lactam antibiotics bind
reversibly to penicillin-binding proteins (PBPs) in the bacteri-
al cytoplasmic membrane that catalyze peptidoglycan cross-
linking. PBP-bound (3-lactam antibiotics act as false analogues
for peptidoglycan strands and result in defective cross-linking
and an unstable cell wall that is lysed by endogenous bacterial
autolysins. PBPs of different bacterial species have varying
affinity for (3-lactam agents and contribute to differential bac-
terial susceptibility to a given (3-lactam antibiotic. Initial diffu-
sion of (3-lactam agents through the outer cell wall is necessary
prior to PBP binding at the inner cytoplasmic membrane. The
peptidoglycan layer of Gram-positive organisms is more easi-
ly penetrated by (3-lactam agents than the complex cell wall of
Gram-negative bacteria and partially accounts for the greater
susceptibility of Gram-positives to (3-lactam antibiotics.

The basic molecular structure of penicillins and cephalo-
sporins is composed of a (3-lactam ring fused with either a
thiazolidine ring or a dihydrothiazine ring, respectively. 58
Different side-chains attached to the basic ring structures con-
fer variable pharmacologic properties including bacterial
specificity, protection from bacterial (3-lactamases, acid stabili-
ty, gastrointestinal absorption, and serum protein binding.
Bioavailability after oral dosing ranges from 30% to 95% for (3-
lactam antibiotics but intravenous administration is generally
preferred in most vascular applications. Serum concentration
peaks immediately after parenteral dosing and serum half-
lives range from 0.5 to 2.0 h for most penicillins and
cephalosporins. 59 (3-Lactam agents provide adequate concen-
trations in soft tissues and have a broad spectrum of activity
against non-(3-lactamase producing, Gram-positive and
-negative bacteria commonly encountered in vascular infec-
tions (e.g. S. aureus, S. epidermidis, E. coli). Antistaphylococcal
penicillins (i.e. methicillin, oxacillin, nafcillin) are relatively
resistant to bacterial (3-lactamases. First- and second-
generation cephalosporins (e.g. cefazolin, cefuroxime, cefox-
itin, cefotetan) have adequate activity against Gram-positive
organisms including staphylococci. Third-generation

cephalosporins (e.g. cefotaxime, ceftizoxime, ceftriaxone, cef-
tazidime) are more active against Gram-negative organisms.
Cefepime, a fourth-generation agent, further extends
Gram-negative activity while providing better coverage of
S. aureus than third-generation cephalosporins. Mezlocillin,
piperacillin, ticarcillin, and ceftazidime are effective (3-lactams
against Pseudomonal species. Hypersensitivity reactions are
commonly manifested by skin rash and rarely by anaphylaxis
and occur in up to 5% of patients taking penicillins. Allergic
reactions to cephalosporins are less common (2%) but the
incidence of cross-sensitivity between penicillin and
cephalosporin agents is reported as 5%. Interstitial nephritis
and associated renal dysfunction has an allergic basis, is not
dose related, and is most common with the antistaphylococcal
penicillins. Organ-specific, dose-related toxicity is rare with (3-
lactam antibiotics.

Additional (3-lactam agents include carbapenems (e.g.
imipenem, meropenem) and monobactams (e.g. aztreonam)
and are used less commonly in vascular applications.
Aztreonam can be used in patients with significant penicillin
and cephalosporin allergies but has activity only against
Gram-negative organisms. Carbapenems possess broad cov-
erage of Gram-positive and -negative bacteria and anaerobes
but are considered as secondary agents for resistance develop-
ing to first-line (3-lactam antibiotics.

Vancomycin is a glycopeptide antibiotic that inhibits bac-
terial synthesis of cell wall components by irreversibly bind-
ing to peptidoglycan precursor molecules. Vancomycin
possesses narrow-spectrum bactericidal activity against pre-
dominantly Gram-positive organisms. It is the antibiotic of
choice for patients allergic to penicillins and cephalosporins,
infections caused by MRSA, and prosthetic graft infections in-
volving S. epidermidis. Nephro- and ototoxicity are rare dose-
related side-effects. Vancomycin is principally excreted by the
kidneys and leads to a prolonged serum half-life in patients
with underlying renal insufficiency. Monitoring of serum lev-
els is therefore recommended in this patient group. Owing to

486

CHAPTER41 Infections and antibiotics in vascular surgery

its large molecular size, vancomycin is not filtered during he-
modialysis and allows infrequent dosing (e.g. once weekly)
with end-stage renal disease.

Antibiotics inhibiting protein synthesis

Aminoglycoside antibiotics (gentamicin, tobramycin,
amikacin, netilmicin) disrupt protein synthesis by irreversible
binding to the 30S subunit of bacterial ribosomes and inhibit-
ing translation of mRNA. The spectrum of activity for amino-
glycosides includes staphylococci and most Gram-negative
species. Aminoglycosides are commonly added to two-drug
regimens including antipseudomonal penicillins or fluoro-
quinolones for synergism against serious Pseudomonas infec-
tions. Drug excretion is principally via the kidneys and serum
half-life after intravenous dosing is 2.5 h with normal renal
function. Once-daily dosing has been shown to be as safe as
conventional multiple daily dose regimens. Renal insuffi-
ciency as well as hypovolemia, hypotension, and prolonged
administration increase serum levels and risk of dose-related
toxicity. Nephrotoxicity due to tubular necrosis and ototoxici-
ty (auditory and vestibular) are not uncommon and mandate
periodic measurement of serum levels. Adjustment of dosing
intervals is also required in patients with renal impairment.

Clindamycin and macrolide antibiotics (e.g. erythromycin,
azithromycin, clarithromycin) both bind to the 50S ribosomal
subunit and inhibit bacterial protein synthesis. Clindamycin
has activity against S. aureus, some streptococci, and anaer-
obes and can be used orally or parenterally in patients with (3-
lactam or vancomycin allergies. Side-effects include diarrhea
due to Clostridium difficile overgrowth and rare pseudomem-
branous colitis. Erythromycin is used orally for mild skin in-
fections due to streptococci or staphylococci but the latter
organisms can become rapidly resistant to the bacteriostatic
macrolides.

Antibiotics inhibiting nucleic acid synthesis

Rifampin, common to anti tubercular regimens, also has broad
activity against staphylococci, streptococci, and some noncon-
form Gram-negative species. A newer role has been proposed
for rifampin in the prophylaxis and treatment of vascular graft
infections caused by S. aureus and S. epidermidis. Rifampin
prevents bacterial synthesis of RNA by inhibition of DNA-
dependent RNA polymerase. It is easily absorbed from the
gastrointestinal tract, is highly fat soluble, and has excellent
tissue penetration. Side-effects are uncommon but potential
hepatotoxicity requires monitoring of liver enzymes. Drug in-
teractions can occur since rifampin accelerates hepatic metab-
olism of other drugs including warfarin, corticosteroids,
(3-blockers, angiotensin-converting enzyme inhibitors, and
oral contraceptives and reduces their systemic effects.

Fluoroquinolones (e.g. ciprofloxacin, ofloxacin, levo-
floxacin) have a broad spectrum of activity against

Gram-positive and -negative bacteria. Specific uses include
treatment of graft infections caused by S. aureus, MRSA, or S.
epidermidis and vascular infections involving multiresistant
Gram-negative organisms. Quinolones inhibit bacterial DNA
gyrase which is responsible for the unwinding of supercoiled
DNA during transcription. Bioavailability is nearly equivalent
between oral and intravenous dosing. Long serum half-lives
(4-7 h) allow less frequent administration (once or twice
daily). Quinolones reach high tissue concentrations owing to
their low protein binding, small molecular size, and rapid dif-
fusion. Primary excretion is by the kidneys and serum levels
and half-life are increased in patients with renal failure
mandating dosage adjustment. Adverse effects are mild and
infrequent although drug interaction with concomitant
warfarin can lead to elevated prothrombin levels and risk of
hemorrhage.

Solutions to antibiotic resistance

Bacterial resistance developing to antimicrobial agents is an
evolving concern. Best characterized for (3-lactam agents,
bacterial resistance to nearly all available antibiotics has
been described. Three known resistance mechanisms exist
for (3-lactam antibiotics. 60 Bacteria can produce genetically
altered PBPs with lower affinity for (3-lactams. Certain bacteria
can alter the permeability of their cell wall to inhibit (3-lactam
diffusion to its site of action at the inner cytoplasmic mem-
brane. The primary mechanism of resistance occurs by bacter-
ial production of (3-lactamases which hydrolyze the (3-lactam
ring and inactivate the antibiotic before it can bind to PBPs. (3-
Lactamases are encoded by genes on bacterial chromosomes
and plasmids. Chromosomal (3-lactamases are present in
most Gram-negative bacteria, preferentially hydrolyze
cephalosporins, and are unfortunately resistant to existing (3-
lactamase inhibitors (clavulanic acid, sulbactam, tazobactam).
Plasmid-mediated (3-lactamases can be transferred between
bacterial strains by phages and are found in both Gram-
positive and -negative organisms. Clavulanate, sulbactam,
and tazobactam readily inactivate nearly all plasmid-encoded
(3-lactamases by irreversible binding and prevention of (3-
lactam hydrolysis. Addition of (3-lactamase inhibitors to
several existing penicillins (oral amoxicillin /clavulanate,
parenteral ampicillin/ sulbactam, ticarcillin/ clavulanate, and
piperacillin /tazobactam) has broadened their spectrum of
action against Gram-positive and -negative and anaerobic
organisms. However, several bacterial isolates have emerged
with resistance to combination (3-lactam/ (3-lactamase
inhibitors. 60

Other families of antibiotics have been plagued by bacterial
resistance mechanisms as well. Resistance is uncommon with
vancomycin but Staphylococcal strains have emerged that de-
velop a plasmid-mediated change in a cell surface receptor
thereby inhibiting antibiotic binding to peptidoglycan.
Plasmid-mediated production of enzymes can inactivate

487

part iv Medical management

Table 41 .7 Suggested empiric antibiotic regimens for perioperative prophylaxis and therapeutic uses pertinent to vascular surgical practice

Clinical use

Anti biotic/dose/d u ration

Surgical prophylaxis
MRSA present/prevalent
(3-Lactam/vanco allergy

Therapeutic uses

Ischemic/diabeticfoot infection
Penicillin allergy

Necrotizing infection

Postoperative wound infections/early

or late prosthetic graft infections
Biofilm prosthetic graft infection

Cefazolin 1-2 g i.v 30 min preopand q8 h for 24-48 horcefuroxime 1 .5 g i.v preopandq 12 h for 24-48 h
Vancomycin 1 g i.v preop and q 1 2 h for 24-48 h
Clindamycin 900 mg i.v preopand q8 h for 24-48 h

Ampicillin/sulbactam 3 g i.v. q 6 h or ticarcillin/clavulanate 3.1 g i.v. q6 horpiperacillin/tazobactam 3.375 g i.v. q6 h
Levofloxacin 500 mg i.v. q day or ciprofloxacin 400 mg i.v. q 12 h or third/fourth-generation cephalosporin and

clindamycin 900 mg i.v. q 8 h
Imipenem 0.5 g i.v. q 6 h or meropenem 1 g i.v. q 8 h and vancomycin 1 g i.v. q 1 2 h
Cefepime2 g i.v. q 8—1 2 h or levofloxacin 500 mg i.v. q day or piperacillin/tazobactam 3.375 g i.v. q6 hand

vancomycin 1 g i.v. q 1 2 h
Vancomycin 1 g i.v. q 1 2 h and rifampin 600 mg p.o. q day (? emergence of resistant S. epi?)

Prophylaxis recommendations adapted in part from the 31 st edition of the Sanford Guide to Antimicrobial Therapy 2001

aminoglycosides by preventing ribosomal binding. Sponta-
neous mutations of DNA gyrase that are not due to bacterial
plasmids can result in fluoroquinolone resistance.

Several newer antibiotics have been developed to combat
emerging bacterial resistance. MRSA and S. epidermidis, and
vancomycin-resistant staphylococcal and enterococcal (VRE)
strains have been particularly problematic. Linezolid, 61 in the
oxazolidinone family, and quinupristin/dalfopristin have
specificity for Gram-positive organisms and demonstrate
in-vitro activity against resistant staphylococcus and VRE.
Oxazolidinones bind bacterial RNA, inhibit formation of the
70S initiation complex, and uniquely interfere with bacterial
ribosomal protein synthesis at an early stage that may confer a
lack of cross-resistance with other antibiotics. Linezolid has
shown equivalent efficacy to vancomycin for treating MRSA
infections and to (3-lactam agents for complex Gram-positive
skin and soft tissue infections in phase III, comparitive clinical
trials. Linezolid has a long half-life allowing twice-daily dos-
ing and has 100% bioavailability after oral dosing, but drug-
related myelosuppression limits the duration of therapy
to less than 2-4 weeks. Quinupristin/dalfopristin has no
current oral formulation, can cause local thromphlebitis with
peripheral infusions, and therefore requires central venous
administration.

Prophylactic use of antibiotics

Antibiotic administration prior to occurrence of bacterial
contamination aims to prevent subsequent development of
infection. Despite their routine use, the efficacy of prophylac-
tic antibiotics has not been definitively shown in vascular
surgery. Several, older, randomized, prospective studies of
aortic and lower extremity arterial reconstructions with pros-
thetic conduits have demonstrated lower wound infection
rates following perioperative antibiotics. 62 ' 63 However, the in-
cidence of infections involving the prosthetic graft has not

been reduced by prophylactic antibiotics in prospective trials
and this association has only been inferred from retrospective
studies. Even acknowledging that reduced bacterial burden in
the surgical wound may potentially lessen the risk of early
graft infection, the existence of multiple factors (i.e. redo
operations, altered host defenses) contributing to late pros-
thetic infection caused by low numbers of opportunistic S. epi-
dermidis may not be significantly influenced by perioperative
antibiosis. 27

A first- or second-generation cephalosporin (cephazolin,
cefuroxime) is recommended as prophylaxis prior to cere-
brovascular, aortic, and lower extremity arterial reconstruc-
tions especially involving prosthetic materials (Table 41. 7). 59
The initial dose should be given 30 min prior to skin incision
and subsequent intraoperative doses given at 4-h intervals
during prolonged procedures or with excessive changes in
blood volume and resuscitation. Prophylaxis may be contin-
ued for 24-48 h postoperatively but longer durations for clean
operations are not supported by available data. Culture-
specific antibiotics should be administered through the
perioperative period for patients undergoing vascular graft
implantation who have coexisting infections of the lower limb
(ischemic lesions or diabetic foot infections) or at other remote
sites (pneumonia, urinary tract infection, intraabdominal).
Prophylactic vancomycin is recommended when an active
MRSA infection is present, and although controversial, should
at least be considered when MRSA colonization exists or in
hospitals with high rates of MRSA infection. Under these
circumstances, cefazolin should be added for groin incisions to
cover Gram-negative bacilli. Clindamycin can be substituted
in patients with vancomycin and cephalosporin allergies.

Therapeutic uses

Several basic principles of antibiotic therapy are pertinent to
infections encountered in vascular practice. These include

488

CHAPTER41 Infections and antibiotics in vascular surgery

choice of a bactericidal antibiotic with organism susceptibility
documented by laboratory culture techniques, use of
synergistic therapy when virulent bacteria are involved (e.g.
Pseudomonas), minimizing risk of superinfection using
narrow-spectrum antibiotics directed at involved bacterial
species, utilizing appropriate dosing to ensure maximal tissue
concentrations above the minimal inhibitory concentration
(MIC) of the bacterial strain being treated, and use of a proper
duration of therapy. Monitoring of serum levels not only is
done to avoid toxicity, but is necessary to assure efficacy
of aminoglycoside antibiotics which exhibit concentration-
dependent killing. Maximal microbial killing occurs with
high aminoglycoside serum concentrations well above MIC
values and is best provided by large doses administered at
daily intervals (or less frequently with renal insufficiency).
On the other hand, (3-lactam antibiotics possess concentration-
independent bacterial killing whereby the duration that
serum levels remain above the MIC is more important for
efficacy than peak concentrations. This is accomplished by
lower doses given at more frequent intervals.

Ischemic lower limb wounds and diabetic foot infections
frequently encountered in patients with underlying arterial
occlusive disease are typically polymicrobial with deep tissue
involvement or osteomyelitis. Severe infections warrant ag-
gressive early debridement prior to revascularization. Broad-
spectrum antibiotics should be administered through both
the early control of foot sepsis and subsequent revasculariza-
tion or the staged definitive amputation if limb salvage can not
be achieved (Table 41.7). Postoperative wound infections are
treated similarly with final antibiotic choice determined from
deep wound cultures. Vancomycin should be started empiri-
cally for significant infections since wound complications fol-
lowing revascularization commonly involve colonization by
S. aureus. Prolonged antibiotic courses are generally required
for early or late-appearing infections involving prosthetic vas-
cular grafts. Prosthetic infections treated by complete graft ex-
cision and ex-situ bypass should be accompanied by at least
4 weeks of parenteral antibiotics. Management by in-situ pros-
thetic replacement or prosthetic graft preservation requires
parenteral antibiotics for 6 weeks followed by 3-6 months of
oral agents. Long-term antibiotic regimens are determined by
cultures of perigraft fluid or graft surfaces. Staphylococcus epi-
dermidis within mucinous biofilm infections involving pros-
thetic grafts is not eradicated with standard intravenous doses
of vancomycin and underscores the importance of adjunctive
perigraft disruption and excisional debridement necessary to
achieve proper healing.

Future directions

Modification of prosthetic vascular grafts with an antibiotic
agent prior to implantation is gaining attention as a further
means to provide local delivery of antimicrobials and prevent
or treat biomaterial infections. In-vitro and in-vivo animal

studies have demonstrated feasibility of the approach but
effective bonding of the antibiotic to graft surfaces is required.
Complex surface coatings including silver nitrate, fibrin glue,
and cyanoacrylate derivatives have been necessary for anti-
biotic bonding to PTFE grafts and only preliminary animal
data are available. Development of collagen and gelatin im-
pregnation techniques for knitted Dacron grafts creates an im-
pervious conduit not needing preclotting and has also
provided a medium for antibiotic binding. Rifampin bonded
with formalin cross-linking to collagen-coated Dacron grafts
during graft fabrication has been associated with improved re-
sistance to hematogenous seeding 64 and with reduced infec-
tion when used as an in-situ replacement in tissues locally
infected with S. aureus in canines. 65 Soaking gelatin-sealed
Dacron grafts in rifampin solutions results in antibiotic reten-
tion via ionic bond formation between the negatively charged
carboxyl groups of gelatin and the positively charged radicals
of rifampin. Grafts treated by this approach possess bacterici-
dal activity above the MIC of S. epidermidis for at least 48 h in
vivo. 66 Duration of bactericidal activity is primarily dependent
on the concentration of rifampin used and 60mg/ml appears
optimal in minimizing development of graft infection after
local staphylococcal challenge. This can be accomplished clin-
ically by soaking gelatin-sealed grafts for 15-20 min in a solu-
tion of rifampin (600 or 1200mg) and 10 or 20ml of saline. Even
at lower concentrations (1 mg/ml), addition of rifampin to
Dacron grafts has resulted in lower staphylococcal infection
rates compared with untreated Dacron grafts in every
reported animal study. Administration of periprocedural
parenteral antibiotics was less effective than an antibiotic-
treated Dacron graft in resisting local bacterial challenge in
animal models, 67 ' 68 but combination of parenteral antibiotics
and rifampin-treated Dacron graft placement was most
effective. 65

Clinical use of rifampin-soaked Dacron grafts for in-situ re-
placement of infected aortic grafts has been encouraging but
has involved under 50 reported cases. 46 ' 69-71 Failure of therapy
and recurrent infection have been associated with graft-
enteric communications and extensive perigraft abscesses at
presentation and perigraft cultures revealing E. coli, MRSA,
and rifampin-resistant S. epidermidis. This experience has mir-
rored laboratory findings where either poor sensitivity of E.
coli to rifampin or early development of "high MIC strains" in-
dicating presence of MRSA becoming resistant to rifampin has
been shown in animal infection models of rifampin-treated
grafts. 72 ' 73 In-vitro data also suggest that the resistant strains
of S. epidermidis develop at high rifampin concentrations
(>1000 x MIC) so that the optimal bactericidal dose of rifampin
minimizing the emergence of resistance appears to be 4-100 x
MIC. 74 These concentrations are generally provided in vivo by
soaking gelatin-sealed Dacron grafts in 1-60 mg/ml rifampin.
Currently, in-situ replacement with a rifampin-soaked Dacron
graft is an acceptable treatment option for biofilm arterial graft
infections caused by S. epidermidis but clinical application to

489

part iv Medical management

graft-enteric communications or MRSA infections is not
recommended. 46 ' 75

Three randomized prospective trials have addressed the
prophylactic role of rifampin-treated Dacron for prevention
of arterial graft infection. 76-78 All studies used lmg/ml
rifampin-soaked, gelatin-sealed Dacron grafts compared with
untreated, commercially available Dacron conduits. The inci-
dence of graft infections was low (<2%) and was not signifi-
cantly reduced by addition of rifampin bonding although the
postoperative wound infection rate was lower with rifampin
in the largest study of over 2500 participants. 78 Longer term
follow-up will be needed to capture late-appearing infections
and fully assess the potential efficacy of prophylactic graft
treatment with rifampin.

References

1. Szilagyi DE, Smith RF, Elliott JP et ah Infection in arterial recon-
struction with synthetic grafts. Ann Surg 1972; 176:321.

2. Bunt TJ. Synthetic vascular graft infections: I. Graft infections.
Surgery 1983; 93:733.

3. Dosluoglu HH, Curl GR, Doerr RJ et ah Stent-related iliac artery
and iliac vein infections: two unreported presentations and re-
view of the literature. / Endovasc Ther 2001; 8 :202.

4. Paget DS, Bukhari RH, Zayyat EJ et ah Infectibility of endovascular
stents following antibiotic prophylaxis or after arterial wall incor-
poration. Am J Surg 1999; 178:219.

5. Parsons RE, Sanchez LA, Marin ML et ah Comparison of endovas-
cular and conventional vascular prostheses in an experimental in-
fection model. / Vase Surg 1996; 24:920.

6. Semba CP, Sakai T, Slonim SM et ah Mycotic aneurysms of the
thoracic aorta: repair with use of endovascular stent-grafts. / Vase
Interv Radiol 1998; 9:33.

7. Deshpande A, Lovelock M, Mossop P et ah Endovascular repair of
an aortoenteric fistula in a high-risk patient. / Endovasc Surg 1999;
6:379.

8. Eskandari MK, Makaroun MS, Abu-Elmagd KM et ah Endovas-
cular repair of an aortoduodenal fistula. / Endovasc Ther 2000;
7:328.

9. Schlensak C, Doenst T, Spillner G et ah Palliative treatment of a
secondary aortoduodenal fistula by stent-graft placement. Thorac
Cardiovasc Surg 2000; 48:41 .

10. Curti T, Freyrie A, Mirelli M et ah Endovascular treatment of an il-
ioenteric fistula : a bridge to aortic homograft. Eur } Vase Endovasc
Surg 2000; 20:204.

11. Chuter TAM, Lukaszewicz GC, Reilly LM et ah Endovascular
repair of a presumed aortoenteric fistula: late failure due to re-
current infection. / Endovasc Ther 2000; 7:240.

12. Towne JB. Complications of the diabetic foot. In: Bernhard VM,
Towne JB, eds. Complications in Vascular Surgery. New York: Grune
&Stratton, 1985:435.

13. Buckels JAC. Management of infection in hemodialysis vascular
access surgery. In: Wilson SE, ed. Vascular Access Surgery. St Louis:
Yearbook Medical Publishers, 1988:305.

14. Golden MA, Hanson SR, Kirkman TR et ah Healing of PTFE arte-
rial grafts is influenced by graft porosity. / Vase Surg 1990; 11:838.

15. Kohler TR, Stratton JR, Kirkman TR et ah Conventional versus
high-porosity PTFE grafts: clinical evaluation. Surgery 1992;
112:901.

16. Szilagyi E, Pfeifer JR, DeRusso FJ. Long-term evaluation of plastic
arterial substitutes: an experimental study. Surgery 1964; 55:165.

17. Back MR, Klingman N, Schultz GS, Seeger JM. Arterial prosthetic
grafts stimulate local production of tumor necrosis factor and
matrix metalloproteases. ACS Surgical Forum 1997; 158:395.

18. Zimmerl W, Waldvogel FA, Vaudaux P et ah Pathogenesis of
foreign body infection: description and characteristics of an
animal model. J Infect Dis 1982; 146:487.

19. Landreneau MD, Raju S. Infections after elective bypass surgery
for lower extremity ischemia: the influence of preoperative
transcutaneous arteriography. Surgery 1981; 90:956.

20. Rubin JR, Malone JM, Goldstone J. The role of the lymphatic
system in acute arterial prosthetic graft infections. / Vase Surg
1985; 2:92.

21 . MacBeth G A, Rubin JR, Mclntyre KE et ah The relevance of arterial
wall microbiology to the treatment of prosthetic graft infections:
graft infection versus arterial infection. / Vase Surg 1984; 1 :750.

22. Ernst CB, Campbell CH Jr, Daugherty ME et ah Incidence and sig-
nificance of intraoperative bacterial cultures during abdominal
aortic aneurysmectomy Ann Surg 1977; 185:626.

23. McAuley CE, Steed DL, Webster MW. Bacterial presence in aortic
thrombus at elective aneurysm resection: is it clinically signifi-
cant? Am } Surg 1984; 147:322.

24. Roon A, Malone JM, Moore WS et ah Bacteremic infectibility: a
function of vascular graft material and design. / Surg Res 1977;
22:489.

25. Moore WS, Malone JM, Keown K. Prosthetic arterial graft ma-
terial: influence on neointimal healing and bacteremic infec-
tibility. Arch Surg 1980; 115:1379.

26. Durham JR, Malone JM, Bernhard VM. The impact of multiple
operations on the importance of arterial wall cultures. / Vase Surg
1987; 5:160.

27. Bandyk DF. Vascular graft infections: epidemiology, micro-
biology, pathogenesis, and prevention. In: Bernhard VM,
Towne JB, eds. Complications in Vascular Surgery, 2nd edn. St Louis:
Quality Medical Publishing, 1991:223.

28. Liekweg WG Jr, Greenfield LJ. Vascular prosthetic infections:
collected experience and results of treatment. Surgery 1977; 81:
335.

29. Coselli JS, Crawford ES, Williams TW et ah Treatment of post-
operative infection of ascending aorta and transverse aortic
arch, including use of viable omentum and muscle flaps.
Ann Thorac Surg 1990; 50:868.

30. Chan FY, Crawford ES, Coselli JS et ah In situ prosthetic graft re-
placement for mycotic aneurysm of the aorta. Ann Thorac Surg
1989;47:193.

31 . Hargrove WC III, Edmunds H Jr. Management of infected thoracic
aortic prosthetic grafts. Ann Thorac Surg 1984; 37:72.

32. Cherry KJ, Roland CF, Pairolero PC et ah Infected femorodistal
bypass: is graft removal mandatory? / Vase Surg 1992; 15:295.

33. Goldstone J, Moore WS. Infections in vascular prostheses: clinical
manifestations and surgical management. Am } Surg 1974; 128:
225.

34. Lorentzen JE, Nielson OM, Arendrup H et ah Vascular graft infec-
tion: an analysis of 62 graft infections in 2411 consecutively im-
planted synthetic vascular grafts. Surgery 1985; 98:81.

490

CHAPTER41 Infections and antibiotics in vascular surgery

35. O'Hara PJ, Hertzer NR, Beven EG et ah Surgical management of
infected abdominal aortic grafts: review of a 25-year experience.
] Vase Surg 1986; 3:725.

36. Quinones-Baldrich WJ, Hernandez JJ, Moore WS. Long-term
results following surgical management of aortic graft infection.
Arch Surg 1991; 126:507.

37. Reilly LM, Stoney RJ, Goldstone J et ah Improved management of
aortic graft infection: the influence of operation sequence and
staging. / Vase Surg 1987; 5:421.

38. Yeager RA, Taylor LM Jr, Moneta GL et ah Improved results with
conventional management of infrarenal aortic graft infection.
} Vase Surg 1999; 30:76.

39. Seeger JM, Pretus HA, Welborn B et ah Long-term outcome after
treatment of aortic graft infection with staged extra-anatomic by-
pass grafting and aortic graft removal. / Vase Surg 2000; 32:451.

40. Nasim A, Thompson MM, Naylor AR et ah The impact of MRS A on
vascular surgery. Eur J Vase Endovasc Surg 2001; 22:211.

41. Naylor AR, Hayes PD, Darke S and Joint Vascular Research
Group. A prospective audit of complex wound and graft infec-
tions in Great Britain and Ireland: the emergence of MRSA. Eur }
Vase Endovasc Surg 2001; 21:289.

42. Bandyk DF, Berni G A, Thiele BL et ah Aortofemoral graft infection
due to Staphylococcus epidermidis. Arch Surg 1984; 119:102.

43. Bergamini TM, Bandyk DF, Govostis D et ah Identification of
Staphylococcus epidermidis vascular graft infections: a comparison
of culture techniques. / Vase Surg 1989; 9:665.

44. Schmitt DD, Bandyk DF, Pequet AJ et ah Bacterial adherence to
vascular prostheses: a determinant of graft infectivity / Vase Surg
1986; 3:732.

45. Bergamini TM, Bandyk DF, Govostis D et ah Infection of vascular
prosthesis caused by bacterial bio films. / Vase Surg 1988; 7:21.

46. Bandyk DF, Novotney ML, Back MR et ah Expanded application of
in situ replacement for prosthetic graft infection. / Vase Surg 2001;
34:411.

47. Schmitt DD, Bandyk DF, Pequet AJ et ah Mucin production by
Staphylococcus epidermidis: a virulence factor promoting adherence
to vascular grafts. Arch Surg 1986; 121:89.

48. Olofsson PA, Auffermann W, Higgins CB et ah Diagnosis of pros-
thetic aortic graft infection by magnetic resonance imaging. / Vase
Surg 1988: 8:99.

49. Lawrence PF, Dries DJ, Alazraki N et ah Indium Ill-labeled leuko-
cyte scanning for detection of prosthetic vascular graft infection.
J Vase Surg 1985; 2:165.

50. O'Hara PJ, Borkowski GP, Hertzer NR et ah Natural history of
periprosthetic air on computerized axial tomographic examina-
tion of the abdomen following abdominal aortic aneurysm repair.
} Vase Surg 1984; 1:429.

51. Calligaro KD, Westcott CJ, Buckley RM et ah Infrainguinal anasto-
motic arterial graft infections treated by selective graft preserva-
tion. Ann Surg 1993; 216:74.

52. Seeger JM, Back MR, Albright JL et ah Influence of patient charac-
teristics and treatment options on outcome of patients with pros-
thetic aortic graft infection. Ann Vase Surg 1999; 13:413.

53. Seeger JM, Wheeler JR, Gregory RT et ah Autogenous graft re-
placement of infected prosthetic grafts in the femoral position.
Surgery 1983; 93:39.

54. Clagett GP, Bowers BL, Lopez- Viego MA et ah Creation of a neo-
aortoiliac system from lower extremity deep and superficial veins.
Ann Surg 1993; 218:239.

55. Nevelsteen A, Lacroix H, Suy R. Autogenous reconstruction with
the lower extremity deep veins: an alternative treatment of pros-
thetic infection after reconstructive surgery for aortoiliac disease.
J Vase Surg 1995; 22:129.

56. Clagett GP, Valentine RJ, Hagino RT. Autogenous aortoiliac/
femoral reconstruction from superficial femoral-popliteal veins:
feasibility and durability. / Vase Surg 1997; 25:255.

57. Kieffer E, Bahnini A, Koskas F et ah In situ allograft replacement of
infected infrarenal aortic prosthetic grafts: results in 43 patients.
] Vase Surg 1993; 17:349.

58. Condon RE, Wittmann DH. The use of antibiotics in general
surgery. Curr Problems Surg 1991; 28:803.

59. Gilbert DN, Moellering RC Jr, Sande MA, eds. The Sanford Guide to
Antimicrobial Therapy, 31st edn. Hyde Park, VT: Antimicrobial
Therapy, Inc., 2001.

60. Bonomo RA, Shales D. Resistance to beta-lactam/beta-lactamase
inhibitor combinations. Infect Med 1992; 9:48.

61. Wilson SE. Clinical trial results with linezolid, an oxazolidinone,
in the treatment of soft tissue and post-operative gram-positive
infections. Surg Infect 2001; 2:25.

62. Kaiser AB, Clayson KR, Mulherin JL Jr et ah Antibiotic prophy-
laxis in vascular surgery. Ann Surg 1978; 188:283.

63. Pitt HA, Postier RG, MacGowan WL et ah Prophylactic antibiotics
in vascular surgery: topical, systemic or both? Ann Surg 1980;
192:356.

64. Chervu A, Moore WS, Gelabert HA et ah Prevention of graft infec-
tion by use of prostheses bonded with a rifampin/ collagen release
system. / Vase Surg 1991; 14:521.

65. Colburn MD, Moore WS, Chvapil M et ah Use of an antibiotic-
bonded graft for in situ reconstruction after prosthetic
graft infections. / Vase Surg 1992; 16:651.

66. Gahtan V, Esses GE, Bandyk DF et ah Antistaphylococcal activity
of rif ampin-bonded gelatin-impregnated Dacron grafts. / Surg Res
1995; 58:105.

67. Shue WB, Worosilo SC, Donetz AP et ah Prevention of vascular
prosthetic with an antibiotic-bonded Dacron graft. / Vase Surg
1988; 8:600.

68. Lachapelle K, Graham AM, Symes JF. Antibacterial activity,
antibiotic retention, and infection resistance of a rifampin-
impregnated gelatin-sealed Dacron graft. / Vase Surg 1994; 19:
675.

69. Torsello G, Sandmann W, Gehrt A et ah In situ replacement of in-
fected vascular prostheses with rifampin soaked vascular grafts:
early results. / Vase Surg 1993; 17:768.

70. Hayes PD, Nasim A, London NJM et ah In situ replacement of in-
fected aortic grafts with rif ampin bonded prostheses: the Leicester
experience (1992-1998). / Vase Surg 1999; 30:92.

71. Young RM, Cherry KJ, Davis PM et ah The results of in situ pros-
thetic replacement for infected aortic grafts. Am } Surg 1999;
178:136.

72. Koshiko S, Sasajima T, Muraki S et ah Limitations in the use of
rifampin-gelatin grafts against virulent organisms. / Vase Surg
2002; 35:779.

73. Brissonniere OG, Leport C, Bacourt F et ah Prevention of vascular
graft infection by rifampin bonding to a gelatin-sealed Dacron
graft. Ann Vase Surg 1991; 5:408.

74. Garrison JR, Henke PK, Smith KR et ah In vitro and in vivo effects
of rifampin on Staphylococcus epidermidis graft infections. ASAIO J
1997; 43:8.

491

part iv Medical management

75. Bandyk DF, Novotney ML, Johnson BL et al. Use of rifampin-
soaked gelatin-sealed polyester grafts for in situ treatment of
primary aortic and vascular prosthetic infections. / Surg Res
2001; 95:44.

76. D'Addato M, Curti T, Freyrie A. Prophylaxis of graft infection
with rifampin-bonded Gelseal graft: 2-year follow-up of a
prospective clinical trial. Cardiovasc Surg 1996; 4:200.

77. Braithwaite BD, Davies B, Heather BP et al. (Joint Vascular Re-

search Group). Early results of a randomized trial of rifampin-
bonded Dacron grafts for extra-anatomic vascular reconstruction.
Br J Surg 1998; 85:1378.
78. Koskas F, Goeau-Brissoniere O, Pechere JC. Prevention of early
wound and graft infection with rifampin-bonded knitted poly-
ester graft: results of the rifampin bonded grafts European trial
(RBGET). Rifampin Bonded Graft European Trial, Paris, France,
May 12-13, 1995.

492

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Endovascular interventions

for vascular disease

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Catheter-based approaches to
the treatment of atheroembolic d

Frank R. Arko
Christine Newman
Thomas J. Fogarty

Arteriosclerosis is recognized as one of the major health prob-
lems of industrialized countries. It is associated with ischemic
heart disease that affects 5 million Americans and is the lead-
ing cause of death in men over 35. It is also associated with
cerebrovascular accidents, cerebral hemorrhage, ischemic
renal disease, and peripheral vascular disease. 1 Arteriosclero-
sis of the extremities is the most common cause of occlusive
arterial disease in patients over 40. 2 With prevalence of
this magnitude it is apparent that any improvement in therapy
for arteriosclerosis has the potential for a significant effect on
the general health. Many of the new techniques for treating
atherosclerotic disease in the periphery are in the form of
catheter-based systems. In order to help understand the
catheter-based therapies for atheroembolic disease it is imp-
ortant to review the structure and consequences of the
atheromatous lesion.

Pathogenesis of in-situ arterial thrombosis

The atheromatous lesion is a fibrofatty plaque consisting
of a raised focal plaque within the intima, a core of lipid, and
a covering fibrinous cap. As the disease advances, the plaque
increases in size and progressively encroaches on the lumen
of the artery as well as the media. This disease process
compromises arterial blood flow, weakens the affected
artery, and often leads to a number of complications such as
ulceration, calcification, aneurysmal dilation, or thrombus
formation. This last complication, thrombus formation, is
one of the most frequently encountered sequelae of atheroscle-
rosis. There are several mechanisms which can lead to this
result. 3

Fully developed atheromas often undergo a series of
changes such as calcification and ulceration. Ulceration of the
luminal surface and rupture of the atheromatous plaque ex-
pose the blood to tissue factors in the vessel wall which trigger
the formation of the thrombus. 4 As subendothelial elements
such as fibrillar collagen become exposed, platelet adherence
begins. Platelets can also be brought into contact with en-

dothelial elements by the alterations in laminar flow induced
by fissured atherosclerotic plaques. Turbulence from laminar
flow can also damage the endothelial surface, yielding the
same result as ulceration. 3 In all cases, platelets are attracted to
the injured site and a platelet nidus develops.

The platelet nidus is composed of platelets adhering to the
damaged vessel wall and to each other, serving to anchor the
developing thrombus to the vessel wall. Small platelet masses,
held together by fibrin strands, may embolize from the grow-
ing thrombus and cause injury to distal arteries. As the platelet
nidus develops, the platelets secrete granules with additional
thrombogenic factors such as fibrinogen, platelet-derived
growth factor, adenosine diphosphate (ADP), antiheparin,
histamine, serotonin, and thromboxane A 2 (TxA 2 ). The secre-
tion of these substances by the original adherent platelets
causes the aggregation of even more platelets at the site. 4

The aggregated platelets subsequently release platelet fac-
tor 3, which initiates the activation of coagulation factors in the
intrinsic pathway. Meanwhile, tissue factor released from the
injured endothelial cells activates the extrinsic coagulation
pathway. At the same time, alterations in smooth blood flow
prevent dilution of the thrombogenic elements, prevent he-
patic clearance of coagulation factors, and retard the inflow of
clotting factor inhibitors. Eventually, the entire atheroma may
be covered with thrombus. In heart chambers and larger arter-
ies, mural thrombi develop and usually do not fill the entire
lumen. In general, these thrombi are grayish-white, friable,
and composed of fibrin and platelets in an interlocking
matrix. 3

Acute clot formation

Development of thrombus over the existing atheroma has ad-
ditional consequences. As mentioned before, clot or plaque
segments may embolize and cause ischemic damage distal to
the origin of the embolus. The thrombus itself, however, ex-
tends into the lumen of the vessel and can contribute signifi-
cantly to the degree of stenosis. Since normal blood flow is

495

pa rt v Endovascular interventions for vascular disease

-» ">r-

Figure 42.1 Simplified progression of occlusive atheroembolic disease
process. (A) Atheroma. (B) Atheroma with in-situ thrombus. (C) Atheroma
with in-situ thrombus and acute thrombotic occlusion.

""N

Figure 42.2 Sequence of use of catheter-based instruments for
reperfusing atheroembolic obstruction. (A) Acute thrombus removed with
conventional embolectomy balloon catheter. (B) Adherent In-situ thrombus
removed with corkscrew-style adherent clot catheter. (C) Underlying
atheroma removed with atherectomy catheter.

disrupted in the stenotic vessel, areas of turbulence and stasis
often develop. Usually, there is a critical level of stenosis after
which the resultant stasis causes further coagulation reactions
in the static blood. Instead of organized thrombus, however,
these reactions have a tendency to produce a clot of a gelati-
nous consistency without fibrin strands, which is not firmly
attached to the underlying vessel wall. This type of clot
can occur in a variety of peripheral vessels and can lead to
limb-threatening ischemia.

Treatment overview

As we consider the treatment for such cases, it is useful to recall
the three layers of obstruction commonly encountered in the
affected vessel (Fig. 42.1). The first is the acutely occlusive
gelatinous thrombus. The second is the more rubbery in-situ
thrombus that grows slowly to eventually precipitate the
acute occlusion. Composed of a fibrin-rich network and
platelets, this thrombus is more difficult to dislodge from the

arterial wall than the acute thrombus. Finally, there is the pri-
mary atheroma, which is more organized than the thrombus
but can be quite variable in terms of consistency and
adherence to the vessel wall. Attempts at restoring flow
need to take these three different elements into account.
Catheter-based treatment is available for each of these three
elements (Fig. 42.2). In recent years, an emphasis on minimal
invasiveness has given rise to the increased consideration
of catheter-based surgical management. Catheter-based ap-
proaches continue to gain acceptance as advances are made
in technology, the instrumentation, and their associated
techniques.

In this chapter, we present a brief overview of the catheter-
based approaches to the treatment of atherosclerosis and
thrombosis. These approaches range from devices that sim-
plify current reconstructive techniques such as autologous
bypass grafting, to devices that offer a practical alternative or
adjunct to bypass, such as the various atherectomy, atheroab-
lation, and graft thrombectomy devices.

496

chapter 42 Catheter-based approaches to the treatment of atheroembolic disease

inflation port

y gate valve

guidewire port

^2>

inflation port

Figure 42.3 Thru-lumen embolectomy
catheter. (A) Deflated, over guide-wire. (B)
Inflated, delivering fluid.

infusion port

valve closed

■*»H£."4'-* v -

Catheter-based treatment of
the atheroembolic site

Removal of acute thrombus

Balloon embolectomy

Removal of clot from the peripheral arterial circulation was ex-
pedited by the introduction of the balloon embolectomy
catheter technique in 1963. 5 The balloon catheters in use today
are quite similar to the early device, consisting of a latex bal-
loon at the distal tip of the catheter and a proximal port for the
introduction of inflation media. In use, the surgeon advances
the tip of the catheter into and beyond the region of the clot.
When the tip is distal to the occlusion, the balloon is inflated to
meet the vessel lumen diameter. The catheter is then with-
drawn, forcing the proximally trapped material toward the
arteriotomy for removal.

Thru-lumen catheter

Since the introduction of the technique in 1965, new products
and visualization technologies have been developed
which further facilitate the use of catheters in the surgical
setting. To take advantage of recent innovations such as
operating room fluoroscopy and advances in lytic therapy
and angioscopy, a simple improvement has been made to the
traditional balloon embolectomy catheter. The Thru-Lumen
Embolectomy Catheter (Edwards Lifesciences, Irvine, CA,
USA) is constructed with an additional lumen that exits at

the distal tip of the catheter (Fig. 42.3). The extra lumen
provides a number of options for the surgeon. The catheter
can be passed over a fluoroscopically monitored guide-wire,
providing access to difficult regions. Also, contrast material,
heparinized saline, or thrombolytic agents can be delivered
through the lumen. Additionally, the inflated balloon can be
used as an occluding device to facilitate localized delivery
of fluid. 6

Removal of in-situ thrombus

Balloon embolectomy catheters, although effective at remov-
ing many occlusive clots, often do not have the gripping
power to remove the more adherent in-situ thrombus. With the
increased use of synthetic grafts and the increase in the num-
ber of patients with atherosclerotic disease, the prevalence
of such adherent in-situ thrombus has increased steadily
over the past 40 years. Also, improvements in visualization
techniques such as angioscopy have provided the surgeon
with additional information regarding the nature and extent
of vascular lesions. These developments have given us a
greater understanding of the limitations of existing proce-
dures to manage adherent thrombotic material. It has become
increasingly evident that a more aggressive approach is
appropriate in many situations.

Adherent clot catheter

The Adherent Clot Catheter (Edwards Lifesciences) is a more
aggressive counterpart to the conventional balloon catheter,
and is designed to remove residual thrombi which the

497

pa rt v Endovascular interventions for vascular disease

6Fr

Handle

=> A

ooec^i

5 mm

$fflhT™ mm

Latex-covered retrieval coil

Figure 42.4 Adherent clot catheter (6 Fr). (A)
Collapsed for introduction (adjusting knob
forward). (B) Partially expanded. (C) Fully
expanded (knob fully retracted).

standard embolectomy catheter has been unable to retrieve
(Fig. 42.4). It has a flexible 4- or 6-Fr body with a distal flexible
cable coiled around a center core wire. The loosely spiraled
outer cable is covered by an elastomeric membrane, which as-
sumes a corkscrew shape due to the structure of the coiled
outer wire. The center core wire runs the length of the catheter
to a knob on a proximal control handle. This handle is used to
expand or contract the distal active portion. As the surgeon ad-
justs the knob on the handle, the pitch of the corkscrew-shaped
retrieval element increases from a fully collapsed 2 mm to an
expanded diameter of up to 10 mm.

The adherent clot catheter is utilized after one or more pass-
es with the conventional balloon embolectomy catheter, and
the technique for use of the two catheters is similar. The sur-
geon advances the tip of the catheter beyond the thrombus
while the balloon is in the low-profile position. Once in place,
the surgeon adjusts the pitch of the spiral wire by feel until the
spiral balloon reaches the proper diameter. Material is then en-
gaged within the spiral sections of the balloon. The surgeon
then slowly withdraws the catheter along the vessel, and re-
moves the material out through the arteriotomy During with-
drawal the pitch can be continuously readjusted to vary the
tractive force, or to accommodate variations in vessel dia-
meter. The mechanism of retrieval is different from a conven-
tional balloon mechanism (Fig. 42.5) because the helical shape
provides greater contact area and a firmer shoulder region for
trapping of material.

Treatment of underlying atheroma

The instruments mentioned thus far are intended for the treat-
ment of soft (acute) and adherent (chronic, in-situ) thrombus.
As mentioned before, however, removing these clots may
not address the underlying pathology which prompted the
thrombosis— the atherosclerotic lesion. Altering the athero-
sclerotic lesion can be an important part of treating the acute
thrombotic occlusions. A number of catheter-based mechani-
cal methods have been proposed (Figs. 42.5 and 42.6).

Balloon dilation

While there have been improvements in balloon and guide-

Acute thrombus

Adherent thrombus

Figure 42.5 Mechanisms of entrapment of thrombus. (A) Compliant
shoulder of embolectomy balloon conforms to adherent material while
withdrawing acute thrombus. (B)The multiple gripping shoulders of the
adherent clot catheter spiral element provide increased traction for removal
of adherent thrombus.

Figure 42.6 Catheter-based tools for addressing the atherosclerotic lesion.
(A) Angioplasty balloon catheter. (B) Simpson directional atherectomy
catheter. (C) Transluminal extraction catheter (TEC device). (D) Auth
Rotablator. (E) Tree-Wright (formerly Kensey) catheter.

498

chapter 42 Catheter-based approaches to the treatment of atheroembolic disease

wire design over the years, the basic procedure has changed
little since cardiologist Andreas Gruntzig first popularized the
approach in 1974. 7 The surgeon advances the deflated balloon
catheter through the artery until it reaches the atherosclerotic
plaque. Inflation of the balloon disrupts the plaque and
stretches the arterial wall which leads to an increased
lumen size.

Although balloon angioplasty has made considerable im-
pact on the treatment of obstructive arterial disease, there are
still limitations of this technique such as abrupt vessel closure,
restenosis, and the creation of intimal flaps. In part, these prob-
lems are due to the fact that angioplasty disrupts atheroscle-
rotic plaque without actually removing it. Because of this
shortcoming, especially in femoral sizes and smaller, tech-
nologies such as atherectomy have been developed to me-
chanically remove the plaque.

Atherectomy

Atherectomy can be generally described as the process of me-
chanically removing obstructive atheroma via a catheter-
based system. In common usage, atheroablation devices also
fall under this category. The atherectomy device as originally
envisioned by cardiologist John Simpson is one that systema-
tically cuts and retrieves atheroma. Atheroablation, on the
other hand, involves the use of a rotating abrasive or cutting
tip to slice or pulverize the plaque from within the artery 8 Both
types of procedure and device are available for clinical use in
selected anatomical applications.

Directional atherectomy

The Directional Atherectomy Catheter (AtheroCath; Devices
for Vascular Intervention, CA, USA) is the most well-known
atherectomy instrument, primarily because of its notoriety in
coronary artery applications. 9 It consists of a flexible catheter
with a cylindrical metal housing at the distal end. The cylinder
has a cutting window on one side and a balloon on the
other which stabilizes the cutter and forces the atheroma
into the cutting window. A steel cutting blade inside the
cylinder is activated by depressing a switch on a hand-held
driver at the proximal end. A trigger/lever on the held-held
driver allows the user to advance the cutter as it spins within
the housing. Atheroma protruding into the opening of the
housing is shaved off the vessel wall and pushed into a collec-
tion chamber at the distal end of the housing for subsequent
withdrawal and pathologic examination. 10 This side-cutting
approach is called directional atherectomy to distinguish it
from other atherectomy catheter designs which are forward
cutting.

TEC system

The major forward-cutting atherectomy device is the Translu-

minal Extraction Catheter (TEC system; Interventional Tech-
nologies, San Diego, CA, USA). It contains a rotating cutter
with microtome-sharp blades at the distal end of the catheter
assembly. An external drive unit rotates the cutter to pare
atheroma while an attached vacuum source aspirates the
excised tissue. In use, the TEC guide-wire is placed across the
obstruction and the catheter is passed over the wire and
positioned at the proximal end of the occluded segment. The
external drive is then energized, vacuum is activated, and the
cutter begins rotating.The forward-cutting catheter is slowly
advanced through the atheromatous material while the va-
cuum provides continuous extraction of the shaved tissue. 11

Atheroablation devices

The two most common ablation-style atherectomy instru-
ments are the Trac-Wright System (Dow Corning Wright,
Arlington, TX, USA) and the Auth Rotablator (Heart Tech-
nologies, Bellevue, WA, USA). The Trac-Wright system is a
flexible catheter with a rotating cam tip driven by an internal
torsion-driven wire. When the catheter tip rotates, a perfusate
sprays radially from behind the catheter tip and acts as a
source of lubrication for the drive cable. A console controls the
rate of fluid flow and speed of the tip, which reportedly can
differentiate between atherosclerotic material and normal
tissue. The Trac-Wright device mechanically pulverizes
the atheromatous lesion and has been postulated to increase
lumen diameter by micropulverizing the atherosclerotic tis-
sue that contacts the catheter tip. 12

The Auth Rotablator consists of a catheter with a distal high-
speed rotary burr that is embedded with fine abrasive dia-
mond chips. At high rotational speeds, the burr effects a fine
particulate ablation of tissue. The high-speed rotation of the
roughened distal burr is designed to grind the atheroma into
embolic particles that are small enough to pass freely through
the capillary bed without obstructing blood flow. The residual
lumen created by the burr reportedly lacks the intimal flaps,
mural cracks or fissures that are often observed after conven-
tional balloon angioplasty. 13

Percutaneous thrombectomy

Trellis peripheral infusion system

The Trellis (Bacchus Vascular, Inc., Santa Clara, CA, USA) infu-
sion system is intended for controlled and selective infusion of
fluids, including thrombolytics, into the peripheral vascula-
ture. The system consists of a catheter with an infusion area
isolated between two occluding balloons. The infusion area is
either 10 cm or 20 cm in length, depending on device configu-
ration (Fig. 42.7). The oscillation drive unit with integrated dis-
persion wire is introduced into the catheter once the Trellis has
been placed in the treatment area. The oscillation drive unit is

499

pa rt v Endovascular interventions for vascular disease

Figure 42.7 Trellis peripheral infusion catheter (6 Fr). The infusion length
for the trellis device is 10-20 cm. There is a proximal and distal balloon to
allow for controlled and selective infusion of thrombolytics. The oscillation
drive unit when activated allows for dispersion of infused fluids.

Figure 42.8 The Fino Thrombectomy Catheter (7 Fr). It is an over-the-wire
catheter with an integral motor drive unit. It has a central hollow drive shaft
connected to an expanding clot maceration and removal system.

then activated, enabling dispersion of infused thrombolytics.
The combination of both localized thrombolytics and the oscil-
latory drive unit allows for the rapid dissolution of acute
thrombus, reducing the time and amount of thrombolytics
given.

Fino thrombectomy catheter

The Fino (Bacchus Vascular, Inc., Santa Clara, CA, USA) is a
single-use, over-the-wire disposable catheter, with an integral
motor drive unit. The Fino has a central drive shaft connected
to an expanding clot maceration and removal system. The clot
removal system is composed of a nitinol macerator and a niti-
nol outer protective basket. The macerator is attached to the
drive shaft and rotates within the stationary protective basket
(Fig. 42.8). The protective basket expands to the lumen diame-
ter and acts to protect the vessel wall from the rotating macera-
tor. The Fino is designed to remove the thrombus material
from the vessel using the vacuum provided by the locking
syringe in conjunction with the mechnical action provided
by the Archimedes screw. The vacuum provided by the
locking syringe is the primary source of aspiration. The me-
chanical movement of the Archimedes screw is designed to aid
in the movement of the clot through the catheter and keep the
catheter from clogging. Aspiration is controlled by a button on
the motor drive unit, which may be activated by the physician
to allow flow out of the Fino.

Percutaneous aspiration thrombectomy

Percutaneous aspiration thrombectomy (PAT), as first de-
scribed by Sniderman and associates more than a decade
ago, 14 is a technique that involves placing an antegrade vascu-
lar sheath into a thrombosed vessel. A guide-wire is intro-
duced and advanced through the thrombus. A 5- to 8-Fr guide

catheter is passed over the wire until the catheter abuts the
leading edge of the thrombus. Aspiration with a syringe or
vacuum container is performed to draw the clot into the
catheter. Firm clot that will not pass into the guide catheter is
withdrawn into the sheath and eliminated.

Three examples of PAT devices that are commercially avail-
able in Europe include the Stark Catheter (Angiomed/Bard;
Covington, GA, USA), SPAT (Bait-Extrusion; Montmorency,
France), and Rotating Aspiration Thrombectomy Device
(RAT) (Angiomed/Bard; Karlsruhe, Germany).

Pullback thrombectomy and trapping

Notable amongst the aspiration thrombectomy devices is the
transvenous embolectomy device introduced by Greenfield
and colleagues in 1969 for the removal of acute pulmonary em-
boli. 15 ' 16 The modern version of the Greenfield Transvenous
Pulmonary Embolectomy Catheter (Medi-tech/ Boston
Scientific, Watertown, MA, USA) consists of a double-lumen
steerable catheter with a vacuum-cup at the distal tip. The
12-Fr catheter is inserted via a jugular or femoral vein through
a venotomy or large sheath and positioned within the pul-
monary artery next to the thrombus using fluoroscopic
guidance and pulmonary angiography. Syringe suction is
applied to aspirate the embolus into the cup where it is held
in place as the catheter is withdrawn. In a series of 32
patients with life-threatening pulmonary embolism, embolec-
tomy was achieved with the device in 29 cases for a technical
success rate of 91% and an overall survival rate of 78%. 15 The
technique was unsuccessful when the embolus was more than
72 hold.

Other percutaneous pullback thrombectomy techniques
include self-expanding sheaths such as the Tulip Sheath
(Schneider Europe, Zurich, Switzerland) and the Ahn
Thrombectomy Catheter (American BioMed; The Woodlands,

500

chapter 42 Catheter-based approaches to the treatment of atheroembolic disease

i
1

4 mm

10 mm

Figure 42.9 Graft thombectomy catheter
(6 Fr). (A) Collapsed for introduction —
adjusting knob forward. (B) Partially expanded.
(C) Fully expanded — knob fully retracted.

/jL

Adjustable spiral retrieval wire

16 mm

TX, USA). The Tulip Sheath consists of a coaxial Fogarty bal-
loon catheter that traps thrombus inside a self-expanding
Wallstent. The thrombus is compressed and removed as the
stent is withdrawn into a 5- to 10-Fr sheath. The Ahn catheter
consists of a thrombectomy catheter with dual silicone bal-
loons. The distal balloon traps clot and prevents blood loss
after the proximal balloon is removed from the artery.

Rotational and hydraulic
recirculation thrombectomy

Recirculation thrombectomy devices create a hydrodynamic
vortex at the tip of the catheter that fragments thrombus and
pulverizes the resulting particles. The hydrodynamic vortex is
generated either by rotational or hydraulic recirculation at the
catheter tip.

Devices that create rotational recirculation all consist of a
catheter with an impeller or basket at the distal tip that rotates
at a high speed (100 000-150 000 rev/min). The resulting fluid
vortex creates a pressure gradient (Venturi effect) that obli-
terates and evacuates thrombus from the vessel lumen. The
"Clot Buster" (Microvena, White Bear Lake, MN, USA) and
"Trac-Wright" Catheter (Dow, Corning, Wright/ Theratek
International, Miami, FL, USA) are examples of rotational
recirculation devices.

Hydraulic recirculation devices make use of high-pressure
fluid jets that create a Venturi effect and pressure differen-
tial that results in pulverization and aspiration of thrombus.
These systems require fluid compression control units. The
Angiojet Rheolytic Thrombectomy System (Possis Medical,
Minneapolis, MN, USA) and the Oasis (Boston Scientific,
Boston, MA, USA) are examples of hydraulic recirculation
devices.

Catheter-based techniques for bypass grafts

Synthetic bypass grafts

Although catheter-based techniques for the treatment of ath-
erosclerosis and atheroembolism are gaining popularity, by-
pass grafting continues to be the dominant form of therapy in

the periphery. Synthetic grafts remain popular and while they
have been useful, they are not without problems. One of the
most common problems encountered in these grafts is adher-
ent thrombus. Thrombus in grafts can be more adherent than
in native vessels. Dacron polyester grafts often exhibit strong
adherence of thrombus and neointimal hyperplasia to graft
walls. At the same time, the toughness of the graft allows the
surgeon to be more aggressive in thrombus removal than
would be possible in native vessels.

Graft thrombectomy catheter

For removal of adherent thrombus in synthetic grafts, there is
the Graft Thrombectomy Catheter (Baxter V Mueller, Niles, IL,
USA), which is similar in design to the adherent clot catheter
mentioned previously, but is more aggressive. It has a 5- or 6-Fr
flexible catheter body, an expandable distal end made of spiral
wires with a flexible tip, and a proximal control handle with a
sliding knob (Fig. 42.9). Compared with the adherent clot
catheter, the spiral-wire retrieval region of the graft thrombec-
tomy catheter is shorter and stiffen These characteristics, com-
bined with the absence of a latex covering, give the graft
thrombectomy catheter more pulling strength than any bal-
loon catheter. The pitch of the retrieval wires can be varied
from 4 to 16 mm in diameter, making it well suited for grafts
within this size range.

As with the adherent clot catheter, the graft thrombectomy
catheter is normally used after multiple passes with a balloon
embolectomy catheter. The surgeon inserts the distal tip past
the thrombus in its low-profile configuration and then ex-
pands the spiral wires to meet the clot. Using the control knob,
the surgeon expands the retrieval wires when they are posi-
tioned within the obstructive material. As the catheter is
drawn out of the graft, the surgeon can continuously adjust the
pitch of the wires as needed. After withdrawal of the graft
thrombectomy catheter, the intraluminal material is removed
from the wires. One can often see the imprint of graft surface
on segments of mature thrombus that are removed. With the
additional tractive power of this catheter, the surgeon is able to
remove all occlusive material and expose the underlying graft
surface, a condition which can be documented by intraopera-
tive angioscopy

501

pa rt v Endovascular interventions for vascular disease

Closing comments

A pathophysiologic examination of the atheroembolic disease
process provides us with a clearer understanding of the need
for different types of therapeutic catheter systems. Conven-
tional balloon embolectomy catheters remove the acute soft
thrombus. More aggressive catheters such as the adherent clot
catheter and the graft thrombectomy catheter are required to
remove the more adherent in-situ thrombosis and graft pan-
nus. Atherectomy devices address the underlying atheroscle-
rotic lesion.

Catheter-based management of atherosclerosis and its
various sequelae holds great promise for the future. As cost-
containment becomes a critical factor in all facets of medicine,
the allure of alternative approaches to treatment is high if the
alternative is less invasive and /or less costly then predomi-
nant methods. Catheter-based procedures, whether for em-
bolectomy, atherectomy, or adjuncts in bypass grafting, hold
the promise of reduced invasiveness, which often leads to
greater economic and medical efficiency. Additionally low-
ered restenosis rates and higher patency can greatly reduce the
need for follow-up procedures and further the cause of cost-
containment while at the same time improving the long-term
health of patients with these problems.

Early results with each of the instruments described here
have been promising, but these results are by no means con-
clusive. The potential of these devices is yet to be reached. Cur-
rently we must rely on our experience and clinical judgment
in order to make the determination of procedure and instru-
mentation. To this end, well controlled prospective studies are
important. It is this ongoing process of technological advance-
ment and practical experience that gives us the best hope for
effectively and efficiently combatting atherosclerosis and its
thrombotic consequences.

References

1. Biermen EL. Atherosclerosis and other forms of arteriosclerosis.

In: Wilson LD, Braunweld E, Isselbecher KJ et al, eds. Harrison's
Principles of Internal Medicine. New York: McGraw-Hill, 1988.

2. Creager MA, Dzau VJ. Vascular diseases of the extremities. In: Wil-
son LD, Braunwald E, Isselbecher KJ et ah, eds. Harrison's Principles
of Internal Medicine. New York: McGraw-Hill, 1988.

3. Cotran RS, Kumar V, Robbins SL. Pathologic Basis of Disease, 4th
edn. Philadelphia: WB Saunders, 1989.

4. Rappaport SI. Introduction to Hematology, 2nd edn. Philadelphia:
Lippincott, 1987.

5. FogartyTJ, CranleyJJ, Krause RJ. A method for extraction of arteri-
al emboli and thrombi. Surg Gynecol Obstet 1963; 116:241.

6. Fogarty TJ, Hermann GD. New techniques for clot extraction and
managing acute thromboembolic limb ischemia. In: Veith FJ, ed.
Current Critical Problems in Vascular Surgery. St Louis: Quality
Medical Publishing, 1991:197.

7. Gruntzig A, Hopff H. Perkutane Rekanalisation chronischer ar-
terieller Verschlusse mit einen neuen Dilatationskatheter: Modifi-
cation der Dotter-technik. DtschMed Wochenschr 1974; 9:2502.

8. White RA, White GH. A Color Atlas of Endovascular Surgery.
Philadelphia: JB Lippincott, 1990.

9. Husten L. Atherectomy lowers restenosis when compared with
angioplasty. Newspaper of Cardiology 1993; Jan:l.

10. Simpson JB, Selmon MR, Robertson GC et al. Transluminal
atherectomy for occlusive peripheral vascular disease. Am }
Cardiol 1988; 61 :96G.

11. Wholey MH, Jarmolowski CR. New reperfusion devices: the
Kensey catheter, athero lytic reperfusion wire device and the TEC.
Radiology 1989; 172:947.

12. Snyder SO Jr, Wheeler JR, Gregory RT, Gayle RG, Mariner DR. The
Kensey catheter: preliminary results with a transluminal atherec-
tomy tool. / Vase Surg 1988; 8:541.

13. Zacca NM, Raizner AE, Noon GP et al. Treatment of symptomatic
peripheral atherosclerotic disease with a rotational atherectomy
device (Rotablator). Am J Cardiol 1989; 63:77.

14. Sniderman KW, Bodner L, Saddekni S, Srur M, Sos TA. Percuta-
neous embolectomy by transcatheter aspiration. Radiology 1984;
150:357.

15. Langham MR Jr, Greenfield LJ. Transvenous catheter embolec-
tomy for life-threatening pulmonary embolism. Infect Surg 1986;
5:694.

16. Greenfield LJ, Proctor MC, Williams DM, Wakefield TW. Long
term experience with transvenous catheter pulmonary embolec-
tomy. / Vase Surg 1993; 18:450.

502

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Balloon angioplasty and transluminal
recanalization devices

Rajesh Subramanian
Stephen R. Ramee

Endovascular therapy has come a long way since Charles
Dotter first described angioplasty with stiff Teflon catheters to
treat atherosclerotic obstructive lesions in the vasculature. 1
While several individuals have contributed to the use and suc-
cess of endovascular therapies, the efforts of Andreas
Gruntzig and John Simpson, with their concepts of expand-
able polyvinyl chloride balloon catheters and steerable
guidewires, respectively, revolutionized the technique of
balloon angioplasty in the 1980s and were responsible for the
rapid technological advances in recent years. 2 ' 3 While assess-
ing the role and impact of endovascular therapies it would be a
fallacy to consider balloon angioplasty and stenting separate-
ly. What follows below is a description of the indications and
technique of balloon angioplasty, which is similar for stenting.
Further, while the results of balloon angioplasty are discussed
below, one should bear in mind that balloon angioplasty with-
out stenting is of historical interest.

General principles

Mechanism of balloon angioplasty

Balloon angioplasty was initially thought to increase the
arterial lumen size by compressing the atherosclerotic plaque
against the arterial wall. 4 Plaque compression is no longer
thought to play a major role. Luminal expansion is now
thought to result from fracturing or breaking of the athero-
sclerotic plaque, along with the creation of intimal flaps and
dissection of the arterial media following balloon inflation. 5
Further, at sites of eccentric plaques, balloon inflation results
in stretching of the normal vessel segment resulting in luminal
expansion. 6 The vessel, almost immediately, responds to this
injury caused by balloon inflation by a process of remodeling. 7
This vessel remodeling process, including elastic recoil and
neointimal hyperplasia, is responsible for restenosis following
successful luminal expansion with balloon angioplasty. 8

Indications

The indication for peripheral vascular intervention is the pres-
ence of symptoms secondary to stenosis or occlusion in the ar-
terial or venous system. The interventionist must insure that
the risk-to-benefit ratio favors intervention. Evaluation of the
patient for angioplasty includes a careful history, physical ex-
amination, and review of noninvasive testing. Most sympto-
matic lesions can be diagnosed without angiography. The role
of angiography is to confirm clinical suspicion and results
of noninvasive testing, determine the number, location, and
morphology of lesions, and to serve as a roadmap for revascu-
larization. The goals of treatment are to relieve symptoms,
preserve organ function, and /or to prolong life. Selection of
patients for revascularization must take into account the
severity of symptoms, the angiographic findings, and the
risk-to-benefit ratio for revascularization. Furthermore, one
should consider the alternative therapies available, including
medical and surgical options. When doing so, the morbidity
and durability of the treatment options should be carefully
assessed. 9

Procedural success

Procedural successes in carefully selected patients undergo-
ing balloon angioplasty in any vascular bed are quite high. For
stenoses and occlusions less than 3 cm long, procedural suc-
cess without a major complication is 99%. In long stenoses and
occlusions more than 3cm in length, the success rate is 80%. 10
In general, short, discrete, concentric, nonostial, stenotic le-
sions without significant calcium are best suited for balloon
angioplasty. 11 The presence of ostial involvement, an eccentric
plaque, or the presence of significant calcium in the lesion
adversely affects the technical success rate for percutaneous
transluminal balloon angioplasty (PTA). 12 Additionally, the
pathology of the lesion influences outcome, with fibromuscu-
lar dysplastic lesions being associated with improved out-
comes compared with atherosclerotic lesions.

The acute and long-term results of balloon angioplasty dif-

503

pa rt v Endovascular interventions for vascular disease

Table 43.1 Sheaths and guiding catheters commonly used in peripheral intervention

Carotid

Vertebral

Subclavian/innominate

Renal/mesenteric

Aortoiliac

Femoral/popliteal
Infrapopliteal

Envoy Guide (6 Fr), Multipurpose Guide (6-8 Fr), Shuttle Sheath (6-8 Fr)
Envoy Guide (6 Fr), Multipurpose Guide (6-8 Fr), JR-4 Guide (6-8 Fr)

Envoy Guide (6 Fr), Multipurpose Guide (6-8 Fr), IMA Guide (6-8 Fr)

IMA Guide (6-8 Fr), Hockey Stick Guide (6-8 Fr)

Regular Sheath (6-8 Fr), Brite tip 35 cm long sheath (6-8 Fr)

Crossover Sheath (6-8 Fr), Arrow Sheath (antegrade femoral access; 6-8 Fr)

Crossover Sheath (6-8 Fr), Multipurpose Guide (6 Fr), Arrow Sheath
(Retrograde femoral access;6-8 Fr); Regular Sheath (antegrade femoral
access; 6-8 Fr)

fer in the different vessels and different lesion morphology.
Long-term outcome depends on clinical and anatomical fac-
tors. For example, restenosis rates are lower in claudicants vs.
in limb salvage, in aortoiliac disease vs. femoropopliteal or
tibioperoneal disease, and with a good distal runoff vs. a poor
distal runoff. 13

Role of stenting

Stenting has broadened the indications for intervention and
dramatically improved the acute and long-term success of
endovascular intervention. This chapter will be restricted to a
discussion of balloon angioplasty indication and techniques
since stenting is the subject of the subsequent chapter. The
reader should keep in mind, however, that any balloon angio-
plasty result that is suboptimal (>30% residual stenosis or
>5-10mm gradient postangioplasty) should be stented to pre-
serve the acute success and organ viability and avoid the need
for emergency bypass surgery. For a complete discussion of
stenting indications and techniques see Chapter 44.

Technique

All patients are pretreated with oral antiplatelet therapy, in-
cluding aspirin (325 mg q.d.) and /or clopidogrel (300-mg load
followed by 75 mg q.d.) 24-48 h before the procedure. Irrespec-
tive of location, balloon angioplasty is performed in a series of
steps.

access site. The common femoral artery is often the preferred
location of vascular access. This is the most common vascular
access site for diagnostic angiography and thus operator fa-
miliarity plays a critical role in its selection for intervention.
Most vascular beds can be approached via a femoral route
with infrainguinal intervention via a retrograde contralateral
or an ipsilateral antegrade approach and suprainguinal, aor-
tic, and that of most aortic branches via a retrograde common
femoral approach. However, other vascular access sites may
be preferred in specific situations. A brachial or radial artery
approach may be preferred when there is the presence of ex-
cessive tortuosity or occlusive disease in the aortoiliac seg-
ment. When planning renal or mesenteric angioplasty a target
vessel with a cephalad takeoff may be better approached from
the arm. Angioplasty of the brachiocephalic or vertebral artery
may also be better approached from the ipsilateral radial or
brachial artery in cases of excessive tortuosity of the subcla-
vian or brachiocephalic artery. There may be other situations
(some of these are discussed below) where a particular ap-
proach may be better suited for a particular target lesion and
hence this key step of obtaining vascular access must be
planned for carefully. While considering issues regarding vas-
cular access, it is important to consider the distance between
the access site and the target vessel as distance may limit deliv-
erability of equipment.

Using the modified Seldinger technique, a needle and wire
are inserted percutaneously and then a sheath is inserted in a
coaxial manner atraumatically. Heparin (3000-5000 U) is
administered by either the intravenous or intraarterial route.

Vascular access

The first and most important step is obtaining vascular access.
The proper choice of vascular access and technical success
of placing a percutaneous sheath is the key to success for
peripheral intervention. Most target arterial lesions may be
approached from more than one vascular access site (see
Table 43.4). Familiarity of the operator, proximity or ease of ap-
proachability to the target vessel, or technical concerns regard-
ing the usual or preferred site dictate the choice of the vascular

Baseline angiography

After obtaining vascular access one then proceeds with
obtaining baseline angiography. An appropriate diagnostic
catheter is used to cannulate the target vessel and imaging is
performed to locate and visualize the target lesion. If diagnos-
tic images were performed previously then one may choose
the angiographic views that best uncovered the stenosis and a
"working view" may then be selected. On occasion multiple
angulations and different angiographic views may be needed

504

chapter 43 Balloon angioplasty and transluminal recanalization devices

to uncover the lesion. It is important to visualize the entire tar-
get vessel including the inflow and outflow. While imaging
one must continue cine-imaging until the venous flow is
reached to establish the presence or absence of collateral circu-
lation. It is equally important to image other vessels that are
known sources of collateral circulation to the culprit vessel
(e.g. imaging all the arch vessels when there is occlusive dis-
ease involving one of the arch vessels or imaging the internal
mammary artery when there is aortoiliac occlusion and there
is a paucity of collaterals from the subdiaphragmatic aortic
branches).

With regard to the choice of radiographic contrast we prefer
the use of low osmolar agents, as these are associated with less
adverse effects and improved patient comfort. Both ionic
(Hexabrix) and nonionic (Omnipaque) agents may be used.
While injecting radiographic contrast one must take into ac-
count the vessel diameter. Radiographic contrast injection of a
sufficient volume and rate should be performed to well visual-
ize the vessel. Insufficient contrast volume or rate of injection
may lead to inadequate visualization of the lesion.

There have been advances in imaging technology that can
be very advantageous for the practicing vascular specialist.
While performing imaging of the peripheral vasculature,
cineangiography is preferred over conventional cut film as
this allows an appreciation of blood flow in addition to
lumenography Digital imaging has significantly improved
image quality. Further utilization of the technique of digital
subtraction angiography (DSA) is extremely useful compared
with standard digital angiography. DSA enables one to image
the vasculature without interference of soft tissue or bony
artifacts. This technique also enables one to utilize less
radiographic contrast. DSA is performed by initially making a
mask of the area of interest followed by imaging with radio-
graphic contrast, the mask of the nonvascular structures is
removed leaving the image of the contrast-filled vasculature.
Another technique that is invaluable to the interventionist is
roadmapping. During the technique of roadmapping a nega-
tive image of the vascular area of interest is superimposed on
the fluoroscopic image. This allows one to "visualize" the ves-
sel as one is advancing the guidewire across the area of steno-
sis or occlusion. This technique increases success in crossing
the lesion as well as decreasing procedural time and thus the
radiation exposure.

Choice of equipment

Once the baseline angiography is performed, it is important to
evaluate the target vessel diameter. This may be performed
with the use of intravascular ultrasound or by comparing the
target vessel diameter with an object of known dimensions.
Depending on the location of the target lesion an appropriate
sheath is selected (6-8 Fr, standard vs. long vs. crossover
sheath) . The size of the sheath is often dictated by the size of the
balloon to be used or stent to be delivered (Table 43.1).

Table 43.2 Guidewires used in peripheral interventions

Carotid

Intracranial
Vertebral

Subclavian
Renal

Mesenteric

Aortoiliac

SFA/popliteal

Infrapopliteal

0.18inx300cm
0. Minx 300 cm

0. 14 in x 300 cm —

0.18inx300cm —
0.14inx300cm —

0.18inx300cm
0.35inx300cm
0.35inx190cm

Benson

0.18inx190cm

0.35inx190cm-

0.14inx190cm

0.35inx190cm

0.14inx190cm —

0.18inx190cm
0.35inx190cm

0.14inx190cm —

0.18inx190cm
0.14inx190cm

0.18inx190cm —

Roadrunner*

Sportt

Balanced medium weightt

Balanced heavy weightt

Platinum Plus*

Choice PT*

Whispert

Roadrunnerwire

Sport

Balanced medium weight

Balanced heavyweight

Platinum Plus

Roadrunner

Amplatz exchange*

Wholey§

Spartacoret Balanced

Middleweight

Steelcoret

Wholey

Spartacoret

Wholey

Amplatz

Glide

Bentsont

Sport

Platinum Plus

Spartacore

Roadrunner

Steelcore

Wholey

Amplatz extra stiff

Glide

Benson

Rosent

Sport

Platinum Plus

Spartacore

Roadrunner

Steelcore

Sport

Platinum Plus

Spartacore

Roadrunner

Steelcore

Manufacturer: *Cook. tGuidant, Temecula, CA, USA. tBoston Scientific,
Natick, MA, USA. §Mallinckrodt, St Louis, MO, USA.

The next step is the choice of guidewire. Peripheral
guidewires come in a range of sizes from 0.014 in to 0.038 in in
diameter (Table 43.2). Guidewires may be either hydrophilic
(e.g. Glide-wire) or nonhydrophilic (e.g. Wholey or Amplatz).
Hydrophilic guidewires are preferred when crossing occlu-
sions or traversing complex lesions. Coronary guidewires

505

pa rt v Endovascular interventions for vascular disease

Table 43.3 Balloon catheters used in peripheral intervention

0.14 in

Guidant, Temecula, CA

Opensail
Crosssail
Powersail

Boston Scientific, Natick, MA

Maveric
NC Monorail
Ranger
NC Bandit

Medtronic, Minneapolis, MN

0.18 in

Guidant, Temecula, CA

Viatrac

Cordis, Miami, FL

Slalom

Medtronic, Minneapolis, MN

Talon

0.35 in

Cordis, Miami, FL

Opta
Powerflex

Boston Scientific, Natick, MA

Marshall

Ultrathin Diamond
Synergy

Table 43.4 Vascular access

Common femoral artery

Retrograde

Antegrade

Contralateral
Popliteal artery
Brachial artery
Radial artery

may also be used, especially in vessels with smaller diameter
like the tibioperoneal or accessory renal arteries. An under-
standing of the characteristic properties of guidewires such as
floppiness, stiffness, and steerability is essential in selecting
the appropriate guidewire. Floppy guidewires such as Who-
ley, Spartacore are less likely to cause vessel trauma and may
be better suited for use in renal angioplasty compared with
stiff guidewires like Glide-wire which is more likely to cause
vessel trauma and is better suited in crossing total occlusions.
A balloon of suitable diameter equal to that of the reference
vessel immediately proximal to the lesion and of suitable
length to cover the length of the stenosis is selected
(Table 43.3). Also, the diameter of the guidewire used to cross
the lesion dictates the specific balloon selected.

Intervention

The target lesion is crossed with the guidewire atraumatically
(the goal is to use finesse over force) and placed beyond the
target lesion. Care is taken to ensure that the tip of the wire is
within the vasculature. The appropriate balloon selected is
placed across the lesion. The balloon is positioned to
completely cover the lesion. This is confirmed by injection of
radiographic contrast to localize the balloon with reference to

the lesion. Balloon inflation is then performed to nominal or
burst pressure rating under fluoroscopic visualization until
the waist of the lesion resolves. A cineangiogram is performed
to document the balloon inflation. The balloon is then deflated
and removed from the target vessel leaving the guidewire in
place beyond the target lesion. A cineangiogram of the target
vessel is performed to evaluate the results of balloon angio-
plasty. Multiple angiographic views may be required to evalu-
ate completely the effectiveness of balloon inflation. The
presence of a dissection flap or inadequate vessel expansion or
recoil with > 30% residual stenosis denotes an inadequate re-
sult and one must consider stenting to treat the suboptimal
balloon angioplasty result. Occasionally angiography alone
may be insufficient to resolve the adequacy of balloon
angioplasty. Intravascular ultrasound may be of use in this
situation. Another useful technique is the use of a 4-Fr multi-
purpose or transit catheter to measure pressure gradients
across the lesion. A residual gradient of > 5-10mmHg may sig-
nify an inadequate result.

Aortoiliac and lower extremity intervention

Aortoiliac angioplasty

Indications

The indications for revascularization in the aortoiliac segment
are lifestyle-limiting claudication, limb-threatening ischemia,
or to maintain femoral access or prevent access complications
from angiography or intervention in other vascular beds.

Vascular access and technique

While considering the technical performance of balloon
angioplasty, the aortoiliac region can be considered as three
contiguous segments: the infrarenal aorta not involving the
aortoiliac bifurcation, the aortailiac bifurcation, and the com-
mon and external iliac artery not involving the common iliac
ostia. Vascular access (Table 43.4) may be obtained in either
common femoral artery for infrarenal aortic angioplasty. Ipsi-
lateral retrograde common femoral access is preferred for iliac
angioplasty not involving the aortic bifurcation. Aortic
bifurcation angioplasty is performed by a kissing balloon
technique by simultaneous balloon inflations with balloon
catheters in both common iliac segments following retrograde
access in both common femoral arteries (Fig. 43.1). Iliac angio-
plasty may also be performed via a brachial artery access.

The ipsilateral or contralateral common femoral artery is
cannulated with a 6- or 7-Fr sheath. The lesion is crossed with a
steerable guidewire. A PTA balloon is chosen that is equal in
diameter to the vessel to be dilated. The reference vessel diam-
eter is obtained by quantitative angiography using a reference
object of known dimensions or by intravascular ultrasound.

506

chapter 43 Balloon angioplasty and transluminal recanalization devices

Figure 43.1 Digital subtraction angiographic
anterior-posterior view of the aortoiliac
bifurcation demonstrating bilateral aortoiliac
stenosis (A). (B) Digital angiographic
(unsubtracted) anterior-posterior view of the
aortoiliac bifurcation demonstrating
simultaneous balloon inflations in the aortoiliac
bifurcation (the kissing balloon technique).
(C) Resolution of stenosis following stenting.

Computed tomography and duplex ultrasound measure-
ments are not accurate and should not be used. The balloon is
inflated until complete balloon inflation is noted and a radi-
ographic record of the dilation is recorded. Nominal balloon
inflation pressure (6-8 ATM) or higher pressure (up to the
rated burst pressure, 12-14 ATM) is used until complete ex-
pansion of the balloon within the lesion is noted. The patient is
observed for abdominal or back pain, which denotes stretch-
ing of the adventitia of the vessel, and is a warning that larger
balloon size and higher inflation pressures should not be used.
A suboptimal balloon result, as manifest by a 30% or greater
stenosis or 5 mm or greater translesion gradient, should be
treated with a stent.

Outcomes

There have been several small series demonstrating the feasi-
bility of balloon angioplasty of the infrarenal aorta, with tech-
nical success ranging from 88% to 100%. 14_17 In the largest
series of 46 patients reported by Elkouri et al., 18 the technical
success rate was 96%, with 83% demonstrating clinical and he-
modynamic improvement. The primary patency rate in this
series at 4 years was 70%, with a higher rate in the absence of
aortoiliac disease than in the presence of aortoiliac involve-
ment (83% vs. 55%). In our experience technical success can be
achieved in over 99% of cases with patency being significantly
improved with the use of intravascular stents. 10 ' 19

Becker et al. 20 analyzed 2697 iliac angioplasty procedures
and found a 92% technical success rate with 2- and 5-year
patency rates of 81% and 70%, respectively. In the only ran-
domized study of surgery vs. angioplasty for iliac angioplasty
there was equivalence between the two strategies with a 73%
3-year patency rate. 21

Femoropopliteal angioplasty

Indications

Femoropopliteal angioplasty (Fig. 43.2) is most often
performed for life-style-limiting claudication. Critical limb
ischemia from isolated chronic occlusive disease in the
femoropopliteal segment is uncommon.

Vascular access and technique

Access is obtained either in the contralateral common femoral
artery, the ipsilateral common femoral artery, or the popliteal
artery. Contralateral femoral access can allow femoropopliteal
intervention after placement of a crossover sheath across the
aortic bifurcation and into the ipsilateral external iliac artery. 22
Antegrade access in the ipsilateral common femoral artery is
preferred especially when there is excessive tortuosity in the
common iliac arteries or presence of ipsilateral distal external
iliac, common femoral, or proximal superficial femoral artery
(SFA) stenosis. It may be problematic for angioplasty of the
proximal or ostial lesions in the SFA. In crossing total occlu-
sions of the SFA a retrograde popliteal artery approach may be
useful if a common femoral artery approach is unsuccessful.
This technique utilizes proximal iliac angiography to define
the popliteal artery location, vessel size, and presence or
absence of access site lesions.

Outcomes

Earlier studies suggested that the long-term outcome of
femoropopliteal angioplasty is comparable to polytetrafluo-
roethylene bypass surgery but less successful compared with

507

pa rt v Endovascular interventions for vascular disease

Figure 43.2 Digital subtraction angiographic
anterior-posterior at the level of the left knee (A)
demonstrating a focal high-grade stenosis of the left
popliteal artery and (B) demonstrating the stenosis
following balloon angioplasty.

venous bypass. 23 Clark et al. 24: recently reported long-term pa-
tency after femoropopliteal angioplasty from a multicenter
registry Primary patency at 12, 24, 36, and 48 months of
follow-up was 87%, 80%, 69%, and 55%, respectively Lesion
length greater than 5 cm, multiple lesions, poor distal vessel,
runoff, and diabetes mellitus have been correlated with de-
creased primary patency 25 In a recent meta-analysis stenting
has been shown to improve patency in patients with more
severe femoropopliteal disease compared with balloon
angioplasty alone. 26 In several recent reports, femoral artery
stenting has shown excellent results that rival those of even
vein graft bypass surgery. 27

Tibioperoneal angioplasty

Indications

Tibioperoneal intervention is usually performed for severe
claudication or critical limb ischemia including ischemic rest
pain, gangrene, or ischemic ulceration.

Vascular access and technique

Angioplasty of the tibioperoneal segment may be performed
from either an ipsilateral antegrade common femoral artery
approach or from the contralateral common femoral artery via
a crossover technique. As in the case of the femoropopliteal
segment, one would use a short 5- or 6-Fr sheath with an ipsi-
lateral antegrade approach and a crossover sheath with the

retrograde contralateral approach. Coronary angioplasty
0.014-in guidewires and low profile balloon catheters are
better suited than peripheral balloons and 0.035-in wires in
these 2.0-4.0-mm diameter vessels.

Outcomes

Kandarpa et al. 28 performed an analysis of the available litera-
ture in 1282 treated limbs, and noted a 93% technical success
rate with a 1-year limb salvage rate of 74%. The primary indi-
cation in 86% of this group was critical limb ischemia. Dorros
et al. 29 published their experience and 5-year follow-up of
tibioperoneal angioplasty in 235 patients with critical limb is-
chemia. They reported a 95% technical and clinical success
rate. At 5 years of follow-up surgical revascularizations were
performed in 8% and amputations in 9% of patients. These
results suggest that tibioperoneal angioplasty is as effective
for limb salvage as is distal bypass surgery.

Mesenteric and renal artery intervention

Renal artery angioplasty

Indications

Atherosclerosis and fibromuscular dysplasia are the most
common causes of renal artery stenosis. Balloon angioplasty
of the renal arteries is most often performed for fibromuscular

508

chapter 43 Balloon angioplasty and transluminal recanalization devices

Figure 43.3 Digital subtraction angiographic
anterior-posterior of the right kidney. (A) The
typical rosary bead appearance of
fibromuscular dysplasia. (B) An excellent
balloon angioplasty result.

dysplasia in patients with suspected renovascular hyper-
tension (Fig. 43.3). PTA has been shown to be ineffective in
treating atherosclerotic renal artery stenosis, especially in
aorto-ostial lesions, and these should generally be treated with
balloon expandable stents. This technique will be discussed in
subsequent chapters.

Vascular access and technique

Renal angioplasty may be performed after obtaining retro-
grade access in the common femoral artery or via the brachial
artery. In these patients, balloon angioplasty can be performed
with or without a guiding catheter from the femoral or
brachial approach. 30 The renal artery ostium is engaged with a
soft, 5- or 6-Fr diagnostic catheter (JR4, Cobra, IMA, or
Simmons) and diagnostic angiography is performed. Bony
landmarks or roadmapping are used to localize the area of
fibromuscular disease. A soft, straight- tipped guidewire such
as a Wholey (Guidant, Santa Clara, CA, USA) or Magic Torque
(Meditech, Billerica, MA, USA) is used to cross the stenosis.
Unlike in atherosclerotic disease, a balloon-to-artery ratio of
1.1 : 1 is used to dilate fibromuscular disease. The balloon is
advanced over the guidewire with or without the aid of a 6- or
7-Fr guiding catheter (hockey stick shape) and inflated to
nominal pressure (6-8 ATM). Again, the patient is monitored
for evidence of back pain signifying that there is stretching of
the adventitia, a warning not to oversize or overinflate the
balloon. Final angiography is then performed to detect the
presence of dissection that may need treatment with a stent.

cular dysplasia as well as nonostial atherosclerotic renal artery
stenosis. While clinical benefit occurred in most, fibromuscu-
lar dysplasia patients were more likely to experience benefit
compared with those with nonostial atherosclerotic lesions
(93% vs. 70% of patients with technically successful proce-
dures). Technical success, clinical benefit, and absence of
restenosis are significantly decreased in the setting of athero-
sclerotic lesions, especially those involving the ostium, and
these patients may be better managed with renal artery stent-
ing. In a representative population of patients with fibromus-
cular dysplasia undergoing balloon angioplasty Tegtmeyer
et al. 31 reported a 100% technical success rate, 87% patency rate
at 10 years, with improvement in renal function in 86% and
cure of hypertension in 39% of 66 patients.

Celiac and mesenteric artery intervention

Indications

Chronic intestinal ischemia is an uncommon manifestation of
atherosclerosis that results most often from atherosclerotic
stenosis or occlusion involving the origin of all three mesen-
teric vessels: the celiac, superior, and inferior mesenteric
arteries (SMA and IMA). In the great majority of patients,
involvement of two or more of the major sphlanchnic vessels is
required to cause symptoms. This is due to an extensive collat-
eral circulation network in the splanchnic vessels. Endovascu-
lar therapy is emerging as a viable therapeutic option in the
management of these patients.

Outcomes

Percutaneous transluminal balloon angioplasty is the treat-
ment of choice for fibromuscular dysplasia. Pooled analysis by
Becker et al. 20 of 1108 patients with renal artery stenosis
demonstrated a technical success rate of 90% in both fibromus-

Vascular access and technique

Mesenteric angioplasty is technically similar to renal angio-
plasty. With the exception of patients with fibromuscular dys-
plasia (FMD), these lesions are atherosclerotic and located
either at the aorto-ostium or in the proximal mesenteric vessel.

509

pa rt v Endovascular interventions for vascular disease

Figure 43.4 Angulated view of the superior
mesenteric artery demonstrating a focal high-
grade stenosis (A). There is a suboptimal
response to balloon angioplasty with a > 50%
residual stenosis (B). (C) Resolution of the
stenosis following stenting.

Figure 43.5 A complex stenotic plaque involving the innominate
bifurcation and the origins of the right subclavian and common carotid

arteries (A). Stents positioned with the kissing balloon technique (B).
Poststent resolution of the stenosis (C).

Atherosclerotic lesions do not respond well to angioplasty,
and do respond well to stenting (unpublished personal experi-
ence). Vascular access for balloon angioplasty may be ob-
tained either retrograde in the common femoral artery or via
the brachial artery Diagnostic catheters from the femoral
artery (Cobra, IMA, or Simmons) or brachial artery (Multipur-
pose) are used to find the ostium of the SMA, celiac, or IMA
and perform diagnostic angiography. The balloon angioplasty
technique is similar to that described above for renal artery
FMD, and again provisional stenting is reserved for flow-lim-
iting dissection following intervention (Fig. 43.4).

Outcomes

The technical success rate in more recent reports is
90-95%. 32_34 Clinical success with relief of symptoms is ob-
tained in 80-90% of patients in follow-up of 2-3 years. Patients
with classical symptoms of abdominal angina obtain the most
benefit.

Aortic arch, subclavian, carotid,
and vertebral intervention

Subclavian intervention

Indications

Endovascular therapy is an excellent treatment strategy
for focal stenosis or occlusive lesions of the aortic arch
vessels (Fig. 43.5). Balloon angioplasty is indicated for relief
of ischemia manifesting as arm claudication, arm weakness,
posterior circulation ischemia from subclavian steal syn-
drome, and coronary ischemia due to inadequate flow to
the internal mammary artery used for coronary artery
bypass. Stenting is used for the majority of atherosclerotic
lesions, with balloon angioplasty reserved for fibromuscular
dysplasia and selected, symmetrical, focal atherosclerotic
stenosis.

510

chapter 43 Balloon angioplasty and transluminal recanalization devices

Figure 43.6 Angulated digital subtracted
image of the intracranial left internal carotid
artery at the level of the carotid siphon
demonstrating a high-grade stenosis (A).
Resolution of stenosis following balloon
angioplasty (B).

Vascular access and technique

Arterial access is obtained in the common femoral artery or
ipsilateral brachial or radial artery In the presence of occlusion
of the subclavian artery or axillary artery one may on occasion
be unable to palpate a brachial pulse. In this situation one may
perform arterial puncture after fluoroscopically visualizing
the brachial artery following injection of contrast dye in the
subclavian artery and waiting for the reconstitution of the
brachial artery via collateral circulation.

Outcomes

Results of balloon angioplasty are very favorable in the upper
extremity and brachiocephalic vessels. The procedural suc-
cess rate is 90-95% for stenosis and lower for total occlu-
sions. 35-37 While there is a paucity of literature regarding
long-term outcome, it is felt to be similar to surgery, with pro-
cedural morbidity and mortality less than 1%. 12 As in renal,
mesenteric, and lower extremity intervention, stenting has
become the mainstay for intervention in atherosclerotic sub-
clavian artery stenosis and will be discussed in subsequent
chapters.

Carotid and vertebral angioplasty

Endovascular therapy for extracranial carotid and vertebral
arteries is almost entirely limited to stenting and is thus not
discussed here.

Intracranial angioplasty

Indications

Intracranial stenosis may account for 5-10% of all ischemic

strokes. 38,39 Intracranial stenosis portends poor prognosis
with a 30-50% risk of stroke. 40 ' 41 Balloon angioplasty is emerg-
ing as an effective therapy for treating symptomatic intra-
cranial stenosis. This is a high-risk, technically challenging
technique that requires extensive knowledge of the intracra-
nial anatomy, physiology, and a high level of technical compe-
tence. We feel this is best managed by a multidisciplinary team
consisting of a neuroradiologist, interventionist familiar with
small vessel intervention and stenting, and a neurologist.

Vascular access and technique

Vascular access is generally obtained in the common femoral
artery. An appropriate guiding catheter is then advanced into
the common carotid (for anterior circulation vessels) or
the vertebral (for posterior circulation vessels) arteries. Coro-
nary guidewires and balloons are preferred due to the small
diameter of these vessels. Typically, balloon inflation times are
short, and maximum inflation pressures low (4-6 ATM) to
prevent prolonged cerebral ischemia. In our experience a sub-
optimal angiographic result (residual stenosis after balloon
angioplasty or nonflow-limiting dissection) may not necessar-
ily be associated with adverse outcome and stents are general-
ly avoided as these may cause side-branch occlusions with
potential catastrophic results (Fig. 43.6).

Outcomes

There are currently small series of patients in whom balloon
angioplasty has been demonstrated to be safe and technically
feasible. 42 Kandarpa et alP evaluated outcomes of all reported
cases of intracranial angioplasty. They noted an increased
technical success associated with a lower rate of complications
for all procedures done after 1997 compared with those done
prior to 1997, suggesting a learning curve. When analyzing

511

pa rt v Endovascular interventions for vascular disease

procedures done after 1997 they noted a technical successful
outcome in 87.7% of patients with a complication rate of
7.2%. 43 Our approach in patients with symptomatic intracra-
nial stenosis has been of provisional stenting. A multidiscipli-
nary team consisting of neuroradiologists, neurologists, and
interventional cardiologists assists by bringing their varied
expertise, ensuring appropriate patient selection and inter-
ventional strategy. Using this approach we reported a 100%
technical success rate with a 1-year freedom from death or
stroke rate of 93.4% in 15 patients. 44 Our experience has since
increased to 29 patients with similar success rates. An interest-
ing observation has been the finding of an unexpected neuro-
logical benefit in over half this cohort.

Bypass graft

Indications

Failure of femoropopliteal bypass grafts within the first week
is usually due to technical factors and is best treated with oper-
ative correction. Balloon angioplasty is, however, useful in the
management of late graft failure due to neointimal hyperpla-
sia at the anastamotic sites. Patients are usually followed by
periodic duplex evaluation to detect a failing graft prior to
complete occlusion. This is one place where intervention is
recommended even in the absence of symptoms to prevent
graft failure, which can be catastrophic.

Vascular access and technique

Contralateral retrograde common femoral access technique
may be the best-suited approach for lower extremity bypass
grafts. The approach and technique are similar to those of the
native vessels.

Outcomes

Balloon angioplasty is an excellent modality to maintain sec-
ondary patency of a failing bypass graft. 45 Goh et al 46 reported
a successful outcome in 39 of 40 patients with a 1- and 5-year
cumulative patency of 79% and 63%, respectively. In general
stenotic lesions in the body of the graft have better outcomes
with balloon angioplasty compared with juxta-anastomotic
lesions. 47

Hemodialysis access angioplasty

Indications

The natural history of hemodialysis access grafts is the loss of
functioning of the majority in the absence of percutaneous or
surgical correction. 48 Failure of hemodialysis access grafts can
result in inadequate dialysis, extremity edema, and access
thrombosis. Graft failure may occur from venous or arterial

anastomotic stenosis or intragraft stenosis. 49 Additionally,
central venous stenosis or occlusion may be caused by neointi-
mal proliferation resulting from vessel injury caused by
temporary central venous catheters and may compromise
the access graft. Percutaneous intervention is an effective
modality to maintain patency and a functioning hemodialysis

access

Vascular access and technique

Hemodialysis accesses are of two common types: endogenous
arteriovenous fistula using the Brescia technique and the poly-
tetrafluoroethylene arteriovenous graft. Vascular access may
be obtained by placing a 6-Fr sheath in the brachial artery in
the case of the endogenous fistula and in the venous limb in the
case of the polytetrafluoroethylene fistula. While evaluating
the hemodialysis access insufficiency it is important to evalu-
ate the venous drainage up to the superior vena cava. Angio-
plasty is performed using high-pressure balloon inflations
(10 ATM or greater) with balloons being sized 1 : 1 with the
reference vessel diameter.

Outcomes

There have been many series of patients treated with balloon
angioplasty of hemodialysis access grafts. A recurring theme
of these studies is that the life of the access graft can be
prolonged by recurrent angioplasty with low attendant
morbidity, though repeated procedures may be required. 51 ' 52

Atherectomy

Directional atherectomy is an obsolete procedure, made that
way by the introduction of stenting. Stenting is technically eas-
ier and associated with fewer complications than directional
atherectomy. Rotational atherectomy is indicated in patients
with calcified and ostial lesions that are undilatable or deemed
unacceptable for simple balloon angioplasty and stenting.
This is a rare situation, but when it happens knowledge
and proficiency in rotational atherectomy can be limb saving.
Trials with the Rotablator (Scimed, Boston Scientific Corp.,
Boston, MA, USA) in inf rainguinal atherectomy have reported
primary, secondary, and clinical success rates of 61%, 67%, and
56% at 12 months with technical success rate of 94%. 53 After
initial enthusiasm the precise role of plaque debulking with
atherectomy devices as an adjunct to balloon angioplasty ex-
cept in undilatable or heavily calcified lesions remains to be
defined.

Laser angioplasty

Laser angioplasty (Fig. 43.7) also grew out of efforts to evalu-
ate therapeutic alternatives to balloon angioplasty for un-
favorable lesions. While atherectomy is performed by shaving

512

chapter 43 Balloon angioplasty and transluminal recanalization devices

Figure 43.7 Total occlusion of the proximal left superficial femoral artery
(A) with reconstitution at the level of the adductor canal (B). Laser
angioplasty catheter crossing the total occlusion (C). Angiography postlaser

angioplasty demonstrating complete recanalization of the left superficial
femoral artery (D,E).

the atherosclerotic plaque, the laser causes photomechanical
ablation by vaporizing the atherosclerotic material. While
several different types of laser devices exist, current efforts re-
volve around the athermic 308-nm wavelength excimer laser
catheter. Luminal expansion with laser angioplasty alone is
inadequate and additional balloon angioplasty is required. As
with atherectomy the precise role for laser angioplasty re-
mains to be defined. Laser angioplasty is currently being stud-
ied for revascularization of chronic total occlusions of the iliac
and femoropopliteal arteries. While the results of the PEL A
(peripheral excimer laser angioplasty) trial should help clarify
its role in chronic SFA occlusions, it would appear that excimer
laser angioplasty is effective in recanalizing long total occlu-
sions with technical success rates of 90% with 1-year primary
and secondary patency rates of 65% and 76%. 54

Complications of angioplasty

Endovascular therapy, while effective, is not benign and is as-
sociated with significant potential for complications. Exten-
sive training and careful planning can decrease or prevent
some but not all complications. A complete understanding is
required for the practicing endovascular specialist as well as
others who are taking care of these patients prior to, during,
and after embarking on the endovascular therapy. The main
complications are related to access site problems, renal dys-
function, and distal embolization (Table 43.5).

Mortality directly related to peripheral angioplasty is low,
occurring in less than 0.5%. 9/55 The most common complica-
tions are related to the access site. These include bleeding, ves-
sel dissection, and arteriovenous fistula formation. Access site

Table 43.5 Major complications of peripheral angioplasty

Access site

Hemostasis/bleeding

Hematoma

Retroperitoneal bleeding

Pseudoaneurysm
Access site infection
Vessel trauma

Dissection

Arteriovenous fistula

Rupture

Target vessel

Dissection
Rupture

Embolic

Stroke
Renal failure
Limb loss

Radiographic

contrast related

Anaphylactoid reaction
Renal failure

bleeding resulting in hematomas occurs in up to 5% of pa-
tients, but most resolve without sequelae. However, infection,
retroperitoneal bleeding, pseudoaneursym formation, need
for transfusion, hypotension, and death can result. In the past,
operative correction of access site complication was required
in 2-3% of these patients. 9 Modern techniques to treat vessel
perforation with covered stents, access site bleeding with bal-
loon tamponade, and pseudoaneurysms with thrombin injec-
tion have decreased the rate of surgical treatment for access
site complications to less than 1% (unpublished Ochsner
Clinic results).

513

pa rt v Endovascular interventions for vascular disease

Renal failure manifesting as a transient or permanent rise of
serum creatinine may occur in 1-5% of patients and risk fac-
tors include diabetes mellitus, preexisting renal dysfunction,
diffuse atherosclerosis, and renal angioplasty. Transient renal
dysfunction results most often are contrast mediated, al-
though a significant proportion of renal dysfunction following
renal angioplasty may be due to distal embolization of athero-
sclerotic plaque debris. Adequate hydration, use of limited
contrast material, and use of adjuvant therapy with N-acetyl-
cysteine may decrease contrast-mediated renal failure. 56 ' 57
Emboli protection devices may have a role in preventing
embolization of atheroemboli during renal angioplasty.

Distal embolization of atherosclerotic debris may result in
stroke, renal failure, gangrene, livido reticularis, and may
occur in up to 1% of patients. 11 The clinical manifestation
varies with the vessel being treated, with renal dysfunction
occurring more often following renal angioplasty and stroke
following cerebrovascular intervention.

Other complications include flow-limiting vessel dissection
that may be treated with stenting. The need for bypass surgery
due to dissection or perforation occurs infrequently.

Conclusion

Balloon angioplasty has come a long way since Charles Dotter
first described it. Endovascular therapy has emerged as an ef-
fective therapy for many if not most patients with arterial oc-
clusive disease. It has proven itself to be a clinically effective
modality in the management of peripheral vascular disease in
most arterial beds. An understanding of its current role and
limitations is important for those taking care of patients with
peripheral arterial disease. While restenosis remains the
Achilles' heel of balloon angioplasty, much progress has been
made reducing this problem with the use of intravascular
stents. While embarking on a treatment strategy it would be ju-
dicious to consider stenting as an extension of balloon angio-
plasty, as these are tools in the interventionist's arsenal with
one used to complement the other. Emerging therapies like
brachytherapy and drug eluting stents should further im-
prove patency rates. Another area of interest is the use of em-
bolic protection devices that have the potential of improving
outcomes by decreasing distal embolization and its associated
complications. Thus endovascular therapy has progressed
from being a therapeutic consideration for patients without
surgical options to surgery being an alternative in the absence
of an endovascular treatment approach.

References

1. Dotter CT, Judkins MR Transluminal treatment of atherosclerotic
obstruction. Description of a new technique and a preliminary re-
port of its application. Circulation 1964; 30:654.

2. Gruntzig A, Hopff H. Percutaneous recanalization after chronic
arterial occlusion with a new dilator-catheter (modification of the
Dotter technique). DtschMed Wochenschr 1974; 99:2502.

3. Simpson J, Bairn D, Robert E et ah A new catheter system for
coronary angioplasty. Am J Cardiol 1982; 49:1216.

4. Dotter CT, Judkins MP, Rosch J. Nonoperative treatment of arteri-
al occlusive disease: a radiologically facilitated technique. Radiol
Clin North Am 1967; 5:531.

5. Waller BR "Crackers, breakers, stretchers, drillers, scrapers,
shavers, burners, welders, melters" — the future treatment of ath-
erosclerotic coronary artery disease? A clinical-morphologic as-
sessment. J Am Coll Cardiol 1989; 13:969.

6. Waller BR Pathology of transluminal balloon angioplasty used
in the treatment of coronary artery disease. Hum Pathol 1987;
18:476.

7. Zarins CK, Lu CT, Gewertz BL et ah Arterial disruption and
remodeling following balloon dilatation. Surgery 1982; 92:1086.

8. Johnson DE, Hinohara T, Selmon MR et ah Primary peripheral
arterial stenosis and restenosis excised by transluminal atherecto-
my: a histopathological study. J Am Coll Cardiol 1990; 15:419.

9. Pentecost MJ, Criqui MH, Dorros G et ah Guidelines for peripheral
percutaneous transluminal angioplasty of the abdominal aorta
and lower extremity vessels. Circulation 1994; 89:511.

10. White CJ, Ramee SR, Collins TJ et ah Initial results of peripheral
vascular angioplasty performed by experienced interventional
cardiologists. Am J Cardiol 1992; 69:1249.

11. O'Keeffe ST, Woods BO, Beckmann CF. Percutaneous translumi-
nal angioplasty of peripheral arteries. Cardiol Clin 1991; 9:515.

12. Standards of Practice Committee of the Society of Cardiovascular
and Interventional Radiology. Guidelines for percutaneous
transluminal angioplasty. Radiology 1990; 177:619.

13. Johnston KW, Rae M, Hogg-Johnston S et ah Five year results of a
prospective study of percutaneous transluminal angioplasty.
Ann Surg 1987; 206:403.

14. Yakes WF, Kumpe DA, Brown SB et ah Percutaneous transluminal
aortic angioplasty: techniques and results. Radiology 1989; 172:965.

15. Odurny A, Colapinto RF, Sniderman KW et ah Percutaneous trans-
luminal angioplasty of abdominal aortic stenoses. Cardiovasc
Intervent Radiol 1989; 12:1.

16. Steinmetz OK, McPhail NV, Hajjar GE et ah Endarterectomy
versus angioplasty in the treatment of localized stenosis of the
abdominal aorta. Can J Surg 1994; 37:385.

1 7. Morag B, Garniek A, Bass A et ah Percutaneous transluminal aortic
angioplasty: early and late results. Cardiovasc Intervent Radiol 1993;
16:37.

18. Elkouri S, Hudon G, Demers P et ah Early and long-term results of
percutaneous transluminal angioplasty of the lower abdominal
aorta. / Vase Surg 1999; 30:679.

19. Martin EC, Katzen BT, Benenati JF et ah Multicenter trial of the
wallstent in the iliac and femoral arteries. / Vase Interv Radiol 1995;
843.

20. Becker GJ, Katzen BT, Dake MD. Noncoronary angioplasty.
Radiology 1989; 170:921.

21. Wilson SE, Wolf GL, Cross AP Percutaneous transluminal angio-
plasty versus operation for peripheral arteriosclerosis. / Vase Surg
1989; 9:1.

22. White CJ, Nguyen M, Ramee SR. Use of a guiding catheter for con-
tralateral femoral artery angioplasty. Cathet Cardiovasc Diag 1990;
21:15.

514

chapter 43 Balloon angioplasty and transluminal recanalization devices

23. Hunink MG, Wong JB, Donaldson MC et ah Patency results of
percutaneous and surgical revascularization for femoropopliteal
arterial disease. Med Decis Making 1994; 14:71.

24. Clark TW, Groffsky JL, Soulen MC. Predictors of long-term paten-
cy after femoropopliteal angioplasty: results from the STAR
registry. / Vase Interv Radiol 2001; 12:923.

25. Capek P, McLean GK, Berkowitz HD. Femoropopliteal angioplas-
ty: factors influencing long-term success. Circulation 1991; 83
(Suppl.I):70.

26. Muradin GS, Bosch JL, Stijnen T, Hunink MG. Balloon dilation
and stent implantation for treatment of femoropopliteal arterial
disease: meta-analysis. Radiology 2001; 221:137 '.

27. Ansel GM, Botti CF Jr, George BS, Kazienko BT; IntraCoil
Femoralpopliteal Stent Trial Investigators. Clinical results for the
training-phase roll-in patients in the Intracoil Femoralpopliteal
Stent Trial. Cathet Cardiovasc Interv 2002; 56:443.

28. Kandarpa K, Becker GJ, Hunink MG et ah Transcatheter interven-
tions for the treatment of peripheral atherosclerotic lesions: part I.
/ Vase Interv Radiol 2001; 12:683.

29. Dorros G, Jaff MR, Dorros AM et ah Tibioperoneal (outflow lesion)
angioplasty can be used as primary treatment in 235 patients with
critical limb ischemia: five-year follow-up. Circulation 2001;
104:2057.

30. White CJ, Ramee SR, Collins TJ et ah Guiding catheter-assisted
renal artery angioplasty. Cathet Cardiovasc Diag 1991; 23:10.

31. Tegtmeyer CJ, Selby JB, Hartwell GD et ah Results and complica-
tions of angioplasty in fibromuscular disease. Circulation 1991; 83
(Suppl. I):155.

32. Odurny A, Sniderman KW, Colapinto RF. Intestinal angina: per-
cutaneous transluminal angioplasty of the celiac and superior
mesenteric arteries. Radiology 1988; 167:59.

33. Allen RC, Martin GH, Rees CR et ah Mesenteric angioplasty in the
treatment of chronic intestinal ischemia. Vase Surg 1996; 24:415.

34. Maspes F, Mazzetti di Pietralata G, Gandini R et ah Percutaneous
transluminal angioplasty in the treatment of chronic mesenteric
ischemia: results and three years of follow-up in 23 patients.
Abdom Imaging 1998; 23:358.

35. Bogey WM, Demasi RJ, Tripp MD et ah Percutaneous transluminal
angioplasty for subclavian artery stenosis. Am Surg 1994; 60:103.

36. Hebrang A, Maskovic J, Tomac B et ah Percutaneous transluminal
angioplasty of the subclavian arteries: long-term results in 52
patients. Am J Roentgenol 1991; 156:1091.

37. Henry M, Amor M, Henry I et ah Percutaneous transluminal
angioplasty of the subclavian arteries. / Endovasc Surg 1999; 6:33.

38. Chimowitz MI, Kokkinos J, Strong J et ah The Warfarin- Aspirin
Symptomatic Intracranial Disease Study. Neurology 1995; 45:1488.

39. Wityk RJ, Lehman D, Klag M et ah Race and sex differences in the
distribution of cerebral atherosclerosis. Stroke 1996; 27:1974.

40. Craig DR, Meguro K, Watridge C et ah Intracranial internal carotid
artery stenosis. Stroke 1982; 13:825.

41. Marzewski DJ, Furlan AJ, St Louis P et ah Intracranial internal
carotid stenosis: long-term prognosis. Stroke 1982; 13:821.

42. Alazzaz A, Thornton J, Aletich VA et ah Intracranial percutaneous
transluminal angioplasty for arteriosclerotic stenosis. Arch Neurol
2000; 57:1625.

43. Kandarpa K, Becker GJ, Ferguson RD et ah Transcatheter interven-
tions for the treatment of peripheral atherosclerotic lesions: part II.
/ Vase Interv Radiol 2001; 12:807.

44. Ramee SR, Dawson R, McKinley KL et ah Provisional stenting for
symptomatic intracranial stenosis using a multidisciplinary ap-
proach: acute results, unexpected benefits, and one-year outcome.
Cathet Cardiovasc Intervent 2001; 52:457.

45. Houghton AD, Todd C, Pardy B et ah Percutaneous angioplasty for
infrainguinal graft-related stenoses. Eur J Vase Endovasc Surg 1997;
14:380.

46. Goh RH, Sniderman KW, Kalman PG. Long-term follow-up of
management of failing in situ saphenous vein bypass grafts using
endovascular intervention techniques. / Vase Interv Radiol 2000;
11:705.

47. Hoksbergen AW, Legemate DA, Reekers JA et ah Percutaneous
transluminal angioplasty of peripheral bypass stenoses. Cardio-
vasc Intervent Radiol 1999; 22:282.

48. Miller PE, Carlton D, Deierhoi MH et ah Natural history of arteri-
ovenous grafts in hemodialysis patients. Am } Kidney Dis 2000;
36:68.

49. Schwab SJ, Harrington JT, Singh A et ah Vascular access for
hemodialysis. Kidney Int 1999; 55:2078.

50. NKF-DOQI clinical practice guidelines for vascular access: Na-
tional Kidney Foundation: Dialysis Quality of Life Initiative. Am J
Kidney Dis 1997; 30:S150.

51. Lay JP, Ashleigh RJ, Tranconi L et ah Result of angioplasty of bres-
cia-cimino hemodialysis fistulae: medium-term follow-up. Clin
Radiol 1998; 53:608.

52. Hingorani A, Ascher E, Kallakuri S et ah Impact of reintervention
for failing upper-extremity arteriovenous autogenous access for
hemodialysis. / Vase Surg 2001; 34:1004.

53. Myers KA, Denton MJ. Infrainguinal atherectomy using the Auth
Rotablator: patency rates and clinical success for 36 procedures.
J Endovasc Surg 1995; 2:67.

54. Scheinert D, Laird JR Jr, Schroder M et ah Excimer laser-assisted re-
canalization of long, chronic superficial femoral artery occlusions.
J Endovasc Ther 2001; 8:156.

55. Belli AM, Cumberland DC, Knox AM et ah The complication rate
of peripheral balloon angioplasty. Clin Radiol 1990; 41:380.

56. Solomon R, Werner C, Mann D et ah Effects of saline, mannitol
and furosemide on acute decreases in renal function induced by
radiocontrast agents. N Engl J Med 1994; 331:1416.

57. Tepel M, van der Giet M, Schwartzfeld C et ah Prevention of
radiographic contrast-agent induced reduction in renal function
by acetylcysteine. N Engl} Med 2000; 343:180.

515

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Endovascular stents

Frank J. Criado
Youssef Rizk
Gregory S. Domer
Hilde Jerius

Ranking just below the breakthrough developments of
Seldinger's catheterization technique, Dotter's angioplasty,
and Gruntzig's balloon catheter, endoluminal stent tech-
nology represents undoubtedly one of the most significant
achievements in the endovascular field. Invented in the 1980s
and further refined in the 1990s, stent devices and stenting
techniques have revolutionized interventional capabilities,
vastly expanding the therapeutic reach of angioplasty (PTA).
In addition, and largely through serendipity, these devices
were also to become the critical components for a whole new
group of very important technologies —the "stent-grafts."

The emergence of endoluminal metallic stents was pro-
pelled mainly by two occurrences, both PTA-related. One was
the rapidly growing realization in the early 1980s that PTA
failures were relatively common and potentially catastrophic,
especially in the coronary circulation. Second, the eventual
elucidation of the mechanisms of angioplasty in the mid-
1980s, leading inevitably to the clear view that vessel "scaf-
folding" would be a necessary tool to avert complications and
improve results. 1 ' 2 Charles Dotter's 1964 premonitions res-
onated powerfully then; he had certainly envisioned (and
written about) the future need for a "splint" to support the ves-
sel lumen during the healing phase following recanalization. 3
The term "stent" did not come into (vascular) use until 1983; its
etymological origin is quite interesting, going back to Dr
Charles Stent, a dentist in mid-19th century London. 4

The need for "bail-out" (or "rescue") during or after angio-
plasty became increasingly obvious. PTA-related thrombosis,
with consequent vessel closure, began to be regarded as a
potentially preventable event through the implantation of an
endovascular device that could prop the vessel open and re-
model the lumen created by the balloon. This is all possible be-
cause stenting results in a relatively smooth and circular flow
channel and unimpeded flow without translesional pressure
gradients, thereby preventing or minimizing the risk of early
thrombosis. Such a "bail-out" role was the intended purpose
and actual raison d'etre for all stents at their inception. More-
over, to this day, "suboptimal angioplasty" remains the only
label indication for all vascular-approved stents. Be that as it

may, it did not take long for clinicians and investigators to see
beyond PTA rescue, understanding that stent devices can have
a much larger role and wider applicability. Worthy of mention
are treatment of arterial dissections (mostly PTA-related), pre-
vention and treatment of restenosis in some cases, resurfacing
of the inner wall, and serving as the skeleton for endoluminal
conduits when combined with fabric covers ("covered stents"
and "stent-grafts").

Numerous stent devices are currently available for possible
use in the peripheral vascular system. It is noteworthy that
only a handful of these have received Food and Drug Admin-
istration (FDA) approval for a vascular (arterial) indication
(Table 44.1). Stents can be classified into two seemingly
distinct categories according with the mode of deployment:
balloon-expandable (Fig. 44.1) and self -expandable (Fig. 44.2). As
a result of Palmaz's early pioneering work, balloon-expand-
able devices received the most attention, predominating dur-
ing the first several years of clinical application. 5 ' 6 They
continue to be popular in the hands of many interventionists
in several vascular beds, mainly iliac and renal arteries. Self-
expandable stents had their beginning early on as well with
the Wallstent device. More recently, this group of stents has
gone through a phase of explosive growth with the advent of
nitinol as the preferred compositional metal. 7 While deploy-
ment characteristics are obviously important, it is design and
composition that represent the signature features of a stent de-
vice. Both conceptually and descriptively, slotted-tube stents
can be distinguished from coil-design devices. The former are
laser-cut from a solid cylindrical tube, and tend to have longi-
tudinal rigidity and high radial strength. The latter are coil or
wire-mesh designs that tend to be more flexible and exert less
radial force on the vessel wall. Such distinctions, however, are
becoming increasingly blurred as newly designed devices
combine features of both groups— this is especially true in the
case of slotted-tube nitinol stents that have excellent flexibility,
good radial strength, and deploy rather precisely.

Stent devices have become critical tools in the intervention-
ist's armamentarium to treat a host of stenotic and occlusive le-
sions in the arterial and venous systems. Intriguingly though,

516

Table 44.1 Vascular stents

chapter 44 Endovascular stents

Deployment mode/type

Manufacturer

Device

Composition

FDA approval

Self-expandable

Bard

Luminexx

Nitinol

Biliary

Conformexx

Nitinol

Biliary

Boston Scientific

Monorail
Wallstent

Elgiloy

Biliary

Wallstent

Elgiloy

Biliary/iliac

Cook

Zilver

Nitinol

Biliary

Cordis

Precise

Nitinol

Biliary

Smart Control

Nitinol

Biliary

ev3

Protege

Nitinol

Biliary

Intracoil

Nitinol

SFA

Guidant

Dynalink

Nitinol

Biliary

Medtronic

Bridge SE

Nitinol

Biliary

Balloon expandables

Cordis

Palmaz

Steel

Iliac, renal, biliary

Palmaz-Genesis

Steel

Biliary

Medtronic AVE

Bridge X3

Steel

Biliary

Bridge Assurant

Steel

Biliary

Boston Scientific

Express Biliary LD

Steel

Biliary

Ni royal

Steel

Biliary

Guidant

RXHerculinkPlus

Steel

Biliary

Omnilink

Steel

Biliary

ev3

Intrastent

Steel

Biliary

IntrastentLP

Steel

Biliary

Intrastent Doublestrut

Steel

Biliary

Intrastent

Steel

Biliary

Doublestrut XS

Intrastent

Steel

Biliary

Doublestrut LD

Intrastent Paramount

Steel

Biliary

Intrastent Paramount XS

Steel

Biliary

Intrastent Mega LD

Steel

Biliary

Intrastent Max LD

Steel

Biliary

Angiodynamics

Omniflex

Platinum/iridium

Biliary

Covered stents

Gore

Viabahn

Nitinol/ePTFE

Tracheo-bronchial

Viabil

Nitinol/ePTFE

Biliary

Viatorr

Nitinol/ePTFE

Biliary

Jomed

Graft Master

Steel/ePTFE

Coronary vein grafts

Vascular Architects

Aspire

Nitinol/ePTFE

SFA

Boston Scientific

Wallgraft

Elgiloy/ePTFE

Tracheo-bronchial

Vascular indication

Figure 44.1 Genesis (Cordis Endovascular, Warren, NJ, USA) balloon-expandable stent.

517

pa rt v Endovascular interventions for vascular disease

Figure 44.2 Deployment of a self-expandable nitinol stent — SMART
(Cordis Endovascular, Warren, NJ, USA).

Table 44.2 Advantages of stenting

Predictable
Simple
Effective
Expeditious

Table 44.3 Disadvantages of stenting

Implant

Cost

Potential complications

there is little if any scientific proof that stent placement is better
than PTA alone. A majority of physicians doing intervention
(ourselves included) would agree that stenting offers
"obvious" practical advantages over PTA (Table 44.2). The
disadvantages (Table 44.3) are often dismissed as irrelevant or
practically unimportant— with the exception of cost. Results
of the Dutch Iliac Stent Trial (published in 1998) are currently
regarded as the most valid and accurate representation of the
precise role of stenting, not just in the iliac arteries, but in many
other vascular beds as well. 8 That is, there is no scientific proof
of benefit when stents are placed primarily (or "routinely") at
the time of angioplasty. Selective stenting for suboptimal
angioplasty is evidence-based and should be adopted as the
standard of practice. Nonetheless, "routine stenting" contin-
ues unabated, beyond science and rationale. It is driven by

various reasons, including interventionists' background and
training, practicality (as discussed), and the near-universal
perception that stent placement is safe and effective. The al-
leged theoretical superiority over PTA alone is based on the
facts that stent placement results in lumen remodeling, pre-
vention of elastic recoil, and the achievement of a larger
lumen. 2

The following is a brief summary of current views and
strategies as they relate to the clinical use of vascular stents
during catheter-based treatment of arterial stenosis and
occlusion:

• Carotid arteries: use of stents (stent-supported angioplasty)
is felt to be mandatory by everyone. Self-expanding stents
are the universal choice at present, nitinol devices in
particular.

• Supra-aortic trunks (innominate, proximal common
carotid, and subclavian arteries): with few exceptions, inter-
ventional specialists use routine stenting during proce-
dures to treat stenoses and occlusions; for the latter in
particular. Opinions vary in terms of the choice between
balloon-expandable and self-expandable nitinol devices.

• Renal and visceral arteries (celiac and superior mesenteric
arteries): most lesions treated by percutaneous intervention
are ostial in location. Stenting is used universally. Balloon-
expandable devices are preferred by almost everyone
(Fig. 44.3).

• Iliac arteries: "purists" use PTA alone, reserving stents for
suboptimal angioplasty. They are in good company: solid
scientific evidence and FDA labeling! But the majority of us
use stents nearly routinely because of the reasons outlined
in Table 44.1. Opinions are divided regarding the choice of
device type (balloon-expandable vs. self-expandable).

• Superficial femoral artery (SFA): this has been a notoriously
difficult area for stents, especially for the first-generation
devices (Palmaz and Wallstent). Present-day nitinol devices
would appear to do much better, but definitive proof will
not be forthcoming for some time. Current trends point in
the direction of more frequent use of self-expanding nitinol
stents during SFA intervention for treatment of stenotic and
occlusive lesions.

• Tibial-peroneal arteries: role and results of stenting in the
infrapopliteal vascular territory are virtually unknown as
there is a lack of meaningful information with the exception
of multiple anecdotal reports and small personal series.
Currently, stent usage is often reserved for PTA rescue only,
with coronary devices being the obvious (and only) choice
for these vessels. Technological developments and clinical
trials in the near future can be anticipated.

Vascular stents: failure modes

Stent failure modes are closely related to the type of implanted
vessel and final expanded diameter. 9 Large-diameter stents

518

chapter 44 Endovascular stents

Figure 44.3 Endovascular treatment of severe
stenosis of proximal superior mesenteric artery:
placement of balloon-expandable stent, left
brachial artery access.

(i.e. aortic stent grafts) are exposed to corrosion which may
lead to loss of structural integrity. Small and medium-sized
devices, on the other hand, are likely to incite a hyperplastic re-
sponse from the vessel wall, with consequent development of
in-stent restenosis. The latter constitutes the most significant
unresolved issue with stent technology and clinical stenting,
both in the coronary and peripheral vascular systems.
Restenosis can be defined as the loss of the luminal diameter
gain achieved at the time of stent placement, caused by neoin-
timal hyperplasia. It tends to occur within 1-3 months after
intervention, but may continue to progress up to 18 months af-
terwards. 9 It has been shown that both the depth of wall
penetration of stent struts and the degree of initial luminal
gain correlate with the magnitude and likelihood of late stent
lumen loss. In other words, and not unexpectedly, there is a
close relationship between the extent of wall (cellular) injury
and neointimal proliferation. These facts point to the possible
counterproductive nature of excessive oversizing at the time
of balloon dilation and stent deployment.

Drug-eluting stents (DES) have been developed in an effort
to minimize or prevent the occurrence of neointimal hyper-
plasia and restenosis. Results so far available from coronary
interventional trials are very promising; 10 similar benefit in
noncoronary vascular beds is hoped for but completely un-
proven at this time. Efforts are also being made with use of
brachytherapy and other adjuncts that may have an impact on
outcome.

The future of vascular stent technology is clearly bright,
even in the face of current unresolved problems related to
hyperplastic restenosis. Perfecting and treating stent metal
surfaces, removing impurities, and improving profile and de-
liverability are just a few aspects among many other areas
where technological progress is sure to occur.

References

1. Criado FJ. Principles of balloon angioplasty. In: Criado FJ, ed.
Endovascular Intervention: Basic Concepts and Techniques. Armonk,
NY: Futura Publishing Co., Inc., 1999.

2. Criado FJ. Vascular stents: basic concepts and designs. In: Criado
FJ, ed. Endovascular Intervention: Basic Concepts and Techniques.
Armonk, NY: Futura Publishing Co., Inc., 1999.

3. Dotter CT, Judkins MP Transluminal treatment of arteriosclerotic
obstruction: description of a new technique and a preliminary
report of its applications. Circulation 1964; 30:654.

4. Sterioff S. Historical vignette: etymology of the word "stent."
Mayo Clin Proc 1997; 72:377.

5. Palmaz JC, Sibbitt RR, Reuter SR et al. Expandable intraluminal
graft: a preliminary study. Radiology 1985; 156:73.

6. Palmaz JC, Laborde JC, Rivera FJ et al. Stenting of the iliac arteries
with the Palmaz stent: experience from a multicenter trial. Cardio-
vasc Intervent Radiol 1992; 15:291.

7. Criado FJ. New developments in nitinol stents. In: Criado FJ, ed.
Endovascular Intervention: Basic Concepts and Techniques. Armonk,
NY: Futura Publishing Co., Inc., 1999.

8. Tetteroo E, van der Graaf Y, Bosch JL et al. Randomised compari-
son of primary stent placement versus primary angioplasty fol-
lowed by selective stent placement in patients with iliac-artery
occlusive disease. Lancet 1998; 351:1153.

9. Palmaz JC, Bailey S, Marton D et al. Influence of stent design and
material composition on procedure outcome. / Vase Surg 2002;
36:1031.

10. Morice MC, Serruys PW, Sousa JE et al. A randomized comparison
of a sirolimus-eluting stent with a standard stent for coronary
revascularization. N Engl J Med 2002; 346:1773.

519

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Endovascular prostheses for repair of
abdominal aortic aneurysms

Carlos E. Donayre

The field of vascular surgery entered an era of change with the
introduction of stent-grafts for the repair of abdominal aortic
aneurysms (AAAs) by Dr. Parodi in 1989. 1 This minimally in-
vasive, catheter-based therapeutic modality was initially ap-
plied to high-risk patients with surprising results. However,
the first endograft repair in the United States was not reported
until years later in 1995 by the group at Montefiore. 2 Endovas-
cular aortic aneurysm repair (EVAR) did not reach center stage
until 1999 when the Food and Drug Administration (FDA)
granted approval to two industry-made devices for clinical
use. Two more industry-made devices have received approval
in the last 2 years, with several devices in the midst of receiving
FDA approval. It has been reported that 50% of all elective
AAAs are now being treated with endografts. 3

The midterm results with 4-6 years of follow-up have
demonstrated efficacy in the prevention of aneurysm rupture
and death from rupture, and equivalent rates of long-term sur-
vival, when compared with open surgical repair. However,
problems have also been encountered such as endoleaks, rup-
ture, migration, fabric tear, and stent fracture. Patients treated
with endografts have had to submit to secondary interven-
tions and extensive and life-long follow-up. Thus, this is a
good time to analyze each of the devices that have received
FDA approval in the United States since 1999 and compare
how EVAR stands against open AAA repair.

Ancure endograft/Guidant

The first industry-produced aortic endograft in the United
States was manufactured by Endovascular Technologies
(EVT) (subsequently acquired by Guidant, Temecular, CA)
based on the Harrison Lazurus patent and concept of securing
an endograft to the aortic wall with the use of an attachment
system consisting of hooks protruding from stents. This initial
device with a tube configuration was successfully deployed
by Dr. Wesley Moore at UCLA Medical Center on 10 February
1993. 4 It was quickly recognized that only a small percentage
of patients with AAAs would be suitable for tube graft repair

owing to the lack of a long distal aortic neck required to
achieve a secure fixation.

The above limitation led to the development of a bifurcated,
partially supported, unibody system by EVT which relocated
the distal fixation site to the iliac arteries. Since a distal aortic
neck was no longer required, EVAR was made available to a
greater number of patients with AAAs. 5

The unique characteristics of this device are its one-piece de-
sign which eliminates the risk of component separation. The
graft material is a woven polyester fabric with crimping ap-
plied to the graft limbs. The graft is unsupported with the ex-
ception of the attachment systems which are present at both
proximal and distal ends. The attachment system relies on ac-
tive fixation provided by self-expanding stents with a series of
incorporated hooks that engage the aortic wall proximally and
the iliac arteries distally (Fig. 45.1). The hook design was de-
veloped to insure longitudinal stability and prevent graft mi-
gration. However, the discovery of pin and /or attachment
system fractures in a few patients led to the voluntary temp-
orary cessation of the tube and bifurcated FDA trials while this
issue could be addressed and rectified. The implant program,
which resumed in 1995 with a newly redesigned attachment
system, avoided the weaknesses that led to the pin fractures.
By 1999, a sufficient number of patients treated with tube
and bifurcated endografts had been accumulated to receive
approval by the FDA (Fig. 45.2). The experimental group
presented to the FDA consisted of 573 patients selected for
placement of an EVT/Guidant bifurcated graft between
22 November 1995 and 12 February 1998 and 111 patients who
were initially acquired when tube endografts were being used
but were not suitable to be treated with that design.

Periprocedural outcome

Endovascular (n = 573) and open (n = 111) groups had similar
ages (72.8 vs. 71.6 years), and similar aneurysm diameters
(50-69 mm; endovascular 60.1 mm/open 61.8mm). A gender
difference was encountered with a male preponderance of
91.5% for the endovascular group and 76.6% for the open

520

chapter 45 Endovascular prostheses for repair of abdominal aortic aneurysms

Figure 45.1 Proximal self-expanding stent with a series of incorporated
hooks designed to actively engage the aortic wall.

Figure 45.2 Food and Drug Administration-approved devices (left to right):
tube graft/unibody bifurcated graft/aorto-uni-iliac configuration.

group (P < 0.001). Of the 573 patients in whom an attempt was
made to place a bifurcated graft, 531 (92.7%) underwent suc-
cessful implantation. The median operating time for implan-
tation was 190 min, and this exceeded the median operating
time of 157 min for open repair in the control group (P < 0.001).
However, the median blood loss in the endovascular group
was 400 mm, one-half of the median blood loss in the open re-

pair group (800 ml; P < 0.001). Shorter length of stay in the in-
tensive care unit (ICU) (24 vs. 27 h) and hospital (2 vs. 6 days)
was also statistically significantly in favor of the endovascular
group, with only 34% of the EVAR patients requiring an ICU
stay compared with 96% of the open group. There was a trend
towards reduced 30-day mortality for the endovascular group
when compared with the open group (1.7 vs. 2.7%). When
morbidity was considered, a major difference was also en-
countered with a complication rate of 28.8% for the EVAR
group and 44.1% for the control group (P = 0.02). Forty- two pa-
tients (7.3%) had to be converted to an open repair early on
(<30 days), in 12 patients (2.1%) the prostheses could not be de-
livered owing to access limitations, and in 30 (5.2%) owing to
failure of accurate placement. The inability to deliver the
device can be attributed to the relatively large profile of this
device (23 Fr in diameter with a delivery sheath 24 Fr in
diameter). This also accounts for the decreased number of fe-
male patients in the EVAR group as they tend to have smaller
vessel diameters.

One-year outcome

No AAA ruptures were encountered at 1 year, and only two
patients demonstrated graft migration. Only two patients re-
quired a late (>30 days) open conversion due to persistent en-
doleak and aneurysm sac enlargement. The unsupported iliac
limbs were problematic, 216 patients (40.6%) requiring inter-
vention because of compromised limb flow. In 31.7% interven-
tions were performed at the initial procedure (stents in 19%,
balloon dilatation in 8%), with 10.3% requiring postprocedure
interventions to treat symptoms (Fig. 45.3). Seventeen of 573
patients with functioning endografts suffered a limb thrombo-
sis (97% limb patency). At time of discharge 37% of patients
had any time of perigraft flow, but this decreased to 28% at
1 year. However, the incidence of type I endoleak was 5.6% at
discharge, dropping to 1.7% at 1 year. 6

Five-year outcome

Patients selected for long-term follow-up included 319 who re-
ceived the bifurcated graft and the 111 who underwent open
repair, in accordance with an agreement between Guidant and
the FDA. During the 60-month follow-up no patient suffered
an aneurysm rupture. The aneurysm sac size had decreased
>5 mm in 53% of the above patients at 1 year, with 78% seeing
such a decrease at 5 years. Only 2% of patients experienced an
increase in sac size during the 5-year follow-up period. Nine
patients (2.8%) underwent conversion to open repair. Indica-
tions for conversion were persistent perigraft flow in three,
endograft infection in two, limb thrombosis in two, AAA en-
largement without evidence of endoleak ("endotension") in
one, and one case of graft migration. No difference was seen in
survival with 70% of the EVAR group and 76% of the open
group alive at 5 years.

521

pa rt v Endovascular interventions for vascular disease

Figure 45.3 (A) Abdominal aortic aneurysm
5.8 cm in diameter with ectatic iliac arteries
and patent IMA. (B) One month after
operation. Aneurysm sac thrombosis with
secure aortic neck fixation and seal achieved in
proximal common iliac arteries bilaterally. (C)
Two years. Persistent type llendoleakfrom
IMA to lumbarcoil embolized. Subsequent
computed tomography scans demonstrated
that the leak persisted with flow appearing to
originate nearthe bifurcation of the
prosthesis, and potentially coming from one of
the iliac limbs. An AneuRx 1 6-mm iliac limb
was added on the right and a 22-mm aortic
cuff on the left achieving a secure distal
fixation. (D) Fouryears. Sac thrombosis with
no evidence of endoleaks.

522

chapter 45 Endovascular prostheses for repair of abdominal aortic aneurysms

Device withdrawal

Despite the excellent early benefits and 5-year outcome similar
to that with open repair, Guidant had to pay $92.4 million to
settle criminal and civil charges. FDA charged that EVT, a sub-
sidiary of Guidant based in Menlo Park, failed to file 2628 re-
ports from 1999 to 2001 describing incidents in which insertion
and deployment of the Ancure Endograf t system may have led
to injuries and death. 7 These incidents were related to prob-
lems in delivering this relatively large profile device through
small or atherosclerotic challenged vessels. The FDA felt that
risk of injury was highest at the time of insertion of the device
and that the 18 000 patients with implanted devices were not at
risk since the long-term performance of the device was not
in question. Nonetheless, Guidant withdrew the Ancure
Endograf t from the market in October 2003.

AneuRx/Medtronic

Device design

The AneuRx stent-graft is a modular bifurcated system com-
posed of a thin-walled, noncrimped, woven polyester graft
supported with a nickel-titanium alloy (nitinol) exoskeleton.
Nitinol has a unique thermal memory, expanding and regain-
ing its original shape when exposed to a warm blood medium.
The nitinol exoskeleton provides passive fixation reliant on
stent oversizing, radial hoop strength, and columnar support.
The primary components are a main bifurcated body with an
integral iliac limb and a separate contralateral iliac limb. Addi-
tional modular components include aortic and iliac extender
cuffs. 8

Since its conception the AneuRx stent-graft has continued
to undergo a slow and methodical evolution. A proximal su-
perstructure composed of a 5-cm long stent and a contralateral
gate with only 2 cm of overlap (deployed in 174 patients) was
replaced with a more flexible proximal configuration of five 1-
cm stents, a longer contralateral gate (4 cm long) and a Dacron
graft with a denser weave (deployed in 1019 patients)
(Fig. 45.4). Following the completion of the clinical trials a
third-generation device, the Xpedient, incorporated a tapered
nosecone with an improved delivery system and a lubricious
coating of the graft cover to reduce friction during nosecone
and runner retrieval. The tapered tip design replaced a stain-
less steel bullet nose to improve ease of device advancement
during delivery. The lubricious coating was added to improve
runner and bullet retraction, a maneuver that on occasions led
to device pull-down, loss of proximal fixation, and require-
ment for the additional deployment of a proximal aortic cuff. A
fourth-generation device with a high-density fabric designed
to reduce porosity, increase abrasion resistance, and enhance
durability is to be introduced in the summer of 2004.

Figure 45.4 AneuRx components: Main body composed of Dacron graft
with nitinol exoskeleton of five 1 -cm nitinol stents proximally, integral iliac
limb, and contralateral gate (4 cm long). Iliac limb extender composed of six
1 -cm nitinol stents. Aortic cuff extender that can be added either proximally
in the aortic location or distally in the iliac location to achieve a secure fixation
and maintain hypogastric patency.

523

pa rt v Endovascular interventions for vascular disease

Clinical trials

The AneuRx stent-graft was introduced in three phases at 19
investigational centers in the United States from 1996 to 1999.
The phase I study consisted of 40 patients with the first de-
ployment taking place at Harbor/UCLA in June 1996. The
phase II study included 424 patients treated with an AneuRx
stent-graft and 66 control patients treated with open repair
from 1997 to 1998. The phase III study included an additional
639 patients treated with EVAR from 1998 to 1999. A cohort
of 90 patients who did not meet inclusion criteria also were
treated but entered into a high-risk study arm.

Phase I

This feasibility trial included 40 patients with infrarenal AAAs
(5.7 ± 0.8 cm) treated at four study centers. An AneuRx
stent-graft was implanted in all patients with successful ex-
clusion of the aneurysm. There were no endoleaks and no sur-
gical conversions. Three patients died (7.5%) in the 30-day
perioperative period, two from chronic obstructive pul-
monary disease and respiratory failure and one from sepsis as
a result of a gangrenous gallbladder. There were four major
complications (10%): iliac limb thrombosis in two patients,
focal stenosis at the site of insertion in one patient, and one pa-
tient had a cardiac arrhythmia. There were no device-related
deaths, no aneurysm ruptures, or late conversions in this
group.

Phase II

Four hundred and sixteen patients receiving a stent-graft
were compared with 66 patients undergoing open repair.
Successful graft deployment was achieved in 98% of
patients, with surgical conversion required in 1.5%. Operative
mortality at 30 days did not differ, 2% in the stent group and
0% in the open group. Once again a 50% reduction in morbidi-
ty was encountered when EVAR was compared with open
AAA repair; a 66% reduction in blood loss and a 63% reduction
in hospital stay were also achieved. Endoleak rate at time of
hospital discharge was 38%, which was reduced to 13% at
1 month. The main cause of endoleak was transgraf t flow seen
at initial time of deployment. Five percent of patients required
a secondary endovascular procedure for endoleak, and 2% for
iliac limb thrombosis. 9

Midterm results

Analysis of patients 3 years after the end of enrollment yields
data for 75% patients followed for at least 2 years, 3-year data
for 43%, and 4-year data for 24%. Intraoperative rupture oc-
curred in two patients at time of initial implantation, three suf-
fered ruptured within 30 days of implantation, and 10 patients
had their aneurysms rupture late (>30 days). Overall rupture

mortality was 60% in this cohort of patients. A total of 53 pa-
tients required surgical conversion to open repair, with 11 in-
traoperative conversions, four within 30 days of implantation,
and 38 late conversions (>30 days). The most common causes
of surgical conversion were endoleak with AAA enlargement
in 18 (34%), rupture in 11 (21%), migration or displacement of
modular component in 11 (21%), failure to access in 13%, and
sac enlargement in two (4%). Endoleak was present in 13% of
patients, AAA sac enlargement was present in 14%, and mi-
gration noted in 9% (Fig. 45.5). At 4 years, freedom from
aneurysm rupture was 98.4%, freedom from surgical conver-
sion was 90.4%, and a 62.4% survival rate was seen. 10

Careful review of these complications identified stent-graft
fixation to the infrarenal aortic neck and iliac arteries, and
junction gate overlap to be more important than endoleak as
the primary cause of failure. This led to revision of morpholog-
ical AAA requirement, 15-mm aortic seal zone, 25-mm iliac
seal zones, and aortic neck angulation <45°. Device selection
required 10-20% oversizing and appropriate and timely pa-
tient follow-up was essential. Results of the AneuRx US clini-
cal trials and worldwide commercial experience as well as the
introduction of device modifications ensure that this user-
friendly device continues to be a safe and effective option for
properly selected patients with AAAs.

Excluder/Gore

Device design

The Excluder endograft is made of expanded polytetrafluoro-
ethylene (e-PTFE) graft material bonded to the inside of a
nitinol exoskeleton and enclosed in a composite film. Angled
wire barbs are located at the proximal end of the main device to
provide additional active fixation to the aortic wall. A radio-
opaque ring marks the contralateral leg opening. This modu-
lar system is composed of one trunk ipsilateral piece and one
contralateral leg piece (Fig. 45.6). Both components have an
attached sleeve made of e-PTFE that is sewn closed around
the prostheses and functions to constrain it. Cuffs designed for
the aorta and iliac arteries are also available. An 18-Fr sheath is
required to deliver the main body piece of this lower profile
device, with only a 12-Fr sheath required for delivery of
the contralateral leg piece. A deployment line is attached to the
e-PTFE sleeve; when pulled the sleeve is released allowing
rapid deployment of the prostheses. 11

Periprocedural outcome

Both endovascular (n = 235) and open (n = 99) groups had sim-
ilar ages (73.0 vs. 70.1 years), and similar aneurysm diameters
(endovascular 55.6 mm/open 58.6mm). A gender difference
was encountered, with a male preponderance of 87% for the
endovascular group and 74% for the open group (P = 0.004).

524

Figure 45.5 AneuRx device 8 years postdeployment. Lateral view
demonstrates abdominal aortic aneurysm (AAA) regression marked by
calcifications parallel to device. (A) Computed tomography (CT) scan of
proximal neck with device migration away from aortic wall as result of neck
dilation and/or elongation. (B) CT scan distal aortic neck with secure
stent/wall apposition and no evidence of endoleak. (C) AAA sac regression

and collapse around iliac limbs. (D-l) Preop to 8 years. AAA regression
following deployment of AneuRx device through a span of 8years. Rigid
superstructure has resulted in device migrating away from renal arteries and
a change in neck configuration but due to AAA regression and lack of
endoleak no further interventions are required.

pa rt v Endovascular interventions for vascular disease

Figure 45.6 Excluder endograft is made of expanded
polytetrafluoroethylene (e-PTFE) graft material bonded to the inside of a
nitinol exoskeleton and enclosed in a composite film. Angled wire barbs are
located at the proximal end of the main device to provide additional active
fixation to the aortic wall. Aortic main body with ipsilateral iliac limb and
contralateral gate opening for docking of iliac limb.

Local or regional anesthesia was utilized in 40% of EVAR pa-
tients but in only 2% of open patients was this possible. All pa-
tients in the EVAR group had a successful device deployment,
33% required one or more extension. The median operating
time for implantation was 221 min, which was less than the
median operating time of 283 min for the open control group
(P < 0.001). The median blood loss in the endovascular group

was only 310 mm, over a liter less than the open repair 1590 ml
(P < 0.001). Shorter length of stay in the ICU (6 vs. 67h) and hos-
pital (2 vs. 9.8 days) was also statistically significantly in favor
of the endovascular group, with only 24% of the EVAR pa-
tients requiring an ICU stay compared with 87% of the open
group. There was no difference in 30-day mortality for the en-
dovascular group when compared with the open group (1.0
vs. 0%). When morbidity was analyzed, again a major differ-
ence was encountered. A major adverse event was seen in 14%
of the EVAR group and 57% of the control group (P < 0.001). No
aneurysm ruptures occurred postimplantation. Only three
patients required conversion to an open repair, all due to
aneurysm enlargement.

Two-year outcomes

Aneurysm reintervention in the EVAR group was necessary in
17 patients (7%) during the first year, and in 14 patients (7%)
during the second year, all but four of the reinterventions were
endovascular in nature with coil embolizations for type II en-
doleaks and sac growth performed in 25. Three of the reinter-
ventions were for major device-related complications. One
patient had immediate trunk migration after deployment that
was dealt with by an aortic extender, one patient had an iliore-
nal bypass for an occluded renal artery, and one patient re-
quired aortic cuff extenders for an increase in proximal neck
angulation and neck diameter enlargement. At 1 year 83% of
patients had no endoleak, and the number remained stable at
2 years with 80% of patients endoleak-free. Type I endoleak
was 1% at 1 year and 3% at 2 years. Type II endoleaks also re-
mained stable at 12% at 1 year and 13% at 2 years. The total en-
doleak rate was 17% at 1 year and 20% at 2 years. Trunk or limb
migration was only 1% at 2 years. 12

Sac enlargement

One of the most puzzling issues for the Excluder endovascular
graft is continuous aneurysm sac enlargement (>5mm). At

1 year this was seen in 7% of patients, increasing to 14% at

2 years, 23% at 3 years, and 32% at 4 years. No aneurysm rup-
tures have occurred, but all open conversions have occurred in
this cohort of patients. 13 Explant analysis has revealed that an
ultraplasma filtrate occurs in the areas in which the prosthesis
is not in contact with an aortic or iliac wall. It thus appears that
the formation of this material allows the transmission of pres-
sure with subsequent sac enlargement. The graft material has
been modified to be impervious to plasma and has already
been given FDA approval. No decision has been reached on
how to deal with patients who continue to experience sac en-
largement in the absence of endoleak. An option would be to
re line the endograft with new aortic cuffs and iliac limbs.

Despite the troublesome issue of sac enlargement, the Ex-
cluder stent-graft has been associated with a reduction in pro-

526

chapter 45 Endovascular prostheses for repair of abdominal aortic aneurysms

> :

) ,

«
M ■

•

Figure 45.7 (A-C) Preop to 1 year. Patient with abdominal aortic aneurysm and calcified atherosclerotic external iliac arteries, an ideal candidate for low-
profile Excluder device. Sac thrombosis with 5 mm sac diameter regression at 1 year.

cedure time, a favorable lower profile making percutaneous
delivery possible, and a striking reduction in patient recovery
time and complication rates (Fig. 45.7).

Zenith AAA endovascular graft/Cook

Device design

The Zenith endovascular graft has evolved from a worldwide
cooperative effort that began in Perth, Australia, under
the guidance of M. Lawrence-Brown and D. Hartley. The
implantable portion of this device has progressed from an
unsupported monoiliac configuration to a fully supported
modular bifurcated system. The main graft is a three-
component device (aortic main body and two iliac legs) com-
posed of woven Dacron fully stented with self-expanding
stainless steel Z-stents. An uncovered stent with staggered
barbs at the top of the graft provides active suprarenal
attachment. A variety of ancillary components (main body
extenders, iliac leg extenders, converters, and occluders)
are available to provide additional length or to convert
a bifurcated graft into an aorto-uni-iliac graft if necessary
(Fig.45.8). 14

Trial design

A total of 352 patients were enrolled prospectively at 15 centers
within the United States. 15 Three endovascular arms were
established. Patients considered candidates for open or en-
dovascular repair made up the standard-risk group (SRG). In-

dividuals at higher physiological risk, potentially unable to
tolerate conventional treatments, made up the high-risk group
(HRG). Finally, each center was allotted a number of patients
to treat before the accumulation of data within the pivotal
study, to gain comfort with the device and procedure. A total of
80 concurrent controls (CG) were enrolled with the intent
of contrasting the morbidity and mortality with the SRG
endovascular group.

A total of 351 patients underwent placement of a Zenith en-
dovascular prosthesis from January 2000 through July 2001.
The mean patient age was 71 years, and approximately 93%
were men. Aneurysms were confined to the infrarenal abdom-
inal aorta in 80% of the patients, whereas 20% had aneurysms
extending from the aorta into one of the common iliac arteries.
The mean aneurysm sizes for the SRG, CG, HRG, and roll-in
group were 56.2, 63.8, 575, and 58.1mm, respectively. All of
the surgical procedures were carried out under general anes-
thesia, whereas 53% of the endovascular grafts were placed
using epidural and 2% using local anesthesia. The median pro-
cedure duration was 140 min for the SRG and 210 min
(P < 0.001) for the CG. Estimated blood loss was greater for the
CG compared with the SRG (1676 ml and 299 ml, respectively,
P < 0.001). The median fluoroscopy time was 25 min. Selected
grafts were <30% oversize when compared with the aortic
neck diameter in ^%, in the remaining 12% the endografts
were oversized >30%.

The 30-day mortality was 0.5% in the SRG (n = 1) and 2.5% in
the CG (n=2). The 12-month mortality regardless of cause was
3.5% in SRG and 3.8% in the CG. AAA-related mortality at
30 days was 0.5% in SRG and 1.3% in CG. Significantly de-
creased morbidity was noted in the SRG at 30 days for cardiac

527

pa rt v Endovascular interventions for vascular disease

Figure 45.8 (A) Aorto/uni-iliac configuration with contralateral iliac
occluder. (B and C) Bifurcated configuration with three components (aortic
main body and two iliac legs) based on self-expanding stainless-steel Z-stents
with staggered barbs at the top of the graft to provide active suprarenal
attachment. Ancillary components (main body extenders, iliac leg extenders,
converters) provide additional length or can convert a bifurcated graft into an
aorto/uni-iliac configuration if necessary.

(P = 0.02), pulmonary (P < 0.001), renal (P = 0.01), and vascular
(P < 0.001) systems. Additionally, diminished blood loss, fewer
transfusion requirements, shorter hospital stay, decreased
ICU time, and faster return to daily activities were associated
with the endovascular procedure. There was no evidence of
deterioration of renal function over the course of the follow-
up. The observed renal infarcts (three in SRG, one in HRG, two
in CG, one in roll-in) were attributed to the coverage or ligation
of accessory renal arteries and occurred after device implanta-
tion or surgical repair.

There were no acute conversions in the study. There were
three late conversions during the first 12 months. The first was
performed for a persistent proximal endoleak, the second for
the development of a supraceliac aneurysm that expanded
rapidly, and the third during a thoracoabdominal repair with
resection of the proximal portion of the prosthesis and place-
ment of an interposition graft. There was one conversion in the
HRG, at 222 days postprocedure, as a result of AAA rupture.

This patient represents the single rupture in the pivotal trial.
The patient underwent successful surgical repair. At the time
of the procedure, the right iliac limb of the endograft was
noted to be within the aneurysm sac. The aneurysm had
contracted rapidly from 6 cm to 4.6 cm over a 6-month period.
Despite a secure proximal fixation, the iliac limb retracted into
the aneurysm, repressurizing the sac and causing the rupture.

The acute (30-day) endoleak rate was 17%. The majority of
these were type II leaks (9.5%), whereas 4.5% were classified as
type I endoleaks. The endoleak rate decreased to 7.4% at
12 months. There were considerably more secondary inter-
ventions in the SRG compared with the CG (11% and 2.5%;
P = 0.03). The treatment of endoleaks constituted the majority
of the secondary interventions (6%) for the SRG. The remain-
der of the secondary interventions was for compromised iliac
limb or arterial flow. The two secondary interventions in the
CG were acutely performed in an effort to control intraabdom-
inal hemorrhage.

At 12 months, 65% had significant sac shrinkage (> 5-mm sac
reduction), 34% remained unchanged, and sac growth was
noted in three patients (two SRG, one HRG; 1.5%). Two were
associated with graft infections and were converted electively.
The third was attributed to a distal endoleak.

The 30-day endoleak rate of 17% was largely composed of
type II leaks that spontaneously thrombosed. Investigators
were encouraged to treat all type I and III endoleaks, leaving a
7.4% incidence of endoleak at 12 months. Aneurysm size
changes were dramatic with this device. A fairly steep rate of
size change was noted. Nearly 70% of the patients in the SRG
experienced a minimum of 5 mm of diameter reduction
within a 12-month period. Only three patients suffered from
aneurysm enlargement, two in the setting of infected grafts
and one as a result of an untreated type I distal endoleak. The
inherent reassurance of decreasing aneurysm size must be
balanced with a careful assessment of follow-up radiographs
to detect potential component migration.

The Zenith device is designed to accommodate larger necks
and iliac arteries than the three commercially available
devices. Proximal necks larger than 28 mm were felt to be
potentially unstable, whereas the current Z-stent design may
provide an inadequate degree of radial force in an overly large
infrarenal neck. The addition of barbs to the suprarenal stent
was undertaken to diminish the incidence of migration. The
main body of the device is intentionally long and optimally de-
signed to place the ostium of the contralateral limb 15 mm
above the aortic bifurcation. This places the ipsilateral limb
in close proximity to the corresponding iliac artery and is
intended to diminish the risk of component separation,
limit the migration effect of blood hitting the bifurcation, and
facilitate limb cannulation.

The placement of the Zenith endovascular graft can be
accomplished with minimal morbidity and mortality. The
patients have relatively short hospital stays, minimal blood
loss, and return to normal function quickly. Graft oversizing

528

chapter 45 Endovascular prostheses for repair of abdominal aortic aneurysms

(>30%) was associated with an increased rate of device migra-
tion at 12 months and with a negative effect on AAA sac re-
gression. Endoleak occurrence and late aortic neck dilation
were impacted. Aneurysm size appears to decrease in the
majority of cases during the follow-up period, and rupture is
extremely rare.

References

1 . Parodi JC, Palmaz JC, Barone HD. Transf emoral intraluminal graft
implantation for abdominal aortic aneurysms. Ann Vase Surg
1991;5:491.

2. Parodi JC, Marin ML, Veith FJ. Transf emoral endovascular stented
graft repair of an abdominal aortic aneurysm. Arch Surg 1995;
130:549.

3. Lee WA, Carter JW, Upchurch G et al. Perioperative outcomes after
open and endovascular repair of intact abdominal aortic repair
aneurysms in the United States during 2001. / Vase Surg 2004;
39:491.

4. Moore WS, Vescera CL. Repair of abdominal aortic repair
aneurysm by transfemoral endovascular graft placement. Ann
Vase Surg 1994; 220:331.

5. Moore WS, Matsumura JS, Makaroun MS et ah for the
EVT/Guidant Investigators. Five-year interim comparison of
Guidant bifurcated endograft with open repair of abdominal
aortic aneurysm. / Vase Surg 2003; 38:46.

6. Moore WS. The Guidant Ancure Bifurcation Endograft: five-year
follow-up. Semin Vase Surg 2003; 16:139.

7. Iwata E. Class action suit coming after Guidant fined $92 million in
cover-up. US Today June 13,2003.

8. Zarins KZ, White RA, Schwarten D et al. for the AneuRx investiga-
tors. AneuRx stent graft versus open surgical repair of abdominal
aortic aneurysms: multicenter prospective trial. / Vase Surg 1999;
29:292.

9. Zarins KZ, White RA, Moll F et al. The AneuRx stent graft: four-
year results and worldwide experience. / Vase Surg 2001; 33:S135.

10. Zarins KZ, for the AneuRx investigators. The US AneuRx Clinical
Trial: 6-year clinical update 2002. / Vase Surg 2002; 37:904.

11 . Bush RL, Najibi S, Lin P et al. Early experience with the bifurcated
Excluder endoprosthesis for treatment of the abdominal aortic
aneurysm. / Vase Surg 2001; 33:497.

12. Matsumura JS, Brewster DC, Mkaroun MS et al. A multicenter
controlled clinical trial of open versus endovascular treatment of
abdominal aortic aneurysm. / Vase Surg 2002; 37:262.

13. Matsumura JS, Brewster DC, Makaroun MS. Mid-term results of a
controlled trial of open versus endovascular treatment of AAA.
S VS Annual Meeting June 6, 2004. Anaheim, California.

14. Van Schie G, Sieunarine K, Lawrence-Brown M et al. The Perth
bifurcated endovascular graft for infrarenal aortic aneurysms.
Semin Interv Radiol 1998; 15:63.

15. Sternbergh WC, Money SR, Greenberg RK et al. for the Zenith In-
vestigators. Influence of endograft oversizing on device migra-
tion, endoleak, aneurysm shrinkage, and aortic neck dilatation:
Results from the Zenith multicenter trial. / Vase Surg 2004; 39:20.

529

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Comparison of conventional
vascular reconstruction and
endovascular techniques

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Surgical and endovascular treatment of
chronic ischemia of the lower limbs

Jean-Paul P. M. de Vries

Frans L. Moll

Jos C. van den Berg

Atherosclerosis is the most common underlying cause of
chronic limb ischemia. It is primarily a disease of the intima
and inner media of the artery. The outer media and adventitia
are usually spared. 1 The pathogenesis of atherosclerosis is still
debated, but its origin lies in a combination of intimal injury
and the accumulation of low-density lipoproteins, cholesterol,
and vessel wall enzymes, in particular proteoglycan. 2 ' 3 Hem-
orrhage into the plaque, or continuing lipid accumulation,
may lead to occlusive plaques. Coexistence of the well known
vascular risk factors such as hypertension, diabetes mellitus,
hyperhomocysteinemia, and the use of tobacco clearly accel-
erates the above-mentioned pathology. 4-6 Although athero-
sclerosis is a generalized disease, the most commonly affected
segments are major arterial bifurcations and the segments
with severe angulation or posterior fixation. 7 Infraguinally,
the superficial femoral and the popliteal arteries are most
prone to atherosclerosis. The transitional zone between the
femoral superficial and the popliteal vessels is especially pre-
disposed to atherosclerosis 8 because of its fixation and oblique
passage in the adductor canal and the offspring of the large
superior genicular branch.

Clinical symptoms due to severe peripheral atherosclerosis
vary from intermittent claudication to rest pain or even tissue
loss. Commonly, ischemic pain or tissue loss is distributed
distal to the stenotic arterial segment(s). Up to 5% of people
60 years of age and older (women less than men) suffer from
intermittent claudication. Intermittent claudication resulting
from superficial femoral artery (SFA) occlusive disease is rela-
tively benign. 9 Only a quarter of people deteriorate to a higher
Fontaine class (III or IV) or incapacitating symptoms (Fontaine
class IIB) and require intervention. The ultimate amputation
rate of the claudicants is limited to 1% per year. 10

In recent decades, modern vascular surgery has made re-
markable progress in the management of chronic ischemia of
the lower extremities. With the introduction of digitized vas-
cular imaging meticulous visualization of the vascular tree has
become routine, and in the majority of patients atherosclerot-
ic-induced disability can be identified. In the wide range of
treatment modalities there has been a change from invasive

bypass operations to more refined techniques such as en-
darterectomy and percutaneous dilation of arterial stenoses.

This chapter reviews the current, state-of-the-art treatment
of chronic lower limb ischemia. Apart from thrombolysis, only
interventional techniques will be discussed for the supra-
genicular and infragenicular arterial segment. The prevention
and conservative treatment of lower extremity sclerotic dis-
ease is beyond our scope, as is acute ischemia due to arterial
thrombosis or embolism. The authors hope that this chapter
will assist the decision-making process of those surgeons or
radiologists who have to deal with lower limb ischemia.

Treatment modalities for
the femoropopliteal segment

Percutaneous techniques

Since the first reports on angioplasty in the 1970s, many
studies have been published about its implementation for
femoropopliteal sclerotic disease. The outcomes of most of
these studies are difficult to relate because of the lack of
standardization. Many variables (e.g. stenotic vs. occluded
segments, differences in length of stenosis, noncomparable
patient groups) greatly influence the primary and secondary
patency. The outcome of treatment can be assessed by sympto-
matic, hemodynamic, and anatomic results. Golledge and
coworkers 11 showed that success by one criterion does not
always predict success for the other two. For these reasons,
we will refer to well defined, prospective studies on
femoropopliteal angioplasty.

Procedure

For diagnostic angiography, arterial access should be con-
tralateral to the symptomatic leg and preferably via the com-
mon femoral artery (CFA). In general, these procedures can be
performed using local anesthetic. Access for percutaneous
transluminal angioplasty (PTA) depends on the location and

533

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

nature of the lesion. A contralateral approach is advocated for
common femoral and proximal deep and superficial femoral
artery stenotic lesions. Occlusions demand direct ipsilateral
femoral puncture, a technique not exempt from complications
or even mortality. 12 ' 13 Most complications are related to the
puncture site (e.g. bleeding, false aneurysm) with an incidence
of 4%. The incidence of distal vessel dissection or embolization
is 2.7%, whereas thrombus formation at the angioplasty site is
seen in 3.5% of PTAs. A retrograde puncture of the ipsilateral
popliteal artery is seldom necessary to perform a successful
PTAoftheSFA.

After the introduction of a 4- to 6-Fr catheter and successful
crossing of the stenosis or occlusion, a balloon angioplasty
catheter is introduced. The lesion is dilated with a balloon that
matches the measured size of the normal arterial segment
above and below the lesion or a deliberately oversized balloon
(up to 10-20%). Usage of low-compliant balloons offers more
dilating force at the sclerotic lesion, combined with a more pre-
dictable diameter and profile retention. The duration of the
balloon inflation should range from 30 to 120 s, whereupon it is
deflated rapidly. The purpose of PTA is to fracture the plaque,
which allows the vessel to dilate (both intima and media), with
permanent loss of elastic fibers. 14

To minimize the occurrence of embolism or thrombosis, a
bolus of heparin (3000-5000 IE) is given through the catheter
before PTA. Because of the PTA-induced vessel wall damage,
patients should be prescribed aspirin (acetylsalicylic acid) for
at least 3 months.

Results

As mentioned above, when dealing with the results of PTA of
the femoropopliteal segment, comparisons are difficult be-
cause of variation in the definitions of success and patency.
Most present-day studies now report clinical success accord-
ing to the grades of clinical response to PTA as proposed by
Rutherford and Becker. 15 The most reliable noninvasive tests
to predict a favorable outcome of angioplasty are the ante-
brachial pressure index (ABPI) and the velocity ratio (VR)
measured by duplex ultrasound. Golledge et al. 11 demonstrat-
ed in their prospective study that patients with an ABPI >0.9
within 24 h after PTA had a significantly lower 1-year resteno-
sis rate than those patients with an ABPI <0.9 (24% vs. 64%). A
VR of 2.5 equates to a hemodynamically significant angio-
graphic stenosis of approximately 60%, and velocity rates
greater than 3 are a predictor of progression of arterial stenosis
to occlusion. 16 In the majority of studies the primary technical
success rate varies from 70% up to 90%. 11/17/18 In Fig. 46.1, a suc-
cessful recanalization of the SFA is shown. The overall cumu-
lative 5-year primary patency rate ranges from 45% to 60%. 17/19
The most significant factor negatively affecting the outcome of
PTA of the SFA is the length of the stenotic or occluded seg-
ment. Lesions > 10 cm have a significantly worse patency than

Figure 46.1 Severe, localized stenosis of the SFA (A) successfully treated
with percutaneoustransluminal angioplasty (B).

shorter lesions, with a disappointing 6-month patency rate of
only 23%. 17/20 From the studies reviewed, other unfavorable
factors on outcome were occlusion instead of stenosis, extent
of outflow disease, extent of Fontaine classification, and the
presence of diabetes mellitus.

To overcome some of these problems, several additional
techniques have been studied in recent years. Percutaneous
transluminal laser angioplasty was claimed to overcome long
occlusive lesions better than PTA alone. 21 This technique,
however, has a considerable failure and complication rate, 22
often leading to surgical intervention.

Vroegindeweij and coauthors 23 evaluated whether en-
dovascular directional atherectomy combined with PTA
would provide better results than conventional balloon
angioplasty alone in symptomatic femoropopliteal disease.
In their prospective study, the outcome after atherectomy
appeared to be worse for lesions > 2 cm and similar for shorter
lesions.

Additional stent placement after PTA of the femoro-
popliteal segment should be avoided. Most reports on the use
of stents in the femoropopliteal area are disappointing com-
pared with stenting of the aortoiliac region. One-year patency
rates of 20-60% are no exception. 24 ' 25 For the greater part, the
Palmaz® (Cordis Johnson and Johnson, Warren, NJ, USA)
balloon-expandable stent and the self-expanding Wallstent®

534

chapter 46 Surgical and endovascular treatment of chronic ischemia of the lower limbs

Figure 46.2 Fractured Wallstent®, several months after placement in the
distal SFA.

(Schneider Boston Scientific, Natick, MA, USA) were used.
There are three randomized trials evaluating the additional
value of stenting after PTA in the SFA, all demonstrating that
stenting leads to a higher initial success. However, long-term
results are similar, or even worse, for SFA stenting. 26-28 The
inferior results of femoropopliteal stents compared with
aortoiliac stents may relate to several variables. The most im-
portant factor appears to be the smaller diameter of the
femoropopliteal vessels. Greater platelet and fibrin deposition
is seen on infrainguinal placed stents because of decreased
flow velocity, higher shear stress, and a relatively more cov-
ered vessel wall. 29 Therefore, patencies of femoropopliteal
stents were only acceptable when the arteries were at least
7 mm in diameter. Another complication of stents placed in the
femoropopliteal section is the possibility of crushing of the
stents, as shown in Fig. 46.2. To master long chronically oc-
cluded SFA segments percutaneously, Bolia et al. 30 introduced
a percutaneous intentional extraluminal recanalization (PIER)
technique. By use of a taper-tip J-wire, an extraluminal dissec-
tion plane is created, extending from just proximal to distal of
the occluded femoropopliteal segment. After this, balloon an-
gioplasty is performed throughout the entire length of the ex-

traluminal passage. Cumulative 3-year symptomatic and he-
modynamic patencies of 46% and 48%, respectively, are men-
tioned. 31 Technical failures are not uncommon (about 20%),
and are presumably caused by extensive medial calcification
which ensures the formation of the extraluminal dissection
plane. To date, the PIER technique seems to be a good alterna-
tive in poor risk patients with long occluded SFA segments for
whom PTA is unsuitable.

Surgical techniques

Remote endarterectomy

Nowadays, extended stenotic or occluded SFA lesions can be
treated by minimally invasive surgical remote endarterecto-
my combined with endoluminal stent implantation. The oper-
ative technique has been thoroughly described by Ho. 32 The
development of a ring strip cutter (Mollring Cutter®; Vascular
Architects, San Jose, CA, USA) makes it possible to perform
endarterectomy of an entirely occluded SFA through a single
groin incision. This ring stripper is a modification of the one
originally described by Cannon in 1955 and Vollmar in 1967.
The metal shaft has a double ring construction at the distal
end, replacing the single ring found on a conventional ring
stripper. Both rings have sharpened cutting edges on the inner
side, mimicking a pair of scissors as the lower ring shears
along the upper ring when a trigger is pulled (Fig. 46.3). After
meticulously dissecting the intimal core of the proximal SFA,
the ring stripper is passed around the intimal core until the
patent PI segment of the popliteal artery is reached. The ring
stripper is then exchanged for the ring strip cutter to cut the
distal part of the atheromatous core, endoluminally. As the
intimal core is simultaneously removed with the Mollring
Cutter®, the disobliterated SFA should be visualized by radio-
logical examination. The distal cut-off point of the intima is
then stented. The initial technical and clinical success rate of
remote endarterectomy is better than PTA taking into account
the length of the occluded SFA segment (10-45 cm). 33 Most
restenoses after remote endarterectomy occur within the first
year of the procedure and can be successfully treated with bal-
loon angioplasty. The cumulative 2-year primary assisted and
secondary patency rates are both 86%. As part of a Food and
Drug Administration trial the authors are using the recently
developed aSpire® Covered Stent (Vascular Architects, San
Jose, CA, USA) to prevent further dissection of the distal tran-
sected intimal flap. This stent (Fig. 46.4) is made of nitinol and
is manufactured in a double spiral configuration. It is then cov-
ered by a thin sleeve of polytetrafluoroethylene (PTFE) to pre-
clude any blood-metal contact. The spiral design is chosen for
better hemodynamic compatibility with the native vessel, and
the concept of partial coverage is intended to inhibit intimal
hyperplasia. Furthermore, the double helix configuration
makes the stent flexible and, therefore, kink and crush

535

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Closed Position

(Insertion and Removal)

pen Kosmon

(During cutting only)

Figure 46.3 The Mollring Cutter® in open
(during cutting only) and closed (insertion and
removal) position (A) and in detail (B).

Figure 46.4 The Aspire® covered stent in unwound position fixed at the
introduction catheter.

resistant while preserving side-branch access and maintain-
ing collaterals (Fig. 46.5). No studies have been published to
date, but the initial experience of the authors with the above-
mentioned technique and aSpire® stent in the distal SFA
section is very promising.

Semiclosed endarterectomy and supragenual
f emoropopliteal bypass surgery

The popularity of the semiclosed endarterectomy for the
femoropopliteal segment has fluctuated during past decades.
After the early promising results in the middle 1950s, several
reports were published at the beginning of the 1970s in which
femoropopliteal bypass grafting seemed to be superior. 34 ' 35
During the last decade, Van der Heijden and coauthors 36
studied retrospectively 231 semiclosed endarterectomies and
found a rather good 5-year overall cumulative patency of 71%,

536

chapter 46 Surgical and endovascular treatment of chronic ischemia of the lower limbs

Figure 46.5 The Aspire® covered stent is kink and crush resistant, here
shown during flexion of the knee with preservation of side-branches.

with acceptable complication and mortality rates (10% and
<1% respectively). However, in contrast, Heider et a/. 37 found a
disappointing 5-year primary patency rate of 44%.

Advocates of the technique favor the following advantages.
If the endarterectomy succeeds, the autologous saphenous
vein is spared and can be used for future revascularizations.
No foreign material is used, which minimizes the risk of peri-
operative infection. In the majority of cases, femoropopliteal
bypass is still possible, even if endarterectomy fails. 38 Last but
not least, endarterectomy spares the collateral circulation and
occluded collaterals can even be opened by endarterectomy.
This preservation of the collateral network affords a better out-
come of endarterectomy failure than femoropopliteal bypass
failure. 39

The overall final amputation rate following endarterectomy
for occlusions is about 5%, which is slightly lower than for
femoropopliteal bypass.

So, semiclosed endarterectomy is technically feasible when
long femoropopliteal segment occlusions have to be over-
come. Nevertheless, the above-described less invasive

procedure of remote endarterectomy with additional stent
placement seems to be a better treatment option than bypass
surgery.

Femoropopliteal bypass surgery should be reserved for se-
vere disabling claudication, limb-threatening ischemia, or to
overcome complicated cases of PTA or endarterectomy. Since
Kunlin 40 performed the first femoropopliteal bypass using au-
tologous vein in 1949, an astonishing number of studies have
been published on the subject of the material used for revascu-
larization. Unfortunately, this literature is confounded by a
multitude of variables which prevents statistically valid com-
parisons. Besides, most of the studies are not randomized or
controlled and did not include more than two revasculariza-
tion options.

Femoropopliteal reconstruction is performed with auto-
logous venous material or prosthetics. Mostly, venous recon-
struction is performed using the great saphenous vein and can
be divided into in-situ, reversed, and nonreversed reconstruc-
tions. In the case of prosthetics, most surgeons prefer PTFE,
ringed or nonringed, or Dacron. Some other alternative con-
duits have been used such as human umbilical vein (HUV) or
homologous denatured saphenous vein; 41 ' 42 both are becom-
ing less popular because they are cumbersome and cause
aneurysmal degeneration, which is a particular problem with
HUV.

Results

Most authors consider the autologous saphenous vein to be
the best conduit for femoropopliteal bypass surgery. The
5-year cumulative patency rates of the autologous veins
(56-76%) are (significantly) higher than prosthetic grafts
(39-61 %). 43-45 In a recent prospective study carried out by
Burger et al., A6 a similar outcome was found.

When the decision is made to use the greater saphenous vein
for supragenual bypass grafting, the question of whether to
perform an in-situ or reversed venous reconstruction is harder
to answer. Several authors 47,48 found no difference in the 5-
year cumulative patency of the two types of venous recon-
struction. Most important is the fact that a greater saphenous
vein diameter <4 cm is likely to halve the expected long-term
patency rate (33% vs. 77%). Furthermore, the in-situ graft
requires more secondary interventions

In the absence of the ipsilateral greater saphenous vein, the
contralateral one is a good alternative. Application of the less-
er saphenous vein or arm veins lowers the primary and sec-
ondary graft patency dramatically, and should therefore be
avoided. In these cases, it is better to use prosthetic material.
Multiple prospective randomized trials could not prove any
significant difference in long-term results between PTFE and
Dacron grafts, 49,50 with 5-year cumulative primary patency
rates of more than 65%. However, it is the individual choice of
the vascular surgeon which stipulates the type of prosthetic
graft used.

537

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Risk factors

In almost every publication on femoropopliteal revasculariza-
tion, the risk factors determining graft failure are evaluated.
Summaries are given below. In the mid-1990s, a multicenter
randomized trial was carried out in the Netherlands to com-
pare the effectiveness of oral anticoagulants with that of
aspirin in preventing venous and prosthetic infrainguinal
bypass grafts from thrombotic events (the Dutch BOA Study).
In total, 1222 supragenicular reconstructions were included.
Univariate risk analysis demonstrated that critical ischemia,
poor run-off (<1 run-off vessel), and use of nonvenous grafts
were associated with higher risk for graft occlusion. 51 Of note,
however, was the fact that treatment of hypertension and /or
hyperlipidemia lowered the risk for graft occlusion (both ve-
nous and prosthetic grafts). In the literature, there is evidence
that the outcome of patients on dialysis is very poor after in-
frainguinal bypass grafting. Because of dramatic 1-year paten-
cy and leg salvage rates of 47% and 37%, respectively, Peltonen
et al. 52 have doubts about lower limb revascularization in dial-
ysis patients at all. Furthermore, the extent of significant
comorbidity (cardiac and pulmonary compromised, diabetes
mellitus, smoking) has a negative influence on the patency of
both venous and prosthetic bypasses.

Treatment modalities for
the infrapopliteal segments

Balloon angioplasty

With the introduction of small-diameter equipment to
perform PTAs of the coronary arteries, endovascular recana-
lization of the infrapopliteal vascular tree has expanded
enormously when an antegrade approach to the ipsilateral
common femoral artery is required. Because of the small ves-
sel size and reduced blood flow, the use of antispasmodics (e.g.
tolazoline, nitroglycerin) is recommended during the proce-
dure. In addition, intraarterial heparin should also be given
periprocedurally. In spite of optimizing the conditions, the
risks of PTA in the infrapopliteal segments are substantially
higher than with the femoropopliteal segment. 53-55 The length
of the occlusion is an important determinant of technical suc-
cess, with a cut-off point at about 5 cm. Lesions > 5 cm are prone
to dissection due to PTA. 56 Bolia and coauthors 57 overcome
this complication by subintimal recanalization of the dissected
segment. They report good technical and clinical success rates
(86% and 79%, respectively), but unfortunately with a median
follow-up of just 4 weeks. In conclusion, PTA of the in-
frapopliteal segments should only be performed in patients
suffering from disabling claudication or Fontaine III and IV
arterial ischemia with short segment stenosis or occlusion
(Fig. 46.6). Another group of patients to profit from balloon an-
gioplasty are those with contraindications to operation.

Bypass surgery

During the last few decades, progressively more distal by-
passes have been carried out in patients with more extensive
comorbidity. Improvement of surgical equipment, technique,
and methods of exposure have made it possible to construct
anastomoses more distally, even as far as the plantar branches.
Increasing use of the in-situ saphenous vein bypass has en-
abled the surgeon to anastomose veins 2.5 mm in diameter
to the crural arteries. If there are no usable veins, better pros-
thetic graft patencies have been realized by improving
distal anastomosis. For example, the creation of an arteriove-
nous fistula at the distal anastomosis of a prosthetic graft
appeared to improve local microcirculatory hemodynamics
and thus the outcome of the operation. 58 The same benefits
are claimed when an additional venous cuff is created at the
distal anastomosis 59 or by the usage of a "shoelike" preformed
prosthetic graft (e.g. Distaflo®; C.R. Bard, Inc., Tempe, AZ,
USA).

Of the 2650 prospectively included infrainguinal bypass
grafts studied in the Dutch BOA Study, 51 about 55% were
infragenicular and 20% were femorocrural. This division of
infrainguinal bypasses is also to be found in other large
studies.

Shah and coworkers 60 studied the long-term results of the
in-situ vein bypasses and found rather good 10-year primary
and secondary patency rates of 60% and 76%, respectively.
Another important and impressive result is the 10-year limb
salvage rate of 90%. Most other studies showed slightly less
impressive results, but nevertheless the mean 5-year primary
patency rate of infrapopliteal venous reconstruction ranged
from 65% to 75%. 61/62 During the first year after infragenicular
venous reconstruction, about 50% of the grafts are at risk.
Early failure (within 1 month) occurs in 10-20% of the grafts 63
and the remaining grafts will develop stenoses thereafter. 64
Reviewing the literature, no consensus about in-situ or re-
versed venous reconstruction can be found. The authors
prefer the in-situ technique only for distal anastomoses
to the posterior tibial artery and for small-caliber crural
arteries.

Compared with supragenicular femoropopliteal bypass
surgery, the primary and secondary patency rates of prosthet-
ic grafts are lower 51 ' 65 than the rates for venous grafts. Three-
year primary patencies in the range of 20-70% have been
described. It is hard to compare results of any published series
because of a mixture of techniques, material, site of distal anas-
tomosis, and patient characteristics. The most important
factors which influence long-term success are the inflow state,
the number of crural vessels, and, most of all, the presence
of straight flow to the foot in combination with the presence
of open pedal vessels. 66

538

chapter 46 Surgical and endovascular treatment of chronic ischemia of the lower limbs

Figure 46.6. Localized, severe stenosis of the
truncus tibiofibularis before (A) and after (B)
percutaneous transluminal angioplasty.

Surveillance of inf rainguinal vascular
reconstruction

The majority of restenoses occur during the first year after en-
dovascular or surgical lower limb revascularization. Peak in-
cidence is in the first month (10-20%), probably because of
technical reasons. Ho 32 showed that as many as 80% of
restenoses after remote endarterectomy occurred in the first
12 months after operation. Only 22% of hemodynamically sig-
nificant restenoses (>50%) were correlated with worsening of
clinical symptoms, change of ankle-brachial index, or both.
The last has been confirmed in more publications. Thus, sur-
veillance of (endo)vascular reconstruction requires a more
reliable test.

In a systematic review, Koelemay. 67 found duplex scanning
(DS) and magnetic resonance angiography (MRA) to be supe-
rior to segmental blood pressure or pulse volume measure-
ments, or Doppler signal analysis for the localization and
gradation of stenoses in the femoropopliteal segment. Intraar-
terial digital subtraction angiography (iaDSA) is the gold stan-
dard for the crural arteries, but the specificity and sensitivity of
DS are also good in these segments. As DS is noninvasive and
relatively cheap, it is the method of choice for surveillance
after lower limb revascularization. The degree of stenosis is

best classified using the peak systolic velocity (PSV) ratio, a
validated criterion published by Legemate et ah 68 A PSV ratio
of >2.5 is considered to relate to a >50% arterial diameter re-
duction with hemodynamic significance. The authors prefer a
follow-up scheme of 3, 6, and 12 months with DS after revascu-
larization, and then annually.

Summary and future perspectives

Peripheral vascular disease of the lower extremities is one of
the most common diseases of mankind, with a 5-year inci-
dence of up to 20% for older (> 60 years) age groups. In combi-
nation with the high prevalence of other vascular comorbidity
(e.g. coronary and cerebrovascular arterial diseases), this pa-
tient group has a high mortality rate. 69 Therefore, it is impor-
tant that these patients are not unnecessarily exposed to
several therapies; initially, the optimal treatment modality
should be chosen. Numerous investigators have studied this
subject. However, it is very apparent that no prospective, ran-
domized, controlled studies can be found in which all treat-
ment modalities (PTA, endarterectomy, and bypass surgery)
are represented with long-term follow-up.

When reviewing the present-day literature, several guide-
lines can be drawn up. In short-segment stenoses or occlu-

539

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

sions, endovascular techniques are preferable. Advantages
of the endovascular techniques are the preservation of auto-
logous veins, the maturation of collaterals, and the post-
ponement of surgery and its related complications. The
turning point to more invasive techniques is 10 cm for the
femoropopliteal segment and 5 cm for the infrageniculate re-
gion. Until recently, additional stent placement was not rec-
ommended after PTA because there is no improvement in
patency and there is a greater likelihood of complications.
With the introduction of new-generation stents, such as the
Aspire® stent, there might be a benefit after all. In the immedi-
ate future, stents coated with thrombolytic agents, antibiotics
(e.g. the Cypherstent®; Cordis, Johnson and Johnson), ra-
dioactive agents, and /or antimitotic agents will all be evaluat-
ed for their effectiveness in counteracting the atherosclerotic
process. This will probably be a step forward in the prevention
of critical lower limb ischemia.

At present, long-segment (>10cm) sclerotic lesions of the
AFS can be treated very successfully using remote endarterec-
tomy with a single groin incision. To perform this procedure, a
patent PI segment of the supragenicular popliteal artery is
necessary. The remote endarterectomy includes placement of
a stent (such as Aspire®) at the distal transition zone. Mid-
term results of this technique are very promising. The advan-
tages of remote endarterectomy over bypass surgery are
similar to the advantages mentioned in the section on PTA.
The more invasive semiclosed endarterectomy is likely to
become obsolete.

Limb-threatening ischemia due to SFA lesions that are un-
suitable for either PTA or endarterectomy (e.g. extended ather-
osclerosis, extension into the popliteal artery) needs surgical
revascularization. Autologous greater saphenous vein by-
passes are the best conduits and are, therefore, preferable. If
the diameter of the autologous (greater or lesser saphenous)
vein is < 4 cm, the primary patency will decrease dramatically
and better results can be achieved by using PTFE or Dacron.
Another advantage of the prosthetic above-knee reconstruc-
tion is the conservation of the greater saphenous vein for
future femorocrural reconstruction. In a prospective study,
Berlakovich et al. 70 showed that only 7% of patients with
a supragenicular bypass needed a secondary below-knee
bypass in the ipsilateral limb during 4 years' follow-up.

In inf ragenicular bypass surgery, a pronounced difference is
seen in the long-term follow-up between venous and prosthet-
ic grafts. If available, the greater saphenous vein should be
used, whether in situ or reversed. Only a venous diameter
< 3-4 mm justifies a prosthetic reconstruction. The outcome of
femorocrural bypasses is strongly influenced by the number of
calf vessels and their continuation to ankle and pedal level.

During the first year after infrainguinal revascularization,
50% of the reconstructions will be at risk owing to (re)stenosis.
Intensive surveillance with noninvasive duplex monitoring is
required to detect these (re)stenoses, most of which occur be-
fore they are clinically obvious. The majority of (re)stenoses

are localized and can be treated successfully by means of min-
imally invasive techniques.

Considering this, it is obvious that the treatment of chronic
lower limb ischemia would benefit from a multidisciplinary
approach from both the vascular surgeon and the interven-
tional radiologist. Both specialisms have to be involved in
the follow-up after infrainguinal revascularization. The basic
principle should be to use the optimal treatment for each indi-
vidual patient, with as little invasion as possible. To improve
the decision-making process of whether to choose an endo-
vascular, endovascular assisted, or surgical technique, well
defined, prospective randomized trials are needed.

References

1. Brown AL, Juergens JL. Arteriosclerosis and atherosclerosis.
In: Fairbairn JF, Juergens JL, Spittell JA, eds. Peripheral Vascular
Disease. Philadelphia: WB Saunders, 1972.

2. Mustard JF Recent advances in molecular pathology: a review:
platelet aggregation, vascular injury and atherosclerosis. Exp Mol
Pathol 1967; 7:366.

3. Kinlough-Rathbone RL, Mustard JF Atherosclerosis: current con-
cepts. Am J Surg 1981; 141:638.

4. Rosen AJ, DePalma RG. Risk factors in peripheral atherosclerosis.
Arch Surg 1973; 107:303.

5. Stamler J, Berkson DM, Lindberg HA. Risk factors: their role in
the etiology and pathogenesis of the atherosclerotic diseases. In:
Wissler RW, Geer JC, eds. The Pathogenesis of Atherosclerosis.
Baltimore: Williams and Wilkins, 1972:41.

6. Ueland PM, Refsum H, Brattstrom L. Plasma homocysteine and
cardiovascular disease. In: Francis RB Jr, ed. Atherosclerotic Cardio-
vascular Disease, Hemostasis, and Endothelial Function. New York:
Marcel Dekker, 1993:183.

7. Texon M, Imperato AM, Helpern M. The role of vascular dynamics
in the development of atherosclerosis. JAMA 1965; 12:1226.

8. Haimovici H. Patterns of arteriosclerotic lesions of the lower ex-
tremity. Arch Surg 1967; 95:918.

9. Imparato AM, Kim GE, Davidson T, Crowely JG. Intermittent
claudication: its natural course. Surgery 1975; 78:795.

10. Walsh DB, Gilbertson JJ, Zwolak RM et al. The natural history of
superficial femoral artery stenoses. / Vase Surg 1991; 14:299.

11. Golledge J, Ferguson K, Ellis M et al. Outcome of femoropopliteal
angioplasty. Ann Surg 1999; 229:146.

12. Lechner G, Jantsch H, Waneck R, Kretschmer G. The relationship
between the common femoral artery, the inguinal crease, and the
inguinal ligament: a guide to accurate angiographic puncture.
Cardiovasc Interv Radiol 1998; 11:165.

13. Pentecost MJ, Criqui MH, Dorros G et al. Guidelines for peripheral
percutaneous transluminal angioplasty of the abdominal aorta
and lower extremity vessels. Circulation 1994; 89:511.

14. Kinney TB, Chin AK, Rurik GW et al. Transluminal angioplasty: a
mechanical-pathological correlation of its physical mechanisms.
Radiology 1984; 153:85.

15. Rutherford RB, Becker GJ. Standards for evaluating and reporting
the results of surgical and percutaneous therapy for peripheral
arterial disease. / Vase Interv Radiol 1991; 2:169.

540

chapter 46 Surgical and endovascular treatment of chronic ischemia of the lower limbs

16. Whyman MR, Ruckley CV, Fowkes FGR. A prospective study
of the natural history of femoropopliteal artery stenosis using
duplex ultrasound. Eur J Vase Surg 1993; 7:444.

17. Capek P, MacLean GK, Berkowitz HD. Femoral angioplasty:
factors influencing long term success. Circulation 1991; 83
(Suppl.I):I70.

18. Johnston KW. Femoral and popliteal arteries: reanalysis of results
of balloon angioplasty. Radiology 1992; 183:767.

19. Hunink MGM, Donaldson MC, Meyerovitz MF et al. Risks and
benefits of femoropopliteal percutaneous balloon angioplasty.
] Vase Surg 1993; 17:183.

20. Murray RR Jr, Hewes RC, White RI Jr et al. Long-segment
femoropopliteal stenoses: is angioplasty a boon or a bust? Radiolo-
gy 1987; 162:473.

21. Pilger E, Lammer J, Bertuch H et al. YAG laser with sapphire tip
combined with balloon angioplasty in peripheral arterial occlu-
sion. Circulation 1991; 83:141.

22. Sanborn TA, Cumberland DC, Greenfield AJ, Welsh CL, Guben JK.
Percutaneous laser thermal angioplasty: initial results and 1-year
follow-up in 129 femoropopliteal lesions. Radiology 1988; 168:121.

23. Vroegindeweij D, Tielbeek AV, Buth J et al. Directional atherectomy
versus balloon angioplasty in segmental femoropopliteal artery
disease: two year follow-up with color-flow duplex scanning.
] Vase Surg 1995; 21:255.

24. Gray B, Sullivan T, Childs M et al. High incidence of restenosis/re-
occlusion of stents in the percutaneous treatment of long-segment
superficial femoral artery disease after suboptimal angioplasty.
] Vase Surg 1997; 25:74.

25. Martin E, Katzen B, Benenati J et al. Multicenter trial of the Wall-
stent in the iliac and femoral arteries. / Vase Interv Radiol 1995;
6:843.

26. Vroegindeweij D, Vos LD, Tielbeek AV, Buth J, vd Bosch HC. Bal-
loon angioplasty combined with primary stenting versus balloon
angioplasty alone in femoropopliteal obstructions: a comparative
study. Cardiovasc Interv Radiol 1997; 20:420.

27. Grimm J, Muller-Hulsbeck S, Jahnke T et al. Randomized study to
compare PTA alone versus PTA with Palmaz stent placement for
femoropopliteal lesions. / Vase Interv Radiol 2001; 12:935.

28. Do-Dai DO, Triller J, Walpoth B et al. A comparison study of
self-expandable stents versus balloon angioplasty alone in
femoropopliteal artery occlusions. Cardiovasc Interv Radiol 1992;
15:306.

29. Palmaz J. Intravascular stents: tissue-stent interaction and design
considerations. Am] Roentgenol 1993; 160:613.

30. Bolia A, Miles KA, Brennan J, Bell PRF. Percutaneous transluminal
angioplasty of occlusions of the femoral and popliteal arteries by
subintimal dissection. Cardiovasc Interv Radiol 1990; 13:357.

31 . London NJM, Srinivasan R, Naylor AR et al. Subintimal angioplas-
ty of femoropopliteal artery occlusions: the long-term results. Eur
] Vase Surg 1994; 8:148.

32. Ho GH. Endovascular remote endarterectomy: initial experience
with a new technique in the treatment of superficial femoral artery
occlusive disease. Thesis, University of Utrecht, the Netherlands,
1999.

33. Ho GH, Moll FL. Remote endarterectomy in SFA occlusive
disease. Eur] Radiol 1998; 28:205.

34. Darling RC, Linton RR. Durability of femoropopliteal reconstruc-
tions. Am J Surg 1972; 123:472.

35. De Weese JA, Barner HB, Mahoney EB, Rob CG. Autogenous ve-

nous bypass grafts and thromboendarterectomy for atheroscle-
rotic lesions of the femoropopliteal arteries. Ann Surg 1966;
163:205.

36. Heijden van der FHWM, Eikelboom BC, Reedt Dortlan van RWH
et al. Endarterectomy of the superficial femoral artery a procedure
worth reconsidering. Eur J Vase Surg 1992; 6:651.

37. Heider P, Hofmann M, Maurer PC, Sommoggy van S. Semi-closed
femoropopliteal thromboendarterectomy: a prospective study.
Eur J Vase Endovasc Surg 1999; 18:43.

38. Heijden van der FHWM. Semi-closed endarterectomy of the
superficial femoral artery. Thesis, ADDIX, Wijk bij Duurstede, the
Netherlands, 1994.

39. Canon JA, Barker WF, Kawakami IG. Femoral popliteal en-
darterectomy in the treatment of obliterative atherosclerotic
disease. Surgery 1958; 43:76.

40. Kunlin J. Le traitement de l'ischemie arterique par la greffe
veineuse longue. Rev Chir 1951; 70:206.

41 . Dardik H, Dardik I. Successful arterial substitution with modified
umbilical vein. Ann Surg 1976; 182:252.

42. Reedt Dortland RWH van, Leeuwen MS van, Steyling JJF,
Theodorides Th, Vroonhoven ThJMV van. Long-term results with
vein homograft in femoro-distal arterial reconstruction. Eur] Vase
Surg 1991; 5:557.

43. Veith FJ, Gupta SK, Ascer E et al. Six-year prospective multicenter
randomized comparison of autologous saphenous vein and
expanded polytetrafluoroethylene in infra-inguinal arterial
reconstructions. / Vase Surg 1986; 3:104.

44. Johnson WC, Lee KK. Comparative evaluation of PTFE, HUV and
saphenous vein bypasses in femoropopliteal above knee vascular
reconstruction. / Vase Surg 2000; 32:268.

45. Kent KC, Donaldson MC, Attinger CE, Couch NP, Mannick JA,
Whittemore AD. Femoropopliteal reconstruction for claudica-
tion. Arch Surg 1988; 123:1196.

46. Burger DHC, Pieter-Kappetein A, Bockel JH van, Breslau PJ. A
prospective randomized trial comparing vein with polytetrafluo-
roethylene in above-knee femoropopliteal bypass grafting. / Vase
Surg 2000; 32:278.

47. Moody AP, Edwards PR, Harris PL. In situ versus reversed
femoropopliteal vein grafts: long-term follow-up of a prospective,
randomized trial. Br J Surg 1992; 79:750.

48. Lawson J A, Tangelder MJD, Algra A, Eikelboom BC. The myth of
the in situ graft: superiority in infrainguinal bypass surgery? Eur J
Vase Endovasc Surg 1999; 18:149.

49. Post S, Kraus T, Muller-Reinartz et al. Dacron versus polytetraflu-
oroethylene grafts for femoropopliteal bypass: a prospective ran-
domised multicentre trial. Eur J Vase Endovasc Surg 2001; 22:226.

50. Rosenthal D, Evans D, McKinsey J et al. Prosthetic above-knee
femoropopliteal bypass for intermittent claudication. / Cardiovasc
Surg 1990; 31:462.

51. Tangelder MJD, Algra A, Lawson JA, Eikelboom BC on behalf of
the Dutch BOA Study Group. Risk factors for occlusion of infrain-
guinal bypass grafts. Eur J Vase Endovasc Surg 2000; 20:118.

52. Peltonen S, Biancari F, Lindgren L, Makisalo H, Honkanen E,
Lepantalo M. Outcome of infrainguinal bypass surgery for critical
leg ischemia in patients with chronic renal failure. Eur } Vase
Endovasc Surg 1998; 15:122.

53. Bakal CW, Sprayregen S, Scheinbaum K et al. Percutaneous
transluminal angioplasty of the infrapopliteal arteries: results in
53 patients. Am J Roentgenol 1990; 154:171.

541

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

54. Leofberg AM, Leorelius LE, Karacgil S et ah The use of below-knee
percutaneous transluminal angioplasty in arterial occlusive dis-
ease causing chronic critical limb ischemia. Cardiovasc Interv
Radiol 1996; 19:317.

55. Treimann GS, Treimann RL, Ichikawa L et ah Should percutaneous
transluminal angioplasty be recommended for treatment of infra-
geniculate popliteal artery or tibioperoneal trunk stenosis? / Vase
Surg 1995; 22:457.

56. Schwarten DE. Clinical and anatomical considerations for nonop-
erative therapy in tibial disease and the results of angioplasty.
Circulation 1991; 83 (Suppl. I):86.

57. Bolia A, Sayers RD, Thompson MM, Bell PRE Subintimal and in-
traluminal recanalisation of occluded crural arteries by percuta-
neous balloon angioplasty. Eur] Vase Surg 1994; 8:214.

58. Jacobs MJHM, Reul GJ, Gregoric ID et ah Creation of a distal
arteriovenous fistula improves microcirculatory hemodynamics
of prosthetic graft bypass in secondary limb salvage procedures. /
Vase Surg 1993; 18:1.

59. Raptis LS, Miller JH. Influence of a vein cuff on PTFE grafts for pri-
mary femoropopliteal bypass. Br] Surg 1995; 82:487.

60. Shah D, Darling R, Chang B et ah Long-term results of in situ
saphenous vein bypass: analysis of 2058 cases. Ann Surg 1995;
222:438.

61 . Donaldson MC, Mannick JA, Whittemore AD. Femoral-distal by-
pass with in situ greater saphenous vein: long-term results using
the Mill valvulotome. Ann Surg 1991; 213:457.

62. Taylor LM, Edwards JM, Porter JM. Present status of reversed vein
bypass grafting: five-year results of a modern series. / Vase Surg
1990; 11:193.

63. Beard JD, Wyatt M, Scott DJA, Baird RN, Horrocks M. The non re-
versed femorodistal bypass graft: a modification of the standard
in situ technique. Eur] Vase Surg 1989; 3:55.

64. Moody P, Gould DA, Harris PL. Vein graft surveillance improves
patency in femoropopliteal bypass. Eur] Vase Surg 1990; 4:117.

65. Tordoir JHM, Plas van der JPL, Jacobs MJHM, Kitselaar PJEHM.
Factors determining the outcome of crural and pedal revasculari-
sation for critical limb ischaemia. Eur] Vase Surg 1993; 7:82.

66. Panayiotopoulos YP, Tyrrell MR, Owen SE, Reidy JF, Taylor PR.
Outcome and cost analysis after femorocrural and femoropedal
grafting for critical limb ischaemia. Br] Surg 1997; 84:207.

67. Koelemay M. Non-invasive assessment of peripheral arterial
occlusive disease. Thesis, PrintPartners Ipskamp, Enschede, the
Netherlands, 2001.

68. Legemate DA, Teeuwen C, Hoeneveld H, Ackerstaff RGA,
Eikelboom BC. Spectral analysis criteria in duplex scanning of
aortoiliac and femoropopliteal arterial disease. Ultrasound Med
Biol 1991; 17:769.

69. O'Riorddain DS, O'Donnel JA. Realistic expectations for the
patients with intermittent claudication. Br] Surg 1991; 78:861.

70. Berlakovich G A, Herbst F, Mittlblock M, Kretschmer G. The choice
of material for above-knee femoropopliteal bypass. A 20-year
experience. Arch Surg 1994; 129:297.

542

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Aortoiliac endovascular recanalization
compared with surgical reconstruction

Peter L. Faries
Michael L. Marin

Endovascular recanalization has emerged as an alternative to
conventional surgical bypass for the treatment of extensive
aortoiliac occlusive disease. 1 Endovascular techniques with
the adjunctive use of stent-grafts for the treatment of long-
segment occlusions of the infrarenal aorta and iliac arteries
offer advantages over both conventional open surgery and
percutaneous transluminal angioplasty (PTA) with or without
stent placement. Conventional surgery requires extensive ab-
dominal and retroperitoneal dissection and is associated with
significant morbidity and mortality, particularly in patients
with comorbid medical conditions. 2 Iliac artery PTA and stent
placement has achieved success in patients with limited areas
of involvement and incomplete occlusion; however, its effec-
tiveness has been severely limited in patients who have more
extensive segments of disease or who have complete occlu-
sions. 3 ' 4 Experience with endovascular recanalization and
stent-graft placement has demonstrated both a high degree
of immediate technical success and good midterm patency
and limb salvage rates while maintaining low morbidity and
mortality rates even in patients with extensive comorbid
illnesses. 5,6

Presentation and natural history

Arterial reconstruction of occluded aortoiliac segments may
be performed for the treatment of severe intermittent claudi-
cation involving the hip, gluteal and thigh musculature, or for
decreased sexual potency in the male patient. These patients
may demonstrate pallor of the extremity on elevation and
rubor on dependency and they may have lower abdominal or
femoral bruits, particular after exercise. Aortoiliac reconstruc-
tion may also be performed for limb-threatening conditions
including rest pain and tissue loss. Patients who present with
limb-threatening ischemia may often have associated infrain-
guinal arterial occlusive disease. In such cases additional in-
frainguinal reconstructions are typically necessary. However,
correction of inflow disease is necessary prior to infrainguinal
procedures and in certain instances correction of inflow

disease alone may be sufficient to relieve the ischemic
symptoms.

Patients who develop symptomatic aortoiliac occlusive dis-
ease frequently exhibit risk factors for generalized atheroscle-
rosis. These most commonly include advanced age, tobacco
use, hyperlipidemia, homocysteinemia, hypertension, dia-
betes mellitus, and renal failure. 7 Of these factors, smoking has
been associated with the greatest risk of developing occlusive
disease in the aortoiliac segment. The risk of developing aor-
toiliac disease for patients who smoke has been calculated to
be three to five times greater than the risk of developing dis-
ease in the more peripheral arterial circulation. 8 Diabetes and
renal failure have been less significantly associated with prox-
imal compared with distal arterial occlusive disease. In addi-
tion, progression of aortoiliac occlusive disease in the diabetic
patient appears to be less rapid than for other risk factors. Of
particular significance, the presence of peripheral arterial
occlusive disease has been associated with decreased life
expectancy compared with similar age-matched patient
cohorts. 9

Conventional surgical therapy

The long-term patency and limb salvage rates reported for
aortofemoral bypass typically exceed 90% at 5 years and 75%
at 10 years. 2 ' 10-12 In addition, these patency rates are most often
achieved without the need for reintervention. As a result
aortofemoral bypass is considered the gold standard for
arterial reconstruction of the aortoiliac segment. The decision
to perform conventional surgical arterial reconstruction
is therefore most commonly based on assessment of the
patient's surgical risk and the extent and distribution of their
disease. In general, aortofemoral bypass is reserved for
patients who do not have significant comorbid medical
conditions that would place them at increased risk for major
vascular surgery. 13

Aortofemoral bypass is typically carried out through a
midline laparotomy incision in conjunction with bilateral

543

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 47.1 Arteriographic image of a right
iliac artery occlusion treated by endovascular
recanalization and stent-graft placement. (A)
The preoperative arteriogram demonstrates
complete occlusion of the right common and
external iliac arteries. (B) After stent-graft
deployment excellent inflow has been
reestablished to the right lower extremity.

groin incisions. Alternatively, a retroperitoneal approach may
be used. The graft may be anastomosed to the aorta in either
an end-to-side or an end-to-end fashion. The end-to-end con-
figuration is thought to provide superior flow characteristics
at the anastomosis and to be easier to cover in the retroperi-
toneum. The end-to-side configuration may aid in maintain-
ing perfusion of the internal iliac and inferior mesenteric
arteries in patients with total occlusion of the external iliac ar-
teries. In constructing the bypass the graft limbs are tunneled
posterior to the ureter and the femoral anastomoses are
created in a manner that ensures direct flow into the profunda
femoris arteries.

Alternative arterial bypass procedures

Extraanatomic bypass may be an alternative for patients with
occlusion of extensive iliac arterial segments and significant
comorbid medical conditions. Options for extraanatomic by-
pass include femorofemoral bypass in patients with unilateral
disease as well as axillofemoral bypass. The physiological im-
pact of the extraanatomic revascularization procedure is less
than that of aortofemoral bypass and the procedure is there-
fore more easily tolerated by patients with significant comor-
bid medical illnesses. 14 The patency and limb salvage rates are
significantly lower for extraanatomic revascularization proce-
dures, particularly for axillofemoral bypass compared with
aortofemoral bypass with patency rates ranging from 35% to
76%. 15 As expected, patency and limb salvage rates are
affected by patient selection and extraanatomic bypass re-
mains a potentially effective treatment option in appropri-
ately selected patients.

Percutaneous transluminal angioplasty

Patients with limited areas of occlusion frequently respond fa-
vorably to less invasive interventions, particularly PTA with
or without stent placement. The effectiveness of angioplasty
procedures is directly influenced by the anatomic characteris-
tics of the occlusive lesion. 4 In particular, reports have con-
firmed that short-segment lesions with incomplete occlusion
maintain higher long-term patency rates than long-segment
total occlusions. Proximal lesions, especially in the common
iliac artery, also respond more favorably to angioplasty than
do lesions located more distally in the arterial tree. 3 Stent
placement has improved the immediate and long-term paten-
cy of PTA in patients who exhibit a residual stenosis or pres-
sure gradient after PTA alone. Stent placement also aids in
managing intimal dissection after angioplasty and serves to
prevent or correct acute occlusive dissections. 16-18

Endovascular recanalization
and stent-graft treatment

Indications

Recanalization of completely occluded iliac arterial segments
using endovascular techniques with the concomitant place-
ment of a stent-graft to reline the dilated tract provides an
effective alternative to conventional bypass procedures 19
(Fig. 47.1). Endovascular stent-graft treatments are less inva-
sive and therefore may be carried out with reduced morbidity
and mortality, decreased patient discomfort, length of hospital

544

chapter 47 Aortoiliac endovascular recanalization compared with surgical reconstruction

stay, and patient convalescence. The midterm results of
stent-graft treatments for iliac occlusive disease compare fa-
vorably with extraanatomic revascularization procedures and
approach those of the aortofemoral bypass. 5 ' 6 Endovascular
iliac artery recanalization and stent-graft placement offer con-
siderable advantage over aortofemoral bypass in patients
with characteristics of a hostile abdomen. In particular, pa-
tients with intraabdominal sepsis particularly with enteric
sources of contamination, irradiation, malignant tumor,
stomas, retroperitoneal fibrosis, multiple abdominal opera-
tions, complex ventral hernia, massive obesity, or ascites may
be more effectively managed with endovascular techniques. 20
Patients at prohibitive risk for general anesthesia or exten-
sive vascular surgery mandate a less invasive approach such
as endovascular recanalization and stent-graft placement.
Physiological factors such as severe coronary artery disease
including recent myocardial infarction, intractable heart
failure, uncorrectable angina pectoris as well as severe
pulmonary insufficiency including forced expiratory volume
<1 1/s, home oxygen dependence and dyspnea at rest as well
as chronic renal failure, morbid obesity, and systemic disease
that limits life expectancy to less than 2 years also provide
strong indications for endovascular rather than conventional
surgical approaches. 21

Historical development of stent-grafts for
arterial occlusive disease

While iliac artery recanalization with or without stent place-
ment has been undertaken in the past, the severe limitation on
long-term patency had prevented widespread adoption. The
first use of a covered stent to treat iliac artery occlusion was
performed by the Ukrainian surgeon, Nicholos Volodos and
coworkers, and was reported in 1986. 22 He utilized a Dacron-
covered Z-stent in the treatment of a long-segment total occlu-
sion of the left iliac artery (Fig. 47.2). This novel approach was
reported in the Russian language literature but was largely
overlooked in the west. The wider use of covered stents or
stent-grafts for the treatment of arterial occlusive disease pro-
gressed with the development of physician-made devices 20 ' 23
(Fig. 47.3). Subsequently, commercially produced devices
have been developed. These devices unite stent technology
with prosthetic vascular graft material to produce covered
stents that may be used for the treatment of arteriovenous fis-
tulae, arterial aneurysms, and pseudoaneurysms in addition
to being used for arterial occlusive disease. 24

Currently, commercially produced covered stents or
stent-grafts which can be deployed percutaneously include
the Viabahn, produced by the WL Gore Corporation (Flagstaff,
AZ, USA) and the Wallgraft produced by the Boston Scientific
Corporation (Natick, MA, USA). Both devices consist of vas-
cular graft material supported throughout its entire length by

Figure 47.2 Original stent-graft utilized by Nicholas Volodos for the
treatment of iliac artery occlusion. (A)The endoluminal prosthesis is
constructed from Dacron fabric supported by Z-shaped stents. (B) Resolution
of the left iliac occlusion was accomplished using the self-fixing
endoprosthesis.

545

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 47.3 (A) The physician-made stent-graft used for treatment of iliac
occlusive disease is composed of a balloon-expandable stent which is
deployed proximally. The PTFE conduit is then used to line the newly
recanalized iliac flow channel. (B) Diagram demonstrating the system used to
deliver the physician-made stent-graft.

^ , . . . - ■

■ - - ■ ■ ■■ •• . ■ ■ ■

■

C *-'

r
i

eft* *

■ 1 h

L . ■

kr--'

Figure 47.4 Deployment of the Viabahn-covered stent. Top: The covered
stent is folded onto itself and constrained by a suture on the delivery catheter.
Middle: As the suture constraining the self-expanding stent is removed, the
covered stent begins to unfold and deploy. Bottom: After the graft is fully
released, it expands to its original size.

metallic stents. The Viabahn utilizes a polytetrafluoroethylene
(PTFE) graft and nitinol stents. It is compressed in diameter
and loaded into its delivery catheter by wrapping the graft
onto itself (Fig. 47.4). In contrast, the Wallgraft is an extension
of the stainless-steel Wallstent which is then covered with
Dacron graft material. The Wallgraft foreshortens in length as
it expands radially during deployment (Fig. 47.5).

Figure 47.5 Wa I Ig raft-covered stent. The Wa I Ig raft-covered stent is
composed of a self-expanding stainless-steel stent covered by Dacron graft
material.

Advantage of stent-graf ts over
uncovered stents

Covered stents or stent-grafts have been demonstrated to pro-
vide a significant advantage in the immediate patency rates
for aortoiliac recanalization procedures. 25,26 Percutaneous an-
gioplasty alone or in combination with uncovered stents com-
monly failed to achieve an adequate arterial flow channel and
as a result often failed to achieve revascularization. By deploy-
ing a covered stent after recanalization, the original investiga-
tors using these devices were more commonly able to achieve
a successful reconstruction (Fig. 47.6). Effective exclusion of
the damaged and dissected arterial flow channel was thought
to contribute to the improved patency. In addition, the success
was due in part to the ability to vigorously postdilate the re-
canalized tract after deployment of the stent-graft. This was
possible due to the protection from perforation provided by
the stent's covering. 20 ' 27

Additional experimental work focusing on stent-grafts
explanted from human subjects after treatment of aortoiliac
and femoral arterial occlusions has suggested that covered
stents result in a reduced proliferative response to injury.
The extent of vascular smooth muscle cell proliferation ob-
served in the displaced atherosclerotic lesions and in the
surrounding media was noted to be limited. 28 This finding
is in direct contrast to the exaggerated hyperplastic response
to injury seen after PTA and uncovered stent placement
(Fig. 47.7).

The mechanisms that may underlie the reduced hyperplas-
tic response have not been fully elucidated. However, some

546

chapter 47 Aortoiliac endovascular recanalization compared with surgical reconstruction

Figure 47.6 Arteriographic image of left iliac artery occlusion treated by
placement of a stent-graft. (A) The preoperative arteriogram demonstrates
complete occlusion of the leftdistal common and external iliacartery. (B)The

occluded segment was successfully recanalized and lined with a stent-graft,
restoring arterial perfusion.

investigators have suggested that elimination of direct
perfusion of the disrupted intima and media by placement of a
covered stent may prevent activation of circulating factors
within the blood such as platelet-derived growth factor. This
in turn may diminish the proliferative response these factors
would otherwise provoke. This relative reduction in the hy-
perplastic response after PTA with stent-graft placement may
in part account for the superior short- and intermediate-term
patency observed when Wallgrafts were compared with
Wallstents in a randomized prospective fashion. 25

Technique

The initial step in the treatment of aortoiliac occlusive disease
with endovascular stent-grafts is to achieve passage of an an-
giographic wire through the occluded arterial segment. This
may be accomplished in an antegrade fashion through the
contralateral iliac artery, using the contralateral femoral artery
for arterial access. The brachial artery may also be used for
arterial access in performing antegrade recanalization.

Figure47.7 Histological appearanceof a human iliac artery treated by
recanalization and stent-graft deployment. The light micrograph
demonstrates extensive native atherosclerotic plaque (P) displaced by the
stent-graft. Inset. The plaque demonstrates only limited PC-1
immunoreactivity (arrows), indicating minimal vascular smooth muscle
proliferation.

547

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 47.8 Schematic diagram indicating the strategy for endovascular
treatment of iliac artery occlusive disease.

Alternatively, retrograde recanalization may be carried out.
Retrograde recanalization is performed using access in the
ipsilateral femoral artery distal to the iliac occlusion. The
use of a hydrophilic coated wire supported by a hydrophilic
directional catheter may facilitate passage of the wire through
the occluded arterial segment.

Wire passage frequently may occur in a subintimal plane
within the arterial wall. In these instances it is necessary to
achieve return passage of the wire into the arterial lumen be-
yond the occlusion in order to allow recanalization and
stent-graft placement. If a recanalization is to be performed
through an arterial cut down, passage of the wire in an ante-
grade fashion is desirable. In these instances reentry into the
luminal plane can be facilitated during open exposure and ar-
teriotomy during the procedure. When retrograde passage of
the hydrophilic wire is utilized, reentry into the arterial lumen
must be achieved proximal to the occlusive lesion, frequently
in the distal aorta.

Once wire passage through the occluded segment has been
accomplished, diffuse dilation of the arterial tract is per-
formed (Fig. 47.8). Angioplasty balloons are used to accom-
plish dilation with the balloon diameter size being determined
by the goal recanalization diameter to be achieved (Fig. 47.9).
An 8-mm balloon is frequently utilized for this purpose. After
diffuse dilation is performed, the stent-graft is advanced
through the recanalized segment. The stent-graft is then de-
ployed in position and postdilation is carried out (Fig. 47.10).
Aggressive postdilation may be used to resolve residual areas
of stenosis completely since the graft covering affords relative
protection from arterial rupture and hemorrhage.

Management of distal endpoint

The management of the distal endpoint of the stent-graft after
recanalization is dependent on the extent of the occlusive dis-
ease and the involvement of additional levels of arterial occlu-
sion. In instances where the iliac disease extends into the
external iliac artery to the level of the inguinal ligament, open

Figure 47.9 Schematic diagram of balloon angioplasty procedure. After a
passage through the occluded segment has been accomplished with the
angiographic wire, diffuse angioplasty is used to recanalize the segment.

exposure of the femoral artery is required. In these instances,
the end of the stent-graft may be terminated in the common
femoral artery if necessary and endoluminal anastomosis per-
formed (Fig. 47.11). Similarly, if large-diameter stent-grafts
are employed a similarly large introducer sheath is required.
This may necessitate direct open exposure of the femoral
artery as well. The use of an arteriotomy for device deploy-
ment also allows the operator to vary the distal anastomotic
site depending on the local pattern of disease. 29 Closure assis-
tance devices may also be employed for larger introducer

548

chapter 47 Aortoiliac endovascular recanalization compared with surgical reconstruction

Figure 47.10 Schematic diagram of stent-graft after deployment through
the recanalized segment. In this instance an open arteriotomy is depicted for
arterial access.

sheaths introduced percutaneously with a considerable de-
gree of success. Conversely, smaller diameter commercially
produced covered stents require smaller introducers (8 or 9 Fr
diameter) and may often be deployed percutaneously

Use with adjunctive inf rainguinal
reconstruction

Additional levels of arterial occlusive disease are commonly
present in patients who present with aortoiliac occlusion and
limb-threatening ischemia. In these cases adjunctive outflow
inf rainguinal reconstruction is often necessary to achieve limb
salvage. Patients with multilevel arterial occlusive disease are
typically older with more advanced disease and additional
manifestations of comorbid medical conditions. They
may therefore be more safely treated using less invasive tech-
niques, particularly endovascular recanalization of the aor-
toiliac segment. The inflow provided by the endoluminal
aortoiliac stent-graft may be effectively used as the source for
the infrainguinal reconstruction. Use of an endovascular
stent-graft provides direct flow from the aorta while avoiding
dissection of alternative arterial beds not involved with occlu-
sive disease such as the axillary or contralateral femoral artery.
After aortoiliac recanalization and stent-graft placement,
outflow bypass may be performed to the appropriate target
artery at the popliteal, tibial, or pedal level (Fig. 47.12). Both
autogenous and prosthetic conduits have been used success-
fully in conjunction with stent-graft recanalization of the
aortoiliac segment. 30 Outflow reconstruction in patients with
infrainguinal occlusive disease is also likely to increase the
long-term patency rates of the inflow procedure as is evi-
denced by the increased limb occlusion rates for aortofemoral
grafts performed to diseased outflow vessels. 13

Figure 47.11 Management of the stent-graft
endpoint. (A) The endpoint of the stent-graft
can be managed by performing an endoluminal
anastomosis in the common femoral artery if no
outflow disease is present. (B) Photograph of
the end of the stent-graft as it emerges from
the femoral arteriotomy.

549

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 47.1 2 Alternative schemes for
management of the stent-graftendpoint. The
aortoiliac stent-graft may be used as the inflow
site for additional vascular reconstruction
including direct anastomosis to the profunda
femoris (A), sequential anastomosis to the
profunda femoris and to a distal target vessel
(B), direct anastomosis to a distal target vessel
(C), or as inflow for a femorofemoral bypass
graft (D).

Results of stent-graft treatments

Reports regarding the use of stent-grafts and covered stents
for the treatment of aortoiliac occlusive disease now encom-
pass over 10 years of experience. 5,6/19,20/22/24/25/27 Initially, the de-
vices utilized for stent-grafting after iliac recanalization were
all physician-made. The large profile of these devices required
open exposure of the femoral artery and arteriotomy to enable
their delivery and deployment. 5 In addition, because limb sal-
vage was the most common indication for intervention in these

initial reports, the iliac recanalization was commonly per-
formed in conjunction with infrainguinal reconstruction. 30

In midterm follow-up of 52 patients treated for limb salvage
with physician-made grafts between 1993 and 1997, the 4-year
primary and secondary patency rates were 66.1% and 72.3%
with an associated limb salvage rate of 88.7% (Fig. 47.13). In
this population, concomitant infrainguinal bypass was per-
formed in 52% of the patients. These results were achieved
in a patient population with extensive comorbid medical
illnesses. 6 The perioperative mortality rate was 6% while the
morbidity rate was 15%. These results were significantly

550

chapter 47 Aortoiliac endovascular recanalization compared with surgical reconstruction

1000
».&
».D
7109

«W

30.D

HhD

100

— » - S*ccnd*r)r Potency

24 3D »

Months

Figure 47.13 Patency, limb salvage, and survival after treatment of
aortoiliac occlusive disease using a physician-made stent-graft.

better than those achieved with angioplasty and stenting
alone for aortoiliac occlusion and added support to the use of
stent-grafts for aortoiliac recanalization.

Increasing use of commercially produced stent-grafts or
covered stents for the treatment of aortoiliac occlusive disease
has been observed as these devices have become more widely
available. 21/24/25 Longer term follow-up has currently been re-
ported for patients treated in European centers where the de-
vices first became available for clinical use. A multicenter trial
using the self-expanding Viabahn or Hemobahn stent-graft
for the treatment of iliac artery occlusion (Fig. 47.14) has been
reported. 5 The investigators used the stent-graft in occlusions
in 61 iliac arteries in patients with either lifestyle-altering clau-
dication or chronic, critical lower limb ischemia. The authors
report 1-year primary patency rates of 91.2% and secondary
patency rates of 94.7% (Fig. 47.15). The higher patency rates in
this study may reflect the use for nonlimb salvage indications
and the absence of need for concomitant infrainguinal bypass.
The morbidity rate was 17% with no perioperative mortality.

The Wallgraft has also been employed in the treatment of
aortoiliac occlusions. 25 In these patients, the primary patency
rate was found to be 86% at 1 year and 82% at 2 years. Longer
lesion length and limited outflow were found to be predictors
of primary failure in this group. Overall these results
have been viewed as encouraging for the continued use of
endovascular techniques for the treatment of aortoiliac
occlusive disease.

Future considerations

Figure 47.14 Endoluminal reconstruction of bilateral external iliac artery
occlusions. (A) The preoperative arteriogram demonstrates complete
occlusion of the external iliac arteries bilaterally. (B) After bilateral
recanalization and Viabahn-covered stent-graft deployment (arrows),
arterial flow is restored.

Additional research to improve further the performance of
stent-grafts for the treatment of arterial occlusive disease is
currently being focused on reducing the profile of the delivery
system. Reduction in device profile will enable wider use of

percutaneous approaches for these patients and will reduce
the requirement for arterial closure-assistance devices. Work
continues to focus on bonding chemically active agents to the
stent-grafts themselves in an effort to reduce the incidence of

551

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

0.E-

p 0,7 H

aft-

o.s

120

I SO

241)

300

PtfiOdays

PP CI

sp 61

Figure 47.1 5 Primary patency (pp) rates and secondary patency (sp) rates
for iliac artery occlusive disease after treatment using the Viabahn-covered
stent.

restenosis after stent-graft placement. In addition, covered
stents and stent-grafts have been used increasingly in the
femoral artery. Continued improvements in technology and
adjuvant pharmacological therapies will probably allow for
increased application to infrainguinal occlusive disease.

Summary

The utilization of endovascular stent-grafts for the treatment
of aortoiliac occlusive disease has increased significantly as
the devices have become more widely available. Stent-grafts
and covered stents can now be used in conjunction with iliac
recanalization with excellent intermediate patency and limb
salvage rates. Combined use with infrainguinal reconstruc-
tion remains an attractive option for patients with multilevel
arterial occlusive disease. Finally, the used of these endovas-
cular techniques has been associated with significant reduc-
tions in morbidity and mortality and has allowed for the
application of this limb-salvaging therapy in patients with
considerable comorbid medical conditions.

References

1. Marin ML, Veith FJ, Sanchez LA et al. Endovascular repair of
aortoiliac occlusive disease. World J Surg 1996; 30:679.

2. Brewster DC. Current controversies in the management of aortoil-
iac occlusive disease. / Vase Surg 1997; 25:367.

3. Johnston KW, Rae M, Hogg-Johnston SA et al. Five-year results of
a prospective study of percutaneous transluminal angioplasty.
Ann Surg 1987; 206:403.

4. Powell RJ, Fillinger M, Walsh DB, Zwolak R, Cronenwett JL. Pre-
dicting outcome of angioplasty and selective stenting of multiseg-
ment iliac artery occlusive disease. / Vase Surg 2000; 32:564.

5. Lammer J, Dake MD, Bleyn J et al. Peripheral arterial obstruction:
prospective study of treatment with a transluminally placed self-
expanding stent graft. Radiology 2000; 217:95.

6. Wain RA, Veith FJ, Marin ML et al. Analysis of endovascular graft
treatment for aortoiliac occlusive disease: what is its role based on
midterm results? Ann Surg 1999; 230:145.

7. Hughson WG, Mann JI, Garrod A. Intermittent claudication:
prevalence and risk factors. Br] Med 1978; 1:1379.

8. Weiss NS. Cigarette smoking and arteriosclerosis obliterans: an
epidemiologic approach. Am] Epidemiol 1972; 95:17.

9. Vogt MT, Wolfson SK, Kuller LH. Segmental arterial disease in the
lower extremities: correlates of disease and relationship to mortal-
ity. / Clin Epidemiol 1993; 46:1267.

10. Crawford ES, Bomberger RA, Glaeser DH et al. Aortoiliac occlu-
sive disease: factors influencing survival and function following
reconstructive operation over a 25-year period. Surgery 1981;
90:1055.

11 . Szilagyi DE, Elliott JR Jr, Smith RF et al. A 30-year survey of the re-
constructive surgical treatment of aortoiliac occlusive disease. /
Vase Surg 1986; 3:421.

12. Nevelsteen A, Wouters L, Suy R. Aortofemoral Dacron reconstruc-
tions for aortoiliac occlusive disease: a 25-year survey. Eur J Vase
Surg 1991; 5:179.

13. Brewster DC. Clinical and anatomic considerations for surgery in
aortoiliac disease and results of surgical treatment. Circulation
1991;83(Suppl.I):42.

14. Rutherford RB, Patt A, Pearce WH. Extra-anatomic bypass: a clos-
er view. / Vase Surg 1987; 6:437.

15. Rutherford RB. Axillobifemoral bypass: current indications, tech-
niques and results. In: Veith FJ, ed. Critical Problems in Vascular
Surgery, Vol. 7. St Louis: Quality Medical, 1996.

16. Palmaz JC, Laborde JC, Rivera FJ et al. Stenting of the iliac arteries
with the Palmaz stent: experience from a multicenter trial. Cardio-
vasc Interv Radiol 1992; 15:291.

17. Martin EC, Katzen BT, Benenati JF et al. Multicenter trial of the
Wallstent in the iliac and femoral arteries. / Vase Interv Radiol 1995;
6:843.

18. Bosch JL, Hunink MGM. Meta-analysis of the results of percuta-
neous transluminal angioplasty and stent placement for aortoiliac
occlusive disease. Radiology 1997; 204:87.

19. Ramaswami G, Marin ML. Stent grafts in occlusive arterial dis-
ease. Surg Clin North Am 1999; 79:597.

20. Marin ML, Veith FJ, Cynamon J et al. Transfemoral endovascular
stented graft treatment of aorto-iliac and femoropopliteal occlu-
sive disease for limb salvage. Am J Surg 1994; 168:156.

21. Gray BH, Sullivan TM. Aortoiliac occlusive disease: surgical
versus interventional therapy. Curr Interv Cardiol Rep 2001;
3:109.

22. Volodos NL, Shekhanin VE, Karpovich IP, Troian VI, Gur'ev IuA.
Self-fixing synthetic prosthesis for endoprosthetics of the vessels.
VestnKhir 1986; 137:123.

23. Cragg AH, Dake MD. Percutaneous femoropopliteal graft place-
ment. / Vase Interv Radiol 1993; 4:455.

24. Rubin BG, Sicard GA. The Hemobahn endoprosthesis: a self-
expanding polytetrafluoroethylene-covered endoprosthesis for
the treatment of peripheral arterial occlusive disease after balloon
angioplasty. / Vase Surg 2001; 33:S124.

25. Krajcer Z, Sioco G, Reynolds T. Comparison of Wallgraft and Wall-
stent for treatment of complex iliac artery stenosis and occlusion.

552

chapter 47 Aortoiliac endovascular recanalization compared with surgical reconstruction

Preliminary results of a prospective randomized study. Tex Heart
Inst J 1997; 24:193.

26. Henry M, Amor M, Ethevenot G, Henry I, Mentre B, Tzvetanov K.
Percutaneous endoluminal treatment of iliac occlusions: long-
term follow-up in 105 patients. JEndovasc Surg 1998; 5:228.

27. Lacroix H, Stockx L, Wilms G, Nevelsteen A. Transfemoral treat-
ment for iliac occlusive disease with endoluminal stent-grafts. Eur
JEndovasc Surg 1997; 14:204.

28. Marin ML, Veith FJ, Cynamon J et al. Human transluminally
placed endovascular stented grafts: preliminary histopathologic

analysis of healing grafts in aortoiliac and femoral artery occlusive
disease. / Vase Surg 1995; 21:595.

29. Wain RA, Lyon RT, Veith FJ et al. Alternative techniques for man-
agement of distal anastomoses of aortofemoral and iliofemoral
endovascular grafts. / Vase Surg 2000; 32:307.

30. Marin ML, Veith FJ, Sanchez LA et al. Endovascular aortoiliac
grafts in combination with standard infrainguinal arterial by-
passes in the management of limb-threatening ischemia: prelimi-
nary report. / Vase Surg 1995; 22:316.

553

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Endovascular stent-graft repair of
thoracic aortic aneurysms and dissections

Jason T. Lee
Rodney A. White

Diseases of the thoracic aorta pose a challenging problem
to vascular and cardiothoracic surgeons. Thoracic aortic
aneurysms (TAAs), acute and chronic type B dissections, pen-
etrating aortic ulcers, and traumatic aortic transection all can
present in the acute or chronic setting. Conventional therapy
requires open thoracotomy, and these patients are typically
elderly and medically debilitated, carrying comorbid
conditions including hypertension, coronary artery disease,
obstructive pulmonary disease, and congestive heart failure.
These comorbidities can significantly worsen surgical out-
comes or even preclude open repair. Owing to the high opera-
tive risk involved with open surgery and with the advent of
endoluminal technology, thoracic endografting has emerged
as a viable and attractive treatment alternative.

The incidence of TAAs is estimated to be as high as 10 cases
per 100 000 people per year, and aortic dissection affects 9000
patients per year in the United States alone. 1-4 Penetrating ul-
cers of the thoracic aorta are a more rare clinical entity and may
develop into a chronic dissection or aneurysm at high risk for
rupture if left untreated. 5 ' 6 Traumatic aortic tears, a disease
typically of younger patients, are encountered in up to 18% of
motor vehicle accidents, with mortality rates of 90% in the
field, and 40-70% in those who survive transport to a trauma
center. 7,8 For each of these thoracic pathologies there are dis-
tinct advantages to the endovascular approach and all patients
with these very demanding illnesses should be considered
candidates for thoracic endografting. In this chapter we will
discuss the natural history and conventional open repair of
thoracic aneurysms and type B dissections, describe the diag-
nostic work-up and operative technique for endovascular in-
terventions in the thoracic aorta, and review the results of
thoracic endografting from the worldwide literature.

Conventional open repair of thoracic
aorta pathology

Descending thoracic aneurysms

The natural history of untreated aneurysmal disease of the

thoracic aorta includes progressive expansion, increasing risk
of rupture, and ultimately death. Actuarial 1- and 5-year sur-
vivals for patients without intervention are estimated to be
60% and 20%, respectively. 9 The yearly risk of any occurrence
of rupture, dissection, or death in a patient with a thoracic
aneurysm >6cm in diameter is over 14%. 10 As is the case with
abdominal aortic aneurysms (AAAs), the risk of rupture in-
creases with size, and the 5-year rupture rate is fivefold higher
for thoracic aneurysms >6 cm in diameter. 11 Because of the risk
of lethal rupture, all patients with descending thoracic
aneurysms should be evaluated for potential operative repair.
Surgical indications include urgent operation in symptomatic
patients who present with signs of rupture, chest or back pain,
hemoptysis, hematemesis, or cardiovascular collapse. In
asymptomatic patients, risk/benefit analysis in a large popu-
lation study supports that thoracic aneurysms >6.5cm be
repaired electively, with the threshold for Marfan's disease
or familial thoracic aortic aneurysm being >6 cm. 10

Traditional open surgical repair of TAAs involves aortic
graft replacement via a left thoracotomy and has been found to
improve survival when compared with medical therapy. 12 De-
spite dramatic advances in the technical expertise for perform-
ing these complex thoracic aortic operations utilizing distal
perfusion methods and spinal cord protection, open surgery
remains a high-risk endeavor, especially given the frequent co-
morbid cardiovascular and pulmonary disease. 13 ' 14 Operative
mortality rates from centers of excellence are reported be-
tween 8% and 20% for elective cases and up to 60% for emer-
gency operations. 15-18 Survivors of open repair of thoracic
aneurysms further suffer from morbidity rates of up to 50% re-
lated to renal, intestinal, and spinal cord ischemia that sub-
stantially limit functional recovery and long-term survival,
with recent 5-year survival rates reported to be about
60-70%. 18 - 20

Type B aortic dissections

Although the surgical indications for TAA repair are relatively
straightforward because they are based on size and symp-

554

chapter 48 Endovascular stent-graft repair of thoracic aortic aneurysms and dissections

toms, the optimal therapy for type B aortic dissection remains
controversial. 21 At most institutions, aggressive antihyperten-
sive medical therapy with (3-blockers is utilized in an intensive
care unit (ICU) setting. Operative intervention for acute type B
dissections is reserved when life-threatening complications
arise, such as progression of the dissection, recent expansion,
end-organ ischemia caused by side-branch compromise, on-
going pain, uncontrollable hypertension, or rupture. Open
surgical interventions are varied and involve a combination of
graft replacement, closure of the open entry site, or fenestra-
tion to create a reentry tear. 22 This selective, or "complication-
specific," approach has yielded early mortality rates of 20% for
medically treated patients, 35% for surgically repaired pa-
tients, and over 50% for operative patients presenting with
end-organ ischemia. 23-25 Long-term survival is limited in
those treated surgically or medically. Actuarial survival rates
are reported by the Stanford group to be 56%, 48%, 29%, and
11% at 1, 5, 10, and 15 years, respectively 26 More recent series
of well selected patients with acute type B dissections and
the aggressive use of distal perfusion, cerebrospinal fluid
drainage, and hypothermic circulatory arrest have slightly im-
proved on early mortality and long-term survival but still
demonstrate significant morbidity (47%). 27 The poor results
for both medical therapy and open surgery and lack of signifi-
cant improvement over the past 30 years have prompted the
search for alternative therapies for acute type B dissections.

Further adding to the treatment dilemma of type B dissec-
tion patients are those that undergo successful nonoperative
antihypertensive therapy and subsequently develop a
subacute or chronic dissection. The natural history of these
asymptomatic patients is that they will develop proximal
aneurysmal dilation and therefore place the patient at risk for
future rupture. There has been some suggestion those patients
should be offered surgical intervention before developing
aortic enlargement. 28,29

Penetrating aortic ulcers

Penetrating ulcers of the thoracic aorta are a rare clinical entity
distinct from classic type B aortic dissection that is character-
ized by rupture of an atherosclerotic plaque through the inter-
nal elastic lamina. The natural history of this ulcerative
process is ill-defined and can lead to pseudoaneurysm forma-
tion, localized dissection, embolization, or rupture. 30 Treat-
ment in the acute setting is similar to that of dissection, namely
antihypertensive medications and afterload-reducing agents
in the ICU setting. Persistent pain, recurrent pain, hemody-
namic collapse, and rapidly expanding aortic diameters are all
indications for surgical intervention, which involves interpo-
sition graft repair. 31 Rupture risks are higher for penetrating
aortic ulcers than for type B dissections (40% vs. 4%) empha-
sizing the need for accurate recognition and treatment of the
pathology 5 Conventional open repair is burdened by the same
morbidity and mortality that faces the aneurysm or dissection

patient undergoing thoracotomy, and the use of endovascular
stents has been suggested to be an attractive alternative
therapy 32

Traumatic aortic rupture

Severe blunt chest trauma can lead to aortic transection or rup-
ture, most commonly occurring in the descending thoracic
aorta at the transition from the mobile distal aortic arch to the
posteriorly bound isthmus. This injury is commonly fatal,
with 85% of patients dying at the scene and 20-30% further
mortality in those patients arriving in emergency room with
vital signs and able to undergo operative intervention. 33 Sig-
nificant morbidity is encountered in patients who survive as a
result of accompanying head injury, abdominal solid-organ
injuries, and pelvic /orthopedic fractures. Paraplegia is the
most devastating complication and is related to aortic cross-
clamp time during the "clamp and sew" technique. High-
volume trauma centers have recently adopted distal perfusion
techniques with passive shunts and partial heart bypass and
have decreased paraplegia rates significantly, but without
much improvement on overall mortality 8

Endovascular stent-graft repair

With the successful development and subsequent refinement
of endovascular techniques to treat AAAs, there has been a
concomitant effort to adapt this technology to the treatment of
thoracic aortic pathology. Endovascular stent-grafting in pa-
tients with descending thoracic aneurysms, type B dissec-
tions, penetrating ulcers, and traumatic aortic rupture has
demonstrated promising short- and midterm results. Pro-
posed advantages of the less invasive endograft approach
include shorter operative time, decreased need for general
anesthesia, lack of aortic cross-clamping, avoidance of car-
diopulmonary bypass, avoidance of major thoracic or thora-
coabdominal incisions, less pain, quicker recuperation, and
shorter hospital and ICU stays. Many patients previously
turned down for open repair because of medical comorbidities
are now routinely referred and treated with endovascular
stent-grafting.

Most reported series, including our own, have documented
high technical success along with major reductions in morbid-
ity and mortality. The improvements in patient outcome are
even more encouraging because the majority of patients in the
early experience had already been declined open surgical
repair because of significant comorbid medical illnesses. The
application of this new technology now is shifting toward
offering endovascular stenting as an attractive treatment
alternative in patients with TAAs, type B dissections, and pen-
etrating ulcers. As experience has been gained at multiple cen-
ters for elective or urgent repair of these pathologies, emergent
stent-graft repair has been performed in the trauma patient

555

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Table 48.1 Criteria for endovascular repair of thoracic aortic disease at Harbor-UCLA Medical Center.

Thoracic aortic aneurysm

Type B aortic dissection

Descending thoracic aneurysm >5.5 cm

Aneurysm 4.5-5.5 cm with increase in size by 0.5 cm in last
6 months or twice normal size

Saccular aneurysm or penetrating ulcer

Nonaneurysmal proximal and distal aortic neck measures between
22 and 40 mm (dependent on device availability)

No extension of aneurysm into abdominal aorta (distal neck at least 2 cm
above celiac)

Acute dissection with intractable pain, uncontrollable hypertension,
progression of dissection, or end-organ ischemia

Chronic dissection with aneurysmal dilation of proximal descending aorta

Chronic dissection with acute symptoms

Entry tear at least 1 cm from left subclavian orifice (potentially 2 cm if plan to
cover subclavian)

No entry site of dissection that is proximal to subclavian or involves arch or
ascending portion of aorta

Devices available that are suitable for patient's anatomy

Patent iliac or femoral arteries that allow introduction of 22-25 Fr delivery

sheath (device dependent)
Life expectancy at least 6 months
Consent for appropriate trials and follow-up protocols

with contained aortic rupture. These patients typically have si-
multaneous intraabdominal or head injuries and may not be
candidates for thoracotomy and cardiopulmonary bypass,
and early results of selected patients treated with a stent-graft
have indicated improved mortality Like any new technology,
however, much remains to be clarified with regard to the
proper applications and long-term durability of thoracic
endografting.

Methods and techniques

All patients referred for endovascular repair of thoracic aortic
lesions at our institution are evaluated using spiral computed
tomography (CT)-angiography including three-dimensional
(3-D) reconstruction in a computerized interactive environ-
ment utilizing Preview™ software [Medical Media Systems
(MMS), West Lebanon, NH, USA]. For elective repair, the ac-
curate preoperative aortic measurements and assessment of
the proximal and distal landing zones aid in the selection of the
proper device. 34 The distal vascular access also needs to be of
sufficient caliber to allow passage of the devices. Both CT-
angiography and the Preview™ reconstruction can allow
rapid determination of the distal vascular access and an esti-
mation of the tortuosity of the aorta. Obviously, time restraints
in an acute rupture, leaking aneurysm or dissection, or trauma
patient occasionally will preclude the typical 24-h turnover
time required by MMS to create the 3-D model.

Current criteria utilized at our institution for identifying
patients able to undergo thoracic endografting are listed in
Table 48.1. These criteria have evolved over the past 5 years
and continue to be updated as there have been improvements
and changes in both stent-graft design and delivery systems.
If the patient is deemed to be a suitable candidate for endovas-

cular repair, he/she is prospectively enrolled into a Food and
Drug Administration-approved Investigational Devices Ex-
emption trial. All protocols as well as the pre- and postopera-
tive surveillance and imaging are in compliance with the
Institutional Review Board of our institution. We have pre-
ferred to use the self-expanding AneuRx and Talent (Medtron-
ic AVE, Santa Rosa, CA, USA) devices, and the Gore Thoracic
Excluder device (WL Gore, Inc., Flagstaff, AZ, USA).

Procedures are performed in a specially designed endo-
vascular operating suite equipped with ceiling-mounted
fluoroscopy capable of digital subtraction angiography. The
operating team consists of two vascular surgeons, an interven-
tional radiologist, and a cardiothoracic surgeon on stand-by.
All patients are placed in the supine position with a right arm
arterial blood pressure monitor and prepped for possible tho-
racotomy and cardiopulmonary bypass. Under local anesthet-
ic with minimal intravenous sedation, the common femoral
artery is exposed and isolated using Roummel tourniquets. If
the vessel is too small to accommodate the introducer sheath,
the common iliac artery is reached via retroperitoneal
approach while converting to general anesthesia. Heparin
(lOOU/kg) is administered before the arterial puncture.
Through the femoral artery puncture, appropriately sized
catheter sheaths are introduced to allow a 0.025 Microvina or
0.035 Amplatz superstiff wire to be guided up to and around
the aortic arch.

We routinely utilize intravascular ultrasound (IVUS) to con-
firm the side-branch anatomy, to measure the diameter and
length of the proximal and distal neck, and to verify optimal
placement of the stent-graft postdeployment. 35 For cases of
aortic dissection, we have found the IVUS with color-flow ca-
pabilities to be invaluable in searching for and confirming cov-
erage of the entry or reentry site (Fig. 48.1; also in colour, see

556

chapter 48 Endovascular stent-graft repair of thoracic aortic aneurysms and dissections

Figure 48.1 Intravascular ultrasound image
of leaking pseudoaneurysm of previously
repaired ruptured thoracic aortic aneurysm.
Note the color flow of the obvious breakdown
of the proximal anastomosis. This patient was
treated with a stent-graft that covered the entry
site, and the patient's symptoms of pain and
hemothorax resolved. See also Plate 7, facing
p. 370.

Figure 48.2 (A) Intravascular ultrasound
images of distal pseudoaneurysm after
previously repaired descending thoracic aortic
aneurysm. The large blow-out of the distal
anastamosis (dashed arrow) is seen clearly on
the pull-back image. (B) Angiogram obtained
after stent-graft placed demonstrating no
further leakage.

Plate 7, facing p. 370). Pull-back reconstructions using IVUS
accurately determine the level of where the blow-out site is at
the neck of an aneurysm or pseudoaneurysm to allow optimal
placement of the stent-graft (Fig. 48.2). Contrast angiography
is kept to a minimum during the procedure, which is obvious-
ly important when treating a patient with renal compromise.
We have also been able to perform an endovascular AAA
repair in a patient with severe iodine contrast allergy under
the sole guidance of IVUS.

After IVUS interrogation and selection of the properly sized
device, we proceed with introduction of the delivery system. It
is carefully advanced to the optimal position under fluoro-
scopic guidance, usually immediately distal to the left subcla-
vian artery. Just before the device is released, adenosine is
administered intravenously (up to 32 mg) to transiently
induce ventricular asystole and reduce the risk that arterial
pulsations can cause the stent-graft to migrate while being de-
ployed. We oversize the proximal neck by 10-20% for cases of
TAAs and rely on at least 20 mm of length in order to achieve

the optimal fixation. In dissections we use the native size of the
normal aorta for sizing of the proximal neck. If a proximal neck
of 20 mm is not possible, we have covered the subclavian artery
with both the stent struts and /or fabric, and postoperatively
follow the patients clinically for any signs of left upper extrem-
ity or vertebrobasilar ischemia. We have not yet had to perform
carotid-subclavian bypass after thoracic endografting since
there is such a rich potential collateral blood supply around
the subclavian artery, and this is in agreement with others. 36
Confirmation angiogram is performed to assure there is no
proximal or distal endoleak and that the device is fully op-
posed to the aortic wall. If such an endoleak is present or if the
device is not fully expanded, balloon dilation of the stent-graft
is performed or an additional cuff is placed. Upon verifying
successful deployment, the sheaths are removed and primary
repair of the common femoral arteriotomy is performed.

At the conclusion of the procedure, the patient is observed
overnight in the ICU for close blood pressure monitoring and
then transferred to the surgical ward for an average length of

557

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

F -jg^

.J. * ^

; *k.

1^.

r j

-*.

^■^|6 mo |

Figure 48.3 Sequential reconstructions of a
symptomatic patient (chest and back pain) with
an 8-cm thoracic aortic aneurysm after
exclusion. By 6-month follow-up, total
aneurysm volume decreased from 487 cm 3 to
282 cm 3 with decrease in diameter and
resolution of symptoms. At 2-year follow-up,
aneurysm volume was no longer decreasing,
and an endoleak was detected atthe junction of
two overlapping pieces. This was treated with
an in-line cuff and the patient remains
asymptomatic. See also Plate 8, facing p. 370.

stay of 2-3 days. Follow-up spiral CTs are obtained immedi-
ately postoperatively, after 1 month, and every 6 months there-
after. Figure 48.3 (also in colour, see Plate 8, facing p. 370)
demonstrates a patient with serial 3-D reconstructions of a
successfully excluded thoracic aneurysm and the midterm
imaging follow-up showing a successfully treated endoleak.

Results of thoracic endograft repair

Descending thoracic aneurysms

From 1992 to 2004 there have been 26 reported series of pa-
tients undergoing endovascular treatment of TAAs and 25 re-
ported series of type B dissections. 37-79 Although the patient
populations are very diverse and there is some amount of se-
lection bias that cannot be avoided because this is an evolving
technology, the early results are extremely favorable. The aim
of stent-graft placement for TAAs is to achieve aneurysmal
exclusion and impede further pressurization of the sac, and
therefore lead to protection from fatal rupture. A summary of
the data from the published trials 37-62 of endovascular repair
of TAAs is presented in Table 48.2.

One of the more striking findings in the review of the world-
wide literature is the high technical success rate reported, es-
pecially in the later series as more experience has been gained.
This highlights the importance of proper patient selection, as
most of these trials had strict entry criteria to ensure that
patients' anatomy would allow stent-graft treatment. Some

of the earlier studies from Stanford, 39 Mount Sinai, 40 and
Cleveland Clinic 43 that did not achieve at least 90% technical
success acknowledge that the failures occurred early in their
experience, when patient selection criteria and technical ex-
pertise were not as streamlined as they are currently. Precise
positioning of the endovascular grafts was cited as the major
difficulty that led to technical failure in the early experience
and adjunctive measures of blood pressure lowering, 52 adeno-
sine arrest, 58 and even induction of controlled ventricular
fibrillation were all experimented with to allow accurate
deployment. 40 This seemed to be a problem mainly of first-
generation and balloon-expandable devices, and the superior
design of second-generation devices has obviated the need for
such intraoperative maneuvers. 47

To maximize further the likelihood of technical success, ac-
curate preoperative imaging and treatment planning is vital.
The important factors are the location and morphology of the
aneurysm, the distal vascular access, and the tortuosity of the
aortic and thoracic aorta. 41 Recognizing unsuitable anatomy
such as excessive tortuosity of the aortic arch or inadequate
distal access has been cited to be the first point of the learning
curve for thoracic endografting. 54 Spiral CT-angiography is
utilized to locate an appropriate proximal and distal landing
zone, as well as note the access in the groins. Severe iliac tortu-
osity or femoral arteries smaller than 7 mm typically prompt
us to consider placing an iliac conduit via retroperitoneal ex-
posure, and in the Arizona Heart Institute series 13% of pa-
tients with TAAs required such an approach. 48 We found in
our series that performing an iliac conduit during AAA endo-

558

chapter 48 Endovascular stent-graft repair of thoracic aortic aneurysms and dissections

Table 48.2 Comparison of 26 series of thoracic aneurysms treated with endovascular devices.

No. of

Procedural

Paraplegia

Length
of stay

30-day

Series

Device

patients

success

Other morbidity

(days)

mortality

Endoleak

F/U

Dake 37 1992-1 994

Homemade

100%

4.8

15%

Ehrlich 38 1997

Talent

100%

10%

20%

Mitchell 39 1992-1 997

Homemade

103

73%

25%

24%

Temudom 40 1997-1998

Vanguard/Gore

78%

14%

2.9

14%

Grabenwoger 41 1996-1999

Talent/Gore

100%

9.5%

9.8

9.5%

n/a

Taylor 42 1 997-2000

AneuRx/Gore

100%

4.3%

8.7%

13%

Greenberg 43 1993-1997

Cook

88%

n/a

25%

12%

Bortone 44 1 999-2000

Gore

100%

n/a

White 45 1997-1 999

AneuRx

94%

12%

Won 46 1 994-1 999

Taewoong

100%

n/a

Cambria 47 1996-2001

Cook/Gore

100%

28%

n/a

5.5%

21%

Thompson 48 2000-2001

Gore

100%

23%

Totaro 49 2000-2001

Gore

100%

30%

Najibi 50 1999-2000

Gore/Talent

95%

16%

Criado 51 1999-2002

Talent

97%

15%

n/a

13%

Herold 52 1999-2001

Talent

100%

Chabbert 53 1997-2001

Talent/Gore

100%

n/a

21%

25%

Fattori 54 1997-2002

Talent

94%

16%

Scharrer-Pamler 55 1997-2002

Talent/Gore

100%

18%

Lamme 56 1998-2002

Gore/Talent

100%

17%

11%

Lepore 57 1999-2001

Gore/Talent

100%

19%

n/a

10%

19%

Krohg-Sorensen 58 2000-2002

Gore/Talent

100%

11%

n/a

11%

Lambrechts 59 2000-2002

Talent/Gore

100%

19%

25%

n/a

Ellozy 60 1998-2002

Talent/Gore

90%

14%

n/a

Czerny 61 1996-2002

Talent/Gore

94%

n/a

29%

Melissano 62 2002

Endomed

100%

11%

n/a

33%

n/a

Other morbidity refers to cardiopulmonary, renal, infectious, and neurologic complications. Endoleaks include those found during follow-up computed
tomography scanning and requiring secondary intervention for resolution. F/U indicates length of mean follow-up in months for each individual series.

grafting did not adversely affect overall procedural success or
long-term morbidity and mortality. 79

Favorable outcome measures in the short term with regards
to morbidity, length of hospital stay, and 30-day mortality are
documented in most of the series. In a controlled investigation
comparing open with endovascular repair, the Vienna group 38
found 30-day mortality was decreased from 31% to 10%, and
that there was significantly decreased operative time (320 min
to 150 min), hospital stay (13 days to 4 days), and rate of neuro-
logic impairment (12% to 0%). In fact, they claimed that all 58
patients in the open surgical group would have been candi-
dates for the endovascular approach but devices were not yet
available in their institution. The Emory group compared
endovascular repair with a historical cohort of open repair
patients that would have been candidates for thoracic
endografting, and found significantly decreased blood loss
(1205 cm 3 to 325 cm 3 ), operative time (255 min to 155 min), ICU
stay (11 days to 1 day), and overall hospital stay (16 days to
6 days). 50

Neurologic complications, mainly paralysis, are one of the
most feared complications of open thoracic aortic repair and
previously occurred in 20-30% of cases, improving recently to
8-15% with numerous adjunctive advances in spinal cord
protection. 80 ' 81 Interestingly, there has been a gratifying lack of
spinal cord ischemic events with thoracic endografting, with
18 out of the 26 series reporting 0% paraplegia rates. The re-
mainder of the series all have less than 7% paraplegia rate,
with a few occurring when there were simultaneous open ab-
dominal aortic interventions performed at the same time. 39,43
This suggests that the neurologic complications after thoracic
aortic surgery are more related to the aortic cross-clamp time
and hypoperfusion during circulatory arrest and cardiopul-
monary bypass. Placing an endovascular stent across even a
long segment of thoracic intercostals does not appear to lead to
neurologic sequelae in the reported series.

Even with the technical success and improved early mor-
bidity and mortality, the long-term durability of TAA repair
with endografting remains to be elucidated. Although

559

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

endoleaks have been extensively studied in abdominal
aneurysms, there is limited literature regarding the thoracic
aorta. The endoleak rate does appear to be lower than in the ab-
dominal aorta, ranging from 10% to 25% in the series reviewed
here. Most were noted on short-term follow-up and related to
proximal, distal, or junctional connections, and secondary
procedures were performed to exclude the endoleak. 45,53 They
do not appear to be due to patent side-branches from inter-
costal or bronchial branches, which would be an analogous
type II leak. Late endoleaks have been observed in the studies
with longer follow-up and typically related to dilation of the
aortic neck. 54 The long-term consequence of endoleaks in the
thoracic aorta is unknown at this time, but our recommenda-
tion is that all proximal and distal endoleaks should be aggres-
sively treated with further stent-graft coverage.

With the possibility of perigraft leakage and the question of
long-term durability, surveillance of all thoracic endograft pa-
tients is mandatory. Most protocols call for spiral CT scans to
be done postoperatively, at 1-month, and 6-month intervals
thereafter. The purpose of such close imaging follow-up is to
search for dilation of adjacent arterial segments or modular
disconnections related to morphologic changes occurring
within the aorta. Aneurysm size by 2-D measurements has
been found to decrease in stented TAA patients without en-
doleak by up to 10%. 53/59 Figure 48.3 (also in colour, see Plate 8,
facing p. 370) demonstrates a patient in our series with
shrinking aneurysm size by diameter and volume that by
2 years showed a modular disconnect and subsequent en-
doleak requiring secondary intervention. We have found that
volume changes measured by computerized 3-D reconstruc-
tion are extremely useful in AAA endograft surveillance to as-
sure that exclusion has been successful, and have routinely
performed such measurements in our TAA patients. 82 With
the increasing number of series being reported that document
midterm outcomes, there will be the inevitable discovery of
other device-related complications. Stent-graft migration,
material deficits, and fatigue fractures of the stent frame have
all been described for AAA repair, and accurate surveillance
for these problems in thoracic endograft patients should be
pursued. 56 As future studies are published along with imaging
data, long-term durability of TAA endografting will be
clarified.

Finally, in order to demonstrate superiority of the endovas-
cular approach over open repair for TAAs, one would have to
show that overall survival is improved in a randomized trial.
Such a trial will be difficult to perform given the ethics and
high morbidity and mortality associated with the disease
process as well as open repair. For now we will have to rely on
continued surveillance of patients treated via endovascular
means and compare their outcomes with historical cohorts un-
dergoing open surgery. The large series from Stanford docu-
mented actuarial survival rates of patients undergoing
thoracic endografting to be 81% survival at 1-year follow-up
and 73% after 2 years. 39,83 This is in comparison with tradition-

al open surgery, where the actuarial survival rates are estimat-
ed to be 70% at 5 years and 40% at 10 years of survivors of the
open repair. 12 ' 15 ' 16 Combining the fact that most of the patients
who underwent stent-graft repair were already turned down
for open repair and yet had similar midterm survival, a case
can certainly be made that the long-term survival will be at
least as good with endografting vs. open surgery.

Midterm results have been documented in the past year
from various high-volume centers. The large series from
Germany with 45 patients treated for TAAs with endografts
shows 2-year survival of 84%, better than their own results
with conventional repair. 55 The Swedish group reported
promising midterm results with 2-year survival of 74% and
freedom from mortality, reoperations, and major complica-
tions of 52% at 1 year. 57 The extensive experience out of 51 pa-
tients treated for TAA at Mt Sinai showed excellent midterm
overall survival of 67% at 40 months with freedom from ad-
verse device-related events, rupture, or endoleak of 74% at
40 months. 60 The large cohort of 54 patients from Austria
documented 53% 3-year event-free survival. In summary,
the midterm durability of stent-graft repair for TAAs is
encouraging based on a low rate of endoleaks requiring
secondary intervention and estimated survivals at least as
good as conventional open surgical patients.

Type B aortic dissections

Centers that began to develop thoracic endografting pro-
grams for TAAs naturally applied their techniques for other le-
sions in the thoracic aorta, namely type B dissections. Over the
past decade stent-graft implantation at the dissection entry
site is gaining recognition as a viable alternative to open repair.
Table 48.3 compares the 25 series in the literature of
stent-grafting for type B dissections. 45 ' 46 ' 48 ' 49 ' 51-54 ' 59 ' 63-78 The
bulk of the series include cases that are acute /subacute
(<2-4 weeks) or have acute complications of a chronic type B
dissection. Mimicking the review of worldwide TAA series,
there are strikingly high technical success rates, with most se-
ries reporting 100% procedural success. Survival data are,
however, severely lacking in all the series reviewed, and only
after clear documentation of improved survival will endovas-
cular therapy become a recommended first-line therapy in the
treatment of acute or chronic type B dissection.

The mechanism of stent-graft treatment of type B dissec-
tions is due to obliteration of the entry tear and ceasing flow
into the false lumen while improving flow to the true lumen.
More than 50% of the mortality from type B dissections results
from rupture of an enlarging false lumen. Endovascular graft-
ing is particularly suited to this solution, since entry sites if
found can be located in short segments even if the dissection is
long. The decrease in flow in the false lumen promotes throm-
bosis, which should prevent the eventual formation of an
aneurysm. 45 For the more significant false lumens seen in
chronic dissections, flow may continue in the perioperative in-

560

chapter 48 Endovascular stent-graft repair of thoracic aortic aneurysms and dissections

Table 48.3 Comparison of 25 series of acute and chronic type B dissections treated with endovascular stent-grafting.

No. of

Percent

Procedural

Paraplegia

30-day

Series

Device

patients

acute

success

Other morbidity

mortality

Thrombosis

F/U

Dake 63 1996-1 998

Homemade

100%

20%

80%

Nienaber 64 1997-1 998

Talent

100%

Czermak 65 1996-1 999

Talent

71%

86%

28%

86%

White 45 1997-1 999

AneuRx

22%

100%

Won 46 1 994-1 999

Taewoong

91%

83%

Kato 66 1997-2000

Cook

100%

89%

Sailer 67 1997-2000

Gore

81%

100%

72%

Bortone 68 1999-2001

Talent/Gore

58%

100%

25%

100%

Shimono 69 1997-2000

Cook

54%

100%

14%

96%

Thompson 48 2000-2001

Gore

n/a

100%

23%

n/a

Totaro 49 2000-2001

Gore

20%

100%

25%

n/a

Hutschala 70 2000-2001

Gore

100%

11%

22%

n/a

Palma 71 1996-2001

Braile

60%

93%

16%

81%

Criado 51 1999-2002

Talent

100%

97%

15%

71%

Shim 72 1994-2001

Homemade

93%

71%

Lepore 73 1999-2001

Gore/Talent

79%

100%

33%

86%

Herold 52 1999-2001

Talent

39%

100%

n/a

Pamler 74 1999-2001

Gore/Talent

29%

100%

28%

n/a

Beregi 75 1997-2000

Talent/Gore

100%

13%

10%

n/a

Chabbert 53 1997-2001

Talent/Gore

33%

100%

11%

77%

Fattori 54 1997-2002

Talent

36%

100%

91%

Grabenwoger 76 1997-2002

Gore

100%

Lopera 77 1999-2001

Homemade

40%

90%

20%

90%

Nienaber 78 1999-2002

Talent

100%

18%

54%

n/a

Lambrechts 59 2000-2002

Talent/Gore

45%

100%

14%

91%

n/a

Other morbidity refers to cardiopulmonary, renal, infectious, and neurologic complications. Thrombosis refers to the percentage of cases that on follow-up
computed tomography scan demonstrated thrombosis of the false lumen, indicating successful coverage of the entry site of the dissection. F/U indicates length
of mean follow-up in months for each individual series.

terval, but the false lumen slowly thromboses with a trend to-
ward decreased flow and volume. Aortic stability is induced
both by thrombosis of the false lumen and from the device it-
self. 64 Postoperative CT-angiogram surveillance is necessary
to visualize thrombosis and to document regression of the
false lumen. Figure 48.4 demonstrates a case of regression of
the false lumen in a case of thoracoabdominal dissection.
Occasionally, however, the true lumen is unable to provide
adequate flow to an ischemic bed after the false lumen has
been removed from the circulation, and other interventional
procedures like fenestration or bypass are necessary. 84 Com-
plete thrombosis of the false lumen is achieved in the majority
of patients reported in the literature, usually >70%. Only the
Austria group reported significantly lower thrombosis rates. 70
At 6 months' follow-up, two out of nine had complete throm-
bosis of the aortic false lumen. The other seven patients had
obliterated false lumens in the region of the thoracic aorta, but
the abdominal region was being perfused via reentry sites.
This may be adequate treatment in the acute patient suffering
from type B dissection, since covering the main entry site near

the takeoff of the left subclavian artery may lead to resolution
of symptoms. 46 ' 72 Even when there is a reentry site at the distal
thoracic aorta or abdominal aorta, diverting the majority of
flow into the true lumen should be enough to cause the false
lumen to no longer expand, and at least protect the patient
from impending thoracic rupture. 74 We have observed this de-
layed regression of the more distal portion of a chronic dissec-
tion (Fig. 48.5; also in colour, see Plate 9, facing p. 370). Longer
term data will obviously be needed to clarify the aortic remod-
eling that occurs after an aortic dissection is subjected to a
stent-graft.

In contrast to open surgical repair for type B dissections,
endovascular stent-grafting shows significantly less serious
short-term morbidity and mortality. Most series document 30-
day mortality rates of <10%, and typically occurring only in
patients with acute presentations and organ ischemia. These
patients are already the highest risk candidates for any form of
operative intervention. Paraplegia rates are extremely low,
with no series having more than one patient (0-11%). Morbid-
ity was minimal even when compared with the TAA series of

561

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

PREOP

1 YEAR

CELIAC

ta^d

■ il

SMA

RENAL

AAA NECK

Figure 48.4 Axial computed tomography images perpendicular to the
centerline of the flow lumen at the celiac artery, superior mesenteric artery
(SMA), lowest renal artery, and the neck of the abdominal portion of a
symptomatic thoracoabdominal dissection presenting with renal failure and
chest, back, and abdominal pain. Note that there is complete obliteration of
the false lumen and no evidence of aneurysmal dilation. This patient currently
remains asymptomatic with normalization of his renal function.

endograf t repair, and certainly when compared with historical
cohorts of open repair for type B dissection. Most authors
report a postimplantation syndrome in up to 70% of patients,
including inflammation, slight fever, leukocytosis, and back
pain that resolves spontaneously in 5-25 days, and antibiotics
are usually not recommended. 71 ' 72

Endovascular complications, such as endoleaks and migra-
tion, are also rarely noted, although follow-up is still in the
short and midterm. 75 ' 85 Of note, in the series from Japan, four
out of the nine patients treated for acute presentations of type
B dissections developed saccular aneurysms in the follow-up
period. 66 They postulated the cause to be a weakened aortic
wall of the dissecting aorta. The other series reviewed did not
observe this complication, and we have not yet witnessed this
in our series. Another complication of note that we have expe-
rienced that has been reported in the literature is retrograde
dissection after stent-graft placement necessitating open
conversion and ascending aortic replacement. 86,87 This does,
however, also occur in 10% of open repairs, and requires
arch exposure and cardiopulmonary bypass. 88

In light of these rare complications and the expected
fragility of the aortic wall, device modifications have been
implemented that should improve the overall safety and
efficacy of the endovascular approach for type B dissections. 74
These changes have included manufacturing smoother
edges, designing more flexible bodies, and building softer
nose-cone tips. Significant improvements have also been
made with regard to the delivery systems, as maneuvering in
the thoracic arch requires precise deployment capabilities
when handling the device. The future of thoracic aortic stent-
grafting for dissections certainly will require that the technol-

Figure 48.5 Computed tomography-angiogram
reconstructions of an 83-year-old man with
symptomatic chronic type B dissection extending
down to his aortic bifurcation. Coverage of the entry
site in the thoracic aorta immediately caused the
thoracic portion of the false lumen to obliterate, and
over time the abdominal portion has slowly
regressed. He currently remains symptom-free
without any chest or back pain. See also Plate 9,
facing p. 370.

562

chapter 48 Endovascular stent-graft repair of thoracic aortic aneurysms and dissections

Figure 48.6 Seventy-nine-year-old man presenting with hemodynamic
collapse and hemoptysis from large expanding pseudoaneurysm due to
penetrating thoracic aortic ulcer. The arrow demonstrates the perforation,
and placement of a stent-graft resolved the leak and the patient's symptoms.
See also Plate 1 0, facing p. 370.

ogy adapt to the long-term findings and complications that are
encountered.

Penetrating aortic ulcers

Only within the past 15 years have penetrating aortic ulcers
been recognized as a distinct clinical entity. 6 Because the le-
sions are focal ruptures in short portions of the thoracic aorta
and can lead to significant morbidity and mortality, they may
be ideal candidates for stent-graft repair (Fig. 48.6; also in
colour, see Plate 10, facing p. 370). A small number of series
have documented encouraging results with > 90% technical
success and low morbidity and mortality. 6 ' 31 ' 89 The only
midterm data are from the Stanford group with survival esti-
mates at 1, 3, and 5 years of 85%, 75%, and 70%, respectively 90
This compares favorably with conventional open repair and
includes patients deemed unfit for thoracotomy, making tho-
racic endografting for penetrating aortic ulcers an attractive
alternative. As with the enthusiasm for stent-grafting for

TAAs and dissection, continued postoperative surveillance
will be necessary to elucidate long-term durability and
complications.

Traumatic aortic rupture

The incidence of concomitant injuries with blunt trauma to the
aorta is high, making conventional open repair with thoraco-
tomy, aortic cross-clamping, single lung ventilation, and sys-
temic anticoagulation potentially risky. With the development
of endografting for TAAs and dissections, devices are more
readily available to place in acute or emergent settings and
there have been numerous small series describing the efficacy
of applying this technology to trauma patients with technical
success of nearly 100%. 91-94 Relative indications for stent-graft
placement in trauma patients include the presence of severe
pulmonary contusion, cardiac risk factors, severe coagulopa-
thy, and closed head injury 94 These patients are extremely
high risk to undergo open repair and the minimally invasive
approach allows for immediate repair of their injured aorta
(Fig. 48.7). Early mortality can be avoided with prompt surgi-
cal intervention, although there are limited data on open vs.
endovascular treatment of traumatic rupture. A retrospective
study from New Mexico of all patients treated for blunt tho-
racic aortic injury showed 92% mortality in the nonoperative
group, 50% in the open surgical group, and 20% in the en-
dovascular group. 95 Midterm survival has been seen out to
21 months, but long-term data and information about device-
related complications at this time are lacking. 91 The minimally
invasive approach to high-risk trauma patients of thoracic en-
dografting for aortic transection is extremely appealing, and
should be considered if available.

Summary

The feasibility, safety, and short- and midterm efficacy of tho-
racic endografting have been delineated in a substantial num-
ber of series of patients treated for TAAs and acute and chronic
type B dissections. Encouraging results have been document-
ed in other diseases of the thoracic aorta including penetrating
ulcers and traumatic disruption. The morbidity and operative
mortality associated with thoracic aortic endovascular repair
are significantly less than with conventional open treatment.
The associated morbidity, including paraplegia, renal failure,
cardiac complications, and pulmonary compromise, is re-
duced because of the avoidance of general anesthesia, aortic
cross-clamping, and a large incision. This allows for a shorter
hospital stay, quicker recovery, and at least as good midterm
survival. Patient selection, accurate imaging, and long-term
surveillance are some of the factors that need to be addressed
in order for this technique to emerge as a standard therapy.
Even at this early state of this technology, where results are af-
fected by the rapid evolution of the technology in extremely

563

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 48.7 Computed tomography (CT)
scan and intravascular ultrasound (IVUS)
findings of 32-year-old man driving at
70 miles/h who struck a parked car. Initial CT-
angiogram of the chest shows loss of smooth
contour of descending thoracic aorta nearthe
isthmus and surrounding hematoma consistent
with aortic injury. Preoperative IVUS
demonstrates dissection flap (dotted arrow) and
aortic wall disruption and hematoma (solid
arrow). Patient underwent placement of
thoracic stent-graft and postoperative images
show excellent apposition of stent-graft to
aortic wall at the level of the injury.

high-risk patients, the dramatic improvements in early
survival and reduced morbidity and mortality make this
approach a very appealing alternative to open surgery In
many patients thoracic endografting is the only alternative
that can be offered as open surgery is often denied to these
prohibitively high-risk lesions.

References

1. Bickerstaff LK, Pairolero PC, Hollier LH et al. Thoracic aortic
aneurysms: a population-based study. Surgery 1982; 92:1103.

2. Crawford ES, Hess KR, Cohen ES, Coselli JS, Safi HJ. Ruptured
aneurysm of the descending thoracic and thoracoabdominal
aorta. Ann Surg 1991; 213:417.

3. Johansson G, Markstrom U, Swedenborg J. Ruptured thoracic aor-
tic aneurysms: a study of incidence and mortality rates. / Vase Surg
1995; 21:985.

4. Fann JI, Miller DC. Aortic dissection. Ann Vase Surg 1995; 9:311.

5. Coady MA, Rizzo JA, Hammond GL, Peirce JG, Kopf GS,
Elefteriades JA. Penetrating ulcer of the thoracic aorta: what is it?
How do we recognize it? How do we manage it? / Vase Surg 1998;
27:1006.

6. Kos X, Bouchard L, Otal P et al. Stent-graft treatment of penetrat-
ing thoracic aortic ulcers. / Endovasc Ther 2002; 9:11-25.

7. Cowley RA, Turney SZ, Hankins JR, Rodriguez A, Attar S,
Shankar BS. Rupture of thoracic aorta caused by blunt trauma. A
fifteen-year experience. / Thorac Cardiovasc Surg 1990; 100:652.

8. Cardarelli MG, McLaughlin JS, Downing SW, Borwn JM, Attar S,
Griffith BP Management of traumatic aortic rupture: a 30-year
experience. Ann Surg 2002; 236:465.

9. Kouchoukos NT, Dougenis D. Surgery of the thoracic aorta. NEngl
J Med 1997; 336:1876.

10. Elefteriades JA. Natural history of thoracic aortic aneurysms. Ann
Thorac Surg 2002; 74:S1877.

11. Perko MJ, Norgaard M, Herzog TM, Olsen PS, Schroeder TV,
Pettersson G. Unoperated aortic aneurysms: a survey of 170
patients. Ann Thorac Surg 1995; 59:1204.

12. Crawford ES, DeNatale RW. Thoracoabdominal aortic aneurysm:
observations regarding the natural course of the disease. / Vase
Surg 1986; 3:578.

13. Safi HJ, Miller CC III, Subramaniam MH et al. Thoracic and thora-
coabdominal aneurysm repair using cardiopulmonary bypass,
profound hypothermia, and circulator arrest via left side of the
chest incision. / Vase Surg 1998; 28:591 .

14. Cambria RP, Davison JK, Carter C et al. Epidural cooling for spinal
cord protection during thoracoabdominal aneurysm repair. / Vase
Surg 2000; 31:1093.

15. DeBakey ME, McCollum CH, Graham JM. Surgical treatment of
aneurysms of the descending thoracic aorta: long-term results in
500 patients. ] Cardiovasc Surg 1978; 19:571.

16. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Experi-
ence with 1509 patients undergoing thoracoabdominal aortic
operations. / Vase Surg 1993; 17:357.

17. Coselli JS, Conklin LD, LeMaire SA. Thoracoabdominal aortic
aneurysm repair: review and update of current strategies. Ann
Thorac Surg 2002;74:S1881.

18. Clouse WD, Hallett JW, Schaff HV, Gayari MM, Ilstrub DM,
Melton LJ. Improved prognosis of thoracic aortic aneurysms. ] Am
Med Assoc 1998; 280:1926.

19. Rectenwald JE, Huber TS, Martin TD et al. Functional outcome
after thoracoabdominal aortic aneurysm repair. / Vase Surg 2002;
35:640.

20. Kouchoukos NT, Dougenis D. Surgery of the thoracic aorta. NEngl
J Med 1997; 336:1876.

21. Umana JP, Lai DT, Mitchell RS, Moore KA, Rodriguez F, Robbins
R. Is medical therapy still the optimal treatment strategy for

564

chapter 48 Endovascular stent-graft repair of thoracic aortic aneurysms and dissections

patients with acute type B aortic dissections? / Thome Cardiovasc
Surg 2002; 124:896.

22. Panneton JM, Teh SH, Cherry KJ et ah Aortic fenestration for acute
or chronic aortic dissection. / Vase Surg 2000; 32:711.

23. Glower DD, Fann JI, Speier RH et ah Comparison of medical and
surgical therapy for uncomplicated descending aortic dissection.
Circulation 1990; 82 (SIV):39.

24. Elefteriades JA, Lovoulos CJ, Coady MA, Tellides G, Kopf GS,
Rizzo JA. Management of descending aortic dissection. Ann
Thorac Surg 1999; 67:2002.

25. Khan IA, Nair CK. Clinical, diagnostic, and management perspec-
tives of aortic dissection. Chest 2002; 122:311.

26. Fann JI, Smith JA, Miller DC et ah Surgical management of aortic
dissection during a 30-year period. Circulation 1995; 92:113.

27. Lansman SL, Hagl C, Fink D et ah Acute type B aortic dissection:
surgical therapy. Ann Thorac Surg 2002; 74:S1833.

28. Kato M. Stent-graft implantation for aortic dissection. In: Liotta D,
ed. Diseases of the Aorta, 2nd edn. Argentina: Editorial Universidad
deMoRon, 2003:285.

29. Marui A, Mochizuki T, Mitsui N, Koyama T, Kimura F, Horibe M.
Toward the best treatment for uncomplicated patients with type B
acute aortic dissection. Circulation 1999; 100:11-275.

30. Stanson AW, Kazmier FJ, Hollier LH et ah Penetrating atheroscle-
rotic ulcers of the thoracic aorta. Ann Vase Surg 1986; 1:15.

31. Harris JA, Bis KG, Glover JL, Bendick PJ, Shetty A, Brown OW.
Penetrating atherosclerotic ulcers of the aorta. / Vase Surg 1994;
19:98.

32. Brittenden J, McBride K, Mclnnes G, Gillespie IN, Bradbury AW.
The use of endovascular stents in the treatment of penetrating
ulcers of the thoracic aorta. / Vase Surg 1999; 30:946.

33. Fabian TC, Richardson JD, Croce MA et ah Prospective study of
blunt aortic injury. Multicenter trial of the AAST. / Trauma 1997;
42:374.

34. Aziz I, Lee JT, Lee J et ah Accuracy of three-dimensional simulation
in the sizing of aortic endoluminal devices. Ann Vase Surg 2003;
17:129.

35. Lee JT, White RA. Basics of intravascular ultrasound: an essential
tool for the endovascular surgeon. Semin Vase Surg 2004; 17:
110.

36. Criado FJ, Barnatan MF, Rizk Y, Clark NS, Wang CF. Technical
strategies to expand stent-graft applicability in the aortic arch
and proximal descending thoracic aorta. / Endovasc Ther 2002;
9:11-32.

37. Dake MD, Miller DC, Semba CP, Mitchell RS, Walker PJ, Liddell
PP. Transluminal placement of endovascular stent-grafts for the
treatment of descending thoracic aortic aneurysms. N Engl} Med
1994; 331:1729.

38. Ehrlich M, Grabenwoeger M, Cartes-Zumelzu F et ah Endovascu-
lar stent graft repair for aneurysms on the descending thoracic
aorta. Ann Thorac Surg 1998; 66:19.

39. Mitchell RS, Miller DC, Dake MD, Semba CP, Moore KA, Sakai T.
Thoracic aortic aneurysm repair with an endovascular stent graft:
the "first generation." Ann Thorac Surg 1999; 67:1971.

40. Temudom T, D'Ayala M, Marin ML et ah Endovascular grafts in
the treatment of thoracic aortic aneurysms and pseudo-
aneurysms. Ann Vase Surg 2000; 14:230.

41. Grabenwoger M, Hutschala D, Ehrlich MP et ah Thoracic aortic
aneurysms: treatment with endovascular self-expandable stent-
grafts. Ann Thorac Surg 2000; 69:441.

42. Taylor PR, Gaines PA, McGuinness CL et ah Thoracic aortic stent
grafts — early commercial experience from two centers using com-
mercially available devices. Eur J Vase Endovasc Surg 2001; 22:70.

43. Greenberg R, Resch T, Nyman U et ah Endovascular repair of
descending thoracic aortic aneurysms: an early experience with
intermediate-term follow-up. J Vase Surg 2000; 31:147.

44. Bortone AS, Schena S, Mannatrizio G et ah Endovascular
stent-graft treatment for diseases of the descending thoracic
aorta. Eur J Cardiovasc Thorac Surg 2001; 20:514.

45. White RA, Donayre CE, Walot I et ah Endovascular exclusion of
descending thoracic aortic aneurysms and chronic dissections:
initial clinical results with the AneuRx device. / Vase Surg 2001;
33:927.

46. Won JY, Lee DY, Shim WH, Chang B, Park SI, Yoon CS. Elective en-
dovascular treatment of descending thoracic aortic aneurysms
and chronic dissections with stent-grafts. / Vase Interv Radiol 2001;
12:575.

47. Cambria RP, Brewster DC, Lauterbach SR et ah Evolving experi-
ence with thoracic aortic stent graft repair. / Vase Surg 2002;
35:1129.

48. Thompson CS, Gaxotte VD, Rodriguez JA et ah Endoluminal stent
grafting of the thoracic aorta: initial experience with the Gore
Excluder. / Vase Surg 2002; 35:1163.

49. Totaro M, Mazzesi G, Marullo AGM, Neri E, Fanelli F, Miraldi F.
Endoluminal stent grafting of the descending thoracic aorta. Hal
Heart J 2002; 3:366.

50. Najibi S, Terramani TT, Weiss VJ et ah Endoluminal versus open
treatment of descending thoracic aortic aneurysms. / Vase Surg
2002; 36:732.

51. Criado FJ, Clark NS, Barnatan MF. Stent graft repair in the aortic
arch and descending thoracic aorta: a 4-year experience. / Vase
Surg 2002; 36:1121.

52. Herold U, Piotrowski J, Baumgart D, Eggebrecht H, Erbel R, Jakob
H. Endoluminal stent graft repair for acute and chronic type B aor-
tic dissection and atherosclerotic aneurysm of the thoracic aorta.
Eur J Cardiothorac Surg 2002; 22:891.

53. Chabbert V, Otal P, Bouchard L et ah Midterm outcomes of thoracic
aortic stent-grafts. J Endovasc Ther 2003; 10:494.

54. Fattori R, Napoli G, Lovato L et ah Descending thoracic aortic dis-
eases: stent-graft repair. Radiology 2003; 229:176.

55. Scharrer-Pamler R, Kotsis T, Kapfer X, Gorich J, Orend KH,
Sunder-Plassmann L. Complications after endovascular treat-
ment of thoracic aortic aneurysms. / Endovasc Ther 2003; 10:711.

56. Lamme B, de Jonge ICDYM, Reekers JA, de Mol BAJM, Balm R.
Endovascular treatment of thoracic aortic pathology. Eur } Vase
Endovasc Surg 2003; 25:532.

57. Lepore V, Lonn L, Delle M, Mellander S, Radberg G, Risberg B.
Treatment of descending thoracic aneurysms by endovascular
stent grafting. / Cardiovasc Surg 203; 18:436.

58. Krohg-Sorensen K, Hafsahl G, Fosse E, Geiran OR. Acceptable
short-term results after endovascular repair of diseases of the
thoracic aorta in high risk patients. Eur J Cardiothorac Surg 2003;
24:379.

59. Lambrechts D, Casselman F, Schroeyers P, De Geest R, D'Haenens
P, Degrieck I. Endovascular treatment of the descending thoracic
aorta. Eur J Vase Endovasc Surg 2003; 26:437.

60. Ellozy SH, Carrocio A, Minor M et ah Challenges of endovascular
tube-graft repair of thoracic aortic aneurysm: midterm follow-up
and lessons learned. / Vase Surg 2003; 38:676.

565

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

61. Czerny M, Cejna M, Hutschala D et ah Stent-graft placement in
atherosclerotic descending thoracic aneurysms: midterm results.
JEndovasc Ther 2004; 11:26.

62. Melissano G, Tshomba Y, Civilini E, Chiesa R. Disappointing re-
sults with a new commercially available thoracic endograft. / Vase
Surg 2004; 39:124.

63. Dake MD, Kato N, Mitchell RS et at. Endovascular stent-graft
placement for the treatment of acute aortic dissection. N Engl J Med
1999; 340:1546.

64. Nienaber CA, Fattori R, Lund G et ah Non-surgical reconstruction
of thoracic aortic dissection by stent-graft placement. N Engl] Med
1999; 340:1539.

65. Czermak BV, Waldenberger P, Fraedrich G et ah Treatment of Stan-
ford type B aortic dissection with stent-grafts: preliminary results.
Radiology 2000; 217:544.

66. Kato N, Hirano T, Kawaguchi T et ah Aneurysmal degeneration of
the aorta after stent-graft repair of acute aortic dissection. / Vase
Surg 2001; 34:513.

67. Sailer J, Peloschek P, Rand T, Grabenwoger M, Thurnher S, Lam-
mer J. Endovascular treatment of aortic type B dissection and pen-
etrating ulcer using commercially available stent-grafts. Am }
Roentgenol 2001; 177:1365.

68. Bortone AS, Schena S, D'Agostino D et ah Immediate versus de-
layed endovascular treatment of post-traumatic aortic pseudo-
aneurysms and type B dissections: retrospective analysis and
premises to the upcoming European trial. Circulation 2002; 106
(Suppl.I):I-234.

69. Shimono T, Kato N, Yasuda F et ah Transluminal stent-graft place-
ments for the treatments of acute onset and chronic aortic dissec-
tions. Circulation 2002; 106 (Suppl. I):I-241.

70. Hutschala D, Fleck T, Czerny M et ah Endoluminal stent-graft
placement in patients with acute aortic dissection type B. Eur J
Car diothorac Surg 2002; 21:964.

71. Palma JH, de Souza JAM, Alves CMR, Carvalho AC, Buffolo E.
Self-expandable aortic stent-grafts for treatment of descending
aortic dissections. Ann Thorac Surg 2002; 73:1138.

72. Shim WH, Koo BK, Yoon YS et ah Treatment of thoracic aortic dis-
section with stent-grafts: midterm results. / Endovasc Ther 2002;
9:817.

73. Lepore V, Lonn L, Delle M et ah Endograft therapy for diseases of
the descending thoracic aorta. J Endovasc Ther 2002; 9:829.

74. Pamler RS, Kotsis T, Gorich J, Kapfer X, Orend KH, Sunder-
Plassmann L. Complications after endovascular repair of type B
aortic dissection. J Endovasc Ther 2002; 9:822.

75. Beregi JP, Haulon S, Otal P et ah Endovascular treatment of acute
complications associated with aortic dissection: midterm results
from a multicenter study. / Endovasc Ther 2003; 10:486.

76. Grabenwoger M, Fleck T, Czerny M et ah Endovascular stent graft
placement in patients with acute thoracic aortic syndromes. Eur]
Cardiothorac Surg 2003; 23:788.

77. Lopera J, Patino JH, Urbina C et ah Endovascular treatment of
complicated type-B aortic dissection with stent-grafts: midterm
results. / Vase Intervent Radiol 2003; 14:195.

78. Nienaber CA, Ince H, Weber F, Rehders T, Petzsch M, Meinertz T.
Emergency stent-graft placement in thoracic aortic dissection and
evolving rupture. / Cardiovasc Surg 2003; 18:464.

79. Hansen C, Lee JT, Lee J et ah Endovascular aneurysm repair in pa-
tients with concomitant occlusive disease complicating access.
Submitted for publication.

80. Safi HJ, Campbell MP, Ferreira ML, Azizzadeh A, Miller CC.
Spinal cord protection in descending thoracic and thoracoabdom-
inal carotid aneurysm repair. Semin Thorac Cardiovasc Surg 1998;
10:41.

81. Hamilton IN, Hollier LH. Adjunctive therapy for spinal cord pro-
tection during thoracoabdominal aortic aneurysm repair. Semin
Thorac Cardiovasc Surg 1998; 10:35.

82. Lee JT, Aziz I, Lee J et ah Volume regression of abdominal aortic
aneurysms and its relation to successful endoluminal exclusion.
J Vase Surg 2003; 38:1254.

83. Fann JI, Miller DC. Endovascular treatment of descending
thoracic aortic aneurysms and dissections. Surg Clin N Am 1999;
79:551.

84. Slonim SM, Nyman U, Semba CP, Miller DC, Mitchell RS, Dake
MD. Aortic dissection: percutaneous management of ischemic
complications with endovascular stents and balloon fenestration.
J Vase Surg 1996; 23:241.

85. Mitchell RS, Ishimaru S, Ehrlich MP et ah First international sum-
mit on thoracic aortic endografting: roundtable on thoracic aortic
dissection as an indication for endografting. / Endovasc Ther 2002;
9:11-98.

86. Totaro M, Miraldi F, Fanmelli F, Mazzesi G. Emergency surgery for
retrograde extension of type B dissection after endovascular stent
graft repair. Eur J Cardiothorac Surg 2001; 20:1057.

87. Bethuyne N, Bove T, Van den Brande P, Goldstein JP. Acute retro-
grade aortic dissection during endovascular repair of a thoracic
aortic aneurysm. Ann Thorac Surg 2003; 75:1967.

88. Jex RK, Schaff HV, Piehler JM et ah Early and late results following
repair of dissections of the descending thoracic aorta. / Vase Surg
1986; 3:226.

89. Faries PL, Lang E, Ramdev P, Hollier LH, Marin ML, Pomposelli
FB. Endovascular stent-graft treatment of a ruptured thoracic aor-
tic ulcer. / Endovasc Ther 2002; 9:11-20.

90. Demers P, Miller DC, Mitchell RS, Kee ST, Chagonjian L, Dake
MD. Stent-graft repair of penetrating atherosclerotic ulcers in the
descending thoracic aorta: mid-term results. Ann Thorac Surg
2004; 77:81.

91. Orford VP, Atkinson NR, Thomson K et ah Blunt traumatic aortic
transection: the endovascular experience. Ann Thorac Surg 2003;
75:106.

92. Marty-Ane CH, Berthet JP, Branchereau P, Mary H, Alric P. En-
dovascular repair for acute traumatic rupture of the thoracic aorta.
Ann Thorac Surg 2003; 75:1803.

93. Orend KH, Pamler R, Kapfer X, Liewald F, Gorich J, Sunder-
Plassmann L. Endovascular repair of traumatic descending aortic
transection. / Endovasc Ther 2002; 9:573.

94. Karmy-Jones R, Hoffer E, Meissner MH, Nicholls S, Mattos M. En-
dovascular stent-grafts and aortic ruptures: a case series. / Trauma
2003; 55:805.

95. Kasirajan K, Heffernan D, Langsfeld M. Acute thoracic aortic
trauma: a comparison of endoluminal stent grafts with open
repair and nonoperative management. Ann Vase Surg 2003; 17:
589.

566

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Brachiocephalic vascular reconstructions
compared with endovascular repair

Edward B. Diethrich

The vascular disorders of the brachiocephalic vessels have a
variety of causes, including atherosclerosis, iatrogenic injury,
inflammation from autoimmune disease, and trauma. Proxi-
mal subclavian or brachiocephalic artery disease has long
been treated using open surgical procedures. Bypass grafting
has been the most commonly used surgical procedure to treat
symptomatic subclavian artery stenoses or occlusions, or to
exclude aneurysms in vessels that supply the upper extremi-
ties. Indeed, carotid-subclavian bypass is relatively simple,
requiring only a short supraclavicular incision, and producing
acute and long-term results that are generally excellent. 1 An
alternative not requiring a prosthetic graft is a transposition
of the left subclavian artery proximal to the vertebral origin
directly to the common carotid artery. 2

While the use of endovascular techniques in the brachio-
cephalic vessels is relatively recent, the results are very en-
couraging, 3-20 leading some authors to suggest that stenting is
the treatment of choice for proximal stenoses and occlusions
of the upper limb vessels. 21 ' 22 Other investigators 23 are less
inclined to recommend endovascular procedures, pointing
to the time-tested results of surgical intervention. In a recent
analysis of the literature regarding endovascular and surgical
treatment of brachiocephalic lesions, 24 one author notes
that "there are no head-to-head trials of one technique
versus another/' Nevertheless, his review suggests good tech-
nical success and low complication rates with stents, noting
higher stroke and death rates in the surgical literature. His
overall conclusion that "percutaneous stenting should be
considered a first-line therapy in treating subclavian or
brachiocephalic obstruction" 24 concurs with our own
experience at the Arizona Heart Institute. 3 ' 4 ' 9 ' 13 Indeed, it is our
own prediction that endovascular approaches to these
pathologies will make classic operative techniques obsolete in
the future.

This chapter describes treatment of brachiocephalic disease
via endovascular procedures, with references and comparison
to open surgical techniques as appropriate.

Treatment of brachiocephalic
vascular disease

Vascular disease is a leading cause of morbidity and mortality
throughout the world, and atherosclerotic lesions are the most
common cause of vascular insufficiency. Extracranial cere-
brovascular disease often manifests itself in brachiocephalic
stenosis and occlusion and is frequently associated with
debilitating symptoms. For example, stenosis or occlusion
of the subclavian artery may eventually result in blockage
that reverses the normal direction of flow in the vertebral
artery, causing a "subclavian steal" from the cerebral circula-
tion. In this regard, occlusive lesions of the origin of the left
subclavian artery are the most prevalent arch vessel patholo-
gies and also the lesions most amenable to successful
treatment by endovascular means.

Endovascular surgical techniques that include balloon
angioplasty, stenting, and endoluminal grafting are now in
frequent use in a variety of vascular regions. Advances in
endovascular device design have yielded low-profile
catheters, hydrophilic catheter coatings, and significant im-
provements in catheter flexibility and in balloon materials.
Additionally, a variety of stents have become available that are
both lower in profile and considerably more flexible than the
original Palmaz designs. All of these advances in catheter-
based technologies now allow treatment with procedures that
are less invasive than classic surgical intervention because
percutaneous approaches are possible in most cases. Indeed,
in some vascular territories, minimally invasive endovascular
procedures have already been associated with reductions in
hospital stay and recovery time. 25

Patient selection

It is important to make a clear distinction between pathologies
of the origin of the great vessels of the aortic arch and those of
the cervical carotid bifurcation cephalad to the arch. At pre-
sent, angioplasty and stenting are used cautiously to treat

567

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 49.1 High-grade stenosis of the brachiocephalic artery requires
either (A) a medial sternotomy with graft insertion to the ascending aorta, or
a small right thoracotomy to expose the (B) ascending aorta for proximal
anastomosis and grafting to the right common carotid or subclavian artery.
Following right brachial artery cannulation (C) a brachiocephalic stenosis may
be accessed using endovascular techniques.

lesions in the latter area, because investigators have not
yet found a perfect device to prevent embolization at the
time of the procedure. While careful patient selection is
clearly very important in avoiding adverse events, further
studies are currently under way and are likely eventually to
elucidate the role of endovascular treatment in the bifurcation
areas. In contrast, however, atherosclerotic lesions at the
arch vessel origin are much less likely to elicit embolic com-
plications. While the plaque is frequently loose and friable at
the bifurcation and is easily disturbed by wire, balloon, and
stent manipulation, lesions in the brachiocephalic, left com-
mon carotid, and subclavian arteries tend to be firm, concen-
tric, localized, and relatively smooth. Given these contributing
characteristics, embolic complications are rare. Our own
experience at the Arizona Heart Institute suggests en-
dovascular approaches are an excellent treatment for occlu-
sive disease in this region. 3/4/9/13 In addition to the low risk
for embolization, the endovascular approach allows treat-
ment of lesions from within the artery and minimizes a
number of complications associated with classic, open
procedures. As an example, a high-grade stenosis of the
brachiocephalic artery (as shown in Fig. 49.1A,B) would
require either a medial sternotomy with graft insertion to the
ascending aorta (Fig. 49.1 A) or a small right thoracotomy to
expose the ascending aorta (Fig. 49. IB) for proximal anasto-
mosis and grafting to the right common carotid or subclavian
artery. Either of these incision "approaches" can be associated
with complications similar to those seen in more complex
open procedures. Elderly patients or those with comorbidities
that make incisions and open surgical procedures a risky
proposition may best be treated with less invasive endovascu-
lar procedures. For example, following right brachial artery
cannulation, a brachiocephalic stenosis maybe accessed using
endovascular techniques (Fig. 49. 1C) rather than a full sternal
or thoracic incision.

Procedural techniques

In the past, there have been many discussions regarding the
need to stent arch vessel lesions following balloon angioplasty.
Our experience with angioplasty alone has been less than
satisfactory due to a high degree of recoil after balloon dila-
tion. Therefore, we have adopted a policy of stenting the vast
majority of cases, and our results have been quite positive,
confirming our position that stents enhance treatment of these
lesions. The advent of new and improved stent designs has
certainly added to our success.

Endovascular procedures can be performed either in an op-
erative suite or catheterization laboratory, but high-resolution
fluoroscopic equipment is mandatory in order to achieve best
results. In most cases, local anesthesia is used, which is an
advantage over the open, surgical procedure. The selection of
anesthesia is generally made by mutual agreement between

568

chapter 49 Brachiocephalic vascular reconstructions compared with endovascular repair

Figure 49.2 The retrograde femoral wire cannot cross the flush occlusion.
Even locating the origin of the artery under these conditions is virtually
impossible. In contrast, a catheter can be directed coaxially and used to
guide a retrograde brachial wire across the occlusion.

Figure 49.3 The technique used with a right brachial-femoral wire that
permits stent deployment even when the artery is tortuous at the angle of
the origin.

the patient and the anesthesiologist. We prefer local anesthesia
with mild sedation because it allows the patient to communi-
cate throughout the procedure. Neurologic changes, although
rare, may be assessed immediately. The major disadvantage to
using local anesthesia is that patient movement during the
procedure disrupts "roadmapping" and catheter guidance
during balloon angioplasty or stent deployment.

A variety of approaches are used to access brachiocephalic
lesions, depending on which vessel is the target and whether
or not the lesion is stenosed or occluded. The primary concern
with the latter is the ability to cross the lesion with a wire. As
illustrated in Fig. 49.2, a flush occlusion of the left subclavian
artery is almost impossible to cross from the retrograde
femoral approach unless it has only recently become
occluded. In these cases, there is potential for the wire to seek
a subintimal position, and given that there is virtually no
"pushability" with the wire, crossing the lesion becomes im-
possible. In contrast, a retrograde brachial approach permits
the wire to be catheter directed in a coaxial position that facili-
tates crossing the lesion. A 6- or 7-Fr sheath allows access of
a 0.035-in angled Glidewire (MediTech/ Boston Scientific,
Natick, MA, USA), and heparin is administered to maintain
the activated coagulation time (ACT) above 250 s. Then, a
straight angiographic catheter or one with a 25-30° angle is
passed over the Glidewire to the occlusion. As the wire is
pushed through the lesion, the catheter is slowly advanced
until it moves into the aortic arch. Using this technique, the
majority of occlusions can be successfully traversed. Fluoro-
scopic guidance, contrast injection, and roadmapping allow
accurate placement of an angioplasty balloon (size range
4-9 mm) for predilation.

Most lesions of brachiocephalic origin are stenotic rather
than occlusive, and therefore are amenable to the retrograde
femoral approach for access. However, depending on the con-
figuration of the aortic arch and the angle of the artery itself, it
can be difficult to deliver a stent— particularly if a guiding
catheter is not used. Again, a right, retrograde, brachial
approach simplifies stent delivery in most cases. We have
encountered situations in which a brachial-femoral wire
configuration was used to deliver the stent in unusual arterial
configurations (Fig. 49.3).

We have never encountered an occlusive lesion at the origin
of the left common carotid artery that required intervention.
These lesions are almost always stenotic, thereby permitting a
stent to be delivered from the retrograde femoral approach.
However, as with the brachiocephalic vessels, it may be diffi-
cult to negotiate certain arch configurations, and therefore
we have introduced a modified open technique. The left
common carotid artery can be easily exposed through a short
supraclavicular incision (Fig. 49.4A). The ballooning
(Fig. 49.4B) and stenting (Fig. 49.4C) can then be accomplished
using a 6- or 7-Fr sheath. A simple purse string suture is used to
close the sheath site (Fig. 49.4D,E).

From a technical standpoint, one of the potential problems
with stenting the origin of the arch vessels is the difficulty in
knowing precisely where the vessel originates from the arch. It
may seem strange that the exact location is not clearly visible
using fluoroscopic and angioscopic imaging, but, in fact, we
have seen many cases in which a stent is deployed too far into
the aortic arch or even misses some of the plaque at the arterial
origin. In general, selection of stent size is based on the
adjacent normal vessel and the length of the lesion, or from

569

Figure 49.4 It may be difficult to negotiate certain arch configurations, and
therefore we have introduced a modified open technique as follows: (A) the
common carotid artery is exposed through a short incision above the clavicle,

and the wire and sheath are introduced, (B) balloon angioplasty is
performed, (C) the stent is deployed, (D and E) the puncture site is closed
with a purse string suture.

chapter 49 Brachiocephalic vascular reconstructions compared with endovascular repair

Figure 49.5 Moving the intravascular ultrasound probe backward and
forward across the origin of the vessel permits the operator to pinpoint
accurately the location where the stent should be deployed.

Figure 49.6 Lesions in the second part of the subclavian (arrow, main
figure) should be evaluated very carefully to ensure that stent placement will
not encroach on the vertebral artery origin (correct stent placement shown at
arrow in the inset figure).

measurements performed with intravascular ultrasound
(IVUS). The use of IVUS is very helpful in these situations, as
the IVUS probe can be introduced either antegrade or retro-
grade, depending on the selected access route. Moving the
probe backward and forward across the origin of the vessel
permits the operator to pinpoint accurately the location where
the stent should be deployed (Fig. 49.5). Our personal prefer-
ence is for the stent to extend several millimeters (but no more)
into the aortic arch. If too much of the stent is allowed to float
unopposed to the artery, there is a risk of it migrating, becom-
ing dislodged, or even lost during subsequent endovascular
manipulation.

There are a variety of stents now available. In January 2002,
Cordis (Warren, NJ, USA) introduced the Palmaz Genesis
stent, which is a new balloon-expandable device incorporat-
ing high radial strength to resist vessel recoil and flexibility
that accommodates tortuous or challenging anatomy. The
stent offers minimal stent shortening and may be delivered
through very low-profile devices; it is available premounted
or unmounted. Also by Cordis is the Smart Stent, which is a
self-expanding nitinol stent that has a flexible, segmented de-
sign. Both these stents are considerably more flexible than the
original Palmaz designs that were commonly used in the past.
Since active extension or flexion of the arm has the potential to
cause malformation of a rigid stent, a flexible stent is a good
choice. The Wallstent (Boston Scientific) is a self-expanding
stent that we have found quite useful in tortuous arteries or in
treating lesions at points of flexion. Alternatively, the VistaFlex
stent (Angiodynamics Inc., Queensbury, NY, USA) is a
balloon-expandable stent composed of platinum in a linked
segment design that is highly visible and magnetic resonance
angiography compatible. Bridge Stents (Medtronic AVE,
Santa Rosa, CA, USA) are also available in extra support and

flexible designs; both are balloon expandable and offered in a
variety of lengths. The Herculink Biliary System (Guidant,
Santa Clara, CA, USA) includes a low-profile, highly trackable
stent that provides differential radial strength. IntraStent
(IntraTherapeutics, Inc., St Paul, MN, USA) is another balloon-
expandable stent that is 6 Fr compatible when used with a
low-profile 5-Fr balloon; its cell structure provides robust
radial force.

Typically, the lesions we encounter in arch vessels are con-
centric and not associated with loose debris. Although they are
quite often calcific in nature, most are easily crossed, bal-
looned, and stented. Disease in the subclavian is most com-
monly found at the arch origins, allowing adequate access for
angioplasty and stenting without subclavian branch compro-
mise. Lesions in the second part of the subclavian artery, how-
ever, should be evaluated very carefully to ensure that stent
placement will not encroach on the vertebral artery origin
because of the risk of occluding or compromising blood flow
with a stent (Fig. 49.6).

One technique that may help ensure appropriate stent
placement involves placing a protective wire across the
vertebral artery orifice. This permits access for ballooning if
the subclavian stent impinges upon the vertebral artery
during the procedure. We have seen cases in which a stent
blocked an artery orifice and it was necessary to thread a wire
through the stent and dilate it to restore flow to the branched
vessel. Certainly, it is prudent to avoid these type of maneu-
vers whenever possible, but one must be prepared for compli-
cations in any surgical or endovascular procedure.

Following stent deployment, an angiographic control
image is taken, and the gradient is recorded. IVUS may be
superior to angiography in detecting inadequate stent
deployment and, when the IVUS images suggest suboptimal

571

Figure 49.7 (A) Intravascular ultrasound image shows incomplete expansion of a stent in the subclavian artery. (B) This is corrected by additional balloon
expansion.

Figure 49.8 (A) Left subclavian artery is occluded just proximal to the
vertebral origin. (B) A retrograde crossing is attempted and fails, with the
wire dissecting in the subadventitial plane. (C) One week later, the patient

returns with severe chest pain secondary to development of a
pseudoaneurysm at the dissection site. (D) In this case, the aneurysm
was successfully excluded with an endoluminal graft.

chapter 49 Brachiocephalic vascular reconstructions compared with endovascular repair

deployment (Fig. 49. 7A), a larger balloon should be used to
expand the stent (Fig. 49. 7B). After completion of the proce-
dure, patients are transferred to the intensive care unit
and monitored. The sheath is withdrawn when the ACT is
below 150 s. The patient is usually discharged the following
day. At some centers, patients with lesions that may be
easily accessed and stented are being treated on an outpatient
basis.

One of the newer indications for subclavian artery stenting
relates to the now almost universal use of the left internal
mammary artery as a conduit for bypassing coronary
artery obstructions. The presence of an obstructed lesion at the
origin of the left subclavian artery (the side usually preferred
with the left internal mammary bypass) compromises flow
and can result in graft failure. Stent deployment at the origin of
the left subclavian can ensure uncompromised flow to the left
internal mammary artery, and since the restenosis rate is so
low, long-term success can be anticipated. In some patients,
angina may resolve initially after a left internal mammary
artery bypass to the left anterior descending artery; later, how-
ever, progressive occlusive disease may develop in the left
subclavian artery and cause a recurrence of symptoms. Stent
deployment is usually successful in restoring patency in these
patients. Transradial blood pressure monitoring is useful for
assessing gradient pressure differences before and after stent
deployment. Residual gradient is an indication for additional
balloon dilation.

While infrequent, formation of a false aneurysm is a compli-
cation that requires a somewhat complex approach to allow
resolution, as illustrated in Fig. 49.8A-D. In this case, an endo-
luminal graft was used to correct a false aneurysm. While
endoluminal grafts are frequently used to treat abdominal
aortic aneurysms, they are not commonly used to correct
pathologies of the subclavian at this time. In the future,
endoluminal graft technology is likely to be used more often
in managing dissections, aneurysms, and traumatic injuries
in this anatomic region.

Overall, our experience with stenting has yielded implanta-
tion success nearing 100% for placement in the subclavian
and innominate arteries. 3,4,9,13 Deployment of stents at our
center has nearly always included ballooning prior to device
implantation. Follow-up ranging up to 4 years with duplex
Doppler scans and /or arteriography has confirmed patency
in the majority of cases.

Summary

Percutaneous management that incorporates angioplasty and
stents is revolutionizing the way vascular interventionists
treat brachiocephalic lesions. In general, the results of
angioplasty and stenting in these vessels have been very
encouraging, with low complication rates and good acute and
long-term patency rates. While there are still cases in which

combined procedures or surgical grafting may be appropriate,
endovascular stenting is now a first-line treatment for the
majority of patients with stenotic or occlusive disease in the
brachiocephalic vessels.

References

1. Diethrich EB, Garrett HE, Ameriso J, Crawford ES, el-Bayar M,
De Bakey ME. Occlusive disease of the common carotid and
subclavian arteries treated by carotid subclavian bypass. Analysis
of 125 cases. Am } Surg 1967; 114:800.

2. Diethrich EB, Koopot R. Simplified operative procedures for
proximal subclavian arterial lesions: direct subclavian-carotid
anastomosis. Am J Surg 1981; 142:416.

3. Diethrich EB. Initial experience with stenting in the innominate,
subclavian, and carotid arteries. JEndovasc Surg 1995; 2:196.

4. Diethrich EB, Cozacov JC. Subclavian stent implantation to
alleviate coronary steal through a patent internal mammary graft.
J Endovasc Surg 1995; 2:77.

5. Queral LA, Criado FJ. The treatment of focal aortic arch branch
lesions with Palmaz stents. / Vase Surg 1996; 23:368.

6. Kumar K, Dorros G, Bates MC, Palmer L, Mathiak L, Dufek C.
Primary stent deployment in occlusive subclavian artery
disease. Cathet Cardiovasc Diagn 1995; 34:281.

7. Motarjeme A. Percutaneous transluminal angioplasty of
supra-aortic vessels. ] Endovasc Surg 1996; 3:171.

8. Reubben A, Tettoni S, Muratore P et ah Feasibility of intraoperative
balloon angioplasty and additional stent placement of isolated
stenosis of the brachiocephalic trunk. / Thorac Cardiovasc Surg
1998; 115:1314.

9. Martinez R, Rodriguez-Lopez J, Torruella L, Ray L, Lopez-Galarza
L, Diethrich E. Stenting for occlusion of subclavian arteries:
technical aspects and follow-up results. Texas Heart J 1997; 24:
23.

10. Link J, Brossman J, Muller-Hulsbeck S, Heller M. PTA of the
brachiocephalic arteries. Aktuelle Radiol 1998; 8:76.

11. Sullivan TM, Gray, BH, Bacharach M et ah Angioplasty and
stenting of the subclavian, innominate, and common carotid
arteries in 82 patients. / Vase Surg 1998; 28:1059.

12. Buth J, Penn O, Tielbeek A, Mersman M. Combined approach to
stent-graft treatment of an aortic arch aneurysm. / Endovasc Surg
1998; 5:329.

13. Rodriguez-Lopez J, Werner A, Martinez R, Torruella LJ, Ray LI,
Diethrich EB. Stenting for atherosclerotic occlusive disease of the
subclavian artery. Ann Vase Surg 1999; 13:254.

14. Vranic M, Vaughn PL, Lobato AC, Rodriguez-Lopez J, Diethrich
EB. Intraoperative subclavian artery stenting to salvage a LIMA
graft. Ann Thorac Surg 1999; 68:2333.

15. Nomura M, Kida S, Yamashima T, Yamashita J, Yoshikawa J,
Matsui O. Percutaneous transluminal angioplasty and stent
placement for subclavian and brachiocephalic artery stenosis in
aortitis syndrome. Cardiovasc Interv Radiol 1999; 22:427.

16. Maskovic J, Jankovic S, Lusic I, Camj-Sapunar L, Mimica Z, Bacic
A. Subclavian artery stenosis caused by non-specific arteritis
(Takayasu disease): treatment with Palmaz stent. Eur } Radiol 1999;
31:193.

17. Al-Mubarak N, Liu MW, Dean LS et ah Immediate and late

573

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

outcomes of subclavian artery stenting. Cathet Cardiovasc Interv
1999;46:169.

18. Korner M, Baumgartner I, Do DD, Mahler F, Scroth G. PTA of the
subclavian and innominate arteries: long-term results. VASA 1999;
28:117.

19. d'Othee BJ, Rousseau H, Otal P, Joffre F. Noncovered stent
placement in a blunt traumatic injury of the subclavian artery.
Cardiovasc Interv Radiol 1999; 22:424.

20. Bruninx G, Wery D, Dubois E et al. Emergency endovascular treat-
ment of an acute traumatic rupture of the thoracic aorta compli-
cated by a distal low-flow syndrome. Cardiovasc Interv Radiol 1999;
22:515.

21. Whitbread T, Cleveland TJ, Beard JD, Gaines PA. A combined
approach to the treatment of proximal arterial occlusions of

the upper limb with endovascular stents. Eur J Vase Endovasc
Surg 1998; 15:29.

22. Hadjipetrou P, Cox SC, Piemonte T, Eisenhauer A. Percutaneous
atherosclerotic revascularization of atherosclerotic obstruction of
aortic arch vessels. J Am Coll Cardiol 1999; 33:1238.

23. Greenberg RK, Waldman D. Endovascular and open surgical
treatment of brachiocephalic arterial disease. Semin Vase Surg
1998; 11:77.

24. Eisenhauer AC. Subclavian and innominate revascularization:
surgical therapy versus catheter-based intervention. Curr Interv
Cardiol Rep 2000; 2:101.

25. Bosch JL, Lester JS, McMahon PM et al. Hospital costs for elective
endovascular and surgical repairs of infrarenal abdominal aortic
aneurysms. Radiology 2001; 220:492.

574

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Carotid endarterectomy compared with
carotid angioplasty and stenting

Mark R. Harrigan
Ricardo A. Hanel
Elad I. Levy
Lee R. Guterman
L. Nelson Hopkins

Carotid endarterectomy (CEA) has an uncommon distinction
among surgical procedures in that it has been shown in ran-
domized clinical trials to reduce significantly the risk of stroke
in selected patients. The efficacy of CEA in patients with
both asymptomatic and symptomatic carotid stenosis has
been proven in the North American Symptomatic Carotid En-
darterectomy Trial (NASCET), 1 the European Carotid Surgery
Trial (ECST), 2 and the Asymptomatic Carotid Atherosclerosis
Study (AC AS). 3 Carotid angioplasty in conjunction with stent-
ing (CAS) has emerged in recent years as a less invasive alter-
native to CEA, and randomized trials are currently in progress
to compare CEA with CAS. In this chapter, the indications for
treatment of cervical carotid disease are reviewed, the surgical
and endovascular techniques used for carotid revasculariza-
tion at the University at Buffalo Department of Neurosurgery
are described, and the advantages and disadvantages of each
approach are discussed.

Carotid endarterectomy

Indications

Symptomatic carotid stenosis

In the NASCET, 2885 patients with transient ischemic attack
(TIA) or minor stroke within the previous 120 days who had a
30-99% ipsilateral internal carotid artery (ICA) stenosis were
randomized to receive either medical therapy (risk factor
modification and aspirin 1300 mg daily) or medical therapy
and CEA. 1 Stenosis was measured on angiography by com-
paring the residual lumen diameter in the most stenotic por-
tion of the ICA with the lumen diameter of the ICA distal to the
stenosis (this method has been used for all randomized trials
of CEA except for the ECST). The arm of the trial for patients
with >70% stenosis was terminated before the end of the study
because an interim analysis showed a considerable advantage
of surgery. For patients with >70% stenosis, the ipsilateral

stroke rate at 2 years was 26% in the medical group but only 9%
in the surgical group, a relative risk reduction of 65%. For these
patients, the benefit persisted for at least 8 years. 4 The risk re-
duction correlated with the degree of stenosis. For moderate
stenosis (50-69%), the 5-year risk of ipsilateral stroke was
15.7% with surgical treatment and 22.2% with medical treat-
ment. 4 In addition, the benefit of surgery included patients
whose only symptom was amaurosis fugax, as well as those
with hemispheric ischemic symptoms. 5

A significant benefit of surgery was also demonstrated by
the ECST in which 3024 patients with TIA, retinal infarction, or
nondisabling stroke within the previous 6 months were ran-
domized to receive either medical therapy (use of aspirin was
permitted but not required) or medical therapy and CEA. 2 In
contrast to the NASCET, stenosis in the ECST was determined
on angiography by comparing the residual stenosis at the
most stenotic portion of the vessel with the probable original
lumen diameter at that site. Consequently, higher degrees of
stenosis were reported in the ECST relative to NASCET angio-
graphic measurements. For instance, 85% stenosis by the
ECST method is approximately 70% stenosis by NASCET cri-
teria. In the ECST, the 3-year risk of major stroke or death in
patients with >80% (approximately >60% by the NASCET
method) symptomatic carotid stenosis was 26.5% in the
medical group and 14.9% in the surgical group, an absolute
risk reduction for surgery of 11.6%. 6 This risk reduction
persisted for at least 10 years after surgery. 7

The Veterans Affairs Cooperative Study on Symptomatic
Stenosis (VACS) compared medical therapy, including aspirin
(325 mg daily), with CEA. 8 One hundred and ninety-seven
men were enrolled in this study before it was prematurely
terminated when NASCET and ECST data were released.
Despite the relatively small size of this study, absolute risk
reductions of 17.7% in patients with >70% stenosis and 11.7%
in patients with >50% stenosis were found.

The benefit of CEA for patients with symptomatic stenosis
was confirmed by a meta-analysis of these three trials and
applied similarly to men and women. 9 CEA appears to reduce

575

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

the risk of stroke in selected patients who have symptomatic
stenosis >50% (measured by the NASCET method). Another
important result of these trials is the finding that outcome after
CEA depends on risk for perioperative complications. In the
NASCET, the 30-day rate of disabling stroke and death was
2.9%. Similarly, in the ECST, the 30-day rate of death was 1.0%,
and the disabling stroke rate was 2.5%. 6 These findings led to a
consensus statement from the American Heart Association
recommending that CEA in symptomatic patients be under-
taken by surgeons whose surgical morbidity and mortality
rate is <6% and in asymptomatic patients by surgeons whose
surgical morbidity and mortality rate is <3%. 10

Asymptomatic carotid stenosis

Four randomized controlled trials have examined the efficacy
of CEA for asymptomatic patients. In the Asymptomatic
Carotid Atherosclerosis Study (ACAS), 1662 patients with
>60% stenosis, defined by either angiography or carotid du-
plex ultrasonography, were randomized to CEA or medical
treatment. 3 All patients received aspirin, 325 mg daily. The
study was stopped prematurely after a median follow-up of
2.7 years because the aggregate risk over 5 years for ipsilateral
stroke and any perioperative stroke or death was estimated to
be 5.1% for the surgical group and 11.0% for the medical group,
a risk reduction of 53%. This benefit was statistically signifi-
cant for men but not women.

In the Veterans Administration Cooperative Asymptomatic
Trial, 444 men with >50% stenosis were randomized to receive
medical therapy (recommended: aspirin 1300 mg daily) or
medical therapy with CEA. 11 After a 4-year follow-up period,
the combined incidence of ipsilateral neurologic events was
significantly lower in the surgical group (8.0% vs. 20.6%). The
high overall mortality rate of 33%, primarily owing to coro-
nary atherosclerosis, suggests that the study population dif-
fered from those in other trials and makes interpretation of
these results difficult. 12 In the Carotid Artery Stenosis
with Asymptomatic Narrowing: Operation Versus Aspirin
(CASANOVA) trial, 410 patients with 50-90% stenosis were
randomized to receive CEA and medical therapy or medical
therapy only and followed for a mean interval of 42 months. 13
All patients received aspirin, 330 mg, and dipyridamole,
75 mg, three times daily. Although no significant difference
was found in stroke rates in the medical and surgical groups,
the small size of the study and the fact that a significant
number of crossovers occurred between the groups obscures
the importance of the findings. In the Mayo Asymptomatic
Carotid Endarterectomy Study, patients were randomized
to receive medical therapy with aspirin or CEA without
aspirin. 14 This study was terminated early owing to the
significantly higher number of myocardial infarctions (Mis)
and transient cerebral ischemic events that occurred in the
surgical group, presumably because this group did not receive
aspirin.

Technique

Carotid surgery is most effective when patients are selected
appropriately. In the three randomized trials for symptomatic
carotid stenosis, patient eligibility was based on angiographic
criteria for evaluating stenosis. Thus, preoperative evaluation
of carotid stenosis must match the accuracy of conventional
angiography. Furthermore, CEA in asymptomatic patients
carries a narrow risk-benefit ratio, making accurate patient
selection essential. Carotid duplex ultrasonography is a
useful screening method, with sensitivity and specificity
in some vascular laboratories reaching 94% and 89%, respec-
tively, for detection of 70-99% stenosis. 15 However, duplex
ultrasonography has several limitations that necessitate
additional confirmatory studies in the preoperative evalua-
tion of patients with carotid stenosis. A significant proportion
of CEAs are performed in general practice settings lacking
designated, accredited vascular laboratories. 16 ' 17 Even accred-
ited, high-volume vascular laboratories may report false-
positive results for carotid stenosis ranging from 20% to 41%. 18
Duplex scanning cannot be used to distinguish accurately pre-
occlusive disease from total occlusion 19 ' 20 or to image the
distal ICA and intracranial vasculature, which is necessary to
identify tandem lesions and other vascular abnormalities. In
addition, duplex scanning does not indicate whether the le-
sion is relatively high in the cervical region, which is informa-
tion that is important for surgical planning. At the authors 7
center, all patients with evidence of carotid stenosis undergo
cerebral angiography. A large volume of cases and recent im-
provements in angiographic technique, such as radial artery
access, 21 and improvements in equipment, such as the
introduction of hydrophilic catheters, have led to a very low
complication rate. 18 Alternatively, magnetic resonance an-
giography 22 ' 23 and computed tomography (CT) angiogra-
phy 24 can also be used to confirm the results of carotid duplex
imaging with a high degree of accuracy.

Medical management

Medical management of carotid artery disease begins with
modification of risk factors, including smoking cessation, con-
trol of diabetes, and reduction of cholesterol. Treatment of
hypertension reduces the risk of stroke but caution should
be exercised in patients with high-grade hemodynamically
carotid stenosis because hypotension can evoke cerebral
ischemia. Platelet antiaggregation therapy with low-dose
aspirin (30-283 mg daily) has been shown to reduce the inci-
dence of stroke in asymptomatic patients with coronary artery
disease 25 and in patients with TLA. 26 ' 27 The importance of as-
pirin use in patients with asymptomatic carotid stenosis was
demonstrated by the aforementioned Mayo Asymptomatic
Carotid Endarterectomy Study, which was terminated be-
cause a significantly higher number of Mis and transient
cerebral ischemic events occurred in the surgical group,

576

chapter 50 CEA compared with CAS

presumably due to the absence of aspirin use in the surgical
group. 14 Although some authors recommend high-dose
aspirin for the prevention of stroke, a recent trial found that
the risk of stroke, MI, and death within 3 months of CEA
was lower for patients taking 81 mg or 325 mg aspirin daily
than for those taking >650 mg daily. 28 The authors 7 preference
is to prescribe aspirin (325 mg daily) for patients with evidence
of carotid artery stenosis.

The cholesterol-lowering medications HMG-CoA reduc-
tase inhibitors have been shown to stabilize carotid
plaques 29,30 and lower stroke risk in patients with coronary
artery disease. 31 Patients placed on these medications should
be informed about myopathy, which can affect a small
percentage of patients, and monitored for signs of this
condition. 32

Surgical technique

A variety of techniques are currently in use for CEA. Excellent
surgical results have been reported using local or general
anesthesia, routine shunting, simple or patch graft closure,
heparin administration, and electroencephalography (EEG)
monitoring. Indeed, in the NASCET, none of these procedures
affected the overall morbidity of the operation. 33 The follow-
ing is the authors' preferred technique for CEA.

The patient is placed under general endotracheal anesthesia
with EEG monitoring. For patients with a relatively high
carotid bifurcation at the level of the C2 vertebra or higher,
nasal intubation is used to optimize exposure behind the angle
of the mandible. The patient is positioned supine, with the
head slightly extended and rotated to the side contralateral to
the lesion that will be operated on. EEG electrodes are placed,
and baseline values are recorded. Blood pressure is main-
tained at the patient's baseline range. The S-shaped incision is
made parallel to the anterior border of the sternocleidomas-
toid muscle, curved posteriorly toward the mastoid and ante-
riorly into a cervical skin crease inferiorly. The anterior edge of
the sternocleidomastoid muscle is dissected free of surround-
ing connective tissue and is retracted laterally. The common
facial vein is ligated after inspection underneath the vein to en-
sure that the hypoglossal nerve is not injured. The internal
jugular vein is mobilized and retracted medially. The carotid
sheath is opened longitudinally; the vagus nerve is identified;
and the common carotid artery (CCA) is exposed. The vessel
is mobilized; and the bifurcation, ICA, external carotid
artery (ECA), and proximal branches of the ECA are exposed
circumferentially

The segment of the carotid artery containing a calcified
plaque can usually be identified with gentle palpation. The
carotid body is infiltrated with 1% lidocaine to inhibit the
baroreceptor response. Prior to the application of temporary
vessel clamps, the operating microscope is placed in the field;
heparin (5000 U) is administered intravenously; the blood
pressure is slightly elevated by adjusting the anesthesia or

adding an intravenous vasopressor; and barbiturates are used
to obtain burst suppression on EEG. Temporary clamps are ap-
plied to the ICA, the CCA, the ECA, and the superior thyroid
artery. Under magnification, an incision is made through the
wall of the CCA, just below the plaque; and Potts scissors are
used to extend the arteriotomy distally into the ICA, just past
the distal extent of the plaque. If EEG changes occur with
clamping (this happens rarely), a shunt is placed. The distal
end of the shunt tubing is placed into the ICA lumen first, and
back-bleeding to flush air and debris from the tubing is
allowed to occur while the proximal end of the shunt tubing
is inserted into the CCA lumen.

In the CCA, a plane between the proximal edge of the plaque
and the vessel wall is identified and developed circumferen-
tially. The plaque is then gently dissected away from the inner
surface of the vessel, working distally into the ICA. The distal
portion of the plaque is removed to allow a smooth, tapered
transition into normal intima. Any remaining intimal flaps
must be trimmed or tacked to the wall with sutures. The entire
exposed vessel surface is irrigated with heparinized saline and
examined and cleared of excess tissue under magnification.
Beginning at the distal apex, the arteriotomy is closed with a
running 6-0 monofilament suture, using the operating micro-
scope and small "bites" to avoid compromise of the ICA
lumen. The normal diameter of the vessel must be preserved.
Prior to completing the closure, each vessel is briefly un-
damped, and debris is flushed from the lumen. Following
closure of the arteriotomy, the superior thyroid artery is left
undamped and the suture line is inspected, with additional
sutures placed if necessary. The ECA and CCA clamps are
removed in sequence, to permit any remaining debris to
be flushed into the ECA. The ICA clamp is removed last.
Microfibrillar collagen hemostat is placed over the suture
line to promote hemostasis, and the wound is closed with
absorbable sutures.

Carotid angioplasty and stenting

Historical background

Endovascular treatment of cervical carotid stenosis has been
made possible by rapid developments in endovascular tech-
nology for other applications in recent decades. Percutaneous
transluminal angioplasty has been established as an alterna-
tive to surgical repair in patients with coronary artery disease
and peripheral vascular disease. Angioplasty for carotid
stenosis was first performed in the early 1980s. 34 ' 35 In contrast
to CEA, in which the atherosclerotic plaque is removed, angio-
plasty within a stenotic lesion results in fracture of the plaque
and stretching of the media. 36 With angioplasty alone, how-
ever, plaque fragments may embolize into the intracranial
circulation; and the resultant irregularities within the plaque
can serve as thrombogenic sites before remodeling and

577

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

endothelialization can occur. 37 ' 38 Indeed, high rates of
postprocedural neurologic events as a consequence of
embolization were described in the initial reports of carotid
angioplasty. 39 ' 40

Intravascular stenting evolved simultaneously with refine-
ments in carotid angioplasty, driven by a need for improve-
ment in coronary balloon angioplasty, in which acute
occlusion and restenosis are problematic. Following the first
report of carotid stenting in 1995, 41 stenting was seen as a nec-
essary adjunct to carotid angioplasty to minimize the risk of
embolization. Stenting in combination with angioplasty has
been rapidly adopted as the endovascular treatment of choice
for carotid stenosis. 42 However, stents did not eliminate the
problem of embolization. An early trial of CAS vs. CEA
was stopped prematurely because of a high stroke rate in the
CAS group. 43 Distal protection techniques have evolved
to prevent embolization during CAS. The first report
of a distal protection technique described a triple coaxial
catheter with a latex balloon mounted at the distal end. 44 The
ICA was occluded during stent placement, and debris was
flushed and aspirated after stent deployment. Since then,
ICA filters and flow-reversal techniques have also been
introduced. The addition of antiplatelet medications such as
clopidogrel and glycoprotein (GP) lib /Ilia inhibitors have
also served to prevent and treat embolization during and after
CAS. 45 ' 46

Preliminary evidence from nonrandomized trials and reg-
istries indicate that there is a similar morbidity rate but an
overall lower cost and shorter hospital stay following CAS
compared with CEA. In a nonrandomized study of patients
undergoing either CEA or CAS, the CAS group had a signifi-
cantly shorter length of stay in the hospital. 47 There was also a
statistical trend toward a more frequent rate of major ipsilater-
al stroke and death in the surgical group compared with the
CAS group (2.9% vs. 0%, P = 0.10). In a multicenter registry of
5210 endovascular carotid stent procedures involving 4757
patients, technical success was achieved in 98.4% of 5129
carotid arteries treated. 48 The periprocedural minor and
major stroke rates were 2.7% and 1.5%, respectively; and the
mortality rate was 0.86%. Restenosis rates after CAS were 2%
and 3.5% at 6 and 12 months, respectively. In a report of more
than 500 carotid stent procedures for both asymptomatic and
symptomatic carotid stenosis, the perioperative major and
minor stroke rates were 1% and 4.8%, respectively 49 The 30-
day stroke and death rate was 7.4%. Over the 5-year study
period, the periprocedural stroke rate improved from 7.1% to
3.1%. Currently, several randomized multicenter trials are
under way to evaluate the efficacy of CAS with adjunctive
distal protection, primarily in high-risk patients (Table 50.1).
Although preliminary reports of CAS are encouraging, wide-
spread use of this procedure should await the results of the
clinical trials. A consensus statement by the American Heart
Association recommended that the use of CAS be limited to
randomized trials. 50

Technique

Medical management with carotid angioplasty
and stenting

All endovascular procedures carry risk of intimal injury and
subsequent thrombosis and vessel occlusion. Stenting may
elevate this risk: angioplasty produces deep arterial injury, 51
and stents are thrombogenic. 52 Therefore, patient preparation
for stenting hinges on adequate antiplatelet and anticoagula-
tion therapy. However, selection and dosing of antithrombotic
medications must also minimize the risk of hemorrhagic
complications. Most information about treatment with these
medications must be gleaned from the cardiac literature
because clinical data in the neurosurgical literature are limited.

Aspirin is a cyclooxygenase-1 inhibitor that irreversibly in-
hibits platelet aggregation but does not impede platelet adhe-
sion or platelet-activated mitogenic activity. Clopidogrel is a
thienopyridine derivative with potent antiplatelet action that
inhibits adenosine phosphate-induced platelet aggregation.
This drug works synergistically with aspirin, and evidence
from the cardiac literature supports the use of combination
antiplatelet regimens. 53,54 Clopidogrel, in combination with
aspirin, has become the standard treatment for patients
undergoing coronary angioplasty and stenting. 55 When possi-
ble, patients should be placed on aspirin (325 mg daily) and
clopidogrel (75 mg daily) for at least 3 days before CAS or be
given a loading dose of clopidogrel (300 mg) early on the day
of the procedure.

For most intracranial stent procedures, an intravenous
bolus dose of heparin (70U/kg) is administered following
catheterization of the CCA. In addition, all saline solutions to
be used for irrigation of catheters should be prepared with
heparin (lU/ml). The activated coagulation time should be
kept between 250 and 300 s for the duration of the procedure.

Platelet GPIIb/IIIa inhibitors, such as abciximab or eptifi-
batide, block the final common pathway of platelet aggrega-
tion by preventing the binding of fibrinogen to platelets and
are the most potent of the antiplatelet drugs. Preliminary data
at the authors' center suggest that patients with chronically is-
chemic brain are at an elevated risk of intracranial hemorrhage
with GPIIb/IIIa inhibitors, therefore these drugs should be
reserved for patients who experience thromboembolic com-
plications during or soon after the procedure. Abciximab can
be given as an initial loading dose of 0.25mg/kg, followed
by a 12-h intravenous infusion at a rate of 10 jig/min. Alterna-
tively, eptifibatide may be administered with a loading dose of
135|ig/kg followed by a 20- to 24-h infusion of 0.5 |ig/kg.
When GPIIb/IIIa inhibitors are used, we recommend
obtaining a CT scan immediately after the procedure to
check for intracerebral hemorrhage before proceeding with
the postprocedure infusion.

Bradycardia occurs occasionally during angioplasty,
particularly when the plaque involves the carotid sinus.

578

chapter 50 CEA compared with CAS

Table 50.1 Carotid angioplasty and stent trials

Clinical characteristics and

Distal

Study

Design

percentage of stenosis

Stent

protection device

ARCHER (Guidant)

Prospective single-arm registry

High risk

Asymptomatic >80%
Symptomatic>50%

Acculink

Accunet

BEACH (Boston Scientific)

Prospective single-arm registry

High risk

Asymptomatic >80%
Symptomatic>50%

Monorail
Wallstent

EPI FilterWire

CABERNET (Boston

Prospective single-arm registry

High risk

NexStent

EPI FilterWire

Scientific and EndoTex)

Asymptomatic >60%
Symptomatic>50%

CARESS (excludes

Prospective comparative

Asymptomatic>75%

Randomized assignment

Randomized assignment to

CREST patients)

consecutive entry

Symptomatic>50%

to available devices

available devices at the

2:1 CEA: CAS

at the center

center

CREST (NIH, Guidant)

Randomized trial

Symptomatic>50% stenosis

Acculink

Accunet

MAVEriC (Medtronic AVE)

Prospective single-arm registry

High risk

Asymptomatic >80%
Symptomatic>50%

MAVEriC

PercuSurge

SAPPHIRE (Cordis)

Prospective registry alongside
a randomized trial

High risk (age >80 years

alone qualifies)
Asymptomatic >80%
Symptomatic>50%

Precise

Angioguard

SECURITY (Perclose)

Prospective single-arm registry

High risk

Asymptomatic >80%
Symptomatic>50%

X.act

NeuroShield

ARCHER, Acculink for Revascularization of Carotids in High Risk Patients; BEACH, Boston Scientific EPI: A Carotid Stent for High Risk Surgical Patients; CAS,
carotid artery angioplasty and stenting; CABERNET, Carotid Artery Revascularization using Boston Scientific EPI FilterWire and Endolex Stent; CARESS, Carotid
Revascularization with Endarterectomy or Stenting Systems; CEA, carotid endarterectomy; CREST, Carotid Revascularization Endarterectomyvs. Stent Trial;
MAVEriC, Evaluation of the Medtronic AVE Self-expanding Carotid Stent System with Distal Protection in the Treatment of Carotid Stenosis; NIH, National
Institutes of Health; SAPPHIRE, Stenting and Angioplasty with Protection in Patients at High-Risk for Endarterectomy.

Adapted with permission from Levy El, Kim SH, Bendok BR eta/. Interventional neuroradiologic therapy. In: MohrJP, Choi DW, GrottaJC, Weir B, Wolf PA, eds.
Stroke: Patholophysiology, Diagnosis, and Management, 4th edn. Philadelphia: Churchill Livingstone; 2004: 1497.

Atropine and a preprepared dopamine solution are kept avail-
able should significant bradycardia and hypotension appear.
Medical management of bradycardia during angioplasty is
usually sufficient; we do not routinely place transvenous
pacemakers before performing CAS.

Following stent placement, heparin therapy is usually dis-
continued but not reversed with protamine. In some situa-
tions, such as when an angiographically visible dissection or
thrombus is present, continued infusion of heparin to main-
tain the activated prothrombin time 1.5-2.3 times the baseline
value is appropriate. Aspirin (325 mg daily) and clopidogrel
(75 mg daily) should be administered for at least 4 weeks to
allow for complete endothelialization of the stent. 56 Aspirin is
continued indefinitely.

Endovascular technique

The technique of CAS varies slightly from case to case, de-

pending on the clinical situation. The following is a general
outline of the procedure used at the authors' center for most
patients.

The procedure is performed in an angiography suite with
biplane digital subtraction and fluoroscopic imaging cap-
abilities. The patient is kept awake, with local anesthesia and
sedatives administered to permit continuous neurologic
assessment. Dorsalis pedis and posterior tibialis pulses are
assessed and marked for later reference, a practice that is par-
ticularly important in patients with coexistent peripheral vas-
cular disease. A Foley catheter and two peripheral intravenous
lines are placed. A 5-Fr sheath is placed in the right femoral
artery, and a three-vessel diagnostic angiogram is obtained
using a 5-Fr Simmons-2 or angled glide catheter. An intracra-
nial angiogram with injection of contrast into the ipsilateral
CCA is necessary for later comparison should thromboem-
bolism within the intracranial circulation be suspected after
angioplasty. Prior to placement of the guide catheter in the

579

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

CCA, the loading dose of heparin is given. When the activated
coagulation time reaches at least 250 s, the diagnostic catheter
is positioned in the CCA or the ECA and is used to place a
0.035-in stiff 300-cm guidewire into the distal ECA. In the
setting of ECA stenosis or occlusion, a "J" wire is placed in
the distal CCA and is used to provide support for the guide
sheath. When the stiff guidewire has been positioned, the 5-Fr
groin sheath is exchanged, over a tapered obturator, for a 7-Fr
sheath (Cook, Bloomington, IN, USA). The distal end of this
system is positioned just proximal to the carotid bifurca-
tion. All catheter systems are flushed continuously with
heparinized saline.

Once the guide catheter is in place, we embark on a four-
stage procedure for CAS. First, the distal protection device is
positioned. Second, prestent deployment angioplasty is per-
formed to enlarge the stenotic region sufficiently to permit
passage of the stent. Third, the stent is deployed; and fourth,
poststent deployment angioplasty is carried out to remodel
and fully expand the stent. After each step, high-resolution
biplanar angiograms are obtained and neurologic exams are
performed to allow for prompt recognition of any changes
from the patient's baseline status.

Three classes of distal protection techniques are currently
practiced. In the retrievable filter technique, a filter designed
to collect debris during CAS without interrupting flow
within the ICAis placed distal to the stenotic region. Examples
of filter devices include the EPI FilterWire (Boston Scientific
Embolic Protection Inc., San Carlos, CA, USA), Accunet
(Guidant, Menlo Park, CA, USA), Angioguard (Cordis
Neurovascular, Miami Lakes, FL, USA), and Mednova
(Abbott Laboratories, Abbott Park, IL, USA). Balloon occlu-
sion techniques involve inflation of a balloon and interruption
of flow in the ICA distal to the stenosis for the duration of
the stenting procedure. An example is the PercuSurge Balloon
(PercuSurge GuardWire; Medtronic AVE, Santa Rosa, CA,
USA). The flow-reversal technique involves placement of
balloons in the ECA and CCA to interrupt flow in these vessels
and cause retrograde flow in the ICA to prevent embolization
into the intracranial circulation. 57

After the guide catheter is positioned, a distal protection
device (the authors' preference is to use a retrievable filter)
mounted on a 0.014-in microguidewire is carefully guided
across the stenotic region, using biplanar roadmapping tech-
nique, and then deployed. High-resolution angiography of
the cervical carotid artery is done; and measurements are
made of the length of the lesion, as well as the diameters of the
carotid artery proximal and distal to the lesion. The distal pro-
tection device is then advanced over the microguidewire and
deployed distal to the stenotic region. An angioplasty balloon
is selected based on the dimensions of the lesion. The balloon
must be long enough to cover the entire length of the lesion,
and the inflation diameter should be undersized to avoid over-
inflation and to open the artery just enough to allow passage of
the stent. After an angiogram of the cervical carotid is obtained

with the distal protection device in place, the angioplasty
balloon is advanced and centered on the lesion. The balloon is
inflated to the manufacturer's recommended nominal pres-
sure for several seconds and then deflated. The blood pressure
cuff is placed on continuous mode during angioplasty to allow
rapid sequential measurement of blood pressure should
bradycardia and hypotension occur.

Most stents currently in use for CAS are self-expanding
stents such as the Wallstent (Boston Scientific Scimed, Maple
Grove, MN, USA), Acculink (Guidant/ Advanced Cardiovas-
cular Systems, Temecula, CA, USA), and Precise (Cordis
Neurovascular) stents. The Wallstent is composed of stainless
steel, and the Acculink and Precise stents are made of nitinol,
a nickel and titanium alloy. Selection of the stent is deter-
mined by lesion length and the normal diameter of the artery.
The stent should be oversized by 1-2 mm more than the nor-
mal arterial caliber and should completely cover the lesion. At
diameters less than full expansion, nitinol stents exert a
chronic outward radial force that serves to maintain apposi-
tion of the stent to the vessel wall after deployment. Often the
stent will extend from the CCA into the ICA, crossing the bi-
furcation and origin of the ECA; in these cases, the stent should
be sized according to the larger caliber of the CCA.

After the stent is in place, poststent deployment angioplasty
is performed. The distal protection device is withdrawn, and
a final series of cervical carotid and intracranial circulation
angiograms is obtained. The catheter systems and femoral
sheath are removed, and a percutaneous closure device such
as the Perclose device (Redwood City, CA, USA) is used to
close the femoral artery puncture site. Following the proce-
dure, the patient is admitted to the intensive care unit for
monitoring overnight. Hourly neurologic checks and close
surveillance of hemodynamic parameters are important. Most
patients are discharged to home on the day after the proce-
dure. As previously mentioned, aspirin and clopidogrel are
prescribed.

Rationale for carotid angioplasty
and stenting

CEA has an established role in the prevention of stroke. How-
ever, CEA does carry significant risk. In the clinical trials for
symptomatic carotid stenosis, morbidity and mortality in the
first month following randomization were higher for patients
in the surgical groups compared with the medically treated
groups. 9 Endovascular treatment of carotid stenosis may be an
attractive alternative to open surgery for patients desiring a
less invasive procedure. Endovascular treatment of carotid
artery disease may also be more cost effective than CEA. In a
comparison of cost and length of stay for CEA vs. CAS, CEA
was both more expensive ($5409 vs. $3417) and associated
with a longer length of stay (3.0 vs. 1.4 days). 47 The rationale
for developing CAS converges on three lines of evidence: (i)

580

the results of the CEA clinical trials have important limita-
tions, (ii) patients at high risk for surgery may benefit from a
less invasive procedure, and (iii) anatomic and other neu-
rovascular considerations in certain patients make CAS more
feasible than surgery

Limitations of CEA clinical trial results

The organizers of the CEA clinical trials sought to eliminate
factors that might obscure the interpretation of the study re-
sults. For instance, in the NASCET, exclusion criteria included
age older than 79 years; a previous ipsilateral endarterectomy;
an intracranial stenosis more severe than the surgically acces-
sible lesion; lung, liver, or renal failure; and lack of angio-
graphic depiction of both carotid arteries and their intracranial
branches. 1 Other exclusion criteria included hypertension;
unstable angina pectoris; MI within the previous 6 months;
contralateral CEA within the previous 4 months; signs of pro-
gressive neurologic dysfunction; or a major surgical proce-
dure within the previous 30 days. The rigid selection criteria
used by these trials make the application of their results to
common practice problematic. In fact, patients with these con-
ditions have been found to have a higher risk for perioperative
complications with CEA. 58-61 Several clinical series of CAS,
consisting mostly of patients who would have been excluded
from the CEA trials, have shown results comparable or
superior to CEA. 42/62

Perioperative complication rates in clinical practice often
exceed the rates obtained in clinical trials. In a study of
Medicare patients undergoing CEA during 1992 and 1993 in
trial hospitals (participating in NASCET and ACAS) and in
nontrial hospitals, the perioperative mortality rate was
significantly greater in the nontrial hospitals. 63 High-volume
centers experience better outcomes than low-volume cen-
ters, 63-65 yet 40-90% of CEAs are done at low-volume cen-
ters. 66-68 In addition, surgeons tend to self-report lower
complication rates than independent observers. In a prospec-
tive series of patients receiving CEA at an academic medical
center and evaluated by neurologists, the 30-day major stroke
or death rates were 11.1% and 5.6% for symptomatic and
asymptomatic patients, respectively 69

High-risk surgical candidates

Patients with significant medical comorbidities are at ele-
vated risk of complications associated with CEA. For instance,
patients with a previous history of MI, angina, or hypertension
are approximately 1.5 times more likely to have medical
complications with CEA than are patients without these
medical problems. 61

Heart disease

Coronary artery disease is the leading cause of death in pa-

CHAPTER 50 CEA compared with CAS

tients with carotid artery disease, 70 ' 71 even for patients who
have undergone CEA. 11 Preexisting coronary artery disease is
associated with cardiac complications 72-74 or death 74 ' 75 in con-
junction with CEA. The incidence of MI in the setting of CEA
ranges from 1% to 4%. 76/77 Patients with coronary artery dis-
ease are also at elevated risk of perioperative stroke and death,
with an incidence as high as 25% 78 or even 40% 79 in some
series. In addition, congestive heart failure is also an indepen-
dent risk factor for stroke or death in conjunction with
CEA. 60/78 Avoiding surgery or general anesthesia for patients
with heart disease by performing CAS may represent a valid
alternative. 42,62 ' 80

Conversely, patients with severe carotid artery disease who
undergo coronary artery bypass graft (CABG) procedures are
at an elevated risk for stroke during cardiopulmonary by-
pass. 81-83 The coexistence of significant carotid artery stenosis
and symptomatic coronary artery disease presents the physi-
cian with a management dilemma. 61 ' 82 Options include per-
formance of a simultaneous procedure or a staged approach in
which one procedure is performed several days after the other.
Published reports on CEA and CABG combined suggest that
the risk of stroke or death ranges from 7.4% to 9.4%, which is
roughly 1.5-2.0 times the risk of each operation alone. 61 In a
meta-analysis of 56 reports of patients undergoing both CEA
and CABG, simultaneous procedures had the best overall re-
sults, with a perioperative stroke rate of 6.2%, a MI rate of 4.7%,
and a mortality rate of 5 .6%. 10 CEA followed by CABG had sig-
nificantly higher rates of MI and death (11.5% and 9.4%, re-
spectively), and CABG followed by CEA had a significantly
higher rate of stroke (10.0%). Preliminary evidence suggests
that CAS followed by CABG may be a safer alternative for pa-
tients who require both procedures. In a series of 49 patients
undergoing CAS prior to CABG, four (8%) patients died of car-
diac arrest and one patient (2%) suffered a major stroke within
30 days after the CABG procedure. 84

Elderly patients

Data from the NASCET indicate that patients aged >75 years
derived a greater benefit from CEA than did those in younger
age groups. 85 However, older patients have higher rates of
perioperative morbidity and mortality with CEA. Mortality
for patients >85 years is three times higher than for those
younger than 70 years. 63 The postoperative stroke or death
rate for patients with asymptomatic carotid stenosis who are
>75 years is 7.5%, compared with 1.8% in patients younger
than 75 years. 60 Similarly, the risk of postoperative MI
associated with CEA was 6.6% in symptomatic patients
>75 years vs. 2.3% in patients younger than 75 years. 59 Carotid
revascularization with a less invasive approach, such as CAS,
may carry a lower perioperative risk for elderly patients
than surgery.

581

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Neurovascular considerations

High cervical stenosis and tandem lesions

Anatomic features that can increase the technical difficulty of
CEA include a carotid bifurcation at or above the level of the
second cervical vertebra, a short or thick neck, cervical
spondylosis that limits rotation of the neck, and a carotid
plaque that extends to the skull base. CAS can avert these
difficulties.

Tandem stenoses in the ICA have been identified in up to
20% of patients with cervical carotid stenosis 86-88 and up to
one-third of patients with symptomatic cervical carotid steno-
sis. 89 The presence of tandem lesions, in which the distal lesion
is more severe than the proximal lesion, was an exclusion crite-
rion for the NASCET. In a review of 1160 CEAs in symptomatic
patients, including 65 patients with ipsilateral carotid siphon
stenosis, there was a statistical trend toward a higher rate of
adverse outcomes in patients with tandem lesions (13.9% vs.
7.9%, P = 0.10). 59 In a series of 11 patients with tandem lesions
who underwent CAS, no perioperative stroke, cardiac, or mor-
tality occurred, suggesting that CAS is a viable alternative to
CEA for these patients. 90

Contralateral carotid occlusion

Approximately 14% of patients with significant carotid steno-
sis have contralateral carotid artery occlusion, 91 and are at
high risk of stroke. In the NASCET, the risk of ipsilateral stroke
in medically treated patients with severe stenosis of the
symptomatic carotid artery and occlusion of the contralateral
carotid artery was 69 .4% at 2 years. 92 Although CEA led to a re-
duction in the risk of stroke in this group of patients, the peri-
operative risk of stroke or death was 14.3%. Reduced cerebral
blood flow can occur during CEA, even when a shunt is used, 93
and this can place patients with limited cerebrovascular re-
serve at elevated risk of ischemia. Endovascular treatment
may minimize or eliminate alterations in cerebral blood flow
during treatment in patients with contralateral carotid occlu-
sion. In a series of 26 patients treated with carotid stenting in
the presence of contralateral carotid occlusion, there was one
minor stroke (3.8%) and no deaths, major strokes, Mis, or
vascular access site complications. 94 In another series, 23
patients with contralateral occlusion underwent CAS with no
perioperative strokes or deaths. 95

Carotid artery stenosis with intraluminal thrombus

CEA in the presence of an intraluminal thrombus superim-
posed on an atherosclerotic plaque can be hazardous. 59,96
Fresh thrombus can dislodge during dissection and clamping
of the carotid artery. In the NASCET, patients with intralumi-
nal thrombus who underwent CEA had a perioperative risk of
stroke or death of 12%. 97 An endovascular approach in this set-

ting affords the opportunity to combine intraarterial throm-
bolysis with CAS. 98 In addition, for patients who experience
embolization of carotid thrombus into the intracranial circula-
tion, intravenous antiplatelet agents such as GPIIb/IIIa
inhibitors can be administered during CAS; this may not be
an option during CEA because of the potential for bleeding
complications.

Radiation-induced carotid stenosis

Radiation therapy concentrated at the cervical region dam-
ages large arteries and leads to atherosclerosis-like stenotic
disease. 99 ' 100 CEA in this situation is impeded by relatively
long lesions, scarring around the vessels, and poorly defined
dissection planes, 101 ' 102 which elevates the risk of periopera-
tive complications. 103 Carotid angioplasty and stent place-
ment can provide a more effective and less morbid approach
in this setting. Several series have reported good results
after CAS for patients with radiation-induced carotid
stenosis. 104-109

Restenosis after carotid endarterectomy

Recurrent carotid artery stenosis takes two forms. Early
restenosis, occurring within 2 years of CEA, is characterized
by myointimal cell proliferation. Diffuse intimal thickening of
the intima and media results in fibrous hypertrophic scarring
throughout the CEA site. Stenosis of this type usually has a
smooth, firm, nonulcerated appearance. Late restenosis is
the result of a reaccumulation of atherosclerotic plaque and is
typically friable and ulcerated in appearance. Reports on the
incidence of recurrent stenosis vary widely, probably because
most studies have relied on single follow-up diagnostic exam-
inations at varying postoperative time points. Also, recurrent
stenosis is likely to remain undetected until symptoms appear.
Similarly, the risk of stroke from recurrent stenosis is unclear
for the same reasons and because of the dual nature of the
pathology. In a meta-analysis of 29 reports, the risk of recurrent
stenosis after CEA was 10% in the first year, 3% in the second,
and 2% in the third. 110 The long-term risk of recurrent stenosis
was about 1% per year. The relative risk of stroke in patients
with recurrent stenosis compared with that in patients with-
out recurrent stenosis ranged from 0.1 to 10. However, most
authors favor treatment for symptomatic patients with recur-
rent stenosis and for asymptomatic patients with high-grade
recurrent stenosis.

Because of postoperative scarring, friability of the recurrent
plaque, and the necessity for more complex surgical tech-
niques, such as interposition grafts, surgery for recurrent
carotid stenosis carries significantly greater risk of morbidity
than surgery for primary stenosis. 111,112 In one series, the rate of
cranial nerve injuries was 17%. 113 Early reports suggest that
CAS is a technically feasible and safe alternative to CEA for
patients with recurrent carotid artery stenosis. 114-116

582

chapter 50 CEA compared with CAS

Future directions

CEA is currently the gold standard for treatment of carotid
stenosis. Considerable improvements in stent design, delivery
devices, distal protection, technique, and medical manage-
ment of carotid disease have occurred since endovascular
treatment of carotid stenosis was introduced some 20 years
ago. A number of randomized clinical trials are under way to
compare the results of CAS with those of CEA. Important
issues that will be resolved by the trials include the incidence
of cerebral embolization associated with CAS and the durabil-
ity of endovascular therapy for carotid stenosis. CAS may
emerge as a valid alternative to CEA once the data from these
trials are available.

References

1. Anonymous. Beneficial effect of carotid endarterectomy in
symptomatic patients with high-grade carotid stenosis. North
American Symptomatic Carotid Endarterectomy Trial
Collaborators. N Engl J Med 1991; 325:445.

2. Anonymous. MRC European Carotid Surgery Trial: interim
results for symptomatic patients with severe (70-99%) or with
mild (0-29%) carotid stenosis. European Carotid Surgery
Trialists' Collaborative Group. Lancet 1991; 337:1235.

3. Anonymous. Endarterectomy for asymptomatic carotid artery
stenosis. Executive Committee for the Asymptomatic Carotid
Atherosclerosis Study JAMA 1995; 273:1421.

4. Barnett HJ, Taylor DW, Eliasziw M et al. Benefit of carotid en-
darterectomy in patients with symptomatic moderate or severe
stenosis. North American Symptomatic Carotid Endarterectomy
Trial Collaborators. N Engl] Med 1998; 339:1415.

5. Streifler JY, Eliasziw M, Benavente OR et al. The risk of stroke in
patients with first-ever retinal vs hemispheric transient ischemic
attacks and high-grade carotid stenosis. North American
Symptomatic Carotid Endarterectomy Trial. Arch Neurol 1995;
52:246.

6. Anonymous. Randomised trial of endarterectomy for recently
symptomatic carotid stenosis: final results of the MRC European
Carotid Surgery Trial (ECST). Lancet 1998; 351:1379.

7. Cunningham EJ, Bond R, Mehta Z, Mayberg MR, Warlow CP,
Rothwell PM. Long-term durability of carotid endarterectomy
for symptomatic stenosis and risk factors for late postoperative
stroke. European Carotid Surgery Trialists' Collaborative Group.
Stroke 2002; 33:2658.

8. Mayberg MR, Wilson SE, Yatsu F et al. Carotid endarterectomy
and prevention of cerebral ischemia in symptomatic carotid
stenosis. Veterans Affairs Cooperative Studies Program 309
Trialist Group. JAMA 1991; 266:3289.

9. Goldstein LB, Hasselblad V, Matchar DB, McCrory DC. Com-
parison and meta-analysis of randomized trials of endarterecto-
my for symptomatic carotid artery stenosis. Neurology 1995;
45:1965.

10. Moore WS, Barnett HJ, Beebe HG et al. Guidelines for carotid
endarterectomy A multidisciplinary consensus statement from

the ad hoc Committee, American Heart Association. Stroke 1995;
26:188.

11. Hobson RW 2nd, Weiss DG, Fields WS et al. Efficacy of carotid
endarterectomy for asymptomatic carotid stenosis. The Veterans
Affairs Cooperative Study Group. N Engl J Med 1993; 328:221 .

12. Tuhrim S, Bederson JB. Patient selection for carotid endarterecto-
my In: Tuhrim S, Bederson J, eds. Treatment of Carotid Disease: A
Practitioner's Manual. Park Ridge: The American Association of
Neurological Surgeons, 1998: 129.

13. Anonymous. Carotid surgery versus medical therapy in asymp-
tomatic carotid stenosis. The CASANOVA Study Group. Stroke
1991; 22:1229.

14. Anonymous. Mayo Asymptomatic Carotid Endarterectomy
Study Group: Results of a randomized controlled trial of carotid
endarterectomy for asymptomatic carotid stenosis. Mayo Clin
Proc 1992; 67:513.

15. Turnipseed WD, Kennell TW, Turski PA, Acher CW, Hoch JR.
Combined use of duplex imaging and magnetic resonance
angiography for evaluation of patients with symptomatic
ipsilateral high-grade carotid stenosis. / Vase Surg 1993; 17:
832.

16. Goldstein LB, Bonito AJ, Matchar DB et al. US national survey of
physician practices for the secondary and tertiary prevention of
ischemic stroke. Design, service availability, and common prac-
tices. Stroke 1995; 26:1607.

17. Chassin MR, Brook RH, Park RE et al. Variations in the use of
medical and surgical services by the Medicare population. N
Engl J Med 1986; 314:285.

18. Qureshi Al, SuriMF, AliZ et al. Role of conventional angiography
in evaluation of patients with carotid artery stenosis demons-
trated by Doppler ultrasound in general practice. Stroke 2001;
32:2287.

19 . Polak JF, Kalina P, Donaldson MC, O'Leary DH, Whittemore AD,
Mannick JA. Carotid endarterectomy: preoperative evaluation
of candidates with combined Doppler sonography and MR
angiography. Work in progress. Radiology 1993; 186:333.

20. Dawson DL, Zierler RE, Strandness DE Jr, Clowes AW, Kohler
TR. The role of duplex scanning and arteriography before carotid
endarterectomy: a prospective study. / Vase Surg 1993; 18:673.

21. Levy EI, Boulos AS, Fessler RD et al. Transradial cerebral angio-
graphy: an alternative route. Neurosurgery 2002; 51:335.

22. Lustgarten JH, Solomon RA, Quest DO, Khanjdi AG, Mohr JP
Carotid endarterectomy after noninvasive evaluation by duplex
ultrasonography and magnetic resonance angiography. Neuro-
surgery 1994; 34:612.

23. Back MR, Wilson JS, Rushing G et al. Magnetic resonance angio-
graphy is an accurate imaging adjunct to duplex ultrasound scan
in patient selection for carotid endarterectomy. / Vase Surg 2000;
32:429.

24. Dillon EH, van Leeuwen MS, Fernandez MA, Eikelboom BC,
Mali WP CT angiography: application to the evaluation of
carotid artery stenosis. Radiology 1993; 189:211.

25. Anonymous. Randomised trial of intravenous streptokinase,
oral aspirin, both, or neither among 17,187 cases of suspected
acute myocardial infarction: ISIS-2. ISIS-2 (Second International
Study of Infarct Survival) Collaborative Group. Lancet 1988;
2:349.

26. Anonymous. Acomparison of two doses of aspirin (30 mg vs. 283
mg a day) in patients after a transient ischemic attack or minor

583

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

ischemic stroke. The Dutch TIA Trial Study Group. N Engl }
Med 1991; 325:1261.

27. Anonymous. Swedish Aspirin Low-Dose Trial (SALT) of 75 mg
aspirin as secondary prophylaxis after cerebrovascular
ischaemic events. The SALT Collaborative Group. Lancet
1991; 338:1345.

28. Taylor DW, Barnett HJ, Haynes RB et al. Low-dose and high-
dose acetylsalicylic acid for patients undergoing carotid
endarterectomy: a randomised controlled trial. ASA and Carotid
Endarterectomy (ACE) Trial Collaborators. Lancet 1999; 353:
2179.

29. Baldassarre D, Veglia F, Gobbi C et al. Intima-media thickness
after pravastatin stabilizes also in patients with moderate to no
reduction in LDL-cholesterol levels: the carotid atherosclerosis
Italian ultrasound study. Atherosclerosis 2000; 151:575.

30. Salonen R, Nyssonen K, Porkkala-Sarataho E, Salonen JT.
The Kuopio Atherosclerosis Prevention Study (KAPS): effect
of pravastatin treatment on lipids, oxidation resistance of
lipoproteins, and atherosclerotic progression. Am J Cardiol 1995;
76:34C.

31. Crouse JR 3rd, Byington RP, Hoen HM, Furberg CD. Reductase
inhibitor monotherapy and stroke prevention. Arch Intern Med
1997; 157:1305.

32. Pasternak RC, Smith SC Jr, Bairey-Merz CN et al ACC/AHA/
NHLBI Clinical Advisory on the Use and Safety of Statins. Stroke
2002; 33:2337.

33. Ferguson GG, Eliasziw M, Barr HW et al. The North American
Symptomatic Carotid Endarterectomy Trial: surgical results in
1415 patients. Stroke 1999; 30:1751.

34. Kerber CW, Cromwell LD, Loehden OL. Catheter dilatation of
proximal carotid stenosis during distal bifurcation endarterecto-
my. Am J Neuroradiol 1980; 1:348.

35. Bockenheimer SA, Mathias K. Percutaneous transluminal
angioplasty in arteriosclerotic internal carotid artery stenosis.
Am J Neuroradiol 1983; 4:791.

36. Castaneda-Zuniga WR, Formanek A, Tadavarthy M et al. The
mechanism of balloon angioplasty. Radiology 1980; 135:565.

37. Block PC, Fallon JT, Elmer D. Experimental angioplasty: lessons
from the laboratory. Am J Roentgenol 1980; 135:907.

38. Zollikofer CL, Salomonowitz E, Sibley R et al. Transluminal
angioplasty evaluated by electron microscopy. Radiology 1984;
153:369.

39. Bergeron P, Chambran P, Hartung O, Bianca S. Cervical carotid
artery stenosis: which technique, balloon angioplasty or
surgery? / Cardiovasc Surg (Torino) 1996; 37:73.

40. Bergeron P, Chambran P, Benichou H, Alessandri C. Recurrent
carotid disease: will stents be an alternative to surgery? /
Endovasc Surg 1996; 3:76.

41. Shawl FA. Emergency percutaneous carotid stenting during
stroke. Lancet 1995; 346:1223.

42. Yadav JS, Roubin GS, Iyer S et al. Elective stenting of the
extracranial carotid arteries. Circulation 1997; 95:376.

43. Naylor AR, Bolia A, Abbott RJ et al. Randomized study of carotid
angioplasty and stenting versus carotid endarterectomy: a
stopped trial. / Vase Surg 1998; 28:326.

44. Theron J, Courtheoux P, Alachkar F, Bouvard G, Maiza D.
New triple coaxial catheter system for carotid angioplasty with
cerebral protection. Am J Neuroradiol 1990; 11:869.

45. Tong FC, Cloft HJ, Joseph GJ, Samuels OB, Dion JE. Abciximab

rescue in acute carotid stent thrombosis. Am J Neuroradiol 2000;
21:1750.

46. Yadav JS. Management practices in carotid stenting. Cerebrovasc
Dis 2001; 11 (Suppl.2):18.

47. Gray WA, White HJ Jr, Barrett DM, Chandran G, Turner R,
Reisman M. Carotid stenting and endarterectomy: a clinical and
cost comparison of revascularization strategies. Stroke 2002;
33:1063.

48. Wholey MH, Wholey M, Mathias K et al. Global experience in
cervical carotid artery stent placement. Catheter Cardiovasc
Inters 2000; 50:160.

49. Roubin GS, New G, Iyer SS et al. Immediate and late clinical out-
comes of carotid artery stenting in patients with symptomatic
and asymptomatic carotid artery stenosis: a 5-year prospective
analysis. Circulation 2001; 103:532.

50. Bettmann MA, Katzen BT, Whisnant J et al. Carotid stenting and
angioplasty: a statement for healthcare professionals from the
Councils on Cardiovascular Radiology, Stroke, Cardio-Thoracic
and Vascular Surgery, Epidemiology and Prevention, and
Clinical Cardiology, American Heart Association. Stroke 1998; 29:
336.

51. Lam JY, Chesebro JH, Steele PM, Dewanjee MK, Badimon L,
Fuster V. Deep arterial injury during experimental angioplasty:
relation to a positive indium-ill-labeled platelet scintigram,
quantitative platelet deposition and mural thrombosis. J Am Coll
Cardiol 1986; 8:1380.

52. Krupski WC, Bass A, Kelly AB, Marzec UM, Hanson SR, Harker
LA. Heparin-resistant thrombus formation by endovascular
stents in baboons. Interruption by a synthetic antithrombin.
Circulation 1990; 82:570.

53. Yusuf S, Zhao F, Mehta SR, Chrolavicius S, Tognoni G, Fox KK.
Effects of clopidogrel in addition to aspirin in patients with
acute coronary syndromes without ST-segment elevation. The
Clopidogrel in Unstable Angina to Prevent Recurrent Events
Trial Investigators. N Engl J Med 2001; 345:494.

54. Mehta SR, Yusuf S, Peters RJ et al. Effects of pretreatment with
clopidogrel and aspirin followed by long-term therapy in
patients undergoing percutaneous coronary intervention: the
PCI-CURE study. Clopidogrel in Unstable angina to prevent
Recurrent Events Trial Investigators. Lancet 2001; 358:527.

55. Gruberg L, Beyar R. Optimized combination of antiplatelet
treatment and anticoagulation for percutaneous coronary
intervention: the final word is not out yet! [letter; comment.].
J Invasive Cardiol 2002; 14:251.

56. Qureshi Al, Luft AR, Sharma M, Guterman LR, Hopkins LN.
Prevention and treatment of thromboembolic and ischemic
complications associated with endovascular procedures: Part
II — Clinical aspects and recommendations. Neurosurgery 2000;
46:1360; discussion 1375.

57. Parodi JC, Schonholz C, Ferreira LM, Mendaro E, Ohki T. "Seat
belt and air bag" technique for cerebral protection during carotid
stenting. / Endovasc Ther 2002; 9:20.

58. Ouriel K, Hertzer NR, Beven EG et al. Preprocedural risk stratifi-
cation: identifying an appropriate population for carotid stent-
ing. / Vase Surg 2001; 33:728.

59. Goldstein LB, McCrory DC, Landsman PB et al. Multicenter
review of preoperative risk factors for carotid endarterectomy
in patients with ipsilateral symptoms. Stroke 1994; 25:1116.

60. Goldstein LB, Samsa GP, Matchar DB, Oddone EZ. Multicenter

584

chapter 50 CEA compared with CAS

review of preoperative risk factors for endarterectomy for
asymptomatic carotid artery stenosis. Stroke 1998; 29:750.

61. Paciaroni M, Eliasziw M, Kappelle LJ, Finan JW, Ferguson GG,
Barnett HJ. Medical complications associated with carotid
endarterectomy. North American Symptomatic Carotid
Endarterectomy Trial (NASCET). Stroke 1999; 30:1759.

62. Babatasi G, Massetti M, Theron J, Khayat A. Asymptomatic
carotid stenosis in patients undergoing major cardiac surgery:
can percutaneous carotid angioplasty be an alternative? Eur J
Cardiothorac Surg 1997; 11:547.

63. Wennberg DE, Lucas FL, Birkmeyer JD, Bredenberg CE, Fisher
ES. Variation in carotid endarterectomy mortality in the
Medicare population: trial hospitals, volume, and patient
characteristics. JAMA 1998; 279:1278.

64. Kucey DS, Bowyer B, Iron K, Austin P, Anderson G, Tu JV. Deter-
minants of outcome after carotid endarterectomy. / Vase Surg
1998; 28:1051.

65. Hannan EL, Popp AJ, Tranmer B, Fuestel P, Waldman J, Shah D.
Relationship between provider volume and mortality for carotid
endarterectomies in New York state. Stroke 1998; 29:2292.

66. Ruby ST, Robinson D, Lynch JT, Mark H. Outcome analysis of
carotid endarterectomy in Connecticut: the impact of volume
and specialty. Ann Vase Surg 1996; 10:22.

67. Morasch MD, Parker MA, Feinglass J, Manheim LM, Pearce WH.
Carotid endarterectomy: characterization of recent increases in
procedure rates. J Vase Surg 2000; 31:901.

68. Cowan JA Jr, Dimick JB, Thompson BG, Stanley JC, Upchurch GR
Jr. Surgeon volume as an indicator of outcomes after carotid
endarterectomy: an effect independent of specialty practice
and hospital volume. J Am Coll Surg 2002; 195:814.

69. Chaturvedi S, Aggarwal R, Murugappan A. Results of carotid
endarterectomy with prospective neurologist follow-up. Neurol-
ogy 2000; 55:769.

70. Dexter DD Jr, Whisnant JP, Connolly DC, O'Fallon WM. The
association of stroke and coronary heart disease: a population
study. Mayo Clin Proc 1987; 62:1077.

71 . Meissner I, Wiebers DO, Whisnant JP, O'Fallon WM. The natural
history of asymptomatic carotid artery occlusive lesions. JAMA
1987; 258:2704.

72. Jordan WD Jr, Alcocer F, Wirthlin DJ et ah High-risk carotid
endarterectomy: challenges for carotid stent protocols. / Vase
Surg 2002; 35:16.

73. Kirshner DL, O'Brien MS, Ricotta JJ. Risk factors in a community
experience with carotid endarterectomy. / Vase Surg 1989; 10:178.

74. Mackey WC, O'Donnell TF Jr, Callow AD. Cardiac risk in pa-
tients undergoing carotid endarterectomy: impact on periopera-
tive and long-term mortality. / Vase Surg 1990; 11:226.

75. Golledge J, Cuming R, Beattie DK, Davies AH, Greenhalgh RM.
Influence of patient-related variables on the outcome of carotid
endarterectomy. / Vase Surg 1996; 24:120.

76. Riles TS, Kopelman I, Imparato AM. Myocardial infarction
following carotid endarterectomy: a review of 683 operations.
Surgery 1979; 85:249.

77. Yeager RA, Moneta GL, McConnell DB, Neuwelt EA, Taylor LM
Jr, Porter JM. Analysis of risk factors for myocardial infarction
following carotid endarterectomy. Arch Surg 1989; 124:1142.

78. Wong JH, Findlay JM, Suarez-Almazor ME. Regional perfor-
mance of carotid endarterectomy. Appropriateness, outcomes,
and risk factors for complications. Stroke 1997; 28:891.

79. McCrory DC, Goldstein LB, Samsa GP et ah Predicting
complications of carotid endarterectomy. Stroke 1993; 24:
1285.

80. Shawl FA, Efstratiou A, Hoff S, Dougherty K. Combined percuta-
neous carotid stenting and coronary angioplasty during acute
ischemic neurologic and coronary syndromes. Am J Cardiol
1996; 77:1109.

81. Faggioli GL, Curl GR, Ricotta JJ. The role of carotid screening
before coronary artery bypass. / Vase Surg 1990; 12:724.

82. Harbaugh RE, Stieg PE, Moayeri N, Hsu L. Carotid-coronary
artery bypass graft conundrum. Neurosurgery 1998; 43:926.

83. Del Sette M, Eliasziw M, Streifler JY, Hachinski VC, Fox AJ,
Barnett HJ. Internal borderzone infarction: a marker for severe
stenosis in patients with symptomatic internal carotid artery dis-
ease. For the North American Symptomatic Carotid Endarterec-
tomy (NASCET) Group. Stroke 2000; 31:631.

84. Lopes DK, Mericle RA, Lanzino G, Wakhloo AK, Guterman LR,
Hopkins LN. Stent placement for the treatment of occlusive
atherosclerotic carotid artery disease in patients with con-
comitant coronary artery disease. / Neurosurg 2002; 96:490.

85. Alamo witch S, Eliasziw M, Algra A, Meldrum H, Barnett HJ.
Risk, causes, and prevention of ischaemic stroke in elderly
patients with symptomatic internal-carotid-artery stenosis.
North American Symptomatic Carotid Endarterectomy Trial
Group. Lancet 2001; 357:1154.

86. Mattos MA, van Bemmelen PS, Hodgson KJ, Barkmeier LD,
Ramsey DE, Sumner DS. The influence of carotid siphon stenosis
on short- and long-term outcome after carotid endarterectomy. /
Vase Surg 1993; 17:902.

87. Griffiths PD, Worthy S, Gholkar A. Incidental intracranial
vascular pathology in patients investigated for carotid stenosis.
Neuroradiology 1996; 38:25.

88. Rouleau PA, Huston J 3rd, Gilbertson J, Brown RD Jr, Meyer FB,
Bower TC. Carotid artery tandem lesions: frequency of angio-
graphic detection and consequences for endarterectomy. Am J
Neuroradiol 1999; 20:621.

89. Kappelle LJ, Eliasziw M, Fox AJ, Sharpe BL, Barnett HJ. Impor-
tance of intracranial atherosclerotic disease in patients with
symptomatic stenosis of the internal carotid artery. The North
American Symptomatic Carotid Endarterectomy Trial. Stroke
1999; 30:282.

90. Kim SH, Mericle RA, Lanzino G, Qureshi AI, Guterman LR,
Hopkins LN. Carotid angioplasty and stent placement in
patients with tandem stenosis (Abstract). Neurosurgery 1998;
43:708A.

91 . Rockman CB, Su W, Lamparello PJ et ah A reassessment of carotid
endarterectomy in the face of contralateral carotid occlusion:
surgical results in symptomatic and asymptomatic patients.
J Vase Surg 2002; 36:668.

92. Gasecki AP, Eliasziw M, Ferguson GG, Hachinski V, Barnett HJ.
Long-term prognosis and effect of endarterectomy in patients
with symptomatic severe carotid stenosis and contralateral
carotid stenosis or occlusion: results from NASCET. North
American Symptomatic Carotid Endarterectomy Trial
(NASCET) Group. J Neurosurg 1995; 83:778.

93. Halsey JH Jr. Risks and benefits of shunting in carotid
endarterectomy. The International Transcranial Doppler
Collaborators. Stroke 1992; 23:1583.

94. Mathur A, Roubin GS, Gomez CR et ah Elective carotid artery

585

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

stenting in the presence of contralateral occlusion. Am J Cardiol
1998; 81:1315.

95. Mericle RA, Kim SH, Lanzino G et al. Carotid artery angioplasty
and use of stents in high-risk patients with contralateral occlu-
sions. JNeurosurg 1999; 90:1031.

96. Barnett HJ, Meldrum HE, Eliasziw M. The appropriate use
of carotid endarterectomy North American Symptomatic
Carotid Endarterectomy Trial collaborators. CMAJ 2002; 166:
1169.

97. Villarreal J, Silva J, Eliasziw M et al. Prognosis of patients with
intraluminal thrombus in the internal carotid artery (for the
North American Symptomatic Carotid Endarterectomy Trial)
(Abstract 18). Stroke 1998; 29:276.

98. Guterman LR, Budny JL, Gibbons KJ, Hopkins LN. Thromboly-
sis of the cervical internal carotid artery before balloon angio-
plasty and stent placement: report of two cases. Neurosurgery
1996; 38:620.

99. Lam WW, Leung SF, So NM et al. Incidence of carotid stenosis in
nasopharyngeal carcinoma patients after radiotherapy. Cancer
2001; 92:2357.

100. Lam WW, Liu KH, Leung SF et al. Sonographic characterisation
of radiation-induced carotid artery stenosis. Cerebrovasc Dis
2002; 13:168.

101. Loftus CM, Biller J, Hart MN, Cornell SH, Hiratzka LF. Manage-
ment of radiation-induced accelerated carotid atherosclerosis.
ArchNeurol 1987; 44:711.

102. Melliere D, Becquemin JP, Berrahal D, Desgranges P, Cavillon A.
Management of radiation-induced occlusive arterial disease: a
reassessment. / Cardiovasc Surg 1997; 38:261.

103. Kashyap VS, Moore WS, Quinones-Baldrich WJ. Carotid artery
repair for radiation-associated atherosclerosis is a safe and
durable procedure. / Vase Surg 1999; 29:90.

104. Al-Mubarak N, Roubin GS, Iyer SS, Gomez CR, Liu MW, Vitek JJ.
Carotid stenting for severe radiation-induced extracranial
carotid artery occlusive disease. / Endovasc Ther 2000; 7:36.

105. Kitamura J, Kuroda S, Ushikoshi S et al. Three cases of successful

stenting for radiation-induced carotid arterial stenosis. No
Shinkei Geka - Neurol Surg 2002; 30:1097.

106. Alric P, Branchereau P, Berthet JP, Mary H, Marty- Ane C. Carotid
artery stenting for stenosis following revascularization or
cervical irradiation. / Endovasc Ther 2002; 9:14.

107. Henry M, Henry I, Klonaris C et al. Benefits of cerebral protection
during carotid stenting with the PercuSurge Guard Wire system:
midterm results. J Endovasc Ther 2002; 9:1.

108. Bonaldi G. Angioplasty and stenting of the cervical carotid
bifurcation: report of a 4-year series. Neuroradiology 2002; 44:164.

109. Houdart E, Mounayer C, Chapot R, Saint-Maurice JP, Merland JJ.
Carotid stenting for radiation-induced stenoses: a report of 7
cases. Stroke 2001; 32:118.

110. Frericks H, Kievit J, van Baalen JM, van Bockel JH. Carotid
recurrent stenosis and risk of ipsilateral stroke: a systematic
review of the literature. Stroke 1998; 29:244.

111. Meyer FB, Piepgras DG, Fode NC. Surgical treatment of recur-
rent carotid artery stenosis. JNeurosurg 1994; 80:781.

112. Hertzer NR, O'Hara PJ, Mascha EJ, Krajewski LP, Sullivan TM,
Beven EG. Early outcome assessment for 2228 consecutive
carotid endarterectomy procedures: the Cleveland Clinic
experience from 1989 to 1995. / Vase Surg 1997; 26:1.

113. AbuRahma AF, Jennings TG, Wulu JT, Tarakji L, Robinson PA.
Redo carotid endarterectomy versus primary carotid endarterec-
tomy. Stroke 2001; 32:2787.

114. Lanzino G, Mericle RA, Lopes DK, Wakhloo AK, Guterman LR,
Hopkins LN. Percutaneous transluminal angioplasty and stent
placement for recurrent carotid artery stenosis. JNeurosurg 1999;
90:688.

115. Diethrich EB, Gordon MH, Lopez-Galarza LA, Rodriguez-Lopez
JA, Casses F. Intraluminal Palmaz stent implantation for treat-
ment of recurrent carotid artery occlusive disease: a plan for the
future. JInterv Cardiol 1995; 8:213.

116. Yadav JS, Roubin GS, King P, Iyer S, Vitek J. Angioplasty and
stenting for restenosis after carotid endarterectomy. Initial
experience. Stroke 1996; 27:2075.

586

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Endovascular intervention for venous
occlusion compared with surgical
reconstruction

Patricia E. Thorpe
Francisco J. Osse

The techniques and tools of minimally invasive therapy have
been successfully adapted to the treatment of occlusive ve-
nous disorders. 1 ' 2 This ranges from catheter-directed throm-
bolysis or mechanical thrombectomy of acute thrombosis to
endovenous reconstruction of chronically occluded axial
veins with metallic stents. Surgical options still include open
thrombectomy, with or without creation of an arteriovenous
fistula, but the evolution of percutaneous procedures has pro-
duced a less invasive option of thrombolysis and stenting. The
endovascular approach may yield comparable or better long-
term results, since the underlying stenosis can be imaged and,
in addition, treated to thrombus removal. Vascular surgeons
readily acknowledge that endovascular techniques play an
important role in new therapy for problems such as chronic
inferior vena cava (IVC) or superior vena cava (SVC)
obstruction, conditions that have never had widely practiced
surgical solutions. 3

When considering application of the endovascular ap-
proach, venous obstruction can be divided into acute and
chronic, depending on the duration of the symptoms. The
clinical presentation of deep vein thrombosis (DVT), how-
ever, may not correlate with the age of all the thrombus in the
affected extremity. Patient selection for the procedures is key
to their success, especially in patients who will require lifelong
monitoring of stents and /or anticoagulation. Patients with
DVT present the most common clinical dilemma. Should
thrombolysis be offered, or should the patient receive low-
molecular-weight heparin and begin a 3- to 6-month course
of warfarin? Furthermore, how should one assess the risk of
recurrent DVT after discontinuation of warfarin at the end
of that period? Studies have shown that catheter-directed
thrombolytic therapy is safe and effective. 4 ' 5 Early removal of
thrombus is associated with fewer post-thrombotic symptoms
and improved quality of life. 6 Early removal of thrombus pre-
serves valves in the deep veins, thereby preventing valve
damage that causes reflux.

This can be accomplished with thrombectomy and throm-
bolysis, or a combination of both techniques. 7 Early thrombus
removal also can preclude pulmonary embolus (PE) and post-

thrombotic syndrome (PTS). The addition of temporary IVC
filters may change practice patterns. Despite these benefits,
experience in the medical community with thrombolysis re-
flects the legacy of bleeding complications associated with
streptokinase use in the 1970s and 1980s. Then, a new surge of
bleeding complications occurred with the use of alteplase in
peripheral obstructions in 2000, when urokinase was not
available. 8 The inability to predict complete lysis and the risk
of bleeding from alteplase led to a cessation of use by many. 9
On the other hand, catheter-directed thrombolytic therapy is
now considered routine, with a continually broadening spec-
trum of indications. 10 The main clinical challenge is not so
much the question of whether thrombolysis will be complete,
but rather, the issue of appropriate patient selection.

It has been shown that acute thrombus less than 10-14 days
old can be totally removed with thrombolytic therapy. 11 In fact,
the endogenous lytic system promotes early regression of
acute thrombosis and this tissue plasminogen activator (t-PA)-
mediated activity continues up to 9 months. However, en-
dogenous fibrinolysis is not equally effective in all patients.
There appear to be individual variations, as well as a differ-
ence between lower extremity venous segments with respect
to the tendency to partially or totally recanalize or remain oc-
cluded. 12 After 63 above-the-knee DVT patients were followed
by sequential duplex for up to 1 year, the study showed that
partial resolution was associated with more reflux. Among 171
sites studied, 71% were initially occluded, and 29% partially
thrombosed. After 1 year, 60% of segments were totally patent,
27% were partially recanalized, and 12% remained occluded.
The femoral vein had the highest incidence of occlusion at
1 year.

What allows us to predict who will effectively resolve
thrombus and who will not? We do not know. Thrombus load
may be a factor, but an individual's genetically determined
ability to autolyse thrombus is not yet well understood. Cer-
tain patients might be ideal candidates for adjunctive throm-
bolytic therapy to compensate for ineffective endogenous
fibrinolysis. Others may recanalize thrombus very effectively.
We cannot predict who will fail heparin and warfarin therapy.

587

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

We are left with a patient selection process that is inconsistent
and one which favors treatment of thrombus less than 2 weeks
old. Phlegmasia and impeding venous gangrene are condi-
tions about which there is consensus regarding immediate
removal of thrombus. 13 Patients with multisegmental acute
thrombosis are generally preferred candidates, if they have no
contraindications to thrombolysis; however, such patients are
often referred only days after failing to improve with heparin
and bed rest. When they are evaluated for thrombolysis, there-
fore, the age of the thrombus may well exceed 10-14 days. Dis-
ability from post-thrombotic syndrome, limb deterioration,
and quality of life becomes a major issue in choosing to inter-
vene for chronic obstruction. Since the incidence of post-
thrombotic syndrome cited in the literature ranges from 27%
to 88%, with 5-34% severely affected and 1-11% developing
ulcers, a large number of patients have this condition. 14

The chapter discussion can be divided between interven-
tions aimed at removing acute thrombus and those designed
to recanalize or bypass chronic thrombus. The techniques vary
little with thrombus location. Upper and lower extremities are
treated with similar tools and techniques that include anti-
coagulation, thrombolysis, thrombectomy, and metallic stents
and temporary filters. Acute thrombus responds quickly to
thrombolysis, but catheter-directed thrombolysis is also a
useful technique in treating chronic obstruction. Although
complete removal of thrombus is unlikely, the thrombolytic
infusion appears to reduce resistance to the guidewire and
catheter manipulations, which are necessary for stent place-
ment. This chapter includes discussion of available interven-
tions and illustrations of clinical applications in treating acute
and chronic occlusions of the upper and lower limbs, includ-
ing the superior and inferior segments of the venae cavae.

Selecting patients for endovenous therapy

There is emerging recognition of a symptomatic condition that
has previously been underdiagnosed or unrecognized and,
in many instances, has led to left leg thrombosis. The
May-Thurner, or iliac-compression, syndrome is caused by
compression of the left common iliac vein by the right iliac
artery 15-21 It is often clinically occult and can cause unilateral
leg or ankle edema and discomfort. The diagnosis is fre-
quently made in association with acute iliofemoral thrombo-
sis; however, the success of endovascular stents has led to
greater awareness of this condition in young women, especial-
ly if they present with unilateral left limb edema without
thrombosis. 22 The diagnosis can be suspected with a history of
notable left leg or ankle symmetry. Often, the duplex examina-
tion, done to rule out DVT, is negative. The subtle findings,
such as flattening of the common femoral wave form, iliac
velocity changes, and widening of the common iliac vein, are
often missed, because the pelvis is not included in the
examination or is difficult to image, due to obesity or bowel

L.O; I c ffl

Figure 51.1 Pelvic venogram of 42-year-old woman with left lower
extremity post-thrombotic syndrome for 2 years. The study performed with
patient prone shows large transpubic collaterals shunting flow left-to-right.
Retroperitoneal collaterals are also present on the left.

gas. Phlebography with contrast injection from the popliteal
and femoral level and intravascular ultrasound (IVUS) are the
best ways to image the abnormality. Transpelvic or trans-
sacral collaterals imply iliac occlusion. This indicates a high
resistance flow pattern due to luminal thrombus or external
compression or wall thickening. The phlebographic appear-
ance may suggest occult thrombosis (Figs. 51.1 and 51.2). Col-
laterals are not always evident, particularly if multiple small
channels persist and the patient is examined in a resting posi-
tion, without valsalva (Fig. 51.3). One should exclude other
causes of iliac thrombosis, such as malignancy or trauma.
Cross-sectional imaging can suggest the condition, but does
not provide information about the hemodynamic significance
of the compression. Everyone has this anatomical relationship
of the vein and artery, but only 20-25% of patients were found
to have focal vessel abnormalities in postmortem studies. 18

Regardless of the age of the thrombus or the number of times
a patient has been hospitalized with DVT, in our experience
symptomatic and hemodynamic improvement can be
achieved and sustained with endovascular therapy. An
important criterion for patient selection is their ability and
willingness to take long-term warfarin. In the future, oral
thrombus inhibitors will be available to facilitate lifelong
anticoagulation. As we gain more long-term data on patients
with iliac stents, we may be able to identify a profile for
patients who do not require lifetime anticoagulation to
maintain patency. Our experience indicates this group in-

588

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

Figure 51 .2 Following 24 h of catheter-directed thrombolysis into the left
common femoral segment, from a popliteal approach, a trace of the residual
iliac lumen is seen along with trans-sacral collaterals. The multiside-hole
catheter was advanced to the common iliac level for additional overnight
thrombolysis before balloon dilation and stent placement (prone position).

eludes patients without malignancy and with one to three
stents in the suprainguinal location, with excellent inflow
from relatively normal distal veins.

The standard contraindications to lytic therapy apply with
chronic DVT patients, especially because the infusions are
longer than with acute DVT. The primary exclusion is hyper-
tension, seizure disorder, and recent trauma. We have
performed uncomplicated, prolonged infusions in patients
shortly after major surgery without bleeding complications.
These included gastric bypass (14 days), abdominal hysterec-
tomy (10 days), and total-knee arthroplasty (5 days). One
patient was treated 14 days after normal vaginal delivery,
and several women have been treated during menses,
without complication.

Renal failure is a relative contraindication to endovenous
reconstruction of chronic thrombosis, due to the need for
repeated use of iodinated contrast. Whereas acute DVT can
be treated with a combination of ultrasound guidance and
monitoring, manipulation of catheters and wires in chronic
occlusions requires fluoroscopic visualization with iodinated
contrast. An alternative contrast agent, carbon dioxide, can be
used to image venous anatomy. If necessary, flow-directed
infusion can be performed with ultrasound assistance for
identifying the location of the saphenous pressure sites, for
disk placement, and for follow-up monitoring.

Figure 51.3 Classic appearance of the "boulevard look" which develops
with May-Thurner or left common iliac vein compression. This is due to a
thickened, abnormal vein wall. The vein is subjected to external pressure in
the anterior-posterior (A-P) dimension. This widens the vein on venography.
Intravascular ultrasound can reveal the extreme A-P narrowing of the
channels. Partially obstructing thrombus is seen in the external iliac segment
(arrow) (patient supine).

Endovenous techniques

When treating an iliac or caval obstruction, the popliteal ap-
proach is generally selected. If necessary, a bilateral approach
is used simultaneously. Occasionally, resistance from old
thrombus will require use of the "pull-through" technique, to
provide sufficient force to advance a catheter through the tight
residual lumen (Fig. 51.4). This requires access from the con-
tralateral or jugular approach. This technique allows one to
keep the wire from buckling or pushing back as the catheter is
advanced. Although the baseline phlebogram is performed
with the patient supine, the endovascular procedure requires
use of a Foley catheter and placement of the patient in a prone
position. Initial deep venous access is acquired, using a combi-
nation of contrast infusion from a pedal vein and ultrasound
guidance of a 21-G needle from a Micropuncture Set (Cook,
Inc., Bloomington, IN, USA) into the back wall of the popliteal
vein or a superficial tributary. Successful entry into a patent

589

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 51 .4 A 1 0-20-mm nitinol snare can be used to secure a 260-cm
exchange length wire in the lower or upper body veins. In this case, the wire
was snared and retracted through the jugular sheath. This provides a
through-and-through wire for enough tension to support pushing a balloon
catheter through a very tight stenosis. Without being able to hold the wire
from both directions, attempts to cross the lesion are frustrated with
repeated buckling of the catheter, even if the wire has traversed the stenosis.

distal vein in continuity with occluded proximal deep venous
segment(s) permits placement of a working sheath for catheter
exchange and contrast injection. Spasms and missed passes
are significantly less with the combination 3-5.5-Fr coaxial
Micropuncture dilators and 0.018-in nitinol wire than with
standard intravenous catheters. An angled 4- to 5-Fr hy-
drophilic catheter, in combination with a 0.035-in Glidewire™
(Boston Scientific, Natick, MA, USA) or 0.035-in Roadrunner™
(Cook, Inc.), is carefully advanced from the popliteal level
through the superficial femoral vein to the common femoral
level.

During the initial 24-48 h, the goal is to place a multiside-
hole catheter along the length of the abnormal segments tra-
versed by the wire. An overnight infusion of a thrombolytic
agent will decrease the intraluminal resistance and set the
stage for subsequent catheter advancement, balloon dilation,
and stent placement. Following passage of a wire into the nor-
mal cava, an exchange is made to position a stiffer exchange-
length wire that is maintained as the working wire during
stent deployment. Sequential predilation in the cava and iliac
vein is conducted from proximal to distal, using 8- to 14-mm
balloons (4-cm length). Balloon selection depends on the esti-
mated size of the native vessel and the observed resistance to
balloon expansion. Overdilation is avoided. Some vessels are
very tenacious, while others dilate with relative ease. Great

care is taken to "feel" the balloon, as one slowly proceeds to
expand the residual lumen.

Techniques of catheter thrombolytic
infusion(s)

The following philosophy governs endovascular therapy for
thrombotic occlusion and especially the chronically occluded
lower extremity. The leg is considered an organ system, in the
sense that certain sections cannot be treated in isolation from
the rest of the system. The venous flow in the femoral segment,
for example, depends on flow from the popliteal and in-
frapopliteal segments. If there is subacute and chronic DVT in
the calf and thigh, treating only the superficial femoral vein
(SFV), without addressing the occlusive thrombus in seg-
ments below, will not produce an effective restoration of deep
venous flow. If flow in the calf preferentially goes to the super-
ficial veins, the pattern will persist, even when the SFV is re-
opened; but if there is too little deep venous flow, the reopened
SFV will not be the path of least resistance, due to persistent
occlusion more distally. This will limit the clinical improve-
ment in chronic venous insufficiency; therefore, the interac-
tion of all segments must be taken into account to achieve
optimal results with lytic therapy. A combination of catheter
techniques and flow-directed therapy is most useful for treat-
ing multisegmental thrombosis. These, in combination with
balloon dilation and stent placement, comprise the main
thrust of endovenous therapy.

Catheter-directed therapy

Venous access for catheter placement to treat iliofemoral and
proximal superficial femoral thrombosis is achieved through
a contralateral femoral vein, or the ipsilateral femoral or
popliteal veins. The right internal jugular approach may be
used, but it can be difficult to manipulate catheter tips from
such a distance. This approach works better for acute throm-
bus. Following placement of a vascular sheath, a multiside-
hole catheter is directed over a guidewire and positioned in the
thrombosed vein. If initial guidewire traversal of the throm-
bosed segment is difficult, a 5-Fr catheter is positioned into the
thrombus as far as possible for initial infusion of lytic therapy.
When a subsequent attempt to pass the guidewire is success-
ful, the catheter can be advanced and strategically positioned
throughout the thrombosed segment. A variety of multiside-
hole 5-Fr catheters can be used, including a predetermined
length (i.e. 10, 20, 30, or 50 cm), or adjustable-length or coaxial
systems are available. The length of the infusion catheter is
selected according to the ability to position the catheter across
the thrombosed area. A long continuous segment of thrombus
is treated with a single catheter, when possible. Extensive
thrombus often requires repositioning of the catheter at the
time of interval follow-up, so an adjustable-length catheter

590

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

will preclude multiple catheter exchanges. The catheter may
be safely advanced over wires in both retrograde and ante-
grade directions. Although it is easier to advance through ve-
nous valves from an antegrade approach, the popliteal vein is
not always the best approach, as it, too, may be obliterated or
occluded. A Roadrunner wire (Cook, Inc.) and an angled
Glidewire (Medi-Tech/ Boston Scientific, Waterstown, MA,
USA) are the instruments of choice for traversing venous
valves, regardless of direction. A 4-Fr or 5-Fr straight or 45°
angle catheter can then be advanced over a wire. An alterna-
tive coaxial system, consisting of a 5-Fr Mewissen multiside-
hole catheter (Cook, Inc.) in combination with a 0.035-in
Katzen (Cook, Inc.) wire, can be useful. With this system, one
can separate the infusions and position the catheter /wire
system in different locations for most effective intrathrombus
delivery.

The AngioDynamics catheter has multiple infusion lengths
and works with an end-occluding wire (AngioDynamics, Inc.,
Queensbury, NY, USA). It works well with the automatic Pulse
Spray™ pump, which we have found very effective in both
acute and chronic occlusions. The coaxial Mewissen-Katzen
system requires two infusion pumps, with a divided dose of
urokinase. The heparin infusion is piggybacked into the
catheter. Urokinase (Abbokinase; Abbott Laboratories, Abbott
Park, IL, USA) is reconstituted by using 1 000 000 IU in 1000 ml
of 0.9% NaCl and delivered at a rate of 1000 IU/cm 3 . The com-
mon rate is 50-100 OOOIU/h. If t-PA is used, the comparable
dose is as follows: 10 mg of Alteplase (Genentec, Inc.) mixed in
1000 cm 3 of normal saline, delivered at a rate of 50-100 cm 3 /h,
for an infusion dose of 0.5-1.0 mg/h. When combined with a
flow-directed infusion from a pedal access site, the total dose
can be divided between the two sites. The amount per infusion
is determined by the initial venous flow rate (seen fluoroscopi-
cally) and the amount and distribution of thrombus. Laborato-
ry monitoring, every 4-6 h (using minisamples of less than
2.5 cm 3 ), is done for evaluation of fibrinogen, prothrombin
time (PT), and partial thromboplastin time (PTT), hemoglobin
and platelet count. The fibrinogen level is maintained at
greater than 25% of baseline, which is usually over 100 mg/dl.
The PTT range is maintained between 50 and 80 s.

Flow-directed infusion

A 22-G intercath placed in a dorsal pedal vein, for the purpose
of performing the baseline venogram, is used for the flow-
directed infusion. A single puncture in a pedal vein below the
ankle is desirable. Although multiple punctures cannot be
avoided in certain patients, we attempt to work from a distal-
to-proximal direction to avoid infusing urokinase below mult-
iple punctures. It is surprising that very little ecchymosis
develops from missed i.v. attempts. A small, clear plastic
dressing is placed over the 22-G catheter to maintain visualiza-
tion of the site throughout the procedure. It is important to
loop the i.v. tubing to prevent inadvertent loss of the site. Infec-

tion and /or bleeding are not problems; but, as always, care is
taken to observe for any extravasation of contrast during each
injection. A short, clear plastic connecting tube is used with a
three-way stopcock to facilitate interval injection of contrast
for follow-up evaluation of the progress of lytic therapy. A
saline infusion is maintained through the pedal site, if uroki-
nase is not being infused. A Velcro-type tourniquet (Tiger
Surgical, Inc., Portland, OR, USA) is placed at the malleolar
level, in combination with a small disk positioned to provide
focal compression of the saphenous vein against the medial
malleolus (Fig. 51.5). Under fluoroscopic visualization, a
small amount of contrast is injected to ascertain focal compres-
sion of the saphenous vein and redirection of flow through a
communicating vein into the deep system. The disk position,
as well as the upper and lower margins of the tourniquet, is
marked on the skin. This allows a nurse to release the tourni-
quet once every hour and replace it in the correct position.
Folded 4x4 gauze is placed under the disk to protect the skin
from pressure. The pedal pulse is marked and monitored with
blood pressure and pulse; the tourniquet provides adequate
redirection of venous flow without any compromise of arterial
flow. It is released lOmin every hour and reapplied at a
specific marked level, which assures proper disk position and
tightness.

Normally, blood flows preferentially from the superficial to
the deep system; however, in the presence of DVT, one may see
contrast reflux through the perforating veins into the superfi-
cial system in the midcalf or above. When this is fluoroscopi-
cally recognized, a second tourniquet is placed at the knee to
compress the greater saphenous vein against the femoral
condyle, promoting redirection of flow into the tibiopopliteal
veins. The thrombolytic infusion includes urokinase,
50-100 000 IU/h, or t-PA at 0.5-1 .0 mg/h, via the pedal IV.

After the course of thrombolytic therapy, additional inter-
vention is performed to treat underlying venous stenosis. Per-
sistently narrowed venous channels can be dilated, but rarely
is the result hemodynamically satisfactory. Improvement does
occur in the lumen, and we have noted that greater widening
of the chronically occluded lumen occurs after overnight use
of the Pulse Spray™ technique vs. drip infusion. In the event of
significant lumen irregularity and residual narrowing of the
iliofemoral segments, one or more self-expanding metallic
stents can be placed to augment outflow. Pullback pressures
are obtained before and after stent placement. We consider a 2-
to 4-mmHg pressure gradient between the IVC and iliac veins
hemodynamically significant. Self-expanding metallic stents
are placed from the ipsilateral or contralateral approach after
maximum thrombolytic therapy. Our rationale for thromboly-
sis is the observation that stents open wider after removal of
any acute or subacute thrombus. Lytic therapy also softens
thrombus and permits catheter traversal, which facilitates
endovascular intervention. Duplex imaging is used to confirm
stent patency and assess flow velocities 1 day postoperatively,
at 1, 3, 6, and 12 months, and yearly thereafter. Flow velocities

591

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 51.5 (A) Contrast venogram of the right foot and ankle showing
how focal compression of the greater saphenous vein results in redirection of
contrast into the perforators and deep venous system. (B) Diagram shows

placement of the tourniquet. The disk is postioned under fluoroscopy to
prevent flow into the superficial system. An additional tourniquet can be
placed at knee level.

in the normal limb can serve as a control for the affected limb
that is undergoing treatment.

Patients are systemically heparinized (PTT greater than 50 s
and greater than 80 s) throughout the period of thrombolysis
and following angioplasty or stent procedures. Oral anticoag-
ulation is tapered before therapy (2 days off Coumadin before
admission) and restarted 1 day prior to stent placement. This
allows removal of the sheath before the International Normal-
ized Ratio (INR) is therapeutic. Upon completion of thrombo-
lysis, heparinization is continued until oral anticoagulation is
consistent with PT greater than 20 s and INR between 2.5 and
3.0. Oral anticoagulation (OAC) is monitored cooperatively
with the referring physician for several months, as inadequate
anticoagulation is the most frequent cause of early failure.
Younger patients with chronic conditions may have a hyper-
coagulable state, making warfarin titration challenging. We
can only emphasize the great importance of diligent monitor-
ing of the PT/INR. Whereas a minimum of 6-12 months of
warfarin is standard in acute DVT, patients with chronic dis-

ease and /or multiple stents are placed on indefinite warfarin.
Compression stockings or leggings are prescribed and fitted
for all patients before discharge. Patients are followed with
clinic visits, duplex examinations, and photoplethysmogra-
phy (PPG) and air plethysmography (APG) upon completion
and at 3-month intervals for 1 year, and yearly thereafter.
When patients become symptomatic with edema or pain,
restenosis can be suspected if there is no evidence of recurrent
thrombosis. Restenosis within pelvic stents may be subtle on
duplex. The stents appear patent. We have found that balloon
dilation of the stents via the right internal jugular approach
(avoiding any insult to the deep venous system of the leg) is a
simple and effective method for treating in-stent hyperplasia.

The role of metallic stents

Percutaneous transluminal angioplasty has been shown to
limit effectiveness in the treatment of venous stenoses. We feel

592

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

local thrombolytic therapy is an important adjunct, prior to
stent placement in patients with chronic venous obstructions
complicated by superimposed acute thrombosis. After acute
thrombus is removed, the chronic lesion should be treated to
expand the lumen. Residual obstruction in the iliac veins can
cause rethrombosis. Stent placement is a simple, percuta-
neous, low-risk procedure and, in any patient, can be a suitable
alternative to venous surgery. Because of its simplicity, it can
also be used in patients who are not suitable candidates for
surgery. In patients with a malignancy, stent placement can
provide desirable palliative treatment. In those with benign
lesions, stent placement promises to be the definitive treat-
ment of large-vein obstruction. 22 ' 23

The endovascular specialist must have a working knowl-
edge of stent properties and be familiar with a wide variety of
commercially available products. The number of marketed
stents has grown significantly over the past decade; more
than 40 different stents are available for coronary use and more
than 20 for use in the periphery. Significant improvement in
patency rates has been observed in iliac lesions treated with
stents and angioplasty. Data from the US National Venous
Registry (NVR) conducted between 1994 and 1997 indicated
that, in 1 year, 74% of limbs treated with stents remained
patent, compared with 53% of limbs not receiving stents
(P<0.001). 24

No stent possesses all the qualities of an ideal stent (e.g.
good radio-opacity, ease of positioning, flexibility, tractability,
fidelity of shape, and low restenosis rate). The issue of resteno-
sis may be eventually decreased with larger, drug-eluting
stents, as in the coronary vessels. Evaluation of the characteris-
tics of the properties of metallic stents is gained with experi-
ence as new stents are introduced and compared with those in
use. Clinical use of stents in the iliac vein was first reported by
Zollikofer et al. in 1988. 25 Prior to that, others had reported use
of stents in the inferior and superior vena cava. 26 Migration
and intimal hyperplasia were documented with early use of
the Gianturrco-Z stent. 27 Modifications were made to prevent
slippage; modular stent units were connected with nylon
suture and small hooks were added. Migration has been a
rare event with all stent designs. 28 When this does occur,
misplaced or migrated Palmaz™ and Wallstents™ can be
effectively retrieved, using endovascular techniques. 29

All stents have been shown to become endothelialized
and/or covered with neointima. Sawada et al. reported this in
all Z-stents observed at autopsy. 27 The neointima appears
rapidly and the endothelialization process occurs within
2-6 weeks. The amount of intimal thickening has been corre-
lated to design. More rigid stents, exerting greater axial force
on the vessel wall, seem to accelerate intimal hyperplasia. 25 In
these animal studies and our experience, the Wallstent™ is
less likely to induce compromising intimal hyperplasia, due to
the small wire size, pliability, and longitudinal flexibility.
Zollikofer and colleagues made two important observations
in animal studies: (i) regression of intimal hyperplasia oc-

curred with time, and (ii) restenosis is more common at a site of
high pressure flow (e.g. arteriovenous fistula). 27 ' 30 These find-
ings appear true for stented and nonstented vein segments.
IVUS has allowed us to examine symptomatic patients and
document restenosis caused by intimal hyperplasia. 22

The majority of interventionalists prefer self-expanding
stents for venous stenting. Although nitinol is a flexible, ac-
ceptable alternative, the construction of the stents permits a
better angiographic appearance with Wallstents™, since the
interstices are smaller. In the common iliac location, with
extrinsic pressure a factor in causing compression, flexibility
and self-expansion may give the Wallstent™ an advantage. A
study comparing the long-term patency of self-expanding vs.
balloon-expandable stents placed in the common iliac vein has
not been done. The choice of lengths and diameters make this
design suitable for 10- to 20-cm vein diameters. The newer
nitinol stents have less foreshortening than the Wallstent™,
but they are somewhat less visible under fluoroscopy, espe-
cially in large patients. Currently available stents are used "off
label/' as the 8- to 16-mm diameter stents are Food and Drug
Administration-approved only for biliary and iliac artery use.
Covered stents, suitable for repair of venous tears (Wallgraft;
Boston Scientific), became available in the United States in
2000; and additional designs are used in Europe, Japan, and
South America. Another stent limitation for venous use is
length. Many venous occlusions are long and reconstruction
requires tandem stent deployment. The fully expanded stent
length ranges from 4.0 to 10.0 cm. Stents remain correspond-
ingly longer if full expansion is not achieved. As in the arterial
system, crossing the origin of branch vessels does not appear
to be associated with problems. We have closely followed iliac
vein Wallstents™ in situ for over 7 years with duplex imaging.
There is no indication of strut fracture or failure. Stents with
larger interstices have a tendency to look more deformed
when deployed in veins damaged with chronic mural throm-
bus. These vessels are not smooth, tubular structures like
normal veins or arteries. The Wallstent™ most consistently
excludes the irregular endothelial surface and creates a
smooth, sufficiently widened diameter. Nitinol stents are also
flexible and suitable for venous application; however, it
is more difficult to see nitinol under fluoroscopy in the
abdominal area.

Thrombectomy

Thrombectomy refers to physical removal of a soft and
fresh thrombus formed inside any vessel. Venous surgical
thrombectomy was first performed by Lawen in Germany in
1937, but introduced in the United States by Homans in 1940.
The surgical approach was the answer, at that time, to all pre-
viously reported poor results and severe chronic venous insuf-
ficiency signs and symptoms observed with conservative
treatment. A comparison of treatment option outcomes shows

593

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

a trend toward a lower incidence of PE and reflux but similar
patency rates at 2 years (Table 51.1).

In the 1980s, endovascular alternatives were introduced.
Concepts and techniques to remove thrombus from a vein
could be accomplished with the advantages of a minimally in-
vasive therapy; however, the risk of thrombolysis discouraged
many. Flow-directed and catheter-directed thrombolysis suf-
fered from the recall of urokinase in late 1999, as well. Al-
though they have become accepted forms of therapy for acute
DVT, the evolution of mechanical thrombectomy devices has
offered an alternative or adjunctive method for thrombus re-
moval. Catheter-directed thrombolysis and stent deployment
are now frequently combined with mechanical thrombecto-
my, in order to shorten the intervention and decrease the risk
of bleeding complications.

For those patients with acute DVT and contraindications for
lysis and surgery, a minimally invasive mechanical device is
the "vacuum" in the medical arsenal that was first filled in by
the Fogarty balloon in the late 1980s. The balloon is a tool for
rapid debulking of the vein, but requires a surgical incision.
Although, over-the-wire and wireless versions are available,
residual mural thrombus persists in most veins and the en-
dothelium can be injured, resulting in permanent valve dam-

Table 51 .1 Outcomes guidelines for iliofemoral thrombectomy

Surgery

Conservative

(with AVF)

Endovascular

Rethrombosis

12%

<10%

>10%

<5%

<1%

Vein patency (2 years)

<35%

>75%

>80%

Valvular competence

<30%

>50%

>65%

(2 years)

Reflux (2 years)

>65%

<50%

<30%

Asymptomatic (2 years)

<15%

>50%

>65%

age or late intimal hyperplasia. Thus, many devices have
emerged to replace the Fogarty catheter, using different re-
sources to remove thrombus with the hope of preserving vein
wall integrity. Unfortunately, none has yet proven to conquer
all the challenges of doing the job without some degree of
vessel injury.

A mechanical thrombectomy device can be classified ac-
cording to how it engages and cleaves the thrombus and how
the thrombus is removed from the vessel. Direct-contact and
negative-pressure-gradient devices are the most accepted sys-
tems in use. The first includes compliant balloons and wire
baskets, and the other category is represented by hydrody-
namic and flow-based devices. Tables 51.1 and 51.2 summa-
rize data gathered from Stainken 31 and a literature review by
the authors.

• Balloons, available from 5 to 14 Fr sizes, can be used with
or without a guidewire. Positive features include speed
and cost-effectiveness, and negative features include a
higher incidence of incomplete thrombectomy and distal
embolization.

• Wire baskets, mostly represented by the Arrow PTD
(percutaneous thrombectomy device), the Bacchus Solera,
Rex Medical Cleaner, and MTI Castaneda Brush, all use a
rotating basket or a fixed one with an inner rotational
structure to fragment the thrombus at higher speeds.

• The hydrodynamic Microvena ATD (Amplatz Throm-
bectomy Device) uses a negative pressure created by a
recirculating vortex coming from antegrade high-pressure
fluid jets along the shaft and retrograde negative pressure at
the tip of the catheter.

• Flow-based devices explore the Venturi/ Bernoulli effect,
which is based on fast-flowing "positive" pressure fluid jets
that are directed to a "negative" pressure exhaustion lumen
of the catheter. The resulting negative gradient pressure
aspirates and causes fragmentation of thrombus. This
category is represented by the Boston Scientific Oasis, the
Possis Angiojet and Expedior, and the Cordis Hydrolyser.

Table 51 .2 Endovascular mechanical thrombectomy devices summary

Thrombectomy device

Size (Fr)

Wire

AVF

Anticoagulation

Venous patency
(1 year)

PErisk

Direct contact

Compliant balloon

5-14

Yes

Arrow PTD

5.5&7

Yes

Bacchus Solera

Yes

Rex Medical Cleaner

Yes

MTI Castaneda Brush

Yes

Hydrodynamic

Microvena Amplatz

7&8

Yes

Flow-based

Possis Espedior

Yes

Boston Sci Oasis

Yes

<90%

>1%

>90%

<1%

>90%

<1%

>90%

<1%

<90%

>1%

>90%

>90%
>90%

>1%

<1%
<1%

594

CHAPTER 51

Endovascular intervention for venous occlusion compared with surgical reconstruction

Figure 51.6 In the past, a simple algorithm
reserved surgical thrombectomy for patients
unable to receive thrombolysis. Newer,
percutaneous thrombectomy devices can be
used alone or in combination with thrombolysis.
Mechanical thrombectomy may include an
agent in the saline solution applied during
operation. Alternatively, mechanical
thrombectomy can debulkthethombus burden
before catheter-directed thrombolysis.

Venous DVT

Yes

Contraindications for
thrombolysis

Surgical thrombectomy with AVF

Mechanical endovascular
thrombectomy

[

Catheter-directed thrombolysis

PTA/STENT

Most of these devices provide a small pilot lumen that leaves
a substantial amount of residual thrombus and does not per-
form well with organized thrombus; however, patients with
multisegmental acute venous obstruction, with contraindica-
tions to lysis and surgery, now have the option of mechanical
restoration of venous outflow, minimizing complications and
leaving an opportunity for optimal venous reconstruction
with adjunctive angioplasty and stent deployment (Fig. 51.6).

Superior vena cava and upper extremity
venous thrombosis

Patients presenting with upper extremity edema due to ve-
nous thrombosis are becoming increasingly common, since
the use of central lines is so widespread. Ports and permanent
dialysis catheters, not to mention PICC lines, are placed in
thousands of patients daily. Not surprisingly, many experi-
ence partial or total venous thrombosis after a foreign body is
placed in the vein. 32 Patients are not always anticoagulated
and many have a neoplastic process that alters coagulability.
Furthermore, underlying venous stenosis, either from tho-
racic inlet compression or previous line placement, is often
present and not excluded before line placement. We see sever-
al groups of patients with upper extremity thrombosis. One
group involves patients with acute line-associated thrombosis
that may be treated with anticoagulation and line removal,
with the expectation that collaterals will form and the main
vein will recanalize. Infection is often suspected and, when
one line is removed, a new, contralateral line is placed. After a
series of these alternating jugular or subclavian lines, the pres-
ence of venous stenosis or mural thrombus is not a surprise.
Another group of patients includes otherwise healthy individ-
uals who develop "effort thrombosis/ 7 due to compression of
the subclavian or innominate segment by the first rib or bands
connecting the first rib to the scalene muscle (Fig. 51.7).

Thrombolysis followed by adjunctive surgical intervention
is an accepted sequence of therapy. 33 This procedure allows
minimally invasive removal of acute and subacute thrombus,
followed by release of the compression, with or without a vein

patch, and exemplifies where endovascular and traditional
surgical intervention work best in sequence. The strength of
each approach is used with the hope of providing the least
trauma to the vein and the best long-term result.

Occasionally, one sees symptomatic jugular vein thrombo-
sis that persists after line removal or occurs without any line
insult. Catheter-directed thrombolysis can be performed from
a femoral approach to lyse the thrombus. The thrombus is
adherent to the wall, so the chance of embolism is low, but
not zero. An underlying stenosis may be present, perhaps
related to thoracic inlet compression, plus or minus a foreign
body. Venous angioplasty alone may yield little improve-
ment, but stent placement in this area is not recommended.
Preservation of the innominate access is desirable and may
require placement of a stent in each segment.

Since the subclavian vein is notorious for restenosis, we are
conservative about placement of stents for benign disease
(Fig. 51.8). Removal of thrombus, anticoagulation, and
surgical removal of extrinsic compression fare better than
endovascular stents in this location. 33 This experience may
change with the introduction of coated and covered stents or
brachytherapy designed for larger veins. Currently, restenosis
occurs rapidly (less than 6 months) and in a large percentage of
patients who receive subclavian and innominate stents. 34
Monitoring with phlebology and reintervention may be nec-
essary to maintain patency. 35 For reasons probably related to
arm motion and muscle activity, restenosis is more aggressive
in the subclavian than the iliac vein. More central placement of
stents is associated with relatively less stenosis; however, if a
dialysis catheter resides in a stented SVC, the combination of
intimal hyperplasia and mural thrombus may be hard to
differentiate.

Recanalization and reconstruction of the SVC require care-
ful use of wires and balloon catheters and judicious placement
of stents. 36 The risk of complications in an area inaccessible to
manual compression demands careful technique and good
imaging equipment. Historically, bypass surgery for caval ob-
struction has not been reserved for highly selected patients.
Stanford et ah elegantly described patterns of central venous
occlusion in the upper body, as well as the spiral-vein bypass

595

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 51.8 (A) Venogram of right arm in a 42-year-old man with a 2-
month history of deep vein thrombosis which remained obstructed following
standard anticoagulation. He received a urokinase infusion for 48 h, ata rate
of 50 000 lU/h. When the wire passed into the superior vena cava, the
subclavian vein was dilated, and 1 0-mm Wallstents were placed. The patient
then underwent first rib resection. (B)The same patient with subsequent
venogram 6 months after stent placement. Intimal hyperplasia causing
luminal narrowing can be retreated with balloon angioplasty. Multiple
reinterventions have been required to maintain this patient's stent patency.
Thestented lumen has been compromised by the amount of hyperplasia. In
February 2004, the patient underwent his fifth dilation with a cutting
balloon.

Figure 51 .7 (Left) (A) Baseline contrast venogram in 38-year-old military
pilot presenting with acute right arm edema. (B) Venogram obtained after
24 h of catheter-directed thrombolysis. (C) The image shows balloon
angioplasty with a 10-mmx4-cm high-pressure balloon. This shows the
impressive strength of the extrinsic bands, which compress the subclavian
vein as it enters the thorax. The patient was taken to the operating room for
band release and first rib resection. The abnormal vein was replaced with a
vein patch and the patient was anticoagulated for 6 months.

596

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

Figure 51.9 (A) Subclavian and superior vena caval obstruction in a 44-
year-old man with tumor mass in the right upper lobe. (B) He presented with
superior vena cava (SVC) syndrome. Treatment included bilateral subclavian
catheter-directed thrombolysis followed by SVC mechanical thrombectomy

with the Amplatz device. (C) Right innominate and SVC Wallstents
(1 4 mm x 4 cm and 1 6 mm x 6 cm) were postitioned and dilated. This
provided excellent palliation of the upper body edema.

technique in 1987. 37 When self-expanding metallic stents
demonstrated enough "hoop strength" to sustain a reestab-
lished caval lumen, stents started being utilized for relief of
malignant caval obstruction in both superior and inferior loca-
tions (Fig. 51.9). The longer life expectancy of patients suffer-

ing from SVC syndrome caused by nonmalignant obstruction
raises the question of longer term patency of stents in
the SVC. Relatively few large series have been reported. Per-
cutaneous procedures are technically feasible, and, like sur-
gical intervention, the long-term patency requires clinical

597

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

surveillance and frequent reintervention. 38 The Mayo group
reported the primary vein bypass patency for benign disease
to be 63% and 53% at 1 and 5 years, respectively The improved
assisted patency is 85% and 80% at 1 and 5 years, respectively,
confirming the value of combining endovascular stents and
balloon interventions with surgery. 39

Generally speaking, the larger stents may stay open longer,
since every person develops some intimal hyperplasia. In
the SVC, stents can range from 12 to 18 mm in diameter; how-
ever, the larger stents require larger delivery systems and are
not as easy to deploy. The largest size of nitinol stents is 14 mm.
The main risk of stent placement in the cava is positioning
above the right atrium. Extension of the stent into the right
atrium may cause an arrhythmia or pericardial perforation.
The caval wall near the right atrial junction is also a potential
site for disruption, if a stent protrudes and angioplasty insults
the wall.

There are several special techniques adapted to the chal-
lenges of recanalizing the SVC. One is the "pull-through" tech-
nique that utilizes tension on each end of the wire, in order to
advance a catheter or balloon through a high-grade stenosis or
obstruction. The femoral approach is favored for advancing a
wire to the open lumen above the obstruction. This may be the
right internal jugular or the basilic-subclavian route. A 55-cm,
7- to 8-Fr sheath or guiding catheter is used coaxially with a 45°
angled catheter and an exchange-length 0.035-in Road Runner
wire (Cook, Inc.) to find and secure the narrow residual lumen
of an apparently obstructed SVC. By injecting contrast from
above and below, a strand of lumen might appear and be cap-
tured with roadmapping. The wire is "inched" cephalad until
it is "free" in the open lumen above the blockage. This can be
done in a reverse direction, as well. A nitinol snare is then used
to secure the wire tip and bring it outside the upper sheath.
A 100-cm, 5-Fr straight catheter is then advanced over the
exchange wire. When the catheter is outside each sheath, a
"working wire" [such as a 260-cm, 0.035-in Amplatz super stiff
(Boston Scientific)] replaces the hydrophilic wire. This system
provides enough support to advance a series of balloons for
incremental dilation of the caval lumen, before placing a
nitinol or Wallstent™ (Boston Scientific).

In the SVC, the advantage of a nitinol stent is the lack of
foreshortening, which assists with accurate placement. If plac-
ing a Wallstent™, we suggest approaching from above. This is
fine, if there is a jugular approach to accommodate a large
sheath size easily (e.g. 10 Fr for a 14- to 16-mm stent). A basilic
approach may limit one to a 12-mm size on a 7-Fr shaft through
an 8-Fr sheath. Basically, the cephalocaudal approach allows
one to flare the Wallstent™ in the right atrium and retract
it to the caval junction, in order to position the proximal edge
accurately in the SVC. We caution against the use of a short, 2-
cm stent in this location, as the stent can "jump" at the last mo-
ment of deployment, as it is released by the constraining
membrane.

Another technique, known as sharp recanalization, has

been described for use with chronic obstruction. 40 The com-
bined upper and lower accesses are used with triangulation
fluoroscopy. The sharp technique employs the back end of
a stiff wire or a truly sharp instrument, such as the TIPS nee-
dle /canula [5-Fr Roach-Uchida set and the 14-Fr Colapinto
needle (Cook, Inc.)]. The disadvantage, if the obstruction is
longer than 10 mm, is the relatively large track made by a stiff
instrument that can veer off-center with tough scar tissue.
An alternative is the long Seldinger [e.g. the 12-G TIPS needle
(Angiodynamics, Queensbury, NY, USA)], which protrudes
beyond the catheter and permits wire passage. Once the
obstruction is traversed, serial dilation and stent placement
can proceed.

IVC occlusion

Among the most symptomatic post-thrombotic patients are
individuals with a combination of iliac vein and IVC obstruc-
tion. Bilateral iliac thromboses can be associated with in-
frarenal IVC occlusion. The severity of symptoms implies
inadequate collateral inflow in the face of multisegmental
obstruction that is poorly recanalized. While not as common
as isolated iliofemoral thrombosis, patients with iliocaval
involvement represented between 1% and 10% in the larger
reported series. 41,42 Although relatively uncommon, the caus-
es of IVC obstruction are varied. In addition to primary caval
malignancy, which is rare, causes of caval thrombosis include
renal cell carcinoma, retroperitoneal fibrosis, radiation thera-
py, aortic aneurysm, ascites, trauma, surgery, and filter place-
ment. 43 Regardless of etiology, affected persons generally
develop significant retroperitoneal and abdominal collaterals
to compensate for occlusion of the infrarenal IVC. The
condition may remain occult when collaterals are adequate,
and becomes clinically evident only upon a subsequent
thrombotic episode.

Patients with IVC obstruction often complain that elevation
does little to relieve extremity edema. Poorly compensated il-
iocaval obstruction can cause unrelenting elevation of venous
pressure, causing a constant sensation of fullness in the groin
area, severe leg discomfort, and stasis ulceration. 44 More se-
vere hemodynamic dysfunction, per clinical class of disease,
can be identified in patients with chronic deep vein obstruc-
tion. 42 Clearly, quality of life can be significantly diminished.
Ultimately, some patients disabled by iliocaval occlusion have
been considered for bypass surgery. Patency results of long
bypass grafts for large-vein obstruction are inconsistent, at
best. 44 Believed to benefit from an arteriovenous fistula to
maintain patency, some bypass grafts fail, due to intimal hy-
perplasia at the anastomosis. Ironically, intimal hyperplasia
may be accelerated by arterial pressures. 41 Endovascular
reconstruction of chronic iliocaval occlusion is feasible. 41
Symptomatic relief is quite remarkable and the long-term
patency, in benign disease, parallels that of iliac stents. While

598

CHAPTER 51

Endovascular intervention for venous occlusion compared with surgical reconstruction

Figure 51.10 (A) Images show the before and after venograms in a 38-
year-old man who had a caval clip placed 1 7 years before presenting with
bilateral stasis ulcers. Following thrombolysis of the iliac and inferior vena
cava(IVC), a series of Wall stents we re placed to open the IVC and iliac veins
on the left. The stents, which were placed in 1 997, extend above the renal

vein confluence. The patient's ulcers remain healed and the stents remain
patent after 7 years without further intervention. (B) Left iliofemoral self-
expanding stents restore excellent flow to the IVC . Stent deployment
occurred after catheter-directed thrombolysis of the acute femoral
thrombosis which occluded the entire deep system of the thigh.

technically more challenging, it is an attractive therapeutic
option for carefully selected patients (Fig. 51.10).

The first endovenous stent was placed in the IVC in 1986
by Zollikofer and colleagues. 30 Since then, the literature con-
tains reports of 89 patients receiving stents to treat benign
(67%) and malignant (33%) IVC obstructions. 41 The Stanford
group reported on a series of 17 consecutive patients with
chronic IVC occlusion treated over a 6-year period. 43 The mean
duration of symptoms was 32 months. Thrombolysis and /or
stents were used with technical success in 15 (88%) patients.
After mean follow-up of 19 months, primary patency rate was
80% and the primary assisted rate was 87% (13/15). There
were no procedure-related complications, although four pa-
tients died during the follow-up period, due to underlying
disease.

Left iliac compression:
May-Thurner syndrome

When Zollikofer et al. 25 first placed a stent in the venous sys-
tem, the indication was treatment of intimal hyperplasia at the

proximal anastomosis of a common-femoral- to-common-iliac
bypass graft. Since then, stents have been shown to be an effec-
tive adjunct to surgery and balloon dilation, particularly in the
left common iliac segment 45 (Fig. 51.11). As in arterial disease,
"culprit" lesions are frequently discovered after removal of
acute thrombus. Verhaeghe et al. A6 reported discovery of an un-
derlying anatomical anomaly or lesion in 13/19 (68%) limbs
treated with catheter-directed rt-PA for iliofemoral thrombo-
sis. Ten of the venous lesions (77%) were uncovered after lysis
and 8/13 (62%) were treated with stent therapy. For the most
part, the condition is diagnosed in association with thrombo-
sis. As Cockett and Thomas observed early on, in acute cases,
"mere removal of the clot does little good in these cases as it
does not deal with the real cause of venous obstruction (i.e. the
stricture)/' 15 Although the majority of large-vein thromboses
recanalize sufficiently, a certain percentage of iliofemoral
DVT patients do not recover satisfactorily. Following standard
therapy of heparin, bed rest, and oral anticoagulation, they re-
main symptomatic with pain and /or edema. Some develop
ulcers. Unfortunately, predicting those patients who develop
a severe post-thrombotic syndrome is not possible. Cockett
and Thomas 15 identified the relative lack of iliac recanalization

599

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 51.11 (A,B) Pelvic venogram of a 58-year-old woman who
presented with an acute left lower limb deep vein thrombosis (DVT) following
a coronary arteriogram. The underlying pathology includes chronic left iliac
thrombosis and minimal residual lumen. The large transpubic collateral had

been visible for years. The patient was in a severe motor vehicle accident in
1 963 and experienced mild post-thrombotic symptoms prior to her first
documented DVT in 1 998.

and associated postinflammatory perivenous scarring as the
main etiologies of serious post-thrombotic sequelae. The
degree of uncompensated residual obstruction causes venous
claudication and the cutaneous changes associated with per-
sistent venous hypertension. Patients in this category were
among those first considered for surgical bypass. 47

The evolution and application of endovascular techniques
for treatment of acute iliofemoral thrombosis led to an
increased awareness of the left common iliac compression
syndrome that was initially described in 1906. 18 Opinions re-
garding whether or not the abnormality represents an ac-
quired venous lesion or residual congenital anomaly seem to
be influenced by observations during surgery vs. postmortem
analysis. 19-21 Between 1906 and 1963, four studies (960
cadaver dissections) were conducted to look at venous
anatomy 19 Compression of the left common iliac vein was
documented in 22%. 19/2 ° Based on these investigations, the

frequency of occurrence cited in adults was between 20%
and 34%, with an incidence of 4-17% in newborns up to
10 months and 10-26% in children between 1 and 8 years of
age. 19 ' 20

Left iliac compression explains why DVT predominantly
affects the left leg (Fig. 51.12). Many persons with occult left
iliac compression or occlusion are, in fact, asymptomatic.
Prethrombotic venous hypertension may be present, but re-
mains largely undiagnosed. Subtle asymmetry in leg or foot
size may be noticed, but is mostly ignored. Iliac compression
patients frequently present with associated femoral thrombo-
sis. In fact, the condition should be suspected in anyone with
acute left extremity DVT and /or symptoms of venous insuffi-
ciency. The youngest patient in whom we have seen venous
claudication and edema, with a normal saphenous vein and
phlebographic confirmation of left iliac compression, was a
12-year-old girl. In addition, we have treated three young

600

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

Figure 51.12 Pelvic venogram of a May-Thurmer compression/occlusion
of the left common iliac vein. Collaterals include the ascending left pericaval
veins and the trans-sacral venous plexus.

women aged between 17 and 24 with iliac stents to alleviate
symptoms caused by May-Thurner compression not yet
complicated by thrombosis. In their study of 94 patients with
suspected iliac vein obstruction, Neglen and Raju 22 have re-
ported on primary stenting of May-Thurner compression.
Forty-three symptomatic limbs with no history or phlebo-
graphy evidence of prior DVT showed better long-term re-
sults than those patients presenting with iliac compression
complicated by thrombosis.

Most published reports of endovascular therapy for iliac
and iliofemoral thrombosis involve single case reports or
small series of patients treated for acute DVT. A review of the
English language literature dealing with endovascular treat-
ment of thrombotic iliac occlusion or May-Thurner syndrome
reveals numerous accounts of limited experience with short
follow-up. 41 In early studies, Mickley and colleagues 48
demonstrated the use of stents after thrombectomy. They
reported that self-expanding stents were used to correct 10
severely stenosed venous segments in 8/30 patients operated
on for iliofemoral thromboses. Stenoses occurred within
3-6 months after thrombectomy and arteriovenous fistula
surgery. The lesions were treated with percutaneous Wall-
stent™ placement. Median follow-up at 17 months (range
3-23 months) showed 100% primary patency. In 1998, Mickley
et al. 49 reported experience stenting left iliac spurs discovered
after thrombectomy. Comparing the results of stented vs. un-
stented iliac spurs reveals rethrombosis in 16/22 (73%) un-

treated spurs, despite adequate anticoagulation, whereas only
1/8 (13%) stented spurs reoccluded (P = 0.01).

In 1994, the Stanford group reported their initial study 11
They treated 27 limbs in 21 patients (20 acute, seven chronic)
with catheter-directed thrombolysis. The average dose of
urokinase was 4.9 million IU (range 1.4-16 million IU) infused
over an average of 30 h (range 15-74 h). Sixteen limbs had un-
derlying stenoses that were treated with angioplasty (two) or
angioplasty and stent. 14 Two chronically occluded iliac veins
could not be traversed with a guidewire. Although primary
patency at 3 months was reported as 11/12 (92%), longer term
group follow-up was not included.

Nazarian and colleagues 50 at Minnesota discussed the
role of metallic stents after failure of balloon angioplasty or
surgery. Over a 65-month period, 55 patients received stents in
the subclavian veins (nine), innominate veins (three), superior
vena cava (four), inferior vena cava (three), iliac veins (29),
femoral veins (five), and portal veins (six). The series included
patients treated for malignant stenoses and benign chronic
iliac occlusions. They noted no significant difference of 1-year
patency between patients with and without a history of DVT,
or relative to the type of stent used (e.g. Gianturrco, Palmaz, or
Wallstent™). Stenotic lesions had a 1-year primary assisted
patency of 74%, compared with 57% for veins with prestent
occlusions (P = 0.15). Among the iliac veins, 13/29 were
initially occluded. Primary assisted patency for iliac veins was
66%, compared with 37% when femoral thrombosis accompa-
nied iliac DVT (P = 0.06). Two-year patency rates were signifi-
cantly lower in patients with no malignancy. Technical
problems were associated with single-module Z-stents that
persistently slipped above or below the stenosis. One external
iliac Z-stent fracture was identified at 5 months with no
adverse outcome.

In 1997, Bjarnason et al. 23 reported on treatment of 86 limbs
in 77 patients. The majority of the patients, mean age 47 years
(range 14-78 years), presented with acute DVT symptoms
of less than 14 days' duration (69/86; 78%), while 9/86 (11%)
had subacute thrombus (14-28 days) and 9/86 (11%) had
thrombus older than 28 days. The mean length of symptoms
prior to thrombolysis was 15 days (range 0-256 days). The
average dose of urokinase was 10.5 million IU (range 0.4-
24 million IU), and the average infusion time was 75 h (range
8-247 h). They reported greater technical success in treating
iliac veins (79%) vs. femoral veins (63%). We have also seen this
pattern. It reflects the fact that subclinical thrombosis is pre-
sent prior to clinical presentation with acute iliofemoral DVT.

Even though the initial technical success was similar
between patients undergoing stent placement vs. those who
did not, it was diminished in those patients with thrombus
older than 4 weeks, compared with those with more acute
conditions. Thrombosed superficial femoral veins are often
poorly recanalized and respond poorly to thrombolysis alone.
Eighty-six stents were placed in 38 (44%) of the 87 limbs treat-

601

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

ed for iliofemoral thrombosis. Seventy-five Wallstents™ were
placed in 36 limbs and 11 Gianturrco stents were placed in two
limbs. Interestingly, they found a lower 6-month primary
patency rate between stented (60%) and nonstented (75%)
iliac veins and 54% vs. 75% at 1 year (P = 0.11). At 1 year, the
secondary patency rate was 76% for stented and 82% for
nonstented vessels (P = 0.46). They hypothesized that patients
requiring stents presented with more severe chronic venous
disease, accounting for the poorer long-term results. Stented
patients were not uniformly maintained on warfarin longer
than 6 months.

An important report in the literature concerns the US NVR
(1994-1997). 24 This multicenter registry collected data on
473 iliofemoral DVT patients treated with endovascular
techniques. The study included 287 patients with adequate
follow-up. The majority of patients had acute presentation of
iliofemoral thrombosis (70%). The average dose of urokinase
was 7.8 million IU and nearly 50% required placement of an
iliac stent. Technical success, including placement of 104
stents, was 97%. Results were reported in terms of lysis grade.
Complete lysis, described as less than 10% residual thrombus,
was achieved in 60% of patients presenting with acute throm-
bus (less than 10 days). Among this group, 90% remained
patent at 12 months compared with 70% of those with less than
complete lysis. Patients were maintained on warfarin for
4-6 months. The study revealed greater 1-year patency in
limbs with iliac stenosis treated with angioplasty plus stent
(74%) vs. angioplasty alone (53%). A remarkably lower
patency rate (20% at 2 months) was observed in the five stents
placed in femoral segments.

Between 1988 and 1999, 84 patients were treated by our
group with combination endovascular therapy for chronic
lower extremity thrombosis at St Joseph's Hospital, in
Omaha, NE, USA. The mean age was 47.5 years (range
12-90 years). Patients received a mean dose of 8.7 million IU
(range 2-27 million IU) with a minimum 24-h infusion of
urokinase (range 24-120 h). Persistent venous stenoses after
thrombolysis required stent placement in 53% (62 patients and
71 limbs). Among these patients, 32 (51% of those stented and
28% of the total) had stents placed in the common femoral
and /or superficial femoral veins. Sixteen patients had a
single iliac stent placed for focal common iliac compression
associated with acute thrombosis (L/15, R/l). In this
subgroup, there was 100% primary patency with medial
follow-up of 24 months (range 8-50 months). Three patients
developed intimal hyperplasia within the stented segment,
causing increase in edema and discomfort, compared with
poststenting. This occurred between 6 and 12 months, and all
were treated with balloon dilation, resulting in resolution of
symptoms. All patients treated for chronic iliofemoral and ilio-
caval occlusion were examined before and after lysis /stenting
with ultrasound. In 16 patients, peak velocities in diseased and
unaffected iliac and femoral veins were suitable for analysis.
We found the median common iliac velocity in the normal

limb to be 44cm/s (n = 16). Comparison of prestent mean
(7.25 cm/s) and poststent mean (41.3 cm/s) common iliac vein
velocity demonstrated a significant difference (P < 0.0001),
whereas the mean stented left iliac vein velocity (41.3 cm/s)
was not significantly different from the untreated right iliac
vein mean velocity (47cm/s) (P = 0.4569). 51 Bjarnason 23 et at.
also reported that velocities of less than 25 cm/s in the stented
iliac segment correlated with poor patency. Further analysis of
stented patients showed 19 required two iliac stents, 20 re-
ceived three stents, and three large patients (greater than
113 kg) had more than three stents placed in the left iliac vein.
Overall, 1-year primary patency of patients with iliac stents,
including those with femoral stents, is 80% (57/71). Rethrom-
bosis occurred within 30 days in 8/71 (11%). Six of these pa-
tients were retreated with thrombolysis and additional stents
to improve inflow. This group included the three large patients
with more than three stents in their long iliac veins. It is impor-
tant to extend the stents from a vein segment with good flow to
the IVC. In our experience, failure to do so results in a higher
incidence of stent thrombosis and recurrent distal DVT
(Fig. 51.13). This simply obeys the principle of placement
of a bypass graft in the arterial system. Without adequate
inflow and outflow, the long-term patency is compromised
(Fig. 51.14). Symptomatic restenosis was documented and
dilated in 6/71 (9%) limbs. In all, the restenosis became clini-
cally apparent between 6 and 12 months. One-year secondary
patency is 94% (67/71).

All patients are discharged in class II compression hosiery
and /or the Velcro CircAid™ legging (CircAid™, San Diego,
CA, USA). Patients are seen in follow-up at 3, 6, and 12 months
and yearly thereafter. In addition to the pertinent physical ex-
amination, duplex imaging and hemodynamic testing (APG)
are repeated at these intervals, or if clinical decline occurs.
Warfarin is usually continued indefinitely, except in patients
with a single iliac stent placed after successful lysis of acute
thrombus. Normal distal veins provide sufficient confluence
of deep venous flow to favor patency. Low-molecular-weight
heparin (LMWH) (lmg/kg, every 12 h, or 1.5mg/kg, every
24 h) has been used in place of unfractionated heparin, with-
out complication. Patients remain at bed rest for at least 24 h
after sheath removal to prevent hematoma. Oral anticoagula-
tion with warfarin is maintained at less than 2.0 or greater than
4.5 s (PT/INR). Low levels can be supplemented with LMWH.
Elevated levels are often associated with antibiotics requiring
temporary adjustments of warfarin dose. Patients undergoing
surgery or dental work are placed on LMWH coverage while
discontinuing warfarin.

Raju et al. 52 have published on a relatively large endovas-
cular patient series in which thrombolysis was not used in
treating chronic venous obstruction. Primary stenting was
performed in a highly selected group of patients with
documented iliac vein stenosis or occlusion. 52 Overall, 118
Wallstents™ were placed in 77 iliac segments, 43 of which
were diagnosed with nonthrombotic iliac occlusion or nar-

602

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

Figure 51.13 (A,B)The right pelvic venogram after 24-h urokinase
infusion shows considerable irregularity and chronic changes. This 62-year-
old woman presented with a 4-week history of right leg edema and pain. She
had been treated with antibiotics for cellulitis before being referred to the
vascular surgeon for further evaluation. The image reveals a pattern that

requires stent placement for optimal venous flow betweeen the thigh and
the inferior vena cava. After successful removal of the acute thrombus, the
stents must extend from the iliac vein to the femoral vein in the upper thigh
to bypass the area of significant chronic disease.

rowing. In the remaining limbs, there was evidence of prior
DVT. As in the other reported large endovenous series, techni-
cal success was high (97%). Eighty-seven limbs were treated
with a 1-year primary patency of 82% and assisted and sec-
ondary patency of 91% and 92%, respectively. Their data
support our finding that focal iliac vein stenoses or occlusions
can be opened effectively and safely stented with good 1-year
patency rates. Clinical improvement usually parallels techni-
cal success; however, in severe chronic DVT involving multi-
ple venous segments, relief of large-vein obstruction can
produce clinical improvement, even when chronic venous in-
sufficiency remains. The technical aspects of venous stenting
and balloon dilation are described in detail.

Bleeding complications can occur with prolonged throm-
bolytic infusions, but they are relatively uncommon in the
post-thrombotic patient population, compared with cardiac
patients. In reviewing the complications reported in the litera-
ture, clearly the most common problem is minor bleeding at

the sheath site. 24 ' 41 Major bleeding, requiring transfusion, has
been reported in 1-5% of patients. This is more likely to occur
in obese patients; the use of ultrasound guidance for initial
puncture is strongly recommended. Pulmonary embolus (PE)
occurred once in our series, and continued thrombolysis re-
solved all symptoms. It has been reported in less than 1% of all
reported cases; therefore, IVC filters are not routinely recom-
mended. Among nearly 1000 iliofemoral venous thrombolysis
patients reported in the literature, death has occurred in less
than 1% due to PE (n = 1), sepsis (n = 2), retroperitoneal
hematoma (n = 1), and intracranial hemorrhage (n = l). 41
Rethrombosis that occurs in less than 30 days is generally
due to poor outflow or inflow and /or subtherapeutic
anticoagulation. When this occurs, retreatment with throm-
bolysis and additional stents is effective. Intimal hyperplasia
causing symptomatic restenosis occurred in approximately
10% of stented veins and can be effectively treated with
angioplasty 53

603

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

Figure 51.14 The tandem 10-mm diameter Wallstents were placed in April
1 993 and remain patent in April 2004 without further intervention.

Neglen and Raju's 53 use of IVUS documented the amount of
recoil occurring in vein segments following angioplasty
(Fig. 51.15). He also performed pressure gradients in all patients
and found a resting gradient of less than 2mmHg in 12/80
(15%). With papaverine injection, 28/80 (35%) demonstrated
a pressure gradient of less than 2mmHg across the stenosis.
Primary, assisted, and secondary patency rates at 1 year were
82%, 91%, and 92%, respectively. Neglen emphasizes the need
to extend the stent into the IVC. Early in our experience
rethrombosis occurred, due to this problem of incomplete
stenting across the compression site. Neglen also notes that re-
flux remains an issue in some, but not all patients. Symptomatic
venous hypertension, due to post-thrombotic valvular injury
or coexisting primary insufficiency, may not respond to com-
pression and may require additional intervention. 52 Meissner
et at. recently reported that, in addition to DVT-associated dam-
age of valves in the deep veins, similar damage can and does
occur in superficial veins when they are thrombosed; but paral-
lel occurrence in the uninvolved limb lends support to Raju's
previous suggestion that poorly functioning valves may, in fact,
be the cause of thrombosis, rather than the result thereof. 44/54

Neglen and colleagues have reported extensively on en-
dovascular therapy for iliac occlusion. Recently, they reported

tf.2 cl

^t* .^^^^^IHQ^^^E

m di

. . .^

■

Mm '

4.9 ii'div

IwuTCFl

Figure 51.15 Intravascular ultrasound imageof the left common iliacvein
in a 43-year-old woman with post-thrombotic syndrome and one episode of
left femoral-popliteal deep vein thrombosis. The images were obtained with
a6-Fr20-mHzprobe. Note the multiple small channelsand the wall
thickening, which is indicated by the echo-dense material separating the
lumina.

a series of 447 limbs stented to treat chronic iliac stenosis
greater than 50%, which represents the largest series in the lit-
erature. 55 They compared patients exhibiting an obstruction
with and without either deep or superficial reflux. The mean
clinical follow-up was 13 months, ± 12 months, and showed a
reduction in swelling and symptoms in both groups and heal-
ing in 55% of ulcerated limbs.

The authors emphasized that, in contrast to the patients in
whom only one stent was placed to treat iliac stenosis, a mean
of three stents was required to span the pelvic vein occlusion
and extend to good flow in the upper thigh. They pointed out
the following:

• Longer occlusive segments (almost one foot) can be
recanalized with stents and have good short-term patency.

• Stenting through an IVC filter is possible and the viscera
seem to tolerate these long stent systems.

• Anticoagulation is not needed; antiplatelet drugs, such as
ketorolac, heparin, and aspirin, should be used.

• The risk is negligible and the benefit, in terms of pain
reduction, is "significant."

• Healing of ulceration occurred in 7/ 12 ulcers, even though
reflux was uncorrected in some patients.

• Physiologic testing with APG and ambulatory venous pres-
sure (AVP) did not show any difference pre- or postinter-
vention because of the collaterals.

604

chapter 51 Endovascular intervention for venous occlusion compared with surgical reconstruction

Discussion

Endovenous therapy for venous occlusive disease is feasible
and generally of value, yet in a state of evolution. In limbs
with acute thrombosis, with or without short-segment iliac
stenosis /occlusion, catheter-directed thrombolysis can effec-
tively remove thrombus and thereby restore flow and limit
valve injury Stenting appears to effectively treat iliac com-
pression and venous stenosis better than angioplasty alone. In
patients with little or no residual obstruction, flow is generally
adequate to promote long-term iliac stent patency, even with-
out warfarin. Chronic venous occlusive disease is technically
more challenging to treat. It takes longer to complete the revas-
cularization, and attention must be focused on creating con-
tinuity of flow within the limb to optimize venous drainage.
Just as success in arterial revascularization procedures often
depends on the state of distal runoff, so the fate of iliac vein
revascularization relies upon inflow. Neglect of distal inflow
status (i.e. tibial and popliteal veins) can result in compro-
mised deep venous flow, jeopardizing stent patency. As Bjar-
nason et al. 23 acknowledged, failure to restore flow adequately
in the superficial femoral vein poses a threat to iliac patency;
however, I share their impressions that superficial femoral
vein lysis is most successful when performed with acute
thrombus during the first episode of DVT. Chronic SFV occlu-
sion is more difficult to correct than chronic iliac obstruction.
The status of the tibial and popliteal veins is a major consider-
ation, since patients remain very symptomatic with distal
disease only, despite normal-appearing proximal venous
segments.

The higher long-term patency rates in our series, despite the
extensive disease in many patients, can be explained in two
ways. First is initial attention to distal inflow. A pedal phlebo-
gram is performed in all patients to assess the flow pattern in
the entire leg. Reestablished flow in the thigh and pelvis can
fail if tibiopopliteal veins remain occluded. Distal thrombotic
obstruction is simultaneously treated with flow-directed
stenting, if they are hemodynamically significant, as demon-
strated by stasis and persistence of dominant collaterals. If
there is a pullback pressure differential of greater than
5mmHg above and below a stenosis, we may consider place-
ment of an additional stent.

Second, patients who are maintained on long-term warfarin
are diligently monitored by the endovascular service. Patients
are encouraged to be involved and to understand anticoagula-
tion. Most of our patients have extensive evidence of distal
post- thrombotic change. When patients have no prior history
or evidence of DVT, iliac stents appear to remain patent with-
out long-term oral anticoagulation. Raju et ah, 52 for example,
maintain their patients on aspirin, not warfarin. At Creighton,
since placement of the first iliac stents in 1993, more complica-
tions (rethrombosis) have occurred due to subtherapeutic
anticoagulation than an elevated INR. One complication

was recorded in a 40-year-old patient taking buproprion
extended-release as an antismoking medication. His elevated
prothrombin time resulted in a spontaneous elbow hemar-
throsis, which, albeit painful, resolved without sequelae.

Endovascular therapy is focused on treating the obstructive
component of venous hypertension by restoring patency with
thrombus removal and /or stent placement. Clinical evalua-
tion of patients and follow-up analysis, using standard report-
ing guidelines, are absent in the literature. Diagnosis of
associated hypercoagulability disorders and valvular reflux
are important, but somewhat ignored issues, as well. Increased
use of LMWH holds promise for improved recanalization. In
the future, we hope to be able to better predict which patients
will have limited autolysis and inadequate recanalization
after extensive thrombosis, to improve patient selection for
early thrombolysis. Perhaps newer antiplatelet and anti-
inflammatory drugs will play a role in preventing thrombosis.
One day, percutaneous placement of substitute valves will be a
treatment option for reflux. If anything, the evolution of en-
dovascular therapy has engendered respect for the complexity
of venous disease. Treatment of chronic peripheral obstructive
disease makes us keenly aware of how much this condition
adversely affects a patient's quality of life. Not surprisingly,
those patients who have undergone successful revasculariza-
tion with the tools and techniques discussed in this chapter are
the strongest advocates for endovascular therapy.

Many in vascular surgery and interventional radiology
have accepted the availability and success of catheter-directed
thrombolytic therapy and thrombectomy devices. The experi-
ence of the last 15 years has, however, failed to alter the stan-
dard therapeutic approach to treatment of acute extremity
thrombosis by most physicians. Patients receive heparin, fol-
lowed by Coumadin. Granted, there are many physicians who
consider the option of minimally invasive therapy for a patient
with phlegmasia; but, in general, the majority of the medical
community await level I data, comparing the risks and long-
term benefits of thrombolytic therapy as a first-line treatment
of occlusive DVT. The difficulties inherent in obtaining such
data include the need for a longitudinal study in an increas-
ingly mobile society, as well as agreeing upon therapy and
finding patients with similar extent and age of thrombosis
who have not had DVT prior to treatment. Although credible
studies addressing quality of life related to post-thrombotic
syndrome advocate early intervention to reduce thrombus,
DVT continues to be widely treated with anticoagulation, rest,
and compression stockings, rather than early ambulation,
thrombus removal, anticoagulation, and stockings.

The standard therapeutic approach of bed rest, anticoagula-
tion, and graduated compression hosiery has given way to
ambulatory therapy with LMWH followed by warfarin. How
do we know who will fail this treatment and be left with a
symptomatic, debilitating post-thrombotic syndrome?
Whereas Strandness et al. 56 predicted that approximately 70%
of DVT patients with obstructive multisegmental thrombus

605

pa rt v I Comparison of conventional vascular reconstruction and endovascular techniques

will be left with some signs or symptoms of post-thrombotic
syndrome, we really do not know how to predict which pa-
tients will autolyse thrombus enough to recover well and pre-
serve valvular function vs. those who will fail to respond to
standard therapy Invariably, a certain number of patients
"fail" heparin and warfarin therapy In our experience, phle-
bography and duplex studies in these individuals demon-
strate minimal deep-system flow due to a poorly recanalized
deep vein, inadequate collateral flow via the saphenous sys-
tem or profunda and transpelvic collaterals, indicating chron-
ic iliac occlusion. When the diagnostic revelations follow
months or years of thrombosis, there are few therapeutic op-
tions. Most patients live their lives with a disability, despite
compression stockings, elevation, and anticoagulation.
Endovascular reconstruction provides an option for selected
patients but there are relatively few medical centers that offer
this therapy, since it requires specialization in endovascular
therapy for nonarterial disease. Long-term patency has not
been documented in very many patients. It appears to be
greatest if only common iliac disease is encountered. 55 Con-
trary to the expectations of many, removal of iliac obstruction
does not result in worsening of post-thrombostic symptoms or
increased axial-vein reflux. 55 Symptomatic limbs dominated
by obstruction, rather than reflux, experience significant clini-
cal improvement by increasing flow, despite the presence of
damaged or stented valves; therefore, although venous
insufficiency most frequently results in valvular reflux, the
presence of a treatable obstructive component in venous
disease may often be overlooked in the clinical evaluation
since we are only beginning to gather long-term follow-up.
The 10-year clinical outcome study at Creighton showed a sta-
tistically significant decrease in post-thrombotic pain reported
by patients who received urokinase for acute DVT vs. those
who received standard heparin therapy without catheter-
directed thromb lysis. 57

Summary

Management of venous disease is becoming a specialty among
vascular interventionalists, but all physicians should recog-
nize it and, unfortunately, they do not. It is also common for
patients to ignore venous disease. The medical community
has not proactively treated venous disorders and patients
often accept leg discomfort and swelling as part of growing
older and not exercising. The lower extremity heaviness and
fatigue develop slowly and become familiar. Patients adjust
their lifestyle and do not always recognize the impact of "leg
problems" on their quality of life. This is seen with both
arterial and venous conditions. As in all areas of vascular in-
tervention, patient selection is a key to good clinical outcomes.
If the clinician recognizes venous disease, even inexpensive,
conservative treatment, such as properly fitting compression
stockings, can significantly help patients with mild condi-

tions. There is opportunity to provide an important clinical
service by providing care for venous problems, since they are
so prevalent in an aging society. Clinicians must observe limbs
for signs and symptoms of venous disease, in order to treat or
refer patients for further care. Given the immense socioeco-
nomic impact of untreated venous disease, it is our obligation
to recognize and treat the many manifestations of this
disorder.

References

1. Wakefield TW. Treatment options for venous thrombosis: Invited
review. / Vase Surg 2000; 31:613.

2. Bergan JJ. Prospects for minimal invasion in venous disease
(expert commentary). / Vase Surg 1999; 33:247.

3. Bergan JJ, Kumins NH, Owens EL, Sparks SR. Surgical and
endovascular treatment of lower extremity venous insufficiency.
/ Vase Interv Radiol 2002; 13:563.

4. Mewissen MW, Seabrook GR, Heissner MH et al. Catheter-
directed thrombolysis for lower extremity deep venous thrombo-
sis: report of a national multicenter registry. Radiology 1999; 211 :39.

5. Semba CP, Dake MD. Iliofemoral deep vein thrombosis: aggres-
sive therapy with catheter-directed thrombolysis. Radiology 1994,
191:487.

6. Comerota AJ, Throm RC, Mathias SD et al. Catheter-directed
thrombolysis for iliofemoral deep venous thrombosis improves
health-related quality of life. / Vase Surg 2000; 32:130.

7. Vedantham S, Vesely TM, Parti N et al. Lower extremity venous
thrombolysis with adjunctive mechanical thrombectomy. / Vase
Interv Radiol 2002; 13:1001.

8. McNamara TO, Dong P, Chen J et al. Bleeding complications asso-
ciated with the use of rt-PA versus r-PA for peripheral arterial and
venous thromboembolic occlusions. Tech Vase Interv Radiol 2001;
4:92.

9. Weitz JI. Limited fibrin specificity of tissue-type plasminogen
activator and its potential link to bleeding. / Vase Interv Radiol
1995;6:19S.

10. Meissner MH. Overview: The management of lower extremity
venous problems. Semin Vase Surg 2002; 15:1.

11. Semba CP, Dake MD. Iliofemoral deep vein thrombosis: aggres-
sive therapy with catheter-directed thrombolysis. Radiology 1994;
191:487.

12. Killewich LA, Macko RF, Cox K et al. Regression of proximal deep
venous thrombosis is associated with fibrinolytic enhancement. /
Vase Surg 1997; 26:861.

13. Hood DB, Weaver FA, Modrall JG, Yellin AE. Advances in the
treatment of phlegmasia cerulea dolens. Am J Surg 1993; 166:
206.

14. Beyth RJ, Cohen AM, Landefeld CS. Long-term outcomes of
deep-vein thrombosis. Arch Intern Med 1995; 155:1031.

15. Cockett FB, Thomas ML. The iliac compression syndrome. Br }
Surg 1965; 52:816.

16. Jaszczak P, Mathiesen FR. The iliac compression syndrome. Acta
ChirScand 1978; 144:133.

17. Moller JW, Eickhoff JH, Buchhardt Hansen HJ et al. The iliac
compression syndrome. Acta Chir Scand 1989; 502:141.

18. McMurrich JP The occurrence of congenital adhesions in the

606

CHAPTER 51

Endovascular intervention for venous occlusion compared with surgical reconstruction

common iliac veins, and their relation to thrombosis of the femoral
and iliac veins. Am J Med Sci 1908; 135:342.

19. Ehrich WE, Krumbhaar EB. A frequent obstructive anomaly of the
mouth of the left common iliac vein. Am Heart } 1943; 26:737.

20. May R, Thurner J. The cause of the predominantly sinistral occur-
rence of thrombosis of the pelvic veins. Angiology 1957; 8:419.

21. Negus D, Fletcher EW, Cockett FB, Thomas ML. Compression and
band formation at the mouth of the left common iliac vein. Br }
Surg 1968; 55:369.

22. Neglen P, Raju S. Balloon dilatation and stenting of chronic iliac
vein obstruction: technical aspects and early clinical outcome. /
Endovasc Ther 2000; 7:79.

23. Bjarnason H, Kruse JR, Asinger DA et ah Iliofemoral deep vein
thrombosis: safety and efficacy during 5 years of catheter-directed
thrombolytic therapy. / Vase Interv Radiol 1997; 8:405.

24. Mewissen MW, Seabrooke GR, Meissner MH et ah Catheter-
directed thrombolysis for lower extremity deep vein thrombosis:
report of a national multi-center registry. Radiology 1999; 211:39.

25. Zollikofer C, Largiader I, Bruhlmann WE Endovascular stenting
of veins and grafts: preliminary clinical experience. Radiology
1988; 167:707.

26. Charnsangavej C, Carrasco CH, Wallace S et ah Stenosis of the
vena cava: preliminary assessment of treatment with expandable
metallic stents. Radiology 1986; 161:295.

27. Sawada S, Fujiwara Y, Koyama T et ah Application of expandable
metallic stents to the venous system. Acta Radiol 1992; 33:156.

28. El Feghaly M, Soula P, Rousseau H et ah Endovascular retrieval of
two migrated venous stents by means of balloon catheters. / Vase
Surg 1998; 28:541.

29. Slonim SM, Dake MD, Razavi MK et ah Management of misplaced
or migrated endovascular stents. / Vase Interv Radiol 1999; 10:851.

30. Zollikofer CL, Antonucci F, Stuckmann Get ah Use of the Wallstent
in the venous system including hemodialysis-related stenoses.
Cardiovasc Interv Radiol 1992;15:335.

31. Stainken BE Mechanical thrombectomy: basic principles, current
devices, and future directions. Tech Vase Interv Radiol 2003; 6:2.

32. Martinelli I, Cattaneo M Panzeri D et ah Risk factors for deep
venous thrombosis of the upper extremities. Ann Intern Med
1997; 126:707.

33. AbuRahma AF, Robinson PA. Effort subclavian vein thrombosis:
evolution of management. / Endovasc Ther 2000; 7:302.

34. Sharafuddin MJ, Sun S, Hoballah JJ. Endovascular management
of venous thrombotic diseases of the upper torso and extremities.
J Vase Interv Radiol 2002; 13:975.

35. Hammer F, Becker D, Goffette P, Mathurin P. Crushed stents in
benign left brachiocephalic vein stenoses. / Vase Surg 2000; 32:392.

36. Qanadli SD, El Hajjam M, Mignon F et ah Subacute and chronic
benign superior vena cava obstructions: endovascular treatment
with self-expanding metallic stents. Am J Roentgenol 1999; 173:159.

37. Stanford W, Jolles H, Ell S, Chiu LC. Superior vena cava obstruc-
tion: a venographic classification. Am} Roentgenol 1987; 148:259.

38. Sprouse II LR, Lesar CJ, Meier III GH et ah Percutaneous treatment
of symptomatic central venous stenosis angioplasty. / Vase Surg
2004; 39:578.

39. Kalra M, Gloviczki P, Andrews JC et ah Open surgical and
endovascular treatment of superior vena cava syndrome caused
by nonmalignant disease. / Vase Surg 2003; 38:215.

40. Farrell T, Lang EV, Barnhart W. Sharp recanalization of central
venous occlusions. / Vase Interv Radiol 1999; 10:149.

41. Thorpe PE, Osse FJ, Dang HP. Endovascular reconstruction for
chronic iliac vein and inferior vena cava obstruction. In: Gloviczki
P, Yao JST, eds. Handbook of Venous Disorders: Guidelines of the
American Venous Forum, 2nd edn. London/New York/New Delhi:
Arnold, 2001:347.

42. Sharafuddin MJ, Sun S, Hoballah JJ et ah Endovascular manage-
ment of venous thrombotic and occlusive diseases of the lower
extremities. / Vase Interv Radiol 2003; 14:405.

43. Razavi MK, Hansch EC, Kee ST et ah Chronically occluded inferior
venae cavae: endovascular treatment insufficiency. In: Ballard JL,
Bergan JJ, eds. Chronic Venous Insufficiency. London: Springer,
1999:179.

44. Raju S. Venous insufficiency of the lower limb and stasis
ulceration: changing concepts and management. Ann Surg 1983;
6:688.

45. Heniford TB, Senleer SO, Olsoffka JM et ah May-Thurner syn-
drome: management with endovascular surgical techniques. Ann
Vase Surg 1998; 12:482.

46. Verhaeghe R, Stockx L, Lacrix H et ah Catheter-directed lysis of
iliofemoral vein thrombosis with use of rt-PA. Eur Radiol 1997;
7:996.

47. Gloviczki P, Pairolero PC, Toomey BJ et ah Reconstruction of large
veins for non-malignant venous occlusive disease. / Vase Surg
1992; 16:750.

48. Mickley V, Friedrich JM, Huttschenreiter S et ah Long-term results
of percutaneous transluminal angioplasty and stent implantation
in venous stenoses following transfemoral thrombectomy. \#1SA
1993; 2:44.

49. Mickley V, Schwagierek R, Rilinger N. Left iliac venous thrombo-
sis caused by venous spur: treatment with thrombectomy and
stent implantation. / Vase Surg 1998; 28:942.

50. Nazarian GK, Austin WT, Wegryn AS. Venous recanalization
by metallic stents after failure of balloon angioplasty or surgery:
four-year experience. Cardiovasc Interv Radiol 1996; 19:227.

51. Thorpe, PE, Osse FJ, Dang HP et ah Duplex velocoties: an impor-
tant tool for assessment in venous occlusive disease. Accepted for
oral presentation at American Venous Forum 13th Annual Meet-
ing, Fort Myers, FL, February 25, 2001.

52. Raju S, Owen Jr S, Neglen P. The clinical impact of iliac venous
stents in the management of chronic venous insufficiency. / Vase
Surg 2002; 35:8.

53. Neglen P, Raju S. In-stent recurrent stenosis in stents placed in the
lower extremity versus outflow tract. / Vase Surg 2004; 39:181.

54. Meissner MH, Caps MT, Zierler BK et ah Deep venous reflux and
superficial venous reflux. / Vase Surg 2000; 32:48.

55. Neglen P, Thrasher TL, Raju S. Venous outflow obstruction: an
underestimated contributor to chronic venous disease. / Vase
Surg 2003; 38:879.

56. Strandness DE, Langlois Y, Cramer M et ah Long-term sequelae of
acute venous thrombosis. JAMA 1983; 250:1289.

57. Thorpe PE, Kaul AF, Bosch R. Abstract No. 27. Thrombolytic ther-
apy versus anticoagulation for treatment of deep vein thrombosis:
10-year clinical observation. 29th Annual Scientific Meeting,
Society of Interventional Radiology. Phoenix, AZ. / Vase Interv
Radiol 2004; Suppl. S143-S302:S153.

607

Vascular Surgery: Basic Science and Clinical Correlations, Second Edition

Index

Note: page numbers in italics refer to

figures: page numbers in bold refer to

tables.

Abbreviations used in this index include:
CAS, carotid angioplasty and stenting
LDL, low-density lipoprotein
MRI, magnetic resonance imaging

abcixamab
CAS 578

intimal hyperplasia 139
abdominal aorta

atherosclerotic lesions 61, 61-62
coarctation 8-10
infrarenal 62

balloon angioplasty 507
abdominal aortic aneurysm (AAA)
classification 259
endovascular prostheses 520-529
Ancure endograft/Guidant 520-523,

522
AneuRx /Medtronic 523-524
Excluder/Gore 524-526
Zenith 527-529
expansion rate 169-170
indications for surgery 171
intravascular ultrasound 412, 413
mortality 171
pain 170

reperfusion injury 248-249
rupture, Ancure endograft/Guidant 521
rupture risk 171
survival 170,170
abdominal vessels, MRI 379, 380
A-P fibers 233, 235, 236
ablation, portal hypertension 283-284
absent inferior vena cava 13
absorption coefficient 405
accessory head gastrocnemius muscle 127
accessory vessel detection 354
Acculink 580
Accunet 580

acetazolamide 329-330, 330
acoustic impedance 316
acoustic interface, intravascular ultrasound

403
acquisition timing 374
activated thromboplastin time (A-PTT),
monitoring of heparin-induced
thrombosis 35

acute limb ischemia

popliteal entrapment syndromes 130
thrombolytic therapy 458-461, 460
acylated plasminogen-streptokinase
activator complex (APSAC)
457-458
adductor canal (Hunter's canal) 128
adductor canal syndrome 128
adenosine

cerebral blood flow 252
endovascular repair thoracic aorta 557
adenosine diphosphate (ADP), hemostasis

28
adenosine triphosphate (ATP), intestinal

ischemia 220
adenylate cyclase (cAMP) messenger

system 47
adhesion molecule expression,

atherosclerosis 44, 45
adrenal glands, angiotensin peptides

188-189
adrenergic receptors, vasospastic disorders

83-86
adrenocorticotrophic hormone secretion

189
adventitia 57

adventitial cystic disease 119-125
angiography 122, 123, 393
autogenous vein grafting 123
diagnosis 122-123
etiology 120-121
histology 119-120
presentation 122
treatment 123-124
afferent arteriolar resistance 186
afferent inhibition 233
age

atherosclerosis 442
Buerger's disease 95
entrapment syndromes 130
thrombosis 196
A-y fibers 233

Ahn Thrombectomy Catheter 500-501
air plethysmography, venous stasis disease

198
albumin trapping, arteriovenous shunts

325
alcohol 448
aliasing 313, 313
allodynia 235, 236
allograft replacement infected grafts 485

alpha 2 -macroglobulin 455

alpha 2 -plasmin inhibitor 455

alpha adrenergic receptors 80-81, 83

type 1 antagonists 111

type 2 81
alprostadil 469
alteplase 458
alveolar dead space 202
amaurosis fugax 255
American College of Rheumatology (ACR)

114
aminoglycoside antibiotics 487
ammonia, encephalopathy 277-278
A-mode devices 320
amputation

endarterectomy 537

level selection/radionuclide scanning
326

peripheral arterial disease 445
anastomosis, angioscopic visualization

430, 430
anatomy

aneurysm 163-164

cerebral circulation 251, 252

entrapment syndromes 127-128

erectile dysfunction 270-272

interscalene triangle 147

lymphatic system 208-209

mesenteric circulation 215-217, 216

mesenteric lymphatics 208, 209

popliteal artery entrapment 127-128

thoracic outlet 147, 149

upper limb lymphatic system 208, 210
ancrod 470

in heparin-induced thrombosis 36
Ancure endograft 520-523

abdominal aortic aneurysm rupture 521

device withdrawal 523

five-year outcome 521

one-year outcome 521

periprocedural outcome 520-521
Anderson Stuart, T.P 126
AneuRx/Medtronic 523, 523-524, 527, 556
aneurysmal disease 162-179

angiography 170

atherosclerosis 166

connective tissue disorders 172-175

epidemiology 162-163

genetics 167-168, 168

histiolytic enzymes 166-167

infected aneurysms 175-176

609

Index

inflammatory aortic aneurysms 167

natural history 168-172

pathogenesis 163-167
aneurysmal sac size

Ancure endograft 521

Excluder endograft 526, 527

Zenith endovascular graft 528
aneurysms

anatomy 163-164

cerebral 358

computed tomography angiography
354, 358

enlargement 170

growth 164

inferior vena cava 14

intracranial /medial fib ro dysplasia 67

locations 162, 163

popliteal artery 131
entrapment 130

repair 170-171, 171

rupture 164, 170, 171

descending thoracic aneurysms 554

sciatic artery 11-12, 12

syphilitic 176

tuberculosis 176

ulnar artery 366

ultrasound evaluation 323, 323
angina, variant 82-83
AngioDynamics catheter 591
angiogenesis 22
angiography 385-400

adventitial cystic disease 122, 123, 393

aneurysmal disease 170

aorta 350

arterial access 385-387

arterial applications 392-393

arterial dissection 390

arterial spasm 390

arteriovenous fistulas 390, 396

arteriovenous malformations 16

axillobrachial approach 387

baseline, for balloon angioplasty
504-505

Buerger's disease 96-97, 392-393

carotid arterial disease 358, 363

cerebral 395

cholesterol emboli 390

circle of Willis 363

complications 389-391, 390

computed tomography 348-370

contraindications 385

contrast media 387-389

diabetes mellitus 391

entrapment syndromes 132

ergotism 107, 107-108

extravasation 391

femoral artery 385-387

femoropopliteal 534, 534

hematoma formation 389-390

indications 385, 386

magnetic resonance 372-373

nephropathy 391

neurovascular applications 395, 397

portal hypertension 278, 279-280

pseudoaneurysm formation 390

pulmonary 334, 391

pulmonary embolism 391

Raynaud's syndrome 392-393

renal arteries 354, 357

retroperitoneal hemorrhage 390

spinal 391

stent deployment 571-573

subclavian steal syndrome 393

superficial femoral artery 387

Takayasu's arteritis 392

technique 385-387

thoracic aorta 557

thoracic outlet syndrome 365

trauma 392, 392

vasculitis 392

vasospasm 390

vasospastic disorders 392-393

vasovagal syncope 390

venous applications 394

vessel patency 325
Angioguard 580
Angiojet Rheolytic Thrombectomy System

501
Angiojet thrombectomy device 594
angioplasty

aortoiliac disease 548

arch vessel lesions 568

atherectomy and 512

carotid 577-580

coronary 411

infrapopliteal vessels 538

intravascular ultrasound 411-412, 412

laser 512-513

see also balloon angioplasty;
percutaneous transluminal
angioplasty (PTA)
Angiopump 426

angioscopy (peripheral vascular surgery)
423-438

blood displacement 425

bypass grafts 432, 432-437

carbon dioxide 425

clinical applications 430-437

equipment 424, 424-425, 428

flow rates 426, 426

graft patency 435

history 423-424

infrainguinal bypass graft monitoring
432-436, 433, 434

interpretation 429-430

irrigation 425,425-427

irrigation fluid volume 426, 426, 427

reoperative surgery for failing grafts
436-437

saline irrigation 425-427

techniques of 425-430

valvulotome injury 430

vasospasm 428

vein conduit preparation 436, 436

vein patency 433

vein quality 433
angiotensin-converting enzyme (ACE)

inhibitors

atherosclerosis 447

renal scintigraphy 332, 333

intimal hyperplasia 137, 139

renovascular hypertension 182

vascular tone 21
angiotensin formation pathways 181, 181
angiotensin II

intestinal blood flow 218

renal autoregulation 187

tubuloglomerular mechanism of
autoregulation 185
angiotensinogen 182

metabolites 181, 182, 183
angiotensin peptides 182

major organ systems 188

myogenic mechanism 184

renal actions 183

renal function 183-184

tubuloglomerular feedback 185, 185-186
angiotensin receptors 182
animal models, intimal hyperplasia

135-139
ankle-brachial pressure index (ABPI)

atherosclerosis 226

blood pressure measurement 303

Buerger's disease 95

entrapment syndromes 131

femoropopliteal disease treatment 534
antegrade femoral puncture 403
Anthony of Egypt, Saint 101-102
antibiotic-bonded prosthetic grafts 485,

489
antibiotics

atherosclerosis 449

graft infection 480

nucleic acid synthesis inhibition 487

prophylaxis 488, 488
graft infection 480
vascular surgery 488, 488

protein synthesis inhibition 487

resistance 487-488

therapeutic uses 488-489

tissue concentration 485

vascular surgery 485-490

see also individual drugs
anticoagulation

alternative, in heparin-induced
thrombosis 36

ergotism 110

hepatic metabolism 31

oral therapy 592

pathways 31
antioxidants

atherosclerosis 53

lipid-lowering therapy 473
antiphospholipid antibodies 33
antiplatelet agents

balloon angioplasty 504

intimal hyperplasia 139
antithrombin III (AT-III) 31

deficiency 32, 37, 194-195
aorta

computed tomography angiography
350

inflammatory aneurysms 167

infrarenal, balloon angioplasty 507

MRI 378
aortic arch

angioplasty 510, 510-512

branch abnormalities 8

developmental abnormalities 8

embryology/development 3, 4, 5
aortic bifurcation 507

angioplasty 506
aortic coarctation 350, 353
aortic dissection 350, 354

intravascular ultrasound 408-409, 409

spin-echo magnetic resonance images
372, 372 '

type B 554-555, 562

endograft treatment 560-563, 561
endovascular repair 556
aortic graft infection, revascularization

timing 483-484
aortic-peripheral system, atherosclerosis

59
aortic ring 8, 9
aortic rupture, traumatic 555, 563, 564

610

Index

aortic stent-grafts, postplacement

evaluation 354
aortic transection 350, 555
aortic ulcers, penetrating 555, 563
aortic wall tension 57, 58
aortobifemoral graft infections 484
aortoduodenal fistula 482
aortofemoral bypass 543-544
aortoiliac angioplasty 506-507
aortoiliac disease

adjunctive infrainguinal reconstruction
549, 550

angioplasty 548

comorbidity 545

endovascular recanalization 544-545
surgical reconstruction vs. 543-553

presentation 543

stenting 547-548
outcomes 550-551

surgical therapy 543-544
aortoiliofemoral bypass grafts, mechanical

erosion 480
apolipoprotein E 52-53
arc of Buhler 11, 11
arc of Riolan 216, 216
area postrema 189
Aristotle 207
arterial access

angiography 385-387

femoropopliteal disease 533-534
arterial anomalies/disease 8-12

flow changes 297-298
arterial dissection, angiography 390
arterial dysplasia, developmental 76-

78
arterial fibrodysplasia

definition 66

developmental arterial dysplasia 76-78,
77,78

intimal fibroplasia 72-75, 75

medial fibrodysplasia 66-68

pathogenesis of 66-79

perimedial dysplasia 68-72
arterial flow 295-297
arterial geometry 296
arterial graft surveillance program 460
arterial injury, intimal hyperplasia 138
arterial insufficiency

assessment 325-326

erectile dysfunction 268
arterial occlusion, intestinal ischemia

219-220
arterial pressure 295
arterial reconstruction, Buerger's disease

98
arterial spasm 390
arterial stenosis

critical 298

energy losses 297-298
arterial system 3-5

structure /function 55-57

susceptible regions 60-62
arterial tangential wall tension 57
arterial thrombosis

detection and presentation 35

heparin-induced 34

in-situ, pathogenesis 495

protein C deficiency 194
arteria lusoria 8
arterial velocities, MRI 379-381
arterial wall

aneurysm formation 164-165, 165

cultures 480

physiological adaptation 57-59
arteriography

erectile dysfunction 272

graft infection 482
arteriolar pressure, compartment

syndrome 242
arteriosclerosis obliterans, Buerger's

disease vs. 96
arteriovenous fistulas

angiography 390, 396

Brescia technique 512

portal hypertension 277
arteriovenous gradient, compartment

syndromes 241
arteriovenous grafts 512
arteriovenous malformations 14, 16-17

angiography 16

clinical syndromes 15
arteriovenous shunts, varicose veins

196
arteritis 114-118

infected aneurysms 175

intimal fibroplasia 74, 76

microbial 175
artery luminal diameter 58
artery wall, intravascular ultrasound 405,

406
artifacts, ultrasound 321, 322, 401
ascites

portal hypertension 278, 286

transjugular intrahepatic portosystemic
shunt 285
Aselli, G. 207
aspiration, cyst, adventitial cystic disease

123
aSpire covered stent 535, 536, 537
aspirin

atherosclerosis 446, 447

balloon angioplasty 504

carotid artery disease 576-577

CAS 578

effect on platelet aggregation 36

heparin-induced thrombosis 36

intimal hyperplasia 139
Asymptomatic Carotid Artery Stenosis

(ACAS) Study 255, 576
atherectomy 499

directional 499, 512

rotational 512
atheroablation 499

devices 499
atheroembolic disease 496

catheter-based approaches see catheter-
based approaches
atherogenesis 56
atheroma 495, 496

primary 496

treatment 498-499

see also atherosclerotic plaques
atherosclerosis

aneurysmal disease 166

carotid circulation 227-228

cellular signaling 46-48

cholesterol esterase transfer protein
(CETi-1 vaccine) 449

coronary arteries 228-229

endothelial permeability 49-50

immune mechanisms 52

lesion localization 55-65

limb ischemia 533

lipoprotein processing 48-49

lower limb 225-227

medical management 441-453

modifiable risk factor treatment 446

molecular aspects 43-54

obesity 444-445

plaque localization 59-60

plaque regression 52-53

renal artery 508

response to injury hypothesis 43-46

risk factors 441-445
irreversible 442
reversible 442-445

secondary prevention 445-449

shear effect 46

superficial femoral artery 62

susceptible regions 60-62

treatment strategies 53
atherosclerotic plaques 495

calcification 58

heart rate 61

morphology 59

necrotic core 57

pathogenesis 495

progression 61

ulceration 495
ATP depletion, intestinal ischemia 220
atrial naturetic factor (ANF), renal

autoregulation 187
attenuation, ultrasound 299
Auth Rotablator 498, 499
autoantibodies, oxidized LDL 52
autogenous vein grafting

adventitial cystic disease 123

entrapment syndromes 132
autologous leukocyte imaging 326
autologous platelets 326
automatic Pulse Spray Pump 591
autonomic nervous system, renal

autoregulation 186-187
autoregulation

cerebral 251-252

renal 186-187
autoregulatory escape 218
axial resolution 315, 316
axillobrachial approach, angiography

387
axillofemoral bypass grafts

aortoiliac disease 544

infection of 484
axillosubclavian vein thrombosis 464
aztreonam 486

bacteremia, graft infection 480
bacteria

antibiotic protein synthesis inhibition
487

antibiotic resistance 487-488

contamination 479

graft adherence 481-482
bacteriology

graft infection 481, 481

infected aneurysms 175
Baker cyst 121
balloon angioplasty 409-410, 503-515

aortoiliac 506-507
outcomes 507

aortoiliac disease 548

arch vessel lesions 568

atherectomy 512

bypass grafts 512

carotid artery 511, 511-512

611

Index

catheters 506, 506

celiac artery 509-510

complications 513, 513-514

equipment choice 505, 505-506, 506

femoral artery 513

femoropopliteal 507-508
outcome 507-508

general principles 503-504

guidewires 505, 505

hemodialysis access 512

iliac 507 "

indications 503

infrapopliteal vessels 538

infrarenal aorta 507

intervention method 506

intracranial 511-512

laser as alternative 512-513

mechanism 503

mesenteric intervention 508-510

mortality 513

procedural success 503-504

renal artery 508-510

sheaths 504,505

subclavian artery 510, 510-511

technique 504-506

tibioperoneal 508

vascular access 504, 506
see also vascular access

vertebral arteries 511-512
balloon dilation

of atheroma 498, 498-499

see also balloon angioplasty
balloon embolectomy 497
balloon-expandable stents 516, 517, 517
balloon occlusion techniques 580
balloon thrombectomy devices 594
Banti's syndrome 283
Barger, G. 101
basal lamina 55

basic fibroblast growth factor (bFGF) 136
Bayliss principle 217
beam steering 316
Bernoulli's principle 295-296
beta-adrenoceptors 81
beta-blockers

aneurysm growth rate 165

Raynaud's syndrome 81
beta-lactam antibiotics 485-486
beta-lactamases 487-488

inhibitors 487-488
bifurcated endografts 520
bifurcated stent-graft evaluation 361,362
bilateral disease

popliteal entrapment syndromes 130

thoracic outlet syndromes 158
bile acid sequestrants 442, 473
biopsy

liver 279

Takayasu's arteritis 116
bleeding

disorders/abnormalities 38

hemostasis 38-39

recurrent 287

thrombolysis complications 587

thrombolytic infusions 603

variceal 284, 287, 287
blood displacement, angioscopy 425
blood flow measurements

Doppler ultrasound 298-300

vasospastic disorders 85
blood flow velocity, critical arterial
stenosis 298, 299

blood pressure measurements

carotid endarterectomy 577

direct 303-304

indirect 303

transradial monitoring 573
blue rubber bleb syndrome 15
B-mode imaging 306, 320
Bohr technique 202
bone hypertrophy, arteriovenous

malformations 16
bone scans, sympathetically maintained

pain 238
bone segmentation 366
boulevard look 589
boundary layer 296

separation 296, 297
brachial approach, retrograde 569
brachial artery, arterial access 548
brachial-femoral wire configuration 569,

569
brachial plexus compression, thoracic
outlet syndrome 148-149, 155
brachiocephalic arteries

arteritis 115

reconstruction 567-574

stenosis 568
brachiocephalic vascular disease

patient selection 567-568

procedural techniques 568-573

treatment 567-573
brachytherapy, intimal hyperplasia

140
bradycardia 578-579
bradykinin

hemostasis 27

renovascular hypertension 188
Brescia technique 512
Bright, R. 180-191
bromocriptine 103, 105
bronchoconstriction, reflex 335
Brook and Eagle method 154
Budd-Chiari syndrome 276-277, 285
Buerger, L. 92
Buerger's color 95
Buerger's disease 92-100

angiography 96-97, 392-393

arteriosclerosis obliterans vs. 96

bypass grafts 98

differential diagnosis 96, 96

epidemiology 92

etiology 92-93

pathology 93-94, 94

presentation 95-96

treatment 97-98

veins 94
Bunt classification 477, 478
bypass grafts

angioplasty 512

angioscopy 432,432-437

aortoiliofemoral 480

axillofemoral 484, 544

Buerger's disease 98

catheter-based techniques 501

caval obstruction 596

femoropopliteal 538

infrapopliteal vessels 538

IVC occlusion 598

platelet sequestration 38

C7 transverse process 148
cafergot see ergotamine tartrate

calcitonin gene-related peptide (CGRP)

Raynaud's syndrome 81

sympathetic dysfunction 237
calcium, intracellular levels/signaling

pathways 46-47
calcium channel blockers

atherosclerosis treatment 53

ergotism 110-111

intimal hyperplasia 139

vasodilatation 469
calf claudication, adventitial cystic disease

122
calf muscle pump 198

venous stasis disease 199
capillary blood flow, compartment

syndrome 241
capillary derecruitment 219
capillary hemangiomas 16
capillary loops, venous disease 201
carbapenems 486
carbon dioxide

angiography contrast 389

angioscopy 425
carbon dioxide tension (P a co 2 ), cerebral

blood flow 252
carbon monoxidemia 443
carboxyhemoglobin levels, Buerger's

disease 92-93
cardiac risk assessment 338-341, 339,
339

preoperative 339, 339-340
cardinal vein derivatives,

embryology/development 6-8
cardioscope 423-424
carnitine 471

carotid angioplasty and stenting (CAS)
255, 577-580

endovascular technique 579-580

history 577-578

rationale for 580-582

technique 578-580

trials 579
carotid arterial disease

assessment 358, 363

computed tomography angiography
358, 363

high cervical stenosis 582

left common and occlusive lesions 569

medical management 576-577

spectral analysis 301, 301

stents 518

tandem lesions 582
carotid artery angioplasty 510, 511-512
carotid artery anomalies 8
carotid artery bifurcation

atheromatous plaque 254

atherosclerosis 60, 60-61

flow field 59

MRI 377, 377
carotid artery stenosis

asymptomatic 576

balloon angioplasty 511-512

contralateral occlusion 582

digital subtraction angiography 511

intraluminal thrombus 582

MRI 377

radiation induced 582

symptomatic 575-576
Carotid Artery Stenosis with

Asymptomatic Narrowing:
Operation Versus Aspirin
(CASANOVA) trial 576

612

Index

carotid-cavernous fistula 395
carotid-cerebral system, atherosclerosis 59
carotid circulation, atherosclerosis 227-228
carotid duplex ultrasonography 576
carotid endarterectomy (CEA)

blood pressure measurements 577

CAS vs. 575-586

cerebral ischemia 255-256

comorbidity 581

coronary artery bypass graft procedure
581

elderly patients 581

indications 575-576

limitations of clinical trial results 581

neurovascular complications 582

radionuclide scanning 328

restenosis 582

surgical technique 577

technique 576
carotid puncture arteriography 387
Carotid Revascularization vs. Stent Trial

(CREST) 255-256
carotid siphon 254, 376
carotid-subclavian bypass 557, 567
carotid system TIAs 255
carotid vessels, MRI 376-378
CASANOVA (Carotid Artery Stenosis with
Asymptomatic Narrowing:
Operation Versus Aspirin) trial 576
catalase 481
catheter(s)

adherent clot 497-498, 498

Ahn Thrombectomy 500-501

angioplasty balloon 498

balloon 506, 506

for balloon angioplasty 504

Directional Atherectomy 499

Fino Thrombectomy 500, 500

fragility 403

graft thrombectomy 501, 501

Greenfield Transvenous Pulmonary
Embolectomy 500

guiding, for balloon angioplasty 504

intravascular ultrasound 403-406
configurations 402-403
removal rate 407

Simpson directional atherectomy 498

TEC system (Transluminal Extraction
Catheter) 498, 499

thrombolytic infusion techniques
590-592

Thru-Lumen 497, 497

Trellis peripheral infusion 500
catheter-based approaches, atheroembolic
disease 495-502

acute thrombus removal 497

atherectomy 499

bypass grafts 501
synthetic 501

in-situ thrombus removal 497-498

percutaneous thrombectomy 499-501
see also thrombectomy

procedures 497-499

sequence of use of instruments 496

underlying atheroma treatment 498-499
catheter-directed thrombolysis

acute limb ischemia 458-461, 459, 460

axillosubclavian vein thrombosis 464

flow-directed infusion 591-592

jugular vein thrombosis 595

May-Thurner syndrome 601

thrombectomy 594

venous recanalization 590-591, 605

see also thrombolytic therapy
causalgia 233
caval clip 599
cavernosal arteries 270
cavernosal artery occlusion pressure

(CAOP) 272
cavernosal nerves 270
cavernosography 272
cavernosometry 272
CD36 50
celiac artery

anatomy 215

angioplasty 509-510

anomalies 10, 10
cell culture, smooth muscle cells 24
cellular signaling, atherosclerosis 46-48
cell wall

structure 485-486,486

synthesis inhibition 485-487
central biasing mechanism 235
central nervous system, angiotensin

188
centrifugal theory 8
centripetal theory 8

cephalosporins, molecular structure 486
cerebral aneurysms 358
cerebral angiography 395
cerebral autoregulation 251-252
cerebral blood flow 251-252, 328
cerebral blood volume 328
cerebral cortex, electrical failure threshold

253
cerebral emboli 254
cerebral hemodynamics 328, 329
cerebral hypoperfusion 255
cerebral ischemia 251-256

carotid endarterectomy 255-256

etiology 253

thresholds 253
cerebral metabolism 251
cerebral oxygen metabolism (rCMR0 2 )

328
cerebral perfusion pressure, compensatory

responses 329
cerebral perfusion reserve 329-330, 330
cerebral viability, radionuclide scanning

329
cerebral vulnerability, radionuclide

scanning 329
cerebrospinal fluid (CSF)

drainage 257-258

spinal cord arterial perfusion 261

spinal cord ischemia 257-258
cerebrovascular disease, radionuclide

scanning 328-331
cervical band compression 146, 595, 596
cervical carotid stenosis 577
cervical rib

clinical correlation 149-150

development 147

thoracic outlet syndrome 148-150

transaxillary view 149
C fibers 233

Chambardel-Dubreuil, L. 126
Child's classification, portal hypertension

279
Chlamydia 445,449

pneumoniae 52
cholesterol emboli 390
cholesterol esterase transfer protein
(CETi-1) vaccine 449

cholesterol levels

carotid artery disease 577

screening 442

see also hypercholesteremia
cholestyramine 473
cholinergic nerves, erectile dysfunction

269
Christmas disease (hemophilia B) 38
chromosome 15 172
chronic mesenteric ischemia (CMI) 215,

216
chronic venous insufficiency (CVI)

definition 192

venous stasis disease 198
chyle 213
chylorrhea 212
chylothorax 212-213
chylous ascites 212
chylous disorders, pathophysiology

212-214
chylous effusion 213-214
chyluria 212

cigarette smoking see smoking
cilostazol 448
cineangiography 505
cinnarizine 469
circle of Willis 251, 252

computed tomography angiography
363

MRI 376, 376
circumaortic renal collar 13-14, 14
cisterna chyli 8, 208

Clatworthy shunt (mesacaval shunts) 282
claudication

calf in adventitial cystic disease 122

intermittent see intermittent claudication

medical management 448
Claviceps purpura 101
clindamycin 487
clinical studies

AneuRx stent-graft 524

Raynaud's phenomenon 80-82
clopidogrel

atherosclerosis 447

balloon angioplasty 504

CAS 578
closure assistance devices

aortoiliac disease 548-549

CAS 580
"Clot Buster" 501
clots 192

acute, formation 495-496

white 34

see also thrombus
clotting factors see coagulation factors
coagulase 481
coagulation 27-42

bleeding disorders 38

deficiencies
see also specific conditions

diabetes mellitus 443-444

endothelium 20-21

normal mechanisms 27-32

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Vascular-Surgery-Books_1aim.net"

See other formats