Emergency Vascular and

Endovascular Surgical
Practice
This page intentionally left blank
Emergency Vascular and
Endovascular Surgical Practice
Second Edition
Edited by
Aires A B Barros D’Sa OBE MD FRCS FRSCEd
Honorary Professor of Vascular Surgery, The Queen’s University of Belfast
Consultant Vascular Surgeon, Regional Vascular Surgery Unit
Royal Victoria Hospital, Belfast, UK
and

Anthony D B Chant BA BSc MS FRCS

Medical Director, British Vascular Foundation
London, UK

Hodder Arnold
A MEMBER OF THE HODDER HEADLINE GROUP
First published in Great Britain in 2005 by
Hodder Education, a member of the Hodder Headline Group,
338 Euston Road, London NW1 3BH

http://www.hoddereducation.com

Distributed in the United States of America by

Oxford University Press Inc.,
198 Madison Avenue, New York, NY10016
Oxford is a registered trademark of Oxford University Press

© 2005 Edward Arnold (Publishers) Ltd

All rights reserved. Apart from any use permitted under UK copyright law,
this publication may only be reproduced, stored or transmitted, in any form,
or by any means with prior permission in writing of the publishers or in the
case of reprographic production in accordance with the terms of licences
issued by the Copyright Licensing Agency. In the United Kingdom such
licences are issued by the Copyright Licensing Agency: 90 Tottenham Court
Road, London W1T 4LP.

Whilst the advice and information in this book are believed to be true and
accurate at the date of going to press, neither the author[s] nor the publisher
can accept any legal responsibility or liability for any errors or omissions
that may be made. In particular, (but without limiting the generality of the
preceding disclaimer) every effort has been made to check drug dosages;
however it is still possible that errors have been missed. Furthermore,
dosage schedules are constantly being revised and new side effects
recognised. For these reasons the reader is strongly urged to consult the
drug companies’ printed instructions before administering any of the drugs
recommended in this book.

British Library Cataloguing in Publication Data

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

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

ISBN-10: 0 340 81012 2

ISBN-13: 978 0 340 81012 5

1 2 3 4 5 6 7 8 9 10

Commissioning Editor: Sarah Burrows

Project Editor: Naomi Wilkinson
Production Controller: Joanna Walker
Cover Design: Sarah Rees
Indexer: Indexing Specialists (UK) Ltd.
Website: www.indexing.co.uk

Typeset in 10/12 Minion by Charon Tec Pvt. Ltd, Chennai, India

www.charontec.com
Printed and bound in the UK by CPI Bath

What do you think about this book? Or any other Hodder Arnold
title? Please visit our website at www.hoddereducation.com
 To
Elizabeth, Vivienne, Lisa, Miranda and Angelina Barros D’Sa
and
Ann, Ben, Harvey and Thomas Chant
This page intentionally left blank
Contents

Contributors xi

Foreword xvii

Foreword to 1st edition xix

Preface xxi

Preface to 1st edition xxiii

Section 1 General Considerations 1

1A Emergency vascular services in the UK 3

Anthony DB Chant
1B Emergency vascular services in Denmark 7
Henrik Sillesen
1C Emergency vascular services in the USA 12
Vincent L Rowe, Juan A Asensio, Fred A Weaver
2 Pathophysiology of acute vascular insufficiency 17
Aires AB Barros D’Sa, Denis W Harkin
3 Intestinal ischaemia in aortic surgery 29
Martin G Björck
4 Ischaemia-reperfusion injury, SIRS and MODS 39
Denis W Harkin, Aires AB Barros D’Sa
5 Pathophysiology of stroke 55
Harvey J Chant
6 Assessing the risk in vascular emergencies 65
Peter R Taylor, David J Gerrard
7A Perioperative care in emergency vascular practice 73
Alasdair Dow, John F Thompson
7B Emergency vascular access 85
Thang D Nguyen, Roy M Fujitani, Samuel E Wilson
8 Imaging for vascular emergencies 95
Peter K Ellis, Barry Kelly, Peter Kennedy, Paul HB Blair
9A Outcomes of emergency vascular procedures. A view from the British Isles 105
Jonothan J Earnshaw
9B Outcomes of emergency vascular procedures in Scandinavia 111
William P Paaske, Mauri Lepäntalo, Thomas Troëng
10 Medico-legal aspects of emergency vascular care in the UK 121
Bruce Campbell

Section 2 Acute Cerebrovascular Syndromes 127

11 The developing stroke 129

Sebastián F Ameriso
viii Contents

12 Role of the vascular surgeon in managing stroke 141

John P Royle, Geoffrey A Donnan, Brian Chambers
13 Surgical experience in evolving stroke arising from the carotid 149
R Huber, RJ Seitz, M Siebler, A Aulich, Wilhelm Sandmann
14 Post-carotid endarterectomy stroke 155
David Rosenthal, Eric D Wellons

Section 3 Acute Lower Limb Ischaemic States 161

15 Acute limb ischaemia: surgical options 163

Clifford P Shearman, Malcolm H Simms
16 Acute limb ischaemia: endovascular options 181
Bernard H Nachbur, Iris Baumgartner, Do D Do, Felix Mahler, Hans-Beat Ris
17 Graft maintenance and graft failure 197
David K Beattie, Alun H Davies
18 Acute ischaemia secondary to occult prosthetic graft infection 211
Linda M Reilly
19 The acute diabetic foot 223
Cameron M Akbari, Frank W LoGerfo

Section 4 The Acutely Swollen Limb 235

20 Deep vein thrombosis 237

Bo GH Eklof, Curtis B Kamida
21 Pulmonary embolism 251
Anh Nguyen, Elaine Imoto, Bo GH Eklof
22 Upper limb vein thrombosis 261
Bengt LT Lindblad, Krassi Ivancev, Simon G Darke

Section 5 Thoracoabdominal Catastrophes 275

23 The emergency aortic aneurysm 277

Jonathan Refson, John HN Wolfe
24 Stenting in acute aortic dissection 287
Michael D Dake
25 Prosthetic aortic graft infection 297
Charles W Bouch, James M Seeger
26 The acutely compromised renal artery 309
Michael P Jenkins, George Hamilton
27 Renal artery aneurysms 315
James C Stanley, Peter K Henke
28 Visceral artery aneurysms 325
Sandra C Carr, William D Turnipseed
29 Mesenteric ischaemia 335
Daryll M Baker, Janice Tsui, Averil O Mansfield

Section 6 Acute Complications of Endovascular Aortic Repair (EVAR) 343

30 Endoleak complicating EVAR 345

Robert J Hinchliffe, Brian R Hopkinson
31 Graft breakdown and migration complicating EVAR 357
S Rao Vallabhaneni, Peter L Harris
32 Atheroembolism complicating EVAR 367
Juan C Parodi, Luis M Ferreira
 Contents ix

Section 7 Regional Vascular Trauma 373

33 Vascular injuries of the limbs 375

Aires AB Barros D’Sa, John M Hood, Paul HB Blair
34 Vascular injuries of the chest 401
Alastair NJ Graham, Kieran G McManus, James A McGuigan
35 Vascular injuries of the neck 419
S Ram Kumar, Fred A Weaver
36 Abdominal vascular injuries 429
John V Robbs
37 Limb replantation 443
Colin M Morrison, Michael D Brennen
Section 8 Iatrogenic Injuries 457

38 Injuries of arterial catheterisation 459

David Bergqvist, Christer Ljungman
39 Injuries of peripheral endovascular procedures 469
Amman Bolia, Peter RF Bell
40 Specialty related iatrogenic vascular injuries 487
Lars Norgren

Section 9 Special Acute Vascular Challenges 493

41 Acute upper limb ischaemic states 495

Kenneth A Myers, Gregory W Self
42 Emergency aspects of Buerger’s disease 511
Sekar Natarajan, Dhanesh Kamerkar
43 Acute limb vascular inflammatory conditions 523
Matthew Waltham, Kevin G Burnand
44 Vascular emergencies caused by substance abuse 531
Ronald A Kline, Ramon Berguer
45 HIV/AIDS related vascular emergencies 545
Jacobus van Marle, Lynne Tudhope
46 Portal hypertension and variceal bleeding 555
Rowan W Parks, Thomas Diamond
47 Cold injury 573
Per-Ola Granberg

Index 591
This page intentionally left blank
Contributors

Cameron M Akbari MD MBA FACS Iris Baumgartner MD

Senior Vascular Surgeon Professor, Division of Angiology
Director Vascular Diagnostic Laboratory Department of Internal Medicine
Washington Hospital Center University of Berne
Washington, DC Berne
USA Switzerland

Sebastián F Ameriso MD
David K Beattie FRCS
Professor of Neurology
Consultant Surgeon
Austral University School of Medicine
Department of Surgery
Chief, Vascular Neurology Division
Charing Cross Hospital
Fundación para la Lucha contra las Enfermedades Neurológicas de
London
la Infancia (FLENI)
UK
Buenos Aires
Argentina
Peter RF Bell KBE MD FRCS
Juan A Asensio MD FACS FCCM Professor of Surgery
Professor and Chief Department of Surgery
Division of Clinical Research in Trauma Surgery University of Leicester
Trauma Surgery and Surgical Critical Care Leicester Royal Infirmary
University of Medicine and Dentistry of New Jersey Leicester
The University Hospital UK
New Jersey
USA David Bergqvist MD PhD FRCS
Professor of Vascular Surgery
A Aulich MD Department of Surgery
Institute of Diagnostic Radiology University Hospital
Heinrich Heine University Uppsala
Duesseldorf Sweden
Germany

Daryll M Baker FRCS Ramon Berguer MD PhD

Department of Vascular Surgery Chief, Division of Vascular Surgery
Royal Free and St Mary’s Hospital Harper University Hospital
London Detroit Medical Center and Wayne State University
UK Detroit, MI
USA
Aires AB Barros D’Sa OBE MD FRCS FRCSEd
Honorary Professor of Vascular Surgery Martin Björck MD PhD
The Queen’s University of Belfast, and Associate Professor
Consultant Vascular Surgeon Consultant Vascular Surgeon
Regional Vascular Surgery Unit Department of Surgery
Royal Victoria Hospital University Hospital
Belfast Uppsala
UK Sweden
xii Contributors

Paul HB Blair MD FRCS Harvey J Chant BSc MD FRCS

Consultant Vascular Surgeon and Director of Trauma Consultant Vascular Surgeon
Vascular Surgery Unit Peninsular Medical School
Royal Victoria Hospital Royal Cornwall Hospital
Belfast Truro
UK UK

Amman Bolia MBChB FRCR Michael D Dake MD

Consultant Vascular Radiologist Associate Professor of Radiology and Medicine
Department of Vascular Radiology Chief, Division of Interventional Radiology
Leicester Royal Infirmary Stanford University School of Medicine
Leicester Stanford, CA
UK USA
Charles W Bouch MD Simon G Darke MS FRCS
Vascular Surgery Fellow Consultant Vascular Surgeon
Division of Vascular Surgery Royal Bournemouth Hospital
University of Florida College of Medicine Bournemouth
Gainesville, FL UK
USA
Alun H Davies MA DM FRCS
Michael D Brennen FRCS
Reader and Consultant Surgeon
Consultant Plastic Surgeon
Department of Surgery
Department of Plastic Surgery
Charing Cross Hospital
Ulster Hospital
London
Belfast
UK
UK

Kevin G Burnand ChM FRCS Thomas Diamond BSc MD FRCS

Professor of Vascular Surgery Consultant Surgeon and Honorary Senior Lecturer
Department of Surgery Mater Hospital
St Thomas’s Hospital Belfast
London UK
UK
Do Dai Do MD
Bruce Campbell MS FRCP FRCS Professor, Division of Angiology
Professor and Consultant Surgeon Department of Internal Medicine
Royal Devon and Exeter Hospital University of Berne
Exeter Berne
UK Switzerland

Sandra C Carr MD Geoffrey A Donnan MD FRACP

Assistant Professor Director, National Stroke Research Institute
Division of Vascular Surgery Professor of Neurology, University of Melbourne
University of Wisonsin Clinical Sciences Director of Neurosciences, Austin and Repatriation Medical Centre
Madison, WI Heidelberg, Victoria
USA Australia

Brian Chambers Alasdair Dow FRCA

Associate Director and Head of Ultrasound Research Consultant Anaesthetist
National Stroke Research Institute Royal Devon and Exeter Hospital
Austin Hospital Exeter
Heidelberg, Victoria UK
Australia
Jonothan J Earnshaw DM FRCS
Anthony DB Chant BA BSc MS FRCS Consultant Surgeon
Medical Director, British Vascular Foundation Gloucestershire Royal Hospital
London Gloucester
UK UK
 Contributors xiii

Bo GH Eklof MD PhD Denis W Harkin MD FRCS

Clinical Professor Consultant Vascular Surgeon
John A Burns School of Medicine Regional Vascular Surgery Unit
University of Hawaii Royal Victoria Hospital
Chief, Vascular Center Belfast
Straub Clinic and Hospital UK
Honolulu, HI Peter L Harris MD FRCS
USA Director and Consultant Vascular Surgeon
Regional Vascular Unit
Peter K Ellis FRCR Royal Liverpool University Hospital
Consultant Radiologist Liverpool
Department of Radiology UK
Royal Victoria Hospital
Belfast Peter K Henke MD
UK Associate Professor of Surgery
University of Michigan Medical Center
Luis M Ferreira MD Ann Arbor, MI
Staff, Vascular Surgery Department USA
Fundación para la Lucha contra las Enfermedades Neurológicas de
Robert J Hinchliffe MB FRCS
la Infancia (FLENI)
Department of Vascular and Endovascular Surgery
Buenos Aires
University Hospital
Argentina
Nottingham
UK
Roy M Fujitani MD FACS
Associate Professor of Surgery
John M Hood MPhil FRCS
Chief, Division of Vascular Surgery
Consultant Vascular Surgeon
UCI Medical Center
Regional Vascular Surgery Unit
Orange, CA
Royal Victoria Hospital
USA
Belfast
UK
David J Gerrard FRCS
Specialist Registrar Brian R Hopkinson MB ChM FRCS
Department of Surgery Emeritus Professor of Vascular Surgery
Guy’s & St Thomas’ Hospital Trust Department of Surgery
London University Hospital
UK Nottingham
UK
Alastair Graham MD FRCS
Consultant Cardiothoracic Surgeon R Huber MD
Department of Thoracic Surgery Department of Vascular Surgery and Kidney Transplantation
Royal Victoria Hospital Heinrich Heine University
Belfast Duesseldorf
UK Germany

Per-Ola Granberg MD PhD Elaine Imoto MD

Emeritus Professor of Endocrine Surgery Department of Chest Diseases
Karolinska Hospital Straub Clinic and Hospital
Stockholm Honolulu, HI
Sweden USA

George Hamilton MD FRCS Krassi Ivancev MD PhD

Professor of Vascular Surgery Professor and Chief, Endovascular Centre
University Department of Surgery Department of Radiology
Royal Free Hospital Malmö University Hospital
London Malmö
UK Sweden
xiv Contributors

Michael P Jenkins BSc MS FRCS Bengt LT Lindblad MD PhD

Consultant Vascular Surgeon Associate Professor
Regional Vascular Unit Department of Vascular Diseases
St Mary’s and Chelsea & Westminster Hospitals University Hospital Malmö
London Malmö
UK Sweden

Dhanesh Kamerkar MB MS(Surgery) Christer Ljungman MD PhD

Consultant Vascular Surgeon Assistant Professor of Vascular Surgery
Ruby Hall Clinic, and Department of Surgery
Associate Honorary Vascular Surgeon University Hospital
Department of Surgery Uppsala
KEM Hospital Sweden
Pune Frank W LoGerfo MD FACS
India Professor and Chief, Division of Vascular Surgery
Beth Israel Deaconess Medical Center
Curtis B Kamida MD
Harvard Medical School
Clinical Assistant Professor
Boston, MA
John A Burns School of Medicine
USA
University of Hawaii
Department of Radiology Felix Mahler
Straub Clinic and Hospital Professor and Chief, Division of Angiology
Honolulu, HI Department of Internal Medicine
USA University of Berne
Berne
Barry Kelly FRCR Switzerland
Consultant Radiologist
Department of Radiology Averil O Mansfield CBE ChM FRCS
Royal Victoria Hospital Former Professor of Surgery
Belfast Imperial College of Science, Technology and Medicine
UK St Mary’s Hospital
London
Peter Kennedy FRCR UK
Consultant Radiologist
James A McGuigan FRCS
Department of Radiology
Consultant Thoracic Surgeon
Royal Victoria Hospital
Department of Thoracic Surgery
Belfast
Royal Victoria Hospital
UK
Belfast
Ronald A Kline MD FACS FAHA UK
Arizona Endovascular Center Kieran G McManus BMedSc FRCS
Medical Director, St Joseph Wound Care Center Consultant Thoracic Surgeon
Tucson, AZ Department of Thoracic Surgery
USA Royal Victoria Hospital
Belfast
S Ram Kumar MD
UK
Resident, Division of Vascular Surgery
Kech School of Medicine Colin M Morrison FRCS
University of Southern California Department of Plastic Surgery
Los Angeles, CA Ulster Hospital
USA Belfast
UK
Mauri Lepäntalo MD PhD
Professor of Vascular Surgery, Helsinki University Kenneth A Myers MS FRACS FACS
Chief, Department of Vascular Surgery Consultant Surgeon
Helsinki University Central Hospital Monash Medical Centre and Epworth Hospital
Helsinki Melbourne, Victoria
Finland Australia
 Contributors xv

Bernard H Nachbur MD FMH Jonathon Refson MS FRCS

Emeritus Professor of Surgery Consultant Surgeon
University of Berne Princess Alexandra Hospital
Berne Harlow
Switzerland UK
Sekar Natarajan MB MS MCh(Vascular) FICS Linda M Reilly MD
Professor and Senior Vascular Surgeon Professor, Department of Surgery
Department of Vascular Surgery University of California San Francisco
Madras Medical College and Research Centre San Francisco, CA
Chennai USA
India
Hans-Beat Ris MD PhD
Anh Nguyen MD Professor and Director
Surgical Resident, John A Burns School of Medicine Department of Thoracic Surgery
University of Hawaii University of Lausanne
Straub Clinic and Hospital Lausanne
Honolulu, HI Switzerland
USA
John V Robbs ChM(CT) FRCS(Ed) FRCPS(Glas) FCS(SA)
Thang D Nguyen MD Professor and Head, Division of Surgery
Department of General Surgery Head, Metropolitan Vascular Services and
University of California, Irvine School of Clinical Sciences
Irvine, CA Nelson R Mandela School of Medicine
USA Durban
South Africa
Lars Norgren MD FRCS
Professor of Surgery David Rosenthal MD
Chairman, Department of Surgery Professor of Surgery, Medical College Georgia
– Chief of Vascular Surgery, Atlanta Medical Center
Orebro University Hospital
– Atlanta, GA
Orebro
Sweden USA

William Paaske MD DrMedSci FRCS FRCSEd FACS Vincent L Rowe MD

Professor of Vascular Surgery, Chief Vascular Surgeon Division of Vascular Surgery
Department of Cardiothoracic and Vascular Surgery T Department of Surgery
Skejby Hospital Keck School of Medicine
Aarhus University Hospital, Skejby Sygehus University of Southern California
Aarhus Los Angeles, CA
Denmark USA

John P Royle FRACS FRCSEd

Rowan W Parks MD FRCSI FRCSEd
Former Director, Vascular Surgery Unit
Senior Lecturer in Surgery and Honorary Consultant Surgeon
Austin & Repatriation Medical Center
Department of Clinical & Surgical Sciences
Heidelberg, Victoria
University of Edinburgh
Australia
Royal Infirmary
Edinburgh Wilhelm Sandmann MD
UK Professor and Chief
Department of Vascular Surgery and Kidney Transplantation
Juan C Parodi MD
University of Duesseldorf
Chief of the Department of Angiology and Vascular Surgery
Duesseldorf
Fundación para la Lucha contra las Enfermedades
Germany
Neurológicas de la Infancia (FLENI)
Buenos Aires James M Seeger MD
Argentina; and Vascular Surgeon
Professor of Surgery and Radiology Professor and Chief, Division of Vascular Surgery
Washington University School of Medicine University of Florida College of Medicine
Saint Louis, MO Gainsville, FL
USA USA
xvi Contributors

RJ Seitz MD J Tsui MA MRCS

Department of Neurology Vascular Research Fellow
Heinrich Heine University Royal Free Hospital
Duesseldorf London
Germany Lynne Tudhope MBChB Mmed
Gregory W Self MB FRACS Consultant Vascular Surgeon
Consultant Vascular Surgeon Unit for Peripheral Vascular Surgery
Monash Medical Centre and Epworth Hospital Pretoria Academic Hospital
Melbourne, Victoria University of Pretoria
Australia South Africa
William D Turnipseed MD
Clifford P Shearman MS FRCS Division of Vascular Surgery
Professor of Vascular Surgery University of Wisconsin Clinical Sciences
University of Southampton Madison, WI
Southampton General Hospital USA
Southampton
S Rao Vallabhaneni FRCS
UK
Endovascular Research Fellow
M Siebler Regional Vascular Unit
Department of Neurology The Royal Liverpool University Hospital
Heinrich Heine University Liverpool
Duesseldorf UK
Germany Jacobus van Marle MBChB Mmed FCS(SA)
Henrik Sillesen MD DMSc Consultant Vascular Surgeon
Chairman, Department of Vascular Surgery Unit for Peripheral Vascular Surgery
Gentofte University Hospital Pretoria Academic Hospital
Hellerup University of Pretoria
Denmark South Africa
Matthew Waltham MA FRCS
Malcolm H Simms MB FRCS Surgical Lecturer
Consultant Vascular Surgeon Guy’s, King’s and St Thomas’ School of Medicine
Department of Vascular Surgery London
University Hospital UK
Birmingham
Fred A Weaver MD
UK
Professor of Surgery
James C Stanley MD Division of Vascular Surgery
Professor of Surgery Department of Surgery
University of Michigan Medical Center Keck School of Medicine
Ann Arbor, MI University of Southern California
USA Los Angeles, CA
USA
Peter R Taylor MA MChir FRCS
Consultant Vascular Surgeon Eric D Wellons MD
London Bridge Hospital Department of Vascular Surgery
London Atlanta Medical Centre
UK Atlanta, GA
USA
John F Thompson MS FRCSEd FRCS Samuel E Wilson MD
Consultant Surgeon Division of Vascular Surgery
Royal Devon & Exeter Hospital UCI Medical Center
Exeter Orange, CA
UK USA
Thomas J Troëng MD PhD John HN Wolfe FRCS
Associate Professor Consultant Vascular Surgeon
Department of Surgery Regional Vascular Unit
Blekinge Hospital St Mary’s Hospital
Karlskrona London
Sweden UK
Foreword

War and strife have long since set the scene for advances in regional vascular surgical centres established in the British
the urgent care of the wounded. Galen, when serving as sur- Isles, his use of early perfusion operative techniques enabled
geon to the school of gladiators at Pergamon, near Troy, more accurate and reliable vascular repairs, and better end
cured a traumatic brachial aneurysm by the simple expedi- results have been the dividend. The second edition of this
ent of prompt and securely maintained local compression. book sets new standards for our specialty.
Ambroise Paré and Baron Larrey both recognised the Endovascular surgery is here to stay, though, predictably,
opportunity of the ‘golden hour’ after injury. Writing from along with its own problems. The potential indications and
the battle zone, Norman Rich documented the memorable techniques for such intervention seem to proliferate. This
achievements of the mobile forward vascular teams in new edition gives authoritative help with these. Today’s
Vietnam, their bounty being that for hundreds of young professional emphasis on risk assessment, the manda-
men the otherwise certain prospect of limb loss was avoided. tory measurement of treatment outcomes and the growing
The emergence of terrorism and violence in Northern awareness of medico-legal consequences and their avoid-
Ireland provided new opportunities for acute, definitive ance are given consideration. New material on stroke
repair of life-threatening vascular wounds. The challenges makes a welcome appearance, as does a rearranged section
faced were quickly appreciated at the Royal Victoria on the acutely swollen limb. The greatly expanded author-
Hospital, Belfast, where Aires Barros D’Sa, heading the next ship, with increased international input, adds further merit
generation of modern vascular surgeons, applied novel and to this distinguished book.
effective strategies, particularly in the testing scenario of
complex limb vascular injuries. In one of the first dedicated H H G Eastcott
This page intentionally left blank
Foreword to the 1st edition

Medical and surgical emergencies involving the vascular variceal bleeding have a place in this book. In tackling vas-
system continue to challenge clinicians and vascular sur- cular injuries, a regional focus is conveniently provided.
geons worldwide. Life as well as limb can be threatened by Iatrogenic vascular injuries as well as limb replantation
a wide variety of diseases or injuries of arteries and veins. which fall within the realm of emergency vascular practice
Emerging new technologies have been associated with both all receive coverage.
diagnostic and therapeutic complications. Some of the subject matter contained in this most wel-
The editors, also contributors to the volume, are leading come book may be found scattered piecemeal in large vol-
vascular surgeons with an international reputation in cur- umes on vascular surgery or may be omitted altogether.
rent vascular surgical practice and research. In this book During this century there have been few efforts to provide
they have mobilized the aggregated experience of other dis- within one volume the comprehensive range of acute vas-
tinguished contributors from Europe, Australia and the cular material which has been collected here. The book con-
United States. They have addressed numerous emergency tains a unique and valued reference for doctors of all
clinical situations which require a mandated emergency disciplines engaged in the management of vascular emer-
response by those with an interest in and an expertise for gencies. The contributions of internationally renowned and
managing vascular disease and injury. experienced physicians and surgeons will ensure its survival
The emergency vascular services and the general support as a unique resource. It is a book which will not be relegated
required in treating the patient are identified. Disease to gather dust on the bookshelf, rather it will be a manual
processes associated with a threat to life ranging from a rup- for frequent perusal by those engaged in emergency vascular
tured aneurysm to a cardiovascular accident are included. practice.
Space is given over to post-surgical complications, includ-
ing infected grafts and the appropriate emergency vascular Norman M Rich MD FACS
response. A combination of medical as well as surgical ther- Professor and Chairman,
apy is incorporated within the armamentarium of response Department of Surgery,
to various vascular emergencies. Challenging clinical emer- Uniformed Services
gencies involving the venous system, ranging from throm- University of the Health Sciences,
bosis and pulmonary embolism to portal hypertension with Bethesda, Maryland, USA
This page intentionally left blank
Preface

Patients presenting with emergency vascular problems, often imaging techniques in refining diagnosis. Indicators of
during antisocial hours, form a substantial percentage ‘best practice’ being absolutely central to modern manage-
of the caseload of a vascular surgeon. Ruptured aortic ment of vascular emergencies, this section ends with chap-
aneurysms, acute limb ischaemia and stroke represent the ters on outcomes of treatment gleaned from the three
core of that emergency practice but the spectrum varies prestigious Scandinavian registries and a more recent UK
from one country to another. In some populations, as has database, followed by an important review of relevant
been true of Northern Ireland for at least a quarter of a medico-legal considerations.
century of terrorist violence, vascular surgeons have also Three further clinical sections follow under the headings
had to deal with life-threatening penetrating injuries. In of acute cerebrovascular syndromes, the acutely swollen
other societies they have had to cope with grave vascular limb and acute lower limb ischaemic states, the latter
emergencies generated by substance abuse, HIV/AIDS or including a superb chapter on the diabetic foot from a well
cold injury. recognised centre. The sheer wealth of topics within the
Between the covers of this book is to be found a com- section on thoracoabdominal catastrophes is exemplified
prehensive range of vascular emergencies affecting the entire by an authoritative contribution on stenting in acute aortic
body, if one excludes those of intracranial and cardiac dissection, a modality of treatment positioned to virtually
origin. Full consideration is given to current practice, evi- displace both the relatively ineffectual conservative med-
dence-based or otherwise, as well as to anticipated develop- ical approach as well as the more perilous surgical option.
ments particularly in the field of minimally invasive Further important sections cover acute complications of
intervention. Endovascular interventions for some emer- endovascular aortic repair (EVAR), regional vascular
gency vascular conditions seem to offer, long-term trial trauma and iatrogenic injuries, the last section including
results pending, speedy resolution, shortened hospital stay first rate chapters on catheterisation and peripheral endo-
and rapid return to an active life. In most centres in the UK vascular injuries. Fresh chapters by recognised world experts
a collaborative team approach, with vascular surgeons and on substance abuse, HIV/AIDS and cold injury give added
radiologists sharing specialist skills, has proved effective weight to the concluding section on special acute vascular
and has largely averted the turf wars with cardiologists and challenges.
neurologists reputedly plaguing those across the Atlantic. With a book such as this devoted solely to the manage-
Each chapter begins by defining the nature and extent ment of vascular emergencies, the vascular specialist look-
of ‘the problem’ when confronted with a particular emer- ing for information is spared the laborious exercise of
gency, emphasising clinical presentation and prudence in delving through burgeoning tomes on vascular surgery in
resorting to time-consuming investigations. In terms of general. For the vascular trainee this book represents
treatment, ‘hands-on’ practical advice is imparted based on essential reading; for the established vascular surgeon,
fundamental principles, identifying pitfalls, offering guide- radiologist and angiologist it is a source to be consulted
lines and giving helpful tips assisted by algorithms on profitably from time to time; for general surgeons, trauma-
available management strategies. The reader will find tologists, emergency physicians and other specialists it
‘boxed’ key points within the text where appropriate, and contains pertinent chapters of interest.
from the extensive bibliography of all chapters, selected Distinguished vascular surgeons, radiologists and other
references are recommended for further reading. specialists around the world have enriched this book with
The text is divided conveniently into nine broad sec- their personal expertise and enthusiasm tempered by
tions, the first on ‘General Considerations’ leading off with sound evidence and mature reflection. I am grateful to
introductory chapters outlining the provision of vascular them for generously taking time in the course of their busy
services in different countries. Focused chapters on the professional lives to contribute invaluably to this book.
pathophysiology of acute limb ischaemia and colonic I must thank Tony Chant, co-editor to the first edition, for
ischaemia, and the systemic sequelae to these insults, pre- staying on board and for his assistance with the prelimi-
cede others on critical care, risk assessment and current nary editing of a few chapters. I would like to register my
xxii Preface

appreciation to the publishers for encouraging a second ensuring that the production of this book is of the highest
edition, in particular Jo Koster, Director of Health Sciences quality. To that end the superb artwork by Simon Lindo,
Subdivision of Hodder Arnold, for moving the project for- Oxford Designers and Illustrators, and the meticulous
ward. I am indebted most of all to Sarah Burrows, Senior scrutiny by Lotika Singha, copy editor, and Andy Anderson,
Development Editor and latterly Commissioning Editor, proofreader, are acknowledged.
for her unfailing kindness, diligent support and professional
expertise during every phase of preparation of the book. Aires A B Barros D’Sa
I also thank Naomi Wilkinson, current Project Editor, for
Preface to the 1st edition

Arteries and veins disrupt or occlude, and if they do so they discuss. We are extremely grateful to them, and to
suddenly an emergency situation arises. Rapid deteriora- Arnold for their help in co-ordinating such an ambitious
tion will threaten the viability of a limb or organ, and venture. Following the three general introductory chap-
indeed the survival of the patient. These emergency vascu- ters, the emergency vascular problems covered in subse-
lar situations call for an attitude of urgency, fine judgement quent chapters are subdivided into three further sections,
and decisiveness if optimal results are to be achieved. namely arteries, veins and trauma. These contributions
Vascular surgical techniques directed at minimizing the come from prominent centres responsible for significant
effects of these acute events have been available since the advances in their respective areas.
early 1960s, and many of the treatment options have changed The authors were especially encouraged to approach their
little during that time. Even relatively new concepts such as topics by first outlining the problem confronting the clini-
thrombolysis have been documented since the late 1960s. cian, and then by examining the pathophysiological sequelae
What has changed, however, is the manner in which these of a particular vascular emergency. Having established this
techniques have been employed: whereas general surgeons starting point, each contributor then outlined the various
previously applied them in a relatively ad hoc manner with strategies employed in the practical management of the con-
occasional spectacular successes, it is now increasingly dition. Many of these vascular emergencies involve other
common for a patient who suffers a vascular catastrophe to medical specialties and therefore require a multidisciplinary
come under the care of a vascular team. approach. It is hoped that access to the broad range of spe-
Our objectives in editing this book were to collect all the cialties fused within the volume will be of value not only to
important vascular emergencies within one volume, the vascular, general and trauma surgeons, but also to radiolo-
chapters of which represent a pooling of intercontinental gists, angiologists, physicians and emergency personnel.
expertise in each of the topics. The very nature of the sub-
ject means that the authors recruited are very busy clini- Anthony D B Chant
cians with a special personal experience of the problems Aires A B Barros D’Sa
This page intentionally left blank
 SECTION
1
General Considerations

1A. Emergency vascular services in the UK 3

1B. Emergency vascular services in Denmark 7

1C. Emergency vascular services in the USA 12

2. Pathophysiology of acute vascular insufficiency 17

3. Intestinal ischaemia in aortic surgery 29

4. Ischaemia-reperfusion injury, SIRS and MODS 39

5. Pathophysiology of stroke 55

6. Assessing the risk in vascular emergencies 65

7A. Perioperative care in emergency vascular practice 73

7B. Emergency vascular access 85

8. Imaging for vascular emergencies 95

9A. Outcomes of emergency vascular procedures. A view from the British Isles 105

9B. Outcomes of emergency vascular procedures in Scandinavia 111

10. Medico-legal aspects of emergency vascular care in the UK 121

This page intentionally left blank
 1A
Emergency Vascular Services in the UK

ANTHONY DB CHANT

The problem 3 Recommendations from the specialty 4

Evolution of emergency vascular services 3 References 5

THE PROBLEM EVOLUTION OF EMERGENCY VASCULAR

SERVICES
A patient suffering a vascular emergency is best served
when under the care of a vascular team working in an Vascular surgery as a specialty in its own right started to
appropriately equipped hospital. The reality, however, is emerge in 1966 when the Vascular Society of Great Britain
often very different. The reasons for this are partly due to and Ireland was founded. Gradually, over a period of some
demographic and socioeconomic conditions but also 40 years this specialty has almost detached itself from gen-
partly the medical culture of the country in question (see eral surgery. The peculiarity of the economics of the British
Chapters 1B and 1C). Perhaps equally important has been National Health Service (NHS), however, has encouraged
our difficulty, as vascular physicians, surgeons and scientists, generalism rather than specialisation and this, in turn,
in establishing vascular surgery as specialty in its own right slowed down the development of the specialty. This was true
and in providing hard evidence with which to convince our with the exception of cities such as Glasgow, Edinburgh
administrations and the myriad healthcare managers that and Belfast where, during the early 1980s, individual sur-
such specialisation is necessary. So-called ‘level 1’ proof, the geons forged ahead independently to establish dedicated
kind that is sometimes obtained from the meta-analysis regional vascular units – the one in Belfast being located at
of prospective trials is, in the vast majority of vascular the main teaching hospital in which also resides a level IA
emergencies, unlikely to be obtained. There have been and trauma centre, with the advantage that patients sustaining
will be opportunities for this approach, particularly in vascular injuries also received optimal vascular care.
adjuvant or supporting therapies. On the whole, however, Nevertheless, in the overall environment, surgeons trained
information has been achieved in a different way. Retro- in vascular surgery found themselves on general surgical
spective and prospective audits have all played an import- emergency intake rotas. Conversely, patients with ruptured
ant role in furthering our knowledge, and reflection on the aneurysms were operated on by, for example gastroenter-
dramatic improvements achieved even over the last ologists or even urologists.
20 years bears witness to their value.1 Neither should we In an effort to paper over this ridiculous divide, serious
be apologetic about the lack of trials. As any aeronautical papers have been published attempting to justify this solu-
scientist will tell you, there is no requirement necessary to tion. Indeed, as recently as July 2000, in the Annals of the
run a controlled trial in order to produce a new aeroplane. Royal College of Surgeons of England, there were no less
A priori thinking, computer modelling, animal experi- than three papers discussing this very topic. The first by
ments together with excellence in the area of audit, both Cook et al.,3 argued that ‘patient outcome alone does not
individual and in combination with other groups, should justify the centralisation of vascular services’. The second,
give us the evidence we need in order to argue the case for by Sutton et al.,4 describes the evolution of vascular surgery
resources necessary to maintain high standards.2 at a district hospital and asks ‘is specialisation inevitable?’.
4 Emergency vascular services in the UK

This particular paper is interesting because it demonstrates that such a unit could care for a population of at least
quite clearly the inevitable rise of the specialty: in 1984 the 500 000. The benefits, of course, would be the concentra-
total number of vascular cases comprised only 3.4 per cent tion of clinical experience and the provision of excellent
of all operations on this particular unit and rose to 33.4 supervision for training.
per cent by 1998, leading to the firmly stated conclusion
that ‘with such a rapidly growing arterial caseload, special-
Intermediate vascular units
isation to vascular surgery is inevitable’. The third paper in
this particular volume of the journal5 discusses regional
These units would service populations of between 300 000
variations in varicose vein operations across England.
and 400 000, but in truth, were they to run efficiently,
A similar disparity was revealed in the provision of carotid
would still require four consultant surgeons plus support
endarterectomy in Wessex.6 The results again confirm the
staff. It was thought, perhaps, that this was not an econom-
inconsistency between subregions in the numbers of oper-
ically viable arrangement.
ations done. In other words, despite the relatively uniform
prevalence of vascular disease across the UK, the actual
number of patients treated depends very much on local Remote vascular units
circumstances.
The kind of arguments that have been previously This description was applied to geographically isolated
advanced by those wishing to retain the status quo and hospitals serving small populations of between 100 000 and
remain generalists as well as doing ‘a bit of vascular surgery’ 250 000. At this point the concept of ‘hub and spoke’ was
go as follows. First, that all but a few pregangrenous legs discussed. The problems associated with this approach
can be treated with analgesics and rehydration overnight have been discussed elsewhere.7 The main criticism of this
and that angiograms ought to be done only in daylight concept is that those specialists at the end of the ‘spoke’,
hours. A further simplistic but plausible argument has also although on appointment well trained in modern vascular
been forwarded regarding abdominal aortic aneurysms: as surgical techniques, in time, and because of the lack of
50 per cent of these patients die before entering hospital, elective operating, inevitably would become less efficient
and as an expert vascular surgeon can salvage only 60 and less tuned in to best practice. Moreover, it is hard to
per cent of the remaining 50 per cent, namely 30 patients, justify, economically, the duplication of support services in
not much is to be gained by having an expert vascular sur- these peripheral units, when for much of the time the
geon; further, because an experienced general surgeon ‘plant’ would remain unused.
might salvage 40 per cent of the remaining 50 per cent, in These initial fairly simple recommendations were fur-
other words, 20 patients, the extra spending on the reten- ther complicated by the fact that the Royal College of
tion of specialist vascular surgeons and resources results in Surgeons changed their Fellowship requirements so that
a net gain of only 10 lives; the money, therefore, could be even quite senior trainees required supervision at all times.
better spent in managing other illnesses. The European Working Time Directive ‘Department of
Health, Hours of Work for Doctors in Training’8 was then
published, this ‘New Deal’ limited the hours that doctors,
RECOMMENDATIONS FROM THE SPECIALTY both senior and junior, could either work or remain on-
call. Trainee vascular surgeons will now receive even less
exposure to vascular emergencies. It was considered wrong,
In October 1998 the Vascular Surgical Society published
for example for a surgeon to be on a more than one in five
the first recommendation in respect of the provision of
emergency rota. Given the complexity of the government
vascular surgery services.1 This very comprehensive docu-
rules and their economic consequences for individual
ment details the problems faced by the British NHS at that
hospitals, it is perhaps not surprising that in a few areas in
time in terms of providing emergency vascular services and
the UK general surgeons still have to look after vascular
goes on to suggest the following solutions.
patients. That said, things have improved markedly and
there is now much closer liaison between both specialist
Major vascular units interventional radiology9 and anaesthetic services. The
repeated emphasis over the years by the National
Wherever possible, vascular services should be provided on Confidential Enquiry into Postoperative Deaths (NCEPOD)
a single site. The recommendation was that four or more of the need for vascular emergencies to be treated by a
consultant surgeons should staff these centres, a number specialist vascular team10 is probably being achieved but
allowing for a workable on-call rota. These consultants, in probably more so as a product of other influences. That
turn, should be supported by interventional radiologists objective, undoubtedly, will be accelerated by the absolutely
and vascular anaesthetists and by facilities such as an inten- unanimous desire of the members of the Vascular Surgical
sive therapy unit, a high-dependency unit and a vascular Society, expressed in November 2003, to seek independent
laboratory. The workload figures at that time suggested specialty status outside general surgery.
 References 5

The most recent recommendations from the Vascular

Surgical Society, the Provision of Emergency Vascular
Conclusion
Services, were published in November 200111 and, realising
The, perhaps slightly optimistic, conclusion to the man-
that the scene was constantly shifting, it was agreed that the
agement of emergencies is that the gradual evolution of
document should be reviewed in 2004. Much of the ration-
comprehensive vascular services in the UK may be in the
ale for the management of vascular emergencies by vascu-
process of being accomplished now that the government
lar surgeons is described above, but two significant new
has given its tacit approval and accepted a measure of
items have been added. First, it notes the introduction of
co-responsibility in the most recent provisions.14 Arriving
so-called ‘clinical governance’ and goes on to state ‘general
at this juncture has been a painful process in many
surgeons who are not vascular specialists undertake no
respects but, to an extent, it illustrates the complexity of
active emergency arterial surgery, and it is difficult for them
having to please everyone. Advancement of the specialty
to justify treating vascular emergencies under the scrutiny
in the UK has been hampered, partly by the remoteness
of clinical governance’. In other words, if a general surgeon
of some parts of these islands, but mainly by under
in the UK now performed an operation for a vascular emer-
resourcing of the NHS and the encouragement, until
gency, and were that patient to die or perhaps lose a leg
very recently, of generalism rather than specialisation. If
unnecessarily, then that surgeon may well be open to liti-
you choose to live or work in a remote area, then it is
gation. It cannot be emphasised enough just how import-
unlikely that you will get the very best of medical treatment.
ant a shift of opinion this is in service terms. The second
That is a reality of life.
important change is that there is at last an admission that
‘there is an accumulation of evidence that outcomes are
better when vascular patients are treated by vascular spe-
cialists’.12,13 Moreover the document11 goes on to state
quite dogmatically that ‘the modern generation of newly Key references
appointed consultant general surgeons is insufficiently
trained and experienced to manage complex emergencies Department of Health. New Deal Doctor Training the European
outside their own specialist field’. Working Hour Directive. London: DHSS, 2002.
While this was all accepted in theory, political negoti- Earnshaw JJ, Ridler BMF (eds). National Vascular Database Report
ations aimed at actually amalgamating units have proved 2000. London: The Vascular Society of Great Britain and
extremely difficult, because, quite certainly, some of the Ireland.
smaller hospitals would lose their vascular services com- The Royal College of Radiologists and the Vascular Society of Great
Britain and Ireland. Provision of Vascular Radiological Services.
pletely. Important situations such as ‘in-house emergen-
Combined Recommendations. London: VSS, April 2003.
cies’ might be difficult to deal with. Does one, for example,
The Provision of Emergency Vascular Services. A document prepared
transfer a diabetic patient who requires a simple ray ampu- for the Vascular Society of Great Britain and Ireland. London:
tation of a toe, 30 or 40 miles (48 or 64 km) to a major unit? VSS, November 2001.
And how does one deal with the dilemma of a patient, who, The Provision of Vascular Services. A document prepared for the
in the process of undergoing a general or other non-vascular Vascular Society of Great Britain and Ireland by the Surgical
surgical operation, sustains an iatrogenic vascular injury Advisory Committee. London: VSS, October 1998.
(see Chapter 40). These and other questions demand sens-
ible, workable and current solutions. Therefore, after a
two-year period of consultation with the membership, the
Vascular Surgical Society, in liaison with, and the approval
of, the departments of health of the four regions of the UK, REFERENCES
published Provision of Vascular Services 2004 in November
2003.14 This document supports local collaborative strategies 1 The provision of vascular services. A document prepared for the
which achieve a preferred minimum emergency on-call rota Vascular Society of Great Britain and Ireland by the Surgical
of one in six, which goes a long way to providing that 24-hour Advisory Committee. London: VSS, October 1998.
cover necessary in dealing with the kind of situations illus- 2 Earnshaw JJ, Ridler BMF (eds). National vascular database report
trated above in this paragraph. Emergency work is onerous 2000. London: The Vascular Society of Great Britain and Ireland.
and time consuming and these facts deserve recognition by a 3 Cook SJ, Rocker MD, Jarvis MR, Whitely MD. Patient outcome
alone does not justify the centralisation of vascular services.
government now committed to the strategies laid down in
Ann R Coll Surg Engl 2000; 82: 268–71.
this document and by hospitals negotiating new contracts 4 Sutton CD, Gilmour JP, Berry DP, Lewis MH. The evolution of
with consultant vascular surgeons. In due course, with a vascular surgeon in a district general hospital. Ann R Coll
increasing participation in the National Vascular Database2 Surg Engl 2000; 82: 272–4.
by the membership of the Vascular Society, solid data should 5 Galland RB, Whatling PJ, Crook TJ, Magee TR. Regional variation
emerge which will assist in refining existing arrangements for in varicose veins operations in England 1989–1996. Ann R Coll
the management of vascular emergencies in the UK. Surg Engl 2000; 82: 275–9.
6 Emergency vascular services in the UK

6 Ferris G, Roderick P, Smithies A, et al. An epidemiological needs Deaths, 1994/95. London: Royal College of Surgeons of
assessment of carotid endarterectomy in an English health England, 1997.
region. BMJ 1998; 317: 447–51. 11 The Provision of Emergency Vascular Services. A document
7 Chant A. Emergency vascular services. In: Chant A, Barros D’Sa prepared for the Vascular Society of Great Britain and Ireland.
AAB. (eds). Emergency Vascular Practice. London: Arnold,1997. London: VSS, November 2001.
8 Department of Health. New deal doctor training the European 12 Michaels J, Brazier J, Palfreys, et al. Cost and outcome
Working Hour Directive. London: DHSS, 2002. implications of the organisation of vascular services. Health
9 The Royal College of Radiologists and the Vascular Society of Technol Assess 2000; 4.
Great Britain and Ireland. Provision of vascular radiological 13 Wolfe J. The delivery of vascular services in the United Kingdom.
services. Combined recommendations. London: VSS, Cardiovasc Surg 1999; 7: 692–3.
April 2003. 14 The provision of vascular services 2004. Document prepared by
10 Gallimore SC, Hoile RW, Ingram GS, Sherry KM. The Report the Working Group of the Vascular Surgical Society of Great
of the National Confidential Enquiry into Perioperative Britain and Ireland. London: VSS, November 2003.
 1B
Emergency Vascular Services in Denmark

HENRIK SILLESEN

The problem and the need for emergency vascular services 7 Political issues 10
Staffing and structuring of the unit and hospital 7 Location of and distance between vascular surgical 11
requirements services
What can an expert unit do? 9 References 11

THE PROBLEM AND THE NEED FOR 10 units staffed solely by specialists in vascular surgery, the
EMERGENCY VASCULAR SERVICES results of surgical treatment of ruptured abdominal aortic
aneurysms (RAAAs) speak for themselves: the mortality in
treating more than 1400 cases of RAAA was 42 per cent
Vascular surgery is a field of expertise demanding some (Table 1B.1).
degree of specialisation and therefore in many countries it
is recognised as a monospecialty. In most of these coun-
tries, specialisation within vascular surgery requires a
STAFFING AND STRUCTURING OF THE UNIT
number of years training within vascular surgery following
AND HOSPITAL REQUIREMENTS
two to three years in general surgical training. It should
follow, therefore, that, in general, results are superior in
larger hospitals with access to vascular surgery rather than In order to provide expert emergency vascular service
in smaller hospitals.1,2 The results of emergency vascular 24 hours a day, 365 days a year, the specialist unit needs to
surgery should also be better if conducted by specialists. be staffed accordingly. Taking all duties into account, in
This difference may be more difficult to observe because addition to vacation and holidays, continuing medical edu-
of a number of factors, including that of smaller numbers cation, scientific meetings and so forth, a minimum of three,
and the difficulty in comparing cases. In Denmark, how- and preferably four to five vascular surgical specialists are
ever, where all vascular surgery is performed within only needed in each unit. Looking at the organisation in Denmark,
a total of only 10 vascular surgical units/departments serves
the population of 5.5 million. All elective as well as emergency
Table 1B.1 Thirty-day mortality following emergency operation cases are treated in these 10 units, staffed, in total, by approx-
for ruptured abdominal aortic aneurysm (RAAA) in Denmark imately 45 consultants.
(www.karbase.dk). With a population of 5.5 million the average
Depending on whether the hospital is a university or a
number of RAAAs is 4.3 per 100 000 per year
regional hospital, the departments may have their own
1997 1998 1999 2000 2001 Total junior staff or share them with a general surgical unit. In
most university hospitals the vascular surgical unit will
No. of cases 259 264 230 205 232 1190 have its own staff on call represented by an intern/resident
Mortality (n) 93 120 105 92 102 512 in-house and a fellow/consultant on call, the latter avail-
able within 30 minutes. When the vascular fellow is on
Mortality 36 46 46 45 44 42
call a consultant is also similarly committed. Thus, when
(per cent)
dealing with an RAAA, an operating team of two to three
8 Emergency vascular services in Denmark

surgeons is available, including a vascular surgical consultant surgery is ideally performed in hospitals where other major
and fellow. Having staff on call leads to long working hours operations on patients with significant competing diseases
and the need for compensatory absence. In Denmark the are undertaken, and therefore, where all the personnel
official working week is 37 hours, and even though some working in the vicinity of the vascular surgeon are cog-
time on call may be compensated financially, a senior nisant of the common challenges.
fellow/consultant will be on compensatory leave, often for A well-equipped intensive case unit (ICU) is mandatory
1 week out of a 4- to 6-week period. With a shortage of con- when performing vascular surgery. Both in elective as well
sultants in almost all specialties in Denmark this poses a as emergency vascular cases, major complications result in
problem, especially for staffing the more ‘remotely’ located failure of almost any organ, most notably the lung, kidney
units. Consequently, a new kind of intercounty collabor- and heart. Similarly, for the staff of the ICU, regular con-
ation is being developed where a university/large regional tacts, preferably daily or at least weekly, with vascular sur-
hospital may cover the neighbouring counties for emergency/ gical patients is mandatory. Access to dialysis is also
acute cases after 4 pm. Vascular surgical activity being important as renal failure is a common complication in
fairly limited beyond daylight hours, more effective use is major vascular surgical emergency cases, in particular rup-
made of a consultant’s skills, but the potential catchment tured aneurysms. Transfer of vascular surgical patients to
population covered may well be a million or more other hospitals for dialysis, where vascular surgeons are not
inhabitants. in attendance, may result in less intensive surveillance of
Naturally, access to an operation room (OR) is manda- coexisting vascular surgical problems. In general, the need
tory and a large theatre is preferable. As in many cases there for dialysis may be a significant prognostic factor for a vas-
is no time for preoperative radiological investigation, the cular patient and, although difficult to document, it is the
surgeon has to resort to intraoperative angiography in experience of the author that moving a vascular surgical
treating lower limb emergencies. Physical space for mobile patient to another hospital for dialysis is a strong predictor
or permanently installed X-ray equipment is necessary, of death. This, of course, is reflected in the high mortality of
along with staff capable of using it at any hour of day or patients with renal failure following major vascular surgery.
night. The nurses and staff in the OR assisting the vascular Having the patient in another hospital where there is no
surgeon should have a thorough knowledge of the instru- vascular surgical expertise, however, may result in decisions
ments and the common vascular reconstructive proced- being made which might not be the best for the patient.
ures. Obviously, the provision of a complete vascular A ward caring only for vascular surgical patients is of
surgical team of assistants, available 24 hours a day, would course the ideal solution. In a number of Danish hospitals
be the ideal situation but this is probably only achievable this is the case. These are relatively small units of 14–20
in very few units. Having a core of nurses mainly in sup- beds staffed with personnel taking care of only vascular
port of the vascular surgeons, makes it possible for at least surgical patients. This allows for special training of the
one of them to be on call most of the time. With a surgical nursing staff, in turn improving the quality of surveillance
volume of a minimum of 400–600 vascular cases per year, and care of vascular surgical patients. These skills include
vascular surgical operations are performed daily, providing experience with ischaemic and venous ulcers, supervision
sufficient numbers for training OR staff. of newly operated patients, ability to measure ankle pres-
This caseload also allows anaesthesia personnel to gain sures, evaluate whether or not a bypass graft is patent and
the necessary experience in dealing with patients with both functioning well, experience with the continuous infusion
central and peripheral arterial disease (PAD). Patients with of thrombolytic agents and management of patients with
PAD generally suffer from other diseases such as concomi- high comorbidities. If a unit solely dedicated to vascular
tant coronary heart disease and chronic obstructive lung surgery is not attainable, a section within a surgical ward
disease. These risk factors in turn account for a high peri- can be a good start. For nursing staff dedicated to vascular
operative morbidity and mortality even in elective lower patients, however, it is vitally necessary in order to main-
limb surgery. The Danish National Register of Vascular tain continued and optimal surveillance and care.
Surgery (www.karbase.dk)3 recorded a 30-day mortality of The radiological department should have a dedicated
approximately 4 per cent in 5012 peripheral bypass oper- angio-suite and one or preferably two vascular radiologists
ations performed during 1996–99. Of course, these patients whose duties are mainly vascular. Whether or not access
did not die directly from the operation as from the conse- to an expert radiological vascular service is necessary at
quences of other competing circulatory or respiratory all times is debatable, however, day-to-day contact is.
diseases. In general, first it should be assumed that any On the other hand, a round-the-clock facility for computed
patient with PAD also suffers from coronary heart disease tomography (CT) is mandatory. It is also advantageous to
even though there are no cardiac symptoms. Second, man- have ultrasound expertise available if the vascular surgeon
aging the major haemodynamic changes during aortic sur- cannot perform this examination on his or her own.
gery following cross-clamping of the aorta requires skill Finally, having a vascular laboratory, ideally located
and experience. Furthermore, the great volume losses dur- within the vascular surgical unit is of great value. Although,
ing RAAA surgery create demanding situations. Vascular many emergency vascular cases may utilise the skills of the
 What can an expert unit do? 9

radiological department, some of the complications can be their general surgical skills up to date at all times, but as all
investigated in the vascular laboratory. The advantages of a vascular surgical units/departments in Denmark are either
vascular laboratory are discussed further below. in university hospitals or regional hospitals, where general
surgical staff are also to be found, a gastroenterological or
urological consultant surgeon, for example, can be called
upon if necessary. An indepth devotion to the specialty,
WHAT CAN AN EXPERT UNIT DO? concentrating on vascular cases only, sharpens a consultant’s
expertise and certainly makes for better outcomes.
In addition to treating emergency vascular cases, vascular The surgical throughput in the vascular surgical units
surgical staff at an expert unit can also provide elective described is sufficient to offer a broad range of experience
services to a region. Units covering an area with in keeping the skills of the consultants and associated staff
400 000–500 000 inhabitants, staffed by three or four con- up to date. Depending on the size of the unit/department,
sultants, should generally have an activity level of around it also allows for one or more resident/fellow appoint-
400–600 arterial cases, 75 per cent being open surgical ments. In Denmark, a ‘small’ unit with three or four con-
cases and 25–35 per cent percutaneous transluminal angio- sultants provides one of the three years of training required
plasties, although the proportion of the latter may be by a fellow for specialisation in vascular surgery. The larger
higher in some countries. Added to the ‘real’ arterial cases units, wherein all kinds of vascular surgical treatments are
is an average of 0.2–0.3 additional, usually secondary oper- undertaken, provide the remaining two years of training.
ations per arterial reconstruction, mainly minor amputa- The larger units also offer resident positions for 6 months
tions, wound revisions and so on, thereby increasing the which are very important in the recruitment of future vas-
operative caseload by 20–30 per cent. cular surgeons.
Vascular surgical activity includes the entire spectrum A consultant is always on call in support of a vascular
of preoperative work-up, diagnostic studies, the operation fellow and to whom the trainee has access for advice and
itself, the care provided during hospitalisation and finally, supervision in the management of emergency cases.
but very importantly, postoperative monitoring and During daylight hours, one vascular trainee working under
indeed follow-up. In Denmark, most departments/units three or four consultants will generally receive very good
offer 30-day and one-year follow-up for all patients. In supervision in the management of elective cases.
addition, patients undergoing bypass surgery are often also Large units facilitate the establishment of a vascular
seen at intervals between the standard re-visits, usually at laboratory located within a vascular surgical unit or outpa-
6–12-week intervals during the first year. Venous disease, tient clinic which not only means that patients do not have
especially the more complex cases of redo varicose vein sur- to attend another clinic but also increases use of expensive
gery, secondary varicose veins and chronic venous obstruc- equipment. For instance, a duplex scan is more frequently
tion can also be treated by vascular surgeons. Although not performed if it can be done immediately and without the
within the scope of this chapter, we find it increasingly time-consuming administrative measures in having it per-
important that vascular surgeons ensure patients are treated formed elsewhere, or even worse, on another day. A vas-
adequately with respect to secondary prevention.4 A special cular laboratory immediately available within a unit also
outpatient clinic dedicated to the organisation of risk factor offers important educational as well as research opportun-
reduction by means of lifestyle changes and medication ities for the vascular surgical trainee.
allows well-trained and supervised vascular nurses to take A senior vascular surgical trainee represents the avail-
the responsibility of improving the total care of the vascular ability of vascular expertise not only for ‘in-house’ emer-
patient. gencies but also for those in another hospital within the
The volume of emergency work is of a reasonable size unit’s intake area. Injured patients brought into the emer-
with an average of 20–25 RAAAs, within a range of 10–35 gency room may turn out to have vascular problems or,
cases treated per unit depending on the size of the catch- occasionally, an iatrogenic vascular injury may complicate
ment area. a non-vascular surgical procedure. If a vascular surgeon is
With a staff of at least three or four consultant vascular not available to deal with such cases, the outcome may be
surgeons per unit, continuing education of staff is possible less than favourable.
in the knowledge that vascular surgical expertise is always On approximately 10 occasions annually, vascular sur-
available during meetings and courses. Given that all vascular geons at Gentofte University Hospital perform emergency
surgical operations in Denmark are performed in the operations at one of the other three hospitals within
aforementioned 10 vascular surgical units/departments, it Copenhagen County, which has a population of 630 000
follows that no Danes are treated by anyone other than inhabitants. For example, the time taken for a vascular sur-
full-time vascular surgical consultants. An organisation in geon to travel to a patient with a ruptured aneurysm in cir-
which all vascular activity is concentrated within a few culatory shock is much less than that needed to stabilise the
centres ensures that vascular surgeons are solely respon- patient for transportation and effect the actual transfer of
sible. It may be argued that vascular surgeons should keep the patient to Gentofte. Using this strategy, the results of
10 Emergency vascular services in Denmark

emergency RAAA operations performed in other hospitals 10

0
on 19 patients from 1996 to 1999 by Gentofte staff show a 0
9 14
14patients
0
8 Average mortality 4
5.0per cent
30-day survival of 47 per cent.5 When compared with the

Mortality (per cent)

0
7
results in-house at Gentofte or with those of the whole 0
6
country, these results are almost as good, despite the fact that 50
in these particular cases the worst possible outcome might 0
4
otherwise have been expected. This raises the question as 30
20
to what is most important when treating a ruptured 10
aneurysm: is time or the organisation of the vascular surgi- 0
cal unit the key issue? It is well recognised that the patient 4 1 2 5 6 3 8 10 7 9
with an RAAA represents a challenge for anaesthetists and
Figure 1B.1 Thirty-day mortality in 10 Danish vascular surgical
it has also been argued that, unless both they and the ICU
departments over a 6-year period (1996–2001); 95 per cent
staff are experienced, results may suffer. In Copenhagen confidence intervals are shown. The numbers below the figure
County, the patients operated outside Gentofte Vascular indicate department number and these are kept confidential; each
Surgical Department are almost always transferred to centre knows its result and what number they are on this figure;
Gentofte Hospital as soon as possible, that is, immediately however, they are unaware of other departments’ numbers (source:
after surgery or the day after. Similarly, for trauma cases www.karbase.dk3)
dealt within Copenhagen County, the availability of a
‘travelling’ expert is important especially as these other
hospitals are all university hospitals with large teaching
anaesthesia departments and well-equipped ICUs. surgeons, case selection and not necessarily the capabilities
It is of course debatable as to whether a ‘travelling ser- of an individual department or hospital (Fig. 1B.1).
vice’ will produce good results in all cases. For example, The five-year mortality rate of 42 per cent, derived
when operating on hypotensive patients with RAAAs in from all RAAA surgery undertaken in Denmark, compares
remote small hospitals inexperienced in major surgical well with similar analyses from other countries. Of the 10
work the outcome is usually poor. A recent Danish study Danish hospitals performing vascular surgery, only five are
from a less densely populated area presented the results of university based and the other five are regional hospitals.
emergency RAAA surgery undertaken in remote hospitals In the USA an analysis of 13 887 cases from 20 per cent of a
by a ‘travelling vascular surgeon’: the 30-day mortality in random sample of all hospital activity during 1996–97
this small series was 70 per cent, and although the authors showed an overall RAAA mortality of 47 per cent. A meta-
concluded that such operations were worthwhile, it could analysis of all English literature quoting RAAA mortality
be argued that the results contradicted this view and that during 1995–98 revealed an overall operative mortality of
patients should be transferred to the expert unit under all 48 per cent.7
circumstances.6 The experience from southern Sweden
shows that the on-table mortality from RAAA is similar
whether patients are operated in a university hospital (12
POLITICAL ISSUES
per cent) or a county hospital (15 per cent), nevertheless,
there was a difference in the 30-day mortality, being 26 and
41 per cent, respectively.2 An understanding of the Danish organisation of emer-
The organisation of vascular services in Denmark, in gency services is based on the recognition of the following
general, has encouraged good results in treating emergen- issues. Almost all healthcare is socialised and paid for by
cies, if the 30-day survival of operations for ruptured public services. The country is divided into 19 counties,
aneurysms is taken as an indicator. Looking at the results each responsible for its own organisation and the econom-
of all cases treated over 5 years (1997–2001), a 30-day ics of healthcare within it. A central administrative author-
mortality rate is 42 per cent in 1190 cases treated (see ity is answerable to a political county council elected by the
Table 1B.1), with only minor variations from one year inhabitants of that county every four years. Virtually all
to another.3 vascular surgery is performed within public hospitals. The
Comparison between departments is difficult on an cost of healthcare in Copenhagen County is paid for by its
annual basis due to the numbers being treated. Over six inhabitants. In total, Denmark has 19 counties, 10 of which
years, however, experience reveals quite similar results in decided to have their own vascular surgical service.
the 10 Danish centres. The difference between the depart- The decision to establish a specialised unit or not is a
ment with the best results and that with the highest local decision for a particular county. If it chooses not to
mortality is 14 per cent (39–53), but using 95 per cent con- have its own specialist unit, it has to transfer vascular sur-
fidence intervals this difference is not statistical significant. gical patients to a hospital in another county and pay for
Even if it were, the differences in mortality might simply their treatment. Given the system of reimbursement
reflect differences in the performance of the attending between counties implemented by the year 2000, and using
 References 11

diagnosis related groups (DRGs), the price for treating the

Key references
individual case is a little higher than the ‘cost’ price, which
makes it fairly attractive to hospitals receiving patients Brown MJ, Sutton AJ, Bell PR, Sayers RD. A meta-analysis of
from another county. 50 years of ruptured aortic aneurysm repair. Br J Surg 2002;
The service a county wishes to provide has a political 89: 714–30.
dimension. Some counties, despite being relatively small, Dimick JB, Stanley AC, Axelrod DA, et al. Variation in death rate
may have a political majority favouring the provision of after abdominal aortic aneurysmectomy in the United States:
all medical speciality services and, consequently, in some impact of hospital volume, gender and age. Ann Surg 2002;
small specialties the cost per case may be very high. In other 235: 579–85.
counties a more pragmatic attitude may be evident where Karbase Landsregister. Website of The Danish Vascular Registry,
economical considerations rule decision making. www.karbase.dk (accessed 11 December 2004).
Sillesen H. Who should treat patients with peripheral
National health authorities define a disease requiring
atherosclerosis – the vascular specialist [editorial]. Eur J Vasc
specialist attention by, for example, stating that aortic
Endovasc Surg 2002; 24: 1–3.
abdominal aneurysms need vascular surgical attention. Zdanowski Z, Danielsson G, Jonung T, et al. Outcome of treatment
Within vascular surgery itself, certain procedures are con- of ruptured abdominal aortic aneurisms depending on the type
sidered so rare that they ought to be concentrated in a few of hospital. Eur J Surg 2002; 168: 96–100.
tertiary units. These cases would include mesenteric and
renal artery obstruction, infected abdominal vascular pros-
thesis and certain supra-aortic emergencies.
REFERENCES

1 Dimick JB, Stanley AC, Axelrod DA, et al. Variation in death rate
LOCATION OF AND DISTANCE BETWEEN after abdominal aortic aneurysmectomy in the United States:
VASCULAR SURGICAL SERVICES impact of hospital volume, gender and age. Ann Surg 2002;
235: 579–85.
2 Zdanowski Z, Danielsson G, Jonung T, et al. Outcome of treatment
Ideally, in the majority of cases, the time needed to trans-
of ruptured abdominal aortic aneurisms depending on the type of
port a patient from home, or from another hospital lacking hospital. Eur J Surg 2002; 168: 96–100.
a specialist vascular service, should not exceed 1–2 hours. 3 Karbase Landsregister. Website of The Danish Vascular Registry,
In Denmark, this is the case almost anywhere in the www.karbase.dk (accessed 11 December 2004).
country, with the exception of people living on small 4 Sillesen H. Who should treat patients with peripheral
remote islands. Even then the possibility of providing atherosclerosis – the vascular specialist [editorial]. Eur J Vasc
transport by helicopter exists. In these cases, however, it is Endovasc Surg 2002; 24: 1–3.
often a matter of transporting a patient to the right hos- 5 Vammen S, Fasting H, Henneberg EW, Lindholdt JS. Karkirurgisk
pital rather than just the nearest one. This, of course, is assisterede operationer for rumperet abdominalt aortaaneurisme
dependent on someone making the correct diagnosis udført på primært modtagende sygehus. Ugeskrift for Læger 1999;
161: 4868–70.
which may not happen until the patient is actually in a hos-
6 Jensen LP, Bækgaard N, Sillesen H. Ruptured aneurysms can
pital which has a vascular surgical service.
be treated by assisted surgery in local hospitals if the condition
does not allow transportation. Ugeskrift for Læger 2000;
162: 198–9.
7 Brown MJ, Sutton AJ, Bell PR, Sayers RD. A meta-analysis of
Conclusion 50 years of ruptured aortic aneurysm repair. Br J Surg 2002;
89: 714–30.
Vascular surgery is a specialty which benefits from cen-
tralisation, partly with respect to quality of treatment,
but also with regard to cost. This is especially important
with regard to patients presenting as emergency cases in
the care of whom the availability of an experienced vas-
cular surgeon is crucial; equally so is the hospital’s abil-
ity and experience in dealing with emergency vascular
surgical cases.
 1C
Emergency Vascular Services in the USA

VINCENT L ROWE, JUAN A ASENSIO, FRED A WEAVER

The problem 12 Operating room 14

Epidemiology 12 References 15
Requirements 13

THE PROBLEM hospital and remain unaccounted for by our current methods
of medical record keeping. An estimate of activity at our
medical centre may provide some insight in this matter.
Emergency vascular surgery services encompass a wide The vascular surgery service at Los Angeles County and
variety of disorders, including aortic aneurysm disease, University of Southern California (LAC  USC) Medical
acute lower limb ischaemia, symptomatic carotid occlusive Centre annually admits over 250 patients and provides
disease and vascular trauma, to name but a few. The variety consultations on over 300 inpatients. Approximately
of disorders encountered in the emergency rooms of hos- 25 per cent of these admissions are of an emergency nature
pitals in the USA necessitates familiarity with evaluation and come through the ER or clinic. Over the past three
and management of vascular disorders in order that the years, our vascular service has performed an average of 355
patient can be cared for expeditiously. operations, 21 per cent of these cases being emergencies.
Delivery of emergency vascular services requires a Interestingly, the same distribution applies to the average
diverse group of healthcare professionals including vas- number of surgical hours spent on vascular elective and
cular and trauma surgeons, emergency room (ER) phys- emergency operations.
icians, nursing personnel, emergency medical technicians, The most common vascular emergency at our institu-
radiological and operating room personnel. Organising tion is that of acute upper and lower extremity ischaemia,
services into a co-ordinated functioning unit requires constituting almost 20 per cent of all emergency vascular
co-operation between surgery, radiology and emergency operations undertaken. While the aetiology of ischaemia in
medicine personnel. the upper extremity is most frequently due to an embolic
This chapter reviews the manner in which vascular ser- source, acute ischaemia of the lower extremity encom-
vices are facilitated and organised in the USA using one passes embolic as well as thrombotic complications in ath-
large public hospital as an example. The most common erosclerotic patients. Given the lower rate of limb salvage
emergency vascular problems are ruptured abdominal in the latter group of patients, the number of amputations
aortic aneurysms (RAAAs), vascular trauma, acute limb performed for ischaemic disease may provide a rough esti-
ischaemia and stroke. mate of this problem throughout the USA.1–3
Despite improvement in care and technological
advancements, the incidence of RAAAs continues to be a
EPIDEMIOLOGY persistent and common problem.4,5 The relative incidence
of ruptured to elective aortic aneurysm repairs for a partic-
The overall incidence of acute vascular surgical emergen- ular medical centre is difficult to quantify given that multi-
cies in the USA is difficult to quantify. Owing to the cata- ple variables are involved. Referral patterns from primary
strophic consequences of delay in correct diagnosis and care practitioners, technological advancements and indeed
treatment of such acute circulatory problems, numer- the reputation of surgical staff, all have an impact on the
ous patients will succumb to vascular emergencies outside incidence of elective aneurysms treated at a particular
 Requirements 13

institution. Similarly, the urgency of the patient’s condi- members of the ACS-COT and strives to set guidelines for
tion and the proximity of the nearest medical facility have the verification of trauma centres. Simultaneously, the des-
an effect on the incidence of RAAAs seen at each medical ignation of trauma centres by ‘levels’ has evolved into the
centre. At the majority of large urban hospitals, a ratio of development of a comprehensive trauma care system, which,
elective to ruptured aortic aneurysm repairs is approxi- in geographical terms, provides care for injured patients
mately 4:1. Here at LAC  USC Medical Centre, however, throughout the USA.
the distribution of major vascular procedures is heavily The LAC  USC Trauma Centre is a level I trauma
weighted towards reconstructions for aortic occlusive dis- centre and considered a regional resource facility caring for
ease. As such, our centre manages an even mix of ruptured the citizens of Los Angeles County, the most populous
and elective abdominal aortic aneurysm repairs, essentially county in the USA. It is one of the largest trauma centres in
unchanged over the past five years. the USA admitting on a yearly basis between 7500 and 8200
The management of patients sustaining vascular injured patients. Being of level I status, in-house trauma sur-
injuries caused by firearms and motor vehicle accidents gery specialists are available 24 hours a day, along with
continues to pose a significant challenge to our institution. 24-hour availability of operating rooms, surgical intensive
Over the past 10 years, the LAC  USC Trauma Centre has care units and a full complement of other surgical and
admitted 967 patients with 1399 vessels injured, a mean non-surgical specialists needed for optimal provision of
revised trauma score (RTS) of 5.44 and a mean injury trauma care.
severity scores (ISS) of 23. Most of these patients were The special expertise and resources available to main-
admitted in haemorrhagic shock with a mean systolic blood tain a verified level I trauma centre are both easily applic-
pressure of 92. The majority of them underwent operative able and transportable to the management of vascular
treatment of their vascular injuries sustaining a mean esti- emergencies. Both injured patients and those admitted
mated blood loss of 3862 mL of blood. Clearly the special with vascular emergencies may arrive at the ER in pro-
resources at our level I trauma centre contributed to the found shock and severe physiological compromise, requir-
excellence in the provision of care for both critically injured ing a rapid mobilisation of all resources. This includes
trauma patients as well as for patients presenting with resuscitation, immediate transport to a surgical suite for
vascular emergencies. definitive lifesaving interventions and immediate availabil-
Cerebrovascular accidents are the third leading cause of ity of large quantities of blood. Immediately postopera-
death in the USA. At our medical centre, approximately tively, individuals trained in surgical critical care continue
750 patients with newly diagnosed strokes are managed the process of resuscitation and administer advanced tech-
annually. Of these patients, the aetiology of the stroke is nological support to patients decompensating into single
ischaemia in over 60 per cent of cases. This is not to men- and/or multiple system organ failure.
tion the additional number of unaccounted patients experi- Trauma patients and those with vascular emergencies
encing transient cerebral ischaemia who are treated and also may have need for multiple staged procedures which
then discharged from the emergency room. To improve require ‘bail out/damage control’. Consequently, the pres-
the timely treatment of this devastating entity, integration ence of a well established ACS verified level I trauma centre
of emergency room personnel, radiology staff, specialty with its multidisciplinary approach to the management of
neurologists and vascular surgeons to form a dedicated complex injuries can also serve a dual purpose in the pro-
acute stroke team is currently in action at our medical vision of excellent care to patients presenting with vascular
centre. emergencies.
Physicians in most emergency rooms make an initial
assessment of a vascular emergency. They provide the ini-
tiative in pursuing an aggressive approach involving
REQUIREMENTS
appropriate specialists and the vascular surgeon in particu-
lar. Familiarity with common vascular emergencies and
The American College of Surgeons (ACS) founded in 1913, the ability to perform a thorough vascular examination is
along with its oldest standing committee, the ACS therefore critical to an emergency physician’s expertise in
Committee on Trauma (ACS-COT) established in 1922, providing an initial assessment and diagnosis on these
have been the leaders in the provision and standardisation patients.
of care for injured patients. Amongst some of the most Emergency physicians are very dependent on emer-
important steps taken by the ACS-COT was the creation of gency services and emergency medical technicians. In the
a document entitled Optimal Hospital Resources for Care of USA, these individuals are trained to stabilise and move the
the Injured Patient, which, over the years, has evolved with patient rapidly from the scene of a non-traumatic problem
the contributions of many trauma surgeons into the to the hospital where treatment can be provided with the
monograph Resources for Optimal Care of the Injured highest degree of sophistication and equipment.
Patient, first published in 1990 and revised in 1993 and The diagnostic approach to extremity trauma has
1999.6,7 This publication is periodically updated by the changed dramatically over the past few decades. Initially, the
14 Emergency vascular services in the USA

influences from combat experience led to the aggressive OPERATING ROOM

approach of mandatory exploration for all penetrating
trauma to an extremity.8,9 The application of this policy to
civilian injuries, however, resulted in negative exploration The availability of OR personnel familiar with emergency
rates as high as 84 per cent in cases of penetrating trauma. vascular procedures, particularly RAAAs, is crucial.
After intense evaluation at our institution, we currently Familiarity with the fundamentals of graft types and vascu-
recommend angiography only in patients with signs of vas- lar instrumentation is necessary for an expeditious oper-
cular injury such as a pulse deficit, bruit or an ankle:brachial ation. With level 1 trauma designation, anaesthesiologists
index of less than 1.00.8–10 and OR staff are present in the hospital 24 hours a day.
Colour-flow duplex (CFD) is gaining widespread Operating room staff not only include OR nurses and tech-
acceptance as an initial tool in the diagnostic work-up of nicians but also perfusionists in cases where significant
vascular trauma. Being non-invasive and lacking side effects haemorrhage is anticipated. Similarly, an OR, with staff
makes it an extremely attractive tool in evaluating patients competent in most areas of management of surgical emer-
with potential vascular injuries. With improvements in gencies, should be accessible 24 hours a day. During the pro-
resolution as well as in operator skills, many authors have cedure, care should be taken to ensure maintenance of core
proposed CFD as a replacement or supplement to angio- body temperature by warming the OR and by using warm
graphy. Reported sensitivities and specificities have ranged fluid intravenously or intracorporeally. The anaesthesia
from 50 to 100 per cent and 99 to 100 per cent, respectively, service should also be well versed in the care of the critically
when compared with angiography.11–13 We are currently ill patient, able to intubate patients with full stomachs safely
evaluating the role of CFD in the management of patients and, in hypotensive patients, provide rapid vascular access
with extremity trauma at our institution, accepting that using large bore central venous lines and arterial lines (see
angiography still remains the gold standard. Chapter 7B). In the OR such staffing and policies expedite
In those instances where an RAAA is being considered, surgical treatment of vascular emergencies.
an expedient history and physical examination are neces-
sary. Currently emergency room physicians and trauma Conclusions
surgeons have the capability of performing fast and rapid
real time ultrasound assessments of the abdominal aorta in What has been presented above is the ideal set up for
those patients in whom the suspicion of rupture is low but emergency vascular services. In our own institution, ‘level 1’
nevertheless has to be excluded. By the same token ultra- trauma centre designation has made such provisions pos-
sound assessment can also provide rapid confirmation of sible. Our medical centre is a very large busy public facility
aneurysm rupture as the patient moves to the operating with the experience of a huge number of ER admissions
room (OR). and emergency vascular procedures. Even in a city as pop-
Immediate and 24-hour availability of computed tomo- ulous as Los Angeles, however, our type of facility is the
graphic (CT) scanning and angiography are essential com- exception and not the norm. Hospitals which are smaller
ponents in the management of vascular emergencies. and without trauma centre designation have difficulty
Technological advances with helical CT scanning configur- providing and executing care for the critically ill vascular
ations have shortened scanning times while providing patient. In these instances, local community hospitals may
enhanced resolution and reconstruction capabilities. At attempt to transfer critical vascular patients from their
the LAC  USC Medical Centre, 24-hour availability of ER to a hospital facility with 24-hour availability of the
these services provides prompt scanning and feedback on type of resources cited above. Transfers of this nature do
the presence or absence of many vascular disorders. occur but only sporadically due to a variety of factors, the
The need for a second surgeon to assist the vascular sur- most important of which is the absence of a recognised
geon during a complex vascular reconstruction cannot be system facilitating transfer. This trauma centre, overseen
underestimated. In those instances where the surgeon is by the ACS, is a successful example of such a system.
working within a university construct with resident train- Expansion of this concept to include vascular surgical
ing, which may be in general surgery or a combination of emergencies as well as other acute life-threatening condi-
general surgery and vascular surgery, valuable surgical tions would substantially improve outcomes.
assistance should readily be available 24 hours a day. For
those surgeons in a community setting, however, the need
for assistance in the operating room is critical when faced
with an entity such as an RAAA. Support from a surgical Key references
partner, trained physician’s assistant, or a certified nurse
Committee on Trauma. Resources for the optimal care of the injured
first assistant is invaluable. A call schedule ensuring the
patient: 1999. Chicago, IL: American College of Surgeons
availability of such personnel within the community for
1999; 23–26.
support is essential.
 References 15

Committee on Trauma. Resources for the optimal care of the injured 5 Department of Health and Human Services, Office of Health
patient: 1999. Chicago, IL: American College of Surgeons Research, Statistics and Technology. Detailed Diagnosis and
1999; 43–46. Surgical Procedures for Patients Discharged from Short Stay
Graves EJ. Detailed Diagnosis and Procedures, National Hospital Hospitals. Hyattsville, MD: National Center for Health Statistics,
Discharge Survey: National Center for Health Statistics. Vital 1982; 30.
Health Stat 1989; 100: 72. 6 Committee on Trauma. Resources for Optimal Care of the Injured
Schwartz MR, Weaver FA, Yellin, et al. Refining the indications for Patient: 1999. Chicago, IL: American College of Surgeons 1999;
arteriography in penetrating extremity trauma: a prospective 23–26.
analysis. J Vasc Surg 1993; 17: 166. 7 Committee on Trauma. Resources for Optimal Care of the Injured
Weaver FA, Papanicolaou G, Yellin AE. Difficult peripheral vascular Patient: 1999. Chicago, IL: American College of Surgeons 1999;
injuries. Surg Clin North Am 1996; 76: 843–59. 43–46.
8 Schwartz MR, Weaver FA, Yellin, et al. Refining the indications
for arteriography in penetrating extremity trauma: a prospective
REFERENCES analysis. J Vasc Surg 1993; 17: 166.
9 Weaver FA, Yellin AE, Bauer M, et al. Is arterial proximity a valid
indication for arteriography in penetrating extremity trauma? A
1 Malone JM, Moore WS, Goldstone J, Malone SJ. Therapeutic and prospective analysis. Arch Surg 1990; 125: 1256.
economic impact of a modern amputation program. Ann Surg 10 Weaver FA, Papanicolaou G, Yellin AE. Difficult peripheral
1979; 189: 798–802. vascular injuries. Surg Clin North Am 1996; 76: 843–59.
2 Krupski WC, Skinner HB, Effeney DJ. Amputation. In: Way LW 11 Bynoe RP, Miles WS, Bell RM, et al. Noninvasive diagnosis of
(ed). Current Surgical Diagnosis and Treatment. San Mateo, CA: vascular trauma by duplex ultrasonography. J Vasc Surg 1991;
Appleton and Lange, 1988; 704–14. 14: 346.
3 Bunt TJ, Malone JM. Revascularization or amputation in the over 12 Fry WR, Smith RS, Sayes DV, et al. The success of duplex
70 year old. Am Surg 1994; 60: 349–52. ultrasonographic scanning in diagnosis of extremity vascular
4 Graves EJ. Detailed Diagnosis and Procedures, National Hospital proximity trauma. Arch Surg 1993; 128:1368.
Discharge Survey: National Center for Health Statistics. Vital 13 Meissner M, Paun M, Johansen K. Duplex scanning for arterial
Health Stat 1989; 100: 72. trauma. Am J Surg 1991; 161: 552.
This page intentionally left blank
 2
Pathophysiology of Acute Vascular Insufficiency

AIRES AB BARROS D’SA, DENIS W HARKIN

The problem 17 Clinical assessment of acute vascular insufficiency 24

Aetiology 17 References 25

THE PROBLEM Table 2.1 Potential causes of acute vascular insufficiency

Embolism Cardiac diseases

Acute vascular insufficiency of the limb is a commonly Extracardiac arterial diseases
encountered surgical emergency. Presenting typically in Intrinsic arterial diseases Occlusive atherosclerosis
elderly patients with widespread atherosclerotic disease, it Atherosclerotic aneurysms
threatens not only the viability of the limb but also the life Arteritis and autoimmune diseases
of the patient if treatment is delayed or unsuccessful. The Fibromuscular dysplasia
Haematological disorders Clotting factor deficiencies
incidence of acute limb ischaemia (ALI) has been esti-
Antiphospholipid antibodies
mated to be approximately 14 episodes per 100 000 popu-
Red cell, white cell and globulin
lation and represents 10–16 per cent of the total vascular disorders
workload.1 In most instances the lower limb is affected, Malignant diseases
either de novo or against a background of chronic arterial Heparin induced thrombosis
disease, and therefore failed treatment with subsequent syndrome
amputation carries substantial social and economic costs Oral contraceptive
for the individual and society at large. Despite modern Repetitive external Thoracic outlet syndrome
advances in investigation and treatment, patients present- trauma Athletic injuries
ing with ALI have a particularly severe short term outlook Axillary crutch injury
in terms of limb loss, with 30-day amputation rates of Trauma to the hands and fingers
Iatrogenic injury Arterial catherisation
10–30 per cent and mortality rates of around 15 per cent.1
Anaesthetic arm blocks
This high mortality reflects the degree of comorbidities and
Intra-arterial injections
associated increased risks from coronary and cerebrovas- Thrombosed axillofemoral bypass
cular events in these patients.2,3 Therefore, when managing Irradiation arteritis
acute vascular insufficiency of the limb, overzealous attempts Ergotism
at revascularisation in the elderly or infirm must be tem- Radial artery fistula steal
pered by the knowledge that they may precipitate systemic Self-inflicted injury Intra-arterial narcotic injection
deterioration and fatality (see Chapters 9A and 9B). Deliberate self-harm
Acute external trauma Blunt injuries
Penetrating injuries
AETIOLOGY
when the presentation is atypical or in a young adult, as
Acute vascular insufficiency of the limb is most commonly listed in Table 2.1. In this section the important features of
caused by thrombosis or embolic occlusion. However, an the commonest aetiologies pertaining to acute vascular
extensive range of aetiologies may be involved, especially insufficiency of the limb are discussed.
18 Pathophysiology of acute vascular insufficiency

Acute thrombosis

A variety of pathological processes may lead to thrombosis Chemokines

in the intact vascular system representing the final com- (MCP-1)
mon pathway to vessel occlusion (see Chapters 15, 16 and Phagooytosis Atherosclerotic
Monocyte (oxLDL) plaque
20). Normal vascular homoeostasis requires integrity of Foam
the vessel wall, a balance between procoagulant and anti- Macrophage cells
coagulant factors in the blood and adequate blood flow.
These factors, individually, may each induce thrombosis,
but more often they act in combination, and for the pur-
Figure 2.1 The pathological development of an atherosclerotic
pose of clarity they are discussed separately as those affect-
plaque. MCP-1, monocyte chemoattractant protein-1
ing the wall, the blood and miscellaneous extrinsic factors.

VESSEL WALL FACTORS PREDISPOSING TO THROMBOSIS

and consequently vessel wall shear stress, also influence
Intrinsic arterial disease, most importantly atherosclerosis, the occurrence and progression of these atherosclerotic
is by far the commonest predisposing factor in acute vas- plaques.10
cular insufficiency and may have a long progressive sub- Acute ischaemic events appear to be related to plaque
clinical phase prior to acute presentation. Also discussed rupture rather than progressive plaque enlargement and
here are the much less common cystic medial necrosis and stenosis.11 The vulnerability of a plaque is dependent on a
inflammatory arteritis. variety of factors including the size of the lipid pool, the
thickness of fibrous cap, the content and metabolic activity
Atherosclerosis and atherothrombosis of lipids12 and finally the activity of macrophages and matrix
Atherosclerosis, previously thought to be the preserve of metalloproteinases.13 Disruption of advanced plaques with
old age, is now recognised in increasingly younger individ- exposure of the highly thrombotic lipid core, combined
uals, indeed autopsy studies have demonstrated that some with high intraluminal levels of tissue factor and platelet
degree of lower limb atherosclerosis is present in most of derived vasoconstrictors, combine to trigger intravascular
the adult population.4 By late middle age, almost 8 per cent thrombosis.14
of males demonstrate significant disease on non-invasive
Cystic medial necrosis
testing and 5 per cent experience symptoms related to
This relatively rare pathological process primarily affects
lower limb arterial insufficiency in the form of intermittent
the aorta and is characterised by myxoid accumulations in
claudication.5 More seriously, studies of hospital patients
the aortic media, accompanied by fragmentation of the
with claudication suggest that 25 per cent ultimately require
elastica, the significance being that cystic medial necrosis
bypass surgery or amputation.6,7 In the upper limb a simi-
can predispose to intimal tears and so precipitate aortic
lar pattern exists,8 although symptoms requiring surgery
dissection. This outcome is thought to represent the end-
are less frequent.
point of a variety of injuries but, in particular, is associated
The central pathology in the development of atheroscler-
with severe or prolonged hypertension. Interestingly, it
osis, namely, the accumulation of plasma lipoproteins,
is also observed in Marfan’s disease and may be seen to
particularly low density lipoproteins (LDLs) in the arterial
represent a connective tissue disorder.
wall, brings about gradual narrowing of the vessel lumen
producing vascular insufficiency. These LDLs undergo oxi- Arteritis, autoimmune and associated disorders
dation (oxLDLs) and are believed to generate and release Arteritis includes a range of inflammatory processes, which
products chemotactic to circulating monocytes and smooth affect the arterial system and lead to acute limb ischaemia
muscle cells (SMCs) in the vessel wall. Monocytes attracted (see Chapter 43). These conditions, listed in Table 2.2, are
to the vessel wall then migrate across the endothelium and often broadly subdivided on the basis of the vessel size they
differentiate to macrophages in the tissues, phagocytose the affect, although significant overlap exists. As systemic inflam-
oxLDLs and become lipid laden foam cells as depicted in matory disorders they commonly present as a non-specific
Fig. 2.1. Multiple growth factors, cytokines and other sub- ‘rheumatic-like’ illness with features such as fever, weight
stances produced by endothelial cells, SMCs, macrophages loss, malaise, anaemia, arthralgia, myalgia and arthritis.15
and T lymphocytes encourage cellular accumulation, pro- Temporal (giant cell) arteritis, is the commonest of the
liferation and inflammatory injury.9 The accumulating arterititides and is characterised by focal granulomatous
atherosclerotic plaque, which is covered by a fibrous cap of inflammation of medium and small arteries, chiefly the
connective tissue (type I and III collagen, elastin and pro- cranial vessels. It most commonly presents clinically in the
teoglycans) synthesised by the SMCs, may then extrude temporal arteries of older white women.16 Headache and
into the vessel lumen compromising cross-sectional diam- ocular symptoms predominate and may be accompanied
eter and blood flow. Subsequent alterations in blood flow, by fever, myalgia, and arthralgia, a syndrome known as
 Aetiology 19

Table 2.2 Some causes of acute vascular insufficiency associated Aneurysmal disease and acute arterial dissection
with arteritis or autoimmune diseases Acute vascular insufficiency of the limb may result from
dilating and dissecting arterial diatheses. Aneurysms, the
Disease Vessels involved Distribution vast majority of which are atherosclerotic in nature, but
occasionally may be mycotic, post-traumatic or congenital,
Takayasu’s disease Aorta, arteries Extremities, cause acute vascular insufficiency due to thrombosis,
head and neck,
embolism and rupture. Aneurysms must be considered a
viscera
Buerger’s disease Arteries, veins, Extremities
potential source of acute or chronic embolism comprom-
nerves ising the distal circulation, especially if a cardiac source is
Giant cell arteritis Muscular arteries Cranial arteries not identified. Moreover, acute intra-abdominal catas-
Polyarteritis nodosa Muscular arteries Any extra-pulmonary trophes such as ruptured abdominal aortic aneurysm and
site, viscera the resultant low flow state may be confused with primary
Hypersensitivity Small venules, All organs and acute limb ischaemia, therefore systemic assessment of the
angiitis capillaries, tissues patient is essential in all cases. The commonest peripheral
arterioles aneurysm, that of the popliteal artery, constitutes a limb-
Kawasaki’s disease Arteries Coronary arteries threatening lesion in that it can be the site of acute thrombo-
embolism or the source of chronic low grade embolism
which progressively occludes the distal arterial network.20
polymyalgia rheumatica. Takayasu’s disease is an idio- Acute aortic dissections often cause life-threatening
pathic systemic inflammatory disease typically involving ischaemia of peripheral arterial territories and the vital end
large arteries such as the aorta and its main branches, and organs they supply. Stanford type B dissections, which com-
sometimes the coronary and pulmonary arteries. Typically mence beyond the left subclavian artery origin and extend
it affects children and young adults, with a marked female distally, affect the iliac vessels either directly by occlusion
preponderance and geographically more commonly seen of their true lumen or indirectly when a dissecting flap closes
in the Far East.17 Granulomatous vasculitis progresses to off the origin of the artery. Compromise of the true aortic
medial fibrosis with resultant stenosis or occlusion. The lumen by an expanding high pressure false lumen may lead
prognosis is poor, with severe hypertension, retinopathy, to the phenomenon of pseudocoarctation resulting in a
aortic regurgitation and aneurysm formation being the resultant low flow state in the distal organs and limbs. Much
four main complications.18 Kawasaki’s disease (mucocu- more rarely, primary dissections of the iliofemoral seg-
taneous lymph node syndrome) is a disease which is ment may cause acute vascular insufficiency of the limb,
endemic in Japan, but occurs much less commonly in and are usually associated with underlying pathology such
Europe. Typically it affects infants and children and pre- as fibromuscular dysplasia.
sents with fever, lymphadenopathy, skin rash, oral and con-
junctival erythema and, in 20 per cent of cases, it leads to
coronary arteritis. Morbidity and even mortality is associ- BLOOD FACTORS PREDISPOSING TO THROMBOSIS
ated with coronary arteritis, which can lead to aneurysm Hypercoagulability or thrombophilia represents a state
formation, thrombosis and acute myocardial injury. of increased risk of developing arterial or venous throm-
Buerger’s disease or thromboangiitis obliterans is an bosis.21 Normally a delicate balance between the natural pro-
inflammatory occlusive disease of the medium- and small- coagulant and anticoagulant pathways prevents unwanted
sized arteries of the extremity presenting with segmental clot formation but a variety of defects, either inherited or
thrombotic occlusions of multiple distal arteries.19 Its preva- acquired, may disturb that equilibrium and cause acute
lence is high in Japan, Israel, and other Asiatic countries thrombosis. Although most available evidence of throm-
and most commonly affects young males who are smokers. bogenesis pertains to the venous system, the same under-
Angiography reveals the pathognomonic features of tapering lying defects are thought to be responsible for thrombosis
of vessels, abrupt occlusions and classically the corkscrew of the native artery or a bypass graft.
configuration of collaterals. The prognosis for the limbs is
generally poor but their chance of survival is improved in Inherited thrombophilia
those who stop smoking (see Chapter 42). Inherited defects, either in the procoagulant, anticoagulant
Polyarteritis nodosa is a systemic disease characterised or fibrinolytic pathways, as shown in Table 2.3, combined
by necrotising inflammation of the medium- and small- with exposure to common risk factors represent a poten-
sized arteries throughout the body, sparing the pulmonary tially increased risk for thrombosis. For instance, although
circulation. The lesions are usually sharply demarcated and the risk of venous thromboembolism below 40 years of age
often induce thrombosis, causing distal ischaemia. Typically is less than 0.5 per cent in the general population, in certain
affecting young adults, especially men, the condition is families, that risk in the same age group rises to 5 per cent.22
accompanied by fever, malaise and weight loss but when One large study of patients with idiopathic venous throm-
the renal vessels are involved morbidity is severe. bosis found that 18 per cent of affected individuals were
20 Pathophysiology of acute vascular insufficiency

Table 2.3 The known inherited thrombophilias arising from in a variety of hyperviscosity syndromes such as poly-
defects in the procoagulant, anticoagulant or fibrinolytic pathways cythaemia, macroglobulinaemia, cryoglobulinaemia and
sickle cell anaemia.
Type Specific defect

Deficiencies Antithrombin Acquired thrombophilia

Protein C
Protein S • Lupus anticoagulant
Plasminogen • Antiphospholipid antibody
Heparin cofactor II
• Myeloproliferative disorders
Defects Antithrombin • Malignancy
Protein C • Prolonged immobilisation
Protein S
Factor V Leiden (activated protein C
• Postoperative states
resistance)
• Nephrotic syndrome
Dysfibrinogenaemias • Birth control pills
Prothrombin mutation 20210 • The known acquired thrombophilias arising from
Thrombomodulin excessive loss or underproduction of anticoagulant
or fibrinolytic factors or secretion of procoagulant
Elevated Prothrombin
factors
procoagulants Hyperhomocysteinaemia
Plasminogen inhibitor activator type 1
Histidine-rich glycoprotein
Phlegmasia caerulea dolens
In certain instances venous thrombosis may directly induce
acute vascular insufficiency. Phlegmasia caerulea dolens, due
heterozygous for factor V Leiden gene point mutation to massive iliofemoral deep venous thrombosis, may present
(activated protein C resistance) and 1.5 per cent of them with a suddenly painful swollen leg (see Chapter 20). Arterial
were homozygous.23 Much interest has also focused on the inflow may be so acutely compromised, due to the venous
presence of prothrombin mutation 20210, which has been outflow obstruction, that it threatens the vascular integrity of
reported to be present in up to 18 per cent of patients with the limb as is the case with phlegmasia alba dolens.
thrombosis.24 Once initiated, coagulation must be con-
trolled by the fibrinolytic system if widespread intravascu-
MISCELLANEOUS EXTRINSIC FACTORS
lar thrombosis is to be avoided. Defects in precursor
PREDISPOSING TO THROMBOSIS
proteins involved in the generation of plasmin, and indeed
deficiencies in plasminogen and plasmin action, have A variety of extrinsic factors which cause acute or chronic
been identified. Metabolic defects, such as hyperhomocys- vessel injury and luminal compromise may induce arterial
teinaemia, are known to be associated with an increased thrombosis and lead to acute vascular insufficiency. Acute
incidence of arterial disease25 and recently have also been arterial trauma, perhaps the most dramatic extrinsic injury
found to be associated with both arterial and venous caused by either blunt or penetrating injury, is discussed in
thromboembolism.26 detail elsewhere (see Chapter 33). A range of less dramatic
but just as significant chronic injuries, however, may cul-
Acquired thrombophilia
minate in acute vascular insufficiency and are discussed
Several systemic factors, as shown below, are associated
here briefly.
with increased thrombogenicity of the blood, and these
include changes in hormonal metabolism, smoking, hyper- Repetitive external trauma
lipidaemia, hyperglycaemia, disseminated cancer, nephrotic Repetitive vessel injury in a few specific syndromes may
syndrome, trauma, oral contraceptives, pregnancy and the result in acute vascular insufficiency, often against a back-
postpartum state.27 Increased plasma fibrinogen, in particu- ground of chronic symptoms.
lar, is an independent risk factor for thrombotic compli- Thoracic outlet syndrome describes a variable group
cations in patients with atherosclerotic disease28 and of symptoms arising from neurovascular compression
trauma.29 Diabetic patients may have a range of defects from either bone or soft tissue structures in the root of the
which increase blood thrombogenicity, due in part to gly- neck, resulting in a clinical picture of chronic pain, neuro-
cosylation of collagen and proteins, increased levels of logical disturbance and, more rarely, vascular symptoms34
plasma fibrinogen and plasminogen activator inhibitor-130 (see Chapter 41). Furthermore, the accompanying post-
and platelet hyperaggregability.31 Recently increased levels stenotic dilation may progress to an aneurysm and either
of circulating tissue factor antigen have been observed in by embolism or acute thrombosis35 present clinically as
patients with cardiovascular disease32 and consumptive acute arm ischaemia. Thoracic outlet syndrome may have
coagulation disorders.33 Thrombosis is also more common acute vascular manifestations due to scalene muscle trauma
 Aetiology 21

in association with hyperextension arm injuries. Further- Table 2.4 Aetiology of acute arterial embolic occlusion
more, repeated external trauma to the axillary artery from a
shoulder crutch, although rare nowadays, may present as Source Aetiology
acute thrombosis or secondary thrombosis complicating
aneurysmal change.36 Popliteal artery entrapment syn- Cardiac Atrial fibrillation
Myocardial infarction (mural thrombus)
drome describes a developmental defect in which the
Infective endocarditis (mycotic
popliteal artery, and occasionally the popliteal vein, is com- vegetations)
pressed either beneath the medial head of the gastrocnemius Prosthetic valve (thrombus)
muscle, a slip of that muscle or more rarely the popliteus Atrial myxoma
muscle. The most widely accepted classification recognises Paradoxical embolus (septal defect)
five types of popliteal entrapment.37 It may be responsible Major vessels Atherosclerotic plaque
for up to 40 per cent of cases in those patients aged less than Proximal aneurysm
30 years presenting with lower limb claudication.38 Mycotic aneurysm
Iatrogenic Arterial catheterisation
Traumatic vessel injury Postangioplasty/stenting
Major trauma resulting in acute vascular insufficiency may Intra-arterial injection
be blunt or penetrating and the vessel injury can range
from subclinical intimal injury to complete vessel disrup-
tion. Compounding factors often include associated arterial pressure cannot withstand even small increases in
multisystem injury, haemorrhage, exposure and delayed compartment pressure.
presentation.39 Although acute traumatic vessel injury is Finally, exertional compartment syndromes may arise
dealt with in subsequent chapters, a few specific situations due to severe or unaccustomed exercise, most commonly
require further mention here. involving the lower extremity and, in particular, the anter-
The incidence of iatrogenic injury has matched the ior compartment of the lower leg. Rarely, an exertional
increasing use of invasive monitoring and endovascular compartment syndrome may present as an emergency.
therapy. Catheterisation of a peripheral artery in the critically
ill is standard practice40 and responsible for acute vascular
insufficiency in up to 4 per cent of cases,41 while bleeding Embolic disease
and pseudoaneurysm formation may occur, especially if
coagulopathy coexists. The increasing use of larger diameter Embolic disease is the second commonest cause of acute
catheters and devices such as intra-aortic balloon pumps has vascular insufficiency of the limb. An embolus is a blood
inevitably heralded a mounting frequency of iatrogenic clot or other foreign body which is formed in or gains
problems in the femoral area (see Chapters 38 and 39). access to the vascular system in one location and is carried
Endovascular therapy such as angioplasty may produce by the flow of blood to another site where it produces vas-
large areas of intimal–medial dehiscence, commonly at the cular obstruction (see Chapters 15, 16 and 21). The lower
site of junction between plaque and disease-free wall.42 extremity vessels are involved approximately five times as
However, only rarely do these dissections result in acute frequently as the upper extremities.46 The commonest lower
vessel closure, and if identified early can be remedied by limb target site is the common femoral bifurcation, which
additional angioplasty or stent placement (see Chapter 39). is also the commonest overall site accounting for 35–50
Angioplasty also significantly stimulates the production of per cent of instances, whereas the commonest upper limb site
hydroperoxy acids, which are known to mediate vessel spasm is the brachial artery.47 The majority of emboli originate in
and inflammation, in turn leading to arterial thrombosis.43 the heart, either in patients with atrial fibrillation or in
Osseofascial compartment hypertension (compart- those with mural thrombosis complicating recent myocar-
ment syndrome) describes a self-propagating cycle, which dial infarction. Less common predisposing cardiac causes
occurs within the confines of the osseofascial compartments are rheumatic heart disease, aortic valve prosthesis, mycotic
of the extremities (see Chapter 33). This syndrome may emboli from subacute bacterial endocarditis and, rarely,
complicate limb trauma, with or without arterial injury, as atrial myxoma (Table 2.4). A significant proportion of
well as after reconstructive arterial surgery for severe or arterial emboli originate from primary arterial diseases
prolonged ischaemia.44 Following a period of ischaemia, proximal to their site of impaction,47 and they include ath-
reperfusion causes capillary endothelial injury which leads erosclerosis, aortitis and aneurysmal disease.
to interstitial oedema and an increase in compartment Once present, the clinical outcome of any embolic event
pressure, further compounding tissue hypoxia. Untreated depends primarily upon the territory of the vessel involved,
it may lead to permanent neurovascular damage, myoglo- the completeness of obstruction and the collateral circula-
binuria, renal failure, sepsis and death.45 Patients with tion available. Skeletal muscle is relatively resistant to
peripheral vascular disease are much more susceptible to ischaemic injury but periods beyond 6 hours will usually
injury from compartment syndrome as their lower resting result in progressive permanent injury. The combination of
22 Pathophysiology of acute vascular insufficiency

a paralysed and insensate limb and the onset of ischaemic Ischaemic injury
neuropathy are ominous signs and demand immediate
revascularisation if the limb is to be salvaged. The early use Adenosine triphosphate (ATP), essential in nearly all body
of therapeutic anticoagulation with heparin helps prevent processes requiring the output of energy, is depleted dur-
secondary clotting or clot propagation, a phenomenon ing the period of ischaemia,53 and persistently low tissue
which causes significant deterioration in untreated embolic levels of ATP can lead directly to cellular death.54 During
occlusions. Furthermore, the clinician must be alert to the the ischaemic period depletion of ATP is associated with
fact that embolic events are often multiple and may involve the conversion of xanthine dehydrogenase to xanthine oxi-
several arterial territories. Occult compromise of the dase, provoking elevations in purine metabolites, hypox-
splanchnic circulation must be suspected early in a patient anthine and xanthine, an environment richly conducive to
who deteriorates systemically if an adverse outcome is to the production of reactive oxygen species.55,56 With more
be avoided. A combination of age, high myocardial risk prolonged periods of ischaemia severe fibre damage con-
and late presentation contribute to high rates of amputa- sisting of myofibrillar derangement, loss of organisation of
tion and mortality.46,48 myofilaments or frank myonecrosis ensues.57

Graft failure Oxidative injury

Failure of an existing arterial bypass graft is a common cause Reperfusion of ischaemic muscle tissue initiates a complex
of acute vascular insufficiency. Most patients with graft sequence of events which includes the generation of react-
occlusion present with a recurrence of previous symptoms ive oxygen metabolites, liberation of proinflammatory
and a smaller number are identified by routine graft sur- mediators and granulocyte adherence, infiltration and acti-
veillance;49 unfortunately, a significant proportion go on vation, which result in microvascular and parenchymal cell
to develop critical ischaemia. Factors within the wall, in the injury.58 The influx of a large amount of molecular oxygen
blood and extrinsic factors may all contribute to graft fail- at reperfusion generates the production of oxygen-derived
ure. Graft failure can be broadly categorised as immediate free radicals (OFR) which play an important role in indu-
(within 30 days), intermediate (between 1 and 18 months), cing postischaemic tissue injury. Oxygen free radicals,
and late (after 18 months). Graft failures in the immediate including superoxide anions (O 2 ), hydrogen peroxide and
phase account for 5–20 per cent of cases, and most failures hydroxyl radicals (OH), arise as byproducts of the enzyme
can be attributed to imperfections in surgical technique xanthine-xanthine oxidase system, the mitochondrial elec-
and in patient selection. Occasionally, in low-flow states, tron transport system, and the nicotinamide adenosine
in situ thrombosis reflects the thrombogenicity of the dinucleotide phosphate (reduced form) (NADPH) oxidase
graft. Intermediate phase graft failures, accounting for 5–10 system of neutrophils. These oxygen-free radicals have a
per cent of cases, are largely caused by neointimal hyper- direct lytic effect on cellular membranes through lipid per-
plasia. Late failures account for the remaining cases and oxidation. Endothelial cell membrane injury is associated
are usually the result of progression of atherosclerotic dis- with increased microvascular permeability,59 indeed leg
ease of inflow and outflow vessels (see Chapter 17). oedema is a common complication after reconstructive
vascular surgery.60 Evidence from animal experiments shows
PATHOPHYSIOLOGY OF ACUTE LIMB ISCHAEMIA that skeletal muscle injury following ischaemia is attenu-
ated by hypoxic reperfusion rather than by oxygenated
Skeletal muscle ischaemia followed by reperfusion produces
blood,61 or by the use of free radical scavengers.62
several functional and morphological changes including
impaired ability to develop tension, mitochondrial swelling,
ANTIOXIDANT DEFENCE AND OXIDATIVE STRESS
disruption of sarcomere organisation, leakage of cytosolic
enzymes into the circulation, endothelial cell swelling and Skeletal muscle has several endogenous defences against
denudation, and the development of the ‘no-reflow’ phenom- oxidative injury caused by reactive oxygen metabolites,
enon. In sublethal skeletal muscle ischaemia-reperfusion including, but not limited to the intracellular enzymes super-
(I/R) injury most of the tissue abnormalities occur not oxide dismutase (SOD), catalase (CAT) and glutathione
during ischaemia but on reperfusion.50 Fortunately, skeletal peroxidase. Superoxide dismutase catalyses the dismuta-
muscle is relatively resistant to ischaemic injury for a time tion of superoxide and hydrogen peroxide to water and
but when it exceeds 2–7 hours,51 the opportunity for sal- molecular oxygen, while glutathione peroxidase can also
vage diminishes. Chronically ischaemic limbs with peri- reduce hydrogen peroxide by catalysing its reaction with
pheral vascular disease will tolerate substantially longer reduced glutathione (GSH) to form oxidised glutathione
periods of ischaemia, in part due to collateralisation but also disulphide and water. The protective effect of exogenous
due to the poorly understood phenomenon of ischaemic SOD and catalase in attenuating ischaemia-reperfusion
tolerance, a sustained ischaemic preconditioning effect, induced skeletal muscle injury has been demonstrated.63,64
which, of course, when induced acutely, can be shown to Unfortunately, to be clinically effective they must be given
prevent reperfusion injury and its systemic effects.52 prior to the ischaemic period, which limits their clinical
 Aetiology 23

Na

Lysosomal enzymes
Ca2
(collagenase, elastase,
Margination cathespin G,
Rolling Degranulation proteoglycanase) H
Adhesion Migration
K
Sarcolemmal
Endothelial distribution
activation Leucocyte Membrane Myoglobin,
activation Reactive oxygen species bound O2 CPK

(O2 , H2O2, OH ) NADPH
Chemotactic Free radicals GOT, GPT,
Lysosomal enzymes
STIMULUS signal oxidase (O •
2 , OH , H2O2) LDH
(collagenase, cathespin,
myeloperoxidase)
Prostaglandins
ACTIVATED SKELETAL
Figure 2.2 The interaction between circulating neutrophils and NEUTROPHIL MUSCLE CELLS
the endothelial walls leads to transendothelial migration and Chemotactic factors
(Leucotrienes, C3a, C5a)
inflammatory tissue injury
Figure 2.3 The process of oxidative cellular injury by an
usefulness.65 Antioxidant vitamins, such as C and E, may activated neutrophil chemoattracted to an ischaemia-reperfusion
also act as free radical scavengers, and have been shown to injured monocyte. CPK, creatine phosphate kinase; GOT, glutamic
prevent experimental ischaemia-reperfusion injury when oxaloacetic transferase; GPT, glutamic pyruvic transferase; LDH,
given prior to the ischaemic injury.66 lactose dehydrogenase; NADPH, nicotinamide adenosine
dinucleotide phosphate (reduced form)
POLYMORPHONUCLEAR LEUCOCYTE–ENDOTHELIAL
INTERACTION postischaemic tissue the accumulation of PMN is associated
with severe morphological injuries manifested as large and
The influx of activated polymorphonuclear leucocytes numerous intramitochondrial dense bodies, mitochondrial
(PMNs) into postischaemic tissues is one of the hallmarks swelling and intermyofibillar oedema.71 While adhering to
of ischaemia-reperfusion injury. Polymorphonuclear leuco- the endothelium, neutrophils damage endothelial cells
cyte adherence to the endothelium is a requirement for and the basement membrane by releasing superoxide anions
these cells to alter the endothelial barrier and so to allow via the NADPH oxidase system, secreting myeloperoxidase
PMN transmigration to the site of injury. The process of which catalyses the production of hypochlorus acid (HOCL),
cell adherence, cell activation and cell migration involves and releasing granular enzymes including elastase, collage-
an interplay between the expression of adhesion molecules nase and cathepsin G,72 as shown in Fig. 2.3. Enhanced
by the endothelial cell, leucocyte activation and local neutrophil adhesion to endothelium is seen in postcapillary
cytokine activity67,68 as illustrated in Fig. 2.2. venules during ischaemia with a more pronounced response
Sequestration of circulating neutrophils in the micro- following reperfusion.73
vessels of skeletal muscle begins with constitutively
expressed neutrophil L-selectin receptors binding to the
‘NO-REFLOW’ PHENOMENON
postcapillary venule endothelial P-selectin and E-selectin
receptors. This is the first phase of neutrophil adhesion, and After an extended period of ischaemia in skeletal muscle,
is manifested by ‘neutrophil rolling’ along the venular lumi- some capillaries fail to perfuse on reinstitution of blood
nal surface. The second phase of neutrophil adhesion occurs flow.74,75 This ‘no-reflow’ phenomenon results in incom-
after activation or ‘upregulation’ of the CD11/CD18 inte- plete and maldistributed perfusion and contributes to
grins and shedding of L-selectin from the neutrophil sur- injury by prolonging hypoxia and exposure of the tissues to
face, caused by release of proinflammatory cytokines, toxic metabolites.76
platelet activating factor (PAF) and eicosanoids from the Maldistribution of blood flow due to the failure of local
damaged endothelial cell. This halts neutrophil rolling by vascular autoregulation is understood to be the cause of the
providing strong neutrophil–endothelial cell adhesion ‘no-reflow’ phenomenon noted after ischaemic injury and
through the interaction of the CD11b/CD18 integrin with tends to compound ischaemic damage in underperfused
the endothelial receptor, intercellular adhesion molecule-1 areas. This maldistribution of flow is not simply based on
(ICAM-1). Finally transmigration occurs, which involves vasoconstriction but represents a failure of vascular smooth
both chemotactic stimuli, such as interleukin-8, and muscle responsiveness to the usual vasoactive mediators
binding to the platelet–endothelial cell adhesion molecule-1 such as PO2, PCO2, pH, lactate, potassium, adenine nucle-
(PECAM-1).69 Attenuation of neutrophil accumulation in otides, bradykinin, prostaglandins, nitric oxide and changes
postischaemic tissues using antibodies directed against the in osmolality.77
neutrophil surface adherence complex in skeletal muscle70 Mechanical obstruction of small vessels due to entrap-
has been successfully demonstrated. After reperfusion of ment of enlarged adherent leucocytes ‘plugging’ the capillary
24 Pathophysiology of acute vascular insufficiency

lumen74 may also contribute to the ‘no-reflow’ phenom- Table 2.5 The Society for Vascular Surgery/International
enon. Oxidative endothelial damage leads to endothelial Society of Cardiovascular Surgery (SVS/ISCVS) grading system
for limb ischaemia
cell swelling which reduces the diameter of the vascular
lumen adding to vascular resistance70 and capillary neu-
Grade Description Treatment
trophil plugging.78 The adhesion of neutrophils to the
venular endothelial cells in turn increases capillary pres-
Class I Viable leg, without Heparin is given and
sure and fluid filtration, through an already damaged impairment of sensory elective treatment,
endothelial barrier, compounding interstitial oedema and or motor function and either conservative or
compression of the microcirculatory system.58 audible Doppler signals interventional, is
arranged
UPREGULATION OF GENE EXPRESSION Class IIa Marginally threatened A delay of 9 hours is
with symptoms limited acceptable to attempt
As our understanding of the intracellular events during to a mild sensory loss treatment by lysis
ischaemia-reperfusion injury improves with advances in (usually in the toes)
molecular biology, we shall soon understand why different Class IIb Immediately at risk Delay is unacceptable
tissues have such disparate tolerances to ischaemic injury. with pronounced and urgent clot
Early work has confirm enhanced gene expression or ‘upreg- sensory loss, mild to extraction by aspiration
ulation’ during ischaemia and perhaps more importantly moderate motor loss or embolectomy is
in early reperfusion.79 Experimental work is already in but audible Doppler required
progress exploring the benefits of new genetic therapies in signals Accelerated thrombolysis
skeletal muscle ischaemia-reperfusion injury.80 Improved may be used
understanding of the role of these genes and their sequence Class III Absent Doppler flow, Attempts to restore
of upregulation may well open a new arena for therapeutic paralysis, total blood flow would lead
intervention. sensory loss and to hyperkalaemia and
irreversible tissue myoglobinuria, so
damage delayed amputation
Systemic response to reperfusion should be performed
after resuscitation
Reperfusion of the acutely ischaemic limb not only causes
local effects but, with revascularization, also carries a var-
iety of proinflammatory stimuli into the systemic circula- guidance because the treatment for each patient has to be
tion. Products of oxidative injury, namely, reactive oxygen individualised by an experienced vascular surgical team.
species, arachidonic acid metabolites and hydroperoxides,
interact with activated endothelial cells and leucocytes in Conclusions
an environment rich in proinflammatory cytokines. These
proinflammatory mediators and activated leucocytes initi- It is clear that acute vascular insufficiency is a common
ate an intravascular cascade of pro- and anti-inflammatory surgical emergency in a high risk patient population
responses resulting in remote vital organ injury.81 This inherently predisposed to cardiovascular and cere-
systemic inflammatory response syndrome (SIRS), which brovascular risk. Despite advances in management it is a
may lead to multiple organ dysfunction syndrome (MODS), condition which continues to be associated with high
is dealt with in Chapter 4. amputation and mortality rates. Diagnosis and treat-
ment require a multidisciplinary approach involving
interventional radiologists, haematologists, immunolo-
CLINICAL ASSESSMENT OF ACUTE gists, clinical biochemists and vascular technicians under
VASCULAR INSUFFICIENCY the supervision of an experienced vascular surgeon.
When revascularisation is indicated the extent and dur-
ation of ischaemia time is the major determinant of out-
The clinician must interpret the patient’s history and come. The clinician must also recognise the settings in
clinical signs to determine the extent and duration of the which revascularisation is detrimental to the patient and
underlying injury, the feasibility of revascularisation and in these instances it may be wiser to adopt a suitably pal-
the wisdom of doing so in a particular setting. A variety of liative approach. The rapid advances in endovascular
clinical grading systems aid the clinician in the decision therapies and our increasing understanding of genetic
making process, perhaps the most widely accepted being and environmental factors predisposing to acute vascu-
that adopted by the Society for Vascular Surgery and the lar insufficiency will combine to ensure that this
International Society of Cardiovascular Surgery (SVS/ISCVS) remains an interesting and evolving clinical field.
and given in Table 2.5. These systems, at best, only provide
 References 25

15 Hoffman GS. Takayasu arteritis: lessons from the American

Key references National Institutes of Health experience. Int J Cardiol 1996;
54(suppl): S99–S102.
Campbell WB, Ridler BM, Szymanska TH. Current management of 16 Paice EW. Giant cell arteritis: difficult decisions in diagnosis,
acute leg ischaemia: results of an audit by the Vascular investigation and treatment. Postgrad Med J 1989; 65: 743–7.
Surgical Society of Great Britain and Ireland. Br J Surg 1998; 17 Rizzi R, Bruno S, Stellacci C, Dammacco R. Takayasu’s arteritis: a
85: 1498–503. cell-mediated large-vessel vasculitis. Int J Clin Lab Res 1999;
Carden DL, Smith JK, Korthuis RJ. Neutrophil-mediated 29: 8–13.
microvascular dysfunction in postischemic canine skeletal 18 Morales E, Pineda C, Martinez-Lavin M. Takayasu’s arteritis in
muscle. Role of granulocyte adherence. Circ Res 1990; 66: children. J Rheumatol 1991; 18: 1081–4.
1436–44. 19 Shionoya S. Buerger’s disease: diagnosis and management.
Dormandy J, Heeck L, Vig S. The natural history of claudication: risk Cardiovasc Surg 1993; 1: 207–14.
to life and limb. Semin Vasc Surg 1999; 12: 123–37. 20 Dawson I, Sie RB, van Bockel JH. Atherosclerotic popliteal
Smith JK, Carden DL, Korthuis RJ. Role of xanthine oxidase in aneurysm. Br J Surg 1997; 84: 293–9.
postischemic microvascular injury in skeletal muscle. Am J 21 Miletich JP. Thrombophilia as a multigenic disorder. Semin
Physiol 1989; 257: H1782–H1789. Thromb Hemost 1998; 24(suppl 1): 13–20.
Smith JK, Grisham MB, Granger DN, Korthuis RJ. Free radical 22 Koster T, Rosendaal FR, Briet E, et al. Protein C deficiency in a
defense mechanisms and neutrophil infiltration in controlled series of unselected outpatients: an infrequent but
postischemic skeletal muscle. Am J Physiol 1989; 256: clear risk factor for venous thrombosis (Leiden Thrombophilia
H789–H793. Study). Blood 1995; 85: 2756–61.
23 FR, Koster T, Vandenbroucke JP, Reitsma PH. High risk of
thrombosis in patients homozygous for factor V Leiden (activated
REFERENCES protein C resistance). Blood 1995; 85: 1504–8.
24 Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common
genetic variation in the 3-untranslated region of the prothrombin
1 Dormandy J, Heeck L, Vig S. Acute limb ischemia. Semin Vasc gene is associated with elevated plasma prothrombin levels and
Surg 1999; 12: 148–53. an increase in venous thrombosis. Blood 1996; 88: 3698–703.
2 Dormandy J, Heeck L, Vig S. The natural history of claudication: 25 De S, V, Finazzi G, Mannucci PM. Inherited thrombophilia:
risk to life and limb. Semin Vasc Surg 1999; 12: 123–37. pathogenesis, clinical syndromes, and management. Blood
3 Leng GC, Fowkes FG, Lee AJ, et al. Use of ankle brachial pressure 1996; 87: 3531–44.
index to predict cardiovascular events and death: a cohort 26 den Heijer M, Koster T, Blom HJ, et al. Hyperhomocysteinemia
study. BMJ 1996; 313: 1440–4. as a risk factor for deep-vein thrombosis. N Engl J Med 1996;
4 Sternby NH. Atherosclerosis in a defined population. An autopsy 334: 759–62.
survey in Malmo, Sweden. Acta Pathol Microbiol Scand 1968; Suppl. 27 Fowkes FG, Housley E, Riemersma RA, et al. Smoking, lipids,
5 Fowkes FG, Housley E, Cawood EH, et al. Edinburgh Artery Study: glucose intolerance, and blood pressure as risk factors for
prevalence of asymptomatic and symptomatic peripheral arterial peripheral atherosclerosis compared with ischemic heart disease
disease in the general population. Int J Epidemiol 1991; 20: 384–92. in the Edinburgh Artery Study. Am J Epidemiol 1992; 135:
6 Fowkes FG. The measurement of atherosclerotic peripheral 331–40.
arterial disease in epidemiological surveys. Int J Epidemiol 1988; 28 Yarnell JW, Baker IA, Sweetnam PM, et al. Fibrinogen, viscosity,
17: 248–54. and white blood cell count are major risk factors for ischemic
7 Fowkes FG. Epidemiology of atherosclerotic arterial disease in heart disease. The Caerphilly and Speedwell collaborative heart
the lower limbs. Eur J Vasc Surg 1988; 2: 283–91. disease studies. Circulation 1991; 83: 836–44.
8 Bergqvist D, Ericsson BF, Konrad P, Bergentz SE. Arterial surgery 29 Dalmon J, Laurent M, Courtois G. The human beta fibrinogen
of the upper extremity. World J Surg 1983; 7: 786–91. promoter contains a hepatocyte nuclear factor 1-dependent
9 Varela O, Martinez-Gonzalez J, Badimon L. The response of interleukin-6-responsive element. Mol Cell Biol 1993; 13:
smooth muscle cells to alpha-thrombin depends on its arterial 1183–93.
origin: comparison among different species. Eur J Clin Invest 30 Meigs JB, Mittleman MA, Nathan DM, et al. Hyperinsulinemia,
1998; 28: 313–23. hyperglycemia, and impaired hemostasis: the Framingham
10 Turitzto VT, Hall CL. Mechanical factors affecting hemostasis Offspring Study. JAMA 2000; 283: 221–8.
and thrombosis. Thromb Res 1998; 92: S25-S31. 31 Rauch U, Crandall J, Osende JI, et al. Increased thrombus
11 Badimon L, Meyer BJ, Badimon JJ. Thrombin in arterial formation relates to ambient blood glucose and leukocyte count
thrombosis. Haemostasis 1994; 24: 69–80. in diabetes mellitus type 2. Am J Cardiol 2000; 86: 246–9.
12 Felton CV, Crook D, Davies MJ, Oliver MF. Relation of plaque lipid 32 Soejima H, Ogawa H, Yasue H, et al. Heightened tissue factor
composition and morphology to the stability of human aortic associated with tissue factor pathway inhibitor and prognosis in
plaques. Arterioscler Thromb Vasc Biol 1997; 17: 1337–45. patients with unstable angina. Circulation 1999; 99: 2908–13.
13 Libby P. Molecular bases of the acute coronary syndromes. 33 Gando S, Nanzaki S, Sasaki S, Kemmotsu O. Significant
Circulation 1995; 91: 2844–50. correlations between tissue factor and thrombin markers in
14 Fernandez-Ortiz A, Badimon JJ, Falk E, et al. Characterization of trauma and septic patients with disseminated intravascular
the relative thrombogenicity of atherosclerotic plaque coagulation. Thromb Haemost 1998; 79: 1111–15.
components: implications for consequences of plaque rupture. 34 Roos DB. The thoracic outlet syndrome is underrated. Arch
J Am Coll Cardiol 1994; 23: 1562–9. Neurol 1990; 47: 327–8.
26 Pathophysiology of acute vascular insufficiency

35 Scher LA, Veith FJ, Samson RH, et al. Vascular complications of 57 Kuzon WM Jr, Walker PM, Mickle DA, et al. An isolated skeletal
thoracic outlet syndrome. J Vasc Surg 1986; 3: 565–8. muscle model suitable for acute ischemia studies. J Surg Res
36 Lee AW, Hopkins SF, Griffen WO Jr. Axillary artery aneurysm as 1986; 41: 24–32.
an occult source of emboli to the upper extremity. Am Surg 58 Korthuis RJ, Granger DN, Townsley MI, Taylor AE. The role of
1987; 53: 485–6. oxygen-derived free radicals in ischemia-induced increases in
37 Rich NM, Collins GJ Jr, McDonald PT, et al. Popliteal vascular canine skeletal muscle vascular permeability. Circ Res 1985;
entrapment. Its increasing interest. Arch Surg 1979; 114: 57: 599–609.
1377–84. 59 Kupinski AM, Shah DM, Bell DR. Permeability changes following
38 Hamming JJ, Vink M. Obstruction of the popliteal artery at an ischemia-reperfusion injury in the rabbit hindlimb. J Cardiovasc
early age. J Cardiovasc Surg (Torino) 1965; 6: 516–24. Surg (Torino) 1992; 33: 690–4.
39 Barros D’Sa AAB. Twenty five years of vascular trauma in 60 Persson NH, Takolander R, Bergqvist D. Lower limb oedema after
Northern Ireland. BMJ 1995; 310: 1–2. arterial reconstructive surgery. Influence of preoperative
40 Clark VL, Kruse JA. Arterial catheterization. Crit Care Clin 1992; ischaemia, type of reconstruction and postoperative outcome.
8: 687–97. Acta Chir Scand 1989; 155: 259–66.
41 Frezza EE, Mezghebe H. Indications and complications of arterial 61 Korthuis RJ, Smith JK, Carden DL. Hypoxic reperfusion attenuates
catheter use in surgical or medical intensive care units: analysis postischemic microvascular injury. Am J Physiol 1989; 256:
of 4932 patients. Am Surg 1998; 64: 127–31. H315–19.
42 Uchida Y, Hasegawa K, Kawamura K, Shibuya I. Angioscopic 62 Smith JK, Grisham MB, Granger DN, Korthuis RJ. Free
observation of the coronary luminal changes induced by radical defense mechanisms and neutrophil infiltration in
percutaneous transluminal coronary angioplasty. Am Heart J postischemic skeletal muscle. Am J Physiol 1989; 256:
1989; 117: 769–76. H789–93.
43 Cragg A, Einzig S, Castaneda-Zuniga W, et al. Vessel wall 63 Reikeras O, Ytrehus K. Oxygen radicals and scavenger enzymes
arachidonate metabolism after angioplasty: possible mediators in ischaemia-reperfusion injury of skeletal muscle. Scand J Clin
of postangioplasty vasospasm. Am J Cardiol 1983; 51: 1441–5. Lab Invest 1992; 52: 113–18.
44 Barros D’Sa AAB. Complex vascular and orthopaedic limb injuries. 64 Ytrehus K, Semb AG, Myhre ES. Ibuprofen abolishes the increase
J Bone Joint Surg Br 1992; 74: 176–8. in leucocyte chemiluminescence observed during ischemic
45 Matsen FA, III, Winquist RA, Krugmire RB Jr. Diagnosis and myocardial failure, but fails to improve hemodynamic function.
management of compartmental syndromes. J Bone Joint Surg Am Basic Res Cardiol 1992; 87: 385–92.
1980; 62: 286–91. 65 Faust KB, Chiantella V, Vinten-Johansen J, Meredith JH. Oxygen-
46 Panetta T, Thompson JE, Talkington CM, et al. Arterial derived free radical scavengers and skeletal muscle
embolectomy: a 34-year experience with 400 cases. Surg Clin ischemic/reperfusion injury. Am Surg 1988; 54: 709–19.
North Am 1986; 66: 339–53. 66 Harkin DW, Barros D’Sa AAB, Yassin MMI, et al. Reperfusion
47 Davies MG, O’Malley K, Feeley M, et al. Upper limb embolus: a injury is greater with delayed restoration of venous outflow in
timely diagnosis. Ann Vasc Surg 1991; 5: 85–7. concurrent arterial and venous limb injury. Br J Surg 2000;
48 Campbell WB, Ridler BM, Szymanska TH. Current management 87: 734–41.
of acute leg ischaemia: results of an audit by the Vascular 67 Adams DH, Shaw S. Leucocyte-endothelial interactions and
Surgical Society of Great Britain and Ireland. Br J Surg 1998; regulation of leucocyte migration. Lancet 1994; 343:
85: 1498–503. 831–6.
49 Moody P, de Cossart LM, Douglas HM, Harris PL. Asymptomatic 68 Davies MG, Hagen PO. The vascular endothelium. A new horizon.
strictures in femoro-popliteal vein grafts. Eur J Vasc Surg 1989; Ann Surg 1993; 218: 593–609.
3: 389–92. 69 Smith CW, Marlin SD, Rothlein R, et al. Cooperative interactions
50 Sjostrom M, Neglen P, Friden J, Eklof B. Human skeletal muscle of LFA-1 and Mac-1 with intercellular adhesion molecule-1 in
metabolism and morphology after temporary incomplete facilitating adherence and transendothelial migration of human
ischaemia. Eur J Clin Invest 1982; 12: 69–79. neutrophils in vitro. J Clin Invest 1989; 83: 2008–17.
51 Karpati G, Carpenter S, Melmed C, Eisen AA. Experimental 70 Carden DL, Smith JK, Korthuis RJ. Neutrophil-mediated
ischemic myopathy. J Neurol Sci 1974; 23: 129–61. microvascular dysfunction in postischemic canine skeletal
52 Harkin DW, Barros D’Sa AAB, McCallion K, et al. Ischemic muscle. Role of granulocyte adherence. Circ Res 1990; 66:
preconditioning before lower limb ischemia–reperfusion protects 1436–44.
against acute lung injury. J Vasc Surg 2002; 35: 1264–73. 71 Formigli L, Lombardo LD, Adembri C, et al. Neutrophils as
53 Hinshaw DB, Armstrong BC, Beals TF, Hyslop PA. A cellular mediators of human skeletal muscle ischemia-reperfusion
model of endothelial cell ischemia. J Surg Res 1988; 44: 527–37. syndrome. Hum Pathol 1992; 23: 627–34.
54 Lindsay TF, Liauw S, Romaschin AD, Walker PM. The effect of 72 Weiss SJ. Tissue destruction by neutrophils. N Engl J Med 1989;
ischemia/reperfusion on adenine nucleotide metabolism and 320: 365–76.
xanthine oxidase production in skeletal muscle. J Vasc Surg 73 Granger DN, Kvietys PR, Perry MA. Leukocyte–endothelial cell
1990; 12: 8–15. adhesion induced by ischemia and reperfusion. Can J Physiol
55 Grisham MB, Hernandez LA, Granger DN. Xanthine oxidase and Pharmacol 1993; 71: 67–75.
neutrophil infiltration in intestinal ischemia. Am J Physiol 1986; 74 Bagge U, Amundson B, Lauritzen C. White blood cell
251: G567–G574. deformability and plugging of skeletal muscle capillaries in
56 Smith JK, Carden DL, Korthuis RJ. Role of xanthine oxidase in hemorrhagic shock. Acta Physiol Scand 1980; 108: 159–63.
postischemic microvascular injury in skeletal muscle. Am J 75 Strock PE, Majno G. Microvascular changes in acutely ischemic
Physiol 1989; 257: H1782–9. rat muscle. Surg Gynecol Obstet 1969; 129: 1213–24.
 References 27

76 Allen DM, Chen LE, Seaber AV, Urbaniak JR. Pathophysiology and 79 Paoni NF, Peale F, Wang F, et al. Time course of skeletal muscle
related studies of the no reflow phenomenon in skeletal muscle. repair and gene expression following acute hind limb ischemia
Clin Orthop 1995; 314: 122–33. in mice. Physiol Genomics 2002; 11: 263–72.
77 Forrest I, Lindsay T, Romaschin A, Walker P. The rate and 80 Brevetti LS, Sarkar R, Chang DS, et al. Administration of
distribution of muscle blood flow after prolonged ischemia. adenoviral vectors induces gangrene in acutely ischemic rat
J Vasc Surg 1989; 10: 83–8. hindlimbs: role of capsid protein-induced inflammation. J Vasc
78 Walden DL, McCutchan HJ, Enquist EG, et al. Neutrophils Surg 2001; 34: 489–96.
accumulate and contribute to skeletal muscle dysfunction 81 Harkin DW, Barros D’Sa AAB, McCallion K, et al. Circulating
after ischemia-reperfusion. Am J Physiol 1990; 259: neutrophil priming and systemic inflammation in limb
H1809–12. ischaemia-reperfusion injury. Int Angiol 2001; 20: 78–89.
This page intentionally left blank
 3
Intestinal Ischaemia in Aortic Surgery

MARTIN G BJÖRCK

The problem 29 Management 34

Incidence and grading of the ischaemic lesion 29 Results 35
Aetiology and pathophysiology 32 References 36
Diagnosis 33

THE PROBLEM aneurysm (AAA),1 its true incidence has long remained
unclear. Differences in case-mix may explain why, during
the very same year of 1960, one group2 reported an inci-
Intestinal ischaemia in the postoperative period after aor-
dence of 10 per cent (12 of 120) while another3 of
toiliac surgery represents a major challenge to the vascular
0.2 per cent (2 of 931); the latter report from Houston of
surgeon. This complication constitutes a lethal threat to
operations on a highly selected group of patients has never
the patient and occurs so frequently that the ability to
been reproduced over the decades right up to the introduc-
manage it will affect the overall outcome.
tion of endovascular repair (EVAR) and illustrates the
The cardinal symptoms, namely, early bloody diarrhoea
continued failure to recognise the existence of colonic
and peritonitis, are more often absent than present. If the
ischaemia. The largest retrospective study from a single cen-
complication is diagnosed merely on the basis of these clin-
tre4 reported an incidence of 1.1 per cent among 2137
ical symptoms and signs over half of these patients would
patients, only 147 (6.8 per cent) of whom required surgery
suffer an unnecessary and potentially lethal delay in treat-
for rupture. The frequency of emergency operations per-
ment. On the other hand, benign mucosal ischaemic lesions
formed and postmortem examinations undertaken, the latter
are common, especially after emergency surgery, and an
being included in only three publications,5–7 has a profound
unnecessary re-laparotomy in such cases would represent a
impact on the reported incidence of the complication.
threat to the patient. Timely recognition and proper grad-
Colonic ischaemia accounts for approximately 95 per cent
ing of the ischaemic lesion, therefore, are essential ingredi-
of cases of intestinal ischaemia complicating aortoiliac sur-
ents for a favourable outcome.
gery.6,8 The resulting ischaemic lesions can be divided into
Understanding the pathophysiology of this condition is
three grades, namely, mucosal ischaemia, mucosal plus
helpful if it is to be prevented. The correlation between
muscular ischaemia and transmural gangrene (Fig. 3.1).9
intestinal ischaemia, ischaemia-reperfusion injury (IRI)
(detailed in Chapter 2), systemic inflammatory response
syndrome (SIRS) and multiple organ dysfunction syndrome Grades of ischaemia
(MODS) (see Chapter 4) and intra-abdominal hypertension
poses a number of questions regarding the general treatment • Grade I – mucosal ischaemia
of patients after major abdominal surgery or trauma. • Grade II – mucosal plus muscular ischaemia
• Grade III – transmural gangrene

INCIDENCE AND GRADING OF THE

Prospective studies, in which all patients were examined
ISCHAEMIC LESION
postoperatively by sigmoidoscopy10–12 or by pHi-guided
sigmoidoscopy,5,13 reported incidences of colon ischaemia
Although colonic ischaemia was reported just a year after of 5–11 per cent after elective AAA surgery and 15–60
the very first successful repair of an abdominal aortic per cent after operation for rupture. Most of these patients,
30 Intestinal ischaemia in aortic surgery

however, had superficial mucosal lesions the clinical The introduction of vascular registries has permitted
relevance of which is a matter of controversy. analyses of the incidence of colonic ischaemia in large
The definition of normal or of pathological appearances non-selected patient populations, undergoing surgery for
at sigmoidoscopy varies. If the findings of diffuse hyper- both elective and ruptured AAA (Table 3.1). In these reports
aemia, submucosal haemorrhagic spots and solitary ero- only those patients with clinically relevant ischaemia were
sions are to be defined as pathological, then no patient included, and most of them (80 per cent) had bowel gan-
undergoing aortic surgery should be regarded as having nor- grene. Patients with grade II lesions often develop
mal colonic mucosa.14 When colonoscopy was combined ischaemic strictures and diarrhoea whereas patients with
with biopsy 1 week after elective aortic surgery (28 for AAA grade III lesions may advance to the clinically problematic
and 28 for occlusive disease) histological signs of colonic manifestation of bloody stools.
ischaemia were observed in 30 per cent of patients.15

(a)

Figure 3.1 (a) Specimen of resected sigmoid colon showing a grade II lesion with distinct differences between normal and ischaemic
mucosa. After a combined operation for abdominal aortic aneurysm (AAA) and renal artery stenosis the patient developed colonic ischaemia
and mucosal gangrene was confirmed on colonoscopy postoperatively. When anuria developed on the fifth day, the patient’s sigmoid colon
was resected. (b) Postmortem specimen of the left colon revealing a grade III lesion with distinct differences between normal and ischaemic
colon at the left flexure. The 80-year-old patient arrived in shock due to a ruptured AAA and 4 hours after the operation developed bloody
diarrhoea; colonic ischaemia was confirmed at sigmoidoscopy and 2 days later the patient died in multiple organ failure
 (b)

Figure 3.1 (Cont’d)

Table 3.1 Incidence of intestinal ischaemia after aortic surgery in vascular registries

Incidence (per cent)

Time Number of Elective

Source Registry period patients Overall operation Rupt. AAA

Amundsen et al.16 All of Norway 1981–1983 444 3.4 2 10

17
Longo et al. VA Registry, USA 1987–1991 4957 1.2 ? ?
Björck et al.6 Swedvasc, Sweden 1987–1993 2930 2.8 1.1 5.9
18
Järvinen et al. Finnvasc, Finland 1991–1993 1752 1.2 0.8 3.1
Norgren* Eurostar 1994–2001 3558 0.2 0.2 –

* L Norgren, personal communication. Data from the Eurostar Registry, August 2001.
32 Intestinal ischaemia in aortic surgery

Incidence after endovascular repair Table 3.2 Independent risk factors for intestinal ischaemia,
identified by multivariate analysis in a cohort of 2824 patients20
Most reports of series of endovascular stent graft repair
Risk factor Relative risk/odds ratio
(EVAR) refer to elective rather than ruptured AAA. The
incidence of bowel ischaemia following open surgery is Blood loss 10 L 6.3
approximately 1 per cent and more than 5000 patients
Patient in shock due to a 3.2
would have to undergo EVAR in order to demonstrate a ruptured AAA
fall in the complication rate to 0.5 per cent.
Ligation of one or both internal 2.6
By August 2001, nine of 3658 patients in the Eurostar
iliac arteries
Registry developed colonic ischaemia, an incidence of 0.2 per
cent (L. Norgren, personal communication; data from the Emergency surgery 2.4
Eurostar Registry, August 2001), which, within 95 per cent Aorto-bifemoral graft 2.4
confidence limits, is lower than the 0.9 per cent (11 cases of Renal insufficiency (creatine 2.3
1160 open elective AAA operations) observed in the Swedvasc 150 mol/L)
Registry.6 Nevertheless, it is important to point out that the Operation at a regional hospital 1.9
Swedvasc data were extensively validated and that the autopsy
rate was 66 per cent; in contrast, such valuable and pertinent Hazard rate
information does not exist in the Eurostar Registry. Age 3.5% per year
Aortic cross-clamp time 1.2% per minute
Importance Operating time 0.9% per minute

The clinical importance of colonic ischaemia in AAA sur-

IMA stump pressures were measured. One patient suffered
gery is evident from an analysis of data available in the
mild, subclinical, ischaemic colitis diagnosed on routine sig-
Swedvasc Registry. We found that 9 per cent of the patients
moidoscopy which resolved completely by the eighth day.
who died after an elective operation and 23 per cent of
This patient had the lowest post-reconstruction stump pres-
those who died after operation of a ruptured AAA suffered
sure of 33 mmHg, and in nine patients pressures of between
this complication.19
40 and 50 mmHg were recorded. From this single observa-
tion far-reaching conclusions were made: it was claimed that
if the IMA had thrombosed or if IMA stump pressure
AETIOLOGY AND PATHOPHYSIOLOGY
exceeded 40 mmHg, ischaemic colitis would not occur and
therefore ligation of the IMA could be safely carried out.
The aetiology of intestinal ischaemia is complex with con- Two other case–control studies13,22 evaluated ligation of
founding factors. For instance, operations for ruptured a patent IMA as a risk factor for intestinal ischaemia and
AAA are associated with prolonged cross-clamping and both concluded that it was not a significant factor. In a
substantial bleeding. We studied 62 patients who suffered prospective study using sigmoid colon pHi measurements,
the complication in a case–control study nested in a cohort patients with an occluded IMA prior to surgery had lower
of 2824 patients.20 For the first time it was possible to study pHi values than those with a patent IMA at all stages, and
the risk factors using multivariate analysis. The independ- the difference became significant on the third post-
ent risk factors identified are shown in Table 3.2. operative day.5 Occlusion of the IMA was associated with
Risk factors, such as preoperative shock or renal insuffi- increasing age, larger aneurysms, longer operations and
ciency are not easily controlled, while others depend on sur- greater volumes of blood loss. In conclusion, occlusion of
gical decision making. Intestinal ischaemia following the IMA prior to surgery is a marker for advanced arterial
surgery was more prevalent in regional than in district hos- disease and there are no data to support reimplantation of
pitals and is probably attributable to the sustained treatment a patent IMA in an attempt to prevent colonic ischaemia.
provided in the former resulting in better survival at 30 days. The fact that the left colon is the segment most liable to
However, no difference in survival was noted between the be affected when bowel ischaemia follows aortoiliac sur-
two groups after 1 year. It appears that in district hospitals gery is explained by the watershed phenomenon. The
staff are more inclined to withdraw treatment when patients combination of aortic cross-clamping, low cardiac output,
deteriorate in the early postoperative period. the low priority given to the colon in terms of critical oxy-
Patients with an occluded inferior mesenteric artery gen delivery and the loss of important collateral pathways
(IMA) had an elevated, though insignificant, risk of devel- due to atherosclerosis or surgical ligation results in the left
oping bowel ischaemia, relative risk (RR)  1.64 (0.62–4.31), colon becoming the segment of that organ most vulnerable
compared with those whose IMA was ligated at surgery. to ischaemia. This mechanism is supported by the actual
In 1978 Ernst et al. published a report21 on 52 patients findings of preoperative shock in ruptured AAA and the
undergoing elective AAA surgery, 13 of whom had an ligation of internal iliac arteries is strongly associated with
occluded IMA and 39 in whom pre- and post-reconstruction the complication.
 Diagnosis 33

Oxygen delivery to the intestine may become critical as Early passage

of stools
a consequence of two different primary events. The collat-
Early diarrhoea
eral circulation may well be adequate initially, but oxygen
delivery may become deficient as a result of central events Haematoschisis
Early stools and
such as postoperative myocardial infarction. Alternatively, haematoschisis
even though the patient may be haemodynamically stable, Early or bloody stools
the collateral circulation may be compromised by the degree Oliguria–anuria
of atherosclerosis and/or the level of surgical intervention.
Septicaemia
Ischaemic bowel injury is the primary event, and organ
impairment may develop as a result of bacterial transloca- Circulatory instability

tion and endotoxin migration through the injured intestinal Thrombocytopenia

wall, but in the clinical situation it is difficult to establish Peritonitis
which mechanism precedes the other. Late diarrhoea
In this context the effects of different vasoactive drugs
0 10 20 30 40 50 60 70
should be evaluated. Patients are subjected to polyphar-
Per cent
macy after aortoiliac surgery, especially after operation for
ruptured AAA. In a randomised trial,23 patients treated Figure 3.2 Frequency of presenting symptoms of intestinal
with low dose dopamine had significantly lower sigmoid ischaemia (Swedvasc)6
colon pHi values than controls.
In the knowledge that ischaemic intestine is vulnerable
to undue pressure, trauma from retractors, especially of DIAGNOSIS
the self-retaining variety, should be avoided.
Although first described by paediatric surgeons in the
1940s24 the importance of intra-abdominal hypertension Clinical
and of the abdominal compartment syndrome (see Chapters
4 and 6) has received little recognition until recently.25–27 The classic clinical triad of intestinal ischaemia is early
We monitored intra-abdominal pressure in all patients bloody diarrhoea and signs of peritonitis. Very few publi-
operated on for aortoiliac disease since 1998 and there are cations report on sufficient numbers of patients to permit
indications of a correlation between intra-abdominal meaningful discussion on the prevalence of presenting
hypertension and colonic ischaemia. symptoms and signs. Longo et al.,17 reporting to the national
Veterans Administration (VA) Registry on the outcome in
4957 patients, identified 49 with bowel ischaemia. The
Aetiology of bowel ischaemia after predominant symptom was postoperative hypotension
endovascular repair (90 mmHg) in 15 patients, diarrhoea in 14, bloody stools
in 11, fever in eight and abdominal pain in one. Thus, in 23
The pathophysiological mechanism of colonic ischaemia patients (47 per cent), none of the three classic symptoms
in the first reported case history after the advent of EVAR28 was a predominant feature.
was attributed to multiple embolism to the foot and via the In a report from Swedvasc6 the picture was similar
internal iliac artery, sometimes designated ‘trash colon’. (Fig. 3.2). A third of the patients lacked all three classic
The latter is a rare cause of bowel ischaemia having been symptoms and signs. In a five-year prospective study we
identified as a possible cause in only one of 63 cases of identified six patients with bowel ischaemia and in two of
embolism following aortoiliac surgery.6 Although them the clinical triad was absent altogether.7 Signs of peri-
embolism may be more common after EVAR than after tonitis develop late, at which point the patients may be
open repair it does not seem to raise the overall incidence beyond salvage. It is also difficult to evaluate the abdomen
of intestinal ischaemia (L. Norgren, personal communica- in the postoperative period, especially if epidural anaesthe-
tion; data from the Eurostar Registry, August 2001). sia has been used. Therefore, if one relies on clinical symp-
Sigmoid colon pHi fell less after EVAR than after open toms and signs alone, many patients will be lost.
repair in one prospective study,29 probably because
haemodynamic impairment using the endovascular tech- Investigations
nique is less of a problem. As EVAR becomes increasingly
used in the treatment of patients with ruptured AAA, Many different investigations have been recommended
in whom the retroperitoneal haematoma remains in the detection of colonic ischaemia during the primary
unevacuated, it would be wise to monitor the surgical intervention. The problem, however, is that colonic
intra-abdominal pressure. Whether the incidence of ischaemia usually manifests itself in the postoperative period
colonic ischaemia rises or falls as EVAR continues to be and is a product of the depth and duration of the ischaemic
applied to patients with ruptured AAA remains an open insult. Of the 63 patients identified in the Swedvasc study6
question. only five (8 per cent) of the cases of ischaemic bowel were
34 Intestinal ischaemia in aortic surgery

evident at the primary operation, all these patients having

been in deep shock due to a ruptured AAA. In a study involv- 43%
ing pH measurements of the distal colon in 34 patients 16%
undergoing aortoiliac surgery5 the lowest pHi values, repre- 18% 29%
senting the most severe degree ischaemia, were observed, not
during the operation, but 4–24 hours postoperatively.
An elevated concentration of plasma lactate
(2.5 mmol/L) has been shown to be a sensitive indicator
of bowel gangrene,30 but the specificity was very low. In
one study D-lactate was more specific than L-lactate.31
60%
It has been hypothesised, though never been tested in this
specific context, that sustained high lactate levels, some- 18%
times referred to as lact-time, might be a better pointer to
bowel gangrene. An elevated D-dimer was found to have a
good sensitivity and specificity in a preliminary study on
patients suspected of having acute thromboembolic occlu-
sion of the superior mesenteric artery,32 but it has not been
investigated in patients operated on for aortoiliac disease. 86%
Leucocytosis is prevalent early in the development of bowel
gangrene and is followed by leucocytopenia, but these are 64%
non-specific indicators. That coagulopathy in general, and
thrombocytopenia in particular, are more common among
patients with this complication,33 is probably a result of
Figure 3.3 Distribution of ischaemic lesions in the colon and
confounding risk factors such as major bleeding and pre-
rectum in 63 patients studied.6 As more than one segment of the
operative shock. When taken as a whole the above estima- colon can be affected in the same patient the sum exceeds 100
tions can be of some value in the sense that a patient with per cent. In 95 per cent of patients some part of the left colon
normal lactate, D-dimer and white cell count is unlikely to and/or rectum was affected
have developed bowel gangrene.
Perioperative intestinal mucosal blood flow has been
measured using a laser Doppler probe34 and theoretically, if
saline-filled balloons has been displaced by continuous air
the probe is left inside the colon with its light beam in close
tonometry (Tonocap; Datex-Ohmeda, Helsinki, Finland)
contact with the mucosa, it can be monitored postopera-
offering practical online measurements.36,37 Several inves-
tively, but bowel movement and faecal content represent
tigators have studied this diagnostic method in elective
major practical problems.
aortoiliac surgery, a situation in which bowel ischaemia
Colonoscopy is eminently repeatable and the findings
occurs in only 1 per cent of cases and, not surprisingly, no
have been extensively reported14 but caution is required to
conclusion on its validity been made. On the other hand,
avoid perforation and abdominal hypertension due to
three studies of patients operated for ruptured AAA,5,7,13
excessive gas insufflation. Faecal contents constitute a prac-
did include patients who developed the complication.
tical problem, especially after emergency operations. The
It is hardly justifiable to monitor sigmoid colon pHi
flushing of large amounts of saline may be necessary to visu-
after routine elective surgery. Since 1992 we have used pHi
alise the mucosa. The investigation should be performed by
monitors on all patients operated for ruptured AAA,
an experienced endoscopist, but it is important that the vas-
occlusive or aneurysmal disease of the internal iliac arteries
cular surgeon is present in order to evaluate the findings.
or in those carrying potentially high risk factors for colonic
The distribution of lesions on endoscopy among the 63
ischaemia.7 In addition to monitoring perfusion of this
patients studied6 is illustrated in Fig. 3.3; an isolated small
very vulnerable organ, pHi monitoring can be used to prevent
bowel ischaemic lesion was present in only one case but the
further ischaemic injury and when it occurs to recognise
left colon and/or rectum were affected in 95 per cent of
it in time and properly grade it so as to avoid unnecessary
cases. Colonoscopy up to the left flexure would have been
re-laparotomies.
diagnostic in most cases. It must be acknowledged, however,
that although colonoscopy will pick up ischaemic lesions, it
cannot differentiate transmural from mucosal ischaemia.35
MANAGEMENT
Sigmoid colon pHi monitoring, developed by Fiddian
Green et al. (see Schiedler et al.,13) relies on the use of a gas-
permeable balloon placed in contact with the mucosa of Management options in the prevention of bowel ischaemia
the sigmoid colon where it detects increased carbon dioxide are summarised below and the sequence of actions to be
production caused by ischaemia. The original method of taken when ischaemia is suspected is illustrated in Fig. 3.4.
 Results 35

Colonic ischaemia suspected RESULTS

Early bloody diarrhoea and signs of peritonitis
Deterioration with circulatory instability, oliguria, etc
In retrospective reviews from single centres the mortality
among those who develop bowel ischaemia has varied
between 50 and 100 per cent. When reviewing this compli-
Prevent further ischaemic injury cation within the Swedvasc Registry we found that among
Prevent or treat hypovolaemia
those who suffered the complication in a mixed group of
Avoid vasoconstriction and abdominal hypertension
patients who underwent emergency or elective aortoiliac
surgery, the overall 30-day mortality had risen from 12 to
41 per cent, and the one-year rate from 18 to 59 per cent.6
Exclude or confirm colonic ischaemia In this group of patients institutional mortality or one-year
Colonoscopy to the left splenic flexure mortality are far more valid measures of outcome than
Repeat examinations may be necessary
classic surgical mortality. Among patients with transmural
colonic gangrene following surgery for ruptured AAA, the
one-year mortality had reached 90 per cent. Results from a
Grade the depth of the ischaemic lesion prospective study with pHi monitoring suggest that mor-
Colonoscopy/pHi measurements/colonoscopy tality can be reduced by timely recognition and treatment.7

Conclusions
Transmural ischaemic lesion is found
Resection with wide margins and colostomy
Colonic ischaemia is a major complication of open
repair of a ruptured AAA. A reduction in its incidence
will improve the overall survival in this group of
Figure 3.4 Algorithm for management of suspected colonic patients. After elective open repair the frequency is
ischaemia after aorto-iliac surgery approximately 1 per cent, but 9 per cent of the patients
who die after elective surgery will have suffered the com-
Of course, when this complication has occurred, timely plication. The incidence may be lower after elective
confirmation and grading of the lesion are prerequisites endovascular repair. Shock, major blood loss and the
to a successful outcome. Effective collaboration with sacrifice of blood flow to the internal iliac arteries are
anaesthetists and the staff in the intensive care unit is the most important risk factors. Investigations per-
important. Excessive fluid administration will lead to pro- formed during the operation offer no solution because
longed mechanical ventilation and increased abdominal 90 per cent of patients develop colonic ischaemia in the
pressure, effects which are as dangerous to a vulnerable postoperative period. The classic clinical triad of early
colonic circulation as are hypovolaemia and the indiscrim- bloody diarrhoea and signs of peritonitis cannot be relied
inate use of inotropic drugs. Once the complication has upon, all these three cardinal features being absent in a
manifested itself as transmural gangrene of the colon, the third of patients. Colonoscopy up to the left flexure will
choice is clear, namely, immediate bowel resection with identify colonic ischaemia in 95 per cent of patients
wide margins and colostomy. affected, and therefore it is an investigation which should
be carried out liberally. Sigmoid colon pH monitoring
offers the advantages of possible prevention, timely
Measures to prevent colonic ischaemia in recognition and proper grading of the ischaemic lesions.
aortoiliac surgery Treatment of transmural gangrene is immediate resection
with wide margins and colostomy.
• Intraoperative measures
– Avoid excessive blood loss, prolonged cross-
clamping and operating time
– Avoid aortobifemoral reconstruction Key references
– Preserve blood flow to the internal iliac arteries
– Avoid embolisation to the internal iliac arteries Björck M, Broman G, Lindberg F, Bergqvist D. pHi-monitoring of the
– Avoid pressure on the colon from retractors sigmoid colon after aortoiliac surgery. A five-year prospective
study. Eur J Vasc Endovasc Surg 2000; 20: 273–80. This study
• Postoperative measures
shows that pHi-monitoring can detect the complication early
– Prevent hypovolaemia and hypotension
and grade the ischaemic lesion properly. Despite that six
– Avoid unnecessary vasoactive drugs patients suffered colonic ischaemia, no mortality was
– Prevent or treat intra-abdominal hypertension associated with the complication.
36 Intestinal ischaemia in aortic surgery

12 Meissner MH, Johansen KH. Colon infarction after ruptured

Björck M, Troëng T, Bergqvist D. Risk factors for intestinal
abdominal aortic aneurysm. Arch Surg 1992; 127: 979–85.
ischaemia after aortoiliac surgery. A combined cohort and
13 Schiedler MG, Cutler BS, Fiddian-Green RG. Sigmoid intramural
case-control study of 2824 operations. Eur J Vasc Endovasc
pH for prediction of ischemic colitis during aortic surgery. Arch
Surg 1997; 13: 531–9. In this combined cohort and case-
Surg 1987; 122: 881–6.
control study independent risk-factors are identified by
14 Scherpenisse J, van Hees PAM. The endoscopic spectrum of
multivariate analysis.
colonic mucosal injury following aortic aneurysm resection.
Björck M, Bergqvist D, Troëng T. Incidence and clinical presentation
Endoscopy 1989; 21: 174–6.
of bowel ischaemia after aortoiliac surgery – 2930 operations
15 Welch M, Baguneid MS, McMahon RF, et al. Histological study of
from a population-based registry in Sweden. Eur J Vasc
colonic ischaemia after aortic surgery. Br J Surg 1998; 85:
Endovasc Surg 1996; 12: 139–49. This study analyses the
1095–8.
incidence and clinical presentation in a non-selected
16 Amundsen S, Trippestad A, Viste A, Søreide O. Abdominal aortic
population.
aneurysms – a national multicentre study. Eur J Vasc Surg 1988;
Moore SW. Resection of the abdominal aorta with defect replaced
2: 239–43.
by homologous graft. Surg Gyn Obst 1954; 99: 745–55. In this
17 Longo WE, Lee TC, Barnett MG, et al. Ischemic colitis
first intelligent case report, the basis of scientific medicine,
complicating abdominal aortic aneurysm surgery in the US
the clinical picture is described and many of the risk factors
Veteran. J Surg Res 1996; 60: 351–4.
are identified.
18 Järvinen O, Laurikka J, Salenius JP, Lepentalo M. Mesenteric
Soong CV, Halliday MI, Barros D’Sa AAB, et al. Effect of low-dose
infarction after aortoiliac surgery on the basis of 1752 operations
dopamine on sigmoid colonic intramucosal pH in patients
from the National Vascular Registry. World J Surg 243–7.
undergoing elective abdominal aortic aneurysm repair. Br
19 Björck M. On intestinal ischaemia after aortoiliac surgery.
J Surg 1995; 82: 912–15. This well performed, randomised,
Epidemiological, clinical and experimental studies.
controlled trial shows that routine administration of dopamine
Comprehensive summaries of Uppsala dissertations from the
results in sigmoid colon ischaemia.
faculty of medicine 740. Uppsala: Acta Universitatis
Upsaliensis,1998.
20 Björck M, Troëng T, Bergqvist D. Risk factors for intestinal
ischaemia after aortoiliac surgery. A combined cohort and case-
REFERENCES control study of 2824 operations. Eur J Vasc Endovasc Surg 1997;
13: 531–9.
21 Ernst CB, Hagihara PF, Daugherty ME, Griffen WO. Inferior
1 Moore SW. Resection of the abdominal aorta with defect mesenteric artery stump pressure: a reliable index for safe IMA
replaced by homologous graft. Surg Gyn Obst 1954; 99: 745–55. ligation during abdominal aortic aneurysmectomy. Ann Surg
2 Smith RH, Szilagyi DE. Ischemia of the colon as a complication in 1978; 187: 641–6.
the surgery of the abdominal aorta. Arch Surg 1960; 80: 806–21. 22 Piotrowski JJ, Ripepi AJ, Yuhas JP, et al. Colonic ischemia: the
3 Ochsner JL, Cooley DA, DeBakey ME. Associated intra-abdominal Achilles heel of ruptured aortic aneurysm repair. Am Surg 1996;
lesions encountered during resection of aortic aneurysms. Dis 62: 557–61.
Colon Rectum 1960; 3: 485–90. 23 Soong CV, Halliday MI, Barros D’Sa AAB, et al. Effect of low-dose
4 Brewster DC, Franklin DP, Cambria RP, et al. Intestinal ischemia dopamine on sigmoid colonic intramucosal pH in patients
complicating abdominal aortic surgery. Surgery 1991; 109: undergoing elective abdominal aortic aneurysm repair. Br J Surg
447–54. 1995; 82: 912–15.
5 Björck M, Hedberg B. Early detection of major complications after 24 Gross R. A new method for surgical treatment of large
abdominal aortic surgery: predictive value of sigmoid colon and omphaloceles. Surgery 1948; 24: 277–92.
gastric intramucosal pH monitoring. Br J Surg 1994; 81: 25–30. 25 Schein M, Ivatury R. Intra-abdominal hypertension and the
6 Björck M, Bergqvist D, Troëng T. Incidence and clinical abdominal compartment syndrome. Br J Surg 1998; 85: 1027–8.
presentation of bowel ischaemia after aortoiliac surgery – 2930 26 Loftus IM, Thompson MM. The abdominal compartment
operations from a population-based registry in Sweden. Eur J syndrome following aortic surgery [review]. Eur J Vasc Endovasc
Vasc Endovasc Surg 1996; 12: 139–49. Surg 2003; 25: 97–109.
7 Björck M, Broman G, Lindberg F, Bergqvist D. pHi-monitoring of 27 Rasmussen TE, Hallet JW, Noel AA, et al. Early abdominal closure
the sigmoid colon after aortoiliac surgery. A five-year with mesh reduces multiple organ failure after ruptured
prospective study. Eur J Vasc Endovasc Surg 2000; 20: 273–80. abdominal aortic aneurysm repair: guidelines from a 10-year
8 Johnson WC, Nabseth DC. Visceral infarction following aortic case-control study. J Vasc Surg 2002; 35: 246–53.
surgery. Ann Surg 1974; 180: 312–18. 28 Sandison AJP, Edmondson RA, Panayitopoulos YP, et al. Fatal
9 Tollefson DJF, Ernst CB. Colon ischemia following aortic colonic ischaemia after stent graft for aortic aneurysm. Eur J
reconstruction. Ann Vasc Surg 1991; 5: 485–9. Vasc Endovasc Surg 1997; 13: 219–20.
10 Hagihara PF, Ernst CB, Griffen WO. Incidence of ischemic colitis 29 Syk I, Brunkwall J, Ivancev K, et al. Postoperative fever, bowel
following abdominal aortic reconstruction. Surg Gyn Obst 1979; ischaemia, and cytokine response to abdominal aortic aneurysm
149: 571–3. repair – a comparison between endovascular and open surgery.
11 Bast TJ, van der Biezen JJ, Scherpenisse J, Eikelboom BC. Eur J Vasc Endovasc Surg 1998; 15: 398–405.
Ischaemic disease of the colon and rectum after surgery for 30 Lange H, Jäckel R. Usefulness of plasma lactate concentration in
abdominal aortic aneurysm: a prospective study of the incidence the diagnosis of acute abdominal disease. Eur J Surg 1994; 160:
and risk factors. Eur J Vasc Surg 1990; 4: 253–7. 381–4.
 References 37

31 Poeze M, Froom AH, Greve JW, Ramsay G. D-lactate as an laser Doppler flowmetry and tissue oximetry to histological
early marker of intestinal ischaemia after ruptured abdominal analysis. Eur J Vasc Surg 1992; 6: 518–24.
aortic aneurysm repair. Br J Surg 1998; 85: 1221–4. 35 Houe T, Thorböll JE, Sigild U, et al. Can colonoscopy diagnose
32 Acosta S, Nilsson TK, Björck M. Elevated D-dimer level could be transmural ischaemic colitis after abdominal aortic surgery? An
a useful early marker for acute bowel ischaemia. A preliminary evidence-based approach. J Vasc Endovasc Surg 2000; 19: 304–7.
study. Br J Surg 2001; 88: 385–8. 36 Heinonen PO, Jousela IT, Blomqvist KA. Validation of air tonometric
33 Lannerstad O, Bergentz SE, Bergqvist D, Takolander R. Ischemic measurement of gastric regional concentrations of CO2 in critically
intestinal complications after aortic reconstructive surgery. ill septic patients. Intensive Care Med 1997; 23: 524–9.
Acta Chir Scand 1985; 151: 599–602. 37 Lebuffe G, Decoene C, Raingeval X, et al. Pilot study with air-
34 Krohg-Sørensen K, Line PD, Haaland T, et al. Intraoperative automated sigmoid capnometry in abdominal aortic aneurysm
prediction of ischaemic injury of the bowel: a comparison of surgery. Eur J Anaesth 2001; 18: 585–92.
This page intentionally left blank
 4
Ischaemia-Reperfusion Injury, SIRS and MODS

DENIS W HARKIN, AIRES AB BARROS D’SA

Introduction 39 Multiple organ dysfunction syndrome 43

The problem 39 Monitoring of organ dysfunction in SIRS 46
Ischaemia-reperfusion injury 39 References 49
Systemic inflammatory response syndrome 40

INTRODUCTION inflict a substantial mortality of 10–20 per cent.4 Although

the majority of early deaths in these patients is due to car-
diovascular disease, which reflects the systemic nature of
The humoral and cell-mediated immune responses to tissue
atherosclerosis, late deaths are mostly due to MODS, which
injury represent the first stages of repair, but if it is excessive
remains the leading cause of death in surgical intensive
local tissue injury may extend to a potentially lethal systemic
care units5 (see Chapters 9A and 9B).
inflammatory response syndrome (SIRS). In certain individ-
uals a heterogeneous variety of insults, including trauma,
haemorrhage, major surgery and ischaemia-reperfusion Clinical implications of premorbid risk factors
injury (IRI), may provoke an overwhelming inflammatory
response ultimately leading to multiple organ dysfunction Perhaps the most important factor influencing outcome in
syndrome (MODS), organ failure and, ultimately, death. emergency cases is the recognition that vascular surgical pro-
While the inflammatory response to tissue injury bears many cedures typically involve the elderly patient in whom either
similarities to the sepsis syndrome, a septic focus is not a occult disease or overt signs of pre-existing organ dysfunc-
prerequisite to SIRS. Paradoxically, when systemic inflam- tion raise the likelihood of perioperative morbidity and mor-
mation becomes established, the essentially sterile IRI often tality. It is now recognised that peripheral vascular disease is
exhibits a septic component which plays a crucial role in a strong pointer to systemic atherosclerosis and therefore
the propagation rather than the resolution of inflamma- coronary and cerebrovascular events occur much more fre-
tion. In this chapter the pathophysiological processes quently in vascular patients.6,7 Renovascular disease, associ-
involved in the evolution of SIRS and MODS in vascular ated with diabetes, hypertension or atherosclerotic renal
patients are discussed. artery stenosis, is relatively more likely to precipitate post-
operative renal dysfunction in vascular patients, and once
established it substantially heightens the risk of mortality,
especially after AAA repair.8 Although, advanced chrono-
THE PROBLEM logical age is predictive of high mortality in some patients,9
good results can be achieved even in octogenarians if they are
Despite modern advances in perioperative care, emergency otherwise physiologically healthy.8 Local ischaemic injury to
vascular surgery continues to exact a severe toll on this the limb can, under certain circumstances, evolve into SIRS
patient population in terms of morbidity and mortality. In potentially leading to MODS and death.
its most florid presentation, surgery for a ruptured abdom-
inal aortic aneurysm (AAA), despite initial success, continues
ISCHAEMIA-REPERFUSION INJURY
to register in-hospital mortality rates of 50–75 per cent in
specialist units.1–3 Of greater concern, however, is the fact
that emergency limb salvage procedures in patients pre- Lower limb IRI is a common clinical sequela of throm-
senting with acute vascular insufficiency of the limb also boembolism, trauma, bypass surgery and low-flow states,
40 Ischaemia-reperfusion injury, SIRS and MODS

classically observed in ruptured AAAs. A vascular emer- reoxygenation (reperfusion) of the tissues is characterised
gency becomes life threatening if accompanied by physio- by local endothelial activation, membrane lipid peroxida-
logically stressful situations: haemorrhage in association tion, cellular oedema, capillary leak and leucocyte recruit-
with a ruptured AAA, multisystem injuries in cases of vascu- ment, activation and tissue infiltration. This local tissue
lar trauma, pre-existing vital organ dysfunction especially injury develops into a systemic phenomenon as metabolic
in the elderly arteriopath. Deoxygenation (ischaemia) and byproducts, oxygen reactive species, arachidonic acid deriva-
tives, cytokines, mediators and activated leucocytes are
flushed into the systemic circulation where they can be
harmful to vital organ systems, as depicted in Fig. 4.1. The
Overspill of pathophysiology of acute limb vascular insufficiency, and
proinflammatory IRI which accompanies it, is covered elsewhere (see
Ischaemia–reperfusion injury mediators into Chapter 2), and therefore the discussion here is focused on
systemic circulation
systemic inflammation and organ dysfunction which
evolve in this setting.

Systemic inflammatory
response syndrome SYSTEMIC INFLAMMATORY RESPONSE
SYNDROME
Myocardial depression/injury Acute renal failure
As proinflammatory mediators overcome native anti-
Gastrointestinal inflammatory pathways a SIRS ensues. Products of oxida-
Acute respiratory failure and tive injury, namely, reactive oxygen species, arachidonic
distress syndrome endotoxaemia acid metabolites and hydroperoxides, interact with acti-
vated endothelial cells and leucocytes in an environment
Multiple organ failure rich in proinflammatory cytokines causing vital organ injury.
In response to IRI the complex cascade of proinflammatory
Figure 4.1 Pictorial representation of the evolution of systemic
inflammation and multiple organ dysfunction after limb IRI
ischaemia-reperfusion injury Inflammatory stimulus

Endothelial activation Neutrophil activation

Aggregation, degranulation, Monocyte/macrophage
PAF, TGF- , nitric oxide
release of O2 radicals Recruitment and activation
and bradykinin

Activation of coagulation Arachidonic acid Platelet

and complement systems metabolism Activation and
Tissue factor release, Thromboxane A2, aggregation T-cell release of IL-2
fibrinolytic activity prostaglandins, leucotrienes and IFN-

Cytokinaemia Circulating neutrophil and

TNF- , IL-1 monocyte activation
IL-6, 8 and 10 Microvascular injury

SIRS
‘Second hit’
inflammatory stimulus
Bacteria, endotoxin and
microbial byproducts Sepsis syndrome

Figure 4.2 The development of the systemic inflammatory response syndrome (SIRS) after the inflammatory stimulus of ischaemia-
reperfusion injury (IRI) represents a complex cascade of proinflammatory mediator and immune cell interactions, which may be further
complicated by the development of a sepsis syndrome. IL, interleukin; IFN, interferon; PAF; platelet activating factor; TGF, transforming
growth factor; TNF, tumour necrosis factor
 Systemic inflammatory response syndrome 41

mediator and immune cell interactions leads to the devel- then converts the local inflammatory injury into a systemic
opment of SIRS, and not infrequently it is complicated by inflammatory response.
the sepsis syndrome, as illustrated in Fig. 4.2.
Definition of systemic inflammatory response
syndrome
Intravascular response to reperfusion
Despite the varied initiating stimuli, SIRS, once established,
At its most severe the intravascular response to limb IRI
follows a common, if often chaotic, pathway leading to
involves the establishment of cytokine and immune cell
widespread increased microvascular permeability and organ
interactions propagating an inflammatory cascade which
injury. The systemic inflammatory response to injury
causes generalised capillary dysfunction and ultimately
essentially represents a spectrum of responses which can be
leads to organ failure. Major vascular surgery has been
mild and self-limiting or severe, leading rapidly to multiple
shown to initiate a systemic inflammatory response char-
organ failure (MOF) and death. This is reflected in the
acterised by increased plasma levels of proinflammatory
descriptive criteria shown in Table 4.1 proposed by Baue.26
cytokines10 and circulating polymorphonuclear (PMN)
It is recognised, however, that the more innocuous first insult
leucocyte, or neutrophil, activation.11 Experimental stud-
is often aggravated by another in the form of sepsis eventu-
ies have shown that activated neutrophils accumulate
ally ushering in MOF and death.27
within remote organs in proportion to the severity of tissue
injury12,13 caused by an acute microvascular insult.14–17
Plasma factors augment neutrophil and endothelial cell SIRS: the propagation or resolution of
activation after revascularisation of ischaemic tissue by the inflammation
activation, or ‘upregulation’, of neutrophil surface adhe-
sion receptor integrins (CD11b/CD18) and endothelial cell A complex cascade of events depicted in Fig. 4.3 leads from
adhesion molecules (ICAM-1),18 in turn leading to the initial inflammatory stimulus to MOF, which is lethal.
degranulation and tissue injury.12 Bone28 classifies SIRS further into the compensatory anti-
Antioxidant defence mechanisms, normally protecting inflammatory response syndrome (CARS) and the mixed
tissues against harmful oxygen reactive species created as antagonistic response syndrome (MARS), which suggests
byproducts of metabolism, respiration or immune medi- that weighting the process in one direction or the other
ated microbial killing are overwhelmed by uncontrolled decides the outcome. As Baue26 points out, however, that
oxidant production. Consumptive systemic depletion of despite the desire to represent inflammation as an orderly
antioxidants has been demonstrated both during ischaemia response, it is in fact neither organised and sequential, nor
and after reperfusion in AAA repair.19 The resultant oxida- coordinated but a rather chaotic process. While the early
tive cell membrane injury leads to the creation of rise in circulating cytokine concentrations could be benefi-
eicosanoids such as prostaglandins, thromboxanes and cial to the host by initiating the acute phase response, their
leucotrienes, which are oxygenated products of 20-carbon uncontrolled production undoubtedly contributes to the
fatty acids such as arachidonic acid. Agonist–receptor medi- development of MODS.29 Paradoxically, depression of the
ated activation of membrane phospholipase A2 leads to cytokine response does not improve outcome; indeed, it is
specific release of arachidonic acid, the substrate for cyclo- associated with increased mortality.30 Systemic inflamma-
oxygenase or lipoxygenase enzymes. Thromboxane (TX) A2, tory response syndrome, once established, is propagated by
a potent vasoconstrictor and platelet aggregator, is a product
of the cyclo-oxygenase initiated pathway and is produced Table 4.1 Criteria for the determination of the systemic
rapidly following aortic clamp release at which point it is inflammatory response syndrome (SIRS) *
related, temporally, to the onset of pulmonary hyperten-
sion.20 Leucotriene (LT) B4, a product of the lipoxygenase Condition† Criterion
initiated pathway is derived from neutrophils, mast cells, Temperature A temperature lower than 36 °C or higher
macrophages and endothelial cells. Both these eicosanoids than 38 °C
have been shown to be highly chemotactic for neutrophils Cardiac A heart rate more than 90 beats per minute
contributing to their endothelial adherence and transmi- Pulmonary A respiratory rate more than 20 breaths
gration.21 Nitric oxide, normally produced in basal amounts per minute or a PaCO2 less than 32 mmHg,
by the vascular endothelium is critical to normal vascular which usually means hypoxic hyperventilation
homoeostasis,21,22 by mediating vascular smooth muscle Haematological A white blood cell count more than
relaxation and inhibiting platelet aggregation and neu- 12.0 109/L or less than 4.0 109/L or the
trophil adherence.23 Endothelial cell production of nitric presence of greater than 10 per cent
oxide is decreased during reperfusion,24 and that is com- immature or band forms
pounded by superoxide anion induced inactivation of * Data from Baue.26
nitric oxide.25 This intravascular proinflammatory milieu †
SIRS requires two or more of the conditions be met.
42 Ischaemia-reperfusion injury, SIRS and MODS

Acute arterial
model of lower limb IRI we have shown that high plasma
insufficiency
IRI levels of IL-6 are associated with MOF and high mortality.39
In vitro studies have shown that IL-6 delays PMN apop-
Decompensation tosis, which may explain why PMNs may exhibit increased
Global tissue MOF functional longevity and a potential for oxidative tissue
hypoxia injury.36 However, IL-6 also has an important immun-
omodulatory role – by directly inhibiting expression of
MODS DEATH proinflammatory TNF- and IL-1 40 and stimulating
macrophage expression of IL-1 receptor antagonist and
Exaggerated/ soluble TNF- receptor. Interleukin-8, produced by a var-
Microvascular-
uncontrolled iety of cell types, is a potent leucocyte chemoattractant and
endothelial SIRS
inflammatory
dysfunction activator41 and highest levels of this interleukin have been
response
recorded in patients who developed MOF after trauma.42
CARS In addition IL-8 has been demonstrated in the bron-
choalveolar lavage fluid of patients with acute respiratory
distress syndrome (ARDS).43 Interleukin-10 is produced
Resolution, repair,
by immunoregulatory cells44 and has the potential to
recovery inhibit the production of various cytokines, including TNF-
and IL-6.45 Despite this an increase in plasma IL-10 levels
Figure 4.3 A schematic representation of the progressive spiral has been associated with septicaemia.46
from a severe inflammatory stimulus to lethal multiple organ
failure. CARS, compensatory anti-inflammatory response
syndrome; MODS, multiple organ dysfunction syndrome; GASTROINTESTINAL BARRIER DYSFUNCTION
MOF, multiple organ failure; IRI, ischaemia-reperfusion injury; The concept of gut origin sepsis, in which dysfunction of
SIRS, systemic inflammatory response syndrome
the intestinal barrier results in the passage of bacteria and
their toxins from the lumen to normally sterile extraintesti-
nal sites, is attributable to Fine.47 Intestinal barrier failure has
a variety of mediators, perhaps most importantly by the been implicated in the pathogenesis of complications
balance between pro- and anti-inflammatory cytokines. following thermal injury,48 trauma49 and major vascular
Similarly, the development of a secondary septic injury or surgery.30
sepsis syndrome, from either exogenous or endogenous
bacterial pathogens, is often a grave indicator of relentless Bacterial translocation
progression to organ failure and death. The gastrointestinal tract is normally inhabited by a large col-
lection of microbial species most notably Gram-negative
bacteria.50 Corson et al.51 showed a link between limb IRI,
CYTOKINES
remote gut injury and increased systemic endotoxin concen-
Cytokines, originating from monocytes, neutrophils, mast trations. Increased intestinal permeability was demonstrated
cells, lymphocytes, tissue macrophages and vascular after major vascular surgery.30 Subsequently, at our centre,
endothelial cells31 have myriad pro- and anti-inflammatory lower limb IRI was confirmed to be associated with remote
properties and play a key a role in an evolving SIRS. gut mucosal injury and increased permeability.52 Indeed,
Postischaemic extremities have been shown to exhibit the research has shown that bacterial translocation can be pro-
immediate release of tumour necrosis factor alpha (TNF- ).32 voked by a variety of injurious stimuli including burns,53
In animal studies TNF elicits many key features of SIRS endotoxaemia54 and haemorrhage.55 Predisposing factors
including fever, proteolysis, shock and organ dysfunction.33 include bacterial overgrowth,54 gut mucosal barrier disrup-
This is in part due to its role in upregulating neutrophil tion56 and abnormal host defences.57
surface adhesion receptors (integrin CD11b/CD18).34
The interleukin (IL) family of cytokines has a variety of Endotoxaemia
pro- and anti-inflammatory effects. Interleukin-1 is an Abdominal aortic aneurysm repair is associated with portal
early proinflammatory cytokine, often synergistically potenti- and systemic endotoxaemia.58,59 Endotoxin, a lipopolysac-
ating TNF- effects, and is produced by macrophages, charide (LPS) component of the cell wall of Gram-negative
monocytes and vascular endothelium.35 Interleukin-6 is a bacteria, is a potent stimulus to cytokine generation, coagu-
pleiotropic cytokine, produced by a wide variety of immune lation and complement activation60 as well as leucocyte
reactive cell types36 and found in high concentrations in activation.61 In humans endotoxin has been implicated in
association with morbidity after AAA repair;37 post- the pathogenesis of SIRS and the development of acute
traumatic elevated IL-6 levels also correlate closely with injury lung injury61,62 but its precise role after vascular surgery
severity scores (ISS) and mortality.38 In an experimental remains controversial.63 In experimental studies at our
 Multiple organ dysfunction syndrome 43

centre it has been demonstrated that lower limb IRI is asso- Table 4.2 Criteria for determination of sepsis syndrome.*
ciated with endotoxaemia64 and that the resulting SIRS is Patients clinically suspected to be septic fulfilled the following
associated with a rise in mortality.65 criteria for sepsis syndrome

Colonic ischaemia System Criteria†

Colonic ischaemia is a feared complication of aortic sur-
Temperature Hyperthermia (38 °C)
gery, with an incidence ranging from 7 per cent after repair Hypothermia (35.5 °C)
of a ruptured AAA to 0.6 per cent after bypass for aortoiliac A proved site of infection
occlusive disease.66 The inferior mesenteric artery, the main Heart rate Tachycardia (90 beats per minute in the
blood supply to the left colon, is often sacrificed at AAA repair absence of -blockade)
and important collateral channels from the hypogastric ves- Respiratory rate Tachypnoea (20 breaths per minute)
sels may be lost if internal iliac arteries are excluded during A requirement for mechanical ventilation
the implantation of a bifurcated aortic graft. The combin- Organ function Evidence of dysfunction of one or more
ation of vascular interruption, hypotension and prolonged end organs, defined as follows:
aortic clamp time magnify the risk to the colon.67 The dam- (a) Plasma lactate greater than
1.2 mmol/L, base deficit greater than
age may range from mild self-limiting mucosal ischaemia
5 mmol/L, or systemic vascular
to frank ischaemic infarction, perforation and sepsis.
resistance less than 800 dyne.s.cm5;
(b) PaO2/fraction inspired oxygen less
HEPATIC RETICULOENDOTHELIAL DYSFUNCTION than 30 kPa or PaO2 less than 9.3 kPa;
(c) Less than 120 mL urine output during
The vast hepatic sinusoidal network is lined with fixed tis- 4 hours; or
sue macrophages, or Kupffer cells, which are strategically (d) Glasgow Coma Scale score less than
located to interact with gut derived endotoxin. In addition 15 in the absence of a neurological lesion.
to clearing bloodborne endotoxin (LPS)68 the Kupffer cells * Data from Bone.28
are a rich source of inflammatory mediators such as TNF, †
Each category requires one or more criteria.
interleukins, platelet activating factor (PAF), and arachi-
donic acid metabolites. Accordingly, the intestinal–hepatic
axis may be crucial in the development of systemic inflam- with overall mortality rates exceeding 70 per cent,74 the rate
mation in patients undergoing major vascular surgery. In rising with the number of organs failing.71 Clinical studies
the critically ill, impairment of Kupffer cell function is have shown that all patients undergoing aortic reconstruc-
associated with overwhelming sepsis and poor prognosis. tion suffer transient organ dysfunction affecting the lung,
gut and kidney.75 Mortality rates of 5 per cent after elective,
and 50 per cent after ruptured AAA repair, are still com-
Sepsis syndrome mon76 and at least 20 per cent of these deaths are attributable
to MODS.77 In the past MODS was assumed to be a com-
The sepsis syndrome, defined in Table 4.2, results from plication of uncontrolled infection or a manifestation of an
activation of the host’s defence mechanisms in response to occult septic focus, but it is now clear that it can develop in
invading microorganisms and their products, including the absence of an identifiable focus of infection.78 This sup-
endotoxin.69 The development of sepsis syndrome is a grave ports the hypothesis that bacteria or endotoxin from the
prognostic indicator, regardless of the origin of the bacter- gut lumen may translocate into the circulation at times of
ial pathogen, whether from an endogenous source such as systemic stress and drive the inflammatory response.79 As
the intestine or exogenous foci such as ventilatory acquired described, an uncontrolled SIRS is implicated in the aeti-
pneumonia or intravascular catheter sepsis. Severe sepsis is ology of MODS.80 The common cellular abnormality in all
a major cause of death in patients admitted to the intensive these organs is increased microvascular permeability with
care unit (ICU) and continues to be responsible for the sequestration of activated inflammatory cells in the
high mortality of 25–58 per cent.70 microvessels of systemic organs.

Immune system
MULTIPLE ORGAN DYSFUNCTION SYNDROME
Deficiencies of the immune system after surgery, burns and
trauma are well documented and may occur in a variety of
Classical definitions of MOF as an ‘all-or-nothing’ phe-
ways. Adverse postinjury events include:
nomenon,71 have been displaced by the consensus view
that MOF, the criteria for which are shown in Table 4.3, is • inhibition of the phagocytic cellular response81
more accurately portrayed as an extension of MODS.73 It • decreased lymphokine (particularly IL-2) generation35
remains the leading cause of death in the surgical ICU, • increased complement activation82
44 Ischaemia-reperfusion injury, SIRS and MODS

Table 4.3 Criteria for the determination of multiple organ failure * coagulopathy.89 Another important haematological path-
way activated after major vascular surgery is that of the
Organ system Criteria †
complement system, the creation of small antigen–
Cardiac Heart rate 54 beats/min antibody complexes in the circulation and the appearance
Mean arterial blood pressure of altered cell surface epitopes which activate the classic and
49 mmHg alternative complement pathways, respectively. Activated
Ventricular tachycardia or fibrillation products of the complement pathway are potent inflamma-
Serum pH 7.24 with PaCO2 tory mediators with myriad effects that include alteration of
49 mmHg blood vessel permeability and tone, leucocyte chemotaxis
Haematological White blood cells 1.0 109/L and the activation of multiple inflammatory cell types.
Platelets 20.0 109/L
Haematocrit 0.20
Hepatic Prothrombin time 4 seconds over control Cardiovascular system
(in the absence of anticoagulation)
Bilirubin 103 mol/L (6 mg/dL)
Depressed cardiac function is common in the critically ill
Central nervous Glasgow Coma Scale 6
system (in the absence of sedation)
and contributes to global hypoperfusion. Supraventricular
Renal Urine output 479 mL/24 hours arrhythmia and impaired myocardial contractility are the
(159 mL/8 hours) most readily evident alterations in cardiac function in the
Urea nitrogen 35.7 mmol/L critically ill. Right ventricular function is particularly affected
(100 mg/dL) as a consequence of increased pulmonary vascular resist-
Creatinine 309 mol/L (3.5 mg/dL) ance90 and an interplay of multiple depressive influences,
(excluding long term dialysis) including catecholamine excess, hypoxia, acidosis91 and
Pulmonary Respiratory rate 5 or 49/min myocardial depressant factors.92,93 A dilated peripheral vas-
PaCO2 50 mmHg cular bed coupled with widespread capillary leakage aggra-
Alveolar–arterial oxygen difference vates cardiac workload and may ultimately leads to
350 mmHg
myocardial ischaemia. This added cardiovascular strain may
Ventilator dependence 3 d with
another organ failed
be critical in those with pre-existing coronary artery disease.

* Data from Knaus et al.72

†
Each category requires one or more criteria. Respiratory system

• raised prostaglandin production (PGE2)83 Acute non-cardiac pulmonary oedema after major vascular
• appearance of immunosuppressive serum factors84 surgery, in particular aneurysm repair, is well recognised.
• alterations in T cell functions including decreased Experimental lower torso ischaemia-reperfusion induces
proliferation57 acute pulmonary injury which is characterised by increased
• relative increase in suppressor cells85 microvascular permeability and neutrophil infiltration,94 a
• alterations in B cell functions such as decreased phenomenon which can be effectively prevented experi-
immunoglobulin production.86 mentally by neutrophil depletion.95 Damage to the capil-
lary endothelium resulting in leakage of fluid and protein
These effects, in combination, render the patient more sus-
is produced by an interaction of inflammatory cells and
ceptible to sepsis and associated complications which con-
mediators including leucocytes, cytokines, oxygen radicals,
tribute to the general proinflammatory milieu.
complement and arachidonate metabolites. The picture is
clinically inseparable from ARDS, which can be of varying
Haematological system intensity as shown in Table 4.4, and is characterised by dif-
fuse pulmonary capillary leak, a common complication in
Critical illness is associated with various haematological the critically ill. The lungs become stiff and less compliant,
abnormalities, a mild anaemia being common along with lung volumes are reduced and as a consequence of alveolar
abnormalities in red cell deformability,87 a high or inappro- atelectasis an extreme level of intrapulmonary shunting
priately low leucocyte count and in some cases even an occurs with characteristic bilateral pulmonary interstitial
absolute lymphopenia. Reductions in platelet count are infiltrates on chest X-ray. Pulmonary hypertension is asso-
common in patients undergoing elective repair of an AAA88 ciated with ARDS96 and may be improved, in part, by
and perhaps more dramatically that of a ruptured AAA.89 locally delivered vasodilators such as nitric oxide. Septic
Later, patients develop hyperfibrinogenaemia and throm- patients are at greater risk of developing ARDS,97 indeed
bocytosis, which may persist for several weeks.88 Activation experimental infusion of LPS elicits a syndrome of acute lung
of the fibrinolytic system, haemorrhage, hepatic dysfunc- injury which closely resembles ARDS.98 Abdominal disten-
tion and massive blood transfusion contribute to the risk of sion contributes to the loss of pulmonary compliance,
 Multiple organ dysfunction syndrome 45

Table 4.4 Postinjury acute respiratory distress syndrome socre (ARDS)

Variables Grade 1 Grade 2 Grade 3 Grade 4

A Pulmonary/radiographic Diffuse, mild interstitial Diffuse, marked interstitial/ Diffuse, moderate air Diffuse, severe air
markings/opacities mild air space opacities space consolidation space consolidation
B PaO2/FiO2 (mmHg) 175–250 125–174 80–124 80
C Minute ventilation (L/min) 11–13 14–16 17–20 20
D PEEP (cmH2O) 6–9 10–13 14–17 17
E Static compliance 40–50 30–39 17–20 20

ARDS SCORE  A  B  C  D  E when PCWP 18 mmHg or when there is no clinical reason to suspect hydrostatic pulmonary oedema.
PEEP, peak end expiratory pressure; PCWP, Pulmonary capillary wedge pressure.

compounding respiratory failure in those who have sus- injury mediated by inflammatory mediators, activated leuco-
tained acute intra-abdominal catastrophes.99 cytes and hypoperfusion.80 The crucial role of gut ischaemia
in the pathogenesis of MODS is indicated by the improved
survival of shocked patients in whom gut ischaemia, as
Renal system determined by gastric tonometry, can be reversed.104
Naturally, the gut has been described as the ‘motor’ driving
Acute renal failure (ARF) complicates both emergency and the systemic inflammatory response.27 Sustained acidosis
elective aortic aneurysm surgery.100 Renal parenchymal of the gastric and sigmoid mucosa has been shown to be a
ischaemic injury is a common sequela of suprarenal aortic highly sensitive predictive indicator of mortality and mor-
clamping and is exacerbated in some by atheroembolism.101 bidity in elective and emergency AAA surgery patients.105
Postoperatively, renal hypoperfusion brought about by pref- For this reason the clinician must remain vigilant to the pos-
erential shunting and falling central arterial blood pressure sible development of non-occlusive mesenteric ischaemia
compounds the effects of renal ischaemia. Furthermore, the (see Chapters 3 and 29). Furthermore, gut mucosal atro-
release of oxygen free radicals, systemic vasoconstrictors, phy during critical illness damages its absorptive ability,
toxic metabolites, myoglobin and activation of neutrophils though that can be prevented, in part, by early enteral
as a consequence of limb reperfusion, act in concert to pro- feeding.
duce acute tubular dysfunction.102 Common radiological
contrast agents also induce renal dysfunction in all too many
vascular patients. Despite advances in providing support for Neuroendocrine system
the critically ill patient the development of ARF remains
quite common and carries a grave prognosis with a mortal- Plasma catecholamine concentrations increase during con-
ity of around 45 per cent, and in those with sepsis it may be ventional surgery,106 and during open AAA repair are asso-
as high as 75 per cent.103 ciated with cardiovascular instability.107 The acute phase
response to injury brings with it a catecholamine surge as
Hepatic system the body tries to maintain its essential functions. Glucocorti-
coids initially provide a welcome anti-inflammatory
The liver may sustain direct ischaemic injury during supra- response, inhibiting the production of cytokines TNF- 108
coeliac aortic clamping and also indirectly in association with and IL-6,109 but their influence, if prolonged, becomes
perioperative hypotension. Hepatocellular injury impairs the deleterious. Failure to satisfy the sustained hypermetabolic
liver’s ability to manufacture clotting factors necessary to sat- demands of critical illness is associated with a poor
isfy the demands of a hypercoagulopathy, to produce albu- outcome.110
min sufficient to maintain intravascular volume in the
presence of capillary leak and finally to metabolise mediators General metabolism
and metabolic byproducts in the presence of ongoing inflam-
mation. Although transient moderate rises in hepatocellular The systemic inflammatory response brings with it a hyper-
enzymes are common after major aortic surgery, a large or catabolic state and as the immune system and major organs
sustained rise must be regarded as a grave sign. attempt to cope with that increased demand the body
switches from aerobic to anaerobic respiration on a mas-
Gastrointestinal tract sive scale. If tissue hypoxia persists, markers of intermediary
metabolism, such as ketone body ratio, point to a process
A variety of factors contribute to gut injury including direct of decompensation heralding MOF and mortality in the
ischaemia by supracoeliac aortic clamping and remote critically ill.111 Postoperative hypermetabolism, as estimated
46 Ischaemia-reperfusion injury, SIRS and MODS

by an increase in baseline oxygen consumption, is mani- as observed in a variety of conditions such as sepsis, myelo-
fested experimentally by higher circulating levels of endo- proliferative disease, pancreatitis, short bowel syndrome
toxin, TNF and IL-6.112 Paradoxically, cytokines such as or induced by drugs such as catecholamines, biguanides,
IL-6 also have a role in depressing cellular metabolic activ- methanol, ethanol and ethylene glycol. In the critically ill, high
ity and that will induce cachexia.113 lactate levels have been shown to be predictive of morbidity
and mortality.117

MONITORING OF ORGAN DYSFUNCTION INDICATORS OF REGIONAL TISSUE OXYGENATION

IN SIRS
Hollow viscus tonometry is increasingly used to monitor
the adequacy of splanchnic tissue perfusion either via the
Although mortality rates from MODS remain high, histor- gastric or the sigmoid route.118 Sustained acidosis of intes-
ical comparisons of critically ill surgical patients suggest tinal mucosa has been shown to be highly sensitive in pre-
the incidence of ARF, ARDS, gastrointestinal stress haem- dicting mortality and morbidity in elective and emergency
orrhage and abdominal abscess formation have decreased. aortic aneurysm surgery.105
The key reason for organ dysfunction in the critically ill is
global tissue hypoxia. Central shunting attempts to main- MONITORING PULMONARY GAS EXCHANGE
tain vital organ perfusion, often at the cost of sacrificing
splanchnic and peripheral circulation. Historically, clin- Arterial blood gas samples with a known fraction of inspired
icians have directed their treatment so as to avert oxygen debt oxygen remains the clinical standard for monitoring pul-
by optimising systemic haemodynamics and tissue oxy- monary gas exchange. Pulse oximetry is useful but does not
genation. Shoemaker et al.114 have gone further and pro- detect impaired tissue oxygenation attributable to a left-
posed targeted therapy to achieve ‘supranormal’ oxygen ward shift of the oxygen–haemoglobin dissociation curve.
delivery in critically ill patients to improve survival. The continued development of fibreoptic blood gas and
pH sensors, at present still experimental, may further refine
continuous monitoring of high risk patients.
Monitoring of global haemodynamics

The pulmonary artery catheter, introduced by Swan et al.,115 Monitoring of renal function
remains the clinical standard for monitoring macrohaemo-
dynamic variables such as cardiac output, pulmonary Diuresis and creatinine clearance currently represent the
artery occlusion pressure, mixed venous oxygen saturation clinically most important variables in monitoring global
and derived variables of oxygen delivery. However, concerns renal function. The goal must be to detect and correct renal
have been raised since Connors et al.116 showed a rise in hypoperfusion before the development of overt signs of
mortality, prolonged stay in intensive care and increased renal failure such as anuria, hyperkalaemia, acidaemia and
treatment costs for those being monitored invasively. azotaemia. In general, renal blood flow is hard to predict
Non-invasive methods for assessing cardiac output remain using simple systemic haemodynamic variables because of
largely experimental and include transthoracic and tran- complex renal-based blood pressure maintaining compen-
soesophageal bioimpedance monitoring, carbon dioxide satory mechanisms, namely, sympathetic activity and the
and soluble gases rebreathing methods and Doppler renin–angiotensin–aldosterone system. Unfortunately, newer
echocardiography. imaging techniques such as greyscale and duplex ultrasono-
graphy and magnetic resonance imaging provide little
assistance in predicting ARF in the critically ill patient.119
Monitoring of tissue oxygenation

Falls in either cardiac output or arterial oxygen content lead Predictive scores in critically ill vascular
to an imbalance whereby oxygen consumption exceeds oxy- patients
gen delivery. This tissue hypoxia in the critically ill patient
is compounded by a heightened catabolic state and anaer- Vascular surgery patients with marked atherosclerotic dis-
obic metabolism ensues. Therefore careful assessment of ease and underlying comorbidities represent a high risk
the quality of tissue oxygenation in these patients is essen- group when compared with most other patients undergoing
tial. Global tissue oxygenation is often assessed indirectly surgery. In many reports over half of those patients under-
by measuring plasma lactate levels which rise in response going elective AAA repair suffer one or more atherosclerosis
to excessive anaerobic metabolism. Elevated lactate levels, related complications.120 Multiorgan failure is the most
however, may reflect either impaired elimination, due to important cause of late death,121,122 the mortality rising with
hepatic failure and renal failure, or increased production, increasing number of failing organ systems.72 Postoperative
 Monitoring of organ dysfunction in SIRS 47

complications have been attributed to preoperative risk during this intensely catabolic process, and ideally via the
factors such as advancing age, poor ventricular function, enteral route. Recent advances in molecular biology have
ischaemic heart disease, chronic pulmonary disease, renal enhanced our understanding of the various actions and
failure and diabetes mellitus.76,123 Therefore, extensive interactions which represent the systemic inflammatory
efforts are being made, using scoring systems, to predict response to tissue injury. Beyond standard support of the
the risk of morbidity and mortality in patients prior to critically ill patient, attention has focused on the use of
commitment to a surgical procedure. either natural or synthetic therapeutic agents aimed at pro-
moting or inhibiting various components of the inflamma-
THE APACHE SCORE tory response. These are outlined in Table 4.5 and are
discussed briefly below.
Acute Physiology And Chronic Health Evaluation
(APACHE) II score72 is perhaps the most widely accepted
predictive score used to assess the severity of critical illness Anticytokine therapies
in ICUs. It is based on 12 physiological and laboratory based
factors (temperature, mean arterial pressure, heart rate, The balance between pro- and anti-inflammatory cytokines
respiratory rate, PO2, arterial pH, serum sodium, serum is crucial to the promotion or resolution of inflammation.
potassium, serum creatinine, haematocrit, white cell count Therapeutic interventions have ranged from purified con-
and Glasgow Coma Score), as well as on age and previous centrates of natural endogenous antibodies and stimulated
health status. In general, APACHE II is a good predictor of donor hyperimmune globulins to monoclonal antibodies
outcome in ruptured AAA repair but its power to predict directed against specific cytokines. In both animal experi-
outcome in any individual patient is limited.124 ments and human trials attention has focused on modula-
tion of the potent proinflammatory cytokine TNF- .
THE POSSUM SCORE Its normal biological activity is usually counterbalanced by
Physiological and Operative Severity Score for the
enUmeration of Mortality (POSSUM) was designed specif- Table 4.5 Novel potential therapeutic targets for the modula-
ically for patients undergoing surgical intervention. It is tion of systemic inflammatory response syndrome (SIRS) and
based on 12 readily available physiological and laboratory prevention of multiple organ dysfunction syndrome (MODS)
variables (age, cardiac signs, respiratory history, systolic
blood pressure, pulse rate, Glasgow Coma Score, haemo- Potential mode of
globin, white cell count, urea, serum sodium, serum potas- Therapies Target action
sium electrocardiogram) as well as on six operative severity
Cytokine TNF- and receptor Anti-inflammatory
parameters. The original POSSUM125 failed to predict out-
IL-1 , 6, 8, and 10 Anti-inflammatory
come sufficiently reliably after ruptured AAAs,124 but the and receptors
modified Portsmouth POSSUM or P-POSSUM126 would
Antiendotoxin Gut decontamination Reduced septic challenge
appear to have greater relevance to vascular patients.
LPS binding protein Improved endotoxin
(LBP) clearance
LPS Direct inhibition
Therapeutic strategies in MODS endotoxin
BPI, recombinant BPI Reduced inflammatory
The management of the critically ill patient with SIRS, sep- response to endotoxin
sis and evolving MODS is complex and best undertaken in Immune cell Leucocyte integrins Inhibits leucocyte
the ICU. The aim of initial resuscitation and supportive adhesion
therapies is to achieve and maintain adequate tissue oxy- Endothelial ICAM-1 Inhibits leucocyte
genation. Hypoxaemia should be managed by increased adhesion
inspired oxygen, where appropriate, to assist non-compliant Steroids Anti-inflammatory
and failing lungs by means of mechanical ventilation and Complement C1, sCR1 Inhibits classical
regular monitoring by blood gas analysis. Cardiovascular pathway
support using a combination of intravenous fluid, inotropes C3 Inhibits alternative
and vasoconstrictors may require invasive haemodynamic pathway
monitoring. Antibiotic therapy should be instituted early C5, C5a, C5aR Inhibits leucocyte
where signs of sepsis exist, initially with broad spectrum activation
antibiotics, and then in a more focused manner depending C5b-9 (MAC) Inhibits cytolysis
on microbiological results. Renal support can be achieved TNF, tumour necrosis factor; IL, interleukin; LPS, lipopolysaccharide;
by optimising renal perfusion, and where this fails, by BPI, bactericidal permeability increasing protein; ICAM, intercellular
means of haemodialysis. Nutrition should be maintained adhesion molecule.
48 Ischaemia-reperfusion injury, SIRS and MODS

natural inhibitors known as TNF- binding proteins and and is protective in animal models of meningococcal
identified as soluble forms of extracellular fragments of the sepsis145 and E. coli sepsis.146 Bactericidal/permeability-
TNF- receptors.127 In high risk patients early postoperative increasing protein (BPI) prevents LPS induced PMN acti-
rises in plasma levels of soluble TNF- receptors, a hallmark vation, TNF- production and, in addition to its ability to
of excessive TNF- production, are associated with a high neutralise LPS, BPI has been shown to alter bacterial mem-
complication rate and a poor prognosis.128 Despite encour- brane permeability and kill Gram-negative organisms.147,148
aging animal studies showing a lowering of mortality from Recently, recombinant BPI has been shown to attenuate
sepsis,129 in one phase II trial a dose dependent increase in systemic inflammation and acute lung injury after experi-
mortality was reported using the potent antagonist soluble mental lower limb IRI.149–151
TNF p75 receptor.130 Another proinflammatory cytokine
and potential target is IL-1 , but animal experiments here Modulation of immune cell function
are also inconclusive with antibodies directed against its
receptor: in small doses protecting those animals from As the key effectors of tissue injury, and the source of many
Klebsiella pneumoniae but in larger doses increasing the of the proinflammatory mediators present in SIRS, immune
lethal power of those organisms.131 The pleiotropic cytokine cells or leucocytes represent an obvious therapeutic target.
IL-6 has also been investigated, and the monoclonal anti- Polymorphonuclear leucocyte depletion has been shown
bodies directed against both IL-6 and its receptor having in animal studies to reduce reperfusion injury effect-
been shown to protect against lethal injury from TNF- , ively.94,95,152 In humans, however, global inhibition of leuco-
LPS and sepsis.132,133 Interestingly, plasma levels of an anti- cyte function is not a viable clinical option so that attempts
inflammatory cytokine IL-10 also rise in patients with septic to reduce their endothelial interaction and transmigration
shock134 and are correlated with an adverse outcome.135 seem a more realistic goal than limiting tissue injury.
Currently, IL-10 inhibition therapy is being explored in a Inhibition of PMN leucocyte adherence and migration,
phase I study in patients undergoing thoracoabdominal using monoclonal antibodies directed against the -chain of
aneurysm repair,136 the results of which are awaited. the CD11/CD18 glycoprotein adherence complex, decreases
PMN adherence to the endothelium and attenuates the
microvascular dysfunction associated with reperfusion of
Antiendotoxin therapy skeletal muscle.153 More recently anti-CD18 monoclonal
antibodies have been used to reduce multiple organ injury in
The sepsis syndrome, secondary to bacteraemia or endo-
an animal model of ruptured AAA.154
toxaemia, remains a leading cause of morbidity and mor-
tality, despite antibiotics and intensive care support. Gut
decontamination has been shown to cause a dramatic Complement inhibition
reduction in colonisation and effectively prevents bacterial
translocation from the gut, but it does not significantly The complement system is activated in a range of inflam-
reduce the mortality rate or hospital stay in critically ill matory states and has various potentially deleterious
patients.137 vasoactive and inflammatory effects. However, it also has
Lipopolysaccharide binding protein (LBP) is an essen- many beneficial effects, especially in the ability of the host
tial factor in the immune system responsible for meeting to resist septic challenge, and therefore, in order to be
bacterial or septic challenges. Initial optimism based on the effective, inhibition must be targeted. Encouraging results
finding that mice, injected with high concentrations of LBP, have been obtained using anti-C5 antibody, and C5a
could survive an otherwise lethal septic challenge138 was receptor (C5aR) antagonist, and the results of ongoing
tempered by studies in LBP-knockout mice which showed human trials are awaited.
that they fare no worse than their wild-type littermates
in response to septic challenge.139 Antiendotoxin immuno-
globulins have also been explored in several studies which Genetic therapies
demonstrated that normal pooled -globulin or immuno-
globulins can improve outcome in sepsis related condi- Quite recently it has become apparent that genetic poly-
tions.140 The Escherichia coli J5-immune plasma and morphisms to many of the previously described cytokines
-globulin study141 revealed some benefit using immune and inflammatory mediators do exist. Some of these poly-
plasma but it was minimal when immune -globulin was morphisms are functional in that it is possible to demon-
used.142 There was more convincing evidence of protection strate differing cytokine responses to a standard stimulus.
from exogenous -globulins in a clinical study using ‘nat- This would suggest the possibility of genetic predisposition
ural’ antiendotoxin antibodies.143 Endotoxin neutralising to increased mortality from sepsis or SIRS in some individ-
protein antilipopolysaccharide factor isolated from the uals. The first such polymorphism to be described in this
amoebocytes of horseshoe crabs, Limulus polyphemus and field was in relation to TNF, but they have also been noted
Limulus tachypleus, is known to bind various endotoxins,144 in regard to IL-1 , IL-1 receptor antagonist and IL-10.155
 References 49

5 Campbell WB, Collin J, Morris PJ. The mortality of abdominal

Conclusions aortic aneurysm. Ann R Coll Surg Engl 1998; 68: 275–8.
6 Dormandy J, Heeck L, Vig S. The natural history of
The lack of improvement in mortality rates in those vas- claudication: risk to life and limb. Semin Vasc Surg 1999; 12:
cular patients who develop MODS probably reflects an 123–37.
increase in the number and complexity of vascular 7 Leng GC, Lee AJ, Fowkes FG, et al. Incidence, natural history
interventions in an increasingly elderly and unfit patient and cardiovascular events in symptomatic and asymptomatic
population rather than a failure of modern intensive peripheral arterial disease in the general population. Int
care support. It is expected that a more enlightened J Epidemiol 1996; 25: 1172–81.
understanding of the mechanisms involved in the devel- 8 Halpern VJ, Kline RG, D’Angelo AJ, Cohen JR. Factors that affect
the survival rate of patients with ruptured abdominal aortic
opment of IRI and its progression via SIRS to MODS
aneurysms. J Vasc Surg 1997; 26: 939–45.
will contribute to improve outcomes. Finally, it cannot
9 Hardman DT, Fisher CM, Patel MI, et al. Ruptured abdominal
be sufficiently emphasised that prevention of MOF by aortic aneurysms: who should be offered surgery? J Vasc Surg
optimising tissue oxygen delivery at all stages is crucial 1996; 23: 123–9.
to achieving better results. Careful assessment of a 10 Groeneveld AB, Raijmakers PG, Rauwerda JA, Hack CE. The
patient’s general medical condition and risk factors inflammatory response to vascular surgery-associated ischaemia
allows suitable patient selection and preventive treat- and reperfusion in man: effect on postoperative pulmonary
ment strategies to be put in place. function. Eur J Vasc Endovasc Surg 1997; 14: 351–9.
11 Barry MC, Kelly C, Burke P, et al. Immunological and
physiological responses to aortic surgery: effect of reperfusion
on neutrophil and monocyte activation and pulmonary function.
Key references Br J Surg 1997; 84: 513–19.
12 Welbourn CR, Goldman G, Paterson IS, et al. Pathophysiology of
Baigrie RJ, Lamont PM, Whiting S, Morris PJ. Portal endotoxin ischaemia reperfusion injury: central role of the neutrophil.
and cytokine responses during abdominal aortic surgery. Br J Surg 1991; 78: 651–5.
Am J Surg 1993; 166: 248–51. 13 Salzman AL, Wang H, Wollert PS, et al. Endotoxin-induced ileal
Baue AE. Multiple organ failure, multiple organ dysfunction mucosal hyperpermeability in pigs: role of tissue acidosis.
syndrome, and systemic inflammatory response syndrome. Am J Physiol 1994; 266: G633–46.
Why no magic bullets? Arch Surg 1997; 132: 703–7. 14 Anner H, Kaufman RP Jr, Valeri CR, et al. Reperfusion of ischemic
Groeneveld AB, Raijmakers PG, Rauwerda JA, Hack CE. The lower limbs increases pulmonary microvascular permeability.
inflammatory response to vascular surgery-associated J Trauma 1988; 28: 607–10.
ischaemia and reperfusion in man: effect on postoperative 15 Welbourn R, Goldman G, Kobzik L, et al. Role of neutrophil
pulmonary function. Eur J Vasc Endovasc Surg 1997; 14: adherence receptors (CD 18) in lung permeability following lower
351–9. torso ischemia. Circ Res 1992; 71: 82–6.
Halpern VJ, Kline RG, D’Angelo AJ, Cohen JR. Factors that affect 16 Gadaleta D, Fantini GA, Silane MF, Davis JM. Neutrophil
the survival rate of patients with ruptured abdominal aortic leukotriene generation and pulmonary dysfunction after
aneurysms. J Vasc Surg 1997; 26: 939–45. abdominal aortic aneurysm repair. Surgery 1994; 116:
Marshall JC, Christou NV, Meakins JL. The gastrointestinal tract. 847–52.
The ‘undrained abscess’ of multiple organ failure. 17 Klausner JM, Anner H, Paterson IS, et al. Lower torso ischemia-
Ann Surg 1993; 218: 111–19. induced lung injury is leukocyte dependent. Ann Surg 1988;
208: 761–7.
18 Barry MC, Wang JH, Kelly CJ, et al. Plasma factors augment
neutrophil and endothelial cell activation during aortic surgery.
REFERENCES Eur J Vasc Endovasc Surg 1997; 13: 381–7.
19 Murphy ME, Kolvenbach R, Aleksis M, et al. Antioxidant depletion
in aortic crossclamping ischemia: increase of the plasma alpha-
1 Adam DJ, Mohan IV, Stuart WP, et al. Community and hospital tocopheryl quinone/alpha-tocopherol ratio. Free Radic Biol Med
outcome from ruptured abdominal aortic aneurysm within the 1992; 13: 95–100.
catchment area of a regional vascular surgical service. J Vasc 20 Paterson IS, Klausner JM, Goldman G, et al. Thromboxane
Surg 1999; 30: 922–8. mediates the ischemia-induced neutrophil oxidative burst.
2 Campbell WB, Ridler BM, Szymanska TH. Current management of Surgery 1989; 106: 224–9.
acute leg ischaemia: results of an audit by the Vascular Surgical 21 Doukas J, Hechtman HB, Shepro D. Endothelial-secreted
Society of Great Britain and Ireland. Br J Surg 1998; 85: arachidonic acid metabolites modulate polymorphonuclear
1498–503. leukocyte chemotaxis and diapedesis in vitro. Blood 1988;
3 Henderson A, Effeney D. Morbidity and mortality after abdominal 71: 771–9.
aortic surgery in a population of patients with high 22 Kuo PC, Schroeder RA. The emerging multifaceted roles of nitric
cardiovascular risk. Aust N Z J Surg 1995; 65: 417–20. oxide. Ann Surg 1995; 221: 220–35.
4 Go LL, Healey PJ, Watkins SC, et al. The effect of endotoxin on 23 McMillen MA, Huribal M, Sumpio B. Common pathway of
intestinal mucosal permeability to bacteria in vitro. Arch Surg endothelial-leukocyte interaction in shock, ischemia, and
1995; 130: 53–8. reperfusion. Am J Surg 1993; 166: 557–62.
50 Ischaemia-reperfusion injury, SIRS and MODS

24 Summers RW, Hayek B. Changes in colonic motility following 42 Botha AJ, Moore FA, Moore EE, et al. Early neutrophil
abdominal irradiation in dogs. Am J Physiol 1993; 264: sequestration after injury: a pathogenic mechanism for multiple
G1024–G1030. organ failure. J Trauma 1995; 39: 411–17.
25 McCall TB, Boughton-Smith NK, Palmer RM, et al. Synthesis of 43 Cohen AB, MacArthur C, Idell S, et al. A peptide from alveolar
nitric oxide from L-arginine by neutrophils. Release and macrophages that releases neutrophil enzymes into the
interaction with superoxide anion. Biochem J 1989; 261: lungs in patients with the adult respiratory distress syndrome.
293–6. Am Rev Respir Dis 1988; 137: 1151–8.
26 Baue AE. Multiple organ failure, multiple organ dysfunction 44 Nickoloff BJ, Fivenson DP, Kunkel SL, et al. Keratinocyte
syndrome, and systemic inflammatory response syndrome. interleukin-10 expression is upregulated in tape-stripped skin,
Why no magic bullets? Arch Surg 1997; 132: 703–7. poison ivy dermatitis, and Sezary syndrome, but not in psoriatic
27 Moore FA, Moore EE. Evolving concepts in the pathogenesis of plaques. Clin Immunol Immunopathol 1994; 73: 63–8.
postinjury multiple organ failure. Surg Clin North Am 1995; 75: 45 Bogdan C, Vodovotz Y, Nathan C. Macrophage deactivation by
257–77. interleukin 10. J Exp Med 1991; 174: 1549–55.
28 Bone RC. Immunologic dissonance: a continuing evolution in our 46 Marchant A, Bruyns C, Vandenabeele P, et al. Interleukin-10
understanding of the systemic inflammatory response syndrome controls interferon-gamma and tumor necrosis factor production
(SIRS) and the multiple organ dysfunction syndrome (MODS) during experimental endotoxemia. Eur J Immunol 1994; 24:
[see comments]. Ann Intern Med 1996; 125: 680–7. 1167–71.
29 Froon AH, Greve JW, Van der Linden CJ, Buurman WA. Increased 47 Fine J. Endotoxaemia in man. Lancet 1972; ii: 181.
concentrations of cytokines and adhesion molecules in patients 48 LeVoyer T, Cioffi WG Jr, Pratt L, et al. Alterations in intestinal
after repair of abdominal aortic aneurysm. Eur J Surg 1996; 162: permeability after thermal injury. Arch Surg 1992; 127: 26–9.
287–96. 49 Pape HC, Dwenger A, Regel G, et al. Increased gut permeability
30 Roumen RM, Hendriks T, van der Ven-Jongekrijg J, et al. Cytokine after multiple trauma. Br J Surg 1994; 81: 850–2.
patterns in patients after major vascular surgery, hemorrhagic 50 van Deventer SJ, ten Cate JW, Tytgat GN. Intestinal endotoxemia.
shock, and severe blunt trauma. Relation with subsequent Clinical significance. Gastroenterology 1988; 94: 825–31.
adult respiratory distress syndrome and multiple organ failure. 51 Corson RJ, Paterson IS, O’Dwyer ST, et al. Lower limb ischaemia
Ann Surg 1993; 218: 769–76. and reperfusion alters gut permeability. Eur J Vasc Surg 1992;
31 May LT, Santhanam U, Tatter SB, et al. Phosphorylation of 6: 158–63.
secreted forms of human beta 2-interferon/hepatocyte 52 Yassin MMI, Barros D’Sa AAB, Parks TG, et al. Lower limb
stimulating factor/interleukin-6. Biochem Biophys Res Commun ischaemia-reperfusion injury alters gastrointestinal structure
1988; 152: 1144–50. and function. Br J Surg 1997; 84: 1425–9.
32 Sternbergh WC 3rd, Tuttle TM, Makhoul RG, et al. Postischemic 53 Deitch EA, Maejima K, Berg R. Effect of oral antibiotics and
extremities exhibit immediate release of tumor necrosis factor. bacterial overgrowth on the translocation of the GI tract
J Vasc Surg 1994; 20: 474–81. microflora in burned rats. J Trauma 1985; 25: 385–92.
33 Tracey KJ, Lowry SF, Fahey TJ 3rd, et al. Cachectin/tumor necrosis 54 Deitch EA, Ma L, Ma WJ, et al. Inhibition of endotoxin-
factor induces lethal shock and stress hormone responses in the induced bacterial translocation in mice. J Clin Invest 1989;
dog. Surg Gynecol Obstet 1987; 164: 415–22. 84: 36–42.
34 Witthaut R, Farhood A, Smith CW, Jaeschke H. Complement and 55 Deitch EA, Bridges W, Baker J, et al. Hemorrhagic shock-induced
tumor necrosis factor-alpha contribute to Mac-1 (CD11b/CD18) bacterial translocation is reduced by xanthine oxidase inhibition
up-regulation and systemic neutrophil activation during or inactivation. Surgery 1988; 104: 191–8.
endotoxemia in vivo. J Leukoc Biol 1994; 55: 105–11. 56 Ambrose NS, Johnson M, Burdon DW, Keighley MR. Incidence of
35 Abraham E. Effects of stress on cytokine production. Methods pathogenic bacteria from mesenteric lymph nodes and ileal
Achiev Exp Pathol 1991; 14: 45–62. serosa during Crohn’s disease surgery. Br J Surg 1984; 71: 623–5.
36 Biffl WL, Moore EE, Moore FA, Peterson VM. Interleukin-6 in the 57 O’Gorman RB, Feliciano DV, Matthews KS, et al. Correlation of
injured patient. Marker of injury or mediator of inflammation? immunologic and nutritional status with infectious
Ann Surg 1996; 224: 647–64. complications after major abdominal trauma. Surgery 1986;
37 Baigrie RJ, Lamont PM, Whiting S, Morris PJ. Portal endotoxin 99: 549–56.
and cytokine responses during abdominal aortic surgery. 58 Soong CV, Blair PH, Halliday MI, et al. Endotoxaemia, the
Am J Surg 1993; 166: 248–51. generation of the cytokines and their relationship to
38 Svoboda P, Kantorova I, Ochmann J. Dynamics of interleukin 1, 2, intramucosal acidosis of the sigmoid colon in elective abdominal
and 6 and tumor necrosis factor alpha in multiple trauma aortic aneurysm repair. Eur J Vasc Surg 1993; 7: 534–9.
patients. J Trauma 1994; 36: 336–40. 59 Woodcock NP, Sudheer V, el Barghouti N, et al. Bacterial
39 Yassin MM, Harkin DW, Barros D’Sa AAB, et al. Lower limb translocation in patients undergoing abdominal aortic aneurysm
ischaemia-reperfusion injury triggers a systemic inflammatory repair. Br J Surg 2000; 87: 439–42.
response and multiple organ dysfunction. World J Surg 2002; 60 van Deventer SJ, Buller HR, ten Cate JW, et al. Experimental
26: 115–21. endotoxemia in humans: analysis of cytokine release and
40 Ulich TR, del Castillo J, Yin SM, Egrie JC. The erythropoietic coagulation, fibrinolytic, and complement pathways. Blood 1990;
effects of interleukin 6 and erythropoietin in vivo. Exp Hematol 76: 2520–6.
1991; 19: 29–34. 61 Foulds S, Cheshire NJ, Schachter M, et al. Endotoxin related
41 Shalaby MR, Halgunset J, Haugen OA, et al. Cytokine-associated early neutrophil activation is associated with outcome after
tissue injury and lethality in mice: a comparative study. thoracoabdominal aortic aneurysm repair. Br J Surg 1997;
Clin Immunol Immunopathol 1991; 61: 69–82. 84: 172–7.
 References 51

62 Roumen RM, Frieling JT, van Tits HW, et al. Endotoxemia after 84 Hakim AA. An immunosuppressive factor from serum of
major vascular operations. J Vasc Surg 1993; 18: 853–7. thermally traumatized patients. J Trauma 1977; 17: 908–19.
63 Roumen RM, Hendriks T, Wevers RA, Goris JA. Intestinal 85 Munster AM. Post-traumatic immunosuppression is due to
permeability after severe trauma and hemorrhagic shock is activation of suppressor T cells. Lancet 1976; 1: 1329–30.
increased without relation to septic complications. Arch Surg 86 Munster AM, Hoagland HC, Pruitt BA Jr. The effect of thermal
1993; 128: 453–7. injury on serum immunoglobulins. Ann Surg 1970; 172: 965–9.
64 Yassin MMI, Barros D’Sa AAB, Parks TG, et al. Lower limb 87 Hurd TC, Dasmahapatra KS, Rush BF Jr, Machiedo GW. Red
ischaemia-reperfusion injury causes endotoxaemia and blood cell deformability in human and experimental sepsis.
endogenous antiendotoxin antibody consumption but not Arch Surg 1988; 123: 217–20.
bacterial translocation. Br J Surg 1998; 85: 785–9. 88 Bradbury A, Adam D, Garrioch M, et al. Changes in platelet
65 Yassin MM, Barros D’Sa AAB, Parks TG, et al. Mortality count, coagulation and fibrinogen associated with elective
following lower limb ischemia-reperfusion: a systemic repair of asymptomatic abdominal aortic aneurysm and aortic
inflammatory response? World J Surg 1996; 20: 961–7. reconstruction for occlusive disease. Eur J Vasc Endovasc Surg
66 Van Damme H, Creemers E, Limet R. Ischaemic colitis 1997; 13: 375–80.
following aortoiliac surgery. Acta Chir Belg 2000; 100: 89 Davies MJ, Murphy WG, Murie JA, et al. Preoperative
21–7. coagulopathy in ruptured abdominal aortic aneurysm predicts
67 Welch M, Baguneid MS, McMahon RF, et al. Histological study poor outcome. Br J Surg 1993; 80: 974–6.
of colonic ischaemia after aortic surgery. Br J Surg 1998; 85: 90 Hoffman MJ, Greenfield LJ, Sugerman HJ, Tatum JL.
1095–8. Unsuspected right ventricular dysfunction in shock and sepsis.
68 Dobbins WO, III. Gut immunophysiology: a gastroenterologist’s Ann Surg 1983; 198: 307–19.
view with emphasis on pathophysiology. Am J Physiol 1982; 91 Levison M. Myocardial failure. Surg Clin North Am 1982; 62:
242: G1–8. 149–56.
69 Glauser MP, Zanetti G, Baumgartner JD, Cohen J. Septic shock: 92 Lefer AM. Role of a myocardial depressant factor in the
pathogenesis. Lancet 1991; 338: 732–6. pathogenesis of circulatory shock. Fed Proc 1970; 29:
70 Bone RC. Why sepsis trials fail. JAMA 1996; 276: 565–6. 1836–47.
71 Fry DE, Pearlstein L, Fulton RL, Polk HC, Jr. Multiple system 93 Haglund U. Systemic mediators released from the gut in critical
organ failure. The role of uncontrolled infection. Arch Surg illness. Crit Care Med 1993; 21: S15–S18.
1980; 115: 136–40. 94 Klausner JM, Anner H, Paterson IS, et al. Lower torso ischemia-
72 Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: induced lung injury is leukocyte dependent. Ann Surg 1988;
a severity of disease classification system. Crit Care Med 1985; 208: 761–7.
13: 818–29. 95 Korthuis RJ, Grisham MB, Granger DN. Leukocyte depletion
73 Bone RC. The sepsis syndrome. Definition and general approach attenuates vascular injury in postischemic skeletal muscle.
to management. Clin Chest Med 1996; 17: 175–81. Am J Physiol 1988; 254: H823–H827.
74 Carrico CJ, Meakins JL, Marshall JC, et al. Multiple-organ- 96 Leeman M. Pulmonary hypertension in acute respiratory
failure syndrome. Arch Surg 1986; 121: 196–208. distress syndrome. Monaldi Arch Chest Dis 1999; 54: 146–9.
75 Paterson IS, Klausner JM, Pugatch R, et al. Noncardiogenic 97 Parsons PE, Worthen GS, Moore EE, et al. The association of
pulmonary edema after abdominal aortic aneurysm surgery. circulating endotoxin with the development of the adult
Ann Surg 1989; 209: 231–6. respiratory distress syndrome. Am Rev Respir Dis 1989; 140:
76 Huber TS, Harward TR, Flynn TC, et al. Operative mortality rates 294–301.
after elective infrarenal aortic reconstructions. J Vasc Surg 98 Wollert PS, Menconi MJ, O’Sullivan BP, et al. LY255283, a novel
1995; 22: 287–93. leukotriene B4 receptor antagonist, limits activation of
77 Harris LM, Faggioli GL, Fiedler R, et al. Ruptured abdominal neutrophils and prevents acute lung injury induced by
aortic aneurysms: factors affecting mortality rates. J Vasc Surg endotoxin in pigs. Surgery 1993; 114: 191–8.
1991; 14: 812–18. 99 Gilroy RJ Jr, Lavietes MH, Loring SH, et al. Respiratory
78 Goris RJ, te Boekhorst TP, Nuytinck JK, Gimbrere JS. Multiple- mechanical effects of abdominal distension. J Appl Physiol
organ failure. Generalized autodestructive inflammation? Arch 1985; 58: 1997–2003.
Surg 1985; 120: 1109–15. 100 Barratt J, Parajasingam R, Sayers RD, Feehally J. Outcome of
79 Marshall JC, Christou NV, Meakins JL. The gastrointestinal tract. Acute Renal Failure Following Surgical Repair of Ruptured
The ‘undrained abscess’ of multiple organ failure. Ann Surg Abdominal Aortic Aneurysms. Eur J Vasc Endovasc Surg 2000;
1993; 218: 111–19. 20: 163–8.
80 Deitch EA. Multiple organ failure. Pathophysiology and potential 101 Gelman S. The pathophysiology of aortic cross-clamping and
future therapy. Ann Surg 1992; 216: 117–34. unclamping. Anesthesiology 1995; 82: 1026–60.
81 Bjornson AB, Bjornson HS, Altemeier WA. Serum-mediated 102 Grace PA. Ischaemia-reperfusion injury. Br J Surg 1994; 81:
inhibition of polymorphonuclear leukocyte function following 637–47.
burn injury. Ann Surg 1981; 194: 568–75. 103 Neveu H, Kleinknecht D, Brivet F, et al. Prognostic factors in
82 Rubin BB, Smith A, Liauw S, et al. Complement activation acute renal failure due to sepsis. Results of a prospective
and white cell sequestration in postischemic skeletal muscle. multicentre study. The French Study Group on Acute Renal
Am J Physiol 1990; 259: H525-H531. Failure. Nephrol Dial Transplant 1996; 11: 293–9.
83 Wood JJ, Grbic JT, Rodrick ML, et al. Suppression of interleukin 104 Gutierrez G, Palizas F, Doglio G, et al. Gastric intramucosal
2 production in an animal model of thermal injury is related to pH as a therapeutic index of tissue oxygenation in critically ill
prostaglandin synthesis. Arch Surg 1987; 122: 179–84. patients. Lancet 1992; 339: 195–9.
52 Ischaemia-reperfusion injury, SIRS and MODS

105 Bjorck M, Hedberg B. Early detection of major complications 124 Lazarides MK, Arvanitis DP, Drista H, et al. POSSUM and
after abdominal aortic surgery: predictive value of sigmoid APACHE II scores do not predict the outcome of ruptured
colon and gastric intramucosal pH monitoring. Br J Surg 1994; infrarenal aortic aneurysms. Ann Vasc Surg 1997; 11: 155–8.
81: 25–30. 125 Copeland GP, Jones D, Walters M. POSSUM: a scoring system
106 Derbyshire DR, Smith G. Sympathoadrenal responses to for surgical audit. Br J Surg 1991; 78: 355–60.
anaesthesia and surgery. Br J Anaesth 1984; 56: 725–39. 126 Prytherch DR, Whiteley MS, Higgins B, et al. POSSUM and
107 Thompson JP, Boyle JR, Thompson MM, et al. Cardiovascular Portsmouth POSSUM for predicting mortality. Physiological and
and catecholamine responses during endovascular and Operative Severity Score for the enUmeration of Mortality and
conventional abdominal aortic aneurysm repair. Eur J Vasc morbidity. Br J Surg 1998; 85: 1217–20.
Endovasc Surg 1999; 17: 326–33. 127 Seckinger P, Isaaz S, Dayer JM. A human inhibitor of tumor
108 Zuckerman SH, Bendele AM. Regulation of serum tumor necrosis factor alpha. J Exp Med 1988; 167: 1511–16.
necrosis factor in glucocorticoid-sensitive and -resistant rodent 128 Pilz G, Kaab S, Kreuzer E, Werdan K. Evaluation of definitions
endotoxin shock models. Infect Immun 1989; 57: 3009–13. and parameters for sepsis assessment in patients after cardiac
109 Ray A, LaForge KS, Sehgal PB. On the mechanism for efficient surgery. Infection 1994; 22: 8–17.
repression of the interleukin-6 promoter by glucocorticoids: 129 Tracey KJ, Fong Y, Hesse DG, et al. Anti-cachectin/TNF
enhancer, TATA box, and RNA start site (Inr motif) occlusion. monoclonal antibodies prevent septic shock during lethal
Mol Cell Biol 1990; 10: 5736–46. bacteraemia. Nature 1987; 330: 662–4.
110 Slag MF, Morley JE, Elson MK, et al. Free thyroxine levels in 130 Fisher CJ Jr, Agosti JM, Opal SM, et al. Treatment of septic
critically ill patients. A comparison of currently available shock with the tumor necrosis factor receptor:Fc fusion protein.
assays. JAMA 1981; 246: 2702–6. The Soluble TNF Receptor Sepsis Study Group. N Engl J Med
111 Ozawa K. [Clinical and biological significance of arterial blood 1996; 334: 1697–702.
ketone body ratio in hepatic surgery] (Article in Japanese). 131 Mancilla J, Garcia P, Dinarello CA. The interleukin-1 receptor
Nippon Geka Gakkai Zasshi 1983; 84: 753–7. antagonist can either reduce or enhance the lethality of
112 Oudemans-van Straaten HM, Scheffer GJ, et al. Oxygen Klebsiella pneumoniae sepsis in newborn rats. Infect Immun
consumption after cardiopulmonary bypass – implications of 1993; 61: 926–32.
different measuring methods. Intensive Care Med 1993; 19: 132 Libert C, Vink A, Coulie P, et al. Limited involvement of
105–10. interleukin-6 in the pathogenesis of lethal septic shock as
113 Oldenburg HS, Rogy MA, Lazarus DD, et al. Cachexia and the revealed by the effect of monoclonal antibodies against
acute-phase protein response in inflammation are regulated interleukin-6 or its receptor in various murine models.
by interleukin-6. Eur J Immunol 1993; 23: 1889–94. Eur J Immunol 1992; 22: 2625–30.
114 Shoemaker WC, Appel PL, Kram HB. Tissue oxygen debt as a 133 Gennari R, Alexander JW. Anti-interleukin-6 antibody
determinant of lethal and nonlethal postoperative organ treatment improves survival during gut- derived sepsis in a
failure. Crit Care Med 1988; 16: 1117–20. time-dependent manner by enhancing host defense. Crit Care
115 Swan HJ, Ganz W, Forrester J, et al. Catheterization of the heart Med 1995; 23: 1945–53.
in man with use of a flow-directed balloon-tipped catheter. 134 Derkx B, Marchant A, Goldman M, et al. High levels of
N Engl J Med 1970; 283: 447–51. interleukin-10 during the initial phase of fulminant
116 Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of meningococcal septic shock. J Infect Dis 1995; 171: 229–32.
right heart catheterization in the initial care of critically ill 135 Neidhardt R, Keel M, Steckholzer U, et al. Relationship of
patients. SUPPORT Investigators. JAMA 1996; 276: 889–97. interleukin-10 plasma levels to severity of injury and clinical
117 Shoemaker WC, Appel PL, Kram HB. Tissue oxygen debt as a outcome in injured patients. J Trauma 1997; 42: 863–70.
determinant of lethal and nonlethal postoperative organ 136 Huber TS, Gaines GC, Welborn MB 3rd, et al.
failure. Crit Care Med 1988; 16: 1117–20. Anticytokine therapies for acute inflammation and the
118 Fiddian-Green RG, Amelin PM, Herrmann JB, et al. Prediction systemic inflammatory response syndrome: IL-10 and
of the development of sigmoid ischemia on the day of aortic ischemia/reperfusion injury as a new paradigm. Shock 2000;
operations. Indirect measurements of intramural pH in the 13: 425–34.
colon. Arch Surg 1986; 121: 654–60. 137 Blair P, Rowlands BJ, Lowry K, et al. Selective decontamination
119 Mucelli RP, Bertolotto M. Imaging techniques in acute renal of the digestive tract: a stratified, randomized, prospective
failure. Kidney Int Suppl 1998; 66: S102-S105. study in a mixed intensive care unit. Surgery 1991; 110:
120 Crawford ES, Saleh SA, Babb JW 3rd, et al. Infrarenal abdominal 303–9.
aortic aneurysm: factors influencing survival after operation 138 Lamping N, Dettmer R, Schroder NW, et al. LPS-binding
performed over a 25-year period. Ann Surg 1981; 193: protein protects mice from septic shock caused by LPS or
699–709. Gram-negative bacteria. J Clin Invest 1998; 101: 2065–71.
121 Gloviczki P, Pairolero PC, Mucha P Jr, et al. Ruptured abdominal 139 Jack RS, Fan X, Bernheiden M, et al. Lipopolysaccharide-binding
aortic aneurysms: repair should not be denied. J Vasc Surg protein is required to combat a murine Gram-negative
1992; 15: 851–7. bacterial infection. Nature 1997; 389: 742–5.
122 Harris KA, Ameli FM, Lally M, et al. Abdominal aortic aneurysm 140 Pilz G, Appel R, Kreuzer E, Werdan K. Comparison of early
resection in patients more than 80 years old. Surg Gynecol IgM-enriched immunoglobulin vs polyvalent IgG administration
Obstet 1986; 162: 536–8. in score-identified postcardiac surgical patients at high risk for
123 Bjerkelund CE, Smith-Erichsen N, Solheim K. Abdominal aortic sepsis. Chest 1997; 111: 419–26.
reconstruction. Prognostic importance of coexistent diseases. 141 Ziegler EJ, Fisher CJ Jr, Sprung CL, et al. Treatment of Gram-
Acta Chir Scand 1986; 152: 111–15. negative bacteremia and septic shock with HA-1A human
 References 53

monoclonal antibody against endotoxin. A randomized, double- factor release induced by the bacteria. J Clin Invest 1992; 90:
blind, placebo-controlled trial. The HA-1A Sepsis Study Group. 1122–30.
N Engl J Med 1991; 324: 429–36. 149 Harkin DW, Barros D’Sa AAB, Yassin MM, et al. Recombinant
142 Cometta A, Baumgartner JD, Lew D, et al. Prospective bactericidal/permeability-increasing protein attenuates
randomized comparison of imipenem monotherapy with the systemic inflammatory response syndrome in lower
imipenem plus netilmicin for treatment of severe infections in limb ischaemia-reperfusion injury. J Vasc Surg 2001;
nonneutropenic patients. Antimicrob Agents Chemother 1994; 33: 840–6.
38: 1309–13. 150 Harkin DW, Barros D’Sa AAB, McCallion K, et al.
143 Fomsgaard A, Baek L, Fomsgaard JS, Engquist A. Preliminary Bactericidal/permeability-increasing protein attenuates
study on treatment of septic shock patients with systemic inflammation and acute lung injury in porcine lower
antilipopolysaccharide IgG from blood donors. Scand J Infect limb ischaemia-reperfusion injury. Ann Surg 2001; 234:
Dis 1989; 21: 697–708. 233–44.
144 Muta T, Miyata T, Tokunaga F, et al. Primary structure of anti- 151 Harkin DW, Barros D’Sa AAB, Yassin MM, et al. Gut mucosal
lipopolysaccharide factor from American horseshoe crab, Limulus injury is attenuated by recombinant bactericidal/permeability-
polyphemus [published erratum appears in J Biochem (Tokyo) increasing protein in hind limb ischaemia-reperfusion injury.
1987; 102:443]. J Biochem (Tokyo) 1987; 101: 1321–30. Ann Vasc Surg 2001; 15: 326–31.
145 Alpert G, Baldwin G, Thompson C, et al. Limulus 152 Belkin M, LaMorte WL, Wright JG, Hobson RW. The role of
antilipopolysaccharide factor protects rabbits from leukocytes in the pathophysiology of skeletal muscle ischemic
meningococcal endotoxin shock. J Infect Dis 1992; 165: 494–500. injury. J Vasc Surg 1989; 10: 14–18.
146 Saladino RA, Stack AM, Thompson C, et al. High-dose 153 Carden DL, Smith JK, Korthuis RJ. Neutrophil-mediated
recombinant endotoxin neutralizing protein improves survival microvascular dysfunction in postischemic canine skeletal
in rabbits, with Escherichia coli sepsis. Crit Care Med 1996; 24: muscle. Role of granulocyte adherence. Circ Res 1990; 66:
1203–7. 1436–44.
147 Marra MN, Wilde CG, Griffith JE, et al. 154 Boyd AJ, Rubin BB, Walker PM, et al. A CD18 monoclonal
Bactericidal/permeability-increasing protein has endotoxin- antibody reduces multiple organ injury in a model of ruptured
neutralizing activity. J Immunol 1990; 144: 662–6. abdominal aortic aneurysm. Am J Physiol 1999; 277:
148 Weiss J, Elsbach P, Shu C, et al. Human H172–82.
bactericidal/permeability-increasing protein and a recombinant 155 Vincent JL. Microvascular endothelial dysfunction: a renewed
NH2-terminal fragment cause killing of serum-resistant gram- appreciation of sepsis pathophysiology. Crit Care 2001;
negative bacteria in whole blood and inhibit tumor necrosis 5(suppl 2): S1–5.
This page intentionally left blank
 5
Pathophysiology of Stroke

HARVEY J CHANT

The problem 55 The concept of the ischaemic penumbra 57

Rationale for acute treatment of ischaemic stroke 56 Clinical consequences of reperfusing ischaemic brain 59
The pathogenesis of cerebral infarction: cell death 56 Logistical problems 61
and apoptosis References 62

THE PROBLEM
The Oxfordshire Community Stroke Project (OCSP)
investigated incidence and cause of first-time strokes and
Stroke can be defined as a rapid onset neurological deficit transient ischaemic attacks (TIAs) in the community.2
of presumed vascular origin lasting for more than 24 hours Using clinical assessment this project classified ischaemic
and which may result in death.1 Each year in the UK, over stroke into four groups: total anterior circulation infarc-
100 000 people are affected by this disease2 and on a world- tion (TACI), partial anterior circulation infarction (PACI),
wide scale, it is estimated that there are 4.5 million deaths lacunar infarction and posterior cerebral infarction. This
from stroke per year.3 Furthermore, with the increasing size classification system is a valuable tool for the studies of stroke
of the elderly population anticipated over the next few years, subtype and incidence. In a large epidemiological survey
this number is likely to rise.3,4 The prevalence of stroke is based in Manchester, UK, Mead et al.8 investigated the
dependent upon case fatality and incidence. In the USA, prevalence of carotid atherosclerosis in the four OCSP
stroke affects 550 000 people per year; of these 150 000 die, stroke categories. Of 305 patients with cerebral infarction,
approximately 350 000 are disabled and there are an esti- 201 (66 per cent) were classified as either PACI or TACI.
mated 3 million survivors from stroke.5 A study of the Furthermore, in the TACI group (100 patients) four
pathophysiology of stroke suggests that it is likely that both patients had ipsilateral severe (70–99 per cent) internal
the survival and the quality of life of stroke survivors might carotid artery (ICA) stenosis and 25 had ipsilateral ICA
be improved by early reperfusion of the area of brain occlusion. In the PACI group, 16 had ipsilateral severe ICA
immediately adjacent to the core of cerebral infarction. stenosis and 11 had ICA occlusion.8 ICA atherosclerosis is
The aetiology of stroke depends on the age of the patient a well recognised cause of stroke.9 Mead’s study quantifies
so that although stroke in the young is rare, and stroke caused the contribution of severe ICA stenosis and ICA occlusion to
by atherosclerosis is exceedingly rare. However, with increas- the epidemiology of stroke and confirms ICA disease as a
ing age, it becomes the predominant cause and because of the major target for stroke prevention and treatment (Fig. 5.1).
greater number of strokes occurring in the elderly, athero- In patients surviving stroke, the personal cost is often
sclerosis overtakes all other causes of stroke.6 It can be diffi- considerable. In the OCSP, 92 of 543 patients with cerebral
cult to establish the cause of cerebral infarction and in up to infarction were classified as TACIs. Of this subgroup only
40 per cent of cases a cause is never identified. Furthermore, 4 per cent of patients were independent and 56 per cent
the relative contribution of each risk factor is difficult to were still dependent at 30 days following stroke.10 The num-
ascertain, for example ischaemic heart disease and atrial fib- ber of patients remaining independent was unchanged at
rillation often coexist making the diagnosis unclear.6 Cerebral one year. Although stroke is primarily a disease of the eld-
infarction accounts for approximately 85 per cent of strokes erly, a significant proportion of cases occur in patients
and the mechanism of infarction is small and large vessel ath- under retirement age. For example in an American study
erosclerosis (45 per cent), cardiogenic embolism (20 per the annual incident rate for first ever or recurrent stroke in
cent), and cryptogenic and unusual causes (35 per cent).7 men between the ages of 55 and 64 years was 458 per
56 Pathophysiology of stroke

It is well recognised that carotid stenosis predisposes to

Stroke
recurrent stroke9,17 and there are series that suggest patients
with severe carotid stenosis may be at a particularly high
risk of early stroke recurrence.18–20 The potential benefit of
Ischaemic stroke (85 per cent)7 early carotid surgery in preventing acute stroke has now
been demonstrated in a trial of 25 patients with PACI and
carotid stenosis (70–99 per cent) randomised to surgery
either within 1 week of stroke or at 2 months. The reinfarction
Anterior circulation stroke (66 per cent)8
rate was significantly lower in the group undergoing urgent
surgery.21 (See Chapters 12 and 13.)
In carotid artery occlusion the situation is less clear. It has
Anterior circulation stroke associated with carotid stenosis previously been assumed that once an ICA has occluded it
or occlusion (18 per cent)8 will be unlikely to cause further embolisation. It is clear,
however, that such patients do have an increased risk of
Figure 5.1 Relative proportions of patients presenting with subsequent stroke either from extension of the original
ischaemic stroke associated with carotid disease thrombus/embolus or by ‘watershed infarction’. Studies of
the natural history of ICA occlusion are complicated by the
100 000.11 This puts great strain on patients and relatives, fact that it may present as a unilateral or bilateral phenom-
with loss of earnings, possible change in lifestyle and enon. Furthermore, there is a variable contribution to cere-
dependency on others. In 1988, Isard and Forbes estimated bral blood flow arriving via the vertebral arteries and other
that the cost to the health service in Scotland for one stroke collaterals.22,23 There is evidence to suggest that embolism
was £6000 but they only considered the cost to the hospital occurs distal to an occluded segment of the ICA24 but it is
and local doctors.12 This is similar to a more recent European not easy to determine whether the emboli originate from
study finding the total cost per patient treated in London the thrombus or arrive via the collateral circulation. In
of approximately £5000.13 However, when other associated the Joint Study of Extracranial Arterial Occlusion, of 368
factors are included, such as community care and loss patients identified as having unilateral ICA occlusion and
of productivity, the cost, in the USA, is estimated to be the having survived the initial stroke, new strokes occurred in
equivalent of £70 000 at 1990 prices.14 The difficulties in at least 25 per cent within the ensuing 44 months.25 Several
assessing the cost of stroke to the community are perhaps other studies confirm these findings and not surprisingly
a reflection of the protean manifestations of this disease, the outlook for bilateral ICA occlusion is worse.26 However,
and that patients often present with other pathologies that all of these studies are subject to selection bias towards poor
may compound the problems of stroke.15 Overall, it is esti- collateral circulation as many patients with good collateral
mated that 4 per cent of UK National Health Service circulation may not present to clinicians in the event of ICA
resources are consumed in treating and managing stroke occlusion.
each year.4,16 The NHS budget for 2002–2003 was in excess In the remainder of this chapter, the pathophysiology of
of £60 billion and therefore the total expenditure on stroke cerebral reperfusion will be described in relation to the
can be estimated to total some £2.4 billion. problems faced by clinicians attempting to provide a cere-
bral reperfusion service (see Chapters 12 and 13).

RATIONALE FOR ACUTE TREATMENT OF

ISCHAEMIC STROKE
THE PATHOGENESIS OF CEREBRAL
INFARCTION: CELL DEATH AND APOPTOSIS
Until recently stroke was considered the least treatable of all
neurological conditions, however, it has become increasingly
‘Ischaemia’ implies complete cessation of blood flow, but
recognised that in certain circumstances, appropriate therapy
in stroke, blood flow is rarely completely absent. The
instituted promptly will reduce the overall impact of
infarction occurs in a nebulous area which ranges between
stroke. There are three mechanisms by which this may occur:
hypoxia and total ischaemia. In the following discussion the
term ischaemia will be used to denote a reduction in the sup-
Measures to reduce the impact of ply of oxygen and nutrients of sufficient severity and dur-
acute stroke ation to cause cerebral infarction. Furthermore, cerebral
infarction involves pan-necrosis where all the elements of
neural tissue, glial and endothelial cells as well as neurones,
• by preventing further strokes
are involved. Complicating the issue of cerebral infarction
• by salvaging ischaemic but not infarcted brain
is the fact that the extent of injury is dependent on both the
• by limiting the damage caused by stroke.
degree and the duration of hypoxia. In the following sections
 The concept of the ischaemic penumbra 57

the mechanisms of cell death will be briefly described fol- As well as the duration of ischaemia, several studies
lowed by a discussion of the effects of depth and duration have demonstrated thresholds of cerebral blood flow upon
of ischaemia. This sets the scene for the introduction of the which neuronal function and survival are critically depend-
concept of the ischaemic penumbra. ent. These are referred to as the threshold of electrical
The brain is extremely sensitive to reductions in oxygen failure and the threshold of membrane failure, respect-
delivery as it depends almost entirely on oxidative phos- ively. The former refers to the rate of blood flow at which
phorylation for energy production. An impairment in the neuronal function ceases and the latter to the point at
supply of oxygen (and other nutrients, especially glucose) which neuronal death ensues.
leads to a failure in maintenance of ionic homoeostasis.
This is followed by cellular depolarisation and the release
The threshold of electrical failure
of excitatory amino acid transmitters from presynaptic
terminals. This damage in turn leads to an increase in
In primates, normal cerebral blood flow is in the range
cytosolic sodium and chloride producing intracellular
of 50–60 mL per 100 g of brain per minute.33 As flow is
oedema followed by brain swelling and further reductions
gradually reduced, the threshold of electrical failure is
in cerebral perfusion. In addition, a rise in intracellular cal-
approached and physiological function is impaired. In a
cium concentration causes the activation of proteolytic
cat model of reversible middle carotid artery (MCA) occlu-
enzymes and the generation of free radical species by cyclo-
sion, Heiss et al. simultaneously recorded single (neuronal)
oxygenase and phospholipase A2. Further structural cellu-
cell activity and local blood flow and were able to demon-
lar damage follows and eventually cell death ensues.27 The
strate that spontaneous electrical activity ceased at flow
above events describe the process of cell necrosis, however,
levels of about 18 mL per 100 g per minute and that normal
some cells are killed by a different mechanism termed
function returned after restoration of flow.34,35 In the
apoptosis. This is an active process of cell death charac-
baboon, Branston et al. demonstrated that evoked cortical
terised by the maintenance of cell membrane and mito-
responses disappeared at similar flow thresholds and
chondrial integrity. The cells reduce in volume and
reappeared on reperfusion.36,37 These findings were subse-
become susceptible to phagocytosis. This occurs prior to
quently corroborated by the results of clinical studies on
membrane rupture and prevents the content of the cell
patients undergoing carotid surgery.38,39 Because elec-
from being discharged into the surrounding milieu.28 Both
troencephalographic activity in the patients undergoing
apoptosis and necrosis occur in the brain: it appears
carotid endarterectomy returned to precross-clamp levels
that necrosis occurs with catastrophic reductions in blood
following restoration of blood flow, it was assumed that
flow, but at times of more controlled flow reduction, or
the neurones were not irreversibly damaged. Clearly, how-
‘relative ischaemia’, apoptosis occurs as a method of ‘dam-
ever, these results need to be interpreted in the light of the
age limitation’.
fact that anaesthetic agents reduce cerebral metabolic
Brief periods of cerebral ischaemia lasting a few minutes
requirements.
cause little or no evidence of histological damage. Ischaemia
lasting more than 5 minutes but less than 1 hour results
in the progressive death of selectively vulnerable neurones. The threshold for membrane failure
Certain groups of neurones are more sensitive than others;
hippocampal CA1 neurones and cerebellar Purkinje fibres The lower limit for cell survival, the threshold for mem-
are more vulnerable to ischaemia compared with other brane failure, was initially estimated by recording the
cells, as demonstrated by their susceptibility to even brief extracellular potassium concentration at varying flow
periods of cardiac arrest.29 The process whereby some rates.40,41 Elevation of extracellular potassium, indicative
brain tissue dies before other areas has been called ‘select- of the collapse of transmembrane potentials, pointed to
ive neuronal necrosis’. However, after 1 hour, depending membrane failure at flow rates below 10–12 mL/100 g per
upon the species, infarction ensues within the zone of minute. The demonstration that hypoxic cell death involved
lowest blood flow progressively enlarging to a maximum the coupled uptake of calcium and efflux of potassium at
volume over 3–4 hours in rodents,30 and 6–8 hours in similar low flow rates was taken as further evidence indica-
non-human primates.31 Until recently it has been hard to tive of membrane failure.42 These thresholds are slightly
define this critical time interval in humans. Observations higher in rats and gerbils perhaps correlating with the higher
of anaesthetised patients undergoing carotid surgery and cerebral metabolic rate of these animals.43
neurosurgery can be misleading as general anaesthetic
reduces cerebral metabolic activity. This has the potential
THE CONCEPT OF THE ISCHAEMIC PENUMBRA
of increasing neuronal survival times. Moreover, with the
advent of trials of thrombolysis it seems clear that in
certain circumstances normal brain function and mor- From the preceding discussion it is clear that focal cerebral
phology can be restored up to and possibly beyond a infarction is highly dependent upon both the duration and
3-hour time window.32 severity of ischaemia. The studies of ischaemic thresholds
58 Pathophysiology of stroke

this model the cells in the penumbral region took up to 24

hours to die.46 In a shorter term experiment using perman-
ent ischaemia in rats, Kohno et al. compared reductions
in blood flow, measured by diffusion-weighted magnetic
resonance imaging (DWI), with areas of biochemical
abnormality, manifested by acidosis and adenosine triphos-
phate (ATP) depletion: after 30 minutes of ischaemia the
area of acidosis and low perfusion was larger than the area of
ATP depletion.47 Over the following two hours, however,
ATP depletion progressed towards the periphery of the
lesion. The authors interpreted this finding as a gradual
depletion of energy in the zone surrounding a central
Figure 5.2 Schematic representation of the ischaemic infarction core of established infarction.
penumbra – a central core of cerebral infarction surrounded by
an area of hypoperfused but viable brain
Patient studies
Computed tomography (CT) is commonly used in the
discussed above have important implications in the devel-
clinical investigation of stroke and is able to demonstrate
opment of cerebral infarction in humans. In 1981, Astrup
haemorrhage, oedema and parenchymal necrosis.
suggested that the ischaemic core of focal infarction is
Unfortunately, it lacks sensitivity for the structural changes
surrounded by a rim of tissue with reduced blood supply,
in the acute stage of stroke. For this reason, and unless
sufficient to maintain transmembrane potentials but insuf-
haemorrhage has to be excluded, CT scanning is usually
ficient to maintain electrical activity.44 It was suggested
delayed for 24 hours or more to ensure a hypodense, i.e.
that the cells in the central core of an infarct were dead but
visible, infarct is identified if present. Olsen et al. per-
were surrounded by a zone of cells in a twilight zone
formed CT and xenon clearance studies on patients with
between electrical failure and membrane failure. The term
acute stroke, demonstrating that blood flow abnormalities
‘ischaemic penumbra’ was based on the Greek word for
make up a significantly larger area than that measured by
partial shadow (Fig. 5.2).44
CT.48 They interpreted this finding as evidence that the tis-
Since this initial description, the ischaemic penumbra
sue surrounding an infarct, i.e. the penumbra, may not
has become the focus of intense research efforts, both in
take on the CT appearances of infarction but is physiolo-
establishing its existence and in verifying it as a target for
gically abnormal and, importantly, a potential therapeutic
therapeutic intervention in stroke. Evidence for the exist-
target.
ence of the penumbra came originally from animal studies
Using single photon emission computed tomography
and later from human studies.
(SPECT) Wise et al. studied the metabolic changes occur-
ring in the tissue surrounding an infarct and found that
Animal studies there was evidence of continuing neuronal death up to
1 week following infarction.49 More comprehensive PET
Several animal studies have shown that the tissue sur- studies, using a shorter time interval, have subsequently
rounding a core of infarction is biochemically distinct confirmed these findings demonstrating metabolically
from both the normal brain and the infarct. In a rat model active tissue at 17 hours following infarction. In the absence
of focal reversible ischaemia, Kristian et al. demonstrated of treatment, this tissue subsequently became metabol-
that calcium homoeostasis is abolished soon after the ically inert.50 In a more comprehensive study, Baron
insult at the core of the infarction but takes up to 6 hours described three patterns of blood flow in untreated stroke
to disappear at the periphery.45 Positron emission tomo- patients in whom an ischaemic penumbra was present.51
graphy (PET) is a useful tool in the study of brain pathol- Firstly, some patients develop early extensive cortical dam-
ogy as it enables direct measurement of cerebral blood flow age, indicated by a rapid decline in the cerebral metabolic
and metabolic parameters such as cerebral glucose, oxygen rate of oxygen, and have no evidence of recovery. A second
consumption and cerebral oxygen extraction. In a non- group showed evidence of hyperperfusion, with preserved
reversible feline model of MCA occlusion, Heiss et al. cerebral metabolic rate of oxygen, indicating complete or
performed serial PET scans up to 24 hours following occlu- almost complete recovery. In the third pattern, patients
sion.46 At the onset of infarction the oxygen extraction was develop severe cortical ischaemia with a relatively well pre-
greatest at the centre of the focus of ischaemia but as served cerebral metabolic rate of oxygen, consistent with
the central cells gradually died, oxygen extraction fell at the the two threshold flow rates already discussed. These
centre and then increased towards the periphery of the patients have a very variable outcome consistent with the
lesion. This process took up to 24 hours suggesting that in presence of hypoperfused but viable tissue.51
 Clinical consequences of reperfusing ischaemic brain 59

Conventional magnetic resonance imaging (MRI) meas- eventual infarction volume, however there are comparatively
ures alterations in tissue water content. This makes the tech- few data confirming the restoration of normal homoeo-
nique of limited value in the early investigation of stroke as stasis in tissue which was previously ‘penumbral’.
urgent treatments should aim to prevent damage which The quiescent cells within the penumbra are precariously
results in fluid shifts. However, there are several modifica- balanced and in the absence of reperfusion the cells eventu-
tions of the technique that have provided a valuable insight ally die. An increase in local blood flow or perfusion pressure
into the existence of the ischaemic penumbra. Diffusion- may be sufficient to restore function but an increase in adja-
weighted magnetic resonance imaging measures the diffu- cent energy demand or an increase in intracranial pressure
sion of water in ‘regions of interest’ and in areas of cerebral may reduce the supply of oxygen and glucose to a level below
ischaemia there is a detectable reduction in the rate of that required to maintain transmembrane potentials.55
diffusion of water molecules. In patients with stroke of Clearly the reperfusion of the brain must be set against
less than 2 hours duration, well circumscribed lesions the disadvantage of delayed reperfusion causing ‘reperfu-
relating to the decrease in molecular water diffusion have sion injury’. This phenomenon is well described in the
been demonstrated.52 Contrast studies also appear useful. peripheral vascular system (see Chapter 4) and is now
Perfusion-weighted scanning obtains images (PWI) by known to be highly relevant in the brain.56
measuring the rate of appearance of contrast in a volume When brain becomes ischaemic, there is an initial eleva-
of brain tissue. Because the rate of appearance of contrast is tion in intracellular calcium, activating proteases and
dependent on blood flow, this technique accurately meas- phospholipases which in turn lead to the breakdown of
ures blood flow at the capillary level. By combining PWI membrane lipids and the production of damaging reactive
and DWI it is possible to measure the volume of the initial oxygen species. In the absence of reperfusion, much of the
infarction (using DWI), the volume of hypoperfused brain initial damage occurs in this fashion. However, following
(PWI) and the final infarction volume (using conventional reperfusion and the restoration of oxygen supply the pro-
MRI). Fisher and Garcia used this technique to show that duction of free radicals is greatly increased. This allows the
the volume of hypoperfused brain is larger than the vol- production of further free radicals by a process of amplifica-
ume of infarcted brain, as would be expected with an tion from several intracerebral sources: the sudden upsurge
infarct surrounded by penumbra, and that the eventual in mitochondrial activity following restoration of blood
infarction volume corresponds to the volume of hypoper- flow brings about the production of arachidonic acid
fused brain, as would be expected if the penumbra slowly released by the phospholipases activated during ischaemia.
died to become complete infarction.53 This fatty acid is converted by cyclo-oxygenase generating
Finally, magnetic resonance spectroscopy studies com- free reactive oxygen species. These molecules cause lipid
plement all of the above studies: with this technique serial peroxidation, leading to neuronal, glial and endothelial
measurements of various key metabolites can be taken membrane dysfunction. The blood–brain barrier becomes
throughout the time course of stroke. Lactate (a marker more permeable and results in the development of ‘vaso-
of anaerobic metabolism), N-acetylaspartate (a neuronal genic’ oedema (see Chapter 13). The neurones, glia and
marker) and creatinine/phosphocreatinine ratio (present in endothelial cells develop ‘cellular’ oedema as a result of cell
both glia and neurones) can be measured spectroscopically. wall dysfunction. Furthermore, the breakdown of the
Abnormal or ischaemic regions can be compared with con- blood–brain barrier allows the migration of macrophages
tralateral healthy brain, and it can be demonstrated that into the cerebral extravascular space and this induces a fur-
neurones die faster than glia. Furthermore, this technique ther cascade of neurotoxic events.27 The understanding of
provides evidence of neuronal death occurring up to 10 days these events at cellular level is helpful in the interpretation
following the initial insult,54 in agreement with the earlier of findings from clinical studies of cerebral reperfusion.
findings of Wise et al. in 1983.49
It is clear from the above studies that there is strong evi-
dence for the existence of an ischaemic penumbra. It appears
CLINICAL CONSEQUENCES OF REPERFUSING
to be a dynamic process which, with time, eventually dis-
ISCHAEMIC BRAIN
appears and is replaced by infarction. Although many of
these studies have indicated that the penumbra may last for
up to several days, it would appear, intuitively, that prompt There are several features of cerebral physiology and
treatment would salvage more brain. anatomy which make the reperfusion of the brain a com-
However, in order to accept the premise that the presence plex process. These factors include the blood–brain bar-
of the penumbra justifies attempts at cerebral reperfusion, we rier, the collateral blood supply, the ventriculo-cisternal
require proof that reperfusion alters the nature of the system containing cerebrospinal fluid, the skull and dura
penumbra; in other words, that restoration of blood flow forming a rigid external boundary, the high metabolic rate
converts inactive neurones at risk of death back to elec- of brain tissue and its almost total dependence upon aer-
tronically functional tissue. There is, in fact, a wealth of obic metabolism. Each factor has a crucial impact on the
evidence to indicate that early reperfusion reduces the specific problems associated with cerebral reperfusion.
60 Pathophysiology of stroke

pressure increases. Eventually, cerebral tissue is forced

Complications of reperfusing ischaemic through the openings of the skull and, in the absence of
brain treatment, the brainstem is forced through the foramen
magnum, so called ‘coning’, and death rapidly ensues.
• Haemorrhagic stroke
• Ischaemic stroke
• Cerebral oedema
Haemorrhagic transformation of ischaemic
infarction

In broad terms however, there are two pathological Between 10 and 20 per cent of strokes are primarily due to
processes that can complicate the reperfusion of ischaemic subarachnoid or intraparenchymal haemorrhage. However,
brain: oedema and haemorrhage. Both have been demon- some ischaemic strokes may undergo haemorrhagic trans-
strated to contribute to the mortality and morbidity of formation.59 This phenomenon ranges from a few petechial
stroke in the presence of physiological (i.e. spontaneous) haemorrhages across the surface of the brain to extensive
and therapeutic reperfusion. intracranial haematoma formation.60 Postmortem studies
give high rates of haemorrhagic transformation, presum-
ably because they are biased towards more severe strokes.
Cerebral oedema However, prospective radiological studies of patients pre-
senting with stroke tend to underestimate the incidence of
Cerebral metabolic disturbances caused by ischaemia and such transformation because, at the time when the research
subsequent reperfusion can lead to the development of was carried out, many scanners did not have the resolution
both cellular and vasogenic oedema. to identify the smaller bleeds.60 Furthermore, to investigate
Cellular oedema is similar to that seen in other organs. the possibility that haemorrhagic transformation may be
Maintenance of the normal cell volume is an energy underdiagnosed, and hence primary haemorrhage overdiag-
dependent mechanism based on the ‘pump–leak’ model.57 nosed, Bogousslavsky et al. describe a series of 15 patients
Both neurones and glia control their volume by the active with an admission CT demonstrating no cerebral haemor-
extrusion of sodium ions against a concentration gradient rhage; subsequent clinical deterioration and CT scanning
favouring inward flux of sodium. Failure of the energy sup- within 18 hours revealed the appearances of intracranial
ply results in a net increase in intracellular sodium fol- haemorrhage arising as haemorrhagic transformation.61
lowed by osmotically obliged water, leading to an increase Although they did not state the size of the population from
in volume. The increase in volume may be clinically which the sample was taken, the work suggests that studies
unnoticed (compensated for by a reduction in extracellular of the aetiology of haemorrhagic stroke may overestimate
fluid volume) but is physiologically desirable (i.e. glial the incidence of primary haemorrhage as the cause.
swelling in compensation for increased local neuronal Clearly, the timing of the CT scan is important, and the
activity), or it may be manifested as a mass effect. use of serial CT scanning has now enabled researchers to
Vasogenic oedema occurs when the integrity of the establish the rate of haemorrhagic transformation over time
blood–brain barrier is interrupted; the flux of water from after onset of stroke. Okada et al. found that of 160 patients
the plasma into the cerebral extracellular fluid results in with well documented cerebral embolism, 65 (41 per cent)
tissue oedema and consequent brain swelling. The causes underwent haemorrhagic transformation within one
of blood–brain barrier disruption are not clear: at the cel- month of the ictus.62 In a series of 65 patients Hornig et al.
lular level free radicals, bradykinin and histamine have been reached similar conclusions; 28 patients (43 per cent)
implicated, as have infection, cranial trauma and stroke.58 developed haemorrhagic transformation within four weeks
In practice, vasogenic and cellular cerebral oedema often of stroke with a peak incidence during the second week.63
coexist and in the context of stroke, this is usually the case. Although there are no reports of studies in this area it seems
Vasogenic oedema necessarily involves an increase in flow of possible that late spontaneous reperfusion may result in an
plasma from the blood. This contains many substances toxic increased risk of haemorrhagic transformation.
to the brain, such as glutamate and potassium, and that in The aetiology of haemorrhagic transformation is not
turn leads to further cerebral damage. As damage con- well understood but the theory proposed by Fisher and
tinues to progress, the microcirculation is disrupted, the Adams in the 1950s is generally accepted: an embolus lodges
tissue swells further to the point at which intracranial pres- in a vessel causing downstream infarction. Subsequently,
sure rises and cerebral perfusion is compromised. As the the embolus autolyses, fragments and, with the force of the
mass expands, the cerebrospinal fluid (CSF) and venous blood behind it, is driven downstream exposing the
volume decrease and healthy parts of brain become com- ischaemic vessels to blood flow at high pressure which
pressed to accommodate the expanding oedematous mass. ruptures the vessel and causes bleeding.64 This hypoth-
Once these compensatory mechanisms fail, the contents of esis appears applicable to haemorrhagic transformation,
the cranial vault become non-compliant and intracranial both spontaneous and that complicating stroke therapies.
 Logistical problems 61

In favour of this theory is the fact that cardioembolism where patients underwent reperfusion regardless of the
increases the rate of haemorrhagic transformation.65 This possibility that they had presented with a cerebral haemor-
suggests that emboli from the heart are able to undergo rhage. Most of the later studies of emergency carotid sur-
lysis and migrate distally as opposed to thromboembolic gery for stroke use CT in the selection criteria to exclude
causes of occlusion. haemorrhage and find surgery advantageous.70
Two studies have found that haemorrhagic transform-
ation is significantly more frequent in large areas of infarc-
tion and in patients with severe neurological deficits.62,63
Furthermore, Okada et al. found that elderly patients (over Future directions in cerebral reperfusion
70 years) were especially at risk.62 Interestingly, in both of
the above studies hypertension was not related to the risk of • Strategies to reduce the time delay between stroke
haemorrhage whereas the two available animal studies, and treatment
designed to assess the role of hypertension in haemorrhagic • Strategies to avoid the complications of reperfusion
transformation, both found that increased blood pressure
was related to an increase in cerebral haemorrhage.66,67
Unfortunately, the species differences between cats, rabbits
and humans make direct comparison of these results The extent of injury following cerebral reperfusion clearly
impossible. Bowes et al. used a rabbit model of stroke and relates to the duration and severity of the initial ischaemic
simply controlled the hypertension caused by induction of insult, such that the sooner reperfusion is established the
stroke66 whereas Saku et al. induced hypertension by aortic less severe the injury. The twofold rationale for early reper-
occlusion.67 fusion of brain following stroke is therefore clear: first,
blood flow must be promptly restored in order to salvage
viable penumbral brain tissue, and second, the extent of
Iatrogenic haemorrhage reperfusion injury avoided or reduced.

The risk of drug induced haemorrhagic transformation is

relevant to potential new therapies because until the advent
LOGISTICAL PROBLEMS
of thrombolysis for stroke, haemorrhagic transformation
held little significance. Now, however, it is seen as a major
risk factor for thrombolysis and consequently a great deal All of the above evidence suggests that the traditional, dog-
of research effort has been channelled into to the cause of matic stance against the treatment of acute stroke is no
haemorrhagic transformation following deliberate cerebral longer tenable. There is a wealth of evidence both from
reperfusion. animal and human studies to indicate that timely reperfu-
In a baboon model of reperfusion, del Zoppo et al. found sion of the brain is possible and may bring reductions in
that there was no difference in the incidence of haemorrhage morbidity and mortality from this disease. However, cere-
and volume of infarction and that none of the three doses of bral reperfusion carries significant risks and approaches to
recombinant tissue plasminogen activator (rtPA) given in reperfusion have been hampered by difficulties in delivering
the study had any increased tendency to haemorrhage the care, the complications of which can be catastrophic.
when given at three hours.68 In a Cochrane review of ran- From the onset of stroke, cerebral tissue is dying and as time
domised controlled, mainly intravenous, trials of throm- progresses the risks associated with reperfusion appear to
bolysis for stroke, death and dependency are reduced at the increase. Safe and effective cerebral reperfusion requires
expense of an increased risk of intracerebral haemorrhage.69 detailed consideration of the following important stages:
However, the important principle from the thrombolysis recognition of the stroke by the patient, rapid transfer to
experience seems to be that the risk of bleeding increases hospital, investigation and treatment.
with the time delay to reperfusion, again supporting the Many studies have investigated the pre-hospital delay of
rationale of early reperfusion, by whatever means (see patients presenting with stroke. Clearly, patients living
Chapters 12 and 13). alone and unable to make contact to seek help are in a grim
Cerebral haemorrhage is a feared complication of carotid position. Suffering a stroke in the presence of people with a
reconstruction for stroke and many studies of emergency knowledge of stroke decreases the time delay from stroke
carotid surgery for stroke include carotid endarterectomy to hospital71,72 whereas the involvement of primary care has
and thromboendarterectomy. The conclusions of early been associated with increased delays in arrival at hos-
studies lean towards avoiding surgery in patients with pital.73–75 Education appears to be the key issue here and
extensive neurological deficits, and it was inferred that the efforts to increase stroke awareness have reduced stroke-
mechanism for haemorrhagic transformation in these to-hospital times in some American centres.74
patients is similar to that proposed by Fisher and Adams.64 The mode of transport to hospital is important. In the
However, most of these studies were from the pre-CT era UK, unpublished audits from Bristol and Manchester
62 Pathophysiology of stroke

showed that the transport used by stroke patients attend-

Key references
ing hospital ranged from emergency ambulance transfer to
the local bus service! Emergency ambulance ‘blue light’ Bamford J, Sandercock P, Dennis M, et al. A prospective study of
transfer in the UK takes less than 8 minutes in 75 per cent acute cerebrovascular disease in the community: Oxfordshire
of patients. If all patients attended via such a service the Community Stroke Project 1981–86. 1. Methodology,
delay may be minimised but there are other potential ‘spin- demography and incident cases of first-ever stroke. J Neurol
offs’. A recent study has demonstrated that ambulance para- Neurosurg Psychiatry 1988; 51: 1373–80.
medics have a high diagnostic accuracy for strokes.75 Bamford J, Sandercock P, Dennis M, et al. Classification and natural
Paramedics are in an ideal position to forewarn hospital history of clinically identifiable subtypes of cerebral infarction.
staff of the imminent arrival of patients who may benefit Lancet 1991; 337: 1521–6.
from early stroke treatment. Dirnagl U, Iadecola C, Moskowitz M. Pathobiology of ischaemic
stroke: an integrated view. Trends Neurosci 1999; 22: 391–7.
A major delay in the management of stroke patients is in
Okada Y, Yamaguchi T, Minematsu K, et al. Haemorrhagic
the availability of appropriate investigations. Over the last 20
transformation in cerebral embolism. Stroke 1989; 20:
years CT scanning has become commonplace in the man- 598–603.
agement of stroke. However, outside the setting of clinical Wise R, Bernadi S, Frackowiak R, et al. Serial observations on the
trials in which strict protocols are observed, there is great pathophysiology of acute stroke. The transition form ischaemia
variability of access to scanning for stroke patients. This is, in to infarction as reflected in regional oxygen extraction. Brain
part, due to the differences in opinion about the usefulness 1983; 106: 197–222.
the timing of the scans. An early CT scan, i.e. within 24 hours,
is of value in excluding haemorrhagic stroke, but it can also
identify thrombus in the middle cerebral artery. The extent REFERENCES
of infarction, however, is not clear at this stage and later
scans are required to define the boundaries of infarction,
seen as hypodense areas on CT. Other reasons for delay in 1 Warlow C, Dennis M, van Gijn J, et al. Stroke: A Practical Guide
CT scanning are a simple lack of resources and, importantly, to Management. Oxford: Blackwell Scientific, 1996.
2 Bamford J, Sandercock P, Dennis M, et al. A prospective study
a pessimistic attitude as to the value of the investigation.
of acute cerebrovascular disease in the community: Oxfordshire
More sophisticated scanning modalities such as MRI
Community Stroke Project 1981–86. 1. Methodology,
and PET are more sensitive in detecting early ischaemic demography and incident cases of first-ever stroke. J Neurol
changes but are even less available than CT scanning. Neurosurg Psychiatry 1988; 51: 1373–80.
Assuming that patients have reached hospital rapidly and 3 Wolfe C. The impact of stroke. Br Med Bull 2000; 56: 275–86.
have been imaged immediately, they then require access to 4 Bergman L, van der Meulin J, Limburg M, Halbema T. Cost of
treatment. Again, outside the context of major trials this is a medical care after first ever stroke in the Netherlands. Stroke
problem: emergency carotid endarterectomy or thrombo- 1995; 26: 1830–6.
endarterectomy is possible, as is thrombolysis, but both 5 Becker R. Incidence of stroke may be on the rise. Stroke 1997;
involve the use of intensive care or high dependency facili- 28: 1657–9.
ties which, in many countries, are scarce. Furthermore, the 6 Chant H, McCollum C. Stroke in young adults: The role of
paradoxical embolism. Thromb Haemost 2001; 85: 23–9.
entire population of the UK is not currently served by vas-
7 Sherman D, Lalonde D. Anticoagulants in stroke treatment.
cular specialists willing to undertake such complex cases.
In: Welch KMA, Kaplan LR, Reis DJ, et al. (eds) Primer on
These problems should be set against the broader picture Cerebrovascular Diseases. London: Academic Press,1997.
of stroke care: an increase in stroke-free or independent 8 Mead G, Shingler H, Farrell A, et al. Carotid disease in acute
stroke survivors has clear long term benefits for both stroke. Age Ageing 1998; 27: 677–82.
society and the individual. 9 European Carotid Surgery Trialists’ Collaborative Group.
Randomised trial of endarterectomy for recently symptomatic
carotid stenosis: final results of the MRC European Carotid
Conclusions Surgery Trial (ECST). Lancet 1991; 351: 1379–87.
10 Bamford J, Sandercock P, Dennis M, et al. Classification and
natural history of clinically identifiable subtypes of cerebral
Acute ischaemic stroke is common and has wide ran-
infarction. Lancet 1991; 337: 1521–6.
ging implications for the individual and society. The bene-
11 Williams G, Jiang J, Matchar D, Samsa G. Incidence and
fit of early treatment of selected patients by intravenous recurrence of total (first-ever and recurrent) stroke. Stroke
thrombolysis has been established and, contrary to the 1999; 30: 2523–8.
nihilistic surgical attitude in most quarters, surgery may 12 Isard P, Forbes J. The cost of stroke to the National Health
also have a role to play. This is a developing area and Service in Scotland. Cerebrovasc Dis 1992; 2: 47–50.
technical advances will need to be paralleled by 13 Grieve R, Hutton J, Bhalla A, et al. on behalf of the Biomed II
improvements in the treatment of stroke patients at all European study of stroke care. A comparison of the costs and
levels from primary to tertiary care. survival of hospital-admitted stroke patients across Europe.
Stroke 2001; 32: 1684–91.
 References 63

14 Taylor T, Davis P, Torner J, et al. Lifetime cost of stroke in the flow following acute middle cerebral artery occlusion in the
United States. Stroke 1996; 27: 1459–66. baboon. Exp Neurol 1974; 45: 195–208.
15 Warlow C. Epidemiology of stroke. Lancet 1998; 352 37 Symon L, Branston N, Strong A, Hope T. The concepts of
(suppl III): 1–4. thresholds of ischaemia in relation to brain structure and
16 Wade DT. Stroke (acute cerebrovascular disease). In: Stevens A, function. J Clin Pathol 1977; (suppl: Roy Coll Pathol)11:
Raftery J. (eds) Health Care Needs Assessment Vol 1. Oxford: 149–54.
Radcliffe Medical Press, 1994. 38 Sharbrough F, Messick J Jr, Sundt T Jr. Correlation of
17 Burn J, Dennis M, Bamford J, et al. Long term risk of recurrent continuous electroencephalograms with cerebral blood flow
stroke. The Oxfordshire community stroke project. Stroke 1994; measurements during carotid endarterectomy. Stroke 1973; 4:
25: 333–7. 674–83.
18 Dosick S, Whalen R, Gale S, Brown O. Carotid endarterectomy 39 Sundt T Jr, Sharbrough P, Anderson R, Michenfelder J. Cerebral
in the stroke patient. Computerised axial tomography to blood flow measurements and electroencephalograms during
determine timing. J Vasc Surg 1985; 2: 214–19. carotid endarterectomy. J Neurosurg 1974; 41: 310–20.
19 Piotrowski J, Bernhard V, Rubin J, et al. The timing of carotid 40 Astrup J, Symon L, Branston N, Lassen N. Cortical evoked
endarterectomy after stroke. J Vasc Surg 1990; 11: 45–52. potentials and extracellular K and H at critical levels of
20 Blaser T, Hofmann K, Buerger T, et al. Risk of stroke, transient brain ischaemia. Stroke 1977; 8: 51–7.
ischaemic attack, and vessel occlusion before carotid 41 Branston N, Strong A, Symon L. Extracellular potassium activity
endarterectomy in patients with symptomatic severe carotid evoked potential and tissue blood flow: relationship during
stenosis. Stroke 2002; 33: 1057–62. progressive ischaemia in baboon cerebral cortex. J Neurol Sci
21 Welsh S, Mead G, Chant H, et al. Early carotid surgery in acute 1977; 32: 305–21.
stroke: a multicentre randomised pilot study. Cerebrovasc Dis 42 Harris R, Symon L, Branston N, Bayhan M. Changes in
2004; 18: 200–5. extracellular calcium activity in cerebral ischaemia. J Cereb
22 Margolis M, Newton T. Collateral pathways between the Blood Flow Metab 1981; 1: 203–9.
cavernous portion of the internal carotid artery and external 43 Siesjo B. Pathophysiology and treatment of focal cerebral
carotid arteries. Radiology 1969; 93: 834–6. ischaemia. Part 1. Pathophysiology. J Neurosurg 1992; 77:
23 Norris J, Krajewski A, Bornstein N. The clinical role of the 169–84.
cerebral collateral circulation in carotid occlusion. J Vasc Surg 44 Astrup J. Thresholds in cerebral ischaemia – the ischaemic
1990; 12: 113–18. penumbra. Stroke 1981; 12: 723–5.
24 Castaigne P, Lhermitte F, Gautier J-C, et al. Internal carotid 45 Kristian T, Gido G, Kuroda S, et al. Calcium metabolism of focal
artery occlusion. A study of 61 instances in 50 patients with and penumbral tissues in rats subjected to transient middle
post mortem data. Brain 1970; 93: 231–58. cerebral artery occlusion. Exp Brain Res 1998; 120: 503–9.
25 Blaisdell F, Hall A, Thomas A. Surgical treatment of chronic 46 Heiss W, Graf R, Weinhard K, et al. Dynamic penumbra
internal carotid artery occlusion by saline endarterectomy. Ann demonstrated by sequential multitracer pet after middle
Surg 1966; 163: 103–11. cerebral artery occlusion in cats. J Cereb Blood Flow Metab
26 Freidman S. Current management of the patient with internal 1994; 14: 892–902.
carotid artery occlusion. Eur J Vasc Surg 1989; 3: 97–101. 47 Kohno K, Hoehn-Berlage M, Mies G, et al. Relationship between
27 Dirnagl U, Iadecola C, Moskowitz M. Pathobiology of ischaemic diffusion-weighted MR images, cerebral blood flow, and energy
stroke: an integrated view. Trends Neurosci 1999; 22: 391–7. state in experimental brain infarction. Mag Res Imag 1995; 13:
28 Leist M, Nictera P. Cell death: Apoptosis versus necrosis. In: 73–80.
Welch K, Kaplan L, Reis D, et al. (eds) Primer on Cerebrovascular 48 Olsen T, Larsen B, Herning M, et al. Blood flow and vascular
Diseases. London: Academic Press, 1997: 101–4. reactivity in collaterally perfused brain tissue: evidence of an
29 Graham D. Hypoxia and vascular disorders. In: Adams J, Duchen L ischaemic penumbra in patients with acute stroke. Stroke 1983;
(eds) Greenfield’s Neuropathology. Oxford: Oxford University 14: 332–41.
Press, 1992. 49 Wise R, Bernadi S, Frackowiak R, et al. Serial observations on
30 Kaplan B, Brint S, Tanabe J, et al. Temporal thresholds for the pathophysiology of acute stroke. The transition form
neocortical infarction in rats subjected to reversible focal ischaemia to infarction as reflected in regional oxygen
cerebral ischaemia. Stroke 1991; 22: 1032–9. extraction. Brain 1983; 106: 197–222.
31 Jones T, Morawetz R, Crowell R, et al. Thresholds of focal cerebral 50 Marchal G, Beaudouin V, Rioux P, et al. Prolonged persistence
ischaemia in awake monkeys. J Neurosurg 1981; 54: 773–82. of substantial volumes of potentially viable brain tissue after
32 Hill M, Hachinski V. Stroke treatment. Time is brain. Lancet stroke. Stroke 1996; 27: 599–606.
1998; 352(suppl III): 10–14. 51 Baron J. Pathophysiology of acute cerebral ischemia: PET
33 Pulsinelli W. Pathophysiology of acute ischaemic stroke. Lancet studies in humans. Cerebrovasc Dis 1991; 1(suppl 1): 22–31.
1992; 339: 533–6. 52 Warach S, Chien D, Li W, et al. Fast magnetic resonance
34 Heiss W, Hayakawa T, Waltz A. Cortical neuronal function during diffusion-weighted imaging of acute human stroke. Neurology
ischaemia. Effects of occlusion of one middle cerebral artery on 1992; 42: 171–3.
single unit activity in cats. Arch Neurol 1976; 33: 813–20. 53 Fisher M, Garcia J. Evolving stroke and the ischaemic
35 Heiss W, Rosner G. Functional recovery of cortical neurones as penumbra. Neurology 1996; 47: 884–8.
related to the degree and duration of ischaemia. Ann Neurol 54 Saunders D, Howe F, van den Boogaart A, et al. Continuing
1983; 14: 294–301. ischaemic damage after acute middle cerebral artery infarction
36 Branston N, Symon L, Crockard H, Pasztor E. Relationship in humans demonstrated by short-echo proton spectroscopy.
between the cortical evoked potential and local cortical blood Stroke 1995; 26: 1007–13.
64 Pathophysiology of stroke

55 Obrenovitch T, Urenjak J, Richards D, et al. Extracellular 66 Bowes M, Zivin J, Thomas G, et al. Acute hypertension, but not
neuroactive amino acids in the rat brain striatum during moderate thrombolysis, increases the incidence and severity of
and severe transient ischaemia. J Neurochem 1993; 61: 178–86. hemorrhagic transformation following experimental stroke in
56 Hallenbeck J, Dutka A. Background review and current concepts rabbits. Exp Neurol 1996; 141: 40–6.
of reperfusion injury. Arch Neurol 1990; 47: 1245–54. 67 Saku Y, Choki J, Waki R, et al. Hemorrhagic infarct induced by
57 Baethmann A, Staub F. Cellular oedema. In: Welch K, Kaplan L, arterial hypertension in cat brain following middle cerebral
Reis D, et al. (eds) Primer on Cerebrovascular Diseases. London: artery occlusion. Stroke 1989; 21: 589–95.
Academic Press, 1997: 153–6. 68 del Zoppo G, Copeland B, Anderchek K, et al. Hemorrhagic
58 Betz L. Vasogenic oedema. In: Welch K, Kaplan L, Reis D, et al. transformation following tissue plasminogen activator
(eds) Primer on Cerebrovascular Diseases. London: Academic in experimental cerebral infarction. Stroke 1990; 21:
Press, 1997: 156–9. 596–601.
59 Lyden P, Zivin J. Hemorrhagic transformation after cerebral 69 Wardlaw J, del Zoppo G, Yamaguchi T. Thrombolysis for acute
ischemia: Mechanisms and incidence. Cerebrovasc Brain Metab ischaemic stroke. In: Cochrane Library, Issue 1, 2002.
Rev 1993; 5: 1–16. 70 Mead G, O’Neill P, McCollum C. Is there a role for carotid
60 Teal P, Pessin M. Haemorrhagic transformation. The spectrum surgery in acute stroke? Eur J Vasc Endovasc Surg 1997; 13:
of ischaemia-related brain haemorrhage. Neurosurg Clin North 112–21.
Am 1992; 3: 601–10. 71 Feldmann E, Gordon N, Brooks J, et al. Factors associated
61 Bogousslavsky J, Regli F, Uske A, Maeder P. Early spontaneous with early presentation of acute stroke. Stroke 1993; 24:
haematoma in cerebral infarct: is primary cerebral 1805–10.
haemorrhage overdiagnosed? Neurology 1991; 41: 837–40. 72 Rosamond W, Gorton R, Hinn A, et al. Rapid response to stroke
62 Okada Y, Yamaguchi T, Minematsu K, et al. Haemorrhagic symptoms: the delay in accessing stroke healthcare (DASH)
transformation in cerebral embolism. Stroke 1989; 20: 598–603. study. Acad Emerg Med 1998; 5: 45–51.
63 Hornig C, Dorndorf W, Agnoli A. Haemorrhagic cerebral 73 Ferro J, Melo T, Oliveira V, et al. An analysis of the
infarction – a prospective study. Stroke 1986; 17: 179–85. admission delay of acute strokes. Cerebrovasc Dis 1994;
64 Fisher C, Adams R. Observations on brain embolism with special 4: 72–5.
reference to the mechanism of haemorrhagic infarction. 74 Barsan W, Brott T, Broderick J, et al. Urgent therapy for acute
J Neuropathol Exp Neurol 1951; 10: 92–4. stroke. Effects of a stroke trial on untreated patients. Stroke
65 Lodder J, Krijne-Kubat B, Broekman J. Cerebral haemorrhagic 1994; 25: 2132–7.
infarction at autopsy: cardiac embolic cause and the 75 Harbison J, Massey A, Barnett L, et al. Rapid ambulance
relationship to the cause of death. Stroke 1986; 17: 626–9. protocol for acute stroke. Lancet 1999; 353: 1935.
 6
Assessing the Risk in Vascular Emergencies

PETER R TAYLOR, DAVID J GERRARD

The problem 65 Abdominal aortic aneurysm repair 67

The concept of risk:benefit ratio 65 Lower limb arterial reconstruction 68
Carotid endarterectomy 66 References 69

THE PROBLEM well to clinical trials because of the clearly defined nature of
the clinical endpoints such as death, stroke or limb loss.
Comparison of the outcome following the experimental
Successful surgical intervention depends on striking a event rate and the control event rate allows the absolute risk
balance between the potential benefit from the procedure reduction rate to be calculated. The number of patients
and the possibility of harm. The majority of elective vascu- ‘needed to treat’ is defined as the number of patients who
lar operations are performed to prevent possible adverse would have to be treated over a specific period of time to
events from lesions which may be asymptomatic or may prevent one bad outcome and this is the inverse of the
have only caused transient symptoms. Elective abdominal absolute risk reduction.1
aortic aneurysm surgery is performed to prevent rupture, The concept has been well demonstrated in the case of
and the majority of patients will have no symptoms related surgery for symptomatic carotid stenoses with regard to
to their aneurysm. Similarly, internal carotid artery stenoses any disabling stroke, fatal stroke or death from any cause
may have caused amaurosis fugax or transient ischaemic following surgery. The impact of surgery on outcome has
attacks (TIAs) which have no lasting effect on the patient, been shown by both the European Carotid Surgery Trial
but may well be the portent of a major stroke. The oper- (ECST) and the North American Symptomatic Carotid
ations of abdominal aortic aneurysm repair and carotid Endarterectomy Trial (NASCET).2,3 The results have recently
endarterectomy are both associated with life-threatening been combined in a meta-analysis which showed that for
complications such as death and stroke, and therefore some internal carotid artery stenoses causing a diameter reduc-
assessment has to be undertaken to identify the risk:benefit tion of between 70 and 99 per cent there is an absolute risk
ratio. Unfortunately, there is a scarcity of good quality reduction of 6.7 per cent with a 95 per cent confidence
publications on this very important subject. The majority limit of 3.2 to 10 per cent when surgery is compared with
of papers which can be used for such an analysis come best medical therapy.4 The number needed to treat is
from randomised clinical trials which almost always involve therefore 15, with confidence limits between 10 and 31.
elective operations. Very few vascular and endovascular The concept is easy to understand: 15 patients have to
non-elective procedures are performed on the basis of undergo carotid endarterectomy in order to prevent one
level 1 evidence, indeed, such trials may be unethical. from having a stroke. The converse is therefore obvious: 14
patients will have no benefit from the procedure. The
paper also showed that patients with 50–69 per cent diam-
THE CONCEPT OF RISK:BENEFIT RATIO eter reduction had an absolute risk reduction of 4.7 per cent
(confidence limits 0.8 to 8.7) with the number needed to
Randomised clinical trials have a great advantage in that treat being 21 (confidence limits 11 to 125). Finally, those
the risk involved in one treatment can be compared with with a stenosis of less than 49 per cent had an absolute risk
another, or indeed with the natural history of the disease increase of 2.2 per cent (confidence limits 0 to 4.4) and
under investigation. Vascular surgery usually lends itself the number needed to harm was 45 (confidence limits 22
66 Assessing the risk in vascular emergencies

to infinity). This last category showed that surgery was be difficult, as the comparator may be a patient who is not
harmful in this group of patients, and gives an example of neurologically normal. Operations undertaken for acute
an intervention which is associated with a worse outcome strokes have this difficulty. Should the final outcome be the
than those patients treated medically. Such an increase is status of the patient before the stroke occurred or that
termed the absolute risk increase, and the number needed immediately before surgery? As some recovery after stroke
to harm is the number of patients, who if they had surgery, is not unusual, should the comparison be made between
would lead to one additional patient being harmed (i.e. the two groups of patients with stroke, one of which had under-
inverse of the absolute risk increase). gone carotid surgery and the other not? This question
There were other factors which were found to be import- clearly muddies the waters in that clearly defined endpoints
ant in assessing the risk:benefit ratio of surgery. In the are not being used.
NASCET study group of patients with 70–99 per cent There is some evidence on the minimum number of
stenoses, those who were male had an increased benefit as operations which should be undertaken either by a hospital
did those over 70 years of age. Other groups who derived or by an individual surgeon each year in order to achieve
more benefit from surgery were those with hemispheric low morbidity and mortality from carotid endarterectomy.
symptoms compared with those who had retinal events, A retrospective study of 9918 carotid operations under-
and those with angiographic evidence of plaque ulceration. taken in Maryland, USA, over 6 years was reported by
The ECST data showed that patients with lacunar infarcts, Perler et al.8 In this series hospitals which performed more
which may not be due to embolism from the internal than 10 but fewer than 50 operations per year had a mortal-
carotid artery plaque, had a relative risk increase from sur- ity of 1.1 per cent and a neurological event rate of 1.3 per cent
gery of 22 per cent (confidence limits 51 to 200 per cent). per year. This compared favourably with hospitals per-
This chapter will concentrate on the evidence for assess- forming 50–100 procedures per year with a death rate of
ing risk in the three index operations for arterial surgery: 0.8 per cent but a slightly higher neurological event rate
carotid endarterectomy, abdominal aortic aneurysm repair of 1.8 per cent. Those hospitals performing fewer than
and lower limb arterial reconstruction. In the absence of 10 carotid endarterectomies per year had a mortality of
good evidence an attempt will be made in each case to 1.9 per cent but a neurological event rate of 6.1 per cent.
extrapolate valid conclusions from the elective to the emer- Another review of 1280 carotid endarterectomies per-
gency situation. formed by eight hospitals in Toronto looked at the number
of cases performed per surgeon each year.9 If more than 12
operations were performed the death rate was 1.2 per cent,
Index operations the stroke rate 4.2 per cent and the stroke and death rate
5.4 per cent. The equivalent figures for those surgeons per-
• Carotid endarterectomy forming between six and 12 operations per year were
• Abdominal aortic aneurysm repair 4.2 per cent, 3.8 per cent and 8 per cent, respectively. Those
• Lower limb arterial reconstruction surgeons doing fewer than six cases per year had no deaths
but a high stroke rate of 18.4 per cent. Figures of 10–50
carotid endarterectomies per year and the individual vas-
cular surgeon responsible for over 12 of these are probably
CAROTID ENDARTERECTOMY better than most vascular surgeons would have predicted.
One further piece of evidence suggests that operations per-
There are many other factors which influence the reported formed by board certified surgeons carry a 15 per cent
outcome following carotid endarterectomy5 and these are lower risk of death or complications compared with those
elaborated elsewhere (see Chapters 12–14). The following undertaken by uncertified surgeons.10 A review of 45 744
summary gives an indication of the difficulties encoun- carotid endarterectomies in Florida, USA, also suggests that
tered in an emergency situation. These include the number a doubling of the operative workload reduced the adverse
and specialty of the surgeons, the type of hospital in which event outcome by 4 per cent.
the operation takes place and the number of operations The technique of carotid endarterectomy is probably
performed by both the hospital and the surgeon each year. important in reducing complications. Most of the data come
However, the presenting symptoms and the findings on from elective operations, as there is so little evidence from
preoperative computed tomography (CT) scans have also non-elective procedures. In a large meta-analysis patching
been found to be predictive of outcome.6 was found to be important in reducing ipsilateral stroke,
There is some evidence on the factors involved in poor perioperative carotid thrombosis, and ipsilateral stroke
outcome after carotid endarterectomy undertaken for urgent and death during follow-up.11 The evidence for shunting was
indications (see Chapter 13). One paper has shown that less strong.12 The importance of avoiding carotid occlusion
urgent surgery for crescendo transient ischaemic attacks is well known to surgeons performing the operation as was
(TIAs) has a worse outcome compared with elective oper- shown by Radak et al.13 reporting a series of 2250 operations:
ation.7 The comparison of the final outcome, however, can 41 patients had an intraoperative stroke with a mortality of
 Abdominal aortic aneurysm repair 67

49 per cent whereas 18 had a postoperative stroke and a development of back pain. Unfortunately, there is no good
mortality of 22 per cent. randomised trial of treatment for symptomatic aneurysms
Why not, therefore, patch all patients undergoing and the natural history of such aneurysms is unknown.
carotid endarterectomy? First, vein patches can rupture; Traditional teaching suggests that infrarenal AAAs associ-
second, synthetic patches can become infected; third, ated with pain should be operated on within 24 hours.
the operation takes longer; and finally patching may not Experienced vascular surgeons know that many patients
abolish technical error. Up to 12 per cent of patched vessels who have tender aneurysms have a contained retroperi-
have major problems, detectable by duplex scanning, which toneal haematoma. Others may have oedema present in
require revision.14 The carotid arteries of women seem to the surrounding tissues and some turn out to be inflam-
be more likely to restenose following primary closure, and matory in nature. Current smoking and poor lung func-
therefore may benefit more from patch angioplasty.15 tion were associated with an increased risk of rupture and
Duplex has the advantage of revealing both anatomical and a higher mortality following surgery in the Small Aneurysm
functional lesions which may require revision,16 and is very Study.21
useful as a teaching tool in making instant feedback avail- The Vascular Surgical Society of Great Britain and Ireland
able to junior staff.17 Some authorities rely on other forms produced a national outcome audit report which attempted
of completion imaging such as arteriography or inspection to predict outcome based upon the patient’s risk factors,
by angioscopy.18 the degree of urgency and the vascular procedure.22 The
The replacement of routine preoperative angiography most important factor in determining outcome was the
by duplex scan, the abandonment of the routine preopera- degree of urgency of the admission, with emergency oper-
tive CT brain scan and the use of high dependency nursing ations carrying the most risk. The patient’s physiology
rather than intensive care postoperatively and early dis- seemed to be of more importance than the factors encoun-
charge, within 24 hours, even after general anaesthesia, can tered at surgery. The ability to compare vascular units and
all be achieved without compromising patient safety.19 These individual vascular surgeons is likely to be based upon such
then are the complicated issues relating to risk–benefit analyses. The Portsmouth group devised P-POSSUM
analysis in the elective situation. Generalisations about (Portsmouth predictor modification of the Physiological
these issues in the emergency situation, therefore, are even and Operative Severity Score for the enUmeration of
more difficult. Any understanding of the subject will largely Mortality and morbidity) for determining outcome fol-
depend on carefully documented series from busy units lowing arterial surgery.23 When this methodology was
such as the Cleveland Clinic in which a series of 314 patients developed further (V-POSSUM) and applied to the
had undergone non-elective carotid endarterectomy.20 Vascular Surgical Society database it accurately predicted
Unfortunately the definition of ‘non-elective’ was wide both mortality and morbidity.24 Further analysis, however,
and included asymptomatic patients (9 per cent). Only suggests that different models should be used to predict
14 per cent had completed strokes and a further 2 per cent outcome following elective and ruptured AAAs,25 as there
developed unstable strokes. The median interval between was a failure of prediction of outcome in ruptured AAAs
presentation and surgery was 2 days with 48 per cent of using both the P-POSSUM and the V-POSSUM models.
operations being performed within 24 hours. The best results Other studies have looked at specific aspects of the emer-
were in the asymptomatic group with a combined stroke gency situation. For example, a prospective study of AAA
and mortality rate of 3.4 per cent; the worst results of repair found a high incidence of multiple sequential organ
14 per cent were found, predictably, in those patients with failure and colonic ischaemia in non-elective operations26
unstable strokes. Women were more likely to have ipsilat- (also see Chapters 3 and 4). A randomised study on the
eral strokes in the follow-up period (risk ratio 2.38, 95 per effect of dopexamine on colonic mucosal ischaemia after
cent confidence limits 1.02 to 5.56). Unfortunately, this elective aortic surgery suggests that it may provide signifi-
study typifies the poor quality of papers on vascular emer- cant histological protection to the colonic mucosa.27 Further
gencies and the inclusion of asymptomatic patients within studies are required to see if this drug can improve the
a non-elective group is difficult to comprehend. mortality in non-elective AAA surgery. There is some evi-
dence that the placement of pulmonary artery catheters and
the use of goal directed therapy may adversely influence
ABDOMINAL AORTIC ANEURYSM REPAIR the outcome after non-elective AAA surgery.28 In another
report a consecutive series of patients was admitted to two
The Small Aneurysm Study performed in the UK, showed different hospitals under the care of a single vascular
that there was no benefit from early surgery for infrarenal surgeon: one unit used pulmonary artery catheters in
abdominal aortic aneurysms (AAAs) measuring between 96 per cent of patients and large volumes of fluid to achieve
4 cm and 5.5 cm on ultrasound.21 This study was conducted specified targets. The other used catheters in only 18 per
on elective patients, and one of the criteria used to remove cent of cases and achieved a significantly lower mortality
patients from the surveillance arm was the development of and a lower incidence of acute renal failure. There has
symptoms such as tenderness over the aneurysm and the been, therefore, a move away from goal directed therapy
68 Assessing the risk in vascular emergencies

in many intensive care units. A randomised study showed

that the use of pulmonary artery catheters with optimisa-
LOWER LIMB ARTERIAL RECONSTRUCTION
tion of the patient’s haemodynamic status provided no
benefit in terms of outcome in patients undergoing elective A number of risk factors need to be considered in patients
vascular surgery.29 presenting with acute lower limb ischaemia who require
Various other factors may be important in predicting urgent reconstruction. The viability of a limb must be
outcome following ruptured AAA surgery and these include established before any attempt is made to revascularise it.
the avoidance of raised intra-abdominal pressure. This The absence of capillary return or the presence of fixed
increased pressure is associated with oedema of the bowel, staining of the tissues are features indicative of irreversible
retroperitoneum and the abdominal wall and is recognised ischaemia. Severe neurological damage sustained in an acci-
as the abdominal compartment syndrome30 (see Chapters dent also causing ischaemia may be a relative contraindi-
3, 4 and 23). Closing the abdominal wound with a mesh cation to vascular surgery. The duration of ischaemia
and delayed primary closure of the wound has been found and ischaemia-reperfusion injury (IRI) is crucial to out-
to be associated with a better outcome. The incidence of come (see Chapter 2). The tolerance of different tissues to
multiple sequential organ failure is lower, but the definitive ischaemia also determines outcome, for instance, periph-
trial has not yet been performed. A recent non-randomised eral nerves and muscle have less resistance than skin to
study has confirmed that patients who had a mesh-based ischaemia. Muscle which has been completely ischaemic
early abdominal closure had a better outcome than those for 4–6 hours is probably damaged irreparably. The
requiring a second operation for abdominal compartment presence of a collateral circulation, however, may help in
syndrome following primary closure.31 This is an important extending the period of ischaemia.
aspect of management which deserves consideration in Local changes may impair restoration of normal flow
Chapter 23. after the circulation has been restored. This has been called
Factors which have been associated with a poor out- the ‘impaired reflow’ or ‘no-reflow’ phenomenon (see
come after ruptured AAA repair include soluble tumour Chapter 2). Swollen cells occluding the lumen, capillary
necrosis factor (TNF) receptors,32 a low platelet count at spasm and the trapping of red and white cells all contribute
the end of the operation33 and low endothelin-1 levels34 to occlusion of the microcirculation. Restoration of the cir-
(see Chapters 3 and 4). A study of coagulation and fibri- culation following acute ischaemia leads to the formation
nolysis factors showed that rupture is associated with an of free radicals and other metabolites which then pass into
inhibition of systemic fibrinolysis and the generation of the systemic circulation where they can damage normal
thrombin.35 This procoagulant state has been postulated to tissues by the activation of neutrophils.40 These neutro-
be the cause of myocardial infarction, multiple organ phils interact with the endothelium and cause damage to
failure and thromboembolism following surgery. distant organs, such as the lungs and kidneys through the
Age has been postulated to be a factor in poor outcome release of free radicals and proteases. Cytokines, including
after a ruptured AAA but there is some evidence that good TNF and interleukins (ILs) such as IL-1, IL-6 and IL-8, are
results can be obtained if biological rather than chrono- thought to be implicated in the development of multiple
logical age is used to determine fitness for surgery.36 The organ failure in vascular patients (see Chapter 4). Thera-
distance travelled by patients remains a controversial sub- peutic measures to counteract these changes include the
ject. Patients surviving the journey to the regional centre administration of free radical scavengers such as mannitol
performing the surgery are likely to survive the operation, and allopurinol41,42 (see Chapters 2, 4 and 33).
while those who are at higher risk select themselves out by Unfortunately, many compounds effective either in
dying en route, as was clearly shown in a Northern Ireland theory or in the experimental situation do not translate
series.37 Distance, therefore probably does not make a huge into the clinical scenario. Monoclonal antibodies against
difference to the outcome.38 IL-1 were shown to be ineffective in a randomised con-
The role of endovascular repair for elective infrarenal trolled trial despite initial enthusiasm.43 Experimental
AAA is currently under investigation in the EndoVascular studies which may help in managing the clinical situation
Aneurysm Repair (EVAR) trials being performed in the in future include the use of technetium-99m-glucarate to
UK. The use of an aorto-uni-iliac device has been shown to identify muscle damage,44 prostaglandins and iloprost to
reduce the mortality of rupture significantly, although this enhance muscle blood flow,45 and the technique of thermal
view is based on small numbers.39 Control can be achieved preconditioning does seem encouraging.46 There is some
by the use of aortic occlusion balloons or by the deploy- human evidence supporting the use of ischaemic precon-
ment of the proximal stent into the neck of the aneurysm. ditioning in the prevention of IRI.47 There is currently
Revascularisation of the contralateral limb is carried out much interest in the use of recombinant human activated
by a femoro-femoral crossover bypass graft. The problems protein C for severe sepsis.48 In a randomised, double blind,
with this technique include the availability and durability placebo controlled multicentre trial, patients with systemic
of stent grafts and the logistical difficulties of mobilising inflammation and organ failure secondary to sepsis had a
both radiological and surgical staff. significantly lower mortality when treated with drotrecogin
 References 69

alfa (activated) (activated protein C) with an absolute risk

Key references
reduction of 6.1 per cent.
The use of thrombolysis is contraindicated in the patient
Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of
with multiple injuries, but may be useful in the acutely recombinant human activated protein C for severe sepsis.
ischaemic limb, especially if it is secondary to graft throm- N Engl J Med 2001; 3444: 699–709.
bosis, but the risk:benefit equation must take into account the Counsell C, Salinas R, Warlow C, Naylor R. Patch angioplasty versus
serious complications associated with thrombolysis. These primary closure for carotid endarterectomy. Cochrane
include a 3–5 per cent risk of death, a 2 per cent risk of stroke Database Syst Rev 2000; 2: CD000160.
and an incidence ranging from 5 to 12.5 per cent of major European Carotid Surgery Trialists’ Collaborative Group. MRC
haemorrhage.49–52 The use of thrombolysis for claudication European Carotid Surgery Trial: interim results for
is contraindicated because of an adverse risk:benefit ratio. symptomatic patients with severe (70–99%) or with mild
The incidence of major amputation of 7 per cent, deaths 14 (0–29%) carotid stenosis. Lancet 1991; 337: 1235–43.
North American Symptomatic Carotid Endarterectomy Trial
per cent and major haemorrhage 5 per cent in a reported
Collaborators. Beneficial effect of carotid endarterectomy in
series is much worse than the natural history of claudica-
symptomatic patients with high-grade carotid stenosis.
tion.53 Age is also associated with a poor outcome for patients N Engl J Med 1991; 325: 445–53.
with acute limb ischaemia treated with thrombolysis.54 The UK Small Aneurysm Participants. Mortality results for
There is evidence that techniques which re-establish blood randomised controlled trial of elective surgery or
flow to ischaemic tissues are beneficial and this is clearly ultrasonographic surveillance for small abdominal aortic
important in acute limb ischaemia caused by trauma. Plastic aneurysms. Lancet 1998; 352: 1649–55.
shunts placed in both arteries and veins in order to arrest tis-
sue hypoxia have been shown to be effective in the Northern
Ireland experience55 (see Chapter 33). Experimental studies REFERENCES
at that centre have confirmed the importance of restoring
venous outflow when both artery and vein are damaged.56 1 Legemate DA, Tijssen JG. Definitions and analysis of
Acute limb ischaemia occurring in a young patient complications in relation to the benefits of treatment. In:
without an obvious cause should arouse concern. Hyper- Branchereau A, Jacobs M (eds). Complications in Vascular and
coaguable states may well be responsible, and any attempt Endovascular Surgery, Part 1. New York: Futura Publishing
at reconstruction may result in amputation. Deficiencies of Company, 2001: 1–5.
2 European Carotid Surgery Trialists’ Collaborative Group. MRC
protein C, protein S, activated protein C or antithrombin
European Carotid Surgery Trial: interim results for symptomatic
III are all responsible for the hypercoaguable state.57 The
patients with severe (70–99%) or with mild (0–29%) carotid
presence of antiphospholipid antibodies and the pro- stenosis. Lancet 1991; 337: 1235–43.
thrombin gene variant (20210 G to A) are also causes of 3 North American Symptomatic Carotid Endarterectomy Trial
arterial thrombosis. Hyperhomocysteinaemia has also Collaborators. Beneficial effect of carotid endarterectomy in
been associated with poor outcome in arterial reconstruc- symptomatic patients with high-grade carotid stenosis.
tion.58–60 The risk:benefit ratio can be altered in favour of a N Engl J Med 1991; 325: 445–53.
successful outcome if conditions such as these are diagnosed 4 Cina CS, Clase CM, Haynes BR. Refining the indications for
and treated before arterial intervention is undertaken. carotid endarterectomy in patients with symptomatic carotid
artery stenosis: a systematic review. J Vasc Surg 1999; 30:
606–17.
Conclusions 5 McGuinness CL, Taylor PR. Difficulties interpreting the reported
results of carotid endarterectomy: the importance of prospective
All interventions carry a risk, particularly those involv- independent audit. Int J Clin Pract 2000; 54: 484–485.
ing the arterial circulation in the emergency setting. 6 Blohme L, Sandstrom V, Hellstrom G, et al. Complications in
Surgeons must be aware of the risk:benefit ratio involved carotid endarterectomy are predicted by qualifying symptoms
in treating patients with vascular emergencies. The clin- and preoperative CT findings. Eur J Vasc Endovasc Surg 1999; 17:
ical picture is often complex and the outcome is depend- 213–18.
ent upon many factors, some of which may be within 7 Gollege J, Cuming R, Beattie DK, et al. Influence of patient
the control of the surgeon. Inevitably, others will have variables on the outcome of carotid endarterectomy. J Vasc Surg
been established well before the surgeon is involved, and 1996; 24: 120–6.
8 Perler BA, Dardik A, Burleyson GP, et al. Influence of age
will remain outside the therapeutic sphere. Surgeons
and hospital volume on the results of carotid endarterectomy:
who deal with such emergencies need to be up to date a statewide analysis of 9918 cases. J Vasc Surg 1998;
with the latest literature to optimise patient care. The 27: 25–31.
evidence for the efficacy of various interventions is slowly 9 Kucey DS, Bowyer B, Iron K, et al. Determinants of outcome after
being accrued but there is much scope for further trials carotid endarterectomy. J Vasc Surg 1998; 28: 1051–8.
to produce convincing evidence in the treatment of 10 Pearce WH, Parker MA, Feinglass J, et al. The importance of
vascular emergencies. surgeon volume and training in outcomes for vascular surgical
procedures. J Vasc Surg 1999; 29: 768–76.
70 Assessing the risk in vascular emergencies

11 Counsell C, Salinas R, Warlow C, Naylor R. Patch angioplasty 29 Bender JS, Smith-Meek MA, Jones CE. Routine pulmonary artery
versus primary closure for carotid endarterectomy. Cochrane catheterization does not reduce morbidity and mortality after
Database Syst Rev 2000; 2: CD000160. elective vascular surgery: results of a prospective randomised
12 Counsell C, Warlow C, Naylor R. Patches of different types for trial. Ann Surg 1997; 226: 229–36.
carotid patch angioplasty. Cochrane Database Syst Rev 2000; 30 Oelschlager BK, Boyle EM Jr, Johansen K, Meissner MH. Delayed
2: CD000071. abdominal closure in the management of ruptured abdominal
13 Radak D, Popovic AD, Radicevic S, et al. Immediate reoperation aortic aneurysms. Am J Surg 1997; 173: 411–15.
for perioperative stroke after 2250 carotid endarterectomies: 31 Rasmussen TE, Hallett JW, Noel AA, et al. Early abdominal closure
differences between intraoperative and early postoperative with mesh reduces multiple organ failure after ruptured
stroke. J Vasc Surg 1999; 30: 245–51. abdominal aortic aneurysm repair: guidelines from a 10-year
14 Seelig MH, Oldenburg WA, Chowla A, Atkinson EJ. Use of case-controlled study. Society for Vascular Surgery Abstract
intraoperative duplex ultrasonography and routing patch Booklet 2001: 86.
angioplasty in patients undergoing carotid endarterectomy. 32 Adam DJ, Lee AJ, Ruckley CV, et al. Elevated levels of soluble
Mayo Clinic Proc 1999; 74: 870–6. tumor necrosis factor receptors are associated with increased
15 Anderson A, Padayachee TS, Sandison AJP, et al. The results of mortality rates in patients who undergo operation for ruptured
routine primary closure in carotid endarterectomy. Cardiovasc abdominal aortic aneurysm. J Vasc Surg 2000; 31: 514–19.
Surg 1999; 7: 50–5. 33 Bradbury AW, Bachoo P, Milne AA, Duncan JL. Platelet count and
16 Padayachee TS, Brooks MD, Modaresi KB, et al. Intraoperative the outcome of operation for ruptured abdominal aortic
high resolution duplex imaging during carotid endarterectomy: aneurysm. J Vasc Surg 1995; 21: 484–91.
which abnormalities require surgical correction? Eur J Vasc Surg 34 Adam DJ, Evans SM, Webb DJ, Bradbury AW. Plasma endothelin
Endovasc Surg 1998; 15: 387–93. levels and outcome in patients undergoing repair of ruptured
17 Padayachee TS, Brooks MD, McGuinness CL, et al. Value of infrarenal abdominal aortic aneurysm. J Vasc Surg 2001; 33:
intraoperative duplex imaging during supervised carotid 1242–6.
endarterectomy. Br J Surg 2001; 88: 389–92. 35 Adam DJ, Ludlam CA, Ruckley CV, Bradbury AW. Coagulation and
18 Naylor AR, Hayes PD, Allroggen H, et al. Reducing the risk of fibrinolysis in patients undergoing operation for ruptured and
carotid surgery: a 7-year audit of the role of monitoring and nonruptured infrarenal abdominal aortic aneurysms. J Vasc Surg
quality control assessment. J Vasc Surg 2000; 32: 750–9. 1999; 30: 641–50.
19 Sandison AJP, Wood CH, Padayachee TS, et al. Cost effective 36 Robson AK, Currie IC, Poskitt KR, et al. Abdominal aortic aneurysm
carotid endarterectomy. Br J Surg 2000; 87: 323–7. repair in the over eighties. Br J Surg 1989; 76: 1018–20.
20 Tretter JF, Hertzer NR, Mascha EJ, et al. Perioperative risk and 37 Barros D’Sa AAB. Optimal travel distance before ruptured aortic
late outcome of nonelective carotid endarterectomy. J Vasc Surg aneurysm repair. In: Greenhalgh RM, Mannick JA (eds). The Cause
1999; 30: 618–31. and Management of Aneurysms. London: WB Saunders 1990;
21 The UK Small Aneurysm Participants. Mortality results for 409–31.
randomised controlled trial of elective surgery or 38 Adam DJ, Mohan IV, Stuart WP, et al. Community and hospital
ultrasonographic surveillance for small abdominal aortic outcome from ruptured abdominal aortic aneurysm within the
aneurysms. Lancet 1998; 352: 1649–55. catchment area of a regional vascular service. J Vasc Surg 1999;
22 Earnshaw JJ, Ridler BMF, Kinsman R on behalf of the 30: 922–8.
Audit Committee of the Vascular Surgical Society of Great 39 Ohki T, Veith FJ. Endovascular grafts and other image-guided
Britain and Ireland. National Outcome Audit Report. London: catheter-based adjuncts to improve the treatment of ruptured
Vascular Surgical Society of Great Britain and Ireland, aortoiliac aneurysms. Ann Surg 2000; 232: 466–79.
May 2000. 40 Pararajasingam R, Nicholson ML, Bell PRF, Sayers RD.
23 Prytherch DR, Whiteley MS, Higgins B, et al. POSSUM and Non-cardiogenic pulmonary oedema in vascular surgery. Eur J
Portsmouth POSSUM for predicting mortality. Br J Surg 1998; Vasc Endovasc Surg 1999; 17: 93–105.
85: 1217–20. 41 Nicholson ML, Baker DM, Hopkinson BR, Wenham PW.
24 Prytherch DR, Beard JD, Ridler BF, Earnshaw JJ. Vascular Surgical Randomised controlled trial of the effects of mannitol on renal
Society operative outcome study: preoperative physiology reperfusion injury during aortic aneurysm surgery. Br J Surg
predicts outcome. Br J Surg 2000; 87: 507–8. 1996; 83: 1230–3.
25 Prytherch DR, Sutton GL, Boyle JR. Portsmouth POSSUM 42 Soong CV, Young IS, Lightbody JH, et al. Reduction of free
models for abdominal aortic aneurysm surgery. Br J Surg 2001; radical generation minimises lower limb swelling following
88: 958–63. femoropopliteal bypass surgery. Eur J Vasc Surg 1994; 8: 435–40.
26 Sandison AJP, Panayiotopoulos YP, Edmondson RC, et al. A four 43 Opal SM, Fisher CJ Jr, Dhainaut JF, et al. Confirmatory
year prospective audit of the cause of death after infrarenal interleukin-1 receptor antagonist trial in severe sepsis: a phase III,
aortic aneurysm surgery. Br J Surg 1996; 83: 1386–9. randomised, double-blind, placebo-controlled, multicenter trial.
27 Baguneid MS, Welch M, Bukhari M, et al. A randomized study to The Interleukin-1 Receptor Antagonist Sepsis Investigator Group.
evaluate the effect of a perioperative infusion of dopexamine on Crit Care Med 1997; 25: 1115–24.
colonic mucosal ischemia after aortic surgery. J Vasc Surg 2001; 44 Wiersema AM, Oyen WJG, Verhofstad AAJ, et al. Early assessment
33: 758–63. of skeletal muscle damage after ischaemia-reperfusion using
28 Sandison AJP, Wyncoll DLA, Edmondson RC, et al. ICU protocol Tc-99m-glucarate. Cardiovasc Surg 2000; 8: 186–91.
may affect the outcome of non-elective abdominal aortic 45 Rowlands TE, Gough MJ, Homer-Vanniasinkam S. Do prostaglandins
aneurysm repair. Eur J Vasc Surg Endovasc Surg 1998; 16: have a salutary role in skeletal muscle ischaemia-reperfusion
356–61. injury? Eur J Vasc Endovasc Surg 1999; 18: 439–44.
 References 71

46 McLaughlin R, Kelly CJ, Kay E, Bouchier-Hayes D. Diaphragmatic 53 Braithwaite BD, Tomlinson MA, Walker SR, et al. Peripheral
dysfunction secondary to experimental lower torso ischaemia- thrombolysis for acute-onset claudication. Thombolysis Study
reperfusion injury is attenuated by thermal preconditioning. Br J Group. Br J Surg 1999; 86: 800–4.
Surg 2000; 87: 201–5. 54 Braithwaite BD, Davies B, Birch PA, et al. Management of acute
47 Kharbanda RK, Peters M, Walton B, Kattenhorn M, et al. Ischemic leg ischaemia in the elderly. Br J Surg 1998; 85: 217–20.
preconditioning prevents endothelial injury and systemic 55 Barros D’Sa AAB. Shunting in complex lower limb vascular
activation during ischemia-reperfusion in humans in vivo. trauma. In: Greenhalgh RM, Hollier LH (eds). Emergency Vascular
Circulation 2001; 103: 1624–30. Surgery. London: WB Saunders, 1992; 331–44.
48 Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety 56 Harkin DW, Barros D’Sa AAB, Yassin MMI, et al. Reperfusion
of recombinant human activated protein C for severe sepsis. injury is greater with delayed restoration of venous outflow in
N Engl J Med 2001; 3444: 699–709. concurrent arterial and venous limb injury. Br J Surg 2000; 87:
49 Ouriel K, Veith FJ, Sasahara AA for the Thrombolysis or Peripheral 734–41.
Arterial Surgery (TOPAS) Investigators. A comparison of 57 Bontempo FA, Kibbe MR, Makaroun MS. Hypercoaguable states
recombinant urokinase with vascular surgery as initial treatment and unexplained vascular thrombosis. In: Branchereau A, Jacobs M
for acute arterial occlusion of the legs. N Engl J Med 1998; 338: (eds). Complications in Vascular Surgery and Endovascular Surgery,
1105–11. Part I. New York: Futura Publishing Company, 2001; 13–22.
50 Weaver FA, Comerota AJ, Youngblood M, et al. and the STILE 58 Neilsen TG, Nordestgaard BG, von Jessen F, et al. Antibodies to
Investigators. Surgical revascularization versus thrombolysis for cardiolipin may increase the risk of failure of peripheral vein
nonembolic lower extremity native artery occlusions: results of a bypasses. Eur J Vasc Endovasc Surg 1997; 14: 177–84.
prospective randomized trial. J Vasc Surg 1996: 24: 513–23. 59 Doggen CJ, Cats VM, Bertina RM, Rosendaal FR. Interaction of
51 Berridge DC, Makin GS, Hopkinson BR. Local low dose coagulation defects and cardiovascular risk factors: increased
intra-arterial thrombolytic therapy: the risk of stroke or major risk of myocardial infarction associated with factor V Leiden or
haemorrhage. Br J Surg 1989; 76: 1230–3. prothrombin 20210 A. Circulation 1998; 97: 1037–41.
52 Comerota AJ, Weaver FA, Hosking JD et al. Results of a 60 Taylor LM Jr, Moneta GL, Sexton GJ, et al. Prospective blinded
prospective, randomised trial of surgery versus thrombolysis for study of the relationship between plasma homocysteine and
occluded lower extremity bypass grafts. Am J Surg 1996; 172: progression of symptomatic peripheral arterial disease. J Vasc
105–12. Surg 1999; 29: 8–19.
This page intentionally left blank
 7A
Perioperative Care in Emergency Vascular Practice

ALASDAIR DOW, JOHN F THOMPSON

The problem 73 Red cell transfusion 77

Initial management 73 Problems associated with massive transfusion 78
‘Preoptimisation’ of emergency vascular patients 73 Component therapy 79
Perioperative cardiac protection 74 Autologous blood transfusion 80
Initial resuscitation 75 Pharmacological agents 81
Resuscitation fluids: how much is enough? 75 Regional block or not for vascular emergencies 81
Intraoperative volume replacement 76 References 82

THE PROBLEM peripheral vasoconstriction and venous access may be dif-

ficult to obtain. The best way to be certain is to anticipate
and to insert wide bore cannulae before initiating proced-
Vascular teams are frequently confronted by patients who ures which may be associated with unexpected blood loss
are rapidly losing blood or have already lost it. Ruptured such as thrombolysis.
aortic aneurysms or trauma involving major vessels are During the management of major trauma lines should
obvious examples but the bleeding may be iatrogenic, for not be inserted in the lower half of the body. There are good
example, vascular injury during interventional radiological reasons for this, such as occult venous injury in the pelvis or
procedures. This chapter concerns immediate resuscita- the need to use lower limb veins for bypass grafting. Most
tion of the shocked patient, blood transfusion and compon- texts recommend the use of arm veins for emergency venous
ent therapy, intraoperative autotransfusion and blood access, but modern catheter systems allow central venous
management in the perioperative period. lines to be inserted with speed and reliability (see Chapter
Intensive care units (ICUs) and high dependency units 7B). Cannulation of the internal jugular vein with the
(HDUs) may improve outcomes for patients with a wide patient in a head-down tilt avoids the risk of pneumothorax
variety of surgical conditions. Recent data suggest that high and is the best approach in patients who are already pro-
risk surgical patients who receive such care prior to surgery foundly hypovolaemic. A large introducer sheath may then
may have reduced mortality rates and a shorter hospital stay. be used for volume resuscitation or a central venous or pul-
A number of current topics, including the choice of monary artery catheter may be inserted. Once intravenous
fluids, timing of surgery, estimation of cardiac output and access is established a urinary catheter should be inserted.
-blockade are relevant and important. The postoperative
care of patients after emergency vascular surgery will also
be considered in this chapter. These problems represent the
most vexing issues for the surgical team and suggestions ‘PREOPTIMISATION’ OF EMERGENCY
for dealing with them will be discussed. VASCULAR PATIENTS

High dependency units and ICUs provide a level of treatment

INITIAL MANAGEMENT not available on the majority of general wards, so-called ‘aug-
mented care’, requiring a number of resources. Improved
Attention to the airway, breathing and circulation must be staffing, often a 1:1 staff to patient ratio is crucially import-
emphasised. The shocked patient has undergone profound ant. Invasive monitoring, such as measurement of left atrial
74 Perioperative care in emergency vascular practice

filling pressure and cardiac output, is also essential. More reduced cardiac morbidity and graft thrombosis in peripheral
recently, mixed venous oxygen saturation (SvO2) monitor- vascular surgery.9 Others, however, were unable to support
ing has been used as an early detector of changes in cardiac this finding.10 Ziegler et al. studied 72 patients undergoing
output and haemoglobin concentration following aortic aortic or limb salvage surgery.11 All were admitted to ICU
surgery.1 This technique is also useful in critically ill preoperatively and randomised to treatment or control.
patients following cardiac surgery.2 The use of complex The 32 patients in the treatment group had their physio-
ventilatory strategies, including invasive and non-invasive logical variables adjusted to improve their SvO2 to above
ventilation, makes a significant difference to the quality of 65 per cent after pulmonary artery catheter insertion. The
care afforded to these patients. Inevitably, the development control group had a similar catheter inserted, but no
of these modern units will prove to be of great benefit. attempt was made to adjust their SvO2. Mortality was 9
per cent in the treatment group and 5 per cent in the control
group, there being no significant difference. It is difficult to
Features of ‘augmented care’ draw conclusions from this small series and there is a
pressing need for a large well-powered study to guide the
use of such expensive preoperative care.
• 1:1 staff: patient ratio
• Invasive monitoring
• Advanced ventilatory strategies
PERIOPERATIVE CARDIAC PROTECTION

There has been considerable debate as to the cost- Vascular surgery involves changes in systemic vascular
effectiveness of such units, but augmented care may resistance (SVR), resulting in alteration in cardiac output.
improve outcome and shorten hospital stay.3 The problem A rise in SVR increases afterload against which the left
is to select those who will benefit and whether time is avail- ventricle has to contract may induce left ventricular failure.
able to do that. The ruptured aneurysm is usually too Conversely, a reduction in SVR may lead to diastolic
urgent a condition to allow anything other than transfer to hypotension, precipitating myocardial ischaemia in patients
theatre, even though patients with leaking/expanding with narrowed coronary arteries. Thus, the manipulation
aortic aneurysms could be considered good candidates. of SVR and cardiac output may be beneficial for emer-
Compared with general surgical emergencies, these gency vascular surgery patients, particularly those with low
patients have a higher incidence of cardiovascular disease, intravascular volume.
spend a prolonged time in theatre and are at risk of renal The gold standard over the past two decades for the
and cardiac complications. estimation of cardiac output has been the Swan–Ganz
There is no agreement on the best method of selecting pulmonary artery catheter which uses thermodilution to
patients who might benefit from ‘preoptimisation’. Echo- estimate cardiac output and software to calculate SVR and
cardiography, either resting or during induced stress, has the other variables. Recent work, however, indicates that the
advantage that it is quicker and less invasive than thallium pulmonary artery catheter may not improve outcome and
perfusion scans and is effective in detecting wall motion may actually contribute to morbidity.12,13
defects in patients with vascular disease.4 Dobutamine There has been recent interest in non-invasive measure-
echocardiography is a good predictor of cardiac risk.5,6 An ment of cardiac output. Doppler ultrasound probes do not
alternative is the use of a scoring system such as POSSUM require central venous cannulation, but their accuracy may
(Physiological and Operative Severity Score for the be more operator dependent. Nevertheless, there is good
enUmeration of Mortality and Morbidity).7 In London, 101 correlation between values obtained from a pulmonary
preoperative general/vascular surgical patients were stud- artery catheter and those calculated by Doppler.14 The
ied. They were ill enough to warrant admission to ICU, but probe is passed into the oesophagus until a mark on the
resources did not permit it. The outcome was predicted by sheath reaches the teeth. The sheath is then adjusted until
POSSUM, and the actual outcomes in those admitted to ICU the ‘best’ signal of descending thoracic aortic velocity is
preoperatively were compared with those admitted post- obtained. To reduce operator error some machines will
operatively: mortality was reduced in patients admitted assist in indicating when that has occurred.
preoperatively.8 Transoesophageal Doppler measures stroke volume,
Which therapies are beneficial in the HDU/ICU? Most cardiac output and systolic flow time corrected for heart
studies are difficult to interpret; and even in emergency rate (FTc), which is an indication of contractility. The
vascular surgery mortality is low, so that the study size majority of studies have looked at the improvement in FTc
needs to be large. However, a theme emerges. Measurement using fluids and/or inotropes. In patients undergoing
of right or left heart filling and improvement in oxygen tis- elective cardiac surgery, Mythen and Webb aimed to show
sue delivery may confer benefit. Berlauk et al. showed that that fluid optimisation would be better using Doppler, as
preoperative improvement of haemodynamic variables demonstrated by better gut perfusion.15 Patients were
 Resuscitation fluids: how much is enough? 75

randomised to a control group receiving standard fluid bleed caused by an aortic tear,18 however, demonstrated
therapy or to the protocol group in which therapy was based the detrimental postoperative effects of high volumes of
on Doppler derived variables. The protocol group had a saline.
shorter ICU and hospital stay with no serious complications, The redistribution of crystalloids through the extravas-
though six in that group did suffer some complications. cular compartment means that they are usually required in
Gan et al. studied 100 patients predicted to have a blood loss volumes five times that of colloid. This ratio increases further
of greater than 500 mL.16 All underwent Doppler probe in situations of trauma including ruptured aneurysms.19
insertion and were then randomised to standard or therapy This redistribution, or ‘third spacing’, is more common
groups. The standard group received fluid boluses according after crystalloid rather than colloid use and can lead to pul-
to deviations from baseline variables not measured by the monary oedema.20 Hydrostatic pressure is more important
Doppler. The therapy group received fluid based on their than colloid osmotic pressure in the movement of water
stroke volume and FTc. The therapy group had a shorter across the pulmonary capillary membrane.
stay, earlier time to diet and less nausea and vomiting. There is little evidence to indicate that colloids reduce
Although this work has not been repeated in vascular sur- mortality or morbidity, despite their theoretical advantages.
gery specifically, the lower complication rate associated with Indeed, doubts were expressed that the use of albumin
Doppler makes it an attractive option in high risk cases. solutions in critically ill patients might be associated with a
There is great interest in the use of -adrenergic blockade higher mortality. A meta-analysis, however, showed that
to reduce the risk of perioperative myocardial infarction in albumin was not associated with alteration in survival.21
vascular surgery. Following several positive observational The studies included in the analysis were not all surgical,
studies, Poldermans reported a randomised trial in 112 but a more complex meta-analysis from the Australian
patients undergoing aortic reconstruction, who had a posi- Cochrane Centre looked at colloid solutions in patients with
tive dobutamine stress test and were not already taking a trauma, burns or after surgery.22 The authors concluded
-blocker. Fifty-nine were randomly assigned to treatment that there was no benefit attributable to any of the colloids
with bisoprolol and 53 to standard care. The combined car- over crystalloids and it was suggested that their extra cost
diac death/non-fatal myocardial infarction rate was 3.4 per could not be justified.
cent with bisoprolol compared to 34 per cent with standard A recent Australian study adds further uncertainty as to
care (P  0.001).17 A prospective trial is underway, but the choice of fluid type. The Saline versus Albumin Fluid
many surgeons are already convinced. Evaluation (SAFE) study used a multicentre prospective
randomised trial method to allocate patients to receive
either 4 per cent albumin or saline for fluid resuscitation.23
INITIAL RESUSCITATION The chosen fluid was employed exclusively during the 28-
day study period for intravascular resuscitation of 6997
patients recruited, of whom 3497 received albumin. The
The precise fluid used for resuscitation is relatively unim- results showed that there were 726 deaths in the albumin
portant. Colloidal solutions, which are starch or gelatin group as against 729 in the saline group (risk ratio 0.99;
derivatives, are relatively expensive and contain potential 95 per cent confidence interval 0.91 to 1.09; P  0.87).
allergens. Hydroxyethyl starch has an average molecular Further, there was no significant difference between the
weight of 450 000 and a half-life of 26 hours. Polygelines number of days spent in the ICU (P  0.44), days in
have a molecular weight of 35 000 and a half life of 2.5 hospital (P  0.30) or days of renal replacement therapy
hours. Hetastarch is the most effective when packed cells are (P  0.41). The power of this study is sufficiently great
transfused in view of its longer half-life. Albumin is no longer to justify the conclusion that in a broad cross-section
regarded as a suitable plasma expander in view of cost and of ICU patients there is a similar outcome following the use
the danger of transmissible disease. With regard to clear of either 4 per cent albumin or saline as a resuscitation
fluids, normal saline contains an excess of chloride, so that fluid.
there is a theoretical risk of hyperchloraemia; also Ringer’s The vascular team is faced with a wealth of data that does
solution contains lactate which may worsen acidosis. not support one fluid type over the other. Until a large-
The ‘crystalloid versus colloid’ argument is based on scale trial favours one type of fluid, it is likely that personal
highly controlled experiments where subjects, usually dogs or institutional preference will predominate.
or pigs, were venesected to induce hypotension and then
resuscitated. It has now become clear that shock increases
microvascular permeability allowing large molecules to
cross into the interstitial space. During recovery they may
RESUSCITATION FLUIDS: HOW MUCH IS
be difficult to remove and may exert an osmotic effect
ENOUGH?
which draws fluid from the intravascular compartment.
The fashion then swung to high volume crystalloid The Advanced Trauma Life Support (ATLS) scheme and
resuscitation. Experimental studies involving a more ‘realistic’ the American College of Surgeons Committee on Trauma
76 Perioperative care in emergency vascular practice

emphasise the importance of aggressive volume resuscitation INTRAOPERATIVE VOLUME REPLACEMENT

in hypotensive patients, and this has been amplified by
recent reports in patients with ruptured abdominal aortic
aneurysms.24 Nevertheless, as long ago as 1928, Canon et al.25 Loss of 20 per cent of the blood volume can generally be
concluded, ‘Haemorrhage in the case of shock may not replaced with clear fluids (Table 7A.1), but in the emer-
have occurred to a marked degree because blood pressure gency situation it may be impossible to estimate blood loss
has been too low and flow too scant to overcome the obs- on clothing or drapes and since rapid haemorrhage involves
tacle offered by a clot. If the pressure is raised before the simultaneous loss of red cells and plasma, the peripheral
surgeon is ready to check any bleeding that may take place, venous haemoglobin concentration or haematocrit do not
blood that is sorely needed may be lost’. fall immediately.
The ‘leaky bucket’ syndrome, where injudicious resus- Pulse, blood pressure and urine output reflect the high
citation leads to further bleeding by disrupting haemostatic pressure side of the circulation, which contains only a frac-
clots within constricted vessels has been demonstrated tion of the blood volume and is maintained by well-known
by Blair et al. who showed an increased rate of re-bleeding neurohormonal reflexes. The greatest part of the intravas-
after upper gastrointestinal haemorrhage in transfused cular volume resides in the venous capacitance vessels of
patients.26 Aggressive resuscitation was questioned by the splanchnic circulation. Adrenergic venoconstriction
Kaweski et al.27 who studied 6855 trauma patients and leads to contraction of this compliant reservoir to maintain
found no correlation between survival and preoperative right atrial filling pressure. Thus, central venous pressure
fluid resuscitation in patients with similar probabilities of may not reflect the state of the blood volume. In the venti-
survival as assessed by TRISS criteria, TRISS being a lated patient, even pulmonary artery wedge pressure may
trauma scoring method based on a combination of the be difficult to determine if positive end expiratory pressure
Revised Trauma Score (RTS) and the Injury Severity Score is used and may not reflect left ventricular filling pressure.30
(ISS). Although hypotension was an overall predictor of In shocked patients the vascular space is contracted and is
poor outcome, the administration of fluids had no influ- not re-expanded even by large quantities of clear fluid and
ence on it. Further evidence that intravenous fluids may blood. This was demonstrated well by Simmons et al.31 in
not be beneficial, but could be harmful, was provided by 29 combat victims, 51Cr-labelled autologous red cells were
Bickell et al.28 A total of 300 consecutive patients with gun- used to estimate red cell mass and plasma volume was
shot or stab wounds with a systolic blood pressure of measured using 125I-labelled albumin. 125I albumin leaves
90 mmHg or less were randomised to either immediate the circulation in shocked patients and therefore blood vol-
intravenous resuscitation by a paramedic team (n  96) or umes were overestimated. Despite this, 13 of the 29 patients
delayed resuscitation (n  81). The latter received no had significantly reduced red cell volume after they had
intravenous fluid until ‘knife to skin’. There were no differ- been ‘adequately’ resuscitated by conventional criteria;
ences in the rate of postoperative complications, but if the reductions in the red cell volume did not correlate well with
data are analysed to only include strict protocol adherents, haemodynamic measurements in shocked patients.
i.e. absolutely no fluid resuscitation, there was a survival In practice, clinical judgement, in other words ‘edu-
advantage. cated guesswork’, and a combination of ‘hard’ measure-
In conclusion, there is convincing evidence against the ments must be used to gauge the volume required to be
use of routine blood pressure elevation by the aggressive transfused. A good technique is to assess the effect of a fluid
administration of intravenous fluids before surgical challenge by rapidly infusing boluses of 200 mL and moni-
haemostasis in patients with trauma.29 toring the effect on various parameters.

Table 7A.1 Pathophysiological features of hypovolaemia

Class I Class II Class III Class IV

Per cent loss 15 15–30 30–40 40

Volume (mL) 750 750–1500 1500–2000 2000
Systolic Unchanged Normal Reduced Low
Pulse (beats per minute) 100 100 120 140
Diastolic Unchanged Raised Reduced Very low
Capillary fill Normal Slow Slow Undetectable
Respiratory rate (per minute) 14–20 20–30 30–40 40
Urine output (mL) 30 20–30 30–40 40
Mental Apprehensive Anxious Confused Lethargic
 Red cell transfusion 77

RED CELL TRANSFUSION very safe in an emergency. O Rhve blood is not always
available and it is permissible to give male patients O Rhve
blood and accept the risk of seroconversion. If female
Red cell transfusion restores oxygen carrying capacity to
Rhve patients are given Rhve blood they can be given
maintain tissue oxygen delivery. Red cells should not be
anti-D serum if they plan to have children in the future. The
used for volume expansion and indeed the adage that
principal reason for avoiding universal group O transfusion
blood loss must be replaced with whole blood is simplistic
in trauma is that group O blood is often in short supply.
and dangerous.
There is often a surplus of group A blood; group AB patients
As with all other aspects of trauma management, com-
can be transfused with group A or B blood, especially if this
munication is vital. The haematology technician on call
is in the form of packed cells.
must be notified of the arrival of the patient and given an
Crossmatching detects reactions between an antibody
estimate of blood which will be required and of the time
present in the recipient’s blood and rare antigens present
scale involved. He or she will usually be involved with
on donor cells. It takes 15–20 minutes to crossmatch blood
other patients and must plan their laboratory work accord-
using modern techniques and it is almost always possible
ingly. Adequate samples should include 5 mL of blood in
to wait before transfusion. Testing of the donor serum
EDTA for full blood count and preferably two 10 mL plain
against the patient’s red cells is no longer necessary, espe-
glass tubes for crossmatch. If there is any doubt regarding
cially if plasma-reduced blood is transfused.
pre-existing coagulopathy, a 5 mL citrate sample should be
Red cell concentrates are now cleared of most of
sent. Coagulopathy is a poor prognostic indicator in patients
their leucocyte content by a process of sedimentation and
with ruptured abdominal aortic aneurysm32 and this prob-
may be further leucodepleted by filtration. This is to
ably applies to other cases of vascular trauma. Results of
reduce the possibility of transmitting leucocyte-associated
coagulation tests may guide early and appropriate compon-
viruses such as cytomegalovirus, other DNA herpes viruses
ent therapy.
and the human T cell group. At the time of writing it is
Accurate labelling is vital, especially in the emergency
known that prion protein has been demonstrated on
situation. The name, date of birth and hospital number
lymphocytes, monocytes and platelets and so there is a
should be checked with wrist band identification whenever
theoretical risk of transmission of variant Creutzfeldt–
possible, as clerical error is still the leading cause of fatal
Jakob disease by blood transfusion. The message for surgi-
transfusion reactions. It is helpful if previous potentially
cal practice is that blood transfusion is life saving but
sensitising episodes such as blood transfusion and preg-
should be carefully monitored to avoid undue exposure to
nancy are mentioned on the request form.
infective risk.
Blood should be infused through sterile giving sets
containing a standard 170 m filter. The set is calibrated
Blood volume replacement so that there are 20 drops of blood to 1 mL. Further
filtration of blood is unnecessary. In trauma, rapid trans-
fusion is probably more important than the benefits of
• Clear fluid infusion – crystalloids versus colloids?
filtration.
How much?
• Whole blood transfusion – type specific/O Rhesus
negative (Rhve) in an emergency
Intraoperative transfusion
• Red cell transfusion – packed cells
• Component therapy – platelets/fresh frozen plasma/
The decision to transfuse blood must be taken with regard
cryoprecipitate
to the overall clinical picture, especially the presence of
• Autologous blood transfusion
coexisting cardiac and respiratory disease, and whether
bleeding is continuing. The optimum haemoglobin level
for tissue oxygen delivery is 10 g/dL because blood viscos-
It is well known that group O Rhve blood is the uni- ity falls with the haematocrit and intracapillary red cell
versal donor but there is a reluctance to use it33 despite flow increases.
prospective studies demonstrating its effectiveness.34 As Portable devices such as the Stat-Crit (Unipath Ltd,
anti A, B or AB antibodies are present in plasma, packed Bedford, UK) or Haemocue (Angelholm, Sweden) enable
cells are safer than whole blood. repeatable, rapid haematocrits or haemoglobin concentra-
In a rapidly bleeding patient, the transfusion of uncross- tions to be determined in theatre to guide blood transfu-
matched blood is not nearly as important as the speed of the sion (Fig. 7A.1). The accepted teaching that a single unit
transfusion. The compatibility of uncrossmatched ran- transfusion is anathema has been challenged by the use of
domly selected blood is 64 per cent. If blood is ABO com- this equipment. Repeated estimations enable the response
patible, this rises to 99.4 per cent, only autologous red cells to a single unit transfusion to be measured; if the agreed
being 100 per cent compatible. Thus, type specific blood is trigger is exceeded, transfusion is stopped.
78 Perioperative care in emergency vascular practice

Chronic anaemia is well tolerated when there are no other

complicating factors. Patients with renal failure frequently
have haematocrits between 20 and 30 per cent and patients
with hookworm infestations have survived haematocrits
below 10 per cent. Czer and Shoemaker investigated peri-
operative mortality based on pretransfusion haematocrit: at a
haematocrit of less than 21 per cent, 68 per cent of a group of
94 critically postoperative patients died; at a haematocrit of
between 27 and 33 per cent, 87 per cent survived, and yet at a
haematocrit of greater than 33 per cent only 50 per cent sur-
vived.35 In a case–control study of 125 Jehovah’s Witnesses,
operative mortality was inversely related to preoperative
haemoglobin level.36 Mortality was 7.1 per cent for patients
with haemoglobin levels above 10 g/dL but rose sharply to
61.5 per cent for those with levels below 6 g/dL. Mortality was
Figure 7A.1 Haemocue device (Angelholm, Sweden) for in-theatre also, not surprisingly, related to perioperative blood loss. In
haemoglobin estimations can be used to direct blood transfusions. this study no patient with a haemoglobin above 8 g/dL and
Seen here next to the Cobe BRAT II cell salvage equipment (Cobe an operative blood loss below 500 mL died.
Cardiovascular, Arvada, CO, USA) during aneurysm repair At haematocrit levels greater than 25 per cent the heart
rate is normal. Stroke volume and therefore cardiac output
Postoperative transfusion increase as long as normovolaemia is maintained. Coronary
blood flow increases due to coronary vasodilatation and
Two important factors lead to an artificially low haemo- increased blood fluidity. Direct electrode studies have
globin level after operation. ‘Preloading’ the circulation with shown that the tissue oxygen concentration of the gut, kid-
clear fluid dilutes the red cells and epidural anaesthesia leads ney, muscle, brain and heart remains normal at haem-
to sympathetic blockade in the lower half of the body. atocrits greater than 20 per cent. Both in animals and
Subsequent vasodilation increases intravascular volume, patients wound healing is not impaired by haematocrits of
which is filled by fluid shift from the interstitial space and by 15–20 per cent provided that blood volume is normal and
exogenous fluid transfusion. In the postoperative period this tissue perfusion is kept at near normal levels.37 In practice
excess intravascular fluid is excreted and redistributed, result- this means that it is probably safe to err on the side of
ing in a rise in the haemoglobin concentration of 1.5–2.0 g/dL. undertransfusion and to use repeated haematocrit estima-
In trauma the true preoperative haemoglobin concentra- tions to guide careful replacement of operative blood loss.
tion is unknown and the red cell mass can only be estimated This strategy of conservative postoperative transfusion,
from body mass. In the clinical situation hypovolaemic studied by Hebert et al.38 showed that both a survival
anaemia can be identified by assessing the response to red advantage as well as a lower complication rate could be
cell transfusion. Hypovolaemic patients accommodate both anticipated in ICU patients.
resuscitation fluids and red cells so that the proportion of
red cells, that is the haematocrit, and the concentration of
haemoglobin do not change significantly. If blood volume is
normal, transfused blood raises the red cell mass and the PROBLEMS ASSOCIATED WITH MASSIVE
plasma volume. Redistribution and subsequent diuresis of TRANSFUSION
the ‘excess’ plasma raises both the haematocrit and haemo-
globin concentrations. Laboratory determination of these There are serious problems associated with high volume
parameters may be inaccurate if performed too soon after blood transfusion (also see Chapter 4). Incompatibility
transfusion, i.e. before redistribution has taken place. and volume overload have been discussed. Hypothermia is
Guidelines for red cell transfusion are available on several common if blood is used straight from the refrigerator.
websites such as http://transfusionguidelines.org.uk and all Countercurrent heat exchangers such as the Level 1 device
follow the general recommendations given in the box below. (Level 1, Inc., Rockland MA, USA) enable fluid to be
infused rapidly at near normal temperature.
Guidelines for red cell transfusion A unit of blood contains 67.5 mL of citrate but only
35 mL is required to chelate calcium to prevent coagulation.
Thus, in cases of rapid transfusion, free citrate may be present
• No transfusion if haemoglobin 10 g/dL and may cause myocardial irritability and even ventricular
• Transfuse if haemoglobin 7 g/dL fibrillation in hypothermic patients. Calcium gluconate is
• Haemoglobin 8–10 g/dL is safe if euvolaemic, even used if transfusion is very rapid.
with cardiopulmonary disease
As greater volumes of bank blood are transfused, mixed
• Transfuse symptomatic anaemic patients venous oxygen tension falls as a consequence of the increased
 Component therapy 79

Table 7A.2 Blood components for the bleeding patient; indications and doses. Remember to warm the patient and
transfuse fluids

Blood component Indication Dose

Platelets Platelet count 50 109 and bleeding 250 109 (adult)
Count 100 109 and serious bleeding
Fresh frozen plasma 1.0 blood volumes transfused four packs for average adult (15 mL/kg)
INR 1.5 and continued bleeding
Cryoprecipitate Only if fibrinogen 1.0 g/L 10 units initially
Use fresh frozen plasma first

INR, international normalised ratio.

oxygen affinity of bank blood. This is manifest as a decrease previous studies.43 The study was a prospective single arm
in the in vivo P50. Although animal experiments have always multicentre trial in the USA and Puerto Rico, and looked at
shown that this does not affect myocardial performance, the 28-day all-cause mortality in patients receiving activated
Weisel et al. demonstrated that arteriopaths are unable to protein C. The study recruited 273 patients and showed
increase cardiac index due to their impaired heart muscle that the 28-day mortality was 26.4 per cent. This was 6
function. The mechanism may involve decreased inorganic per cent lower than the mortality of the placebo group in
phosphate or decreased ionised calcium associated with the Protein C Worldwide Evaluation in Severe Sepsis
blood transfusion.39 (PROWESS) and Secretory Phospholipase A2 Inhibitor
Coagulopathy following massive transfusion generally (sPLA2I) trials, two trials studying similar groups of
occurs after more than 15 units have been given. Miller et al. patients who had been randomised to activated protein C
studied the effect on battle casualties in Vietnam and or placebo. Further, the PROWESS trial had a treatment
observed only one episode of clinical bleeding before 20 group mortality of 24.4 per cent, similar to the figure for
units were given.40 All of their patients developed coagu- the ENHANCE study. The ENHANCE trial serves to con-
lopathy after 30 units. The mechanism was multifactorial firm previous data and suggests that activated protein C is
and bleeding was corrected by fresh blood and platelets but a beneficial therapy in patients with septic shock. The
not fresh frozen plasma. major side effect with reference to vascular patients is that
Following massive transfusion some patients develop of excessive bleeding, and the current advice is that acti-
coagulopathy whereas others do not. The volume of blood vated protein C should not be given until 12 hours following
transfused does not always correlate with the extent of the surgery. Further, if unexpected surgery is required during
coagulopathy but it does correlate very well with the dur- treatment, the infusion should be stopped, and recom-
ation and depth of the shock period.41 It is the disease, menced 12 hours after return from the operating theatre.
namely, hypoperfusion, and not the treatment, i.e. transfu-
sion, that causes coagulopathy. Even after an exchange
transfusion, clotting factor levels remain at 30 per cent of
COMPONENT THERAPY
their original value and several studies emphasise the rela-
tive unimportance of fresh frozen plasma transfusion.42
Septicaemia remains one of the greatest challenges in The separation of blood into its components of red cells,
the postoperative patient. Patients who have undergone plasma and platelets has enabled clinicians to undertake
emergency vascular surgery are at particular risk, because specific therapy aimed at the treatment of deficiencies of
of associated gut ischaemia and also because of acute lung any one of these factors. The indications are summarised in
injury/adult respiratory distress syndrome complicating Table 7A.2.
massive transfusion. The recognition that some patients Platelet concentrates are either prepared from whole
with septic shock may have low levels of protein C resulted blood, with a volume of 50–70 mL and platelet count of
in the introduction of recombinant human activated 0.5–1.1 1011/L, or by using a continuous flow cell separ-
protein C, drotrecogin alpha. This recent advance in ator which yields a volume of 20–500 mL and a platelet
biotechnology has provided some impressive results in the content of 2.6–3.0 1011/L. Indications for platelet trans-
reduction of mortality from septic shock, and it is gaining fusion in surgical patients are either dilutional thrombo-
rapid acceptance in ICU practice. cytopenia or acquired platelet dysfunction. Dilutional
One of the most recent evaluations, under the acronym thrombocytopenia secondary to massive transfusion
ENHANCE US (Extended evaluation of recombinant should be treated with platelet transfusions only if the
human activated protein C United States Trial), compared platelet count is less than about 50 109/L in association
the effect on mortality of activated protein C with that in with active oozing from capillaries. Platelet function is
80 Perioperative care in emergency vascular practice

highly dependent on temperature and it may be necessary activated clotting and other factors, there have been no
to warm the patient to achieve adequate haemostasis. reports of adverse consequences using this method. The
Ideally, ABO and Rh specific platelet concentrates Solcotrans orthopaedic system (CR Bard, NJ, USA) uses a
should be transfused but in an emergency incompatible moderate vacuum to drain blood and is useful following
platelets may be given for the reasons discussed above in arthroplasty. Various devices such as the Sorensen system
relation to red cell transfusion. Giving group O platelet con- can be used to collect chest drain or mediastinal drainage
centrates to group A or B subjects may result in an acute blood for reinfusion. The utility of these devices depends
haemolytic reaction and rhesus sensitisation is possible via on the rate of bleeding and the haematocrit of the salvaged
red cells present in platelet concentrate. blood. If blood loss is rapid it has no time to be lysed and
During cardiopulmonary bypass and high volume sal- therefore clots. If blood loss is very slow the transfusate is
vage autotransfusion, platelet function is depressed due to mainly tissue fluid with a low red cell content.
platelet activation in the extracorporeal circuit. Platelet During the Vietnam war, Klebanoff and Watkins
transfusion may be necessary in these situations. developed a roller pump technique to reinfuse blood aspir-
There are very few indications for fresh frozen plasma ated from the surgical field.46 Unfortunately, reports of air
transfusion. In the surgical patient it is useful for emer- embolism led to its withdrawal from the market. The Cell
gency reversal of oral anticoagulants. There is no evidence Saver was a descendent of centrifugal blood separators used
to support the routine use of fresh frozen plasma in severe to produce -globulin and albumin from whole blood for
liver disease, disseminated intravascular coagulation (DIC) the treatment of burns. Originally the process was discon-
and notably massive transfusion. There is no justification tinuous but the invention of a rotary seal by Latham in 1947
for the use of plasma as a volume expander as colloids enabled continuous flow separation to be used. The operat-
are more effective, cheaper and much safer. When trans- ing principle of the Cell Saver (Haemonetics UK Ltd, Leeds,
fusion exceeds 1–1.5 blood volumes and there is clinical UK) is now well known: a centrifuge is used to separate the
non-surgical bleeding, plasma may be required. red cells while anticoagulant and plasma are eluted. The
The haemostatic system is extraordinarily resilient to technique is particularly useful in trauma (Fig. 7A.2a)
loss of clotting factors during haemorrhage. Counts et al. because the patient does not have to be heparinised. Modern
studied 27 massively transfused patients prospectively dur- equipment can be set up in minutes and the equipment does
ing resuscitation, using modified whole blood which had not require a specialised perfusionist (Fig. 7A.2b).
the platelets and cryoprecipitate removed.44 Despite high There are few studies reporting the use of cell centrifuge
volume haemorrhage and transfusion, it was unusual for devices in trauma. Goulet et al. reported a 42 per cent
measured coagulation factors to decrease to dangerous reduction in homologous blood requirement for revision
levels. Non-surgical bleeding was found to be due to hip arthroplasty and found that autotransfused patients
thrombocytopenia and also DIC. There was no justification who had sustained spinal trauma required 33 per cent less
for the routine administration of supplemental plasma in blood than historical controls.47 Cell salvage devices are
this and other studies. very useful in ruptured aortic aneurysm surgery, but, hav-
ing said that, small contained retroperitoneal ruptures may
result in a hypercoagulable state. Platelets and clotting fac-
tors have not been consumed to a great degree and the
AUTOLOGOUS BLOOD TRANSFUSION patient has activated platelets and elevated factor VIII as a
result of the stress. This may make salvage impossible as
The perceived dangers of third party blood transfusion and the machinery fills with clot. If, on the other hand, there
the cost of screening donor blood have rekindled interest has been a large bleed and platelet and clotting factors have
in blood conservation. Autologous blood can be pre- been consumed the patient becomes auto-anticoagulated
deposited, withdrawn using isovolaemic haemodilution in and salvage is particularly useful.
the anaesthetic room, salvaged during surgery with imme- A further area of concern is the use of cell washing in a
diate reinfusion or collected from wound drains. The last contaminated field. Although the process does not elim-
two techniques are relevant to trauma surgery. inate bacterial contamination, reinfusion of contaminated
Blood which collects in serosal cavities such as the chest blood has been reported without significant complica-
or peritoneum is exposed to tissue plasminogen activator, tions.48 In trauma surgery intraoperative autotransfusion
undergoes clotting and then lysis. Reinfusion of chest combined with broad spectrum antibiotic cover can be life
drainage in trauma patients with haemothoraces was first saving if no other source of blood is available. The only
reported in 1917.45 This technique is still useful if chest absolute contraindication to cell saving is faecal contam-
drains are inserted while in the field. The blood can be ination because of the danger of physical blockage of the
reinfused during transfer to the trauma centre. filters. There is no evidence that the use of intraoperative
Postoperatively, blood which has been collected from autotransfusion devices leads to acquired bleeding tenden-
wound drains may be reinfused without processing. Despite cies either due to thrombocytopenia or disseminated
theoretical worries about the consequence of infusing intravascular coagulopathy.49
 Regional block or not for vascular emergencies 81

there has been a surge of interest in the serine protease

inhibitor, aprotinin (Trasylol, Baylor, UK). Although apro-
tinin reduces blood loss following cardiopulmonary bypass,
a double blind placebo-controlled study showed it to be
ineffective during elective aortic surgery.50 A similar trial in
ruptured aortic aneurysm patients failed to show benefit51
and it is likely that aprotinin is only of benefit in hyperfibri-
nolytic states. Recombinant factor VII is at present undergo-
ing randomised clinical trials and although there are a few
anecdotal case reports from enthusiasts there is no convinc-
ing evidence to support this very expensive drug.
(a)

REGIONAL BLOCK OR NOT FOR VASCULAR

EMERGENCIES

The role of regional blocks in elective vascular surgery is

either as an adjunct to general anaesthesia or as the sole
method of anaesthesia for limb surgery or carotid endarterec-
tomy. Regional techniques are just as applicable to emergency
as well as elective situations, provided that time and normal
coagulation variables permit. Two questions have to be con-
sidered in the application of regional anaesthesia:

• Is there a benefit of regional block in addition to

general anaesthesia?
• How does abnormal coagulation influence the timing
of such blocks?

Yeager et al. studied 53 high risk patients undergoing

elective vascular surgery: they were randomised to receive
either general anaesthesia  epidural (n  28) or general
anaesthesia  postoperative intravenous opiates (n  25).52
Patients were matched for surgical risk and there were no
other major differences between the groups. The epidural
group had a lower incidence of postoperative complica-
tions, cardiac failure, major infectious complications and
lower hospital costs; all differences were significant at
P  0.05. Subsequent studies have confirmed improved
cardiovascular and pulmonary function in patients receiv-
ing epidural analgesia following vascular surgery but have
not shown an improvement in morbidity. This may be
because of an inadequate number of small studies, which
prevented closer interpretation by meta-analysis.53 However,
(b)
a large-scale trial currently in progress is addressing the
question of cardiac morbidity and mortality with epidural
Figure 7A.2 Life-saving cell salvage devices can be life saving in analgesia.
vascular emergencies (a) A motor cyclist in a road accident There are other benefits of epidural anaesthesia, either
presented with a right kidney avulsed at the renal artery and vein as a sole technique or combined with general anaesthesia.
and tearing the vena cava. (b) Six litres of blood were recycled A meta-analysis showed that patients undergoing surgery
during surgery and he made a good recovery
for hip fracture had a lower incidence of venous throm-
boembolism under regional block, than with general
PHARMACOLOGICAL AGENTS
anaesthesia alone.54 However, this benefit did not extend
to reduced mortality. In lower limb revascularisation pro-
There are no clear data to support the use of pharmacological cedures, there is a reduction in early graft thrombosis in
agents in the reduction of traumatic blood loss. However, patients receiving epidural analgesia.55,56
82 Perioperative care in emergency vascular practice

Abdominal procedures are complicated by deterior-

Key references
ation in pulmonary function in the postoperative period.
Epidural analgesia reduces this problem and improves Christopherson R, Beattie C, Frank SM, et al. Peri-operative
oxygenation.57 A large meta-analysis of randomised con- morbidity in patients randomised to epidural or general
trolled trials found that an opioid/local anaesthetic mix- anaesthesia for lower limb vascular surgery. Anaesthesiology
ture via a thoracic epidural decreased atelectasis, pulmonary 1993; 79: 422–34.
infection and hypoxaemia when compared with systemic Finfer S, Bellomo R, Boyce N et al. A comparison of albumin and
opioids. saline for fluid resuscitation in the intensive care unit.
The dilemma with the emergency vascular patient is N Engl J Med 2004; 350: 2294–6.
no longer whether an epidural be performed, but when. Gan TJ, Soppitt A, Maroof M, et al. Goal-directed intraoperative
In ruptured aortic aneurysms, those patients who are fluid administration reduces length of hospital stay after
major surgery. Anesthesiology 2002; 97: 820–6.
unstable and require volume expansion should have their
Ranaboldo CJ, Thompson JF, Davies JN, Shutt A, et al. Aprotinin in
epidural insertion delayed until after surgery has been
elective aortic reconstruction: a double blind randomised
completed. All others should have a brief explanation control trial. Br J Surg 1997; 84: 1110–13.
about the benefits and disadvantages of epidural analgesia, Treasure T, Bennett D. Reducing the risk of major elective surgery.
and if agreed, an epidural catheter should be placed before BMJ 1999; 318: 1087–8.
surgery.
The issue with those who have not had an epidural
catheter placed before surgery relates mostly to placement
and coagulopathy. This is particularly relevant in emergency
vascular patients, who may have received large volumes of
REFERENCES
fluids and heparin intraoperatively. No large-scale trial has
considered this issue and it is unlikely that such a trial would 1 Pestana D, Garcia de Lorenzo A, Madero R. Relationship between
be feasible because of the low incidence of the most feared mixed venous saturation and cardiac index, haemoglobin and
complication, namely, epidural haematoma. An audit of 912 oxygen consumption in aortic surgery. Rev Esp Anesthesiol
vascular patients with an indwelling epidural catheter, and Reanim 1998; 45: 136–40.
2 Vedrinne C, Bastien O, De Varax R, et al. Predictive factors for
who received full anticoagulation peroperatively, found no
usefulness of fibreoptic pulmonary artery catheter for
occurrence of symptomatic epidural haematomas.58 The
continuous oxygen saturation in mixed venous blood monitoring
main criticisms of this study, apart from the design and in cardiac surgery. Anaesth Analg 1997; 85: 2–10.
small sample size, are that the catheter was already sited and 3 Treasure T, Bennett D. Reducing the risk of major elective
also that asymptomatic haematomas were not excluded. A surgery. BMJ 1999; 318: 1087–8.
second study in patients who had received low molecular 4 Henein MY, Anagnostopoulos C, Das SK, et al. Left ventricular
weight heparin thromboprophylaxis prior to spinal or long axis disturbances as predictors for thallium perfusion
epidural analgesia found no cases of epidural haematoma,59 defects in patients with known peripheral vascular disease. Heart
and concluded that the practice was safe. 1988; 79: 295–300.
In emergency practice the patient will usually require an 5 Shafritz R, Ciocca RG, Gosin JS, et al. The utility of dobutamine
epidural after surgery. The evidence suggests that this is still echocardiography in preoperative evaluation for elective aortic
surgery. Am J Surg 1997; 174: 121–5.
likely to be very safe, but platelet function and coagulation
6 Ryckwaert F, Leclerq F, Colson P. Dobutamine echocardiography
studies should be normalised before the introduction of
for the pre-operative evaluation of patients for surgery of the
such a block. This may delay extubation after surgery, but it abdominal aorta. Ann Fr Anaesth Reanim 1998; 17: 13–18.
is justified by the benefits of epidural analgesia. 7 Copeland GP, Jones D, Walters M. POSSUM. A scoring system for
surgical audit. Br J Surg 1991; 78: 355–60.
8 Curran JE, Grounds RM. Ward versus intensive care management
Conclusions of high-risk surgical patients. Br J Surg 1998; 85: 956–61.
9 Berlauk JF, Abrams JH, Gilmour IJ, et al. Preoperative optimisation
Recent studies have provided new evidence that closely of cardiovascular haemodynamics in improves outcome in
monitored limited early resuscitation, early surgical peripheral vascular surgery. Ann Surg 1991; 214: 289–97.
10 Bender JS, Smith-Meek MA, Jones CE. Routine pulmonary artery
control of bleeding and careful use of blood transfusion
catheterisation does not reduce morbidity and mortality of
can improve the outlook in patients with massive bleed-
elective vascular surgery: results of a prospective randomised
ing. Preoptimising patients, if time allows, and the use trial. Ann Surg 1997; 226: 229–36.
of HDU/ICU is essential. Epidural anaesthesia and tech- 11 Ziegler DW, Wright JG, Choban PS, Flancbaum L. A prospective
niques for avoiding postoperative coagulopathy, espe- randomised trial of preoperative optimisation of cardiac function
cially warming, and the use of autologous blood, are in patients undergoing elective peripheral vascular surgery.
further refinements which should be part of routine Surgery 1997; 122: 584–92.
emergency surgical practice. 12 Dalen JE, Bone RC. Is it time to pull the pulmonary artery
catheter? [editorial] JAMA 1996; 276: 916–18.
 References 83

13 Robin ED. Death by pulmonary artery flow-directed catheter. 34 Schwab CW, Shayne JP, Turner J. Immediate trauma resuscitation
Time for a moratorium? Chest 1987; 92: 727–31. with type O uncrossmatched blood: a two year prospective
14 DiCorte CJ, Latham P, Greilich P. Pulmonary artery catheter vs. experience. J Trauma 1986; 26: 897–902.
esophageal Doppler monitor: measurement of cardiac output 35 Czer SC, Shumaker WC. Optimal haematocrit value in critically
and left ventricular filling during cardiac surgery [abstract]. ill post-operative patients. Surg Gynecol Obstet 1978; 147:
Anaesth Analg 1999; 88: SCA37. 363–8.
15 Mythen MG, Webb AR. Peri-operative plasma volume expansion 36 Carson JL, Poses RM, Spence RK, Bonavita G. Severity of anaemia
reduces the incidence of gut mucosal hypoperfusion during and operative mortality and morbidity. Lancet 1988; i: 727–9.
cardiac surgery. Arch Surg 1995; 130: 423–9. 37 Hunt TK, Rabkin J, Von Smitten K. Effects of oedema and
16 Gan TJ, Soppitt A, Maroof M, et al. Goal-directed intraoperative anaemia on wound healing and infection. Curr Stud Haematol
fluid administration reduces length of hospital stay after major Blood Transf 1986; 53: 101–11.
surgery. Anesthesiology 2002; 97: 820–6. 38 Hebert PC, Wells G, Blajchman MA, et al. A multicenter,
17 Poldermans D for the Dutch Echocardiographic Cardiac Risk randomized, controlled clinical trial of transfusion requirements
Evaluation Applying Stress Echocardiography Study Group. The in critical care. Transfusion Requirements in Critical Care
effect of bisoprolol on perioperative mortality and myocardial Investigators, Canadian Critical Care Trials Group. N Engl J Med.
infarction in high risk patients undergoing vascular surgery. N 1999; 340: 409–17.
Engl J Med 1999; 341: 1789–94. 39 Weisel RD, Dennis RC, Manny J, et al. Adverse effects of
18 Bickel WH, Bruttig SP, Millnamov GA, et al. The detrimental transfusion therapy during abdominal aortic aneurysetomy.
affects of intravenous crystalloid after aortotomy in swine. Surgery 1978; 83: 682–90.
Surgery 1991; 110: 529–36. 40 Miller RD, Robins TO, Tong MJ, Barton SL. Coagulation defects
19 Bock JC, Barker BC, Clinton AG, et al. Post-traumatic changes in, associated with massive blood transfusion. Ann Surg 1971; 174:
and effect of colloid oncotic pressure on the distribution of body 781–94.
water. Ann Surg 1989; 210: 395–405. 41 Harke H, Rahman S. Haemostatic disorders in massive
20 Astiz ME, Galera-Santiago A, Rackow EC. Intravascular volume transfusion. Bibl Haematol 1980; 46: 179–88.
and fluid therapy for severe sepsis. New Horizons 1993; 1: 42 Collins JA. Recent developments in the area of massive
127–36. transfusion. World J Surg 1987; 11: 75–81.
21 Wilkes MM, Navickis RJ. Patient survival after human albumin 43 Bernard GR, Margolis BD, Shanies HM, et al. Extended evaluation
administration. A meta-analysis of randomised, controlled trials. of recombinant human activated protein C United States Trial
Ann Intern Med 2001; 135: 205–8. (ENHANCE US): a single-arm, phase 3B, multicenter study of
22 Alderson P, Schierhout G, Roberts I, Bunn F. Colloids versus drotrecogin alfa (activated) in severe sepsis. Chest 2004; 125:
crystalloids for fluid resuscitation in critically patients. Cochrane 2206–16.
Database Syst Rev 2000; 2. 44 Counts RB, Haeisch C, Simon TL, et al. Haemostasis in massively
23 Finfer S, Bellomo R, Boyce N et al. A comparison of albumin and transfused trauma patients. Ann Surg 1979; 190: 91.
saline for fluid resuscitation in the intensive care unit. N Engl J 45 Elmendorf A. Uber wieder infusion nach punktion eines frischen
Med 2004; 350: 2294–6. hamathorax. Munch Med Wochenschr 1917; 64: 36–7.
24 Johansen K, Kohler TR, Nicholls SC, et al. Ruptured abdominal 46 Klebanoff G, Watkins D. A disposable auto-transfusion unit. Ann
aortic aneurysm: the Harborview experience. J Vasc Surg 1991; J Surg 1968; 116: 475–6.
13: 245–7. 47 Goulet JA, Bray TJ, Timoman LA, et al. Intraoperative auto-
25 Cannon WB, Fraser J, Kelwell EM. The preventative treatment of transfusion in orthopaedic patients. J Bone Joint Surg 1989; 71a:
wound shock. JAMA 1928; 70: 618–21. 3–7.
26 Blair SD, Janverin SB, McCollum CN, Greenhalgh RM. Effects 48 Timberlake GA, McSwain NE. Auto-transfusion of blood
of early blood transfusion on gastrointestinal haemorrhage. contaminated by enteric contents: a potentially life saving
Br J Surg 1986; 73: 783–5. measure in the massively haemorrhaging trauma patient. J
27 Kaweski SM, Sise MJ, Virgilio RW. The effect of pre hospital fluids Trauma 1988; 28: 855–7.
on survival in trauma patients. J Trauma 1990; 30: 1215–19. 49 Thompson JF. Intra-operative auto-transfusion. Curr Pract Surg
28 Bickell WH, Shafton GW, Mattox KL. Intravenous fluid 1993; 5: 137–41.
administration and uncontrolled haemorrhage. J Trauma 1989; 50 Ranaboldo CJ, Thompson JF, Davies JN, Shutt A, et al. Aprotinin
29: 409. in elective aortic reconstruction: a double blind randomised
29 Assalia A, Schein M. Resuscitation for haemorrhagic shock. control trial. Br J Surg 1997; 84: 1110–13.
Br J Surg 1993; 80: 213. 51 Robinson J, Nawaz S, Beard J. Randomised multi centre
30 Shippey CR, Appel PL, Shumaker WC. Reliability of clinical double blind placebo controlled trial of the use of aprotinin
monitoring to assess blood volume in critically ill patients. in ruptured abdominal aortic aneurysm. Br J Surg 2000; 87:
Crit Care Med 1984; 12: 107–12. 754–7.
31 Simmons RL, Heisterkamp CA, Mosely RV, Doty DB. Post 52 Yeager MP, Glass DD, Neff RK, Brinck-Johnsen T. Epidural
resuscitative blood volumes in combat casualties. Surg Gynecol anaesthesia and analgesia in high-risk surgical patients.
Obstet 1969; 128: 1193–201. Anaesthesiology 1987; 66: 729–36.
32 Davies MJ, Murphy WG, Murie JA, et al. Pre-operative 53 Buggy DJ, Smith G. Epidural anaesthesia and analgesia: better
coagulopathy in ruptured abdominal aortic aneurysm predicts outcome after major surgery? BMJ 1999; 319: 530–1.
poor outcome. Br J Surg 1993; 80: 974–6. 54 Sorenson RM, Pace NL. Anaesthetic techniques during surgical
33 Barnes A. Transfusion of universal donor and uncrossmatched repair of femoral neck fractures. A meta-analysis.
blood. Bibl Haematol 1980; 46: 132–42. Anaesthesiology 1992; 77: 1095–104.
84 Perioperative care in emergency vascular practice

55 Tuman KJ, McCarthy RJ, March RJ, et al. Effects of epidural at rest and during mobilization after major abdominal surgery.
anaesthesia and analgesia on coagulation and outcome after Anaesth Analg 1992; 74: 362–5.
major vascular surgery. Anaesth Analg 1991; 73: 696–704. 58 Baron HC, LaRaja RD, Rossi G, Atkinson D. Continuous epidural
56 Christopherson R, Beattie C, Frank SM, et al. Peri-operative anaesthesia in the heparinized vascular surgical patient: a
morbidity in patients randomised to epidural or general retrospective review of 912 patients. J Vasc Surg 1987; 6:
anaesthesia for lower limb vascular surgery. Anaesthesiology 144–6.
1993; 79: 422–34. 59 Bergqvist D, Lindblad B, Matzsch T. Low molecular weight
57 Dahl JB, Rosenberg J, Hansen B, et al. Differential analgesic heparin for thromboprophylaxis and epidural/spinal anaesthesia –
effects of low-dose epidural morphine and morphine-bupivacaine is there a risk? Acta Anaesthesiol Scand 1992; 36: 605–9.
 7B
Emergency Vascular Access

THANG D NGUYEN, ROY M FUJITANI, SAMUEL E WILSON

The problem 85 Femoral vein 89

The haemodialysis patient 85 The internal jugular vein 90
Indications for haemodialysis 86 Subclavian line placement 92
Acute access 88 References 93
Seldinger technique 88

THE PROBLEM required to preserve potential access sites. A poorly planned

‘quick fix’ in these patients can ruin an entire limb for future
dialysis.
Intravenous (IV) catheters have become essential tools in In this chapter we will discuss the vascular access
the management of hospital patients. Over the years, options available during emergencies. Selected procedures,
developments in catheters and insertion techniques have particularly central line placements along with the inser-
allowed reliable and rapid access to the central venous sys- tion techniques and the management of their complica-
tem. Techniques have become simplified to the point that tions will be discussed. Emergency vascular access for the
bedside insertions are now commonplace. Indications for purpose of haemodialysis requires specific evaluation and
central line access have evolved to encompass parental this subject deserves special consideration.
nutrition, administration of irritating or caustic solutions,
haemodynamic monitoring and temporary haemodialysis.
The ability to attain rapid central access without the need
for operating suites makes these types of access useful tools
THE HAEMODIALYSIS PATIENT
during medical and surgical emergencies. In spite of the
simplified techniques and the high frequency with which
we use them, complications ranging from minor to fatal Chronic renal insufficiency: patients
can occur.1–9 Respect for central access and understanding anticipating haemodialysis
its limitations remain essential when seeking central access.
The subject of emergency vascular access encompasses a The chronic renal failure patient who anticipates haemodial-
wide array of clinical settings. During a critical injury where ysis is usually referred by nephrologists and is frequently
the priority of IV access ranks second only to securing air- encountered in outpatient settings. These patients are
way, two large-bore peripheral IVs are often sufficient. In medically stable and do not commonly require immediate
the setting of a hypotensive patient in vascular collapse access. They will, however, require a permanent arteriove-
peripheral IVs can be challenging. These patients require nous (AV) fistula. The preferred site is the non-dominant
two large calibre central catheters, one on either side of the distal upper extremity, typically the radial artery to the
diaphragm to facilitate resuscitation. Similarly, patients who nearest suitable vein. Whenever possible, an autologous
require rapid fluid resuscitation may require multilumen fistula of the Brescia–Cimino type is preferred over the
central catheters to accommodate multiple IV medications. synthetic polytetrafluoroethylene (PTFE) graft because of
In these circumstances the central venous catheter may be its superior primary patency and lower revision rate10
inserted in any available site. Haemodialysis patients, on the (see Chapter 41). Since native vein shunts typically require
other hand, require additional considerations. When these a 2–4-week maturation period, autologous fistulae fre-
patients present for emergency access, strategic planning is quently require the simultaneous placement of an indwelling
86 Emergency vascular access

haemodialysis catheter in the contralateral internal jugular to evaluate outflow obstruction. In the presence of proximal
vein. This allows the patient to undergo haemodialysis stenosis a distal AV graft will be at risk for early failure. Some
while the fistula matures. The catheter can be removed vascular surgeons advocate a duplex ultrasound study
after the functional status of the Cimino fistula is established. whenever an AV graft in the distal extremity thromboses.
In the event that the native vein fistula is not possible due Pertinent laboratory studies include a basic metabolic
to a small calibre or tortuous radial vein, a PTFE straight panel and a complete blood count. Marked elevation in
graft from the radial artery to the antecubital vein should blood urea nitrogen can result in uraemic encephalopathy
be considered. Given that PTFE grafts do not customarily as well as platelet dysfunction. The latter may be an import-
require a maturation period, the simultaneous placement ant consideration when contemplating surgery or tempor-
of a temporary catheter is not necessary. ary catheter placement. Marked fluid overload or severe
hyperkalaemia requires emergency haemodialysis. This
may be necessary prior to the availability of a functional
Acute renal failure and the thrombosed graft. In these circumstances, the placement of a temporary
arteriovenous graft haemodialysis catheter is appropriate. The site for perman-
ent access should be assessed prior to percutaneous can-
End stage renal patients require lifelong haemodialysis. nulation. When using the internal jugular vein for
Unfortunately, AV grafts do not last a lifetime. There are a temporary access, the catheter should be inserted con-
number of modalities available to address failed AV grafts, tralateral to the site of the planned permanent access.13
but none can be considered superior. Interventional radi- In general, acute vascular access can be accommodated
ology offers a variety of percutaneous techniques which in one of two ways: a double lumen central venous catheter
achieve satisfactory thrombolysis and thrombectomy under and a bridge graft AV shunt. Whereas the former is appro-
fluoroscopic guidance. In appropriate circumstances, inter- priate for haemodialysis commencing immediately, the
ventional radiologists can perform percutaneous translumi- latter can be available for dialysis commencing within 24
nal (balloon) angioplasty of compromised outflow or inflow hours. If dialysis is deemed an emergency, and will be
vessels. Several series report promising results which rival needed for two or more weeks, a soft tunnelled silastic
surgical outcome.11,12 Secondary patency rates, however, catheter with a Dacron cuff should be used. A soft catheter
remain similar. For many patients with acutely thrombosed carries a lower incidence of vein stenosis and the Dacron
AV grafts, this modality offers a minimally invasive alterna- cuff decreases the likelihood of line sepsis by serving as a
tive to surgical revisions. Grafts with multiple revisions or barrier against bacterial migration. Whereas bridge graft
extensive intimal hyperplasia require surgical intervention. AV shunts must be placed in the operating room (OR), the
When these patients present for surgical repair it is percutaneous and tunnelled catheters can be inserted in
important to determine the relative urgency of the matter. the OR or in the interventional radiology suite. Only the
In the presence of fluid overload, uraemic encephalopathy percutaneous catheter can be inserted at the bedside.
or hyperkalaemia, the need for AV access should be con-
sidered as an emergency. The proper evaluation of patients
in this condition mandates a thorough history and physical
examination. Important historical elements include a his-
INDICATIONS FOR HAEMODIALYSIS
tory of previous central catheter placements, hypercoagula-
ble states, immune compromise conditions and the date With the exception of complete renal failure, no two
of the last dialysis. A thrombosed graft identified 1 day after patients have the same degree of renal insufficiency.
effective dialysis would not require emergency repair Furthermore, the susceptibility to renal failure complica-
whereas one identified 3 days later may do so. Physical tions varies from patient to patient. One factor which can
examination must include the patient’s vital signs, weight, be fatal in all renal failure patients, however, is hyper-
mental status and neurological status. If a new graft involving kalaemia. The elderly patient with pre-existing heart
the radial artery is contemplated, an evaluation of the ipsilat- disease may be more susceptible to cardiac arrhythmias
eral ulnar artery supply is warranted. A weak or absent ulnar from hyperkalaemia than one without a cardiac history.
artery can be demonstrated by Allen’s test (see Chapter 41). Additionally, a patient with chronic renal failure may tol-
A graft involving the radial artery in the presence of a weak erate a higher degree of hyperkalaemia than one in acute
or absent ulnar pulse may result in a steal phenomenon renal failure (ARF). Consequently, there is no consensus
causing ischaemia in the ipsilateral hand. Fever and leuco- on the laboratory values which ought to trigger emergency
cytosis are signs of infection precluding the placement of a haemodialysis. Most physicians will agree that symptom-
PTFE graft. A thrombosed AV graft site should be examined atic uraemia or abnormal potassium levels along with elec-
for evidence of infection necessitating graft removal. The trocardiographic (ECG) changes warrant urgent dialysis.
ipsilateral extremity should be examined for evidence of Frequently, medical management of hyperkalaemia must be
venous hypertension, the presence of which warrants a initiated while awaiting haemodialysis. This is particularly
duplex ultrasound of the axillary–subclavian venous system important in the hyperkalaemic patient with metabolic
 Indications for haemodialysis 87

acidosis. This combination is potentially fatal and necessi- intubated patient who cannot compensate for metabolic
tates immediate intervention. acidosis by respiratory efforts. The threshold for arrhyth-
When failing kidneys lose the ability to excrete bodily mia arising from hyperkalaemia depends on the patient
acids, blood pH must be maintained by other compensa- and can occur abruptly without uraemic symptoms. Thus,
tory means. One mechanism involves the uptake of excess dialysis access in these patients may be more urgent than
extracellular hydrogen in exchange for intracellular potas- symptoms would suggest. The priority of management in
sium; the result is homoeostatic pH at the expense of worsen- these patients must be directed toward protecting the
ing hyperkalaemia. This can be further exacerbated in an heart. An ECG must be obtained to evaluate hyper-
kalaemia induced changes. Table 7B.1 lists the ECG signs
of hyperkalaemia. Calcium infusion must be administered
Table 7B.1 Electrocardiographic (ECG) changes in to stabilise the myocardium. Serum potassium can be
hyperkalaemia reduced by the following manoeuvres: intravenous admin-
istration of glucose and insulin, intravenous bicarbonate
ECG changes Potassium levels
and enteral Kayexalate. The use of loop diuretics, such as
T waves tenting 5.7–6.9 mEq/L furosemide, can often be effective, but may not be useful in
the setting of renal failure. These therapeutic options are
P wave amplitude decreases, 7.0–8.3 mEq/L
widen PR interval
initiated in addition to establishing vascular access in
preparation for haemodialysis.
P wave flat 8.4–8.9 mEq/L
Table 7B.2 lists signs and symptoms of uraemia. Table
QRS widening 9.0–11 mEq/L 7B.3 lists absolute and relative indications for haemodialy-
Ventricular fibrillation 12 mEq/L sis. In general, chronic dialysis therapy is indicated when
glomerular filtration rate falls below10 mL/min. More
acute indications include fluid overload, congestive heart
failure, hyperkalaemia, metabolic acidosis, hypertension
Table 7B.2 Signs and symptoms of uraemia
uncontrolled by conservative measures and uraemia induced
Signs Symptoms conditions such as encephalopathy, neuropathy, pericard-
itis and bleeding diathesis. Occasionally a fluid-restricted
Pericardial friction rub Nausea/vomiting anuric patient requires blood transfusions or parenteral
Refractory pulmonary oedema Anorexia feeds. Haemodialysis may be the only means of removing
excess fluids in these patients. Patients with ARF who may
Metabolic acidosis Fatigue
not require haemodialysis are those developing ARF from
Foot/wrist drop Diminished sensorium
reversible conditions. Such conditions include dehydra-
Asterixis tion, urinary tract infection, urinary obstruction, hyper-
catabolic states, hypercalcaemia and low cardiac output
states. The decision to initiate haemodialysis therapy
requires an evaluation of the patient’s clinical condition as
Table 7B.3 Indications for haemodialysis
well as of serum potassium levels. The urgency with which
Relative indications Absolute indications dialysis is required will determine whether temporary or
permanent access is more appropriate.
Blood urea nitrogen >100 mg/dL Volume overload
Total parental nutrition or blood Hyperkalaemia Types of catheter
transfusions in acute renal failure
Uraemic coagulopathy Metabolic acidosis Table 7B.4 lists the various types of central venous catheter
Drug intoxication Uraemia and their sizes. There are a multitudes of catheters avail-
able and not all are appropriate for haemodialysis.

Table 7B.4 Types of catheters

Type Indications Insertion site Size

Cordis, triple lumen Acute blood loss, dehydration Femoral, internal jugular, subclavian 9 Fr cordis/7 Fr triple lumen
HD Catheter Haemodialysis Femoral 12–13 Fr/16–19 cm
HD Catheter Haemodialysis Internal jugular vein 12–13 Fr/13–16 cm
HD Catheter Plasmaphoresis Femoral 12–13 Fr/16–19 cm
88 Emergency vascular access

Renal failure patients for

acute dialysis access

Uraemic or fluid overload Serum K 6.0 mEq/L

Administer Kayexalate,
6.0 mEq/L K 5.0 mEq/L
insulin/glucose, or bicarbonate

Figure 7B.1 Algorithm for acute

vascular access for haemodialysis. Immediate dialysis
Temporary catheters are appropriate Serum K 6.0 Serum K 5.0
or in less than 24 hours
for haemodialysis commencing
immediately. A bridge graft
Dialysis in
arteriovenous (AV) shunt should 24 hours
be placed if dialysis commences in
24 hours. If the immediate need for
haemodialysis is anticipated to exceed Anticipated
Anticipated duration of duration of
two weeks, a soft tunnelled silastic dialysis
dialysis
catheter with a Dacron cuff should be
used. In asymptomatic patients, serum
potassium (K) levels of 6 mEq/L or
greater requires immediate medical Short term Long term
intervention in the form of Kayexalate,
insulin/glucose infusion or bicarbonate.
Potassium levels between 5 and 6 mEq/L Less than two More than two
AV graft
may be treated medically. More than weeks weeks
one dose may be necessary. Refractive
hyperkalaemia (6 mEq/L) necessitates Percutaneous Silastic
immediate dialysis. IR, interventional temporary tunnelled
catheter catheter by IR
radiography

When contemplating a temporary catheter for haemodialy- exceedingly difficult and even dangerous. In recent years,
sis, both diameter and length must be considered. The unit portable real time ultrasound has made the identification of
known as French (Fr) describes the catheter diameter, and suitable targets easier and increased the rate of successful
three Fr units are equivalent to 1 mm. The length of central cannulation. Not surprisingly, central line placement under
venous catheters is measured in centimetres. Catheters ultrasound guidance is associated with fewer complica-
specifically designed for haemodialysis are typically 12 or 13 tions.14–19 When central access is contemplated for patients
Fr (4 mm). Whereas the femoral vein can accommodate a with previous multiple central lines, real time ultrasound is
long catheter, the internal jugular catheter length is limited recommended.
by the position of the heart. In the femoral vein a 16–19 cm
catheter is required whereas in the internal jugular vein a
13–16 cm catheter would be appropriate for most patients. SELDINGER TECHNIQUE
On an upright chest X-ray the internal jugular catheter tip
situated at the junction of the superior vena cava (SVC) and
In principle, the Seldinger technique provides a reliable
the right atrium would be considered ideal.
method for locating and percutaneously cannulating a ves-
Figure 7B.1 provides an algorithm for acute vascular
sel. This method makes use of a soft guidewire threaded
access for haemodialysis.
through an 18 gauge needle placed in a vein. The wire then
serves as a guide over which a softer catheter can be
inserted. As with any procedure, proper preparation helps
ACUTE ACCESS to minimise complications. The site of insertion must be
prepped widely with povidone iodine. The patient as well
At the bedside, central access is attained percutaneously. This as the surrounding work area is draped with sterile towels.
method relies on anatomical landmarks to identify suitable The physician is capped, masked, gowned and gloved in a
veins. Occasionally, anatomical variation or pre-existing sterile fashion. All components of the catheter must be
thrombosis makes the ‘blind’ insertion of central catheters confined to a sterile field and readily accessible.
 Femoral vein 89

The skin and subcutaneous tissue at the insertion site

are anaesthetised with 3–5 mL of 1 per cent lidocaine
(Xylocaine) using a 25 gauge needle. A thin walled 18 Inguinal ligament
gauge needle is introduced into the vein guided by land-
marks and/or real time ultrasound. The needle enters at a
45 degree angle with the skin, aspirating as the needle
Femoral vein
advances. The return of dark non-pulsatile blood indicates
insertion site
venepuncture. The syringe is then removed with the needle
still in place. The hub of the needle should be covered at this
point to avoid introducing air into the vascular system. A
sterile guidewire is passed into the vein through this needle
Femoral
which is then removed over the wire. A number 11 scalpel nerve, artery
blade extends the insertion site by 2–3 mm at the skin. A and vein
dilator is passed over the wire through the cutaneous tissue
into the vein and subsequently removed. A catheter is then
placed over the wire into the vein. The wire is withdrawn
with the catheter still in place. The functions of all ports are
Adductor
evaluated by withdrawing blood from each of them, fol- Sartorius
longus
lowed by a sterile heparinised saline flush. The catheter is
ultimately sutured in place and dressed in sterile fashion.

FEMORAL VEIN Figure 7B.2 The femoral triangle is defined by the inguinal
ligament superiorly, the sartorius muscle laterally and the
adductor longus medially. The femoral nerve, artery and vein
The femoral vein lies medial to the usually palpable
course through this triangle. The femoral vein lies medial to the
femoral artery located in the femoral triangle (Fig. 7B.2).
artery. The palpable artery in this triangle helps to identify the vein.
The insertion site is approximately 1 cm medial to the The arrow indicates site of catheter insertion
femoral artery and 3–4 cm below the inguinal ligament.
The patient is positioned supine with the lower extremity
slightly externally rotated. The Seldinger technique is then Disadvantages of the femoral line
used to complete the insertion. Resistance at the femoral
site suggests either that the wire has failed to enter the vein The femoral site is in close proximity to the rectum and
or that the patient is not supine. In the former case, the genitalia. Bacteria migration facilitated by bodily fluids from
wire must be withdrawn and the femoral vein is sought this region increases the likelihood of line infections. In
again with a needle and syringe, whereas in the latter the non-emergency settings, femoral access may interfere with
patient is repositioned lying supine but avoiding flexion at mobility and is itself thrombogenic. Nonetheless, femoral
the hip so that the wire does not have difficulty negotiating lines are used in ambulatory dialysis settings and have been
a bend. Ascites can sometimes cause resistance at the reported safe and effective for as long as 14 days.14 In the set-
femoral site. In no circumstances should the wire be ting of trauma where hip fracture is a possibility the femoral
forced. An alternative site should be considered if resist- site may not be the best choice. Local oedema and tender-
ance persists despite the above manoeuvres. ness interfere with proper positioning and exposure. If the
line is placed for the purpose of haemodialysis, the femoral
site is not always practical. The sitting position produces a
Advantages of the femoral site
bend in the femoral catheter thereby reducing flow during
dialysis. The patient is thus required to remain supine for
The femoral veins are relatively far from the heart and lungs.
the duration of the haemodialysis session lasting 2–4 hours.
This makes the femoral site ideal when a central line is
Furthermore, the femoral IV can be particularly cumber-
required during cardiopulmonary resuscitation (CPR). From
some in the OR because anesthesiologists, positioned at
a technical viewpoint, the femoral vein is easily accessible and
the head of bed, may find it difficult to gain access to the
its landmarks are readily apparent. Alternative sites such as
catheter.
the internal jugular or subclavian vein increases the risk of
pulmonary injury. Thus, patients with poor ventilation, pul-
monary oedema or congestive heart failure may be better Complications associated with femoral access
served with femoral line access.1,2 The femoral catheter
does not require a postprocedural X-ray to verify position. While the placing of femoral lines is technically simple, and
A successfully placed femoral line is therefore immediately provides a rapid means of establishing vascular access, it is
functional and extremely useful during emergencies. not without complications. Femoral pseudoaneurysms
90 Emergency vascular access

can result from attempted femoral vein catheterisation and THE INTERNAL JUGULAR VEIN
is attributed to inadvertent artery puncture, the risk being
higher in the anticoagulated patient (see Chapter 38). The
patient commonly presents with local pain with an associ- Catheterising the internal jugular vein is also achieved by
ated palpable mass at the site of puncture. Auscultation the Seldinger technique. When possible, the patient should
over the mass may reveal a bruit. Ultrasound shows be adequately sedated, and placed under external cardiac
extravasated blood contained within the surrounding tis- monitoring. The patient is place in a 15–20 degree Trende-
sues. Direct pressure usually results in total resolution but lenburg position with the head turned away from the inser-
can be painful and may take up to 37 minutes to tion side. A towel roll placed horizontally behind the
resolve.20,21 In a small percentage of cases, compression shoulders extends the neck and accentuates landmarks. The
therapy fails, requiring ultrasound guided thrombin injec- internal jugular can be accessed either by a middle or poster-
tion. Obliteration of these false aneurysms with thrombin ior approach2 (Figs 7B.3 and 7B.4). The safest method to
injections is reportedly immediate and pain free. Failure of locate the IJ vein is by way of a 25 gauge seeker needle. Once
thrombin therapy may be related to the size of the found, an 18 gauge needle can be introduced into the vein
pseudoaneurysm. The principal concern is the rupture of following a course parallel to that of the seeker needle.
the pseudoaneurysm resulting in uncontained blood loss.13
Surgical repair may be necessary.
Another femoral catheter complication is the formation
of an AV fistula (see Chapter 38). Diversion of blood from
artery to vein has the potential of causing ipsilateral venous
hypertension, distal arterial insufficiency and high output
heart failure.13 A possible mechanism involves inadvertent
arterial puncture during attempts at central access. The
subsequent venepuncture on the ipsilateral side sets up an
AV communication. Duplex ultrasound is often diagnostic
but an angiogram may be warranted in the presence of
arterial insufficiency and inconclusive ultrasound. Large
symptomatic fistulae usually require surgical repair.
Another reported complication of the femoral catheter
is the entanglement of the guidewire with an existing infer-
ior vena cava (IVC) filter22 (see Chapter 20). Careful atten-
tion to the history helps avoid this particular complication. Sternocleido- Carotid
Guidewire entanglement with an IVC filter has also been mastoid artery
reported during catheter placement in the right internal jugu- (SCM)
lar vein6 and subclavian vein.22–26 The interventional radi- Internal
External jugular
ologist can, under fluoroscopic guidance, facilitate removal jugular vein
of the wire. In the presence of an IVC filter, it is probably vein
best to perform central access under fluoroscopic guidance. Sternal
head of
SCM

Potential complications at sites used

in venous access

• Femoral Subclavian
vein
– Pseudoaneurysm Clavicular head
– AV fistula of SCM
– Guidewire entanglement in IVC filter
• Internal jugular SVC
– Pneumothorax
Figure 7B.3 Muscles and vasculature of the neck. The vessel
– Arrhythmias
coloured red represents the carotid artery whereas those coloured
– Thrombosis
blue represent the internal jugular vein, the subclavian vein and
– Guidewire migration the superior vena cava (SVC). The clavicular and sternal heads
• Subclavian of the sternocleidomastoid (SCM) join superior to the clavicle
– Stenosis and serve as a landmark in locating the internal jugular vein.
– Venous hypertension and thrombosis The clavicle and suprasternal notch are landmarks for the
subclavian approach
 The internal jugular vein 91

The middle approach syringe is removed with the needle remaining in place. An
18 gauge thin wall needle then follows a parallel course
The internal jugular lies deep to the sternocleidomastoid entering the skin at a 30–45 degree angle. When dark non-
(SCM), approximately 1 cm posterolateral to the palpable pulsatile blood returns, the Seldinger technique is used to
internal carotid artery, coursing most superficially between complete catheter insertion.
the clavicular and sternal heads of the SCM2 (see Figs 7B.3
and 7B.4). The middle approach cannulates the internal The posterior approach
jugular at this site. With one finger on the carotid pulse
and a 25 gauge needle attached to a 5 ml syringe in the The posterior approach enters the internal jugular from the
opposite hand, the internal jugular vein is sought by pene- posterior edge of the SCM superior to the point where it
trating the skin lateral to the pulse in the direction of the meets the external jugular vein. A seeker needle is directed
ipsilateral nipple, aspirating as the needle advances. If redir- deep to the SCM and towards the suprasternal notch. Once
ection is required, the needle should be withdrawn and located, an introducer needle follows the course of the seeker
reinserted in a more lateral direction. Once found, the needle in parallel fashion. The internal jugular vein can then
be cannulated as described by the Seldinger technique.
An upright chest X-ray is essential to verify catheter tip
position and to rule out pulmonary injury. The preferred
tip position is in the SVC approximately 1–2 cm above the
SVC–atrium junction. Most catheters have depth markers
that help guide proper placement. In a 70 kg patient, the
catheter should be inserted to the 17 cm mark from the left
internal jugular vein. When cannulating the right internal
jugular vein, the catheter should be inserted to the 14 cm
mark. Soft catheters can be placed as far as the right atrium.
Stiff catheters such as those used for temporary haemo-
dialysis will irritate the myocardium causing arrhythmia.1

Advantages of the internal jugular catheter

In the vast majority of cases the internal jugular site is read-

ily accessible. The landmarks are readily apparent, and the
carotid artery can be easily compressed should it be nicked
Posterior
approach during the search for the vein. Central catheters in this
Middle
approach position are well tolerated. Mobility of the neck is min-
imally affected, and the patient may walk without affecting
the function or position of the catheter. Being relatively
remote from the rectum and genitalia keeping this site
clean is less problematic than it is for femoral catheters,
thus allowing the line to be kept in place for a relatively
longer period of time. The right internal jugular leads
directly to the SVC without much curvature so that a rigid
catheter can be placed easily without provoking the stenos-
ing effect associated with subclavian central lines.1–3,13,27,28
Subclavian approach

Figure 7B.4 Central line insertion sites at the neck. The arrows Disadvantages of internal jugular catheters
indicate sites of insertion. Two widely used sites for the internal
jugular vein are illustrated, namely, the middle and the posterior Insertion of a large bore needle into the anterior triangle of
approach. The middle approach uses the most superficial segment the neck in search of a carotid sheath structure carries inher-
of the internal jugular vein, which lies at the junction of the two
ent risks. Local injuries ranging from haematomas, infec-
heads of the sternocleidomastoid (SCM). The posterior method
approaches the internal jugular from behind the lateral edge of the
tions and nerve injuries to ipsilateral lung collapse are all
SCM just superior to its junction with the external jugular vein. documented complications. In addition, there are a variety
Whereas risks of arterial puncture are higher from the posterior of conditions that can make the internal jugular vein virtu-
approach, the middle approach is associated with a higher ally inaccessible. In the trauma setting, a hard cervical collar
incidence of pneumothorax. Also illustrated is the subclavian site. often precludes the placement of a central line at this site. In
The risk of pneumothorax is greatest with this approach an obese patient with a short neck the anatomical landmarks
92 Emergency vascular access

may be distorted making identification of the internal jugu- advocate continuous cardiac monitoring during all internal
lar vein difficult. A patient in pulmonary oedema or conges- jugular and subclavian central access procedures.
tive heart failure would not tolerate the Trendelenburg Central vein thrombosis and thrombosis within the
position required for the cannulation of vessels in the neck. atrium have been described in association with central
Cardiopulmonary resuscitation in progress, particularly venous catheters.1 Although not often symptomatic, central
chest compressions, makes it virtually impossible to place vein thrombosis presents with swelling of the ipsilateral
a line in the internal jugular vein. Compared with other extremity with associated pain and tenderness. Ultrasound
central sites, the internal jugular carries the highest rate of venography is diagnostic and the treatment is catheter
anatomical variability resulting in difficult venous access. In removal and anticoagulation. A thrombus in the right
a patient without previous central lines, anatomical variabil- atrium is rare but life threatening. Removal of the catheter
ity can be as high as 5 per cent.1 In a patient who has not pre- under real time sonography with lytic therapy may be neces-
viously had a central line inserted thrombosis or variability sary to avoid large scale pulmonary embolism.1
ranges from 16 to 27 per cent.1,8,9,14,15,17 Whereas real time A less frequent but real complication of the internal
ultrasound can guide the operator toward a suitable vein in jugular line is losing control of the wire and allowing it to
41 per cent of these cases, in the remaining cases the veins be dislodged entirely into the vascular system.29 Retrieval
were found to be either too small for cannulation, throm- requires the help of interventional radiology using real time
bosed or nonexistent.9 When confronted with a difficult ultrasound guidance. Migration of the ‘out of control’ wire
access it is important to keep in mind the fact that the risk of into the heart can result in fatal arrhythmias. The inability
pneumothorax increases with the number of attempts. An to remove the inciting factor quickly in this situation is
ultrasound guided technique may therefore be helpful. potentially life threatening. The solution is prevention and
therefore the wire must be kept under control at all times.

Complications of internal jugular catheters

SUBCLAVIAN LINE PLACEMENT
The lung apex is in the direct path of the introducer needle
as it penetrates the internal jugular from either side and
therefore pneumothorax represents an inherent risk of The union of the cephalic and axillary veins marks the
internal jugular vein catheterisation. The rate of this com- beginning of the sublcavian vein. It enters the thorax, cours-
plication in association with internal jugular insertions is ing deep to the clavicle and superficial to the anterolateral
within the range of 0 to 1.8 per cent,1 occurring much more aspect of the first rib. The subclavian artery lies deep to the
frequently with the middle approach.2 Pneumothorax may vein. The anterior scalene muscle separates the two vessels,
be recognised during the procedure by the aspiration of air, and the lung apex lies medial and posterior to the sub-
respiratory distress and decreased ipsilateral breath sounds. clavian vein. Landmarks for subclavian vein access include
Contralateral tracheal shift and hypotension are signs of a the clavicle, acromioclavicular joint and suprasternal notch2
tension pneumothorax requiring immediate decompres- (see Figs 7B.3 and 7B.4).
sion. An upright chest X-ray confirms the diagnosis, and in The patient is placed in a 20–30 degree Trendelenburg
most patients, particularly those on positive pressure venti- position with the head turned away from the side of inser-
lation, air has to be evacuated via thoracostomy tubes. tion. A towel roll is placed vertically between the scapulae
The risk of pneumothorax can be minimised by attention to accentuate landmarks. The insertion site is widely
to a few technical considerations. First, the introducer nee- prepped with antiseptic solution and the surrounding
dle should enter the skin at a 30–45 degree angle to the neck, work area is draped in a sterile fashion. As with any central
particularly when using the middle approach. A lower angle access procedure, the physician is capped, gowned and
places the needle in the direct path of the lung apex. Second, gloved in sterile fashion.
the incidence of pneumothorax rises with the number of The insertion site is 1–2 cm below the inferior margin
attempts. Given the high anatomical variability and abnor- of the clavicle at the junction of its distal and medial thirds.
mality in patients, repeated blind attempts are never recom- The skin, subcutaneous tissue and the clavicular perios-
mended. Real time sonographic guidance can be a useful teum are anaesthetised with 1 per cent lidocaine using a 25
adjunct when landmarks fail. gauge needle. With the index finger on the sternal notch
Cardiac problems are not infrequently encountered dur- and the thumb on the clavicle, an 18 gauge introducer nee-
ing attempts at central access on patients in the intensive dle enters the anaesthetised skin with the bevel cephalad,
care unit. Continuous cardiac monitoring reveals premature aiming at the sternal notch and intentionally hitting the
ventricular contractions which occasionally lead to ventricu- clavicle. When the clavicle is encountered, march down the
lar tachycardia. The cause is often myocardial irritation clavicle until the needle is flush with the inferior surface of
caused by the guidewire. When this occurs, the wire is par- the clavicle. The needle is advanced slowly with the needle
tially withdrawn until the arrhythmia subsides. Sometimes, ‘hugging’ the inferior margin of the clavicle and aspirating
infusion of 1 mg/kg of injectable lidocaine is necessary to simultaneously. When dark non-pulsatile blood returns,
abort this arrhythmia. Consequently, some institutions the needle is rotated 180 degrees so that the needle bevel
 References 93

now faces caudal to facilitate guidewire insertion. The of vascular access. Stenosis at this central site can comprom-
Seldinger method is used to complete the cannulation. ise outflow from the ipsilateral arm. Consequently, AV
shunts distal to the lesion ultimately fail secondary to
venous hypertension. This observation led to a strategic
Advantages of subclavian catheters approach in the management of dialysis patients who
require temporary catheters. Clearly, the first choice for a
Although potentially dangerous and clearly requiring skill,
temporary catheter site should be the right internal jugular
many physicians favour the subclavian site because it offers
vein. Equally clear is the site of last resort, namely, the sub-
constant anatomy. The clavicle serves as a landmark regu-
clavian vein. Choices falling in between are the left internal
larly directing the physician to the subclavian vein. This
jugular and the femoral site. Whereas the left internal jugu-
landmark is usually palpable even in obese patients, regard-
lar has a tortuous path to the SVC, the femoral veins are
less of the length of the neck. In the trauma setting where
associated with higher risks of line infection.
cervical collars preclude internal jugular placement, sub-
clavian catheterisation remains a viable option. Once in place,
the catheter can be secured below the clavicle, hidden away Conclusions
from visible areas. The mobility of the patient is entirely
unaffected, and this site can be kept clean with relative ease. The need for emergency vascular access is encountered in
a wide array of clinical settings. In recent years, temporary
Disadvantages of subclavian catheters venous catheters have become indispensable tools within
hospital. The ease of insertion has made these devices use-
Along with the many advantages of subclavian catheters, ful during resuscitation efforts in cases of trauma, dehy-
there are definite pitfalls. Like internal jugular lines, sub- dration, haemorrhage and ‘code blue’ situations. For
clavian catheters are difficult if not impossible to place during patients requiring emergency haemodialysis, these
chest compressions. Like the internal jugular site, subclavian catheters have become life-saving temporising measures.
line placements are most successful when patient is in the In spite of the many short term advantages, immediate
Trendelenburg position. Patients unable to tolerate the flat and life-threatening complications such as haemopneu-
or head-down position may not be suitable candidates for mothorax, haemorrhage, infection, arterial laceration, as
subclavian central lines. Also as with internal jugular well as the long term consequences of thrombosis and
catheterisation, pneumothorax is an associated risk. The risk stenosis are associated with temporary catheters. These
of pneumothorax is higher during subclavian line placement complications can occur even in the hands of experienced
(2–5 per cent)2 compared with that that of internal jugular physicians. Respect for the potential complications and
catheterisation. In addition, the clavicle is an effective rigid an understanding of the physician’s own limits are essen-
barrier over the subclavian vessels making direct compres- tial when seeking central access.
sion impossible should the subclavian artery be punctured
inadvertently.
Key references
Complications of subclavian catheters
Agee R, Kim, Balk AR. Central venous catheterization in the
critically ill patient. Crit Care Clin 1992; 8: 677–86.
Subclavian and internal jugular access procedures share
Bambauer R, Inniger R, Pirrung KJ, et al. Complications and side
many common complications but those unique to the sub- effects associated with large-bore catheters in the subclavian
clavian catheter deserve separate discussion. Subclavian and internal jugular veins. Artificial Organs 1993; 18: 318–21.
haemodialysis catheters are associated with subclavian vein Cameron LJ. Current Surgical Therapy, 6th edn. Vascular Access, 833–7.
stenosis ranging from 42 to 50 per cent.1,2,30 Significantly, McIntyre AS, Levison RA, Wood S, et al. Duplex Doppler ultrasound
the same type of catheter in the internal jugular site is asso- identifies veins suitable for insertion of central feeding
ciated with a markedly reduced internal jugular stenosis catheters. J Parenter Enteral Nutr 1992; 16: 264–7.
rate (0 to 10 per cent),1,22,31 thought to be attributable to Schwab S, Beathard G. The hemodialysis catheter conundrum: hate
stresses exerted by the rigid haemodialysis catheter on the living with them, but can’t live without them. Kidney Int 1999;
subclavian vein as it negotiates the brachiocephalic trunk. 53: 1–17.
The point of insertion becomes a focal point of stress
aggravated by transmitted irritation from the beating
heart13 and endothelial irritation eventually developing REFERENCES
into a stenotic lesion. The relatively direct path from the
right internal jugular vein to the SVC requires no bend in 1 Schwab S, Beathard G. The hemodialysis catheter conundrum:
the catheter. The result is that less stenosis is observed with hate living with them, but can’t live without them. Kidney Int
haemodialysis catheters at the internal jugular vein.1 1999; 53: 1–17.
This is particularly important to the chronic renal dialy- 2 Agee R, Kim, Balk AR. Central venous catheterization in the
sis patient whose continued dialysis depends on availability critically ill patient. Crit Care Clin 1992; 8: 677–86.
94 Emergency vascular access

3 Bambauer R, Inniger R, Pirrung KJ, et al. Complications and side 18 Farrell J, Gellens M. Ultrasound guided cannulation versus the
effects associated with large-bore catheters in the subclavian landmark-guided technique for acute haemodialysis access.
and internal jugular veins. Artificial Organs 1993; 18: 318–21. Nephrol Dial Transplant 1997; 12: 1234–7.
4 Barrera R, Mina B, Huang Y, Groeger JS. Acute complications of 19 Kwon TH, Kim YL, Cho DK. Ultrasound guided cannulation of the
central line placement in profoundly thrombocytopenic cancer femoral vein for acute haemodialysis access. Nephrol Dial
patients. Cancer 1996; 78: 2025–30. Transplant 1997; 12: 1009–12.
5 De Moor B, Vanholder R, Ringoir S. Subclavian vein hemodialysis 20 Taylor BS, Rhee RY, Muluk S, et al. Thrombin injection versus
catheters: advantages and disadvantages. Artificial Organs 1993; compression of femoral artery pseudoaneurysms. J Vasc Surg
18: 293–7. 1999; 30: 1052–9.
6 Duong MH, Jensen WA, Kirsch CM, et al. An unusual complication 21 Eisenberg L, Paulson EK, Kliewer MA, et al. Sonographically
during central catheter placement. J Clin Anesth 2001; 13: 131–2. guided compression repair of pseudoaneurysms: further
7 Wood KE, Reedy JS, Pozniak MA, Coursin DB. Phlegmasia cerulea experience from a single institution. AJR Am J Roentgenol 2000;
dolens with compartment syndrome: a complication of femoral 174: 1788–9.
vein catheterization. Crit Care Med 2000; 28: 1626–30. 22 Loesberg A, Taylor FC, Awh MH. Dislodgment of inferior vena
8 Farrell K, Walshe J, Gellens M, Martin KJ. Complications caval filters during ‘blind’ insertion of central venous catheters.
associated with insertion of jugular venous catheters for AJR Am J Roentgenol 1993; 161: 637–8.
hemodialysis: the value of postprocedural radiograph. Am J 23 Granke K, Abraham FM, McDowell DE. Vena cava filter disruption
Kidney Dis 1997; 30: 690–2. and central migration due to accidental guide-wire
9 Denys BG, Uretsky BF. Anatomical variations of internal jugular manipulation: a case report. Ann Vasc Surg 1996; 10: 49–53.
vein location: Impact on central venous access. Crit Care Med 24 Marelich GP, Tharratt RS. Greenfield inferior vena cava filter
1991; 19: 1516–19. dislodged during central venous catheter placement. Chest 1994;
10 Gibson KD, Gillen DL, Kohler TR, et al. Vascular access survival 106: 957–9.
and incidence of revisions: A comparison of prosthetic grafts, 25 Ellis PK, Deutsch LS, Kidney DD. Interventional radiological
simple autogenous fistulas, and venous transposition fistulas retrieval of a guide-wire entrapped in a greenfield filter –
from the United States Renal Data System Dialysis Morbidity and treatment of an avoidable complication of central venous access
Mortality Study. J Vasc Surg 2001; 34: 694–700. procedure. Clin Radiol 2000; 55: 238–9.
11 Schwartz CI, McBrayer CV, Sloan JH, et al. Thrombosed dialysis 26 Uppot RN, Garcia M, Gheyi V, et al. Entanglement of guide wires
graphs: comparison of treatment with transluminal angioplasty by vena cava filters during central venous catheter insertion:
and surgical revision. Radiology 1995; 194: 337–41. report of three cases and a review of the literature. Del Med J
12 Beathard GA. Thrombolysis versus surgery for the treatment of 2000; 72: 69–73.
thrombosed dialysis access grafts. J Am Soc Nephrol 1995; 6: 27 Wendt RJ. Cannulation of the right internal jugular vein is
1619–24. preferable to that of the left internal jugular vein. JAMA 1986;
13 Cameron LJ. Current Surgical Therapy, 6th edn. Vascular Access, 255: 1140.
833–7. 28 Rello J, Campistol JM, Almirall J, Revert LI. Vascular access for
14 McIntyre AS, Levison RA, Wood S, et al. Duplex Doppler haemodialysis. Lancet 1989; i: 379.
ultrasound identifies veins suitable for insertion of central 29 Akazawa S, Nakaigawa Y, Hotta K, et al. Unrecognized migration
feeding catheters. J Parenter Enteral Nutr 1992; 16: 264–7. of an entire guide-wire on insertion of a central venous catheter
15 Funaki B, Zaleski GX, Leef JA, et al. Radiologic placement of into the cardiovascular system. Anesthesiology 1996; 84: 241–2.
tunneled hemodialysis catheters in occluded neck, chest, or small 30 Beenen L, van Leusen R, Deenik B, Bosch FH. The Incidence of
thyrocervical collateral veins in central venous occlusion. Ann subclavian vein stenosis using silicone catheters for
Emerg Med 1999; 34: 711–14. hemodialysis. Artificial Organs 1993; 18: 289–92.
16 Sadler DJ, Gordon AC, Klassen J, et al. Image-guided central venous 31 Schillinger F, Schilleinger D, Montagnag R, Milcent T. Post
catheters for apheresis. Bone Marrow Transplant 1999; 23: 179–82. catheterization vein stenosis in haemodialysis: comparative
17 Hatfield A., Bodenham A. Portable ultrasound for difficult central angiographic study of 50 subclavian and 50 internal jugular
venous access. Br J Anaesth 1999; 83: 964. accesses. Nephrol Dial Transplant 1991; 6: 722–4.
 8
Imaging for Vascular Emergencies

PETER K ELLIS, BARRY KELLY, PETER KENNEDY, PAUL HB BLAIR

Introduction 95 Venous imaging 100

Vascular imaging techniques 95 References 104
Imaging in specific vascular territories 97

INTRODUCTION establishing priorities, particularly if an imaging procedure

has to be abandoned and the patient taken to the operating
theatre expeditiously.
Vascular imaging techniques have improved significantly As a result of recent endovascular developments, some
in the past decade and a wide range of invasive and non- centres have excellent imaging facilities in the operating
invasive modalities are now available to the clinician. It should theatre. This can reduce delays and facilitate endovascular
be remembered that when dealing with vascular emergencies, intervention.
time is of critical importance in preventing life-threatening It is important that a good working relationship is main-
haemorrhage and restoring vascular continuity and tissue tained between vascular surgeons and radiologists and that
perfusion. Although there has been an overall reduction in multidisciplinary guidelines based on local resources and
the time required to obtain high quality images, significant clinical expertise are agreed and implemented in the man-
delays during investigations still occur which may have a dis- agement of vascular emergencies.
astrous consequence for the patient. Liberal use of radio- This chapter provides a broad overview of the range of
logical investigations is never a substitute for good clinical imaging techniques available in managing vascular emer-
acumen. A careful history and detailed examination should gencies. The advantages and disadvantages of imaging
facilitate selection of the appropriate imaging technique. techniques in specific anatomical areas are discussed.
In certain situations, unstable patients in the accident Individual authors outline their preferred imaging tech-
and emergency department benefit from direct transfer to niques in managing specific vascular emergencies in the
the operating room for life- or limb-saving surgery. Recent relevant chapters of this volume.
developments in endovascular surgery have improved the
quality and range of imaging techniques available in the
operating room for such patients.
Alternatively, following diagnostic imaging, some patients
VASCULAR IMAGING TECHNIQUES
may benefit from radiological intervention such as vessel
embolisation or stent placement. In this situation it is Digital subtraction angiography
important that patients are monitored appropriately in
the imaging suite while retaining easy access for possible The modern angiographic suite is equipped with a large
anaesthetic and surgical intervention. A patient presenting diameter image intensifier which offers excellent real time
with a vascular emergency is best served by the early involve- fluoroscopic image quality with a minimised radiation
ment of both a senior vascular surgeon and radiologist. In dose. Many strategies are employed to keep the radiation
addition to established skills in performing complex inter- dose to the operator to a minimum such as pulsed fluo-
ventional procedures an experienced radiologist can often roscopy, last image hold, undertable lead curtains and ceiling
detect subtle signs from initial plain films or non-invasive hung lead glass which can be positioned optimally for each
tests. Similarly, during an evolving vascular emergency, the acquisition. The suite will often provide an operating the-
presence of an experienced vascular surgeon will help in atre environment with at least 20 air exchanges per hour,
96 Imaging for vascular emergencies

quality floor and wall coverings and a minimum of hori- and some detailed vascular studies can be time consuming,
zontal surfaces. Image acquisition can be achieved at high thus limiting its role in emergency situations. Its advan-
frame rates, typically up to 12 frames per second with a tages include the lack of ionising radiation, the relatively
1024 matrix. Postprocessing tools such as automatic pixel low cost of equipment and its mobility. The recent devel-
shift, rewindow and remask will often enable diagnostic opment of small portable scanners is likely to increase the
images to be retrieved from series acquisition even follow- role of emergency duplex sonography.
ing significant patient movement. An example of a periph-
eral leg run-off study is shown in Fig. 8.1a.
Computed tomography

Duplex ultrasound Computed tomography (CT) is widely available and in the

context of vascular imaging is relatively easy to carry out
Duplex ultrasonography is the combination of grey scale and to interpret. Imaging is performed in the arterial phase
ultrasound imaging with Doppler sonography allowing an following intravenous injection of contrast (approximately
evaluation of vessel morphology and blood flow character- 20–50 seconds). Multiplanar reconstruction and various
istics. The technique is particularly useful in the diagnosis display techniques such as surface shaded display and max-
of carotid artery disease and deep vein thrombosis (DVT). imum intensity projection provide different formats with
The modality is somewhat operator dependent, however, which to view the acquired data. Computed tomography

Figure 8.1 Normal lower limb

angiogram: (a) digital subtraction
angiogram, (b) computed
(a) (b) (c) tomography angiogram and
(c) magnetic resonance angiogram
 Imaging in specific vascular territories 97

provides superb contrast resolution and visualisation of IMAGING IN SPECIFIC VASCULAR

adjacent non-vascular structures thus giving a multisystem TERRITORIES
evaluation from one acquisition and is particularly useful
in the evaluation of a trauma patient. The development of
multislice technology has further expanded the indications Head and neck
for CT angiography (CTA) as accurate images of run-off
vessels (Fig. 8.1b) are now also available with the advan- The gold standard technique for imaging the carotid and
tages of reduced examination time, reduced contrast load vertebral arteries is selective angiography. The most import-
and increased contrast and spatial resolution.1 There is a ant complication of cerebral angiography is stroke occurring
radiation dose penalty but advances in detector technology as a result of emboli dislodged from atherosclerotic plaques
have minimised this disadvantage. The occasional draw- or the accidental injection of thrombus or air. The incidence
backs are the requirement for long and cumbersome post- of permanent cerebral ischaemic episodes is approximately
processing and the presence of calcified plaques which 0.5 per cent.10 Other complications include vessel dissection
further complicate the evaluation.2 and puncture site haematoma.
Duplex ultrasonography of the extracranial carotid arter-
ies provides exquisite images of the carotid bifurcation but
Magnetic resonance angiography falls short in the evaluation of high internal carotid artery
lesions and in characterising very high grade stenoses.
Magnetic resonance imaging (MRI) has some drawbacks Recent advances in MRA have resulted in high quality,
in the acute setting: emergency access is limited in many high resolution, non-invasive imaging which is likely to
institutions due to a large scheduled workload; it may be replace catheter directed angiography.
difficult to determine the presence of any contraindica-
tions to imaging for each patient; transfer of the intubated Imaging of thoracic vessels
patient into the magnetic field requires MRI compatible
anaesthetic equipment.3 Intravenous injection of gadolin- The thoracic aorta can be studied using a catheter passed to
ium-based contrast reduces the T1 value of blood so that the level of the ascending aorta. Several images per second
short TR (time of repetition) imaging results in high signal should be obtained using a number of radiographic projec-
from the blood, providing excellent contrast resolution.4 tions, in particular the steep right posterior oblique which
Most sequences used are three-dimensional gradient unfolds the aortic arch and the great vessels best (Fig. 8.2).
recalled echo techniques with heavy T1 weighting. Low Complications (1–3 per cent) are rare and are most
TR values ensure suppression of background structures,
while high intra-arterial concentration of contrast material
ensures good vessel-to-background contrast.5 Short imag-
ing sequences obtained during a single breath-hold min-
imise respiratory artefact. The use of contrast to directly
image the blood avoids the limitations encountered when
the imaging signal is derived from blood flow, such as long
imaging times, in-plane saturation effects and loss of signal
at sites of stenosis due to turbulent flow.6 This last limitation
often results in an overestimation of the degree of vessel
stenosis, although the presence of occlusion is detected very
accurately by both methods. When higher doses of gadolin-
ium are used results can match those of intra-arterial digi-
tal subtraction angiography (DSA), even for the clinically
important stenoses of 70–99 per cent.7
Additional techniques are being developed constantly
to improve image quality. Several authors have reported the
value of image subtraction, comparable to DSA, in which
an image is obtained immediately before administration of
gadolinium and used as a mask image to subtract from that
obtained after gadolinium is injected (Fig. 8.1c). Thus
background tissues, in particular fat, are removed and as a
result the quality of the image is improved.8 Magnetic res- Figure 8.2 Flush arch angiogram in right posterior oblique
onance angiography (MRA) has replaced invasive preoper- projection demonstrating the aortic arch and origins of great vessels
ative imaging in many centres, and it has been used as the in a patient with blunt chest trauma. There is a pseudoaneurysm
imaging modality for planning endovascular interventions.9 indicating injury of the proximal innominate artery
98 Imaging for vascular emergencies

commonly seen at the puncture site. Others include arte- hypertension but the overall mortality from this complica-
rial dissection, embolisation resulting in cerebral ischaemia tion is around 0.3 per cent.
and major artery thrombosis. In general terms, inexperi-
ence on the part of the operator and prolonged procedure Abdominal aorta
times compound the risk of complications.
Transoesophageal echocardiography (TOE) in the man- Where possible angiography of the aorta is performed via a
agement of the critically injured patient is of value in transfemoral approach, although in the situation of aortic
detecting possible injury to the thoracic aorta and in exclud- stenosis or occlusion a 4 Fr catheter placed via the brachial
ing the presence of blood in the pericardial sac. This is a or axillary artery is effective. In current practice direct
particularly useful investigation in unstable patients in the translumbar aortography is very rarely performed.
operating room or intensive care unit who cannot be trans- Volumetric CT scanning is the most commonly used
ported to the radiology suite. technique in the evaluation of aortic aneurysms as it pro-
Spiral CT images, derived during the pulmonary arter- vides accurate information concerning the size of the
ial phase in a single breath-hold following the administra- aneurysm and in cases of rupture by demonstrating with
tion of intravenous contrast, have been shown to be useful high sensitivity the presence of retroperitoneal haemor-
in the diagnosis of pulmonary embolism.11 The options for rhage. The use of multiplanar reconstruction allows accur-
reconstruction and display, and the more rapid imaging ate measurement of aneurysm dimensions essential in
afforded by multislice scanning, will no doubt increase the planning for aortic stent grafting.
utility of this modality in coming years. Magnetic reso- In the assessment of aortic aneurysms MR imaging can
nance imaging can also be used to visualise the pulmonary provide excellent depiction without the requirement for
arteries, (Fig. 8.3) and, in addition, MR velocity mapping iodinated contrast, again multiplanar reformatting is pos-
can confirm reverse diastolic flow in patients with pul- sible providing advantages similar to those with CT scan-
monary hypertension. ning. Ultrasound is useful particularly in screening for aortic
Pulmonary angiography today is a safe and simple pro- aneurysms; its sensitivity in the detection of retroperi-
cedure and is still considered the gold standard for imaging toneal haemorrhage, however, is significantly less than that
these vessels. Access is usually obtained via a femoral vein of CT scanning. The use of Doppler ultrasound, particu-
although the jugular or brachial routes are alternatives. larly when combined with ultrasound contrast agents, has
Complications are rare and usually involve cardiac been shown to be extremely sensitive in the detection of
arrhythmias while the catheter is within the heart. Acute endoleaks post-aortic stent grafting.12
cor pulmonale may occur in patients with pulmonary
Renal arteries

The renal arteries can be assessed angiographically by the

placement of a flush catheter within the aorta just above
their origins. More than one view is normally obtained as
renal arteries have a variable site of origin from the aortic
wall. Carbon dioxide is a useful adjunct to the procedure in
a setting in which the patient has renal impairment and
where radiographic contrast is relatively contraindicated.
Contrast-enhanced MRA now offers an alternative. Early
sequences using time-of-flight (TOF) methods or phased-
contrast techniques were unfortunately limited by poor
spatial resolution and in-plane flow saturation. Three-
dimensional acquisitions using gadolinium enhancement
with a single breath-hold overcome these problems and allow
satisfactory imaging in most instances (Fig. 8.4). Sensitiv-
ities of 95–100 per cent have been regularly reported.13

Visceral aortic branches

Angiography is still the gold standard for assessment of the

coeliac axis and mesenteric arteries. Vessel origins are best
displayed using a lateral or lateral oblique projection of a
Figure 8.3 Three-dimensional gadolinium-enhanced magnetic flush aortogram. Selective injections can be undertaken,
resonance angiogram of the chest demonstrating pulmonary and particularly when looking for distal pathology such as acute
systemic vessels (courtesy of Siemens Medical Imaging) bleeding (Fig. 8.5). A variety of precurved catheters are
 Imaging in specific vascular territories 99

available for this purpose and the optimal choice will depend particularly useful in the assessment of the portal vein,
on the configuration of a vessel. Very distal positions within angiographic definition of which is generally achieved
the mesenteric or coeliac territories can be achieved using indirectly. It is also useful in the assessment of hepatic
coaxial catheters where, for example, a 3 Fr catheter is arterial flow volumes and in liver transplantation where
passed through the lumen of the 4 or 5 Fr catheter placed in stenosis of the hepatic arterial anastomosis can lead to loss
the origin of the vessel. This technique is particularly use- of the graft.
ful where acute distal bleeding points are identified in var- Computed tomography arteriography is also useful and
ious settings: trauma, acute gastrointestinal haemorrhage the various methods of image reconstruction described
and wherever embolisation is contemplated. above can be used to demonstrate very adequately the prox-
Doppler ultrasound can be used to assess the volume of imal branches of the superior mesenteric artery and coeliac
flow in the superior mesenteric artery by measuring the axis. Distal active bleeding points can also be detected as a
luminal area and the mean velocity of flow. In general, the blush of contrast extravasation. Mesenteric occlusive disease
waveform in this vessel demonstrates high impedance flow can also be assessed using MR angiography. A single breath-
with reversal of flow in diastole. Doppler ultrasound is hold, three-dimensional gradient echo sequence using
intravenous gadolinium can be used to avoid the artefacts
produced using older techniques. In addition, real time
imaging of up to 20 images per second can be achieved.14
Quantification of blood flow is also achievable using
several MR techniques.

Pelvic imaging

Haemorrhage is a common complication of pelvic trauma.

Computed tomography can very adequately delineate
pelvic haematoma and with dynamic scanning extravasa-
tion of contrast will point to an arterial bleeding source.
Frequently CT fails to show the exact source of bleeding
but that objective is best achieved using arteriography.
Figure 8.4 Three-dimensional gadolinium-enhanced magnetic
Flush angiography from a site just above the aortic bifurca-
resonance angiogram of renal arteries (courtesy of Siemens
Medical Imaging)
tion will frequently demonstrate an actively bleeding source
in the pelvic vessels (Fig. 8.6). Should that fail, selective

Figure 8.5 Selective arteriogram of distal jejunal arcade in a Figure 8.6 Flush arteriogram in a patient with multiple pelvic
patient with acute gastrointestinal haemorrhage. Note active fractures. Note extravasation from an actively bleeding right
extravasation from a jejunal bleeding point obturator artery
100 Imaging for vascular emergencies

injection, particularly of both internal iliac arteries, will be provide equivalent, if not superior, diagnostic information.
helpful using a variety of precurved catheters. This should Studies have shown comparable, or in some cases, slightly
allow identification of bleeding sites and subsequently better visualisation of distal vessel patency using MR
haemorrhage control by transcatheter embolisation using angiography.17
a variety of embolic materials.15
The decision to attempt pelvic embolisation can be
extremely difficult and requires an informed discussion Upper extremity
between the clinicians involved. It is important to exclude
other sources of abdominal haemorrhage and to achieve Angiographic studies of the upper extremity are normally
some form of pelvic bony stability before embarking on performed via a transfemoral approach. Injection in the
attempts at embolisation. aortic arch is necessary for the evaluation of the origins of
the subclavian arteries. In order to visualise the more distal
vasculature, selective catheterisation of the artery and its
Lower limb vessels branches can be carried out. Direct catheterisation is asso-
ciated with a higher complication rate.
The imaging gold standard for the assessment of lower limb The distal vasculature of the upper extremity, superfi-
arteries is probably angiography (Fig. 8.7). A non-invasive cially located as it is, is eminently suited to Doppler ultra-
method of assessment is duplex ultrasonography. This sound examination. The peripheral arteries of the arm
technique is time consuming and requires considerable show the same triphasic pattern as the lower extremity
experience, but it has a sensitivity and specificity exceeding arteries. Magnetic resonance angiography using various
90 per cent in the detection of arterial lesions.16 techniques can provide excellent images of the upper
A wide variety of MR angiographic pulse sequences can be extremity, including detailed angiography of the hand and
used to demonstrate adequate blood flow. Two-dimensional wrist. In addition, in the diagnosis of thoracic outlet syn-
time-of-flight imaging (with saturation banding) can be drome, MRI offers the capability of providing angio-
employed to detect flow in a single direction but it normally graphic imaging as well as useful anatomical detail of
takes a relatively long time and currently gadolinium- cervical ribs or bands and the associated deviation or dis-
enhanced three-dimensional angiography is preferred in tortion of the brachial plexus and subclavian artery.
many centres. The latter technique, however, demands the
exact timing of contrast enhancement and a moving table
is also necessary. Magnetic resonance angiography may well
replace catheter directed angiography but to do that it must VENOUS IMAGING

Lower limb deep venous thrombosis

The two commonly used imaging methods for identifying

lower limb DVT are ascending venography and ultrasound.

Ascending venography

Ascending venography is a method of delineating the

venous anatomy of a territory by opacifying the vessels
with iodinated contrast, typically a low osmolar water-
soluble contrast agent. After placing a tourniquet around
the ankle, a vein on the dorsum of the foot is cannulated
with a 21 gauge needle and a further tourniquet is
placed above the level of the knee. The function of
both tourniquets is to divert venous return from the super-
ficial veins into the deep veins so as to opacify them with
contrast.
The patient is placed on a tilting fluoroscopic table and
50 mL of contrast is injected. Fluoroscopic imaging of the
leg is performed from the level of the ankle joint, through
Figure 8.7 Arteriogram of the left femoral arteries. Note abrupt the knee, thigh and then the pelvis. The progress of the
‘cut-off’ of the profunda femoris artery due to transection by a contrast is screened and ‘spot-films’ are taken; in normal
bullet circumstances the veins opacify uniformly. When veins are
 Venous imaging 101

ultrasound is now routinely used as the primary imaging

method for suspected DVT.

Ultrasound venography

Ultrasound imaging utilises the production of high fre-

quency sound waves, typically of megahertz frequency, to
identify tissue planes spatially within the body. Ultrasound
is particularly effective in the evaluation of non-osseous,
non-gaseous soft tissues and also fluid-filled structures.
These attributes render it particularly effective in the evalu-
ation of venous anatomy. A piezoelectric crystal is used
to generate the sound waves and the image generated can
be displayed on a television monitor, providing a two-
dimensional view of the veins: this is known as real time or
B-mode sonography. Not only can the physical dimensions
of veins be estimated but, using the principle of Doppler
ultrasound, information on the presence, speed, and direc-
tion of blood flow can also be derived.
The Doppler principle utilises red blood cells as reflec-
tors of pulses of sound waves emitted and received by the
ultrasound probe. The blood’s speed and direction can be
encoded as either a spectral trace or a colour image. In the
latter case the colour denotes the direction of blood flow
away from or towards the probe and the hue and bright-
ness correspond to the speed of the flow. Absence of
colour-coded flow or persistent underfilling of a venous
segment with isoechoic or echogenic thrombus is clear
evidence of thrombus.18
The combination of the real time image and spectral
trace or colour map represents duplex ultrasound and all
three appearing on the same image is known as triplex
ultrasound.

Ultrasound imaging in deep venous

Figure 8.8 Leg venogram: thrombus within the femoral vein thrombosis

Essentially, two components are incorporated in the ultra-

thrombosed a small track of contrast is seen at the margins sound evaluation of the lower limb with DVT. The first of
of the vessel while the central thrombosed section of the these is venous compression, and the second, a series of
vessel remains unopacified, an appearance which is termed adjunctive manoeuvres to provide supplementary infor-
‘tramlining’ (Fig. 8.8). mation based on the phasic flow of blood in the veins.
As a rule the imaging proceeds cranially to the level of
the iliac vein and the lower part of the inferior vena cava.
VEIN COMPRESSIBILITY
Streaming of contrast at the level of the iliac vein can
reduce resolution and therefore a key observation in the When the soft tissues overlying normal healthy veins are
diagnosis of iliac thrombosis is the demonstration of an compressed with an ultrasound probe, their thin walls are
extensive collateral network from the affected side across brought together obliterating the vein lumen completely.
to the contralateral normal side as small venous channels The walls of an artery, however, being more muscular,
enlarge to permit venous drainage. cannot be easily compressed, which is a useful differentiat-
Contrast ascending venography is still recognised as the ing feature.
gold standard for obtaining images of lower limb DVT but In the presence of thrombosis the vein lumen will
it requires patient exposure to ionising radiation in addition not compress fully (Fig. 8.9), and this failure is an import-
to intravenous injection of iodinated contrast. Consequently, ant finding. Although established thrombus is echogenic
102 Imaging for vascular emergencies

and easily visualised, a recent thrombus may be sonolu- circles (Fig. 8.10). This makes a convenient reference
cent, i.e. black, and, on real time ultrasound examination, point. The two veins should compress freely but the artery
may be indistinguishable from normal sonolucent blood will not.19 Transverse compression is therefore applied
within a vein. The normal vein is, of course, easily from the level of the groin to the knee at 1 cm intervals.
compressible. Cranial to the level of the knee joint, the femoral vessels
Imaging begins at the level of the groin. Three vessels, enter the adductor canal and move posteriorly. Direct
the long saphenous vein, common femoral vein and common compression in this area can be problematic. It may be
femoral artery are identified as three black overlapping necessary to turn the patient into the decubitus or erect

Figure 8.9 Transverse ultrasound of

groin: deep venous thrombosis

Figure 8.10 Transverse ultrasound

image of groin: normal vascular
structures
 Venous imaging 103

position to evaluate the venous segments in the adductor augmented by demonstration of phasic flow, will exclude
canal and indeed the popliteal fossa. significant thrombus.
The sensitivity for diagnosing venous thrombus from the
popliteal venous level upwards is of the order of 98 per cent.20 CENTRAL AND CAVAL THROMBUS
Studies have demonstrated that the sensitivity falls below
popliteal level and therefore calf veins are not often formally In a significant number of cases thrombosis of the superior
evaluated with ultrasound. Currently there is significant vena cava and brachiocephalic veins is associated with
inter-physician variability on the treatment algorithm for sinister underlying pathology. Lymphadenopathy and neo-
infrapopliteal DVT and, in addition, because of the lack of plasia, particularly bronchogenic carcinoma and lymphoma
ionising radiation, it is perfectly feasible to perform serial should be excluded. Formal evaluation of the frontal chest
ultrasound examinations of the lower limb to exclude the radiograph is essential to ensure that the mediastinal con-
possibility of thrombus propagation extending from the tours are normal.
calf into the popliteal veins.21 When the superior vena cava and the brachiocephalic
veins are being evaluated, ultrasound has only a very limit-
ed role. The imaging method of choice is contrast venog-
SUPPLEMENTARY ULTRASOUND TECHNIQUES raphy. To opacify the brachiocephalic veins and the superior
Unlike the pulsatile arterial system, flow within the venous vena cava, synchronous injections of contrast are performed
system is phasic, i.e. flow which varies with the phases of by simultaneously injecting 50 mL of water-soluble contrast
respiration. This phasic flow can be observed by imaging a medium through 21 gauge needles positioned at the level
venous segment and using Doppler ultrasound to identify of the elbow. Fluoroscopic evaluation of the veins is then
either colour or spectral frequency changes as the patient performed and spot images taken.
breathes normally.
Clearly, if cessation of normal phasic activity can be
demonstrated, for example, by asking the patient to breath- Cross-sectional imaging in central venous
hold, this will demonstrate that there is no obstruction to thrombosis
normal venous return between the area being evaluated,
e.g. the groin, and the right side of the heart. It is therefore COMPUTED TOMOGRAPHY
possible to indirectly ascertain whether or not there is Computed tomography is very accurate at identifying
occlusion of the inferior vena cava and iliac veins. In add- mediastinal masses which are causing compression or
ition, if the patient actively increases intra-abdominal pres- effacement of the superior vena cava. Low attenuation within
sure, e.g. with a Valsalva manoeuvre, further proof is the SVC, however, can result from the mixing of contrast
provided that there is no venous obstruction or throm- enhanced blood draining from the arm which has been
bosis proximally. cannulated, and unopacified blood from the contralateral
When the deep veins are imaged between the femoral limb. This artefact can be mistaken in normal individuals
and popliteal levels and the calf is squeezed distally by the for thrombus. The presence of delayed or collateral venous
sonologist, blood is propelled forward in a cephalad direc- filling provides further evidence of central thrombus. Com-
tion. This sudden relatively high velocity venous blood puted tomography also has the added advantage in that if
is detectable again either by using colour flow or spectral an associated underlying malignancy is present, it can be
Doppler. Confirming a normal calf squeeze response indi- accurately staged simultaneously.22
cates that there is a patent venous system distal to the point Helical and multidetector CT angiography enable
of the ultrasound probe. The inclusion of the calf squeeze volumetric acquisition of data in a single breath-hold
and Valsalva manoeuvres allows the complete evaluation and this reduces motion artefact and respiratory mis-
of the venous system from the level of the calf veins to the registration, improving accuracy of detection of venous
inferior vena cava. thrombosis.23

MAGNETIC RESONANCE ANGIOGRAPHY

Venous thrombosis in the upper limbs Magnetic resonance angiography has an increasing role in
the first line investigation of vascular pathology – MR has
ISOLATED ARM VEIN THROMBOSIS
the advantages of multiplanar imaging and does not use
Once again the same imaging methods are used. With con- ionising radiation. Like CT, MRA can also delineate the
trast venography, a tourniquet is placed at the level of the underlying aetiological process.
elbow to occlude the superficial cephalic vein and allow Magnetic resonance venography may also permit the
the contrast to be directed preferentially through the deeper assessment of vascular flow patterns and thereby determine
upper limb veins. With Doppler sonography, compressibility, the evolution of thromboembolic events.24
104 Imaging for vascular emergencies

compared with conventional angiography – initial experience.

Conclusions Radiology 2000; 216: 909–15.
7 Loewe C, Schoder M, Rand T, et al. Peripheral vascular occlusive
Imaging of vascular disease continues to evolve. Increas- disease: evaluation with contrast-enhanced moving-bed MR
ingly non-invasive alternatives to percutaneous angio- angiography versus digital subtraction angiography in 106
graphic and venographic techniques such as duplex patients. Am J Roentgenol 2002; 179: 1013–21.
ultrasound, CT and MRI are used in the management of 8 Ho KYJAM, Haan MW de, Kessels AGH, et al. Peripheral vascular
vascular emergencies and are gaining acceptance in most tree stenoses: detection with subtracted and nonsubtracted MR
vascular centres. It is likely that these imaging modalities angiography. Radiology 1998; 206: 673–81.
9 Mistretta CA. Relative characteristics of MR angiography and
will eventually replace invasive diagnostic imaging, the
competing vascular imaging modalities. J Magn Reson Imag
latter being largely reserved for therapeutic intervention.
1993; 3: 685–98.
The principles, advantages and drawbacks of these tech- 10 Willinsky RA, Taylor SM, terBrugge BAK, et al. Neurologic
niques when applied to specific anatomical areas have to complications of cerebral angiography: prospective analysis of
be considered. In addition, the field of vascular radiology 2899 procedures and review of the literature. Radiology 2003;
is evolving and now exercises a rapidly expanding thera- 227: 522–8.
peutic role. Continued innovation can only be of benefit 11 Remy-Jardin M, Remy J, Wattine L, et al. Central pulmonary
to patients and those involved in the treatment of vascu- thromboembolism: diagnosis with spiral volumetric CT with
lar diseases. single breathhold technique – comparison with pulmonary
angiography. Radiology 1992; 185: 381–7.
12 McWilliams RG, Martin J, White D, et al. Use of contrast-
enhanced ultrasound in follow up after endovascular aortic
aneurysm repair. J Vasc Interv Radiol 1999; 10: 1107–14.
Key references 13 Bakker J, Beek FJ, Beutler JJ, et al. Renal artery stenosis and
accessory renal arteries: accuracy of detection and visualisation
Baum RA, Rutter CM, Sunshine JH, et al. Multicenter trial to
with gadolinium enhanced breath-hold MR angiography.
evaluate magnetic resonance angiography of the lower
Radiology 1998; 207: 497–504.
extremity. JAMA 1995; 174: 875–80.
14 Kerr AB, Pauly JM, Hu BS, et al. Real-time interactive MRI on a
Baxter GM. The role of ultrasound in deep venous thrombosis.
conventional scanner. Magn Reson Med 1997; 38: 355–67.
Clin Radiol 1997; 52: 1–3.
15 Ayella RJ, DuPriest RW Jr, Khanega SC, et al. Transcatheter
Miller N, Satin R, Tousignant L, Sheiner NM. A prospective study
embolisation of autologous clot in the management of bleeding
comparing duplex scan and venography for the diagnosis of
associated with fractures of the pelvis. Surg Gynecol Obstet
lower-extremity deep venous thrombosis. Cardiovasc Surg
1978; 147: 849.
1996; 4: 505–8.
16 Cossman DV, Ellison JE, Wagner JH, et al. Comparison of
Prokop M. Multislice CT angiography. Eur J Radiol 2000; 36: 86–96.
contrast arteriography to arterial mapping with colour-flow
Remy-Jardin M, Remy J, Wattine L, et al. Central pulmonary
duplex imaging in the lower extremities. J Vasc Surg 1989;
thromboembolism: diagnosis with spiral volumetric CT with
10: 522–9.
single breathhold technique – comparison with pulmonary
17 Baum RA, Rutter CM, Sunshine JH, et al. Multicenter trial to
angiography. Radiology 1992; 185: 381–7.
evaluate magnetic resonance angiography of the lower
extremity. JAMA 1995; 174: 875–80.
18 Baxter GM. The role of ultrasound in deep venous thrombosis.
Clin Radiol 1997; 52: 1–3.
REFERENCES
19 Kelly BE, McArdle CS. Imaging the Acutely Ill Patient: A Clinician’s
Guide. London: WB Saunders, 2001; 112–15.
1 Rubin GD, Shiau MC, Leung AN, et al. Aorta and iliac arteries: 20 Miller N, Satin R, Tousignant L, Sheiner NM. A prospective study
single versus multiple detector-row helical CT angiography. comparing duplex scan and venography for the diagnosis of
Radiology 2000; 215: 670–6. lower-extremity deep venous thrombosis. Cardiovasc Surg 1996;
2 Prokop M. Multislice CT angiography. Eur J Radiol 2000; 36: 86–96. 4: 505–8.
3 Chaljub G, Kramer LA, Johnston RF, et al. Projectile cylinder 21 Baud JM, Stephas L, Ribadeau-Dumas C, et al. Short and medium
accidents resulting from the presence of ferromagnetic nitrous term duplex sonography follow up of deep venous thrombosis of
oxide or oxygen tanks in the MR suite. AJR Am J Roentgenol 2001; the lower limbs. J Clin Ultrasound 1998; 26: 7–13.
177: 27–30. 22 Webb WR, Brant WE, Helms CA. Fundamentals of Body CT, 2nd
4 Prince MR. Contrast-enhanced MR angiography: theory and edn. Philadelphia: WB Saunders, 1998: 35–7.
optimization. Magn Reson Imaging Clin N Am 1998; 6: 257–67. 23 Bradbury MS, Kavanagh PV, Bechtold RE, et al. Mesenteric
5 Ho KY, Leiner T, de Haan MW, van Engelshoven JM. Peripheral venous thrombosis: diagnosis and non-invasive imaging.
MR angiography. Eur Radiol 1999; 9: 1765–74. Radiographics 2002; 22: 527–41.
6 Mitsuzaki K, Yamashita Y, Sakaguchi T, et al. Abdomen, pelvis and 24 Butty S, Hagspiel KD, Leung DA, et al. Body MR venography.
extremities: diagnostic contrast-enhanced turbo MR angiography Radiol Clin North Am 2002; 40: 899–919.
 9A
Outcomes of Emergency Vascular Procedures.
A View from the British Isles

JONOTHAN J EARNSHAW

Introduction 105 Towards a national vascular database in the UK 107

Results of emergency arterial surgery in the 106 and Ireland
British Isles References 109

INTRODUCTION surgeon may plan to deal with urgent operations in day-

light hours, the fact is that they often have to be squeezed
The results of elective arterial procedures are well docu- into a schedule that is already busy. In many hospitals
mented, but that is not the case in emergency vascular where the vascular services are under pressure, it could be
surgery. Yet, out-of-hours operating constitutes a large argued that emergency patients receive less than ideal care.
proportion of the workload of a vascular surgeon, in con- In the UK, arrangements for the care of vascular patients
trast to the other specialties grouped under ‘general sur- remain varied. Increasingly vascular surgeons are gather-
gery’. Reports are sparse because the data are harder to ing into major units where 24-hour cover is offered with
collect. It is generally believed that the outcome from on-call specialist staff. This has only been possible in large
emergency arterial surgery is worse than for elective oper- conurbations with populations of over 600 000, where four
ation which might explain the natural reluctance to detail or five surgeons can manage an on-call rota. Other models,
those results. Much of the emergency surgery in the UK including cross-cover between adjacent units, have enabled
and Ireland used to be performed by trainees, though this an increasing number of vascular on-call rotas.1 This does
is changing as a result of recent training reforms. mean that a patient with a vascular emergency may need to
Arterial surgery falls into three broad categories: carotid travel some distance for specialist treatment. For most vas-
disease, which is seldom performed urgently, aortic cular patients, the delay is immaterial; indeed several studies
aneurysms, where up to 50 per cent of the workload of an have suggested that even for those with a ruptured aortic
individual unit may include urgent or ruptured aneurysms, aneurysm, there is no correlation between distance trav-
and the management of leg ischaemia, the traumatic causes elled to hospital and outcome.2 Many surgeons believe that
of which will not be discussed here. The management of this is a process of natural selection, because only those
acute leg ischaemia represents one of the most difficult likely to survive the operation survive the journey.
challenges for a vascular surgeon. Management decisions There is some evidence that the outcome of vascular
are complex and treatment selection can significantly affect surgery is better when carried out by a specialist, particu-
outcome, which is often poor, with high amputation and larly for acute leg ischemia, the results of which improved
mortality rates. Although a large proportion of vascular with the establishment of a regional vascular unit.3 In the
operations are performed out of hours, it is often possible UK, there remain a number of units with insufficient sur-
to avoid doing so under emergency conditions. Only true geons to man an on-call vascular rota. Although most elect-
ruptured aortic aneurysms and acute total leg ischaemia ive arterial operations in the UK are performed by trained
demand immediate intervention; other conditions may vascular surgeons, the arrangements for emergency cover
permit a period of observation and semi-elective surgery. are variable. In some hospitals, dedicated vascular sur-
Although this sounds attractive and suggests that a vascular geons provide ad hoc cover, but in others, general surgeons
106 Outcomes of emergency vascular procedures: British Isles

provide emergency care. As might be expected this can exag- to define a clear outcome for this condition. One reason is
gerate the variation in the outcome from emergency vascu- that most patients never actually reach hospital, and once
lar surgery, though this may not simply be the result of there, a significant number are rejected as unfit. It is
surgical skill. In a study from Wales, the outcome of patients estimated that only 15 per cent of patients with a ruptured
with a ruptured aneurysm was similar if they had surgery abdominal aortic aneurysm (AAA) survive. Surgeons may
by a general or a vascular surgeon, but general surgeons have less influence over the outcome of the condition than
were more likely to turn a patient down for operation.4 they think. It is often difficult to obtain complete data on
patients with a ruptured AAA. Those who die in the acci-
dent and emergency department or in the preoperative bay
RESULTS OF EMERGENCY ARTERIAL usually end up with scanty hospital records and the data
SURGERY IN THE BRITISH ISLES somehow never seem to reach a surgical database.
Classification difficulties can also confuse the issue.
A shocked patient with a ruptured aneurysm is a clear clinical
Most publications on outcome derive from the practices of
entity, however, some patients with a contained rupture may
enthusiasts or major centres. In general, only good results
remain well for hours or even days. Some patients present
are reported and that imparts a degree of bias to the true
with a tender aneurysm and have an urgent operative repair.
outcome for any disease or treatment. Outcomes derived
It is clear that there is a spectrum of outcomes for these dif-
from prospectively collected data, preferably from several
ferent situations. Surgeons may influence the results they
sources in a multicentre trial, or results from ran-
report by choosing which groups to include.9 For example,
domised trials are more likely to be nearer the truth. Some
the best results for aortic surgery are obtained by including
Scandinavian countries have large incident databases from
only patients who have an elective admission and operation.
which both process and outcome data can be obtained
Results for emergency AAA repair can be improved by
(Swedvasc and Finnvasc). In the UK, a few multicentre
counting all those who have an out-of-hours operation,
studies have been carried out. The Audit and Research
including patients with a tender non-ruptured aneurysm.
Committee of the Vascular Society of Great Britain
There are few good reports of outcome from ruptured
and Ireland (VSGBI) has conducted prospective and
AAA. Meta-analysis of all English language publications
retrospective audits on some of the principal vascular
suggested that the average mortality rate was 48 per cent
operations.
and that the results have improved slowly over the last
50 years.10 In the UK, the VSGBI pilot national database
Carotid surgery included 276 patients who had AAA surgery, but only
43 who had emergency surgery. The mortality rate was
Most carotid surgery is done electively. In the audit per- 23 per cent for urgent AAA repair and 51 per cent for rup-
formed by the VSSGBI, there were no emergency pro- ture.11 In Wales a prospective audit of operations per-
cedures.5 Urgent carotid endarterectomy for recent stroke formed by both general and vascular surgeons revealed a
became unpopular in the 1970s following the publication survival rate of only 36 per cent.4 The results for the two
of dismal results.6 This may be explained by the fact that a groups were similar, though vascular surgeons operated on
proportion of the operations were done on patients with 82/92 (89 per cent) of patients admitted under their care
cerebral hemorrhage rather than embolic disease. Recently, with a 39 per cent survival rate, compared with general sur-
it has been argued that a more selective approach using geons who operated on 51/141 (36 per cent) patients with
computed tomography to identify appropriate patients a 31 per cent survival rate.4 Longitudinal data from large
with a thromboembolic stroke means that more patients vascular units such as the one in Edinburgh have shown little
could be offered urgent surgery. In a pilot trial, however, change in survival over the last 20 years.12 Mortality rates
few suitable patients were identified (16 from 593 reviewed), after ruptured AAA actually rose in the second decade
although surgical results were reasonable.7 Perhaps the from 35 to 40 per cent.
most frequent indication for emergency intervention is when Operative outcome can be improved by case selection.
a patient develops a stroke after carotid endarterectomy The more unfit patients are refused surgery, the better will
(see Chapter 14). Most surgeons reoperate for an early be the results of an individual surgeon. Surgeons have agon-
carotid occlusion as it is felt intuitively that this should ised about the indications for repair of a ruptured AAA in
improve outcome. There is little evidence, even in recent an unfit patient. Hardman et al. described a number of cri-
studies, that this is the case.8 teria that can help make this difficult decision.13 There is
also the possibility that the Physiological and Operative
Severity Score for the enUmeration of Mortality and mor-
Aortic surgery bidity (POSSUM) physiology scoring could give a preoper-
ative prediction of outcome.14 The Portsmouth POSSUM
Although ruptured aortic aneurysm is the most frequent rea- model was developed from 213 emergency aneurysm repairs,
son for an emergency vascular operation, it is very difficult with a mortality rate of 40 per cent.15
 Towards a national vascular database in the UK and Ireland 107

As the published results have barely changed in a gener- define the patients most likely to benefit from throm-
ation, it may be time to think of alternative means of bolytic therapy. Factors affecting amputation-free survival
reducing the community mortality of aneurysm disease. included the severity of leg ischaemia and the type of ves-
One way would be to commence a screening programme sel occluded, namely native artery (72 per cent) or graft
to detect AAAs while asymptomatic and to perform elec- (78 per cent). The main anxiety in thrombolysis is the risk
tive repair before rupture occurs.16 An alternative might be of stroke which occurred in 2.4 per cent of patients on the
to consider endovascular repair for ruptured AAAs. Small database although only half of these were haemorrhagic,
series reports of this treatment exist,17 but regular inter- and many occurred several days after initiating thrombolysis.
ventions would require an investment in infrastructure
and a change in culture. On the other hand, that might be
the only way of bringing about a significant improvement
TOWARDS A NATIONAL VASCULAR
in outcome for ruptured AAAs.
DATABASE IN THE UK AND IRELAND

Acute leg ischaemia With the exception of Northern Ireland, which has had a
vascular database (Northern Ireland Vascular Registry or
A prospective audit of acute leg ischaemia in Gloucestershire NIVASC) for the last seven years, surgeons in the rest of
suggested that a county with a population of half a million the UK have looked enviously at the well ordered national
would treat approximately 75 patients a year.18 The overall databases of Scandinavian countries. There remains a
30-day outcome was limb salvage 67 per cent, amputation problem of scale, however, in that expectations of collecting
7 per cent and deaths 26 per cent. A number of patients routine data from over 400 individual consultant members
were not suitable for intervention and when they were of the VSGBI would require resources beyond any current
excluded, the results for patients treated actively were limb medical organisation. Existing databases, however, have
salvage 78 per cent, amputation 6 per cent and deaths really only collected data on process and mortality, and
16 per cent.18 surgeons in tertiary referral centres or deprived inner city
The VSGBI also completed a prospective audit on acute areas may feel disadvantaged if their clinical performance
ischaemia: 539 episodes were reported in 474 patients.19 At were to be judged solely on the basis of crude mortality
30 days 70 per cent of legs were definitely viable, 16 per cent rates. In fact such assessments might influence the surgeon
had been amputated and the overall mortality was 22 per offering surgical treatment to a patient. A surgeon worry-
cent, these results confirming the high risk attached to the ing about personal results may be reluctant to operate on a
treatment of acute leg ischaemia.19 A subsequent paper high risk patient. Finally, there is the problem of verifica-
demonstrated the poor intermediate term outcome for this tion. Most registries rely on the honesty of their partici-
condition, a further 35 per cent of initial survivors having pants. A follow-up of patients who had surgery but were
died within 2 years.20 Recurrent leg ischaemia and subse- not included in the database, whether accidentally or delib-
quent amputation risk appeared to be reduced by warfarin erately, revealed that these are a very high risk group.22
anticoagulation. The study also highlighted the variety of The VSGBI has adopted the attitude that any national
different methods of treatment available. This difficult database should take account of the case-mix of patients.
condition requires decisions to be taken at senior level, For the past few years there has been an investigation into
such as making the choice between surgery or throm- various scoring systems that could be used to incorporate
bolytic therapy. An easy embolectomy can be done by an an allowance for case-mix into a comparison of outcomes
experienced trainee, but when the operation is not pro- among individual surgeons. UK vascular surgeons have
ceeding well a consultant’s experience is needed, particu- performed two large trials collecting POSSUM data items
larly if a full range of vascular procedures may have to be on their patients. POSSUM scoring involves 12 preopera-
considered. Thrombolytic therapy is a high risk option and tive physiological data items and eight operative items.
demands the active involvement of trained radiology staff It has been shown that both POSSUM and Bayes analytical
(see Chapters 8, 16 and 39). techniques used on this dataset can predict outcome. The
The Thrombolysis Study Group has collected data on first VSGBI study involved nearly 1500 patients from 121
intra-arterial thrombolysis over the past decade.21 Over surgeons in 93 hospitals.23 They collected data on consecu-
1000 episodes of lysis have been evaluated. The initial tive arterial procedures over a two-month interval in 1998.
outcome after intra-arterial lysis was a complete lysis The statisticians introduced a new POSSUM regression
rate of 41 per cent, a partial lysis rate of 28 per cent, and equation, known as V-POSSUM, which predicted outcome
a failed lysis, or lysis but no run-off, rate of 29 per cent. across the range of arterial procedures and hospitals. Indeed
At 30 days the outcome was amputation-free survival in it was possible to do so based on the preoperative physi-
75.2 per cent, major limb amputation 12.5 per cent and ology scores alone, meaning that in future, a preoperative
deaths 12.4 per cent. Although most of this work has been outcome prediction might be possible for an individual
observational, the Group is now in a position to try to patient.23
108 Outcomes of emergency vascular procedures: British Isles

Table 9A.1 Results of POSSUM analysis on abdominal aortic aneurysms (AAA) from the Vascular
Surgical Society of Great Britain and Ireland (VSGBI) Pilot National Vascular Database 2001

Per cent range Mean per cent

predicted risk predicted risk No. Predicted Reported 2

(a) All abdominal aortic aneurysms *

Mortality: V-POSSUM physiology and operative scores
0 to  7 3.84 352 14 19 2.33
7 to  15 10.61 204 22 31 4.52
15 to  30 20.91 118 25 31 2.05
30 to 100 48.47 99 48 35 6.82
0 to 100 13.95 773 108 116 15.72
2  15.720 P(4 df)  0.00 Evidence of lack of fit

Mortality: V-POSSUM physiology only scores

0 to 7 3.89 446 17 48 56.42
7 to 15 9.91 196 19 31 7.67
15 to 30 20.54 85 17 16 0.15
30 to 100 43.15 46 20 21 0.12
0 to 100 9.58 773 74 116 64.36
2  64.360 P(4 df)  0.00 Evidence of lack of fit
(b) Ruptured abdominal aortic aneurysms †
Mortality: physiology data only
0 to 30 22.61 30 7 11 3.39
30 to 40 35.01 34 12 16 2.17
40 to 50 45.04 23 10 11 0.07
50 to 100 64.93 35 23 19 1.74
0 to 100 42.44 122 52 57 7.37
2  7.369 P(4 df)  0.12 No evidence of lack of fit

Mortality: physiology data only

0 to 30 24.20 24 6 10 3.99
30 to 40 34.22 31 11 11 0.02
40 to 50 44.54 24 11 13 0.90
50 to 100 64.33 43 28 23 2.20
0 to 100 44.89 122 55 57 7.12
2  7.117 P(4 df)  0.13 No evidence of lack of fit

*Using the V-POSSUM score, statistical analysis shows ‘lack of fit’, i.e. the method does not predict outcome accurately.
†
Analysed using Portsmouth ruptured AAA equation (Portsmouth data excluded from VSGBI data). There is no
‘lack of fit’, i.e. the method predicts outcome accurately.
POSSUM, physiological and Operative Severity Score for the EnUmeration of Mortality and Morbidity.

These results were translated into a second phase. It was surgeons.24 The analysis has shown that it is possible to
decided to refine the data collection to include just three predict outcomes for individual surgeons performing quite
index vascular procedures – carotid endarterectomy, aortic small numbers of index procedures; data from the 1999
aneurysm repair (divided into ruptured and non- registry indicate that the average VSGBI member performs
ruptured) and infrainguinal bypass. Participating vascular 71 index procedures per annum (18 carotid endarterec-
surgeons were asked to include all patients having these tomies, 25 aneurysms (9 ruptured, 16 non-ruptured) and
procedures under their care. A decision was made to col- 28 infrainguinal bypass procedures). The best models for
lect data only electronically using the POSSUM dataset. predicted versus expected outcome were gained by using
Although electronic data collection has limited the number procedure specific databases and formulae. For example,
of surgeons able to participate thus far, a package of formal POSSUM analysis worked best for ruptured
supporting devices including an Access database, Excel aneurysms using a specific POSSUM regression equation
spreadsheet and bespoke data collection software has been developed in Portsmouth15 (Table 9A.1). Mortality for the
made available to encourage them. The most recent work 122 ruptured aneurysms in this latest series was 47 per cent.
involved over 12 000 procedures collected from 149 VSGBI It was also possible to produce comparative data for all
 References 109

SMRs for AAAs using V-POSSUM physiology

and operative probability as the predictor for than those commonly quoted in the scientific literature.
consultant performing 14 procedures (n  371) These figures are often used to justify surgical proce-
5 dures. Accurate data about the general risks of surgery can
4 help place the value of a procedure to society. Accurate
data about the specific risks of surgery to an individual
3 are crucial to obtaining informed consent from a patient.
SMR

1
Key references
0
H8C1

H13C1

H2C2

H51C3

H0C1

H6C4

H7C2

H6C19

H1C1

H0C2

H6C18

H7C1

H13C2
Bown MJ, Sutton AJ, Micholson ML, et al. A meta-analysis of 50
years of ruptured abdominal aortic aneurysm repair. Br J Surg
Consultant 2002; 89: 714–30.
O/E SMR  1 Neary WD, Crow P, Foy C, et al. Comparison of POSSUM scoring
and the Hardman Index in selection of patients for repair of
ruptured abdominal aortic aneurysm. Br J Surg 2003; 90:
Figure 9A.1 Data from the Pilot National Vascular Database
421–5.
of the Vascular Society of Great Britain and Ireland (2001).
Elfstrom J, Stubberod A, Troeng T. Patients not included in medical
Standardised mortality ratios (SMRs) for individual consultant
audit have a worse outcome than those included. Int J Qual
vascular surgeons who have entered at least 14 procedures
Health Care 1996; 8: 153–7.
onto the Database. Includes both ruptured and non-ruptured
Prytherch DR, Ridler BMF, Beard JD, Earnshaw JJ on behalf of the
aneurysms. If the vertical error bars include the SMR of one, the
Audit and Research Committee of the Vascular Surgical
individual surgeon’s results are not significantly better or worse
Society of Great Britain and Ireland. A model for national
than the whole group. AAA, abdominal aortic aneurysm
outcome audit in vascular surgery. Eur J Vasc Endovasc Surg
2001; 21: 477–83.
aneurysms, both ruptured and non-ruptured, using Bayes The Vascular Society of Great Britain and Ireland. Fourth National
analysis of the same dataset. Cardiac surgeons in the UK Vascular Database Report 2004. Oxford: Dendrite Clinical
currently publish an annual report including outcomes Systems, 2005.
after cardiac surgery,25 and this could soon also be done by
vascular surgeons (Fig. 9A.1).
REFERENCES
Conclusions
1 Baird RN, Baker AR, Hine C, et al. Interhospital provision of
emergency vascular services for a large population: early
Results of emergency vascular procedures are worse than
outcomes and clinical results. Br J Surg 2001; 88: A620–1.
for elective operations. It might be anticipated that this 2 Cassar K, Godden DJ, Duncan JL. Community mortality after
is where the skill of a vascular surgeon could make a real ruptured abdominal aortic aneurysm is unrelated to the
difference. However, these are often sick patients, near- distance from the surgical centre. Br J Surg 2001; 88: 1341–3.
ing the end of life and decisions as to whether to inter- 3 Clason AE, Stonebridge PA, Duncan AJ, et al. Acute ischaemia
vene or not are complex. More data are required so that of the lower limb: the effect of centralizing vascular surgical
patients and their surgeons can make informed choices services on morbidity and mortality. Br J Surg 1989; 76: 592–3.
about management. 4 Basnyat PS, Biffin AHB, Moseley LG, et al. Mortality from
A good surgeon guide has already been published in a ruptured abdominal aortic aneurysm in Wales. Br J Surg 1999;
UK national newspaper using nationally available statis- 86: 765–70.
5 McCollum PT, da Silva A, Ridler BD, de Cossart L. Carotid
tics. Cardiac units were ranked by outcome. Indeed, it
endarterectomy in the UK and Ireland: audit of 30 day
has also been possible to do this analysis for aortic sur-
outcome. The Audit Committee for the Vascular Surgical
gery.26 Unless surgeons are closely involved in data col- Society. Eur J Vasc Endovasc Surg 1977; 14: 386–91.
lection and analysis, comparison among surgeons and 6 Wylie EJ, Hein MF, Adams JE. Intracranial haemorrhage
units will be done using crude mortality rates based on following surgical revascularization for treatment of acute
unreliable hospital information systems. In the UK, sur- strokes. J Neurosurg 1964; 21: 212–15.
geons can now submit their data for central comparison 7 Mead GE, Murray H, Farrell A, et al. Pilot study of carotid
against their peers and expect to receive a formal endarterectomy for acute stroke. Br J Surg 1997; 84: 990–2.
acknowledgement from their Vascular Society. 8 Stewart AHR, Cole SAE, Smith FCT, et al. Re-operation for
The system should prove invaluable for surgeons who neurological complications following carotid endarterectomy.
wish to pass this information on to their patients. It is Br J Surg 2003; 90: 832–7.
9 Campbell WB. Mortality statistics for elective aortic aneurysms.
suspected that the results of some procedures are worse
Eur J Vasc Surg 1991; 5: 111–13.
110 Outcomes of emergency vascular procedures: British Isles

10 Bown MJ, Sutton AJ, Micholson ML, et al. A meta-analysis an audit by the Vascular Surgical Society of Great Britain and
of 50 years of ruptured abdominal aortic aneurysm repair. Ireland. Br J Surg 1998; 85: 1498–503.
Br J Surg 2002; 89: 714–30. 20 Campbell WB, Ridler BM, Szymanska T on behalf of the Audit
11 The Vascular Surgical Society of Great Britain and Ireland. Committee of the Vascular Surgical Society of Great Britain
National Outcome Audit Report 1999. Oxford: Dendrite and Ireland. Two year follow-up after acute thromboembolic
Clinical Systems, 1999. limb ischaemia: the importance of anticoagulation. Eur J Vasc
12 Bradbury AW, Adam DJ, Makhdoomi KR, et al. A 21-year Endovasc Surg 2000; 19: 169–73.
experience of abdominal aortic aneurysm operations in 21 Earnshaw JJ, Whitman B, Foy C on behalf of the Thrombolysis
Edinburgh. Br J Surg 1998; 85: 645–7. Study Group. National Audit of Thrombolysis for Acute Leg
13 Hardman DT, Fisher CM, Patel MI, et al. Ruptured abdominal Ischaemia database: final clinical analysis. Br J Surg 2003;
aortic aneurysms: who should be offered surgery? J Vasc Surg 90: A504.
1996; 23: 123–9. 22 Elfstrom J, Stubberod A, Troeng T. Patients not included in
14 Neary WD, Crow P, Foy C, et al. Comparison of POSSUM scoring medical audit have a worse outcome than those included. Int
and the Hardman Index in selection of patients for repair of J Qual Health Care 1996; 8: 153–7.
ruptured abdominal aortic aneurysm. Br J Surg 2003; 90: 421–5. 23 Prytherch DR, Ridler BMF, Beard JD, Earnshaw JJ on behalf of the
15 Prytherch DR, Sutton GL, Boyle JR. Portsmouth POSSUM models Audit and Research Committee of the Vascular Surgical Society
for abdominal aortic aneurysm surgery. Br J Surg 2001; 88: of Great Britain and Ireland. A model for national outcome audit
958–63. in vascular surgery. Eur J Vasc Endovasc Surg 2001; 21:
16 Beard JD. Screening for abdominal aortic aneurysm. Br J Surg 477–83.
2003; 90: 515–16. 24 The Vascular Society of Great Britain and Ireland. Fourth
17 Hinchcliffe RJ, Braithwaite BD, Hopkinson BR. The endovascular National Vascular Database Report 2004. Oxford: Dendrite
management of ruptured aortic aneurysm. Eur J Vasc Endovasc Clinical Systems, 2005.
Surg 2003; 25: 191–201. 25 The Society of Cardiothoracic Surgeons of Great Britain and
18 Davies B, Braithwaite BD, Birch PA, et al. Acute leg ischaemia in Ireland. National Adult Cardiac Surgical Database Report
Gloucestershire. Br J Surg 1997; 84: 504–8. 2000–2001. Oxford: Dendrite Clinical Systems, 2001.
19 Campbell WB, Ridler BMF, Szymanska TH on behalf of the Audit 26 Rigby KA, Palfreyman S, Michaels JA. Performance indicators
Committee of the Vascular Surgical Society of Great Britain and from routine hospital data: death following aortic surgery as a
Ireland. Current management of acute leg ischaemia: results of potential measure of quality of care. Br J Surg 2001; 88: 964–8.
 9B
Outcomes of Emergency Vascular Procedures in
Scandinavia

WILLIAM P PAASKE, MAURI LEPÄNTALO, THOMAS TROËNG

Introduction: vascular surgery and vascular registries 111 Surgery for acute ischaemia 115
in Scandinavia Vascular trauma 117
Emergency aneurysm surgery 113 References 120

INTRODUCTION: VASCULAR SURGERY AND Denmark

VASCULAR REGISTRIES IN SCANDINAVIA
WORKFORCE
Norway, Finland and Sweden are countries characterised Vascular surgery was recognised as a monospecialty in
by widely varying regional population densities, which has 1983. Currently, serving in the 11 services of vascular sur-
implications for the organisation of vascular surgical ser- gery, are 41 specialists in vascular surgery employed as
vices. That is less true of Denmark where the population is consultants supported by four staff specialists and 12
more evenly distributed. Another Scandinavian feature is senior registrars or vascular fellows (see Chapter 1B).
the ageing population: in Denmark on New Year’s Day
2001, 1.06 million (19.8 per cent) were over 60 years of age VASCULAR REGISTRY
with a projected increase of 18.5 per cent to 1.25 million by
2010. The corresponding figures for those over 65 years of The Danish Vascular Registry (Karbase) has been oper-
age in Finland are 0.77 rising to 0.90 million by 2010 and ational since 1990. Complete national data are available
1.35 million by 2030. Table 9B.1 presents selected data on from 1996 onwards: in 1996, 5798 procedures were carried
demographics and organisation of vascular registries. out of which 4129 (71.2 per cent) were primary operations
and 484 (8.3 per cent) reoperations. The corresponding
figures for 2000 were: 6508 procedures, 4106 (63.1 per cent)
Table 9B.1 Vascular surgery in Denmark, Finland and Sweden primary and 545 (8.3 per cent) reoperations, which trans-
lates into a frequency of 77 primary procedures in a popu-
Denmark Finland Sweden lation of 100 000 (77/105).
Arterial reconstructions are not performed outside vas-
Population (million) 5.35 5.19 8.9
Vascular centres 11 23 50
cular surgical departments and in 2000 were responsible
Vascular surgeons * 41 42 130 for 891 venous procedures. During that year one depart-
Registrars 12 12 NA ment with a single surgeon performed 49 procedures while
Monospecialty 1983 1999 – other institutions, staffed with three to five consultants,
Registry initiated 1990 1989 1985 had an arterial and endovascular surgery volume ranging
Registry covering 1996 1991–1995 1994 from 303 to 998. Also during 2000, Karbase registered
the whole country 1107 percutaneous transluminal angioplasties (PTAs) and
192 arterial thrombolysis treatments. The number of index
* In Finland mostly thoracic and vascular surgeons; in Sweden general
surgeons with a vascular profile; in Denmark all are specialists in vascular procedures in 1999 as defined by the European Board
surgery. of Vascular Surgery1 has been published,2 the contribu-
NA, data not available. tors being those with established registries, namely the
112 Outcomes of emergency vascular procedures: Scandinavia

Scandinavian countries, Northern Ireland, Slovakia, New 1999 data from registries and other sources 11 835 oper-
Zealand and single regions in Spain and Russia. ations were performed for peripheral vessels, both arteries
and veins, and 4846 day-case procedures for simple venous
problems. As for endovascular activity, 4160 procedures
Finland were recorded in 2000, 292 of these being thrombolysis
treatments.
WORKFORCE
Finland has 21 central hospital regions served by five uni-
Sweden
versity and 16 central hospitals. Within central hospital
regions there are a number of district hospitals but their
WORKFORCE
role as active vascular units is diminishing; currently,
23 hospitals provide an arterial surgery service. Vascular surgery used to be performed in 92 units, but the
In conjunction with harmonising Finnish specialist number of vascular centres has fallen to 50. Sweden has
training within the European Union, the previously allied nine university hospitals and four of these have separate
specialty of thoracic and vascular surgery was divided into departments of vascular surgery or combined cardiovascu-
independent specialties, namely, vascular surgery and car- lar departments. The 22 county hospitals all have vascular
diothoracic surgery. The vascular workload is borne by surgical units. There is also vascular activity in some 15 of
36 thoracic and vascular surgeons, 29 of them with a strong the smaller district hospitals. Vascular surgery is not a spe-
vascular profile, six formally trained vascular surgeons, cialty in its own right in Sweden as it is in Denmark and
three staff specialists and 12 vascular surgical trainees. Finland; it remains the responsibility of 130 general sur-
In smaller centres most of these surgeons also perform lung geons with a vascular profile of whom around 30 are
surgery while others also practice general surgery. At the accredited as vascular surgeons by the Swedish Society for
end of year 2003 there were six formally trained vascular Vascular Surgery.
surgeons and 12 vascular surgical trainees. The anticipated
population growth by 2010 will demand an estimated VASCULAR REGISTRY
increase in the number of vascular surgeons to 60 and an
expansion of resources required for the diagnosis and Sweden has pioneered the development of vascular regis-
treatment of venous disease. tries in Scandinavia. Inspired by American experiences
during the 1970s the planning for a regional vascular registry
began in southern Sweden in late 1985.5
VASCULAR REGISTRY
The aims of the registry were to monitor practice,
The Finnvasc registry commenced in 1989 and a pilot changes and outcomes of vascular surgery undertaken in
scheme progressing to systematic data collection was initi- the routine care of a population, to safeguard a high level of
ated in 1991. All hospitals with a vascular surgical service data quality, to establish a basis and a starting point for sci-
participated fully in this exercise up to 1995, but thereafter entific studies, and to develop techniques for quality devel-
their involvement fluctuated between 68 and 81 per cent. opment. An additional aim was to compare the results of
The diminution in interest is attributable to the extra work- emerging vascular units in county hospitals with those of
load for which there was no remuneration, no support established centres at university level.
from health authorities and data collection and reporting After a year of preparation, including a month of pilot
were delayed. Most significantly, the very strict interpret- testing, the Vascular Registry in Southern Sweden (VRISS)
ation of the new personal registry law at the beginning of started regular activities in January 1987 in the South Health
2000 paralysed this important activity. At present, registra- Care Region of Sweden and some neighbouring areas. The
tion practices are being reviewed with the aim of establish- initial 17 participating centres, including two university,
ing computerised real time recording and reporting systems, six county and nine district hospitals, covered a population
at least at local level. of 1.9 million.
In comparing Finnvasc registry figures with those from Over the years increasing numbers of surgeons found it
hospital records the mean percentage of missing vascular useful to participate but by 1990, when half the country’s
procedures in the former was 19 per cent.3 Similarly, a surgeons had joined in, the registry was renamed the Swedish
comparison of data from Statistics Finland, the national Vascular Registry (Swedvasc). From early 1994 surgeons at
statistical bureau, and the Finnvasc registry showed that all 50 hospitals performing peripheral vascular surgery were
18 per cent of patients operated for ruptured abdominal members of the registry, covering practically the whole
aortic aneurysm (RAAA) and subsequently died were Swedish population of 8.8 million and included the 10
missing from the Finnvasc registry.4 university/regional centres and the 25 county hospitals in
The vascular units were staffed with between one and the country. Around 130 surgeons take part regularly
eight consultants with an annual arterial and endovascular in this completely volunteered, professional and non-
throughput ranging from 12 to 1755. According to the authoritarian exercise. The day to day work of the registry
 Emergency aneurysm surgery 113

is led by a steering committee representing both academic Table 9B.2 Emergency aneurysm surgery in Scandinavian
and community hospitals. Each year more than 9000 pro- vascular registry data
cedures are registered and, to date, close to 100 000 cases
Denmark Finland Sweden
have been registered.
Data from years 1996–2000 1991–1997 1996–2001
Number of patients 1242 1152 1905
EMERGENCY ANEURYSM SURGERY Mortality (per cent)
Symptomatic but NA 11.6 7.6
not ruptured
Denmark RAAA/stable NA 24.8 22.0
RAAA/shock NA 59.8 45.7
During the 5-year period 1996–2000, a total of 1242 30-day mortality 43 46 38
patients were admitted to Danish specialist departments Total hospital 55 * 68 * NA
mortality
with RAAA and underwent surgery. This corresponds to
an operation frequency of 4.9 per 105 inhabitants of all * Includes 2.5 per cent of non-operated patients from Denmark and
ages per year. The male:female ratio was 7:4 (1094 men, 30 per cent non-operated patients from Finland.
88.1 per cent; 148 women, 11.9 per cent) and the median NA, data not available; RAAA, ruptured abdominal aortic aneurysm.
age was 72 years (range 29–91). Although a total of 1276
patients were recorded with this diagnosis, most centres
14 per cent, pneumonia or atelectasis 12.5 per cent, assisted
had excluded patients who had not been operated. Not all
ventilation for more than 2 days 9.9 per cent, double or
patients with RAAA were admitted to hospitals with vascu-
higher rises in creatinine 9.5 per cent, dialysis required 8.0
lar surgical departments, especially those in whom the
per cent and intensive care unit stay of more than 3 days
diagnosis was unclear, and there were cases where, in con-
11.1 per cent. A ‘re-do’ or additional surgical procedure or
sultation with a vascular surgeon, the emergency pro-
arterial reconstruction was necessary in 17.7 per cent. Half
cedure was deemed unreasonable or hopeless.
of the patients were discharged to their own homes and the
The circumstances associated with the admission of a
other half to another institution.
RAAA do not permit detailed preoperative questioning of
risk factors, comorbidities and social circumstances, and in
around a quarter or more of all admissions this information Finland
is not available. Given that limitation the data elicited were
as follows: cerebrovascular history 29.6 per cent, hyperten- In Finland, the reporting format allowed differentiation
sion 52.3 per cent, cardiac problems 51.3 per cent, pul- between three categories of emergency AAAs: emergency,
monary complaints 41.5 per cent, smoking 9.3 per cent, stable with rupture and unstable or in shock with rupture. In
retired, old age pensioners or on sick leave 66.0 per cent, the seven years from 1991 to 1997, the 30-day mortality rates
previous vascular surgery 8.9 per cent, previous amputa- were as follows: 474 symptomatic cases without confirma-
tion 1.1 per cent. tion of rupture, 55 deaths (11.6 per cent); 270 with stable
Statistics on the operation itself were as follows: a mid- rupture, 67 deaths (24.8 per cent); 418 patients with shock
line laparotomy was used in 91.4 per cent, duration of pro- due to rupture, 250 deaths (59.8 per cent) (see Table 9B.2).
cedure was 164 minutes (maximum 480 minutes), blood The exact frequency of the various conditions expressed in
loss 4.6 L (maximum 42 L), hospital stay 10.3 days (range per cent per population unit per year, however, is difficult to
0–119). In terms of outcome these were the figures: total estimate for this 7-year period due to decreasing and varying
hospital mortality 55 per cent, 30-day operative mortality reporting activity from 1996 and 1997. Based on the 1995
43 per cent (Table 9B.2). The causes of death were judged figures, which cover the entire country, the frequency of
to be: cardiac 15.3 per cent, cerebrovascular 0.5 per cent, procedures per 105 inhabitants per year for the three presen-
renal 1.1 per cent, haemorrhage 8.6 per cent, multiorgan tations was 1.6 symptomatic non-ruptured, 0.9 ruptured
failure 10.5 per cent, bowel ischaemia 1.2 per cent, other and stable, and 1.2 ruptured and shock, or 3.7/105 in all.
1.8 per cent, unknown 15.7 per cent. A detailed analysis of mortality in RAAA for the years
The complications directly related to the operation were: 1991–94 was undertaken as a cross-sectional study based
amputations 1 per cent, wound infection 6.2 per cent (deep on Finnvasc and the national cause of death registry.4
1.8 per cent), bleeding necessitating a further procedure 7.5 A total of 610 emergency repairs for RAAA were identified
per cent, wound dehiscence 3.3 per cent, intestinal obstruc- in Finnvasc corresponding to 2.9 procedures per 105
tion 1.1 per cent, bowel ischaemia requiring surgery 5.7 per inhabitants per year (2.9/105 per year). Of these, 454 opera-
cent, surgery for peripheral embolisation 2.3 per cent. In all, tions were for rupture (2.2/105 per year) and 156 for emer-
23.2 per cent of patients had a complication demanding gency cases without rupture (0.8/105 per year). In addition,
operative treatment. General complications observed were: it was possible to identify 293 operations for rupture in an
cardiac (myocardial infarction, pump failure, arrhythmia) additional 18 of the 23 hospitals performing operations for
114 Outcomes of emergency vascular procedures: Scandinavia

rupture. The true frequency of surgery for rupture was Mortality within 30 days (per cent)
found to be around 3.6–3.8/105 per year, accounting for 75 60
per cent of the activity in Denmark. A projected increase of 50
at least 50 per cent in RAAA interventions can be expected
in Finland during the next two decades.6 40
The 30-day postoperative mortality as registered in 30
Finnvasc was 46 per cent after surgery for rupture and 13.5
20
per cent for acute non-ruptured cases. Based on national
statistics, the mortality was 54 per cent. Total hospital 10
mortality including all patients brought alive to the emer- 0
gency unit was 68 per cent. The total number of deaths 1994 1995 1996 1997 1998 1999 2000
from rupture based on national statistical information was Non-ruptured Rupture, no shock
1004, and as 245 had survived rupture, the overall survival Rupture, shock
frequency was 19.6 per cent. The death rate of rupture was
4.9/105 per year. If the 1.2/105 per year patients surviving Figure 9B.1 Mortality within 30 days in emergency abdominal
from Finnvasc are added, the total incidence of RAAA aortic surgery in Sweden
comes to 6.1/105 per year. It is noteworthy that there was
no correlation between hospital volume and operative
mortality in rupture repair, but there was an inverse asso- Table 9B.3 Characteristics of patients operated for aortic
ciation between hospital volume and total hospital rupture aneurysms as emergency procedures in Sweden. Except for age all
mortality. The operative activity of RAAA ranged from 37 data are percentages
to 88 per cent among Finnish centres.4
The quality of life (QoL) after survival following repair Emergency RAAA
of a RAAA has been an issue. A recent study by Korhonen not
ruptured stable shock
et al.7 showed that survivors after repair of RAAA had
almost the same QoL as the norms of an age and sex adjusted
Males 78.1 84 85.6 P  0.002
general population. This further justifies an aggressive
Median age,
operative policy in RAAA. The Glasgow Aneurysm Score, years
as a potential predictor of the immediate outcome after Males 72 73 74 NS
surgery for RAAA, could only provide information which Females 76 77 77 NS
is supplementary to clinical decision making.8 Cerebrovascular 15.3 11.6 13.6 NS
disease
Diabetes 6.4 6.4 5.8 NS
Sweden Cardiac disease 51.4 45.4 44.8 NS
Hypertension 46.1 51.7 38.9 001
Hyperlipidaemia 6.1 4.6 4.3 NS
The total number recorded in Swedvasc increased rapidly
Pulmonary disease 16.0 19.9 13.1 P  0.003
during 1987–93, but during 1994–98 that figure stayed at a
Renal disease 9.7 10.6 11.4 NS
fairly constant level with the 1999 data. It is noteworthy that Previous vascular 16.3 11.2 9.7 P  0.002
since Swedvasc began, emergency procedures, as a propor- surgery
tion of all operations for AAAs, have, with some regional Recorded smoker 31.0 34.9 25.0 P  0.001
variations, remained unchanged at around 40 per cent.
Interestingly, postoperative mortality has not improved sig- NS, not significant. RAAA, ruptured abdominal aortic aneurysm;
nificantly since 1994 when all centres had become partici-
pants of Swedvasc (Fig. 9B.1).
The clinical features and outcomes of emergency surgery graft 12.5 per cent, aortobifemoral graft 20 per cent; for
for AAAs registered during 1996–2001 are presented in rupture without shock it was 20 per cent and for the three
detail in Tables 9B.2–9B.6. For two decades AAAs in Sweden types of graft 19.1, 11.5 and 45.5 per cent, respectively; for
treated by surgery have been classified according to rupture with shock it was 40.1 per cent and the breakdown
Eriksson.9 Elective procedures can be done for asymptomatic was 37.6, 34.9 and 54.2 per cent, respectively. It is obvious
or for symptomatic aneurysms. Emergency procedures are that the clinical presentation has an important bearing on
done for non-ruptured, ruptured without shock, or rupture the outcome and this dimension of case-mix ought to be
including shock. These five groups differ considerably in included in comparisons between centres.
terms of postoperative 30-day mortality. Patient characteristics differ little among these groups
The 30-day mortality for non-ruptured cases was but in the more urgent cases such data are often missing;
8.7 per cent and it rose with the degree of complexity of the interestingly, women seem to arrive in time for surgery
graft implanted: aortic tube graft 6.3 per cent, aortoiliac before AAA rupture more frequently than do men.
 Surgery for acute ischaemia 115

Table 9B.4 Complications after emergency aneurysm surgery in Table 9B.6 The age distribution of 30-day mortality
Sweden as related to the severity of the preoperative state (per cent) expressed in per cent in Sweden, 1999

Emergency RAAA RAAA

Age
not
stable shock (years) No rupture stable shock
ruptured
50–59 33.3 10.0
Any surgical 15.8 24.3 32.4 P  0.0001
complications 60–69 10.5 20.8 20.0
Any general 25.4 41.5 70.4 P  0.001 70–79 5.0 15.0 46.2
complication 80–89 33.3 27.8 55.0
Any non-vascular 5.9 9.3 14.3 P  0.001
reoperation RAAA, ruptured abdominal aortic aneurysm.

RAAA, ruptured abdominal aortic aneurysm.

These outcomes were further affected by age (Table 9B.6).
Table 9B.5 Complications after emergency surgery in Sweden The results are better than those in other Scandinavian
reports, but the total hospital mortality is not available.
Complications Per cent

Most common surgical complications SURGERY FOR ACUTE ISCHAEMIA

Haemorrhage/haematoma 9.0
Intestinal ischaemia 3.2 Denmark
Occlusion/thrombosis 2.5
Superficial infection 2.2 The database structure allows identification of patients who
Wound rupture 1.7 were admitted to the specialist departments and underwent
Deep/graft infection 1.2 surgery for acute ischaemia. During the 5-year period
Distal embolisation 0.9 1996–2000, 2808 patients were admitted 3438 times for
various procedures (Table 9B.7). The male:female ratio
Most common general complications was 0.98 (male 1390 (49.5 per cent):female 1418 (50.5 per
ICU 5 days 14.3 cent)) with a median age of 70 years (2–100) and an inci-
Cardiac 12.7 dence of 11.0/105 per year.
Multiorgan 11.13 In contrast to patients with RAAA, information on risk
Pulmonary 10.8 factors and social history for acute ischaemia was elicited in
Renal 8.8
over 95 per cent of cases: diabetes 15.4 per cent, cerebrovas-
cular disease 17 per cent, hypertension 33.2 per cent, car-
Cerebrovascular 2.2
diac disease 47.2 per cent, pulmonary disease 25.1 per cent,
Sepsis 1.9 previous vascular surgical procedures in 42.5 per cent, pre-
Most common non-vascular reoperations vious amputation 3.7 per cent, smoking 77.6 per cent,
Intestinal resection 3.2 retired or old age pensioners 73.6 per cent. The median
Laparotomy 3.1
delay from admission to procedure was 1.2 days (0–76) and
median length of hospital stay was 7.1 days (0–101).
Fasciotomy 2.8
The most frequent procedure was embolectomy/
Amputation (above knee) 0.9 thrombectomy in 47.2 per cent of cases: 14.8 per cent in
Amputation (below knee) 0.2 previously implanted grafts and 85.2 per cent in native
ICU, intensive care unit. arteries. The 30-day occlusion frequency was 14.8 per cent
in grafts and 10.6 per cent in native arteries. Amputation
was necessary in 5.9 per cent of those with occluded grafts
Complications are common after emergency AAA sur- and 3.8 per cent of those with blocked native arteries. The
gery (Tables 9B.4 and 9B.5), and are clearly associated with other outcomes of graft thrombectomy were: wound prob-
the severity of presentation (Table 9B.4). A targeted analy- lems 11.3 per cent, wound infection 3.5 per cent, vascular
sis of 1999 data disclosed 336 patients who had open surgical complications 2.3 per cent, general complications
repairs, of whom 182 presented with shock, 85 without 8.6 per cent; 30-day mortality was 5.5 per cent. Similarly,
shock and 69 without rupture. The total 30-day mortality outcomes of operations on native arteries were: wound
was 28.6 per cent and the corresponding subgroup mor- problems 4.1 per cent, wound infection 0.9 per cent, vas-
tality rates were: non-rupture 8.7 per cent, rupture without cular surgical complications 2.3 per cent, general compli-
shock 20 per cent, and rupture with shock 40.1 per cent. cations 10.1 per cent; 30-day mortality was 16.6 per cent.
116 Outcomes of emergency vascular procedures: Scandinavia

Table 9B.7 Characteristics of patients treated for acute limb ischaemia in Denmark and Sweden.
The Danish data are from 2808 patients treated during 1996–2000 and underwent 3438 procedures.
The Swedish data are from 7496 primary lower limb procedures carried out during 1987–2000. All
data are percentages except for age

Denmark Sweden

7.8

Number of procedures/ Comparison between

100 000 per year 11.0 Embolism Thrombosis Swedish E v T data

Males 49.5 43.1 55.7 P  0.0001

Median age 70 years 80 years 75 years P  0.0001
Cerebrovascular disease 17 26.4 17.7 P  0.0001
Diabetes 15.4 16.6 19.5 P  0.002
Cardiac disease 47.2 76.6 55.9 P  0.0001
Hypertension 33.2 35.0 37.8 P  0.014
Hyperlipidaemia – 3.0 4.5 P  0.0001
Pulmonary disease 25.1 10.3 11.2 NS
Renal disease – 5.7 7.1 P  0.013
Previous vascular surgery 42.5 14.8 39.6 P  0.0001
Previous amputation 3.7 – –
Recorded smoker 77.6 15.1 34.7 P  0.0001

NS, not significant.

Arterial thrombolysis was the second most common occlusion are old and often have other complicating dis-
procedure accounting for 488 treatments (13.4 per cent) eases. The traditional distinction between embolic and
with the following outcomes: 30-day occlusion 13.1 per cent, thrombotic occlusion has been questioned but the clinical
amputations 11.3 per cent, wound complications 6.8 classification used by surgeons seems to define disparate
per cent, wound infection 0.4 per cent, vascular surgical populations with differences between countries.
complications requiring surgery 3.5 per cent, general com- The frequency of occurrence of risk factors in patients
plications 8.4 per cent; 30-day mortality was 4.9 per cent. defined as having had an embolus on the one hand and
Percutaneous transluminal angioplasty was the primary thrombosis on the other differs significantly in several
procedure in 4.8 per cent carrying a 30-day occlusion rate aspects (see Table 9B.7).
of 14.9 per cent, an amputation rate of 10.3 per cent and a The classic treatment for acute ischaemia is an emer-
30-day mortality of 4.6 per cent. gency thromboembolectomy. Around 20 per cent of
arterial occlusions occur in the upper extremity, and a
Finland small proportion involve the visceral arteries; in both
instances surgical embolectomy still appears to be the first
A total of 2108 surgical revascularisations were performed option. The choice of procedure in the lower extremity,
for acute ischaemia in Finland during 1991–97 with a however, has changed considerably during the past decade
mean annual throughput of 387 cases with complete data with endovascular techniques, notably thrombolysis, being
(7.6/105 per year). These cases represented 9.8 per cent of chosen increasingly (Fig. 9B.2). During the five years
the total vascular surgical workload. At the same time only 1996–2000, 28 per cent of the procedures used were
363 endovascular procedures, mainly thrombolysis, were endovascular with a significant difference between indica-
performed which represent 3.2 per cent of the endovascular tions for treatment: embolism 16.9 per cent, thrombosis
workload. The use of thrombolysis, however, has increased 34.7 per cent. It also appears as if different therapeutic
threefold during the last five years. Operations most often techniques were chosen for different patient groups (Table
missing from the Finnvasc database were emergencies and 9B.8). According to an analysis using logistic regression of
endovascular procedures;3 detailed analysis of Finnish data risk factor differences, the endovascular technique was
on acute ischaemia would therefore be misleading. preferred in younger patients with a lower prevalence of
cerebrovascular disease. In patients with thrombosis the
Sweden picture is more complex in that individual risk factors have
distinct influences, for example, age was less important
Patients treated for acute ischaemia of the leg constitute a het- while a history of previous vascular surgery was of greater
erogeneous group. Most patients treated for acute arterial significance.
 Vascular trauma 117

Techniques used for arterial embolism (per cent) Not surprisingly, outcome in the short term is also dif-
100 ferent between the two groups (Table 9B.9). The 30-day
90 outcome was better (P  0.001) after endovascular inter-
80
vention in managing embolism and thrombosis. Most
70
60 impressive, however, is the difference in mortality: in both
50 groups the risk is at least halved when the endovascular
40 technique is used. Whether this is a true difference in
30 outcome, or simply due to differences in patient selection
20
or degrees of operative trauma, remains to be studied. In
10
0 those patients for whom thrombolysis is considered pos-
sible or suitable, the outcome is certainly not worse than
88

90

92

94

96

98

00
(a)
19

19

19

19

19

19

20
that after open surgery.
Bypass Embolectomy
Other Thrombolysis
VASCULAR TRAUMA
Techniques used for arterial thrombosis (per cent)
100 The incidence and spectrum of vascular trauma has
90
80
changed in Scandinavia during the past 10 years. The rising
70 rate of iatrogenic vascular injuries is directly related to the
60 mounting number of interventional vascular procedures,
50 and, to a lesser degree, of laparoscopic techniques. The
40 majority of blunt injuries are caused by traffic accidents. In
30
20
penetrating trauma the most common wounding agents
10 are knives and shards of glass, gunshot wounds being rare
0 in Scandinavia, whereas in Finland at least, most patients
when injured are under the influence of alcohol.10
88

90

92

94

96

98

00

(b)
19

19

19

19

19

19

20

A striking feature in Scandinavian practice is that most

Bypass Embolectomy surgeons on call handling emergency vascular trauma
Other Thrombolysis have little or no experience in vascular surgery. This is
true of all hospitals with the exception of those in Swedish
Figure 9B.2 Proportions (per cent) of different therapies for (a) and Finnish university hospitals and larger vascular
arterial embolism and (b) arterial thrombosis in Sweden centres.

Table 9B.8 Univariate comparison of risk factors among patients presenting with acute ischaemia and
treated by different techniques in Sweden

Embolism Thrombosis
Endo Open Difference Endo Open Difference

Number 428 2108 1114 1861

Males 49.1 39.5 51.1 53.9
Median age, years 78 81 75 74
Cerebrovascular disease 19.2 26.8 0.001 15.5 18.9 0.02
Diabetes 16.4 14.9 NS 20.8 17.3 0.018
Cardiac disease 76.2 77.7 NS 54.3 56.0 NS
Hypertension 42.5 37.4 0.048 41.0 34.8 NS
Hyperlipidaemia 5.1 3.6 NS 7.1 5.0 0.018
Pulmonary disease 9.1 10.6 NS 10.0 13.6 0.004
Renal disease 6.5 6.0 NS 4.8 8.2 0.0001
Previous vascular surgery 19.6 23.2 NS 39.4 57.3 0.0001
Recorded smoker 18.5 13.5 0.007 33.6 32.6 NS

Endo, endovascular; NS, not significant; open, open surgery.

118 Outcomes of emergency vascular procedures: Scandinavia

Table 9B.9 30-day outcome in patients presenting with acute ischaemia in Sweden (per cent)

Embolism Thrombosis
Open surgery Endovascular Open surgery Endovascular

Alive, improved 65.5 77.1 54.7 69.2

Alive, unchanged 16.9 13.8 20.4 17.6
Alive, amputated 4.2 2.3 10.5 6.6
Dead 13.4 6.8 14.4 6.6
Number 2108 428 1863 1114

Table 9B.10 Mechanisms of vascular injuries in Finland and Sweden

Trauma mechanism (per cent)

No. of
Country Years patients Penetrating Blunt Iatrogenic

Finland 1991–1999 503 39 19 42

Sweden 1987–2000 1000 31 23 46

Denmark (31 v 23 per cent) (Table 9B.10). In Scandinavia, aortic and

inferior vena cava injuries are caused by stabbing rather
Arterial trauma is rare and during the five-year period than by gunshot wounds, in Finland three times more
1996–2000, 140 patients (male 100 (71.4 per cent), female commonly.15,16
40 (28.6 per cent), median age 35 years (3–90)), 57.3 The anatomical location of vascular injuries in Finland
per cent actively employed, underwent 146 treatment pro- and Sweden was similar (Table 9B.11). The relative distri-
cedures, and taking all ages together represented an inci- bution of injuries of vessel injuries was: lower and upper
dence of only 0.5/105 per year; the median hospital stay was extremity arteries 45 per cent and 33 per cent, respectively;
3.6 days and 89.5 per cent survived. great thoracic vessels 4 per cent; abdominal vessels includ-
ing the iliac arteries 10 per cent; cervical vessels 3 per cent;
veins of the limbs and neck 7 per cent.
Finland and Sweden The five most common vascular repairs according to
Finnvasc (n  597) were suture repair or ligation 41 per cent,
The accuracy of emergency case data in the Finnish and interposition graft 21 per cent, end-to-end anastomosis
Swedish vascular registries is limited by underreporting,3 10 per cent, thromboembolectomy 6 per cent, patch repair
a fact especially true of vascular injuries which are not 4 per cent. The corresponding figures for Swedvasc (n  999)
always treated by vascular surgeons. Based on Finnvasc were 32, 33, 9, 6 and 8 per cent, respectively. The mortality
covering the years 1991–99, the annual incidence of vascu- rates for Swedvasc (n  931) were: penetrating trauma
lar injuries was 1.3/105 per year (range 0.9–2.0). During the 1.7 per cent, blunt 1.4 per cent, iatrogenic injuries 4.5 per cent.
past 30 years the total incidence in Sweden had increased The overall mortality in Finnvasc (n  503) was 2.8 per cent.
from 1.1 to 2.26/105 per year, mainly due to the rise in In Finland the majority of patients with vascular injuries
iatrogenic vascular injuries.11 are managed by vascular surgeons in university and central
In many European countries iatrogenic injuries, as a hospital emergency departments, by general surgeons in
proportion of vascular trauma, exceeds 40 per cent.12 In smaller hospitals and by cardiothoracic or vascular sur-
Sweden, the increase in iatrogenic vascular trauma seems geons in larger hospitals. In Finland the historic association
to be associated with the introduction of PTA.13 No iatro- of orthopaedics and traumatology might have hampered
genic injuries were reported in Helsinki in 1985, that being the recognition of vascular injuries and caused delay in
the preangioplasty era,10 but 10 years later the incidence treatment. A new training programme initiated at the
was 42 per cent.12 The incidence of iatrogenic vascular University of Helsinki Medical School is aimed at produc-
injuries in gynaecological interventions was found to be ing general surgeons capable of managing traumatic and
0.76/104 per year for laparotomies and 0.93/104 per year for other surgical emergencies of soft tissues. The trend in
laparoscopies, together responsible for 3 per cent of all vas- Sweden, applying the generally accepted Advanced Trauma
cular injuries in Sweden during the reported period.14 Life Support (ATLS) concept, is to centralise management
The incidence of penetrating trauma exceeds blunt of major trauma in county or university hospitals, a princi-
injuries in both Finland (39 v 19 per cent) and Sweden ple also valid for endovascular procedures. Only university
 Vascular trauma 119

Table 9B.11 Anatomic location of civilian vascular injuries in

data on the fate of patients not reaching the operating
Finland and Sweden. Data are number (per cent)
theatre are omitted from the registries. Overall case
Finland Sweden Total fatality, that is including all patients with RAAA, was
(n  510) (n  998) (n  1508) 80 per cent in Finland,4 67 per cent in Viborg County,
Denmark, and 88 per cent in Sweden.17 Total hospital
Carotid artery 17 (3) 29 (3) 46 (3) mortality, which includes patients entering the emer-
Vertebral artery – 2 (0.2) 2 (0.1) gency room alive, is an objective indicator of the ability
Axillary artery 23 (5) 39 (4) 62 (4) of any given hospital to handle RAAA. Finland’s high
Brachial artery 53 (10) 138 (14) 191 (13)
volume centres with an active repair policy had better
Radial or ulnar 65 (13) 107 (11) 172 (11)
results than did smaller units. Similarly, the Danish
artery
Thoracic great 17 (3) 37 (4) 54 (4) experience18 of improved results with RAAAs reflects
vessel the establishment of vascular centres, centralised emer-
Abdominal great 13 (3) 29 (3) 42 (3) gency vascular services and standard criteria for RAAA
vessel repair. The potential of mass AAA screening in improv-
Iliac artery 33 (6) 83 (8) 116 (7) ing overall mortality is dominating the literature.
Femoral artery 159 (31) 304 (30) 463 (31) Endovascular techniques have improved outcome
Popliteal artery 32 (7) 84 (8) 116 (8) and lowered short term mortality in comparison with
Crural artery 56 (11) 41 (4) 97 (6) open surgery in acute ischaemic states. Whether this is
Venous injuries 42 (8) 90 (9) 132 (9) the effect of patient selection or of thrombolysis, with or
Miscellaneous – 15 (2) 15 (1)
without PTA, remains to be seen. In differentiating
between embolic and thrombotic aetiology Swedvasc
data may be overdiagnosing embolism; this is not
hospitals have specialised vascular surgeons on call, but by attempted in Karbase. From a clinical point of view,
voluntary agreements, surgeons who have vascular experi- acute ischaemia should be assumed to be thrombotic
ence can be summoned to the emergency departments of unless proved otherwise, and that stance is reflected in
county hospitals when vascular trauma is encountered. the increased use of thrombolysis.
In recent years vascular trauma, in some instances, has Almost half of the vascular injuries in Scandinavia are
been the subject of malpractice claims. Most of these cases iatrogenic, affecting older patients with complex coex-
have concerned vascular occlusion or disruption associated isting diseases, and carry a higher mortality (4–5 per cent)
with orthopaedic or endovascular interventions, the injury than injuries from other causes (1–2 per cent). It under-
having been either overlooked or a diagnosis delayed. lines the need for better teamworking between vascular
surgeons and interventional radiologists.

Conclusions
ACKNOWLEDGEMENTS
Scandinavian vascular registries depend on information
voluntarily submitted and managed by committees We are grateful to the following: from Denmark, Leif
under national vascular societies. Contributing vascular Panduro Jensen and Jesper Laustsen, Karbase; from Finland,
surgeons are, in principle, civil servants employed by Juha-Pekka Salenius and Maarit Heikkinen, Finnvasc, and
hospital authorities, and although the legal position of Ari Leppäniemi, Helsinki University Central Hospital for
ownership of and access to the database has to be clari- trauma expertise, Anita Mäkeläfor secretarial assistance;
fied, it remains open to scrutiny by national data inspect- from Sweden, David Bergqvist, Martin Bjö rck, Claes
orates. Although national reporting formats differ, data Forssell, Johan Elfströ
m, Tommy Skau, Lars Norgren, K-G
conforming to index procedures, as defined by the Ljungström.
European Board of Vascular Surgery (Eurovasc),1 can be
extracted. The new collaborative working group, Vascunet,
representing established registries in Europe,2 should
Key references
contribute to a convergence of structures and formats as
essential requirements for scientifically valid compara- Bengtsson H, Bergqvist D. Ruptured abdominal aortic aneurysm:
tive studies. a population-based study. J Vasc Surg 1993; 18: 74–80.
Data collection on risk factors in RAAA are prone to Kantonen E, Lepäntalo M, Brommels M, Luther M, Salenius J-P,
be erroneous and incomplete but that on operative pro- Ylönen K and the Finnvasc Study Group. Mortality in ruptured
cedures and outcome are much more reliable. The true abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 1999;
efficacy of surgery in these cases cannot be assessed if 17: 208–212.
120 Outcomes of emergency vascular procedures: Scandinavia

Kantonen I, Lepäntalo M, Salenius JP, Forström E, Hakkarainen T, 7 Korhonen SJ, Kantonen I, Pettilä V, et al. Long-term survival and
et al. Auditing a nationwide vascular registry – the 4-year health-related quality of life of patients with ruptured
Finnvasc experience. Eur J Vasc Endovasc Surg 1997; 14: abdominal aortic aneurysm. Eur J Vasc Endovasc Surg 2003; 25:
468–74. 350–3.
Paaske WP. Eurovasc Report 1999: Vascular and endovascular 8 Korhonen SJ, Ylonen K, Biancari F, et al. Finnvasc Study Group.
surgical activity in Denmark, Finland, Galicia Region Spain, Glasgow Aneurysm Score as a predictor of immediate outcome
New Zealand, Northern Ireland, Slovakia, St Petersburg Region after surgery for ruptured abdominal aortic aneurysm. Br J Surg
Russia, and Sweden. Eur J Vasc Endovasc Surg 2001; 22: 282. 2004; 91: 1449–52.
Swedvasc. Auditing Surgical outcome. Ten years with The Swedish 9 Eriksson I, Hallén A, Simonsson N, Aberg T. Surgical classification
Vascular Registry-Swedvasc. Eur J Surg 1998; 164(suppl 581): of abdominal aortic aneurysms. Acta Chir Scand 1979; 145:
1–48. 455–8.
10 Lepäntalo M, Tukiainen E, Böstman O, et al. Peripheral
vascular injuries in Helsinki 1985. Finn J Orthop Traumatol 1987;
10: 38–9.
REFERENCES 11 Bergqvist D, Helfer M, Jensen N, Tägil M. Trends in civilian
vascular trauma during 30 years. Acta Chir Scand 1987; 153:
1 Paaske WP. Index operations for procedural activity in vascular 417–22.
surgery. In: Bastounis EA. (ed.) Proceedings of the 13th Congress 12 Fingerhut A, Leppäniemi AK, Androulakis GA, et al. The European
of the European Chapter of the International Union of Angiology. experience with vascular injuries. Surg Clin North Am 2002; 82:
Milan: Monduzzi 1999; 241–4. 175–88.
2 Paaske WP. Eurovasc Report 1999: Vascular and endovascular 13 Bergqvist D, Jonsson K, Weibull H. An analysis of complications
surgical activity in Denmark, Finland, Galicia Region Spain, to percutaneous transluminal angioplasty (PTA) of extremity and
New Zealand, Northern Ireland, Slovakia, St Petersburg Region renal arteries. Acta Radiol 1987; 28: 3–12.
Russia, and Sweden. Eur J Vasc Endovasc Surg 2001; 22: 282. 14 Bergqvist D, Bergqvist A. Vascular injuries during gynecologic
3 Kantonen I, Lepäntalo M, Salenius JP, et al. Auditing a surgery. Acta Obstetr Gynecol Scand 1987; 66: 19–23.
nationwide vascular registry – the 4-year Finnvasc experience. 15 Jousi M, Leppäniemi A. Management and outcome of traumatic
Eur J Vasc Endovasc Surg 1997; 14: 468–74. aortic injuries. Ann Chir Gynaecol 2000; 89: 89–92.
4 Kantonen E, Lepäntalo M, Brommels M, et al. and the Finnvasc 16 Leppäniemi AK, Savolainen HO, Salo JA. Traumatic inferior
Study Group. Mortality in ruptured abdominal aortic aneurysms. vena caval injuries. Scand J Thorac Cardiovasc Surg 1994; 28:
Eur J Vasc Endovasc Surg 1999; 17: 208–12. 103–8.
5 Swedvasc. Auditing surgical outcome. Ten years with The 17 Bengtsson H, Bergqvist D. Ruptured abdominal aortic
Swedish Vascular Registry – Swedvasc. Eur J Surg 1998; aneurysm: a population-based study. J Vasc Surg 1993; 18:
164(suppl 581): 1–48. 74–80.
6 Heikkinen M, Salenius JP, Auvinen O. Ruptured abdominal aortic 18 Lindholt JS, Henneberg EW, Fasting H. Decreased mortality of
aneurysm in a well defined geographical area. J Vasc Surg 2002: abdominal aortic aneurysms in a peripheral county. Eur J Vasc
36: 191–296. Endovasc Surg 1995; 10: 466–9.
 10
Medico-Legal Aspects of Emergency
Vascular Care in the UK

BRUCE CAMPBELL

Introduction 121 The team approach and guidelines 124

The basis of medico-legal actions 122 Omission of prophylaxis in the emergency setting 124
Missed or delayed diagnosis 122 Written records and notes 124
Who should be responsible for 122 Incident reporting 125
emergency vascular problems? Clinical audit 125
Decisions about withholding treatment 123 References 125
Informed consent 123

INTRODUCTION
events observed in vascular surgical practice were poten-
tially preventable. Despite the variation in these figures,
There has been a burgeoning of medico-legal activity in the there is a clear message that many clinical problems after
UK1 and Europe in recent years, which has followed, but by arterial operations may be preventable.
no means caught up with, a similar trend in the USA. This With regard to the incidence of medico-legal claims,
has affected vascular surgery rather less than many other dis- collated figures from the National Health Service Litigation
ciplines such as obstetrics, and these differences are reflected Authority (which has indemnified Health Service hospitals
in professional insurance premiums. The area of vascular in England and Wales since 1995) and from the Medical
work most often associated with medico-legal problems is Defence Union (MDU) (reflecting private practice in the
the treatment of varicose veins,2,3 but such cases are invari- UK from 1990 to 1999) include 36 claims against surgeons
ably elective, rather than emergencies, and have often been for failure to diagnose or treat ischaemia, and 45 for com-
dealt with by surgeons who are not vascular specialists.4 plications of aortic surgery.4 The MDU database for gen-
Arterial surgery is associated with substantial numbers eral practice for the same period contained 299 claims,
of adverse events, and these often follow emergency or urgent all related either to alleged mismanagement of limb
presentations. In the USA, the Harvard Medical Practice ischaemia (n  197) or diagnosis and treatment of aneurysms
Study,5 involving 30 121 hospital records, found a higher (n  102).
proportion of adverse events in vascular surgery, largely arter- In elective arterial practice there is usually ample oppor-
ial, than any other specialty (16 per cent), although a smaller tunity for assessment, counselling and record keeping, which
percentage of these were judged to be due to negligence are particularly important in guarding against subsequent
than in other specialties (18 per cent). A study of 15 000 complaints or litigation.8 Emergency vascular practice poses
hospital admissions in Utah and Colorado,6 documented a greater chance of medico-legal problems: decisions often
aortic aneurysm repair and lower limb bypass grafting as need to be made quickly and without all the information
the two operations with the highest adverse event rates of clinicians would like; there may be little opportunity for
all, 19 per cent and 14 per cent, respectively, with 8 per cent counselling and the immediate writing of good notes; and
and 11 per cent of these adverse events judged as prevent- the potential for loss of life or limb is high. In addition,
able. In Australia, the Quality in Australian Healthcare Study7 these patients frequently present to doctors who are not vas-
on 14 000 admissions claimed that 49 per cent of the adverse cular specialists, and the diagnosis may be missed or delayed,
122 Medico-legal aspects of emergency vascular care in the UK

with serious consequences for the patient and difficulties presentations and may also become targets for the general
for the receiving vascular consultant. dissatisfaction about the management of the patient.
This chapter presents a number of aspects of emergency Certain conditions have special potential for delayed or
vascular practice which may be associated with complaints missed diagnosis.
or litigation, and proposes suggestions about how to avoid Spontaneous acute ischaemia due to embolism or throm-
being sued. It certainly does not provide all the answers, bosis may be confused with other causes of limb pain, for
but hopefully offers food for thought. example sciatica, or may simply not be noticed in uncom-
municative or bedridden patients who are in hospital wards
or nursing homes. Traumatic arterial damage may escape
detection in a setting of multiple trauma or after injuries
THE BASIS OF MEDICO-LEGAL ACTIONS
which are not typically associated with vascular involve-
ment. The swelling and generalised pallor often associated
It is important, briefly, to rehearse the law of negligence, with trauma can make assessment difficult, and a high index
which governs medico-legal action against doctors in the of suspicion is the key to diagnosis. Late complications such
UK. There are three fundamental principles: as false aneurysms can also result in accusations of negli-
gence. Failing to recognise compartment syndromes asso-
1 To establish that the doctor had a duty of care to the
ciated with acute trauma and/or acute ischaemia is a matter
patient: as a rule this poses no problems, although it
of particular concern for vascular surgeons.
may do so in the setting of emergency vascular practice.
The diagnosis of leaking abdominal aneurysm is not
2 To show that the doctor has breached that duty of
infrequently delayed or missed by doctors both in primary
care to the patient: therefore he has been negligent
care and in hospital. Leaking aneurysms are well known to
(liability).
present with atypical features, being easily confused with
3 To prove that the doctor’s actions, or omissions, have
other conditions causing abdominal pain such as ureteric
caused damage to the patient (causation).
colic or diverticular disease, with acute back problems or
A longstanding and important precedent exists from with conditions causing hypotension such as myocardial
the case of Bolam (Bolam v Friern Hospital Management infarction. Aortoenteric fistula is another condition which
Committee 1957), namely, ‘a doctor is not guilty of negli- may be rapidly fatal but which is sometimes not recog-
gence if he acted in accordance with the practice accepted nised: even when suspected in a patient with an aortic graft
as proper by a responsible body of medical men skilled in the diagnosis may be difficult to prove.
that particular art’, but the professional opinion called to If there is concern about the possibility of any of these
support him must be capable of ‘withstanding logical analy- conditions, but the diagnosis is not clear, then good record
sis’ (from Bolitho v City and Hackney Health Authority keeping is fundamental from the risk management point of
1997). Judgments about what was, or was not, likely to view. Written notes that a particular diagnosis was considered,
have happened are based on the balance of probabilities the reasons for uncertainty and the steps taken to investigate,
(greater or lesser than 50 per cent probability) and not on are a powerful defence if that diagnosis was later found to
the principle of ‘beyond reasonable doubt’ used in crim- have been ‘missed’. It is equally important not to accuse or
inal cases. Increasingly in medico-legal judgments, the condemn other doctors in writing after a referral has been
concept is being used of what a ‘reasonable man’ would made with a delayed diagnosis: the full circumstances may
have done rather than simply considering the views of not be known to the receiving specialist, and written accu-
medical specialists. sations of negligence can be the cause of serious and need-
It is important to remember that only a small propor- less problems.
tion of complaints against doctors develop into medico-
legal claims: proceedings are issued in only a few of these
cases, and fewer still go to court. However, the stress and
WHO SHOULD BE RESPONSIBLE FOR
emotional impact of medical errors9 and medico-legal
EMERGENCY VASCULAR PROBLEMS?
threats can be oppressive for those involved, and the time
involved in dealing with the issues may be considerable.
This question has assumed importance with the increasing
specialisation in general surgery. The days have long gone
when acute ischaemia was dealt with by any general surgeon
MISSED OR DELAYED DIAGNOSIS and associated trainees,10 and decreasing numbers of non-
vascular consultants feel competent to operate on leaking
This is a major cause of medico-legal activity. The doctors aortic aneurysms. In addition, there is a rising public expect-
held responsible for missing emergency vascular conditions ation fuelled by the press that all treatment will be delivered by
are most often not vascular surgeons, but vascular surgeons ‘specialists’. When all does not go well with emergency vascu-
may then find themselves presented with difficult late lar cases, ‘generalists’ may become increasingly vulnerable to
 Informed consent 123

medico-legal criticism. The underlying problem, particularly of life. This is an area of potential medico-legal concern,
apparent in the healthcare system of the UK, is one of insuffi- although, in response to a questionnaire in 1998,12 only 22
cient vascular specialists in many hospitals, especially those per cent of vascular surgeons in the UK and Ireland stated
outside large urban areas. The history and reasons for this are that medico-legal concerns ever influenced their decisions
complex, but the result is that many hospitals do not have a about not operating on patients with leaking aortic
formal emergency vascular surgical, or vascular radiology aneurysms. Quite a different response would probably have
rota, and as a consequence general surgeons with little or no been received from surgeons in the USA, where there are
regular experience of elective vascular work may have to take generally higher expectations from both patients and rela-
on emergency cases. tives for heroic treatment of the very elderly with little chance
Where formal vascular rotas do not exist, there may be of survival.
a special medico-legal threat for vascular surgeons, who When deciding on palliative care rather than surgical
frequently make themselves available as much as they are intervention, it is vital that all staff caring for the patient
able for emergencies.11 If they are contacted by telephone are in tune with the decision, and that they are invited to
and are unable or unwilling to attend because they are not voice any concern or disagreement. Not only is this part of
formally ‘on call’, what would be their liability if the out- teamwork and common sense, but it also guards against
come is unfavourable for a patient who is dealt with by any disaffected member of the care team later claiming to
generalist colleagues or trainees, or who is transferred to have been an unwilling party to a wrong or negligent deci-
another hospital? What is their clinical responsibility and sion. The patient’s family must be counselled sensitively,
duty of care if they are not contractually on call? Offering and must concur with the decision, but not be made to feel
advice over the telephone is usually taken to imply clinical responsible for it: it is they who might later take legal action
responsibility, at least to some degree. This is a difficult if the situation was not dealt with well. Whenever practical,
area but one which surgeons and managers need to con- patients themselves should be involved in the decision.
sider. Each hospital should have clear understandings about These discussions should always involve a senior member
which consultant is in charge of the patient’s care at any of the vascular surgical team.13
given time: admitting general surgeons should not assume A particular problem in emergency situations is that lit-
that their vascular colleagues have taken over care of an tle may be known about the patient’s pre-existing medical
emergency admission until this has been explicitly agreed. condition, and relatives may not be immediately to hand;
This is also important medico-legally with regard to the ques- patients with leaking aortic aneurysms who have house-
tion of who supervises trainees dealing with the patient, bound spouses with disabilities are a case in point. When a
particularly when the duty trainees are not part of the vas- decision about palliative care is particularly difficult it may
cular team, currently a common situation. be an advantage to involve a senior anaesthetist, both to
Another potentially difficult medico-legal issue relates give an opinion and to record their opinion in the notes;
to transfer of patients, which may be required if there is no this can be especially helpful when relatives press for active
vascular specialist available in a hospital, or if adequate facil- treatment in a patient whose anaesthetic risks are excessive.
ities are not available, for example, no intensive care bed With sensitive explanation and counselling, as well as
for a leaking aneurysm. Might there be a medico-legal chal- good written records, thoughtful decisions to withhold active
lenge if the treatment of such a patient is delayed, and they surgical treatment will hopefully remain an area which sel-
die or lose a limb as a result? It is a legal principle that hos- dom gives rise to medico-legal action, but the potential for
pitals should not offer services which they cannot provide problems needs to be kept clearly in mind.
to a good standard, and this could be taken to imply that it
is better to transfer an emergency vascular patient, rather
than to treat them in a substandard way. The principles of
INFORMED CONSENT
transfer should be the subject of agreement by health
authorities and hospital trusts, following discussion from a
‘public health’ perspective. Treatment of vascular emergencies is fraught with poten-
tial complications, but the urgency of the situation and the
condition of many patients can make thorough counselling
about risks both impractical and unkind. For example, the
DECISIONS ABOUT WITHHOLDING collapsed patient with a ruptured aortic aneurysm is in no
TREATMENT state to receive any amount of information, except the mes-
sage that he needs major emergency surgery for a leaking
Decisions about withholding active treatment are required artery. The patient’s condition may be so grave that signing
quite often in emergency vascular practice, for example, in of a consent form is not a sensible expectation, but this
patients with leaking aortic aneurysms12 or unsalvageable should never deter surgeons from doing what they believe
acute limb ischaemia,13 who are very elderly, who have ser- to be in the patient’s best interests as this is their duty from
ious, and often multiple, comorbidities and poor quality both an ethical and a legal point of view.
124 Medico-legal aspects of emergency vascular care in the UK

When the situation permits more explicit informed con- OMISSION OF PROPHYLAXIS IN THE
sent, then there is a medico-legal expectation that patients EMERGENCY SETTING
should be told about relevant risks, both very serious prob-
lems which occur occasionally, and lesser problems which
occur more frequently.14–18 Just how fully the frightened eld- It is easy for important prophylactic measures to be forgot-
erly patient should be confronted with any risks beyond the ten in the turmoil of emergency treatment. Prophylactic
threat to an acutely ischaemic limb is a matter of individual antibiotics are especially important when grafting ruptured
judgement. Whatever they, and their close relatives and espe- aortic aneurysms: there can be no defence if they are not
cially carers, are told, it is important that the conversation is given at the time of aortic grafting and the graft becomes
documented and this is dealt with further below. While it is infected. The need is especially great in the context of asep-
very helpful, and I believe essential, to have a good standard tic precautions which may be less rigorous than usual, as is
information booklet about each common operation in the sometimes the case when dealing with collapsed patients,
elective setting, this is more difficult for emergency proced- and extra doses of antibiotics are wise at the end of the
ures. However, when there is the opportunity to give the operation when there has been massive blood loss.
patient a booklet, for example about ‘thrombolysis’ or Failure to provide prophylaxis against venous throm-
‘bypass grafts to the limb’, then this should certainly be done. boembolism has become a regular cause of medico-legal
Occasionally patients may say that they do not want to action. In the treatment of acute ischaemia there may also
be informed in detail about proposed procedures or risks. be a need to consider special measures to prevent further
That is their right, but if they are not informed because of arterial thromboembolism, using anticoagulants or dextran
this kind of request then this should be recorded clearly. It 40. Tetanus prophylaxis should not be forgotten in cases of
should always be the presumption that every patient wants trauma.
to be well informed.

WRITTEN RECORDS AND NOTES

THE TEAM APPROACH AND GUIDELINES
The importance of good records cannot be overempha-
Although the vascular surgeon bears ultimate responsibil- sised in any discussion about medico-legal matters.3,8,18,21
ity for the vascular surgical patient, other specialists and dis- When dealing with vascular emergencies, for example, rup-
ciplines are also important, not only for their contributions tured aneurysms, pressure of time often makes writing of
to management of the patient, but also in the risk manage- thorough preoperative notes impossible. Written notes may
ment process and avoidance of litigation. This is particu- be confined to bare working details when time is pressing,
larly true of vascular radiologists who play the leading role but after an emergency operation there is no excuse for the
in thrombolysis for acute ischaemia, and there needs to be absence of an adequate record of everything that happened.
very clear communication between them and their surgical I always dictate an operation note, which is then typed, and
colleagues about the decision for thrombolysis, who coun- this is a good opportunity to describe the initial presenta-
sels and consents the patient, and their specific responsibil- tion, the discussions and the reasons for decisions. The typed
ities while lysis is in progress. There is a potential for serious note is also a useful place to record personal thoughts about
medico-legal consequences if a patient has a stroke during the case, prognosis and plans, together with what has been
thrombolysis without a record of explicit informed consent, said to the patient’s relatives about the likely outcome. The
or if they develop bleeding which is not dealt with appro- details are not only helpful to those subsequently involved
priately because of poorly defined responsibilities. in the patient’s care, but they are also invaluable if there is
Thrombolysis is a particular area of emergency vascular any subsequent complaint or medico-legal claim.
practice in which written guidelines are very valuable, partic- It is vital for doctors to understand that the records
ularly for trainees and nursing staff, and they should describe they write or dictate provide the only evidence of what was
the action to be taken if problems occur. Many doctors worry thought, said and done. A patient or relatives may make
about the medico-legal disadvantages of guidelines,19 but assertions later which are completely untrue, but if no record
thrombolysis is a good example of a treatment for which exists to contradict them then it is difficult to refute their
their advantages far outweigh any possible disadvantage. It claims. It is worth remembering the cynical point of view:
is relatively complex treatment which involves several dis- ‘If you didn’t write it down, it didn’t happen’. It is tediously
ciplines, with which trainees and less experienced nurses basic, but absolutely mandatory from the medico-legal stand-
may not be very familiar, and which requires efficient moni- point, that each page of the clinical record has the patient’s
toring with a clear understanding of what to do if things go name and reference number on it, that each entry is prop-
wrong. It may be helpful to use published consensus docu- erly dated including the year and not just the month, and
ments20 when constructing guidelines, but for practical pur- that entries, especially for emergency and urgent records,
poses they must reflect local circumstances and preferences. have a time, the 24-hour clock being best.
 References 125

INCIDENT REPORTING Conclusion

All hospitals should now have systems in place for immedi- Emergency vascular work poses many possibilities for
ate reporting of any incidents which might result in com- medico-legal challenge. However, the surgeon who makes
plaint or medico-legal proceedings.18,21,22 If clinicians do thoughtful decisions, who communicates them sympa-
this conscientiously, then problems may be averted or min- thetically to patients and their relatives, and who keeps
imised. Incident reporting is a huge subject in its own right thorough records stands a good chance of avoiding both
but all doctors need to know about their local incident complaints and successful claims.
reporting systems and how to use them.

CLINICAL AUDIT Key references

Campbell B, Callum K, Peacock N. Operating Within the Law.
There are two particular medico-legal implications of clin- Kemberton: tfm publishing, 2001.
ical audit. First, it is probably advantageous to be able to Campbell B, France F, Goodwin H, on behalf of the Research and
report that any case with an adverse outcome was discussed Audit Committee of the Vascular Surgical Society of Great
at a peer group audit meeting. Second, audit of individual Britain and Ireland. An analysis of medico-legal claims in
results may help to support a surgeon’s reputation and vascular surgery for the National Health Service and private
sector in the United Kingdom. Ann R Coll Surg Engl 2002;
credibility in a particular area of work. The National Vascular
84: 181–4.
Database being developed by the Vascular Society of
Fenn P, Diacon S, Gray A, et al. Current cost of medical negligence
Great Britain and Ireland may well come to represent an in NHS hospitals: analysis of claims database. BMJ 2000;
important tool by which vascular specialists can confirm 320: 1567–71.
that their results conform to national norms.23 Contributing NHS Executive. Risk Management in the NHS. London: Department
data about procedures to national and international regi- of Health, 1994.
stries, for example aortic stent grafts used to treat leaking Seeking Patients’ Consent: The Ethical Considerations. London:
aneurysms, is a hallmark of good practice and may help General Medical Council, 1998.
support a clinician, particularly in the context of relatively
new procedures.
The extent to which the results of past performance REFERENCES
should be required or offered in medico-legal proceedings is
a matter for debate. Clinical governance now demands that 1 Fenn P, Diacon S, Gray A, et al. Current cost of medical
such records be made, and it seems inevitable that lawyers negligence in NHS hospitals: analysis of claims database.
will increasingly ask for the figures. Demonstration of regu- BMJ 2000; 320: 1567–71.
lar, careful audit, with appropriate changes in practice is 2 Giordano JM. Malpractice and the vascular surgeon. J Vasc
likely to contribute to the defence to a surgeon accused of Surg 1993; 18: 901–4.
negligence, particularly if the results are good. By the same 3 Goodwin H. Litigation and surgical practice in the UK. Br J
token surgeons are understandably worried that audit might Surg 2000; 87: 977–9.
4 Campbell B, France F, Goodwin H, on behalf of the Research and
also be used against them. Their results might be compared
Audit Committee of the Vascular Surgical Society of Great
unfavourably with published series, which often describe Britain and Ireland. An analysis of medico-legal claims in
much better outcomes than the norm.16 In addition, audit vascular surgery for the National Health Service and private
data might be used to criticise them for doing a specific pro- sector in the United Kingdom. Ann R Coll Surg Engl 2002; 84:
cedure infrequently: for example a surgeon who had had 181–4.
poor outcomes from just two or three cases of vascular 5 Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse
trauma might find this fact used against them despite good events and negligence in hospitalized patients. N Engl J Med
reasons for these results, and a good track record of elective 1991; 324: 370–6.
arterial reconstructions. Although regular audit and involve- 6 Gawande AA, Thomas EJ, Zinner MJ, Brennan TA. The incidence
ment in national registries certainly are important, these con- and nature of surgical adverse events in Colorado and Utah in
siderations may discourage recourse to audit figures as a 1992. Surgery 1999; 126: 66–75.
7 Wilson RM, Runciman WB, Gibberd RW, et al. The Quality in
routine in medico-legal proceedings.
Australian Health Care Study. Med J Aust 1995; 163: 458–71.
Clinicians involved in research on procedures for emer- 8 Baird RN. The vascular patient as litigant. Ann R Coll Surg
gency vascular conditions should take particular care in Engl 1996; 78: 278–82.
advising patients and their relatives, in record keeping, and 9 Wu AW. Medical error: the second victim. BMJ 2000; 320:
in all aspects of research governance. Medico-legal action 276–7.
resulting from an ‘experimental’ procedure requires particu- 10 Nachbur B. Treatment of acute ischaemia: every general
larly robust defence. surgeon’s business? Eur J Vasc Surg 1988; 2: 281–2.
126 Medico-legal aspects of emergency vascular care in the UK

11 Campbell WB, Ridler BMF, Thompson JF. Providing an acute 17 Reference guide to consent for examination or treatment.
vascular service: two years experience in a district general London: Department of Health, 2001.
hospital. Ann R Coll Surg Engl 1996; 78: 185–9. 18 Campbell B, Callum K, Peacock N. Operating Within the Law.
12 Hewin DF, Campbell WB. Ruptured aortic aneurysm: the decision Kemberton: tfm publishing, 2001.
not to operate. Ann R Coll Surg Engl 1998; 80: 19 Hurwitz B. Legal and political considerations of clinical
221–5. practice guidelines. BMJ 1999; 318: 661–4.
13 Campbell WB, Verfaillie P, Ridler BMF, Thompson JF. Non- 20 Working Party on Thrombolysis in the Management of Limb
operative treatment of advanced limb ischaemia: the decision for Ischemia. Thrombolysis in the management of lower limb
palliative care. Eur J Vasc Endovasc Surg 2000; 19: peripheral arterial occlusion – a consensus document. Am J
246–9. Cardiol 1998; 81: 207–18.
14 Marshall JE, Baker PN. Informed consent – legal and ethical 21 NHS Executive. Risk Management in the NHS. London:
issues. Health Care Risk Report 1999: 12–14. Department of Health, 1994.
15 Gladstone J, Campbell B. A model for auditing informed consent. 22 Roberts G. Untoward incident reporting: quality improvement
J Clin Effectiveness 2000; 1: 247–50. and control. Clin Risk 1995; 1: 168–70.
16 Seeking patients’ consent: the ethical considerations. London: 23 Campbell B, Earnshaw J. Getting governance to work in
General Medical Council, 1998. surgery. Ann R Coll Surg Engl 2001; 83(suppl): 56–7.
 SECTION
2
Acute Cerebrovascular Syndromes

11. The developing stroke 129

12. Role of the vascular surgeon in managing stroke 141

13. Surgical experience in evolving stroke arising from the carotid 149

14. Post-carotid endarterectomy stroke 155

This page intentionally left blank
 11
The Developing Stroke

SEBASTIÁN F AMERISO

The problem 129 General management 132

Aetiology/pathophysiology 129 Specific treatment 135
Clinical aspects/diagnosis 131 References 137

THE PROBLEM AETIOLOGY/PATHOPHYSIOLOGY

Stroke is defined as the acute onset of a focal neurological Classification

deficit caused by alterations in blood circulation in the vas-
cular territory of the central nervous system. Cerebral There are two types of stroke: ischaemic and haemor-
ischaemia can also affect the brain diffusely during pro- rhagic. Ischaemic strokes or cerebral infarcts account for
longed systemic hypotension thereby causing extensive approximately 85 per cent of all strokes.3,4
infarcts.
Stroke is the third leading cause of death in North
America and the most common cause of prolonged dis- Stroke classification
ability in the industrialised world.1 The absolute number
of stroke patients is likely to increase in ageing popula- • Ischaemic strokes
tions. Until recently, the management of acute cerebral – Large artery atherosclerosis
infarction was restricted to supportive measures and – Cardioembolism
control of risk factors in secondary stroke preven- – Small vessel occlusion
tion. Acute ischaemic stroke is now viewed as a medical – Other aetiologies (some unknown)
emergency with a narrow therapeutic window clinically • Haemorrhagic strokes
comparable to acute myocardial infarction.2 In the vast – Intraparenchymal haemorrhages: hypertensive;
majority of patients neuronal damage occurs within amyloid angiopathy; other causes, such as
a few hours of the initial event and the term ‘brain tumours, blood dyscrasias, etc.
attack’ emphasises the need for rapid diagnosis and – Subarachnoid haemorrhage: aneurysmal;
treatment. traumatic; other causes
The purpose of this chapter is to discuss the diagnosis – Rupture of arteriovenous malformations
and management of acute ischaemic stroke, beginning in
an emergency room setting. The supportive and ancillary
treatment measures will be underlined and the strategies Athero-thromboembolic or large vessel strokes originate
aimed at reperfusing and protecting the brain while limit- in atheromatous lesions localised within the large cerebral
ing the area of infarction discussed. Special attention will vessels, both extracranial and intracranial. Typically infarcts
also be given to the prevention of common iatrogenic are large cortical and subcortical lesions in the territory of
complications of these patients. one of the major cerebral vessels. Cardioembolic strokes
130 The developing stroke

occur in conditions such as atrial fibrillation, myocardial

infarction, congestive failure, valve disease, congenital mal- • Modifiable risk factors
– Hypertension
formations and aortic arch plaque.5,6 The role of mitral
– Smoking
valve prolapse and patent foramen ovale in embolism
– Diabetes mellitus
remains controversial. Lacunar or small vessel strokes are
– Dyslipidaemia
small subcortical infarcts caused by occlusion of small,
– Cardiac disease
deeper placed vessels affected by the pathological process of
– History of stroke/TIA
lipohyalinosis. Other aetiologies are vasculitis, procoagulant
– Other factors: haemorrheological factors (i.e.
conditions, drug use and venous occlusion.7 Haemorrhagic
elevated haematocrit and fibrinogen), procoagulant
strokes comprise approximately 15 per cent of acute cere-
conditions, hyperhomocysteinaemia, genetic
brovascular events, including intracerebral haemorrhage and
factors, recent infection/inflammation
rupture of aneurysms and arteriovenous malformations.
Strokes can also be classified according to the temporal
evolution of symptoms. Transient ischaemic attacks (TIAs)
are focal neurological deficits with complete resolution Pathophysiology
within 24 hours. In most cases TIAs last less than a minute
or two and certainly well below 1 hour. A reversible The basic mechanisms of cerebral ischaemia are: (i) in situ
ischaemic neurological deficit (RIND) or minor stroke is thrombosis resulting in narrowing of the vessel lumen and
an event lasting more than 24 hours with complete or near reduction in flow or (ii) fragmentation and dislodgement of
complete resolution within 3–7 days. Infarcts with transient a clot in the heart or in an artery with subsequent embolisa-
neurological symptoms (ITNS) cannot be distinguished tion to a cerebral vessel. When blood flow to an area of the
clinically from TIAs, but computed tomography (CT) or brain is reduced to a critical level through either one of
magnetic resonance imaging (MRI) will demonstrate a these mechanisms the result is infarction9 (see Chapter 5).
recent ischaemic lesion in the affected territory. Strokes in
evolution are neurological events that progress after onset. ISCHAEMIC PENUMBRA
The deterioration usually occurs during the initial hours Experimental and clinical evidence have demonstrated the
of the episode and is caused, at least in part, by progression existence of an area surrounding the infarcted tissue in
of the underlying thrombotic process. Completed strokes which blood flow is maintained by collateral circulation at
are permanent neurological deficits without major changes a level below 50 per cent of normal. This viable area,
occurring after the initial assessment. known as the ischaemic penumbra, remains in danger of
sustaining progressive damage and necrosis.10 The fact that
Temporal evolution of acute ischaemic irreversible damage can occur within a short period of time
cerebrovascular events provides the rationale for early treatment of stroke either
by prompt reperfusion or by administering agents which
• Transient ischaemic attacks protect the brain from further damage, in other words
• Reversible ischaemic neurological deficit or pharmacological neuroprotection. The time between the
minor stroke initial insult and the completion of neuronal death in the
• Infarcts with transient neurological symptoms penumbra represents the therapeutic window of opportun-
• Stroke in evolution ity for treatment of ischaemic stroke.11 The duration of this
• Completed stroke time window is probably no longer than 3–6 hours and the
response to different therapeutic modalities may vary from
patient to patient.
Stroke risk factors
CEREBROVASCULAR AUTOREGULATION
Similar to patients with coronary disease, stroke patients
usually have conditions that predispose them to the occur- This is a normal physiological mechanism which allows for
rence of strokes.8 These stroke risk factors can be modifi- constant maintenance of cerebral blood flow (CBF)
able or non-modifiable. through a wide range of perfusion pressures.12 The normal
range of mean arterial pressure (MAP) regulation is
60–130 mmHg and when it goes outside that range
Stroke risk factors autoregulation is lost and cerebral blood flow follows
changes in MAP. Chronic hypertension is associated with
• Non-modifiable risk factors
displacement of the autoregulatory curve to the right.
– Advanced age
Cerebrovascular autoregulation is defective immediately
– Male sex
after a stroke and the ischaemic brain tissue is entirely
– Genetic factors
dependent on available collateral surrounding flow.
 Clinical aspects/diagnosis 131

Cerebral autoregulation normally maintains CBF at a

constant level across a wide range of mean perfusion pressures
(mean arterial pressure 60–150 mmHg). Autoregulation is
impaired following stroke, and ischaemic brain tissue
becomes passively dependent on surrounding collateral
blood supply.12 Rapid reduction of blood pressure to nor-
motensive levels may compromise collateral supply and
precipitate acute neurological deterioration.13

CLINICAL ASPECTS/DIAGNOSIS

Acute ischaemic stroke is a medical emergency. Early evalu-

ation of patients allows the proper planning of manage-
ment strategies. Patients with new onset focal neurological
deficit, either transient or established, should be admitted
to institutions with the resources capable of managing acute
cerebrovascular events. That would include the rapid avail-
ability of neurologists, vascular surgeons, neurosurgeons,
internists and radiologists, backed up by appropriate diag-
Figure 11.1 Non-contrast brain computed tomography in a
nostic devices available on a 24-hour basis.
patient with acute onset of right hemiparesis and aphasia. Note
hyperdense middle cerebral artery signal suggesting acute
occlusion in the vessel (arrow)
Elements raising suspicion of stroke

• Focal neurological deficit reaching maximal severity 24 hours after an ischaemic stroke but permits the diagno-
within minutes to hours sis of haemorrhagic stroke and other conditions sometimes
• Subject older than 55 years misdiagnosed as ischaemic strokes.14 The use of fibrinolytic
• Presence of vascular risk factors drugs in the first few hours of stroke and their potential for
severe haemorrhagic complications has stimulated the
identification of early changes on CT which suggest the
Patients should be evaluated immediately upon arrival. presence of infarcted tissue (see box and Fig. 11.1).
A detailed medical history will often require help from rela-
tives or other witnesses, particularly if there are signs of
language disturbance or a diminished level of awareness. Early CT changes in acute ischaemic stroke
Those with severe deficits or other serious medical condi-
tions should be admitted to an intensive care unit. It is • Hyperdense middle cerebral artery signal
important to determine promptly the need for orotracheal • Acute hypodensity
intubation. Careful and frequent evaluation of clinical and • Mass effect
neurological status is of value in prevention, diagnosis and • Loss of grey/white matter interface
treatment of complications. • Loss of sulci
Special attention should be given to ascertaining the • Obscured basal ganglia
timing and characteristics of symptoms at onset, past neuro- • Loss of insular ribbon
logical status, medication and history of important stroke
risk factors such as hypertension, diabetes, smoking, heart
disease, previous stroke or TIAs, drug abuse and family his- In recent years, MRI using the diffusion/perfusion tech-
tory. A complete physical examination is necessary keeping nique (DWI/PWI) has helped in the early identification of
in mind cardiac arrhythmias and murmurs, carotid bruits ischaemic changes and, in cases with DWI/PWI mismatch,
and evidence of pulmonary aspiration. Neurological exam- the delineation of an area of penumbra which will recover
ination will determine the characteristics of the deficit and if reperfusion can be initiated promptly15 (Fig. 11.2).
the localisation of the lesion. The affected vascular territory Basic laboratory studies, namely, blood count, coagula-
can usually be established based on the spectrum of symp- tion assays, electrolytes, glucose, renal, and liver function
toms and signs. tests, electrocardiogram and chest X-ray should be under-
After a brief but thorough initial physical and neuro- taken on admission.
logical examination a non-contrast brain CT should be After initial evaluation a strategy of management should
obtained. This study is often normal during the first be established, with diagnostic evaluation focused on
132 The developing stroke

DWI PWI stroke include antiphospholipid antibodies, elevated fib-

rinogen, erythrocytosis, hyperhomocysteinaemia, activated
protein C resistance and, less frequently, protein C and S
congenital deficiencies and antithrombin III deficiency.
Cerebrospinal fluid examination is presently reserved for
cases of dubious aetiology, especially in young subjects and
when there is suspicion of some infectious or inflammatory
brain process.
Approximately one-third of patients with ischaemic
stroke deteriorate after the initial event: during the first
24 hours (or later in posterior circulation strokes) it represents
a progression of the thrombotic process, in other words,
stroke in evolution.16 Cerebral oedema usually accounts for
(a)
the deterioration occurring between the second and fifth
day, especially when patients have sustained large infarcts.
DWI PWI Neurological deterioration can also be related to a haem-
orrhagic transformation of the infarct, a particular danger in
patients treated with fibrinolytic or anticoagulant drugs, but
it may follow fresh episodes of embolism in patients with
embolic sources, or medical complications such as infection,
metabolic disturbance and haemodynamic failure.

GENERAL MANAGEMENT

The principles of emergency management of stroke, i.e.

(b) control of potential life-threatening complications and
specific treatment of the vascular event,17 are summarised
Figure 11.2 Brain magnetic resonance imaging with in the algorithm given in Fig. 11.3.
diffusion/perfusion weighted (DWI/PWI) technique. (a) DWI/PWI
mismatch. Perfusion defect (light blue arrow) is larger than the
diffusion defect (yellow arrow). (b) Perfusion and diffusion defects Bed rest
are similar suggesting absence of a penumbra area
Bed rest is recommended for the first 48 hours, especially
for those with orthostatic hypotension, commonly observed
assessing the aetiology of the event and locating disease in among diabetics, elderly patients and those receiving anti-
large or small vessels, identifying cardioembolic sources, hypertensive medication. Neurological deterioration may
prothrombotic states and other conditions. This process result from premature mobilisation of patients with ortho-
will refine acute management and facilitate planning the static hypotension and impaired cerebral autoregulation.
most appropriate strategy for secondary prevention.
Vascular ultrasound and magnetic resonance angiography
have proved useful in the detection of severe stenosis and Airway and breathing
occlusion in the carotid artery. Cerebral angiography is
performed transarterially when intra-arterial thrombolysis Ensure airway patency, particularly in patients with decreased
is planned or when the information to be obtained is likely alertness. Hypoxia may alter neurological status and con-
to influence therapeutic decisions. The possibility of a car- tribute to the neurological deficit but supplemental oxygen
diac embolic source is excluded by detailed cardiological is not needed routinely. Pulse oximetry will identify the
examination, chest X-ray and electrocardiogram (ECG) patient with obstructive breathing patterns or desaturation
and if still suspected by transthoracic and/or trans- during sleep.18 Hypercapnia can also be detrimental because
oesophageal echocardiography, 24-hour Holter-ECG, it raises intracranial pressure.
myocardial perfusion studies and coronary angiography.
Transoesophageal echocardiography allows the determina- Circulation
tion of size and motility of cardiac chambers, presence of
intracavitary thrombi (a negative study does not rule out The management of hypertension following acute ischaemic
their existence), valvular disease and atheromatous disease stroke remains a somewhat controversial issue.19 Over
of the aortic arch. Prothrombotic disorders associated with two-thirds of patients with acute stroke have elevated initial
 General management 133

Clinical evidence of acute focal neurological deficit

Clinical and neurological exam in ER Airway/breathing

Establish time of onset Circulation
Fluids
Intracranial pressure
Laboratory
Glycaemic control
Chest X-ray
Temperature
ECG
Skin care
Bladder management
Non-contrast CT DVT prophylaxis
Physical therapy

Changes consistent Intracranial Surgical

with ischaemic stroke haemorrhage consultation

Eligible for IV r-tPA use?

Parenchymal Subarachnoid haemorrhage

cerebral or arteriovenous
YES NO haemorrhage malformation rupture

Personnel with expertise Consider surgical Cerebral

in IV r-tPA use available? evacuation in angiography
selected cases

NO YES Consider surgical Surgery

consultation in large Endovascular treatment
hemispheric and Radiosurgery
cerebellar infarctions
Consider referral Start
to stroke centre IV r-tPA
Complete diagnostic work-up
Continue ancillary treatment
Start rehabilitation
Start secondary prevention measures

Figure 11.3 Algorithm of the principles of emergency management of stroke. CT, computed tomography; DVT, deep vein thrombosis; ECG,
electrocardiogram; ER, emergency room; IV, intravenous; r-tPA, recombinant tissue plasminogen activator

blood pressures (170/100 mmHg), yet there is no clear stenosis and of course the elderly.22,23 Chronic hypertension
relation between hypertension and neurological wors- is common in stroke patients and may be associated with
ening or outcome.20 Blood pressures tend to normalise spon- an upward shift of the lower autoregulatory limit for CBF,
taneously following stroke, with decreases of about say a mean of 85–150 mmHg; thus even modest blood
20 mmHg systolic and 10 mmHg diastolic during hospital- pressure reductions may compromise collateral supply to
isation.20 The rate of decline is most rapid during the initial dependent regions of the brain.24
four days and greatest in subjects with higher initial values. Knowledge of stroke pathophysiology and the benign
Moderate hypertension followed by normalisation is thus course of post-stroke hypertension support the axiom that
expected following ischaemic stroke; moderate hypertension hypertension should be left untreated early on, unless it is
may even be beneficial early on, as has been demonstrated in dangerously elevated or sustained25 (Fig. 11.1). Strokes in the
some animal models.21 setting of aortic dissection, symptomatic congestive heart
Certain individuals are particularly susceptible to the failure, acute myocardial infarction or other hypertensive
dangers of rapid blood pressure reduction, including those organ failure represent exceptions which require more
with prior chronic hypertension or high grade arterial rapid management.
134 The developing stroke

Treatment of hypertension, a cornerstone of secondary Swallowing, hydration and nutrition

stroke prevention, should therefore be initiated cautiously
maintaining vigilance for orthostatic side effects. While a Acute ischaemic stroke patients are at high risk of aspiration,
graduated reduction of blood pressure is of greater import- a devastating complication in most instances. Oral fluids or
ance than the antihypertensive medication selected, we food should not be given until a formal swallowing exam-
generally avoid nifedipine, diuretics and clonidine which ination has been performed.35 The risk of aspiration is partic-
cause rapid and sharp falls in blood pressure, volume ularly high in patients with absent gag reflex, coughing or
contraction, lowered CBF and changes in mental status.26 choking when attempting to drink, dysarthria and in those
Labetalol, a combined -adrenergic blocker, has proved who are obtunded.36 For patients with swallowing difficulties,
to be safe and effective in accelerated hypertension when we recommend placement of a duodenal tube on the second
administered orally or intravenously to patients as long as day after the stroke as nasogastric tube feeding also carries
they do not have asthma, chronic obstructive pulmonary the risk of aspiration. Duodenal tube fluid administration
disease, bradycardia, heart block or failure.27 should be by continuous infusion to prevent regurgitation,
Hypotension may occur in patients with severe coexist- acute gastric retention and osmotic diarrhoea. The head of
ent heart disease marked by arrhythmias, heart failure or the bed should be elevated at least 30 degrees at all times.
acute myocardial infarction, or if they are dehydrated or The volume and caloric content can be increased gradually
septic. Cerebral blood flow may be impaired in this situa- over the first 48–72 hours to meet daily requirements.
tion and urgent correction is advised.

Glycaemic control
Fluids
An association between elevated blood glucose levels and
An intravenous line is required for management of fluids poor stroke outcome, perhaps due to the effect of lactic aci-
and electrolytes and for drug administration. We rec- dosis on infarcted brain, has been reported.37 Some stud-
ommend normal saline infusion for most patients, the total ies, however, have shown that there is no direct association
fluid intake not exceeding 2500 mL/day for the first four between hyperglycaemia and stroke outcome.38 Diabetic
days; fluid restriction minimises the development of brain patients are best managed initially using a glucose sliding
oedema in those with suspected or proved large infarcts. scale and corrections with regular insulin followed by diet
Approximately 15 per cent of patients develop hypona- and either oral hypoglycaemic agents or insulin depending
traemia with or without dehydration. Sodium correction on severity and requirements.
must be timely and executed with caution if central nervous
system lesions, for example central pontine myelinolysis, are
to be avoided. Hypernatraemia is less common but it too Pyrexia
must also be corrected cautiously to avoid cerebral oedema.
Hyperthermia, even mild, is associated with increase of
infarct size and higher morbidity and mortality.39,40 Pre-
Raised intracranial pressure ceding or concomitant infection is common in ischaemic
stroke patients.41,42 Experimental studies in animals and
Oedema is caused by intracellular retention of fluid in the humans have shown that moderate hypothermia might be
ischaemic area and reaches its peak 48–72 hours after the beneficial in acute stroke.43,44 Until the results of further
stroke.28 In patients with large hemispheric and cerebellar trials of cooling are published it would be prudent to keep
infarctions mass effect may result in clinical deterioration body temperature under control in these patients.
and death (see Chapter 5). Early signs of progressing oedema
include decreased alertness and pupillary changes.29 Treat-
ment of elevated intracranial pressure following ischaemic Pressure areas
stroke is difficult. Hyperventilation produces hypocapnia
and vasoconstriction, with potential worsening of cerebral Skin care and prevention of decubitus ulcers is achieved by
ischaemia.30 Steroids are ineffective in ischaemic stroke.31 the use of special mattresses and by mobilising the patient
We recommend mannitol at doses of 0.25–0.50 g/kg at every 2 hours.45
4–6-hour intervals, carefully managing electrolytes and
maintaining serum osmolality below 300 mOsm.32 In severe
cases intravenous furosemide may be combined with manni- Bladder management
tol.33 Patients with cerebellar infarction developing brain
stem compression may benefit from shunting and posterior Incontinence of urine is common in the first few days.
fossa surgical decompression, an interesting but still Indwelling urinary catheters should be avoided if possible
unproved option in those with large hemispheric strokes.34 to prevent urinary tract infections.46
 Specific treatment 135

Venous thromboembolism prophylaxis when used within 3 hours of onset of symptoms in highly
selected patients.53,55 The number of patients who make an
Deep vein thrombosis (DVT) is common in stroke patients47 excellent recovery is significantly higher in those who
and even more so in patients with dense paralysis of the receive r-tPA although there is no substantial difference in
affected limb. Older subjects are at increased risk, particu- mortality and this is also true of ischaemic stroke subtypes.
larly those with obesity, congestive heart failure, varicose Studies using streptokinase in acute ischaemic stroke
veins, history of thromboembolism and hypercoagulable showed that it was unhelpful and ill advised.56 The safety of
states.47 However, clinically apparent DVT occurs in less using r-tPA continues to be a major concern, brain haem-
than 10 per cent of subjects and pulmonary embolism in orrhage having been observed 10 times more frequently in
less than 2 per cent. Prophylactic measures for patients at patients receiving this treatment (6.4 v 0.6 per cent).
risk include early mobilisation, external pneumatic com- Criteria for r-tPA use are listed below. Strict adherence
pression, antiplatelet therapy and antithrombotic drug ther- to these criteria is essential if catastrophic haemorrhagic
apy using low dose subcutaneous heparin or low molecular complications are to be prevented. Inadequate control of
weight heparinoids.48–50 blood pressure, stroke severity, and evidence of early signs
of infarction on initial CT are important predisposing
factors for haemorrhagic complications.
Epileptic seizures

Seizures are frequent in patients with large infarctions

especially during the first 48 hours.51 Acute cerebrovascu- Criteria for treatment with r-tPA
lar events are the most common cause of seizures in sub-
jects older than 65 years. The response to anticonvulsive • Stroke onset 3 hours
therapy is usually satisfactory and there is no evidence to • Intensive care available for 24 hours post-treatment
support the prophylactic use of antiepileptic drugs. monitoring
• Patients already receiving antiplatelet agents prior to
stroke are still eligible for r-tPA
Rehabilitation
• As the use of thrombolytic drugs carries a real risk of
major bleeding, whenever possible the risks and
Physical and occupational therapy and assessment of speech potential benefits of r-tPA should be discussed with
difficulties should be initiated soon after admission. the patient and his or her family before treatment is
Rehabilitation programmes must be customised for each initiated
individual and should start as soon as the patient is clini-
cally stable.
• Exclusion criteria
– Current use of oral anticoagulants or a
prothrombin time greater than 15 seconds –
international normalised ratio (INR) greater
SPECIFIC TREATMENT
than 1.7
– Use of heparin in the previous 48 hours and a
Stroke is a medical emergency with a narrow therapeutic prolonged partial thromboplastin time
window. Minimising the time interval between the onset of – A platelet count less than 100 000/mm3
symptoms and the initiation of specific treatment is key to – Another stroke or a serious head injury in the
the patient’s chances of a satisfactory recovery.52 previous 3 months
– Major surgery within the preceding 14 days
– Pretreatment systolic blood pressure greater than
Thrombolytic therapy
185 mmHg or diastolic blood pressure greater
than 110 mmHg
Multicentre, double blind, placebo controlled randomised
– Rapidly improving neurological signs
trials in the USA and Europe have demonstrated the effi-
– Caution is advised before giving r-tPA to persons
cacy of early administration of activated recombinant tis-
with severe stroke (NIH Stroke Scale Score greater
sue plasminogen activator (r-tPA) in patients with acute
than 22)
ischaemic stroke.53–55 Nevertheless, the drug is underused
– Isolated, mild neurological deficits, such as ataxia
in most centres in spite of the lack of alternative strategies
alone, sensory loss alone, dysarthria alone or
because of delays in presentation to the emergency room,
minimal weakness
fears of haemorrhagic complications and the inadequacy
– Prior intracranial hemorrhage
of an infrastructure to manage stroke.
– Blood glucose less than 50 mg/dL or greater than
Intravenous r-tPA is effective at a dose of 0.9 mg/kg
400 mg/dL
given intravenously over 60 minutes (10 per cent of dose
– Seizure at the onset of stroke
as intravenous bolus over 1 minute; maximal dose 90 mg)
136 The developing stroke

– Gastrointestinal or urinary bleeding within the Treatment of thrombolysis related bleeding

preceding 21 days
– Recent myocardial infarction • Discontinue ongoing infusion of thrombolytic drug
• r-tPA should be avoided in patients with evidence of • Obtain blood samples for coagulation tests
mass effect or oedema on CT scan or early changes • Obtain surgical and haematological consultations
indicating involvement  1/3 of the middle cerebral • Consider transfusion, cryoprecipitate and platelets
artery territory. An expert must interpret the
CT scan
The Prolyse in Acute Cerebral Thromboembolism II
(PROACT II) Study58 showed that stroke patients treated
Thrombolytic therapy is not recommended unless a with intra-arterial prourokinase within 6 hours of onset of
physician with expertise in stroke management establishes symptoms were more likely to live independently after the
the diagnosis aided by a CT of the brain. If the CT demon- stroke. The use of intra-arterial thrombolysis for patients
strates early changes of recent major infarction such as sulcal with acute ischaemic stroke remains under continued
effacement, mass effect, oedema or possibly haemorrhage, review by the Food and Drug Administration (FDA) and
then thrombolytic therapy should be avoided. Exclusion has not yet been approved.
criteria in the National Institute of Neurological Disorders Ancrod, a potent fibrinogenolytic agent derived from
and Stroke r-tPA Stroke Study (NINDS) must be considered snake venom, effective in the Stroke Treatment with
to be contraindications for r-tPA use. Important measures Ancrod Trial (STAT) trial, is also awaiting FDA approval.59
in ancillary care, during and after administration of r-tPA
are described below. Anticoagulation/antiaggregation

Anticoagulation started in the first day or two after stroke

Ancillary care during and after may reduce the risk of DVT and pulmonary embolism but
administration of r-tPA there is no evidence of other short or long term neurologi-
cal benefit.60,61 There are a few exceptions to these criteria
in which use of intravenous unfractionated heparin or
• Admission to a skilled care facility – intensive care
subcutaneous low molecular weight heparins may be
unit or acute stroke care unit
considered.62,63
• Careful management of arterial blood pressure,
avoiding excessively high blood pressure and
excessive lowering of blood pressure
• Central venous access and arterial punctures are Potential indications for anticoagulation in
restricted during the first 24 hours acute cerebrovascular disease
• Placement of an indwelling bladder catheter should
be avoided during drug infusion and for at least 30 • Recent arterial dissection
minutes after infusion ends • Cerebral venous thrombosis
• Insertion of a nasogastric tube should be avoided, • Frequent or imminent cardiac embolism
if possible, during the first 24 hours after • Stroke-in-evolution specially in the posterior
treatment circulation
• Crescendo TIA
• High grade symptomatic stenosis specially in the
posterior intracranial circulation
Intracranial and systemic bleeding are the most cata-
strophic complications after r-tPA use.57 Thrombolytic
therapy should not be used unless facilities to handle these
bleeding complications are readily available. Bleeding The efficacy of antiaggregants and anticoagulants in
should be presumed to be the cause of any neurological some acute stroke patients can probably be attributed to
worsening after the use of a thrombolytic drug until con- the prevention of early recurrence in subjects at risk.64–66
firmed by CT, which should be obtained immediately. The Certain drugs such as warfarin and ticlopidine have latency
management of thrombolysis related bleeding depends on effects limiting their use during the acute event. New
the location and size of the haematoma, the potential for agents such as clopidogrel at high doses of 300 mg, hirudin
controlling the bleeding mechanically, the neurological and antagonists of GPIIa/IIIb receptors have rapid
risk, the interval between administration of the drug and antithrombotic effects and are being evaluated in clinical
the onset of haemorrhage and the thrombolytic drug used. trials. In patients with acute ischaemic stroke and a proved
Appropriate measures are listed below. cardioembolic source intravenous heparin followed by
 References 137

warfarin is recommended. This strategy is intended for

Acute stroke should be considered a medical emer-
acute secondary prevention and is unlikely to be beneficial
gency with a narrow therapeutic window. Fibrinolytic
for the already completed event. The timing for initiation
therapy is relatively safe and effective for patients with
of treatment is controversial. An acceptable approach
acute ischaemic stroke when administered early in the
would be to start anticoagulation 48–72 hours after the
course of the disease and within very strict eligibility
event in those cases in which the absolute contraindica-
guidelines. Its use should be reserved for institutions
tions of uncontrolled hypertension or haemorrhagic trans-
with adequate infrastructure and experience in the man-
formation in a follow-up CT do not apply. In patients with
agement of the condition.
large infarcts anticoagulation is often delayed for 1–2
weeks. In patients with atrial fibrillation the risk of early
recurrence is not very high and anticoagulation may be
delayed for a few days. Key references
Low to medium dose aspirin (160–300 mg) started in the
acute phase of an ischaemic stroke slightly reduces morbid- Adams H, Adams R, Del Zappo G, et al. Guidelines for the early
ity and mortality probably because of the earlier initiation management of patients with ischemic stroke. 2005
of secondary prevention of stroke and other thrombotic Guidelines update. A scientific statement from the Stroke
complications.64,65 Council of the American Heart Association/American Stroke
Association. Stroke 2005; 36: 916–21.
Adams HP Jr, Brott TG, Furlan AJ, et al. Guidelines for thrombolytic
Neuroprotective agents therapy for acute stroke: a supplement to the guidelines for
the management of patients with acute ischemic stroke.
Although over 100 compounds have been shown to dimin- Circulation 1996; 94: 1167–74.
ish the extent of ischaemic damage in various laboratory Adams HP, Bendixen BH, Kappelle LJ, et al. Classification of subtype
models, none has yet been shown convincingly to benefit of acute ischemic stroke. Definitions for use in a multicenter
subjects with stroke.67 clinical trial. Stroke 1993; 24: 35–41.
Brott T, Bogousslavsky J. Drug therapy: treatment of acute ischemic
stroke. N Engl J Med 2000; 343: 710–22.
Treatment of haemorrhagic stroke The National Institute of Neurological Disorders and Stroke rt-PA
Stroke Study Group. Tissue plasminogen activator for acute
Medical management of increased intracranial pressure ischemic stroke. N Engl J Med. 1995; 333: 1581–7.
includes the use of osmotic agents such as mannitol, urea or
glycerol as well as steroids and hyperventilation.68 Occa-
sionally, it is necessary to place an intraparenchymal or
intraventricular catheter to directly measure intracranial REFERENCES
pressure. Investigation and correction or reversal of any
coagulation abnormality is mandatory especially when 1 American Heart Association. 1991 Heart and Stroke Facts. Dallas,
bleeding occurs after a surgical procedure. TX: American Heart Association, 1991.
Surgical evacuation of the haematoma is often considered 2 Alberts MJ, Perry A, Dawson DV, Bertels C. Effects of public and
although data from the literature are controversial.68,69 It is professional education on reducing the delay in presentation and
possible that minimally invasive techniques such as stereo- referral of stroke patients. Stroke 1992; 23: 352–6.
tactic and endoscopic aspiration will allow early and safe 3 Foulkes MA, Wolf PA, Price TR, et al. The stroke data bank: design,
methods, and baseline characteristics. Stroke 1988; 19: 547–54.
removal of the clot. Large lobar and cerebellar haematomas
4 Adams HP, Bendixen BH, Kappelle LJ, et al. Classification of
are the most frequent indications for surgery.70 subtype of acute ischemic stroke. Definitions for use in a
multicenter clinical trial. Stroke 1993; 24: 35–41.
5 Hart RG. Cardiogenic embolism to the brain. Lancet 1992; 339:
Conclusions 589–94.
6 Amarenco P, Cohen A, Tzourio C, et al. Atherosclerotic disease of
Acute stroke is common and carries a high morbidity the aortic arch and the risk of ischemic stroke. N Engl J Med 1994;
and mortality rate. Recent advances in its treatment dur- 331: 1474–9.
ing the acute phase may help reduce its devastating con- 7 Markus HS, Hambley H. Neurology and the blood: haematological
sequences. The complexity of modern acute stroke care abnormalities in ischaemic stroke. J Neurol Neurosurg Psychiatry
and the risks posed by the different strategies used have 1998; 64: 150–9.
prompted the organisation of stroke services staffed by 8 Sandercock PA, Warlow CP, Jones LN, Starkey IR. Predisposing
professionals with specialised training in the manage- factors for cerebral infarction: the Oxfordshire community stroke
ment of the disease and the potential complications of project. BMJ 1989; 298: 75–80.
9 Siesjo BK. Pathophysiology and treatment of focal cerebral
treatment. Specialised multidisciplinary stroke units
ischemia. Part I: Pathophysiology. J Neurosurg 1992; 77:
may reduce morbidity and mortality for stroke patients.
169–84.
138 The developing stroke

10 Hossmann K-A. Viability thresholds and the penumbra of focal 34 Schwab S, Steiner T, Aschoff A, et al. Early hemicraniectomy in
cerebral ischemia. Ann Neurol 1994; 36: 557–65. patients with complete middle cerebral artery infarction. Stroke
11 Pulsinelli WA. The therapeutic window in ischemic brain injury. 1998; 29: 1888–93.
Curr Opin Neurol 1995; 8: 3–5. 35 Gordon C, Hewer RL, Wade DT. Dysphagia in acute stroke. BMJ
12 Meyer JS, Shimazu K, Fukuuchi Y, et al. Impaired neurogenic 1987; 295: 411–14.
cerebrovascular control and dysautoregulation after stroke. 36 Horner J, Massey EW, Riski JE, et al. Aspiration following stroke:
Stroke 1973; 4: 169–86. clinical correlates and outcome. Neurology 1988; 38: 1359–62.
13 Fischberg GM, Lozano E, Rajamani K, et al. Stroke precipitated by 37 Pulsinelli WA, Levy DE, Sigsbee B, et al. Increased damage after
moderate blood pressure reduction. J Emerg Med 2000; 19: 339–46 ischemic stroke in patients with hyperglycemia with or without
14 Libman R, Wirowski E, Alvir J, Rao T. Conditions that mimic established diabetes mellitus. Am J Med 1983; 74: 540–4.
stroke in the emergency department. Arch Neurol 1995; 52: 38 Matchar DB, Divine GW, Heyman A, Feussner JR. The influence of
1119–22. hyperglycemia on outcome of cerebral infarction. Ann Int Med
15 Fisher M, Prichard J, Warach S. New magnetic resonance 1992; 117: 449–56.
techniques for acute ischemic stroke. JAMA 1995; 274: 908–11. 39 Chen H, Chopp M, Welch KM. Effect of mild hyperthermia on the
16 Gautier JC. Stroke-in-progression. Stroke 1985; 16: 729–33. ischemic infarct volume after middle cerebral artery occlusion in
17 Brott T, Bogousslavsky J. Drug therapy: treatment of acute the rat. Neurology 1991; 41: 1133–5.
ischemic stroke. N Engl J Med 2000; 343: 710–22. 40 Ginsberg MD, Busto R. Combating hyperthermia in acute stroke.
18 Rout MW, Lane DJ, Wollner L. Prognosis in acute cerebrovascular A significant clinical concern. Stroke 1998; 29: 529–34.
accidents in relation to respiratory pattern and blood gas 41 Ameriso SF, Wong VLY, Quismorio FP, Fisher M.
tensions. BMJ 1971; 3: 7–9. Immunohematologic characteristics of infection-associated
19 Potter JF. What should we do about blood pressure and stroke? cerebral infarction. Stroke 1991; 22: 1004–9.
Q J Med 1999; 92: 63–6. 42 Grau A, Buggle F, Heindl S, et al. Recent infection as a risk factor
20 Wallace JD, Levy LL. Blood pressure after stroke. JAMA 1981; for cerebrovascular ischemia. Stroke 1995; 26: 373–9.
246: 2177–80. 43 Schwab S, Schwarz S, Spranger M, et al. Moderate hypothermia
21 Cole DJ, Matsumura JS, Drummond JC, Schell RM. Focal cerebral in the treatment of patients with severe middle cerebral artery
ischemia in rats: Effects of induced hypertension, during infarction. Stroke 1998; 29: 2461–6.
reperfusion, on CBF. J Cereb Blood Flow Metab 1992; 1264–9. 44 Yanamoto H, Nagata I, Niitsu Y, et al. Prolonged mild
22 Jansen PAF, Schulte BPM, Meybook RHB, Gribnau FWJ. hypothermia therapy protects the brain against permanent focal
Antihypertensive treatment as a possible cause of stroke in the ischemia. Stroke 2001; 32: 232–9.
elderly. Age Ageing 1986; 15: 129–38. 45 Nuffield Institute for Health. Effective health care. The
23 Ruff RL, Talman WT, Petito F. Transient ischemic attacks prevention and treatment of pressure sores. Effective Health Care
associated with hypotension in hypertensive patients with 1995; 2: 1–16.
carotid artery stenosis. Stroke 1981; 12: 353–5. 46 Nakayama H, Jorgensen HS, Pedersen PM, et al. Prevalence and
24 Paulson OB, Waldamer G, Schmidt JF, Strandgaard S. Cerebral risk factors of incontinence after stroke. The Copenhagen stroke
circulation under normal and pathologic conditions. Am J Cardiol study. Stroke 1997; 28: 58–62.
1989; 63: 2C–5C. 47 Sioson ER. Deep vein thrombosis in stroke patients: an overview.
25 Yatsu FM, Zivin J. Hypertension in acute ischemic strokes. Not to J Stroke Cerebrovasc Dis 1992; 2: 74–9.
treat. Arch Neurol 1985; 42: 999–1000. 48 Turpie A, Gallus A, Beattie WS, Hirsh J. Prevention of venous
26 Bertel O, Conen D, Radu EW, et al. Nifedipine in hypertensive thrombosis in patients with intracranial disease by intermittent
emergencies. BMJ 1983; 286: 19–21. pneumatic compression of the calf. Neurology 1977; 27: 435–8.
27 Gonzalez ER, Peterson MA, Racht EM, et al. Dose response 49 Antiplatelet Trialists’ Collaboration. Collaborative overview of
evaluation of oral labetalol in patients presenting to the randomized trials of antiplatelet therapy. III: Reduction in venous
emergency department with accelerated hypertension. thrombosis and pulmonary embolism by antiplatelet prophylaxis
Ann Emerg Med 1991; 20: 333–8. among surgical and medical patients. BMJ 1994; 308: 235–46.
28 Garcia JH, Ho K-L, Pantoni L. Pathology. In: Barnett HJM, 50 Turpie AG, Gent M, Cote R, et al. A low-molecular weight
Mohr JP, Stein BM, Yatsu FM (eds). Stroke. Pathophysiology, heparinoid compared with unfractionated heparin in the
diagnosis, and management. New York NY: Churchill Livingstone, prevention of deep vein thrombosis in patients with acute
1998: 139–57. ischemic stroke. Ann Intern Med 1992; 117: 353–7.
29 Ropper AH, Shafran B. Brain edema after stroke. Clinical syndrome 51 Burn J, Dennis M, Bamford J, et al. Epileptic seizures after a first
and intracranial pressure. Arch Neurol 1984; 41: 26–29. stroke: the Oxfordshire community stroke project. BMJ 1997;
30 Raichle M, Plum F. Hyperventilation and cerebral blood flow. 315: 1582–7.
Stroke 1972; 3: 566–75. 52 Adams H, Adams R, Del Zappo G, et al. Guidelines for the early
31 Norris JW, Hachinski VC. High dose steroid treatment in cerebral management of patients with ischemic stroke. 2005 Guidelines
infarction. BMJ 1986; 292: 21–3. update. A scientific statement from the Stroke Council of the
32 Schwarz S, Schwab S, Bertram M, et al. Effects of hypertonic American Heart Association/American Stroke Association. Stroke
saline hydroxyethyl starch solution and mannitol in patients with 2005; 36: 916–21.
increased intracranial pressure after stroke. Stroke 1998; 29: 53 The National Institute of Neurological Disorders and Stroke
1550–5. rt-PA Stroke Study Group. Tissue plasminogen activator for acute
33 Pollay M, Fullenwider C, Roberts A, Stevens A. Effect of mannitol ischemic stroke. N Engl J Med. 1995; 333: 1581–7.
and furosemide on blood-brain osmotic gradient and intracranial 54 Hacke W, Kaste M, Fieschi C, et al. Randomised double-blind
pressure. J Neurosurg 1983; 59: 945–50. placebo-controlled trial of thrombolytic therapy with
 References 139

intravenous alteplase in acute ischaemic stroke (ECASS II). 63 Saver JL, Easton JD. Dissections and trauma of cervicocerebral
Lancet 1998; 352: 1245–51. arteries. In: Barnett HJM, Mohr JP, Stein BM, et al. (eds) Stroke:
55 Adams HP Jr, Brott TG, Furlan AJ, et al. Guidelines for pathophysiology, diagnosis and management. New York:
thrombolytic therapy for acute stroke: a supplement to the Churchill Livingstone, 1998: 769–86.
guidelines for the management of patients with acute ischemic 64 CAST (Chinese acute stroke trial) Collaborative Group. CAST:
stroke. Circulation 1996; 94: 1167–74. randomized placebo-controlled trial of early aspirin use in
56 Donnan GA, Davis SM, Chambers BR, et al. Trials of streptokinase in 20 000 patients with acute ischaemic stroke. Lancet 1997; 349:
severe acute ischaemic stroke. Lancet 1995; 345: 578–9. 1641–9.
57 The National Institute of Neurological Disorders and Stroke rt-PA 65 International Stroke Trial Collaborative Group. The International
Stroke Study Group. Intracerebral hemorrhage after intravenous Stroke Trial (IST). A randomised trial of aspirin, subcutaneous
t-PA therapy for ischemic stroke. Stroke 1997; 28: 2109–18. heparin, both, or neither among 19,435 patients with acute
58 Furlan A, Higashida R, Wechsler L, et al. for the PROACT ischaemic stroke. Lancet 1997; 349: 1569–81.
Investigators. Intra-arterial prourokinase for acute ischemic 66 Kay R, Wong KS, Yu YL, et al. Low molecular weight heparin for
stroke: the PROACT II study: a randomized controlled trial. JAMA the treatment of acute ischemic stroke. N Engl J Med 1995; 333:
1999; 282: 2003–11. 1588–92.
59 Sherman DG, Atkinson RP, Chippendale T, et al. Intravenous 67 Devuyst G, Bogousslavsky J. Clinical trial update: neuroprotection
ancrod for treatment of acute ischemic stroke: the STAT study: a against acute ischaemic stroke. Curr Opin Neurol 1999; 12:
randomized controlled trial. JAMA 2000; 282: 2395–403. 73–9.
60 Bousser MG. Aspirin or heparin immediately after a stroke? 68 Qureshi AI, Tuhrim S, Broderick JP, et al. Medical progress:
Lancet 1997; 349: 1564–5. spontaneous intracerebral hemorrhage. N Engl J Med 2001; 344:
61 The Publications Committee for the Trial of Org 10172 in Acute 1450–60.
Stroke Treatment (TOAST) Investigators. Low molecular weight 69 Hankey GJ, Hon C. Surgery for primary intracerebral hemorrhage:
heparinoid, Org 10172 (danaparoid), and outcome after acute is it safe and effective? A systematic review of case series and
ischemic stroke: a randomized controlled trial. JAMA 1998; 279: randomized trials. Stroke 1997; 28: 2126–32.
1265–72. 70 Gerritsen van der Hoop R, Vermeulen M, Van Gijn J. Cerebellar
62 Einhaupl KM, Villringer A, Meister W, et al. Heparin treatment in hemorrhage: diagnosis and treatment. Surg Neurol 1988;
sinus venous thrombosis. Lancet 1991; 338: 597–600. 29: 6–10.
This page intentionally left blank
 12
Role of the Vascular Surgeon in Managing Stroke

JOHN P ROYLE, GEOFFREY A DONNAN, BRIAN CHAMBERS

The problem 141 Technical issues 144

Selection for urgent surgery 141 Operative technique 144
Investigations 142 References 146
Management 142

THE PROBLEM SELECTION FOR URGENT SURGERY

In 1954, Eastcott et al.1 reported a successful carotid artery The basis for urgent surgery follows the knowledge that a
reconstruction in a patient with transient cerebral ischaemia. stroke may occur quickly after a TIA. A patient with an intact
Following this report attention was focused on the carotid brain after a TIA has everything to save. Delay in surgery may
artery in the neck as a cause of stroke. It was realised that a result in a potentially preventable stroke. The 1977 study of
stroke from carotid artery disease could occur due to a reduc- Cartlidge et al. from the Mayo Clinic5 showed that the risk of
tion in flow, or due to emboli, and transient ischaemic attacks stroke was high in the first month after a TIA, and probably
(TIAs) could, similarly, arise from both of these causes. the greatest risk was in the first few days after the initial event.
There then developed a worldwide controversy over The two most important aspects of selection of patients
the place of carotid endarterectomy. With the results of for surgery are diagnosis and timing. For this reason it is
two randomised trials of carotid artery surgery, the place essential that neurologists and vascular surgeons work in
of this operation was finally confirmed. In the European close collaboration.
Carotid Surgery Trial (ECST)2 and in the North American
Symptomatic Carotid Endarterectomy Trial (NASCET),3 it
was firmly established that when a stenosis of 70–99 per Wrong diagnosis
cent was present in patients with symptomatic carotid dis-
ease, carotid artery surgery was clearly superior to best med- One of the most important aspects of selection of patients
ical treatment. When the stenosis was 50–70 per cent, the for carotid endarterectomy is to be certain that a carotid
results were, initially, unclear but it is now recognised that artery lesion is indeed responsible for the ischaemic deficit.
there is also some advantage of surgery over best medical An embolus may come from the heart, aortic valve, athero-
therapy in this group of patients.4 matous aortic arch or carotid artery. There may occasion-
More recently, carotid angioplasty and stenting has been ally be difficulty in distinguishing posterior from anterior
used as an alternative to carotid endarterectomy, particu- circulation symptoms ipsilateral to a demonstrated carotid
larly in patients in whom there are relative contraindica- stenosis. Alternatively, although symptoms may relate to
tions to surgery. Routine deployment of a mesh funnel or the anterior circulation, the pathology may be in small
other distal protection device reduces the risk of dislodging penetrating vessels quite unrelated to an ipsilateral carotid
plaque material during the procedure, and stroke compli- artery stenosis. The most common example of this is in a
cations are now lower than when the technique was first patient with lacunar TIA or minor stroke due to in situ sin-
used. Currently there are trials underway comparing carotid gle penetrator vessel disease and not caused by emboli
endarterectomy and carotid stenting. If stenting proves more from the heart or carotid artery. Lacunar TIAs may occa-
efficacious than endarterectomy, most of the material pre- sionally present in crescendo form with repeated bursts of
sented in this chapter still applies. hemiplegia, the so-called capsular warning syndrome.6
142 Role of the vascular surgeon in managing stroke

Cerebral haemorrhage of duplex scanning, the latter is now the usual investigation
employed in the demonstration of carotid stenosis. At our
Where the initial TIA has in fact been due to a subdural institution, we have found that transcranial Doppler gives
haematoma, intracerebral tumour or cerebral haemorrhage, valuable information about the status of the intracranial cir-
a carotid endarterectomy is strongly contraindicated. The culation and collaterals. Angiography is reserved for patients
relief of a tight stenosis will, by increasing the blood pres- in whom a duplex scan presents technical difficulties, as
sure at the site of the bleed, substantially increase the risk of sometimes occurs with heavily calcified lesions or when
further haemorrhage. A computed tomography (CT) scan there is doubt about the result of a duplex scan. With experi-
performed early will exclude these entities. enced ultrasonographers, the percentage of patients in
whom there will be a doubt is now very low. Gadolinium-
enhanced MRI or CT angiography can be used instead of
Timing
conventional angiography in selected cases.
There is anecdotal evidence that when an enhancing infarct
is demonstrated on a CT scan there is an increased risk of
converting a recent infarct into a haemorrhagic infarct. MANAGEMENT
Therefore, operation is best delayed for at least 2–3 weeks
after the stroke.
Current routine at the Austin and
Repatriation Medical Centre (see Fig. 12.1)
INVESTIGATIONS
Patients present to the emergency department in three ways:
A CT scan is required to determine whether there is any • a local general practitioner has phoned the admitting
cerebral haemorrhage present and to determine whether officer regarding the patient
there is any infarct present. However, when a CT scan is • the relatives have brought the patient directly to the
performed within 24 hours after an infarct, it may appear emergency department without first attending a local
normal and an infarct may only be demonstrated on follow- doctor
up CT a few days later. When there is an established neu- • an ambulance has brought the patient following a
rological deficit it is assumed that an infarct is present, ‘collapse’.
even though the CT scan may not demonstrate it initially. When the admitting officer is forewarned of the patient’s
As magnetic resonance imaging (MRI) becomes more arrival, the patient will be shown directly to an assessment
widely available in acute care hospitals, its superiority over room, but when the patient comes unannounced, the triage
CT in evaluation of acute stroke is appreciated more and sister will perform this task.
more. Provided the scanning protocol includes a gradient When a patient has a neurological deficit immediate
echo sequence, cerebral haemorrhage is readily diagnosed. admission is arranged. The emergency department medical
The main advantage over CT, however, is the use of diffu- staff make their initial assessment. A CT scan is usually
sion-weighted imaging (DWI) to demonstrate acute brain arranged. Further assessment is done by a neurology regis-
infarction as early as 1 hour after the onset of symptoms. trar and/or neurologist concerning the need for a duplex
The sensitivity of MRI in detecting cerebral infarction is scan. Patients with a gross deficit may have a duplex scan at
much greater than even delayed CT (see Chapter 13). a later time; as they will not be considered for surgery on an
Therefore MRI can identify those individuals in whom it may urgent basis there is no need for this to be expedited.
be better to delay carotid endarterectomy. In patients with Patients without a deficit, or with a doubtful diagnosis,
infarction, the location of the infarct, e.g. anterior versus pos- are usually assessed in the emergency department by the
terior circulation or cortical versus lacunar, helps determine neurology registrar. If necessary, a CT scan and duplex study
whether or not carotid disease is implicated. Also, magnetic of carotid arteries may be performed immediately. An MRI
resonance angiography (MRA) of intracranial, and if neces- scan may also be expedited. If these show a carotid stenosis,
sary, extracranial vessels provides invaluable information admission and surgery are expedited. If the duplex scan is
concerning vascular pathology without injection of contrast doubtful, then angiography may be arranged, although this
media, and often, if performed within the first few hours after is not common nowadays. Some patients with no carotid
the onset of symptoms, embolic occlusion of the middle pathology may be admitted for further detailed evaluation,
cerebral artery may be demonstrated. looking for alternative sources of emboli. Others will be
sent home either with or without aspirin and may have fur-
Duplex scan versus angiography ther evaluation on an outpatient basis.
In many instances, patients with a minor neurological
A diagnosis of carotid stenosis used to be made by angio- deficit follow the same path as patients who have had a
graphy, but with improvements of equipment and technique TIA and whose neurological signs have resolved completely,
 Management 143

GP referral Ambulance Self-referral

Emergency department

Triage nurse

Assessment room, emergency medical staff

CT scan

Neurology/stroke registrar assessment

With deficit Without deficit or minor deficit

Admit to Stroke Unit Definite TIA Non-TIA

Investigate underlying cause Duplex scan

(Duplex, MRI/A, TOE etc.)
Duplex ve
ve ve Home  aspirin

tPA if 3 hours Supportive Consider Investigate

Aspirin Surgery
post stroke treatment delayed CEA alternative cause

Figure 12.1 Emergency management of patients with stroke at the Austin and Repatriation Medical Centre, Melbourne, Australia. CEA,
carotid endarterectomy; CT, computed tomography; GP, general practitioner; MRI/A, magnetic resonance imaging/angiography; TIA, transient
ischaemic attack; TOE, transoesophageal echocardiography; tPA, tissue plasminogen activator

i.e. they have an urgent duplex scan, which, if it confirms a endarterectomy is delayed for a month unless a further TIA
carotid stenosis, is then followed by an urgent operation. If occurs. However, as mentioned earlier, if fluctuations of
the duplex scan is negative, there is no need for further neurological deficit are occurring with almost complete reso-
consideration for carotid endarterectomy. lution between, in spite of adequate medical therapy, consid-
In difficult cases the CT scan is particularly important. eration should be given to emergency endarterectomy. Such
It will exclude a haemorrhage, even a small one. In patients patients should be otherwise medically fit with a relatively
where CT scan is performed within 4 hours of the onset of minor neurological deficit between fluctuations.
an ischaemic deficit, only subtle changes of infarction are The circle of Willis provides a theoretical collateral path-
occasionally seen, but the majority of infarcts will show way for cerebral blood supply. In some patients this is con-
some changes within the first 48 hours. The use of MRI/MRA, genitally incomplete. It has been shown that there is an added
when available, will usually demonstrate infarction even in risk of stroke in patients with contralateral tight stenosis
early cases. or occlusion,7 associated intracranial disease,8 or lack of
One area where a duplex scan may be in error is when intracranial collaterals.9 In each of these situations, carotid
the duplex scan shows complete occlusion when in fact endarterectomy improves the outcome.
‘trickle flow’ may be present. An angiogram will demon-
strate a ‘carotid string sign’. This is a thin sliver of dye seen
connecting the common carotid artery with an intracranial Timing of urgent carotid endarterectomy
normal patent internal carotid artery. The appearance occurs
when there is extremely low flow in the carotid artery and The ‘emergency’ operation is usually undertaken in the next
should not be confused with complete occlusion. In the sta- operating session, rather than in the middle of the night.
ble patient, the risk of stroke is considered to be low as for This obviates problems that may occur when non-regular
complete occlusion, but there are patients with recurring staff are used to help. Similarly, when a patient, for whom
or crescendo TIAs in whom carotid endarterectomy, which a delay in surgery has been advised because of a mild
is not possible with complete occlusion, can be performed. deficit, sustains another TIA, operation is scheduled for the
The management of minor strokes is more controversial. next available session. When the unit has an operating ses-
Usually, if a patient has had a mild cerebral deficit, carotid sion on each day, as we do, there is little delay. However, in
144 Role of the vascular surgeon in managing stroke

these circumstances, when an urgent operation is decided Surgery

on a Friday, we prefer to do the operation on Saturday morn-
ing rather than leave the patient until the following Monday.
In situations where patients are having multiple TIAs, Urgent Delayed
namely, those with so-called crescendo TIAs, a carotid
endarterectomy is performed as a true emergency, if neces-
sary in the middle of the night. In general, the mechanism Next session Immediate
of cerebral ischaemia in these patients is haemodynamic,
secondary to a very tight carotid stenosis. Alternatively, an
extensive haemorrhagic plaque repeatedly dislodging small Anaesthesia
emboli may be present. An angiogram may even show intra-
luminal clot on the surface of such plaques. In these cases
intravenous heparin and/or plasma expanders have nor- LA GA  LA
mally been commenced in order to reduce the risk of stroke.
This therapy is ceased immediately prior to surgery. We do Shunt if deficit Routine
not use vasodilators, and hypertension is treated very cau- appears shunt
tiously indeed. We have seen three patients, each with a
very tight stenosis, presenting with a TIA and marked hyper-
tension, in whom the blood pressure had been reduced to Patch all
120 mmHg systolic, that promptly had a devastating
stroke; the hypertension had been necessary to maintain
flow past the tight stenosis. External carotid Fabric

TECHNICAL ISSUES Completion angiogram or ultrasound

BP Monitor
• When a non-specialist anaesthetist, inexperienced
theatre staff or inexperienced assistants are used, the
risks of operation rise.
Unstable Stable
• When there is a very tight stenosis, relief of the stenosis
may result in loss of autoregulatory mechanisms,
with resultant cerebral haemorrhage (so-called Inpatient until stable Home
hyperperfusion syndrome – 1.3 per cent in our
experience). Figure 12.2 Operative management and technique followed at
the Austin and Repatriation Medical Centre, Melbourne, Australia.
• When there is a haemorrhagic plaque with intralumi-
BP, blood pressure; L/GA, local/general anaesthesia
nal thrombus which may be loose, that thrombus may
be easily dislodged at the time of surgery.
are very anxious, those who have a short bull neck or those
This last group often contains those patients who are with a chronic cough.
neurologically unstable. Even so, our experience would sug- When the operation is performed under local anaesthe-
gest that carefully performed endarterectomy is still associ- sia, there is no need to use a shunt routinely. However, the
ated with low morbidity and mortality. The emphasis is on response to carotid clamping is quite variable. In some
‘carefully’ performed endarterectomy.10 When transcra- patients deficits or even loss of consciousness will occur
nial Doppler evaluation was first available we used it as a within 10 seconds of clamping. On occasions the applica-
routine during carotid endarterectomy. This demonstrated tion of the clamp causes seizures; it may then be difficult to
that an increase in microembolic signals often occurred get the shunt in, and operation in these circumstances is
when the site of the plaque was dissected. certainly not for the inexperienced surgeon. When the oper-
ation is performed under general anaesthesia, a shunt is
used as a routine and it is usually possible to get the shunt
OPERATIVE TECHNIQUE (Fig. 12.2) into position within one and a half minutes.

In patients that are unstable we prefer to perform the oper- Monitoring

ation under local anaesthesia. However, this is inadvisable
in the following: those who may not be able to understand The blood pressure and electrocardiogram (ECG) of the
the instructions because of language difficulties, those who patient are continually monitored. In particular, under
 Operative technique 145

general anaesthesia, great care has to be taken with induc- Sometimes rotation of the head and neck to the contralat-
tion of anaesthesia. An intra-arterial blood pressure line is eral side, to provide optimal exposure at the time of
required before induction of anaesthesia, as this is frequently operation, compromises the collateral circulation and will
when the most marked fall in blood pressure occurs. If produce a neurological disturbance. Thus, this should be
hypotension occurs, it is routine for the anaesthetist checked before the patient is actually anaesthetised.
to use vasopressor drugs such as metaraminol bitartrate The operative technique used by the author closely resem-
(Aramine). The blood pressure should not be allowed to bles that described by Lord.10 There are, however, several
fall below 100 mmHg systolic during operation or below points relevant to the urgent situation which should be
120 mmHg systolic at the times of carotid clamping. If the stressed.
blood pressure becomes very high before or during induc- It is important to dissect the patient away from the artery.
tion of anaesthesia, agents such as intravenous lidocaine The artery must be handled as little as possible, particularly
(Xylocard) 1 mg/kg can be administered. The vocal cords at the site of the plaque. If a haemorrhagic plaque with
are routinely sprayed with local anaesthetic (4 per cent loose clots is present, some of the latter may easily become
lidocaine), and before closure of the neck wound at the end dislodged when pulling on the artery. If further dissection
of the procedure the surgeon injects 0.5 per cent bupiva- is required at this point, it is best done after application of
caine (Marcain) into the wound edges. The resulting the clamps. The effect of rough handling of a badly diseased
analgesia reduces restlessness and thus helps to lessen rises carotid artery is very dramatically demonstrated by the use
in blood pressure which may occur at the end of the anaes- of a transcranial Doppler during surgery.
thetic. It also avoids the need to give narcotic analgesia If a vessel loop is placed around the internal carotid artery,
postoperatively. as is our routine, no clamp should be placed on the loop in
The induction of anaesthesia is achieved by intravenous case the weight of the clamp inadvertently pulls on the
administration of 100 g fentanyl and 2–3 mg midazolam internal carotid artery and occludes it with its low pressure.
while the patient breathes oxygen, followed by propofol The internal carotid artery pressure is not measured. If the
(Diprivan). When Diprivan is given after fentanyl and operation is being undertaken under local anaesthesia, the
midazolam, a lower dose is required than when Diprivan is patient acts as his or her own cerebral monitor. In patients
used alone, and consequently the hypotensive effects of undergoing carotid endarterectomy under general anaes-
Diprivan are minimised. thesia, a shunt is used routinely and placed in position as
Relaxant anaesthesia is used with nitrous oxide, oxygen soon as possible.
isoflurane (Forthane) or seroflurane (Serorane). Dissection It is important to anaesthetise the carotid sinus before
around the carotid sinus nerve or vagus nerve may induce application of clamps. Occasionally, the sinus is very sensi-
a bradycardia that can be corrected with atropine before tive to clamp application; a disturbance of blood pressure
any effect on cardiac output occurs. If during the proced- in these circumstances at this time can be disastrous. It is
ure hypertension occurs, increasing the dose of isoflurane essential to have the shunt and shunt instruments readily
or seroflurane, if either of these is being used to maintain available and to check these with the scrub nurse before the
anaesthesia, will bring the blood pressure down. Failing clamps are applied. There should be very little time gap
this, 5–10 mg hydralazine can be used. If tachycardia ensues, between application of the clamp on the external carotid
the -blocker atenolol is used. Very rarely, sodium nitro- artery, the internal carotid artery and the common carotid
prusside as a continuous infusion may be required. However, artery, as the external carotid artery may well be providing
it is difficult to control swings of blood pressure with nitro- major collateral blood flow.
prusside and we therefore reserve this for patients in whom Normally, if a patient presents with a TIA or minor
other measures have failed. Postoperative hypertension stroke and angiography reveals complete occlusion of the
similarly can be treated with hydralazine 4–40 mg, given in internal carotid artery, no operation is performed. If com-
5 mg increments with or without a -blocker, or rarely, with plete occlusion is found at operation and the duplex scan
sodium nitroprusside if elevation is otherwise uncontrolled. or angiogram performed shortly before operation had
Postoperative hypotension may be treated by elevation shown a patent vessel, the vessel is opened. Gentle attempts
of the foot of the bed, and judicious use of intravenous to extract the clot are made, but if this is not successful,
fluids such as Haemacel. If these simple measures fail, we strenuous attempts are not made because of the risk of dis-
use 5 mL 0.5 per cent bupivacaine as an injection into the lodging clot and subsequent cerebral embolism.
Redivac drain tube. Very occasionally, a dopamine infusion When closing, it is once again important not to have a
(2.5–5 g/kg per minute) may be required. long delay before flow is restored. A suture is commenced
from each end. By this means, and by leaving a few loose
sutures adjacent to the shunt on each side, it is possible to
Operation remove the shunt, flush out from the external carotid,
internal carotid and common carotid arteries, wash the
It is important to check the position of the patient and the endarterectomy site with heparinised saline, complete the
patient’s head on the operating table before anaesthesia. arteriotomy and restore flow, once again within about one
146 Role of the vascular surgeon in managing stroke

and a half minutes. Flow is normally restored to the exter- deaths (1.9 per cent) comprising seven strokes and one
nal carotid artery first, so that if there are any loose platelet myocardial infarct. An independent audit showed that there
aggregations inadvertently left at the reconstruction site, were 22 (5.2 per cent) non-fatal strokes, a stroke being
despite the saline washing, they will embolise to the external defined as a neurological deficit present for more than
rather than the internal artery. It is our routine to perform 24 hours. Thus stroke or death occurred in 7.1 per cent.
a postoperative angiogram immediately after restoration of
flow and before closure of the wound.
In the past we did not use a patch routinely but we do so Conclusions
now. If the external carotid artery runs parallel to the internal
carotid artery, the external carotid artery is used as a patch We showed previously that when there was a high grade
following the technique described by Leather (see Bufo contralateral carotid stenosis the procedure of carotid
et al.11). Otherwise a fabric patch is used. endarterectomy presented an increased risk. These patients
Dextran is a polysaccharide compound commonly used of course have an increased risk of stroke anyway.7
as a volume expander but it also has antiplatelet and rheo- Patients with crescendo TIAs are clearly in a very high
logic properties. Although many vascular surgeons use dex- risk group. One has to accept, and explain to the patient
tran in the belief that it reduces perioperative stroke, no and relatives, that whatever course of action is under-
prospective randomised controlled trials have been per- taken, be it conservative or operative, the patient’s clini-
formed. Our hospital has initiated such a trial. Patients ran- cal condition is critical. We believe that, despite this,
domized to dextran receive an intravenous bolus of 20 mL of surgery produces a lower complication rate than non-
dextran 1 (Promit) 2 minutes before skin incision to avoid operative treatment.
anaphylaxis. This is followed by an intravenous infusion of Urgent carotid endarterectomy is one of the most sat-
1000 mL of 10 per cent dextran 40 in normal saline com- isfying operations in vascular surgery. However, to obtain
menced at the time of skin incision. The first 500 mL is good results, surgical technique must be impeccable,
administered over 4 hours (125 mL/hour) and the second patient selection is vital and of utmost importance is the
500 mL over the next 12 hours (42 mL/hour). Preliminary general organisation of hospital services. Patients who
studies using transcranial Doppler monitoring have demon- need such an operation require streamlined resources
strated that dextran reduces postoperative microemboli in capable of recognising the urgency of the situation and
the ipsilateral middle cerebral artery.12 The Dextran in treating it in appropriate and timely fashion.
Carotid Endarterectomy (DICE) Trial continues in order to
determine the effect of dextran on clinical outcome.
Key references
Postoperative surveillance Barnett HJM, Taylor DW, Eliasziw M, et al. Benefit of carotid
endarterectomy in patients with symptomatic moderate or
It is important that the patient’s neurological status, and in severe stenosis. N Engl J Med 1998; 339: 1415–25.
particular the blood pressure, are carefully monitored post- Donnan GA, O’Malley H, Hurley S, et al. The capsular warning
operatively. Any hypertension should be promptly treated, syndrome: pathogenesis and clinical features. Neurology
as described earlier. Hypertension as a postoperative prob- 1993; 43: 957–62.
lem is usually evident very soon after surgery. Bourke and European Carotid Surgery Trialists’ Collaborative Group. MRC
Crimmins have shown that in the absence of postoperative European Carotid Surgery Trial: interim results for
symptomatic patients with severe (70–99%) or mild (0–29%)
hypertension within 12 hours of surgery, it is safe to send
carotid stenosis. Lancet 1991; 337: 1235–43.
the patient home on the next day.13 All of our patients are
Kappelle LJ, Eliasziw M, Fox AJ, et al. Importance of intracranial
followed for life by clinical surveillance and duplex scan- atherosclerotic disease in patients with symptomatic stenosis
ning. Early restenosis, within two years of operation is usu- of the internal carotid artery. Stroke 1999; 30: 282–6.
ally due to fibrointimal hyperplasia, which occurs in 2 per Levi CR, Stork JL, Chambers BR, et al. Dextran reduces embolic signals
cent of patients. The risk of later restenosis is very low and after carotid endarterectomy. Ann Neurol 2001; 50: 544–7.
seldom requires reoperation, the risk of subsequent prob-
lems being very low.
Patients who have a small infarct have undergone oper-
ation without undue risk. However, when a small infarct is REFERENCES
seen on CT or MRI, we still advocate a delay of 4–6 weeks
before surgery. 1 Eastcott HHG, Pickering GW, Rob CG. Reconstruction of internal
carotid artery in a patient with intermittent attacks
of hemiplegia. Lancet 1954; 2: 994–6.
Recent results 2 European Carotid Surgery Trialists’ Collaborative Group. MRC
European Carotid Surgery Trial: interim results for symptomatic
Of 412 consecutive carotid endarterectomies performed patients with severe (70–99%) or mild (0–29%) carotid stenosis.
between January 1999 and November 2001, there were eight Lancet 1991; 337: 1235–43.
 References 147

3 North American Symptomatic Carotid Endarterectomy Trial 8 Kappelle LJ, Eliasziw M, Fox AJ, et al. Importance of
Collaborators. Beneficial effect of carotid endarterectomy in intracranial atherosclerotic disease in patients with symptomatic
symptomatic patients with high-grade carotid stenosis. stenosis of the internal carotid artery. Stroke 1999; 30:
N Engl J Med 1992; 325: 445–53. 282–6.
4 Barnett HJM, Taylor DW, Eliasziw M, et al. Benefit of carotid 9 Henderson RD, Eliasziw M, Fox AJ, et al. Angiographically defined
endarterectomy in patients with symptomatic moderate or collateral circulation and risk of stroke in patients with severe
severe stenosis. N Engl J Med 1998; 339: 1415–25. carotid artery stenosis. Stroke 2000; 31: 128–32.
5 Cartlidge NEF, Whisnant JPR, Elveback LR. Carotid and vertebral- 10 Lord RSA. Surgery of occlusive cerebrovascular disease.
basilar transient cerebral ischaemic attacks: a community study, St Louis: CV Mosby, 1986.
Rochester, Minnesota. Mayo Clin Proc 1977; 52: 117–20. 11 Bufo AJ, Shaf DM, Chang BB, Leather RP. Carotid
6 Donnan GA, O’Malley H, Hurley S, et al. The capsular warning bifurcationplasty: an alternative to patching. J Cardiovasc
syndrome: pathogenesis and clinical features. Neurology Surg 1992; 33: 308–10.
1993; 43: 957–62. 12 Levi CR, Stork JL, Chambers BR, et al. Dextran reduces embolic
7 Gasecki AP, Ferguson GG, Barnett HJM. Long-term prognosis and signals after carotid endarterectomy. Ann Neurol 2001; 50:
effect of endarterectomy in patients with symptomatic severe 544–7.
carotid stenosis and contralateral carotid stenosis or occlusion: 13 Bourke BM, Crimmins DC. Overnight hospital stay for carotid
results from NASCET. J Neurosurg 1995; 83: 778–82. endarterectomy. Med J Aust 1998; 168: 149–50.
This page intentionally left blank
 13
Surgical Experience in Evolving Stroke
Arising From the Carotid

R HUBER, RJ SEITZ, M SIEBLER, A AULICH, WILHELM SANDMANN

The problem 149 Discussion 152

Material and methods 149 References 154
Results 151

THE PROBLEM MATERIAL AND METHODS

The Joint Study of Extracranial Carotid Occlusion docu-

mented in 1969 the uselessness of emergency revascularisa- Selection based on clinical criteria
tion in patients with acute stroke and carotid occlusion.1,2
Only recently have neurologists accepted indications for • Conscious patient
urgent carotid surgery3,4 based on new diagnostic tech- • Acute/recurrent stroke
niques5–8 and close clinical observation during care in the • Recurrent/crescendo transient ischaemic attacks (TIAs)
stroke unit.9,10 Both the new and the old goals are preven-
tion of recurrent or disabling stroke and prevention and
Between 1 November 1997 and 31 December 2002 we
reduction of the extent of brain infarction. For decades
prospectively monitored all patients diagnosed in the stroke
surgeons undertook carotid endarterectomy to remove the
unit at our centre and selected patients were transferred to our
embolic source and to restore internal carotid artery (ICA)
operating theatre for urgent carotid surgery. Neurologists,
blood flow, but the indications for urgent reconstruction
neuroradiologists and vascular surgeons had initially col-
of the ICA were not defined precisely at that time. Since
laborated closely in arriving at a consensus on the criteria
then the diagnostic armamentarium, the surgical safety
for patient selection: (i) radiologically, the presence of acute
measures and the drugs available to discourage thrombosis
extracranial carotid occlusive disease, a patent intracranial
have changed substantially allowing new strategies to
component of the ICA, evidence of perfusion of the middle
develop in dealing with these acute cases. Magnetic
cerebral artery (MCA); and (ii) clinically, evidence of acute
resonance imaging (MRI) can reveal structural damage by
stroke, recurrent stroke or recent TIAs, possibly aggravated or
diffusion-weighted imaging (DWI) within 1 hour of onset
crescendo in nature within the previous week. Unconscious
of symptoms, and by comparing DWI and perfusion-
patients who required ventilation were excluded.
weighted imaging (PWI) the size of the infarct as well
as the tissue at risk can be determined at a very early
stage.6,7 By using this approach, patients who have a
Investigations: radiological/ultrasound
dangerous degree of ICA occlusive disease, but who also
have a substantial amount of potentially rescuable brain
tissue, in other words that forming the penumbra, can be • Duplex scan
identified. • Computed tomography (CT) scan of brain stem
150 Evolving stroke arising from the carotid

carotid artery before dissection of the bifurcation and the

• Magnetic resonance angiography (MRA)
distal ICA. Shunting was applied routinely after thrombec-
• Intra-arterial digital subtraction arteriography (DSA)
tomy of the distal ICA on the basis of earlier studies in
• Transcranial Doppler sonography (TCD)
which a patient with a history of previous stroke or CT
• Magnetic resonance imaging (MRI) (DWI/PWI)
findings positive for stroke, regardless of the hemisphere
involved, tended to have a better outcome with shunting.14
Patients in the stroke unit were placed on heparin which
Doppler and duplex sonography confirmed a diagnosis was maintained during surgery and continued at a low
of stenosis or acute occlusion. In cases of multivessel dosage postoperatively, followed by aspirin 100 mg daily.
involvement magnetic resonance arteriography and/or A total of 166 patients underwent 173 urgent carotid
intra-arterial digital subtraction arteriography (DSA) was operations (seven bilateral under the same anaesthetic),
performed. Intracranial circulation and embolism were representing 12.9 per cent of all carotid artery operations
assessed by transcranial Doppler sonography (TCD) and (n  1285) performed during the period in question.
high intensity transient signal (HITS) detection, respe- These patients included 119 men (71.7 per cent) and
ctively.5 A patent middle cerebral artery (MCA) was docu- 47 women (28.3 per cent), median age 64.7 years (range
mented by TCD, MRA or intra-arterial DSA. 31–86). Risk factors for both occlusive arterial disease and
vascular surgery were: hypertension in 133 (80.1 per cent),
smoking in 105 (63.3 per cent), cardiac disease in 84
Selection based on investigative findings (50.6 per cent), diabetes in 38 (22.9 per cent) and coagulation
disorders in 7 (4.2 per cent).
All patients underwent preoperative Doppler and
• Acute extracranial ICA disease
duplex sonography. Transcranial Doppler was performed
• Patent intracranial ICA
in 104 patients (62.7 per cent) and in 46 (44.2 per cent) of
• Perfusing MCA
these cases microemboli were detected. Intra-arterial DSA
• HITS
was performed in 83 of cases (50 per cent) and MRA in
• Evidence of penumbra
91 cases (54.8 per cent). Preoperative vascular findings were
ipsilateral carotid occlusion in 45 (27.1 per cent), tight
stenosis in 111 (66.9 per cent) and pseudo-occlusion in 10
Magnetic resonance imaging procedures were used to (6 per cent). Findings in the contralateral carotid artery
assess acute reversible and irreversible ischaemic brain were: normal in 116 (69.9 per cent), occluded in 27 (16.3
damage (see Chapter 12).6–8 In order to evaluate actual per cent), stenotic in 23 (13.8 per cent).
brain tissue damage CT or whenever possible MRI with Preoperative clinical findings were stroke in 98
DWI and PWI was performed. The cerebral area of critical (59.1 per cent) and TIA in 67 (40.3 per cent); one patient
ischaemia was visualised by means of colour-coded PWI. (0.6 per cent) had an asymptomatic acute ICA occlusion,
As shown earlier, a time-to-peak delay in perfusion imag- the progression from stenosis to carotid occlusion having
ing of greater than 4 seconds compared with the contralat- been discovered accidentally on repeat duplex sonography
eral hemisphere had to be considered ‘tissue at risk’.6,8 within 12 hours. Clinical presentations were: stroke with a
Because quantitative volumetry of the infarct, compared stable neurological deficit in 34 (20.5 per cent), stroke with
with the area of ‘misery perfusion’ or penumbra measured stuttering symptoms in 27 (16.3 per cent), progressive stroke
by DWI and PWI, is time consuming and not always avail- in 20 (12 per cent), stroke after TIA in 17 (10.2 per cent),
able outside normal working hours, it was only performed crescendo TIA in 40 (24.1 per cent), TIA with positive
in a small number of patients selected on precise clinical microemboli detected in 11 (6.6 per cent), crescendo TIA
criteria.11 Measurements were done before and 8 days after with microemboli detected in 11 (6.6 per cent) and recent
surgery. Impairment was assessed clinically using the TIA in 5 (3 per cent). In these patients angiographic find-
Rankin impairment scale12 and the Barthel index13 before ings were critical: small mobile thrombus in one, subtotal
surgery, 8 days afterwards and during follow-up. carotid artery stenosis in two, bilateral high grade stenosis
in two. None of the patients underwent surgery without a
CT scan or MRI. Extended MRI evaluation by means of
Objective clinical measurements DWI and PWI was undertaken in 104 patients (62.6 per
cent). Acute lesions with a large perfusion/diffusion mis-
• Rankin impairment scale match were found in 47 cases (45.2 per cent) among which
• Barthel index were 15 patients with TIA. Median time between onset of
first symptoms and operation was 4.5 days (range 4 hours
to 141 days).
Surgery was performed under general anesthesia and Reconstructive techniques applied were conventional
was standardised with early clamping of the common carotid endarterectomy with vein patch closure in 129
 Results 151

patients (77.7 per cent) and eversion endarterectomy15 in Mean age: 62  10 years
22 (13.3 per cent). In six cases (3.6 per cent) a vein inter- Neurological deficit preop: 71  15 (ESS*)

position graft had to be inserted. Additional thrombectomies Neurological outcome: 76  16 (ESS*)

70
of the petrous portion of the ICA were necessary in
29 patients. In nine patients (5.4 per cent) the carotid 60

Lesion volume (mL)

artery could not be reconstructed because backflow after 50

thrombectomy could not be achieved due to thrombosis 40

P  0.02
extending into the siphon, and the wound was closed after 30
thromboendarterectomy of the external carotid artery. In 20
seven additional cases, the presence of bilateral subtotal 10
occlusion led to contralateral ICA reconstruction under 0
the same anaesthetic. PWI, pre DWI, post
Severe ischaemia: Cytotoxic oedema
rCBF: 28  5 mL/100 g per minute of brain infarction

RESULTS Figure 13.1 Rescue of ‘tissue at risk’ in acute symptomatic

patients (emergent carotid endarterectomy). Patients were studied
pre- and post-operatively with perfusion-weighted magnetic
Subsequent to surgery 53 patients (31.9 per cent) remained resonance imaging (PWI) (n  17). *ESS, European Stroke Scale
free of symptoms. According to the Rankin scale12 clinical (0, severely abnormal; 100, normal). DWI, diffusion-weighted
disability improved in 63 (38 per cent) of the stroke patients. imaging
Thirty four stroke patients (20.5 per cent) remained stable,
although 11 did so with severe disability (Rankin 4, 5). Rankin (postoperative)
Sixteen patients (9.6 per cent) deteriorated: two had pre- 0 1 2 3 4 5 6 Total
sented with TIAs and progressed to complete stroke and 14
(8.4 per cent) showed worsening of stroke symptoms. In 0 53 2 55
Rankin (preoperative)

the TIA group one stroke occurred intraoperatively in a 1 6 12 18

patient with contralateral ICA occlusion having presented
primarily with crescendo TIAs and highly positive detec- 2 2 12 6 1 21
tion of microemboli, and in a second patient a large DWI
3 1 5 12 5 1 4 1 29
lesion present before surgery progressed during a hyper-
tensive crisis 6 days after surgery. Nine patients experi- 4 1 4 5 10 8 5 33
enced aggravation of MCA infarction symptoms, and one
5 1 4 3 2 10
of them with contralateral occlusion died 2 days post-
operatively. Three patients in whom flow through the ICA Total 63 33 24 18 13 12 3
could not be restored developed brain oedema which
resolved spontaneously in two of them. In the third patient Figure 13.2 Outcome of carotid surgery in acute symptomatic
a midline shift necessitated craniotomy. Another patient patients (Black: stable (n  34), free of symptoms (n  53); red:
with contralateral occlusion developed secondary infarc- improved and Rankin 0–3 (n  63), blue: deteriorated and
tion with brain oedema 5 days postoperatively and died 24 Rankin ⬎3 (n  16))
days later. One patient sustained rupture of the carotid
patch during gastroscopy 5 days postoperatively and died
12 days later. Of the 166 patients, only one suffered sec- Median follow-up until 31 December 2002 was 1.3 years
ondary intracerebral bleeding, detected on routine CT and was complete for all but nine patients. During this
scanning, but without manifesting any new symptoms. The time period all patients were on aspirin 100 mg/day. Ten
30-day mortality was 2.4 per cent (4/166). Postoperative patients died. The causes of death were: contralateral
neurological morbidity was 9.6 per cent (n  16), proce- intracerebral bleeding with ventricular rupture (n  1),
dure related peripheral nerve lesions (n  9) delayed cerebral infarction on the contralateral side (n  1), myocar-
wound healing (n  6) and reoperation for haematoma dial infarction (n  3), pulmonary embolism (n  1) and
(n  6) was 12.6 per cent. cancer (n  4). Survival after 1 year was 93.5 per cent and
The volume of ischaemic brain lesions decreased in after 2 years 91.4 per cent (Fig. 13.3).
17 patients until day eight as assessed by quantitative lesion All surviving patients are living in their own homes,
volumetry11 and was accompanied by improvement of the seven dependent on help from their families. No further
Barthel index13 (Fig. 13.1). Early postoperative neurological strokes or TIAs occurred. Three patients developed grand
assessment showed that 139 patients (83.7 per cent) bene- mal seizures after stroke and remain on anticonvulsive
fitted from carotid surgery and 27 (16.2 per cent) patients medication. One patient was reoperated for asymptomatic
remained severely disabled or deteriorated (Fig. 13.2). recurrent stenosis and has remained asymptomatic.
152 Evolving stroke arising from the carotid

haemorrhage occurred in one patient only and even then

1,0
no new symptoms had developed.
The results of the European Carotid Surgery Trial16,17
0,8
and the North American Symptomatic Carotid Endarterec-
0,6
tomy Trial18,19 and the therapeutic demands made by stroke
Survival

units9,10,20 in the wake of new imaging techniques have

0,4 changed attitudes to carotid surgery in acutely symptomatic
patients. The results of these studies and our growing experi-
0,2 ence show that urgent surgery can be performed in carefully
selected patients at risk of developing stroke, recurrent
0,0 stroke and/or progression of stenosis to complete occlusion,
if nothing other than anticoagulant therapy is given to alter
0 500 1000 1500
the spontaneous course of the disease. We have to ask our-
Overall survival (days)
selves: Why have the results of urgent surgery improved?
Figure 13.3 Overall survival (Kaplan–Meier) after carotid Although the reasons for intraoperative or postoperative
endarterectomy in acute symptomatic patients (n  166) deterioration are often a matter for speculation, they gener-
ally fall into three categories: (i) morphological abnormali-
ties of the luminal surface of the ICA itself, (ii) extracranial
DISCUSSION
and intracranial multivessel involvement and (iii) structural
brain tissue lesions.
We have come a long way since the days of condemnation Specific morphological abnormalities of the ICA which
of carotid artery reconstruction in patients presenting with test the technical skills of the surgeon include occlusive dis-
symptoms of acute stroke1 to the current recognition that ease located near the base of the skull, siphon involvement,
restoration of carotid flow in the acute case is of value.3,4 thrombus which is loose or not adherent to the vessel wall,
Although some surgeons have always insisted that surgery continuous embolisation and even dissection. In those
is indicated for this particular presentation, no clear guide- cases in our series in which deterioration followed attempts
lines and/or criteria defining the objectives of urgent at carotid reconstruction, the ICA, contrary to preopera-
carotid surgery have been spelt out. In earlier times how- tive angiographic appearances, was found at surgery to be
ever, clinicians were equipped with the patient’s history, occluded, the blockage extending intracranially. In those
the results of a meticulous neurological examination, an patients who deteriorated, embolisation during thrombec-
arteriogram of reasonably acceptable quality, an echo- tomy manoeuvres was the most probable cause. Nonetheless,
encephalogram for imaging of brain oedema and finally, a in our opinion, an acute carotid artery occlusion should
surgeon willing to perform the operation. always be disobliterated if the probability of infarction or
The major concern with urgent surgery was the pos- reinfarction is considered to be very high.21,22
sibility of intracranial haemorrhage thought to result from
rupture of the fragile capillaries in the infarction territory.
It became evident from CT scanning, that localised Causes of perioperative deterioration
oedema and small areas of bleeding can develop not only
around ischaemic infarcts but also within the brain of • Morphological abnormalities in the ICA
patients suffering from TIAs, but this process would not • Multisegmental vessel disease
necessarily progress to massive haemorrhage even though • Brain tissue lesions
anticoagulation was being used more aggressively. It was
felt that as long as the blood–brain barrier had not closed, Other reasons for aggressive intervention are: first, the
reconstruction of the ICA was too dangerous an undertak- probability of spontaneous embolisation into an intracranial
ing (see Chapter 5).2 At our institution we began to disre- artery,23,24 and second, the limitations of collateral circula-
gard the phenomenon of the unclosed blood–brain barrier tion in acute infarction, either because of haemodynamic
and our neurologists exhorted us to reconstitute carotid impairment within the territory of the exposed ICA, or, in
blood flow urgently. The indications were clinical instabil- the case of bilateral ICA occlusion, of the contralateral hemi-
ity in the presence of haemodynamic and/or angiographic sphere as well. Furthermore, outcome studies of re-explo-
evidence of deterioration of the occluding lesion and to ration have shown good long term results with the ICA
our surprise intracranial haemorrhage did not occur. remaining patent with a low risk of stroke recurrence.25 In
Today, with the availability of MRI derived assessment of patients with contralateral ICA occlusion and multivessel
structural and perfusion deficit, the decision to operate is involvement, clamping the common carotid artery can
made on the assessment of the volume of tissue at risk and cause haemodynamic impairment even if a temporary
the extent to which the patient might improve after elim- shunt is used, unless unrestricted flow through the shunt
inating the perfusion deficit. In our series intracranial can be ascertained. Placement of a shunt per se does not
 Discussion 153

guarantee flow and therefore we recommend continuous The results of surgery will always be classified clinically,
shunt flow measurement and neuromonitoring using but the influence of carotid flow in restoring the DWI/PWI
somatosensory evoked potentials. Additional stenosis or deficit, thereby salvaging brain tissue at risk, is an equally
occlusion of the MCA, undiagnosed preoperatively, may important and challenging issue. In eight patients in whom
present a further problem which we encountered only we were able to obtain preoperative and postoperative
once: a patient with concomitant contralateral internal MRIs, a decrease in the volume of the lesion and neurolog-
carotid occlusion, in whom shunting of a small calibre ICA ical improvement was evident using the Barthel score of
was impossible, had a shunt inserted into the external daily activities.13
carotid artery which probably helped to a degree in pre- During the spontaneous course of the disease early rein-
venting perfusion ischaemia. The MCA was diseased in farction rates of 1.29 per cent,28 as well as reinfarction rates
two cases of intraoperative embolisation in one of which of 12 per cent29 and mortality rates of 19 per cent30 within
evidence of embolism into the MCA was confirmed by the first year of onset of symptoms have to be considered.
comparing preoperative and intraoperative angiograms. In our patients, reinfarction within the former symptom-
The key challenge presented by structural intracerebral atic hemisphere related to the reconstructed ICA did not
lesions is estimating how much of the area recognised as occur, the one-year mortality was 6.5 per cent and two-
malperfused is actually irreversibly damaged using modern year mortality 8.6 per cent. During follow-up restenosis
imaging techniques. Magnetic resonance imaging reveals was detected only in one asymptomatic patient who was
structural damage in DWI within 1–2 hours after onset reoperated and has remained asymptomatic ever since.
of symptoms and is an early marker for brain infarction.6,7 Finally we pose this question: What might have happened
By means of parametric colour-coded PWI the cerebral area to this subset of patients presenting with symptoms of
of critical ischaemia is visualised. As shown earlier, a time- acute stroke symptoms prior to having redefined the case
to-peak delay in perfusion imaging of greater than four as for urgent carotid endarterectomy? Clearly, the sponta-
compared with the contralateral hemisphere, has to be neous course would have been significantly worse, other-
treated as ‘tissue at risk’.6,8 We worked on the basis that wise our neurology colleagues would not have considered
patients in whom the PWI area was significantly more those patients to be good candidates for carotid surgery.
extensive than the DWI lesion, reflecting the PWI/DWI area
mismatch, were ideal candidates for urgent carotid artery
reconstruction; this was because the volume mismatch of Conclusions
the two areas was anticipated to be the penumbra.26,27
In our series, clinical deterioration was observed either in On one hand, the traditional criteria for patient
patients with a large intracerebral structural deficit already selection in urgent carotid surgery such as evidence of
present in DWI at the time of operation, or in patients with extracranial carotid artery disease, neurological deficit,
smaller structural deficits in DWI but a longer interval loss of consciousness or not, will continue to be consider-
between onset of symptoms and surgery. The latter was the ed. On the other hand, the preservation of an open
case in two patients in whom the interval was 8 days and MCA, the status of embolisation, and PWI/DWI mis-
16 days. This leads to the critical issue of the optimal time match must also be taken into account. Although an
interval between the onset of symptoms and the operation. expert stroke team should be able to acquire the neces-
A balanced approach is required in stabilising the patient in sary information within hours, the timeframe which
the stroke unit, gathering the complex morphological, obtains in practice, even in a university hospital, is still
haemodynamic and clinical data needed, and scheduling the too long: at our centre a median of 4 days elapsed
patient for surgery. Our current experience suggests a period from onset of symptoms until surgery but recently that
of 36 hours after onset of symptoms as the time frame within interval has shortened. A neurological morbidity rate of
which the procedure should be undertaken. 9.6 per cent, a procedure related morbidity of 12.6 per cent
Another important issue remains to be discussed: the as well as a 30-day mortality of 2.7 per cent still remain
new and sensitive MRI techniques of DWI and PWI have higher than that observed in the 2262 patients who
shown, even more clearly than CT, that clinical classifica- underwent elective carotid surgery at our hospital from
tion of brain ischaemia cannot be correlated reliably to 1990 to 1999,31 namely, morbidity 4.8 per cent, 30-day
brain tissue status. A patient thought to have a TIA clin- mortality 0.9 per cent. Nevertheless, these outcome fig-
ically may easily be found to have a brain infarction as ures no longer represent an objection to urgent carotid
determined by MRI criteria. Two of our TIA patients deteri- surgery, because the alternative course is much worse.
orated preoperatively and in both cases, just before opera- The improvement in our series was achieved by detailed
tion, microembolism was detected and larger structural preoperative diagnosis and observing the indications for
brain lesions were confirmed on MRI. In our opinion, this carotid surgery based on clinical assessment, Doppler
underlines the fact that embolisation renders carotid sur- and duplex sonography, TCD for microemboli detec-
gery even more urgent in TIA patients, especially when tion, multimodal MRI and DSA.
HITS are detected by TCD despite heparin therapy.
154 Evolving stroke arising from the carotid

14 Sandmann W, Willeke F, Kolvenbach R, et al. Shunting and

Key references neuromonitoring: A prospective randomized study. In: Greenhalgh
RM (ed). Surgery for Stroke. London: WB Saunders, 1993: 287–96.
Baird AE, Benfield A, Schlaug G, et al. Enlargement of human cerebral 15 Raithel D. Carotid eversion endarterectomy: a better technique
ischemic lesion volumes measured by diffusion weighted than the standard operation? Cardiovasc Surg 1999; 5: 471–2.
magnetic resonance imaging. Ann Neurol 1997; 41: 581–9. 16 European Carotid Surgery Trialists’ Collaborative Group.
Futrell N, Millikan CH. Stroke is an emergency. Dis Mon 1996; 42: Randomised trial of endarterectomy for recently symptomatic
199–264. carotid stenosis: final results of the MRC European Carotid
Mahoney FI, Barthel DW. Functional evaluation: the Barthel index. Surgery Trial (ECST). Lancet 1998; 351: 1379–86.
Md State Med J 1965; 14: 61–5. 17 European Carotid Surgery Trialists’ Collaboratory Group. MRC
Neumann-Haefelin T, Wittsack HJ, Wenserski F, et al. Diffusion- European Carotid Surgery Trial: Interim results for symptomatic
and perfusion weighted MRI. The DWI/PWI mismatch region in patients with severe (70–90%) or with mild (0–29%) carotid
acute stroke. Stroke 1999; 30: 1591–7. stenosis. Lancet 1991; 337: 1235–43.
Sandmann W, Willeke F, Kolvenbach R, et al. Shunting and 18 North American Symptomatic Carotid Endarterectomy Trial
neuromonitoring: A prospective randomized study. In: Collaborators Group. Beneficial effect of carotid endarterectomy
Greenhalgh RM (ed). Surgery for stroke. London: WB Saunders, in symptomatic patients with high-grade carotid stenosis.
1993: 287–96. N Engl J Med 1991; 325: 445–53.
19 Gasecki AP, Ferguson GG, Eliasziw M, et al. Early endarterectomy
for severe carotid artery stenosis after nondisabling stroke:
REFERENCES results from the North American Symptomatic Carotid
Endarterectomy Trial. J Vasc Surg 1994; 27: 288–95.
20 Hacke W, Kaste M, Olsen TS, et al. Acute treatment of ischemic
1 Blaisdell WF, Clauss RH, Galbraith, et al. Joint Study of
stroke. Cerebrovasc Dis 2000; 10(suppl 3): 22–33.
Extracranial Carotid Occlusion IV. A review of surgical
21 Hennerici M, Hulsbomer HB, Rautenberg W, Hefter H.
considerations. JAMA 1969; 209: 1889–95.
Spontaneous history of asymptomatic internal carotid occlusion.
2 Thompson JE, Austin DJ, Patman PD. Endarterectomy of the
Stroke 1986; 17: 718–22.
totally occluded internal carotid artery for stroke, results in 100
22 Klijn CJM, Kappelle JK, van der Grond J, et al. Magnetic
operations. Arch Surg 1967; 95: 791–801.
resonance techniques for the identification of patients with
3 Pritz MB. Timing of carotid endarterectomy after stroke. Stroke
symptomatic carotidartery occlusion at high risk of cerebral
1997; 28: 2563–7.
ischemic events. Stroke 2000; 31: 3001–7.
4 Whittemore AD, Ruby ST, Couch NP, Mannick JA. Early carotid
23 Barber PA, Davis SM, Darby DG, et al. Absent middle cerebral
endarterectomy in patients with small, fixed neurologic deficit.
artery flow predicts the presence and evolution of the ischemic
J Vasc Surg 1984; 1: 795–9.
penumbra. Neurology 1999; 52: 1125–32.
5 Ringelstein EB, Droste DW, Babikian VL, et al. Consensus on
24 Kniemeyer HW, Aulich A, Schlachetzki F, et al. Pseudo- and
microembolus detection by TCD. International Consensus Group
segmental occlusion of the internal carotid artery: a new
on Microembolus Detection. Stroke 1998; 29: 725–9.
classification, surgical treatment and results. Eur J Vasc Endovasc
6 Neumann-Haefelin T, Wittsack HJ, Wenserski F, et al. Diffusion-
Surg 1996; 12: 310–20.
and perfusion weighted MRI. The DWI/PWI mismatch region in
25 von Arbin M, Britton M, de Faire U. Mortality and recurrences
acute stroke. Stroke 1999; 30: 1591–7.
during eight years following stroke. J Intern Med 1992; 231:
7 Baird AE, Benfield A, Schlaug G, et al. Enlargement of human
43–8.
cerebral ischemic lesion volumes measured by diffusion weighted
26 Astrup J, Siesjö BK, Symon L. Thresholds of cerebral ischemia: the
magnetic resonance imaging. Ann Neurol 1997; 41: 581–9.
ischemic penumbra. Stroke 1981; 12: 723–5.
8 Neumann-Haefelin T, Wittsack HJ, Fink GR, et al. Diffusion- and
27 Lassen NA. Pathophysiology of brain ischemia as it relates to
perfusion weighted MRI. Influence of severe carotid artery
the therapy of acute ischemic stroke. Clin Neuropharmacol 1990;
stenosis on the DWI/PWI mismatch in acute stroke. Stroke 2000;
13: S1.
31: 1311–17.
28 Sandercock P, Tangkanakul C. Very early prevention of stroke
9 Bath P, Butterworth RJ, Soo J, Kerr JE. The King’s College
recurrence. Cerebrovasc Dis 1997; 7(suppl 1): 10–15.
Hospital Acute Stroke Unit. J R Coll Phys Lond 1996; 30: 13–17.
29 Easton DJ. Epidemiology of stroke recurrence. Cerebrovasc Dis
10 Futrell N, Millikan CH. Stroke is an emergency. Dis Mon 1996;
1997; 7(suppl 1): 2–4.
42: 199–264.
30 Dennis MS, Burn JPS, Sandercock PAG, et al. Long term survival
11 Wittsack HJ, Ritzl A, Fink GR, et al. Magnetic resonance imaging
after first ever stroke: the oxfordshire community stroke project.
in acute stroke: DWI and PWI. parameters predicting infarct size.
Stroke 1993; 24: 796–800.
Radiology 2002; 222: 397–403.
31 Ommer A, Dziewanowski M, Pillny M, et al. Langzeitergebnisse
12 Rankin J. Cerebral vascular accidents in patients over the age of
nach Carotisrekonstruktion im Defektstadium der
60. Prognosis. Scottish Med J 1957; 2: 200–15.
cerebrovaskulären Verschlu krankheit. Dtsch Med Wochenschr
13 Mahoney FI, Barthel DW. Functional evaluation: the Barthel
2002; 127: 370–5.
index. Md State Med J 1965; 14: 61–5.
 14
Post-Carotid Endarterectomy Stroke

DAVID ROSENTHAL, ERIC D WELLONS

The problem 155 Late neurological deficits 158

Early postoperative deficits 155 References 159
Discussion 156

THE PROBLEM 66.5 years (range 40–86). A history consistent with cor-
onary artery heart disease documented by electrocardio-
gram (ECG) was present in 716 (66 per cent), hypertension
After nearly half a century of carotid artery surgery, the in 662 (61 per cent), diabetes mellitus in 391 (36 per cent)
pathogenesis and management of the postoperative neuro- abnormal lipid profiles in 434 (40 per cent) and a history of
logical deficit remains controversial. Several reports1–5 cigarette smoking was elicited in 521 (48 per cent).
believe reperfusion injury or technical error resulting in Postoperative neurological deficits were classified into
thromboembolic events to be the cause, whereas others6–8 three categories: (i) a focal episode of neurological dys-
attribute cerebral ischaemia during carotid occlusion to function which resolved within 24 hours was defined as
the pathogenesis of postoperative deficits. It is often diffi- a TIA; (ii) a neurological deficit which lasted more than
cult to determine the exact aetiology of a postoperative 24 hours, yet resolved completely within 3 weeks was des-
deficit, but each of these factors may play some role. ignated a reversible ischaemic neurological deficit (RIND),
The purpose of this chapter is to discuss the most likely whereas (iii) a fixed non-progressive neurological deficit
causes of early and late neurologic deficits after carotid lasting less than 24 hours and caused by cerebral infarction
endarterectomy (CEA) and to evaluate the safest and most was characterised as a stroke.9
efficient means of managing such patients in order to min- The incidence of postoperative neurological deficit was
imise permanent neurological impairment. not significantly different if patients underwent CEA with
routine shunting, routine non-shunting or selective shunt-
ing based on electroencephalogram (EEG) criteria). Of the
EARLY POSTOPERATIVE DEFICITS
patients who underwent CEA with a shunt (n  512), a
transient deficit occurred in 15 (2.9 per cent) and a perman-
Between 1980 and 1999, 1085 patients underwent 1238 ent deficit in 11 (2.1 per cent). When CEA was performed
CEAs. The indications for operation included hemispheric without a shunt (n  274), a transient deficit occurred in
transient ischaemic attacks (TIAs) (n  579), symptoms of eight patients (2.9 per cent) and a permanent deficit
vertebrobasilar insufficiency (n  248), reversible ischaemic occurred in nine (3.3 per cent). Transient deficit after
neurological deficit (RIND) or stroke (n  165), prophy- CEA monitored by EEG surveillance (n  452) occurred
lactic CEA (n  111), amaurosis fugax (n  108) and in 12 (2.7 per cent) and permanent deficit occurred in
stroke in evolution (n  27). 11 patients (2.4 per cent). Overall, 35 patients experienced
Based on the different operative techniques employed, a transient postoperative deficit and 31 a permanent deficit.
the patients were categorised into three groups: (i) CEA Although the incidence of postoperative stroke (3.3 per cent)
performed with a shunt (n  512); (ii) CEA without a was slightly higher in the routine non-shunting group,
shunt (n  274); and (iii) CEA monitored by electroenceph- there was no significant statistical difference when the dif-
alogram (EEG) surveillance (n  452). Of the 1085 ferent methods of cerebral protection were compared
patients, 705 (65 per cent) were men. The mean age was (P  6.25).
156 Post-carotid endarterectomy stroke

The preoperative arteriograms of all patients experiencing more pronounced deficit had a computed tomography
a postoperative neurological deficit were reviewed. Of 35 (CT) brain scan within 72 hours of operation: all scans
patients who experienced a transient postoperative deficit, 25 were normal and all patients regained neurological func-
patients (72 per cent) had ulcerated plaque disease, as did 10 tion within 1 month of operation.
of 31 (32 per cent) who sustained a postoperative stroke. This A total of 35 patients with immediate profound post-
suggests that embolisation from the ulcer bed may have been operative deficit and a suspected stroke required emergency
the cause of the postoperative deficit. Intracranial occlusive operation. In 31 cases a deficit was identified upon awaken-
disease, namely, stenosis at the siphon or stenoses within the ing from anaesthesia and four deficits occurred in the recov-
circle of Willis, was also noted in 29 patients. ery room within 3 hours of operation. A patent CEA site
The incidence of postoperative neurological complica- was identified in 15 patients and arteriography verified an
tions was evaluated on the basis of indications for opera- intracranial embolic shower in 12, and three arteriograms
tion. It is of interest to note that patients with ‘stable’ were normal. In 20 patients a thrombosed carotid artery was
preoperative symptoms such as TIAs, amaurosis fugax and found caused by technical errors: an intimal flap was identi-
vertebrobasilar insufficiency had nearly a fourfold increased fied in 11, a lateral tear in three and a residual plaque in two.
incidence of transient and permanent deficits over patients Four other patients underwent thrombectomy of platelet–
whose neurological status was ‘unstable’, namely those pre- fibrin aggregates or ‘white clot’. Despite arterial re-exploration
senting with stroke in evolution, RIND or stroke (Table and arteriography a cause could not be found and therefore it
14.1), and these features have been observed by others.10,11 was assumed to be due to heparin induced thrombocy-
After a RIND or stroke or during a stroke in evolution, topenic thrombosis. After thrombectomy and correction of
a zone of ischaemic brain tissue is present which may the technical error, a patch graft was constructed in 14
be more vulnerable to diminished perfusion during patients and in six the carotid bifurcation was replaced
carotid cross-clamping than normal brain tissue.12,13 This with saphenous vein. Of these 35 patients, three showed
ischaemic zone is supplied by highly resistant collateral ves- immediate return of neurological function, eight improved
sels and a drop in perfusion pressure during carotid cross- slowly, 20 were unchanged and four died of stroke related
clamping may cause further ischaemia. In order to diminish causes.
the potential for a postoperative neurological deficit in
these neurologically ‘unstable’ patients, the wisest course of
action is to shunt the carotids in order to avoid any drop in
DISCUSSION
perfusion during CEA.
In the immediate postoperative period (12 hours),
35 patients experienced a transient neurological deficit. Early neurological deficits after CEA occur infrequently and
Seventeen patients experienced a focal minor deficit, which both the causative factors and the management of these
resolved rapidly, and 18 patients experienced a more pro- patients remains undefined. A comparison of the post-
nounced deficit such as contralateral sensory and motor operative neurological deficit with the preoperative arteri-
changes of the face and extremities upon awakening from ograms reveals two findings of interest. First, most
anaesthesia. Fifteen were reanaesthetised preparatory to complications (72 per cent transient deficit, 32 per cent per-
immediate operative arteriography. Carotid re-exploration manent deficit) occurred in patients who had ulcerative
was necessary in four patients, and the others had normal plaque disease identified at arteriography. It seems likely
arteriograms. Each of the 18 patients who experienced a that patients with ulcerated plaque disease are more prone
to embolic events during carotid artery mobilisation, where
Table 14.1 Postoperative deficit/indication for operation
intraluminal cellular debris is not adherent to the ulcer
bed, compared with patients who have calcific high grade
Transient Permanent obstructive lesions. Second, of the 65 patients who suffered a
Number Per cent (n) Per cent (n) postoperative deficit, 29 had intracranial occlusive disease
represented by stenoses within the siphon or circle of Willis.
Neurologically ‘stable’ Severe intracranial arterial disease has been previously
TIA 579 2.0 (12) 2.2 (13) demonstrated to place patients at high risk of developing
VBI 248 2.8 (7) 2.3 (6) postoperative neurological complications, while neither
Prophylactic 111 1.8 (2) 1.8 (2)
recurrent stenosis nor contralateral occlusion appear to
Amaurosis fugax 108 2.7 (3) 0.9 (1)
increase the stroke risk.14–16 Extracranial arterial occlusive
Total 2.3 1.8
disease indicates the severity of cerebrovascular insufficiency
Neurologically ‘unstable’ present and the surgeon may selectively shunt this patient,
Stroke in evolution 27 48.8 (4) 11.0 (3) but intracranial arterial occlusive disease is the limiting fac-
RIND/stroke 165 4.2 (7) 3.6 (6) tor of cerebral ischaemia during carotid cross-clamping. In
Total 9.5 7.3
the absence of intraoperative EEG surveillance in the patient
TIA, transient ischaemic attack; VBI, vertebrobasilar insufficiency; with severe intracranial arterial occlusive disease, the better
RIND, reversible ischaemic neurological deficit. part of surgical wisdom would be to use a shunt.
 Discussion 157

Since most surgeons have little experience with neurolog- be remembered, however, that postoperative neurological
ical deficit after endarterectomy, a succinct management deficits caused by ‘embolic showers’ are likely to be made
schema is necessary (Fig. 14.1). When a focal, minor tran- up of cholesterol–platelet–fibrin aggregates, and in this set-
sient deficit or suspected TIA occurs and resolves within ting thrombolytic agents may have little benefit.
minutes, supportive non-operative treatment is most appro- Continued investigation of this therapeutic approach is,
priate. This is the patient in whom urgent, carotid colour- nevertheless, warranted.
flow duplex ultrasonography should be performed to When a pulseless, thrombosed endarterectomy site is
evaluate the operative site. If, however, there is progression of found, the common and external carotid arteries are cross-
the deficit, symptoms that wax and wane, the surgeon is clamped, the endarterectomy site is opened, and the internal
unsure or the ultrasound examination is abnormal, then the carotid artery is allowed to back-bleed freely in the hope of
safest and most expeditious means of managing the patient is washing out any thromboembolic material. If no back-
prompt return to the operating room for neck exploration. bleeding occurs, thromboembolectomy using a no. 2
Fogarty catheter, and very gently executed to avoid cavernous
sinus injury, is appropriate. Once back-bleeding is estab-
• A neurological deficit immediately after operation lished, a temporary shunt is inserted to ensure restoration of
calls for emergency re-exploration cerebral blood flow. Technical errors should be corrected
but, unfortunately, these may not always be found. The arte-
riotomy should be closed with a patch graft or the artery
At re-exploration if a pulsatile artery is found, intraop- replaced with saphenous vein depending on the condition of
erative arteriography is performed through a common the artery. A completion arteriogram is then mandatory.
carotid puncture proximal to the endarterectomy site. In As mentioned above, four endarterectomy sites were
this manner, the common carotid clamp site as well as the thrombosed by white clot composed of platelet–fibrin
extracranial and intracranial internal carotid components aggregates. This was probably heparin induced thrombo-
can be visualised. If no defects are identified, the patient is cytopenic thrombosis caused by either heparin dependent
simply kept under observation. If, however, there is evi- platelet membrane antibodies, which induce platelet aggre-
dence of intracranial embolism current wisdom would gation in the present of heparin,21,22 or a disequilibrium in
indicate the administration of local intra-arterial throm- the balance of the prostaglandin systems (thromboxane A2
bolysis. Urokinase (1 MU (million units) in 100 mL saline and prostacyclin), affecting platelet proaggregation and
over 1 hour) or tissue plasminogen activator (tPA) disaggregation activity. Not a great deal of information
(10 mg in 100 mL saline over 1 hour) may be adminis- exists on the ‘white clot syndrome’ after CEA, but faced
tered through a microcatheter inserted via the internal with such a problem the wisest course of action may be sys-
carotid artery intracranially up to middle cerebral artery temic infusion of low molecular weight dextran, replace-
level or via an indwelling shunt.17–20 These reports are ment of the endarterectomised segment with saphenous
anecdotal, but offer encouraging results (Fig. 14.2). It must vein and adjunctive use of clopidogrel or ticlopidine.

Transient deficit Profound deficit

(suspected TIA) (suspected stroke)

Rapidly resolving Progressive Operating room Pulseless artery

symptoms/signs deficit

Pulsatile artery Thromboembolectomy

Ultrasonography Abnormal

Intraoperative arteriography Technical error

Normal

Intracranial emboli Patch/replace artery

Observe Figure 14.1 Management

Urokinase/tPA? Urokinase/tPA? schema in post-carotid
endarterectomy stroke.
TIA, transient ischaemic attack;
tPA, tissue plasminogen
Support ‘Completion’ arteriogram
activator
158 Post-carotid endarterectomy stroke

(a) (b) (c)

Figure 14.2 (a) Thrombosed intracranial internal carotid artery (ICA) after left carotid endarterectomy; (b) ICA 20 minutes after tissue
plasminogen activator (tPA) infusion; and (c) ICA 60 minutes after tPA infusion

When a profound deficit in the form of a suspected LATE NEUROLOGICAL DEFICITS

stroke occurs upon ‘awakening’ from anaesthesia, the
patient is immediately reanaesthetised and the wound
reopened. If the deficit becomes apparent in the recovery Late neurological deficit after CEA, which occurs in a very
room, the patient is immediately returned to the operating small percentage of patients, is represented by two patho-
room and the steps shown in the algorithm in Fig. 14.1 are logical entities, namely, the cerebral hyperperfusion syn-
followed. The ‘take home’ message from the algorithm is drome and intracerebral haemorrhage. Risk factors which
that the safest and most expeditious step in managing these place patients at risk of developing hyperperfusion syn-
patients is immediate wound exploration in the operating dromes are as follows: the relief of a high grade, preocclu-
room. Once in the operating room, arteriography, carotid sive stenosis in the presence of a contralateral occlusion or
exploration, thrombolytic therapy or any combination of contralateral high grade stenosis, recent (3 months) con-
these may be performed without delay. tralateral CEA, a history of hypertension, postoperative use
of systemic anticoagulants, an increase in cerebral blood
flow of 100 per cent or more measured at the time of oper-
• Technical errors account for most postoperative ation and patterns of cerebrovascular chronic hyperperfu-
strokes sion.25,26 In areas of chronic cerebral ischaemia the
‘autoregulatory mechanism’ involving the intracerebral
In our experience the incidence of postoperative neuro- arterial system is known to be impaired. Brain tissue nor-
logical deficits when CEA was performed with a shunt, mally has the ability to resist changes in blood flow by the
without a shunt or under EEG surveillance has not been reflex contraction and relaxation of the intracerebral ves-
statistically different (see above). The concept of inade- sels as blood pressure increases. In the chronically
quate cerebral collateral flow during CEA, therefore, can- ischaemic brain, this autoregulatory mechanism is lost and
not be incriminated as the cause of the postoperative the arteriolar vessels tend to remain maximally dilated.
neurological complications. Technical errors causing With restoration of flow into a preocclusive internal
carotid thrombosis or cerebral embolism rather than carotid artery the delicate intracerebral small vasculature is
inadequate collateral cerebral flow, account for most cases suddenly exposed to an uncontrolled high pressure head.
of neurological deficit after CEA. When an early postoper-
ative neurological deficit does occur, however, the most
appropriate management is the immediate return of the • The late development of a neurological deficit points
to a diagnosis of cerebral hyperperfusion syndrome
patient to the operating room, which provides the opportu-
or intracerebral haemorrhage
nity to confirm the diagnosis and to take appropriate
action. Although immediate reoperation is mandatory, the
clinical outcome may not be significantly altered by this The onset of symptoms associated with a hyperperfu-
course of action.23,24 sion syndrome is heralded by a hemicranial vascular
 References 159

headache ipsilateral to the side of the endarterectomy. of incidence of intracerebral haemorrhage after CEA is less
Additionally, the patient may experience uncontrollable than 0.5 per cent.31
hypertension, seizures and migrainous episodes associated In summary, the symptom of severe hemicranial
with visual disturbances such as flashing lights or sco- headache after CEA mandates swift evaluation and CT,
tomata. The treatment for these symptoms is appropriate MRI, TCD, anticoagulation profile and an EEG are appro-
blood pressure control, possibly steroids to help stabilise priate. If the EEG reveals lateralising paroxysmal activity, the
the blood–brain barrier and antiseizure medication when TCD demonstrates consistently elevated velocities or the
indicated.27 Considering the possible risk of intracerebral CT/MRI shows patchy oedema, the patient may be started
haemorrhage, it is safest to avoid any type of analgesic on anticonvulsants; any antiplatelet or anticoagulant med-
which has anticoagulant or antiplatelet effects. ication is stopped and blood pressure is carefully controlled.

• Severe hemicranial headaches after CEA often herald Conclusions

the cerebral hyperfusion syndrome
Stroke after CEA is a devastating event and a succinct,
Cerebral CT scan and EEG studies may be of benefit for systematic management schema is necessary. A
establishing a diagnosis of the cerebral hyperperfusion syn- postoperative major deficit mandates immediate return
drome. A postoperative CT scan may demonstrate patchy to the operating room where arteriography, carotid
oedema, and an EEG demonstrates paroxysmal lateralising exploration, thrombolytic therapy or a combination of
epileptiform discharges (PLEDs) ipsilateral to the side of these may be performed without delay. A minor or tran-
endarterectomy, indicating a localised focus of irritability. sient deficit, which resolves in minutes, necessitates
A transient neurological deficit is occasionally seen with a urgent colour-flow duplex ultrasonography. If, however,
hyperperfusion syndrome and is almost exclusively a pos- there is any progression of the deficit, prompt return to
tictal paresis.27 Transcranial Doppler (TCD) and magnetic the operating room is indicated. Late neurological events
resonance imaging (MRI) may prove useful in establishing after CEA, which occur in a very small percentage of
a definitive diagnosis of the hyperperfusion syndrome. patients, are the cerebral hyperperfusion syndrome and
Transcranial Doppler normally shows a return of velocities intracerebral haemorrhage. These are generally heralded
in the middle cerebral artery to the 40 mL/s range after a by hypertension and severe hemicranial headache ipsilat-
period of reactive hyperaemia. A greater than 100 per cent eral to the side of CEA and are best managed medically.
increase in peak velocity or pulsatility index of the middle
cerebral artery compared with baseline values is consistent
with the hyperperfusion syndrome.28 Meticulous blood Key references
pressure control in these patients is essential.
Comerota AJ, Eze AR. Intraoperative high-dose regional urokinase
infusion for cerebrovascular occlusion after carotid
• A delayed neurological deficit should be evaluated endarterectomy. J Vasc Surg 1996; 24: 1008–16.
systematically to establish cause: is it a technical Pomposelli FB, Lamparello PJ, Riles TS, et al. Intracranial hemorrhage
error or an incorrigible event? after carotid endarterectomy. J Vasc Surg 1988; 7: 240–7.
Reigel MM, Hollier LH, Sundt TM Jr, et al. Cerebral hypoperfusion
syndrome: a cause of neurologic dysfunction after carotid
Intracerebral haemorrhage is the most feared sequela of endarterectomy. J Vasc Surg 1987; 5: 628–34.
the hyperperfusion syndrome. Risk factors associated with Rockman CB, Jacobowitz GR, Lamparello PJ, et al. Immediate
intracerebral haemorrhage are virtually identical to the reexploration for the perioperative neurologic event after
hyperperfusion syndrome and include recent cerebral carotid endarterectomy: is it worthwhile? J Vasc Surg 2000;
infarction, hypertension, relief of a high grade, preocclusive 32: 1062–70.
stenosis associated with drastic increases in cerebral blood Rosenthal D, Zeichner WD, Pano LA, Stanton PE Jr. Neurologic
flow and the use of anticoagulants. Presenting symptoms of deficit after carotid endarterectomy; pathogenesis and
management. Surgery 1983; 94: 776–80.
intracerebral haemorrhage, which also occurs 3–7 days after
endarterectomy are severe hemicranial headache, seizures,
severe unremitting hypertension and a dilated unilateral
pupil.29,30
Treatment is again symptomatic, but the most appro- REFERENCES
priate approach ought to be preventive, first, by recognis-
ing the symptoms and second, by controlling the patient’s 1 Riles TS, Imparato AM, Jacobowitz GR, et al. The cause of
blood pressure effectively, because once haemorrhage perioperative stroke after carotid endarterectomy. J Vasc Surg
occurs, herniation and death are all too common. The rate 1994; 19: 206–16.
160 Post-carotid endarterectomy stroke

2 Rosenthal D, Zeichner WD, Pano LA, Stanton PE Jr. Neurologic 16 Domenig C, Hamdan AD, Belfield AK, et al. Recurrent stenosis
deficit after carotid endarterectomy; pathogenesis and and contralateral occlusion: high-risk situations in carotid
management. Surgery 1983; 94: 776–80. endarterectomy? Ann Vasc Surg 2003; 17: 622–8 [epub ahead
3 Hertzer NR, Beven EG, Greenstreet RL, Humphries AW. Internal of print 23 October 2003].
carotid artery back pressure, intraoperative shunting, ulcerated 17 Comerota AJ, Eze AR. Intraoperative high-dose regional
atheromata, and the incidence of stroke during carotid urokinase infusion for cerebrovascular occlusion after carotid
endarterectomy. Surgery 1978; 83: 306–12. endarterectomy. J Vasc Surg 1996; 24: 1008–16.
4 Hingorani A, Ascher E, Tsemekhim B, et al. Causes of early post 18 Barr JD, Harowitz MB, Mathis JM, et al. Intraoperative urokinase
carotid endarterectomy stroke in a recent series: the increasing infusion for embolic stroke during carotid endarterectomy.
importance of hyperperfusion syndrome. Acta Chir Belg 2002; Neurosurgery 1995; 36: 606–11.
102: 435–8. 19 Del Zoppo GJ. Investigational use of tPA in acute stroke.
5 Laman DM, Wieneke GH, van Duijin H, van Huffelen AC. High Ann Emerg Med 1988; 11: 1196–201.
embolic rate early after carotid endarterectomy is associated 20 Chalela JA, Katzan I, Liebskind DS, et al. Safety of intra-arterial
with early cerebrovascular complications, especially in women. thrombolysis in the postoperative stroke. Stroke 2001; 32:
J Vasc Surg 2002; 36: 278–84. 6:1365–9.
6 Frawley JE, Hicks RG, Beardon M, Woodey R. Hemodynamic 21 Kapsch D, Silver D. Heparin-induced thrombocytopenia and
ischemic stroke during carotid endarterectomy; an appraisal of hemorrhage. Arch Surg 1981; 116: 1423–9.
risk and cerebral protection. J Vasc Surg 1997; 25: 611–19. 22 Adams JG, Humphrey LJ, Zhang X, Silver D. Do patients with
7 Archie JP Jr. Technique and clinical results of carotid stump back heparin-induced thrombocytopenia syndrome have heparin
pressure to determine selective shunting during carotid specific antibodies? J Vasc Surg 1995; 21: 247–54.
endarterectomy. J Vasc Surg 1991; 13: 319–27. 23 Stewart AH, McGrath CM, Cole SE, et al. Reoperation for
8 Owens MC, Wilson SE. Prevention of neurologic complications of neurological complications following carotid endarterectomy.
carotid endarterectomy. Arch Surg 1982; 117: 551–5. Br J Surg 2003; 90: 832–7.
9 Moore WS, Barnett HJM, Beebe HG, et al. Guidelines for carotid 24 Findlay JM, Marchak BE. Reoperation for acute hemispheric
endarterectomy; a multidisciplinary consensus statement from stroke after carotid endarterectomy: is there any value?
the ad hoc committee, American Heart Association. Stroke 1995; Neurosurgery 2002; 50: 486–92.
26: 188–201. 25 Waltz AG. Effect of blood pressure on blood flow in ischemic and
10 Whittemore AD, Ruby ST, Couch NP, et al. Early carotid in non-ischemic cerebral cortex. Neurology 1968; 18: 613–21.
endarterectomy in patients with small fixed neurologic deficits. 26 Ascher E, Markevich N, Schutzer RW, et al. Cerebral
J Vasc Surg 1984; 1: 795–9. hyperperfusion syndrome after carotid endarterectomy: predictive
11 Giodano JM, Trout HH, Kozloff L, DePalma RG. Timing Carotid factors and hemodynamic changes. J Vasc Surg 2003; 3: 769–77.
Arterial Endarterectomy Surgery after Stroke. J Vasc Surg 1985; 27 Reigel MM, Hollier LH, Sundt TM Jr, et al. Cerebral hypoperfusion
2: 250–5. syndrome: a cause of neurologic dysfunction after carotid
12 Pomposelli FB, Lamparello PJ, Riles TS, et al. Intracranial endarterectomy. J Vasc Surg 1987; 5: 628–34.
hemorrhage after carotid endarterectomy. J Vasc Surg 1988; 28 Zanette EM, Fieschi C, Bozzao I, et al. Comparison of cerebral
7: 240–7. angiography and transcranial doppler sonography in acute
13 Rothwell PM, Slattery J, Warlow CP. A systematic comparison stroke. Stroke 1989; 20: 899–903.
of the risks of stroke and death due to carotid endarterectomy 29 Piepgras DG, Morgan MK, Sundt TF Jr, et al. Intracerebral
for symptomatic and asymptomatic stenosis. Stroke 1996; hemorrhage after carotid endarterectomy. J Neurosurg 1988;
27: 266–9. 68: 532–36.
14 Thompson JE, Talkington CM. Carotid endarterectomy. Adv Surg 30 Pomposelli FB, Lamparello PJ, Riles TS, et al. Intracranial hemorrhage
1993; 26: 99–131. after carotid endarterectomy. J Vasc Surg 1985; 51: 114–15.
15 Tu JV, Wang H, Bowyer B, et al. Risk factors for death or stroke 31 Rockman CB, Jacobowitz GR, Lamparello PJ, et al. Immediate
after carotid endarterectomy. Observations from the Ontario reexploration for the perioperative neurologic event after carotid
Carotid Endarterectomy Registry. Stroke 2003; 34: 2568–73 endarterectomy: is it worthwhile? J Vasc Surg 2000; 32:
[epub ahead of print 2 October 2003]. 1062–70.
 SECTION
3
Acute Lower Limb Ischaemic States

15. Acute limb ischaemia: surgical options 163

16. Acute limb ischaemia: endovascular options 181

17. Graft maintenance and graft failure 197

18. Acute ischaemia secondary to occult prosthetic graft infection 211

19. The acute diabetic foot 223

This page intentionally left blank
 15
Acute Limb Ischaemia: Surgical Options

CLIFFORD P SHEARMAN, MALCOLM H SIMMS

The problem 163 Specific techniques 176

Diagnosis 164 Postoperative management 178
Patient and treatment selection 167 Results 178
Surgical intervention 170 References 178

THE PROBLEM changes in symptoms and signs may be subtle and escape
notice. While early recognition of impending gangrene is
essential, it is just as important for the surgeon to recognise
Each year between 500 and 1000 per million of the population the onset of irreversible tissue death so that futile and
of northern Europe will suffer limb-threatening ischaemia potentially hazardous revascularisation procedures are
due to atherosclerotic vascular disease.1 The majority of avoided when amputation is the appropriate management.
these patients present with progressive deterioration in their Acute ischaemia and gangrene may be precipitated by
condition and can be offered elective vascular reconstruction, spontaneous thromboembolism, acute deterioration in
resulting in limb salvage in up to 90 per cent.2 Although chronic ischaemia, failure of an established arterial bypass
demanding in terms of hospital resources, revascularisation graft (see Chapter 2) or the consequences of either accidental,
currently offers the best therapeutic chance of the patient intentional or iatrogenic arterial trauma (see Chapters 38, 39
remaining ambulant and independent. Importantly, it is and 40). Arterial trauma is discussed elsewhere in this book
highly cost-effective when compared with amputation.3,4 (see Chapter 33). Intradermal ischaemia producing digital
Of the cohort of patients suffering severe limb ischae- and cutaneous gangrene is characteristic of extensive arterio-
mia, a certain proportion, 20 per cent in most centres, lar thrombosis secondary to diffuse endothelial injury, as seen
present with a precipitate worsening of the condition in some forms of septicaemia, immune complex deposition,
requiring urgent or emergency intervention. When limb toxic injection and poorly controlled diabetes (see Chapter
blood flow becomes insufficient to support resting tissue 19). Such cases of small vessel occlusion are not amenable to
metabolism, a cascade of changes commences at the cellular surgical intervention and treatment is supportive.
level, which, if not reversed, will lead inevitably to tissue
necrosis5 (see Chapter 2). Early detection is therefore vital
if revascularisation is to succeed. Acute limb-threatening Aetiology of acute ischaemia
ischaemia appears in a variety of guises and its recognition
still remains a matter for clinical judgement rather than • Spontaneous thromboembolism
physiological measurement. In the classic presentation the • Acute deterioration in chronic ischaemia
patient reports that the limb has become painful, cold, • Failure of arterial bypass graft
immobile and numb, and the surgeon recognises that it is • Accidental/intentional trauma
pale and pulseless. Pain, however, may not be reported if • Iatrogenic trauma
the patient is confused, demented or drugged. Similarly, • Diffuse endothelial injury – septicaemia, immune
physical signs may be obscured in a traumatised or oedema- complex deposition, toxic injection, poorly
tous limb. When acute deterioration occurs against a back- controlled diabetes
ground of severe chronic ischaemia, the accompanying
164 Acute limb ischaemia: surgical options

It is important to note a shift in the spectrum of presen- Clinical examination

tation of acute ischaemia. Recent decades have seen a decline
in valvular heart disease, with its potential for dispatch- Patients should be examined supine and in a comfortable,
ing emboli into a previously disease-free arterial bed. warm environment. In assessing the severity of the periph-
This decline has been offset by an increased prevalence eral perfusion deficit, it is important to take into consider-
of ischaemic heart disease that is frequently associated ation reversible systemic factors such as dehydration,
with peripheral vascular disease. Thus, even when the cause anaemia, lung disease, heart failure and hypothermia.
of the acute ischaemia is an embolus of cardiac origin, it is
likely that this will lodge in an already diseased arterial
tree, making simple measures such as balloon embolectomy Indicators for urgent revascularisation
ineffective.6,7 Acute-on-chronic ischaemia is therefore an
increasingly common mode of presentation, affecting an • Peripheral neurosensory loss
enlarging geriatric and diabetic population, while bypass • Inability to flex/extend ankle
graft related problems have become more frequent as recon- • Loss of sensation distal to mid-calf
struction rather than primary amputation becomes the pre-
ferred treatment for critical limb ischaemia. The steady
Examination of the affected limb, particularly in com-
increase in the number and complexity of endovascular
parison with its pair, will help to determine the urgency of
interventions undertaken in patients with advanced limb
the situation. Peripheral neurosensory loss in an ischaemic
ischaemia has been accompanied inevitably by a proportion
limb implies that its viability is compromised and that
of complications that can lead to a precipitate worsening of
revascularisation needs to be achieved without delay to
ischaemia. The outcome for such patients presenting with
prevent irreversible damage. In severe cases the patient is
acute and acute-on-chronic ischaemia is in general less
completely unable to flex or extend the ankle and there is
favourable than in chronic ischaemia, and it is the manage-
no preservation of sensation distal to the mid-calf. Waxy
ment of this subgroup that we seek to address here.
pallor of the extremity which persists in full dependency of
the limb and is accompanied by coldness, numbness and
paralysis indicates a near complete loss of arterial perfu-
DIAGNOSIS sion, requiring urgent correction. Tenderness of the calf
muscles suggests the onset of tissue damage whereas indura-
The objectives of diagnosis are to ascertain the cause and tion denotes irreversibility. Whenever there is evidence of
severity of ischaemia, determine as far as possible the state of muscle paralysis and tenderness, particularly if this is asso-
the arteries involved, and consider the patient’s general med- ciated with pain on passive stretching, it is likely that cellu-
ical condition, including the possible role of environmental lar ischaemia is established and that muscle reperfusion
factors. Obtaining a clear history followed by careful clinical will be followed by cellular swelling and interstitial oedema,
examination provides most of the necessary information, resulting in raised pressure in vulnerable muscle compart-
and it is upon this that vital decisions regarding reconstruc- ments (see Chapter 2). In such cases revascularisation pro-
tion or amputation are taken. In the context of impending cedures should be followed by fasciotomy.
gangrene, vascular imaging through ultrasound and angiog-
raphy does not play its usual pivotal role. The reasons for this
are twofold: first, the urgency of the situation may not allow
Extreme indicators for urgent
sufficient time for vascular imaging to be performed, and
revascularisation
second, a stagnant vascular bed does not facilitate the provi-
sion of satisfactory angiographic or ultrasonic images. • Fixed blotchy cyanosis of skin limited to forefoot
and toes
• Waxy pallor persisting with full dependency of
Clinical history extremity
• Coldness, numbness and paralysis of calf muscles
The aetiology of the ischaemic episode may be suggested • Tenderness of calf muscles
by the clinical history.8 Sudden, spontaneous and unher-
alded acute ischaemia favours a thromboembolic aetiol-
ogy, whereas a prior history of claudication and rest pain If in addition to these signs the muscles have a stiff and
suggests acute-on-chronic deterioration. Any history of ‘putty-like’ consistency and the overlying and distal skin
previous vascular intervention warrants detailed interroga- exhibits fixed cyanosis, the limb should be considered non-
tion and scrutiny of operative scars and medical records viable and amputation at the appropriate level should be
when available. Pain and sensory loss in the limb is usually offered. When the distribution of fixed blotchy staining of
the predominant symptom but may give way to loss of skin is limited to the forefoot and toes, such a finding in
limb function as the effects of ischaemia progress. isolation should not deter attempted revascularisation.
 Diagnosis 165

Reperfusion of ischaemic skin and digits carries a low risk

of systemic complications and superficial or peripheral
necrosis of the foot is compatible with recovery. Varying
degrees of distal necrosis or superficial ulceration may
complicate diverse disorders such as chronic ischaemia,
neuropathy, tissue trauma and venous disease and may be
regarded as potentially reversible. Digital and cutaneous
changes should therefore not influence the decision to
revascularise unless they appear sufficiently extensive as to
preclude functional rehabilitation.

Indicators of non-viability of leg

• Fixed cyanosis of overlying and distal skin

• Loss of movement
• Stiff and putty-like consistency of muscles

When placement of the ischaemic foot in a position of

maximum dependency produces slow capillary return and
there remains some active movement in the calf muscles,
then, whether or not the foot remains anaesthetic, there is
less immediate urgency to restore perfusion. Assuming facil-
ities are available, there is time for angiography to be under-
taken, with a view to evaluating the comparative merits of
Figure 15.1 Irreversibly ischaemic leg with fixed staining of the
thrombolysis, angioplasty or reconstruction. skin and a waxy, dough-like feeling of the calf muscles at which
Some patients will present with a limb that already exhibits time the limb was painless
signs of established irreversible ischaemia. Loss of both move-
ment and sensation, fixed cyanotic staining and tense muscle
compartments indicate that the limb is beyond salvage.
Attempts to revascularise such a limb are pointless and Vigilance is required when assessing patients suffering
expose the patient to the systemic complications of reper- from vasospastic and neurological disorders that affect the
fusion injury (see Chapter 4), notably pulmonary dysfunction microvasculature of the extremities. Patients with para-
and renal failure (Fig. 15.1) (see Chapter 4). Amputation is plegia, stroke and poliomyelitis may exhibit changes in skin
then the therapeutic option of choice. colour and temperature that bear a superficial resemblance
Acute limb ischaemia may occur as an agonal event in to those of critical ischaemia. Careful assessment of ankle
some patients, especially in those with advanced malig- pulses and Doppler-derived ankle systolic pressure should
nancy or other severe debilitating disease. The decision to prevent misdiagnosis. Similarly, chronic oedema can make
intervene with the limb then has to be balanced against the assessment of peripheral arterial perfusion particularly dif-
patient’s overall prognosis and quality of life. We believe ficult, especially when it is due to a combination of venous
that if the patient has potential to use the limb and is likely and arterial insufficiency.
to survive, even for only a few months, then revascularisa-
tion procedures should be considered. The worst scenario
is for a patient to spend the last year of his or her life strug- Evaluation of arterial status
gling not only with their primary disease but also with the
challenge of rehabilitating from an amputation. In patients exhibiting severe acute ischaemia, inspection of
Assessment of the severity of limb ischaemia should be the limb with pulse palpation may provide all the necessary
accompanied by thorough general examination to detect evidence to determine the cause and extent of the problem.
and evaluate comorbid conditions, particularly those likely Comparison of the affected and unaffected limbs is of
to influence oxygen delivery such as shock, dehydration particular value in distinguishing embolic events from
and cardiac or pulmonary insufficiency. Cardiovascular the thrombotic complications of peripheral atherosclerosis.
examination should concentrate on the identification of Palpation at the usual sites, i.e. abdomen, groins, knees,
possible sources of embolism and on assessing the condi- ankles and feet, should distinguish not only pulse volume
tion of the arteries by palpation, auscultation and Doppler but also note any evidence of mural calcification or aneurys-
insonation. mal dilatation. Palpation of the superficial femoral artery in
166 Acute limb ischaemia: surgical options

the mid-thigh of a thin patient can occasionally localise the arteries can be obtained by enlisting gravity in order to max-
level of a femoropopliteal block. Detection of an arterial imise distal perfusion pressure. The patient should be repos-
thrill or bruit can assist in the localisation of stenoses. itioned sitting up with both legs hanging in dependency over
the edge of the bed. All three crural arteries should be
insonated in the lower third of the calf, the peroneal artery
Portable Doppler ultrasound examination being located by compressing the probe into the soft tissues
posteromedial to the fibula. Under conditions of impending
With the patient in the supine position, insonation of the leg gangrene arterial signals may be so damped as to resemble
arteries with a handheld Doppler probe (8 MHz) can yield venous signals. A distinction can be made, however, by con-
valuable information. In the groin the presence of a tapping tinuing insonation while delivering a sharp squeeze to the
signal is indicative of downstream occlusion in the common, foot. Augmentation of the signal suggests it is of venous ori-
superficial or profunda femoris arteries and the level of gin and diminution arterial.
occlusion may be detected by insonation of the superficial An understanding of the anatomy and interrelationships
femoral artery along the subsartorial canal and into the of the crural and pedal arteries is essential in the optimal
popliteal fossa. The presence of a palpable popliteal pulse surgical management of the ischaemic foot. This under-
with a ‘pistol shot’ Doppler signal is suggestive of a recent standing is enhanced by habitual use of dependent Doppler
thromboembolic occlusion at the trifurcation of the popliteal insonation to map these arteries in both health and disease.
artery (Fig. 15.2). Insonation of the tibial arteries is easiest at At the level of the ankle joint the dorsalis pedis artery
the malleolar level and allows calf sphygmomanometry to be lies immediately deep to the tendon of extensor hallucis
employed to derive the ankle systolic blood pressure. longus but as it enters the foot the artery emerges lateral to
Ankle pressure values obtained by calf compression, this tendon to lie on the dorsal convexity of the foot. The
however, can be elevated erroneously by loss of arterial wall
compliance. This occurs most often in patients with
diabetes or with chronic renal failure but it also occurs to
some extent in all patients with peripheral vascular disease.9
When calf derived ankle systolic blood pressure appears to
be at variance with the clinical picture, the limb should be
elevated (Fig. 15.3) while insonating one of the ankle arter-
ies with the Doppler probe.10 If the signal disappears repro-
ducibly at a certain height of elevation above the heart, this
indicates that the perfusion pressure at this level is equiva-
lent to the height of elevation expressed as cm of water
(1 mm pressure of mercury is equivalent to 1.36 cm water).
In any limb the maximum recording of elevation pressure is
limited by its length and therefore pressures exceeding
around 50 mmHg are unrecordable.
In the presence of symptomatic ischaemia, more detailed
information concerning the state of the infrapopliteal

Figure 15.3 Limb being elevated and ankle vessel being

Figure 15.2 Doppler waveform from a vessel proximal to an insonated with an ultrasound probe; the signal will disappear if the
acute occlusion perfusion pressure is exceeded (the ‘pole test’)
 Patient and treatment selection 167

dorsalis pedis artery ends by entering the first metatarsal it is important to be realistic in selecting patients for sur-
space to become the deep plantar artery. gery. In a one-year prospective audit in two centres
The posterior tibial artery is insonated about 1 cm poste- we found that factors associated with a poor outcome
rior to the tibial border and when patent can be traced under were previous bypass surgery, absence of suitable vein
the medial malleolus into the foot. As it runs deep to the belly for a bypass conduit, extensive ulceration of the foot
of adductor hallucis muscle it divides into its medial and lat- and absence of an identifiable pedal arch with dependent
eral plantar branches, which are often difficult to trace. Doppler. Diabetes mellitus was not associated with any
The peroneal artery divides into its anterior and pos- worse outcome. A good outcome in terms of rehabilitation
terior communicating branches 2–3 cm proximal to the following either reconstruction or amputation could be
ankle joint. The anterior branch penetrates the interosseous predicted prior to intervention from the level of social sup-
membrane and runs distally onto the foot on the anterior port available, such as that from a partner living with the
surface of the lateral malleolus, where it is normally palpa- patient, and from the mental state of the patient.
ble. In the foot it communicates with the dorsalis pedis The role of subintimal angioplasty (SIA) remains contro-
artery before anastomosing with the lateral end of the versial. Most published reports fail to provide sufficient
arcade that links the bases of the metatarsals. The posterior haemodynamic, anatomical or follow-up detail to enable
peroneal artery branch passes distally about 1 cm parallel to valid comparisons to be made with surgical series. In units
the posterior and inferior border of the lateral malleolus. It where access to both SIA and infrapopliteal bypass is unre-
gives off a posterior communicating branch which passes stricted, selection of a preferred option can usually be made
between the back of the ankle joint and the tendo Achilles on the basis of clinical and angiographic criteria. In a sig-
to join the posterior tibial artery before descending into the nificant proportion of patients presenting with acute
foot, where under normal conditions it may be palpated ischaemia SIA is impracticable, for instance in multisegment
subcutaneously on the lateral aspect of the calcaneum. It iliofemoral occlusion, aneurysmal disease or severe acute
then winds under the arch of the calcaneum to enter the ischaemia with fresh thrombus. When there is apparent
foot where it terminates by anastomosing with branches of equipoise in the potential application of the two techniques
the lateral plantar artery in the sole of the foot. their relative merits have not yet been compared. This issue is
The major component of the pedal arch is the deep now the subject of a randomised prospective study in the
plantar artery which links the dorsal and plantar metatarsal UK. Most reports seem to agree that failed SIA does not com-
arcades through the first intermetatarsal space. This artery promise subsequent attempts at bypass surgery so a rational
should be insonated with the probe placed over the base policy when both alternatives are applicable would be either
of the first intermetatarsal space. If an arterial flow signal to randomise or to attempt SIA as the first line of treatment?
is obtained, digital compression of the ankle arteries one
by one should determine which is the dominant vessel
supplying the pedal arch. When the dominant artery is Questions in decision making
compressed the pedal arch signal will disappear then return
on release. Conversely, when the non-dominant artery is • Revascularisation or amputation?
compressed the strength of the deep plantar signal may • Will limb loss lead to patient’s deterioration?
be enhanced. When neither the anterior nor the posterior • Is effective treatment available to save the leg?
tibial appears dominant the peroneal trunk should be • Potential success/difficulty of chosen method of
occluded by digital compression posteromedial to the revascularisation?
fibula. Whichever crural artery seems to provide dominant • Fitness for anaesthesia either by GA or LA?
inflow to the deep plantar should be considered the best
outflow when distal bypass is required.11
Unfortunately, in acute and severe ischaemia arterial The decision on whether to proceed with any vascular
Doppler signals are often unobtainable even with the limb intervention rests on the answers to two questions: first,
dependent and the patient warm and hydrated. This situ- whether the patient’s overall situation will deteriorate if the
ation provides a useful index of the severity of ischaemia limb is lost, and second, whether there is an effective treat-
but no clue as to run-off anatomy. ment available. If the answer to either question is negative
then the intervention should not be considered. If both
answers are affirmative it still remains to weigh the diffi-
culty and potential success of the intervention as well as its
PATIENT AND TREATMENT SELECTION
predictable mortality and morbidity. The alternatives of
palliative therapy, percutaneous recanalisation, surgical
Surgical intervention carries risk for the patient and is bypass or primary amputation may all merit consideration.
expensive in time and materials. The worst outcome for Of these, amputation remains the only irrevocable step
the patient and the least economical for the community is although when it offers the patient the best prospect for
a failed bypass that results in amputation. For these reasons recovery it should be adopted in a spirit of optimism. The
168 Acute limb ischaemia: surgical options

concept of the ‘unreconstructable leg’ is flexible and open

to biased interpretation and therefore it should legitimately
be restricted to patients in whom arteriography has shown
extensive infrapopliteal occlusion with good distal contrast
filling but no axial arteries shown in the calf or the foot.
Most patients in this category will be suffering from dia-
betes mellitus or Buerger’s disease or thromboangiitis oblit-
erans (see Chapter 42). Fitness for anaesthesia should not
be used as an argument for amputation in preference to
bypass; regional and local anaesthetic techniques are applic-
able to both. The limited life expectancy of the vascular
patient has been deployed, with some justice, as an argu-
ment in favour of SIA, which often confers only short term
benefit. Distal bypass is undoubtedly a more durable option
but the potential for perioperative complication is greater.

The case for preintervention investigation

When initial clinical and Doppler assessment suggest that Figure 15.4 Angiogram of a 68-year-old woman with a
revascularisation is appropriate, further investigation should three-year history of intermittent claudication and sudden
be considered with the objective of clarifying pathological deterioration; despite clinical expectations she had a saddle embolus,
anatomy and determining interventional strategy, whether which was removed by bilateral femoral balloon embolectomy
thrombolysis, angioplasty or reconstruction. If, despite a
normal femoral pulse, the foot remains cold and anaesthetic facilitate vein harvesting and also aids strategy in those
and ankle Doppler signals are unobtainable in dependency patients in whom the vein is absent or unsuitable and alter-
then the two points mentioned previously should be native sources of vein need to be identified.
recalled: first, that there may be much less than a 12-hour
window before the onset of muscle necrosis and second,
that transfemoral arteriography is very unlikely to yield Angiography
diagnostic images of the infrapopliteal arteries. In a signifi-
cant proportion of patients presenting with impending If time and the condition of the limb allows, angiography
gangrene, however, there is some evidence of continuing should be arranged. On the basis of the history and clinical
distal perfusion. In these cases vascular imaging is helpful in examination it is often difficult to discern whether the
planning operative intervention and essential whenever patient has suffered an embolic or thrombotic event. If
percutaneous intervention is considered feasible. angiography reveals a simple embolus, the decision whether
to proceed to catheter directed thrombolysis (CDT) or to
surgical removal of clot will depend on clinical urgency (sur-
Duplex ultrasonography
gery is usually the quicker option), the relative availability of
surgeon or radiologist and the presence or otherwise of con-
Colour flow duplex ultrasound scanning has been proved as
traindications to thrombolysis (see Chapter 16). Balloon
a reliable means of determining the extent of femoro-
embolectomy for acute limb ischaemia can be undertaken in
popliteal occlusive disease. It, therefore, has a major role in
most cases with reasonable expectation of a good result (Fig.
the non-invasive selection of cases that appear suitable for
15.4). Conversely, if atherosclerotic occlusive disease is dis-
angioplasty. Claims have been made for the superiority of
covered, complex reconstruction may be required.
duplex over dependent Doppler for the evaluation of the
infrapopliteal and pedal arteries. We remain unconvinced
and continue to be swayed by the accessibility and simplic-
Investigation and assessment
ity of Doppler. Duplex ultrasound scanning requires special
technical skills so may not be available out of hours.
Aneurysms of the femoral and popliteal arteries can be • Doppler pressures/waveforms – simple and
accessible
assessed with accuracy and the presence and extent of arter-
ial and venous thrombus can be recorded. Duplex ultra- • Colour flow duplex scanning – not always available
out of hours
sound is less reliable in the assessment of the aortoiliac
segment, particularly in obese patients. • Ultrasound mapping of long saphenous vein
In all but the thinnest legs it is helpful to request ultra- • Angiography – transfemoral (ipsilateral)
orthograde  CDT of embolus, (contralateral)
sound mapping of the long saphenous vein. This may
 Patient and treatment selection 169

retrograde  saddle embolectomy or inflow iliac

percutaneous transluminal angioplasty (PTA),
transbrachial, intravenous digital subtraction
angiography (DSA)
• Magnetic resonance angiography
• Intraoperative angiography

If the patient has a palpable femoral pulse on the side of

the affected limb, ipsilateral orthograde transfemoral angiog-
raphy should be undertaken, since its greater selectivity
reduces contrast load and facilitates more detailed distal
imaging. It is also the preferred point of access if percuta-
neous intervention is contemplated. Retrograde arteriog-
raphy via the contralateral groin is an effective alternative.
If neither femoral pulse is palpable, we favour brachial
catheterisation and in over 250 cases have had only one
episode of brachial artery injury. Intravenous DSA has in Figure 15.5 Magnetic resonance angiogram of patient with
our experience not been particularly helpful in the emer- femoropopliteal disease
gency situation as unfit patients are seldom able to tolerate
the large volumes of contrast necessary.
If the angiogram reveals underlying thrombus, a
guidewire is passed into the thrombus, which is agitated, condition, both arterial and general, to the resources, skills
and a catheter is then passed along the guidewire. We cur- and experience of the vascular department, in order to
rently use tissue plasminogen activator (tPA) delivered as a achieve the best outcome for the individual.
5 mg bolus three times 30 minutes apart or as a low dose
infusion (0.5 mg/hour). We found that the local complica-
tion rate became unacceptable after the catheter had been Magnetic resonance angiography
in place more than 24 hours and now no longer continue
infusions beyond that period. If lysis reveals a lesion suit- With improvements in magnetic resonance technology
able to angioplasty, this is carried out at the same time. we find this an increasingly useful tool, particularly for
Often, however, the angiogram will reveal extensive distal imaging the distal vessels (Fig. 15.5). It is ideal in patients
disease that is only amenable to bypass surgery. These with impaired renal function or those in whom arterial
issues are dealt with in greater detail in Chapter 16. access is limited. The investigation can be disturbing to
some elderly patients and access to the equipment may be
difficult, especially in the emergency situation.
Angioplasty for inflow occlusion

If a stenosis or short occlusion of the iliac artery is dis- Peroperative angiography

covered in association with extensive distal disease, we
favour angioplasty of the iliac lesion at the time of angiog- In some patients with advanced ischaemia of the limb, time
raphy, proceeding to distal surgery when necessary. If and haemodynamics do not favour a preoperative angiogram
percutaneous access is difficult or the haemodynamic sig- and in such cases intraoperative prebypass arteriography
nificance of the lesion is difficult to ascertain, peroperative may be considered. In the presence of severe acute ischaemia
angioplasty (PTA) can also be undertaken (see below). distal perfusion may be so stagnant that direct intrafemoral
Extensive aortofemoral atheroma may warrant surgical injection of contrast fails to provide diagnostic images of the
treatment, either bypass or endarterectomy, as described infrapopliteal segment. Under these circumstances it is
below. preferable to clear femoral thrombus before embarking
The choice between angioplasty and bypass has gener- on angiography, or alternatively to choose a more distal
ated a shifting debate in recent years, although in the emer- injection point such as the below-knee popliteal artery.
gency situation the decision may be dictated by manpower Fortunately, in most patients residual distal perfusion is suf-
logistics, reflecting the scarcity of skilled interventional ficient to permit the distribution of contrast and so achieve
radiologists in many centres. The planning of emergency satisfactory images via the common femoral artery. A 19
revascularisation procedures involves matching the patient’s gauge butterfly needle is inserted into the artery, and 40 mL
170 Acute limb ischaemia: surgical options

non-ionic X-ray contrast (Ultravist, Schering AG, Berlin, SURGICAL INTERVENTION

Germany) is injected by hand as rapidly as possible with the
femoral artery cross-clamped above the needle. If an image
intensifier is not available an X-ray plate is wrapped and General supportive measures
placed beneath the calf and a single film obtained 5–10
seconds after injecting the contrast; the more severe and The main concern should be to restore adequate blood
distal the ischaemia, the longer the delay. Some external rota- supply to the limb as soon as possible. Most patients fortu-
tion of the calf improves the perspective. This technique sel- nately present with a cold painful foot in which there is
dom fails to reveal the distal run-off vessels. Failure to obtain some preservation of calf muscle function and only partial
clear images implies either a technical failure with the con- loss of sensation. This provides a window of a few hours for
trast injection or an underestimate of the severity of distal assessment and preparation during which time optimisa-
ischaemia. tion of the patient’s general condition may be achieved.
Intraoperative arteriography can be of value when Marginal improvement may occur after admission to hos-
a distal vessel has been explored but it is unclear whether pital with adequate analgesia and correction of any fluid
the outflow bed is satisfactory. At the popliteal artery depletion. Simple measures such as nursing the foot in
level, we use the same technique as for the femoral but dependency, intravenous dextran 40 (500 mL in normal
reduce the X-ray contrast to 20 mL. Similarly, at the level saline over 4 hours) and inspired oxygen (28 per cent) may
of the calf and ankle, 10 mL of contrast is injected through allow a well-developed collateral circulation to compen-
a 21 gauge cannula (Venflon) via an arteriotomy. This sate. This may gain time for preintervention investigation
technique is useful in seeking to confirm that the selected of the patient.
run-off vessel communicates with the pedal arch
(Fig. 15.6). Anaesthesia

Patients at imminent risk of limb gangrene are likely to be

systemically unwell and are at increased risk of ischaemic
heart disease, so that the early involvement of an experi-
enced anaesthetist is advisable, both to help prepare the
patient and to select the appropriate method of anaesthesia.
In our practice epidural anaesthesia is the favoured
approach, particularly in patients with chronic obstructive
airways disease. Spinal anaesthesia has the advantage of
speed and simplicity and is useful for amputation proce-
dures. Because its duration is limited to 2 hours it is gener-
ally unsuited to the uncertainties of revascularisation
surgery. The ease of prolongation of epidural anaesthesia is
a distinct advantage in lower limb vascular reconstruction
and provides the option of maintaining good postopera-
tive analgesia without suppressing respiratory function.
When used in this way, bladder catheterisation and vigilant
nursing care of pressure areas are essential. Epidural anaes-
thesia must be used with particular care in patients with
ischaemic heart disease, in whom blood pressure must be
rigorously controlled to avoid coronary hypoperfusion.
Epidural anaesthesia also carries a small risk of spinal
haematoma so should be avoided or used with caution
whenever there is a possibility of impaired coagulation.
Inhalational anaesthesia is acceptable for many patients
and can be used in combination with epidural anaesthesia,
when analgesic requirements are reduced. There remains a
difficult cohort of patients in whom neither neuraxial nor
Figure 15.6 Operative angiogram via proximal posterior tibial
artery showing disease in the posterior tibial vessel at the ankle, inhalational anaesthesia is safe or practicable. These include
which supplies the pedal arch and fills the peroneal artery via the patients with myocardial instability, severe metabolic and
posterior and anterior communicating branches; the dorsalis pedis renal dysfunction, coagulopathy and spinal deformity.
artery is occluded; the proximal peroneal artery was used for distal Fortunately, nearly all infrainguinal reconstructions
run-off with a good result can be carried out under local anaesthesia.12 Expertise in
 Surgical intervention 171

carrying out percutaneous femoral, obturator and sciatic Operation: general considerations
nerve block is useful but not essential and our favoured
technique requires no special training or equipment. Systemic anticoagulant cover is not used routinely except
Prilocaine 0.5 per cent is used to infiltrate the skin and in patients with suspected thromboembolic disease, when
subcutaneous tissues in the groin; deeper infiltration lat- intravenous heparin is given by bolus and infusion to
eral to the femoral pulse should block the medial and achieve perioperatively a partial thromboplastin ratio of
anterior sensory branches of the femoral nerve, so that it is around 2 control. In general we favour the intraoperative
rarely necessary to perform supplementary infiltration in use of a local flush comprising Hartmann’s solution with
order to expose the long saphenous vein in the thigh and heparin added at a concentration of 10 units/mL. Prophy-
calf. Next the skin and subcutaneous tissues of the distal lactic antibiotic cover is administered routinely intra-
medial thigh are infiltrated to allow a medial approach to venously, commencing on induction of anaesthesia and
the above-knee popliteal artery. Some additional infiltra- continuing with two or more postoperative doses accord-
tion of muscle fascia is usually required in order to dissect ing to clinical circumstance. Our usual regimen is flu-
with a finger deep and posterior to the artery until the sci- cloxacillin 500 mg, metronidazole 400 mg and gentamicin
atic nerve is palpated, lying on the surface of the biceps 120 mg, unless the patient has renal impairment, but com-
femoris muscle. The nerve is then hooked round the fore- bination and dosage can vary.
finger and injected slowly and gently with about 10 mL of The first step in most patients is exposure of the femoral
0.5 per cent prilocaine so that the nerve trunk is felt to dis- artery and its branches, affording an opportunity for intra-
tend. This injection usually produces momentary discom- operative arteriography when appropriate. If the femoral
fort but within a minute will induce complete distal pulse is of normal volume, attention is directed towards
anaesthesia for 2–4 hours, with no adverse neurological distal reconstruction. If the common femoral pulse is
sequelae. Provided that this manoeuvre is performed suffi- inadequate due to local atheroma then local femoral
ciently proximal to include the lateral popliteal nerve, as is endarterectomy should be planned.
nearly always the case, no further anaesthesia is required Some patients may present with impending gangrene
and it is notable that patients who have been suffering despite the presence of a palpable popliteal or superficial
severe ischaemic pain will often relax and sleep. Adjunctive femoral pulse. In such cases, as well as those in whom the
sedation is not necessary with this form of anaesthesia and cause of ischaemia is known to be an aneurysm of the
may produce disinhibition and restlessness. By this means popliteal artery, the level of the initial exposure should be
it should be possible to achieve sufficient anaesthesia for as dictated by the clinical picture. Whether at femoral or
any infrainguinal reconstruction with around 80 mL of 0.5 popliteal level, the object of the exposure should be to
per cent prilocaine. assess the condition of the artery wall, to gain all possible
When only infrapopliteal reconstruction is required the insight into the cause of ischaemia and to obtain sufficient
below-knee popliteal artery can be exposed using local control to facilitate all potential manoeuvres, whether
infiltration of the skin, fat and fascia with 0.5 per cent diagnostic or therapeutic.
prilocaine. The posterior tibial nerve will be sighted as it Care should be taken during arterial exposure to preserve
disappears under the soleus arch and it is there injected the long saphenous vein and to consider the most appropri-
with 5–10 mL prilocaine to produce effective anaesthesia of ate orientation of the proposed arteriotomy, whether trans-
the posteromedial aspect of the leg. Anaesthesia of the verse, oblique or, as is usually the case, longitudinal. The use
anterior tibial compartment will require separate percutan- of transverse or oblique arteriotomy should be restricted to
eous infiltration of the lateral popliteal nerve as it lies simple embolectomy procedures in arteries that appear free
superficial to the neck of the fibula. from atheroma. In this situation they have the advantage of
Occasionally revascularisation at ankle level only is easy closure without recourse to patching, and the use of
required, usually as a result of distal thromboembolism interrupted suture technique will avoid any risk of stenosis.
complicating a more proximal surgical or endovascular
intervention. Microtibial embolectomy can then be accom-
plished under formal ankle block or by simple subcuta- Inflow procedures
neous infiltration over the malleolar portions of the anterior
and posterior tibial arteries.13 If the external iliac pulse is weak and it is unclear whether
Extra-anatomical procedures such as cross-femoral and downflow will prove sufficient to support a distal recon-
axillofemoral bypass can be carried out with the help of struction, then femoral arterial pressure can be measured
local infiltration anaesthesia. However, the length of the by direct cannulation with a 19 gauge needle connected to
subcutaneous tunnel in axillofemoral bypass is too great a pressure transducer. If this shows a normal arterial pres-
for local anaesthesia alone. Temporary supplementation sure compared with upper limb pressure, 30 mg papaver-
using inhalational agents or ketamine will enable the tun- ine is injected. A fall in pressure of greater than 30 mmHg
nelling procedure to be covered. is deemed an indication for an inflow procedure.
172 Acute limb ischaemia: surgical options

If adequate in-theatre imaging is available on-table Acute ischaemia associated with loss of both femoral pulses
angioplasty can be carried out. This can be difficult to is much more likely to result from thrombotic occlusion of
achieve, especially if the femoral artery has been opened. In a grossly atheromatous aortoiliac segment. Recanalisation
extreme situations a catheter may be passed up the iliac or bypass is usually required. Aortofemoral endarterec-
artery and withdrawn so as to measure the pressure gradient, tomy has the reputation of being bloody and unreliable
permitting blind angioplasty to be performed at the site of whereas bypass procedures are tried and tested. The choice
pressure fall. or aortic or axillary artery inflow will balance the better
When an unexpectedly weak femoral pulse persists after long term patency of aortofemoral reconstruction against
local disobliteration of the common femoral artery and, if the reduced anaesthetic requirements and lower mortality
necessary, retrograde external iliac endarterectomy, the of axillofemoral bypass. When utilising aortic inflow,
problem is most likely to involve the whole length of the access to the infrarenal aorta may be either direct or
external iliac artery and may extend into the common iliac. retroperitoneal. The latter trades some limitation of access
If the contralateral femoral pulse is of normal volume then for a reduction in postoperative ileus and wound pain.
cross-femoral bypass may be the best and safest operation. As with all suprainguinal inflow procedures, long term
However when the patient is under a general or epidural patency demands unimpeded outflow beyond the com-
anaesthetic and there is a satisfactory pulse at the aortic mon femoral bifurcation. Achieving this may necessitate
bifurcation the various options of iliofemoral revasculari- the addition of measures such as profundaplasty, the tack-
sation may be preferred. An ipsilateral oblique iliac inci- ing of superficial femoral plaque or immediate sequential
sion provides good extraperitoneal exposure of the iliac infrainguinal bypass.
bifurcation and palpation will convey a good impression of
the density of atheroma and degree and extent of arterial
calcification. In the majority of patients it will be possible Operation: inflow procedures
to use external finger fracture to dissect and dislodge the
core of atheroma from the external iliac artery and to • On-table iliac angioplasty
squeeze it out through the open common femoral. If there • Retrograde iliac embolectomy/endarterectomy
is significant common iliac atheroma or thrombus this can • Iliofemoral bypass
be treated in similar fashion provided that the aortic • Femoro-femoral crossover bypass
bifurcation is judged to be compressible. The internal iliac • Aortofemoral endarterectomy
artery should be clamped during the process of finger frac- • Aortofemoral bypass
ture to protect it from embolisation. The core of common • Axillofemoral bypass
iliac atheroma/thrombus will require fragmentation in
order for it to be extruded down the external iliac. This
process of pulsion endarterectomy should be continued
until all luminal obstruction is eliminated and a uniformly Staged in situ bypass
soft, pulsatile segment can be palpated down to the com-
mon femoral. Proximal clamping should not be required, When revascularisation at femoral level is achieved,
although intermittent distal clamping is required for whether as a consequence of inflow reconstruction or fol-
haemostasis; the arterial pulse then aids extrusion of lowing a femoral embolectomy, the adequacy of distal per-
atheromatous fragments. In the presence of significant fusion can be predicted from observation of the femoral
arterial calcification this technique is inappropriate and outflow or from knowledge of the status of the distal vas-
conventional iliofemoral bypass using a prosthetic graft is culature. On rare occasions this prediction may be difficult
preferred. The advantage of pulsion endarterectomy is its and the surgeon is then faced with the dilemma as to
speed and convenience through a limited exposure and the whether or not it is necessary to proceed to immediate dis-
avoidance of prosthetic material, which carries increased tal bypass in order to achieve limb salvage. A policy of ‘wait
potential for infective complications. and see’ may lead to persisting distal ischaemia, which may
When clinical or angiographic assessment indicates that in turn lead to compartment syndrome, tissue loss (see
there is significant stenosis or occlusion at or proximal to Chapter 2) or an untimely rush back to the operating the-
the aortic bifurcation then iliofemoral or femoro-femoral atre to reopen the groin incision and add a distal bypass.
procedures will be insufficient. Endovascular treatment is, A useful compromise in this situation is to mobilise the
at present, unlikely to be effective in these circumstances. proximal end of the long saphenous vein and to use it to
Surgical options are likely to involve bypass reconstruc- close the femoral arteriotomy, as for in situ bypass. Any
tion, obtaining inflow from the abdominal aortic or axil- accessible valves in the vein should be destroyed first by
lary artery level. using scissors or by probing in order to procure some limited
It is rare for removal of a saddle embolus from the aor- fistulous outflow. In this way the option of proceeding to
tic bifurcation to be achieved successfully by the retrograde distal bypass is reserved and can be implemented at any time,
transfemoral deployment of balloon embolectomy catheters. using local anaesthesia, without recourse to reopening the
 Surgical intervention 173

groin incision; downflow is achieved by retrograde deploy- decision to abandon bypass in favour of primary amputa-
ment of a valvulotome into the pulsating proximal seg- tion may have to be taken. Ideally such situations should be
ment of long saphenous vein. If in the event no further foreseen and the patient counselled and consented in
bypass is found necessary, the persisting controlled saphe- advance. In general the prospect of salvaging the limb by
nous vein fistula has no serious disadvantage and can be bypass reconstruction should not be abandoned until the
left alone or tied off as desired. peroneal artery has been explored in the distal calf or the
dorsalis pedis and lateral plantar arteries in the foot. When
faced with a difficult intraoperative dilemma, the surgeon
Distal procedures should keep in mind the basic principle that correction of
critical ischaemia demands the restoration of pulsatile flow
Once satisfactory inflow to the femoral artery has been across two anatomical levels. Thus if the femoral pulse was
confirmed and distal bypass is judged essential, we prefer to palpable at the outset, the graft will have to cross the
use the long saphenous vein in a non-reversed configur- adductor hiatus and the popliteal trifurcation to perfuse a
ation, ensuring that the widest and most proximal section crural artery. If the popliteal pulse was still palpable, the
of the vein is available for anastomosis to the common bypass will have to cross the popliteal trifurcation and the
femoral arteriotomy. This may necessitate side-clamping malleolar anastomosis to reach a pedal artery.
the femoral vein in order to harvest the long saphenous vein Once the distal outflow site is selected, the vein graft can
flush with the common femoral vein, closing the resulting be cut to length and downflow established. When using the
venotomy with a running suture. When a synchronous long saphenous in our favoured non-reversed configur-
inflow procedure has been performed we employ a side-by- ation we proceed as follows: on completion of the proximal
side technique for the junctional anastomosis, where the anastomosis a valvulotome (Hall, Cardial, Le Maitre, etc)
inflow graft and the proximal end of the distal bypass are is passed up the distal end of the vein to the femoral artery
anastomosed alongside one another on the common anastomosis and then withdrawn. Each time the valve cut-
femoral arteriotomy. Any accessible valves in the proximal ter engages the valve cusp, its position on the skin is
long saphenous vein should be incised with scissors before marked. When all valve cusps have been cut the vein graft
the proximal anastomosis is performed. On completion of is explored through short, discontinuous incisions at the
the proximal anastomosis the femoral clamps are removed sites of the valves. We have found that the majority of valve
to restore flow into the profunda femoris. Palpating the branches can be identified at these points and ligated.
pulsating proximal portion of the arterialised long saphe- Following this a retrograde irrigation test or ‘squirt’ test is
nous vein provides some index of the quality of the inflow. performed. The thumb and finger of the assistant occlude
Spontaneous thrombosis of this column of blood, which the vein graft through the incisions in the limb, and retro-
must remain stagnant until distal outflow is achieved, is grade irrigation using heparinised Hartmann solution is
rare and usually points to some thrombotic complication carried out. If it is not possible to irrigate and the assistant
either proximally or in the vein itself. feels transmitted pressure waves between the thumb and
It now remains to prepare the distal graft and the out- finger, it is assumed that no branches exist, and the assist-
flow artery prior to completion of the distal anastomosis; if ant moves up to the next incision until the femoral anasto-
two surgeons are collaborating these tasks can be under- mosis is reached.14 We believe that the use of short,
taken synchronously. We no longer believe that it is discontinuous incisions produces fewer wound complica-
important to leave the long saphenous vein undisturbed in tions, particularly in frail elderly patients with advanced
its original bed, although every effort should be made to ischaemia.
minimise its exposure to trauma and ischaemia. When, as For crural and pedal anastomosis, haemostasis is best
is usually the case, the distal anastomosis is to lie below the achieved with microvascular clamps or silastic slings and
knee joint, we mobilise the vein fully in order to re-route it we have not found any advantage in the use of tourniquets.
through a deep anatomical tunnel, alongside the neurovas- Usually 6-0 or 7-0 monofilament polypropylene material
cular bundle. In this position the graft is unlikely to be on an 8 mm curved atraumatic needle is suitable but 8-0
affected by cutaneous wound problems and should not be may be preferable for some delicate pedal anastomoses.
subject to abrupt angulation or tissue compression. Loupe magnification is desirable ( 2.5 or 3.5) and a
The artery selected for distal run-off is exposed, taking continuous suture line in a ‘short parachute’ configur-
care to avoid damage to venous collaterals by the use of ation, commencing at the heel and completing in the mid-
loupe magnification. Hopefully the artery will be disease dle of the second panel, is standard.
free and compressible and if pricked gently with the point On release of the clamps some confirmation that graft
of a fine blade it will bleed briskly. When these ideals are flow is adequate and outflow resistance is low should be
knowingly compromised the surgeon should be satisfied that sought. Palpation of a distal pulse is reassuring, as is the
the artery selected is the best available. Occasionally, when detection below the anastomosis of a biphasic Doppler
arteriography and informed exploration fail to discover signal, with continuing flow in diastole. Graft flow in
any artery capable of supporting a graft, an intraoperative mL/min can be measured by various methods including
174 Acute limb ischaemia: surgical options

electromagnetic flowmetry or controlled pressure infusion encountered with the long saphenous except that with arm
but the most widely used is the ‘Op-Dop’ ultrasound vein the calibre does not usually favour either a reversed or
device (SciMed, Bristol, UK). If pressure in the bypass graft non-reversed configuration. In general we prefer to use the
is measured using a cannula inserted through a side branch non-reversed configuration. The proximal anastomosis is
of the vein graft, peripheral resistance can be calculated. In performed first and the valves lysed with a valvulotome as
grafts with a peripheral resistance of over 1.5 and flow rates usual so that the graft can be tunnelled while pulsating in
below 80 mL/min, a bolus injection of 30 mg papaverine is order to reduce the risk of kinks, twists and compression. If
given into the graft. If the flow and resistance do not using the reversed configuration we prefer to perform the
improve, an angiogram is taken to determine the cause of distal anastomosis first in order to discourage the forma-
this poor haemodynamic performance. tion of intraluminal thrombus. In the event of graft occlu-
sion, however, the residual valves tend to complicate
thromboembolectomy procedures.
Alternatives to the long saphenous vein

In many patients the long saphenous vein will be found to

Alternatives to ipsilateral long
be inadequate either in its entirety or over short segments,
saphenous vein
especially around the knee. In others the entire vein may be
absent because of prior excision or grafting, in which case
it is prudent to request duplex ultrasound scanning of the • Branches of long saphenous vein
arms in advance. For short deficits, branches of long or • Short saphenous vein
short saphenous vein segments can be substituted. However, • Arm vein – cephalic, ulnar/basilic, antecubital,
figure of eight combinations of above, upper arm
in order to avoid further incisions being made on the
loop
ischaemic limb we have a low threshold for utilising arm
vein, which has been shown to yield results comparable • Contralateral long saphenous vein
with long saphenous vein grafts.15 This may be harvested in
a variety of configurations. The full length of the cephalic
vein from wrist to shoulder will suffice for grafting to the Arm vein is our first alternative to the ipsilateral long
proximal calf. Similar lengths can be obtained from the saphenous vein, the contralateral long saphenous vein is
ulnar/basilic system or by figure of eight combinations of our second. The reasons for this order of priority are that
ulnar, cephalic and basilic vein linked by the antecubital donor site complications are more frequent and severe in
vein. Another option is the upper arm loop, comprising the leg than in the arm and that there is a definite risk that
cephalic, antecubital and basilic veins, although it should the contralateral leg may develop occlusive disease in the
be remembered that in this configuration the valves are future, requiring the vein for grafting.
orientated in opposite directions in the two main sections If it proves impossible to obtain sufficient vein from any
of the graft. source to construct a femorodistal graft, other options may
In general, the more distal the arm vein the more robust be considered. Grafts should never be made longer than
is its wall but the more likely that it has been injured by pre- necessary and if the femoropopliteal segment is patent then
vious cannulation. Although the proximal cephalic vein can it should be used as the inflow site for the graft. Sometimes
seem alarmingly flimsy it usually improves in appearance percutaneous angioplasty, either transluminal or subinti-
with gentle dilatation and will withstand arterial pressures mal, will recanalise the femoropopliteal sufficiently to sup-
when implanted. The basilic vein in the medial upper arm is port a popliteal to distal bypass. Similarly we have had
normally of good calibre but of limited length and in the occasional success with restoring popliteal inflow by semi-
interest of avoiding arm swelling is best not harvested prox- closed superficial femoral endarterectomy, using ring
imal to its confluence with the axillary vein. In general there cutters to dissect and remove the core of atheroma.
is minimal morbidity associated with arm vein harvest, with Unfortunately, as with angioplasty, there is a strong ten-
cutaneous neuropraxia and minor haematoma being the dency towards restenosis. We have found it more reliable in
only problems we have experienced. Discontinuous skin this situation to use a thin-walled polytetrafluoroethylene
incisions should be used over the elbow to avoid contrac- (PTFE) graft, either 6 mm or 8 mm diameter, as a bypass
tures. Harvest under local anaesthetic infiltration is accept- from the common femoral to the above-knee popliteal
able and easily performed in patients undergoing neuraxial artery. Ideally there should be angiographic evidence of an
anaesthesia. isolated patent popliteal segment but we have found when
Veno-venous anastomoses may have to be constructed this is not the case that local endarterectomy can restore
to obtain significant length. We have not found this to be a sufficient lumen to receive an anastomosis and may reopen
particular problem and it is not in our experience associ- previously occluded geniculate collaterals.16 A vein jump
ated with an increased incidence of graft stenosis. The graft is then taken from a composite anastomosis alongside
problem of the venous valves is little different from that the PTFE graft on the popliteal artery to the distal run-off
 Surgical intervention 175

Figure 15.7 Diagram illustrating

composite sequential anastomosis:
polytetrafluoroethylene (PTFE)
bypass to blind popliteal segment with
vein bypass to distal crural vessel; this
technique can be used when there is
insufficient vein for full length of bypass

vessel, utilising a deep anatomical tunnel to traverse the

knee joint. This technique of composite sequential grafting
(Fig. 15.7) significantly reduces the amount of vein required
and has provided results that are comparable with those of
primary long saphenous vein bypass. We endorse the
widely held consensus that regards prosthetic bypass to dis-
tal arteries as a last resort, with or without the addition of
venous cuffs for the distal anastomosis. However, although
patency rates for such grafts are consistently low, late fail-
(a) (b)
ure, after healing of ischaemic lesions has taken place, may
not necessarily precipitate recurrence of critical ischaemia.
Thus each case must be considered individually.

Multiple outflow techniques

The aim of distal reconstruction is to achieve durable
restoration of pulsatile flow to a healthy artery in the leg or
foot which, through its direct or anastomotic communica-
tions, is capable of perfusing all previously ischaemic areas.
In practice, this is usually an artery with direct access to the
pedal arch. There are, however, a number of cases where Figure 15.8 Alternative configurations for multiple outflow in
identification of such an artery is not obvious. There may distal bypass: (a) bifurcated tibial outflow and (b) sequential
be two patent calf arteries below an occluded popliteal popliteal-tibial outflow
which appear to share equal access to the pedal arch, or an
isolated popliteal segment may appear to be a more obvi-
ous recipient for a distal graft than a distal calf vessel, All that can be said as a consequence is that a second out-
which seems to perfuse little more than the foot. In such flow will increase flow in the graft trunk by an average of
cases it is possible to consider constructing a graft with two 30–40 per cent and that there appears to be no increase in
outflows, either in a bifurcated configuration (Fig 15.8a) or morbidity associated with the exercise (which is only
by utilising a single trunk with sequential anastomoses at undertaken when conditions are favourable). The fact that
two levels, the first side to side and the second end to side long term patency in this group appears better than would
(Fig. 15.8b). We have taken an interest in bifurcated and be expected with grafts having single artery outflow may be
sequential grafts over a number of years and have attempted attributed to the fact that limbs with greater options for
to derive relevant data on flow and resistance where possible. graft outflow constitute an intrinsically privileged subset.
176 Acute limb ischaemia: surgical options

acute lower limb ischaemia. It may be detected angio-

Distal revascularisation procedures graphically, sometimes as a complication of angioplasty or
thrombolysis and is sometimes seen as a result of emboli-
• Staged in situ femorodistal bypass, preferably
sation from aneurysms of the aorta or of the femoral or
unreversed (valvulotomy) vein to popliteal
popliteal arteries. The term ‘trash foot’ implies distal
(isolated?)/crural/pedal artery
microembolisation developing during aortic reconstruct-
• Femoral endarterectomy/percutaneous angiography
ive procedures. The condition may also develop as a con-
(transluminal or subintimal)  popliteal-distal
sequence of coagulopathy associated with malignancy or
bypass
certain types of bacteraemia. The foot is observed to be
• PTFE femoropopliteal bypass  popliteal
persistently cold, blotchy and cyanotic despite the presence
endarterectomy
of an apparently satisfactory popliteal pulse. In these cir-
• Composite femorodistal bypass: PTFE to
cumstances neither transpopliteal embolectomy nor sys-
popliteal  vein jump graft to distal artery
temic anticoagulation is effective and limb salvage depends
• PTFE femoral/popliteal bypass to distal
on physical removal of the thrombotic material from the
artery  vein cuff
pedal arch. The anterior and posterior tibial arteries are
• Multiple outflow bypass – dual bifurcated
exposed just above the ankle joint using local infiltration
configuration/single sequential to distal arteries
anaesthesia and proximal and distal embolectomy is per-
formed through small transverse arteriotomies using a 2 Fr
balloon catheter. It is necessary to partially deflate the
SPECIFIC TECHNIQUES balloon just prior to its emergence to avoid splitting the
arteriotomy. Closure of the arteriotomies with interrupted
8-0 suture material avoids stenosis without recourse to
Graft routing patching. The liberal use of heparinised flushing solutions
and of topical antispasmodics such as papaverine helps to
We have already mentioned our preference for routing prevent rethrombosis. Good results can be achieved when
vein grafts across the knee in a deep anatomical tunnel the cause of ischaemia is embolic but patients with spontan-
alongside the popliteal neurovascular bundle, in the inter- eous intravascular thrombosis secondary to malignancy or
ests of avoiding kinking and compression and the potential infection tend to do badly.
consequences of wound failure. In distal bypass surgery,
about 40 per cent of grafts are anastomosed to the anterior
tibial or dorsalis pedis arteries and so have to cross from Fasciotomy
the popliteal fossa to the anterior tibial compartment. We
prefer to make this tunnel through the upper third of the Successful revascularisation of the severely ischaemic limb
interosseous membrane, close to the route of the anterior will result inevitably in some degree of reactive tissue
tibial artery itself, running the vein graft distally alongside swelling. Influential variables, apart from the severity and
the anterior tibial vessels. A long curved clamp such as a duration of ischaemia, include the state of venous
Crafoord aortic clamp is suitable for this purpose but is too drainage, the position of the limb, the extent of operative
rigid to cross the ankle joint and pass under the extensor trauma and blood loss and the presence of infection.
retinaculum into the foot. When bypassing to the dorsalis Increases in tissue compartment pressure above 20 mmHg
pedis we use a flexible tunnelling device such as the non- are abnormal, and above 40 mm Hg fasciotomy is manda-
barbed head of the Cardial valvulotome or a stiffened bal- tory. When calf muscles are found to be paralysed and
loon embolectomy catheter. This can be either used to pull tender prior to revascularisation, the likelihood of com-
down the graft directly or used to place a ligature with partment syndrome developing is such as to justify pro-
which the graft can be pulled through. phylactic fasciotomy. In all other cases it is important to
It is occasionally necessary to undertake a re-do bypass keep a careful watch on the limb in the first 12 hours after
from the common femoral to the proximal anterior tibial revascularisation. Swelling and induration of the calf mus-
artery using a free vein graft. For this purpose a lateral cles combined with the elicitation of calf muscle pain on
subcutaneous tunnel is more direct and avoids scarred tis- passive stretching constitute grounds for intervention.
sue planes but it is important to cross the knee in the line of Compartment pressure can be measured quickly at the
the lateral ligament in order to avoid traction during joint bedside using very simple equipment if electronic manom-
movement. etry is not available (Fig. 15.9).
Although semiclosed fasciotomy can be used as a
Microtibial embolectomy9 prophylactic measure, the correct treatment for elevated
compartment pressure is open fasciotomy (Fig. 15.10),
Occlusion of the arteries of the ankle and foot by throm- which will require a regional or general anaesthetic. In the
boembolic material is a justifiably feared manifestation of recently operated calf the location of existing incisions may
 Specific techniques 177

unlikely to expose vascular grafts. In addition it lends

itself to delayed primary closure by means of a running
polypropylene subcuticular suture, which serves initially to
prevent excessive skin retraction and can then be closed by
simple traction under analgesia once swelling has reduced
sufficiently, usually in 4–5 days.
6

A B C D E
5
4
3
2

Reintervention
1
0

Early graft failure may be attributable to inappropriate

operative strategy, technical error or systemic factors such
as hypotension or hypercoagulability. Success in reinter-
vention depends upon the cause of failure being identified,
which in turn requires observation, deduction and experi-
ence. The correction of technical problems such as retained
Figure 15.9 A simple technique for measurement of compartment
valve cusps, graft twists and residual intraluminal throm-
pressure: with all taps open the syringe is compressed slowly; the
sphygmomanometer reading at the moment the meniscus of dyed bus can often be undertaken under local anaesthesia with
saline moves towards the muscle is the compartment pressure. A, minimal morbidity and excellent outcome. More complex
sphygmomanometer; B, air-filled syringe and three-way tap; C, revisions and repeated failures call for careful risk/benefit
drip tube and thumb clamp; D, dyed saline meniscus; E, no. 1 needle review.
in muscle

Wound management

Wounds should be planned to minimise trauma and

ischaemia and should be closed in a way that will protect
underlying structures, avoid fluid collections and promote
tissue viability. Closed suction drainage is useful in selected
cases. We favour discontinuous incisions for vein harvest-
ing and find that in very obese patients the groin incision is
best made along the skin crease. Closure of the groin
wound in layers including a subcuticular skin stitch, using
absorbable synthetic material throughout, has given the
best results in our experience. In the thigh and proximal
calf we leave fascia open and again use buried absorbable
synthetics for the fat and skin. In the distal calf the absence
of fat and the uncertain vascularity favours the use of inter-
rupted monofilament material, the finer the better.
Figure 15.10 Open fasciotomy Because of the poor viability of the skin of the distal calf
in patients with lower limb ischaemia we try to minimise
dissection in this area. In popliteal to pedal bypass, the
proximal portion of the long saphenous vein should be
influence the surgical approach. However in most cases a harvested and the vein transposed distally, rather than
single lateral incision will decompress all four musculofas- using a strict in situ technique. In some cases, arm vein seg-
cial compartments with minimal morbidity. With this ments can be used in preference to local vein. When paral-
approach the fascia between the peroneal muscles is lel skin incisions in the lower leg are necessary, the incision
incised to expose the full length of the fibula, taking care to overlying the graft and distal anastomosis should be closed
protect the lateral popliteal nerve proximally. There is no and the other left to be covered with split skin.
need to excise the fibula, and the peroneal vessels which lie
medially should not be exposed. Using scissor dissection,
all tense fascial coverings over the anterior and posterior Amputations
compartments are then split longitudinally until all muscle
bundles are soft and bulging. This single lateral approach is It is a general principle of vascular surgical management
simple and effective, incurs minimal morbidity and is that prior to reconstruction, infection should be controlled
178 Acute limb ischaemia: surgical options

by debridement and appropriate antibiotic therapy (see with a mortality rate of 28.6 per cent and a 30-day limb
Chapters 18 and 33). Following reconstruction, definitive salvage/survival of 62 per cent. The remainder required
amputation should ideally be delayed until demarcation is surgical bypass, 22 per cent an inflow procedure, but the
established and the operative wounds have healed. In the majority an infrainguinal reconstruction. The 30-day
context of acute or severe ischaemia, with imminent, pro- mortality rate for surgical bypass was 19.6 per cent and
gressive or established gangrene these principles may have the overall limb salvage/survival rate 52 per cent.
to be adapted but they should not be ignored. Clearly, the
priority is to restore arterial perfusion and prevent or halt
ischaemic necrosis. Once this is achieved, however, steps
must be taken to control infection and if this requires Conclusions
drainage of pus and/or excision of necrotic tissue then this
should be undertaken at the time of reconstruction. It will The management of the acutely ischaemic leg remains a
hardly ever be appropriate to attempt primary closure after major challenge to the vascular surgeon. The key lies in
this type of surgery and the patient should be made aware good preintervention diagnostic information, which
of the possible need for further excision or amputation allows treatment to be undertaken with a reasonable
once demarcation is established. expectation of a good outcome. In many cases this can
Sadly, amputation remains a possible outcome after be achieved by thrombolytic therapy or simple balloon
emergency surgery for severe lower limb ischaemia. Evidence embolectomy. In those patients in whom an advanced
that failed attempts at vascular reconstruction prejudice degree of ischaemia precludes thrombolysis, however,
the level of subsequent amputation is contested. However, or in those patients with atheromatous vascular disease
as far as is possible, incisions for infrainguinal reconstructive that makes embolectomy inappropriate, urgent surgical
procedures should be placed so as to minimise interference bypass will have to be undertaken. A number of simple
with a potential below-knee amputation. manoeuvres can be performed which will aid the sur-
geon in undertaking these difficult procedures. Such
techniques can produce acceptable limb salvage and sur-
POSTOPERATIVE MANAGEMENT vival in this challenging group of patients.

We do not routinely anticoagulate these patients, who are

often elderly and in whom the risk of anticoagulation may
exceed any theoretical benefit. Unless contraindicated, they Key references
all receive 75 mg aspirin daily. Patients are followed up in
the surveillance programme with duplex scanning of the Hickey NC, Thomson IA, Shearman CP, Simms MH. Aggressive
arterial reconstruction for critical lower limb ischaemia.
graft at 6 weeks and then subsequently at 3-monthly inter-
Br J Surg 1991; 78: 1476–8.
vals. Any patient who has a stenosis discovered in their graft
Mahmood A, Garnham A, Sintler M, et al. Composite sequential
with a more than 50 per cent velocity shift over 1 cm is grafts for femorocrural bypass reconstruction: experience with
referred for balloon angioplasty or, if this proves ineffective a modified technique. J Vasc Surg 2002; 36: 772–8.
or of transient benefit, for patch angioplasty. Patients who McKay C, Razik WA, Simms MH. Local anaesthetic for lower limb
are found to have failing grafts due to distal anastomotic revascularisation in high risk patients. Br J Surg 1997; 84:
stenosis or progression of disease in the native artery are 1096–8.
treated with jump grafts. We have found percutaneous bal- Second European consensus document on chronic critical leg
loon angioplasty of some of the long grafts particularly dif- ischaemia. Eur J Vasc Surg 1992: 6(A).
ficult owing to the long length of catheter needed. In these Shearman CP, Gwynn BR, Curran FT, et al. Non-invasive
patients the vein graft can be dissected and controlled prox- femoro-popliteal assessment: is that angiogram really
necessary? BMJ 1986; 293: 1086–9.
imal to the stenosis under local anaesthetic and an on-table
balloon angioplasty carried out.

RESULTS
REFERENCES
We adopted the above principles for the management of
1 Dormandy J (ed). European Working Group on Critical Limb
acute ischaemia of the lower limb over a 4-year period in Ischaemia. European Consensus Document on Critical Limb
two acute care hospitals. During that time 126 limbs were Ischaemia. Berlin: Springer Verlag, 1989.
considered for surgical intervention. Fourteen patients were 2 Hickey NC, Thomson IA, Shearman CP, Simms MH. Aggressive
either too moribund for intervention or required primary arterial reconstruction for critical lower limb ischaemia. Br J Surg
amputation. Of the rest 42 were treated with embolectomy, 1991; 78: 1476–8.
 References 179

3 Shearman CP, Ashley EMC, Gwynn BR, Simms MH. Rehabilitation 10 Smith FCT, Shearman CP, Simms MH, Gwynn BR. Falsely elevated
of patients after vascular reconstruction for critical lower limb ankle pressures in severe leg ischaemia: the pole test, an
ischaemia. Br J Surg 1991; 77: A346. alternative approach. Eur J Vasc Surg 1994; 8: 408–12.
4 Cheshire NJ, Wolfe JHN, Noone MA, et al. The economics of 11 Shearman CP, Gwynn BR, Curran FT, et al. Non-invasive
femorocrural reconstruction for critical leg ischaemia with and femoro-popliteal assessment: is that angiogram really necessary?
without autologous vein. J Vasc Surg 1992; 15: 167–75. BMJ 1986; 293: 1086–9.
5 Second European consensus document on chronic critical leg 12 McKay C, Razik WA, Simms MH. Local anaesthetic for lower limb
ischaemia. Eur J Vasc Surg 1992: 6(A). revascularisation in high risk patients. Br J Surg 1997; 84:
6 Fiorani P, Taurino M, Novelli G, et al. Acute occlusion of the lower 1096–8.
limbs. In: Greenhalgh RM, Hollier LH (eds). Emergency vascular 13 Mahmood A, Hardy R, Garnham A, et al. Microtibial
surgery. London: WB Saunders, 1992: 387–99. embolectomy. Eur J Vasc Endovasc Surg 2003; 25: 35–9.
7 Hight DW, Tilney NL, Couch NP. Changing clinical trends in 14 Shearman CP, Gannon MX, Gwynn BR, Simms MH. A clinical
patients with peripheral arterial emboli. Surgery 1976; 79: 172–6. method for the detection of arteriovenous fistulas during in situ
8 Jivegard LE, Arfivdisson B, Holm J, Schersten T. Selective great saphenous vein bypass. J Vasc Surg 1986; 4: 578–81.
conservative and routine early operative treatment in acute lower 15 Andros G, Harris RW, Salles-Cunha SX. Arm veins for arterial
limb ischaemia. Br J Surg 1987; 74: 263–71. revascularisation of the leg: arteriographic and clinical
9 Faris IB, Duncan HJ. The assessment of critical skin ischaemia. In: observations. J Vasc Surg 1986; 4: 416.
Greenhalgh RM, Jamieson CW, Nicolaides AN (eds). Vascular 16 Mahmood A, Garnham A, Sintler M, et al. Composite sequential
Surgery: Issues in Current Practice. London: Grune and Stratton, grafts for femorocrural bypass reconstruction: experience with a
1986: 91–6. modified technique. J Vasc Surg 2002; 36: 772–8.
This page intentionally left blank
 16
Acute Limb Ischaemia: Endovascular Options

BERNARD H NACHBUR, IRIS BAUMGARTNER, DO D DO, FELIX MAHLER, HANS-BEAT RIS

The problem 181 Validation of the combined endovascular catheter 185

Diagnosis 182 technique
Selection for treatment 183 Discussion of the new evidence 189
Dangers of reversing ischaemia 183 When should endovascular options be recommended 190
What is new about endovascular options for 183 for acute limb ischaemia?
acute limb ischaemia? What developments does the future hold? 192
Technique of combined endovascular catheter therapy 185 References 194

usually found at the level of arterial bifurcations where

THE PROBLEM they can be readily located either clinically, radiologically
or by duplex ultrasonography. It is important to realise
Acute lower limb ischaemia poses a threat to both limb and that they can also split into two or more fragments at a
life. Ischaemia of the upper limb is less frequent, and in the bifurcation and thus occlude more than just one artery.
normal course of events more dangerous when associated Therefore radiological visualisation is mandatory for
with vascular trauma of the shoulder region than when the exact assessment of sites of occlusion. Figure 16.1 illus-
caused by thromboembolic occlusion. Embolic occlusions trates an example of such multiple splitting of an embolus.
of the axillary artery or around the elbow at the bifurcation It shows the angiogram of a 17-year-old who collapsed on
of the cubital artery do not occur infrequently as a result of her way to school because of sudden paralysis of the lower
atrial fibrillation. The occluded native artery is practically limbs. The multiple occlusions of the iliac, femoral and
always normal and surgical removal is so easy that endovas- distal vessels causing sudden muscular weakness had ori-
cular options are unnecessary, disadvantageous and in fact ginated from a large myxoma of the left atrium and were
impractical. This chapter will deal with the endovascular removed surgically, followed by open heart surgery for
treatment options for ischaemia of the lower limb. removal of the myxoma. This emphasises the importance
Acute ischaemia is by definition a sudden event which of harvesting material for histopathological examination.
can therefore be timed quite exactly, most particularly so in More than 50 per cent of cases with acute lower limb
the case of embolism into a previously normal arterial tree. ischaemia are probably not of embolic origin neither are they
Such emboli usually originate in the heart as a consequence restricted to arterial bifurcations; they are caused by in situ
of atrial fibrillation or mural thrombosis following myo- thrombosis of a pre-existing atherosclerotic plaque, so-called
cardial infarction. They might occasionally originate from acute-on-chronic ischaemia.1 The arrest of arterial perfu-
aortic ulcers or aneurysms, popliteal aneurysms being the sion in the residual artery leads to propagation of throm-
most frequent source of arterio-arterial embolism. Mural bus distal to the occluding site. Angiographically, the
thrombi from subclavian arteries exposed to repeated com- difference between embolic occlusion and in situ throm-
pression at the thoracic outlet produce similar effects (see bosis can be difficult but is possible in most cases. Duplex
Chapter 41). sonography can also be of help if the quality of the arterial
Emboli have a certain given diameter and are swept wall can be defined. This is relevant because embolic occlu-
distally to a point where the calibre of the vasculature sions at the level of the aortic, iliac or femoral bifurcation
prevents further migration. This is why embolic clots are can readily be treated by a general surgeon who has sufficient
182 Acute limb ischaemia: endovascular options

Table. 16.1 Predictive power of factors in favour of embolism

versus thrombosis

Factor Predictive power

1 (embolism)
Atrial fibrillation/acute myocardial infarction 0.82
Duration of symptoms * 0.56
Age† 0.51
Diabetes 0.27
Systemic malignancy 0.20
1 (thrombosis)

* The shorter the duration of symptoms, the greater the likelihood of

embolism.
† The older the patient, the greater the likelihood of thrombosis.
Reproduced with permission from Goldstone J (ed). Perspectives in
vascular surgery. Vol. 2. St Louis, MO: Quality Medical Publishing,
1989: 11–17.

presence of pre-existing collaterals. A clear distinction

between embolic occlusion and in situ thrombosis, or acute-
on-chronic occlusive disease, cannot be made with suffi-
cient reliability on clinical grounds alone; neither is the
objective documentation of ischaemia using such investi-
gations as Doppler ultrasound measurements or even
duplex sonography decisive.
Imaging of the arterial tree is therefore of cardinal
importance and for which, today, there are no contra-
Figure 16.1 A 17-year-old patient with multiple bilateral indications. The nephrotoxic effect of angiography with
embolic occlusions of the external iliac and common femoral conventional contrast medium can be avoided by contrast-
arteries and their branches enhanced magnetic resonance imaging (MRI) using low
osmolar gadolinium as a viable alternative contrast agent
for angiographic investigation. Moreover, the reduction in
renal function induced by radiographic contrast agents
knowledge, a special interest in vascular surgery and is
such as iopromide, a non-ionic, low osmolality contrast
conversant with modern techniques.2 On the other hand,
agent, can now also be offset by the prophylactic adminis-
surgical thrombectomy in patients with acute-on-chronic
tration of the antioxidant acetylcysteine, along with
occlusive disease can be hazardous and may actually
hydration, in patients with chronic renal insufficiency.4
worsen the degree of ischaemia if not dealt with optimally
Knowledge of the exact location of the occlusion as well as
by an experienced vascular surgeon who has access to
the extent of thrombosis will influence the tactical
endovascular options and can call upon the support of
approach. It is relevant to note that in a prospective con-
ancillary investigators. The potential of these endovascular
trolled study the serum creatinine concentration decreased
options has been subject to significant evolution in the past
significantly in the acetylcysteine group.5 This is but one of
few years as a consequence of the discrete and diminishing
the major recent developments. For practical purposes dig-
role played by thrombolysis in the treatment of acute
ital subtraction angiography (DSA) is the diagnostic test of
ischaemia. The examination of that potential is the object-
choice and is a prerequisite in the management of acute
ive of this chapter.
ischaemia. It is also an indispensable method of forewarn-
ing the surgeon or physician in charge of the coexistence of
chronic occlusive disease before surgical, lytic or other
DIAGNOSIS
endovascular options are decided upon.
The non-opacification of peripheral run-off in an intra-
Some of the major factors and their discriminating power arterial digital subtraction angiogram, however, can often
in the diagnosis of embolism and thrombosis are outlined be misleading. It has been shown that by applying pulse
in Table 16.1.3 The degree of ischaemia in acute-on- generated run-off techniques,5 a significantly better func-
chronic occlusive disease is usually less severe than in tional picture can be obtained compared with conven-
patients with sudden complete occlusion because of the tional intra-arterial DSA alone. This can also be
 What is new about endovascular options for acute limb ischaemia? 183

demonstrated by intraoperative angiography after exposure which acute and total ischaemia has led to a situation
of the infragenicular popliteal artery. Non-opacification of which is aptly described by the French expression of
the calf vessels by intra-arterial femoral DSA therefore does ischémie dépassée. The clinical picture is characterised by
not preclude reconstructive procedures, be they surgical or the grotesque mottled appearance of the skin of the
endovascular. Endovascular techniques can be applied to involved extremity, total loss of sensitivity and motility of
the crural arteries intraoperatively following surgical expo- the foot and incipient or even well-developed rigor mortis
sure of a distal segment of the popliteal artery.6–9 of the muscles. Rigor mortis of the calf is easily diagnosed
Previous teaching has emphasised run-off as a signifi- by examining passive motility of the toes and foot. If the
cant discriminator between success and failure of recon- knee joint has stiffened, occlusion has probably occurred at
structive procedures in both acute and subacute occlusions.10 the level of the iliac arteries.
The fallacy of this contention has been convincingly In this situation the ischaemic damage done to soft tis-
demonstrated by Do et al.11 who point out both the advan- sue is so extensive that any attempts at revascularisation
tages provided by pulse generated examination and intra- might endanger the life of the patient, whether the limb can
operative angiography of the peripheral arterial tree, and be salvaged or not. Following successful disobliteration in
also the immense potential of intraoperative or catheter such patients, it is not unusual for them to die within a few
intraclot lysis. hours, death usually being attributable to reperfusion
injury (see Chapters 2 and 4). It is important to understand
that, just as with surgical revascularisation, concepts such
as rhabdomyolysis11 and its associated metabolic disturb-
SELECTION FOR TREATMENT12 ances, such as hyperkalaemia,12,13 the closed compartment
syndrome elevated serum phosphokinase, uraemia, hyper-
The patient may present with a paralysed anaesthetic limb phosphataemia, hypocalcaemia and hypercalcaemia in
and the first signs of cutaneous mottling with, in a worst association with anuria become critical parameters.14
case scenario, rigor mortis. Further delay cannot be toler- Indeed, because combined endovascular treatment options
ated and immediate and decisive treatment is mandatory if are so successful in the smaller arteries, such changes are
amputation is to be avoided. When the condition is this likely to be more marked and acute. The danger of revascu-
serious there is no time for preoperative angiography. The larisation must therefore be assessed by taking biochemical
situation might call for surgical intervention, but com- and individual clinical factors, such as renal insufficiency,
bined endovascular catheter treatment can now be con- into account.
sidered to have equal potential. In general, however, it is
the patient with acute-on-chronic occlusive disease who
benefits most from endovascular treatment options. In
some patients the collateral circulation opens up after ini-
WHAT IS NEW ABOUT ENDOVASCULAR
tial treatment with analgesics and heparin and may even
OPTIONS FOR ACUTE LIMB ISCHAEMIA?
regain normal use of the extremity with little or no sign of
anaesthesia. Indeed, pain may subside so as to justify a cer- Percutaneous, catheter directed low dose infusion of throm-
tain deferment of invasive action. Such an improvement bolytic agents has been used for many years since its intro-
allows time for decisions as to whether to treat with antico- duction by Hans Hess et al. in 1982,15 and has found
agulants alone or to pursue either the surgical or endovas- widespread use as an alternative to open surgery for treat-
cular option in a more leisurely fashion. Because of the ment of acute arterial occlusion of the legs.16–26 The safety
need for ancillary diagnostic and therapeutic options (see and efficacy of the procedure, however, continues to give
Chapter 15), it is self-evident that treatment of acute- rise to concern. For example, in the Thrombolysis Or
on-subacute ischaemia with impending gangrene should Peripheral Artery Surgery (TOPAS) trial,26 the latest most
not be the business of a general surgeon with casual expe- authoritarian prospective study comparing catheter directed
rience in the treatment of acute vascular occlusion. thrombolysis with open surgery, serious bleeding problems
Referral to a specialist vascular unit, if that is possible, is as well as intracranial haemorrhage were associated with
important.12 local thrombolysis requiring a considerable number of
blood transfusions. Furthermore, in other studies reported
hitherto, blood flow was restored more slowly than by
immediate surgical revascularisation. Therefore, for endovas-
DANGERS OF REVERSING ISCHAEMIA13 cular techniques to be as efficient as the surgical treatment of
acute limb-threatening ischaemia, the ideal catheter directed
As with surgery it is important to identify those patients in method must reduce haemorrhage and, if possible, reduce
whom reversal is contraindicated and life threatening. the time necessary for restoring arterial flow comparable to
Thus on occasions the attending physician, who should be the duration of a surgical procedure. Ideally, catheter
a vascular surgeon, will be confronted with an extremity in directed restoration of patency should be achieved within
184 Acute limb ischaemia: endovascular options

Figure 16.2 Illustration of the

efficacy of thrombus aspiration. (a) this
patient had acute occlusion of the
popliteal trifurcation with motor and
sensory deficit of 20 hours duration. (b)
following an intraclot instillation of
60 000 units of urokinase for 20
minutes there was only minor clot
dissolution without clinical
improvement. (c) following
percutaneous catheter thrombus
aspiration complete restoration of
(a) (b) (c) patency was achieved within
15 minutes

the same time span as an open surgical procedure. That this accept catheter directed procedures, especially in the treat-
is a viable option will be shown later in this chapter. ment of acute limb ischaemia.
In the TOPAS trial reported by Ouriel et al. in 199826 it It is now well established, however, that percutaneous
was shown that intra-arterial infusion of urokinase reduced catheter thrombus extraction can work, and by itself is
the need for open surgical procedures with no significantly capable of restoring patency following acute thrombotic or
increased risk of amputation or death. Intra-arterial throm- thromboembolic occlusion of the arteries of the leg.28,29
bolysis, however, was associated with a significantly higher The advantages and the potential which this regimen offers
number of major haemorrhages (12.5 per cent) than the in terms of the significant reduction in the dosage of lytic
surgery group (5.5 per cent; P  0.005) and in both groups agent used and in the duration of the procedure in restor-
the amputation rate was surprisingly high. Moreover, of ing flow remains to be fully recognised. The important
particular concern in the TOPAS study, are the four place of thrombus aspiration, in fact its superiority over
episodes of intracranial haemorrhage in the urokinase lytic therapy alone, was recognised a number of years ago
group (1.6 per cent), one of which was fatal. The develop- and is illustrated in Fig. 16.2.
ment of these bleeding complications has been criticised by It is the aim of this chapter, illustrated by a retrospective
Porter.27 In the surgical group there were, as expected, no study, to demonstrate and to prove that by performing
episodes of intracranial haemorrhage. Therefore, although combined endovascular therapy either by percutaneous
intra-arterial infusion of urokinase has significantly reduced catheter aspiration alone or in association with thrombol-
the need for open surgical procedures, the safety of ysis and percutaneous transluminal angioplasty (PTA),
catheter thrombolysis necessitating the use of large amount high rates of primary success and limb preservation can be
of lytic agents has remained an awesome and serious threat achieved both for acute and subacute ischaemia. In a
to the patient. Furthermore, in the studies reported hith- majority of cases thrombus aspiration alone works and
erto in the literature, blood flow was restored more slowly thrombolysis can be avoided altogether or reduced to
than by immediate surgical revascularisation, and tissue harmless levels. Where necessary, thrombus infiltration
ischaemia may progress to necrosis before thrombolysis with small doses of urokinase can loosen the thrombus and
has taken effect; this is another reason for the reluctance to allow continuation of thrombus aspiration. In our study30
 Validation of the combined endovascular catheter technique 185

in only one out of five cases was it necessary to apply a cases (20 per cent) will need modest amounts of lytic agent
modest dose of intraclot urokinase. Thus the doses of to obtain vascular patency. Hospital time is reduced and
urokinase are minimised and the duration of the inter- patient comfort considerably enhanced.
vention shortened substantially. If a culprit plaque is
unmasked in the course of the intervention it can be
treated by PTA. On balance then, the benefits of percutan- TECHNIQUE OF COMBINED ENDOVASCULAR
eous catheter therapy should prevail over open surgical CATHETER THERAPY30
procedures. In fact amputation-free survival rates at 6 and
12 months are comparable to or better than those of open
Direct anterograde catheterisation of the common femoral
surgical procedures reported in the TOPAS trial and at no
artery is performed under local anaesthesia. Applying the
extra cost or risk of serious haemorrhage, death or second-
Seldinger technique a 6–8 Fr sheath with a haemostatic
ary intervention.
valve and a side-port for flushing is introduced and a bolus
In summary, the new achievements of endovascular
of 5000 units of heparin is injected. In the presence of fresh
treatment options areas follows:
thromboembolic material a thin walled 6–8 Fr catheter
1 By combining percutaneous clot aspiration with with an end-hole is positioned very closely to or barely into
modest doses of thrombolytic agent only when the proximal end of the thrombus, whereupon the clot is
necessary, and PTA of underlying stenotic plaques sucked and extracted with a 60 mL syringe. For extraction
when unmasked, endovascular treatment can be of larger pieces of clot, the haemostatic valve has to be
shortened to periods of 35–160 minutes – comparable removed. By repeated suction all the non-adherent mater-
to the customary time spent in surgical interventions. ial is removed, while residual thrombotic material adher-
This is achieved without serious bleeding risks and ing to the vessel wall is loosened with a wire loop30 and
transfusions. For our unit endovascular options have subsequently also sucked away. In the crural arteries a 5 Fr
become a valid and preferable option in managing aspiration catheter is used.
cases of acute limb ischaemia including those with Should complete removal of the clot prove completely
associated sensory and motor deficit. impossible, then local catheter thrombolysis is performed at
2 Endovascular treatment implies repeated injection the same session. To achieve that end a 0.035 inch (0.9 mm)
of small amounts of contrast medium. This can be guidewire is used to introduce a microporous balloon
hazardous in patients with chronic renal insufficiency. catheter down into the proximal part of the thrombus taking
With prophylactic oral administration of the care not to pass beyond the clot so as to avoid peripheral
antioxidant acetylcysteine or the use of low osmolar embolism. Through the balloon catheter 10 000–20 000
gadodiamide for diagnostic and angiographic units/cm of urokinase is infiltrated into the thrombus.
interventions, along with hydration, a fall in renal Under fluoroscopic supervision the balloon catheter is
function induced by contrast medium can be offset advanced centimetre by centimetre until the distal end of the
entirely4,5 and in patients with chronic renal thrombus is reached. Clot material loosened by partial lysis
insufficiency serum creatinine levels even improve. is removed by repeated suction (Fig. 16.3). Angiography is
performed at this time to confirm free peripheral outflow
During our preliminary experience we were impressed
and underlying stenotic lesions are treated by PTA at the
by the fact that with a minimal dose of lytic agent followed
same session in keeping with the ‘all-in-one’ principle.
by percutaneous clot extraction, complete restoration of
Completion angiography is performed before removal of all
patency of an occluded popliteal trifurcation can be
instruments. The duration of the intervention and the total
achieved within an hour (Fig. 16.2). With the passage of
dose of the lytic agent are recorded.
time and experience we eliminated the use of preliminary
thrombolytic agents entirely and performed clot aspiration
alone. In only one out of five cases was it necessary to apply
small but safe doses of urokinase in order to loosen up
VALIDATION OF THE COMBINED
residual clot and facilitate complete aspiration. In the pres-
ENDOVASCULAR CATHETER TECHNIQUE30
ence of an underlying stenotic plaque identified as the
cause of thrombotic or thromboembolic occlusion con- Patient selection and immediate results
ventional PTA is performed immediately.
Given this experience it is our contention that the com- In the single centre study30 performed by us between
bined catheter approach as outlined above can be recom- January 1995 and May 1997 there were 89 consecutive
mended as first line treatment for the group of patients patients (42 men and 47 women; mean age 70.7  14.9
with acute or subacute occlusions in infrainguinal arteries. years, range 29–100), 93 legs with acute or subacute
The intervention proposed allows for clot aspiration, mild thromboembolic occlusion of native femoropopliteal and
thrombolysis and PTA as an ‘all-in-one’ procedure and is crural arteries meeting the guidelines for reversible limb-
performed under local anaesthesia. Only a minority of threatening ischaemia. The indication and rationale for
186 Acute limb ischaemia: endovascular options

Table 16.2 Baseline patient characteristics

Demographics
Age (years  SD) 70.7 (14.9)
Number of patients 89
Gender
Men 42
Hemostasis
Women 47
Sheath
valve with
side port Risk factors: n (per cent)
Catheter
extension History of smoking 39 (44)
removed Suction Diabetes 14 (16)
Hypertension 49 (56)
Sheath 9 FR
Hypercholesterolaemia 17 (19)
Catheter for
lysis and/or Comorbidity: n (per cent)
clot aspiration Cardiomyopathy 45 (50.5)
Coronary heart disease 41 (46.5)
Atrial fibrillation 19 (21.5)
Clot Transient cerebral ischaemia or stroke 14 (16)
Clot in dissolution Chronic obstructive pulmonary disease 11 (12.5)
Pulmonary embolism 9 (10)
Multifocal arterial embolism 7 (8)
Renal insufficiency 9 (10)

Total number (per cent) of patients 68 (76)

with polymorbidity

Figure 16.3 The principle of clot extraction. The haemostatic

valve with the side port extension (for heparinisation) has been
removed and the radio-opaque sheath 8 or 9 Fr is left in place to
protect the vessel wall from injury due to catheter manipulation.
The catheter used for extraction is advanced into the clot which, in Table 16.3 Clinical features and occlusion sites
one out of five cases, has to be loosened up with modest doses of
urokinase. The clots are readily aspirated. The wider the sheath, the Clinical presentation
larger the clots which can be extracted but by the same token the Duration of symptoms (days) 6.2 (7.3)
danger of intimal damage increases Pain at rest 89 (100 per cent)
Motor loss 7 (7.7 per cent)
Sensory loss 73 (81 per cent)
Embolism likely 55 (61.7 per cent)
Acute on chronic occlusive disease 40 (49 per cent)
catheter therapy was critical ischaemia in all 69 patients
presenting with rest pain and a further 20 patients with Site of arterial occlusion
trophic lesions. The therapeutic decision was reached Superficial and profunda femoral 24
jointly by the vascular surgeon and the interventional Superficial femoral, popliteal, crurals 32
Crurals 26
angiologist. Percutaneous catheter directed treatment was
Superficial and profunda femoral, 9
restricted strictly to infrainguinal native arteries. Patients
popliteal
with occlusions of upper limb arteries, iliac arteries, the Profunda femoral 2
femoral bifurcation or bypass grafts were treated by open Superficial femoral 1
surgery. In the entire group of 89 patients with 93 treated
legs, the minimum observation time was 12 months for all
cases, but three patients were lost to follow-up after initial
catheter therapy.
Catheter therapy had to be performed urgently on between onset of the acute occlusive episode and
the same day or within 24 hours of onset of symptoms catheter therapy was 6.2 days (7.3). A motor deficit was
in 33 cases (37.5 per cent), whereas in 11 patients (12.5 present in seven patients and a sensory deficit in 73; 20 legs
per cent) the acute episode of occlusive disease was more presented with trophic lesions (21.5 per cent). Baseline
than 14 days old (range 15–28 days). In the remaining patient characteristics (Table 16.2) and clinical presenta-
45 patients the acute onset of arterial occlusion was less tion and sites of arterial occlusion (Table 16.3) were
than 14 days old. For the whole group the mean duration recorded.
 Validation of the combined endovascular catheter technique 187

Figure 16.4 (a) Acute occlusion,

probably of embolic origin, of
the popliteal trifurcation with
non-opacification of the crural
arteries presenting with motor and
sensory deficit. (b) After clot aspiration,
patency is achieved in all three crural
(a) (b)
arteries within 25 minutes

Figure 16.5 (a) Acute-on-chronic

occlusion of the infrapopliteal crural
arteries with sensory deficit. (b) After
thrombus aspiration combined with
(a) (b)
percutaneous angioplasty the
infrapopliteal arteries are patent

Procedures (21.4 per cent) was it necessary in the last resort to add a
modest amount of urokinase to initiate thrombus dissol-
Percutaneous thrombus aspiration was performed as a pri- ution and facilitate continuation of aspiration. In this last
mary procedure in all patients and as the single therapeutic group all three therapeutic modalities were combined,
act in 30 legs (32.2 per cent). In 44 legs (47 per cent) throm- namely, clot aspiration, thrombolysis and PTA (Figs 16.4
bus aspiration was followed by balloon angioplasty of and 16.5). The duration of all percutaneous catheter inter-
underlying atherosclerotic lesions. In only 20 patients/legs ventions was between 35 and 160 minutes.
188 Acute limb ischaemia: endovascular options

Outcome criteria Table 16.4 Results of combined endovascular catheter

intervention
The primary endpoints of this study were mortality, ampu-
Number Per cent
tation at 30 days, 6 and 12 months and amputation-free
survival of the entire patient group at 6 and 12 months.
Other endpoints were primary removal of occluding Procedures 93 100
thrombus, i.e. primary success and patency rate achieved, Aspiration only 29 31
Thrombolysis 20 22
and the number and percentage of secondary reinterven-
Percutaneous transluminal angioplasty 67 69
tions necessary, whether open surgical or by catheter.
Surgical reinterventions included femorodistal bypasses or Primary success 84 90
thromboendarterectomy. Catheter reinterventions included
repeat thrombus aspiration with and without adjuvant Secondary procedures 28 30
Endovascular 14
lytic therapy and PTA and the placement of a stent in one
Surgical 16
instance.
Secondary endpoints were the ankle systolic pressure Amputation (6 months) 8 9
index (ASPI) achieved by percutaneous thrombus aspir- Below-knee 6
ation with or without complementary PTA and thrombolysis Through-knee 1
and the occurrence of adverse effects of treatment. Special Above-knee 2
attention was given to bleeding complications including Mortality
cerebral haemorrhage with or without associated mortality. 30 days 8
The extent of clot lysis attempted was based on comple- 6 months 16
tion arteriography immediately after catheter therapy in all 12 months 19
patients. The ASPI was obtained during admission or
from the referring hospital if treatment was ambulatory.
Mortality was recorded at 30 days and 6 and 12 months;
these outcome measures were obtained either from our
own records or those of the referring hospitals and phys- (90.3 per cent), primary failure with ASPIs of 0.27 or
icians. For a definition of episodes of major haemorrhage less, immeasurable due to incomplete clot removal or
we adhered closely to the standards set in the TOPAS re-occlusion in nine legs (9.7 per cent).
study. These were as follows: severe blood loss greater Eight patients underwent major amputation, two
than 500 mL within 1 week after therapy (14 days in the above-knee, five below-knee and one through-knee with
TOPAS study), blood loss requiring surgery, transfusion or an amputation-free survival of 82.5 per cent at 6 months
causing hypotension or intracranial haemorrhage. False and 77.9 per cent at 12 months (Fig. 16.6). The mean ASPI in
aneurysms were also excluded. 81 legs treated successfully and measured 1–7 days post
The total dose of urokinase used for direct individual intervention was 0.98  0.23. The 30-day mortality was
thrombus infiltration in the 21 patients treated with lytic 7.8 per cent (seven patients). At 6 months 14 patients
agents was measured and considered in the light of pos- had died (mortality 15.7 per cent) and at 12 months,
sible bleeding complications. 17 patients (mortality 19 per cent). In 28 cases (30.1 per
cent) a secondary intervention was required: in 14 cases
this was an endovascular procedure (suction, lysis, PTA)
and in 16 cases an open surgical procedure. Some patients
Statistical analysis
had to undergo two reinterventions.
A total of 20 legs in 19 patients (21.5 per cent) did have
Quantitative data were expressed as mean 1 SD. Times to
thrombolytic therapy in the form of urokinase, not as the
events were analysed by Kaplan–Meier analysis (Stata,
main line of treatment but as an adjunct to facilitate disso-
Release 5, Stata Corporation, College Station, TX, USA,
lution and aspiration; therefore, only modest doses were
1997). Group differences were assessed with Student’s t test
required, the lytic agent being sprayed within the clot. The
and with the Mann–Whitney U test for non-parametric
individual doses were small and varied between 50 000 and
groups. A P value less than 0.05 was considered significant.
250 000 IU (mean: 112 500  55 901.7 IU) thereby avoid-
ing significant systemic fibrinolytic activity. Accordingly
there were no major bleeding episodes and no haem-
Clinical outcomes atomas at the puncture site with the exception of one case
of false aneurysm which was treated successfully with local
The results are shown in Table 16.4. Primary success with digital compression. No surgical interventions or blood
complete dissolution and complete restoration of patency transfusions were necessary and, of utmost importance,
as proved by angiography was achieved in 84/93 legs not a single case of intracranial haemorrhage occurrred.
 Discussion of the new evidence 189

1.00 with its associated hazards is unnecessary in four out of five

87 cases of acute limb ischaemia. Moreover, in those 20 legs
0.75
(21.4 per cent) where urokinase was actually used the doses
70
were small (112 500  55 900 IU) and not meant to dis-
solve the thrombus entirely but merely to loosen up the
0.50
clot and to facilitate continued aspiration. The mean dose
of urokinase in the TOPAS study26 was, in contrast,
0.25 3.5  1.8 million IU and the mean duration (SE) of
urokinase infusion 24.4  14.2 hours. The mortality rates
0.00 in our study and in those of the TOPAS study as well as the
0 100 200 300 400 baseline patient characteristics indicate that the popula-
Days tions are comparable at least with respect to their comor-
bidity. In our study, the two mechanisms, minimal
Figure 16.6 Kaplan–Meier curve showing 12-month thrombolysis when necessary and clot suction in every case
amputation-free survival in 77.9 per cent of patients
work in synergy. Small doses of lytic agents facilitate clot
aspiration and conversely percutaneous clot removal
0.5 Aspiration alone decreases the need for urokinase. The result is a remarkable
Aspiration  PTA reduction of time necessary to obtain patency and certainly
0.4 Aspiration  PTA well within that taken for surgical intervention.
 lysis
Eight patients (9 per cent) in our series underwent
0.3 major amputation, six below-knee, two above-knee and one
through-knee. In the TOPAS study the amputation-free sur-
0.2 vival rates in patients treated with urokinase were comparable
to those achieved in the surgically treated group. Nevertheless,
0.1
58 major amputations were performed in the urokinase
group (21 per cent), 33 below-knee and 25 above-knee. The
percentage of amputees in the TOPAS study was therefore
0
significantly greater than that in our series (P  0.05); the
Figure 16.7 Percutaneous catheter thrombus suction was proportion of above-knee amputations was also higher.
performed in all patients in this series of 93 legs. It was the only The 6-month and 12-month amputation-free survival
form of therapy in 31 per cent of limbs, it was combined with rates in our retrospective study were 82.5 per cent and
percutaneous angioplasty (PTA) in 47 per cent and in only 77.5 per cent, respectively. In the first prospective ran-
22 per cent was it necessary to add modest doses of urokinase to domised study by Ouriel et al.,18 the Rochester trial, the
produce clot disintegration to facilitate clot aspiration amputation-free survival rates at 1 year were 75 per cent
for the patients assigned to thrombolysis with urokinase
For the entire group of 89 patients the acute occlusive and 52 per cent for those assigned to surgical treatment.
event had occurred within 6.2  7.3 days. In the group of While those figures are similar, both in the Rochester study
19 patients (20 legs) requiring adjunctive thrombolysis to and in our study there is a highly significant difference with
facilitate thrombus suction the process had taken longer respect to outcome and complications of catheter therapy.
(11.15  9.8 days; t  2.56; P ⬇ 0.01). In the 70 patients The Rochester study reported an 11 per cent rate of serious
who did not need lytic therapy the duration of symptoms bleeding complications in patients receiving urokinase
before treatment was shorter (4.7  5.6 days; Mann–Whitney with one death due to haemorrhage.
U test P  0.0045). The Surgery versus Thrombolysis for Ischaemia of the
Percutaneous catheter thrombus aspiration was per- Lower Extremity (STILE) study17 reported a 5.6 per cent
formed in all patients in our series of 93 legs. It was the only rate of serious haemorrhage in patients receiving throm-
form of therapy in 31 per cent and was combined with bolytic agents, with low fibrinogen values identified as a
adjunctive PTA in 47 per cent (Fig. 16.7). In 22 per cent risk factor. In the recent report by Ouriel et al.26 on the
only was it necessary to add modest doses of urokinase to TOPAS study there was a 12.5 per cent incidence of major
produce clot disintegration to facilitate clot aspiration. bleeding in the urokinase group, increasing when patients
received heparin. In four cases these haemorrhages were
intracranial and in one case fatal. The concomitant use of
DISCUSSION OF THE NEW EVIDENCE
heparin had to be restricted and discontinued during the
course of the study. Transfusions of more than 1 unit of
The aim of our single centre retrospective study30 compris- packed red cells were necessary in 92 patients in the urokinase
ing 93 legs in 89 patents was to show that by performing group of the TOPAS study (33.8 per cent). As stated
percutaneous catheter thrombus aspiration, thrombolysis earlier, there were no major haemorrhages at all in our
190 Acute limb ischaemia: endovascular options

single centre study, accordingly transfusions or surgical anaesthesia if open surgery is considered, but it does not
interventions for bleeding complications were unnecessary. interfere with percutaneous clot aspiration, which is per-
These figures underline why it is so important to get away formed under local anaesthesia and offers extra benefit over
from using lytic agents and, if unavoidable, to minimise the open surgery by allowing for PTA and mild fibrinolysis – all
dosage so as not to lower systemic fibrinogen levels by more in one. Hospital stay is reduced and patient comfort con-
than 5–10 per cent. This can be achieved by percutaneous siderably enhanced.
clot suction. Before the introduction of suction therapy in a
study combining short and long term catheter thromboly-
sis Do et al.11 found that fibrinogen levels merely dropped
WHEN SHOULD ENDOVASCULAR OPTIONS
from 2.6  0.7 to 2.5  1.2 g/L when urokinase was used.
BE RECOMMENDED FOR ACUTE LIMB
In this study there were 28 patients (30.1 per cent) who
ISCHAEMIA?
needed a secondary intervention: in 14 cases endovascular,
in 16 cases surgical. In contrast, in the TOPAS study uroki-
nase group26 the percentage of secondary interventions Native arteries
was 55 per cent, suggesting that routine primary catheter
clot extraction, apart from avoiding major haemorrhage Open surgical treatment for acute ischaemia requires either
and shortening the procedure, has the added advantage of general or regional anaesthesia, the latter precluding prior
providing better patency results. administration of anticoagulant therapy which is an unfor-
Although Dotter is credited with the recognition of the tunate requirement. Soft tissues are invariably injured at
potential of selective clot lysis with low dose streptokinase, surgery and may delay wound healing. Surgical treatment
it was Hess et al.15 who were most instrumental in confirm- involves the repeated use of Fogarty balloon catheters
ing and establishing this procedure worldwide. Although which can damage the intima, burst or break and occasion-
Starck et al.,28 Schneider and Hoffmann29 and Mahler31 ally perforate the vessel wall. The degree of patency
repeatedly demonstrated that percutaneous catheter clot achieved cannot be monitored during the process of dis-
extraction was a valuable addition to the armamentarium obliteration but can be assessed by completion angiogra-
of catheter therapy, its real potential was never fully recog- phy. Combined endovascular catheter therapy avoids all
nised elsewhere. Reports on thrombolytic therapy prolifer- these drawbacks. Whereas surgical treatment is expeditious,
ated with the sights constantly set on finding an ever better straightforward and therefore preferable for embolic occlu-
lytic agent. Once it was recognised that streptokinase had its sions of the aorta, iliac and femoral bifurcations, it does
drawbacks and limitations, especially with regard to dur- require an inguinal incision and there is the additional like-
ation of treatment, haemorrhagic complications and aller- lihood of having to treat the ischaemia-reperfusion syn-
gic reactions, attention turned to urokinase and then to drome (see Chapter 2). Endovascular options should
recombinant tissue plasminogen activator (r-tPA)32 in the definitely be given preference in infrainguinal thromboem-
hope of discovering a more powerful agent with less undesir- bolic occlusions, which in practice means those distal to the
able side effects. r-tPA has been shown to be faster acting level of Hunter’s canal, and they are particularly relevant
and fibrinogen sparing and to have an improved safety when the popliteal trifurcation is occluded.
record when compared with other agents, but for practical The earlier caveats of lytic therapy and the absolute and
reasons it has not hitherto replaced urokinase.31 Further relative contraindications of using large doses no longer
improvement of local lytic treatment was sought by play a decisive role when our combined treatment mode is
improving and altering the mode of local delivery either by employed. Nevertheless, these are listed here for those
means of pulsed delivery33 or intraclot spray.34 therapists still using lytic therapy alone with the inevitably
Surprisingly however, the benefits of our regimen in large doses of urokinase or r-tPA. Despite intraclot instilla-
limiting the dosage of, and in the majority of cases render- tion there can be leakage into the general vascular system
ing superfluous, the lytic agent has not received the atten- causing episodes of internal bleeding. The following situa-
tion that it deserves. The shorter duration of aspiration, tions therefore represent absolute or relative contraindica-
the adjunctive interventions and the limited exposure to tions: operative interventions upon the central nervous
thrombolytic agents when they become necessary are the system, lumbar puncture, severe trauma, gastrointestinal
factors we consider responsible for the lower rate of serious and urogenital haemorrhages, uncontrolled hypertension
complications. As the procedure does not take more time (200/100 mmHg), bleeding disorders, aortic aneurysms,
than a surgical intervention we recommend it as first line severe hepatic and/or renal failure, pregnancy (before the
treatment for acute or subacute infrainguinal arterial occlu- third and after the seventh month), bacterial endocarditis,
sions. It can be combined with PTA for the treatment of proliferative diabetic retinopathy and suspicion of throm-
underlying stenotic lesions. Only in a minority of roughly bus in the left heart with its attendant danger of cerebral
20 per cent of cases will a modest amount of lytic agent be infarction. None of these contraindications, however, is a
necessary to facilitate clot disintegration and aspiration. reason for not performing combined endovascular catheter
Preoperative systemic heparinisation may preclude spinal therapy using only clot suction or extraction.
 When should endovascular options be recommended for acute limb ischaemia? 191

Our studies have shown that in those 70 legs where

thrombolysis was not necessary the acute occlusive event
was less than 4.7  5.6 days old and significantly shorter
than in those 20 legs in need of adjunctive thrombolysis
(6.2  7.3 days; Mann–Whitney U test, P  0.0045). In
other words the sooner acute limb ischaemia is treated the
greater the chances that percutaneous clot aspiration alone
will suffice.

Occluded synthetic grafts

Special awareness is justified when contemplating lysis of

Dacron grafts as most lytic agents are capable of rendering
such grafts porous. While positioning of catheters in
thromboembolic occlusions of native arteries is successful
in the vast majority of cases this is not always the case in
autologous vein or synthetic bypass grafts. Catheter
directed clot aspiration, however, is possible in some cases,
as illustrated in Fig. 16.7. This is exemplified in Fig. 16.8 in
which an occluded Dardik graft is disobliterated by percu-
taneous clot extraction. Figure 16.9 shows the thrombotic
material sucked out of the graft. Occluded axillofemoral
Dacron bypasses are readily amenable to surgical thrombec-
tomy using a Fogarty balloon catheter, but the same proced-
ure is not nearly as successful in graft bypasses in the
infrainguinal position, usually because of intimal hyper-
plasia at the anastomotic site. Balloon dilatation of anasto-
motic stenosis does not have a lasting effect and revision
surgery is usually envisaged in our experience. Comerota
et al.23 reported that in 39 per cent of patients randomised
to lysis catheter placement failed and surgical treatment
was required. Overall, a significantly better composite clin-
ical outcome at 30 days and at 1 year was observed in the
surgical group compared with lysis.

Autologous venous grafts

Here again the problems and difficulties are the same as

those for synthetic grafts. In the case of acute ischaemia a
re-do reconstruction is probably the best alternative.

(a) (b)

Figure 16.8 (a) Acute ischaemia caused by occlusion of an

umbilical Dardik graft. (b) Patency restored after transcutaneous
catheter clot extraction but the graft channel remains irregular
(with kind permission of Prof J Largiadèr and Dr E Schneider, Figure 16.9 Clots extracted percutaneously from occluded Dardik
Zürich) graft in Fig. 16.8 (with kind permission of Dr E Schneider, Zürich)
192 Acute limb ischaemia: endovascular options

Combined endovascular catheter directed treatment is not an which time they have remained unchanged in practice. The
easy option unless started immediately after graft occlusion. impact and potential of IOL in lowering the dosage of lytic
agent and in accelerating the restoration of blood flow has
hitherto not been fully recognised and appreci-
Intraoperative intra-arterial local thrombolysis ated.30,31,36–38 Further debate on advances in this area of
and clot aspiration management of acute lower limb ischaemia has been high-
lighted in the recent literature.36–44
Since 1990 we have adopted a method of intra-arterial
intraoperative lysis (IOL) adhering to the principles and
experimental work of Quinones-Baldrich et al.8 for those WHAT DEVELOPMENTS DOES THE FUTURE
desperate cases in which viability of the limb is seriously HOLD?
threatened and the entire length of arterial vasculature
cannot be opacified angiographically. Other authors have
adopted a similar policy.33–38 If, on exploration, the femoral, The latest development in the treatment of acute and sub-
popliteal and all three crural arteries prove to be occluded, acute thrombotic arterial occlusions has been the advent of
amputation is the usual outcome.
Rewarding results, however, can be obtained by bypass-
ing the occluded femoropopliteal segment with a poly-
tetrafluoroethylene (PTFE) graft and introducing catheters
into all three crural arteries simultaneously (Fig. 16.10); up
to 175 000 units of urokinase is injected into each vessel to
a total dose of 500 000 units over a 25–30-minute period,
with intermittent clot aspiration to hasten disobliteration
and increase the efficacy of lysis. Simultaneously, another
member of the operating team can prepare the femoral
artery bifurcation for the upper anastomosis of a PTFE
graft. In Fig. 16. 11 a thrombolised distal artery offering the
necessary run-off can be seen, a prerequisite for femorop-
opliteal bypass surgery. Figure 16.12 shows the result of IOL
of the posterior tibial artery behind the medial malleolus.
This treatment protocol for IOL is similar to the
University of California at Los Angeles (UCLA) protocol,
which in turn is based on investigations of experimentally
induced thrombosis:8 the angiographic improvement fol-
lowing thrombectomy was 20 per cent reaching 80 per cent
if thrombectomy was followed by IOL with streptokinase
and heparin for 30 minutes. Our present protocol uses
250 000 units of urokinase dissolved in 100 mL 0.9 per cent
sodium chloride with 1000 units of heparin administered
into the clot for 30 minutes as described above. The results
first published in 19939 are shown in Table 16.5, since
Figure 16.11 Complete
occlusion of the entire arterial
vasculature from groin to foot.
Local thrombectomy of the
distal part of the popliteal artery
has been performed for distal
anastomosis to a
polyfluorotetraethylene (PTFE)
graft. From here on downwards
intraoperative lysis using 500 000
units of urokinase for 25 minutes
and clot suction of the crural
arteries was performed reopening
the previously occluded posterior
Figure 16.10 Schematic drawing showing the principle of tibial artery. There now is
introducing three catheters from an infragenicular approach into sufficient run-off to enable
the crural arteries for simultaneous intraoperative thrombolysis patency of the femoropopliteal
and clot suction bypass graft
 What developments does the future hold? 193

a rotational thrombectomy device, the Stator Rotarex 5 mm from the tip rotate over two corresponding open slits
System (Straub Medical AG, Wangs, Switzerland; www. in the internally placed stationary Stator. The Rotarex cut-
straubmedical.com) which has been clinically tested and ting head with its internally placed blades rotates around
improved over the past three years.43,44 This Rotarex rota- the Stator at 40 000 rotations per minute driven by a spiral
tional thrombus debulking device is made of stainless steel which also has a transporting function; the thrombotic
fashioned into a wedge-shaped cutting head ground down material is drawn by the suction force through the inter-
to form a blunt rounded tip with a central opening for the nally placed cutting edges and is broken down and
guidewire over which the catheter passes. Two open slits removed through the middle of the catheter leaving no
residual detached material in the lumen. The catheter fol-
lows the direction given by the guidewire in the blood ves-
sel and allows controlled movements of the rotating head
to be made.
This device has been designed for removing both fresh
thrombus and organised thrombotic occlusions of up to 6
months but is not designed to deal with calcified plaque.
When underlying plaques are uncovered they should be
treated by balloon angioplasty. The diameter of the
recanalised artery is approximately double that of the
catheter used. The catheters presently available are 8 Fr
antegrade, 8 Fr crossover and 6 Fr antegrade, all of which
can be adapted to the motor unit. The aspiration capacity
in fresh thrombus is 0.5–1 cm/s. For the sake of safety,
mainly to avoid the danger of perforation, it is recom-
mended that the Rotarex system not be used beyond the
tibioperoneal trunk.
Figure 16.12 A case of complete ischaemia of the right foot. This mechanical ‘four-in-one’ device incorporates the
Local fibrinolysis via the posterior tibial artery provides excellent combined functions of thrombus detachment, suction,
visualisation of the pedal circulation cutting and transport of debris out of the vessel. It has
enormous potential, acknowledging the aforementioned
danger of perforation and the costs, one catheter currently
Table 16.5 Results of intraoperative clot lysis combined with priced at SF1500. Nevertheless, with commercially avail-
clot suction for total ischaemia undertaken from 1 January 1990 able devices such as this, the time spent in reopening
to 31 July 1992 occluded arteries is being significantly shortened and soon
the use of thrombolytic agents may be dispensed with
Number of cases 34 entirely.
Limb salvage rate 86 per cent The algorithm in Fig. 16.13 provides guidelines for the
Patency rate 81 per cent
modern management of acute and subacute ischaemia of
30-day mortality 14 per cent
the lower limbs indicating optimal methods of dealing
Wound haematoma 4 per cent
with occlusions at different levels of the arterial tree.

Acute arterial occlusion

Surgical thrombectomy with Fogarty

Clinical and radiological examination
balloon catheter/PTA/stent

Occlusion above
Rotarex thrombectomy  PTA
inguinal level

Occlusion at femoral Open surgery and thrombectomy 

bifurcation endarterectomy/intimectomy and patch

Occlusion of superficial
Rotarex thrombectomy
femoral Figure 16.13 Algorithm of the
current treatment of acute and
Rotarex thrombectomy of femoropopliteal subacute ischaemia of the lower
Occlusion of superficial femoropopliteal
axis and ‘conventional’ catheter aspiration extremities. PTA, percutaneous
axis and popliteal trifurcation
of crural arteries angioplasty
194 Acute limb ischaemia: endovascular options

Nicolaides AN (eds). Limb Salvage and Amputation for Vascular

Conclusions Disease. London: WB Saunders, 1988: 105–12.
2 Srinivasan R, Cooper G, Bell PRF. Popliteal embolectomy; does
Acute and subacute thrombotic and embolic occlusions it still have a role? Eur J Vasc Surg 1992; 6: 424–6.
of the femoral, popliteal and infrapopliteal arteries are 3 Dinkel HP, Hoppe H, Baumgartner I, et al. Low-osmolar
especially amenable to local endovascular catheter gadodiamide as alternative contrast agent for diagnostic
therapy. A new concept, combining clot aspiration with angiography and angiographic interventions in patients with
PTA whenever atherosclerotic plaques are uncovered, renal insufficiency. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb
together with clot suction and modest doses of lytic Verfahr 2002; 174: 56–61, (comment in Rofo Fortschr Geb
Rontgenstr Neuen Bildgeb Verfahr 2003; 1, author reply 571–2).
agents when necessary, heralds a fresh era in the treat-
4 Tepel M, Van der Giet M, Schwarzfeld C, et al. Prevention of
ment of acute ischaemia of the limbs. The combined
radiographic-contrast-agent induced reductions in renal
procedures shorten interventional disobliteration and at function by acetylcysteine. N Engl J Med 2000; 343: 180–4.
the same time avoid systemic thrombolysis with the 5 O’Brian RS, Thomas H, Crow A, Lamont PM. Calf vessel
inherent dangers of haemorrhage. We have shown that preservation in peripheral vascular disease – angiography versus
a 90 per cent primary patency rate and a 77 per cent pulse generated run-off. Eur J Vasc Surg 1993; 7: 177–9.
amputation-free 12 month survival rate can be achieved 6 Comerota AJ, White JV, Gosh JD. Intraoperative intra-arterial
with little risk of bleeding complications. Transfusions thrombolytic therapy for salvage of limbs in patients with distal
are unnecessary. This treatment modality works arterial thrombosis. Surg Gynecol Obstret 1989; 169: 283–9.
adequately with most degrees of acute ischaemia 7 Alavaikka A. Local thrombolytic therapy as a support for catheter
because of the rapidity of the procedure. It is therefore embolectomy in limb arterial occlusions. Ann Chir Gynaecol
1991; 80: 357–62.
recommended as first line treatment for acute as well as
8 Quinones-Baldrich WJ, Baker JD, Busuttil RW, et al.
subacute ischaemia caused by infrainguinal occlusions.
Intraoperative infusion of lytic drugs for thrombotic complication
Currently, at the Vascular Department of the of revascularization. J Vasc Surg 1989; 10: 408–17.
University of Berne, this approach is frequently being 9 Knaus J, Ris HB, Do D, Stirnemann P. Intraoperative thrombolysis
combined with Straub Rotarex catheter aspiration and as an adjunct to surgical revascularisation for infrainguinal limb-
significantly accelerating the reopening of occluded ves- threatening ischemia. Eur J Vasc Surg 1993; 7: 507–12.
sels. The success of this new technique has been such 10 Scott DJ, Wyatt MG, Wilson YG, et al. Intraarterial streptokinase
that thrombolytic agents are being dispensed with com- infusion in acute lower limb ischaemia. Br J Surg 1991; 78:
pletely. It has already undergone extensive clinical trial 732–4.
and holds significant promise for the future. 11 Do DD, Mahler F, Triller J, Nachbur B. Combination of short and
long-term catheter thrombolysis for peripheral arterial occlusion.
Eur J Radiol 1987; 7: 235–8.
12 Nachbur B. Treatment of acute ischaemia; every general
Key references surgeon’s business? Eur J Vasc Surg 1988; 2: 281–2.
13 Bywaters EGL, Beall C. Crush injuries with impairment of renal
Do DD, Mahler F, Triller J, Nachbur B. Combination of short and function. BMJ 1941; 1: 427.
long-term catheter thrombolysis for peripheral arterial 14 Nachbur B, Horber F, Sigrist S. Metabolic disorders in acute limb
occlusion. Eur J Radiol 1987; 7: 235–8. ischemia. In: Bergan JJ and Yao JS (eds). Vascular Surgical
Nachbur B. Treatment of acute ischaemia; every general surgeon’s Emergencies. Orlando, FL: Grune and Stratton, 1987.
business? Eur J Vasc Surg 1988; 2: 281–2. 15 Hess H, MietaschK, Rath H. Local low-dose thrombolytic therapy
Ouriel K, Veith FJ, Sasahara AA, for the TOPAS investigators. of peripheral arterial occlusions. N Engl J Med 1982; 307:
A comparison of recombinant urokinase with vascular surgery 1627–30.
as initial treatment for acute arterial occlusions of the legs. 16 Ouriel K. Comparison of surgical and thrombolytic treatment of
N Engl J Med 1998; 338: 1105–11. peripheral arterial disease. Rev Cardiovasc Med 2002; 3(suppl 2):
Zehnder T, Birrer M, Do DD, et al. Percutaneous catheter thrombus S7–16.
aspiration for acute or subacute arterial occlusion of the legs: 17 Graor RA (Study Chairman). Results of a prospective randomized
how much thrombolysis is needed? Eur J Vasc Endovasc Surg trial evaluating surgery versus thrombolysis for ischaemia
2000; 20: 41–6. of the lower extremity. The STILE trial. Ann Surg 1994; 220:
Zeller T, Frank U, Burgelin K, et al. Early experience with a 251–68.
rotational thrombectomy device for treatment of acute and 18 Ouriel K, Shortell CK, DeWeese JA, et al. A comparison of
subacute infra-aortic arterial occlusion. J Endovasc Ther 2003; thrombolytic therapy with operative revascularization in the
10: 322–31. initial treatment of acute peripheral arterial ischemia. J Vasc
Surg 1994; 19: 1021–30.
19 Swischuk JL, Fox PF, Young K, et al. Transcatheter intraarterial
REFERENCES infusion of rt-PA for acute lower limb ischemia: results and
complications, J Vasc Interv Radiol 2001; 12: 423–30.
20 Braithwaite BD, Petrik PV, Ritschie AWS, Earnshaw JJ.
1 Jamieson CW. Is it important to differentiate between acute and Computerized angiographic analysis of the outcome of
acute on chronic ischaemia? In: Greenhalgh RM, Jamieson CW, peripheral thrombolysis. Am J Surg 1995; 170: 131–5.
 References 195

21 Braithwaite BD, Quinones-Baldrich WJ. Lower limb intraarterial 33 Buckenham TM, George CD, Chester JF, et al. Accelerated
thrombolysis as an adjunct to the management of arterial and thrombolysis using pulsed intra-thrombus recombinant human
graft occlusions. World J Surg 1996; 20: 649–54. tissue type plasminogen activator (rt-PA). Eur J Vasc Surg 1992;
22 Ouriel K, Veith FJ, Sasahara AA ,for the TOPAS investigators. 6: 237–40.
Thrombolysis or peripheral arterial surgery: phase I results. 34 Yusuf SW, Whitaker SC, Gregson HS, et al. Prospective
J Vasc Surg 1996; 23: 64–73. randomised comparative study of pulse spray and conventional
23 Comerota AJ, Weaver FA, Hosking JD, et al. Results of a local thrombolysis. Eur J Vasc Endovasc Surg 1995; 10: 136–41.
prospective, randomized trial of surgery versus thrombolysis for 35 Parent FN, Piotrowski JJ, Bernhard VM, et al. Outcome of
occluded lower extremity bypass grafts. Am J Surg 1996; 172: intraarterial urokinase for acute vascular occlusion. J Cardiovasc
105–12. Surg 1991; 32: 680–9.
24 Armon MP, Yusuf SC, Whitaker RH, et al. Results of 100 cases of 36 Hopfner W, Bohndorf K, Vicol C, Loeprecht H. Percutaneous
pulse-spray thrombolysis for acute and subacute leg ischaemia. hydromechanical thrombectomy in acute and subacute lower
Br J Surg 1997; 84: 47–50. limb ischemia. Rofo Fortschr Geb Rontgenstr neuen Bildgeb
25 Davidian MM, Powell A, Benenati JF et al. Initial results of Verfahr 2001; 173: 229–35.
reteplasma in the treatment of acute lower extremity arterial 37 Canova GR, Schneider E, Fischer L, et al. Long-term results of
occlusions. J Vasc Interv Radiol 2000; 11: 289–94. percutaneous thrombo-embolectomy in patients with
26 Ouriel K, Veith FJ, Sasahara AA, for the TOPAS investigators. infrainguinal embolic occlusions. Int Angiol 2001; 20: 66–73.
A comparison of recombinant urokinase with vascular surgery 38 Wang HJ, Kao HL, Liau CS, Lee YT. Export aspiration catheter
as initial treatment for acute arterial occlusions of the legs. thrombosuction before actual angioplasty in primary coronary
N Engl J Med 1998; 338: 1105–11. intervention for acute myocardial infarction. Katheter Cardiovasc
27 Porter JM. Thrombolysis for acute arterial occlusion of the legs Interv 2002; 57: 332–9.
[editorial]. N Engl J Med 1998; 338: 1148–9. 39 Ouriel K. Comparison of surgical and thrombolytic treatment
28 Starck E, McDermott J, Crummy A, et al. Die perkutane of peripheral arterial. Rev Cardiovasc Med 2002; Suppl 2:
Aspirations-Thromboembolektomie: eine weitere transluminale S7–S16.
Angioplastiemethode. Deutsche Med Wochenschr 1986; 111: 40 Carlson GA, Hobollah JJ, Sharp WJ. Surgical thrombectomy:
167–72. current role in thromboembolic occlusions. Tech Vasc Interv
29 Schneider E, Hoffmann U. Perkutane lokale Lysetherapie und Radiol 2003; 6: 14–21.
Thrombenextraktion bei Verschlüssen der Extremitätenarterien. 41 Kalinowski M, Wagner HJ. Adjunctive techniques in
Internist 1996; 37: 607–12. percutaenous mechanical thrombectomy. Tech Vasc Interv Radiol
30 Zehnder T, Birrer M, Do DD, et al. Percutaneous catheter 2003; 6: 6–13.
thrombus aspiration for acute or subacute arterial occlusion of 42 Henke PK. Approach to the patient with acute limb ischemia:
the legs: how much thrombolysis is needed? Eur J Vasc Endovasc diagnosis and therapeutic modalities. Cardiol Clin 2002; 20:
Surg 2000; 20: 41–6. 513–20.
31 Mahler F. Lokale Katheterthrombolyse und Katheterthrombektomie. 43 Zeller T, Frank U, Burgelin K, et al. Early experience with a
In: Mahler F (ed.) Katheterinterventionen in der Angiologie. rotational thrombectomy device for treatment of acute and
Stuttgart: Thieme, 1990: 113–29. subacute infra-aortic arterial occlusion. J Endovasc Ther 2003;
32 Verstraete M, Hess H, Mahler F, et al. Femoro-popliteal artery 10: 322–31.
thrombolysis with intra-arterial infusion of recombinant tissue- 44 Zeller T, Frank U, Burgelin K, et al. Treatment of acute embolic
type plasminogen activator, report of a pilot trial. Eur J Vasc Surg occlusions of the subclavian and axillary arteries using a
1988; 2: 155–9. rotational thrombectomy device. Vasa 2003; 32: 11–16.
This page intentionally left blank
 17
Graft Maintenance and Graft Failure

DAVID K BEATTIE, ALUN H DAVIES

The problem of graft failure 197 Management 202

Graft failure: aetiology and pathophysiology 197 Postoperative care 205
The maintenance of graft patency and diagnosis of the 199 Results 206
failing graft References 207
Diagnosis of the failing graft 199

THE PROBLEM OF GRAFT FAILURE GRAFT FAILURE: AETIOLOGY AND

PATHOPHYSIOLOGY
The advent of infrainguinal arterial reconstruction brought
with it a new problem, that of graft stenosis and graft fail- Graft failure is conventionally described as occurring in one
ure, irrespective of the conduit used for bypass. There has of three postoperative periods: early graft failure develops
been no significant improvement in graft patency rates fol- within 30 days of surgery, while the majority of failures
lowing such reconstructions since Szilagyi et al.1 described occur during an intermediate period from 1 month to
the phenomena nearly 30 years ago, with most series report- 1 year and late failures are those that come to light after
ing that 20–30 per cent of all grafts develop a stenosis or 1 year.
fail, usually within 1 year of surgery.2–5 Between 5 and 15 per cent of all infrainguinal grafts will
The potential benefits of identifying a failing graft before fail within 1 month of surgery.13 Continued early graft
it fails are clear. In one series 79 per cent of patients oper- patency is dependent, given appropriate patient selection,
ated on for limb salvage were again at risk of limb loss upon technical factors and, where vein is used, vein quality.
following graft occlusion, and 91 per cent of claudicants Technical problems occur in up to 15 per cent of femorodis-
were either worse or similar to their preoperative status.6 tal bypasses and identification of these is important to graft
Furthermore, the revision of surveillance detected graft patency. Twisting and kinking of the vein graft are the usual
stenoses using either surgical or endovascular techniques causes of graft malfunction. The former is more common
has been shown to give excellent primary assisted patency in reversed and fully mobilised grafts, though the latter is
rates in excess of 80 per cent at 5 years7–9 whereas revascu- less common than in synthetic grafts.14–16 Insertion of the
larisation following graft occlusion yields poor long term graft under tension may also jeopardise flow.
limb salvage with secondary graft patency rates as low as 43 While the use of an in situ technique decreases the risk
per cent at 5 years.10–12 In the majority of cases it is of malalignment, the risk of persistent valves and tributar-
technically easier to revise a graft before rather than after ies is increased. The former may not always be immediately
it fails. detectable, even with on-table arteriography, and failure to
This chapter addresses the question of both the detec- adequately disrupt the valves correlates with occlusion. There
tion and management of the ‘at risk’ graft and the manage- is evidence that, even in patients with reversed vein grafts,
ment of the failed graft. However, it is important to stress competent valves may create a pressure trap effect with
that, where a graft has already failed, reintervention is only segmental hypertension, flow stagnation and graft thicken-
undertaken when there is renewed critical ischaemia or ing, and stenosis.17 Non-division of saphenous tributaries
severe symptoms due to ischaemia. causes arteriovenous fistulae with shunting of blood to the
198 Graft maintenance and graft failure

Vascular endothelial cell and smooth muscle cell (SMC)

injury. Release of endogenous intracellular mitogens and
Immediate to minutes
growth factors. Induction of early response proto-oncogenes.
Platelet aggregation and polymorph and monocyte adherence

Synthesis and release of upregulated growth factors and

Day 1
cytokines. Polymorph and macrophage infiltration

DNA synthesis by medial SMCs. Proliferation of medial

Day 2–7 SMCs and migration to the intima. Phenotypic change to the
synthetic state

Second SMC proliferative phase. Production of chondroitin

Day 7–10
sulphate in the subendothelial space

Increased production of matrix including heparin proteoglycan

synthesis. SMC proliferation becomes maximal at 14 days and
Day 10 onwards then declines. Eventually SMCs return to a contractile
phenotype and collagen is formed. Thickness of intimal
hyperplasia peaks at about 1 month

Figure 17.1 The time course and events in the formation of intimal hyperplasia

deep veins and a resultant decrease in distal graft flow;18

this may occur in 20 per cent of in situ grafts. Occasionally
fistulae enlarge causing graft failure.
Anastomoses are particularly prone to technical error,
and the problem is magnified as the bypass becomes more
distal and the recipient vessel therefore smaller. The most
common problems are intimal dissection and luminal nar-
rowing, and a number of techniques are used to prevent
these complications. These include the use of fine sutures,
passage of the suture from the intimal surface through to
the adventitia, the fashioning of anastomoses over catheters
and maintenance of an optimal ratio between cross-sectional
area of the graft and artery. However, early graft patency is
also substantially dependent on graft inflow and outflow.
Poor cardiac output and proximal atherosclerotic disease Figure 17.2 Cross-section of a stenosed vein graft (alpha-actin
stain 40). Note the marked luminal narrowing secondary
are harbingers of a poor outcome. Nevertheless, a run-off
to the development of florid neointimal hyperplasia
deficiency is the most common cause of early graft failure
and this is dependent upon graft site and the degree of pre-
operative ischaemia. For instance, it has been demonstrated which can range from mild endothelial denudation to severe
that graft outcome correlates with tibial vessel patency19 and medial disruption. Histologically, medial smooth muscle cells
that an intact pedal arch is important. alter phenotype, proliferate and migrate to the intima. In the
Up to 75 per cent of failures occur during the inter- intima a second phase of proliferation occurs and the cells
mediate period, by far the most common cause being the become embedded in an extracellular matrix comprising
development of neointimal hyperplasia with resultant, usu- mainly collagen.20 Lesions develop quickly, often within 2
ally single, discrete stenoses, though valve cusp fibrosis and weeks of surgery; the time course of these events is shown in
aneurysmal dilatation may also be responsible. Neointimal Fig. 17.1. A cross-section through a vein graft stenosis sec-
hyperplasia is a uniform vascular response to vessel injury, ondary to marked intimal hyperplasia can be seen in Fig. 17.2.
 Diagnosis of the failing graft 199

Late graft failure is significantly less common than

intermediate or early failure. The primary cause is disease
progression in the native inflow, or more commonly run-off,
vessels compromising graft flow and leading to thrombosis
which threatens survival of the limb (see Chapter 2). The act
of grafting itself may increase the rate of progression of native
vessel disease, though failure may also be due to the devel-
opment of atheroma within the graft. Rarely vein grafts may
fail due to other disease processes such as aneurysmal dilata-
tion with subsequent thrombosis, embolism or rupture.
(a) (b)

THE MAINTENANCE OF GRAFT PATENCY Figure 17.3 Cross-section of a segment of saphenous vein taken
AND DIAGNOSIS OF THE FAILING GRAFT at the time of bypass surgery. (a) Haematoxylin and eosin stained
section of saphenous vein 40. (b) Alpha-actin stain. Note
The task of ensuring continued graft patency following the marked existing intimal hyperplasia present even before the
vein has been arterialised
infrainguinal reconstruction begins preoperatively. Preopera-
tive assessment involves adequate assessment of both the
clamping the patient should be heparinised. On completion
arterial lesion and, if vein is to be used, the proposed donor
of the bypass a check for patency and technical errors should
vein. Traditionally, recourse to arteriography, preferably
be made. This may involve the use of intraoperative Doppler,
biplanar, has been necessary. Similarly, arteriography is
on-table arteriography or angioscopy, though this has not
used to assess outflow, particularly in the presence of distal
found routine acceptance. A check on haemodynamic
disease, though it may fail to demonstrate patent distal ves-
improvement should also be made prior to the patient
sels. Subtraction techniques may provide additional infor-
leaving hospital. This should involve at least re-estimation
mation. Duplex, however, has become the mainstay of
of the ankle:brachial pressure index (ABPI), but preferably
non-invasive investigation, and Doppler insonation and
duplex examination.
pulse generated run-off are superior to arteriography in
identifying patent distal vessels.21
The importance of assessing the donor vein prior to DIAGNOSIS OF THE FAILING GRAFT
bypass is increasingly being recognised. Preoperatively, an
assessment of vein anatomy, compliance and internal diam-
eter is feasible. Previously, venography was the gold stand- There is strong evidence that bypass graft surveillance
ard but in many units duplex Doppler is now the mainstay increases the detection of flow limiting graft stenoses and
for preoperative venous assessment, providing information that treatment of these prior to occlusion, i.e. when the
on vein wall morphology,22 blood flow, calibre, varicosities graft is ‘at risk’ rather than ‘failed’ leads to improved long
and tributaries, as well as cusps.23 Duplex accurately predicts term graft patency.31 The primary methods of graft surveil-
vein location, size and quality24 and is better than venogra- lance include clinical examination, ABPI estimation, duplex
phy in assessing calf vein diameter.25 In one series, duplex imaging and arteriography. Of these, duplex scanning is
directly diagnosed pre-existing venous disease, with respect increasingly recognised as the most valuable, though con-
to wall thickness, calcification and occlusion, in 62 per cent siderable debate surrounding its use and efficacy still exists.
of cases.26 More recently, these pre-existing changes have
been correlated with the development of vein graft steno- Arteriography
sis.27,28 Figure 17.3 shows a cross-section of a biopsy taken
from the donor saphenous vein at the time of bypass sur- Biplanar arteriography remains the gold standard in arter-
gery. Both preoperative internal vein diameter and compli- ial assessment and is able to detect both native vessel and
ance have been shown to correlate with graft survival.5,29,30 graft stenoses. It has advantages over non-invasive methods
The importance of treating the patient medically as well as in that image resolution is not compromised proximal to
surgically at the time of operation, or well before it, is well the inguinal ligament, pressure gradients across iliac
documented, both to protect the graft and to limit the pro- lesions can be measured and new techniques allow multi-
gression of native vessel disease. Issues to be addressed planar imaging. However, as a means of surveillance it is
include the cessation of smoking, adequate treatment of unacceptable due to its invasive nature and the associated
hypertension and diabetes, the identification and treatment morbidity and mortality, the costs involved and the lim-
of hyperlipidaemias and hyperhomocysteinaemia and the ited resource supply. Invasive arteriography no longer
institution of antiplatelet therapy where not contraindicated. has a place in the routine surveillance of grafts postopera-
There are important peroperative steps to improve the tively, but remains paramount in the confirmation of non-
immediate outcome of bypass surgery. Prior to arterial invasive findings and in planning revision procedures.
200 Graft maintenance and graft failure

Figure 17.4 (a) Arteriogram showing

intimal vein graft stenosis initially
(a) (b)
detected on duplex scanning. (b)
Arteriogram after interposition grafting

Figure 17.4 shows a vein graft stenosis initially detected on DUPLEX GRAFT SURVEILLANCE
duplex scanning and here confirmed by arteriography at
The rationale behind duplex graft surveillance is based
the time of balloon angioplasty.
upon the supposition that the detection and correction of
graft stenoses, with consequent improvements in patency
rates will result in improved limb survival. This, however,
Non-invasive surveillance has never been proved in a large randomised trial. In fact, a
large summation analysis of 6649 grafts comparing out-
ANKLE:BRACHIAL PRESSURE INDICES
comes for grafts entered into a duplex surveillance pro-
The use of resting ABPI estimation in the detection of gramme with those where surveillance was not performed
graft stenoses has been accepted for some time. There found that, while surveillance appeared to improve graft
are, however, considerable limitations to the technique. patency rates, there was no impact whatsoever upon limb
Ankle:brachial pressure index values may vary by as much as salvage and amputation rates.31 Furthermore, it has never
0.1, without any change in the status of the patient. A fall of been proved adequately that stenosis inevitably means
0.15 in the ABPI is generally accepted as being significant, either subsequent occlusion or potential limb loss. There
but such a fall requires the presence of a stenosis of at least are data, however, which suggest quite the opposite. Those
50 per cent. At this level graft occlusion is likely to be sudden from Bristol relating to 275 grafts showed that there was no
and hence the opportunity to revise the graft before failure is difference in the 12-month cumulative patency rates between
lost. Several studies have shown that even a fall in the resting grafts with treated and those with untreated lesions.35 In
ABPI of at least 0.2 does not predict grafts that will fail.32,33 one study of 80 grafts followed angiographically, 22 grafts
Treadmill testing may however increase the sensitivity of were shown to have a stenosis, five of which occluded, but
ABPI measurements in the detection of stenosis, just as it four in the group of 58 with no stenosis also occluded.36
has been shown to do for native vessel disease. A significant Others have doubted that the non-invasive detection of
number of patients with stenosis and a stable resting ABPI graft stenoses identifies those grafts at risk of occlusion.32,37,38
have been shown to have a fall in the ABPI after exercise.34 Even with the presumption that graft stenosis predisposes
Ankle:brachial pressure index measurements may be of little to graft occlusion and limb loss, a number of studies failed
use in diabetic patients due to vessel calcification. to find any benefit to duplex surveillance.
 Diagnosis of the failing graft 201

A randomised controlled Finnish study examining 185

vein grafts studied with either duplex surveillance or ABPI
measurements failed to find any benefit of duplex surveil-
lance.39 Any differences in outcome were apparent during
the first postoperative month, prior to the commencement
of surveillance.
Another group looking at a mixture of 85 vein, poly-
tetrafluoroethylene (PTFE) and composite grafts found no
support for the efficacy of prophylactic graft revision for
grafts identified as failing by currently accepted duplex cri-
teria.40 This perhaps reflects, at least in part, the increas-
ingly accepted belief that duplex surveillance has a place in
the surveillance of vein grafts but not prosthetic grafts. In a
series of 69 prosthetic grafts from the Leicester group, for
example, 14 failed after 30 days, 12 of which were not pre-
dicted by the surveillance programme.41
The results of an ongoing large randomised multicentre
trial are awaited to provide direct evidence of the benefits
of a duplex surveillance programme for vein grafts in terms
of limb salvage, quality of life and cost benefit.42
The above evidence must call into question the place of
duplex in graft surveillance but even if the concept of duplex
graft surveillance is accepted there are problems. The gen-
erally accepted duplex criteria for the detection of a failing
graft are shown in the box below, and Fig. 17.5 shows a
duplex detected vein graft stenosis. A peak systolic velocity
(PSV) greater than 150 cm/s implies a diameter reduction
of at least 80 per cent.43 The use of standardised criteria
means that the precision and reproducibility of duplex in
graft surveillance from laboratory to laboratory must be
excellent. In a blinded trial of interobserver agreement in Figure 17.5 Duplex scan of a tight vein graft stenosis. The
the duplex scanning of vein grafts the kappa statistic was just duplex performed here demonstrates an elevated peak systolic
0.69. This signifies ‘good’ agreement, but is it good enough?44 velocity and hence stenosis
The PSV criteria are open to question and there are doubts
about the PSV ratio (PSVR) at which grafts should be revised.
For example, there is evidence that graft stenoses with a
PSVR of less than 3.0 can be safely left, provided that Duplex criteria for the identification of
duplex surveillance is performed every 3 months.45 This the failing graft
suggests that current duplex criteria for the detection of
stenosis may be too sensitive, a viewpoint supported by a • ABPI fall 0.2
series of 46 grafts in which duplex detected abnormalities, • PSV 45 cm/s
as defined by a PSVR of more than 3.0, were followed up. • Increase in PSV to 150 cm/s
Only 14 grafts were eventually revised, and only three • PSVR across a stenosis 2.0
occluded while being followed. All three showed a PSVR in
excess of 7.0 prior to occlusion.46 In the same group, and in
a trial using the same criteria, a comparison of the outcome There is also ongoing controversy as to the ideal follow-
of failing grafts which underwent prophylactic revision up time for graft surveillance. On one hand there are the
with those in which no revision was performed, found that advocates of long term duplex surveillance. For example,
primary and secondary patency and limb salvage rates were Lundell et al. in their randomised trial of graft surveillance
not significantly different.47 One suggestion, based on a versus non-surveillance found that significant differences
study of 121 patients demonstrating that graft diameter in patencies were found only by year 2.49 This stance is
and the location of the distal anastomosis significantly shared by the Leicester group which feels that vein graft
affect the flow velocity with the graft, is that currently surveillance should be lifelong.50 Mills et al., however, found
established criteria for graft revision on the basis of veloc- that of 91 grafts that were normal at 3 months only two
ity parameters may be improved if they can be modified to subsequently developed de novo stenoses.51 All grafts pro-
account for graft diameter and outflow.48 gressing to high grade stenosis had a duplex abnormality at
202 Graft maintenance and graft failure

Table 17.1 The Transatlantic inter-Society Consensus (TASC) has published surveillance guidelines as shown in Table 17.1.
guidelines for the surveillance of infrainguinal bypass grafts It is noted that there is a need for establishing the cost effec-
tiveness and optimal duration for surveillance, but that sur-
Recommended surveillance programme for vein bypass grafts veillance should be for at least 2 years.
Patients undergoing vein bypass graft placement in the lower
extremity for the treatment of claudication or limb-threatening
ischaemia should be entered into a surveillance programme.
This programme should consist of: MANAGEMENT
• Interval history (new symptoms)
• Vascular examination of the leg with palpation of inflow, An algorithm (Fig. 17.6) illustrates pathways of manage-
graft and outflow pulses ment, the treatment options available, the chances of bypass
• Periodic measurement of resting and, if possible, post- graft failure and surveillance to thwart that outcome.
exercise ankle:brachial indices
• Duplex scanning of the entire length of the graft, with
calculation of the peak systolic velocities and the velocity
ratios across all identified lesions The failing graft
• Surveillance should be performed in the immediate
postoperative period and at regular intervals for at least Intervention in the failing graft has been shown to result in
2 years excellent primary assisted patency rates.31 There is consider-
able inconsistency in the literature as to the level of stenosis,
Recommended surveillance programme for prosthetic particularly with respect to vein grafts, at which intervention
bypass grafts should be considered. It is generally agreed however that a
Patients undergoing prosthetic femoropopliteal or femorotibial rapidly progressive lesion, or a lesion deemed to be advanced
bypass for the treatment of claudication or limb-threatening on the basis of velocity criteria, should be revised. In our
ischaemia should be entered into a surveillance programme. institution a PSVR in excess of 2.0, indicating a reduction in
This programme should consist of:
diameter of 50 per cent and a cross-sectional reduction of
• Interval history (new symptoms) 70 per cent, is considered to be haemodynamically significant.
• Vascular examination of the leg with palpation of inflow, Many moderately severe lesions are detected on duplex
graft and outflow pulses
scanning in the early postoperative period. Most will remain
• Periodic measurement of resting and, if possible, post-
exercise ankle:brachial indices
stable and will not require revision. Where intervention is
• Surveillance should be performed in the immediate deemed necessary options include percutaneous angio-
postoperative period and at regular intervals for at least plasty, vein patch angioplasty and interposition or jump
2 years bypass grafting. Although some studies have shown equiva-
lent results for surgery and balloon angioplasty for all
stenoses,56 in practice, short lesions such as isolated stenoses
and webs tend to be addressed with balloon angioplasty, or
6 weeks. This suggests that grafts with a normal early scan increasingly, vein patch angioplasty, with more extensive
need much less intensive surveillance. Similarly, in a pros- graft revision being reserved for long lesions and grafts
pective trial of 300 patients undergoing duplex surveillance exhibiting multiple lesions.
after the first year of operation it was concluded that the Where a prosthetic graft is felt to be at risk by virtue of
duration of surveillance may be restricted to the first six falling ABPIs, arteriography is usually required as duplex
months in those who have a normal bypass in that time.52 does not image the surface of the graft sufficiently, particu-
What is not disputed is the increased workload required larly where grafts are externally supported. Furthermore, a
to maintain vein graft patency consequent upon a graft detailed assessment of the inflow and outflow vasculature
surveillance programme, much of which is radiological.53 is needed. Inflow and outflow lesions thus identified can
Despite this, however, revision of a duplex identified steno- be addressed by standard techniques. Where there is a
sis has been shown to be significantly cheaper than revision lesion within the graft, the graft usually requires revision.55
after thrombosis has occurred. Similarly, limb salvage is more Restenosis at the site of a previous percutaneous or vein
cost effective than amputation and, despite concern over the patch angioplasty usually requires resection and interpos-
high frequency with which revisions have been required in ition grafting or jump grafting.
some series, the expense of a duplex surveillance programme,
together with the increased workload generated, appears to
be justified.54 The failed graft – graft thrombosis
It is apparent that much controversy remains surrounding
the use of duplex-based graft surveillance. Nevertheless, the Despite the use of surveillance programmes, a signifi-
TransAtlantic inter-Society Consensus55 (TASC) Working cant number of grafts will not be identified as failing and
Group on the Management of Peripheral Arterial Disease will instead present as thrombosed grafts. A number of
 Management 203

Infrainguinal Per/postoperative quality Entry to graft

Failed graft – graft
graft reconstruction check prior to discharge surveillance
occlusion
Vein/prosthetic (duplex/ABPI) programme

Prosthetic

Ensure best No recurrence of Recurrence of critical ischaemia

medical treatment critical ischaemia or or severe ischaemic symptoms
severe ischaemic
symptoms

Six-week check: Six-week check: Vein

interval history as prosthetic plus
examination including pulses duplex of inflow, graft Conservative Assess severity of ischaemia by
ABPI and outflow as per management SVI/ISCS guidelines
as per TASC guidelines TASC guidelines

Consider
Abnormality: Abnormality: No abnormality Class I/IIa Class IIb/III
amputation
clinical change clinical change
altered pulses altered pulses
ABPI fall 0.2 ABPI fall 0.2
PSV 45 cm/s Continue surveillance at Revision of Thrombolysis Graft Revision of
PSVR 2.0 3, 6, 9, 12, 18, 24 entire graft ? in recent thrombectomy entire graft
duplex-detected months postoperatively occlusion only
Duplex stenosis

Correction of
Return to surveillance underlying lesion:
No abnormality PTA or vein patch
Diagnostic (? with increased frequency
arteriography if only abnormality is PSVR interposition graft
2.0–4.5) jump graft

Continue surveillance:
3, 6, 9, 12, 18, 24 months
postoperatively Graft failing
Return to surveillance

Intervention:
Return to surveillance total graft revision
? increased frequency PTA or vein patch
angioplasty interposition
grafting jump grafting

Figure 17.6 Algorithm illustrating pathways of management. ABPI, ankle:brachial pressure index; PSV, peak systolic velocity; PSVR, PSV
ratio; PTA, percutaneous transluminal angioplasty; SVI/ISCS, Society for Vascular Surgery/International Society of Cardiovascular Surgery;
TASC, TransAtlantic inter-Society Consensus

Table 17.2 The Society for Vascular Surgery/International Most patients presenting with an occluded bypass graft
Society of Cardiovascular Surgery (SVS/ISCVS) grading system will require secondary intervention. A small number of grafts,
for limb ischaemia however, do occlude asymptomatically due to the develop-
ment of collaterals or healing of the tissue necrosis for which
Class I Viable. Not immediately threatened. Arterial Doppler
the original bypass was performed.57 Such patients should
signals audible, no sensory loss and no muscle
not be subjected to further intervention. Furthermore, up to
weakness
a quarter of patients who undergo surgery for critical limb
Class IIa Marginally threatened. Salvageable if promptly
ischaemia will have an effective functioning limb after occlu-
treated. Arterial Doppler signal often inaudible,
minimal sensory loss and no muscle weakness
sion, particularly if some time has elapsed since the original
procedure.
Class IIb Immediately threatened. Salvageable with
Practically, where graft occlusion has occurred, the most
immediate revascularisation, arterial Doppler signal
usually inaudible, sensory loss with rest pain in more vital initial assessment is that of the severity of ischaemia,
than toes, mild to moderate muscle weakness and with it an assessment of the urgency of treatment (see
Class III Irreversible. Major tissue loss or permanent nerve
Chapters 15 and 16). Although there is no substitute for
damage inevitable if there is significant delay before experience in this situation, the Society for Vascular Surgery/
operation. Arterial and venous Doppler signal International Society of Cardiovascular Surgery (SVS/ISCVS)
inaudible, profound limb anaesthesia and paralysis grading system is accepted and is shown in Table 17.2.58
Those patients whose limbs exhibit Class II or worse
considerations dictate the most appropriate management ischaemia and hence require urgent intervention should be
course. These include the nature of the symptoms exhib- anticoagulated with heparin on presentation. The initial
ited by the patient, the severity of the ischaemia, the time treatment options then rest between thrombolysis, usually
from occlusion to presentation, and the time period since with subsequent revision of any underlying lesion, and sur-
the original procedure. gical revision.
204 Graft maintenance and graft failure

Thrombolysis Table 17.3 Guidelines for thrombolysis in the presence of an

occluded lower limb bypass graft
It has been widely held that patients presenting with Class
IIa ischaemia are suitable patients for thrombolysis; there Thrombolysis may be considered as an option in those where
is sufficient time to allow this, and a lesion may be unmasked bypass graft occlusion is a recent event, and where the limb is
not immediately threatened
which can be treated with limited intervention. The role of
thrombolysis in the treatment of graft failure, however, The intravenous route for the administration of high dose
requires closer scrutiny (see Chapter 16). Thrombolysis thrombolytic agents is associated with a high incidence of
has a minimal role in graft occlusion59 as it often unmasks morbidity and mortality and should no longer be used
a lesion requiring definitive treatment and hence merely Where an intra-arterial infusion method is used there is still a
delays an inevitable intervention. significant morbidity and mortality
At least four trials have failed to show any advantage to Adverse events include:
thrombolysis over surgery in the context of either acute
• Intracerebral bleeding (1.2–2.1 per cent in the STILE and
lower limb ischaemia or an occluded bypass graft.60–63
TOPAS trials)
Thrombolysis however, has been shown to be more effect-
• Haematoma formation
ive in terms of mortality and amputation where the • Distal embolisation
occlusion is less than 14 days old.60,63 The Surgery versus • Septicaemia
Thrombolysis for Ischemia of the Lower Extremity • Renal failure
(STILE) trial61 was a multicentre study randomising patients • Pseudoaneurysm formation
with acute lower limb ischaemia to treatment with surgery • Angina
or one of two thrombolytic agents. Both vein and pros- • Ulnar neuropathy
thetic bypass grafts were included. A disadvantage of the Where lysis is successful, adjuvant procedures are usually
study was the use of a large number of clinical parameters required to maintain patency. This may, however, be a lesser
to establish a primary endpoint. Overall there was a highly procedure than would otherwise be required
significant difference in favour of surgery, but subgroup
analysis showed that, where the duration of ischaemia was STILE, Surgery versus Thrombolysis for Ischaemia of the Lower
Extremity;61 TOPAS, Thrombolysis Or Peripheral Arterial Surgery.60
less than 2 weeks, amputation rates were three times higher
for surgery than for thrombolysis; after that period ampu-
tation rates were four times higher in the lytic group. quickly realised however that this was usually unrewarding
STILE established surgery as a more effective and safer without the identification and correction of any underly-
treatment than lytic therapy, and demonstrated that any ing lesion.
benefit associated with thrombolysis is only gained if it is Progression of native disease in either the inflow or out-
used within two weeks of the onset of symptoms. flow vessels is treated by angioplasty or bypass according to
The Thrombolysis Or Peripheral Arterial Surgery standard principles (see Chapter 16). Vein patch angioplasty
(TOPAS) study60 demonstrated similar amputation and was the initial favoured option where vein grafts develop
mortality rates at one year for thrombolysis and surgery. As discrete stenoses, but secondary patency rates below 25 per
with the STILE trial better results were claimed for graft cent were reported65,66 It is now accepted that secondary
occlusion than for native vessel occlusions. Furthermore, autogenous replacement of the graft gives the best results,
success was found to be predictable upon traversing the with secondary patency rates in excess of 50 per cent
occlusion with a guidewire and lysis was more likely to be reported.67,68 Where localised problems are evident, such as
successful in prosthetic rather than vein grafts. a short discrete stenosis or a valve cusp stenosis, open repair
A recent four-centre study has questioned the justification is acceptable and yields better patency rates than dilata-
for thrombolysis of occluded grafts having demonstrated tion.9 Involvement of the distal anastomosis, or extension of
a 20 per cent patency after thrombolysis.64 Interestingly native vessel disease, may require extension of the bypass
an additional procedure such as percutaneous or vein graft. This is especially true where PTFE grafts develop dis-
patch angioplasty appeared to make no difference to the tal anastomotic strictures as these lesions are less amenable
patency rate. Table 17.3 shows guidelines with respect to to balloon dilatation. In treating graft failure, the recommen-
thrombolysis in the presence of an occluded lower limb dation of the TASC group should be considered, namely
bypass graft. that, after the relief of acute limb ischaemia and the estab-
lishment of its aetiology, the choice of treatment of the
underlying lesion should be similar to the treatment choice
Surgical intervention in the setting of chronic limb ischaemia and not linked to
the method of clot removal.55 It has certainly been shown
Open surgical procedures form the basis for the traditional that a pragmatic approach, where the revision strategy
approach to graft occlusion. This tradition started in 1963 favours vein patch angioplasty for graft body lesions and
with the introduction of the Fogarty balloon embolec- jump grafts for distal anastomotic lesions, yields acceptable
tomy catheter and with it simple graft thrombectomy. It was assisted primary patency rates.69
 Postoperative care 205

POSTOPERATIVE CARE permeability and vasodilatation. Growth factor and growth

factor receptor expression is upregulated by hypoxia and
ischaemia, thus allowing a targeted response and limiting
Immediate postoperative care does not differ substantively pathological angiogenesis.
from those patients who are undergoing primary arterial Recent studies have shown that recombinant angiogenic
reconstruction. Nevertheless, two points that have already growth factors augment collateral development in animal
been made are worth reiterating. First, all patients at the time models of hind limb ischaemia.75 It has also been shown that
of operation, and preferably considerably before, should be VEGF can be used in an animal model to successfully modu-
receiving best medical therapy to address smoking, hyper- late the disturbed endothelium dependent blood flow in the
tension, diabetes, hypercholesterolaemia, hyperhomocys- arterial circulation supplied by collateral vessels, thus imply-
teinaemia and cardiac failure and all should be receiving ing a physiological as well as an anatomical advantage.75 In a
antiplatelet medication. Second, prior to discharge, all rabbit model of ischaemic hind limb the administration of
patients should undergo investigation to ensure that the VEGF caused a significant reduction in the haemodynamic
haemodynamic integrity of the affected limb has been deficit of the limb at 7 days and had positive effects upon
restored; ideally all patients should undergo postoperative both smooth muscle cell and endothelial cell proliferation.76
duplex scanning, although in practice ABPI improvement This supports the concept that cellular augmentation con-
may be considered sufficient. tributes to enhanced collateral formation when therapeutic
angiogenesis is employed.
The failed revascularisation Vascular endothelial growth factor is effective following
direct gene transfer. In a rat model, ischaemia was induced
In many cases a failed attempt at graft salvage will lead followed by transfection of the plasmid encoding VEGF
inevitably to amputation (see Chapter 2). There is evidence, into the ischaemic limbs .77 Gene transfer produced a sig-
however, that a small number of patients with critical limb nificantly more extensive, though morphologically similar,
ischaemia following a failed revision will benefit from non- pattern of collaterals to that seen in control limbs. A simi-
surgical treatments. lar effect was seen when plasmid encoding each of the three
Iloprost, an extensively studied stable analogue of prosta- main human VEGF isoforms was delivered on angioplasty
cyclin, has been shown to have beneficial effects on mortality balloons to the ischaemic hind legs of rabbits78 and it thus
and limb salvage at 3–6 months in three trials.70–72 Its effects appears that the transfer of naked DNA encoding a
include prevention of platelet and leucocyte activation and secreted cytokine is an alternative to the administration of
vasodilatation, hence it works at a microcirculatory level. recombinant protein in therapeutic angiogenesis. Striated
The administration regimens are all labour intensive and muscle is capable of taking up and expressing foreign genes
costly, requiring intravenous administration of iloprost for transferred in the form of naked plasmid DNA, albeit with
between 2 and 4 weeks. Patients receiving iloprost, however, low levels of gene expression. The efficacy of this technique
were significantly more likely to be alive and to have avoided when applied to VEGF augmented collateral development
major amputation during the follow-up period than those in the ischaemic limb has been shown79 and intramuscular
having placebo, with a 16 per cent and 20 per cent risk transfection of genes encoding angiogenic factors is an
reduction, respectively. Side effects may limit the efficacy of exciting prospect in those with peripheral vascular disease.
iloprost with flushing, headache, nausea, abdominal pain, Trials utilising this method of transfection have been
hypotension and angina being reported. It does seem rea- performed with very small numbers of patients.80 These
sonable given the current evidence to consider the use of ilo- appear to support the hypothesis that intramuscular injec-
prost where surgery has failed or is not feasible, and the tion of naked plasmid DNA leads to constitutive over-
alternative would be amputation. Trials on a recently devel- expression of VEGF sufficient to induce therapeutic
oped oral form of iloprost are awaited and, if successful, may angiogenesis in selected patients with critical limb
have a profound effect upon the use of this therapy. A num- ischaemia. Phase I trials with intravenous VEGF have been
ber of other agents have undergone trial in the treatment of completed, but the results have not yet been published.
critical ischaemia but are of little or no proven benefit. These Although it is notable that no pharmacological interven-
include heparin, defibrinating agents, L-arginine, pentoxi- tion shown to be successful in the laboratory or in animal
fylline and naftidrofuryl. models has subsequently been shown to be effective in
An exciting new approach to the problem is that of ther- humans, the recent dramatic increase in the number of tri-
apeutic angiogenesis which represents the clinical use of als of tissue angiogenesis reflects the success of the preclin-
growth factors to enhance or promote the development of ical models and the favourable early clinical results.
collateral blood vessels in ischaemic tissue.73 Vascular
endothelial growth factor (VEGF) is an angiogenic growth
factor existing in four isoforms which vary in permeability The successful revascularisation
and heparin binding properties.74 It stimulates the growth,
proliferation and migration of endothelial cells by binding There is little available at present, besides that already men-
to receptors; physiological effects include enhanced vascular tioned, to augment a technically adequate revascularisation
206 Graft maintenance and graft failure

Table 17.4 Published primary, primary assisted and secondary patency rates following infrainguinal bypass surgery

Primary
Follow- Primary assisted Secondary Limb
Number of up patency patency patency salvage
Authors Year grafts Nature of grafts (months) (per cent) (per cent) (per cent) (per cent)

Shah et al.91 1995 2058 In situ vein 60 81 95

120 70 90
92
Byrne et al. 1999 165 Above-knee, popliteal – vein 48 62 64
150 Below-knee, popliteal – vein 48 77 81
94 Tibial – vein 48 86 82 90
93
Eugster et al. 2001 280 Infrainguinal non-reversed 60 63 85
vein
107 Infrainguinal in situ vein 60 58 83
94
Jamsen et al. 2001 263 Femoropopliteal/distal 12 70 63 82
263 Femoropopliteal/distal 60 52 65 77
Ihlberg et al.39 1998 185 Infrainguinal vein – duplex 12 74 71 81
surveillance
Infrainguinal vein – ABPI 12 84 88
surveillance
Belkin et al.95 1996 189 Infrainguinal non-reversed 60 65 74 82
vein
568 Infrainguinal in situ vein 60 72 82 90
ABPI, ankle: brachial pressure index.

operation. Two very different interventions may however arterialised vein and thus improve the patency of vein
offer hope in future for improving graft patency where vein grafts.88
has been used. Finally, there may be a role for gene therapy in the
It has been shown previously that the long term adapta- reduction of intimal hyperplasia in bypass grafts. Vein
tion of vein to the arterial circulation can be modified by grafts are well suited for gene therapy as there is direct access
external stenting. Kohler et al.81 postulated that wall stress to the graft during vein preparation. Recent advances include
regulates wall structure as vein graft thickening stops when adeno-associated and modifed adenovirus gene transfer,
the ratio of lumen radius to wall thickness equals that of a allowing prolonged expression in vivo. Targets for gene trans-
normal artery.82 They found that, in rabbits, restrictive fer have included tissue plasminogen activator, prostacyclin
external support of vein grafts caused reduction of the wall synthase and tissue inhibitors of metalloproteinases.89 One
area. Angelini et al.,83 however, proposed that restrictive example of such work has used an organ culture of human
stenting, while reducing medial thickening, promoted vein to demonstrate reduced intimal hyperplasia following
neointimal thickening, thus leading to a reduction, not adenovirus mediated transfer of nitric oxide synthase.90
increase, in final luminal diameter. This may be due to
reduced fluid flux around the graft, or vessel wall hypoxia
due to adventitial disruption.84 They therefore used a non- RESULTS
restrictive and porous external stent in a pig model of arte-
riovenous bypass grafting to reduce tangential wall stress,
Examples of published primary, primary assisted and second-
and found a dramatic decrease in both medial thickening
ary patency rates following infrainguinal arterial reconstruc-
and neointimal hyperplasia 4 weeks after implantation.
tion are shown in Table 17.4, where available limb salvage
The degree of oversize of the stent appears not to be impor-
rates are also shown.
tant.85 Subsequent work suggested a role for prostacyclin
(PG12) and nitric oxide in promoting microangiogenesis
in the adventitia of stented grafts, which may limit graft
hypoxia.86 Notably, stenting also seemed to result in a sig-
Conclusions
nificant reduction in the expression of platelet derived
The graft failure rate following infrainguinal reconstruc-
growth factor (PDGF) by the graft .87 Most recently, exter-
tion remains high and the consequences potentially ser-
nal stenting in a rat model has been shown to limit intimal
ious both for the patient and financially. Graft revision
hyperplasia and cellular proliferation, with the conclusion
surgery can be technically demanding. Prior to any
that external stenting will reduce atherosclerotic change in
 References 207

9 Nehler MR, Moneta GL, Yeager RA, et al. Surgical treatment

intervention the most meticulous attention must be paid
of threatened reversed infrainguinal vein grafts. J Vasc Surg
to both the overall condition of the patient and to the 1994; 20: 558–63.
degree of ischaemia of the affected limb. Within the con- 10 Robinson KD, Sato DT, Gregory RT, et al. Long-term outcome after
straints of urgency, vigorous attempts to gain as much early infrainguinal graft failure. J Vasc Surg 1997; 26: 425–37.
anatomical and haemodynamic information as possible 11 Bartlett ST, Olinde AJ, Flinn WR, et al. The reoperative potential
must be made. Despite these measures, many revasculari- of infrainguinal bypass: long-term limb and patient survival.
sations will fail. It remains to be seen whether the newer J Vasc Surg 1997; 5: 170–9.
therapies currently in the experimental stage discussed 12 Belkin M, Donaldson MC, Whittemore AD, et al. Observations on
above will lead to improved limb salvage rates, even when the use of thrombolytic agents for thrombotic occlusion of
reintervention has not been successful. infrainguinal vein grafts. J Vasc Surg 1990; 11: 289–94.
13 Budd JS, Brennan J, Beard JD, et al. Infra-inguinal bypass surgery;
Factors determining late graft patency. Br J Surg 1990; 77: 1382–7.
14 Taylor RS, McFarland RJ, Cox MI. An investigation into the causes
of failure of PTFE grafts. Eur J Vasc Surg 1987; 1: 335–43.
Key references 15 Stept LL, Flinn WR, McCarthy WJ, et al. Technical defects as a
cause of early graft failure after femorodistal bypass surgery.
Davies MG, Hagen PO. Pathobiology of intimal hyperplasia. Br J Arch Surg 1987; 122: 599–604.
Surg 1994; 81: 1254–69. 16 Kehler M, Albrechtsson U, Alwmark A, et al. Intra-operative
Ihlberg L, Luther M, Tierala E, Lapantalo M. The utility of duplex digital angiography as a control of the in-situ saphenous vein
scanning in infrainguinal vein graft surveillance: results from bypass graft. Acta Radiol 1988; 29: 645–8.
a randomised controlled study. Eur J Vasc Endovasc Surg 1998; 17 Robicsek F, Thubrikar MJ, Fokin A, et al. Pressure traps in femoro-
16: 19–27. popliteal reversed vein grafts. Are valves culprits?
Ouriel K, Veith FJ, Sasahara AA. Thrombolysis or peripheral arterial J Cardiovasc Surg (Torino) 1999; 40: 683–9.
surgery (TOPAS): Phase I results. TOPAS Investigators. J Vasc 18 Gannon MX, Goldman MD, Simms MH, et al. Peri-operative
Surg 1996; 23: 64–75. complications of in-situ vein bypass. Ann R Coll Surg Engl
Results of a prospective randomised trial evaluating surgery versus 1987; 74: 252–5.
thrombolysis for ischemia of the lower extremity. The STILE 19 Szilagyi DE, Hageman JH, Smith RF, et al. Autogenous vein
trial. Ann Surg 1994; 220: 251–68. grafting in femoropopliteal atherosclerosis: the limit of its
Szilagyi DE, Elliot JP, Hageman JH, et al. Biologic fate of effectiveness. Surgery 1979; 86: 836–51.
autogenous vein implants as arterial substitutes – clinical, 20 Davies MG, Hagen PO. Pathobiology of intimal hyperplasia. Br
angiographic and histologic observations in femoro-popliteal J Surg 1994; 81: 1254–69.
operations for atherosclerosis. Ann Surg 1973; 178: 232–46. 21 Currie IC, Wilson YG, Davies AH, et al. Pulse generated run-off
versus dependent Doppler ultrasonography for assessment of calf
vessel patency. Br J Surg 1994; 81: 1448–50.
22 Kupinsky AM, Evans JM, Khan AM, et al. Ultrasonic characteristics
REFERENCES
of the saphenous vein. Cardiovasc Surg 1993; 1: 513–17.
23 Katz ML, Pilla TS, Comerota AJ. Technical aspects of venous
1 Szilagyi DE, Elliot JP, Hageman JH, et al. Biologic fate of duplex imaging. J Vasc Technol 1988; 12: 100–2.
autogenous vein implants as arterial substitutes – clinical, 24 Seeger JM, Schmidt JH, Flynn TC. Preoperative saphenous and
angiographic and histologic observations in femoro-popliteal cephalic vein mapping as an adjunct to reconstructive arterial
operations for atherosclerosis. Ann Surg 1973; 178: 232–46. surgery. Ann Surg 1987; 205: 733–9.
2 Mamode N, Scott RN. Graft type for femoro-popliteal bypass 25 Leopold PW, Shandall AA, Corson JD, et al. Initial experience
surgery. Cochrane Database Syst Rev 2000; 2: CD001487. comparing B-mode imaging and venography of the saphenous
3 Englund R, Harris J, May J. Factors affecting the outcome of 463 vein before in-situ bypass. Am J Surg 1986; 152: 206–10.
femorotibial reconstructions. Aust N Z J Surg 1988; 58: 279–83. 26 Panetta TF, Marin ML, Veith FJ, et al. Unsuspected pre-existing
4 Luther M, Lepantalo M. Femorotibial reconstructions for chronic saphenous vein disease: an unrecognised cause of vein bypass
critical leg ischaemia: influence on outcome by diabetes, gender failure. J Vasc Surg 1992; 15: 102–12.
and age. Eur J Vasc Endovasc Surg 1997; 13: 569–77. 27 Davies AH. Vein factors that affect the outcome of femorodistal
5 Idu MM, Buth J, Hop WC, et al. Factors influencing the bypass. Ann R Coll Surg Engl 1995; 77: 63–66.
development of vein graft stenosis and their significance for 28 Beattie DK, Sian M, Greenhalgh RM, Davies AH. Influence of
clinical management. Eur J Vasc Endovasc Surg 1999; 17: 15–21. systemic factors on pre-existing intimal hyperplasia and their
6 Brewster DC, Lasalle AJ, Robinson JG, et al. Femoropoliteal effect on the outcome of infrainguinal arterial reconstruction
graft failures: Clinical consequences of success of secondary with vein. Br J Surg 1999; 86: 1441–7.
reconstructions. Arch Surg 1983; 118: 1043–7. 29 Davies AH, Magee TR, Baird RN, et al. Vein compliance: a pre-
7 Avino AJ, Bandyk DF, Gonsalves AJ, et al. Surgical and operative indicator of vein morphology and of veins at risk
endovascular intervention for infrainguinal vein graft stenosis. of vascular graft stenosis. Br J Surg 1992; 79: 1019–21.
J Vasc Surg 1999; 29: 60–70. 30 Davies AH, Magee TR, Sheffield E, et al. The aetiology of vein
8 Berkowitz HD, Fox AD, Deaton DH. Reversed vein graft stenosis: graft stenoses. Eur J Vasc Surg 1994; 8: 389–94.
early diagnosis and management. J Vasc Surg 1992; 15: 31 Golledge J, Beattie DK, Greenhalgh RM, Davies AH. Have the
130–42. results of infrainguinal bypass improved with the widespread
208 Graft maintenance and graft failure

utilisation of postoperative surveillance? Eur J Vasc Endovasc 50 McCarthy MJ, Olojugba D, Loftus IM, et al. Lower limb
Surg 1996; 11: 388–92. surveillance following autologous vein bypass should be
32 Barnes RW, Thompson BW, MacDonald CM, et al. Serial non- life long. Br J Surg 1998; 85: 1369–72.
invasive studies do not herald post-operative failure of 51 Mills JL, Bandyk DF, Gahtan V, Esses G. The origin of infrainguinal
femoropopliteal or femorotibial bypass grafts. Ann Surg 1989; vein graft stenosis: a prospective randomised trial based on
210: 486–93. duplex surveillance. J Vasc Surg 1995; 21: 16–25.
33 Laborde AL, Synn AY, Worsey MJ, et al. A prospective comparison 52 Idu MM, Buth J, Cuypers P, et al. Economising vein-graft sur-
of ankle/brachial indices and colour duplex imaging in veillance programmes. Eur J Vasc Endovasc Surg 1998; 15: 432–8.
surveillance of the in situ saphenous vein bypass. J Cardiovasc 53 Loftus IM, Reid A, Thompson MM, et al. The increasing workload
Surg 1992; 33: 420–5. required to maintain infra inguinal graft patency. Eur J Vasc
34 Tong Y, Somjen G, Teeuwsen W, Royle JP. Percutaneous Endovasc Surg 1998; 15: 37–41.
transluminal angioplasty: follow-up with treadmill exercise 54 Wixon CL, Mills JL, Westerband A, et al. An economic appraisal of
testing. Cardiovasc Surg 1994: 2: 503–7. lower extremity bypass graft maintenance. J Vasc Surg
35 Wilson YG, Davies AH, Currie IC, et al. Vein graft stenosis; 2000; 32: 1–12.
incidence and intervention. Eur J Vasc Endovasc Surg 1996; 55 TASC Working Group. Management of peripheral arterial disease.
11: 164–9. TransAtlantic Inter-Society Consensus. Eur J Vasc Endovasc Surg
36 Moody P, DeCossart LM, Douglas HM, Harris PL. Asymptomatic 2000; 19(suppl A): S217–218.
strictures in femoropopliteal vein grafts. Eur J Vasc Surg 1989; 3: 56 Tonnesen KH, Holstein P, Rordam L, et al. Early results of
389–92. percutaneous transluminal angioplasty of failing below-knee
37 Mattos MA, Van Bremmelen PS, Hodgson KJ, et al. Does bypass grafts. Eur J Vasc Endovasc Surg 1998; 15: 51–6.
correction of stenoses identified with colour duplex scanning 57 Brewster DC, Lasalle AJ, Robinson JG, et al. Femoropoliteal
improve infrainguinal graft patency? J Vasc Surg 1993; 17: graft failures: clinical consequences of success of secondary
54–66. reconstructions. Arch Surg 1983; 118: 1043–7.
38 Ihlberg L, Luther M, Alback A, et al. Does a completely 58 Rutherford RB, Baker JD, Ernst C, et al. Recommended standards
accomplished duplex-based surveillance programme prevent for reports dealing with lower extremity ischaemia: revised
vein-graft failure? Eur J Vasc Endovasc Surg 1999; 18: 395–400. version. J Vasc Surg 1997; 26: 517–38.
39 Ihlberg L, Luther M, Tierala E, Lapantalo M. The utility of duplex 59 Lacroix H, Suy R, Nevelsteen A, et al. Local thrombolysis for
scanning in infrainguinal vein graft surveillance: results from occluded arterial grafts: is the yield worth the effort?
a randomised controlled study. Eur J Vasc Endovasc Surg 1998; J Cardiovasc Surg 1994; 35: 187–91.
16: 19–27. 60 Ouriel K, Veith FJ, Sasahara AA. Thrombolysis or peripheral
40 Dougherty MJ, Calligaro KD, DeLaurentis DA. Revision of arterial surgery (TOPAS): Phase I results. TOPAS Investigators.
failing lower extremity bypass grafts. Am J Surg 1998; 176: J Vasc Surg 1996; 23: 64–75.
126–30. 61 Results of a prospective randomised trial evaluating surgery
41 Dunlop P, Sayers RD, Naylor AR, et al. The effect of a surveillance versus thrombolysis for ischemia of the lower extremity. The
programme on the patency of synthetic infrainguinal bypass STILE trial. Ann Surg 1994; 220: 251–68.
grafts. Eur J Vasc Endovasc Surg 1996; 11: 441–5. 62 Ouriel K, Shortell CK, DeWeese JA, et al. A comparison of
42 Kirby PL, Brady AR, Thompson SG, et al. The vein graft thrombolytic therapy with operative revascularisation in
surveillance trial: rationale, design and methods. VGST the initial treatment of acute peripheral arterial ischaemia.
participants. Eur J Vasc Endovasc Surg 1999; 18: 469–74. J Vasc Surg 1994; 19: 1021–30.
43 Buth J, Disselhoff B, Sommeling C, Stam L. Color-flow duplex 63 Comerota AJ, Weaver FA, Hosking JD, et al. Results of a
criteria for grading stenosis in infrainguinal vein grafts. J Vasc prospective randomised trial of surgery versus thrombolysis
Surg 1991; 14: 716–28. for occluded lower extremity bypass grafts. Am J Surg 1996;
44 Ihlberg L, Alback A, Roth WD, et al. Interobserver agreement 172: 105–12.
in duplex scanning for vein grafts. Eur J Vasc Endovasc Surg 64 Galland RB, Magee TR, Whitman B, et al. Patency following
2000; 19: 504–8. successful thrombolysis of occluded vascular grafts. Eur J Vasc
45 Olojugba DH, McCarthy MJ, Naylor AR, et al. At what peak Endovasc Surg 2001; 22: 157–60.
velocity ratio should duplex-detected vein graft stenoses be 65 Whittemore AD, Clowes AW, Couch NP, Mannick JA. Secondary
revised? Eur J Vasc Endovasc Surg 1998; 15: 258–60. femoropopliteal reconstruction. Ann Surg 1981; 193: 35–42.
46 Dougherty MJ, Calligaro KD, DeLaurentis DA. The natural history 66 Cohen JR, Mannick JA, Couch NP, Whittemore AD. Recognition
of ‘failing’ arterial bypass grafts in a duplex surveillance protocol. and management of impending vein graft failure. Ann Surg
Ann Vasc Surg 1998; 12: 255–9. 1986; 121: 758–9.
47 Dougherty MJ, Calligaro KD, DeLaurentis DA. Revision of 67 DeWeese JA, Leather R, Porter J. Practical guidelines: lower
failing lower extremity bypass grafts. Am J Surg 1998; 176: extremity revascularisation. J Vasc Surg 1993; 18: 280–94.
126–30. 68 Belkin M, Conte MS, Donaldson MC. Preferred strategies for
48 Treiman GS, Lawrence PF, Bhirangi K, Gazak CE. Effect of outflow secondary infrainguinal bypass: lessons learned from 300
level and maximum graft diameter on the velocity parameters of consecutive operations. J Vasc Surg 1995; 21: 282–95.
reversed vein bypass grafts. J Vasc Surg 1999; 30: 16–25. 69 Sullivan TR, Welch HJ, Iafrati MD, et al. Clinical results of
49 Lundell A, Lindblad B, Bergqvist D, Hansen F. Femoropopliteal- common strategies used to revise infrainguinal vein grafts. J Vasc
crural graft patency is improved by an intensive graft Surg 1996; 24: 909–17.
surveillance programme: a prospective randomised study. 70 Norgren L, Alwmark A, Angqvist KA, et al. A stable prostacyclin
J Vasc Surg 1995; 21: 26–34. analogue (iloprost) in the treatment of ischaemic ulcers of the
 References 209

lower limb. A Scandinavian-Polish placebo controlled, 83 Angelini GD, Izzat MB, Bryan AJ, Newby AC. External stenting
randomised multicentre study. Eur J Vasc Surg 1990; 4: 463–7. reduces early medial and neointimal thickening in a pig model of
71 UK Severe Limb ischaemia Study Group. Treatment of limb arteriovenous bypass grafting. J Thorac Cardiovasc Surg 1996;
threatening ischaemia with intravenous iloprost: a randomised 112: 79–84.
double blind placebo controlled study. Eur J Vasc Surg 1991; 84 Barker SGE, Talbert A, Cottam S, et al. Arterial intimal
5: 511–16. hyperplasia after occlusion of the adventitial vasa vasorum in
72 Guilmot JL, Diot E. Treatment of lower limb ischaemia due to the pig. Arterioscler Thromb 1993; 13: 70–7.
atherosclerosis in diabetic and non -diabetic patients with iloprost, 85 Izzat MB, Mehta D, Bryan AJ, et al. Influence of external stent
a stable analogue of prostacyclin. Drug Invest 1991; 3: 351–9. size on early medial and neointimal thickening in a pig model
73 Henry TD. Therapeutic angiogenesis. BMJ 1999; 318: 1536–9. of saphenous vein bypass grafting. Circulation 1996; 94:
74 Ferrara N, Davis-Smyth T. The biology of vascular endothelial 1741–5.
growth factor. Endocri Rev 1997; 18: 1–22. 86 Jeremy JY, Dashwood MR, Mehta D, et al. Nitric oxide,
75 Bauters C, Asahara T, Zheng LP, et al. Recovery of disturbed prostacyclin and cyclic nucleotide formation in externally
endothelium-dependent flow in the collateral-perfused rabbit stented porcine vein grafts. Atherosclerosis 1998; 141: 297–305.
ischemic hindlimb after administration of vascular endothelial 87 Mehta D, George SJ, Jeremy JY, et al. External stenting reduces
growth factor. Circulation 1995; 91: 2802–9. long-term medial and neointimal thickening and platelet derived
76 Takeshita S, Rossow ST, Kearney M, et al. Time course on growth factor expression in a pig model of arteriovenous bypass
increased cellular proliferation in collateral arteries after grafting. Nat Med 1998; 4: 235–9.
administration of vascular endothelial growth factor in a 88 Meguro T, Nakashima H, Kawada S, et al. Effects of external
rabbit model of lower limb vascular insufficiency. Am J Pathol stenting and systemic hypertension on intimal hyperplasia in
1995; 147: 1649–60. vein grafts. Neurosurgery 2000; 46: 963–9.
77 Takeshita S, Isshiki T, Mori H, et al. Microangiographic 89 Newby AC, Baker AH. Targets for gene therapy of vein grafts.
assessment of collateral vessel formation following direct Curr Opin Cardiol 1999; 14: 489–94.
gene transfer of vascular endothelial growth factor in rats. 90 Cable DG, Caccitolo JA, Caplice N, et al. The role of gene therapy
Cardiovasc Res 1997; 35: 547–52. for intimal hyperplasia of bypass grafts. Circulation 1999;
78 Takeshita S, Tsurumi Y, Couffinahl T, et al. Gene transfer of naked 100(19 suppl): II392–6.
DNA encoding for three isoforms of vascular endothelial growth 91 Shah DM, Darling RC, Chang BB, et al. Long term results of
factor stimulates collateral development in vivo. Lab Invest 1996; in situ saphenous vein bypass. Analysis of 2058 cases. Ann Surg
75: 487–501. 1995; 222: 438–46.
79 Tsurumi Y, Takeshita S, Chen D, et al. Direct intramuscular 92 Byrne J, Darling RC, Chang BB, et al. Vascular surgical society of
gene transfer of naked DNA encoding vascular endothelial Great Britain and Ireland: review of 94 tibial bypasses for
growth factor augments collateral development and tissue intermittent claudication. Br J Surg 1999; 86: 706–7.
perfusion. Circulation 1996; 94: 3281–90. 93 Eugster T, Stierli P, Aeberhard P. Infrainguinal arterial
80 Baumgartner I, Pieczek A, Manor O, et al. Constitutive expression reconstruction with autogenous vein grafts: are the results
of phVEGF165 after intramuscular gene transfer promotes for the in situ technique better than those of non-reversed
collateral vessel development in patients with bypass? A long-term follow-up study. J Cardiovasc Surg
critical limb ischaemia. Circulation 1998; 97: 1114–23. (Torino) 2001; 42: 221–6.
81 Kohler TR, Kirkman TR, Clowes AW. The effect of rigid external 94 Jamsen T, Tulla H, Manninen H, et al. Results of infrainguinal
support on vein graft adaptation to the arterial circulation. bypass surgery: an analysis of 263 consecutive operations.
J Vasc Surg 1989; 9: 277–85. Ann Chir Gynaecol 2001; 90: 92–9.
82 Zwolak RM, Adams MC, Clowes AW. Kinetics of vein graft 95 Belkin M, Knox J, Donaldson MC, et al. Infrainguinal arterial
hyperplasia: association with tangential stress. J Vasc Surg 1987; reconstruction with nonreversed greater saphenous vein. J Vasc
5: 126–36. Surg 1996; 24: 957–62.
This page intentionally left blank
 18
Acute Ischaemia Secondary to Occult Prosthetic
Graft Infection

LINDA M REILLY

The problem 211 Management 212

Epidemiology 211 Results 218
Diagnosis 211 References 219
Bacteriology 212

THE PROBLEM clinical evidence to suggest an infection, the vascular sur-

geon usually discovers the infection incidentally during treat-
ment focused on relieving ischaemia. The challenge presented
The principles of treatment of prosthetic graft infection are to the vascular surgeon, therefore, is the management of
straightforward: eliminate the infection while preserving occult graft infection discovered during the treatment of
perfusion to the vascular bed supplied by the infected graft. graft or graft limb occlusion.
Two circumstances, however, pose significant challenges
to successful treatment, namely, associated bleeding and
associated ischaemia. In each case the urgency of the asso-
ciated condition imposes an immediacy which may limit DIAGNOSIS
the options for evaluation, preparation and treatment, as
well as have an adverse impact on outcome.
As mentioned, these patients usually lack the clinical signs
and symptoms associated with vascular graft infection,
namely, fever, malaise, persistent/recurrent wound fluid
EPIDEMIOLOGY
collection, cellulitis, drainage, chronic sinus tract or septi-
caemia.6 The detected abnormalities, such as elevated white
While prosthetic vascular graft infection occurs in 2–4 blood cell count and elevated sedimentation rate, are mild
per cent of patients1–3 it is an uncommon cause of acute and quite reasonably attributed to the physiological effects
limb ischaemia. In our experience and that of others, less of the ischaemia. The suspicion of a possible graft infection
than 10 per cent of patients with graft infection present is first aroused by the intraoperative findings at explo-
with graft or graft limb occlusion.4,5 While the incidence of ration to treat the ischaemia. Typically, these include a lack
graft infection among all patients presenting with graft or of graft incorporation by the surrounding tissue, the pres-
graft limb occlusion is unknown, clearly most of these ence of perigraft fluid, especially if within the perigraft cap-
occlusions result from haemodynamic causes and not from sule and white blood cells on Gram stain of that fluid. It is
underlying graft infection. rare to see organisms on Gram stain of the fluid, but if
In most cases of graft occlusion and associated infection found they are pathognomonic of graft infection. The stand-
of the prosthesis, graft infection is occult and the ischaemia ard approaches to the diagnosis of graft infection using
is not limb threatening. It is less likely for the infection to imaging studies, nuclear medicine scans, aspiration of peri-
be evident or for the ischaemia to limb threatening. As the graft fluid or sinus tract injection will not have been applied
usual clinical setting is one of graft occlusion without any in the setting of an occult graft infection.
212 Acute ischaemia secondary to occult prosthetic graft infection

BACTERIOLOGY spectrum of activity, which at the least, includes skin

organisms. If tissue ischaemia is extensive, for example
affecting both lower extremities or if it is advanced, i.e.
The absence of organisms on Gram stain and the common present for some time, and especially with associated neuro-
finding of no growth on subsequent cultures is consistent muscular deficit, then renal protection against myoglo-
with the low virulence of the most common causative organ- bin damage should be initiated by means of hydration and
ism, Staphylococcus epidermidis.7 Central to the pathogenesis alkalinisation of the urine with a continuous infusion of
of this organism is the production of a polysaccharide bicarbonate. With extensive or advanced ischaemia the
biofilm, commonly referred to as ‘slime’, which allows the S. patient will also need volume resuscitation to correct the
epidermidis organisms to adhere to the surface and the inter- considerable sequestration of fluid accompanying this
stices of the graft material in a protected environment which event. Renal protection is particularly important because
white cells, antibodies and antibiotics have difficulty pene- administration of intravenous contrast will almost cer-
trating effectively. The resultant tissue destruction is rela- tainly be necessary at some point in evaluation and man-
tively confined to the perigraft area and produces a muted agement of the patient. Electrolyte abnormalities should
host response. It is this limited pathogenicity which allows also be corrected. These basic elements of treatment must
the graft infection to remain occult, pursuing a chronic, be accomplished expeditiously and should not delay rapid
indolent course over many months or years. Although there intervention to correct the ischaemia.
are no specific data regarding the bacteriology of occult graft
infections presenting with graft or graft limb occlusion, in
comparison to those in other manifestations of graft infec- Initial management
tion, infection is less likely to be occult when it results from
more virulent organisms such as Staphylococcus aureus and • Anticoagulant treatment
Gram-negative organisms. The latter bacteria are more • Antibiotic therapy
likely to cause significant tissue destruction resulting in a • Renal protection
more obvious host response, manifesting itself as cellulitis, – hydration
abscess, anastomotic disruption with either false aneurysm – alkalinisation of the urine
formation or actual bleeding. – against intravenous contrast
• Volume resuscitation
• Correction of electrolyte abnormalities
MANAGEMENT • Intervention to correct ischaemia

In general, the management of a patient with a graft infec-

tion requires evidence of that infection, definition of rele- Diagnosis
vant arterial anatomy, determination of the extent of 1 Computed tomography (CT)
2 Nuclear medicine scan
involvement of the graft by infection, and then selection of 3 Image-guided perigraft fluid aspiration
the best approach to treatment (Fig. 18.1). In the specific 4 Sinus tract injection
clinical situation where the patient presents with graft
thrombosis, however, that approach will be determined by
Define arterial anatomy
the severity of tissue ischaemia, the level of suspicion, if
1 CT angiography
any, of an underlying graft infection, the infecting organ- 2 Magnetic resonance angiography
ism, once it has been identified, and the patient’s associ- 3 Conventional arteriography
ated comorbidities.
Establish extent of graft involvement
Factors determining approach 1 CT
2 Nuclear medicine scan
3 Sinus tract injection
• Severity of tissue ischaemia 4 Operative exploration
• Level of suspicion of infection
• Infecting organism
• Associated comorbidities Select definitive treatment

Excision Revascularisation Conduit

As the presenting feature in these patients is ischaemia,
1 Partial 1 Extra-anatomical 1 Prosthetic
anticoagulation should be initiated promptly. In addition, 2 Complete 2 In situ 2 Autograft
antibiotics should be administered, not because of any sus- 3 Allograft
picion of underlying graft infection, but in anticipation of
invasive intervention. The antibiotic selected should have a Figure 18.1 Algorithm of management of graft infection
 Management 213

Graft occlusion with occult infection the setting of advanced ischemia because of the associated
delay in achieving reperfusion.
In the most common clinical setting, namely, graft occlusion A particular risk of thrombolysis in this setting is the
with no clinical evidence or suspicion of graft infection, the limited opportunity to diagnose graft infection. Occasionally,
vascular surgeon will be concentrating on treating the perigraft fluid will be detected during puncture of the graft to
ischaemia. Therefore, the treatment chosen will be deter- establish access for thrombolysis. If that does not occur,
mined by the severity of the ischaemia (Fig. 18.2). If the however, and if lysis is successful in restoring perfusion, and
patient has mild or moderate ischaemia, with no associated the lesion(s) responsible for the graft occlusion is(are) then
neuromuscular deficit, or at most a very mild sensory deficit, managed endoluminally, it is likely that the graft infection
then either thrombolysis or thrombectomy are appropriate will go undiagnosed. This is of particular concern if endo-
treatment options. The vascular anatomy can be defined by luminal treatment involves placing stents or stent-grafts
contrast arteriography performed just prior to initiating through a graft in which infection is present but has not been
thrombolysis or performing thrombectomy, although com- recognised. The potential for contaminating the new endo-
plete definition of the anatomy may not be possible in the prosthesis and for spreading the infection to other previously
very poorly perfused ischaemic bed. Thrombolysis, if effect- uninvolved areas of the graft or arterial tree is an alarming
ive, has the potential benefit of averting an open procedure prospect. One approach, which would prevent such a missed
especially if extensive or even multiple procedures might diagnosis and would have no impact on the risk of spreading
have been necessary. Some consideration, however, should infection, is to define the arterial anatomy at the conclusion
be given to the possibility that thrombolysis may increase of successful thrombolysis and endoluminal treatment using
the embolic potential of a pseudointima already destabilised fine-cut computed tomography (CT) with three-dimensional
by graft infection.6 Embolisation of contents from an arterial reconstruction. This will demonstrate the perigraft
infected, occluded graft may create extensive soft tissue soft tissue and detect any of the typical abnormalities associ-
infection and destruction leading to limb loss. In those even ated with graft infection, i.e. perigraft fluid, air, indistinct
rarer circumstances when the graft infection has presented tissue planes, abscess or false aneurysm. Artefacts from endo-
with septic emboli at the time of occlusion, thrombolysis is luminal prostheses, however, can make interpretation of the
contraindicated. Thrombolysis is obviously not advisable in study difficult. If thrombolysis is unsuccessful or if treatment

Appropriate for mild to moderate

ischaemia without neuromuscular deficit
Avoids additional open procedure
Thrombolysis Concern about potential for septic embolisation
May miss diagnosis of graft infection Advanced ischaemia
Thrombectomy Thrombolysis not feasible

Successful Unsuccessful
Dependent upon availability
Avoids additional open procedure Most likely approach
Concern about potential for Confine to body of graft
Mechanical graft or anastomotic Operative Avoid anastomosis if possible
disruption May precipitate definitive repair

Successful Unsuccessful Successful Unsuccessful

Define arterial anatomy Confine graft revision to the lesion

responsible for failure of simple Graft revision
(if not already done) thrombectomy
Do not perform remote stenting or
angioplasty through the infected graft
Do not place prosthetic extensions
of the infected graft

Establish extent of graft Successful Unsuccessful

involvement by infection
If simple thrombectomy and limited graft revision are
unsuccessful, must be prepared to proceed with all or
part of the definitive repair, depending on graft anatomy (see text).
May require:
Select and perform Intraoperative definition of arterial anatomy
Availability of alternative tissue conduits
definitive treatment Reprep and redrape to begin work in non-infected fields

Figure 18.2 Algorithm of management of graft occlusion with occult infection

214 Acute ischaemia secondary to occult prosthetic graft infection

of the causative lesion requires open operation, then diagno- deployment site is remote from the infected graft. Local
sis of the occult graft infection will not be missed. anastomotic revision, for example, patch angioplasty to treat
Thrombectomy is the treatment of choice in the setting an anastomotic stenosis, should be limited to the shortest
of advanced or extensive ischaemia, or if thrombolysis is patch that will correct the stenosis. Prosthetic extensions of
not safe or technically feasible. Thrombectomy may be per- the infected graft such as proximal or distal jump-grafts are
formed using the percutaneous mechanical approach (see to be avoided, as they will certainly complicate the subse-
Chapter 16) or the standard open approach (see Chapter quent definitive treatment of the graft infection. In short, if
15). Again, the percutaneous approach has the same advan- graft revision is needed to ensure successful perfusion and
tages and disadvantages as thrombolysis, except that it can reversal of ischaemia, it should be confined to the minimum
provide more rapid restoration of flow. Given that graft that is needed and should be completed in a manner that
infection can result in structural weakening of the graft does not make the subsequent definitive treatment of the
and its attachments, mechanical thrombectomy carries the graft infection any more complicated.
potential for disrupting the conduit or the anastomoses, Occasionally, open thrombectomy will have the effect
and that must be kept in mind. Finally, the use of mechan- of precipitating immediate definitive treatment of the graft
ical thrombectomy is limited by the availability of the neces- infection. The most common reasons for this are inability
sary devices. to re-establish adequate perfusion through the infected
Open thrombectomy is the most common approach in graft and/or bleeding as a result of disruption of an anasto-
patients with extensive or advanced ischaemia, and will mosis or the conduit itself during attempts at thrombec-
also be needed for a proportion of those in whom throm- tomy or focal revision. The specific approach to the graft
bolysis fails. When an occult graft infection is encountered infection is individualised in each case but some examples
at the time of graft exploration for thrombectomy several will illustrate general principles.
principles should be followed. An adequate sample of any
perigraft fluid should be sent for immediate Gram staining
as well as for culture and sensitivity. The goal of this oper- Situations governing specific approaches to
ation should be to restore perfusion, allowing adequate time infection
for appropriate assessment and planning for definitive
treatment of the graft infection. Accordingly, the proce- • Occult infection of aortofemoral bypass graft limb
dure should begin by transgraft thrombectomy avoiding • Occult infection of femoropopliteal bypass graft
exposure of any anastomosis. To do so increases the risk of • Graft occlusion with suspected infection
disruption of an anastomosis already weakened by infec-
tion. Anastomotic disruption may precipitate immediate
definitive treatment of the graft infection, which is most
Occult infection of aortofemoral graft limb
undesirable at this point because the extent of graft involve-
ment is not known and the pertinent arterial anatomy may
INADEQUATE OUTFLOW
not be completely defined. The lack of such information
may result in either an unnecessarily extensive or an inap- If the patient presents with occlusion of one limb of an
propriately limited procedure, either of which can adversely aortobifemoral bypass graft (AFBG) and occult infection
affect outcome. Adequate drainage of the infection should is detected at open thrombectomy, the first step must be
be established, but extensive debridement of the soft tissue transgraft thrombectomy. If inflow is restored but there is a
involved should not be performed at this time. If thrombec- focal stenosis at the femoral anastomosis, the appropriate
tomy is successful in restoring flow, the procedure is then options would include prosthetic patch angioplasty or focal
terminated. The patient should then undergo expeditious common femoral artery (CFA) endarterectomy (Fig. 18.3).
definition of the arterial anatomy and determination of the The prosthetic patch angioplasty needs to be short and should
extent to which the graft is involved in the infection. With not be extended for any distance onto either the superficial
this information the surgeon can select an appropriate femoral artery (SFA) or the profunda femoris artery (PFA).
approach to definitive treatment. Common femoral endarterectomy is only feasible if there is
If simple transgraft thrombectomy is not successful in no suggestion of anastomotic weakening. Inappropriate
restoring perfusion to the ischaemic bed, graft revision will options include prosthetic femoropopliteal bypass or a
be required. Careful consideration must be given to the prosthetic extension graft from the AFBG limb to the SFA
technical aspects of this revision. Endoluminal angioplasty or PFA. If the SFA is occluded and the PFA does not provide
of stenoses limiting inflow or outflow can be utilised pro- sufficient outflow to maintain patency of the AFBG limb, a
vided that access is through a non-infected site and does not femoropopliteal bypass may be constructed using vein
involve traversing the infected graft. Placement of stents or conduit only, with the graft originating from the PFA.
stent-grafts to treat any flow-limiting, graft-threatening Thereafter, subsequent resection of the infected AFBG limb
stenosis is not advisable in the setting of an infection, even if and reconstruction, either extra-anatomical or in-line,
access is through a non-infected site and the planned can be undertaken without having to revise the proximal
 Management 215

Inadequate inflow Inadequate outflow

Close incision
reprep and redrape
Anastomotic Occluded SFA Diseased PFA
stenosis

Expose opposite
AFBG limb Appropriate Appropriate Appropriate
• Short patch angioplasty – do Femoropopliteal bypass • Detach AFBG limb
NOT extend onto SFA or PFA • Use vein only • Open PFA TEA
• Perianastomotic • Originate from PFA • Autogenous patch
endarterectomy – if no • Reimplant AFBG limb
anastomotic deterioration Inappropriate
• Endoluminal treatment Inappropriate
Infected Not infected
Inappropriate through infected access • Prosthetic graft
• Prosthetic femoropopliteal • Stent or stent-graft extensions
bypass insertion
• Prosthetic graft extension to • Prosthetic femoro
Anastomose Anastomose popliteal bypass
the SFA or PFA
conduit to conduit to
PFA or SFA AFBG limb Determine extent of graft
involved by infection

Define arterial anatomy

Complete half of suprapubic tunnel Detach infected AFBG limb, (if not already done)
Bring conduit out onto skin through resect segment and oversew
small mid-tunnel incision Anastomose vein conduit

Resect remaining graft

Reopen original incision involved by infection
Close and seal incision
Complete tunnel and reconstruct

Figure 18.3 Algorithm of management of occult infection of aortobifemoral graft limb. AFBG, aortobifemoral bypass graft; PFA, profundus
femoris artery; SFA, superficial femoral artery; TEA, thromboendarterectomy

anastomosis of the newly placed femoropopliteal graft. If not use another prosthetic conduit in this situation and
the only outflow is the PFA and if that vessel has an extended prefer either the patient’s superficial femoral vein har-
lesion, then the appropriate technique is to detach the vested from the non-infected side, a cryopreserved superfi-
AFBG limb, extend the open arteriotomy down the PFA, cial femoral vein or a cryopreserved arterial segment of
perform an open endarterectomy, patch the arteriotomy appropriate size. The chosen conduit is anastomosed to the
with either vein or an endarterectomised segment of the uninfected AFBG limb, tunnelled halfway across a supra-
occluded SFA and then re-anastomose the AFBG limb into pubic tunnel and, maintaining orientation within it, brought
the patch or into the CFA just proximal to the patch. out through a short incision in the skin. The incision used
to expose the non-infected AFBG limb and to perform that
anastomosis is closed and sealed.
INADEQUATE INFLOW
The original incision used to expose the infected graft limb
If inflow is not restored by transgraft thrombectomy, then is now re-opened and the vein conduit tunnelled the remain-
an alternative inflow source is needed (see Fig. 18.3). It will der of the way across the suprapubic region. The infected
be necessary to close the incision used to explore the AFBG limb is now detached from the femoral artery and the
occluded, infected graft limb and begin again with a clean vein conduit is anastomosed to the CFA. If necessary, the
field and clean instruments. An incision is made exposing anastomosis of this vein conduit may be brought well down
the contralateral patent, and presumably non-infected, onto the PFA or SFA to treat any femoral anastomotic sten-
AFBG limb. If that graft limb shows no evidence of infec- osis. The ample size of the conduit facilitates a lengthy anasto-
tion, the perigraft space is swabbed and a Gram stain per- mosis if it is needed. The infected AFBG limb is resected back
formed. If no white cells are seen on Gram stain, the AFBG to a point just proximal to the inguinal ligament and over-
limb can be used for inflow to the ischaemic limb. We do sewn. After this procedure has successfully restored flow to
216 Acute ischaemia secondary to occult prosthetic graft infection

the ischaemic limb, appropriate investigations can be under- performed through the graft, without risk of spreading
taken to determine how much of the AFGB is involved by the infection because the graft is already involved. This may not
infection. A second procedure will be needed to remove the be an appropriate intervention, however, if the graft infec-
remaining infected segments of the graft. tion seems localised, for example, confined to a segment of
If there is any concern that the contralateral AFBG limb is the body of the graft. Alternatively, a short focal patch angio-
involved by the infection, or that the body of the AFBG may plasty may be performed as long as it does not extend to any
be involved, then the vein conduit is not anastomosed to the major tibial branch point. Inappropriate options would be
contralateral AFBG limb. Instead the vein conduit is anasto- prosthetic extension of the infected graft to a more distal
mosed either to the contralateral SFA or, if the SFA is segment of the popliteal artery or to a tibial artery.
occluded, to the PFA. If the entire AFBG subsequently If there is significant disease in the run-off arteries and
proves to be involved by the infection, it can be removed adequate outflow can only be obtained by treating this dis-
without necessitating revision of the vein conduit anastomo- ease or extending the graft to a more distal arterial seg-
sis. If the entire AFBG requires removal, our preference is to ment, such as the below-knee popliteal artery or a tibial
perform in-line reconstruction, again using either the artery, the entire infected graft should be removed and
patient’s superficial femoral vein or, more commonly, either replaced. In this situation, we would not place another
cryopreserved superficial femoral vein or cryopreserved aor- prosthetic graft but choose instead the patient’s greater
toiliac segments (see Chapter 25). The chosen conduit is saphenous vein if it is still available, the other choices being
anastomosed proximally to the infrarenal aorta and then dis- the lesser saphenous vein, an arm vein or a cryopreserved
tally to the recently placed cross-femoral vein conduit. The saphenous vein. An inappropriate option would be an
site of the distal anastomosis to the vein conduit is optional. angioplasty or stenting of the distal disease, there being no
endoluminal approach to those lesions which does not
involve traversing the infected graft. Excision of the infected
Occult infection femoropopliteal graft graft without revascularisation is an acceptable option only
if the patient has no symptoms of acute ischaemia, a clinical
INADEQUATE OUTFLOW
scenario not covered in this chapter.
If the patient presents with occlusion of a femoropopliteal
graft and at open thrombectomy an occult infection is
INADEQUATE INFLOW
detected, the first step is to proceed to transgraft thrombec-
tomy (Fig. 18.4). If inflow is sufficient, but there is a distal If inflow to femoral level is insufficient to maintain patency
anastomotic stenosis, focal balloon angioplasty may be of the femoropopliteal graft, there are several other options

Inadequate inflow Inadequate outflow

Distal anastomotic Diseased

Endoluminal Open operative stenosis run-off

Appropriate Appropriate Appropriate Appropriate

Iliofemoral angioplasty Iliofemoral TEA Short patch angioplasty – do Remove entire graft
(use non-infected Iliofemoral bypass NOT extend to major branches Replace with tissue conduit
access) Femoro-femoral bypass Transgraft endoluminal
Inappropriate
Inappropriate (use tissue conduits for anastomotic angioplasty
Prosthetic graft extension
all bypasses) Inappropriate
Stent placement to distal popliteal or tibial
Stent-graft placement Inappropriate Prosthetic graft extension to arteries
Prosthetic conduits distal popliteal or tibial arteries Outflow artery angioplasty
Outflow artery angioplasty or or stenting through the
stenting through the infected infected graft
graft
Successful Unsuccessful

Resect remaining graft

Determine extent of graft Define arterial anatomy
involved by infection
involved by infection (if not already done)
and reconstruct

Figure 18.4 Algorithm of management of occult infection of femoropopliteal graft. TEA, thromboendarterectomy
 Management 217

(Fig. 18.4). Balloon angioplasty can be performed to treat graft. Subsequent removal of the infected femoropopliteal
any inflow lesion, provided that access is obtained through a conduit can then be performed without revising the distal
non-infected site, for example via a contralateral retrograde anastomosis of the recently placed inflow graft.
femoral approach or an antegrade brachial approach. Stent Re-establishing perfusion to the ischaemic limb allows
placement should be avoided because of the risk of infection time for expeditious assessment of the extent of involve-
even though there is no direct continuity between the prox- ment of the graft by the infection, the relevant arterial
imal stent and the infected femoropopliteal graft. Stent-graft anatomy and any associated comorbidities. Appropriate
placement should not be performed, even in a remote loca- definitive treatment can then be selected. Occasionally an
tion. If ipsilateral inflow through the native iliofemoral infection will involve only the body of a femoropopliteal
arteries is insufficient and cannot be adequately improved graft, but more commonly the entire graft is infected.
by angioplasty, then an open approach to establish adequate Although some investigators have recommended preserva-
inflow is needed. Appropriate options include ipsilateral tion or partial preservation of an infected lower extremity
iliofemoral endarterectomy, with or without patch angio- graft as the optimal treatment,8–10 we prefer to remove the
plasty, ipsilateral iliofemoral bypass or cross-femoral bypass. infected graft and replace it with a tissue conduit in the
We would not use prosthetic material for any patch or form of either the patient’s vein (greater saphenous, lesser
another prosthetic conduit for the iliofemoral or cross- saphenous, or arm vein) or cryopreserved saphenous vein.
femoral bypass, instead preferring the patient’s superficial We generally do not use cryopreserved arterial conduits for
femoral vein harvested from the uninfected leg, cryopre- replacement of infected lower extremity grafts because the
served superficial femoral vein or cryopreserved arterial seg- size is generally not well matched to the native arteries.
ments of an appropriate diameter. The patient’s greater
saphenous vein may be used if it is of adequate calibre, and Graft occlusion with suspected infection
we prefer to preserve this conduit and use it to replace the
infected femoropopliteal graft. The distal anastomosis of When graft occlusion occurs and graft infection is sus-
the inflow iliofemoral bypass or cross-femoral bypass is to the pected, the severity of tissue ischaemia still determines the
native CFA or PFA and not to the infected femoropopliteal approach to treatment (Fig. 18.5). Thrombolysis, however,

Lack of limb-threatening
Mild/moderate ischaemia Advanced ischaemia
ischaemia allows standard
approach to management
of graft infection

Thrombectomy

Define arterial anatomy Dependent upon availability

Avoids additional Most likely approach
open procedure Mechanical Confine to body of graft
Concern about Operative Avoid anastomosis if possible
potential graft or May precipitate definitive repair
anastomotic
disruption
Establish extent of graft
involvement by infection
Successful Unsuccessful

Successful Unsuccessful

Select and perform Confine graft revision to the lesion responsible for Graft revision
definitive treatment failure of simple thrombectomy
Do not perform remote stenting or angioplasty
through the infected graft
Do not place prosthetic extensions of
the infected graft

Successful Unsuccessful

If simple thrombectomy and limited graft revision are

unsuccessful, must be prepared to proceed with all or part of the
definitive repair, depending on the graft anatomy (see text).
May require:
Intraoperative definition of arterial anatomy
Availability of alternative tissue conduits
Reprep and redrape to begin work in non-infected fields

Figure 18.5 Algorithm of management of graft occlusion with suspected infection

218 Acute ischaemia secondary to occult prosthetic graft infection

is no longer an appropriate consideration. The inevitability within the setting of acute graft or graft limb occlusion, one
of an open intervention eliminates the potential for total can use the available outcome data for generic prosthetic
endoluminal management, the principal benefit of throm- graft infection to establish a range of obtainable outcomes.
bolysis. Meanwhile, all of the potential disadvantages of In general, the outcome of graft infection has improved as
thrombolysis remain, namely, delay in revascularisation, the principles of management have become better defined
risk of bleeding and risk of septic embolisation. Therefore, and the treatment options have expanded. Investigators usu-
patients with mild or moderate ischaemia will undergo ally acknowledge that the presence of the foreign body, i.e.
expeditious assessment of the extent of infection, delin- the prosthetic graft, potentiates the infection. Therefore,
eation of arterial anatomy and then definitive treatment of complete removal of the infected prosthesis and avoidance
the infection. Patients with advanced ischaemia will most of tissue planes by using extra-anatomical routes in any sub-
often undergo open thrombectomy. The approach to open sequent revascularisation procedure offers the best chance of
thrombectomy and graft revision, if simple transgraft eradicating the infection. Almost all investigators, however,
thrombectomy fails, is the same as that for totally occult also acknowledge that the parallel goal of maintenance of
graft infection discovered incidentally during treatment of perfusion to the distal bed is not optimally achieved with
graft occlusion. Again, the basic principles to be empha- extra-anatomical bypass. The standard approach to the man-
sised are to perform as minimal a procedure as possible agement of infected grafts, i.e. complete graft excision with
which will successfully restore perfusion without further revascularisation through remote non-infected fields,5,11–17
complicating subsequent definitive treatment of the graft has been supplemented with in situ revascularisation utilising
infection. After restoration of flow and reversal of ischaemia, a variety of conduits including prostheses such as Dacron
the patient will undergo routine assessment and manage- and expanded polytetrafluoroethylene (PTFE),7,18–33 antibi-
ment of the graft infection. otic (rifampin)-bonded prosthetic conduits,34–38 autologous
vein,39–47 arterial allografts48–60 and venous allografts.
The potential benefits of in situ revascularisation are that
it is a simplified single operative procedure with lower ampu-
RESULTS tation rates and possibly even lower mortality. The potential
disadvantages are the risk of persistent or recurrent infection,
In general, patients who present with limb-threatening particularly when another prosthetic conduit is placed in
ischaemia of any aetiology have a higher rate of limb loss the tissue bed contaminated by graft infection. The use of
than patients who present with non-limb-threatening autogenous deep venous conduits is associated with a risk of
ischaemia. In addition mortality and morbidity rates are significant limb oedema and even the venous compartment
higher in this patient group, consistent with both the physi- syndrome.39,45 Allograft use is associated with conduit failure
ological effects and the need for an urgent, if not emergency, either because of aneurysmal degeneration or stenosis/occlu-
intervention. One would expect these observations also to sion, possibly reflecting low grade chronic rejection.
apply to patients with acute ischaemia in the setting of occult Overall, the available results of these treatment approaches
or suspected graft infection. Although there are no outcome published during the past 15 years suggest that reason-
data specific to the clinical situation of occult graft infection able outcomes may be achieved (Table 18.1). Many of the

Table 18.1 Outcome of prosthetic vascular graft infection: selected series 1990–2004

Recurrent/
Conduit Conduit persistent Conduit
Death Limb loss disruption failure infection dilation
Conduit Number (per cent) (per cent) (per cent) (per cent) (per cent) (per cent)

Venous autograft 62 10–33 0–20 0–33 0–17

Arterial allograft 248 13–30 0–6 3–17 5–44 7–24 5–22
In situ prosthetic 72 0–50 0–40 0–7 0 10–60
In situ prosthetic- 36 11–36 0 0–18 9–36 0–18
antibiotic bonded
Extra-anatomical bypass 276 19 14 13–27 3–13
and infected graft
removal – others
Extra-anatomical bypass and 55 24 15 0 25 5
infected graft removal – UCSF
Venous allograft – UCSF 20 15 0 4 0 5 5

UCSF, University of California, San Francisco.

 References 219

reported series, however, are small, the patients having

been selected rather than taken consecutively, accumulated
Conclusions
over long time intervals and during which there have Acute ischaemia secondary to graft occlusion with an
been many changes in patient management. Outcome underlying occult graft infection is a rare occurrence.
data beyond the perioperative interval are often limited or There are few specific data to guide treatment or establish
not available. Finally, most investigators triage their choice expected outcomes. In general, the degree of ischemia will
of the in-line conduit according to the clinical scenario; dictate the initial treatment approach. In the most com-
prosthetic in-line conduits tend to be used almost exclu- mon setting non-limb-threatening ischaemia, caused by
sively in the setting of aortoenteric fistula, whereas tissue the occluded graft, results in operative exploration and
in-line conduits tend to be used in the setting of chronic incidental discovery of graft infection. The principles
infection. guiding treatment are the relief of ischaemia using the
In general, in situ reconstruction is recommended for simplest approach in restoring flow, avoiding any proced-
consideration in infections which are culture-negative or ure which risks spreading infection or which makes sub-
when there are no systemic signs of it. Such treatment is sequent definitive treatment of the graft infection more
also appropriate when significant occlusive disease severely difficult.
disadvantages the haemodynamic success of an extra- Once ischaemia has been relieved expeditiously, the
anatomical bypass, when the degree of occlusive disease arterial anatomy is delineated, the extent of graft
itself demands treatment, and finally, and when major involved by the infection is established and definitive
branches are involved in the infected graft, the last treatment is chosen and executed. In situ graft replace-
two situations being particularly relevant to infected ment is employed only for non-invasive, non-aggressive
aortic grafts. infections. The conduit chosen for in situ graft replace-
ment is, in order of preference, vein autograft, vein
allograft, arterial allograft, antibiotic-bonded prosthetic
graft or a simple prosthesis. In following these guidelines
In situ reconstruction the best possible combined outcome endpoints of patient
survival, eradication of infection and limb salvage can be
• Recommended achieved.
– Culture-negative infections
– No signs of systemic infection
– Likely compromise of extra-anatomical bypass by
Key references
significant occlusive disease
– Significant disease in major (especially aortic) Bandyk DF, Novotny ML, Back MR, et al. Expanded application of in
branches requiring treatment situ replacement for prosthetic graft infection. J Vasc Surg
– Infection involving major (especially aortic) 2001; 34: 411–20.
branches Brown PM, Kim VB, Lalikos JF, et al. Autologous superficial femoral
• Not recommended vein for aortic reconstruction in infected fields. Ann Vasc Surg
– Abscess 1999; 13: 32–6.
– Gross, virulent or invasive infection Chiesa R, Astore D, Picolo G and the Italian Collaborative Vascular
– Gram-negative, S. aureus or culture-positive Homograft Group. Fresh and Cryopreserved arterial
infection homographs in the treatment of prosthetic graft infections:
– Systemic manifestations of infection experience of the Italian Collaborative Group. Ann Vasc Surg
1998; 12: 457–62.
Hayes PD, Nasim A, London NJM, et al. In situ replacement of
infected aortic grafts with rifampicin-bonded prostheses: The
Leicester experience (1992–98). J Vasc Surg 1999; 30: 92–8.
In general, in situ reconstruction is not recommended Seeger JM, Pretus HA, Welborn MB, et al. Long-term outcome after
treatment of aortic graft infection with staged extra-anatomic
for consideration when there is an abscess or when infec-
bypass grafting and aortic graft removal. J Vasc Surg 2000;
tions are gross, culture-positive, caused by Gram-negative
32: 451–61.
organisms, by S. aureus, particularly if methicillin sensitive
or resistant (MRSA), and if there are systemic, virulent or
invasive manifestations.
For the most part occult graft infection, manifesting as
REFERENCES
graft or graft limb occlusion, will not have any of the above
features contraindicating in situ replacement of the infected 1 Lorentzen JE, Nielsen OM, Arendup H, et al. Vascular graft infection:
graft, as almost all of them reflect an aggressive and invasive an analysis of sixty-two graft infections in 2411 consecutively
infection which is not likely to remain occult. implanted synthetic vascular grafts. Surgery 1985; 98: 81–6.
220 Acute ischaemia secondary to occult prosthetic graft infection

2 Bunt TJ. Synthetic vascular graft infections. I: Graft infections. 23 Jensen LJ, Kimose HH. Prosthetic graft infections: a review of
Surgery 1983; 93: 733–46. 720 arterial prosthetic reconstructions. Thorac Cardiovasc
3 Seeger JM. Management of patients with prosthetic graft Surgeon 1985; 33: 389–91.
infection. Am Surg 2000; 66: 166–77. 24 Young RM, Cherry KJ Jr, Davis PM, et al. The results of in situ
4 Reilly LM, Lusby RJ, Altman H, Kersh RA, et al. Late results prosthetic replacement for infected aortic grafts. Am J Surg
following surgical management of vascular graft infection. 1999; 178: 136–40.
J Vasc Surg 1984; 1: 36–44. 25 Fiorani P, Speziale F, Rizzo L, et al. Long-term follow-up after
5 Seeger JM, Back MR, Albright JL, et al. Influence of patient in situ graft replacement in patients with aortofemoral graft
characteristics and treatment options on outcome of patients infections. Eur J Vasc Endovasc Surg 1997; 14(suppl A): 111–14.
with prosthetic aortic graft infection. Ann Vasc Surg 1999; 26 Miller JH. Partial replacement of an infected arterial graft by a
13: 413–20. new prosthetic polytetrafluoroethylene segment: a new
6 Wilson SE. New alternatives in management of the infected therapeutic option. J Vasc Surg 1993; 17: 546–48.
vascular prosthesis. Surg Infect 2001; 2: 171–7. 27 Henke PK, Bergamini TM, Rose SM, Richardson JD. Current
7 Bandyk DF, Esses GE. Prosthetic graft infection. Surg Clin North options in prosthetic vascular graft infection. Am Surg 1998; 64:
Am 1994; 74: 571–90. 39–46.
8 Calligaro KD, Veith FJ, Yuan JG, et al. Intre-abdominal aortic 28 Thomas WEG, Baird RN. Secondary aorto-enteric fistulae:
graft infection: complete or partial graft preservation in patients Towards a more conservative approach. Br J Surg 1986; 73:
at very high risk. J Vasc Surg 2003; 38: 1199–205. 875–8.
9 Calligaro KD, Veith FJ, Schwartz ML, et al. Selective preservation 29 Jacobs MJHM, Reul GJ, Gregoric I, Cooley DA. In-situ
of infected prosthetic arterial grafts: analysis of a 20-year replacement and extra-anatomic bypass for the treatment of
experience with 120 extracavitary infected grafts. Ann Surg infected abdominal aortic grafts. Eur J Vasc Endovasc Surg 1991;
1994; 220: 461–71. 5: 83–6.
10 Calligaro KD, Veith FJ, Schwartz ML, et al. Are Gram-negative 30 Vollmar JF, Kogel H. Aortoenteric fistulas as postoperative
bacteria a contraindication to selective preservation of infected complication. J Cardiovasc Surg 1987; 28: 479–84.
prosthetic grafts. J Vasc Surg 1992; 16: 337–46. 31 Higgins RSD, Steed DL, Julian TB, et al. The management of
11 Reilly LM, Stoney RJ, Goldstone J, Ehrenfeld WK. Improved aortoenteric and paraprosthetic fistulae. J Cardiovasc Surg 1990;
management of aortic graft infection: the influence of operation 31: 81–6.
sequence and staging. J Vasc Surg 1987; 5: 421–31. 32 Sorensen S, Lorentzen JE. Case report: recurrent graft-enteric
12 Kuestner L, Reilly LM, Jicha D, et al. Secondary aortoenteric fistulae. Eur J Vasc Endovasc Surg 1989; 3: 583–5.
fistulas: Contemporary outcome using extra-anatomic bypass 33 Walker WE, Cooley DA, Duncan JM, et al. The management of
and graft excision. J Vasc Surg 1995; 21: 184–96. aortoduodenal fistula by in situ replacement of the infected
13 O’Hara PJ, Hertzer NR, Beven EG, Krajewski LP. Surgical abdominal aortic graft. Ann Surg 1987; 205: 727–32.
management of infected abdominal aortic grafts: review of a 34 Nasim A, Hayes P, London N, et al. In situ replacement of
25-year experience. J Vasc Surg 1986; 3: 725–31. infected aortic grafts with rifampicin-bonded prostheses. Br J
14 Schmitt DD, Seabrook GR, Bandyk DF, Towne JB. Graft excision Surg 1999; 86: 695.
and extra-anatomic revascularization: the treatment of choice for 35 Hayes PD, Nasim A, London NJM, et al. In situ replacement of
the septic aortic prosthesis. J Cardiovasc Surg 1990; 31: 327–32. infected aortic grafts with rifampicin-bonded prostheses: the
15 Yeager RA, Taylor LM Jr, Moneta GL, et al. Improved results with Leicester experience (1992–98). J Vasc Surg 1999; 30: 92–8.
conventional management of infrarenal aortic infection. J Vasc 36 Torsello G, Sandmann W. Use of antibiotic-bonded grafts in
Surg 1999; 30: 76–83. vascular graft infection. Eur J Vasc Endovasc Surg 1997;
16 Seeger JM, Pretus HA, Welborn MB, et al. Long-term outcome 14(suppl A): 84–87.
after treatment of aortic graft infection with staged extra- 37 Naylor AR, Clark S, London NJM, et al. Treatment of major aortic
anatomic bypass grafting and aortic graft removal. J Vasc Surg graft infection: preliminary experience with total graft excision
2000; 32: 451–61. and in situ replacement with a rifampicin bonded-prosthesis.
17 Ricotta JJ, Faggioli GL, Stella A, et al. Total excision and extra- Eur J Vasc Endovasc Surg 1995; 9: 252–6.
anatomic bypass for aortic graft infection. Am J Surg 1991; 162: 38 Batt M, Magne J-L, Alric P, et al. In situ revascularization with
145–9. silver-coated polyester grafts to treat aortic infection: early and
18 Robinson JA, Johansen K. Aortic sepsis: is there a role for in situ midterm results. J Vasc Surg 2003; 38: 983–9.
graft reconstruction? J Vasc Surg 1991; 13: 677–84. 39 Brown PM, Kim VB, Lalikos JF, et al. Autologous superficial
19 Bandyk DF, Novotny ML, Back MR, et al. Expanded application of femoral vein for aortic reconstruction in infected fields. Ann Vasc
in situ replacement for prosthetic graft infection. J Vasc Surg Surg 1999; 13: 32–6.
2001; 34: 411–20. 40 Nevelsteen A, Suy R. Autogenous venous reconstruction in the
20 Menawat SS, Gloviczki P, Serry RD, et al. Management of aortic treatment of aortobifemoral prosthetic infection. J Cardiovasc
graft-enteric fistulae. Eur J Vasc Endovasc Surg 1997; 14(suppl A): Surg 1988; 29: 315–17.
74–81. 41 Nevelsteen A, Lacroix H, Suy R. Autogenous reconstruction with
21 Towne JB, Seabrook GR, Bandyk D, et al. In situ replacement of the lower extremity deep veins: an alternative treatment of
arterial prosthesis infected by bacterial biofilms: long-term prosthetic infection after reconstructive surgery for aortoiliac
follow-up. J Vasc Surg 1994; 19: 226–35. disease. J Vasc Surg 1995; 22: 129–34.
22 Bandyk DF, Bergamini TM, Kinney EV, et al. In situ replacement of 42 Franke S, Voit R. The superficial femoral vein as arterial
vascular prostheses infected by bacterial biofilms. J Vasc Surg substitute in infections of the aortoiliac region. Ann Vasc Surg
1991; 13: 575–83. 1997; 11: 406–12.
 References 221

43 Sicard GA, Reilly JM, Doblas M, et al. Autologous vein 52 Chiesa R, Astore D, Piccolo G and the Italian Collaborative
reconstruction in prosthetic graft infections. Eur J Vasc Endovasc Vascular Homograft Group. Fresh and cryopreserved arterial
Surg 1997; 14(suppl A): 93–8. homografts in the treatment of prosthetic graft infections:
44 Nevelsteen A, Lacroix H, Suy R. Infrarenal aortic graft experience of the Italian Collaborative Vascular Homograft
infection: in situ aortoiliofemoral reconstruction with the Group. Ann Vasc Surg 1998; 12: 457–62.
lower extremity deep veins. Eur J Vasc Endovasc Surg 1997; 53 Desgranges P, Beaujan F, Brunet S, et al. Cryopreserved arterial
14(suppl A): 88–92. allografts used for the treatment of infected vascular grafts.
45 Benjamin ME, Cohn EJ Jr, Purtil WA, et al. Arterial reconstruction Ann Vasc Surg 1998; 12: 583–8.
with deep leg veins for the treatment of mycotic aneurysms. 54 Agrifoglio G, Bonalumi F, Scalamogna M, et al. Aortic allograft
J Vasc Surg 1999; 30: 1004–15. replacement: North Italy Transplant Programme (NITp). Eur J Vasc
46 Clagett GP, Bowers BL, Lopez-Viego MA, et al. Creation of a neo- Endovasc Surg 1997; 14(suppl A): 108–10.
aortoiliac system from lower extremity deep and superficial 55 Vogt PR, von Segesser LK, Goffin Y, et al. Eradication of aortic
veins. Ann Surg 1993; 218: 239–49. infection with the use of cryporeserved arterial allografts.
47 Clagett GP, Valentine RJ, Hagino RT. Autogenous aortoiliac/ Ann Thorac Surg 1996; 62: 640–5.
femoral reconstruction from superficial femoral-popliteal 56 Vogt PR, Brunner-La Rocca HP, Carrel T, et al. Cryopreserved
veins: feasibility and durability. J Vasc Surg 1997; 25: arterial allografts in the treatment of major vascular infection:
255–70. a comparison with conventional surgical techniques. J Thorac
48 Ruotolo C, Plissonnier D, Bahnini A, et al. In situ arterial Cardiovasc Surg 1998; 116: 965–72.
allografts: a new treatment for aortic prosthetic infection. Eur J 57 Lehalle B, Geschier C, Fieve G, Stolz JF. Early rupture and
Vasc Endovasc Surg 1997; 14(suppl A): 102–107. degeneration of cryopreserved arterial allografts. J Vasc Surg
49 Knosalla C, Goeau-Brissonniere O, Leflon V, et al. Treatment of 1997; 25: 751–2.
vascular graft infection by in situ replacement with 58 Kieffer E, Gomes D, Plissonnier D, et al. Current use of allografts for
cryopreserved aortic allografts: an experimental study. J Vasc infrarenal aortic graft infection. In: Yao JSY, Pearce WH (eds).
Surg 1998; 27: 689–98. Modern Vascular Surgery. New York: McGraw Hill, 2000: 297–308.
50 Kieffer E, Bahnini A, Koskas F, et al. In situ allograft replacement 59 Mestres CA, Mulet J, Pomar JL. Large-caliber cryopreserved
of infected infrarenal aortic prosthetic grafts: results in forty- arterial allografts in vascular reconstructive operations: early
three patients. J Vasc Surg 1993; 17: 349–56. experience. Ann Thorac Surg 1995; 60: S105–S107.
51 Bahnini A, Ruotolo C, Koskas F, Kieffer E. In situ fresh allograft 60 Verhelst R, Lacroix V, Vraux H, et al. Use of cryopreserved arterial
replacement of an infected aortic prosthetic graft; eighteen homografts for management of infected prosthetic grafts:
months’ follow-up. J Vasc Surg 1991; 14: 98–102. A multicentric study. Ann Vasc Surg 2000; 14: 602-7.
This page intentionally left blank
 19
The Acute Diabetic Foot

CAMERON M AKBARI, FRANK W LOGERFO

The problem 223 Principles of arterial reconstruction in the 231

Pathology of the diabetic foot 223 diabetic foot
Clinical evaluation 226 Results 232
Treatment 229 References 233

THE PROBLEM Acting synergistically, they contribute to the sequence

of tissue ulceration, necrosis and eventually gangrene.
Prevention and treatment of diabetic foot problems should
With the rising incidence and prevalence of diabetes melli- be tailored to these pathogenic factors, either solely or in
tus worldwide, there has been an even greater increase in combination.
the prevalence of diabetes related complications. Foot
ulceration is one of the most common and formidable
complications of diabetes, affecting almost 20 per cent of Neuropathy
all diabetic patients during their lifetime. Despite advances
in management, foot problems continue to be the most Peripheral neuropathy is a common complication of dia-
common reason for hospitalisation among these patients.1 betes mellitus, affecting as many as 60 per cent of all
Diabetes remains the single strongest risk factor for limb patients in their lifetime and up to 80 per cent of those pre-
loss, contributing to half of all lower extremity amputa- senting with foot lesions.3 Broadly classified as focal and
tions in the USA. The relative risk for leg amputation is diffuse, the latter is more common and includes both the
40 times greater among diabetic patients. Moreover, up to autonomic neuropathy and chronic sensorimotor polyneu-
50 per cent of diabetic amputees will undergo a second ropathies implicated in foot ulceration.
leg amputation within 5 years of the first one.2 Sensorimotor neuropathy initially involves the distal
The spectrum of diabetic foot disease ranges from the lower extremities; it progresses centrally and tends to
asymptomatic patient, who may require only preventive be symmetrical. Sensory nerve fibre involvement leads to
foot care, to the unstable and critically ill patient in whom loss of the protective sensation of pain, while motor
both loss of life and limb are imminent threats. Indeed, the nerve fibre loss results in small muscle atrophy in the foot,
variety of presentations of the diabetic foot often con- flexion of the metatarsals and subsequent prominence of
tributes to the clinical confusion and diagnostic and treat- the metatarsal heads and clawing of the toes. This in
ment delays which unfortunately lead to limb loss. It turn results in the development of abnormal pressure
follows that the clinician needs to develop an orderly points on the plantar bony prominence which lacks pro-
approach, grounded on firm pathophysiological prin- tective sensation culminating in ulceration at these pres-
ciples, in order to prevent such disasters. sure points. Loss of intrinsic muscle function also results
in digital contractures, hammer toe and clawed toe deform-
ities, pes cavus and ulceration. Because the motor neuro-
PATHOLOGY OF THE DIABETIC FOOT pathy may affect the extensor musculature of the leg,
deformities can also involve the ankle joint, with result-
The pathogenic mechanisms involved in diabetic foot ant so-called equinus deformity and abnormal bending
disease include ischaemia, neuropathy and infection. forces.
224 The acute diabetic foot

Sensorimotor neuropathy Infection

The unique anatomy of the foot has implications for the

• Tends to be symmetrical
presentation and treatment of infection. Most infections
• Sensory – loss of protective sensation of pain
progress within the plantar aspect of the foot consisting of
• Motor – small muscle atrophy, flexion contractures
three compartments: medial, central and lateral.5 The floor
• Deformities – callosities, hammer toe, clawed toes,
of each compartment is the rigid plantar fascia and the roof
pes cavus, equinus
is composed of the metatarsal bones and interosseous
• Necrosis and ulceration at pressure points
fascia. A thick medial intermuscular septum, extending
from the medial calcaneal tuberosity to the head of the first
metatarsal, defines the medial and central compartments.
Autonomic denervation leads to loss of sympathetic The intrinsic muscles of the great toe are in the medial
tone and increased arteriovenous shunting in the foot with compartment whereas the central compartment is composed
defective nutrient flow. Impaired autonomic regulation of of the intrinsic muscles of the second, third and fourth
the sweat glands leads to anhidrosis and cracking of dry toes as well as the extensor and flexor tendons of those toes,
skin, which creates a predisposition to skin breakdown and the medial and lateral plantar nerves, and the plantar vas-
ulceration. Because of sympathetic innervation to the cular structures. The lateral intermuscular septum, from
bone, autonomic neuropathy may also result in an increase the calcaneus to the fifth metatarsal, delineates the lateral
in bone blood flow and subsequent osteopenia and ‘bone compartment, which contains the intrinsic muscles of the
washout’. This osteoarthropathy associated with diabetes is fifth toe.
more commonly known as Charcot’s foot, once histor- Diabetic foot infection may result from a simple punc-
ically associated with syphilitic tabes dorsalis, and now ture wound, a neuropathic ulcer, the nail plate, or from
almost exclusively a complication of diabetes. the interdigital web space. Because the intrinsic muscles of
each digit are essentially confined within the respective
plantar compartment, untreated distal phalangeal infection
Autonomic neuropathy may progress to a plantar abscess. Infection within these
rigid anatomical compartments also creates high intra-
compartmental pressures, which subsequently impairs
• Loss of sympathetic tone
capillary blood flow and leads to progressive tissue ischaemia
• Increased arteriovenous shunting – defective flow
and necrosis. Because the roof of each compartment
• Anhidrosis – skin cracking and breakdown
is composed of bone and fascia, deep space infections
• Increased bone blood flow – osteopenia or ‘bone
show deceptively little abnormality on the dorsal foot.
washout’
Left untreated, ongoing infection and cellulitis progresses
to bacterial spread from one compartment to another
through direct perforation of the medial or lateral inter-
The aetiology of Charcot’s foot deformity includes a muscular septum (Fig. 19.1), causing the development of
combination of both sensorimotor and autonomic neu-
ropathies. Continued walking on an insensate joint, com-
bined with muscle imbalance and atrophy from motor
neuropathy and limited joint mobility, leads to joint instabi-
lity, loss of joint architecture and ultimately bone and joint
destruction.4 This destructive process leads to increased
bone blood flow, concomitant resorption and softening of
normal bone, which in turn leads to further bone destruc-
tion and ultimately fracture and Charcot’s joint.

Pathology of Charcot’s joint

• Aetiology – sensorimotor and autonomic neuropathy

• Insensate joint, muscle atrophy, contracture –
reduced mobility
• Joint instability, loss of joint architecture
• Weight-bearing and walking – bone and joint Figure 19.1 Infection of the foot from a small plantar ulcer
which has progressed from the central to medial compartments
destruction
with proximal extension as well
 Pathology of the diabetic foot 225

proximal foot and ankle abscesses at the calcaneal conver- response following injury increases the susceptibility of the
gence of the septum and eventually leading to an unsalvage- diabetic foot to infection.14,15 These changes, however, do
able foot. not lead to narrowing of the capillary lumen, and arteriolar
The microbiology of the diabetic foot infection is blood flow may be normal or even increased despite these
dependent on the patient’s environment, whether as outpa- changes.16
tient or hospitalised, and the severity of the infection itself.6 A variety of other microvascular abnormalities may be
Mild localised and superficial ulcerations, particularly demonstrated in the diabetic foot.17 Both capillary blood
among outpatients, are usually caused by aerobic Gram- flow and the maximal hyperaemic response to stimuli are
positive cocci such as Staphylococcus aureus or streptococci. reduced in the diabetic foot, suggesting that functional
In contrast, deeper ulcers and more generalised, limb- microvascular impairment is a major contributing factor in
threatening infections are usually polymicrobial in content. diabetic foot problems. Neurogenic vasodilatation is also
In addition to Gram-positive cocci, the causative organisms impaired in the diabetic foot.18 Normally, injury-mediated
in this latter group may include Gram-negative bacilli nociceptive C fibre stimulation results in adjacent neuro-
(Escherichia coli, Klebsiella, Enterobacter aerogenes, Proteus genic release of vasoactive peptides, which subsequently lead
mirabilis and Pseudomonas aeruginosa) and anaerobes to vasodilation and increased blood flow to the area of injury.
(Bacteroides fragilis and peptostreptococci). Enterococci Absence of this axon reflex further reduces the inflammatory
may also be isolated from the wound, notably among hos- hyperaemic response to injury in the diabetic foot.
pitalised patients, and these, in the absence of other cul-
tured virulent organisms, should probably be considered
pathogenic. Microvascular functional impairment

• Endothelial dysfunction
Ischaemia – microvascular and macrovascular • Thickening of the basement membrane
considerations • Impairment of migration of leucocytes
• Impaired inflammatory hyperaemic response
Two distinct types of vascular disease are seen in patients • Susceptibility to sepsis
with diabetes.7 The first is a non-occlusive microcirculatory • Arterial flow may be normal or increased
impairment, characteristically involving the capillaries and • Neurogenic axon reflex mediated vasodilatation is
arterioles of the kidneys (nephropathy), eye (retinopathy) and impaired
peripheral nerves (neuropathy) with significant effects in the
diabetic foot. The second is a macroangiopathy, characterised
As noted earlier, macrovascular lower extremity disease in
by atherosclerotic lesions of the coronary and peripheral
the diabetic patient is morphologically similar to that in the
arterial circulation, which is morphologically and function-
non-diabetic patient. The major difference between these
ally similar in both non-diabetic and diabetic patients.
two populations of patients is the pattern and location of the
In the context of the diabetic foot and microvascular dys-
occlusive lesions.19 Whereas occlusive lesions of the superfi-
function, the so-called ‘small vessel disease’ of diabetes is an
cial femoral and popliteal segment are commonly found in
inaccurate term, since it suggests an untreatable occlusive
the non-diabetic patient with limb ischaemia, diabetic
lesion in the microcirculation. Prospective anatomical8 and
patients commonly have occlusive disease involving the
physiological studies have demonstrated that there is no such
infrageniculate, or tibial, arteries. However, the foot arteries
microvascular occlusive disease. Dispelling the notion of ‘small
are almost invariably patent, which allows for extreme distal
vessel disease’ is fundamental to the principles of limb salvage
arterial reconstruction despite extensive tibial or even prox-
in patients with diabetes, since arterial reconstruction is
imal multisegmental disease. In addition, the popliteal, super-
almost always possible and successful in these patients.9
ficial femoral and more proximal arteries are less likely to be
While there is no occlusive disease in the microcircula-
affected by atherosclerosis, which happily allows these vessels
tion, multiple structural and physiological abnormalities
to serve as an inflow source for distal arterial bypass grafts.
result in functional microvascular impairment.10 Endothelial
dysfunction and the response to nitric oxide is diminished in
patients with diabetes, neuropathy and vascular disease.11 Macrovascular disease
Hyperglycaemia alone may lead to at least some of these
abnormal responses.12 Thickening of the capillary basement
membrane is the dominant structural change in both neur-
• Femoropopliteal system less affected by
atherosclerosis
opathy and retinopathy whereas alterations of the basement
membrane are likely to contribute to albuminuria and the
• Occlusive disease often in infrageniculate and tibial
arteries
progression of diabetic nephropathy.13 In the diabetic foot,
capillary basement membrane thickening may theoretically
• Medial calcinosis of tibial arteries

impair the migration of leucocytes, and the hyperaemic

• Foot arteries invariably patent
226 The acute diabetic foot

CLINICAL EVALUATION strongly suggestive of ischaemia. Certainly, an ulcer or

gangrenous area which has been present for several months
is unlikely to heal without some type of further additional
As with all other disease processes, the initial evaluation of treatment, whether it be off-loading of weight-bearing areas,
the patient with any diabetic foot problem begins with a treatment of infection, or, most commonly, the correction
complete history and careful physical examination. Broadly of arterial insufficiency. Had the present ulcer healed pre-
classified, this bedside assessment includes the healing viously, and is the current episode a relapsing problem? A
potential of the foot, the details of the foot problem, e.g. history of intermittent healing followed by relapse should
ulcer, gangrene, infection, osteomyelitis, etc., the systemic raise suspicion of underlying untreated infection, such as
consequences of diabetes, and any immediate threats to life recurrent osteomyelitis, or uncorrected architectural abnor-
and/or limb.20 mality such as a bony pressure point or varus deformity.
It is helpful to consider past opinions and treatments,
History while still formulating an objective treatment plan
based on presenting data. Many diabetic patients with cor-
The history of the foot problem itself can give valuable rectable foot ulceration and limb ischaemia have been told
insight as to the potential for healing, the presence of coex- that the only option is limb amputation, usually due to
isting infection or arterial occlusive disease and the need ‘inherited pessimism’ and inadequate knowledge of the
for further treatment. Any patient presenting with foot advances made in limb and foot salvage. In these circum-
ulceration or gangrene should immediately arouse suspi- stances, when seeking an additional treatment opinion, it is
cion of underlying arterial insufficiency, even if neuropa- best to ‘start at the beginning’ rather than blindly concur
thy or infection is present. In the patient with diabetes and with previous actions.
arterial insufficiency, the development of a non-healing The past history should be first directed to previous foot
foot ulcer may be a seemingly benign event such as cutting and limb problems. Recent ipsilateral ulceration or foot
a toenail, soaking the foot in a warm bath or using a heat- surgery, healed in timely fashion and with uncomplicated
ing pad (Fig. 19.2). course, may suggest adequate arterial supply. When the
The duration of ulceration also provides important clues history is more remote, such information becomes less
to the extent that a long-standing, non-healing ulcer is useful. A history of previous leg revascularisation, includ-
ing percutaneous therapies, also provides an important
clue as to underlying arterial insufficiency. Other cardio-
vascular risk factors, such as cigarette smoking or hyperlip-
idaemia, should also be considered, as their presence
increases the likelihood that ischaemia is contributing to
the present foot problem.
Although claudication or rest pain has traditionally been
associated with vascular disease, diabetic neuropathy may
obscure those symptoms, and their absence in the diabetic
patient certainly does not rule out ischaemia. Because even
moderate ischaemia will preclude healing in the diabetic foot,
the absence of rest pain is not a reliable indicator of adequate
arterial blood supply; moreover, many patients may not walk
a sufficient distance to develop true vasculogenic claudica-
tion. Conversely, some patients with true ischaemic rest pain
are dismissed for years as having ‘painful neuropathy’.
Because unrecognised infection in the diabetic patient may
rapidly progress to a life-threatening condition, attention
should be directed toward detecting the subtle manifest-
ations of an infected foot ulcer. Worsening hyperglycaemia,
recent erratic blood glucose control and higher insulin
requirements all suggest untreated infection. Due to micro-
vascular and neuropathic abnormalities in the diabetic foot,
the classic symptoms of infection such as chills or pain are
often absent, and hyperglycaemia is often the sole presenting
symptom of undrained infection. As infection and hyper-
glycaemia progress, impending ketoacidosis or non-ketotic
Figure 19.2 Gangrene and non-healing following a ‘simple’ hyperglycaemic hyperosmolar coma may develop with symp-
arthroplasty toms of weakness, confusion and altered mental status.
 Clinical evaluation 227

The history should also include a comprehensive assess- maximal perfusion to heal. Inadequate or faulty assessment
ment of the patient’s overall health, to help stratify peri- and treatment of underlying ischaemia will lead to failure of
operative risk should some type of operative intervention limb salvage in the patient with diabetic foot ulceration.
be needed. Knowledge of previous cardiac events, such as Inspection of the leg and foot, including the ulcer itself will
myocardial infarction or revascularisation, and present car- often provide suggestive clues. For example, distal ulcer-
diac status, anginal severity and heart failure symptoms are ation at the tip of a digit, an ulcer unassociated with an
all mandatory components of the history taking. Similarly, exostosis or weight-bearing area, and the presence of gan-
in the patient with suspected infection and ischaemia, a his- grene, are all strongly consistent with underlying ischaemia.
tory of worsening renal function or impending need for Multiple ulcers or gangrenous areas on the foot, the absence
haemodialysis will help determine the dose and choice of of granulation tissue or lack of bleeding with debridement
antibiotics and may alter plans for standard contrast arteri- of the ulcer should immediately raise concerns of under-
ography. Functional status also becomes an important con- lying arterial insufficiency (Fig. 19.3). Other signs suggestive
sideration at this point, and the history should carefully of ischaemia include pallor with elevation, fissures, particu-
determine the ambulatory and rehabilitative potential of larly at the heel and absent hair growth. Although poor skin
the patient, so that appropriate decisions may be made for condition and hyperkeratosis may not always be good indi-
limb salvage or amputation. cators of arterial disease, they should be noted, as they may
help confirm initial clinical impressions.
Pulse examination, most notably the status of the foot
Physical examination pulses, is the single most important component of the phys-
ical exam, since, as has been emphasised, ischaemia is always
Initial evaluation of the diabetic foot ulcer should include a presumed to be present in the absence of a palpable foot pulse.
strong suspicion and thorough search for infection. In the As such, special attention should be directed toward the foot
patient with cellulitis, the entire foot, including the web pulses, which requires a knowledge of the usual location of
spaces and nail beds, should be examined for any potential the native arteries. The dorsalis pedis artery is located
portals of entry, such as a puncture wound or an interdigital between the first and second metatarsal bones, just lateral to
‘kissing’ ulcer. Encrusted and heavily calloused areas over the extensor hallucis longus tendon, and its pulse is palpated
the ulceration should be deroofed, and the wound thor-
oughly inspected to determine the extent of involvement.
An apparently benign dry gangrenous eschar can often hide
an undrained infectious collection. Cultures should be taken
from the base of the ulcer, avoiding superficial swabs which
may only yield colonising organisms. Findings consistent
with infection might include purulent drainage, crepitus,
tenderness, mild erythema and sinus formation, although
these findings may be entirely absent in the neuropathic
foot. Close inspection of the ulcer and the use of a sterile
probe may also confirm the presence of osteomyelitis, which
occurs commonly even in benign appearing ulcers; if bone
is detected with gentle probing, osteomyelitis is presumed
to be present. Although not always present, fever and tachy-
cardia are strongly suggestive of deep or undrained infection
with impending or already established sepsis.
Because of its prevalence and causative role in diabetic
foot problems neuropathy should be assessed in every dia-
betic patient, and appropriate preventive measures taken
to ensure against foot ulceration in the high risk neuro-
pathic foot. Protective sensation may be assessed with a
Semmes–Weinstein 5.07 strength monofilament; inability
to feel the monofilament when pressed to the skin correl-
ates well with an increased risk of foot ulceration. Advanced
sensorimotor neuropathy will lead to the presence of a
‘claw’ foot due to gradual atrophy of the intrinsic muscles.
Charcot’s deformity, with bone and joint destruction of
the mid-foot, may also be seen. Figure 19.3 Non-healing and lack of granulation tissue at the
Assessment of arterial perfusion in the diabetic foot is a bed of a third toe open amputation site with extension of gangrene
fundamental consideration, since the diabetic foot needs to the fourth toe
228 The acute diabetic foot

by the pads of the fingers as the hand is partially wrapped normal arterial perfusion at that level, in the occasional
around the foot. If the pulse cannot be palpated, the fingers patient, the waveform may be triphasic at the ankle even
may be moved a few millimetres in each direction, as the though occlusive disease may be present more distally. In
artery may occasionally have a slightly aberrant course. addition, the quality of the waveforms is affected by periph-
A common mistake is to place a single finger at one location eral oedema and in that sense is technically dependent.
on the dorsum of the foot. The posterior tibial artery is typ- Similarly, plethysmography and pulsed volume recordings
ically located in the hollow curve just behind the medial have several shortcomings, primary among them being that
malleolus, approximately halfway between the malleolus it frequently underestimates the severity of proximal arte-
and Achilles’ tendon. The examiner’s hand should be con- rial disease mainly due to the presence of collateral vessels.
tralateral to the examined foot, i.e. the right hand should be As with segmental Doppler waveforms, the quality of the
used to palpate the left foot and vice versa, so as to allow the test is affected by several variables, including room temper-
curvature of the hand to follow the ankle. ature, since temperature differences in the air cuff can
change the pressure measured by the air-filled plethysmo-
graph; other factors are peripheral oedema and obesity.
Non-invasive arterial evaluation Finally, extensive casts or bandages preclude accurate wave-
form measurement.
Non-invasive arterial testing has a limited role in the diabetic Despite the invaluable role of duplex ultrasound in the
patient with foot ulceration, and should not be used in place diagnosis of carotid artery disease and in postoperative
of the bedside evaluation. In selected patients, and in con- graft surveillance, multiple limitations apply to its use for
junction with the clinical findings, non-invasive testing the diagnosis of lower extremity arterial disease. A large
may provide useful information. Some examples might variation in the range of ‘normal’ velocities for leg arteries
include the diabetic patient with absent foot pulses and a may lead to a significant stenosis being misinterpreted.
superficial ulcer showing evidence of healing and a previ- Although the femoral and popliteal vessels may be visu-
ous history of a healed foot ulcer, or the patient without alised relatively easily, the tibial vessels are more cumber-
any foot lesions scheduled to undergo elective foot surgery. some to scan, and the velocities in the tibial arteries may be
In the patient with poor healing or gangrene and absent even more difficult to interpret. When one considers the
foot pulses, however, non-invasive testing will add little usual pattern of diabetic vascular disease in diabetes, with
additional information to the initial clinical evaluation and a predilection toward atherosclerotic involvement of the
will serve only to further delay vascular reconstruction. tibial vessels, the limitations of duplex in the diagnosis of
All the non-invasive arterial tests have limitations in the arterial insufficiency in the diabetic patient can be under-
presence of diabetes.21 Medial arterial calcinosis occurs fre- stood. Because the study depends on accurate sonographic
quently and unpredictably in patients with diabetes, and localisation of the vessel, duplex is quite ‘operator depend-
its presence renders the arteries non-compressible resulting ent’. Finally, multiple other variables can influence the
in artificially elevated segmental systolic pressures and quality of the image, including medial arterial calcification,
ankle:brachial indices (ABI). Therefore, a ‘normal’ ABI in a which can cause artefactual shadowing, obesity and periph-
patient with diabetes should be interpreted with caution. eral oedema, so compromising the imaging of tibial vessels.
Measurements may also be affected by the use of cuffs of Regional transcutaneous oximetry (TcPO2) measure-
inappropriate size. For example, too narrow a cuff will result ments are also unaffected by medial calcinosis, and some
in artefactually high pressures, and the so called ‘narrow cuff studies have noted its reliability in predicting healing of
artefact’ is often associated with obese patients. Lower levels ulcers and amputation levels.22 Because haemodynamics
of calcification in the toe vessels support the use of toe sys- are not measured, the test is immune to many of the prob-
tolic pressures as a surrogate measure of healing potential, lems facing other non-invasive tests in the presence of dia-
but despite its advantages in the presence of calcified vessels, betes, such as non-compressible vessels. Due to the unique
toe pressure measurements also have several limitations. considerations of the diabetic foot, however, TcPO2 meas-
The presence of a bandage or toe ulcer often precludes urements are not entirely reliable in the diabetic patient with
placement of the cuff. In addition, as a plethysmograph is foot ulceration. Although values less than 20 and greater
used to detect the pressure at which volume increases, the than 60 can be predictive, there is a large ‘grey area’ of inter-
quality of the tracing may be affected by any vasoconstricted mediate values which are of little clinical use. Additionally,
states caused by cold weather such as a cold room, a nervous patients with diabetes develop foot ulceration at higher
patient, etc. Finally, both the volume and photoplethysmo- TcPO2 values in comparison with the non-diabetic popula-
graphs require close calibration, and poor contact of the tion and, due to the effects of arteriovenous shunting and
photocell with the skin will yield poor results. microvascular dysfunction, a higher TcPO2 value may not
Segmental Doppler waveforms and pulsed volume correlate with healing potential in the diabetic patient. Even
recordings are unaffected by medial calcification, but evalu- in the patient with a ‘normal’ TcPO2 value, the measurement
ation of these waveforms is primarily qualitative and not may not accurately reflect healing potential at the target
quantitative. Although a triphasic Doppler waveform suggests area. Because the probe is placed typically at the proximal
 Treatment 229

dorsal foot, i.e. near the ankle, more distal ischaemia due to radiographic tests may confirm initial clinical suspicions,
possible distal tibial and paramalleolar occlusive disease the determination of the severity of infection is almost
may not be identified. Technical problems, such as poor always made based on the clinical findings, and should be
probe placement, lack of equipment standardisation and made without undue delay. Infection in the diabetic foot
user variability and lack of familiarity, may also undermine may range from a minimal superficial infection to fulmin-
the reliability of the study. ant sepsis with extensive necrosis and destruction of the
foot. Accordingly, the assessment of infection and the sub-
sequent treatment plan should consider the following:
TREATMENT choice of antibiotic, which requires knowledge of the
microbiology, the need for drainage, local or even guillo-
tine amputation and the medical condition of the patient.
Once the initial bedside history and physical evaluation has
In the compliant patient with no evidence of deep space
been completed, the astute clinician will have formulated a
involvement or systemic infection, treatment may be per-
plan determining the type and urgency of subsequent
formed as an outpatient, and consists of an oral antibiotic,
treatment and diagnostic tests. Specifically, this timely
pending culture results, and non-weight bearing of the
assessment focuses on the presence and severity of infec-
involved extremity. Because most pathogens in this group of
tion, whether the limb is salvageable, which includes an
patients are either Staphylococcus or Streptococcus, an oral
assessment of the underlying medical condition of the
penicillin or first generation cephalosporin is usually ade-
patient, and the presence of ischaemia. These considera-
quate. The patient should be instructed on appropriate dress-
tions are summarised in the algorithm in Fig. 19.4.
ing changes to the wound, frequent follow-up and guidelines
to recognise determining improvement or worsening.
Infection Unfortunately, a more common presentation is the
patient with ulceration or gangrene and a deep infection
Evaluation and treatment of infection assumes first priority involving tendon or bone and possible systemic involve-
in the management of any diabetic foot problem.23 Although ment. These patients require immediate hospitalisation,

Initial evaluation – history and physical examination

Yes
Septic/medically unstable? Evaluate for infection – is it present?
Yes No
No
Needs guillotine amputation? Start antibiotics, correct metabolic
abnormalities, decide need for drainage
Yes
Correct metabolic abnormalities,
stabilise patient, guillotine amputation Needs drainage?

Formal below- or above-knee amputation Drain, debride

No
Foot salvageable?
Yes
Yes
Arteriography Evaluate for ischaemia – ischaemia present?
No
Continued wound care –
Revascularisation dressings, debride as necessary,
continue antibiotics as necessary No

Need secondary (definitive) foot surgery?

Yes
Proceed with foot operation
Figure 19.4 Algorithm
Yes No detailing evaluation and
Preventive foot care Healed?
treatment of the diabetic foot
230 The acute diabetic foot

bed rest with elevation of the infected foot, correction of haemodynamic instability, bacteraemia and severe acid–base
any systemic abnormalities and broad spectrum intra- and electrolyte abnormalities. Such a patient should undergo
venous antibiotics, the range of which is narrowed once prompt open guillotine below-knee amputation. This type
culture results are complete. As noted earlier, the clinical of amputation is usually performed at ankle level to remove
findings of impending sepsis may be subtle, and therefore the septic source and to allow for subsequent revision and
these patients should have a complete laboratory work-up closure at a later date. Intravenous antibiotics, correction of
to detect and correct electrolyte and acid–base imbalances. dehydration and electrolyte abnormalities and continuous
Duration and choice of antibiotic therapy is dependent cardiac monitoring are absolutely essential throughout the
on the extent of infection. As noted earlier, deep or chronic, treatment process.
recurrent ulcers are typically polymicrobic, and appropriate Once the infection has been drained and the tissues
empirical antibiotic choices for those infections which are debrided, continued wound inspection and management
not life-threatening might include the following: clin- is essential. Persistent necrosis should raise suspicion of
damycin plus a fluoroquinolone; clindamycin plus a third or undrained infection or untreated ischaemia and further
fourth generation cephalosporin; an antipseudomonal peni- debridement and treatments may be necessary. Wounds
cillin. Subsequent culture results will then dictate any fur- should be kept moist, avoiding caustic solutions, soaks or
ther antibiotic coverage. In the absence of osteomyelitis, whirlpool therapy. Attention should also be focused on
antibiotics should be continued, typically 10–14 days, until avoiding any weight-bearing on the affected foot, while
the wound appears clean and all surrounding cellulitis has also maximising nutrition and controlling hyperglycaemia.
resolved. If osteomyelitis is present, treatment should
include both surgical debridement and a prolonged course,
Limb salvageability
say 4–6 weeks, of antibiotic therapy, though it may be abbre-
viated if the entire infected bone has been removed, as with
While the infection is being treated and controlled, the sur-
digital or transmetatarsal amputation. Heel lesions will often
geon should determine the chances and feasibility of limb
present with some degree of calcaneal destruction, and the
salvage. Assessment should include the patient’s functional
diagnosis of osteomyelitis may be made by either clinical
status and the viability of the foot. For example, primary
examination alone or in conjunction with other radio-
limb amputation may be considered in a non-ambulatory,
graphic tests such as plain X-rays or magnetic resonance
bedridden patient or in a patient with severe Charcot
imaging (MRI).
destruction and degeneration and for whom no further
In the presence of an abscess or deep space infection, imme-
reconstructive foot procedures are possible. ‘Poor’ medical
diate drainage, of all infected tissue planes, is mandatory.
condition is not necessarily an indication for primary limb
Incisions should be chosen with consideration of the normal
amputation, considering the higher perioperative morbid-
anatomy, including the compartments, of the foot and the
ity associated with limb amputation. Moreover, in many
need for subsequent secondary foot salvage procedures.
patients, optimisation of the underling medical comor-
These incisions should be placed to allow for dependent and
bidities may be accomplished during treatment of infec-
complete drainage and all necrotic tissue must be debrided. If
tion and while evaluating for ischaemia.
incision and drainage is adequate, placement of drains, for
The assessment for limb salvage should be performed as
example of the Penrose type, is unnecessary, and reliance on
the infection is being treated, as appropriate drainage and
such drains should be avoided. Repeat cultures, both aerobic
antibiotics can dramatically change the appearance and
and anaerobic, of the deep tissues should be obtained. Care is
perceived viability of the foot. If, however, limb salvage is
taken to avoid drainage incisions upwards to the dorsum of
not deemed possible, the patient should undergo formal
the foot. Abscesses in the medial, central or lateral compart-
below-knee or above-knee amputation.
ments should be drained using longitudinal incisions in the
direction of the neurovascular bundles and extending the
entire length of the abscess. The medial and central compart- Ischaemia
ments should be drained through a medial incision and the
lateral compartment should be drained through a lateral The surgeon’s initial evaluation, based on the history and
incision, just above the plantar surface of the forefoot. physical examination, should offer a fairly accurate impres-
Similarly, web space infections may be drained through the sion of the arterial circulation to the foot. That assessment
plantar aspect of the foot. In some instances, an open toe or should be directed at the treatment goal, namely, the restor-
transmetatarsal amputation may be necessary to allow com- ation of maximal, pulsatile arterial perfusion to the foot. The
plete drainage and resection of necrotic tissue. Strict adher- limitations of non-invasive vascular testing in diabetic
ence to textbook amputations may lead to unnecessary soft patients with foot ulceration emphasise the continued
tissue removal and the possible need for a higher amputation importance of thorough bedside evaluation and clinical
in the future. Therefore all viable tissue should be conserved. judgement. To reiterate, the status of the foot pulses is the
With continuing undrained infection, the patient may most important aspect of the physical examination and if
present with an unsalvageable foot and fulminant sepsis, impalpable, occlusive disease is present. As the restoration
 Principles of arterial reconstruction in the diabetic foot 231

of pulsatile flow maximises the chances of healing in the bypass to the popliteal or proximal tibioperoneal arteries
diabetic foot, the absence of foot pulses must be an indica- may restore foot pulses, more distal revascularisation is
tion for contrast arteriography. This has particular rele- often needed to achieve this goal, again owing to the pattern
vance in the clinical setting of tissue loss, poor healing and of occlusive disease in the diabetic patient. Similarly,
gangrene, even if neuropathy may have been the antecedent although excellent results have been reported with peroneal
cause of skin breakdown or ulceration. Arteriography is artery bypass, this artery is not in continuity with the foot
essential in determining and planning the type of arterial vessels and may not achieve the maximal flow required for
reconstruction likely to restore the foot pulses. healing, particularly at forefoot level. We therefore reserve
Concern about contrast induced renal dysfunction in the peroneal artery bypass for those rare circumstances in which
presence of diabetes should not contraindicate high quality there is no dorsalis pedis or posterior tibial artery in conti-
arteriography of the entire distal circulation. Several prospect- nuity with the foot, or when limited venous conduit length
ive studies showing that the incidence of contrast induced mitigates against more distal bypass.
nephropathy is no higher in the diabetic patient free from Autogenous vein grafting to the dorsalis pedis, distal
pre-existing renal disease, especially if it is undertaken with posterior tibial and plantar arteries incorporates our
the judicious use of hydration and renal protective agents.24 knowledge of the anatomical pattern of diabetic vascular
N-acetylcysteine in a 600 mg dose twice daily should be disease, satisfies the fundamental goal of restoration of the
started the day before the arteriogram and continued for 48 foot pulse and provides durable and effective limb salvage.28
hours,25 and intravenous hydration with 0.45 per cent nor- Indeed, extensive experience with arterial reconstruction to
mal saline should be run at a rate of 1 mg/kg per hour, begin- the pedal vessels has established the efficacy, durability and
ning 12 hours prior to the scheduled arteriogram. In selected safety of these procedures and improved limb salvage rates
patients, magnetic resonance angiography (MRA), carbon in the diabetic patient may be directly attributed to the
dioxide angiography or both may be used, either in conjunc- increasing use of pedal bypass.29 The choice of outflow
tion with or in place of contrast arteriography. artery should be based on availability of conduit, the loca-
Whatever preoperative imaging modality is chosen tion of the foot ulcer and the quality of the outflow vessel.
prior to arterial reconstruction, it is mandatory that con- The dorsalis pedis artery is exposed through a longitu-
sideration be given to the pattern of lower extremity vascu- dinal incision 1 cm distal to an imaginary transverse line
lar disease in patients with diabetes and that the complete between the distal malleoli and 1 cm lateral to the extensor
infrapopliteal circulation be incorporated, including that hallucis longus tendon. A continuous-wave Doppler probe
of the foot vessels. It is essential that arteriograms should may be used preoperatively to mark the location of the
not end at mid-tibial level because foot vessels are often artery. Experience has shown that the artery is less likely to
spared in the atherosclerotic occlusive process, even when be calcified at the intermalleolar level. The incision is deep-
tibial arteries are occluded. ened through the fascia and the artery is found just beneath
the inferior margin of the extensor retinaculum. The dor-
salis pedis artery gives off a lateral tarsal branch proximally
and then bifurcates into deep plantar and dorsal metatarsal
PRINCIPLES OF ARTERIAL RECONSTRUCTION arteries distally.
IN THE DIABETIC FOOT In the patient with an ischemic heel ulcer, first consid-
eration should be given to the posterior tibial or plantar
The treatment of ischaemia in the diabetic foot is aimed at arteries if they are patent on preoperative imaging. The dis-
restoring maximal perfusion to the foot and ideally to tal posterior tibial artery is found in the hollow just behind
restore a palpable foot pulse.26 Possible approaches include the medial malleolus, halfway between the bone and Achilles’
the endovascular techniques of angioplasty and stenting, tendon. Again, a Doppler probe may be used to mark
bypass grafting using autogenous or prosthetic grafts, or a its location preoperatively. A longitudinal incision is car-
combination of the two.27 Ultimately, the choice of proced- ried down through the fascia, upon which the artery may
ure should be tailored to the patient’s anatomy, comor- be visualised. If more distal dissection is required, such as
bidities and preoperative assessment, the goal being to for plantar artery exposure and bypass, the distal dense fas-
provide the most durable procedure with the least risk. For cia and flexor retinaculum are divided. This exposes a tri-
example, angioplasty alone may be of benefit in the patient furcation, with a medial calcaneal branch running directly
with an isolated iliac artery stenosis or a focal lesion in the posteriorly toward the heel, followed by the medial and lat-
superficial femoral artery, but it may also be used in com- eral plantar arteries approximately 1 cm beyond the cal-
bination with an infrainguinal bypass in the diabetic caneal branch. If further exposure of the plantar arteries is
patient with multilevel disease. required, the muscle fibres of the abductor hallucis muscle
In most patients, restoration of the foot pulse in the may be divided.
ischaemic diabetic foot is usually achieved by infrainguinal The distal location of the dorsalis pedis and posterior
arterial bypass grafting to an outflow artery in direct con- tibial arteries theoretically necessitates a long venous con-
tinuity with the foot (see Chapter 15). Although proximal duit, which is often not attainable. By using the popliteal or
232 The acute diabetic foot

distal superficial femoral artery as an inflow site, however, a an obvious alternative, several considerations limit its use in
shorter length of vein may be used, with excellent long term the diabetic patient. A contralateral leg vein may not always
patency. This is particularly true in the diabetic patient, be present in this population of patients who often require
again due to the pattern of atherosclerotic disease. This multiple cardiovascular interventions. More importantly,
avoids dissection in the groin and upper thigh, a common diabetes is a strong risk factor for subsequent contralateral
location for wound complications. In addition, the shorter limb bypass, and almost 60 per cent of patients require such
length of saphenous vein obviates the need to extend the a bypass at 3 years. Therefore, our approach has been to use
vein harvest incision into the foot, which is parallel to the arm vein grafts as the first alternative in the absence of ipsi-
one required to expose the paramalleolar and inframalleo- lateral saphenous vein.30 The cephalic, basilic, or upper arm
lar arteries; this avoids the resultant skin bridge which may basilic-cephalic loop vein grafts may be harvested. Once the
occasionally become ischaemic from undue tension. If a vein has been harvested, angioscopic evaluation is crucial, as
skin bridge is created, as with an in situ bypass, care must be many of these patients have undergone multiple venipunc-
taken to create the subcutaneous graft tunnel proximal and tures and cannulations with resultant scarring and web-like
not through the skin bridge. synechiae. Angioscopy allows for detection and correction
The vein graft can be prepared as an in situ, reversed, or of many of these areas, and allows for precise valve lysis
non-reversed graft, without any significant difference in within these thin-walled veins.31
outcome. We therefore advocate a flexible approach, taking Active infection in the foot is not a contraindication to
advantage of the technical strategy best suited to individual dorsalis pedis bypass, as long as the infectious process is
vascular anatomy. The in situ technique minimises size mis- controlled.32 Adequate control implies resolution of cel-
match between the graft and native arteries, eliminates the lulitis, lymphangitis and oedema, especially in areas of pro-
need to completely mobilise the vein and may prevent inad- posed incisions required to expose the distal artery or
vertent twisting of the graft. We have used the in situ tech- saphenous vein. Occasionally, severe circumferential calci-
nique for most infrageniculate distal bypass grafts originating fication of the distal artery may also be encountered.
from the common femoral artery. Although the valves in Strategies include the use of special intraluminal bulb-
the vein graft may be lysed blindly, we prefer to cut the tipped vessel occluders or tourniquet occlusion, with or
valves under direct angioscopic guidance using a flexible without attempts made at endarterectomy or ‘cracking’ the
valvulotome (Fig. 19.5) This also allows for assessment of plaque. Results of bypasses to calcified vessels are compar-
the saphenous vein to detect intraluminal abnormalities able to non-calcified vessels.33
and can help direct endoluminal interventions which upgrade
the quality of the conduit and improve patency.
The absence of an ipsilateral greater saphenous vein is not a
RESULTS
contraindication for pedal bypass, as comparable results may
be attained using arm vein or lesser saphenous vein grafts.
Prosthetic material should seldom, if ever, be used for Several reports have summarised the results of the dorsalis
extreme distal bypass grafting. When ipsilateral saphenous pedis artery bypass, of which the most recent describes a
vein is not available, several alternatives exist for an autogen- decade-long experience of more than 1000 cases.34 In that
ous conduit. Although the contralateral saphenous vein is series, 5-year primary patency rate was 57 per cent with a

Valve Valvulotome Lysed valve

Figure 19.5 Angioscopic evaluation allows for valve identification and precise lysis
 References 233

limb salvage rate of almost 80 per cent, confirming the effi- 7 Akbari CM, LoGerfo FW. Diabetes and peripheral vascular disease.
cacy and durability of these procedures. Additionally, con- J Vasc Surg 1999; 30: 373–84.
cern regarding perioperative morbidity and long term 8 Strandness DE Jr, Priest RE, Gibbons GE. Combined clinical and
outcome in diabetic patients has also been dispelled.35,36 pathologic study of diabetic and nondiabetic peripheral arterial
disease. Diabetes 1964; 13: 366–72.
9 LoGerfo FW, Coffman JD. Vascular and microvascular disease of
the foot in diabetes. N Engl J Med 1984; 311: 1615–19.
Conclusions 10 LoGerfo FW. Vascular disease, matrix abnormalities, and
neuropathy: implications for limb salvage in diabetes mellitus.
These superior results are due the continued application J Vasc Surg 1987; 5: 793–6.
of sound anatomical and pathological principles to a 11 Veves A, Akbari CM, Primavera J, et al. Endothelial dysfunction
coordinated, systematic and aggressive approach in and the expression of endothelial nitric oxide synthetase in
managing the diabetic foot. This, coupled with advances diabetic neuropathy, vascular disease, and foot ulceration.
in technology and more sophisticated techniques in Diabetes 1997; 47: 457–63.
endovascular approaches, has undoubtedly led to even 12 Akbari CM, Saouaf R, Barnhill DF, et al. Endothelium-dependent
higher limb salvage rates among patients with diabetes. vasodilation is impaired in both micro- and macrocirculation
during acute hyperglycemia. J Vasc Surg 1998: 28: 687–94.
13 Morgensen CE, Schmitz A, Christensen CR. Comparative renal
pathophysiology relevant to IDDM and NIDDM patients. Diabetes
Metab Rev 1988; 4: 453–83.
Key references 14 Flynn MD, Tooke JE. Aetiology of diabetic foot ulceration: A role
for the microcirculation? Diabet Med 1992; 8: 320–9.
Akbari CM, Macsata R, Smith BM, Sidawy AN. Overview of the 15 Rayman G, Williams SA, Spencer PD, et al. Impaired
diabetic foot. Semin Vasc Surg 2003; 16: 3–11. microvascular hyperaemic response to minor skin trauma in Type
Akbari CM, LoGerfo FW. Distal bypasses in the diabetic patient. In: I diabetes. BMJ 1986; 292: 1295–8.
Yao JST, Pearce WH (eds). Current Techniques in Vascular 16 Parving HH, Viberti GC, Keen H, et al. Hemodynamic factors in
Surgery. New York: McGraw-Hill, 2001: 285–96. the genesis of diabetic microangiopathy. Metabolism 1983; 32:
Akbari CM, LoGerfo FW. The diabetic foot. In: Wilmore DW, Souba 943–9.
WW, Fink MP, et al. (eds). ACS Surgery: Principles and Practice. 17 Akbari CM, LoGerfo FW. Microvascular changes in the diabetic
New York: WebMD Professional Publishing, 2003: 1–11. foot. In: Veves A, Giurini JM, LoGerfo FW (eds). The Diabetic
LoGerfo FW, Gibbons GW, Pomposelli FB Jr, et al. Trends in the care Foot: Medical and Surgical Management. Totowa, NJ:
of the diabetic foot: Expanded role of arterial reconstruction. Humana Press, 2002: 99–112.
Arch Surg 1992; 127: 617–21. 18 Parkhouse N, LeQueen PM. Impaired neurogenic vascular
Pomposelli FB Jr, Kansal N, Hamdan AD, et al. A decade of response in patients with diabetes and neuropathic foot lesions.
experience with dorsalis pedis artery bypass: analysis of N Engl J Med 1988; 318: 1306–9.
outcome in more than 1000 cases. J Vasc Surg 2003; 37: 19 Menzoian JO, LaMorte WW, Paniszyn CC, et al. Symptomatology
307–15. and anatomic patterns of peripheral vascular disease: differing
impact of smoking and diabetes. Ann Vasc Surg 1989; 3: 224.
20 Akbari CM, LoGerfo FW. The diabetic foot. In: Wilmore DW, Souba
WW, Fink MP, et al. (eds). ACS Surgery: Principles and Practice.
New York: WebMD Professional Publishing, 2003: 1–11.
REFERENCES 21 Akbari CM, LoGerfo FW. Peripheral vascular disease in the person
with diabetes. In: Porte D, Sherwin RS, Baron AD (eds). Ellenberg
1 Reiber GE, Boyko EJ, Smith DG. Lower extremity foot ulcers and and Rifkin’s Diabetes Mellitus, 6th edn. New York: McGraw-Hill,
amputations in diabetes. In: Diabetes in America, 2nd edn, 2003: 845–57.
National Institute of Diabetes and Digestive Kidney Diseases NIH 22 Ballard JL, Eke CC, Bunt TJ, et al. A prospective evaluation of
Publication No. 95–1468. National Diabetes Data Group. transcutaneous oxygen measurements in the management of
Bethesda, MD: National Institutes of Health, 1995, 409–28. diabetic foot problems. J Vasc Surg 1995; 22: 485–92.
2 Nathan DM. Long-term complications of diabetes mellitus. N 23 Akbari CM, Pomposelli FB Jr. Diabetes and diseases of the foot.
Engl J Med 1993; 328: 1676–85. Intern Med 2000; 21: 10–17.
3 Grunfeld C. Diabetic foot ulcers: etiology, treatment, and 24 Solomon R, Werner C, Mann D, et al. Effects of saline, mannitol,
prevention. Adv Intern Med 1992; 37: 103–32. and furosemide to prevent acute decreases in renal function induced
4 Frykberg RG, Kozak GP. The diabetic Charcot foot. In: Kozak GP, by radiocontrast agents. N Engl J Med 1994; 331: 1416–20.
Campbell DR, Frykberg RG, Habershaw GM (eds). Management of 25 Tepel M, van der Giet M, Schwarzfeld C, et al. Prevention of
Diabetic Foot Problems, 2nd edn. Philadelphia, PA: WB Saunders, radiographic-contrast-agent-induced reductions in renal
1995: 88–97. function by acetylcysteine. N Engl J Med 2000; 343: 180–4.
5 Akbari CM, Macsata R, Smith BM, Sidawy AN. Overview of the 26 Akbari CM, LoGerfo FW. Distal bypasses in the diabetic patient.
diabetic foot. Semin Vasc Surg 2003; 16: 3–11. In: Yao JST, Pearce WH (eds). Current Techniques in Vascular
6 Joshi N, Caputo GM, Weitekamp MR, Karchmer AW. Infections in Surgery. New York: McGraw-Hill, 2001: 285–96.
patients with diabetes mellitus. N Engl J Med 1999; 341: 27 Faries PL, Brophy D, LoGerfo FW, et al. Combined iliac
1906–12. angioplasty and infrainguinal revascularization surgery are
234 The acute diabetic foot

effective in diabetic patients with multilevel arterial disease. Ann 32 Tannenbaum GA, Pomposelli FB Jr, Marcaccio EJ, et al. Safety of
Vasc Surg 2001; 15: 67–72. vein bypass grafting to the dorsal pedal artery in diabetic
28 Akbari CM, LoGerfo FW. Saphenous vein bypass to pedal arteries patients with foot infections. J Vasc Surg 1992; 15: 982–90.
in diabetic patients. In: Yao JST, Pearce WH (eds). Techniques in 33 Misare BD, Pomposelli FB Jr, Gibbons GW, et al. Infrapopliteal
Vascular and Endovascular Surgery. Norwalk, CT: Appleton and bypasses to severely calcified outflow arteries: two year results.
Lange, 1998: 227–32. J Vasc Surg 1996; 24: 6–16.
29 LoGerfo FW, Gibbons GW, Pomposelli FB Jr, et al. Trends in the 34 Pomposelli FB Jr, Kansal N, Hamdan AD, et al. A decade of
care of the diabetic foot: Expanded role of arterial experience with dorsalis pedis artery bypass: analysis of outcome
reconstruction. Arch Surg 1992; 127: 617–21. in more than 1000 cases. J Vasc Surg 2003; 37: 307–15.
30 Faries PL, Arora S, Pomposelli FB Jr, et al. The use of arm vein in 35 Hamdan AD, Saltzberg SS, Sheahan M, et al. Lack of association
lower-extremity revascularization: results of 520 procedures of diabetes with increased postoperative mortality and cardiac
performed in eight years. J Vasc Surg 2000; 31: 50–9. morbidity. Arch Surg 2002; 137: 417–21.
31 Akbari CM, LoGerfo FW. Value of arm vein in femoral distal 36 Akbari CM, Pomposelli FB Jr, Gibbons GW, et al. Lower extremity
bypass. In: Yao JST, Pearce WH (eds). Advances in Vascular revascularization in diabetes: late observations. Arch Surg 2000;
Surgery. New York: McGraw-Hill, 2001: 261–9. 135: 452-6.
 SECTION
4
The Acutely Swollen Limb

20. Deep vein thrombosis 237

21. Pulmonary embolism 251

22. Upper limb vein thrombosis 261

This page intentionally left blank
 20
Deep Vein Thrombosis

BO GH EKLOF, CURTIS B KAMIDA

The problem 237 Conservative treatment 239

Predisposing factors 237 The controversy of calf vein thrombosis 240
Risk factors 238 Percutaneous endovascular therapy 240
Pathophysiology of DVT 238 Thrombectomy 243
Diagnosis 238 Case illustrating our current approach in proximal DVT 246
Treatment 239 References 249

THE PROBLEM PREDISPOSING FACTORS

Deep vein thrombosis (DVT) and pulmonary embolism Rudolph Virchow presented his classical triad in 1856 for
(PE) are major health hazards which have a great impact the aetiology of DVT: changes in the blood elements pro-
on healthcare costs. In the general population the rate of ducing a hypercoagulable state, reduced blood flow vel-
DVT is about 160 per 100 000 and the rate of fatal PE is 60 ocity causing stasis and vein wall injury resulting in
per 100 000. Pulmonary embolism is approximately half as endothelial damage. An increased risk of thrombosis is
common as myocardial infarction and about three times as demonstrated in association with an increase in procoagu-
common as cerebrovascular accident (see Chapter 21). lant activity in the plasma, including increases in platelet
About one-half to two-thirds of patients who die from PE count and adhesiveness, changes in the coagulation cascade
do so within 1 hour of onset, too short a period of time to and endogenous fibrinolytic activity. Additionally, defi-
establish the diagnosis and institute specific treatment.1 ciencies of antithrombin III, protein C, protein S and
Prophylactic anticoagulant treatment might have pre- resistance to activated protein C, as well as the presence of
vented the development of DVT and some subsequent a circulating lupus anticoagulant, indicate either primary
deaths. Therefore, primary prevention of DVT in defined or secondary hypercoagulable states. Stasis is generally
risk groups is important. When DVT occurs, the secondary accepted as a major factor causing DVT, and there is evi-
objectives are to prevent: dence that stasis occurs, for example, on the operating
table. It is logical that reduced flow might prolong the con-
• extension of the thrombus and fatal PE tact time of activated platelets and clotting factors with the
• progressive swelling of the leg and increased vein wall, thereby permitting thrombus formation. To date,
compartmental pressure, which can lead to phlegmasia no study has shown stasis alone to be causally related to
caerulea dolens, venous gangrene and limb loss DVT.2 Virchow’s third factor, endothelial damage, has now
• later development of severe post-thrombotic syndrome received more attention with experimental and clinical
(PTS) by preservation of the venous outflow and valve data.3 The data show an intraoperative venodilatation that
function leads to endothelial damage and exposure of the suben-
• chronic pulmonary hypertension. dothelial collagen, which is highly thrombogenic. The lesions
238 Deep vein thrombosis

were infiltrated with leucocytes and platelets, trapping the completely occluded the thrombus begins to adhere to the
blood cells and stimulating fibrin deposition. It is postu- endothelium.
lated that the venodilatation is induced by products of tis- The process of organisation, namely, invasion with
sue injury released at the operative site and which gain granulation tissue and replacement of the fibrin by fibrous
entry into the circulation. tissue, occurs wherever the thrombus is adherent. Where
the thrombus remains loose within the lumen, the poly-
merisation and maturation of the fibrin within the throm-
RISK FACTORS bus cause it to retract. Thrombus retraction and organisation
eventually lead to recanalisation and re-endothelialisa-
tion. This process destroys all the valves in the affected seg-
Surgical patients who sustain major trauma or undergo
ment of vein and is accompanied by enlargement of the
prolonged operative procedures are at risk of developing
collateral venous channels. There is a considerable danger
venous thromboembolic disease. The degree of risk is
of embolism until a non-adherent, non-occlusive throm-
increased by age, obesity, malignancy, prior history of
bus begins to contract. Contraction usually occurs 5–10
thromboembolism, varicose veins, recent operative proced-
days after thrombus formation. If a thrombus has not frag-
ures and thrombophilic states. These factors are further
mented by this stage, it usually adheres to one side of the
modified by general care, including duration of operation,
vein and organisation occurs as if the thrombus were fully
type of anaesthesia, preoperative and postoperative immo-
adherent. In most cases DVT develops in the calf veins,
bility, level of hydration and presence of sepsis.4 In medical
extends proximally and can lead to secondary iliofemoral
patients the risk is increased after acute myocardial infarc-
DVT. The soleal sinusoids and the valves of the calf veins
tion and cerebrovascular accidents, and in immobile gen-
are the common sites of origin of DVT. Primary iliofemoral
eral medical patients.4 Pulmonary embolism is the leading
DVT starts in the iliac vein, mainly owing to iliac vein com-
cause of death in cases following surgical procedures for
pression, and extends distally.
gynaecological cancer and is a major contributor to maternal
New data on the inflammatory response to DVT are
mortality. There is an increased risk of DVT in women tak-
appearing. Downing et al. in Ann Arbor have shown that the
ing oral contraceptives (OCPs) containing 50 g or more
leucocyte adhesion molecule P-selectin activates the leuco-
of oestrogen. The third generation OCPs in combination
cytes emigrating into the venous wall, creating inflammation
with the hypercoagulable state (APC resistance) is reported
which destroys the venous wall and the valves.6 Using an
to increase the risk of DVT 30–40 times. The same risk seems
experimental model See-Tho et al. in Stanford showed that if
to involve women on hormone replacement therapy. With
the thrombus were to be removed early then the inflamma-
this knowledge and with the appearance of new risk factors
tory changes are reversible.7 Caps et al. in Seattle had shown
for the hypercoagulable state, screening of larger groups of
previously that the remaining thrombotic proximal occlu-
people at risk is increasingly favoured. In our own practice
sion in human leg veins will lead to progressive distal valvu-
we perform hypercoagulation screening in all young patients
lar incompetence.8 The lesson from this new information is
with DVT and in all patients with recurrent DVT before we
to remove the thrombus as early and as quickly as possible.
start treatment. Finally, all of the above factors may be
compounded by long periods of immobility, such as in air-
line travel, to produce DVT.
DIAGNOSIS

PATHOPHYSIOLOGY OF DVT A patient, who, over a few hours, develops aching pain and
swelling of one leg accompanied by pleuritic chest pain,
The development of a vein thrombus is well described in a shortness of breath and haemoptysis most certainly has
text on vein diseases.5 The initial platelet cluster on the ves- DVT with PE.
sel wall is followed by coralline thrombus, which is pro- This clinical presentation, however, is quite rare. In
duced, presumably in response to adenosine diphosphate most cases the presentation is atypical and objective meth-
or thromboxane release, by the deposition of more platelets. ods of diagnosis are necessary. In a patient with acute
The thrombus then grows towards the centre of the vessel iliofemoral vein thrombosis (IFVT) and phlegmasia alba
lumen, alternate layers of fibrin and red cells being trapped dolens, the clinical diagnosis is easy and usually correct.
between layers consisting mainly of platelets. As the throm- Given the abundance of differential diagnoses, clinical
bus grows out into the blood stream, it bends in the diagnosis in a patient with pain and swelling in the calf can
direction of the flow and extends across the lumen. The be difficult. The Doppler ultrasound probe has high sensi-
flow beyond it becomes turbulent and gradually decreases. tivity and specificity for mainstem DVT but it is less accurate
A red thrombus, a mixture of fibrin and red cells, extends in calf DVT. The method of choice today is duplex scan-
in the direction of the flow and grows rapidly when ning, particularly using colour flow imaging.9 Phlebography
flow falls to critical levels. When the vein becomes is considered the gold standard for detection of DVT, but
 Conservative treatment 239

this position has been eroded for various reasons: high and peroneal veins, but also the soleal and the gastrocnemial
costs, painful and sometimes impossible access to foot veins can be interrogated. In cases of extension of DVT
veins, contrast induced phlebitis and allergic reactions. into the iliac vein without visualisation of the upper end of
The advantages of duplex scanning are several. It is non- the thrombus, a femoral phlebogram from the contralat-
invasive, easily repeated and permits imaging as well as eral side is performed. This will show the IVC and localise
flow studies. It is applicable to the venous system from the the top of the thrombus in the involved iliac vein. With the
level of the inferior vena cava (IVC) down to the ankle. It advent of duplex scanning, indirect methods such as
distinguishes total from partial venous occlusion, estimates plethysmography and thermography have become obsolete
the age of the thrombus and can be used to follow the nat- and indications for isotope uptake tests are rare. The role of
ural history and recanalisation of the thrombus. the D-dimer test for screening of DVT and PE is still con-
troversial. A negative test may in the future reduce the need
Advantages of duplex scanning for duplex scans to eliminate the diagnosis of DVT and PE
in patients in whom the diagnosis is suspected.
• Non-invasive, easily repeated
• Images system from IVC down to ankle
TREATMENT
• Gives data on both morphology and flow
• Distinguishes total from partial occlusion
• Estimates age of thrombus Standard hospital treatment for patients with acute venous
• Allows monitoring of natural history of thrombosis thromboembolism (VTE) in the USA is bedrest with leg
and recanalisation elevation and anticoagulation using intravenous unfrac-
tionated heparin (UFH) for at least 5 days. Oral warfarin is
By the same token there are a number of disadvantages. started simultaneously and continued for 3 months. Heparin
Admittedly, duplex scanning equipment is expensive but therapy is monitored by activated partial thromboplastin
as it replaces phlebography the investment argument is time (APTT) and the heparin infusion dose is adjusted to
clearly valid. Nevertheless, it has to be acknowledged that maintain the APTT ratio at 1.5–2.5 times control to minimise
duplex scanning remains a time consuming and operator recurrent thrombotic events and risk of bleeding. The lower
dependent process. Technically, the quality of study is lim- level should be reached within 24 hours. Warfarin therapy
ited by gross oedema, the presence of plaster casts, etc. It is is monitored by the international normalised ratio (INR),
sometimes difficult to visualise the upper limits of the the therapeutic range of which is defined by an INR of
thrombus in the abdomen and pelvis and equally calf veins 2.0–3.0.
and duplicated vessels lower down. It is difficult, of course, There are new actors on the scene and they will change
to detect recurrent DVT in a chronically damaged vein. our behaviour in the treatment of VTE: low molecular
weight heparin (LMWH), catheter directed thrombolysis
(CDT) with adjunctive angioplasty and stenting and
Disadvantages of duplex scanning venous thrombectomy (TE) with temporary arteriovenous
fistula (AVF). Low molecular weight heparin has been used
• The equipment is expensive for several years in the treatment of VTE in Australia,
• Studies are time consuming and operator dependent Canada and Europe based on numerous level I clinical
• Technical limitations – gross oedema, plaster casts trials; in 1999 it was approved by the Food and Drug
• Difficulty visualising upper limit of thrombus, calf Administration (FDA) for treatment in the USA. Catheter
veins, duplicated vessels directed thrombolysis is theoretically the method of choice
• Difficult detecting recurrent DVT in chronically for timely dissolution of the acute thrombus and preserv-
damaged vein ing patency of the vein and competency of the valves,
thereby preventing fatal PE (see Chapter 21) and the dis-
abling PTS (for the use of CDT in acute arterial throm-
When DVT is suspected on the basis of history, clinical
boembolic disease see Chapter 16). When there are
examination and Doppler ultrasonography, it is routine to
contraindications to thrombolysis or when it fails, the
proceed to duplex scanning using colour flow imaging.
combination of TE with AVF is a valid alternative.
There are different protocols as to how a proper duplex
scan in a patient with a suspected DVT should be per-
formed. Many vascular laboratories only look at the com-
CONSERVATIVE TREATMENT
mon femoral and popliteal veins, disregarding the iliac and
calf veins. There are studies showing that with such a proto-
col more than 30 per cent of clots will be missed in the The advent of LMWH has opened a new exciting chapter
iliac and calf veins. With improved technology and method- in the treatment of VTE. All LMWHs are easily absorbed
ology using colour and power Doppler, not only the tibial from subcutaneous tissue and they have a much longer
240 Deep vein thrombosis

plasma half-life and better bioavailability at low doses than THE CONTROVERSY OF CALF VEIN
UFH as well as a more predictable dose–response relation- THROMBOSIS
ship. These properties allow LMWHs to be administered
once or twice daily without laboratory monitoring of coagu-
lation factors. Although LMWHs seem to have less platelet The clinical significance of calf vein thrombosis is contro-
interaction, the danger of heparin induced thrombocyto- versial. Isolated calf vein DVT is estimated to be found in
penia is still in the 1 per cent range, and therefore platelets 5 per cent to 33 per cent of all cases of DVT diagnosed.13,14
should be monitored during therapy. Reports on propagation in the absence of treatment vary
The results of 14 major randomised studies10 have from 8 to 28 per cent.15 The direct relationship of isolated
shown that LMWHs are highly effective in the initial treat- calf DVT and PE is controversial (see Chapter 21). In most
ment of established DVT and are superior to standard series where calf DVT and PE were shown to have a high
UFH in the following areas: superior or equal thrombolysis association, cases were selected by identifying those who
on repeat venography, fewer bleeding complications, presented with PE and were then subsequently scanned
reduced mortality at 90 days particularly in patients with and found to have calf DVT. The question remains in these
cancer, reduced recurrence of DVT and no requirement reports as to whether a larger, more proximal clot in the
for monitoring. Hence the potential for treatment in the femoral or popliteal segment could have embolised before
outpatient setting. Partsch of Vienna11 reported on 212 discovery of the calf clot. The development of the post-
patients with IFVT treated with LMWH, compression thrombotic sequelae is also controversial. Passman et al.
bandage and ambulation: PE at admission was 45 per cent, reported a significant incidence of late onset abnormal
new PE after 10 days was 7 per cent and one patient (0.2 venous haemodynamics, but only 5 per cent presented
per cent) suffered a fatal PE. The treatment recommended with swelling and 3 per cent with skin pigmentation and
in an ambulatory patient with IFVT was LMWH given ulcer.16 In our own study we found that the most common
subcutaneously once a day based on body weight, starting site for calf DVT was the peroneal vein (76 per cent) and
warfarin therapy and continuing ambulation with a com- there was no case of anterior tibial vein involvement. In the
pression bandage. untreated group the incidence of propagation into prox-
A MEDLINE search showed that seven studies12 have imal veins was 8 per cent with no case of clinical PE. At
examined cost, and each found the efficacy of LMWHs and three-year follow-up 95 per cent were asymptomatic,
UFH to be comparable but the costs of the LMWHs to be 5 per cent had discoloration of the leg but no ulcers. Based
less. Two studies involved home therapy and reported cost on these results we recommended duplex scan surveillance
savings of up to US$1100 per patient over traditional in- for at least 2 weeks without treatment.17 Passman et al.
hospital treatment. The drawbacks of LMWH treatment recommended for all patients treatment with LMWH
are that we cannot measure the effects on coagulation and followed by oral anticoagulants for 3 months.16 At the dis-
that reversal of anticoagulation is more difficult. If one cussion at the Third Pacific Vascular Symposium on
believes in the concept of early removal of the thrombus, Venous Disease in Hawaii in 1999 the compromise sug-
particularly in extensive DVT, then it is a step backwards to gested was that most patients with calf vein thrombosis
treat all patients with LMWH given that there are no stud- should be treated, while those at risk of complications from
ies on the long term effect of LMWH on the development anticoagulation should be followed with duplex scans.
of the PTS.

PERCUTANEOUS ENDOVASCULAR THERAPY

Advantages of LMWH over UFH
A variety of endovascular techniques can be used in the
• Longer plasma half-life and better bioavailability at treatment of acute DVT, and most of them were initially
low doses developed for the treatment of limb ischaemia. They can be
• More predictable dose–response relationship divided into procedures designed to remove obstructing
• Does not require laboratory monitoring – ideal for thrombus and those which correct any underlying anatom-
outpatient care ical or structural abnormality. The procedures used to
• Less platelet interaction, but it should be checked remove obstructing thrombus can be further divided into
during treatment pharmacological and mechanical methods. In addition, the
• Highly effective thrombolysis at initial treatment of use of IVC filters will be discussed in this section.
established DVT
• Fewer bleeding complications
Clot removal – catheter directed thrombolysis
• Lower mortality at 90 days, especially in cancer
patients
Twenty years after Dotter et al.’s description of CDT in the
• Lower recurrence rate of DVT
treatment of arterial thromboembolism,18 Semba and Dake
 Percutaneous endovascular therapy 241

published their experience in 22 patients with DVT treated Contraindications to the use of any thrombolytic drug
with CDT.19 The rationale for CDT is that by directly include recent surgery, recent major trauma or biopsy,
administering thrombolytic drug into an offending clot, active or recent gastrointestinal bleeding, pregnancy, and
the drug dosage needed would be lower than the amount recent stroke or the presence of intracranial neoplasm. Our
used in systemic lytic therapy. This lower dose would still own experience suggests that almost every other patient
produce a drug concentration immediately adjacent to the has a relative contraindication to thrombolysis. This reflects
clot higher than that achieved by intravenous systemic the underlying condition which predisposed the patient to
therapy (for details on CDT in acute arterial thromboem- developing thrombosis in the first place. Examples include
bolism see Chapter 15). This method would lead to a DVT after joint replacement surgery and immobility sub-
decrease in drug cost, treatment time, and haemorrhagic sequent to extensive trauma.
complications while increasing treatment efficacy.
Until 1998, the drug of choice for thrombolytic therapy
TECHNIQUE
in the peripheral arterial and venous systems was urokinase
(Abbokinase). Because of concerns regarding the risk of Catheter directed thrombolysis involves the insertion of an
transmission of infectious disease inherent in the produc- angiographic catheter into the venous system, positioned
tion of urokinase from human tissue, the FDA banned its immediately adjacent to the clot or embedded within the
sale in mid 1998.20 Consequently, physicians in the USA clot itself. The entry site for venous access does not seem to
turned to alteplase (recombinant tissue plasminogen acti- affect lytic success, as long as the catheter tip can be manipu-
vator (r-tPA)) and reteplase (r-PA), human plasminogen lated into the appropriate position.22 We generally do not
activators currently approved for intravenous use. The two use additional mechanical methods such as pulse spray or
drugs differ from urokinase and from each other in several injection of lytic agent via a power injector. Our concern is
respects, including half-life, fibrin specificity and clot pene- that the time advantage gained is outweighed by the risk of
tration. Both drugs appear to be acceptable alternatives to clot fragmentation, dislodgement and PE formation.
urokinase, but no large clinical trials have been performed For iliofemoral thrombosis which does not involve the
and extensive clinical experience has yet to be obtained. popliteal vein, the usual site of venepuncture and catheter
The half-life of reteplase is between 13 and 16 minutes insertion is the popliteal vein. The latter is approached
while the half-life of alteplase is much shorter, probably using ultrasound guidance or fluoroscopic assistance while
less than 5 minutes. The fibrin specificity of reteplase is less simultaneously injecting X-ray contrast into a pedal vein.23
than that with alteplase. Theoretically, this may allow the With this approach, however, the site for catheter entry is
molecule, less bound as it is with the fibrin on the surface above the major popliteal vein valves and therefore these
of the clot, to penetrate deeper into the thrombus bring valves cannot be directly exposed to the lytic agent. To
about faster clot lysis. This is because the fibrin-bound solve this problem without resorting to retrograde manipu-
plasminogen inside the clot is better activated with r-PA, lation of a catheter through a valve cusp, Cragg has develop-
but this has not been demonstrated convincingly and ed a technique for gaining access into the venous system
remains unconfirmed. In the USA, r-tPA is probably the below the popliteal vein.24 This aspect of the technique is
dominant agent used to replace urokinase although r-PA is important because preservation of the integrity of the
being investigated and used in many centres, including popliteal vein valves has been shown to be a major deter-
ours. Until mid 2000, there were no published reports on minant in reducing the incidence of development of PTS.25
the use of r-PA in catheter directed lytic therapy. As mentioned previously, the most efficacious dose of
Most of the data discussed in the following sections the newer thrombolytic agents is still under investigation.
were obtained using urokinase as the thrombolytic agent. Reteplase doses used currently range from 0.5 to 2.0 units
We feel that the method of CDT therapy is more important per hour. We use 1.0 unit per hour with slight adjustment
than the actual drug used within a reasonable margin of for patient size. A dose greater than 2.0 units per hour
safety, but obviously that has to be confirmed. probably increases the rate of bleeding complications with-
out improving the lytic effect to any significant degree.26
Alteplase doses currently used are in the range of 0.5–1.0 mg
INDICATIONS AND CONTRAINDICATIONS
per hour. Semba found that lysis tends to be faster and
The rapid removal of clot in the acute phase of DVT will more complete when using alteplase compared with uroki-
obviously reduce the risk of development of PE. In add- nase23 but there is always a ‘balancing act’ between speed of
ition, phlegmasia and its symptoms can be alleviated by the lysis and haemorrhagic complications.
removal of the offending thrombus obstructing venous The use of heparin is controversial. Some fear that full
blood return. The long term sequelae of venous insuffi- heparinisation will lead to an increase in bleeding compli-
ciency involve the inadequacy of valve function. Removing cations and consequently tend to use subtherapeutic doses
thrombus before the recanalisation process sets in leads to or none at all.23 A recent study with reteplase found that
valve destruction and so thrombolytic therapy aids the therapeutic levels of heparin do not seem to increase the
maintenance of normal valve competence.21 rate of haemorrhagic complications26 but the authors still
242 Deep vein thrombosis

advocate caution. Heparin should not be added directly to Adjunctive procedures

alteplase solutions because of the formation of precipitates
and also because the degree of dilution seems to affect drug Once the acute clot has been removed, a large percentage
efficacy. Complete details and protocols have been pub- of patients will require angioplasty or stenting of under-
lished and are available for the interested reader.27 lying anatomical or structural abnormalities. In the report
of the National Venous Registry, 99 stent procedures were
RESULTS necessary to reestablish iliac venous outflow in 221
Recent small studies suggest that the efficacy of the newer patients.22 In particular, patients with strictures from radi-
thrombolytic agents will be similar to urokinase.26 A clin- ation therapy, compression by pelvic malignancies and
ically successful thrombolytic result should be expected in those sustained postoperatively will usually show better
80–90 per cent of patients28 when using urokinase. Acute patency with metal stents than with angioplasty alone.36
clot (10 days) or patients with acute symptoms should One entity deserves special mention. The ‘iliac vein
respond to lysis faster and better than those with chronic compression syndrome’, also known as the May–Thurner
disease or chronic disease with acute symptomatology.22 or Cockett–Thomas syndrome, is often encountered after
A health related quality of life survey was undertaken on successful thrombolytic therapy. It was first described in
patients who had lytic therapy as well as those treated with 1851 by Virchow who noted that thrombosis in the pelvic
heparin alone. Improved quality of life scores were obtained veins occurs more frequently on the left side than on the
in the lytic group when compared with those on anticoag- right.37 He postulated that the right common iliac artery
ulant treatment alone. Patients who failed to respond to resting on the left common iliac vein leads to compression
lysis had outcomes similar to those treated with heparin and venous obstruction. This prolonged extrinsic com-
alone. In addition, fewer post-thrombotic symptoms were pression ultimately leads to intraluminal abnormalities
elicited in those treated successfully with thrombolysis.29 within the left iliac vein, consisting of webs and synechiae
This reduction in post-thrombotic complications due to which often do not respond to angioplasty alone and do
preservation of valve function has been seen in other large require stent placement.34
series where urokinase30 was used. The creation of an AVF to help patency after endovascu-
lar intervention is an option whose role is not clearly
defined. Theoretically, an AVF helps to increase venous
Percutaneous endovascular therapies blood flow but in our practice we do not usually place a fis-
tula when the lesion is short, has a reasonably large diameter,
• Clot removal – catheter directed therapy
as in the case of an iliac vein, or when venous inflow is good.
• Clot removal – mechanical thrombectomy devices
Nevertheless, we have created an AVF in patients with poor
• Adjunctive procedures – stents, AV fistulae
venous inflow is caused by multiple webs and synechiae
• IVC filter placement – permanent, temporary
found in a post-thrombotic superficial femoral vein.

Clot removal – mechanical thrombectomy Inferior vena caval filter

devices
The indications for the placement of an inferior vena caval
In those patients not eligible for thrombolytic therapy but filter (VCF) include contraindications to or failure of anti-
who would benefit greatly from clot removal, a number of coagulant therapy, the failure of a previous device to pre-
percutaneous mechanical devices have been developed which vent PE by extension of thrombus through a filter,
are of theoretical benefit.31–33 We have not found any of them recurrent embolism and, in some instances, as a prophy-
to be effective as a ‘stand alone’ method for venous clot lactic measure.38 Indications for prophylactic filter inser-
removal but have used them in the treatment of small or per- tion include a free-floating caval thrombus, DVT with
sistent residual clot after lytic therapy. The lack of directional compromised cardiac or pulmonary function or an
control, noted by others,34 can sometimes be overcome by impending orthopaedic procedure.39 In general, we do not
placing the device through a guiding catheter with an appro- place a permanent filter prior to CDT because of the low
priate bend at the tip but this does necessitate placing a larger rate of complicating PE in patients undergoing this treat-
sheath than normally used. Anatomical and the histological ment.22 The use of a temporary filter, however, might be
integrity of vein valves does not seem to be compromised by reasonable in cases of a large clot burden. These filters are
the use of at least one of the more vigorous devices35 but most not currently approved by the FDA and are not widely
of them have not been tested in this regard. available except in Europe. A recent multicentre registry of
At the time of writing, none of the devices had FDA current practice of temporary VCF insertion has been pub-
approval for the treatment of iliofemoral or native vein lished.40 These filters can be left in place for up to 14 days.
thrombosis. It should be re-emphasised that such approval A total of 188 patients were described: in 53.1 per cent
is awaited for both reteplase and alteplase. the indication was systemic, not catheter directed, lytic
 Thrombectomy 243

therapy; four patients died from PE but none died from THROMBECTOMY
filter induced complications.
Percutaneous techniques can be used to place any of the
several permanent VCFs. Specific techniques differ accord- Historical background
ing to the device and all of the manufacturers supply
detailed instructions for placement. These can be placed The history of TE in the USA is quite interesting. At the
from a femoral, jugular, or in the case of the Simon Nitinol annual meeting of the New England Surgical Society in
filter (Nitinol Medical Technologies, Worburn, MA, USA), Poland Spring, ME, on 28 September 1940, John Homans44
from an antecubital vein approach. There are currently five from Tufts-New England Medical Centre presented the
permanent devices available in the USA possessing rela- paper, ‘Exploration and division of the femoral and iliac
tively equivalent efficacy rates and long term outcomes.41 veins in the treatment of thrombophlebitis of the leg’.
For the critically ill patient who cannot be safely trans- Homans, who was an advocate for division of the femoral
ported from the intensive care unit to the angiography vein to prevent PE, made many suggestions and raised
suite, a VCF can be placed using portable C-arm fluo- questions that are pertinent today: ‘I believe that in the
roscopy.42 Filter insertion using duplex ultrasonography future, instead of at once dividing the various femoral veins,
alone is feasible and has been described.43 This is highly it might be permissible to repair the vein and institute for
operator dependent, patient size and the ability to visualise the next few days a vigorous heparinization. Such a proced-
the IVC through bowel gas being complicating factors. ure is probably, in skilled hands, less hazardous than non-
Congenital anomalies such as a double IVC, a circumaortic operative treatment’. He also advocated division of the
or retroaortic left renal vein and a double renal vein may be femoral or iliac veins to prevent reflux if these vessels were
extremely difficult to visualise with ultrasound. When pos- affected by previous thrombophlebitis. This ‘will always do
sible, the filter should be inserted under direct and adequate good, and never harm’. In this paper Homans discussed
fluoroscopic guidance with a cavogram performed prior to indications for TE with or without ligation of the femoral
filter insertion because the diameter of the IVC will affect vein, the technique, the complications and the importance
the choice of filter type. Currently, only the Bird’s Nest of preventing reflux.
device (Cook Group, Bloomington, IN, USA) can be placed The modern era of TE in the USA started with Howard
in a large IVC or mega cava, i.e. with a diameter greater Mahorner’s paper,45 ‘New management for thrombosis of
than 30 cm. deep veins of extremities’ in 1954 in which he advocated
TE followed by restoration of vein lumen and regional
heparinisation. He presented six patients, five of whom
Indications for placement of VCFs had an excellent result with rapid disappearance of leg
swelling, very little late morbidity and minimal leg oedema.
• Contraindications to anticoagulant therapy There was no instance of PE prior or subsequent to sur-
• Failure of anticoagulant therapy gery. He claimed that this method restores vein function
• Thrombus extension and PE through previously with preservation of the vein lumen and vein valves. In a
inserted device paper in 195746 he reported 16 patients in whom TE had
• Recurrent PE been performed on 14 legs and two arms with excellent
• Prophylactically results in 12, good in two and poor in two patients.
– Free floating thrombus The wave of enthusiasm created by Mahorner’s paper,
– DVT with compromised cardiac function boosted by the report by Haller and Abrams in 1963,47 was
– Impending orthopaedic procedure effectively quelled by Lansing and Davis in 1968 presenting
the results of a five-year follow-up of Haller and Abrams’
patients.48 Haller and Abrams had presented 45 patients
Summary of endovascular techniques with IFVT who underwent TE; of 34 patients with a short
history (10 days) excellent bidirectional flow was estab-
Up until mid-1998, the mainstay of CDT had been the use of lished in 31 patients (91 per cent). At follow-up after an
urokinase as the pharmacological agent. Since its withdrawal average of 18 months, 26 out of these 31 patients (84 per cent)
from the market, alteplase and reteplase have been used for had normal legs and ascending venography permitted by
the purpose of thrombolysis in acute DVT. The ideal dosing 13 patients showed normal patency of the deep venous
regimen and the use of additional heparin is still being inves- system in 11 (85 per cent). Lansing and Davis reported the
tigated. Historical results obtained with urokinase may not five-year follow-up results of those 34 patients with a short
strictly apply to the other thrombolytic agents but the history, and of 17 patients (50 per cent) interviewed
method of drug delivery remains the same. Catheter directed 16 were found to have swelling of the leg requiring stock-
thrombolysis and the use of adjunctive procedures such as ings; one patient had developed an ulcer. Ascending
stenting can relieve venous obstruction, preserve valve func- venography in the supine position in 15 patients showed
tion and hopefully reduce the risk of PE (see Chapter 21). patent veins in most patients but ‘the involved area of
244 Deep vein thrombosis

the deep venous system was found to be incompetent in all Surgical technique
cases and there were no functioning valves’. This study is
flawed because they did not study the iliac vein to prove The first thrombectomy for IFVT was performed by Lawen
patency and their interpretation of incompetence of the in Germany in 1937.52 Surgery today is performed under
valves in the femoral and popliteal veins cannot reliably be intubation anaesthesia, 10 cm water positive end-expiratory
drawn from only an ascending venographic study in the pressure (PEEP) having been added during manipulation
supine patient. of the thrombus, to prevent perioperative PE. The involved
Lansing and Davis’ paper was presented at the annual leg and abdomen are prepared. A longitudinal incision is
meeting of the American Surgical Association in Boston on made in the groin to expose the long saphenous vein (LSV)
18 April 1968, where Hanlon thought it was ‘an important to its confluence with the common femoral vein (CFV) dis-
paper despite the rather melancholy message which it sected up to the inguinal ligament. The superficial femoral
brings, reversing some previous optimistic reports’ and artery 3–4 cm below the femoral bifurcation is prepared for
requested the need ‘to have a series of patients followed construction of the AVF. Further dissection depends upon
objectively with clinical and radiographic data over a long the aetiology of the IFVT.
period of time after two treatment regimens, operative and In primary IFVT with subsequent distal progression of
non-operative’. At the annual meeting of the Southern the thrombus, a longitudinal venotomy is made in the CFV
Surgical Association at Hot Springs, VA, on 8–10 December and a venous Fogarty thrombectomy catheter is passed
1969, Edwards et al.49 presented a paper, ‘Iliofemoral venous through the thrombus into the IVC. The balloon is inflated
thrombosis: reappraisal of thrombectomy’ where he and repeated exercises with the Fogarty catheter are per-
argued with Lansing’s results and concluded that: ‘venous formed until no more thrombotic material is extracted.
TE offers an effective and safe method of restoring flow in With the balloon inflated in the common iliac vein a suc-
the deep venous system; when the thrombus is less than 10 tion catheter is introduced to the level of the internal iliac
days in duration and is of the iliofemoral segment, TE is vein to evacuate thrombus. Backflow is not a reliable sign of
recommended; venograms at operation to determine the clearance as a proximal valve in the external iliac vein may
patency of the deep venous system will aid in complete be present in 25 per cent of cases preventing retrograde flow
removal of the thrombus and give a basis for later compari- in a cleared vein. In contrast, backflow can be excellent
son and evaluation of long-term patency’. In the discus- from the internal iliac vein and its tributaries despite resid-
sion Lansing repeated his findings from the five-year ual occlusion of the common iliac vein. An intraoperative
follow-up and questioned the value of TE. Haller stated completion venogram, therefore, is mandatory. Alterna-
that he was never consulted about the follow-up report. At tively, an angioscope, which enables removal of residual
a recent visit to Louisville, he had studied 17 patients in thrombus, may be used under direct vision. The distal
whom total removal of the thrombus had been possible, thrombus in the leg is removed by manual massage of the
none with any residual oedema. Despite some optimism leg starting at the foot. The Fogarty catheter can sometimes
for TE at this meeting, the impact of Lansing’s report was be advanced gently in retrograde fashion, the aim being to
striking, and few papers have since been published from remove all fresh thrombi from the leg.
the USA but they all showed very good clinical results In IFVT secondary to ascending thrombosis from the
above 75 per cent. calf, the thrombus in the superficial femoral vein is often
Two reports basically abolished TE in the USA: Karp old and adherent to the venous wall, by which stage the
and Wylie’s50 one-page short paper on 10 patients of battle of the valves has been lost. The objective is to restore
whom eight had re-occlusion of the femoral vein before patency and preserve valvular function. If iliac patency is
discharge, and Lansing and Davis’48 skewed paper based on established but the thrombus in the femoral vein is too old
a third of the original material, using questionable to be removed, ligation of the superficial femoral vein is
methods to reach their verdict and arrived at without com- preferable. Recanalisation will otherwise lead to valvular
municating with the original investigators. However, there incompetence and subsequent reflux. In a 13-year follow-up
seems to be renewed interest in TE judging by current after superficial femoral vein ligation Masuda et al.53 found
American textbooks in vascular surgery. This revival is excellent clinical and physiological results without evi-
mainly based on positive reports from Europe. dence of PTS. If normal flow cannot be re-established in
Modern venous reconstructive surgery using valvulo- the superficial femoral vein, we recommend extending the
plasty can give good long term results in primary venous incision distally and exploring the orifices of the deep
disease with severe reflux while the results of vein segment femoral branches where thrombus is usually isolated and
transfer and autologous vein transplantation in secondary, venous flow can be restored with a small calibre Fogarty
or post-thrombotic, venous disease are much less promis- catheter. The superficial femoral vein is ligated. The veno-
ing.51 It is, therefore, important to treat thrombosis of the tomy is closed with continuous suture and an AVF is cre-
leg early and successfully to avoid obstruction of the ated using the long saphenous vein, anastomosing it
venous outflow tract and preserve valvular function in end-to-side to the superficial femoral artery. An intraoper-
order to prevent the development of a severe PTS. ative venogram is performed through a catheter inserted in
 Thrombectomy 245

a branch of the AVF and, if satisfactory, the wound is pulmonary fibrosis, the autopsy excluding fresh PE; an
closed in layers without drainage. If there are signs of iliac autopsy on the other patient showed preoperatively
vein compression, which can occur in about 50 per cent of undetected cirrhosis of the liver and IVC extension of the
left-sided IFVT, we recommend intraoperative angioplasty thrombus, the patient dying in multiorgan failure on the
and stenting. 32nd postoperative day due to intra-abdominal haemor-
If phlegmasia caerulea dolens or venous gangrene is rhage and severe shock caused by over-anticoagulation.
present, we start the operation with fasciotomy of the calf
compartments in order to release pressure and re-establish PULMONARY EMBOLISM
the circulation. If there is extension of the thrombus into
In our experience there were no cases of fatal PE in the
the IVC, the cava is approached transperitoneally through
perioperative period. To avoid this problem it is of utmost
a subcostal incision. The IVC is exposed by reflecting the
importance to demand a preoperative venogram to exclude
ascending colon and duodenum medially (see Chapter 36).
extension of thrombus into the IVC as it may fracture dur-
Depending upon the venographic findings relative to the
ing manipulation with the Fogarty catheter. We do not use
upper end of the thrombus, the IVC is controlled, usually
a separate balloon catheter to occlude the IVC but rou-
just below the renal veins. The IVC is opened and the
tinely ask the anaesthetist to apply PEEP during the proced-
thrombus is removed by massage, especially of the iliac
ure. In a prospective randomised study from Sweden,55 we
venous system. If the iliofemoral segment is involved, the
operation is continued into the groin as described above. found positive perfusion scans at admission in 45 per cent
When laparotomy is contraindicated in patients in poor of all patients; additional defects were seen after 1 and
condition, a caval filter of the Greenfield type can be intro- 4 weeks in the conservatively treated group in 11 per cent
duced prior to TE in order to protect against fatal PE (see and 12 per cent, respectively, and in the thrombectomised
Chapter 21). group in 20 per cent and 0 per cent, respectively. Mavor
Heparin is continued at least 5 days postoperatively and and Galloway56 demonstrated that incomplete clearance of
warfarin, started on the first postoperative day, is contin- the thrombus in the iliac vein increased the incidence of
ued routinely for 6 months. The patient is ambulant the re-thrombosis and PE. In the Swedish series no additional
day after the operation wearing a compression stocking perfusion defects developed after the first postoperative
and is usually discharged on the tenth postoperative day to week following thrombectomy and AVF. Since the AVF
return after 6 weeks for closure of the fistula. The object- effectively prevented re-thrombosis it is reasonable to
assume that the fistula was one reason for the low inci-
ives of a temporary AVF are to increase blood flow in
dence of postoperative PE.
the thrombectomised segment to prevent immediate
Earlier reports of high mortality due to fatal PE have not
re-thrombosis, to allow time for healing of the endothelium
been borne out in our experience. Several reasons may
and to promote the development of collaterals in case of
account for the decreased risk of developing significant
incomplete clearance or immediate re-thrombosis of the
symptomatic and fatal PE with the present technique: care-
iliac segment. A new percutaneous technique for fistula
ful selection of patients; preoperative venographic demon-
closure was developed by Endrys et al. in Kuwait.54 Through
stration of extension of thrombus and its upper limit,
a puncture of the femoral artery on the opposite, surgically
requiring an extended surgical approach if the IVC is
untouched, side a catheter is inserted and positioned at the
involved; use of PEEP during surgery; intraoperative venog-
level of the fistula. Prior to inflation and release of the bal-
raphy or venoscopy to prove clearance of the iliac vein;
loon or coil, an arteriovenogram can be performed to evalu-
creation of an AVF.
ate the patency of the iliac and caval veins, a study which is
also of prognostic value. Despite initial successful surgery
EARLY MORBIDITY AFTER THROMBECTOMY
more than 10 per cent of patients have been shown to have
significant residual stenosis of the iliac vein. A transvenous The rate of early re-thrombosis of the iliac vein varies. In a
percutaneous angioplasty and stenting can be performed retrospective study from Hawaii57 it was 34 per cent in
under the protection of the AVF, which is then closed primary IFVT (8/24) and 18 per cent in secondary IFVT
4 weeks later after repeat arteriovenography. (6/33) without the use of a temporary AVF. In a prospect-
ive randomised Swedish study58 of TE using an AVF, 13
per cent developed early re-thrombosis of the iliac vein
Complications of surgical management despite the temporary AVF. This re-thrombosis rate is cor-
roborated in a series of 555 patients59 in whom 12 per cent
MORTALITY
developed early re-thrombosis. This complication can be
One of the many reasons inducing surgeons to abandon minimised by avoiding surgery in those patients with
TE in the 1960s was the high mortality. Surgery still bears a symptoms of iliac obstruction of more than 7 days.
risk, but given the present perioperative precautions, results A Fogarty catheter to clear the external and the common
have improved. In our series of over 200 patients two died: iliac veins, giving special consideration to the internal iliac
one succumbed from acute respiratory failure due to chronic vein, and a direct caval approach when the IVC is involved
246 Deep vein thrombosis

is recommended. Intraoperative venography or venoscopy ‘Late’ results

will confirm clearance of the iliac venous system. Early and
liberal use of a temporary AVF and fasciotomy in patients There are few studies giving ‘long term’ results of TE with
with phlegmasia cerulea dolens are worthwhile adjunctive AVF. In eight studies comprising 521 patients, with more
procedures. Early ambulation, compression stockings and than two years’ follow-up, ‘clinical success’ is claimed in
carefully monitored postoperative anticoagulation are all 62 per cent.59 In five studies of iliac vein patency involving
very helpful measures. 247 patients with more than two years’ follow-up the figure
was 82 per cent (range 77–88 per cent).59 In five studies on
Guidelines for reducing morbidity and femoro-popliteal valvular competence in 259 patients alto-
mortality after TE gether, with more than two years’ follow-up, competence
was 60 per cent (range 36–84 per cent).59
In the prospective, randomised study from Sweden we
• Careful patient selection
found a highly significant difference in the number of
• Avoiding thrombectomy when symptoms of iliac
asymptomatic patients after 6 months, 42 per cent in the
obstruction exceed 7 days
surgical group and 7 per cent in the conservatively treated
• Venographic demonstration of upper limit of
group.58 At 5 years 37 per cent of the operated patients
thrombus (into IVC?)
were asymptomatic compared with 18 per cent in the con-
• Direct caval approach when IVC involved
servative group.60 At 10 years 54 per cent in the surgical
• Use of PEEP during surgery
group were basically asymptomatic (class 0–2 using the
• Venographic/venoscopic proof of clearance of iliac
CEAP classification) compared with 23 per cent in the con-
thrombus
servative group, but the difference was not a significant.61
• Early and liberal use of an AVF
Iliac vein patency demonstrated by venography at
• Carefully monitored postoperative anticoagulant
6 months was 76 per cent in the surgical group compared
therapy
with 35 per cent in the conservative group,58 a significant dif-
• Compression stockings and early ambulation
ference upheld at 5 years and 10 years with 77 per cent and
• Evacuation of groin haematoma
77 per cent patency, respectively, in the surgical group, and
30 per cent and 47 per cent, respectively, in the conservative
Postoperative bleeding with haematoma formation in group.60,61 Femoro-popliteal valvular competence studied by
the groin was not uncommon despite drainage of the descending venography using Valsalva at 6 months was 52
wound, as full anticoagulation with heparin was continued per cent in the surgical group compared with 26 per cent in
for 5 days after operation. Haematomas should be evacu- the conservatively treated group, which is a significant differ-
ated to avoid compression of the vein and the risk of ence.58 After 5 years, the patients who underwent TE had
thrombosis and infection. significantly lower ambulatory venous pressures, improved
Early re-thrombosis, previously common, is now down venous emptying, as shown by plethysmography, and better
to 12 per cent with the use of a temporary AVF. In cases of calf pump function with less reflux, as measured by foot
immediate re-occlusion of the iliac vein, re-exploration volumetry. Combining the results of all functional tests,
followed by angioplasty and stenting or a femoro-femoral 36 per cent of surgical patients had normal venous function
crossover bypass graft will help to prevent retrograde compared with 11 per cent of the conservatively treated
thrombosis and subsequent valve insufficiency. group but the differences were not statistically significant due
Venous gangrene is very rare, in most cases caused by to the loss of patients at follow-up.60 At 10 years using duplex
underlying malignancy and can be prevented in phlegma- scanning, popliteal reflux was found in 32 per cent of the
sia caerulea dolens by fasciotomy. Groin infection was very surgical group compared with 67 per cent of the conser-
common until we improved preoperative hygiene and vative group. Six patients who had a successful TE 10 years
began using prophylactic antibiotics. We still see lymphatic previously, and in whom the iliac vein was not obstructed
leakage but it usually stops within 2–3 weeks. In two after surgery, were all asymptomatic with patent iliac veins,
patients from the Swedish series the AVF caused high out- and 50 per cent of them had competent popliteal veins.61
put cardiac failure, returning to normal immediately after Successful TE seems to be beneficial in the long term.
the AVF was closed. Both operations had been performed
in elderly patients with previously known compromised car-
diac function; therefore, careful patient selection is essen-
CASE ILLUSTRATING OUR CURRENT
tial. One objection to the AVF is the resulting rise in
APPROACH IN PROXIMAL DVT
venous pressure and swelling of the lower limb. We did not
observe any rise in iliac vein pressure when outflow was
normal. Stenosis of the vein cephalad to the fistula pro- A 13-year-old Chinese-American boy was admitted from
duced higher pressures and stresses the importance of another hospital with superficial thrombophlebitis involv-
clearing the proximal vein. ing the right long saphenous vein extending into the
 Case illustrating our current approach in proximal DVT 247

Figure 20.1 Venogram via the left common femoral vein with
Figure 20.2 Follow-up venogram 27 hours after catheter directed
the tip into the right external iliac vein shows thrombus partially
thrombolysis shows marked improvement, but with residual filling
occluding the common femoral, external and common iliac veins
defect in the common femoral vein; attempts at percutaneous,
extending 2 cm into the inferior vena cava
mechanical destruction of the blood clot failed, and surgical
common femoral vein, the iliac vein and protruding into thrombectomy with a temporary AVF was performed successfully
the IVC. Five months before transfer, he was diagnosed
with systemic lupus erythematosus (SLE) for which he was Catheter directed thrombolysis with urokinase was
treated with steroids and hydroxychloroquine sulphate. started with an infusion of 100 000 IU/hour for 20 hours.
For about 5 weeks before admission he had suffered from A follow-up venogram showed marked improvement with
recurrent episodes of severe pain and swelling of the right a filling defect still present in the common femoral vein,
lower leg. Duplex scanning of the leg at the onset of symp- which remained despite another 7 hours of treatment.
toms did not reveal any evidence of DVT. A repeat duplex Attempts were made to mechanically destroy and aspirate
scan from the groin to the knee 3 days before admission the clot attached to the vein wall (Fig. 20.2). Surgical explor-
was still negative for DVT. The day before referral, how- ation of the common femoral vein on the same day
ever, a further scan revealed thrombus in the right com- revealed organised thrombus, which was removed, the wall
mon femoral vein. Upon admission there was slight of the common femoral vein being severely inflamed and
swelling of the right calf and thigh, no chest symptoms or 1.5 mm thick. The long saphenous vein was also diseased
signs and the platelet count was 33 000. A repeat duplex with evidence of active thrombophlebitis. An AVF was cre-
scan showed thrombosis of the right common femoral vein ated anastomosing the long saphenous vein end-to-side to
occluding the long saphenous vein and extending up the the superficial femoral artery. An intraoperative venogram
iliac vein into the IVC but the deep veins of the leg were showed complete removal of the thrombus and a patent
free from thrombus. A venogram via the left common common femoral vein and iliac vein.
femoral vein revealed no thrombus on the left side, but After some postoperative problems the patient rapidly
confirmed complete occlusion the right common femoral improved and still remains on anticoagulation and other
vein and iliac vein extending 2 cm into the IVC (Fig. 20.1). medication for his SLE. Three years later he has no swelling
248 Deep vein thrombosis

Patient with known iliofemoral deep vein thrombosis

Evaluate activity level, risk, and life expectancy

Poor Reasonable

Anticoagulation therapy, leg elevation Yes Iliocavogram

Progressive symptoms (oedema, pain)? Contraindication for thrombolysis?

No
Yes No

Free-floating caval thrombosis?

Venous thrombectomy

No Yes

Fails Vena caval filter

Catheter directed thrombolysis

Succeeds

Treat iliac vein stenosis if present

Compartment syndrome?

No Yes

Chronic anticoagulation therapy

Fasciotomy
Compression therapy

Figure 20.3 Algorithm for managing patients with proved iliofemoral vein thrombosis (IFVT) (redrawn from Comerota AJ. Iliofemoral deep
vein thrombosis. In: Cronenwett JL (ed). Decision Making in Vascular Surgery. Philadelphia: WB Saunders, 2001: 282–5)

or pain in his leg. Repeated duplex scans have shown a

CDT with or without adjunctive procedures such as
patent deep venous system without any reflux. The AVF
angioplasty and stenting. When there are contraindica-
closed spontaneously after about 1 year. Three years after
tions to thrombolysis or when it fails, TE with a tempor-
the extensive DVT he is fully active and playing basketball.
ary AVF is a valid alternative. Both interventions should
The first line of treatment was CDT and although only
be followed by anticoagulation. These aggressive inter-
partially successful, immediate TE with AVF resulted in a
ventions are not justified in chronically ill, bedridden,
patent and competent deep venous system and, to date,
high risk or aged patients, or those with serious inter-
had prevented the potential development of a severe PTS.
current disease and/or limited life expectancy. In this
An algorithm (Fig. 20.3) provides pathways of manage-
group of patients these interventions can only be justi-
ment of a patient with proved IFVT.
fied for limb salvage in phlegmasia caerulea dolens, in
which conservative treatment does not prevent the
Conclusions development of an acute compartment syndrome and
venous gangrene. The role of LMWH in acute iliofemoral
Early and quick removal of the thrombus is indicated to DVT in preventing the PTS should be tested in a prospect-
avoid the late PTS in active patients with acute ive randomised trial comparing it with the interventional
iliofemoral DVT. The first line of treatment should be approach.
 References 249

initial treatment of deep venous thrombosis: an updated

Key references meta-analysis. Drugs 1996; 52(suppl 7): 30–7.
13 Markel A, Manzo RA, Bergelin RO. Strandness DE. Pattern and
Eklof B, Kamida CB, Kistner RL, Masuda EM. Contemporary distribution of thrombi in acute venous thrombosis. Arch Surg
treatment of iliofemoral deep vein thrombosis. Perspect Vasc 1992; 127: 305–9.
Surg 1999; 11: 1–27. 14 Mattos MA, Melendres G, Sumner DS, et al. Prevalence and
Hull RD, Pineo GF. Prophylaxis of deep venous thrombosis and distribution of calf vein thrombosis in patients with symptomatic
pulmonary embolism: current recommendations. Med Clin deep venous thrombosis: a color-flow duplex study. J Vasc Surg
North Am 1998; 82: 477–93. 1996; 24: 738–44.
Kalodiki E, Nicolaides AN. Low-molecular-weight heparin in the 15 Lohr JM, James KV, Deshmukh RM, Hasselfeld KA. Calf vein
treatment of deep vein thrombosis. In: Pifarre R (ed). New thrombi are not a benign finding. Am J Surg 1995; 170:
Anticoagulants for the Cardiovascular Patient. Philadelphia: 86–90.
Hanley & Belfus, Inc, 1997: 609–19. 16 Passman MA, Moneta GL, Taylor LM, et al. Pulmonary embolism
Mewissen MW, Seabrook GR, Meissner MN, et al. Catheter-directed is associated with the combination of isolated calf vein
thrombolysis for lower extremity deep venous thrombosis: thrombosis and respiratory symptoms. J Vasc Surg 1997; 25:
Report of a national multi-center registry. Radiology 1999; 39–45.
211: 39–49. 17 Masuda EM, Kessler DM, Kistner RL, et al. The natural history of
Semba CP, Dake MD. Iliofemoral deep venous thrombosis: calf vein thrombosis: lysis of thrombi and development of reflux.
aggressive therapy with catheter directed thrombolysis. J Vasc Surg 1998; 28: 67–74.
Radiology 1994; 191: 487–94. 18 Dotter C J, Rosch J, Seaqman AJ. Selective clot lysis with
low-dose streptokinase. Radiology 1974; 111: 31–7.
19 Semba CP, Dake MD. Iliofemoral deep venous thrombosis:
Aggressive therapy with catheter directed thrombolysis.
REFERENCES
Radiology 1994; 191: 487–94.
20 Hartnell GG, Gates J. The case of Abbokinase and the FDA: the
1 Miller G. Pulmonary embolism. In: Browse NL, Burnand KG, events leading to the suspension of Abbokinase supplies in the
Lea TM (eds). Diseases of the Veins: Pathology, Diagnosis and United States. J Vasc Interv Radiol 2000; 11: 841–7.
Treatment. London: Edward Arnold, 1988:558–80. 21 Eklof B, Kamida CB, Kistner RL, Masuda EM. Contemporary
2 Comerota A. Causes of deep vein thrombosis. Proceedings of the treatment of iliofemoral deep vein thrombosis. Perspect Vasc
Straub Pacific Health Foundation 1993; 57: 2. Surg 1999; 11: 1–27.
3 Comerota A, Stewart GJ. Operative venous dilation and its 22 Mewissen MW, Seabrook GR, Meissner MN, et al. Catheter-
relation to postoperative deep venous thrombosis. In: Goldhaber SZ directed thrombolysis for lower extremity deep venous
(ed). Prevention of Venous Thromboembolism. New York: Dekker thrombosis: report of a national multi-center registry. Radiology
1993: 25–50. 1999; 211: 39–49.
4 Nicolaides AN, Arcelus J, Belcaro G, et al. Prevention of venous 23 Semba CP. Venous thrombolysis in the post-urokinase era. Tech
thromboembolism. European consensus statement, 1–5 Vasc Interv Radiol 2000; 3: 2–11.
November 1991, developed at Oakley Court Hotel, Windsor, UK. 24 Cragg AH. Lower extremity deep venous thrombolysis: a new
Int Angiol 1992; 11: 151–9. approach to obtaining access. J Vasc Interv Radiol 1996; 7:
5 Browse NL, Burnand KG, Lea TM (eds). Diseases of the Veins: 283–6.
Pathology, Diagnosis and Treatment. London: Edward Arnold, 25 Shull KC, Nicolaides AN, Fernandes E, et al. Significance of
1988: 603–25. popliteal reflux in relation to ambulatory venous pressure and
6 Downing LJ, Wakefield TW, Strieter RM, et al. Anti-P-selectin ulceration. Arch Surg 1979; 114: 1304–6.
antibody decreases inflammation and thrombus formation in 26 Ouriel K, Katzen B, Mewissen MW, et al. Reteplase in the
venous thrombosis. J Vasc Surg 1997; 25: 816–28. treatment of peripheral arterial and venous occlusions. A pilot
7 See-Tho K, Harris EJJ. Thrombosis with outflow obstruction study. J Vasc Interv Radiol 2000; 11: 849–54.
delays thrombolysis and results in chronic wall thickening of rat 27 Valji K. Evolving strategies for thrombolytic therapy of
veins. J Vasc Surg 1998; 28: 115–22. peripheral vascular occlusion. J Vasc Interv Radiol 2000; 11:
8 Caps MT, Manzo RA, Bergelin RO, et al. Venous valvular reflux in 411–29.
veins not involved at the time of acute deep vein thrombosis. 28 Grossman C, McPherson S. Safety and efficacy of catheter-
J Vasc Surg 1995; 22: 524–31. directed thrombolysis for iliofemoral venous thrombosis. AJR Am
9 Douglas MG, Sumner DS. Duplex scanning for deep vein J Radiol 1999; 172: 667–72.
thrombosis: Has it replaced both phlebography and noninvasive 29 Comerota AJ, Throm RC, Mathias SD, et al. Catheter-directed
testing? Semin Vasc Surg 1996; 9: 3–12. thrombolysis for iliofemoral deep venous thrombosis improves
10 Kalodiki E, Nicolaides AN. Low-molecular-weight heparin in the health-related quality of life. J Vasc Surg 2000; 32: 130–7.
treatment of deep vein thrombosis. In: Pifarre R (ed). New 30 Bjarnason H, Kruse JR, Asinger DA, et al. Iliofemoral deep vein
Anticoagulants for the Cardiovascular Patient. Philadelphia: thrombosis: Safety and efficacy during five years of catheter
Hanley & Belfus, Inc, 1997: 609–19. directed thrombolytic therapy. J Vasc Interv Radiol 1997; 8:
11 Partsch H. Acute iliofemoral DVT: results of ambulatory 405–18.
treatment. Vasc Surg 1997; 31: 307–9. 31 Sharafuddin MJA, Hicks ME. Current status of percutaneous
12 Leizorovicz A. Comparison of the efficacy and safety of low mechanical thrombectomy. Part I. General principles. J Vasc
molecular weight heparins and unfractionated heparin in the Interv Radiol 1997; 8: 911–21.
250 Deep vein thrombosis

32 Sharafuddin MJA, Hicks ME. Current status of percutaneous 46 Mahorner H, Castleberry JW, Coleman WO. Attempts to restore
mechanical thrombectomy. Part II. Devices and mechanisms of function in major veins which are the site of massive thrombosis.
action. J Vasc Interv Radiol 1998; 9: 15–31. Ann Surg 1957; 146: 510–22.
33 Sharafuddin MJA, Hicks ME. Current status of percutaneous 47 Haller JAJ, Abrams BL. Use of thrombectomy in the treatment of
mechanical thrombectomy. Part III. Present and future acute iliofemoral venous thrombosis in forty-five patients. Ann
applications. J Vasc Interv Radiol 1998; 9: 209–34. Surg 1963; 158: 561–9.
34 O’Sullivan GJ, Semba CP, Bittner CA, et al. Endovascular 48 Lansing AM, Davis WM. Five-year follow-up study of iliofemoral
management of iliac vein compression (May-Thurner) syndrome. venous thrombectomy. Ann Surg 1968; 168: 620–8.
J Vasc Interv Radiol 2000; 11: 823–36. 49 Edwards WH, Sawyers JL, Foster JH. Iliofemoral venous
35 Sharafuddin MJA, Gu X, Urness M, Amplatz K. Lack of acute thrombosis: reappraisal of thrombectomy. Ann Surg 1970; 171:
injury to venous valves by the Amplatz Thrombectomy Device 961–70.
during experimental antegrade venous thrombectomy. J Vasc 50 Karp RB, Wylie EJ. Recurrent thrombosis after iliofemoral venous
Interv Radiol 1998; 9(suppl): 203. thrombectomy. Surg Forum 1966; 17: 147.
36 Nazarian GK, Bjarnason H, Dietz CA Jr, et al. Iliofemoral venous 51 Kistner RL. Valve repair and segment transposition in primary
stenoses: Effectiveness of treatment with metallic endovascular valvular insufficiency. In: Bergan JJ, Yao JST (eds). Venous
stents. Radiology 1996; 200: 193–9. Disorders. Philadelphia: WB Saunders, 1991: 261–72.
37 Virchow R. Uber die Erweiterung kleiner Gefasse. Arch Path Anat 52 Lawen A. Uber Thrombektomie bei Venenthrombose und
1851; 3: 427. Arteriospamus. Zentralbl Chir 1937; 64: 961–8.
38 Greenfield LJ, Rutherford RB. Recommended reporting standards 53 Masuda EM, Kistner RL, Ferris EB. Long-term effects of
for vena caval filter placement and patient follow-up. J Vasc superficial femoral vein ligation: thirteen-year follow-up. J Vasc
Interv Radiol 1999; 10: 1013–19; Erratum in: J Vasc Interv Radiol Surg 1992; 16: 741–9.
1999; 10: 1270. 54 Endrys J, Eklof B, Neglen P, et al. Percutaneous balloon occlusion
39 Hull RD, Pineo GF. Prophylaxis of deep venous thrombosis and of surgical arteriovenous fistulae following venous
pulmonary embolism: current recommendations. Med Clin North thrombectomy. Cardiovasc Intervent Radiol 1989; 12: 226–9.
Am 1998; 82: 477–93. 55 Plate G, Ohlin P, Eklof B. Pulmonary embolism in acute
40 Lorch H, Welger D, Wagner V, et al. Current practice of iliofemoral venous thrombosis. Br J Surg 1985; 72: 912–15.
temporary vena cava filter insertion: a multi-center registry. 56 Mavor GE, Galloway JMD. The iliofemoral venous segment as a
J Vasc Interv Radiol 2000; 11: 83–8. source of pulmonary emboli. Lancet 1967; i: 871.
41 Athanasoulis C, Kaufman J, Halpern E, et al. Inferior vena caval 57 Kistner RL, Sparkuhl MD. Surgery in acute and chronic venous
filters: review of a 26-year, single-center clinical experience. disease. Surgery 1979; 85: 31–43.
Radiology 2000; 216: 54–66. 58 Plate G, Einarsson E, Ohlin P, et al. Thrombectomy with
42 Rose SC, Kinney TB, Valji K, Winchell RJ. Placement of inferior temporary arteriovenous fistula: the treatment of choice in acute
vena caval filters in the intensive care unit. J Vasc Interv Radiol iliofemoral venous thrombosis. J Vasc Surg 1984; 1: 867–76.
1997; 8: 61–4. 59 Eklof B, Kistner RL. Is there a role for thrombectomy in
43 Neuzil DF, Garrard CL, Berkman RA, et al. Duplex-directed vena iliofemoral venous thrombosis? Semin Vasc Surg 1996; 9:
caval filter placement: report of initial experience. Surgery 1998; 34–45.
123: 470–4. 60 Plate G, Akesson H, Einarsson E, et al. Long-term results of
44 Homans J. Exploration and division of the femoral and iliac veins venous thrombectomy combined with a temporary arterio-
in the treatment of thrombophlebitis of the leg. JAMA 1941; venous fistula. Eur J Vasc Surg 1990; 4: 483–9.
224: 179–86. 61 Plate G, Eklof B, Norgren L, et al. Venous thrombectomy for
45 Mahorner H. New management for thrombosis of deep veins of iliofemoral vein thrombosis. 10-year results of a prospective
extremities. Am Surgeon 1954; 20: 487–98. randomized study. Eur J Vasc Endovasc Surg 1997; 14: 367–74.
 21
Pulmonary Embolism

ANH NGUYEN, ELAINE IMOTO, BO GH EKLOF

The problem 251 Investigations 254

Clinical presentation 251 Further investigations 255
Risk factors 251 Treatment 256
Clinical signs 253 References 259
Differential diagnosis 254

THE PROBLEM ‘moderate’ with isolated dyspnoea syndrome and to ‘severe’

with circulatory collapse.4 In assessing the patient with an
acute presentation, the clinician must be able to distinguish
Pulmonary embolism (PE) is a potentially fatal complica- between small and massive PE, as the latter is life threaten-
tion of deep venous thrombosis (DVT). While common, it ing and must be diagnosed and treated promptly.5
remains underdiagnosed. The antemortem diagnosis of Some or none of the following symptoms may be present
PE was made in only 30 per cent of patients confirmed as in the patient with a submassive or massive PE presenting
having had PE at autopsy.1 Eighty per cent of patients with with circulatory collapse at risk of death in the next three
angio-proven PE had bilateral leg venography which hours:4,6 dyspnoea, pleuritic pain, haemoptysis, wheeze,
showed evidence of DVT in 82 per cent, symptomatic in anterior chest pain, cough, and leg swelling and pain. The
only 42 per cent.2,3 classic triad of symptoms of dyspnoea, haemoptysis and
The astute physician must combine a high clinical suspi- pleuritic chest pain occurred in only 28 per cent of the
cion for PE, based on presenting features and appropriate 160 patients in the Urokinase Pulmonary Embolism Trial
risk factors, with timely investigations and empirical therapy. (UPET) and in only 20 per cent of patients with massive
The majority of PEs are caused by portions of thrombi origi- embolism.7 None of the five patients in the Prospective
nating in the upper leg and pelvic veins breaking off and trav- Investigation of Pulmonary Embolism Diagnosis (PIOPED)
elling to the lungs. The focus of this chapter therefore will be study with circulatory collapse from PE, loss of conscious-
the diagnosis and therapy of thombotic emboli. At all times ness or blood pressure 80 mmHg, had haemoptysis, pleur-
less frequent causes of PE such as upper extremity thrombi, itic pain, palpitations or angina-like pain.4
air, bone marrow, arthroplasty cement, amniotic fluid, talc,
fat, septic and tumour emboli should be kept in mind.
Symptoms of PE
CLINICAL PRESENTATION • Dyspnoea, wheeze
• Pleuritic pain
The clinical presentation is quite diverse, depending on the • Cough, haemoptysis
size of the vessel occluded and the patient’s cardiorespira- • Leg swelling and pain
tory reserve. A normal person may tolerate total occlusion
of a unilateral pulmonary artery with minimal symptoms.
A patient with pre-existing cardiopulmonary compromise
RISK FACTORS
may die from the same occlusion. Findings in PE can be
understood in terms of the severity of PE as it increases As symptoms and signs vary between presentations, the
from ‘mild’ with the pulmonary infarction syndrome, to presence of risk factor(s) will add to the suspicion of PE and
252 Pulmonary embolism

(a) (b)

(c) (d)

Figure 21.1 A 72-year-old male developed pain and swelling of the right leg with shortness of breath 3 months after a lengthy flight from
Africa. A duplex scan showed extensive deep vein thrombosis (DVT) in the right leg. (a) Computed tomography (CT) angiogram confirming
right pulmonary artery embolism with the doughnut sign (arrow) despite having been on warfarin (Coumadin) treatment for an aortic
valve prosthesis. After adding low molecular weight heparin an inferior vena cava (IVC) filter was placed below the renal veins. (b) A
ventilation/perfusion (V/Q) scan 5 days later confirming pulmonary embolism (PE) of the right lung. He was discharged 2 days thereafter to
continue on warfarin. (c) A chest radiograph on readmission following a syncopal attack and hypoxia on the day after discharge showing
that the heart was not enlarged and pulmonary vascularity was normal. Several vague ‘nodules’ were seen projecting over the lung bases
suspicious of metastatic disease. (d) A repeat V/Q scan showing increased embolism to the lungs. Lytic therapy in the form of tissue
plasminogen activator (tPA) was commenced in the intensive care unit (ICU) but stopped in 24 hours because of gastrointestinal bleeding. He
also developed severe swelling of the left leg and another duplex scan showed DVT in both legs extending from the tibial veins all the way up
to the IVC filter. Following a fresh syncopal attack 2 days after readmission a further V/Q scan revealed more embolism to the right lung.

help to decide whether further investigation is warranted.8,9 therapy (HRT), resistance to activated protein C, antiphos-
They all revolve around the thrombotic triad elucidated by pholipid antibodies and mild-to-moderate hyperhomocys-
Virchow in 1856: local trauma to the vessel wall, hypercoag- teinaemia.11 In the Nurses’ Health Study,12 nurses 60 years
ulability and stasis. Stasis and local augmentation of the or older in the highest quintile of body mass index had the
coagulation process are the most important factors in highest rate of PE. Heavy cigarette smoking and high blood
thrombus formation making recognition of acquired risk pressure were identified as independent risk factors.
factors (see Chapter 2) critical for therapy.10 Acquired risk Genetic predisposition appears to explain one-fifth of
factors include recent major surgery or trauma, immobilisa- cases of PE. The family history is critical if the patient is
tion, as for instance on long flights, advancing age, malig- young and if there is a possibility of lifetime anticoagula-
nancy (Fig. 21.1), previous thrombosis, pregnancy and tion and avoidance of additional acquired risk factors.11
puerperium, use of contraceptives or hormone replacement The most common inherited risk factors are activated
 Clinical signs 253

Figure 21.1 (Continued)

(e) A CT angiogram 2 days later
showing extension of IVC thrombus
2 cm above the filter ending about
1 cm below the renal vein. The IVC was
ligated after removal of thrombus
below the renal veins. He was
discharged on warfarin, recovered and
made it to the golf course. He died
suddenly 3 months later with
generalised metastatic cancer, the
primary site still unknown

(e)

protein C resistance due to a mutation in the factor V gene,

– Homozygous C677T mutation in the
namely, factor V Leiden, with a 5 per cent prevalence in the
methylene-tetrahydrofolate reductase gene
Caucasian population, G20210A mutation in the pro-
– Homozygous homocysteinuria
thrombin (factor II) gene and homozygous C677T muta-
tion in the methylene-tetrahydrofolate reductase gene.13
Antithrombin, protein C and protein S deficiencies are
rare while homozygous homocysteinuria is very rare. CLINICAL SIGNS

Signs of PE are often absent, and, even when present tend to

be non-specific. The concurrent presence of DVT (30 per
Risk factors cent) must be investigated as supportive evidence for PE and
as an indication for anticoagulation. The most common
• Acquired signs of PE14,15 include cyanosis, tachypnoea, the use of
– Recent major surgery/trauma accessory muscles of respiration, tachycardia, hypotension,
– Prolonged immobilisation elevated jugular venous pressure (JVP), left parasternal
– Advancing age heave, chest wall tenderness, gallop rhythm on cardiac aus-
– Malignancy cultation, pleural rub, crackles, thrombophlebitis, fever and
– Previous thrombosis sweating. Signs of right heart insufficiency should prompt
– Pregnancy/puerperium early action. Elevated JVP, low blood pressure, gallop
– Contraceptives/HRT rhythm and cyanosis all suggest pulmonary hypertension as
– Antiphospholipid antibodies a result of massive PE.
• Inherited
– Antithrombin deficiency
– Protein C deficiency Signs of PE
– Protein S deficiency
– Factor V Leiden (Activated protein C resistance) • Cyanosis, fever, sweating
– G20210A mutation prothrombin (F II) gene • Tachypnoea, use of accessory muscles of respiration
254 Pulmonary embolism

wedge-shaped defect or the Hampton hump. The most

• Tachycardia, hypotension, raised JVP
common X-ray finding is simple cardiomegaly.18
• Left parasternal heave, gallop rhythm
The ECG will help to exclude other diagnoses, especially
• Chest wall tenderness, pleural rub, crackles
acute MI and tachyarrhythmias. As with chest films, ECGs
• Thrombophlebitis
may suggest PE but they are usually not diagnostic.19
The classic ECG changes of S1Q3T3 are present in only 25
per cent of patients with PE. The ECG will frequently reveal
DIFFERENTIAL DIAGNOSIS non-specific tachycardia or atrial fibrillation. Evidence of
right ventricular insufficiency may exist in the form of right
bundle branch block, P pulmonale, right axis deviation, and
The signs and symptoms of PE are often non-specific
T wave inversion in leads V1–3.5 These findings, however,
but they should be considered in a differential diagnosis.16
are present only after significant obstruction (⬃ 50
The possible diagnoses include acute myocardial infarc-
per cent) of the pulmonary vasculature has occurred in an
tion (MI), acute pulmonary oedema, pneumonia, asthma,
otherwise healthy patient, but they are more pronounced in
pneumothorax, pleurisy, tachyarrhythmia, pericardial
patients with previous cardiopulmonary disease.20
tamponade, musculoskeletal pain, rib fracture, lobar col-
lapse, lung cancer, primary pulmonary hypertension, cos-
Plasma D-dimer
tochondritis, aortic dissection or anxiety. Overdiagnosis is
as likely as underdiagnosis. A detailed history, whenever
This is a sensitive but a non-specific test of venous throm-
possible, physical examination and selective testing will
boembolic disease.21–23 D-dimers are crosslinks which are
improve diagnostic accuracy.
cleaved and then released into the circulation by the action of
plasmin during fibrinolysis. This event occurs within one
hour of thrombus formation. D-dimer values ⬍500 ␮g/mL
Differential diagnoses of PE reliably exclude PE in patients with intermediate and low prob-
ability lung scans.21,24,25 They do not, however, exclude PE in
• Cardiovascular – acute MI, acute pulmonary
high probability scans. D-dimers also have no positive predic-
oedema, tachyarrhythmia, pericardial tamponade,
tive value as a variety of reasons cause them to rise e.g. recent
aortic dissection
surgery, trauma, MI, congestive heart failure, pneumonia,
• Pulmonary – pneumonia, asthma, pneumothorax,
cancer and sepsis. The D-dimer assay is useful in patients who
pleurisy, lobar collapse, lung cancer, primary pul-
present to the emergency department or physician’s office
monary hypertension
without other systemic illnesses. If clinical likelihood is low, a
• Chest wall – musculoskeletal pain, rib fracture,
D-dimer assay and duplex scan of the leg veins may be useful
costochondritis
in ruling out PE/DVT. Because it is not 100 per cent sensitive,
• Anxiety
D-dimer cannot be used to rule out the diagnosis in patients
in whom there is a high clinical suspicion of PE.26,27
INVESTIGATIONS Other laboratory tests are helpful in delineating disease
processes such as elevated cardiac enzymes in MI and an ele-
vated leucocyte count in infection. Leucocytosis, however,
Laboratory tests for the diagnosis of PE should include an can also be present in a stressed state with lung infarction.
arterial blood gas (ABG), chest X-ray (CXR), electrocardi- The decision to investigate further should be based on
ography (ECG) and D-dimer. Arterial blood gases are clinical suspicion as to whether a PE has occurred. The most
checked if the patient has hypoxia detected on pulse oxime- common scenario is the patient with a risk factor who
try or if the patient is tachypnoeic. Increased minute venti- becomes breathless suddenly and has a normal CXR and per-
lation is a sensitive indicator of larger PE. A respiratory haps mild hypoxia, without any obvious cause. Assessment
alkalosis may be detected and hypoxaemia can be con- of the patient allows the clinical probability of a diagnosis of
firmed. In the case of a small embolus, however, these find- PE to be expressed as high, intermediate, low or unlikely,
ings may not be present, so that the absence of findings on depending on the predictive value ascribed to a particular
ABG does not exclude the possibility of an embolus.5,17 presentation.28
Frequently obtained, a CXR helps to exclude other diag-
noses (Fig. 21.1c). Occasional, non-specific CXR findings
in PE include an elevated hemidiaphragm, subsegmental
Clinical probability of diagnosis of
atelectasis, small pleural effusions and patchy infiltrates due
PE (per cent chance)
to associated oedema, haemorrhage or infarction. Large
emboli can cause dilatation of a proximal pulmonary artery • High* (80–100 per cent)
– Presence of a risk factor
with oligaemia in the distant lung field known as the
– CXR, ABG or D-dimer consistent finding; lack
Westermark sign. More often than not, the CXR in PE is
of evidence for another explanation
more likely to appear normal than to show the classic
 Further investigations 255

per cent of PE.35–38 Spiral CT and V/Q scans are comparable

• Intermediate* (20–80 per cent)
as first line testing for PE.24 Multidetector CT (MDCT) pro-
– Neither high nor low probability
vides the next step in diagnostic accuracy with the ability to
• Low (1–19 per cent)
obtain thin sections 2–3 mm at narrow intervals of 2–3 mm
– No risk factor
– Clinical symptoms and signs explainable by and yet perform the study within a 10–15 second breath-
other causes hold. Although no article with specific statistics has been
– ABG/CXR explainable by other causes published, the advantages are clear-cut for PE and have been
– D-dimer positive addressed by studies on helical CT.39 An additional advan-
tage of MDCT is the ability to combine a CT venogram of
• Unlikely (1 per cent)
the deep venous system with the CT pulmonary angiogram,
– Low clinical probability
– D-dimer negative two tests being performed together in a timely manner for
* diagnosis of PE with a good look at the inferior vena cava
Further investigations should be undertaken in high and
(IVC) and iliac veins when, in selected cases, consideration
intermediate probability cases
is given to inserting an IVC filter (Fig. 21.1e). This may be
the single examination of choice when more institutions
have MDCT installed.40
FURTHER INVESTIGATIONS
Further investigations
Ventilation/perfusion lung scanning

This remains the first line investigation of possible PE when

• V/Q lung scanning

multidetector helical computed tomography (CT) scanning

• Spiral CT of the lungs

is unavailable.25 The sensitivity and specificity of ventila-

• Echocardiography

tion/perfusion (V/Q) scans has been investigated in the

• Pulmonary angiography

PIOPED multicentre trial. They can be performed rapidly

• Duplex scanning of the legs

with minimal risk to the patient (see Fig. 21.1b, d). The
interpretation and use of V/Q scans seems most effective in Echocardiography
conjunction with clinical probability assignment.21,29–31 V/Q
scan results should be interpreted as normal, or demonstrat- V/Q scans have shown good results as seen in the PIOPED
ing low, intermediate or high probability for PE. A normal study,31,41 PE being present in those scans indicating high
scan excludes PE with 96 per cent accuracy and a high probability. Unfortunately, only a small number of
probability scan with a risk factor is diagnostic of PE with patients with PE have high probability scans.41
86–92 per cent accuracy. A low probability scan with no risk Echocardiography has been looked at for its value in the
factors still requires further investigation to exclude PE but diagnosis of PE.21,42 It may be useful after a large PE in a
8 per cent of patients with low probability lung scans and haemodynamically compromised patient,35 as it can reveal
negative lower extremity Doppler ultrasound have PE.32 the presence of right heart dysfunction, the occasional
Patients with intermediate scans and those with incon- intracardiac thrombus and increased pulmonary artery
gruity between the V/Q result and clinical suspicion require pressure readings with dilated pulmonary arteries. The
further investigation.25,29 presence and degree of right ventricular pressure overload
can be important for diagnostic and prognostic purposes.
These findings may be documented using either the
Spiral CT of the lungs
transthoracic or transoesophageal approach. One limita-
tion to consider is the presence of chronic obstructive pul-
Depending on radiological availability and cooperation at
monary disease, which can make the differential diagnosis
individual institutions, V/Q scans are not often performed
between acute PE and chronic cor pulmonale difficult.43
until the day after presentation or after a weekend. This
delay has prompted many clinicians to take advantage of
spiral CT, if available, as the first line investigation, which Pulmonary angiography
it often is when a large PE is suspected and early diagnosis
is needed (Fig. 21.1a) or when a patient has pre-existing Pulmonary angiography remains the gold standard investi-
cardiopulmonary disease which would limit interpretation gation for PE where MDCT is not available. It is invasive,
of the V/Q scan results.33–35 expensive, not readily available, time consuming, and labour
The speed of spiral CT allows the pulmonary vasculature intensive.5,25 It is most reliable if performed immediately
to be examined with the use of peripherally administered after embolism has taken place and becomes less reliable
contrast during a single breath-hold. It is sensitive and spe- with the passage of time because thrombolysis occurs rap-
cific for central and segmental vessels, but is not as good at idly within the pulmonary circulation. It should be con-
detecting peripheral emboli, which may account for up to 20 sidered if non-invasive tests are inconclusive, if there is a
256 Pulmonary embolism

high clinical suspicion and if MDCT is not available. The change of dosage, then daily once the APTT level is stable
mortality rate associated with pulmonary angiography in with a target range of 1.5–2.5 times normal. Weight based
the PIOPED study was less than 1 per cent, major morbidity nomograms may assist in rapid achievement of target
was 1 per cent and non-major complications 5 per cent.41 APTT.45 If clinical suspicion of PE is low, treatment does
not have to be commenced until after pulmonary imaging.
Duplex scanning of the legs On the other hand, in the absence of specific contraindica-
tions such as an active bleed, patients with a moderate or
A positive duplex scan along with an intermediate prob- high likelihood of PE should receive intensive anticoagula-
ability lung scan provide strong evidence for PE.5,29 Duplex tion while undergoing further investigation.16 Heparin
scanning is a powerful adjunct to lung scanning and can be does not reduce acute mortality but significantly reduces
performed with minimal delay. further untoward events. Low molecular weight heparin
is now first line treatment for DVT and is as effective as
intravenous heparin in PE.44,46–51 Figure 21.2 offers an
TREATMENT algorithm on the management of patients with possible PE
but who are haemodynamically stable.
With massive PE, the presentation may be quite dramatic,
The goals of therapy are to support and maintain life during
and immediate, aggressive institution of supportive care is
the acute episode, to stop the spread of the thromboembolus,
crucial, consisting of attempts to increase venous return and
to foster spontaneous or induce fibrinolytic removal of the
maintaining filling pressures within the heart to ensure car-
thromboembolus and to prevent recurrence. These goals can
diac output. Approximately 10 per cent of patients with PE
be further categorised as prevention, primary treatment and
do not survive the initial embolic event.52 Specific therapy
secondary prophylaxis, involving supportive care, anticoagu-
consists of anticoagulation and lytic therapy. When there is
lation, thrombolysis, and/or embolectomy. With early diag-
cardiac compromise and a significantly unstable haemody-
nosis and swift aggressive therapy many patients will survive
namic state, thrombolysis may be instituted.48 Many patients
PE with a low rate of recurrence.
die from massive PE because of right ventricular failure sec-
ondary to the sudden increase in pulmonary vascular resist-
Prevention ance caused by embolism. Thrombolysis offers a theoretical
advantage over heparin, achieving faster lysis and resolution
Identifying patients with risk factors is essential, as the of thrombus and resulting in a lowering of pulmonary hyper-
incidence of venous thrombosis and PE can be reduced by tension and improvement in cardiac output. The long term
limiting venous stasis or by administering drugs to inhibit outcomes, however, are similar.50,53 In addition, thromboly-
coagulation.44 Elastic stockings and pneumatic compression sis has potentially devastating side effects and therefore its use
devices increase venous return, and the pneumatic pump is restricted to life-threatening cardiac compromise.
may increase fibrinolytic activity with its rhythmic compres- Faced with this severe situation and if thrombolysis has
sions. Prophylaxis with low molecular weight heparin failed or if there are contraindications to this treatment, pul-
(LMWH) by once-daily subcutaneous injection is simple and monary embolectomy, either as suction catheter embolec-
does not need monitoring as its effects are predictable and tomy or as open surgical embolectomy on cardiopulmonary
weight dependent. Perioperative use of low dose subcuta- bypass can be contemplated. This has an average periopera-
neous heparin can prevent about half of PEs and about two- tive morbidity rate of 26 per cent. Transfemoral removal
thirds of DVTs, with a significant reduction in fatal episodes.5 of emboli was described by Greenfield in 1981 using a
specially designed bell-mouthed suction catheter passed
Primary therapy into the pulmonary artery under fluoroscopic control. In
1908 Trendelenburg had first described open pulmonary
The response to pulmonary embolisation is determined by embolectomy, that today involves cardiopulmonary bypass
the size of the embolus and the patient’s cardiopulmonary and embolectomy using embolectomy catheters and manual
status. If the PE is small and the patient is reasonably well, compression of the affected lung. To avoid re-embolisation,
then oxygen administration by mask or nasal prong may an IVC filter may be placed concurrently with the
be all that is needed for supportive care.5 Tracheal intuba- procedure.47,49,54,55 Figure 21.3 offers an algorithm for the
tion and respiratory assistance may be necessary if the management of patients with possible PE but who are
hypoxaemia is profound. haemodynamically unstable.
Specific treatment consists of intravenous heparin infu-
sion following an initial bolus. When sufficient amounts
of heparin are given to stabilise and stop the clotting Secondary prophylaxis
process, the activated partial thromboplastin time (APTT)
becomes about twice that of control. The APTT should be Secondary treatment comprises oral anticoagulants in the
monitored 6 hours after initiation and 6 hours after any form of subcutaneous heparin and insertion of an IVC
 Treatment 257

Stable haemodynamics

High clinical probability Intermediate clinical probability Low clinical probability

Contraindication to anticoagulant therapy? D-dimer

Yes No  

Start heparin pending tests Follow

MDCT  CT venogram
or
V/Q scan  duplex scan

 for PE/DVT  PE  femoro-ilio-caval DVT

Follow Anticoagulation Contraindication for CDT Figure 21.2 Algorithm of management of

patients with possible pulmonary embolism (PE)
who are haemodynamically stable. AVF,
Recurrent PE Yes No arteriovenous fistula; CDT, catheter directed
thrombolysis; DVT, deep vein thrombosis;
IVC, inferior vena cava; MDCT, multidetector
computed tomography; TE, thrombectomy;
IVC filter TE  AVF CDT
V/Q, ventilation/perfusion

filter.48 Warfarin is started in hospital and heparin con- contraindicated or has been ineffective, are good candi-
tinued until the prothrombin time is therapeutically ele- dates for IVC interruption to prevent fatal PE.50 The IVC
vated, with an international normalised ratio (INR) of two filters are inserted under radiological guidance via the
to three times normal.47,49,50,56 This level must be main- femoral or jugular veins and are lodged below the renal
tained for 24 hours before heparin is discontinued, usually veins. Complications of insertion are low but can be
within 5–7 days. Warfarin can be started at the same time extremely significant, including events threatening life and
as heparin, as there is no advantage to prolonging heparin limb. Transvenous insertion of a Greenfield titanium vena
therapy beyond the few days needed to establish a goal INR caval filter protects 97 per cent of patients from PE while
with warfarin. Heparin is continued for a 24 hour overlap maintaining patency of the IVC in 100 per cent of them.55
period with warfarin. Subcutaneous LMWH can be given In emergency situations, IVC filters can be placed under
in cases where warfarin is contraindicated, for example in duplex ultrasound guidance. The rare complication of IVC
pregnancy. filter failure may warrant ligation of the IVC.
After the first episode of PE, treatment is recommended
for at least 3 months.47,49,57 With repeated episodes of PE,
permanent anticoagulation is implemented unless there is an Recurrent thromboembolic pulmonary disease
obvious reversible cause. Repeat duplex scans of the legs or
repeat V/Q scans have been used to confirm complete reso- Recurrent pulmonary emboli, often occurring over many
lution of thrombus before stopping anticoagulation. Elastic years, eventually lead to irreversible pulmonary hyperten-
stockings and avoidance of positions of venous stasis are pre- sion with resultant right heart failure and cor pulmonale.5
scribed to reduce the incidence of recurrent embolism. Dyspnoea is chronic and insidious, often with stepwise
Patients with recurrent thromboembolic disease, and progression. Cyanosis and peripheral oedema is common.
especially those in whom anticoagulant treatment is The JVP is elevated and S3, S4 heart sounds are often
258 Pulmonary embolism

Unstable haemodynamics

Supportive treatment
Vasopressors, inotropes

Stable haemodynamics Unstable haemodynamics

Contraindication to
anticoagulant therapy
Yes No

Support
Yes No

IVC filter Anticoagulation

PE confirmed with
MDCT  CT venogram
or
V/Q scan  duplex scan

No Yes

Support Contraindication to
thrombolytic therapy

Figure 21.3 Algorithm for management of

No Yes Thrombectomy
patients with possible pulmonary embolism
who are haemodynamically unstable. CT,
computed tomography; IVC, inferior vena cava; Consider thrombolysis
 anticoagulation
V/Q, ventilation/perfusion

present. The pulmonary component of S2 is loud. Systolic as shown by autopsy studies: the diagnosis is established
pulmonary artery pressure is often greater than 70 mmHg. in only a third of patients. The astute physician must
Treatment is aimed at preventing further progression, combine a high clinical suspicion for PE, based on
but symptoms are unlikely to be reversed. Intervention presenting features and appropriate risk factors, with
consists of anticoagulation, domiciliary oxygen and treat- early diagnosis and swift, aggressive therapy involving
ment of peripheral oedema with diuretics, balanced with supportive care, anticoagulation, thrombolysis, and/or
maintaining adequate filling pressures. Embolectomy embolectomy. Many patients will survive pulmonary
evaluation should be done at specialised centres before con- embolism with a low incidence of morbidity and
sidering heart–lung transplantation in appropriate patients. recurrence.

Conclusions
Key references
Pulmonary embolism is a potentially fatal complication Anderson DR, Wells PS. Improvements in the diagnostic approach
of DVT, in most cases originating in the upper leg and for patients with suspected deep vein thrombosis or
pelvic veins. While common, it remains underdiagnosed pulmonary embolism. Thromb Haemost 1999; 82: 878–86.
 References 259

Bates S, Hirsch J. Treatment of venous thromboembolism. Thromb 18 Elliot CG, Goldhaber SZ, Visani L, DeRosa M. Chest radiographs
Haemost 1999; 82: 870–7. in acute pulmonary embolism. Chest 2000; 118: 33–8.
Gray HH, Firoozan S. Management of pulmonary embolism. Thorax 19 Stein PD, Dalen JE, McIntyre KM, et al. The electrocardiogram in
1992; 47: 825–32. acute pulmonary embolism. Prog Cardiovasc Dis 1975; 17:
Michiels JJ. Rational diagnosis of pulmonary embolism (RADIA PE) in 247–57.
symptomatic patients with suspected PE: an improved strategy 20 McIntyre KM, Sasahara AA, Littmann D. Relation of the
to exclude or diagnose venous thromboembolism by the electrocardiogram to hemodynamic alterations in pulmonary
sequential use of a clinical model, rapid ELISA D-dimer test, embolism. Am J Cardiol 1972; 30: 205–10.
perfusion lung scan, ultrasonography, spiral CT and pulmonary 21 Anderson DR, Wells PS. Improvements in the diagnostic approach
angiography. Semin Thromb Hemost 1998; 24: 413–18.
for patients with suspected deep vein thrombosis or pulmonary
PIOPED Investigators. Value of the ventilation/perfusion scan in
acute pulmonary embolism: results of the Prospective embolism. Thromb Haemost 1999; 82: 878–86.
Investigation of Pulmonary Embolism Diagnosis (PIOPED). 22 Ginsberg JS, Wells PS, Kearon C, et al. Sensitivity and specificity
JAMA 1990; 263: 2753–9. of a rapid whole-blood assay for D-dimer in the diagnosis of
pulmonary embolism. Ann Intern Med 1998; 129: 1006–11.
23 Perrrier A, Bounameaux H, Morabia A, et al. Contribution of
REFERENCES D-dimer plasma measurements and lower-limb venous
ultrasound to the diagnosis of pulmonary embolism: a decision
analysis model. Am Heart J 1994; 127: 624–35.
1 Stein PD, Henry JW. Prevalence of acute pulmonary embolism 24 Indik J, Alpert J. Detection of pulmonary embolism by D-dimer
among patients in a general hospital and at autopsy. Chest 1995; assay, spiral computed tomography and magnetic resonance
108: 978–81. imaging. Prog Cardiovasc Dis 2000; 42: 261–72.
2 Girard P, Musset D, Parent F, et al. High prevalence of detectable 25 Michiels JJ. Rational diagnosis of pulmonary embolism (RADIA
deep venous thrombosis in patients with acute pulmonary PE) in symptomatic patients with suspected PE: an improved
embolism. Chest 1999; 116: 903–8. strategy to exclude or diagnose venous thromboembolism by the
3 Lindblad B, Eriksson A, Bergqvist D. Autopsy-verified pulmonary sequential use of a clinical model, rapid ELISA D-dimer test,
embolism in a surgical department: analysis of the period from perfusion lung scan, ultrasonography, spiral CT and pulmonary
1951 to 1988. Br J Surg 1991; 78: 849–52. angiography. Semin Thromb Hemost 1998; 24: 413–18.
4 Stein PD, Henry JW. Clinical characteristics of patients with 26 Bounameaux H, de Moerloose P, Perrier A, Miron MJ. D-dimer
acute pulmonary embolism stratified according to their testing in suspected venous thromboembolism: an update.
presenting syndromes. Chest 1997; 112: 974–9. QJM 1997; 90: 437–42.
5 Gray HH, Firoozan S. Management of pulmonary embolism. 27 Kline JA, Johns KL, Colucciello SA, Israel EG. New diagnostic tests
Thorax 1992; 47: 825–32. for pulmonary embolism. Ann Emerg Med 2000; 35: 168–80.
6 Bell WR, Simon TL, Demets DL. The clinical features of submassive 28 Hyers T. Diagnosis of pulmonary embolism. Thorax 1995; 50: 930.
and massive pulmonary emboli. Am J Med 1977; 62: 355. 29 Miniati M, Marini C, Allescia G, et al. Non-invasive diagnosis
7 Urokinase pulmonary embolism trial: Phase 1 results: of pulmonary embolism. Int J Cardiol 1998; 65(suppl 1):
a cooperative study. JAMA 1970; 214: 2163. S83–6.
8 Coon WW. Risk factors of pulmonary embolism. Surg Gynecol 30 Wells PS, Ginsberg JS, Anderson DR, et al. Use of a clinical model
Obstet 1976; 143: 385. for safe management of patients with suspected pulmonary
9 Goldhaber SZ. Optimizing anticoagulant therapy in the embolism. Ann Intern Med 1998; 129: 997–1005.
management of pulmonary embolism. Semin Thromb Hemost 31 Worsley DF, Alavi A. Comprehensive analysis of the PIOPED
1999; 25(suppl 3): 129–33. study. J Nucl Med 1995; 36: 2380–7.
10 Dahl O. Mechanism of hypercoagulability. Thromb Haemost 32 Meyerovitz MF, Mannting F, Polak JF, Goldhaber SZ. Frequency of
1999; 82: 902–6. pulmonary embolism in patients with low-probability lung scan
11 Seligsohn U, Lubetsky A. Genetic susceptibility to venous and negative lower extremity venous ultrasound. Chest 1999;
thrombosis. N Engl J Med 2001; 344: 1222–31. 115: 980–2.
12 Goldhaber SZ, Grodstein F, Stampfer MJ, et al. A prospective 33 Cross JJ, Kemp PM, Walsh CG, et al. A randomized trial of spiral
study of risk factors for pulmonary embolism in women. JAMA CT and ventilation perfusion scintigraphy for the diagnosis of
1997; 277: 642–5. pulmonary embolism. Clin Radiol 1998; 53: 177–82.
13 De Stefano V, Chiusolo P, Piciaroni K, Leone G. Epidemiology of 34 Kim KI, Muller NL, Mayo JR. Clinically suspected pulmonary
factor V Leiden: clinical implications. Semin Thromb Hemost embolism: utility of spiral CT. Radiology 1999; 210: 693–7.
1998; 24: 367–79. 35 Torbicki A. Imaging venous thromboembolism with emphasis on
14 Hull RD, Hirsch J, Carter CJ, et al. Pulmonary angiography, ultrasound, chest CT, angiography and echocardiography. Thromb
ventilation lung scanning and venography for clinically Haemost 1999; 82: 907–12.
suspected pulmonary embolism with abnormal perfusion lung 36 Grenier PA, Beigelman C. Sprial computed tomographic scanning
scan. Ann Intern Med 1983; 98: 891–9. and magnetic resonance angiography for the diagnosis of
15 Sasahara AA, Sharma GV, Barsamian EM, et al. Pulmonary pulmonary embolism. Thorax 1998; 53(suppl 2): S25–31.
Thromboembolism. JAMA 1983; 249: 2945–50. 37 de Monye W, van Strijen MJ, Huisman MV, et al. Suspected
16 Goldhaber SZ. Pulmonary embolism. N Engl J Med 1998; 339: pulmonary embolism: prevalence and anatomic distribution in
93–104. 487 consecutive patients. Advances in New Technologies
17 Cvitanic O, Marino PL. Improved use of arterial blood gas analysis Evaluating the Localization of Pulmonary Embolism (ANTELOPE)
in suspected pulmonary embolism. Chest 1989; 95: 48–51. Group. Radiology 2000; 215: 184–8.
260 Pulmonary embolism

38 Mullins MD, Becker DM, Hagspiel KD, Philbrick JT. The role of 47 Bates S, Hirsch J. Treatment of venous thromboembolism.
spiral volumetric computed tomography in the diagnosis of Thromb Haemost 1999; 82: 870–7.
pulmonary embolism. Arch Intern Med 2000; 160: 293–8. 48 Goldhaber SZ. Contemporary pulmonary embolism thrombolysis.
39 Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosis of Int J Cardiol 1998; 65(suppl 1): S91–3.
pulmonary embolism with spiral CT: comparison with pulmonary 49 Hyers T, Hull R, Weg J. Antithrombotic therapy for venous
angiography and scintigraphy. Radiology 1996; 200: 699–706. thromboembolic disease. Chest 1995; 108(suppl 4): S335–51.
40 Cham MD, Yankelevitz DF, Shaham D, et al. Deep venous 50 Kearon C. Initial treatment of venous thromboembolism. Thromb
thrombosis: detection by using indirect CT venography. Haemost 1999; 82: 887–91.
Radiology 2000; 261: 744–51. 51 Simonneau G, Sors H, Charbonnier B, et al. A comparison of low-
41 PIOPED Investigators. Value of the ventilation/perfusion scan in molecular-weight heparin with unfractionated heparin for acute
acute pulmonary embolism: results of the Prospective pulmonary embolism. The THESEE Study Group. N Engl J Med
Investigation of Pulmonary Embolism Diagnosis (PIOPED). 1997; 337: 663–9.
JAMA 1990; 263: 2753–9. 52 Matsumoto A, Tegtmeyer C. Contemporary diagnostic approaches
42 Pavan D, Nicolosi GL, Antonini-Canterin F, Zanuttini D. to acute pulmonary emboli. Radiol Clin North Am 1995; 33:
Echocardiography in pulmonary embolism disease. Int J Cardiol 167–83.
1998; 65(suppl 1): S87–90. 53 Urokinase-streptokinase embolism trial: Phase 2 results:
43 Grifoni S, Olivotto I, Cecchini P, et al. Utility of an integrated a cooperative study. JAMA 1974; 229: 1606.
clinical, echocardiographic, and venous ultrasonographic 54 Doerge H, Schoendube FA, Voss M, et al. Surgical therapy of
approach for triage of patients with suspected pulmonary fulminant pulmonary embolism: early and late results. Thorac
embolism. Am J Cardiol 1998; 82: 1230–5. Cardiovasc Surg 1999; 47: 9–13.
44 Goldhaber SZ. Venous thromboembolism prophylaxis in medical 55 Greenfield LJ. Caval interruption procedures. In: Rutherford RB
patients. Thromb Haemost 1999; 82: 899–901. (ed). Vascular Surgery, 5th edn. Philadelphia: WB Saunders,
45 Raschke RA, Reilly BM, Guidry JR, et al. The weight-based 2000: 1968–78.
heparin dosing nomogram compared with a ‘standard care’ 56 Prins MH, Hutten BA, Koopman MM, Buller HR. Long-term
nomogram: a randomized controlled trial. Ann Intern Med 1993; treatment of venous thromboembolic disease. Thromb Haemost
119: 874–81. 1999; 82: 892–8.
46 Agnelli G, Sonaglia F. Anticoagulant agents in the management 57 British Thoracic Society Research Committee. Optimum duration
of pulmonary embolism. Int J Cardiol 1998; 65(suppl 1): of treatment for deep-vein thrombosis and pulmonary embolism.
S95–8. Lancet 1992; 340: 873–6.
 22
Upper Limb Vein Thrombosis

BENGT LT LINDBLAD, KRASSI IVANCEV, SIMON G DARKE

The problem 261 Clinical presentation 267

History 261 Investigations 267
Pathogenesis, nomenclature and classification 261 Complications 267
Anatomical considerations 262 Treatment alternatives 269
Venographic and haemodynamic findings 263 References 272
Classification 263

THE PROBLEM HISTORY

Upper limb venous thrombosis (ULVT) is fairly uncommon. Perhaps the most exotic, dramatic and at the same time
It is estimated to represent 1–3 per cent of recognised deep the first described case of probable ULVT was that of
vein thromboses.1–3 Cases of ULVT are divided into primary, Henry of Navarre (1553–1610) who, as King Henry IV of
when no certain explanation for development of the throm- France, led his army into the Battle of Ivry (1590). He used
bosis of the axillary/subclavian vein is found, and secondary, his sword arm to such excess that he could not move
when an obvious factor has caused the development. it for 6 weeks.8 The first description of venous compromise
Anticoagulation has been widely used and gives of the upper extremity is credited to Sir James Paget in
favourable results preventing re-thrombosis and pulmonary 1875. His perception, however, was that the underlying
embolism (PE); it also lowers the frequency of patients hav- morphology was a type of ‘gouty arthritis’.9 The true
ing severe long term arm disability.1–5 During the past decade venous cause was recognised nine years later by von
many reports have been published suggesting more aggres- Schroetter,10 hence the eponymous term Paget–von
sive treatment with thrombolysis with or without thoracic Schroetter syndrome.
outlet decompression, i.e. first rib resection, scalenectomy
and venolysis.6,7 Most of these reports concern primary
thrombosis and the subgroup of young patients with so-
PATHOGENESIS, NOMENCLATURE AND
called effort induced ULVT. As a result of the low incidence
CLASSIFICATION
of ULVT, the experience of any one clinician is limited; this
compounds the problems and the opportunities of finding
an optimal therapeutic strategy. Furthermore, the disease Upper limb vein thrombosis has a multifactorial aetiology.
fails to fall naturally into any particular medical specialty. The classification most used differentiates between pri-
This again dilutes experience and the expertise of any one mary and secondary axillary/subclavian vein thrombosis.1–5
individual or department and reduces the potential for Primary thrombosis may occur spontaneously or follow-
prospective controlled trials of treatment options. This ing effort while secondary thrombosis may be iatrogenic
review includes a suggestion of treatment pathways depicted or complicate malignancy and other miscellaneous causes.
in an algorithm (Fig. 22.1), but it is essential to be aware that Before considering this in further detail it is of value to
this is not based on solid evidence-based facts, but reflects the address the anatomy and the venographic appearances of
authors’ opinion. these conditions.
262 Upper limb vein thrombosis

Suspicion of axillary/subclavian vein thrombosis

Venography, duplex No thrombosis

Primary thrombosis Secondary thrombosis

Effort induced Spontaneous or unclear

Thrombolysis Symptoms for more than 10–14 Anticoagulation and 3–12

 anticoagulation days or minor symptoms months oral anticoagulation

Venography at Severe long term symptoms

1 month

Venography, MR
Remaining stenosis
or major positional
stenosis

Thoracic outlet Thoracic outlet Conservative

decompression decompression treatment

Figure 22.1 An algorithm of the management of suspected axillary/subclavian vein thrombosis. MR, magnetic resonance (scan)

ANATOMICAL CONSIDERATIONS Clavicle

Anterior scalene muscle

‘Thoracic outlet syndrome’ is a loose term which describes Subclavius muscle

compression of the artery, vein or plexus of nerves as they
First
pass through the neck to the upper extremity. In anatom- rib
ical terms the problem seems to present in three distinct
areas.1,7,11 The first two of these, which do not in general
concern us here, might be called ‘axillo-pectoral compres-
sion’ and ‘the costo-scalene tunnel’. In axillo-pectoral com-
pression intermittent compression of the axillary vein can
be seen between the pectoral muscles and the rib cage. In Subclavian
the costo-scalene tunnel the artery or possibly the nerves Subclavian vein
artery Costoclavicular
become compressed between a cervical rib, a congenital
ligament
fibrous band or the posterior surface of the scalenus anter-
ior muscle. Removal of such constrictions (Fig. 22.2) Figure 22.2 Anatomy of the thoracic outlet. Note that the vein
decompresses the artery but normally not the vein. is most likely to be compressed by the subclavius muscle and
In contrast, the vein is caught more medially in the costo- costoclavicular ligament, and the artery by the anterior scalene
clavicular tunnel. The axillary vein becomes the subclavian muscle
vein at the outer border of the first rib and passes through
a potentially narrow space bounded inferiorly by the first What then are the morphological features of apparently
rib and pleura. The vein is restricted, anteriorly, by the normal structures within this space which predispose to
clavicle and subclavius muscle. Posteriorly, it is constrained venous compression and thrombosis? There are two princi-
by the anterior border of the scalenus anterior muscle and pal components and in most cases both contribute to a rela-
medially by the costoclavicular joint and ligament.11 It is tively greater or lesser extent. The first of these is that the gap
obvious, therefore, that cervical ribs, bands or associated quite simply is congenitally narrow6 and can be detected at
structures have no relevance to subclavian venous problems. operation.11 The second factor is related to physical activity.
 Classification 263

(a)
(b)

Figure 22.3 (a, b) A short subclavian vein occlusion with collaterals which later fill central veins. During surgical venolysis some
of these collaterals may be injured and occlude. Re-thrombosis may then give an even more symptomatic long term outcome

This may be an acute and isolated episode of exceptional is obtained with a 30 degree abduction. This compression
exertion, or it may be a consequence of prolonged activity is due to the closeness between the pectoral muscles and
such as training, sporting pursuits and similar avoca- the rib cage.18
tions.6,7,10,11 Here it is thought that the critically situated Venous arm pressure recordings also show that com-
subclavius muscle becomes engorged or hypertrophied (see pression of the veins occurs in different positions of the
Fig. 22.2). Thrombosis, however, occurs clearly in only a few arms. In the relaxed position, arm venous pressure is
selected individuals who undertake these activities. It seems around 6–7 mmHg and is much the same when comparing
logical, therefore, that if the vein is to be harmed by physical healthy volunteers and patients with arm and shoulder dis-
activity, an element of anatomical narrowing referred to ability with or without earlier ULVT.19 In hyperabduction
above must almost certainly coexist. It follows that in all or the ‘hurray position’, and especially in a ‘military pos-
cases of primary thrombosis this idiosyncratic anatomical ition’ with the shoulders drawn backwards, the pressure
narrowing is present. increases up to around 10 mmHg in normal individuals. In
patients with chronic non-recanalised subclavians, venous
pressure increased more prominently (艐20 mmHg) in
Sites of thoracic outlet syndromes and hyperabduction and especially in the ‘military position’.
vulnerable structures Thus it appears that collaterals passing the narrow costo-
clavicular area are also affected by arm position.
• Axillo-pectoral compression – vein In a number of studies evaluating haemodynamics using
• Costo-scalene tunnel – artery, nerve plethysmography, it has been shown that slightly reduced
• Costoclavicular tunnel – vein venous capacity and maximal venous outflow are observed in
patients with chronic non-recanalised ULVT.19 In veno-
graphic terms, two main patterns of ULVT are seen, namely,
a short subclavian vein occlusion (Fig. 22.3) or a more exten-
VENOGRAPHIC AND HAEMODYNAMIC
sive thrombotic process involving the axillary vein.14,16–18
FINDINGS

CLASSIFICATION
A key issue in the strategy of treatment is the incidence of
this anatomical narrowing, both in an unselected popula-
tion and among those suffering from ULVT. There are a Primary thrombosis, i.e. where no evidence is found to
few venographic studies showing that 70 per cent of account for the development of axillary or subclavian vein
unselected individuals do have a degree of stenosis of the thrombosis, is reported with very different frequencies in
venous outflow, especially with the arm in hyperabduc- different reports (Table 22.1).3,4,10,20–32 One reason for this
tion.12,13 Similarly, Stevenson and Parry reported that in is the manner in which population details are collected and
patients with ULVT they could almost always document a another is whether they are representative of the popula-
compression abnormality of the vein in the contralateral tion,3,4 21,25 the other being that more severely symptomatic
arm in hyperabduction.14 This has also been reported by patients are referred.6,7,26,27 In some patients a history of
others with a frequency ranging from 56 to 80 per cent.6,15–17 unusual upper limb activity precedes the development of
With the arm resting along the body compression of the ULVT, so-called effort induced thrombosis. This accounts
axillary vein is not uncommon and best venographic filling for 5–10 per cent of patients in an unselected population.
264 Upper limb vein thrombosis

A larger group of about 30 per cent are considered to be Primary thrombosis

spontaneous cases of primary ULVT; in most cases the aeti-
ology is multifactorial. The term ‘primary thrombosis’ includes spontaneous,
The number of cases of secondary thrombosis has now idiopathic, effort induced, positional, strain and Paget–von
increased to approximately 60 per cent of those detected Schroetter syndrome. It even includes the term ‘traumatic’
(Table 22.1).3 This is due to an increased number of when it has been used to describe probable anatomical
instances of iatrogenic thrombosis, such as those caused by abnormalities causing trauma to and compression of the
central venous lines and pacemaker leads. subclavian vein.2,6,7 It is obvious that the ‘normal’ narrow-
In an earlier series (1971–86) from MalmöUniversity ing of the costoclavicular space is a pathogenic factor in
Hospital,4 and representative of the population of the city, most of these patients; the problem is establishing how
61 per cent were considered to have primary thrombosis. crucial this and the other factors are in causing the devel-
The majority of these had no well-defined factors con- opment of thrombosis.
tributing to the development of thrombosis. Effort or
physical activity related thrombosis was probably evident
in only 6 per cent. A proved hypercoagulable state was seen Effort thrombosis
in 9 per cent, to which should be added 17 per cent on con-
traceptive pills and those with previous thromboembolic This presentation is an obviously recognisable and distinct
episodes. In a recent analysis of the period 1987–2001 and entity.2,3,6,7,10 It is of relevance not only because it can be
within the same catchment area, only 3 per cent out of the defined as effort induced by the obvious sequence of events,
total of 155 cases were effort induced while 30 per cent but also because it relates to a possible long term outcome,
were probably cases of spontaneous ULVT. a factor which must influence management strategy.
Secondary thrombosis, in which a major contributory Effort induced thrombosis occurs as a result of excessive
factor is found, comprised about 40 per cent of the patients or prolonged physical activity in combination with the nar-
with subclavian vein thrombosis in our series from 1971 to row space in the costoclavicular region through which the
1986.4 In the recent series recorded from 1987 to 2001 the vein passes. A spectacular variety of provocative activities
incidence of thrombosis has increased to 67 per cent: of has been reported: throwing balls, playing handball or ten-
these 44 per cent are due to central venous lines or pace- nis, swimming, chopping wood, painting ceilings, pulling a
maker leads, 20 per cent to malignancy and another 3 per propeller, rowing a boat, washing walls, starting an out-
cent are considered to be miscellaneous cases. Just how board motor, lifting weights, carrying a suitcase, wearing a
many patients have undiagnosed secondary upper arm figure-of-eight splint, falling asleep with one’s arm over the
thrombosis may only be speculated upon. back of the chair and windsurfing.1–7,10

Table 22.1 Classification of upper extremity deep venous thrombosis

Primary (per cent) Secondary (per cent)

Authors Year No. Effort Spontaneous Iatrogenic Malignant Others

Hughes10 1949 320 49 34 0 0 17

Adams et al.20 1965 25 52 32 0 0 16
Coon and Willis21 1967 60 23 5 38 10 24
Tilney et al.22 1970 48 21 21 17 6 35
Donayre et al.23 1986 41 22 5 51 12 10
Lindblad et al.4 1988 120 6 55 25 11 3
Hauptli et al.25 1989 96 3 24 34 28 10
Horattas et al.24 1988 33 6 12 40 24 18
AbuRahma et al.26 1991 23 100 0 0 0 0
Malcynski et al.27 1993 23 52 0 35 9 4
Martin and Brors28 1995 91 13 39 13 9 26
Adelman et al.29 1997 38 16 20 [...................................64....................................]
Beygui et al.30 1997 32 42 6 29 29
Hingorani et al.3* 1997 170 1 21 65 37 11
Marie et al.31 1998 49 10 14 22 33 21
Sakakibara et al.32 1999 12 9 9 67 17 0
Zell et al.33 2001 82 34 4 [...................................62...................................]
Malmö University Hospital 2001 155 3 30 44 20 3

* 33 per cent of the patients had multiple risk factors.

 Classification 265

A variant of this category occurs when thrombosis fol- Iatrogenic thrombosis

lows excessive long term activities. These include excessive
weight training, javelin throwing and occupations such as Iatrogenic thrombosis may account for up to two-thirds of
window cleaning.11 In these cases contralateral venography cases of ULVT nowadays.3 A large proportion of these are
often reveals a bilateral abnormality and previously unde- subclinical and undetected and therefore not treated.
tected and asymptomatic thrombosis on the opposite
side.11 CENTRAL VENOUS CATHETERS
The incidence of clinical symptomatic thrombosis caused by
Classification of ULVT indwelling central venous lines is low at 1–6 per cent (Table
22.2).39–55 If carefully investigated, however, a large propor-
tion of, if not all, catheters will have a fibrin sleeve as a coating
• Primary
over the foreign material (Table 22.3).39,42,43,48,49,56–59 Mural
– Spontaneous, idiopathic
– Effort, positional, strain (Paget–von Schroetter or occlusive thrombosis is not uncommon and is found in
syndrome) 12–68 per cent of patients undergoing prospective veno-
graphic examination (Table 22.2). Many patients with central
• Secondary
venous cannula induced thrombosis remain undiagnosed.
– Iatrogenic
– Malignancy The stiffness of the catheter, the lumen size, the length of
– Miscellaneous the catheter and the length of catheterisation required are
all reflected in the frequency of thrombosis.40,55,56 The type
of infusion given, i.e. total parenteral nutrition, hyperos-
Hypercoagulable states molarity solutions or chemotherapy and the reason for the
use of a central venous line, e.g. malignancy, are likewise
Many reports have reported different coagulation and fibri- important. Normally short catheterisation periods will not
nolytic defects in patients developing ULVT. 4,34–37 In some induce significant thrombosis.41
reports 9–42 per cent of patients had a hypercoagulable state Studies of the long term use of central venous catheters
(also see Chapters 2 and 20). Heron and coworkers analysed and prevention of induced thrombosis confirm the efficacy
51 patients with primary thrombosis: comparing the spon-
Table 22.2 Central venous catheters: mural and occlusive
taneous thrombosis group (n  31) with the effort induced
thrombosis
thrombosis group (n  20) there was a higher frequency of
a family history of DVT, 42 per cent and 15 per cent, respect- Frequency Diagnostic
ively, and at least one coagulation abnormality, 42 per cent Authors Year No. (per cent) method
and 15 per cent, respectively.35 Leebeek and associates found
a 27 per cent frequency of antiphospholipid antibodies.36 In Hoshal et al.39 1971 55 4 PM
an analysis of 60 patients collected from our hospital in Ryan et al.40 1974 34 24 PM
Malmöduring the 1970s, a decreased fibrinolytic defence Lindblad et al.41 1981 28 32 PM
mechanism was seen in 49 per cent, F VIII:C activity was Ahmed42 1976 63 20 POP
increased in 48 per cent and F VIII:R Ag was elevated in 20 Bennegård et al.43 1982 22 68 POP
per cent.34 Activated protein C resistance, polymorphism in
Braun et al.44 1970 50 23 CP
prothrombin fragments 1 and 2, antithrombin, protein C Fassolt and Braun45 1978 50 12 CP
and protein S deficiency were coagulation defects seen in Ladefoged et al.46 1979 32 49 CP
just a small percentage of patients.37 Lindblad et al.41 1981 95 12 CP
Reports relating thrombosis to ingestion of the oestro- Magalhaes et al.47 1986 51 23 CP
gen containing contraceptive pill may well be coincidental. Balestreri et al.48 1995 57 46 CP
However, within some larger series there are a small num- Martin et al.49 1999 60 12 CP
ber of patients in whom this appears to be a contributory Allen et al.50 2000 137 24 CP
factor.4,6,22–24,26,38 Male et al.51 2003 85 34 CP  Duplex
 MRI
Frank et al.52 2000 319 35 RNP
Secondary thrombosis Wu et al. 53
1999 81 25 Duplex
54
A large and increasing proportion of patients are diag- Soto-Velasco et al. 1984 1311 0.8 Clinical
nosed with secondary upper arm thrombosis. In this group Schwarz et al.55 2000 791 6.5 Clinical
of patients anatomical narrowing in the costoclavicular PM, postmortem; POP, pull-out phlebography; CP, conventional
space is not a major cause but local trauma or compression phlebography; MRI, magnetic resonance imaging; RNP, radionuclide
from malignancy is of major aetiological importance. phlebography.
266 Upper limb vein thrombosis

of standard heparin, low molecular weight heparin (LMWH) venography reveals a 30–50 per cent incidence.66,67 The
and low dose warfarin therapy.60–64 The number of studies discrepancy between clinically documented and prospec-
available is not sufficient for evidence-based recommenda- tively recorded thrombosis is significant, with most asymp-
tions, but they warrant further analysis. If possible, such pre- tomatic cases going undetected.
vention ought to be used more frequently than most centres
do today. Prevention will not only reduce the risk of throm-
bosis but also that of pulmonary embolism, a not too infre- Iatrogenic causes of ULVT
quent sequela to secondary ULVT.
• Central venous catheters, type of infusate
PACEMAKER LEADS • Pacemaker and defibrillator leads – single/double

In spite of improved pacemaker lines during the last decade,

• Haemodialysis catheters, arteriovenous (AV) fistula
and late central vein stenosis
more complex implants such as double-lead pacemakers
and defibrillators have been developed. Clinically symp-
tomatic thrombosis is not as common,65 but routine
HAEMODIALYSIS CATHETERS
These are large, often double lumen, catheters that pose a
Table 22.3 Central venous catheters: fibrin sleeve formation high risk for thrombosis (11–40 per cent).68 Later, when
Frequency Diagnostic
patients are evaluated for angio-access, occluded or stenotic
Authors Year No. (per cent) method central veins are often found68–71 (see Chapter 7B). Few of
these patients have any symptoms in their arms and many
Scholz and Loewe55 1969 60 55 PM centres routinely perform venography before an angio-access
Hoshal et al.39 1971 55 100 PM operation if the patient has had a haemodialysis catheter on
that side.69 A stenosis or an occlusion in the venous outflow
Hoshal et al.39 1971 31 100 POP
will reduce patency of peripherally constructed AV fistulae.
Ahmed42 1976 63 75 POP
Central vein stenosis can develop several months after an
Brismar et al.57 1980 35 66 POP
Andersson et al.58 1981 60 85 POP AV fistula operation resulting in higher resistance to venous
Bennegård et al.43 1982 22 100 POP return. Angioplasty has shown acceptable results at 6
Starkhammar et al.59 1992 16 100 POP months, primary patency being 31–42 per cent and second-
ary patency 71–80 per cent, but restenosis often develops
Balestreri et al.48 1995 57 78 CP requiring repeated dilatations.68,70 Stenting of such veins
Martin et al.49 1999 60 58 CP has been used and seems a reasonable alternative (Fig. 22.4
PM, postmortem; POP, pull-out phlebography; CP, conventional a, b),68,72 but repeated endovascular procedures are needed
phlebography. at approximately 3-month intervals and if considerable

(a) (b)

Figure 22.4 (a, b) An enlarged cephalic vein resulting from a distal arteriovenous fistula with stenosis of the most proximal portion of the
cephalic vein as well as of the subclavian vein, both of which were stented. Restenosis made repeated dilatations necessary about every
third month and patency was maintained for over 4 years
 Complications 267

narrowing in the costoclavicular tunnel is noted, the risk of unusual.3,4,10,25,30 In many reports a male preponderance
stent breakage must be considered.73 has been reported.10,21,26 In the Malmöseries 4 and others38
no such difference was observed between the sexes; nei-
Malignant disease ther was there any predominant involvement of the
right upper limb compared with the left. Venous gangrene
In the majority of these cases the cause is direct pressure is a rare finding and underlying malignancy should be
by tumour mass. They account for 10–30 per cent of considered.
ULVT.4,25,33 Less frequently the thrombosis is ascribable to
a hypercoagulable state adding to the other risk factors.3
Local radiotherapy may also be a contributory factor in the INVESTIGATIONS
development of thrombosis (see Chapter 33).
Normally, a conventional arm venogram will establish the
Miscellaneous causes diagnosis of ULVT. Alternative techniques are duplex,
ultrasound, spiral computed tomography (CT) and mag-
Other causes that have been reported are congestive heart netic resonance angiography (MRA).1–5,53 Each of these
failure, myocardial infarction, congenital anomaly, vascular techniques has limitations and needs to be evaluated based
malformations, drug abuse (see Chapter 44) and trauma.1–5,31 on the experience of the examiner. With the arm held by
There are even case reports of patients with vasculitis and the side of the body it is not unusual for the axillary vein to
vasospasm. In some cases severe illness, for example sepsis, be compressed by surrounding muscles.18 If the contrast is
ulcerative colitis, filariasis, otitis media, etc., is associated with not injected mainly into the basilic vein, the venogram will
ULVT and these are considered secondary occurrences, but be of poor quality16 as precise timing is required in assess-
the main causative factor may be difficult to identify. ing the central veins.17 Duplex ultrasound has the advan-
Congenital webs are known to cause thrombosis at tage of being non-invasive, but a large collateral cannot be
other sites as in the iliac vein web and Budd–Chiari syn- easily distinguished from the axillary/subclavian veins.
drome but there must be doubt as to whether this occurs in New imaging techniques such as spiral CT and MR imag-
the axillary/subclavian vein. Thus, an intraluminal ‘web’ is ing are most promising.51
likely to be a secondary phenomenon due to incomplete Routine laboratory testing should be performed and in
recanalisation after previous thrombosis or a function of cases of primary thrombosis a coagulation screen is recom-
local trauma. mended. A chest X-ray will rule out osteal malformation of
the clavicle or an unknown pulmonary malignancy.
Miscellaneous causes of ULVT

• Congestive heart failure, myocardial infarction COMPLICATIONS

• Congenital venous anomalies – webs,
malformations
• Vasculitides Pulmonary embolism
• Drug abuse
Pulmonary embolism (PE) is considered first because it
• Trauma
threatens the patient’s life (see Chapter 21). Table 22.4 docu-
• Sepsis – focal or systemic
ments the frequency of selected series reporting on non-fatal
and fatal pulmonary embolism.3,4,6,7,10,20–22,24,25,27,29,31,32,38,
51,74–85
For patients with ULVT classified as primary, a 1 per
CLINICAL PRESENTATION cent (10/1017) incidence of non-fatal PE has been reported.
A 7 per cent incidence of PE, 10 fatal and 49 non-fatal,
The characteristic findings are swelling of the affected arm, has been documented followed secondary ULVT among
distension of superficial veins, especially in the shoulder 807 patients. In this group, major associated aetio-
region, mild cyanosis, and pain.1–5,10,38 In our earlier reported logical factors were taken into account in many of the
series4 oedema was seen in 97 per cent, pain in 50 per cent, patients who developed PE. In a prospective study
distended collateral veins in 35 per cent, acrocyanosis in 10 Monreal and co-workers79 followed 86 patients with cen-
per cent and erythema in 10 per cent. Few had altered skin tral venous cannula induced upper extremity thrombosis.
temperature and only 4 per cent had fever. Importantly, Anticoagulant treatment was started and perfusion/venti-
about half of the patients with symptoms and suspicion of lation scintigraphy revealed pulmonary embolism in 15
ULVT had normal venograms.4 per cent of patients, of whom two died of recurrent pul-
Normally, patients present with a history of 2–3 days monary embolism in spite of prophylactic anticoagulant
of symptoms, but a delay of up to several months is not therapy.
268 Upper limb vein thrombosis

Table 22.4 Fatal and non-fatal Pulmonary embolism costoclavicular space, and if there is an anatomical narrow-
ing, these collaterals may also be dependent on arm position.
Primary Secondary
This narrowing is usually visible on venography with the
Authors Year No. NFPE No. NFPE FPE arm in different positions16–19 and is also verifiable with
pressure recordings.19
Hughes10 1949 269 1 51 2 It is not possible to predict long term sequelae on the
Adams et al.20 1965 21 3 4 1 basis of presenting symptomatology.4,10,19 Even in patients
Coon and Willis21 1967 17 0 43 3 whose occluded veins recanalise, normally possible in
Swinton et al.744 1968 23 1 10–20 per cent, a clear reduction in long term disability is
Tilney et al.22 1970 17 0 31 0
not consistently found.2,4,26,38 Of those referred from out-
Campbell et al.75 1977 9 0 10 2
Harley et al.76 1984 3 1 11 4
side the normal recruitment area of our university depart-
Horattas et al.24 1988 4 0 29 4 ment, there is a preponderance of young patients with
Lindblad et al.4 1988 73 0 47 2 3 primary thrombosis and severe symptomatology. Often the
Hauptli et al.25 1989 26 0 70 1 time from onset of symptoms until referral is more than
Reed et al.77 1992 5 0 7–14 days. Each aetiological group must be considered sep-
Thompson et al.78 1992 6 0 arately because, inherently, the outcome may be different.
Malcynski et al.27 1993 12 0 11 0
Monreal et al.79 1994 86 11 2
Machleder6 1993 50 0 Primary subclavian thrombosis
Molina80 1995 18 0
Adelman et al.29 1997 14 0 24 0
Long term symptoms are difficult to quantify and evaluate
Hingorani et al.3 1997 2 0 168 10 2
Azakie et al.81 1998 33 2
objectively. Protagonists of more aggressive surgery such as
Marie et al.31 1998 12 1 37 5 first rib resection are likely to take a pessimistic view of the
Sakakibara et al.32 1999 2 0 10 0 unoperated outcome. An element of bias can easily cloud
Urschel and Razzuk7 2000 294 0 the issue. Thus, in the literature different opinions exist on
Angle et al.82 2001 18 0 the frequency of long term disability, especially after primary
Feugier et al.83 2001 10 0 ULVT (20–82 per cent),1–7,10,38 this difference is also due to
Kreienberg et al.85 2001 23 0 differences in the selection of patients reported upon. Most
Lokanathan et al.85 2001 28 1 authors report that a proportion of patients will have mild
Sabeti et al.38 2002 33 0 85 7 symptoms with occasional swelling, increased superficial
Male et al.51 2003 85 0 networks of veins over the shoulder and pain, but few
Total 1017 10 807 49 10
patients develop more severe disability requiring, for
FPE, fatal pulmonary embolism; NFPE, non-fatal pulmonary embolism. example, a change of occupation.1,2,4,25,28,31 Patients usually
complain of difficulties working for long periods with the
arm in an elevated position.
The analysis from Malmö4 covering the entire population
Complications of ULVT and the outcome of all symptomatic cases, comes as close as
it is possible to get an accurate estimate of the long term out-
• Pulmonary embolism come. An attempt was made to quantify the severity of the
• Residual vein thrombosis outcome. ‘Mild’ sequelae were defined as occasional pain
• Recurrent vein thrombosis and some restrictions in arm capacity, ‘moderate’ as symp-
• Chronic vein stenosis toms limiting certain activities and ‘severe’ as daily symp-
• Disabilities of restricted venous outflow toms with severe limitation, often leading to a change of
occupation. Of the 73 patients with primary thrombosis 15
(21 per cent) had mild long term symptoms and in a further
three (4 per cent) they were moderate after a median of
Long term symptoms 6 years of follow-up.
The small group of patients with effort induced ULVT
Although less dramatic than PE, long term symptoms are seems to have more frequent long term disability,6,7,26,27,29,
30,80–85
the crucial area, in practical terms, on which management is but other reports show lower frequencies of long
based. The widening of existing collaterals and/or recanali- term sequelae.2–5,34,38,86 Many reports, especially from North
sation is normally seen during the first six months after axil- American centres, seem to include cases of spontaneous pri-
lary/subclavian vein thrombosis. Collaterals can be seen mary thrombosis. In those patients with what can be clearly
from the arm/shoulder extending to the chest wall, the classified as effort induced thrombosis, up to 50 per cent
anterior/posterior neck veins and contralateral neck veins complain of moderate long term sequelae, and in 25 per cent
as well. Many of these collaterals also pass through the this leads to working disability.6,7,26
 Treatment alternatives 269

Secondary subclavian thrombosis For patients with secondary ULVT, anticoagulation in the
form of heparin or LMWH for 5–7 days should be the first
In many patients any disability arising from ULVT will be treatment alternative.2 The risk of PE speaks in favour of oral
hidden by other serious diseases.2–4 Most authors agree that anticoagulation for 3–12 months. In patients with catheter
the frequency of long term sequelae is low, that severe and induced thrombosis, it may be wiser not to withdraw the can-
even moderate symptoms are unusual and that mild symp- nulae immediately after diagnosis. One can speculate as to
toms occur in 30–74 per cent.2–4,25,31 In iatrogenic ULVT late whether anticoagulant treatment should be initiated and
problems are uncommon and the majority may remain whether the thrombus should be a little more organised so as
totally symptom free.41,54 In the Malmöseries, 47 patients to minimise the risk of PE at removal of the catheter.
had secondary thrombosis, 14 (30 per cent) had mild and
two (4 per cent) had moderate long term sequelae.4 If one
analyses different treatment modalities in our series, the out- Thrombolysis
come was worse in the more aggressively treated patients, but
these presented with more severe symptoms. What then is Although thrombolytic therapy is of no proven benefit in
the evidence that treatment influences long term disability? the lower limb, there is increasing interest in its employ-
In a retrospective review by Hicken and Ameli2 persistent ment in axillary/subclavian vein thrombosis and many
symptoms were reported in 52 per cent of conservatively reports, together including 569 patients, have been pub-
treated patients, in 32 per cent of those anticoagulated and in lished, especially during the last decade.6,7,26,27,29,30,77–86,88–91
26 per cent of those treated by thrombolysis. Most reports on thrombolysis have used catheter directed
low dose therapy, and in general, treatment has been
required for 1–2 days. Three or four non-fatal cases of PE
have been reported during treatment.2,38 The success of
TREATMENT ALTERNATIVES initial recanalisation is reported to be high (40–100 per cent),
the results seeming better if the treatment time span from
the onset of signs of thrombosis to the commencement
Many patients with ULVT were treated conservatively using of thrombolysis is less than 14 days.7,26,81
arm elevation and rest with favorable outcomes.10 More The re-occlusion frequency is high (34 per cent),6,7,26,38
modern alternatives employed were anticoagulation and but it has been reduced by endovascular and surgical meas-
thrombolysis with or without thoracic outlet decompres- ures aimed at decompression of the costoclavicular narrow-
sion. An algorithm (see Fig. 22.1, page 262) provides treat- ing. The majority of patients treated with thrombolysis are
ment options for ULVT. The plethora of published young patients with effort induced thrombosis,6,7 but many
algorithms and surgical approaches in achieving decom- reports also include patients with spontaneous primary
pression of the thoracic outlet underlines our limited under- thrombosis. In a retrospective comparison of systemic
standing of the aetiology of ULVT. thrombolysis and anticoagulation more veins remained open
after thrombolysis and long term problems in the arm were
Anticoagulation infrequent, there being no difference between the groups.38
The effectiveness of thrombolysis in cases of thrombosis of
In many reports and reviews the beneficial effect of antico- central venous cannulae has also been reported.30,31
agulation has been advocated,1–5,20,23,25,28,31,38 but there are Most reports have used urokinase in treatment but
no comparative studies of non-medical conservative treat- streptokinase and tissue plasminogen activator (tPA) are
ment10 and anticoagulation. Extrapolating from studies on reported to be as effective.6,7,77–79,81,83–86,88,89 Currently, in
deep venous thrombosis in the lower extremity, the value selected cases, we use catheter directed low dose thromboly-
of anticoagulation in ULVT has also been established. sis into the thrombus with repeated venography and adjust-
Intravenous heparin or LMWH1–5,79,86,87 combined with ment of the catheter position every 12 hours. We normally
3–12 months of oral anticoagulation is safe and reliable use tissue plasminogen activators at a dose of 1–2 mg/hour.
and that has been evaluated reasonably scientifically. This Most centres today use this form of local thrombolytic ther-
view is further endorsed by the circumstantial evidence apy but the risk of potentially dangerous bleeding complica-
that anticoagulation prevents occlusion of collaterals tions must be considered, and it should only be advocated in
and may favourably influence the potential for long selected cases. In our selected series of over 550 patients
term symptoms. There are a few reviews compiling data receiving thrombolytic therapy, however, no major bleeding
on the effectiveness of anticoagulant treatment and which complications were reported although bleeding and peri-
record a low frequency of treatment complications.1,2,5,30 catheter thrombosis at the access site were observed. Our
Pulmonary embolism has also been reported, but there patients are selected on non-evidence-based criteria, such as
have been no fatal cases.2,5–7,21,26 Early re-thrombosis is the probable factors responsible for the development of
infrequent, and more often seen after anticoagulation thrombosis, the magnitude of presenting symptoms and
ceases.2,5–7,21,26 the collateral pattern on diagnostic venography prior to
270 Upper limb vein thrombosis

recommending specific therapy. All these factors, of course, stenting. The likelihood of restenosis seems to be high, cer-
are biased by the authors’ judgement of each individual case. tainly higher than normally reported.68–71
Currently, about 10 per cent of patients with ULVT would be
recommended for thrombolysis.
Normally in secondary axillary/subclavian thrombosis the Surgical thoracic outlet decompression
long term outcome of thrombolysis is less disabling but more
risky than anticoagulation and therefore should be carefully Several reports, especially from North America, recommend
evaluated before being recommended. A few patients will thoracic outlet decompression including first rib resection,
have disabling long term arm symptoms having failed to scalenectomy, thrombectomy, venolysis and even medial
develop collaterals wide enough for acceptable venous out- claviculectomy to reduce long term disability, especially in
flow, but it is impossible to select them at the outset. the subgroup of patients with effort induced throm-
bosis.6,7,26,78 Using more aggressive surgical decompression
treatment in acute cases, i.e. those with a history of less than
Treatment options 14 days, only 25 per cent are reported to suffer from chronic
arm disabilities.6,7,26,28,81,84 The case series described, how-
• Anticoagulant therapy – intravenous heparin, ever, differ and not only include patients with effort induced
LMWH, oral (longer term) thrombosis but also what we would classify as spontaneous
• Catheter directed thrombolysis – urokinase, tPA, primary thrombosis. The anatomical narrowing, justifying
streptokinase thoracic outlet decompression, often also exists in the con-
• Angioplasty  stenting tralateral arm. Nevertheless, development of bilateral throm-
• Thoracic outlet decompression – first rib resection, bosis is only seen in a small subset of patients (6 per cent).7
scalenectomy The decision to decompress is often based on venographic
• Venolysis – excision medial evidence of narrowing after thrombolysis, representing the
clavicle/subclavius/costoclavicular ligament phlebitic reaction around the vein in which thrombus had
• Jugular vein transposition or saphenous vein been dissolved. Initial symptoms, or a venographically exist-
bypass  AV fistula ing stenosis, are not useful pointers in predicting long term
arm disability.
In performing a first rib resection care should be given to
Angioplasty and stenting excise its most medial part for optimal vein decompres-
sion.7,81,84 Medial claviculectomy may be an effective way of
One advantage of thrombolytic therapy is that completion thoracic decompression, especially if more medial compon-
venography will show residual stenosis, compression or ents, such as the subclavius muscle, costoclavicular ligament
fibrosis. These lesions might indicate a risk of subsequent and the most medial part of the first rib are responsible.92
re-thrombosis and a less favourable long term outcome. Transaxillary resection of the first rib, because it approaches
Based on this, further intervention may be proposed. In the problem from below, carries the advantage of minimising
primary axillary/subclavian vein thrombosis angioplasty is the risk of interfering with existing collateral veins.
often used after thrombolysis.6,84,85 Other centres are more First rib resection has been advocated after a suitable
restrictive in their use of angioplasty.7,82,86 The restenosis period subsequent to thrombolysis, if symptoms or morpho-
and re-thrombosis rate is high if not combined with surgi- logical evidence of residual subclavian vein stenosis persists
cal thoracic outlet decompression.6 Supplementary stent- (Fig. 22.5).6,78,83,85,89 Equally, many recommend decom-
ing should not be undertaken as many reports have shown pression at the time of initial presentation after throm-
that the narrow anatomical space in the costoclavicular bolysis.7,80,82,84,90 Both policies appear to give durable
region will frequently cause stent fracture, distortion and symptomatic relief, and few complications have been
re-thrombosis.91 Currently, angioplasty can be used in pri- reported in both early and more selective decompression
mary thrombosis of the axillary/subclavian vein, but the later.7,80 Advocates of immediate exploration feel that to
benefits of such a procedure are restricted to short remain- delay runs a risk of re-thrombosis and missing the window of
ing lesions. Stenting should not be employed except in opportunity to achieve patency and a durable outcome. On
association with decompression of the thoracic outlet,84 the other hand, a delayed operation is more selective: patients
although we have seen stents break despite decompression. with effort induced thrombosis who benefit from thrombol-
Recanalisation of chronically stenosed or occluded seg- ysis and have no residual stenosis at follow-up, do not require
ments could be an alternative if secondary thrombosis has surgery as compression is not a major causative factor.
not been caused by narrowing of the costoclavicular space.80 The value of venolysis, normally requiring a supraclavic-
In patients with previously inserted haemodialysis catheters, ular approach, is apparent in some studies but is not utilised
a high number of central vein abnormalities is often seen.68 routinely by other experts.6,7,28,78,81 Such an approach may
These stenoses or occlusions are caused by indwelling large jeopardise the collaterals. It is important to be aware, how-
lumen catheters and could be treated by angioplasty or ever, that most recommendations of aggressive treatment
 Treatment alternatives 271

(a) (b)

Figure 22.5 (a–c) A 20-year-old man working as a butcher, who, 12

months after developing a probable effort induced thrombosis, had
remaining severe arm symptoms in spite of recanalisation of the
axillary/subclavian vein. Note the fibrous trabeculae. After first rib
(c)
resection the vein was stented but symptoms were only partly
alleviated

modalities are based on uncontrolled case series. Some stud- chronic symptoms can be relieved by aggressive decompres-
ies, however, are based on the subgroup of young patients sion and an endovascular treatment modality in about 50
with effort induced thrombosis in whom late arm sequelae per cent of cases.7,80
are reduced by more aggressive therapy and therefore, such
an approach is to be advocated.6,7,80,82,90 Nevertheless, it is
difficult to base one’s treatment strategy on these reports. Conclusions
The frequent venographic finding of stenosis of the sub-
clavian vein in ‘normal’ individuals as well as in the con- Optimal management of subclavian/axillary vein throm-
tralateral arm of patients with axillary/subclavian thrombosis bosis remains contentious. There are numerous under-
but no higher incidence of bilateral upper arm thrombosis,7 lying causes and multiple treatment options but no
argues strongly against thoracic outlet compression being comparative trials. The aetiology falls naturally into two
the major cause of primary subclavian thrombosis. Thoracic categories: primary, due to ‘effort’ or ‘spontaneous’
decompression can be a valuable tool, but the role of antico- thrombosis, and secondary, due to catheter/pacemaker
agulation must not be forgotten. placement, malignancy or drug abuse. Recent experience
has seen a relative increase in the secondary type.
Swelling is present in nearly all cases at presentation. Pain
Chronic occlusion and superficial vein dilatation occur in up to half of the
cases. Pulmonary embolism occurs in both primary (2
Venous reconstruction has been attempted in extreme cir-
per cent), and more frequently, secondary thrombosis (7
cumstances where distressing symptoms persist.80,83,90 The
per cent). It may be fatal. Long term symptoms, follow-
use of the ipsilateral internal jugular vein, which is divided
ing primary or secondary thrombosis, occur in about a
high in the neck and anastomosed to the distal and unoc-
quarter of the patients, and a small but definite propor-
cluded subclavian vein, seems to have been technically suc-
tion is significantly incapacitated. This is probably more
cessful.93 This procedure may be combined with an AV fistula
frequent among patients with effort induced thrombosis.
and/or excision of the clavicle. Bypasses using the saphenous
Anticoagulants should be given in the form of heparin
vein have had mixed results.
or low molecular weight heparin and then oral anticoagu-
In some severe chronic cases, often referred from other
lation to address the risks of PE. They may also improve
centres to our specialised unit in Malmö , thoracic outlet
the long term symptomatic outcome. Local thrombolysis
decompression combined with venous recanalisation and
has been shown to have a high technical success rate. Overall
stenting has given relief of symptoms but it has not become
the justification for more aggressive primary treatment
standard treatment. Other series have shown that those with
272 Upper limb vein thrombosis

6 Machleder HI. Evaluation of a new treatment strategy for

such as thrombectomy, thrombolysis, stenting, first rib Paget-Schroetter syndrome: spontaneous thrombosis of the
resection and venolysis remains controversial. Only a axillary-subclavian vein. J Vasc Surg 1993; 17: 305–15;
prospective randomised trial, however, will determine discussion 316–17.
whether its routine use in preference or in addition to 7 Urschel HC Jr, Razzuk MA. Paget-Schroetter syndrome: what
anticoagulants is justified. First rib resection may be is the best management? Ann Thorac Surg 2000; 69: 1663–8;
employed in the few seriously incapacitated symptomatic discussion 1668–9.
patients with effort induced thrombosis. The essential 8 Bailey H, Love M. King Henry IV of France – Battle of Ivry in
dilemma is that there remains no way of knowing which 1590. In: Rains AJ, Ritchie HD (eds). Bailey & Love’s Short
Practice of Surgery. London: Lewis, HK, 1981: 175–84.
patient would be one of the few to be left with truly per-
9 Paget J. Clinical Lectures and Essays. London: Longman’s
sisting debilitating symptoms if treated with anticoagula-
Green, 1875.
tion alone. The use of surgical thoracic decompression as a 10 Hughes ESR. Venous obstruction in the upper extremity
primary procedure at the time of presentation, inevitably (Paget-Schroetter’s syndrome). A review of 320 cases. Int
means unnecessary operations in the majority. Surgical Abs Surg 1949; 88: 89–127.
bypass procedures, angioplasty and stenting have been 11 Pittam MR, Darke SG. The place of first rib resection in the
reported with mixed results; they should be reserved for management of axillary-subclavian vein thrombosis. Eur J
the rare and exceptional case. Vasc Surg 1987; 1: 5–10.
12 Dunant JH. Subclavian vein obstruction in thoracic outlet
syndrome. Int Angiol 1984; 3: 157–9.
13 Broome A, Eklof B, Gothlin J, Hallbook T. [Phlebography and
plethysmography in venous obstruction of the arm]. Radiology
Key references 1971; 11: 155–61.
14 Stevenson IM, Parry EW. Radiological study of the aetiological
Hicken GJ, Ameli FM. Management of subclavian-axillary vein factors in venous obstruction of the upper limb. J Cardiovasc
thrombosis: a review. Can J Surg 1998; 41: 13–25. Surg (Torino) 1975; 16: 580–5.
Hingorani A, Ascher E, Lorenson E, et al. Upper extremity deep venous 15 Galea MH, Berridge DC, Gregson RHS, et al. Axillary/subclavian
thrombosis and its impact on morbidity and mortality rates in vein thrombosis: a clinical and radiological evaluation of
a hospital-based population. J Vasc Surg 1997; 26: 853–60. conservative management. Phlebology 1990; 5: 193–9.
Lindblad B, Tengborn L, Bergqvist D. Deep vein thrombosis of the 16 Barker NW, Nygaard KK, Watters W. Statistical study of
axillary-subclavian veins: epidemiologic data, effects of postoperative venous thrombosis and pulmonary embolism.
different types of treatment and late sequelae. Eur J Vasc Location of thrombosis. Relation of thrombosis and embolism.
Surg 1988; 2: 161–5. Proc Mayo Clin 1941; 16: 33–7.
Monreal M, Alastrue A, Rull M, et al. Upper extremity deep venous 17 Kunkel JM, Machleder HI. Treatment of Paget-Schroetter
thrombosis in cancer patients with venous access devices- syndrome. A staged, multidisciplinary approach. Arch Surg
prophylaxis with a low molecular weight heparin (Fragmin). 1989; 124: 1153–7; discussion 1157–8.
Thromb Haemost 1996; 75: 251–3. 18 Kerr TM, Lutter KS, Moeller DM, et al. Upper extremity venous
Thompson RW, Schneider PA, Nelken NA, et al. Circumferential thrombosis diagnosed by duplex scanning. Am J Surg 1990;
venolysis and paraclavicular thoracic outlet decompression 160: 202–6.
for ‘effort thrombosis’ of the subclavian vein. J Vasc Surg 19 Lindblad B, Bornmyr S, Kullendorff B, Bergqvist D. Venous
1992; 16: 723–32. haemodynamics of the upper extremity after subclavian vein
thrombosis. Vasa 1990; 19: 218–22.
20 Adams JT, McEvoy RK, DeWeese JA. Primary deep venous
thrombosis of upper extremity. Arch Surg 1965; 91: 29–42.
21 Coon WW, Willis PW 3rd. Thrombosis of axillary and
REFERENCES subclavian veins. Arch Surg 1967; 94: 657–63.
22 Tilney NL, Griffiths HJG, Edwards EA. Natural history of major
1 Becker DM, Philbrick JT, Walker FBT. Axillary and subclavian venous thrombosis of the upper extremity. Arch Surg 1970;
venous thrombosis. Prognosis and treatment. Arch Intern 101: 792–6.
Med 1991; 151: 1934–43. 23 Donayre CE, White GH, Mehringer SM, Wilson SE. Pathogenesis
2 Hicken GJ, Ameli FM. Management of subclavian-axillary determines late morbidity of axillosubclavian vein thrombosis.
vein thrombosis: a review. Can J Surg 1998; 41: 13–25. Am J Surg 1986; 152: 179–84.
3 Hingorani A, Ascher E, Lorenson E, et al. Upper extremity deep 24 Horattas MC, Wright DJ, Fenton AH, et al. Changing concepts
venous thrombosis and its impact on morbidity and mortality rates of deep venous thrombosis of the upper extremity – report of
in a hospital-based population. J Vasc Surg 1997; 26: 853–60. a series and review of the literature. Surgery 1988; 104: 561–7.
4 Lindblad B, Tengborn L, Bergqvist D. Deep vein thrombosis of 25 Hauptli W, Schmitt HE, Huber P, et al. [Etiology and long-term
the axillary-subclavian veins: epidemiologic data, effects of course of subclavian vein thrombosis with reference to acute
different types of treatment and late sequelae. Eur J Vasc therapy]. Schweiz Med Wochenschr 1989; 119: 647–52.
Surg 1988; 2: 161–5. 26 AbuRahma AF, Sadler D, Stuart P, et al. Conventional versus
5 Prandoni P, Bernardi E. Upper extremity deep vein thrombosis. thrombolytic therapy in spontaneous (effort) axillary-subclavian
Curr Opin Pulm Med 1999; 5: 222–6. vein thrombosis. Am J Surg 1991; 161: 459–65.
 References 273

27 Malcynski J, O’Donnell TF Jr, Mackey WC, Millan VA. Long-term 46 Ladefoged K, Davidsen HG, Efsen F, et al. [Complications of
results of treatment for axillary subclavian vein thrombosis. prolonged parenteral nutrition. Experiences from an 11-year
Can J Surg 1993; 36: 365–71. period]. Ugeskr Laeger 1979; 141: 3356–60.
28 Martin M, Brors G. [Subclavian vein thrombosis: epidemiologic 47 Magalhaes P, Humair L, Jacot C, de Torrente A. [Venous
data of the PHLEKO (phlebothrombosis conservative treatment) thrombosis in central venous catheterization: incidence and
Study]. Vasa 1995; 24: 120–5. promoting factors]. Schweiz Med Wochenschr 1986; 116:
29 Adelman MA, Stone DH, Riles TS, et al. A multidisciplinary 579–82.
approach to the treatment of Paget-Schroetter syndrome. 48 Balestreri L, De Cicco M, Matovic M, et al. Central venous
Ann Vasc Surg 1997; 11: 149–54. catheter-related thrombosis in clinically asymptomatic oncologic
30 Beygui RE, Olcott Ct, Dalman RL. Subclavian vein thrombosis: patients: a phlebographic study. Eur J Radiol 1995; 20: 108–11.
outcome analysis based on etiology and modality of treatment. 49 Martin C, Viviand X, Saux P, Gouin F. Upper-extremity deep
Ann Vasc Surg 1997; 11: 247–55. vein thrombosis after central venous catheterization via the
31 Marie I, Levesque H, Cailleux N, et al. [Deep venous thrombosis axillary vein. Crit Care Med 1999; 27: 2626–9.
of the upper limbs. Apropos of 49 cases]. Rev Med Interne 50 Allen AW, Megargell JL, Brown DB, et al. Venous thrombosis
1998; 19: 399–408. associated with the placement of peripherally inserted central
32 Sakakibara Y, Shigeta O, Ishikawa S, et al. Upper extremity catheters. J Vasc Interv Radiol 2000; 11: 1309–14.
vein thrombosis: etiologic categories, precipitating causes, 51 Male C, Chait P, Andrew M, Hanna K, Julian J, Mitchell L, for
and management. Angiology 1999; 50: 547–53. the PARKAA investigators. Central venous line-related
33 Zell L, Scheffler P, Heger M, et al. The Paget-Schroetter thrombosis in children: association with central venous line
syndrome: work accident and occupational disease. Ann location and insertion technique. Blood 2003; 101: 4273–8.
Acad Med Singapore 2001; 30: 481–4. 52 Frank DA, Meuse J, Hirsch D, et al. The treatment and outcome
34 Sundqvist SB, Hedner U, Kullenberg HK, Bergentz SE. Deep of cancer patients with thromboses on central venous
venous thrombosis of the arm: a study of coagulation and catheters. J Thromb Thrombolysis 2000; 10: 271–5.
fibrinolysis. BMJ (Clin Res Ed) 1981; 283: 265–7. 53 Wu X, Studer W, Skarvan K, Seeberger MD. High incidence of
35 Heron E, Lozinguez O, Alhenc-Gelas M, et al. Hypercoagulable intravenous thrombi after short-term central venous catheteri-
states in primary upper-extremity deep vein thrombosis. zation of the internal jugular vein. J Clin Anesth 1999; 11:
Arch Intern Med 2000; 160: 382–6. 482–5.
36 Leebeek FW, Stadhouders NA, van Stein D, et al. 54 Soto-Velasco JM, Steiger E, Rombeau JL, et al. Subclavian vein
Hypercoagulability states in upper-extremity deep venous thrombophlebitis: complication of total parenteral nutrition
thrombosis. Am J Hematol 2001; 67: 15–19. (TPN). Cleve Clin Q 1984; 51: 159–66.
37 Fijnheer R, Paijmans B, Verdonck LF, et al. Factor V Leiden in 55 Schwarz RE, Coit DG, Groeger JS. Transcutaneously tunneled
central venous catheter-associated thrombosis. Br J Haematol central venous lines in cancer patients: an analysis of device-
2002; 118: 267–70. related morbidity factors based on prospective data collection.
38 Sabeti S, Schillinger M, Mlekusch W, Haumer M, et al. Ann Surg Oncol 2000; 7: 441–9.
Treatment of subclavian-axillary vein thrombosis: long-term 56 Scholz G, Loewe KR. [Puncture of the subclavian vein and its
outcome of anticoagulation versus systemic thrombolysis. complications from the pathological and anatomical
Thromb Res 2002; 108: 279–85. viewpoint]. Med Welt 1969; 41: 2248–51.
39 Hoshal VL Jr, Ause RG, Hoskins PA. Fibrin sleeve formation on 57 Brismar B, Hardstedt C, Malmborg AS. Bacteriology and
indwelling subclavian central venous catheters. Arch Surg phlebography in catheterization for parenteral nutrition. A
1971; 102: 253–8. prospective study. Acta Chir Scand 1980; 146: 115–19.
40 Ryan JA Jr, Abel RM, Abbott WM, et al. Catheter complications 58 Andersson T, Brodin M, Lindberg E, Wickbom G. Frequency of
in total parenteral nutrition. A prospective study of 200 thrombosis in patients with central venous catheter. Svensk
consecutive patients. N Engl J Med 1974; 290: 757–61. Kirurgi 1981; 36: 189.
41 Lindblad B, Efsing HO, Mark J, Wolff T. Central venous 59 Starkhammar H, Bengtsson M, Morales O. Fibrin sleeve
catheters and thromboembolic complications. A comparison formation after long term brachial catheterisation with an
of three different materials. Thromb Haemost 1981; 46: implantable port device. A prospective venographic study.
318. Eur J Surg 1992; 158: 481–4.
42 Ahmed N. Thrombosis after central venous cannulation. Med 60 Cobos E, Dixon S, Keung YK. Prevention and management of
J Aust 1976; 1: 217–20. central venous catheter thrombosis. Curr Opin Hematol 1998;
43 Bennegård K, Curelaru I, Gustavsson B, et al. Material 5: 355–9.
thrombogenicity in central venous catheterization. I. A 61 Randolph AG, Cook DJ, Gonzales CA, Andrew M. Benefit of
comparison between uncoated and heparin-coated, long heparin in central venous and pulmonary artery catheters: a
antebrachial, polyethylene catheters. Acta Anaesthesiol meta-analysis of randomized controlled trials. Chest 1998;
Scand 1982; 26: 112–20. 113: 165–71.
44 Braun U, Fassolt A, Teske HJ. [Phlebographic studies on 62 Boraks P, Seale J, Price J, et al. Prevention of central venous
foreign body induced thrombosis caused by infraclavicular catheter associated thrombosis using minidose warfarin in
subclavian catheterization]. Anaesthesist 1970; 19: patients with haematological malignancies. Br J Haematol
432–7. 1998; 101: 483–6.
45 Fassolt A, Braun U. [The prophylaxis of thrombosis due to 63 Bern MM, Bothe A Jr, Bistrian B, et al. Prophylaxis against
central venous catheter with low-dose-heparin (author’s central vein thrombosis with low-dose warfarin. Surgery
transl)]. Schweiz Rundsch Med Prax 1978; 67: 57–60. 1986; 99: 216–21.
274 Upper limb vein thrombosis

64 Bern MM, Lokich JJ, Wallach SR, et al. Very low doses of warfarin devices – prophylaxis with a low molecular weight heparin
can prevent thrombosis in central venous catheters. A (Fragmin). Thromb Haemost 1996; 75: 251–3.
randomized prospective trial. Ann Intern Med 1990; 112: 423–8. 80 Molina JE. Need for emergency treatment in subclavian vein
65 Kar AK, Ghosh S, Majumdar A, et al. Venous obstruction after effort thrombosis. J Am Coll Surg 1995; 181: 414–20.
permanent pacing. Indian Heart J 2000; 52: 431–3. 81 Azakie A, McElhinney DB, Thompson RW, et al. Surgical
66 Sticherling C, Chough SP, Baker RL, et al. Prevalence of central management of subclavian-vein effort thrombosis as a result
venous occlusion in patients with chronic defibrillator leads. of thoracic outlet compression. J Vasc Surg 1998; 28: 777–86.
Am Heart J 2001; 141: 813–16. 82 Angle N, Gelabert HA, Farooq MM, et al. Safety and efficacy
67 Stoney WS, Addlestone RB, Alford WC Jr, et al. The incidence of early surgical decompression of the thoracic outlet for
of venous thrombosis following long-term transvenous pacing. Paget-Schroetter syndrome. Ann Vasc Surg 2001; 15:
Ann Thorac Surg 1976; 22: 166–70. 37–42.
68 Quinn SF, Schuman ES, Demlow TA, et al. Percutaneous 83 Feugier P, Aleksic I, Salari R, et al. Long-term results of venous
transluminal angioplasty versus endovascular stent placement revascularization for Paget-Schroetter syndrome in athletes.
in the treatment of venous stenoses in patients undergoing Ann Vasc Surg 2001; 15: 212–18.
hemodialysis: intermediate results. Vasc Interv Radiol 1995; 84 Kreienberg PB, Chang BB, Darling RC, et al. Long-term results
6: 851–5. in patients treated with thrombolysis, thoracic inlet
69 Surratt RS, Picus D, Hicks ME, et al. The importance of decompression, and subclavian vein stenting for Paget-
preoperative evaluation of the subclavian vein in dialysis Schroetter syndrome. J Vasc Surg 2001; 33: S100-5.
access planning. Am J Roentgenol 1991; 156: 623–5. 85 Lokanathan R, Salvian AJ, Chen JC, et al. Outcome after
70 Lumsden AB, MacDonald MJ, Isiklar H, et al. Central venous thrombolysis and selective thoracic outlet decompression for
stenosis in the hemodialysis patient: incidence and efficacy of primary axillary vein thrombosis. J Vasc Surg 2001; 33:
endovascular treatment. Cardiovasc Surg 1997; 5: 504–9. 783–8.
71 Jean G, Vanel T, Chazot C, et al. Prevalence of stenosis and 86 Lee WA, Hill BB, Harris EJ Jr, et al. Surgical intervention is not
thrombosis of central veins in hemodialysis after a tunneled required for all patients with subclavian vein thrombosis. J
jugular catheter. Nephrologie 2001; 22: 501–4. Vasc Surg 2000; 32: 57–67.
72 Gray RJ, Horton KM, Dolmatch BL, et al. Use of Wallstents for 87 Savage KJ, Wells PS, Schulz V, et al. Outpatient use of low
hemodialysis access-related venous stenoses and occlusions molecular weight heparin (Dalteparin) for the treatment of
untreatable with balloon angioplasty. Radiology 1995; 195: deep vein thrombosis of the upper extremity. Thromb Haemost
479–84. 1999; 82: 1008–10.
73 Verstandig AG, Bloom AI, Sasson T, et al. Shortening and 88 Theiss W, Wirtzfeld A. [Fibrinolytic treatment of acute and
migration of Wallstents after stenting of central venous subacute thromboses of the deep veins of the shoulder girdle.]
stenosis in hemodialysis patients. Cardiovasc Intervent Radiol Dtsch Med Wochenschr 1982; 107: 933–6.
2003; 26: 58–64. 89 Strange-Vognsen HH, Hauch O, Andersen J, Struckmann J.
74 Swinton NW, Edgett JW, Hall RJ. Primary subclavian-axillary Resection of the first rib, following deep arm vein thrombolysis
vein thrombosis. Circulation 1968; 38: 737–45. in patients with thoracic outlet syndrome. J Cardiovasc Surg
75 Campbell CB, Chandler JG, Tegtmeyer CJ, Bernstein EF. Axillary, 1989; 30: 430–3.
subclavian and brachiocephalic vein obstruction. Surgery 90 Lee MC, Grassi CJ, Belkin M, et al. Early operative intervention
1977; 82: 816–26. after thrombolytic therapy for primary subclavian vein
76 Harley DP, White RA, Nelson RJ. Pulmonary embolism secondary thrombosis: an effective treatment approach. J Vasc Surg
to thrombosis of the arm. Am J Surg 1984; 141: 221–4. 1998; 27: 1101–8.
77 Reed JD, Harman JT, Harris V. Regional fibrinolytic therapy for 91 Meier GH, Pollak JS, Rosenblatt M, et al. Initial experience
iatrogenic subclavian vein thrombosis. Semin Interv Radiol with venous stents in exertional axillary-subclavian vein
1992; 9: 183–9. thrombosis. J Vasc Surg 1996; 24: 974–83.
78 Thompson RW, Schneider PA, Nelken NA, et al. Circumferential 92 Green RM, Waldman D, Ouriel K, et al. Claviculectomy for
venolysis and paraclavicular thoracic outlet decompression subclavian venous repair: long-term functional results. J Vasc
for ‘effort thrombosis’ of the subclavian vein. J Vasc Surg Surg 2000; 32: 315–21.
1992; 16: 723–32. 93 De Weese JA. Results of surgical treatment of axillary
79 Monreal M, Alastrue A, Rull M, et al. Upper extremity deep subclavian venous thrombosis. In: Bergan JJ, Yao JST (eds).
venous thrombosis in cancer patients with venous access Venous Disorders. London: Saunders WB, 1991: 421–33.
 SECTION
5
Thoracoabdominal Catastrophes

23. The emergency aortic aneurysm 277

24. Stenting in acute aortic dissection 287

25. Prosthetic aortic graft infection 297

26. The acutely compromised renal artery 309

27. Renal artery aneurysms 315

28. Visceral artery aneurysms 325

29. Mesenteric ischaemia 335

This page intentionally left blank
 23
The Emergency Aortic Aneurysm

JONATHAN REFSON, JOHN HN WOLFE

The problem 277 Operative technique 279

Diagnosis 277 Special preoperative problems 281
Differential diagnosis 278 Postoperative care 283
Patient selection 278 Recent advances 283
Patient transfer 279 References 284
Consent 279

THE PROBLEM alert the clinician to the possibility of a ruptured AAA, par-
ticularly where the patient is known to have such an
aneurysm. On examination, the patient may be obviously
Emergency aneurysms pose a difficult healthcare problem, shocked, pale, cold, clammy and agitated. Abdominal exam-
as health resources are currently unable to provide the neces- ination may reveal a pulsatile mass, but it may be difficult to
sary expertise to all patients at the point of presentation. palpate, particularly in an obese individual; it is also import-
Therefore, diagnosis of this problem needs to be accurate ant to feel the peripheral pulses. The investigations of these
with rapid passage of the patient from admitting unit to patients should be kept to a minimum and an electrocardio-
operating theatre, whether in the same hospital or after gram (ECG) is really all that is required to rule out possible
transfer. The diagnosis, management and common prob- myocardial infarction, which may be a differential diagnosis.
lems of this condition are discussed. In the haemodynamically challenged patient radiological
The ratio of planned aneurysm surgery to urgent and investigation is contraindicated, in particular computed
emergency aneurysms has remained remarkably constant at tomography (CT) scanning, as it will only delay the patient’s
roughly 2:1 over the past 30 years.1,2 Emergency aneurysms transfer to the operating theatre for definitive treatment.
can present with a history of back, loin or abdominal pain Resuscitation should be simple and minimal; if the
associated with an episode of circulatory collapse or persist- patient is talking sensibly there is enough blood pressure
ent hypotension; approximately 70 per cent will not be to perfuse the vital organs. High flow oxygen should be
known to have an abdominal aortic aneurysm (AAA).3 The delivered by facemask, venous access should be achieved
level of urgency dictates the management of the ‘emergency with two large bore cannulae (see Chapter 7B) and no fluid
aneurysm’; has the aneurysm ruptured or is the rupture need be given. If fluid is administered to restore the
impending? In order to make the discussion of the subject patient’s blood pressure to normality, further bleeding may
easier, we have drawn a distinction between urgent and occur with catastrophic consequences. Blood should be taken
emergency aneurysms. Urgent aneurysms are those where for baseline investigations, namely blood count, clotting
the aneurysm has obviously become symptomatic but has studies, typing and crossmatching 10 units of blood and
not caused the patient to be hypotensive. renal function. It is advisable to ensure that fresh frozen
plasma and platelets are also available, as the coagulopathy
associated with both haemorrhage and massive transfu-
DIAGNOSIS
sion causes troublesome and life-threatening bleeding. No
further intervention is required in the emergency depart-
The history of sudden back or abdominal pain with cardio- ment and the patient should be transferred directly to the
vascular collapse in a middle aged or elderly patient should operating theatre.
278 The emergency aortic aneurysm

Suspected ruptured/leaking
AAA

Cardiovascular stability History of loss of

No hypotension consciousness/collapse
No history of collapse and or hypotension
Diagnostic uncertainty

Go directly to theatre with patient

contact anaesthetist /arrange
Urgent CT with contrast
blood crossmatch, platelets and
clotting factors

AAA with no leak No AAA, other AAA with leak

pathology

Treat accordingly

No tenderness Tender AAA AAA size already known

Scan suggests sudden
rapid growth

Figure 23.1 Algorithm of management

of suspected ruptured abdominal aortic Arrange elective Repair on next
aneurysms (AAAs). CT, computed repair according to available elective
size list
tomography

In the scenario of the urgent aneurysm, the patient aneurysms in their late teens and early twenties. Other con-
may complain of pain or tenderness but hypotension will ditions that can be difficult to distinguish from the rup-
not have been documented. Therefore, it is acceptable to tured AAA are myocardial infarction, pancreatitis and
perform an urgent CT scan to assess the extent of the renal stone disease, but a quick and careful history, exam-
aneurysm and to establish whether or not it has leaked. It ination and simple investigations should help to establish
will also identify small leaks that have not caused circula- the diagnosis. It is important to note that renal stones
tory collapse and some thoracoabdominal aneurysms that rarely present for the first time in the over-fifties.
should be transferred to a specialist centre. This time win-
dow allows a more indepth assessment of the patient’s
PATIENT SELECTION
overall fitness for surgery, but in the event that a leak is
diagnosed on CT scanning, surgery should not be delayed.
Where there has been hypotension, however, time is of the An untreated leaking aneurysm has a mortality approach-
essence in getting the patient to theatre for surgery. An ing 100 per cent.4 Not to operate will, therefore, almost
algorithm (Fig. 23.1) provides pathways of care in manag- certainly result in the patient’s demise. In these circum-
ing patients with suspected AAA rupture. stances the patient and relatives should be advised that if no
surgery were to be undertaken then the patient would die,
but also that operative intervention carries a 50/50 chance
DIFFERENTIAL DIAGNOSIS of survival. Unless the patient flatly refuses surgery, or has
an extensive thoracoabdominal (type II) aneurysm, we
Age is important in distinguishing patients with aneurys- offer surgery to all-comers. Even if patients have such ser-
mal disease. As this is a problem occurring almost invari- ious comorbidity that they have been turned down for an
ably in the over-fifties, a patient under this age is unlikely elective procedure it is appropriate, now that the AAA has
to have leaked or ruptured an aneurysm. There is, how- ruptured, to offer intervention. This approach is rein-
ever, an important rare group of exceptions to this rule forced by aggressive initial support in the intensive care
including such conditions as Marfan’s and Ehlers Danlos unit during the first 24–48 hours, but unless there are clear
syndromes. Congenital connective tissue disorders are the signs of recovery that support may then be withdrawn.
pathological basis of these conditions and patients develop This is because preoperative assessment is very difficult in
 Operative technique 279

the absence of prior knowledge about the patient, indeed So far, we have dealt with a scenario where the expertise
some surgeons will not operate on anuric patients with low needed to manage the patient is available in the hospital to
blood pressure.5 which the patient has been admitted. Not infrequently,
however, an aneurysm presents to a team without the neces-
sary expertise to deal with the patient’s problem. If this is
Contraindications to surgery the case the patient needs to be transferred to a unit able to
perform the surgery, with the necessary intensive post-
• Patient alert but flatly refuses operative support. The logistical problems surrounding this
• Extensive thoracoabdominal (type II) aneurysm exercise are beyond the scope of this chapter, but ideally
the patient should be stable for transfer and the receiving
hospital should have a vascular surgeon and an intensive
In the urgent scenario certain risk factors strongly influ- care bed available. In the ambulance the patient should
ence outcome for both elective and urgent aneurysm sur- have a nurse and medical escort. In the scenario where the
gery. Age seems to be the strongest predictor with patients patient remains unstable, and in spite of the potential dan-
over the age of 70 having a mortality rate between five and gers, it is probably preferable to transfer rather than to
seven times that of those under the age of 70.6,7 attempt repair.9
Renal function also relates to outcome; patients with This situation of insufficient cover is being addressed
elevated creatinine beyond the normal range, i.e. over and some hospital groups have developed collaborative
50 per cent loss of nephron mass, have a mortality rate two networks to provide full vascular cover 24 hours a day,
to three times normal creatinine values. 7 days a week (see Chapters 1A–1C). In this situation hos-
Poor respiratory function, especially in the form of pitals provide a collective on-call rota and one surgeon
chronic obstructive airways disease (COAD), is another covers all the hospitals. The surgeon, therefore, may have
predictor of poor outcome.8 Many patients with aneurys- to travel to the patient rather than the other way round.
mal disease are smokers and, as such, will have a chronic This ensures that patients need not be exposed to a risky
respiratory problem. This does not pose an immediate transfer.
threat, but makes the postoperative management and
weaning from ventilation much more difficult and predis-
poses the patient to hospital acquired pneumonia. Serious
CONSENT
consideration should be given to early tracheostomy to aid
bronchial toilet and in weaning the patient off the ventilator
and avoiding the deleterious effects of sedation. Placing an If the patient is fully conscious it is important to explain
epidural catheter for analgesia postoperatively may aid the nature of the procedure in simple terms, stressing the
respiratory function and, provided that the patient’s clotting urgent need for surgery. In the circumstances, a detailed
function is not deranged, should be done as early as possible. discussion of all the possible complications is less relevant.
If family members are present it is important that they
understand the situation and they should be included in
Predictors of poor outcome in ruptured AAAs any discussion. If the patient is not able to give his/her own
consent the procedure can continue on the basis that it is in
• Raised creatinine levels, 50 per cent loss of the patient’s best interests. With the advent of living wills,
nephron mass – mortality 2–3 times normal however, it is prudent to enquire whether the patient has
• Chronic respiratory failure (COAD) carries a high expressed an objection to surgery.
mortality rate
• Advancing age – beyond 70 years – mortality 5–7
times normal
OPERATIVE TECHNIQUE

The patient is placed supine on the operating table with the

PATIENT TRANSFER arms abducted. This allows the anaesthetist access for
insertion of arterial and venous cannulae. Two wide bore
In hospitals fully equipped to look after ruptured cannulae should be inserted for rapid transfusion access
aneurysms there should be no delay in making the transfer (see Chapter 7B). The patient is prepared and draped
from the accident and emergency department to the oper- awake and the anaesthetist should only induce anaesthesia
ating theatre once the diagnosis is made. The vascular sur- when the surgeon is ready to make the incision. This is
gical team must ensure that all the relevant teams are because the abdominal wall musculature provides tam-
contacted: operating theatre staff, anaesthetists, transfu- ponade, and paralysis of the muscles is often followed by a
sion technicians and the intensive care unit. precipitous fall in blood pressure.
280 The emergency aortic aneurysm

Ideally, three surgeons should be available for the particularly the ascending colic artery. Third, the aneurys-
procedure, and in our case the principal surgeon stands on mal process may compromise the internal iliac arteries and
the left of the patient with the scrub nurse opposite. The if both of these are occluded, as well as the IMA, then sig-
first assistant stands opposite the surgeon and the other moid and pelvic ischaemia may ensue. Fourth, mobilisa-
provides retraction to his or her right. This can be difficult tion of the iliac artery bifurcation lateral to the sigmoid
with the arm abducted. A long midline incision is made colon mesentery may also damage sigmoid arteries.
from the xiphisternum to a point halfway between umbil- These points must be taken into consideration when
icus and pubic symphysis allowing good exposure to the performing the emergency procedure and if the IMA is
abdominal cavity. In the event of a retroperitoneal rupture patent with poor back bleeding the surgeon should be
there will be blood-stained fluid in the peritoneal cavity alerted to the possible need for IMA reimplantation. This
and a retroperitoneal haematoma. reimplantation may also be necessary if one internal iliac
The mobilisation of the aortic neck should continue as artery is aneurysmal and therefore has to be ligated, parti-
for an elective or unruptured case, with evisceration of the cularly if the orifice of the other is compromised, in which
transverse colon and omentum cranially and the small case the surgeon cannot predict continuing internal iliac
bowel to the patient’s right. It is tempting to rush the mobil- artery flow with confidence.
isation of the aorta, but a few extra minutes spent carefully On opening the sac and removing the thrombus sac
dissecting around the neck of the aneurysm will save much dark venous blood may well up copiously; this is usually
time later trying to control venous bleeding from the infer- due to the presence of an aortocaval fistula. In these cir-
ior mesenteric vein, left gonadal vein and renal vein. It is cumstances, external compression by the assistant should
often venous bleeding that leads to exsanguination, not arter- allow the surgeon to oversew the defect from within the
ial. In fact, the retroperitoneal haematoma often facilitates aortic sac using large bites with 2-0 Prolene. At this stage
dissection by lifting the peritoneal layers off the front of the haemostasis is of more importance than the potential for
aorta. Having incised the peritoneum over the aneurysm the narrowing the inferior vena cava. No attempt should be
sac is easily found and should be followed cranially to the left made to mobilise the cava and delay only leads to further
of the duodenum until the neck is found and a suitable dramatic blood loss. At this point haemostasis should have
clamp can then be applied anteroposteriorly on the aorta. been achieved and the anaesthetist should have corrected
The practice of fully mobilising the back of the aorta and the hypovolaemia. The next step is to suture in the graft. It
clamping transversely is not necessary and may cause avul- should be noted that heparin is not given in cases of rup-
sion of a lumbar artery. ture as it may further aggravate a coagulopathy.
Whether the patient is rapidly exsanguinating or if the An appropriately sized graft is chosen. A 120 cm 3-0
leak is contained, iliac artery control is necessary. This is Prolene suture is ideal for the proximal anastomosis. If the
usually straightforward as the ectatic and elongated arteries anastomosis is sound, except for a single bleeding point,
tend to be looped forward facilitating clamping. Extending then a carefully placed 4-0 suture should suffice. If, on the
the peritoneal incision caudally and dissecting along the other hand, entire segments of the back wall appear to have
surface of the aneurysm sac exposes the common iliac torn away then the anastomosis should be fully revised
arteries. No attempt should be made to dissect around and without further delay. Wasting time on a number of poorly
sling the iliac arteries, as they are densely adherent to the placed sutures that also tear out could lead to disastrous
iliac veins, which are easily damaged. If the common iliac further haemorrhage. Occasionally a Teflon felt patch can
arteries are aneurysmal then the external iliac artery should be used to buttress the proximal anastomosis, and equally
be mobilised and clamps should be applied at this level. If a 2 mm button of Teflon can be extremely useful in ensur-
mobilisation is difficult, for example, because it is an ing that a single stitch to a solitary bleeding point in a fri-
inflammatory aneurysm, then balloons can be inserted to able aortic wall does not tear through.
occlude flow, rather than resorting to the use of clamps. Once the proximal anastomosis is secure the surgeon
Having isolated the aneurysm proximally and distally, can concentrate on the distal anastomosis; the latter can
the sac should be opened along its right anterolateral also be performed using 3-0 Prolene and slightly dilated
aspect and thrombus removed. At this point bleeding from iliac arteries should not deter the surgeon from using a
the lumbar arteries is usually encountered. This can be tube graft as this will expedite the procedure. It is advisable
controlled by direct suture with 3-0 Prolene. The inferior to release the clamped iliac arteries at this point and check
mesenteric artery (IMA) may back-bleed and this vessel, for back flow. As no heparin has been given it is quite likely
too, can be controlled with sutures from within the sac. that clot will have been formed. Provided that back flow is
Colonic ischaemia is a potential problem, particularly good a 50 mL bladder syringe charged with heparinised
in emergency aneurysm repair. This may be due to a num- saline should be flushed down each iliac artery and the ves-
ber of reasons. First, clumsy manhandling of the sac may sels reclamped. If back flow is poor and if femoral pulses
lead to ‘trash’ embolisation of the colon as well as more were present preoperatively a suction catheter should be gen-
distal tissues. Second, ligation of the IMA through the tly inserted into the iliac artery and clot aspirated until good
haematoma may cause damage to the colonic arteries, back flow has been restored; if not, balloon thrombectomy
 Special preoperative problems 281

such as pooled platelets and fresh frozen plasma. While

haemostasis is being ensured the feet should be inspected
to assess perfusion. The foot pulses may not be present but
the feet should be pink and blanch on pressure, with rea-
sonably brisk capillary refill.
Once this has been achieved the remaining aneurysm
sac can be closed over the graft with an absorbable suture.
The sac should be plicated to avoid dead space, which then
fills with haematoma. It is important to ensure that the
duodenum is not in contact with the graft as this may lead
to erosion through the duodenum and a subsequent aor-
toenteric fistula. Closing retroperitoneal fat behind the
duodenum using a tongue of omentum helps to avoid this
complication. Care should be taken that the duodenum is
not tethered down anterior to the graft. Any clots should
now be evacuated from the abdominal cavity and a brief
laparotomy performed as in most cases there is no time to
do so prior to aneurysm repair. Care is taken to ensure that
the nasogastric tube is correctly placed. The abdominal
musculature can now be closed with a continuous mass
closure, using a monofilament suture such as nylon.
In the case of free peritoneal rupture exsanguination
occurs rapidly and therefore the technique has to be modi-
fied. On opening the abdomen the supracoeliac aorta can
be compressed manually by an assistant while the surgeon
places their left hand in the open aneurysm sac and plugs
the neck of the aneurysm with the thumb. The aorta can
then be cannulated through the site of rupture with a spig-
Figure 23.2 Creating a ‘swallow tail’ at the lower end of a tube oted 24 Fr Foley catheter and a 50 mL syringe connected to
graft to facilitate iliac anastomosis
the balloon. This is advanced to the suprarenal aorta and
the balloon inflated, so gaining proximal control. Alter-
is indicated. The artery should then be flushed with natively, the thumb allows the right hand to rapidly per-
heparinised saline as described above. Prior to completion form quite accurate blunt dissection around the neck of
of the distal anastomosis back flow should be rechecked. It the aorta. Once this is exposed a clamp can be placed. It
should also be ensured that there is good forward flow in is also possible to temporarily clamp the aorta at the
the graft and that it has been flushed free of any clot. diaphragm: the lesser omentum is divided as is the left crus
Cutting the graft into a ‘swallow tail’ (Fig. 23.2) ensures of the diaphragm, and by pushing the oesophagus to the
that there is sufficient width to splay across the orifices of left, the aorta is approached and then clamped. Whatever
the aneurysmal aorta.10 Without this technique the distal technique is employed, the aim should be to dissect out the
Dacron may be too narrow for the separated, dilated ori- neck as quickly and as safely as possible. If the clamp can-
fices and attempts to pull the orifices together with the not be applied below the renal arteries then using either the
anastomotic stitch may result in stenosis of the origins of Foley catheter or a suprarenal clamp the proximal anasto-
the arteries. mosis is carried out. Once complete, the clamp is moved
At this point the anaesthetist should be warned that the down to the graft, reinstating renal flow.
surgeon is about to release the distal clamp and reperfuse
the legs, and that should be done slowly. A drop in systolic
SPECIAL PREOPERATIVE PROBLEMS
blood pressure will indicate that the legs are being reper-
fused, as will a rise in end tidal carbon dioxide, as the accu-
mulated ischaemic metabolites are cleared from the leg Aortocaval fistula
(see Chapter 2). There is often further bleeding from lum-
bar arteries as they are reperfused from the iliac circulation An aortocaval fistula can be diagnosed preoperatively if
and they should be underrun with a 3-0 Prolene suture. It one or more of the following signs are found: a loud
is important to check that a femoral pulse is palpable on machinery murmur over the aneurysm, pulsatile leg veins
both sides. or high central venous pressure in the presence of hypoten-
Patience is now required to ensure good haemostasis sion. At operation a small volume of arterial blood should
and this may require the administration of clotting factors, come out of the sac but if venous blood continues to well
282 The emergency aortic aneurysm

Figure 23.4 Pearly white appearance of an inflammatory

aneurysm

Figure 23.3 Computed tomography scan demonstrating a

horseshoe kidney
level, but once it has been swept aside the plane and areo-
lar tissue around the aorta can be cleared rapidly and a
out from it then a fistula is almost certainly present. This clamp applied. Having done this the sac can be opened to
loss can be controlled by pressure on the vena cava having the left of the densely adherent duodenum and the prox-
warned the anaesthetist that there will be a fall in venous imal anastomosis can be done. Once the latter has been
return. Repair of the fistula is best effected by suturing it completed the clamp may be moved on to the graft, thus
from within the aorta with large bites of a large suture reperfusing the visceral and renal arteries. It is neither safe
which is remarkably easy. No attempt should be made to nor sensible to attempt to dissect the duodenum off the
disconnect the fistula or repair the cava directly. aortic wall unless the inflammatory process is in its very
early stages and a plane is readily identified.

Horseshoe kidney
Special problems discovered at operation
A horseshoe kidney is extremely rare. If a CT scan (Fig.
23.3) has been performed then the diagnosis will be clear- • Aortocaval fistula
cut allowing a retroperitoneal operative approach to be • Horseshoe kidney
used and the kidney is not involved in the operation. In • Inflammatory aneurysm
most cases the surgeon faced with the unexpected anatomy • Thoracoabdominal aneurysm
associated with this condition. Once the proximal neck has • Other abdominal pathology – gallstones,
been controlled the upper sac is opened and the proximal malignancy
anastomosis is performed. The graft is then tunnelled
through the unopened sac to the iliac arteries. Access to
bleeding lumbar arteries within the sac can be difficult.11
Thoracoabdominal aneurysm

Inflammatory aneurysms In an unstable patient with a ruptured aneurysm there will

be little opportunity to diagnose a thoracoabdominal
These account for about 5 per cent of all AAAs. If diag- aneurysm. Fortunately, the most common site for rupture
nosed preoperatively they can be repaired using a of these aneurysms is in the abdominal element (Fig. 23.5).
retroperitoneal approach. At operation they can be recog- Under these very difficult circumstances an attempt should
nised by the glistening white, pearly carapace involving the be made to control the aneurysm at a narrow point, either
aortic wall (Fig. 23.4). in the pararenal area or the supracoeliac aorta through the
When confronted with the situation in an emergency lesser sac. Once control has been achieved the proximal
the safest approach is to expose the supracoeliac aorta, anastomosis will have to be performed to a baggy aorta,
which is rarely involved in the process. This can be done and this is best achieved by plicating the arterial wall into a
through the lesser sac by cutting the right crus of the large diameter graft. If this anastomosis holds the situation
diaphragm. There is a significant amount of muscle at this can be salvaged and the operation completed.
 Recent advances 283

POSTOPERATIVE CARE

The majority of patients coming to surgery survive the

operative procedure but many of these elderly patients
then succumb to the multiorgan failure precipitated by
massive blood loss and transfusion (see Chapters 4 and
7A). Meticulous attention to postoperative patient care is
crucial if survival is to be improved. The systemic problems
frequently encountered following emergency surgery are
renal failure requiring support in the form of haemofiltra-
tion and respiratory failure requiring ventilatory support.
(a) An intensive care facility is essential with good cooperation
between surgeon and intensivist. If an elderly patient with
extensive comorbidity survives initial emergency surgery,
but is obviously not thriving, then the clinical team must
take stock.
During the first 48 hours intensive support should be
given to the patient and discussions should take place with
the family so that they are fully aware of the situation.
Within this period it will become apparent that some of
these patients are most unlikely to survive. Under these cir-
cumstances the surgeon, intensivist and nursing staff
should discuss among themselves, and then with the fam-
ily, whether escalating support should be withdrawn.

(b)
RECENT ADVANCES
Figure 23.5 (a) Type B aortic dissection, (b) the aneurysmal
abdominal component of which had leaked, necessitating a type IV
thoracoabdominal aneurysm repair Screening

Recent advances in the fields of screening and surveillance of

small aneurysms have refined our knowledge as to whether
In the postoperative phase the full extent of the aneurysm to offer surgery to patients with abdominal aortic aneurysms.
can be assessed and a judgement made as to whether further The Small Aneurysm Study demonstrated that it was safe to
surgery is appropriate. Fortunately, a significant number of pursue a policy of surveillance of aneurysms up to 5.5 cm in
these aneurysms become symptomatic and the surgeon is diameter.13 Some screening studies have suggested that sur-
in a position to investigate the aneurysm prior to surgery. veillance may be safe in an aneurysm of up to 6 cm diam-
If a thoracoabdominal aneurysm is diagnosed under these eter.14 Screening has also demonstrated a 50–60 per cent
circumstances then referral to an appropriate specialist reduction in the incidence of ruptured AAAs in a screened
unit is sensible. That unit will then perform further appro- population.15,16 Currently in the UK 33 per cent of all
priate investigations, including cardiac, respiratory and aneurysm surgery is undertaken on an emergency basis,1,17
renal assessment, prior to elective repair using standard which is nearly double the European average.17
techniques.

Stenting
Associated abdominal pathology
Recent reports have demonstrated that endovascular tech-
Other coexisting pathological conditions are encountered niques can be successfully employed in managing emer-
in about 5 per cent of patients undergoing aneurysm sur- gency AAAs. Others18–20 have demonstrated a 92 per cent
gery.12 With the exception of gallstone disease synchron- success rate in managing leaking aneurysms in cases where
ous procedures should be avoided because of the risk of the anatomy of the aorta was suitable for endovascular
graft sepsis. In the scenario of the ruptured aneurysm the repair. These results compare favourably with open repair.
priority is haemostasis and restoration of circulation and The endovascular technique is currently available to only a
all secondary procedures should be left to a later date, even small minority of centres and should be used by units
if malignancy is found. which regularly perform elective aneurysm stenting.
284 The emergency aortic aneurysm

Conclusions Scott RA, Ashton HA, Lamparelli MJ, et al. A 14-year experience
with 6 cm as a criterion for surgical treatment of abdominal
aortic aneurysm. Br J Surg 1999; 86: 1317–21.
Death from abdominal aortic aneurysmal disease is the
Wilmink AB, Quick CR, Hubbard CS, Day NE. Effectiveness and cost
tenth commonest cause of death in the UK and accounts of screening for abdominal aortic aneurysm: results of a
for 1 per cent of all male deaths.21 The rate of ruptured population screening program. J Vasc Surg 2003; 38: 72–7.
aneurysms across Europe varies from 12 to over 55 per
cent of all aneurysm repairs.17 The reasons for this dis-
parity are not clear, but may relate to healthcare fund-
ing. The survival of all operated ruptures, however, is REFERENCES
remarkably constant across both Europe and the USA at
35–60 per cent.23,24 Dramatic improvements in surgical
1 Castleden WM, Mercer JC. Abdominal aortic aneurysms in
and anaesthetic techniques, as well as intensive care sup- Western Australia: descriptive epidemiology and patterns of
port, have had little impact. rupture. Br J Surg 1985; 72: 109–12.
The distinction between urgent and emergency 2 Samy AK, MacBain G. Abdominal aortic aneurysm: ten years’
aneurysms has been set out in terms of the level of hospital population study in the city of Glasgow. Eur J Vasc Surg
urgency of definitive care necessary, and hypotension is 1993; 7: 561–6
employed as the discriminating factor. This parameter 3 Sterpetti AV, Cavallari N, Allegrucci P, et al. Seasonal variation in
has been shown to determine outcome with ‘urgent’ the incidence of ruptured abdominal aortic aneurysm. J R Coll
aneurysms having a morality rate of 10–20 per cent, while Surg Edinb 1995; 40: 14–15.
emergency repair has a mortality rate of 35–60 per cent. 4 Walker EM, Hopkinson BR, Makin GS. Unoperated abdominal
aortic aneurysm: presentation and natural history. Ann R Coll
The risk of elective infrarenal AAA repair is approximately
Surg Engl 1983; 65: 311–13.
5 per cent, whereas the mortality rate of operations for
5 Hewin DF, Campbell WB. Ruptured aortic aneurysm: the decision
ruptured aneurysms remains stubbornly at 35–60 per not to operate. Ann R Coll Surg Engl 1998; 80: 221–5.
cent.23 The true mortality rate of a ruptured aneurysm is 6 Refson JS, Wilmink A, Kerle M, et al. Age, renal dysfunction and
in fact 80–90 per cent since many of the patients do not aneurysm outcome. Br J Surg 2001; 88: 611.
reach hospital.25 7 Bayly PJ, Matthews JN, Dobson PM, et al. In-hospital mortality
The increasing enthusiasm for screening programmes from abdominal aortic surgery in Great Britain and Ireland:
may, in the future, reduce the need for emergency Vascular Anaesthesia Society audit. Br J Surg 2001; 88:
aneurysm repair in the same way as advances in the 687–92.
management of peptic ulcers have reduced the need for 8 Brady AR, Fowkes FG, Greenhalgh RM, et al. Risk factors for
repair of perforated peptic ulcers. postoperative death following elective surgical repair of
abdominal aortic aneurysm: results from the UK Small Aneurysm
There is much discussion about the use of stents for
Trial. On behalf of the UK Small Aneurysm Trial participants. Br J
elective infrarenal aneurysm repair as reintervention
Surg 2000; 87: 742–9.
rates are high. If, however, the technique is successful in 9 Adam DJ, Mohan IV, Stuart WP, et al. Community and hospital
occluding the rupture and maintaining flow to organs outcome from ruptured abdominal aortic aneurysm within the
and limbs, then it has an important role in the emer- catchment area of a regional vascular surgical service. J Vasc
gency situation. Once the patient has survived emer- Surg 1999; 30: 922–8.
gency surgery further investigations can be undertaken 10 Gilling-Smith GL, Wolfe JH. The tailed aortic graft: a technique
electively, with considerably reduced attendant risks. for widening the distal orifice of an aortic tube graft. Br J Surg
1986; 73: 208
11 Stroosma OB, Kootstra G, Schurink GW. Management of aortic
aneurysm in the presence of a horseshoe kidney. Br J Surg 2001;
88: 500–9
12 Szilagyi DE, Elliott JP, Berguer R. Coincidental malignancy and
Key references abdominal aortic aneurysm. Problems of management. Arch Surg
1967; 95: 402–12.
13 Mortality results for randomised controlled trial of early elective
Bown MJ, Sutton AJ, Bell PR, Sayers RD. A meta-analysis of
surgery or ultrasonographic surveillance for small abdominal
50 years of ruptured abdominal aortic aneurysm repair
aortic aneurysms. The UK Small Aneurysm Trial Participants.
[review]. Br J Surg 2002; 89: 714–30.
Lancet 1998; 352: 1649–55.
Brady AR, Fowkes FG, Greenhalgh RM, et al. Risk factors for
14 Scott RA, Ashton HA, Lamparelli MJ, et al. A 14-year experience
postoperative death following elective surgical repair of
with 6 cm as a criterion for surgical treatment of abdominal
abdominal aortic aneurysm: results from the UK Small
aortic aneurysm. Br J Surg 1999; 86: 1317–21.
Aneurysm Trial. On behalf of the UK Small Aneurysm Trial
15 Wilmink AB, Quick CR, Hubbard CS, Day NE. Effectiveness
participants. Br J Surg 2000; 87: 742–9.
and cost of screening for abdominal aortic aneurysm:
Hewin DF, Campbell WB. Ruptured aortic aneurysm: the decision
results of a population screening program. J Vasc Surg
not to operate. Ann R Coll Surg Engl 1998; 80: 221–5.
2003; 38: 72–7.
 References 285

16 Lindholt JS, Juul S, Fasting H, Henneberg EW. Hospital costs and 21 Office of National Statistics. Mortality Statistics. Cause 1993.
benefits of screening for abdominal aortic aneurysms. Results London: HMSO, 1995.
from a randomised population screening trial. Eur J Vasc 22 Neary WD, Crow P, Foy C, et al. Comparison of POSSUM scoring
Endovasc Surg 2002; 23: 55–60. and the Hardman Index in selection of patients for repair
17 Emergency AAA. Vascular news. 2001; 1. of ruptured abdominal aortic aneurysm. Br J Surg 2003; 90:
18 Veith FJ, Ohki T. Endovascular approaches to ruptured infrarenal 421–5.
aorto-iliac aneurysms. J Cardiovasc Surg (Torino) 2002; 43: 23 Bown MJ, Sutton AJ, Bell PR, Sayers RD. A meta-analysis of 50
369–78. years of ruptured abdominal aortic aneurysm repair [review]. Br J
19 Hinchliffe RJ, Braithwaite BD, Hopkinson BR. The endovascular Surg 2002; 89: 714–30.
management of ruptured abdominal aortic aneurysms. Eur J Vasc 24 Podlaha J, Gregor Z, Roubal P, et al. Ruptured abdominal aortic
Endovasc Surg 2003; 25: 191–201. aneurysm – outcomes in the last ten years. Bratisl Lek Listy 2000;
20 Peppelenbosch N, Yilmaz N, van Marrewijk C, et al. Emergency 101: 191–3.
treatment of acute symptomatic or ruptured abdominal aortic 25 Buskens FG. Incidence, risk and operability of abdominal aortic
aneurysm. Outcome of a prospective intent-to-treat by EVAR aneurysm in the elderly patient [in Dutch]. Tijdschr Gerontol
protocol. Eur J Vasc Endovasc Surg 2003; 26: 303–10. Geriatr 1990; 21: 169–71.
This page intentionally left blank
 24
Stenting in Acute Aortic Dissection

MICHAEL D DAKE

The problem 287 Devices 291

Pathophysiology 287 Stent-graft placement and effects 294
Classification and rationale 287 Results 295
Management 287 Technical challenges 295
Endovascular stent-graft repair 288 References 296

THE PROBLEM CLASSIFICATION AND RATIONALE

Acute dissection is the most frequent catastrophe affecting The process of aortic dissection is traditionally categorised
the aorta. The mean age of patients at the time of diagnosis by the extent of disease. The most widely adopted classifica-
of acute aortic dissection is 60 years, a decade younger than tion schemes rely on the ability to diagnose ascending aor-
that observed in other thoracic aortic diseases such as tic dissection and distinguish it from isolated involvement
aneurysm, penetrating aortic ulcer or aortic intramural of the aortic arch and/or descending aorta. The rationale for
haematoma. Although hypertension is clearly recognised this critical distinction is based on well-established prog-
as a sine qua non for its occurrence, a number of other clin- nostic implications and, consequently, therapeutic consid-
ical factors are associated with its aetiology. erations associated with dissection of the ascending aorta.
Presently, without immediate surgical intervention, the
prognosis for aortic dissection of the ascending aorta man-
PATHOPHYSIOLOGY aged with medical therapy alone is extremely grave, and the
risk of aortic rupture, pericardial tamponade or cardiac
death from proximal extension into the coronary arteries is
Aortic dissection occurs when blood enters the aortic wall
considerable. Thus, standard therapy for this pattern of dis-
through a disruption of the intimal lining. The tear in the
ease is open surgical repair of the ascending aorta. The most
intima allows flowing blood to cleave a plane within the
commonly used classification scheme is the Stanford cat-
medial layer of the wall. The intimal entry is associated with
egorisation. This straightforward system is predicated on
intramural extension of blood which creates a longitudinal
the determination as to whether there is involvement of the
delamination of the aortic wall. The dissecting blood or
ascending aorta irrespective of the distal extent of the dis-
thrombus may progress from the entry tear in a proximal
section or location of the entry tear. If the ascending aorta is
and/or distal direction. As the process extends, a dissection
involved the dissection is defined as type A; exclusive dis-
flap or septum, consisting of the cleaved lamellar layer of
section distal to the ascending aorta is type B.
intima and partial thickness of media is formed separating
the original ‘true’ lumen lined by intima from the newly
formed ‘false’ intramural channel. The trajectory of flap
progression results in dissection with a distinct morpho-
MANAGEMENT
logical features in each case of dissection. The anatomical
relationships between the flap, true lumen, and false lumen Currently, conventional therapeutic strategies for type B
and importantly, the aortic branch vessels they engage, are disease are medical management with pharmacological
idiosyncratic to the individual concerned. -blockade and antihypertensive agents for uncomplicated
288 Stenting in acute aortic dissection

cases; surgical repair is reserved for dissection complicated involvement, thrombosing the aortic false lumen, and
by ischaemic abdominal aortic branch vessel compromise holds the attractive potential of reducing the frequency of
or aortic rupture. In this regard, a new alternative for thoracic false lumen aneurysm formation. In this regard,
managing complicated acute type B aortic dissection has the false lumen in type B dissection has the propensity to
recently emerged. Endovascular stent-grafts combine become aneurysmal in patients who are treated medically.
self-expanding stent technology, originally designed to The frequency of this complication ranges from 30 to 40 per
treat patients with arterial occlusive disease, with vascular cent at 3 years after the initial diagnosis. It is important to
graft materials used in vascular bypass or replacement sur- note that the management of chronic dissection with false
gery. Initially, these devices were used to repair abdominal lumen aneurysm formation continues to represent a formi-
aortic aneurysms and subsequently, thoracic aortic and dable surgical challenge.
peripheral arterial aneurysms. Recently, the spectrum of Clinical evaluation of stent-grafts, for complicated
applications for this technology has expanded to include or uncomplicated acute type B dissection, selected patterns
peripheral arterial occlusive disease, failing haemodialysis of type A dissection as well as chronic dissection with
access conduits, venous and arterial traumatic injuries, false lumen aneurysm formation, is in progress at a grow-
transjugular intrahepatic portal-systemic stent shunts ing number of institutions around the world. This chap-
(TIPSS), and malignant biliary, oesophageal and tracheo- ter will focus on treatment considerations and early
bronchial lesions. results related to stent-graft management of acute aortic
dissection.
The particular manifestations of dissection which con-
stitute imperatives for intervention in the setting of acute
ENDOVASCULAR STENT-GRAFT REPAIR type B disease require an evaluation of many clinical symp-
toms, physiological effects and the findings of imaging
The concept of endovascular stent-graft repair of aortic dis- studies. After weighing these considerations, however, and
section is predicated on successful placement of the device discarding many of them, it is possible to narrow the
over the intimal entry tear in order to obliterate blood flow criteria down to three distinctive reasons for intervention.
into the false lumen. The intent is to mimic the effect of suc- These include: early aortic expansion within 30 days of the
cessful operative repair of dissection with isolation of the initial onset of symptoms, associated with persistent pain or
false lumen from the circulation and redirection of blood other findings suggestive of impending rupture (Fig. 24.1);
flow into the true lumen. This endovascular surrogate for aortic rupture, free or contained (Figs 24.2 and 24.3); and
open surgery is capable of reversing ischaemic branch vessel peripheral branch vessel involvement with correlative

(a)

Figure 24.1 Early dilatation of false lumen in type B dissection in a 68-year-old man with acute back pain. (a) Computed tomography (CT)
scan shows type B aortic dissection.
(b)

(c)

Figure 24.1 (b) Follow-up CT 5 weeks

after previous scan shows marked
increase in diameter of descending aorta
at the axial level of the pulmonary
arteries associated with development of
partial thrombosis of the false lumen.
(c) Thoracic aortogram shows multiple
communications between true and false
lumens through tears in the flap located
in the proximal and mid-descending
aorta. (d) Aortogram and three-dimensional
CT angiogram after stent-graft placement
show obliteration of the tears and
(d) thrombosis of the thoracic aortic false
lumen over the length of the stent-graft
 (a)

(b) (c)

(d)

(e)

Figure 24.2 Acute type B aortic dissection and rupture in an 84-year-old man with acute back pain and shock. (a) Computed tomography
(CT) scan done elsewhere at 6 pm shows type B dissection with substantial mediastinal haemorrhage. (b) After helicopter transfer, chest
X-ray at 8:14 pm shows complete opacification of left hemithorax with marked displacement of the trachea. (c) Initial aortogram demonstrates
entry tear and false lumen filling but no evidence of the exact site of rupture. (d) Endovascular placement of a 20 cm stent-graft centred on
the entry tear and aortogram at 8:43 pm shows no false lumen, opacification or evidence or extravascular contrast medium. (e) Follow-up CT
scan 6 weeks after stent-graft placement shows only true lumen filling, with marked reduction in the volumes of the false lumen and
mediastinal haematoma
 Devices 291

(b)

(a)

(c) (d)

Figure 24.3 Acute type A dissection with rupture and entry tear distal to left subclavian artery in a 74-year-old man with acute back pain
and shock. (a) Computed tomography (CT) scan shows aortic rupture associated with type A dissection. (b) Aortography outlines entry tear
in the proximal descending aorta. (c) Aortography shows stent-graft placement covering entry tear with no flow in the false lumen and no
extravascular flow. (d) Follow-up CT scan 5 weeks after stenting shows marked resolution of the mediastinal haematoma and reduction in
size of the thrombosed aortic false lumen

evidence of end organ or tissue ischaemia (Fig. 24.4). In dissection (acute v chronic), extent of aortic disease (type
any of these clinical settings, the mortality and morbidity A v type B), indications for intervention (uncomplicated v
associated with open surgical intervention are high and complicated), patient symptoms and involvement of
dramatically greater than similar results for elective thoracic branch vessels. Nonetheless, valuable lessons from this early
aneurysm repair. Consequently, the universally acknowl- experience have served to fuel progress in our understand-
edged high risk diagnostic group of complicated acute type B ing of the disease process as well as its management by less
aortic dissection, represents both a challenge and an opportu- invasive means.
nity for stent-grafts to prove they are better, less invasive
alternatives to the current default strategy of open repair.
DEVICES
Acute type B dissection – interventional
Currently, there are three commercially manufactured
imperatives
thoracic stent-grafts that are widely available and marketed
outside the USA. The TAG (WL Gore & Associates, Flagstaff,
• Early aortic expansion or evidence of impending
AZ, USA), the Talent device (AVE/Medtronic Inc., Santa
rupture
Rosa, CA, USA), and the TX2 (Cook, Inc., Bloomington, IN,
• Aortic rupture (free or contained)
USA) are the most commonly implanted thoracic prostheses
• Symptomatic aortic branch vessel involvement
with a worldwide aggregate experience approaching 5000
patients. At the present time, however, not a single thoracic
endograft has been approved by the Food and Drug
Initial results from published reports of experience with Administration (FDA) for general clinical use in the USA.
stent-graft technology in treating patients with aortic dissec- Nevertheless, all three devices mentioned above are currently
tion are encouraging. Unfortunately, however, it is often dif- undergoing phase II patient trials.
ficult to determine from these publications the precise Published reports are also available of experiences with
characteristics of the treatment group, in terms of the age of less widely distributed devices and of expectations with
292 Stenting in acute aortic dissection

(a)

(c) (b)

(d)

(e)
(f)

Figure 24.4 Acute type B dissection with mesenteric ischaemia associated with dynamic branch vessel involvement and aortic true lumen
collapse in a 75-year-old woman with abdominal pain. (a) Chest computed tomography (CT) scan shows type B dissection. (b) Abdominal CT
scan reveals near obliteration of anteriorly positioned aortic true lumen. (c) Aortography shows marked compression of aortic true lumen
distal to coeliac trunk. (d) Arch aortogram outlines type B dissection. (e) Aortogram after stent-graft deployment shows incremental true
lumen size and no evidence of false lumen flow. (f) CT evaluation at discharge from hospital demonstrates thrombosed thoracic aortic false
lumen with retained patency of abdominal false lumen and marked increase in diameter of aortic true lumen

several newly developed endografts specifically targeting and feature considerable improvements over those of the
the challenges presented by thoracic aortic diseases. In previous generation. The individual benefits of these new
all cases, however, the condition most commonly treated products are detailed below, but in general, manufacturers
is descending thoracic aortic aneurysm rather than aortic are focused on improved durability, deliverability and
dissection. reliability.
Detailed discussion of the various first generation home- The Gore TAG is composed of a self-expanding nitinol
made devices is only of modest historical interest. Suffice stent lined with polytetrafluoroethylene (PTFE) graft mater-
to say, most of these prostheses prototypes were self- ial. Previously, the first generation of this device, the TAG
expanding and based on a combination of a polyester graft Excluder, was lined with ultra-thin wall PTFE graft with a
with a modified type of Gianturco Z stent. Most delivery 30  internodal distance similar to the pore size of conven-
systems were large (24–27 Fr), relatively rigid, and depend- tional PTFE vascular grafts. The new TAG uses a different
ing upon the anatomy, as well as the length and diameter of proprietary multilayer composition which, in preclinical
the device, difficult to target and deploy due to marked fric- in vitro comparison tests, was found to be more durable
tional resistance encountered during withdrawal of the and scratch resistant than the graft material previously used.
outer device containing sheath. Detailed descriptions of The ends of the device have a scalloped contour to
individual systems, components, fabrication processes and enhance graft contact with the aortic wall over a wide range
deployment techniques are provided elsewhere. of aortic tortuosities and angulations. The scalloped pro-
The current designs of the three most frequently used jections are covered with PTFE and their length is directly
stent-grafts represent the second generation of each device proportional to the diameter of the graft. The device is very
 Devices 293

flexible radially and longitudinally. The new TAG does not self-expanding stent-graft occurs. After deployment, the
have the S-shaped stabilisation wires that were anchored delivery catheter is removed. The procedure is complete if
180 degrees apart and spanning the length of its previous the desired position of the graft is achieved without arteri-
counterparts. Worryingly, fracture(s) of these longitudin- ographic or transoesophageal ultrasound imaging evidence
ally oriented wires has been observed after implantation in of inadequate obliteration of the entry tear with persisting
a number of patients. This complication occurred in 10–30 flow through it into the false lumen. If, however, the device
per cent of the implants and appeared to be related to the is poorly positioned or not fully expanded and there is a
length of time since implantation, tortuosity of the device perigraft leak, supplemental manoeuvres including place-
once implanted and, perhaps, the presence of overlapping ment of additional stent-graft(s) or gentle balloon expan-
endografts. Fortunately, no deaths related to this compli- sion over the segment proximal to the entry opening,
cation were reported to or observed by the independent where the leak is suspected, may prove beneficial. As a gen-
core laboratory monitoring the follow-up of US clinical eral rule, if no false lumen filling is noted via the entry tear
trials. There were, however, five untoward events caused during post-placement aortography, balloon expansion of
by spine fractures. Two patients in the USA developed a the device is discouraged in order to avoid the creation of sec-
type 3 endoleak through the graft material, presumably via ondary tears in the dissection septum. Resumption of the
a puncture hole created by an end of the fractured spine: sales of the new TAG device by Gore began in March 2004
one was managed successfully by adding another TAG and the results of treatment of patients with a variety of
device to line the damaged prosthesis while the second led thoracic aortic diseases are anxiously awaited.
to a rupture of the aneurysm necessitating surgical conver- The second device currently available for treatment of
sion and explantation of the device. The other three serious patients with descending thoracic aortic aneurysm is the
events occurred in Europe: in two cases, the type 3 endoleak Talent endoprosthesis. The design of this product has
was managed endovascularly by placement of an additional evolved since its introduction into clinical practice. The
device, while in the third, elective surgical conversion with current version has a lower profile and more flexible deliv-
removal of the device was performed. ery catheter than the original system. The prosthesis is
In view of the potential for clinical sequelae, Gore volun- composed of sinusoidal nitinol stent elements sandwiched
tarily suspended enrolment in its phase II US trial, withdrew between thin layers of polyester graft material. The indi-
the product from the global market and suspended sales in vidual stent forms are secured in place with oversewn
2001. In the current newly redesigned TAG, the graft mater- sutures to prevent migration, but they are not connected to
ial is engineered to perform the same function as that pro- one another and there are segments of unsupported graft
vided by the former longitudinal wires, that is, to stabilise the interposed between stents. This design allows independent
implant and prevent any longitudinal compression during stent motion and confers a degree of longitudinal flexibil-
deployment. The graft is axially compressed onto the end of ity. As with the earlier generation of the TAG Excluder
the delivery catheter and constrained by a PTFE corset laced device, the Talent utilises two longitudinal wires to provide
with PTFE suture. The suture runs down the length of the stabilisation and prevent longitudinal compression.
catheter and is attached to a deployment knob at the oppos- A unique aspect of the device is its proximal margin
ite end. Grafts are available in a range of diameters between with broad-based nitinol wire scallops. The wide, uncovered
26 and 40 mm, and a selection of lengths between 7.5 and interstices may be placed across the origin of the left sub-
40 cm. The sizes of the delivery system and compatible clavian artery in cases where there is a short proximal neck
introducer sheaths vary according to the diameter of the of between 10 and 20 mm. In this setting, placement of the
device and scale over a range of 20–24 Fr. uncovered stent across the left subclavian artery helps to
After preliminary arteriography studies have defined orient the graft optimally, stabilise its position and during
the preferred proximal and distal stent-graft ‘landing deployment, secure precise targeting of the graft material
zones’, and these targets have been confirmed by trans- at the distal subclavian margin.
oesophageal ultrasound, a suitably sized 30 cm long intro- The stent-grafts are available in a wide range of diam-
ducer sheath is advanced over a guidewire to the infrarenal eters and lengths. In addition, custom fabrication of a pros-
aorta. Alternatively, in certain situations, the catheter deliv- thesis based upon an individual patient’s anatomy is
ery system may be introduced over the guidewire without possible within 3 weeks. The delivery profile for Talent is
the use of a sheath. In either case, the device catheter is between 22 and 27 Fr depending on the diameter and length
tracked over the wire until it reaches the selected target in a of the device. The delivery catheter has a flexible, conical
manner that bridges the entry tear. tip. Set back from the tip is an integrated balloon used for
After final positioning of the device has been achieved, smoothing the graft material and promoting adequate stent
deployment is effected by pulling the knob adjacent to the expansion following deployment of the self-expanding
hub of the catheter. As the knob is smoothly retracted, the device. The prosthesis is collapsed over the distal segment of
attached suture is withdrawn and opening of the corset the delivery catheter and maintained in this packed config-
occurs initially in the middle and then proceeds towards uration by an overlying transparent sheath. Proximal to the
both ends. An instantaneous release of the underlying, loaded stent-graft is a blunt metal stopper functioning as a
294 Stenting in acute aortic dissection

brace to maintain the device position as the constraining The modular design of the TX2 endograft requires that
sheath is withdrawn. two components, proximal and distal pieces, be placed, in
Once the stent-graft is properly positioned, usually each case, irrespective of the length of diseased aorta to be
1–3 cm proximal to the optimal landing zone to mitigate treated. The intent behind this concept is to confer increased
against inadvertent downstream drift during deployment, flexibility upon the endograft and to allow for optimal
the overlying sheath is slowly withdrawn. As the initial accommodation within potential changes that may be
uncovered stent elements cantilever open, gentle retraction likely to occur in the long term. At the time of writing this
of the device is applied until the exact desired position of chapter, the use of the TX2 in the management of patients
the proximal graft margin is contained. After the device is with aortic dissection is limited.
fully deployed, the balloon may be withdrawn and, if nece-
ssary, expanded within the proximal segments to fully
expand the prosthesis. Subsequently, the final result is docu-
STENT-GRAFT PLACEMENT AND EFFECTS
mented angiographically and the arteriotomy repaired
surgically.
In an analysis of the relative merits of the devices, it is There is growing worldwide experience with endovascular
important to note that both devices described so far have stent-graft placement in treating acute type B dissections as
established records of technical and clinical successes. In well as chronic dissections with coexisting descending aor-
certain cases, however, the particular disease process and tic false lumen aneurysm.1–5
anatomy may recommend one device over the other. The In both pathologies, successful management is predi-
marked flexibility of the Excluder device and its delivery cated on the obliteration of the primary entry tear of the dis-
system, as well as its smaller introduction profile, make it section by placement of the prosthesis within the true lumen
better suited for patients with severely angled aortic across the entry tear. Stent-graft coverage of the entry site
anatomy or those with small, calcified, or tortuous closes the primary communication to the false lumen and its
iliofemoral conduits. In aortic dissection cases, with short flow is markedly reduced or choked off completely. In acute
(15 mm) proximal necks where the primary entry tear is type B dissection, the true lumen immediately increases in
very close to the left subclavian artery, the Talent device diameter without a corresponding incremental change in
may be preferred because of its leading segment of uncov- the overall aortic diameter. Downstream, any dynamic com-
ered stent. In terms of ease of use, the Excluder has some promise of the branch arteries of the abdominal aorta by the
advantages. The maximum graft length available is 20 cm dissection process is expeditiously reversed within seconds
compared with 13 cm for an individual, non-custom of stent-graft placement.
Talent device. This, in combination with its simple and In cases of both acute and chronic dissection, stagnant
straight forward deployment, make it more efficient in blood in the false lumen of the thoracic aorta will clot. In the
extending the treatment zone. In aortic dissection, the rel- majority of patients, progressive thrombosis of the false
ative radial force exerted by the prosthesis may be a con- lumen proceeds from the point of proximal involvement of
sideration. In the acute setting, the lower hoop strength of the thoracic aorta distally, irrespective of the location
the Excluder may allow adequate coverage of the entry site of the primary tear. The tempo of this process varies and
without causing an iatrogenic secondary tear in the thin, is influenced by certain factors: the size of the false
fragile dissection flap. In this regard, there are reports that lumen, abdominal branch vessel distribution off the aortic
the relatively rigid leading bare metal proximal segment of lumens, the amount of residual flow through the false lumen
the Talent device can injure the aortic wall adjacent to the via uncovered additional tears in the dissection flap, retro-
entry tear and create a retrograde type A dissection. On the grade false lumen branch vessel flow via collaterals, retro-
other hand, the greater radial force of the Talent may be grade perfusion from the abdominal aortic false lumen, etc.
beneficial in cases of chronic dissection to displace a thick
and resistant dissection septum and thus enhance the true
lumen diameter. Acute type B dissection – branch vessel
The third device, the Cook TX2 thoracic endograft is involvement
made up of two components and is currently undergoing
FDA phase II clinical trial in the USA. The predecessor of • Static – direct extension of flap into aortic branch
this device, the TX1, was a single piece device sold exclu- with or without distal re-entry tear
sively outside the USA. The current modular device and the • Dynamic – true lumen collapse or obliteration with
uni-body former version both incorporate a distal uncov- prolapse of aortic septum over ostia of abdominal
ered stent-body, in conjunction with graft material extend- branches with true lumen origin
ing to the proximal margin of the endograft. In both cases, • Both – co-existing static and dynamic processes
fixation barbs are employed around the proximal aspect affecting an individual branch
(pointing distally) and distal extent (pointing proximally).
 Technical challenges 295

It is expected that isolated sections of the abdominal the approximate size of the device most frequently used in
aortic false lumen will be kept patent via natural fenestra- current practice. Obviously, if there is retrograde proximal
tions in the dissection flap at levels corresponding to the extension of the dissection from the entry site, other plan-
abdominal branches coming off the false lumen. This phe- ning steps must be taken. These include calculation of the
nomenon permits sufficient perfusion of the false lumen mean true lumen diameter from measurements of the
branches via true lumen trans-septal flow after stent-graft maximum and minimum true lumen dimensions, selec-
obliteration of the primary communication to the thoracic tions of an arbitrary diameter corresponding to a value
aortic false lumen. In cases of acute dissection, if throm- larger than the true lumen but smaller than the overall aor-
bosis occurs in the false lumen after stent-graft placement, tic diameter, etc.
progressive false lumen resolution may occur with a cor- In terms of the device length, most investigators
responding gradual enlargement of the true lumen. In this implant devices which are clearly longer than the entry tear
regard, follow-up imaging at 1 year has shown apparent and usually in the range of 10–15 cm. After implantation
‘healing’ of the dissection in a number of cases, there being this added length confers a more normal anatomical
no computed tomography (CT) evidence of a residual thor- appearance to the aortic morphology, especially in the
acic aortic false lumen or dissection flap.1,2 arch, than that observed following placement of a short
device focally over the entry tear. In addition, the longer
device promotes accelerated thrombosis within the prox-
RESULTS imal thoracic aortic false lumen. Further extension of the
overall length of the device into the distal third of the
descending thoracic aorta, however, should be avoided in
Early results from clinical series of stent-graft management
this setting because of an associated increased risk of spinal
in limited numbers of patients with acute type B and acute
cord ischaemia.
type A aortic dissection, where the primary tear is identi-
In those cases of aortic dissection in which the primary
fied distal to the left subclavian artery, are encouraging.1,5–9
entry site is located classically at the isthmus, the tear may
Obliteration of flow through the entry tear into the false
be within 10 mm of the left subclavian artery. In this situ-
lumen was achieved in greater than 90 per cent of cases,
ation, a device with a proximal segment consisting of a
with associated complete thrombosis of the proximal thor-
bare stent can be placed across the left subclavian artery
acic aortic false lumen segment apparent in 80–100 per cent
effectively maximising the length of graft contact with the
and distal thoracic segment thrombosis noted less fre-
aortic wall cephalad to the tear. When retrograde proximal
quently. Progressive false lumen shrinkage at 1, 6, and 12
extension of the dissection occurs from the tear, however,
months follow-up imaging was observed in most cases.
it may be necessary to place the graft over the origin of the
Complications, including paraplegia, rupture and iatro-
subclavian artery with its leading margin between the left
genic extension of the dissection into the ascending aorta,
carotid and subclavian arteries. In addition to carefully
have been reported anecdotally in the early experience with
monitoring the patient post procedure for ischaemic
stent-graft placement in acute dissection.6–9
symptoms in the left arm caused by a ‘covered’ left sub-
clavian, it is important to image the thoracic aorta carefully
to exclude persistent perfusion of the false lumen via retro-
TECHNICAL CHALLENGES
grade subclavian flow around the device into the arch.
As with acute type B dissection, experience is mounting
In terms of the procedure, some technical challenges in treating patients with chronic aortic dissection and false
related to a specific set of morphological manifestations of lumen aneurysm using endografts as an alternative to open
aortic dissection are often discussed when stent-graft ther- surgical repair. In this regard, multiple studies record rates
apy is considered. A common issue for consideration is the of aneurysm thrombosis and subsequent false lumen
optimal method of selecting the diameter and the length of shrinkage mirroring those in reported series of acute
the prosthesis. As the diameter of the true thoracic aortic dissection.2–11 One controlled investigation comparing
lumen represents a fraction of that of the overall vessel stent-graft therapy with open surgery in matched groups of
trunk, and is rarely cylindrical in shape, choosing the patients with chronic type B dissection reported improved
‘right’ device dimensions is a unique logistical dilemma. survival and decreased neurological complications with the
Most practitioners base their selection on more than one less invasive procedure.2
measurement. Perhaps, the most compelling is the diam- The therapeutic options available in managing type A or
eter of the non-dissected aorta immediately proximal to type B acute aortic dissections, depending on the entry tear
the entry tear. This is a good estimate of the original size of location of the former and the status of involvement of the
the proximal involved segment prior to dissection. This latter, taking into account the level of acceptability as well
measurement, plus an oversize factor of 20 per cent to as the risk attached to each of those choices, are illustrated
ensure secure anchoring and a tight circumferential seal, is in the algorithm provided (Fig. 24.5).
296 Stenting in acute aortic dissection

Acute aortic dissection

Type A Type B

Entry tear location Status of involvement

Ascending Aortic Descending Uncomplicated Complicated

aorta arch aorta

Medical Operative
therapy risk
Operative
risk

Acceptable High Acceptable High Acceptable High Acceptable High

Open Open or Stent-graft or Medical Open Combination Open Stent-graft Stent-graft Open Stent-graft Stent-graft
placement or or
surgery surgery therapy surgery bypass/ surgery placement placement surgery placement placement
in selected stent-graft
patients with procedure
suitable entry
tear anatomy

Figure 24.5 Algorithm of therapeutic options available in managing acute aortic dissection: the degree of acceptability and the level of
risk attached to choices made in treating type A dissections with different entry tear locations, and in dealing with complicated or
uncomplicated type B dissections, are shown

REFERENCES
Conclusions

The recent development of endovascular stent-graft 1 Dake MD, Kato N, Mitchell RS, et al. Endovascular stent-graft
technology and its application as an alternative manage- placement for the treatment of acute aortic dissection. N Engl
ment strategy to medical therapy or open surgical treat- J Med 1999; 340: 1546–52.
2 Nienaber CA, Fattori R, Lund G, et al. Non-surgical reconstruction
ment of patients with aortic dissection is an exciting and
of thoracic aortic dissection by stent-graft placement. N Engl
potentially valuable advance. The next major challenge
J Med 1999; 340: 1539–45.
that interventionists are facing is the important task of 3 Sakai T, Dake MD, Semba CP, et al. Descending thoracic aortic
objectively elucidating the ‘real’ benefits, risks and com- aneurysm: thoracic CT findings after endovascular stent-graft
plication of thoracic stent-grafts through rigorous, placement. Radiology 1999; 212: 169–74.
prospective controlled investigations of each possible 4 Kato N, Hirano T, Shimono T, et al. Treatment of chronic type B
anatomical and clinical manifestation of dissection. aortic dissection with endovascular stent-graft placement.
Only after this type of scientific scrutiny is performed Cardiovasc Intervent Radiol 2000; 23: 60–2.
can clinicians confidently counsel patients with accurate 5 Czermak BV, Waldenberger P, Fraedrich G, et al. Treatment of
information regarding their therapeutic options. Stanford type B aortic dissection with stent-grafts: preliminary
results. Radiology 2000; 217: 544–50.
6 Ehrlich MP. Thoracic aorta endografting: the Austrian experience.
The First International Summit on Thoracic Aorta Endografting,
Key references Tokyo, Japan, 2001.
7 Fattori R. Endovascular treatment of the thoracic aorta. The First
Czermak BV, Waldenberger P, Fraedrich G, et al. Treatment of International Summit on Thoracic Aorta Endografting, Tokyo,
Stanford type B aortic dissection with stent-grafts: Japan, 2001.
preliminary results. Radiology 2000; 217: 544–50. 8 Ishimaru S. Thoracic aorta grafting; the reliable treatment
Dake MD, Kato N, Mitchell RS, et al. Endovascular stent-graft option. The First International Summit on Thoracic Aorta
placement for the treatment of acute aortic dissection. N Engl Endografting, Tokyo, Japan, 2001.
J Med 1999; 340: 1546–52. 9 Lauterjung L. Endovascular stent-grafting for the thoracic aorta.
Kato N, Hirano T, Shimono T, et al. Treatment of chronic type B The First International Summit on Thoracic Aorta Endografting,
aortic dissection with endovascular stent-graft placement. Tokyo, Japan 2001.
Cardiovasc Intervent Radiol 2000; 23: 60–2. 10 Dake MD. The advent of thoracic aortic endografting. The First
Nienaber CA, Fattori R, Lund G, et al. Non-surgical reconstruction International Summit on Thoracic Aorta Endografting, Tokyo,
of thoracic aortic dissection by stent-graft placement. N Engl Japan, 2001.
J Med 1999; 340: 1539–45. 11 Iwase T. Endovascular repair for chronic aortic dissection by
Sakai T, Dake MD, Semba CP, et al. Descending thoracic aortic Inoue stent-graft system. The First International Summit on
aneurysm: thoracic CT findings after endovascular stent-graft Thoracic Aorta Endografting, Tokyo, Japan, 2001.
placement. Radiology 1999; 212: 169–74.
 25
Prosthetic Aortic Graft Infection

CHARLES W BOUCH, JAMES M SEEGER

The problem 297 Surgical treatment 302

Incidence, aetiology and bacteriology 297 Management of prosthetic graft erosions/fistulae 305
Diagnosis 298 Late complications 307
Preoperative management 302 References 308

THE PROBLEM INCIDENCE, AETIOLOGY AND BACTERIOLOGY

Management of patients with infected prosthetic aortic Fortunately, infection of prosthetic aortic grafts is very
grafts is one of the most difficult challenges faced by the uncommon. Although prospective series assessing the pre-
vascular surgeon. Patients often present with non-specific cise risk of aortic graft infections are not available, retro-
symptoms and confirmation of the diagnosis of an aortic spective review of patients followed for at least 5 years after
graft infection can be difficult. Delay in treatment of an graft implantation suggests that the incidence of prosthetic
infected aortic graft, however, can lead to life-threatening vascular graft infection is between 0.5 and 2.5 per cent.1
sepsis and/or haemorrhage. Patients with infected aortic The risk of infection varies with the position of the graft,
grafts are also elderly and have multiple medical problems with the incidence of infection for aortic grafts confined to
so that sepsis, haemorrhage and major surgical reconstruct- the abdomen being 0.5 to 1 per cent, and for aortofemoral
ive procedures are associated with significant morbidity grafts being 1.5 to 2.5 per cent. The incidence of infection
and mortality. Furthermore, the infected aortic graft is in in thoracic and thoracoabdominal grafts is less well known,
the most favourable anatomical position for revascularisa- but it appears to be similar to that of prosthetic aortic
tion, making maintenance of adequate lower extremity grafts limited to the abdomen.
perfusion after graft removal difficult. Because of all of Although it has not been definitively established, it is likely
these factors, elimination of the infection, preservation of that infection occurs in most cases at the time of graft
limb perfusion and long term survival are achieved in at implantation, at the time of a subsequent procedure which
most 70–80 per cent of patients presenting with prosthetic involves the graft, e.g. revision or repair of a graft, arteriogra-
aortic graft infection. phy through an aortofemoral graft or an infrainguinal bypass
This chapter will discuss diagnosis and treatment of originating from it, or from involvement of the graft by an
patients with prosthetic aortic graft infection, emphasising adjacent infection. Late infection from bacteraemia, although
the basic principles necessary for successful management uncommon, is also a possible cause. The risk of development
of this complex problem. Essential information about the of aortic graft infection increases when there is a break in
incidence, aetiology and bacteriology of such vascular graft sterile technique, when leg infection is present at the time of
infections will also be reviewed briefly. The pros and cons graft placement or when a postoperative wound infection
of the various treatments of aortic graft infection and the develops, even if it appears to be only superficial. In addition,
factors that lead to selection of a particular type of treat- the thrombus within an aortic aneurysm sac contains bacteria
ment for an individual patient will also be discussed as one in approximately 13 per cent of cases2 and the atherosclerotic
form of treatment is not applicable to all patients with this plaque in the arterial wall at the graft–artery anastomosis has
difficult problem. been shown to harbour bacteria in up to 40 per cent of cases.2
298 Prosthetic aortic graft infection

Factors influencing prosthetic aortic Natural history of prosthetic aortic

graft infection graft infection

• Break in sterile technique at time of graft • Low grade sepsis

implantation • ‘Dis-incorporation’ of graft
• Leg infection present at time of graft placement • Systemic sepsis
• Intra-aneurysmal thrombus containing bacteria • Erosion into adjacent structures
(approximately 13 per cent) • Graft occlusion
• Plaque at the graft–artery anastomosis harbouring • Anastomotic disruption/life-threatening
bacteria (up to 40 per cent) haemorrhage
• Postoperative wound infection
• Revision or repair of graft
• Arteriography through aortofemoral graft or
linked infrainguinal bypass
• Late infection from bacteraemia

Organisms responsible for prosthetic graft infections

include Staphylococcus aureus, Staphylococcus epidermidis
and Escherichia coli in approximately 60 per cent of
cases, with Pseudomonas, Klebsiella, Proteus, Enterobacter,
Bacteroides, Enterococcus or other non-haemolytic strepto-
cocci and occasionally yeast accounting for most of the
remainder.2,3 Two or more organisms may be isolated
from 10–37 per cent of infected aortic grafts, particularly in
cases of graft-enteric erosion4–6 and an alarming rise in the
frequency of graft infection by methicillin resistant
S. aureus (MRSA) has been observed over the past decade. Figure 25.1 Computed tomography (CT) scan of a patient
Use of perioperative antibiotics reduces the risk of pros- showing obvious infection surrounding an aortic graft used to
thetic vascular graft infection: prophylaxis against S. aureus repair an abdominal aortic aneurysm 6 months previously.
in most cases and prophylaxis against MRSA in recently Extensive perigraft inflammation containing air can be seen in
hospitalised patients. In addition, small anecdotal reports the midline adjacent to and above the aortic graft
suggest that omental coverage of the graft may reduce the risk
of late infection, although the superiority of this approach
over peritoneal coverage alone has not been documented.
DIAGNOSIS
Identification of the organisms responsible for the
infection is important in selecting appropriate antibiotic
therapy and, at times, in deciding on appropriate treat- When an aortofemoral bypass graft limb is seen in an
ment of the infected vascular graft, e.g. in situ graft replace- infected groin wound, visualised on computed tomography
ment versus graft removal. No organisms, however, can be (CT) or magnetic resonance imaging (MRI) or clearly out-
cultured from up to a quarter of patients with an obviously lined by a sinogram, the diagnosis of vascular graft infection
infected prosthetic vascular graft, unless special methods is obvious. In Fig. 25.1 an aortic prosthesis is seen within an
for recovering S. epidermidis from the graft material are abscess. In all other circumstances, some degree of diagnos-
used. In such instances, so called ‘dis-incorporation’ of the tic uncertainty exists until the graft is explored surgically
prosthetic vascular graft can be used to describe infection and/or cultured. This is because most signs and symptoms
of the vascular graft as over 70 per cent of these grafts will of prosthetic aortic graft infection are non-specific: malaise,
have positive bacterial cultures using advanced culture low grade fever, unexplained weight loss and anorexia.
techniques.7 Furthermore, the natural history of a dis- On investigation, the findings associated with prosthetic vas-
incorporated, culture negative prosthetic vascular graft is cular graft infection tend to be subtle. As previously noted,
the same as for one that is culture positive. This can range a history of leg sepsis prior to graft implantation, postoper-
from initially low grade sepsis if the site can drain and if an ative groin wound infection or systemic sepsis, a recent
anastomosis is not involved in the long term to systemic invasive procedure involving the graft or an episode of sep-
sepsis, erosion into adjacent structures, graft occlusion ticaemia, increases the likelihood of vascular graft infection.
and/or anastomotic disruption with life-threatening These historical findings, however, are far from definitive.
haemorrhage. Similarly, the systemic signs of sepsis, namely, fever, elevated
 Diagnosis 299

white blood cell count, elevated sedimentation rate, with-

out specific local signs of graft infection such as an abscess
adjacent to the prosthetic graft, a pseudoaneurysm, inflam-
matory groin mass, etc., suggest vascular graft infection
only when more common causes of sepsis have been
excluded. This is particularly true in the early postoperative
period after graft implantation when aortic graft infection is
a rare cause of systemic signs of sepsis.
In contrast, localised sepsis adjacent to a prosthetic aor-
tic graft such as a late groin abscess or purulent drainage
from a previously healed groin wound must be assumed to
be related to infection of the aortic graft unless it can be
definitively proved otherwise. This assumption is strength-
ened significantly when what initially appears to be a
superficial cellulitis in the groin fails to resolve with antibi-
otic therapy or recurs immediately after it has been discon-
tinued. Early postoperative groin erythema following
aortofemoral bypass, however, may simply represent the
presence of a groin haematoma or seroma and early (a)
re-exploration provides diagnostic certainty, evacuation
and treatment of the haematoma/seroma and groin closure.
Similarly, the presence of multiple anastomotic pseudoa-
neurysms strongly suggests the presence of aortic graft
infection while a single anastomotic pseudoaneurysm is
usually not associated with infection but should prompt
speedy evaluation of the other vascular graft anastomoses.
Aortofemoral limb thrombosis is also usually due to pro-
gressive infrainguinal arterial occlusive disease but aortic
graft thrombosis may be associated with infection in up to
25 per cent of cases.8 Patients with infected prosthetic aor-
tic grafts can also present with symptoms of obstruction of
surrounding structures such as the ureter (Fig. 25.2) and
such obstruction occurring late after aortic graft implant-
ation should be considered to be caused by aortic graft
infection until proved otherwise. Haemoptysis and
haematemesis are signs also warranting graft evaluation in
patients with thoracic or thoracoabdominal grafts and may
be due to aortobronchial or aorto-oesophageal erosions. In
(b)
rare instances, patients may present with septic emboli or
hypertrophic osteoarthropathy, each of which strongly
Figure 25.2 Computed tomography (CT) scan of a patient
suggests the presence of vascular graft infection.
showing (a) bilateral ureteric obstruction and (b) an infected
Gastrointestinal haemorrhage in a patient with a pros- aortobifemoral prosthetic graft, surrounding both limbs of which
thetic aortic graft is of special concern; although patients fluid can be seen. These findings were confirmed at surgical
who require aortic grafts are usually elderly and, therefore, exploration, when the graft was removed followed by in situ graft
at risk of more common causes of gastrointestinal bleeding, replacement constructed from superficial femoral vein
including diverticulosis, peptic ulcer disease, arteriovenous
malformation, malignancy, portal hypertension and
varices, etc. Gastrointestinal haemorrhage due to graft-
enteric erosion, particularly involving a suture line, is such increases the likelihood of graft-enteric erosion being the
a catastrophic event that it is a possibility which must be cause. In the absence of other signs of graft infection, how-
excluded when gastrointestinal bleeding occurs in any ever, when a patient with gastrointestinal bleeding, mani-
patient with a prosthetic aortic graft. Bleeding can be slow fested as haematemesis, melaena or haematochezia, has a
and intermittent when the graft-enteric erosion does not history of implantation of a thoracic or intra-abdominal
involve an anastomosis, or it may be massive when anasto- vascular graft, all efforts must be made to establish the
motic disruption from infection has occurred. Bleeding, in cause of the bleeding and/or prove that an infected vascu-
conjunction with other signs of sepsis or graft infection, lar graft and a graft-enteric erosion do not exist.
300 Prosthetic aortic graft infection

Clinical diagnosis of prosthetic aortic

graft infection

• Non-specific symptoms and signs of sepsis

• Inflammatory groin mass, seroma, cellulitis,
abscess, discharge
• Graft visible within groin abscess
• Haematemesis, melaena, haemoptysis/shock
• Pseudoaneurysm at femoral anastomosis
• Septic embolism/hypertrophic osteoarthropathy
• Aortofemoral limb thrombosis

Diagnostic studies can usually confirm or exclude aortic

graft infection and/or graft-enteric fistula. At times, how-
ever, the diagnosis can only be made by surgical exploration
of either a groin abscess to determine the presence and
extent of a vascular graft infection or of an aortic graft to
confirm or exclude graft-enteric erosion. Initially, the aortic
graft should be imaged using either CT or MRI. Findings Figure 25.3 Computed tomography (CT) scan of a patient with
upper gastrointestinal haemorrhage showing a previously inserted
suggestive of vascular graft infection include perigraft fluid,
aortic graft. The inflammatory changes and air adjacent to the
ectopic gas, swelling of soft tissue adjacent to the graft and
area of the proximal anastomosis of the aortic graft were strongly
pseudoaneurysm formation (Fig. 25.3). Computed tomog- suggestive of an aortoenteric fistula, which was confirmed at
raphy imaging may also be used to guide aspiration of peri- surgical exploration
graft fluid for culture to help establish the diagnosis of
vascular graft infection, although this is seldom necessary in
our experience. The sensitivity and specificity of CT scan- of gallium-67 scanning (Fig. 25.4) to be 94 per cent which is
ning in the diagnosis of vascular graft infection has been superior to that of CT scanning.14 Nevertheless, both tech-
reported to be approximately 90 per cent in experienced niques are time consuming, can produce false positive stud-
hands,9 and that has certainly been our experience at the ies due to cross-labelling of platelets or inflammation from
University of Florida. Findings of vascular graft infection other causes and can also be made falsely negative by antibi-
on CT scanning, however, can be very subtle, especially otic therapy. This is particularly true early after graft implant-
when minimal fluid is present around the graft such as is ation and in patients with minimal evidence of infection by
seen with infection by S. epidermidis and therefore the graft CT or MRI scanning. Technetium-99m exametazime has
must be clearly visualised and normal throughout its course been used more recently to label leucocytes with repeatedly
if aortic graft infection to be excluded. Computed tomogra- improved results and a small study with surgical/
phy scanning is also of limited value in the diagnosis of vas- bacteriological confirmation of findings demonstrated
cular graft infection during the early postoperative period 100 per cent sensitivity, 94 per cent specificity, 90 per cent
because the normal perigraft changes seen on CT scanning positive predictive value and 100 per cent negative predic-
during the first three months after graft implantation are tive value.15 Regardless, at least in our experience, these
similar to the findings associated with vascular graft infec- radionuclide studies add little to CT or MRI scanning in the
tion.10 Magnetic resonance imaging has been reported to be evaluation of patients suspected of having vascular graft
potentially superior to CT in detecting small amounts of infection and the exclusion of infection, in patients in whom
fluid and limited inflammatory changes surrounding an CT and MRI studies are inconclusive, remains difficult.
infected vascular graft, as would be seen in a patient with
S. epidermidis infection. Experience with MRI scanning in
patients with vascular graft infections, however, is limited at Diagnosis of aortic prosthetic infection on
present. In addition, MRI imaging cannot differentiate investigation
between sterile and infected fluid or between the normal
early inflammatory changes associated with graft implanta-
tion and those due to infection.11
• Fluid collection in groin for culture

Radionuclide imaging studies have been extensively used

• Positive sinogram of groin

in the diagnosis of infected prosthetic aortic vascular grafts.

• CT/MRI findings indicative of infection of body of
graft and perigraft tissues, ureteric obstruction
The overall accuracy of indium-111 white blood cell scans (? bilateral)
has been reported to be 83–88 per cent12,13 and the specificity
 Diagnosis 301

Gallium whole body

24-hour anterior

Gallium SPECT
Coronal slices from anterior ⎯→ posterior

? ?
?

wel
Bo

Figure 25.4 Whole-body gallium and selected computed tomography (CT) scan images of a patient with an infected aortic graft. Uptake in
the body of the graft (arrows) can be seen on the 24-hour gallium single photon emission computed tomography (SPECT) images. Selected
CT images (lower three) from the same area show inflammatory changes and air adjacent to the graft (reproduced with permission from
Drane WE (ed). Nuclear Medicine: The basic, the advanced and the controversial. Gainesville, FL: NMNF, Inc)

undergo imaging of the graft as described above to deter-

• CT-guided aspiration of aortic perigraft fluid for mine the presence of signs of infection of the aortic graft.
culture Demonstration of an anastomotic pseudoaneurysm, par-
• Angiographic evidence of pseudoaneurysm ticularly at the proximal aortic anastomosis, is pathogno-
(multiple) at aortic/femoral anastomosis, graft limb monic and any evidence of aortic graft infection also
thrombosis essentially confirms the diagnosis (Fig. 25.5). In the
• Radionuclide imaging using Tc-99m hexametazime, absence of evidence of vascular graft infection on the imaging
gallium-67 or indium-111, but they add little to studies, upper and lower gastrointestinal endoscopy
CT/MRI findings should be done to search for other causes of the bleeding.
• Upper gastrointestinal endoscopy – negative study Upper gastrointestinal endoscopy should also be done in a
unreliable patient who is bleeding more rapidly, haemodynamic sta-
• Surgical exploration and culture of biofilm bility permitting, to exclude other causes such as peptic
ulcer disease or oesophageal varices. Furthermore, a nor-
mal study except for bleeding from the third or fourth por-
The appropriate use of diagnostic studies in a patient tion of the duodenum essentially establishes the diagnosis
suspected of having an aortic graft-enteric erosion depends of aortic graft-enteric erosions in such patients. In con-
on the degree of the gastrointestinal haemorrhage and trast, a normal study without evidence of bleeding does not
the haemodynamic stability of the patient. When the exclude the diagnosis of graft-enteric erosion as anasto-
bleeding is slow and intermittent the patient should motic fistulous connections can occur to the stomach,
302 Prosthetic aortic graft infection

Antibiotic therapy should be initiated based on graft and/or

wound cultures, if available, or on the pathogens most likely
to be present, depending on the location and timing of the
graft infection. Plans should also be made for the adminis-
tration of intravenous antibiotics for 4–6 weeks after graft
removal and possibly oral antibiotics for an additional 6–12
weeks. Importantly, operative records of implantation of
the prosthetic graft now infected should be reviewed to
determine the precise anatomy of the infected vascular graft
so that graft removal and reconstruction can be planned
appropriately.
Clearly, in this serious situation, the patient and family
members must be counselled preoperatively regarding
the extreme gravity of the situation in order to avoid false
expectations. Taken overall, postoperative mortality ranges
from 10 to 30 per cent, amputation from 0 to 15 per cent,
reinfection post reconstruction 0 to 20 per cent and 5-year
survival averages 50 per cent. Furthermore, multiple surgi-
cal procedures are frequently required to preserve lower
extremity perfusion.
Figure 25.5 Computed tomography (CT) scan of patient with an
aortic graft infection subsequently found to have a graft-enteric
erosion into the small bowel. Although the air seen adjacent to the
graft is suggestive of graft-enteric erosion, infection by a gas SURGICAL TREATMENT
producing organism may also be responsible
Infection of a prosthetic aortic graft is a grave problem. In
small bowel or colon. Because of this, in the face of persist- our experience, only 58 per cent of patients with this serious
ent, intermittent bleeding or massive bleeding without a problem are alive and still have their limbs at 30 months
definitive cause, surgical exploration is mandatory, graft- after treatment though this varies with the original indica-
enteric erosion being excluded only when the bowel is tion for placement of the infected aortic graft: 48 per cent in
completely freed from the retroperitoneum. Complete patients whose grafts were placed for occlusive disease, 65
graft incorporation in this setting effectively rules out per cent for patients whose grafts were placed for aneurys-
erosion or infection. mal disease.16 Types of treatment currently used in patients
with prosthetic aortic graft infection include first, staged
graft excision and extra-anatomical bypass and second,
PREOPERATIVE MANAGEMENT simultaneous graft excision and in situ graft replacement
using a variety of types of graft, namely a new prosthetic
graft, an autogenous graft constructed from deep vein or a
As previously noted, patients with infected prosthetic aortic
homograft. Graft preservation may also be attempted in
grafts are usually elderly and have multiple medical prob-
selected patients with presumed graft contamination rather
lems. In addition, they may be debilitated by chronic sepsis
than obvious infection and in those with evidence of low
and at times by acute or chronic blood loss. Treatment of
grade infection and severe life-limiting medical comorbidi-
these problems is therefore essential in reducing morbidity
ties. Each of these approaches has advantages and disadvan-
and mortality associated with management of the infected
tages and no one type of therapy is applicable to all patients.
vascular graft. Fortunately, time for possible correction or
Therefore, in the individual patient with an aortic graft
improvement of such problems and preparing the patient
infection, many things must be considered to determine the
for surgical treatment, is usually available, as most proced-
treatment most likely to achieve eradication of the infection
ures are urgent rather than emergent in nature. Specific
and long term limb salvage with the lowest risk (Fig. 25.6).
preparation for treatment of vascular graft infection
includes bilateral non-invasive vascular testing to assess
limb haemodynamics and arteriography of both lower and Aims of treatment
upper extremities, including the aortic arch when appropri-
ate. This defines the anatomy of arteries adjacent to the • Control bleeding and save life
infected vascular graft and identifies potential extra- • Restore flow
anatomical bypass inflow sites. Venous duplex ultrasound • Eradicate infection
should also be done to assess potential venous conduits, • Prevent amputation(s)
including deep leg veins, available for arterial reconstruction.
 Surgical treatment 303

Suspected aortic graft infection

CT scan
Thin cut with contrast

Signs of infection Signs of infection No signs of aortic graft infection

thoracic/ infrarenal grafts
thoracoabdominal graft
Other source of sepsis identified
Angiogram/non- Yes No
Angiogram/non- invastive testing
invasive testing
Treat sepsis/close Other findings associated with graft
follow-up infection/recent episodes of sepsis
In situ homograft
replacement Yes No

Angiogram/non- Follow-up with

Older, high risk Younger, low risk
invasive testing serial CT scans

Suitable SFV
Explore

Suitable proximal SFA/ No Yes

profunda infrarenal aortic cuff*

Yes No In situ Obvious graft infection Localised Graft welt

autogenous perigraft infection incorporated

Staged extra- bypass

Homograft Younger, low risk Elderly, high risk
anatomical replacement
replacement adequate SFV Segmental Close and follow
bypass/graft replacement with
excision antibiotic bonded
Close, staged
In situ autogenous treatment of prosthetic graft or SFV
bypass replacement infected aortic graft

Figure 25.6 Algorithm for the diagnosis and management of aortic graft infections. *Adequate length of infrarenal aorta for secure
stump closure, patent SFA or profunda with adequate run-off. SFA, superficial femoral artery; SFV, superior femoral vein

Infection can occur in prosthetic vascular grafts in any infected aortic grafts. Thirty-day mortality was 6.8 per cent
portion of the aorta. Fortunately, most infected aortic grafts for patients without aortoenteric fistulae (AEF) and
encountered by the vascular surgeon are in the infrarenal 65 per cent for patients with AEF. Again, there were signifi-
aorta while infected suprarenal and thoracic aortic grafts cant allograft complications in almost one-third of the
are very uncommon. Graft excision and in situ prosthetic patients: allograft rupture in 10, thrombosis in 9, dilatation
graft or homograft replacement is the usual treatment for in 7 and stenosis in 4. The mortality rate as a direct result of
infected thoracic and suprarenal aortic grafts as graft exci- early and late homograft failure was 21 per cent.
sion and extra-anatomical bypass is generally not possible. Hayes et al.3 used new rifampicin-bonded polyester grafts
Only limited reports of this approach in patients with this to replace infected prosthetic aortic grafts in 11 patients
exceedingly complex problem are available, but average with a mortality of 18.2 per cent and no early amputations.
mortality rates of 10–36 per cent, and of complication rates Unfortunately, once again late catastrophic complications
such as 5 per cent for amputation and approximately 20 were common with two patients dying of recurrent graft
per cent for early graft failure/infection, appear acceptable. infections within 30 months. Finally, Bandyk et al.19 treated
In situ graft replacement has also been used in the man- 40 patients with prosthetic vascular grafts infected by
agement of selected patients with infrarenal prosthetic aor- biofilms caused by Staphylococcus spp., including 25 with
tic graft infection. Kieffer et al.17 used aortic allografts to infected aortic grafts, by graft excision and in situ graft
replace infected infrarenal prosthetic grafts in 43 patients, replacement using rifampin-bonded polyester and poly-
including 36 with aortic graft infections, with a mortality tetrafluoroethylene (PTFE) grafts with a 2.5 per cent 30-day
of 12 per cent and no early amputations. There were, mortality and no limb loss. No late graft failures were seen
however, early allograft complications in two patients and at an average follow-up of 18 months but 10 per cent devel-
late allograft complications in nine patients, i.e. 26 per cent oped recurrent graft infection. Early reinfection due to
overall, including one death thought to be due to allograft rapid development of rifampin-resistant S. epidermidis has
graft rupture. Verhelst et al.18 reported results from a mul- also been reported by Bandyk et al.20 It seems likely that
ticentre trial of the use of cryopreserved homografts in 90 rifampin has limited efficacy against Gram negative and
patients with vascular graft infections, including 41 with MRSA organisms in vivo. Because of these findings,
304 Prosthetic aortic graft infection

we limit the use of in situ replacement of infected prosthetic in situ aortic graft replacement using one constructed from
aortic grafts, using either allografts or rifampin-bonded superficial femoral vein remains a demanding procedure
polyester grafts, to patients with suprarenal or thoracic for both patient and surgeon.
aortic infections, a select group of patients with AEF, as will At present, we use both staged extra-anatomical bypass
be discussed later, and those few patients not appropriate with graft excision, and in situ graft replacement using a
for treatment with either of the two main options. graft constructed from superficial femoral vein, in the treat-
Excision and extra-anatomical bypass using a prosthetic ment of aortic graft infection. The treatment used in an indi-
graft has been the most commonly used treatment of vidual patient is selected based on the location of the
infrarenal aortic graft infection and results of treatment of infected graft, the degree of occlusive disease present and the
aortic infection using this approach have gradually patient’s age and other medical problems. If the infected
improved since it was introduced by Blaisdell et al. in aortic graft is confined to the abdomen, staged axillofemoral
1970.21 This was particularly true following the observa- bypass to the common femoral artery followed by graft exci-
tion by Reilly et al.22 that staged extra-anatomical bypass sion appears to be the simplest method of managing a
followed by graft excision was associated with lower mor- patient with this problem. The procedure is staged because
tality and improved initial limb salvage. Postoperative such an approach clearly limits the physiologic stress com-
mortality and amputation rates after treatment of an aortic pared with simultaneous extra-anatomical bypass and graft
graft infection with this approach now average 10–12 per cent removal and, surprisingly, the staged approach is also
and 0–10 per cent, respectively, and Yeager et al.23 recently associated with a lower rate of graft failure, amputation and
reported a 73 per cent primary and 92 per cent secondary infection of the extra-anatomical bypass.22
extra-anatomical bypass graft patency at 5 years in 50
patients treated for aortic graft infection in this manner. They
also demonstrated that late aortic stump disruption, which Operative approaches
previously had been a common cause of late death, was now
uncommon. Seeger et al.24 reported similar results in 36 • Staged graft excision and extra-anatomical bypass
patients with 64 per cent primary and 100 per cent secondary using axillofemoral, obturator foramen bypass or
patency at 5 years with one case of aortic stump disruption. autogenous vein iliofemoral or femoro-femoral bypass
Thus, modern results of this approach to treatment of aortic • Simultaneous graft excision and in situ graft
graft infection appear acceptable. The long term patency of replacement using new prosthetic graft (rifampicin-
extra-anatomical bypass after aortic graft excision, however, bonded polyester or PTFE), autogenous graft
remains problematic and this has led to the investigation of constructed from deep vein and/or
in situ replacement of infected aortic grafts with a new pros- endarterectomised aortoiliac and superficial artery
thetic graft or a homograft, as discussed above, or a graft con- segments, homograft
structed of infection-resistant autogenous tissue. • Graft preservation
Ehrenfeld et al.25 and Seeger et al.26 used grafts con- • Other procedures – closure of bowel, fasciotomies,
structed of endarterectomised aortoiliac and superficial amputation(s)
artery segments combined with portions of saphenous vein • Endovascular option – stent-graft to arrest bleeding
to replace 15 and 10 infected aortobifemoral grafts, with plus one of the above approaches
early mortalities of 20 per cent and 9 per cent and early
amputation rates of 7 per cent and 9 per cent, respectively.
Unfortunately, long term graft failure, due primarily to Similarly, if the infection appears to be limited to the dis-
stenosis of the saphenous vein portion of the autogenous tal portion of one limb of an aortobifemoral bypass graft, we
graft, occurred in two-thirds of the patients in the report by use staged extra-anatomical bypass, i.e. axillary to profunda
Seeger et al.26 More recently, Clagett et al.4 and Nevelsteen femoris artery bypass or axillary to superficial femoral artery
et al.27 reported the results of the use of autogenous bypass (Fig. 25.7), and excision of the infected portion of the
grafts constructed using deep femoral veins to treat patients graft. If the ipsilateral profunda femoris and superficial
with infected aortic grafts. Postoperative mortality rates in femoral arteries are unsuitable as distal targets for an extra-
these studies were 10 per cent and 7 per cent, and early anatomical bypass in this setting, an obturator foramen
amputation rates 5 per cent and 7 per cent, respectively. bypass or an autogenous graft iliofemoral or femoro-
Furthermore, in the Clagett study, primary graft patency femoral may be used to provide lower extremity perfusion
at 5 years was 83 per cent, secondary graft patency 100 per after the infected graft limb has been excised. The above
cent and significant lower extremity oedema was uncom- knee popliteal artery, however, must be patent for an obtur-
mon. However, major morbidity occurred in 49 per cent of ator foramen bypass to be successful, the use of an in situ
the patients in Clagett’s series, including compartment syn- autogenous graft replacement risks spread of infection to the
drome in 12 per cent and limb paralysis in 7.5 per cent. non-infected portion of the graft through the autogenous
More recently, these two complications appear to have been graft tunnel. Axillopopliteal bypass is not used because of
reduced by the liberal use of prophylactic fasciotomies, but extremely poor patency of this type of bypass in this setting,
 Management of prosthetic graft erosions/fistulae 305

secondary patency being 33 per cent at 12 months.16

Alternatively, if a localised biofilm infection is present, graft
excision and in situ prosthetic PTFE graft replacement is
reported to have equivalent results to either graft excision
and extra-anatomical bypass or in situ graft replacement
with deep vein, although published results concerning this
approach are limited.
When the infection involves the body or both limbs of a
prosthetic aortobifemoral bypass graft, thus involving both
common femoral arteries with infection, the choice of treat-
ments is between extra-anatomical bypass followed by graft
excision and simultaneous graft excision and in situ replace-
ment with an autogenous graft constructed from superficial
femoral vein. As noted previously, reported postoperative
mortality and early risk of amputation are essentially the
same after these two procedures. Nevertheless, the incidence
of major postoperative morbidity after in situ replacement
of an infected aortobifemoral graft with an autogenous graft
made of superficial femoral vein was as high as 49 per cent in
(a) early reports4 while the long term risk of graft failure and
subsequent graft infection is lower than after extra-anatomi-
cal bypass and aortic graft excision. Based on these results, we
choose graft excision and autogenous graft replacement for
younger, healthier patients who would benefit from a more
definitive procedure and those patients whose profunda
femoris and proximal superficial femoral arteries are unsuit-
able as distal targets for extra-anatomical bypasses (Fig. 25.8).
In contrast, we use staged extra-anatomical bypass and graft
excision in patients with infected aortobifemoral bypass
prostheses.

MANAGEMENT OF PROSTHETIC GRAFT

EROSIONS/FISTULAE

Patients with slow intermittent gastrointestinal bleed-

ing from prosthetic graft-enteric erosions but who are
haemodynamically stable can be managed in a manner simi-
lar to those with such infections but no evidence of gas-
trointestinal bleeding (Fig. 25.9). In fact, such patients are
usually thought to have graft infection alone and the graft-
enteric erosion is found incidentally at the time of removal
of the infected graft. In a patient who is actively bleeding
with a known or strongly suspected aortoenteric fistula,
the traditional approach has been control of the fistula and
removal of the aortic graft followed by extra-anatomical
bypass for lower limb revascularisation. Alternatively, fistula
(b)
repair and in situ graft replacement with a new prosthetic
graft has been described by Walker et al.28 and repair of the
Figure 25.7 Arteriogram showing the distal anastomoses in a
patient treated for aortic graft infection by staged bilateral
fistula alone has been reported by Eastcott.29 Reported
axillofemoral bypasses and graft excision. (a) The right bypass is to mortality rates for AEF using the traditional approach to
the profunda femoris artery, the right superficial femoral artery treatment are up to 40 per cent with an early amputation
having occluded, and (b) the left bypass is to the proximal rate of up to 10 per cent. In contrast, mortality after fistula
superficial femoral artery, the left profunda femoris branching repair and in situ prosthetic graft replacement was reported
early and being of relatively small calibre to be 22 per cent with an amputation rate of 0 per cent; after
 (a) (b)

Figure 25.8 Arteriogram of a patient previously treated for aortic graft infection with graft removal and in situ graft replacement constructed
from superficial femoral vein showing (a) the proximal aortic anastomosis and body of the graft and (b) the distal anastomoses. This approach was
chosen because first, both profunda femoris arteries were occluded and second, there was severe superficial femoral artery occlusive disease

GI bleeding in patients with aortic

graft (suspected AEF, AEE)

Haemocynamically STABLE
Yes No

CT Scan – Thin cut with contrast Endoscopy in OR if

possible
Evidence of No evidence of
graft infection graft infection
Endoscopy to exclude Endoscopy Explore
other sites of bleeding AEF GI source/
no AEF

Treat as infected graft in GI source No source Repair fistula Treat as appropriate

an expedited manner identified identified
(Fig. 25.6)
Close, resuscitate,
Treat as Intermittent bleeding treat as infected graft
appropriate continues in expedited manner
(Fig. 25.6)

Exploration
Fistula Negative

Graft incorporated Close and follow

Yes No

Repair fistula/ Repair fistula

local graft
replacement
Close, resuscitate,
treat as infected graft
Follow with serial CT in expedited manner
(Fig. 25.6)

Figure 25.9 Algorithm for the diagnosis and management of aortoenteric fistulae or erosions. AEE, aortoenteric erosion; AEF,
aortoenteric fistula; CT, computed tomography; GI, gastrointestinal; OR, operating room
 Late complications 307

fistula repair alone mortality was 0 per cent with an ampu-

tation rate of 0 per cent. After fistula repair and in situ
Late complications
prosthetic graft replacement, three patients (17 per cent)
had recurrent graft infection and rupture of the proximal • Stenosis/failure of extra-anatomical or in situ bypass
anastomosis, causing death in two, while after fistula repair • Degeneration/rupture of graft
alone, four patients (40 per cent) had evidence of recurrent • Aortic stump disruption
infection, all of whom subsequently underwent graft exci-
sion and extra-anatomical bypass. Aortic stump disruption, although uncommon, does
Based on the results of these studies, we have recently occur and, in our experience, patients who require repo-
adopted a modified approach to treating the patient with an sitioning of the renal arteries, and sometimes of the superior
actively bleeding AEF, similar to that suggested by Stoney:30 mesenteric artery to allow adequate aortic debridement, are
the patient is rapidly explored, the bleeding is controlled, at high risk. Homograft degeneration occurs in at least 25
the fistula is identified and the bowel is closed. If the major- per cent of patients, commonly resulting in graft rupture if
ity of the aortic graft is well incorporated, the graft is cul- left untreated. Durability of autogenous grafts constructed
tured, the proximal portion of the graft is replaced with a from superficial femoral vein appears good at present but
PTFE graft, the patient is treated with 6 weeks of antibiotics follow-up has been limited and graft or anastomotic sten-
and is then followed carefully with serial CT scanning. osis has been reported.4 Because of all of these factors, we fol-
Alternatively, if the graft is obviously infected, the bowel and low patients after successful treatment of infected prosthetic
fistula are repaired, the abdomen is closed and the patient is vascular grafts at 3, 6, 12, 18 and 24 months and yearly there-
allowed to recover for a few days after which the aortic graft after for haemodynamic testing and CT imaging at each
infection is dealt with in a less urgent setting. Our experience visit. Long term suppressive antibiotics, beyond the previ-
with this modified protocol for the treatment of AEF is ously described 3–4 months after graft removal, are not used
limited and the follow-up brief but the results have been as there is no evidence to support the value of such an
good: 0 per cent mortality, 0 per cent amputation rate. approach. Only to those patients in whom an attempt is
A novel, endovascular approach to the management of made to preserve the infected aortic graft are suppressive
AEF has also been recently described in three case reports.31–33 antibiotics given.
In each case, the infrarenal aortic graft had eroded into the
duodenum resulting in haemorrhage, and deployment of a
covered aortic stent-graft within the old graft successfully Conclusions
arrested the bleeding. In one case, further erosion occurred
during follow-up requiring subsequent staged extra- Treatment of infected prosthetic vascular grafts remains
anatomical bypass and graft excision.32 Nevertheless, these challenging. Familiarity with multiple types of treatment,
reports suggest a possible role for an aortic stent-graft as a experience with complex surgical reconstructions and
method of controlling active bleeding in patients with AEF, careful long term follow-up is necessary to achieve opti-
potentially with less physiological stress. More experience mal results in patients presenting with this grave problem.
with these new approaches will be required to characterise Fortunately, prosthetic vascular graft infection is uncom-
their role in the management of AEF and only a larger mon but by the same token significant experience in the
experience and longer follow-up will determine whether management of patients with this problem is also com-
the change in our approach to these patients is of value. paratively rare. For this reason, it is important that patients
with prosthetic vascular graft infection are referred to cen-
tres with significant experience in this area. With appro-
LATE COMPLICATIONS priate application of the techniques currently available for
treatment of prosthetic aortic graft infection, long term
Patients must be carefully followed up, essentially for life, elimination of infection and limb preservation is possible
after initially successful treatment of an infected aortic graft. and can be achieved in the great majority of patients.
Depending on the type of treatment used, they are at risk of
developing complications associated with extra-anatomical
or in situ bypass graft stenosis/failure, aortic stump disrup-
tion, recurrent infection of new in situ prosthetic grafts or Key references
extra-anatomical bypass grafts and degeneration/rupture of
homograft or autogenous grafts. Extra-anatomical bypass Bandyk DF, Novotney ML, Back MR, et al. Expanded application of
graft failure occurs in approximately 25 per cent of in situ replacement for prosthetic graft infection. J Vasc Surg
patients23 and can be treated with either repair of the failed 2001; 34: 411–20.
graft or re-do aortic grafting as long as 6–12 months have Clagett GP, Valentine RJ, Hagino RT. Autogenous aortoiliac/femoral
reconstruction from superficial femoral-popliteal veins:
passed after removal of the infected aortic graft, as described
feasibility and durability. J Vasc Surg 1997; 25: 255–70.
by Dimuzio et al.34
308 Prosthetic aortic graft infection

Low RN, Wall SD, Jeffery RB Jr, et al. Aortic enteric fistula and 16 Seeger JM, Back MR, Albright JL, et al. Influence of patient
perigraft infection; evaluation by CT. Radiology 1990; 175: characteristics and treatment options on outcome of patients
157–62. with prosthetic aortic graft infection. Ann Vasc Surg 1999; 13:
Seeger JM, Pretus HA, Welborn MB, et al. Long term outcome after 413–20.
treatment of aortic graft infection with staged extra-anatomic 17 Kieffer E, Bahnini A, Koskas F, et al. in situ allograft replacement
bypass grafting and aortic graft removal. J Vasc Surg 2000; of infected infrarenal aortic prosthetic grafts: results in forty-
32: 451–61. three patients. J Vasc Surg 1993; 17: 349–56.
Verhelst R, Lacroix V, Varaux H, et al. Use of cryopreserved arterial 18 Verhelst R, Lacroix V, Varaux H, et al. Use of cryopreserved
homografts for management of infected prosthetic grafts: arterial homografts for management of infected prosthetic
a multicentric study. Ann Vasc Surg 2000; 14: 602–7. grafts: a multicentric study. Ann Vasc Surg 2000; 14: 602–7.
19 Bandyk DF, Novotney ML, Back MR, et al. Expanded application
of in situ replacement for prosthetic graft infection. J Vasc Surg
REFERENCES 2001; 34: 411–20.
20 Bandyk DF, Novotney ML, Johnson BL, et al. Use of rifampin-soaked
gelatin-sealed polyester grafts for in situ treatment of primary
1 Liekweg WG Jr, Greenfield LJ. Vascular prosthetic infections: aortic and vascular prosthetic infections. J Surg Res 2001; 95: 44–9.
collected results of treatment. Surgery 1977; 81: 335–42. 21 Blaisdell FW, Hall AD, Lim RC Jr, Moore WC. Aorto-iliac arterial
2 Yeager RA, Porter JM. Arterial and prosthetic graft infection. Ann substitution utilizing subcutaneous grafts. Ann Surg 1970; 172:
Vasc Surg 1992; 6: 485–91. 775–80.
3 Hayes PD, Nasim A, London NJM, et al. in situ replacement of 22 Reilly LM, Stoney RJ, Goldstone J, Ehrenfeld WK. Improved
infected aortic grafts with rifampicin-bonded prostheses: the management of aortic graft infection: the influence of operation
Leicester experience (1992 to 1998). J Vasc Surg 1999; 30: 92–8. sequence and staging. J Vasc Surg 1987; 5: 421–31.
4 Clagett GP, Valentine RJ, Hagino RT. Autogenous aortoiliac/ 23 Yeager RA, Taylor LM, Moneta GL, et al. Improved results with
femoral reconstruction from superficial femoral-popliteal veins: conventional management of infrarenal aortic infection. J Vasc
feasibility and durability. J Vasc Surg 1997; 25: 255–70. Surg 1990; 30: 76–83.
5 Vogt PR, Brunner-La Rocca HP, Carrel T, et al. Cryopreserved 24 Seeger JM, Pretus HA, Welborn MB, et al. Long term outcome
arterial allografts in the treatment of major vascular infection: after treatment of aortic graft infection with staged extra-
a comparison with conventional surgical techniques. J Thorac anatomic bypass grafting and aortic graft removal. J Vasc Surg
Cardiovasc Surg 1998; 116: 965–72. 2000; 32: 451–61.
6 Chiesa R, Astore D, Piccolo G, et al. Fresh and cryopreserved 25 Ehrenfeld WK, Wilbur BG, Olcott CN IV, Stoney RJ. Autogenous
arterial homografts in the treatment of prosthetic graft tissue reconstruction in the management of infected prosthetic
infections: experience of the Italian Collaborative Vascular grafts. Surgery 1979; 85: 82–92.
Homograft Group. Ann Vasc Surg 1998; 12: 457–62. 26 Seeger JM, Wheeler JR, Gregory RT, et al. Autogenous graft
7 Padberg FT, Smith SM, Eng RHK. Accuracy of disincorporation replacement of infected prosthetic grafts in the femoral position.
for identification of vascular graft infection. Arch Surg 1995; Surgery 1983; 93: 39–45.
130: 183–7. 27 Nevelsteen A, Lacroix H, Suy R. Autogenous reconstruction with
8 Southern FN, Eidt JF, Barnes RW, Moursi MM. Thrombosis and the lower extremity deep veins: an alternative treatment of
infection in aorta femoral bypass grafts. Presented at the 23rd prosthetic infection after reconstruction surgery for aortoiliac
Annual Meeting of the Southern Association for Vascular disease. J Vasc Surg 1995; 22: 129–34.
Surgery, January 1999. 28 Walker WE, Cooley DA, Duncan JM, et al. The management of
9 Low RN, Wall SD, Jeffery RB Jr, et al. Aortic enteric fistula and aortoduodenal fistula by in situ replacement of the infected
perigraft infection; evaluation by CT. Radiology 1990; 175: 157–62. abdominal aortic graft. Ann Surg 1987; 205: 727–32.
10 Qvarfortdt PG, Reilly LM, Mark AS, et al. Computerized 29 Eastcott HHG. Aortoenteric fistula. possibilities for direct repair.
tomographic assessment of graft incorporation after aortic In: Greenhalgh RM, ed. Extra-anatomic and Secondary Arterial
reconstruction. Am J Surg 1985; 150: 227–31. Reconstruction. London: Pitman, 1982: 58.
11 Spartera C, Morettini G, Petrassi C, et al. Role of magnetic resonance 30 Stoney RJ. Discussion of Peck JJ, Eidemiller LR: Aortoenteric
imaging in the evaluation of aortic graft healing, perigraft fluid Fistula. Arch Surg 1992; 127: 1191–4.
collection and graft infection. Eur J Vasc Surg 1990; 4: 69–73. 31 Deshpande A, Lovelock M, Mossop P, et al. Endovascular repair of
12 Brunner MC, Mitchell RS, Baldwin JC, et al. prosthetic graft an aortoenteric fistula in a high risk patient. J Endovasc Surg
infection: limitations of indium white cell scanning. J Vasc Surg 1999; 6: 379–84.
1986; 3: 42–8. 32 Chuter TA, Lukaszewicz GC, Reilly LM, et al. endovascular repair
13 Reilly DP, Grigg MJ, Cunningham DA, et al. Vascular graft infection: of a presumed aortoenteric fistula: late failure due to recurrent
the role of indium scanning. Eur J Vasc Surg 1989; 3: 393–7. infection. J Endovasc Ther 2000; 7: 240–4.
14 Johnson KK, Russ PD, Bair JH, Friefeld GD. Diagnosis of synthetic 33 Grabs AJ, Irvine CD, Lusby RJ. Stent-graft treatment for bleeding
vascular graft infection: comparison of CT and gallium scans. from a presumed aortoenteric fistula. J Endovasc Ther 2000; 7:
Am J Roentgenol 1990; 154: 405–9. 236–9.
15 Fiorani P, Speziale F, Rizzo L, et al. Detection of aortic graft 34 Dimuzio PJ, Reilly LM, Stoney RJ. Redo aortic grafting after
infection with leukocytes labeled with technetium treatment of aortic graft infection. J Vasc Surg 1996; 24:
99m-hexametazime. J Vasc Surg 1993; 17: 87–95. 328–37.
 26
The Acutely Compromised Renal Artery

MICHAEL P JENKINS, GEORGE HAMILTON

The problem 309 Trauma 312

Iatrogenic injury in renal artery interventions 309 Problems during aortic surgery 313
Embolism and atheroembolism 312 References 314

THE PROBLEM chapter is based on data from reported case series, first
principles and personal experience.
A kidney is in imminent danger once its blood supply is
interrupted for whatever reason. A previously healthy kid-
IATROGENIC INJURY IN RENAL ARTERY
ney without prior vascular compromise will survive for only
INTERVENTIONS
40–50 minutes without a blood supply. In the presence of
chronic renal artery stenosis (RAS) the kidney is precondi-
tioned to ischaemia, thereby allowing it to survive without Thrombosis
a blood supply for up to 60 minutes and sometimes even
90 minutes before irreversible damage occurs. This time Despite the advent of computed tomography (CT) and
between interruption of arterial supply and renal infarc- magnetic resonance (MR)-based diagnostic imaging, any
tion is known as the warm ischaemia time during which it renal artery intervention will, as a prerequisite, involve
is mandatory to achieve reperfusion or employ renal cool- selective catheterisation. Even prior to attempted balloon
ing if kidney survival is to be ensured. angioplasty or stent placement, catheter manipulation can
Renal blood flow may be compromised either by lead to plaque disruption or dissection, both of which can
rupture or occlusion of the main renal artery and these progress to renal artery thrombosis. Although primary
scenarios will be considered in turn. Trauma and iatro- stenting is increasingly used for treatment of ostial athero-
genic injury are the commonest causes of acute injury, sclerotic stenoses, more distal lesions and those secondary
but in the latter situation, pre-existing atherosclerotic or to fibromuscular dysplasia are more likely to be treated by
fibromuscular disease is common. Increasing use of angioplasty without stenting. At present there is no good
catheter-based diagnostic and therapeutic interventions evidence to suggest that renal artery stenting has any long
result in manipulation in the region of the renal ostia. term beneficial effect in terms of hypertension control or
These procedures may jeopardise renal perfusion even preservation of renal function over angioplasty alone.1
without selective catheterisation or renal artery interven- In non-ostial stenoses there is therefore a place for
tion per se. keeping stent placement in reserve to treat extensive dis-
In this chapter the more commonly encountered causes section or occlusion, albeit accepting that such complica-
of renal artery compromise will be documented, the thera- tions are less likely to occur away from the ostium. The
peutic options available discussed and authors’ preferential most comprehensive review of thrombotic complications
treatment modality highlighted. Individual experience of in percutaneous renal artery interventions (as seen in Fig.
such situations is unlikely to be great and the often unique 26.1) suggests an incidence of 2.3 per cent.2
set of circumstances involved and need for urgency, If possible, an acute dissection or thrombosis secondary
exclude any hope of meaningful trials contributing to our to plaque rupture should be treated immediately with
current knowledge. Thus, much of the advice within this the placement of a stent. If successful, this is obviously an
310 The acutely compromised renal artery

(a) (b) (c)

Figure 26.1 Angiogram showing thrombosis of the left renal artery following balloon angioplasty prior to the use of stents.
(a) Preintervention, (b) during balloon inflation and (c) post angioplasty. Today, this situation may be averted by primary stenting or
rescued by the introduction of a stent post angioplasty

attractive proposition as it maintains the minimally inva- extend the renal ischaemia time and if there is concern
sive nature of the procedure and obviates the time delay regarding renal viability, the kidney should be cooled with
involved in transferring a patient to theatre for open sur- topical ice and the renal artery perfused with ice-cold per-
gery. Under such conditions, secondary stenting for angio- fusate. Ideally such a patient would have been better
plasty failure or complications thereof has been reported treated by surgery in the first place as extensive aortic occlu-
to offer equivalent outcome results to primary angioplasty, sive or aneurysmal disease is a relative contraindication to
but at a cost of a 9.1–21.0 per cent major complication a percutaneous approach.
rate.3,4 If unable to place a stent or in the face of a more dis- If the aorta is in good condition a Dacron or polytetra-
tal thrombosis, urgent surgery is the treatment of choice, fluoroethylene (PTFE) bypass could be taken directly from
although in some cases there may be a role either for the infrarenal aorta. This is probably the fastest approach
thrombolysis or for doing nothing, both of which will be in an acutely ischaemic kidney. It also has the attraction of
discussed later. not requiring a complete aortic cross-clamp in a patient
In our unit, there is prearranged surgical cover for every who may not have had full preoperative preparation.
renal angioplasty. Notwithstanding the aforementioned In the presence of an unattractive infrarenal aorta, the
warm ischaemia times, a chronically stenosed renal artery supracoeliac portion often remains soft and well preserved
has often promoted collateralisation, allowing a window of despite extensive atheroma elsewhere. Access to this seg-
several hours in which to salvage a kidney even if the main ment is achieved via the lesser sac between the crura and
renal artery occludes. Wong et al. have reported a series of again controlled with a side-biting clamp obviating com-
51 consecutive patients salvaged operatively following failed plete aortic cross clamping. A prosthetic bypass can then
renal artery angioplasty.5 Although this included a mixed be fashioned to one or both kidneys with good long term
bag of indications for initial and emergency intervention, results, the latest reported series suggesting a 96 per cent
among the atherosclerotic cases, surgical revascularisation secondary patency rate at 5 years.6
was performed with a mortality of 9.4 per cent. In the pres- Other options include a splenorenal bypass to the left
ence of an on-table renal artery thrombosis, the operative renal artery or a hepatorenal to the right using either the
decision is based on anatomical criteria, i.e. the total blood gastroduodenal artery or a reversed saphenous vein bypass
supply of the kidney and whether any segmental branches from the common hepatic artery (see Fig. 26.2). Both these
remain patent in addition to preintervention physiological options require a healthy coeliac trunk and the dissection
factors. These include the overall renal function, differen- involved would be a relative contraindication in the acute
tial function of the contralateral kidney, angioplasty indi- situation.
cation and the fitness of the patient. Surgical options for Thrombolysis has been used to retrieve an iatrogeni-
revascularisation then depend on the condition of the cally thrombosed renal artery but there are few reports in
infrarenal aorta, whether it is aneurysmal or not, and the the literature, although Salem et al. have described a small
presence or absence of disease within the coeliac trunk. series7 and there are many anecdotal reports of success.
If the aorta is aneurysmal or contains a large amount of It would seem prudent to employ the fastest lysis technique
atheromatous plaque, then it should be replaced in the available and at present this would involve using a high dose
usual way and a 6–8 mm side branch sutured end-to-end ‘pulse spray’ technique. This may buy time to allow a further
to the renal artery. Obviously, such an approach will endovascular approach, with or without stent placement,
 Iatrogenic injury in renal artery interventions 311

perfusion of both kidneys. With a short neck, asymmetrical

renal take-off or a distal separate lower pole artery, comprom-
ises are sometimes made. A catheter should be retained
in the juxtarenal aorta to allow an angiogram after partial
deployment in order that the final position with respect to the
renal arteries can be checked prior to complete deployment.
The incidence of renal occlusion by the covered portion
of the stent-graft during EVAR is largely unknown, but in
Hopkinson’s series (Kalliafas et al.8) of 204 patients over
4 years, five (including one bilateral) renal artery occlusions
were reported. Although efforts should be made to reposi-
tion the graft, in reality this is often unsuccessful and the
situation is often managed expectantly. It is therefore vital
to know as much as possible about renal size and differen-
tial function preoperatively in order that a measured deci-
sion can be made at a time of crisis.
Contemporary issues surrounding EVAR and renal per-
fusion involve the ‘short proximal neck problem’ and long
Figure 26.2 Diagrammatic illustration of extra-anatomic
term renal artery patency with different stent-graft types. One
bypass procedure either to revascularise the right kidney from the
common hepatic artery (in approximately 40% of cases via the
novel method of ensuring adequate proximal stent fixation,
gastroduodenal branch) or to the left kidney using the splenic in a neck of inadequate length, is to perform an iliorenal
artery (the spleen remains in situ receiving blood from splenic bypass prior to stent deployment over the renal ostium.9
collaterals and the short gastric arteries). Alternatively, in either A more elegant approach uses the newly available fenestrated
case, a vein graft may be used. stent-grafts, allowing suprarenal deployment with preci-
sion alignment to allow flow through the fenestrations into
or allow the opportunity to arrange a formal surgical recon- the renal artery. This would allow many more patients with
struction. Thrombolysis should not be commenced as a difficult or short necks to be treated by EVAR.
‘knee-jerk’ reaction to an angiographic occlusion and it is Current stent-grafts effectively fall into two main types
important to weigh up the estimated time to clot dissol- with respect to proximal fixation technology: those with
ution against renal ischaemia time. uncovered suprarenal segments and those without. Although
If the contralateral kidney contributes the majority of the message from the RETA and EUROSTAR registries is
overall renal function and the indication for angioplasty is unclear, anecdotal evidence would suggest that migration
hypertension, then, in the presence of an on-table occlusion, problems are reduced in the former, but at the expense of
there is no mandate for insisting on revascularisation espe- an incidence of renal artery thrombosis during long term
cially in an unfit patient. Leaving the renal artery occluded follow up. At present it is unclear how common this is10,11
will remove the renin drive to uncontrolled hypertension (see Chapter 31).
without greatly affecting overall renal function.

Renal artery perforation

• Before embarking on renal angioplasty, ensure the
availability of an appropriate stent and surgical A perforation during catheter manipulation or more com-
cover. Stent placement should be attempted as the monly balloon dilatation is easily detected by the presence
first response to on-table thrombosis of extra-arterial contrast. Although dramatic, bleeding can
• Despite predicted ischaemia times, surgical often be stopped by a 10–15 minute period of balloon tam-
revascularisation can be worthwhile even after a ponade and heparinisation reversed with protamine while
few hours. If the situation is not salvageable blood is being cross-matched. If a longer period of occlu-
radiologically, it remains the preferred option sion is required, distal perfusion with ice-cold saline can be
• Only attempt revascularisation if the kidney employed. If this fails, a covered stent may be deployed to
contributes significantly to overall renal function seal the perforation and it is advisable to keep appropri-
ately sized stents on the shelf for just this purpose. Once
haemorrhage has been arrested the patient should be
Occlusion during endovascular aortic repair observed closely and re-imaged to exclude an expanding
retroperitoneal haematoma.
In endovascular aortic aneurysm repair (EVAR), accurate siz- If endovascular measures fail, surgical exploration is
ing and placement should allow a stent-graft to be deployed mandatory and during transport to theatre the angioplasty
sufficiently distal to the renal ostia to allow adequate balloon should remain deployed to reduce blood loss.
312 The acutely compromised renal artery

renal failure, but as stated above, the prognosis for renal

• Attempt catheter balloon occlusion and arrange for
recovery under these circumstances is poor.
availability of cross-matched blood
• Appropriately sized covered stents can be used very
effectively if available TRAUMA
• If surgery becomes necessary, do not remove the
balloon catheter as this can provide essential
tamponade during operative exposure Traumatic disruption of the renal pedicle is usually associ-
ated with high velocity and deceleration injuries typically
seen in road traffic accidents (see Chapter 36). They are
thus rarely seen in isolation and often occur in patients
EMBOLISM AND ATHEROEMBOLISM with significant multiple injuries. The emphasis in manage-
ment is therefore skewed towards arrest of haemorrhage and
preservation of life rather than renal reperfusion.
In the majority of cases renal emboli emanate from the heart Here, more than in any other circumstance, each case
and unlike the situation in occlusion following a long period must be treated based on its individual merit, but some
of RAS, acute occlusion of a previously normal renal artery principles will be outlined below. The first decision to
supplying a kidney without a collateral circulation rapidly make depends on the stability of the patient. In the face
lead to irreversible ischaemic damage. The insult may be of an unstable patient with a haemoperitoneum, urgent
silent or manifest itself late with signs of infarction: loin laparotomy is required. If a torn renal pedicle is encoun-
pain and haematuria. Either way it is a difficult condition tered, it should be ligated and the kidney sacrificed as the
to diagnose early and time is of the essence. priority lies with stabilising the patient and transferring
Once suspected, urgent angiography is the investigation him or her to the intensive care unit or dealing with the
of choice as it allows catheter placement for thrombolysis. other injuries. Assuming that a patient can be stabilised,
A number of reports in the literature suggest promising imaging can be used to confirm the presence of a functioning
results from local intra-arterial lysis, some with clot lysis contralateral kidney and whether there is any ongoing
achieved even after a 7-day period of thrombosis.12 haemorrhage. A decision can then be made to manage a
However, immediate radiological success does not neces- retroperitoneal hematoma conservatively or to embolise a
sarily translate into preservation of renal function and kidney in the presence of ongoing active bleeding. In a cold
longer term scintigraphy studies have shown lack of recov- multiply injured patient with a coagulopathy this approach
ery in many examples of complete occlusion despite appar- has many attractions compared with open surgery.
ent complete lysis. This is probably explained by the Revascularisation via an operative approach may be
existence of irreversible nephron ischaemia or reperfusion used in more isolated injuries associated with stabbing or
injury (see Chapters 2 and 4). low velocity ballistic injuries (see Chapter 36). Any of
In view of the time required to achieve complete lysis the revascularisation methods described above may be
and the known renal warm ischaemia time, one could employed with the proviso that in a contaminated wound,
make a case for urgent surgery. However, such patients a prosthetic conduit should be avoided.
are often very poorly, with a high incidence of ischaemic In the patient with a less severe initial injury managed
heart disease, and often a recent myocardial infarct as the conservatively, false aneurysm formation is not an uncom-
aetiology behind the embolus, so it is not surprising mon sequela. If involving the main stem renal artery, these
that surgical revascularisation carries a mortality of up to can often be treated with a covered stent, but more distal
25 per cent.13,14 lesions are problematic. A recent report16 suggests that
In the presence of a high burden of atherosclerotic soft microembolisation techniques can be very successful in
plaque commonly seen in this area of the aorta, athero- excluding the false aneurysm cavity without adversely
embolism is a very real risk. Indeed significant athero- affecting renal perfusion or function.
embolism leading to acute renal failure has a universally
poor prognosis. Cholesterol embolisation occurs acutely at
the time of catheterisation, but the resulting damage man-
ifests itself insidiously over the next few weeks. In one of
• The patient should be resuscitated and initial
management prioritised along Advanced Trauma
the largest reviews of histologically proven cholesterol Life Support (ATLS) principles
embolisation, Fine et al. reported that 17 per cent of 221
cases had undergone recent angiography.15
• Urgent laparotomy is required in an unstable
patient with a haemoperitoneum
The mainstay of treatment is supportive. Heparinisation
is instituted, but there is no evidence for any additional
• A conservative approach plus or minus endovascular
intervention is recommended in patients without
benefit from thrombolysis. Haemofiltration is required in life-threatening haemorrhage, especially in the
the face of the development of acid–base, electrolyte and presence of other injuries
fluid overload complications as in any other cause of acute
 Problems during aortic surgery 313

PROBLEMS DURING AORTIC SURGERY atherosclerosis. If this area is unattractive or difficult

to access safely, a supracoeliac clamp position is
employed to perform the proximal anastomosis
It cannot be stressed enough that the majority of aortic and the clamp moved down to the graft once the
aneurysm/renal artery configurations can be anticipated
from a good quality spiral CT and therefore a strategy for
renal protection or revascularisation should be worked out
in advance. However, this is not always possible in the case
of a very angulated or tortuous neck where it can be diffi-
cult to define the exact morphology and, of course, in the
majority of ruptured aneurysms imaging is not available.
The following scenarios can cause problems:

• Large lower pole renal artery: It is not uncommon for

multiple renal arteries to be present, but the majority
are close together and proximal to the aneurysm
neck. Less commonly, there is a large separate lower
polar artery which contributes significantly to renal
perfusion (Fig, 26.3) or even a low single renal artery
(Fig. 26.4). These should always be implanted
expeditiously using a Carrel patch technique. If the
proximal aortic anastomosis is predicted to be difficult
and time consuming, cooling can again be employed.
• Insufficient aortic neck: Clamp placement has to
ensure a sufficient segment of aortic neck below the
renal arteries to fashion an anastomosis. If the neck is
too short for this, the clamp can be placed above one
or both renal arteries and under such circumstances
Figure 26.4 Arteriogram illustrating asymmetrical renal artery
it is prudent to commence dopamine and mannitol.
take-off. The right renal artery was reimplanted into the Dacron graft
Such a situation should be predictable from a preoper-
ative CT scan (as seen in Fig. 26.5). If the left renal
vein is fully mobilised and slung, it is often possible to
avoid ligating and dividing it. However, if division is
essential to allow access, it should be performed as
there is no good evidence to suggest it has any deleteri-
ous effect on renal function.17 The suprarenal aorta can
then be inspected to assess a suitable site for cross-
clamping avoiding heavy calcification and extensive

Figure 26.5 Computed tomography reconstruction from which

it should be anticipated that aortic cross-clamping at renal artery
level would be difficult. In this case the aorta was clamped just
Figure 26.3 Angiogram illustrating a significant lower pole below the diaphragm and the clamp moved down once the
left renal artery which was reimplanted at the time of surgery proximal anastomosis had been completed
314 The acutely compromised renal artery

anastomosis is complete. The supracoeliac aorta is REFERENCES

best approached through the lesser sac, dividing the
crura if necessary, where it often remains relatively soft
1 van de Venn PJ, Kaatee R, Beutler JJ, et al. Arterial stenting and
even in the presence of disease above and below it.
balloon angioplasty in ostial atherosclerotic renovascular
disease: a randomised trial. Lancet 1999; 353: 282–6.
In a very diseased aorta, clamp placement abutting the
2 Bergentz S-E, Bergqvist D. Iatrogenic Vascular Injuries. Berlin:
renal arteries can risk plaque disruption leading to renal Springer Verlag, 1989.
artery occlusion. In such situations, the renal arteries 3 Bush RL, Najibi S, MacDonald J, et al. Endovascular
should be exposed and slung to allow an assessment of revascularization of renal artery stenosis: technical and clinical
renal blood flow on clamp removal. This can be done with results. J Vasc Surg 2001; 33: 1041–9.
digital palpation but is best done with a Doppler probe. 4 Boisclair C, Therasse E, Oliva VL, et al. Treatment of renal
If there is doubt regarding the adequacy of perfusion a direct angioplasty failure by percutaneous renal artery stenting with
puncture with a 21 gauge needle attached to a pressure Palmaz stents: midterm technical and clinical results. Am J
manometer will reveal any gradient with respect to the aortic Roentgenol 1997; 168: 245–51.
pressure. Renal artery endarterectomy with patch closure 5 Wong JM, Hansen KJ, Oskin TC, et al. Surgery after failed
percutaneous renal artery angioplasty. J Vasc Surg 1999;
would be the intervention of choice if necessary.
30: 468–82.
6 Paty PSK, Darling RC, Lee D, et al. Is prosthetic renal artery
reconstruction a durable procedure? An analysis of 489 bypass
Conclusions grafts. J Vasc Surg 2001; 34: 127–32.
7 Salem TA, Lumsden AB, Martin LG. Local infusion of fibrinolytic
Although the situations described above represent agents for renal artery thromboembolism: report of ten cases.
diverse scenarios, some basic principles can be applied Ann Vasc Surg 1993; 7: 21–6.
8 Kalliafas S, Albertini JN, Macierewicz J, et al. Incidence and
universally.
treatment of intra-operative problems during endovascular
1 The viability of the renal parenchyma should be repair of complex abdominal aortic aneurysms. J Vasc Surg 2000;
assessed prior to making any therapeutic decisions. 31: 1185–92.
2 The renal arterial anatomy must be defined 9 Lin PH, Madsen K, Bush RL, et al. Iliorenal artery bypass grafting
radiologically. The presence of multiple renal to facilitate endovascular abdominal aortic aneurysm repair. J
arteries, stenoses, low accessory lower pole arteries Vasc Surg 2003; 38: 183–5.
and the configuration of the neck of the aneurysm, 10 Lau LL, Hakaim AG, Oldenburg WA, et al. Effect of suprarenal
versus infrarenal aortic fixation on renal function and renal
if present, must be clarified before intervention is
artery patency: a comparative study with intermediate follow up.
planned. J Vasc Surg 2003; 37: 1162–8.
3 Surgical cover is mandatory for all endovascular 11 Bove PG, Long GW, Shanley CJ, et al. Transrenal fixation of
renal artery procedures. endovascular stent-grafts for infrarenal aortic aneurysm repair:
4 Endovascular complications should ideally be mid-term results. J Vasc Surg 2003; 37: 938–42.
rescued by immediate further endovascular 12 Fischer P, Konnak JW, Cho KJ, et al. Renal artery embolism:
intervention. therapy with intra-arterial streptokinase infusion. J Urol 1981;
5 During operative approaches, flexibility is crucial 125: 402–4.
and the many sources of possible inflow 13 Lacombe M. Surgical versus medical treatment of renal artery
remembered, while not forgetting the option of embolism. J Cardiovasc Surg 1977; 18: 281–90.
nephrectomy if appropriate. 14 Nicholas GC, De Muth WE. Treatment of renal artery embolism.
Arch Surg 1984; 119: 278–81.
15 Fine MJ, Kapoor W, Falanga V. Cholesterol crystal embolization:
a review of 221 cases in the English literature. Angiology 1987;
Key references 38: 769–84.
16 Cantasdemir M, Adaletli I, Kantarci F, et al. Emergency
Dyet JF, Ettles DF, Nicholson AA, Wilson SE. Textbook of endovascular embolization of traumatic intrarenal arterial
Endovascular Procedures. Philadelphia, PA: Churchill pseudoaneurysms with N-butyl cyanoacrylate. Clin Radiol 2003;
Livingstone, 2000: 151–73. 58: 560–5.
Earnshaw JJ, Murie JA. The Evidence for Vascular Surgery. 17 Elsharawy MA, Cheatle TR, Clarke JM, et al. Effect of left renal
Cheltenham: TFM Publishing Ltd, 1999: 165–72. vein division during aortic surgery on renal function. Ann R Coll
Novick A, Scoble J, Hamilton G. Renal Vascular Disease. London: Surg Engl 2000; 82: 417–20.
WB Saunders, 1996.
Textor SC, Wilcox CS. Renal artery stenosis: a common treatable
cause of renal failure? Annu Rev Med 2001; 52: 421–42.
 27
Renal Artery Aneurysms

JAMES C STANLEY, PETER K HENKE

The problem 315 Dissecting renal artery aneurysms 319

True renal artery aneurysms 315 References 322

THE PROBLEM aneurysm rupture is fatal in 10 per cent of cases, reflecting

the seriousness of this complication. This mortality rate is
less than suggested in earlier reports.4,8,16,17 Nevertheless,
Aneurysms of the renal artery are an uncommon vascular loss of the involved kidney continues to be a nearly univer-
disease, the clinical importance of which is a matter of con- sal outcome of renal artery aneurysm rupture.12,17
troversy.1–13 The emergency surgical treatment of renal In the largest reported experience with renal artery
artery aneurysms represents an ill-defined practice, with aneurysms, overt extraparenchymal rupture (Figs 27.1 and
most reported experiences being anecdotal. The two most 27.2) occurred in 1.8 per cent of patients harbouring these
common complications leading to emergency operations lesions, and covert rupture (Fig. 27.3) caused renal arteri-
are aneurysm rupture and thrombosis with peripheral ovenous fistulae in an additional 1.2 per cent.5 The com-
embolism. Occurrences of these complications are often bined 3 per cent frequency of overt and covert rupture in
overestimated in that most published reports describe the latter review is higher than that reported in non-surgical
operative rather than population-based experiences. These series.7,13 The rate of rupture of renal artery aneurysms
emergency scenarios require careful clinical judgement during pregnancy is unknown, but the sequelae are often
and skilled surgical intervention if optimal care is to be catastrophic, causing foetal death in nearly 85 per cent of
achieved. The two most relevant renal artery macro- cases and maternal death in over half of them.18–21
aneurysms, namely, true aneurysms and those associated Increased risks of renal artery aneurysm rupture have
with dissections, deserve individual discussion. been attributed to large size, absence of calcification and
elevated blood pressure, but these factors are not always
TRUE RENAL ARTERY ANEURYSMS relevant. In fact, overt rupture often occurs in normoten-
sive patients as well as in patients with calcific aneurysms.12
When rupture does occur it usually involves the base of the
The prevalence of true renal artery aneurysms in the gen- aneurysm and not the aneurysmal dome as might be
eral population approaches 0.09 per cent.12 The group of expected. A large aneurysm, in contrast to a small one, has
patients being studied bears greatly on the reported fre- an inconsistent but relatively greater potential for rupture.
quency of these lesions. Macroaneurysms were identified Aneurysm rupture during pregnancy does not appear to be
in 0.7 per cent of arteriographic studies performed in related to age or number of prior pregnancies.18
patients suspected of renal disease,14 2.5 per cent of those Overt renal artery aneurysm rupture usually causes
studies having been undertaken for suspected renovascular unremitting costovertebral, flank and abdominal pain of
hypertension,12 and 9.2 per cent of studies performed in varying intensity. In addition to pain, perirenal bleeding
hypertensive adults with renal artery fibrodysplasia.15 into the retroperitoneal space is usually associated with an
ileus and abdominal distension. Nausea and vomiting are
Clinical manifestations common in this setting. Most patients develop micro-
haematuria in association with rupture. Loss of retroperi-
Rupture is the most serious complication attending renal toneal containment with free bleeding into the peritoneal
artery aneurysms. Exsanguinating haemorrhage with cavity may result in life-threatening shock.
316 Renal artery aneurysms

(a)
(a)

(b)

Figure 27.1 Overt renal artery aneurysm rupture. (b)

(a) Aortographic demonstration of renal artery aneurysm
rupture with no evidence of distal parenchymal vessels. (b) Later Figure 27.3 Covert renal artery aneurysm rupture. (a) Arterial
peripelvic collection of contrast medium (from Stanley JC, phase digital subtraction arteriography demonstrating large
Whitehouse WM Jr. Renal artery macroaneurysms. In: Bergan JJ, inferior pole arterial aneurysm communication with an adjacent
Yao JST (eds). Aneurysms. Orlando, FL: WB Saunders, 1982: vein. (b) Venous phase demonstrating rapid filling of the inferior
417–31) vena cava with contrast (from Henke PK, Cardneau JD, Welling TH
III, et al. Renal artery aneurysms: A 35-year clinical experience
with 252 aneurysms in 168 patients. Ann Surg 2001; 234: 454–63)

A renal artery aneurysm may rupture covertly into an

adjacent renal vein producing an arteriovenous fistula (see
Fig. 27.3). These patients may initially experience vague
flank pain, but more often than not they are asymptomatic.
As these fistulae expand in size, nearly half of them become
associated with hypertension and chronic microhaema-
turia. It is uncommon for an arteriovenous fistula of the
renal vessels to present as a surgical emergency.
Thromboembolism of dislodged material originating in
an aneurysm results in renal ischaemia and infarction.
Atheromatous plaque rupture and ulceration within large
aneurysmal sacs may predispose to these embolic compli-
cations (Fig. 27.4). Small aneurysms without calcific athero-
sclerosis may also be the source of embolic renal arterial
Figure 27.2 Gross specimen of ruptured renal artery aneurysm occlusion (Fig. 27.5). In a study reported from our institu-
depicted in arteriographic study in Fig. 27.1 (from Stanley JC, tion, embolism was a clear complication of renal artery
Whitehouse WM Jr. Renal artery macroaneurysms. In: Bergan JJ, aneurysms in only three of 118 patients.12 This complica-
Yao JST (eds). Aneurysms. Orlando, FL: WB Saunders, 1982: tion, manifested by hypertension alone, is not considered a
417–31) surgical emergency.
 True renal artery aneurysms 317

Figure 27.4 (a) Large

arteriosclerotic renal artery
aneurysm containing thrombus.
(b) Deep cortical infarct (arrow)
secondary to thromboembolism
from the hilar aneurysm (from
Stanley JC, Whitehouse WM Jr.
Renal artery macroaneurysms.
In: Bergan JJ, Yao JST (eds).
(a) (b) Aneurysms. Orlando, FL: WB
Saunders, 1982: 417–31)

mediated provoked by profoundly ischaemic renal tissue.

Eventually, with scarring and the loss of the affected kidney
parenchyma, the blood pressure usually returns to normal.
A small number of individuals will have such severe dis-
comfort with renal infarction that urgent operative inter-
vention will be required. Similarly, in rare instances, the
infarcted segment will rupture into the renal pelvis causing
massive haematuria. These patients require emergency
nephrectomy.
Renal infarction usually results in a rise in the concentra-
tions of a number of serum enzymes. Although somewhat
non-specific, these elevations may be useful in supporting a
diagnosis of renal infarction. Immediate increases in alanine
aminotransferase are often observed followed 1 or 2 days
later by peak levels of lactose dehydrogenase. Later, 3–5 days
after infarction, peak elevations in alkaline phosphatase will
Figure 27.5 Small non-atherosclerotic intraparenchymal be noted.
aneurysm associated with segmental thromboembolic renal
Compression of a renal artery aneurysm or kinking of an
ischaemia and cortical infarct (from Stanley JC, Whitehouse WM
Jr. Renal artery macroaneurysms. In: Bergan JJ, Yao JST (eds). adjacent segment causing renal ischaemia and consequent
Aneurysms. Orlando, FL: WB Saunders, 1982: 417–31) severe uncontrolled renovascular hypertension has been
documented in the literature.22,23 Unless the patient develops
malignant hypertension, which is unlikely, this complica-
tion is not considered a surgical emergency. A patient pre-
Aneurysm related thromboembolism may present with
senting with marked blood pressure elevations usually has
vague flank discomfort caused by an inflammatory peri-
coexisting renal artery stenotic disease in the vicinity of the
nephritis in the case of a small renal infarct, or with severe
aneurysm accounting for the secondary hypertension.
back, flank and abdominal pain following a large renal
infarct. This is an uncommon but recognised presenta-
tion as a surgical emergency. The patient is usually febrile, Aetiology
exhibits a leucocytosis and develops haematuria and
albuminuria. Severe hypertension, lasting for a few days, Renal artery aneurysms usually arise as a result of a con-
may accompany acute renal infarction. It is often attrib- genital elastic tissue defect or medial degenerative process.
uted to pain experienced, but is more likely to be renin Fragmentation of the internal elastic lamina and in the
318 Renal artery aneurysms

media a paucity of elastic tissue and loss of recognisable kidney with a revascularisation procedure is intended.
smooth muscle are characteristic features of these aneurysms. Gadolinium-enhanced magnetic resonance arteriography24
The well-known discontinuity of internal elastic lamina at and three-dimensional reconstructed computed tomog-
the bifurcation of all muscular arteries further comprom- raphy (CT) scanning25 have a potential but unproved role in
ises the structural integrity of the renal artery and, together the anatomical delineation of these aneurysms.
with the elevated blood pressure observed in 40–80 Intraoperative diagnosis of a ruptured aneurysm should
per cent of these patients, contributes to the development be entertained when a large retroperitoneal haematoma
of an aneurysm. surrounding the kidney is noted during an emergency
Complicated atherosclerotic manifestations such as cal- operation for pain or suspected haemorrhage. In this set-
cium deposition, collections of cholesterol, necrotic debris, ting, the exact diagnosis is usually evident upon examin-
haemorrhage and a matrix of fibrous tissue are often pre- ation of the excised kidney. On rare occasions, lesser degrees
sent in the walls of larger aneurysms. These changes are of bleeding or thromboembolism may facilitate dissection
considered part of a secondary rather than a primary aeti- and exposure of the aneurysm followed by revascularisa-
ological event. The observation that atherosclerotic changes tion of the kidney.
are present in some, but not all, aneurysms in a patient
with multiple aneurysms, supports the tenet that a non- Management
atherosclerotic cause is responsible for most renal artery
aneurysms.12 Nevertheless, secondary atherosclerosis con- The objective of surgical therapy is to eliminate the renal
tributes to further vessel wall weakness. artery aneurysm without removing the kidney or comprom-
ising its function.5,12,26–29 This objective is rarely met in
emergency procedures. Nephrectomy is the usual outcome
True renal artery aneurysms: presentation in managing most ruptured aneurysms or extensive
and manifestations thromboembolic complications of these aneurysms. As far
as overall health of the patient is concerned, nephrectomy
• Rare – less than 1 in 1000 individuals
may be well tolerated. In a recent report on the manage-
• Aneurysms are usually discovered incidentally, often
ment of renal artery aneurysms with a follow-up of nearly
in association with hypertension
a decade, planned nephrectomy, and in a few instances
• Risk of aneurysm rupture – not solely dependent on
unplanned nephrectomy, did not result in renal failure.5
size, significantly higher during pregnancy
Nevertheless, arterial reconstruction should be considered
• Renal infarction may result from thromboembolism
whenever the kidney does not appear to have been
irreparably injured from ischaemia.
In attempting to salvage a kidney after an aneurysm has
Diagnosis ruptured, it is important to recognise that renal function is
impaired after blood flow has been interrupted for 40 min-
The clinical manifestations of renal artery aneurysm rup- utes. After 60 minutes of warm ischaemia, retrieval of renal
ture or acute thromboembolism are quite protean. The function becomes unlikely. If prolonged renal ischaemia is
diagnosis is usually not immediately clear, the differential anticipated during a reconstructive procedure, renal hypo-
diagnoses including other vascular emergencies such as thermia using cold (4 °C) hypertonic electrolyte solution
ruptured or expanding aortic or splanchnic artery aneurysms. infusions should be undertaken to protect the kidney.
Acute renal colic with the passage of kidney stones may Longer periods of interruption of renal blood flow may be
also mimic the complications of certain aneurysms. Inflam- tolerated in those patients with pre-existing stenotic dis-
matory processes such as diverticulitis with abscess forma- ease and the presence of collateral vessels to the kidney.
tion, pancreatitis and acute biliary tract diseases may lead Partial nephrectomy may be possible when aneurysmal
to urgent surgical intervention at which point a ruptured erosion has occurred into an adjacent vein causing an arte-
renal artery aneurysm may be discovered. riovenous fistula or when thromboembolism has caused a
Imaging studies in some of the settings noted above limited area of segmental infarctions. Acute arteriovenous
may reveal renal artery aneurysm calcification, and distor- fistulae may be treated occasionally by endovascular
tion of the kidney substance and collecting system by the means, namely, with transcatheter instillation of absolute
aneurysm or a haematoma if rupture has occurred. If a alcohol to eliminate diseased tissue or embolisation of par-
renal artery aneurysm is suspected in a haemodynamically ticulate matter or coils to obliterate the fistula, selectively
stable patient based on the history, physical examination or infarcting small areas of the kidney.
non-specific imaging studies, then arteriography becomes Most renal artery aneurysms are best approached via a
an essential diagnostic study.5,12 transabdominal, extraperitoneal exposure of the renal vas-
Arteriographic documentation of renal arterial anatomy culature, displacing the overlying colon and foregut viscera
in the region of an aneurysm is especially necessary in medially. In the situation of active bleeding from a rup-
planning emergency operative intervention if salvage of the tured aneurysm, the renal artery is compressed or clamped
 Dissecting renal artery aneurysms 319

proximally near its aortic origin before attempting to dis- DISSECTING RENAL ARTERY ANEURYSMS
sect distal tissue around the aneurysm. The patient’s over-
all haemodynamic state and estimated renal ischaemia
time will dictate whether a simple nephrectomy is per- Isolated renal artery dissection causing an aneurysm
formed or an arterial repair and renal salvage is attempted. is rare.37–40 Dissections are usually classified into two
Large aneurysms affecting the main renal artery bifurca- types: the first type is due to blunt abdominal trauma or
tion can usually be excised with a simple angioplastic closure. intraluminal catheter induced injury and the second
Excision of smaller aneurysms may require arterial closure occurs spontaneously.
with a vein patch. More extensive renal artery reconstruc-
tions using autogenous saphenous vein or internal iliac
artery as aortorenal grafts are favoured for those aneurysms Clinical manifestations
associated with functionally important stenoses of the renal
artery.5,12,29 Aneurysmectomy with reimplantation of the Flank and back pain, haematuria, ileus and hypertension
involved vessel or vessels into a normal adjacent or proxi- frequently accompany acute dissections regardless of the
mal renal artery is appropriate for treating many first and cause.38,41,42 Acute dissections often present as emergencies
second order branch aneurysms. These procedures are usu- with excruciating pain, associated nausea and vomiting
ally undertaken in situ, although ex vivo reconstructions suggestive of an acute abdomen. It is uncommon for renal
may be preferred in certain cases,30–32 especially with coex- artery dissection to result in vascular disruption and
istent segmental renal artery stenotic disease. These arterial uncontrolled haemorrhage. Chronic renal artery dissec-
reconstructive procedures are often lengthy and are usually tion, when clinically relevant, is usually associated with
only undertaken for treatment of symptomatic intact renovascular hypertension or impaired renal function but
aneurysms. Extensive tissue disruption and blood staining it does not immediately threaten life or the kidney con-
associated with aneurysm rupture may preclude completion cerned, very rarely presenting as a surgical emergency.
of a complex renal artery reconstruction.
Renal artery aneurysms are not usually amenable to
endovascular intervention and are even less so when they Aetiology
present as emergencies. Bleeding aneurysms may be an
exception. Certainly if life-threatening haemorrhage is evi- Dissection of the renal artery affects men nearly 10 times
dent at angiography, coil or balloon occlusion of the renal more often than women,43 indeed men have a greater like-
artery may be quite acceptable and appropriate. In anec- lihood of developing trauma induced dissections.
dotal reports, aneurysms of the main renal artery in a Although an overall predilection for right renal artery
non-emergency setting have been successfully excluded by involvement exists, trauma related dissection more com-
stent graft placement and branch aneurysms successfully monly affects the left renal artery.
embolised.33–36 Embolisation of intact intraparenchymal Blunt trauma contributes to renal artery dissection by
aneurysms is a reasonable alternative to partial nephrec- two specific mechanisms. The first mechanism is violent
tomy in selected symptomatic patients. Endovascular displacement of the kidney, causing the renal artery to
treatment of renal artery aneurysms may become more stretch, fracturing the intima and resulting in a subintimal
common as the technology improves in the future. dissection; this happens most frequently in deceleration
injuries (see Chapters 26 and 36). The second relates to
traumatic compression of the renal artery against the ver-
True renal artery aneurysms: diagnosis and tebra, causing haemorrhage within the deeper media, false
management aneurysm formation and vessel wall disruption. Both
forms of trauma are most commonly associated with
motor vehicle accidents.
• Non-specific abdominal complaints uncommon
Iatrogenic catheter related injury during diagnostic
with intact aneurysms, most are asymptomatic
arteriography is an uncommon cause of renal artery dissec-
• Ruptured aneurysms are associated with severe flank
tion (Fig. 27.6). In an earlier series from our institution,
pain, ileus and haemodynamic instability
only four catheter related renal artery dissections were
• Confirmed rupture in unstable patients justifies
encountered in more than 11 000 abdominal diagnostic
emergency operation, usually nephrectomy
arteriographic examinations, including more than 2200
• Suspected rupture in haemodynamically stable
selective renal arteriograms.38 Iatrogenic dissections of this
patients warrants urgent arteriography and an
operative attempt to salvage the kidney type usually occur within the inner media or subintimal
tissues of the renal artery. Dissections accompanying thera-
• Consider endovascular embolisation for rupture of a
peutic balloon angioplasty are very common, although
segmental renal artery branch aneurysm in
haemodynamically stable patients only a few cause critical narrowing or occlusion of the renal
artery.44 In those instances of critical stenoses, stent placement
320 Renal artery aneurysms

(a)

Figure 27.7 Spontaneous saccular dissecting renal artery

aneurysm in a patient with coexisting fibrodysplasia (from
Gewertz BL, Stanley JC, Fry WJ. Renal artery dissections. Arch Surg
1977; 112: 409–14)

(b)

Figure 27.6 (a) Catheter induced dissecting renal artery

aneurysm. (b) Postoperative appearance following
aneurysmectomy and arterial reconstruction with an
aortorenal bypass (from Gewertz BL, Stanley JC, Fry WJ. Figure 27.8 Dissection exhibiting deep mural haematoma and
Renal artery dissections. Arch Surg 1977; 112: 409–14) compression of adjacent lumen (haematoxylin eosin stain, original
magnification 60) (from Stanley JC. Pathologic basis of
macrovascular renal artery disease. In: Stanley JC, Ernst CB, Fry WJ
at the moment when the dissection is recognised will usu- (eds). Renovascular Hypertension. Philadelphia, PA: WB Saunders,
ally restore the renal artery lumen.45 1984: 46–74)
Spontaneous dissection causing a pseudoaneurysm
affects the renal arteries more than any other peripheral
artery. Most of these lesions are associated with coexisting Diagnosis
atherosclerotic or fibrodysplastic renovascular disease (Fig.
27.7). These dissections usually occur within the outer A correct initial clinical diagnosis of a renal artery dissec-
media adjacent to the external elastic lamina (Fig. 27.8). tion is uncommon, occurring in less than half of these
Spontaneous renal artery dissection most often affects the patients.42 Intravenous pyelography has been advocated in
proximal vessel and terminates at its branching. evaluating serious renal hilar injuries, including dissecting
 Dissecting renal artery aneurysms 321

aneurysms, but in view of the high incidence of false nega- Local angioplastic procedures, often undertaken in the
tive and false positive studies, such examination should be treatment of true renal artery aneurysms, are inappropriate
deferred in favour of arteriography. in the treatment of dissections, regardless of the cause. In
Arteriography is necessary to diagnose as well as define this situation the affected arterial segment usually demands
the extent of a renal artery dissection and is essential in plan- replacement or bypass with a graft. Renal artery dissections
ning operative therapy. The features of dissection include: with preserved renal blood flow allow time for a repair to
be planned and executed in a semi-elective fashion. Stand-
• luminal irregularities with fusiform aneurysmal
ard arterial reconstruction in the form of an aortorenal
dilatation or saccular outpouchings associated with
bypass using autogenous saphenous vein or hypogastric
segmental stenoses
artery (see Fig. 27.6), and ex vivo repairs in selected cases,
• extension of the dissection to the first renal artery
provide reasonable kidney salvage rates.38,39
branching
• cuffing at branchings
• variable degrees of reversibility documented on serial
Conclusions
studies.
In the case of renal artery occlusion with infarction of A patient with a renal artery aneurysm, regardless of
the kidney, increases occur in the levels of the same array aetiology, may present as an emergency, if the aneurysm
of enzymes previously described when true aneurysms are ruptures causing severe haemorrhage and shock or if
complicated by thromboembolism. thromboembolism results in segmental infarction of the
kidney. In both circumstances, and given the commoner
Dissecting renal artery aneurysms: acute abdominal conditions which mimic those two sce-
presentation and diagnosis narios, the clinical diagnosis may not become immediately
apparent. Arteriography is the investigation of choice in
• Often present with severe abdominal and back pain establishing the diagnosis. Ideally, the objectives of sur-
and nausea gical treatment are to eliminate the aneurysm, reconstruct
• Blunt trauma is a more common cause than the artery and salvage the kidney. In reality, nephrectomy
spontaneous dissection and a more common cause is often the outcome, as it is when endovascular tech-
than catheter related dissecting aneurysm niques are used to occlude the renal artery as a life-saving
• Arteriography will reveal the diagnosis measure. A patient with acute dissection of the renal
artery, and the possible (pseudo)aneurysms complica-
ting it, whether caused by blunt trauma, balloon angio-
plasty or if it occurs spontaneously, is also likely to be
Management hypertensive. In such a case, operative replacement or
bypass of the renal artery should remedy the problem
Emergency arterial reconstruction of trauma induced dis- while also preserving the kidney.
section is vital in haemodynamically significant narrowing
of the main renal artery or a major segmental branch.38,39
Spontaneous dissecting aneurysms, when acute, are tech-
nically easier to treat than traumatic lesions and once diag-
nosed most should be dealt with urgently.
Operative intervention is indicated in chronic trauma Key references
related and spontaneous dissections associated with severe
Cohen JR, Shamash FS. Ruptured renal artery aneurysms during
renovascular hypertension, and in some instances deteri-
pregnancy. J Vasc Surg 1987; 6: 51–9. Comprehensive review
orating renal function (see Chapter 26); these latter cir-
of renal artery aneurysm rupture during pregnancy.
cumstances, however, do not often constitute surgical Henke PK, Cardneau JD, Welling TH III, et al. Renal artery
emergencies. Endovascular stent graft placement may be aneurysms: a 35-year clinical experience with 252 aneurysms
an appropriate alternative for short proximal renal artery in 168 patients. Ann Surg 2001; 234: 454–63. Largest
dissections with a defined distal endpoint, but there is little reported series of renal artery aneurysms, from a surgical
in the literature pertaining to the long term durability of perspective.
such treatment. Reilly LM, Cuningham CG, Maggisano R, et al. The role of arterial
Kidney preservation must be the paramount concern reconstruction in spontaneous renal artery dissection. J Vasc
when treating renal artery dissection, particularly as con- Surg 1991; 14: 468–77. Comprehensive review of non-
tralateral renal artery dissection occurs in a third of cases, iatrogenic, spontaneous renal artery dissections.
Schorn B, Falk V, Dalichau H, Mohr FW. Kidney salvage in a case of
and in addition the contralateral kidney may be diseased in
ruptured renal artery aneurysm: case report and literature
half of the patients sustaining blunt abdominal trauma.38
review. Cardiovasc Surg 1997; 5: 134–6. Brief review of
Nephrectomy, under such circumstances, should be avoided.
322 Renal artery aneurysms

patient survival and kidney salvage after renal artery aneurysm 19 Lacroix H, Bernaerts P, Nevelsteen A, Hanssens M. Ruptured renal
rupture. artery aneurysm during pregnancy: Successful ex situ repair and
Stanley JC, Messina LM, Wakefield TW, Zelenock GB. Renal artery autotransplantation. J Vasc Surg 2001; 33: 188–90.
reconstruction. In: Bergan JJ, Yao JST (eds). Techniques in 20 Love WK, Robinette MA, Vernon CP. Renal artery aneurysm
Arterial Surgery. Philadelphia, PA: WB Saunders, 1990: 247–63. rupture in pregnancy. J Urol 1981; 126: 809–11.
Depiction of various techniques available for the surgical 21 Rijbroek A, van Dijk HA, Roex AJM. Rupture of renal artery
treatment of renal artery aneurysms. aneurysm during pregnancy. Eur J Vasc Surg 1994; 8: 375–6.
22 Reiher L, Grabitz K, Sandmann W. Reconstruction for renal artery
aneurysm and its effect on hypertension. Eur J Endovasc Surg
REFERENCES 2000;20;454–6.
23 Youkey JR, Collins GJ, Orecchia PM, et al. Saccular renal artery
aneurysm as a cause of hypertension. Surgery 1985; 97: 498–501.
1 Bastounis W, Pikoulis E, Georgopoulos S, et al. Surgery for renal 24 Prince MR, Narasimham DL, Stanley JC, et al. Breath-hold
artery aneurysms: A combined series of two large centers. Eur gadolinium-enhanced MR angiography of the abdominal aorta
Urol 1998; 33: 22–7. and its major branches. Radiology 1995; 197: 785–92.
2 Bulbul MA, Farrow GA. Renal artery aneurysms. Urology 1992; 25 Cikrit DF, Harris VJ, Hemmer CG, et al. Comparison of spiral CT
40: 124–6. scan and arteriography for evaluation of renal and visceral
3 Dzsinich C, Gloviczki P, McKusick MA, et al. Surgical arteries. Ann Vasc Surg 1996; 10: 109–16.
management of renal artery aneurysm. Cardiovasc Surg 1993; 26 Forbes TL, Abraham CZ, Pudupakkam S. Repair of ruptured giant
3: 243–7. renal artery aneurysm with kidney salvage. Eur J Vasc Endovasc
4 Hageman JH, Smith RF, Szilagyi DE, Elliott JP. Aneurysms of the Surg 2001; 22: 278–9.
renal artery: problems of prognosis and surgical management. 27 Hupp, T, Allenberg JR, Post K, et al. Renal artery aneurysms:
Surgery 1978; 84: 563–72. surgical indications and results. Eur J Vasc Surg 1992; 6: 477–86.
5 Henke PK, Cardneau JD, Welling TH III, et al. Renal artery 28 Mercier C, Piquet P, Piligian F, Ferdani M. Aneurysms of the renal
aneurysms: A 35-year clinical experience with 252 aneurysms in artery and its branches. Ann Vasc Surg 1986; 1: 321–7.
168 patients. Ann Surg 2001; 234: 454–63. 29 Stanley JC, Messina LM, Wakefield TW, Zelenock GB. Renal artery
6 Henriksson C, Bjorkerud S, Nilson AE, Pettersson S. Natural reconstruction. In: Bergan JJ, Yao JST (eds). Techniques in Arterial
history of renal artery aneurysm elucidated by repeated Surgery. Philadelphia, PA: WB Saunders, 1990:247–63.
angiography and pathoanatomical studies. Eur Urol 1985; 11: 30 Belzer FO, Raczkowski A. Ex vivo renal artery reconstruction with
244–8. autotransplantation. Surgery 1982; 92: 642–5.
7 Henriksson C, Lukes P, Nilson AE, Pettersson S. Angiographically 31 Bugge-Asperheim B, Sdal G, Flatmark A. Renal artery aneurysm:
discovered, non-operated renal artery aneurysms. Scand J Urol ex vivo repair and autotransplantation. Scand J Urol Nephrol
Nephrol 1984; 18: 59–62. 1984; 18: 63–6.
8 Hubert JP Jr, Pairolero PC, Kazmier FJ. Solitary renal artery 32 Dubernard JM, Martin X, Gelet A, Mongin D. Aneurysms of the
aneurysm. Surgery 1980; 88: 557–65. renal artery: surgical management with special reference to
9 Lumsden AB, Salam TA, Walton KG. Renal artery aneurysm: extracorporeal surgery and autotransplantation. Eur Urol 1985;
a report of 28 cases. Cardiovasc Surg 1996; 4: 185–9. 11: 26–30.
10 Martin RS III, Meacham PW, Ditesheim JA, et al. Renal artery 33 Bui BT, Oliva VL, Leclerc G, et al. Renal artery aneurysm:
aneurysm: selective treatment for hypertension and prevention Treatment with percutaneous placement of a stent-graft.
of rupture. J Vasc Surg 1989; 9: 26–34. Radiology 1995; 195: 181–2.
11 Soussou ID, Starr DS, Lawrie GM, Morris GC. Renal artery 34 Centenera LV, Hirsch JA, Choi IS, et al. Wide-necked saccular
aneurysm: long-term relief of renovascular hypertension by renal artery aneurysm: endovascular embolization with the
in situ operative correction. Arch Surg 1979; 114: 1410–15. Guglielmi detachable coil and temporary balloon occlusion of the
12 Stanley JC, Rhodes EL, Gewertz BL, et al. Renal artery aneurysms: aneurysm neck. J Vasc Interv Radiol 1998; 9: 513–516.
significance of macroaneurysms exclusive of dissections and 35 Karkos CD, D’Souza SP, Thompson GJ, et al. Renal artery
fibrodysplastic mural dilations. Arch Surg 1975; 110: aneurysm: endovascular treatment by coil embolization with
1327–333. preservation of renal blood flow. Eur J Vasc Endovasc Surg 2000;
13 Tham G, Ekelund L, Herrlin K, et al. Renal artery aneurysms: 19: 214–16.
natural history and prognosis. Ann Surg 1983; 197: 348–52. 36 Tateno T, Kubota Y, Sasagawa I, et al. Successful embolization of
14 Edsman G. Angiography and suprarenal angiography. Acta Radiol a renal artery aneurysm with preservation of renal blood flow. Int
1965; Suppl 155: 104. Urol Nephrol 1996; 28: 283–7.
15 Stanley JC, Gewertz BL, Bove EL, et al. Arterial fibrodysplasia: 37 Edwards BS, Stanson AW, Holley KE, Sheps SG. Isolated renal
histopathologic character and current etiologic concepts. Arch artery dissection: presentation, evaluation, management and
Surg 1975; 110: 561–6. pathology. Mayo Clin Proc 1982; 57: 564–71.
16 Hidai H, Kinoshita Y, Murayama T, et al. Rupture of renal artery 38 Gewertz BL, Stanley JC, Fry WJ. Renal artery dissections. Arch
aneurysm. Eur Urol 1985; 11: 249–53. Surg 1977; 112:409–14.
17 Schorn B, Falk V, Dalichau H, Mohr FW. Kidney salvage in a case 39 Reilly LM, Cuningham CG, Maggisano R, et al. The role of arterial
of ruptured renal artery aneurysm: case report and literature reconstruction in spontaneous renal artery dissection. J Vasc
review. Cardiovasc Surg 1997; 5: 134–6. Surg 1991; 14: 468–77.
18 Cohen JR, Shamash FS. Ruptured renal artery aneurysms during 40 Smith BM, Holcomb GW 3rd, Richie RE, Dean RH. Renal artery
pregnancy. J Vasc Surg 1987; 6: 51–9. dissection. Ann Surg 1984; 200: 134–46.
 References 323

41 Hare WS, Kincaid-Smith P. Dissecting aneurysm of the renal 44 Stanley JC. Surgery of failed percutaneous transluminal renal
artery. Radiology 1970; 97: 255–63. artery angioplasty. In: Bergan JJ, Yao JST (eds). Reoperative
42 Rao CN, Blaivas JG. Primary renal artery dissecting aneurysm: a Arterial Surgery. Orlando, FL: Grune & Stratton, 1986: 441–54.
review. J Urol 1977; 118: 716–19. 45 Mali WP, Geyskes GG, Thalman R. Dissecting renal artery
43 Bakir AA, Patel K, Schwartz MM, Lewis EJ. Isolated dissecting aneurysm: Treatment with an endovascular stent. Am J Radiol
aneurysm of the renal artery. Am Heart J 1978; 96: 92–6. 1989; 153: 623–4.
This page intentionally left blank
 28
Visceral Artery Aneurysms

SANDRA C CARR, WILLIAM D TURNIPSEED

The problem 325 Treatment options: surgical and endovascular 327

Presenting symptoms and signs 325 Management strategies and recommendations 330
Diagnostic studies 326 References 332

THE PROBLEM liver trauma is the most common cause of intrahepatic

pseudoaneurysms. True aneurysms of the splenic artery are
found in multiparous women and in patients with portal
Aneurysms of the visceral arteries (VAAs) represent an hypertension and splenomegaly. Mycotic aneurysms most
uncommon but potentially lethal form of vascular disease. commonly occur in the coeliac and superior mesenteric
These lesions are rare, with an incidence of 0.01–0.2 per cent arteries and are associated with subacute bacterial endo-
in routine autopsies.1 The most commonly involved vessels carditis, Salmonella infection or syphilis. Patients with
include the splenic, hepatic, superior mesenteric (SMA) and fibromuscular dysplasia, connective tissue disorders such as
coeliac arteries. The gastric-gastroepiploic, jejunal-ilieal-colic, Ehlers–Danlos syndrome, and inflammatory conditions
pancreaticoduodenal-pancreatic, gastroduodenal and infe- such as polyarteritis nodosa, systemic lupus erythematosus,
rior mesenteric arteries are less often involved. Behçet’s disease and Takayasu’s disease are also at increased
In the past, most VAAs were discovered at autopsy, rup- risk of developing visceral aneurysms.
ture often being the cause of death. With the widespread The clinical findings associated with VAAs vary depend-
use of computed tomography (CT) and angiography many ing upon the artery involved. Splenic artery aneurysms
asymptomatic aneurysms are now discovered. Despite this (SAAs) are more common in women (78 per cent) than in
many of these patients still present acutely with rupture. men (21 per cent)3,4 (Fig. 28.1). More than 90 per cent of
Recent literature documents mortality ranging from 21 SAAs are asymptomatic having been discovered in plain
per cent to nearly 100 per cent,2 depending upon the ves- films, CT scan or magnetic resonance imaging (MRI)
sel involved. As these aneurysms frequently present as life- performed for other reasons.4 Multiple SAAs are found in
threatening clinical emergencies, it is essential that vascular 5 per cent of patients.4 Only 4.5 per cent present with rup-
surgeons know how to diagnose and treat VAAs. ture. Patients taking -blockade medication may be at
decreased risk of rupture.4 The most common symptoms
are abdominal pain, epigastric or left upper quadrant or
PRESENTING SYMPTOMS AND SIGNS back pain.2 Patients with ruptured aneurysms may present
with shock or gastrointestinal haemorrhage. Bleeding may
Making the correct diagnosis requires an appropriate be confined initially to the lesser sac but continued haem-
index of suspicion and knowledge of various associated con- orrhage therefrom through the foramen of Winslow into
ditions. Visceral pseudoaneurysms result from inflamma- the peritoneal cavity results in haemorrhagic shock; this is
tion and/or trauma to the artery. Splenic aneurysms are known as the ‘double rupture phenomenon’.5
commonly associated with pancreatitis and/or pseudocyst. Unlike SAAs, those of the hepatic artery (HAAs) occur
Similarly, aneurysms of the pancreaticoduodenal, pancreatic, more frequently in men.2 The routine use of abdominal CT
gastroduodenal, gastric and gastroepiploic arteries often scan for blunt liver trauma and the increasing numbers of
result from pancreatic inflammation. Blunt or penetrating percutaneous biliary procedures has led to an increase in
326 Visceral artery aneurysms

occur into the peritoneal cavity, retroperitoneum, common

bile duct, gallbladder, duodenum or portal vein. Erosion into
the bile duct may cause haemobilia, and erosion into the
duodenum will result in gastrointestinal haemorrhage.5
About half of the symptomatic patients will have abdominal
pain and half will present with gastrointestinal haemorrhage
or haemobilia. Biliary colic and jaundice may occur in up to
10 per cent of patients because of extrinsic compression of
the bile duct by the aneurysm.2 The ‘classic triad’ of pain,
haemobilia and obstructive jaundice is seen in less than 30
per cent of patients. Less commonly, patients may present
with a right upper quadrant mass.
The majority of patients with SMA or coeliac artery
aneurysms are symptomatic at the time of presentation,
38 per cent of them having ruptured aneurysms with a
resulting mortality of 30–40 per cent.2,8 As with SAAs, there
may be a decreased risk of rupture of SMA aneurysms in
patients taking -blockade medication.4 Most patients have
abdominal pain which may be postprandial and mimic the
symptoms of chronic mesenteric occlusive disease. Other
Figure 28.1 Multiple splenic artery aneurysms in a 52-year-old
signs and symptoms include a palpable abdominal mass,
multiparous woman. This patient was treated with splenectomy
nausea, vomiting, jaundice or anaemia.3 Rupture may result
and resection of the aneurysms
in gastrointestinal haemorrhage, haematemesis, shock
or bowel infarction. Approximately 5 per cent of SMA
aneurysms are mycotic, associated with systemic infection
or subacute bacterial endocarditis.4
Aneurysms of the gastroduodenal, pancreatoduodenal
and pancreatic arteries are often associated with pancreatitis
and its complications. Up to 68 per cent of these aneurysms
present with rupture and mortality as high as 50 per cent.3
More than 80 per cent of these aneurysms are symptomatic
causing abdominal pain at the time of diagnosis. Rupture at
the time of diagnosis has been observed in 56 per cent of
gastroduodenal aneurysms and 68 per cent of pancreatoduo-
denal aneurysms. Aneurysms may erode into an adjacent
portion of the gastrointestinal tract, causing gastrointestinal
haemorrhage or may rupture into the peritoneal cavity. Gas-
trointestinal haemorrhage or haemobilia occurs in approxi-
mately 50 per cent of patients. Jaundice occurs in an
additional 14–31 per cent.2,9
Gastric and gastroepiploic aneurysms are rare, comprising
only 4 per cent of all VAAs. The vast majority (90 per cent)
of these aneurysms present as emergencies with rupture
and gastrointestinal or intraperitoneal haemorrhage.
Figure 28.2 Multiple intrahepatic artery aneurysms in a young
male patient following blunt liver trauma Visceral aneurysms may also occur in the jejunal, ileal,
colic or inferior mesenteric arteries. In a report by Tessier
the discovery of HAAs. Intrahepatic pseudoaneurysms may et al., rupture occurred in 25 per cent of the mesenteric
be associated with percutaneous biliary procedures, biopsy, branch artery aneurysms and all were in colic arteries.10
abdominal trauma or pancreatitis2 (Fig. 28.2). Aneurysms of These patients may present with abdominal pain or haem-
the hepatic artery have been discovered recently in patients orrhagic shock and a haematoma in the mesentery.10
who had previously undergone previous liver transplant-
ation.6 In a current series by Abbas et al., 78 per cent of the
HAAs were extrahepatic and 25 per cent of the HAAs were
DIAGNOSTIC STUDIES
symptomatic, including 14 per cent presenting with rupture.7
The most common symptoms are right upper quadrant pain, With the increasing use of CT and angiography, more
radiating to the back and unrelated to meals. Rupture may asymptomatic visceral artery aneurysms are being discovered.
 Treatment options: surgical and endovascular 327

Many VAAs contain calcification in the wall appearing as a and therefore should be treated urgently. Treatment is
curvilinear or signet ring shaped density on plain abdomi- also indicated in patients with symptomatic or enlarging
nal film or intravenous pyelogram (IVP). Characteristic aneurysms.5,14 Conversely, small asymptomatic aneurysms,
‘egg shell’ patterns of calcification may be seen in the epi- less than 2–3 cm in diameter, can be safely observed, espe-
gastrium or upper abdominal quadrants. cially in older women.4,15
Ultrasound or CT scans can be used to size these Operative treatment most frequently consists of splenec-
aneurysms accurately. Typical ultrasound findings include tomy and removal of that portion of the splenic artery con-
cystic or solid masses with sonolucent or mixed echo signals taining the aneurysm. Surgical exposure is obtained though
but ultrasound is very technician dependent. Contrast- the lesser sac where the splenic artery can be easily con-
enhanced CT is more appropriate in delineating and estab- trolled. Proximal aneurysms can be excised or ligated, often
lishing the patency of the artery from which the visceral with preservation of the spleen,4,15 whereas those in the
aneurysm arises. Non-contrast CT scans may reveal a low midportion of the vessel can be excluded with proximal and
attenuation mass with or without peripheral rim calcifica- distal ligation or reconstructed with an end-to-end anasto-
tion, haematomas around the porta hepatis, liver or in the mosis.16 Inflammatory aneurysms associated with pancrea-
retroperitoneal space. A haematoma in the lesser sac is titis are more difficult to treat, as the aneurysm may be
often associated with a ruptured SAA. Contrast-enhanced embedded within the pancreas. Pseudoaneurysms associ-
CT scans will increase diagnostic accuracy. Visceral artery ated with pancreatitis are most safely treated by direct
aneurysms will appear as a brightly enhanced lesion with a ligation from within the aneurysm sac. Distal SAAs may
variable amount of luminal thrombus. Spiral CT angiog- necessitate partial pancreatectomy along with resection of
raphy is an additional technique which can be of value in the aneurysm.15 Laparoscopic ligation of the splenic artery
identifying VAAs and their arteries or origin.8 proximal and distal to the aneurysm provides a minimally
Magnetic resonance imaging is another imaging modal- invasive treatment option.17
ity that can be used to evaluate visceral artery aneurysms. Many patients with SAAs have acute or chronic medical
Non-iodinated contrast agents such as gadolinium used in conditions, such as pancreatitis, which put them at high
MR angiography (MRA) are not nephrotoxic. Magnetic risk for open surgical intervention.18 Advances in guidewire
resonance angiography can define the vessels of origin and techniques and microvascular instruments have encour-
provide information on collateral flow. Despite advances aged the development of percutaneous embolisation as an
in spiral CT angiography and MRA, conventional digital alternative treatment of SAAs and other VAAs. Although
subtraction angiography is still the most commonly used most SAAs are treated surgically, transcatheter embolisa-
diagnostic test. Contrast angiography provides accurate tion (TCE) is becoming a more commonly used form of
preoperative information and can also be used for thera- treatment.18,19 Selective catheterisation of the splenic artery
peutic embolisation. or the aneurysm sac is followed by the introduction of
coils, Gelfoam particles, or detachable balloons into the
vessel. In the case of a saccular aneurysm, it is possible to
TREATMENT OPTIONS: SURGICAL AND fill the aneurysm sac with coils, also known as the ‘packing’
ENDOVASCULAR method, while maintaining flow within the splenic artery
and thus preserving the spleen. A possible complication of
this form of therapy is embolic ischaemia of the spleen,
Most visceral aneurysms, excluding coeliac and SMA lesions,
infarction and even abscess formation19 (Fig. 28.3). Other
can be treated with surgical ligation. Endovascular methods
complications include pain, fever, embolisation to other
of thromboembolisation are appropriate for treating
visceral arteries, incomplete occlusion and recanalisa-
selected visceral aneurysms.11
tion.20 The postembolisation syndrome, consisting of
abdominal pain, fever, slowed transit and elevation of pan-
Splenic artery aneurysms creatic enzymes, can occur in up to 30 per cent of cases and
generally resolves over 3–5 days.18 Staged TCE, inducing
The risk of rupture of SAAs, in cases not associated with arterial occlusion over a few days, may reduce the risk of
pregnancy, is low at only 2–5 per cent. However, the mor- splenic infarction. Recently, stent grafts, such as the
tality associated with rupture is high, being approximately Wallgraft endoprosthesis and the Jostent stent graft have
30–40 per cent.4 Splenic artery aneurysms can be especially been used to treat large visceral aneurysms and preserve
dangerous in the pregnant patient, the majority of which splenic flow.21
rupture during the last trimester. Rupture during pregnancy
is associated with a maternal mortality of 70 per cent and a
foetal death rate of 95 per cent.12,13 Therefore, women of Hepatic artery aneurysms
childbearing age, even with small aneurysms, should be
treated when diagnosed. Pseudoaneurysms associated with About 34 per cent of HAAs are intrahepatic, many of them
inflammation or trauma are fragile and likely to rupture, small and multiple, making surgical exposure difficult.
328 Visceral artery aneurysms

Figure 28.3 A splenic artery aneurysm was successfully

thrombosed with percutaneous transcatheter embolisation. The
patient developed ischaemic infarction of the spleen with abscess (a)
formation

(b)

Figure 28.5 (a) A 56-year-old man developed haemobilia

following common bile duct resection and placement of a
Figure 28.4 Recurrence of this hepatic artery aneurysm
transhepatic biliary catheter. Arteriography demonstrated a
occurred after embolisation of the right hepatic artery. Successful
pseudoaneurysm originating from the right hepatic artery. (b) The
thrombosis was obtained using a combined percutaneous
aneurysm was treated with coil embolisation followed by the
transhepatic and transarterial approach. Note the needle
placement of a Viabond stent graft (Gore)
approaching the aneurysm for direct puncture and occlusion

Although partial liver resection is a therapeutic option, this recurrence.22 Recanalisation or incomplete thrombosis can
operation is performed for HAA much less commonly occur and therefore it is necessary to obtain follow-up
today. Most of these intrahepatic aneurysms are treated angiography. Often a staged approach is required to
with TCE.3 Selective catheterisation of the hepatic artery achieve complete thrombosis. Transcatheter embolisation
and its branches allows for TCE and thrombosis of the can be combined with direct percutaneous puncture of the
aneurysm sac. The feeding vessel should be embolised HAA in some cases15 (Fig. 28.4). As with SAAs, stent grafts
as distally as possible because occlusion of a more proxi- may be used in the common hepatic artery or its proximal
mal vessel may result in incomplete thrombosis and branches21 (Fig. 28.5).
 Treatment options: surgical and endovascular 329

Figure 28.6 Intraoperative photograph of an aneurysm of the

common hepatic artery. The aneurysm was treated with excision (a)
and revascularisation using autogenous saphenous vein

Larger, extrahepatic aneurysms usually require surgical

ligation with or without revascularisation (Fig 28.6).
Aneurysms proximal to the gastroduodenal artery can be
treated by excision or exclusion without reconstruction
because of collateral flow from the SMA via the gastroduo-
denal artery. In cases where collateral flow is insufficient or
if the gastroduodenal artery is involved arterial reconstruc-
tion is required. Surgical options for vascular reconstruc-
tion include interposition graft, aorto-hepatic artery bypass
or spleno-hepatic bypass using saphenous vein or hypogas-
tric artery.3,14 Non-operative management may be appro-
priate in the following situations: patients at high operative (b)
risk, those with a life expectancy less than 2 years, when
HAAs are asymptomatic and of less than 2 cm diameter.
Careful observation with serial radiological examinations
is recommended.7

Superior mesenteric and coeliac artery

aneurysms

The potential for SMA and coeliac aneurysms to produce

life-threatening haemorrhage or bowel ischaemia makes
intervention with ligation or resection of the aneurysm
entirely appropriate. For proximally located aneurysms,
vascular reconstruction using aorto-mesenteric bypass or (c)
reimplantation is required.16 Aneurysms of the midportion
of the SMA may be treated with excision or ligation and an Figure 28.7 A large superior mesenteric artery aneurysm was
interposition graft3,19,23 (Fig. 28.7). discovered in a 32-year-old woman who presented with
Similarly, patients with coeliac artery aneurysms are intermittent abdominal pain in the postpartum period. (a) Visceral
most often treated with excision and vascular reconstruc- arteriogram demonstrating the superior mesenteric artery (SMA)
tion. Exposure of the coeliac artery is obtained through a aneurysm. (b) Intraoperative view of the aneurysm. (c)
transabdominal incision with medial rotation of the vis- Interposition autogenous saphenous vein graft replacing the
excised SMA aneurysm
cera or directly through the lesser sac (also see Chapter 36).
In a life-threatening emergency, simple ligation of the
coeliac artery aneurysm may be indicated. Ligation, how- In high risk patients, TCE may be a useful alternative to
ever, may result in hepatic necrosis in patients with insuffi- open surgery.23 This may be particularly useful for saccular
cient collateral flow and should be used with caution in aneurysms arising from the side of the SMA and for some
those with pre-existing liver disease.24 pseudoaneurysms. Aneurysms of the first branch of the
330 Visceral artery aneurysms

SMA with sufficient collaterals may also be treated with

TCE.25 Slow growing aneurysms with enlarged collateral
vessels present may be successfully occluded without vascular
reconstruction. Embolisation, however, is not recommended
for fusiform aneurysms of the SMA trunk. Because of the
risk of intestinal infarction, TCE should be reserved for
high risk patients with favourable anatomy and adequate
collateral flow. Embolisation of the coeliac artery is a
potential treatment option, but little has been published
regarding this form of therapy. Aneurysms may also occur
in the coeliac axis for which resection of the aneurysm and
revascularisation will be necessary.26,27

Gastroduodenal, pancreaticoduodenal and

pancreatic artery aneurysms
Surgery for aneurysms of the gastroduodenal, pancreatico-
duodenal and pancreatic arteries involves ligation and may
also necessitate treatment of associated pancreatitis or
pseudocyst.28,29 Aneurysms in the middle or distal region
of the pancreas call for distal pancreatic resection. These
aneurysms often have multiple communicating vessels Figure 28.8 Visceral arteriogram demonstrating an aneurysm of
making simple ligation difficult. Endoaneurysmectomy the pancreaticoduodenal artery (PDA) in a patient with arterial
and suture ligation of the feeding vessels from within the occlusive disease of the coeliac and superior mesenteric arteries.
sac will prevent recurrence. The PDA aneurysm developed in an important collateral vessel.
The patient was treated with aorto-coeliac and aorta-to-superior
Open surgery in patients with these aneurysms can be
mesenteric artery bypass, using autogenous saphenous vein,
especially dangerous with mortality of 13 per cent or higher.
followed by ligation of the aneurysm
Percutaneous treatment with TCE has become more popu-
lar and is becoming the preferred treatment for most gas-
troduodenal, pancreaticoduodenal and pancreatic artery excision of the aneurysm. In these situations TCE repre-
aneurysms 28,30 The availability of wires and catheters for sents a good treatment option and is being increasingly
super-selective catheterisation of small feeding vessels per- employed.24
mits the precise localisation and treatment of these difficult
aneurysms. The collateral circulation between the coeliac Mesenteric branch artery aneurysms
artery and SMA ensures that satisfactory blood flow remains
even after embolisation of the parent artery in both the distal Surgical treatment of aneurysms of the jejunal, ileal and
and proximal portion of the aneurysm.31,32 These aneurysms, colic arteries usually involves ligation or resection of the
and especially pseudoaneurysms, can be very fragile and aneurysm, sometimes along with a portion of the bowel
embolisation may result in secondary rupture.30,32 which it supplies.14 Super-selective catheterisation of the
There have been several case reports of aneurysms of the aneurysm with the injection of embolic material has been
GDA or the pancreaticoduodenal artery associated with reported as a possible option10,25 (Fig. 28.9). Transcatheter
coeliac artery occlusive disease. The increased blood flow in embolisation must be used with caution in these lesions
these enlarged collateral vessels is thought to be the causative because of the risk of infarcting the segment of bowel sup-
factor. Coeliac artery occlusive disease may result from ath- plied by the parent artery.24
erosclerosis or from compression by the median arcuate lig-
ament. Transcatheter embolisation of these aneurysms via
the SMA can be especially useful in the treatment of a bleed- MANAGEMENT STRATEGIES AND
ing aneurysm.33 Most ruptured aneurysms are still treated RECOMMENDATIONS
surgically with mesenteric revascularisation of the coeliac
artery and ligation or resection of the aneurysm34 (Fig. 28.8).
The various types of VAAs differ somewhat in their natural
history and treatment options. The surgeon must decide
Gastric and gastroepiploic artery aneurysms whether or not to treat an aneurysm based on their know-
ledge of the natural history of the disease and the risks asso-
Gastric and gastroepiploic artery aneurysms occurring in ciated with a particular intervention. Unlike abdominal
an extragastric location may be treated with ligation or aortic aneurysms, the natural history of visceral artery
 Management strategies and recommendations 331

treatment options for a particular type of VAA. The under-

standable lack of a large patient series precludes consensus
on best practice but recommendations can be made based
upon published literature and current experience.
The CT scan is the best test for detecting a VAA. As mul-
tiple VAAs are often seen, all patients with known VAAs
should undergo contrast arteriography. Haemodynamically
unstable patients are best taken directly to the operating
room for emergency exploration and ligation of the bleeding
vessel.35,36 If the patient is stable enough to undergo arteri-
ography, information obtained radiologically will assist in
planning an operative procedure or in considering treat-
ment either by TCE or the insertion of a stent graft.
Most SAAs which present with rupture are at least 2 cm
in diameter. Patients with enlarging or symptomatic
aneurysms should be treated. The tendency for these
aneurysms to expand and rupture during pregnancy requires
that women of childbearing age should have them treated
(a) electively. Aneurysms found during pregnancy should be
dealt with before the third trimester. Patients with portal
hypertension or those who have a portocaval shunt are also
at increased risk of rupture. These aneurysms may rupture
following liver transplantation and therefore aneurysm
repair at the time of transplantation, or in a staged approach,
should be considered.37 Although SAAs in young patients
should be treated, smaller aneurysms in elderly or high risk
patients may be safely observed.15 Open surgery is the pre-
ferred treatment for ruptured SAAs. Although in the elect-
ive setting preservation of the spleen is preferred in order
to avoid postsplenectomy sepsis, in the case of rupture,
splenectomy and resection of the aneurysm is necessary. In
patients presenting with pancreatitis or portal hyperten-
sion the operative mortality for open surgery can be quite
high. Transcatheter embolisation or stent graft treatment
may be a better option for patients who are not good surgi-
cal candidates. Percutaneous therapy is especially attractive
for aneurysms associated with pancreatic inflammation or
the ‘hostile abdomen’.19,36
The optimal size at which HAAs should be treated has
not been clearly established. There is evidence to suggest
(b) that some small aneurysms of less than 2 cm diameter
may be safely observed.7 Although more HAAs are being
Figure 28.9 An aneurysm of a branch of the superior mesenteric discovered in asymptomatic patients, many still present
artery in a high risk patient with a ‘hostile abdomen’ successfully with rupture. The known mortality rate from rupture of
treated using transcatheter embolisation. (a) The aneurysm
approximately 30 per cent supports an aggressive approach
(arrowed) originating from a branch of the superior mesenteric
artery. (b) Successful thrombosis with coils
to incidentally discovered lesions in the patient with a low
risk for surgical intervention.7 Risk factors for rupture
include multiple HAAs and those of non-atherosclerotic
aneurysms is mostly unknown. At our institution, 42 per origin, in particular aneurysms associated with polyarteritis
cent of the VAAs that were diagnosed presented with rup- nodosa.7 Pseudoaneurysms may also be at increased risk of
ture. In this series the mortality rate from rupture was rupture. Intrahepatic artery aneurysms are most commonly
25 per cent.19 Although ruptured VAAs must be treated, treated with transarterial or TCE techniques. Aneurysms of
controversy surrounds management when they are found the common hepatic artery should be treated by open sur-
incidentally and remain asymptomatic. Unlike aortic gery and arterial reconstruction.15,24 Endovascular stent
aneurysms, the relation between VAA aneurysm size and graft placement offers another option for aneurysms of the
risk of rupture is not clear. There may be several appropriate common hepatic artery in patients at high risk for surgery.10
332 Visceral artery aneurysms

Many SMA and coeliac artery aneurysms are symptomatic 2 Shanley CJ, Shah NL, Messina BS, Messina LM. Common
at the time of diagnosis. The dangers of bowel ischaemia or splanchnic artery aneurysms: splenic, hepatic, and celiac. Ann
life-threatening haemorrhage are persuasive reasons for Vasc Surg 1996; 10: 315–22.
treating them, except for small asymptomatic aneurysms in 3 Messina LM, Shanley CJ. Mesenteric ischemia. Surg Clin North
Am 1997; 77: 425–43.
the high risk patient. Ligation or resection of the aneurysm
4 Abbas MH, Stone WM, Fowl RJ, et al. Splenic artery aneurysm:
with arterial revascularisation is the treatment of choice. two decades experience at Mayo Clinic. Ann Vasc Surg 2002;
Faced with aneurysms of the gastroepiploic, pancreati- 16: 442–9.
coduodenal or pancreatic arteries the best treatment is 5 Rokko S, Amundsen S, Bjerke-Larssen T, Jensen D. Review: the
TCE.38,39 Open surgery is indicated when a pseudocyst or diagnosis and management of splanchnic artery aneurysms.
other intra-abdominal pathology is present, when percu- Scand J Gastroenterol 1996; 31: 737–43.
taneous treatment fails or in the low risk patient.40 Bleeding 6 Leelaudomlipi S, Bramhall SR, Gunson BK, et al. Hepatic-artery
aneurysms of the gastric and gastroepiploic arteries are aneurysm in adult transplantation. Transpl Int 2003; 16: 257–61.
best treated with TCE. As at least 90 per cent of these 7 Abbas MH, Fowl RJ, Stone WM, et al. Hepatic artery aneurysm
aneurysms present with rupture, incidentally discovered factors that predict complications. J Vasc Surg 2003; 38: 41–5.
gastric or gastroepiploic artery aneurysms should be 8 Stone WM, Abbas M, Cherry KJ, et al. Superior mesenteric artery
aneurysm: is presence an indication for intervention? J Vasc Surg
treated electively with TCE. Mesenteric branch artery
2002; 36: 234–7.
aneurysms are best treated with ligation of the aneurysm 9 Shanley CJ, Shah NL, Messina BS, Messina LM. Uncommon
and resection of the adjacent portion of the bowel.10 splanchnic artery aneurysms: pancreaticoduodenal,
gastroduodenal, superior mesenteric, inferior mesenteric, and
colic. Ann Vasc Surg 1996; 10: 506–15.
Conclusion 10 Tessier DJ, Abbas MH, Flowl RJ, et al. Management of rare
mesenteric arterial branch aneurysms. Ann Vasc Surg 2002; 16:
An aggressive approach is indicated in the treatment of 586–90.
most VAAs. In our experience, these aneurysms usually 11 Hossain A, Reis ED, Dave SP, et al. Visceral artery aneurysms:
tend to expand, produce symptoms and rupture, some- experience in a tertiary-care center. Am Surgeon 2001; 67:
times with fatal results. Once rupture occurs, ligation of 432–7.
the bleeding vessel is most often sufficient, but revascular- 12 Asokan S, Chew EK, Ng KY, et al. Post partum splenic artery
isation is sometimes indicated because of lack of adequate aneurysm rupture. J Obstet Gynecol Res 2000; 26: 199–201.
collateral flow. Endovascular interventional choices for 13 Herbeck M, Horbach T, Putzenlechner C, et al. Ruptured splenic
VAAs should take into account the location of the artery aneurysm during pregnancy: a rare case with both
maternal and fetal survival. Am J Obstet Gynecol 1999; 181:
aneurysm, the perceived risk of rupture and the antici-
763–4.
pated morbidity and mortality of the intended treatment.
14 de Perrot M, Buhler L, Deleaval J, et al. Management of true
aneurysms of the splenic artery. Am J Surg 1998; 175: 466–8.
15 Carr SC, Pearce WH, Vogelzang RL, et al. Current management
Key references of visceral artery aneurysms. Surgery 1996; 120: 627–33.
16 Grego FG, Lepidi S, Ragazzi R, et al. Visceral artery aneurysms: a
Abbas MH, Fowl RJ, Stone WM, et al. Hepatic artery aneurysm single center experience. Cardiovasc Surg 2003; 1: 19–25.
factors that predict complications. J Vasc Surg 2003; 38: 17 Arca MJ, Gagner M, Hentford BT, et al. Splenic artery aneurysms:
41–5. methods of laparoscopic repair. J Vasc Surg 1999 30: 184–8.
Abbas MH, Stone WM, Fowl RJ, et al. Splenic artery aneurysm: two 18 Guillon R, Garcier JM, Abergel A, et al. Management of splenic
decades experience at Mayo Clinic. Ann Vasc Surg 2002; 16: artery aneurysms and false aneurysms with endovascular
442–9. treatment in 12 patients. Cardiovasc Intervent Radiol 2003; 26:
Carr SC, Pearce WH. Management of visceral artery aneurysms. 256–60.
Practical Vasc Surg 1999: 241–58. 19 Carr SC, Mahvi DM, Hoch JR, et al. Visceral artery aneurysm
Shanley CJ, Shah NL, Messina BS, Messina LM. Common splanchnic rupture. J Vasc Surg 2001; 33: 806–11.
artery aneurysms: splenic, hepatic, and celiac. Ann Vasc Surg 20 Melissano G, Chlesa R. Successful surgical treatment of visceral
1996; 10: 315–22. artery aneurysms after failure of percutaneous treatment. Tex
Tessier DJ, Abbas MH, Flowl RJ, et al. Management of rare Heart Inst J 1998; 25: 75–8.
mesenteric arterial branch aneurysms. Ann Vasc Surg 2002; 21 Larson RA, Solomon J, Carpenter JP. Stent graft repair of visceral
16: 586–90. artery aneurysms. J Vasc Surg 2002; 36: 1260–3.
22 Tarazov PG, Ryzhkov VK, Polysavov VN, et al. Extraorganic
hepatic artery aneurysm: failure of transcatheter embolization.
REFERENCES HPB Surg 1998; 11: 55–60.
23 Zimmerman-Klima PM, Wixon CL, Bogey WM Jr, et al.
Considerations in the management of aneurysms of the superior
1 Rokko S, Amundsen S, Bjerke-Larssen T, et al. The diagnosis and mesenteric artery. Ann Vasc Surg 2000; 14: 410–14.
management of splanchnic artery aneurysms. Scan J 24 Carr SC, Pearce WH. Management of visceral artery aneurysms.
Gastroenterol 1996; 31: 737–42. Practical Vasc Surg 1999: 241–58.
 References 333

25 Lorelli DR, Cambria RA, Seabrook GR, Towne JB. Diagnosis and 33 Kobayashi S, Yamaguchi A, Isogai M, et al. Successful
management of aneurysms involving the superior mesenteric transcatheter embolization of a pancreaticoduodenal artery
artery and its branches. A report of four cases. Vasc Endovasc aneurysm in association with celiac axis occlusion: a case report.
Surg 2003; 37: 59–66. Hepatogastroenterology 1999; 46: 2991–4.
26 Veraldi GF, Dorrucci V, deManzoni G, et al. Aneurysm of the 34 Suzuki K, Kashimura H, Sato M, et al. Pancreaticoduodenal artery
celiac trunk: diagnosis with US-color-doppler. Presentation of a aneurysms associated with celiac axis stenosis due to
new case and review of the literature. Hepatogastroenterology compression by median arcuate ligament and celiac plexus.
1999; 46: 781–3. J Gastreoenterol 1998; 33: 434–8.
27 Detroux M, Anidjar S, Nottin R, Robinson LP. Aneurysm of a 35 Wagner WH, Allins AD, Treiman RL, et al. Ruptured visceral artery
common celiomesenteric trunk. Ann Vasc Surg 1998; 12: 78–82. aneurysms. Ann Vasc Surg 1997; 11: 342–7.
28 Coll DP, Ierardi R, Kerstein MD, et al. Aneurysms of the 36 Carmeci C, McClenathan J. Visceral artery aneurysms as seen in a
pancreaticoduodenal arteries: a change in management. Ann community hospital. Am J Surg 2000; 179: 486–9.
Vasc Surg 1998; 12: 286–91. 37 Jovine E, Mazziotti A, Grazi GL, et al. Rupture of splenic artery
29 Konstantakos AK, Coogan S, Husni EA, Raaf JH. Aneurysm of the aneurysm after liver transplantation. Clin Transplantation 1996;
gastroduodenal artery: an unusual cause of obstructive jaundice. 10: 451–4.
Am Surgeon 2000; 66: 695–8. 38 deWeerth A, Buggisch P, Nicolas V, Maas R. Pancreaticoduodenal
30 Yamagami T, Arai Y, Sueyoshi S, et al. Letter to editor re: artery aneurysm – a life-threatening cause of gastrointestinal
embolization of ruptured pancreaticoduodenal artery aneurysm: hemorrhage: case report and review of the literature.
report of two cases. Cardiovasc Intervent Radiol 1999; 22: Hepatogastroenterology 1998; 45: 1651–4.
440–2. 39 Neschis DG, Safford SD, Golden MA. Management of
31 Kasirajan K, Greenberg RK, Clair D, Ouriel K. Endovascular pancreaticoduodenal artery aneurysms presenting as
management of visceral artery aneurysm. J Endovasc Ther 2001; catastrophic intraabdominal bleeding. Surgery 1998; 123:
8: 150–5. 8–12.
32 de Perrot M, Berney T, Deleaval J, et al. Management of true 40 Yeh TS, Jan YY, Jeng LB, et al. Massive extra-enteric
aneurysm of the pancreaticoduodenal arteries. Ann Surg 1999; gastrointestinal hemorrhage secondary to splanchnic artery
229: 416–20. aneurysms. Hepatogastroenterology 1997; 44: 1152–6.
This page intentionally left blank
 29
Mesenteric Ischaemia

DARYLL M BAKER, JANICE TSUI, AVERIL O MANSFIELD

The problem 335 Postoperative care 339

Aetiology 335 Complications 340
Pathophysiology 335 Results 340
Diagnosis 336 References 340
Treatment 337

THE PROBLEM Table 29.1 Aetiology of acute mesenteric ischaemia

Causes Frequency (per cent)6

Mesenteric ischaemia can be classified into acute and chronic
disease, which present very differently clinically and require Acute mesenteric arterial embolus 50
different management approaches. Here, the focus will be on Acute mesenteric arterial thrombosis 25
acute mesenteric ischaemia, which presents as a catastrophic Non-occlusive mesenteric ischaemia 20
surgical emergency with high morbidity and mortality. It Acute mesenteric venous thrombosis 5
Miscellaneous causes Rare
represents 0.1 per cent of hospital admissions1 with reported
mortality rates of 60–100 per cent.1–4
Prompt diagnosis and aggressive management are
required. Correct diagnosis in the early stages, however, is
often difficult due to the non-specific nature of symptoms level, high energy phosphates such as adenosine 5 triphos-
resulting in delay in diagnosis.5 phate (ATP) are depleted, with a build-up of catabolic prod-
ucts and lactate.7 Mucosal permeability increases and tissue
injury, which is initially reversible, occurs (see Chapter 3).
AETIOLOGY The severity of the injury depends on the duration and sever-
ity of ischaemia, and ranges from mucosal infarction where
The commonest causes of mesenteric ischaemia are mesen- the lesion does not extend deeper than the muscularis
teric arterial occlusion due to thrombus or embolus, mucosae, to mural infarction involving the mucosa and sub-
mesenteric venous occlusion and non-occlusive mesenteric mucosa and to transmural infarction where all visceral layers
ischaemia.1 Rare causes include mechanical causes, haema- are affected8 (Fig. 29.1).
tological, vasculitic and endocrine conditions and drug Subsequent reperfusion may lead to ischaemia-
induced events. The relative frequencies of these causes are reperfusion injury resulting in further tissue damage, which
shown in Table 29.1. is in part due to free radical production via the xanthine
oxidase pathway9 (see Chapter 2). Further tissue necrosis
occurs, with increased capillary permeability, protein leak
into the gut lumen and translocation of bacteria, endotoxin
PATHOPHYSIOLOGY and gut enzymes into the portal circulation,10 and systemic
complications.11 This could lead to the development of mul-
Acute insufficiency of the blood supply to the small bowel tiple organ dysfunction syndrome with its associated high
and/or right colon causes hypoxic insult to the gut. At cellular mortality12 (see Chapter 4).
336 Mesenteric ischaemia

DIAGNOSIS

Clinical presentation

Patients present with severe abdominal pain which is fre-

quently out of proportion to the physical findings.6 This
may be accompanied by sudden gut emptying, by vomiting
and/or defaecation. Symptoms may occur suddenly and
progress within a few hours, usually due to acute mesen-
teric ischaemia secondary to embolisation, but may also
present more insidiously with progression of symptoms
over a few days.
On physical examination, few signs are present in the early
phases. The abdomen may be distended but bowel sounds are
present. As bowel ischaemia progresses to infarction, peri-
Figure 29.1 Mesenteric infarction toneal signs develop. The abdomen becomes severely tender
and grossly distended with absent bowel sounds.
Acute mesenteric arterial embolism

The usual sources of emboli are left atrial or ventricular

Investigations
mural thrombi or cardiac valvular lesions.1 The majority of
Investigations should not lead to unnecessary delays in
emboli impact on the superior mesenteric artery (SMA)
revascularisation. Laboratory investigations are non-
just distal to the origin of the middle colic artery. Reactive
specific and abnormal findings tend to develop late. They
mesenteric vasoconstriction may also occur, reducing col-
include a rise in the haemoglobin and haematocrit consist-
lateral flow and exacerbating ischaemic injury.
ent with haemoconcentration, leucocytosis with left shift,
elevated liver enzymes, serum amylase and lactic dehydro-
Acute mesenteric arterial thrombosis genase, and metabolic acidosis.6
Plain abdominal X-rays are useful in excluding other
Thrombosis of the SMA or coeliac axis usually occurs at causes of abdominal pain such as bowel obstruction and
regions of pre-existing atherosclerotic plaques. This usu- perforation. An ileus or bowel wall thickening due to sub-
ally occurs at the origin of the vessel.1 mucosal oedema or haemorrhage may be seen in mesenteric
ischaemia, but in 25 per cent of these patients completely
normal plain abdominal X-rays have been reported.15 In the
Non-occlusive mesenteric ischaemia
late stages, pneumatosis of the bowel wall may be detected
and portal vein gas indicates an extremely poor prognosis.1
Low mesenteric flow may result in mesenteric ischaemia in
Duplex ultrasonography is a simple, non-invasive test
the absence of anatomical arterial or venous obstruction.
which can help in assessing the patency of the mesenteric
Causes of low-flow states include cardiogenic shock, sepsis
vessels.16 However, technical expertise is required and the
and administration of vasoconstrictors such as digoxin or
presence of dilated loops of bowel in many patients makes
-adrenergic agents. Excessive sympathetic activity is stimu-
it impossible to image the coeliac axis and the SMA.
lated during these conditions in an attempt to maintain
Abdominal computed tomography (CT) scan may show
cardiac and cerebral perfusion, and may cause mesenteric
bowel wall thickening in subacute mesenteric ischaemia. In
vasospasm.13 Underlying atherosclerotic arterial disease is
mesenteric vein thrombosis, thrombus or lack of opacifica-
usually also present.
tion of the mesenteric veins with intravenous contrast may
be observed,17 whereas non-enhancement of the mesenteric
Acute mesenteric venous thrombosis arteries following timed intravenous contrast injections
may identify mesenteric arterial thrombosis or embolisa-
Conditions predisposing to acute mesenteric venous throm- tion. Pneumatosis and portal vein gas can also be visualised.
bosis include hypercoagulability, cirrhosis, splenomegaly, Mesenteric angiography remains the definitive diagnos-
malignancy, infection, trauma, pancreatitis and diverticular tic test. It is performed via a transfemoral approach and both
disease.14 Bowel wall oedema, increased outflow resistance anteroposterior and lateral views are required for adequate
and increased blood viscosity impede arterial flow resulting assessment of the vessels. Angiography not only enables
in bowel infarction.1 Massive fluid influx into the bowel visualisation of the mesenteric vessels with identification of
wall and lumen can occur, resulting in hypovolaemia and the cause of ischaemia, but also detects any underlying ath-
haemoconcentration. erosclerotic disease, which is important in the planning
 Treatment 337

of revascularisation procedures. Intra-arterial infusion of prevent further thrombus propagation. Antibiotics effective
vasodilators can be also commenced and endovascular pro- against bowel flora should be administered. Nasogastric
cedures carried out if appropriate (see below). decompression should be instituted to avoid aspiration.

ANGIOGRAPHIC FINDINGS
Endovascular procedures
Superior mesenteric artery emboli appear as sharp rounded
filling defects on the angiogram but they may also occlude SELECTIVE INTRA-ARTERIAL VASODILATOR INFUSIONS
the origin of the SMA and be mistaken for thrombosis.
At the time of angiography, selective SMA catheter infusion
Distal vessels are poorly visualised due to poor collateral
of vasodilating agents such as papaverine can be adminis-
flow and intense vasospasm but minimal atherosclerotic
tered. This is the primary therapy for non-occlusive mesen-
changes are usually seen in other vessels. Acute SMA throm-
teric ischaemia, but is also useful in reducing visceral
bosis appears as an abrupt cut-off of the vessel at or within
vasoconstriction present in all forms of acute mesenteric
2 cm of the origin. Atherosclerotic narrowing of other ves-
ischaemia. Heparin is also administered to prevent throm-
sels and vasospasm are often seen. Large collaterals from the
bosis in the cannulated vessel but must be infused through
coeliac axis or inferior mesentery artery suggest underlying
a separate line to avoid precipitation when mixed with
chronic mesenteric ischaemia.
papaverine. Other vasodilators such as phenoxybenzamine
In non-occlusive mesenteric ischaemia, SMA occlusion
and prostaglandin E1 have also been used in this way but with
is not seen but narrowing of the SMA branch origins
less success than papaverine. Local complications of intra-
and irregularities and impaired filling of vessels indica-
arterial vasodilator infusions include pericatheter throm-
tive of vasospasm are observed. Partial occlusion or non-
bosis and puncture site haematoma and haemorrhage.
opacification of the superior mesenteric and portal veins
Systemic hypotension is rare unless the catheter becomes
occurs in mesenteric vein thrombosis. Delayed arterial emp-
dislodged from the SMA.18
tying and vasospasm may also be seen.

THROMBOLYSIS AND ANGIOPLASTY

TREATMENT In highly selected patients with an early diagnosis of SMA
embolus with no signs of peritonism, intra-arterial throm-
General principles bolytic therapy delivered at the time of angiography may
be used in specialised units.19,20 During therapy, patients
The patient should be resuscitated aggressively, intravenous must be closely monitored and development of any peri-
anticoagulation commenced to prevent thrombus propaga- toneal signs or failure of lysis within 4 hours are indications
tion and angiography performed promptly if acute mesen- for immediate surgical exploration.1 Percutaneous angio-
teric ischaemia is suspected. Endovascular procedures may plasty of atherosclerotic plaques at the SMA origin
be suitable in selected patients with no peritoneal signs, pro- unmasked by thrombolysis has also been reported.21,22
ceeding to surgery if symptoms deteriorate or success is not However, this is technically difficult and restenosis rates of
evident within 4 hours. Surgical management involves 25–50 per cent have been reported.
revascularisation and resection of necrotic bowel. Frankly
necrotic bowel should be resected and the extent of bowel
resection minimised by planning a second look laparotomy Surgery
in 12–24 hours to reassess bowel viability and for further
resection if necessary. Bowel continuity can be restored pri- Most patients with acute mesenteric ischaemia require
marily or stomas exteriorised and anastomosed at a later surgical exploration when the diagnosis is confirmed, other
stage. pathologies excluded, the bowel revascularised and infarcted
bowel removed. The operating room should be kept as warm
as possible to prevent further vasoconstriction. A midline
Initial resuscitation incision is used and the abdomen fully explored. The extent
of bowel ischaemia and necrosis is assessed (see below).
Resuscitation should be commenced immediately under Revascularisation is carried out before resecting necrotic
adequate monitoring with a urinary catheter, central venous bowel, but if the extent of bowel necrosis is very extensive,
access or pulmonary artery catheter where necessary. Admis- endarterectomy or palliative treatment are the only options.
sion to the intensive care unit may be required. Volume
deficit and metabolic disturbances should be corrected with
SURGICAL EXPOSURE OF THE SMA
intravenous crystalloids and blood products where neces-
sary. Causative factors such as cardiac failure should be The proximal SMA is exposed by first retracting the trans-
treated. Intravenous heparin should be commenced to verse colon superiorly and the small bowel inferiorly and
338 Mesenteric ischaemia

Figure 29.2 Superior mesenteric artery exposed Figure 29.3 Rationale for ‘second look’: a segment of bowel had
undergone resection but at the ‘second look’ exploration there
to the right. The small bowel mesentery is incised at a point are further areas of necrosis requiring further resection
in the root of the transverse mesocolon where the middle
colic artery ascends, and the proximal SMA is dissected performed. More commonly, several regions of the bowel are
free between the pancreas and the fourth part of the duo- affected and the degree of ischaemia is unclear. Obviously
denum. To expose both the SMA and the coeliac axis, the necrotic areas are resected and bowel ends are exteriorised as
lesser sac is opened lateral to the oesophagus. The supra- stomas. A ‘second-look’ laparotomy is undertaken 12–24
coeliac aorta is exposed by dividing the median arcuate liga- hours later to assess bowel of questionable viability (Fig.
ment and carefully dissected caudally until the coeliac axis 29.3) when non-viable bowel is resected and the ends may be
is reached. The SMA is then exposed under the upper bor- anastomosed.24 This allows a conservative approach to bowel
der of the pancreas (Fig. 29.2). resection at the initial laparotomy preserving as much bowel
as possible. Once the decision to perform a ‘second look’
ASSESSMENT OF BOWEL ISCHAEMIA laparotomy has been made, it should be carried out regard-
Necrotic bowel must be resected to prevent perforation less of the patient’s clinical condition because a significant
and sepsis. It is important, however, to preserve as much proportion of patients not showing signs of clinical deteri-
viable bowel as possible to avert the complications of the oration at this time do require further bowel resection.5
short bowel syndrome. Where there is extensive small bowel infarction, the
Intraoperatively, bowel viability is assessed visually patient’s prognosis is extremely poor, and the decision to
according to its colour, sheen, peristaltic activity and pres- carry out extensive bowel resection or to manage the patient
ence of pulses, where dark, dull bowel with no visible peri- palliatively has to be made. The premorbid medical condi-
stalsis and absent pulses indicate non-viable bowel. These tion of the patient is important because those with other ath-
signs, however, are not always reliable. Doppler ultrasonog- erosclerotic manifestations such as severe ischaemic heart
raphy can be used to aid in the assessment of bowel viability disease and previous debilitating cerebrovascular episodes
by detecting arterial and venous flow in the mesentery. This are poor candidates for massive bowel resection. Bowel per-
is a simple, inexpensive method but it requires experience to forations, sepsis and multiple organ dysfunction syndrome
obtain reliable results. are indicators of high mortality (see Chapters 3 and 4).
An alternative or additional method is the administration Where extensive small bowel resection is carried out, life-
of intravenous fluorescein followed by illumination with long total parenteral nutrition will be required in surviving
Wood’s ultraviolet light to confirm bowel perfusion. Normal patients. An algorithm (Fig. 29.4) provides a pathway of care
bowel shows a homogeneous yellow-green pattern while in suspected acute mesenteric ischaemia.
patchy fluorescence or areas of non-fluorescence are not
viable. Other techniques have been investigated, but Doppler Definitive operative treatment
ultrasonography and fluorescein dye remain the most useful
although limited in their sensitivity and specificity.23 Operative procedures are designed to meet specific mani-
festations of acute mesenteric ischaemia.
MANAGEMENT OF BOWEL ISCHAEMIA
ACUTE MESENTERIC ARTERIAL EMBOLISM
The management of infarcted bowel depends on the degree
and extent of tissue necrosis. A single short segment of The site and extent of the embolus can be determined by
non-viable bowel can be resected and a primary anastomosis direct palpation of the SMA, where the proximal pulse is
 Postoperative care 339

Acute abdomen

Clinical suspicion of acute mesenteric ischaemia

Peritonitis

Mesenteric
angiography

Laparotomy Endovascular
Revascularisation Anticoagulation treatment
Bowel resection

Second look laparotomy

Figure 29.4 Algorithm providing a pathway of care for patients Figure 29.5 Saphenous vein bypass onto the superior
with suspected acute mesenteric infarction mesenteric artery

usually palpable. The SMA is then mobilised at or just dis- Operative procedures for acute mesenteric
tal to the level of the obstruction. Vascular loops are used ischaemia
for proximal and distal control. Heparin should have been
administered by this stage. A small transverse or a longitu- • SMA embolism – embolectomy
dinal arteriotomy is made. The embolus is removed and a • SMA thrombosis – aorta or iliac artery to SMA
balloon embolectomy catheter passed proximally and dis- bypass using vein, Dacron or PTFE, supracoeliac
tally in the artery to retrieve residual embolus. Once aorta to SMA
adequate inflow and back-bleeding have been established, • Coeliac axis/SMA thrombosis: bifurcated graft
the vessel is flushed with heparin saline. Transverse arteri- bypass
otomies are closed primarily in a transverse fashion and • Ischaemic bowel: resection and planned ‘second
longitudinal arteriotomies are closed with a vein patch to look’
prevent narrowing of the vessel.

NON-OCCLUSIVE MESENTERIC ISCHAEMIA

ACUTE MESENTERIC ARTERIAL THROMBOSIS
As mentioned above, selective arterial administration of
The site of thrombosis is most commonly in the proximal papaverine is the primary therapy in non-occlusive mesen-
1–3 cm of the SMA and therefore the proximal pulse is not teric ischaemia. In addition, the underlying cause should
palpable. Thrombosis usually occurs at the site of athero- be treated and any vasoconstricting agents such as -agonists
sclerotic narrowing. Revascularisation is usually required stopped. Surgical exploration is limited to patients with
as thrombectomy and endarterectomy is technically diffi- peritonitis indicating bowel necrosis and the need for bowel
cult, requiring exposure of the supracoeliac aorta and the resection.
proximal mesenteric vessels25 and early re-occlusion rates
are high.18 MESENTERIC VEIN THROMBOSIS
Retrograde or antegrade bypasses can be used. A short
side-to-side bypass graft from the aorta to the SMA distal Patients with no signs of peritonitis or any other evidence of
to the obstruction is the simplest procedure.18 However, bowel necrosis are managed conservatively with anticoagu-
extensive aortic atherosclerosis may preclude its use and a lation. Intravenous heparin should be started at the time of
retrograde iliac to SMA bypass may be more suitable. diagnosis and patients should be maintained on lifelong
Alternatively, an antegrade bypass from the supracoeliac therapy with warfarin to reduce the chances of recurrence.
aorta to both the SMA and the coeliac axis can be performed Surgical treatment is limited to resection of necrotic bowel
if both vessels are diseased. A temporary arterial shunt may as venous thrombectomy is difficult and rarely successful.
be inserted to allow restoration of blood flow during bypass
grafting.25 An autologous conduit, most commonly the
POSTOPERATIVE CARE
reversed long saphenous vein, should be used in the pres-
ence of abdominal contamination from necrotic bowel6
(Fig. 29.5). If no suitable vein is available, polytetrafluoro- Patients should be managed on an intensive care unit with
ethylene (PTFE) or Dacron grafts may be used. close monitoring and support to prevent and manage
340 Mesenteric ischaemia

multiple organ failure (see Chapter 4). Fluid balance and The average hospital stay for patients who survive surgery
serum electrolytes must be monitored carefully and cor- is 2 months. Most patients return to their premorbid level of
rected. Intravenous heparin should be continued. Broad independent life18 but many will suffer from gastrointestinal
spectrum antibiotics commenced preoperatively should symptoms such as weight loss, loss of appetite, abdominal
also be continued until culture results are available and pain, diarrhoea, constipation, nausea and vomiting. Most of
adjusted accordingly. these symptoms, however, will be mild.2 The reported inci-
Alimentation should be introduced as soon as possible, dence of clinically apparent chronic short bowel syndrome
the exact timing and type given depending on the procedure ranges from 20 per cent to 60 per cent.2
untaken. Parental nutrition is often required until normal Recurrent arterial mesenteric ischaemia is rare, with a
oral intake is resumed. Long term parenteral nutrition higher (5 per cent) incidence of recurrent mesenteric vein
should be managed by specialist teams. Further investiga- thrombosis which increases to 25 per cent if the patient is not
tions into any underlying causes of acute mesenteric anticoagulated.18 Despite the low incidence of recurrent
ischaemia should be carried out, e.g. in patients sustaining mesenteric ischaemia, the life expectancy of these patients is
mesenteric embolism, cardiac echocardiography is required. low. In a retrospective study of the outcome of 31 patients
Anticoagulation should be continued in the post- discharged following surgery for acute mesenteric ischaemia,
operative period in all patients unless there are contraindica- the 2-year survival rate was less than 70 per cent and the
tions. Patients with mesenteric venous thrombosis, arterial 5-year survival rate was 50 per cent.2 This was mainly due to
embolism and proved haematological disorders should be cardiovascular comorbidity, in particular myocardial infarc-
commenced on warfarin, whereas those with mesenteric tion and cerebrovascular events, and malignancies.
arterial thrombosis and non-occlusive disease should be
treated with antiplatelet agents.2 The decision for lifelong
anticoagulation must be made after balancing the low risk Conclusions
of recurrent mesenteric ischaemia with the risk of haemor-
rhage. In patients where the risks of anticoagulation are Acute mesenteric ischaemia is a rare but catastrophic acute
low, therapy should probably be continued. surgical emergency. It represents a diagnostic challenge
and therefore a high index of suspicion is required as
patients often present without obvious physical signs.
Angiography should be carried out promptly if it is sus-
COMPLICATIONS pected, but revascularisation should not be delayed by
investigations. Aggressive resuscitation, intraoperative and
Multiple organ failure is an important and common com- postoperative management are all important.
plication following acute mesenteric ischaemia and must
be treated aggressively (see Chapter 4). Other early specific
complications include thrombosis at arteriotomy sites or of
bypass grafts. This results in early recurrence of symptoms Key references
and requires angiography and re-exploration. Haemorrhage
Baker DM, Mansfield AO. Acute mesenteric ischaemia. In: Monson JM,
may also occur from the arteriotomy or graft anastomosis
Duthie G, O’Malley K (eds). Surgical Emergencies. Oxford:
site and is often due to infection. Later, persistent diarrhoea Blackwell Science, 1999: 305–15.
is a problem in many patients, even in those without short Endean ED, Barnes SL, Kwolek CJ, et al. Surgical management of
bowel syndrome. Gastrointestinal bleeding may occur due thrombotic acute intestinal ischemia. Ann Surg 2001; 233:
to ulceration or rarely from a SMA intestinal fistula. Late 801–8.
arterial strictures may occur at arteriotomy or anastomotic Klempnauer J, Grothues F, Bektas H, Pichlmayr R. Long-term results
sites with recurrence of symptoms. after surgery for acute mesenteric ischemia. Surgery 1997;
121: 239–43.
Mansour MA. Management of acute mesenteric ischemia. Arch
Surg 1999; 134: 328–30.
RESULTS McKinsey JF, Gewertz BL. Acute mesenteric ischemia. Surg Clin
North Am 1997; 77: 307–18.

Despite increased awareness and better understanding of the

pathophysiology of acute mesenteric ischaemia, morbidity
and mortality have remained high over the past 30 years. REFERENCES
Overall mortality rates of 60–100 per cent have been
reported.1–4 There is evidence that mesenteric arterial occlu- 1 McKinsey JF, Gewertz BL. Acute mesenteric ischemia. Surg
sion and non-occlusive mesenteric ischaemia carry a worse Clin North Am 1997; 77: 307–18.
prognosis, with mortality rates of 70–90 per cent,2 whereas 2 Klempnauer J, Grothues F, Bektas H, Pichlmayr R. Long-term
patients with mesenteric vein thrombosis have a better results after surgery for acute mesenteric ischemia. Surgery
chance of survival, with mortality rates of 20–70 per cent.26,27 1997; 121: 239–43.
 References 341

3 Montgomery RA, Venbrux AC, Bulkley GB. Mesenteric vascular 16 Nicoloff AD, Williamson WK, Moneta GL, et al. Duplex
insufficiency. Curr Probl Surg 1997; 34: 941–1025. ultrasonography in evaluation of splanchnic artery stenosis.
4 Stoney RJ, Cunningham CG. Acute mesenteric ischemia. Surgery Surg Clin North Am 1997; 77: 339–55.
1993; 114: 489–90. 17 Rosen A, Korobkin M, Silverman PM, et al. Mesenteric vein
5 Endean ED, Barnes SL, Kwolek CJ, et al. Surgical management thrombosis: CT identification. AJR Am J Roentgenol 1984;
of thrombotic acute intestinal ischemia. Ann Surg 2001; 143: 83–6.
233: 801–8. 18 Baker DM, Mansfield AO. Acute mesenteric ischaemia. In:
6 Mansour MA. Management of acute mesenteric ischemia. Monson JM, Duthie G, O’Malley K (eds). Surgical Emergencies.
Arch Surg 1999; 134: 328–30. Oxford: Blackwell Science, 1999: 305–15.
7 Parums DV. The pathology of ischaemic-reperfusion injury. In: 19 McBride KD, Gaines PA. Thrombolysis of a partially occluding
Grace PA, Mathie RT (eds). Ischaemia-Reperfusion Injury. Oxford superior mesenteric artery thromboembolus by infusion of
Blackwell Science, 1999: 3–19. streptokinase. Cardiovasc Intervent Radiol 1994; 17: 164–6.
8 Crawford JM. The oral cavity and gastrointestinal tract. In: 20 Rivitz SM, Geller SC, Hahn C, Waltman AC. Treatment of acute
Kumar V, Contran RS, Robbins SL (eds). Basic Pathology. mesenteric venous thrombosis with transjugular intramesenteric
Philadelphia, PA: Saunders, 1997: 470–515. urokinase infusion. J Vasc Intervent Radiol 1995; 6: 219–23.
9 Parks DA, Granger DN. Contributions of ischemia and reperfusion 21 Hallisey MJ, Deschaine J, Illescas FF, et al. Angioplasty for the
to mucosal lesion formation. Am J Physiol 1986; 250: treatment of visceral ischemia. J Vasc Intervent Radiol 1995;
G749–53. 6: 785–91.
10 Stahl GL, Pan HL, Longhurst JC. Activation of ischemia- and 22 VanDeinse WH, Zawacki JK, Phillips D. Treatment of acute
reperfusion-sensitive abdominal visceral C fiber afferents. Role mesenteric ischemia by percutaneous transluminal angioplasty.
of hydrogen peroxide and hydroxyl radicals. Circ Res 1993; Gastroenterology 1986; 91: 475–8.
72: 1266–75. 23 Ballard JL, Stone WM, Hallett JW, et al. A critical analysis of
11 Haglund U, Osterberg J. Local consequences of reperfusion in adjuvant techniques used to assess bowel viability in acute
the gut. In: Grace PA, Mathie RT (eds). Ischaemia-Reperfusion mesenteric ischemia. Am Surg 1993; 59: 309–11.
Injury. Oxford: Blackwell Science, 1999: 65–70. 24 Levy PJ, Krausz MM, Manny J. Acute mesenteric ischemia:
12 Deitch EA. Multiple organ failure. Pathophysiology and potential improved results – a retrospective analysis of ninety-two
future therapy. Ann Surg 1992; 216: 117–34. patients. Surgery 1990; 107: 372–80.
13 Bassiouny HS. Nonocclusive mesenteric ischemia. Surg Clin 25 Whitehill T, Rutherford R. Acute mesenteric ischaemia caused
North Am 1997; 77: 319–26. by arterial occlusions: optimal management to improve survival.
14 Rhee RY, Gloviczki P. Mesenteric venous thrombosis. Surg Clin Semin Vasc Surg 1990; 3: 149–55.
North Am 1997; 77: 327–38. 26 Anane-Sefah JC, Blair E, Reckler S. Primary mesenteric venous
15 Smerud MJ, Johnson CD, Stephens DH. Diagnosis of bowel occlusive disease. Surg Gynecol Obstet 1975; 141: 740–2.
infarction: a comparison of plain films and CT scans in 23 27 Gertsch P, Matthews J, Lerut J, et al. Acute thrombosis of the
cases. AJR Am J Roentgenol 1990; 154: 99–103. splanchnic veins. Arch Surg 1993; 128: 341–5.
This page intentionally left blank
 SECTION
6
Acute Complications of Endovascular
Aortic Repair (EVAR)

30. Endoleak complicating EVAR 345

31. Graft breakdown and migration complicating EVAR 357

32. Atheroembolism complicating EVAR 367

This page intentionally left blank
 30
Endoleak Complicating EVAR

ROBERT J HINCHLIFFE, BRIAN R HOPKINSON

The problem 345 Management 350

Incidence 345 Postoperative care 353
Aetiology/pathophysiology 345 Results 353
Diagnosis 349 References 354

THE PROBLEM been unable to confirm this report. The absence of endoleak
does not reliably predict treatment success.7–9
The feasibility of endovascular repair of infrarenal abdom-
inal aortic aneurysm (EVAR) was demonstrated over a INCIDENCE
decade ago by Parodi, Volodos and coworkers.1,2 The tech-
nology associated with EVAR has evolved at a rapid pace in
parallel with recognition of the factors required for success- The incidence of endoleak from the literature varies
ful endovascular aneurysm exclusion. Enthusiasm for this tremendously from values as high as 44 per cent to 10 per
technique has, however, been tempered by reports of signifi- cent or less.10–12 Critical analysis of these figures serves to
cant device and non-device related complications. These illustrate the multifactorial aetiology of endoleak following
complications have included aneurysm rupture, prevention endografting. As will be discussed later, the presence of
of which is the basis of treatment.3 In fact, a cumulative risk endoleak depends upon patient selection criteria, endovas-
of 1 per cent per year has been reported by the European cular graft (EVG) type, operator experience, imaging
collaborators on Stent-graft Techniques for abdominal aortic modality, classification, timing of investigation and dur-
Aneurysm Repair (EUROSTAR) using first and second ation of patient follow-up.
generation endovascular grafts.4 The incidence of primary endoleak differs from the sec-
Of all possible complications the one that has attracted ondary rate. Primary endoleak occurs in the perioperative
most attention has been endoleak, which is peculiar to period (within 30 days of the EVAR), whereas secondary
EVAR. The term ‘endoleak’ was coined by White et al.: endoleak manifests later. The Sidney experience discov-
‘a condition associated with endoluminal vascular grafts, ered 12 per cent of cases with secondary endoleak requir-
defined by the presence of blood flow outside the lumen of ing intervention.12 The secondary endoleak prevalence is
the endoluminal graft, but within an aneurysm sac or an adja- likely to increase in time. It is interesting to note that the
cent vascular segment being treated by the graft’.5 Although intervention rate for endoleak in the Sydney experience was
all endoleaks share a common feature of blood flow outside double the quantity of interventions required for other
the stent-graft it soon became apparent that a variety of types complications. These figures clearly show the scale of the
of endoleak with differing aetiology and natural history problem.
existed.
The importance of endoleak is its ability to predict
success or failure of endografting and, crucially, whether
AETIOLOGY/PATHOPHYSIOLOGY
it predisposes to aneurysm rupture. Early reports that
endoleak may act as a reliable indicator of outcome were Endoleak is an important consideration when talking about
encouraging.6 Recent evidence with longer follow-up has the success of EVAR. Endoleaks may transmit pressure to
346 Endoleak complicating EVAR

Table 30.1 Classification of endoleaks

Type Cause

1 Attachment sites (or occluder in aorto-uni-iliac

endovascular graft)
2 Retrograde sac filling by patent aortic side branches
3 Graft tear, disintegration or modular limb dislocation
4 Angiographic ‘blush’ experienced on completion
angiography with certain thin walled devices.
? Self-limiting
‘5’ ‘Endotension’. Evidence of raised intrasac pressure
(usually aneurysm sac expansion) in the absence of
radiological evidence of endoleak

Figure 30.2 Type 1 endoleak (proximal)

thrombus has been well recognised in open surgery for a

number of years (Fig. 30.1). Endoleaks may be present
singly or multiply and each may act as an inflow or outflow
channel with respect to the aneurysm sac.

Type 1 endoleak

Type 1 endoleak15 (Fig. 30.2) is due to inadequate seal at

the interface between stent-graft and arterial wall at the
attachment zones. It occurs in up to 10 per cent of patients
undergoing EVAR. Early type 1 endoleak (occurring
within 30 days of operation) has two major causative fac-
tors. The first factor, ‘bad planning’, is related to patient
selection criteria. Accurate preoperative imaging of
aneurysm morphology is mandatory and a prerequisite of
successful aneurysm exclusion. The proximal aneurysmal
neck is a particularly important consideration to ensure a
Figure 30.1 ‘Endotension’: rupture of a thrombosed abdominal good seal.16 Consideration of the aneurysm neck must take
aortic aneurysm in a patient who had endovascular repair five into account its length, diameter, angulation, shape and
years previously; he subsequently developed graft migration,
wall constitution. Essentially all these factors simply relate
kinking and occlusion and was eventually admitted with aneurysm
rupture
to the amount of covered stent providing a seal with the
aortic wall. Unfortunately, there is insufficient evidence to
implicate all these factors in type 1 endoleak. Indeed, contro-
versially, a recent study of 100 patients found no statistically
the aneurysm sac and therefore predispose to aneurysm significant association between any of the above factors
rupture. Four types of endoleak (1–4) have been described and the subsequent development of type 1 endoleak.17
(Table 30.1). The classification is further dependent upon Experimental aneurysm flow models have, however, found
anatomical site (a, b, c) and the presence of an outflow significantly increased proximal perigraft endoleak flow for
channel. Subclassification of the four types of endoleak is proximal neck angulation  30 degrees.18 In vivo studies
not generally undertaken in everyday practice, instead an have also confirmed the importance of neck angulation
anatomical description is preferred to prevent confusion. and to a lesser degree the width of the aortic neck in the
Before describing endoleak in detail two factors must causation of type 1 endoleak.19 The concept of stent-graft
be considered. Pressure transmitted by an endoleak chan- oversizing is now widely believed to be a relevant factor in
nel is not related to the size of the endoleak (diameter or stent-graft planning. Oversizing of 10–20 per cent of the
length).13 Thrombosis of the endoleak channel may not original aortic diameter has been found to significantly
abolish pressure transmission.14 Pressure transmission via reduce the incidence of endoleak.20
 Aetiology/pathophysiology 347

(a) (b)

Figure 30.3 Type 2 endoleak at angiogram: (a) lumbar vessels providing aneurysm inflow, (b) inferior mesenteric artery outflow

The second factor in the development of primary type 1 expansion. Identifying this group of patients preopera-
endoeak is ‘bad deployment’. Operator error, deploying tively is not yet possible.
the graft too low in the aortic neck or even in the aneurysm
sac will result in endoleak. A particular hazard encoun-
tered in angulated necks is a parallax error which results in Type 2 endoleak
maldeployment.
In general there has been a reduction in the develop- Type 2 endoleak arises from retrograde flow in patent aor-
ment of early type 1 endoleak brought about by improve- tic side branches.15 The most commonly involved vessels
ments in preoperative imaging, an awareness of aneurysm are the inferior mesenteric artery (IMA) and one or more
morphology required for successful EVAR, the necessity of lumbar arteries, although other vessels such as the gonadal or
oversizing and increased technical expertise. accessory renal vessels may be responsible (Fig. 30.3).
Finally, it is important to appreciate the differences Accepted type 2 endoleak rates are in the order of 10–25 per
among commercially available stent-grafts. Each graft has cent.5,25 There appear to be no reliable factors, which predis-
different deployment characteristics and sizes, making some pose to the subsequent development of type 2 endoleak,26,27
aneurysms suitable for one stent-graft but unsuitable for although Armon et al.’s results suggested patterns of
another. The differences are most marked in the accom- thrombus distribution may be able to predict patients at
modation of neck morphology, particularly width. risk from persistent endoleak via lumbar vessels.28 Only a
Secondary development of type 1 endoleak has a differ- quarter of patent IMAs will subsequently perfuse the
ent pattern from its early counterpart. Of course, if the aneurysm sac. And there is little evidence to suggest that graft
aneurysm is excluded from the circulation via a tenuous design has any bearing on the incidence of type 2 endoleak.
seal due to ‘bad planning’ or ‘bad deployment’ then a leak Approximately two-thirds of side branch endoleaks seen
may be expected. Unsurprisingly EVGs suffer from the on completion angiography will spontaneously thrombose
same lack of incorporation into the aortic tissues as an in the early postoperative period.29 A small proportion will
open graft.21 They are also subject to the forces exerted on persist.
it by aortic blood flow and elastic recoil. The pattern of One technique that has been found to be a useful indicator
endoleak which emerges in medium and long term follow- of subsequent development of type 2 endoleak is the intra-
up is related to changing aneurysm morphology and graft operative ‘sacogram’ (Fig. 30.4). This technique involves
migration. Significant graft migration is most commonly injection of contrast into the aneurysm sac alongside the
seen 2–3 years post implantation (see Chapter 31). There deployed EVG. In a series of patients in Nottingham under-
are intimate relations between migration and changing going EVAR a negative sacogram was found to predict a
morphology. The fate of the proximal aneurysm neck group in whom the subsequent risk of development of type
following EVAR, like its open counterpart is controversial, 2 endoleak was low.30
with reports of both stability and expansion.22–24 There The natural history of type 2 endoleak remains undeter-
appears to be a subgroup of patients prone to neck mined. Early work on type 2 endoleak discovered a rather
348 Endoleak complicating EVAR

benign course.31 Indeed, the EUROSTAR database, the patients with persistent collaterals subsequently presented
largest accumulation of data currently available, revealed with sac rupture.34
no relation between type 2 endoleak and sac rupture, this In vivo pressure manometry in patients with type 2
being in contrast to types 1 and 3 endoleak. More recently endoleak has recorded systemic pressures and pulsatile
there have been worrying reports of aneurysm sac expan- waveforms.35 We cannot reliably identify which type 2
sion (Fig. 30.5) and even rupture.32,33 It is not surprising endoleaks result in aneurysm expansion or rupture; suffice
that type 2 endoleak is associated with adverse outcome in to say the reports of type 2 rupture have all been pre-
some patients. In patients with aneurysms treated by surgi- dictable events.
cal ligation and bypass a 2 per cent aneurysm sac patency
via ‘type 2 endoleaks’ was recorded despite intraoperative Type 3 endoleak
ligation of collateral vessels. Almost a quarter of the
Type 3 endoleak36 is primarily a device related complica-
tion. It has a well documented association with aneurysm
rupture. It encompasses any form of graft disruption lead-
ing to leak of blood in to the aneurysm sac and includes
graft tears and iliac limb disconnection (modular EVGs).
Many of these problems have been attributed to first and
second generation EVGs. The underlying problems have
been investigated and corrected. Two particular EVGs have
been implicated in type 3 endoleak, the Stentor (Mintec,
LaCiotat, France) and the Vanguard (Boston Scientific,
Natick, MA, USA). The former experienced endoleak due
to seam defects and the latter due to metallic wear of the
fabric.
An interesting observation has been made by the
Liverpool group in patients with successful aneurysm
exclusion. They noted a number of graft kinks, precipitat-
ing type 3 endoleak in modular grafts, which they were
unable to attribute to graft migration (see Chapter 31).
They attributed this complication to longitudinal shrink-
age of the aneurysm sac. This observation highlights the
requirement for ongoing graft surveillance to ensure
patency of the graft.37 Of course graft migration without
Figure 30.4 Intraoperative sacogram revealing patent aortic
side branch, a previously unrecognised left accessory renal artery longitudinal sac shrinkage may precipitate either type 1 or

(a) (b)

Figure 30.5 Neck of aneurysm: (a) at time of EVAR, and (b) showing dilation post endografting
 Diagnosis 349

type 3 endoleak. Modular grafts clearly remain at higher mean arterial pressure in the sac equals that in the body.
risk of developing type 3 endoleak when compared with This oscillation of blood can be identified on duplex scan-
their unitary counterparts. ning with a good ‘window’ but is missed by computed
tomographic angiography (CTA).
In general the importance of endoleak has moved away
Type 4 endoleak
from flow and concentrated on pressure. This has resulted
from the realisation that thrombus may transmit pressure
Type 4 endoleak36 is seen as an ‘angiographic blush’ on
just as well as blood.
completion angiography. The placement of ultrathin
endovascular grafts, in part conceived to reduce sheath
diameter and employ a percutaneous introduction, were
DIAGNOSIS
associated with this observation. The AneuRx graft allowed
leakage of contrast at suture holes and from fabric interstices.
The leaks have all been reported to be sealed at 1 month. Clinical
Type 4 endoleak appears to be a self-limiting leak with, as
yet, no recorded adverse sequelae.38 The long term durabil- Endoleak is invariably asymptomatic. Patients experience
ity of thin walled porous grafts remains undetermined. symptoms only when the endoleak is associated with sud-
den aneurysm expansion or rupture. As endoleak was first
described and essentially remains a radiological diagnosis,
Type 5 (endotension)
little work has been done to assess its relation with clinical
examination. Clinical examination hinges upon the detec-
Type 5 (endotension) was first introduced as a concept
tion of pulsatile wall motion of the sac, the absence of
because a number of aneurysm sacs were undergoing expan-
which, it is hoped, indicates successful aneurysm exclusion.
sion without evidence of endoleak.39,40 In essence pressure
Greenberg and Green have incorporated physical examin-
is transmitted to the aneurysm sac in the absence of
ation in their EVAR follow-up protocol.45 Convincing evi-
detectable blood flow. There has been no clear explanation
dence of its usefulness in the diagnosis and management of
for this phenomenon, however, it is thought that inter-
patients with endoleak is awaited.
mittent endoleak,41 missed endoleak42 or thrombosed
endoleak are responsible either singly or in combination.
Recent evidence that thrombus may transmit pressure Investigations
has been presented in a variety of both experimental43 and
clinical reports.44 This property of thrombus has been recog- The purpose of any investigation with respect to endoleak
nised in open surgery for many years (see Fig. 30.1 showing is to:
rupture of a thrombosed abdominal aortic aneurysm).
Experimental data have, however, explained why type 1 • identify the presence of an endoleak
endoleaks remain dangerous even when thrombosed • classify the endoleak (1–4)
whereas a thrombosed type 2 endoleak rarely causes prob- • identify the anatomical site of the endoleak
lems. Pressure transmission is greater by short thrombosed • identify endotension/identify successful aneurysm
exclusion (shrinking of aneurysm sac, intrasac
endoleak channels with a large cross-sectional area than it
pressure reduction, reduced pulsatile wall motion).
is by long ones with a smaller cross-sectional area,45 whereas
the pressure transmission of a patent endoleak channel is Management of endoleak is dependent upon answers to
unaffected by either cross-sectional area or length.13 These these investigations. Unfortunately, no single currently avail-
results may explain why coil embolisation of endoleak able investigation has the ability to fulfil all these criteria.
channels does not always effectively reduce pressure trans- Plain abdominal X-ray (AXR) has little place in the
mission despite successful thrombosis.14 armamentarium for investigation of endoleak. It is a useful
tool for the detection of stent fracture, disruption or
migration (see Chapter 31). These signs may infer the pres-
Three grades of endotension ence of an endoleak. The detection of migration with AXR
is probably less accurate than computed tomography (CT)
• Grade 1 – high flow due to the problems of parallax error. Lateral films help to
• Grade 2 – low flow reduce this problem.
• Grade 3 – ‘no flow – no detectable’ endoleak Spiral CTA (SCTA) remains the gold standard for detec-
tion of endoleak. It is minimally invasive but, importantly,
is limited by its inability to reliably distinguish the type of
Endotension may also be due to an endoleak with no out- endoleak. The use of delayed acquisition scans may increase
flow where the blood is allowed in during systole and then the identification of endoleak by up to 11 per cent.46 The
will oscillate in the endoleak channel at the point where the optimum technique of SCTA to diagnose endoleak has yet
350 Endoleak complicating EVAR

to be determined. Current problems with SCTA are related Measurement of long term intrasac pressure appears to
to blood flow. Low flow endoleaks or thrombi are not be the holy grail of successful endovascular aneurysm
detected on current imaging but, significantly, maintain the exclusion. One-off and short duration postoperative
ability to transmit pressure.47 The use of ‘blood pool’ con- manometry has demonstrated quite clearly that successful
trast agents and platelets labelled with radioisotope may EVAR is associated with a significant reduction of intrasac
offer further assistance in the detection of low flow endoleak pressure.51 The presence of endoleak has been associ-
or fresh thrombus transmitting pressure to the sac. ated with raised intrasac pressure.35 Permanent intrasac
Worries over the invasiveness, cost and dose of radiation manometry appears desirable and in itself will raise many
during follow-up with angiography and CT have increased questions. These are likely to include the relevance of the
the demands for other techniques. Ultrasound is a technique site of pressure readings from within the sac (there is evi-
gaining popularity for its use in EVG surveillance. Duplex dence that an aneurysm sac may be compartmentalised
ultrasonography is observer dependent and may be tech- so that a decrease in pressure in one part of the sac does
nically difficult, especially in the early postoperative period not necessarily equal complete depressurisation of the
where the aorta is often obscured by bowel gas. The tech- entire sac42), the absolute level of pressure reduction
nique is able to identify the anatomical site of endoleak but required to prevent rupture and the significance of the
critically requires flow to determine the presence of an pressure waveform and the level of pressure transmitted
endoleak. Its use as a diagnostic tool may be enhanced with from the endoleak channel to the wall of the aneurysm.
intravascular ultrasound contrast agents.48 Importantly, the pressure readings will not tell us which
Digital subtraction angiography (DSA) is frequently used sacs are going to rupture because all sacs are not the same.
to identify the type and anatomical site of an endoleak. It Experience from the management of abdominal aortic
is usually a second line investigation due to its invasive aneurysms has revealed that some will rupture at 5 cm
nature. Selective views allow assessment of a variety of whereas others will not do so until they reach 9 cm. More
anatomical sites including the lumbar and inferior mesen- confusingly still, Baum et al. measured systemic pressures
teric arteries. The technique can differentiate inflow and which were transmitted to the wall of the sac in a number
outflow tracts and may permit therapeutic procedures such of type 2 endoleaks.35 Despite this evidence the majority of
as embolisation. Digital subtraction angiography is fre- type 2 endoleaks generally follow a benign course. The
quently used intraoperatively at completion of endograft development of a chronic intrasac pressure transducer is
deployment, i.e. completion angiography, where it remains awaited.
a useful tool in identifying primary endoleak. Detection of Current methods of imaging are severely limited by the
patent side branches can be facilitated by waiting for time delay between scans. Interval ruptures may be pre-
delayed filling or by using a ‘sacogram’ achieved by inject- ventable if an early warning sign is available to allow timely
ing contrast into the aneurysm sac post deployment. intervention of endoleak. Intrasac manometry with readily
Magnetic resonance angiography has proved effective as available or continuous pressure recordings is certainly an
a sole imaging modality for the assessment of the suitabil- attractive proposition.
ity of the aorta for endografting. Its role in the investiga-
tion of endoleak appears attractive but is restricted to
stent-grafts containing non-ferrous metals. MANAGEMENT
The detection of endotension usefully predicts those
patients who will require intervention to prevent abdom-
Management strategies are outlined in Table 30.2. It is gen-
inal aortic aneurysm rupture. Endoleak may remain
erally accepted that type 1 and 3 endoleak require inter-
undetected despite the presence of endotension. Signs of
vention to prevent rupture. The same cannot be said of
endotension include an increasing sac diameter or volume,
type 2 endoleak. It is not yet clear which type 2 endoleaks
pulsatile wall motion and raised intrasac pressure. Both
require intervention to prevent rupture. Consequently, the
duplex and CTA are able to reproduce aneurysm sac diam-
EUROSTAR database found type 2 endoleak a risk factor
eter. More recently, however, sac volume has been found
for secondary intervention but not rupture.52
to be a more accurate predictor of successful aneurysm
exclusion.49 Unfortunately, this technique remains rather
cumbersome for everyday practice. Table 30.2 Management strategies for endoleaks
The effect of endoleak on the aneurysm wall can be inves-
tigated in a dynamic fashion by ultrasound. Ultrasound Type 1 Urgent treatment
tracking documents aneurysm wall motion. Malina et al. Type 2 Observation unless associated with aneurysm
found the absence of pulsatile wall motion to be an indicator expansion
of successful aneurysm exclusion.50 Interestingly, in Resnikoff Type 3 Urgent treatment
Type 4 Conservative
et al.’s experience of open aneurysm exclusion and bypass,
Type 5 Try to identify cause. If no cause found consider
all aneurysms which subsequently ruptured exhibited pul- open conversion
satile wall motion on echo-tracking ultrasound.34
 Management 351

Prevention of endoleak however, to apply it on a more selective basis where

the sacogram reveals patent side branches. A negative
Primary, type 1, 3 and 4 endoleaks can be prevented by sacogram has been found to be a useful predictor. Type 2
appropriate case selection, graft oversizing and accurate endoleak development is unlikely in patients who have had
deployment. The endoleak rate increases markedly when a negative sacogram.30
operating on patients with adverse aneurysm morph-
ology.53 Uncovered suprarenal stent fixation and grafts Treatment of endoleak
with increased columnar strength may prevent migration
and subsequent endoleak but the evidence for their efficacy Management of endoleak requires a full medical assess-
is awaited in the long term (Fig. 30.6). ment of the patient. A medically high risk, frail patient may
Some institutions elect to embolise patent aortic side be better served by non-operative management. That said,
branches preoperatively. Embolisation is usually performed the majority of interventions can be performed by the
at the preoperative angiogram with insertion of metallic less invasive endovascular route, open treatment being
coils to induce vessel thrombosis and subsequent fibrosis. reserved for cases refractory to endovascular methods.55
As already alluded to, there really is no evidence to support Intraoperative management of type 1 endoleak is widely
this manoeuvre and it is not performed in our unit.54 The believed to be necessary to prevent early complications,
procedure of intraoperative filling of the aneurysm sac including rupture56 (Table 30.3). A number of the com-
with thrombogenic sponge, as it is performed in our unit, mercially available self-expanding stent-grafts are supplied
may be undertaken in all cases (Fig. 30.7). It is preferable, with a low pressure balloon in order to ‘mould’ the graft to
the wall and allow it to conform to the aorta. If the low
pressure balloon fails then an angioplasty balloon may
offer some benefit.

(a)

Figure 30.7 Aneurysm sac packing using polyvinylalcohol

sponge (Netcell, Skipton, UK) for the prevention of type 2 endoleak

Table 30.3 Management of type 1 endoleak

Cause Endovascular/open treatment

EVG landed too low/migration Extension piece

Poor EVG/aortic wall interface Balloon angioplasty/Palmaz
stent
Angulated neck Palmaz stent (straighten aortic
(b)
neck)
Dilated neck Periaortic ligature

Figure 30.6 (a) Suprarenal and (b) infrarenal stent fixation EVG, endovascular graft.
352 Endoleak complicating EVAR

Table 30.4 Management of type 2 endoleaks graft. If an occluding device fails the best policy is to ligate
the common iliac artery or the external and internal iliac
Preoperative Coil embolisation, laparoscopic ligation
arteries separately.
Intraoperative Aneurysm sac packing (thrombogenic
sponge), coil embolisation, laparoscopic Open transperitoneal procedures for the treatment of
ligation endoleak vary from ligation of side branch vessels to con-
Postoperative Coil embolisation, laparoscopic ligation, open version to traditional aneurysmorrhaphy.52 They have a
ligation or conversion role, particularly where less invasive techniques have failed.
In the case of endovascular embolisation procedures this is
usually due to access failure or the presence of multiple
A stent deployed too low in the proximal aortic neck is endoleak channels. Although an open ligation of side
rectified by the insertion of a covered extension graft. branch endoleak appears particularly invasive it is not
Deployment as close to the renals as possible is desirable. nearly so stressful for the patient as a conventional open
For patients with difficult aortic necks, especially those that repair, the latter requiring prolonged clamping of the aorta
are angulated, the giant Palmaz stent is an ideal adjunct. with its attendant sequelae. In some patients with type 2
These stainless steel balloon expandable stents straighten endoleak open methods are reassuring in their exclusion of
out angulated aortic necks and improve the seal at the other causes of endoleak, for example, a proximal type 1
stent-graft/aortic wall interface. Failure of the Palmaz endoleak providing inflow to the aneurysmal sac and a type
stent to eliminate an endoleak requires the placement of 2 endoleak the outflow channel.
periaortic ligatures.57 The periaortic ligatures are applied at The techniques for open conversion have been reviewed
mini-laparotomy or laparoscopy following dissection of thoroughly by May et al.60 Patients in whom the device has
the aortic neck. Placement of the periaortic ligatures is migrated entirely into the aneurysm sac may be treated by
facilitated by the Palmaz stent. The rigid stent provides conventional aneurysm repair. In patients who have had
a firm structure on which to ‘snug’ the nylon periaortic the device deployed suboptimally, the technique required
tapes, in so doing preventing stenosis or occlusion of the is a modified conventional repair suturing the top of the
aorta or kinking of the endograft. graft to the aneurysm neck. In addition to proximal and
A variety of methods have been used for the treatment distal clamping the graft is clamped in order to reduce
of type 2 endoleak (Table 30.4). Endovascular embolisa- blood loss and facilitate removal of proximal and distal
tion with metallic coils or other embolic agents including stents from the infrarenal aorta and iliac arteries, respec-
liquids has been described.58 There are concerns about the tively. Surprisingly little damage is encountered in these
use of liquid embolic agents, especially when treating lumbar vessels post-endografting.
arteries because of their proximity to the spinal circulation. Patients who have had a suprarenal endograft that needs
Embolisation of endoleak channels has also been per- to be removed, usually require supracoeliac clamping or the
formed via a direct translumbar approach and in some use of a large (30 mm) angioplasty occlusion balloon placed
cases has been combined with embolisation of the aneurysm in the aorta above the renal arteries. At secondary conversion
sac itself. There have even been suggestions to fill the the metal suprarenal frame is usually left in situ because it is
aneurysm sac with a non-absorbable medium, creating a frequently well incorporated. The suprarenal component
permanent solid sac. The direct approach is useful where is incorporated into the anastomosis.
endovascular access is difficult or there are fears regarding the
disturbance of an endograft iliac limb with wires or catheters.
Opponents of ‘open’ techniques warn of the dangers of Treatment of complications of endoleak
introducing infection and the potential bleeding complica- (rupture)
tions subsequent to creating a hole in an aneurysm sac
which may be under systemic pulsatile pressure. Aneurysm rupture is the complication of endoleak. The
Laparoscopic ligation of vessels is an attractive method feasibility of endovascular repair of ruptured abdominal
of treating type 2 endoleak. The aorta may be approached aortic aneurysm has been demonstrated in a number of
either by the transperitoneal or the retroperitoneal route. It institutions. Endovascular repair of rupture post endo-
is an intervention best undertaken at the same sitting as grafting is possible but clearly depends upon the cause of
EVAR or in the early postoperative period. The inflamma- the rupture and the patient’s fitness and stability to tolerate
tory reaction which tends to develop in the perioperative a preoperative delay for investigations. A great many of the
period often makes periaortic dissection and identification ruptures experienced in the early years of endografting
of feeding vessel difficult.59 required open repair. Frequently no endoleak could be
Endoleak from the iliac limbs or occluding devices in identified at laparotomy. A number of risk factors have
the case of aorto-uni-iliac prostheses, is usually managed been identified for rupture. If the cause of the leak/rupture
best by open methods via an extraperitoneal approach. can be identified preoperatively then endovascular treat-
Extension cuffs can, of course, be used if the iliac limb has ment may salvage the situation. If the anatomical site of the
landed short or if there is a type 3 endoleak in a modular leak cannot be identified then open repair is required.
 Results 353

1800s and published by Peacock and Brown in 1968.66

Identified risk factors for rupture52 Many are sceptical of the efficacy of this technique.
Similarly, embolisation for type 1 and type 3 endoleak has
• Type 1 endoleak
proved disappointing. An analysis of experimental data
• Type 2 endoleak
shows why type 1 and 3 endoleaks do not appear to be suc-
• Endograft migration
cessfully treated using embolisation techniques, whereas
• Postoperative endograft kinking
their type 2 counterparts do.43 Both type 1 and 3 endoleaks
tend to be of large cross-sectional area and short length,
and therefore even when thrombosed, either naturally or
POSTOPERATIVE CARE induced by embolic material, they may be expected to
transmit a significant level of pressure. Type 2 endoleaks
The majority of secondary endovascular procedures can be are usually longer and of smaller cross-sectional area and
performed under local anaesthesia. Where instrumenta- consequently dampen pressure transmission.
tion of the iliac arteries or aorta is required and large
sheaths are used then edpidural or spinal anaesthesia with
sedation or general anaesthesia is more comfortable for the Conclusions
patient and reduces movement artefact. The recovery time
for elective cases is short and major complications are rare. Types 1 and 3 endoleak should be regarded as dangerous
and accordingly managed with a degree of urgency. If
these lesions are recognised intraoperatively then they
RESULTS are best treated at the same sitting. The outcome of type
2 endoleaks is generally benign although some are asso-
ciated with aneurysm expansion. There is insufficient
Endovascular aneurysm surgery is in its infancy and there-
evidence to suggest that treatment of type 2 endoleaks is
fore there are very few long term follow-up results. Many
required to prevent rupture. Those endoleaks resulting
of the early endoleaks occurred when the requirements for
in aneurysm expansion require intervention. Further
successful endografting were not understood and the
investigation is required to predict the type 2 endoleaks
devices were first and second generation.
which will result in aneurysm expansion.
It was initially suggested by May et al. that a stent-graft
Early endoleak can be prevented by accurate assess-
may confer some benefit to the patient if sac rupture
ment of aneurysm morphology, appropriate patient
occurs,61 an effect presumed to be due to the attenuated
selection and precise endograft deployment. Whereas
haemodynamic disturbance. A more recent and contradict-
early endoleak is now largely avoidable, the problems of
ory report from a larger experience compiled by the
late endoleak remain. Late endoleak occurs as a result of
EUROSTAR database found the mortality to be 50 per cent
a previously unrecognised or missed endoleak, graft fail-
(similar to that of open repair of de novo rupture).62 Open
ure, changing aneurysm morphology and endograft
conversion carries an 18 per cent mortality when performed
migration. Improved graft durability will hopefully
primarily and 27 per cent at secondary conversion. Primary
reduce the incidence of type 3 endoleak and improved
conversion is usually performed for access failure or device
graft design, including more secure fixation, may prevent
migration. Secondary conversion is generally required for
graft migration and type 1 endoleak. Changing aneurysm
rupture or persistent endoleak.52 These high mortality rates
morphology, and neck dilatation in particular, is likely to
highlight the importance of endoleak prevention and early
require alternative techniques to ensure adequate seal
endovascular intervention to prevent open conversion.
and fixation of the proximal endograft. New technology
The use of Palmaz stents and periaortic ligatures has
such as an endovascular staple is in development.
been confined to experimental models, case reports and
Our ability to detect endoleaks is largely dependent
small series with relatively short duration of follow-up.63,64
upon significant blood flow. Low flow endoleaks and
They are certainly feasible techniques and appear effica-
the pressure transmitted by thrombus are more difficult
cious in the short term although their effect on device
to detect. Current methods of investigation can only
durability is unknown.
reliably detect the secondary effects of raised pressure on
It has been claimed that coil embolisation of endoleak
the sac, i.e. an expanding aneurysm sac or pulsatile wall
channels may not effectively seal lower intra-aneurysmal
motion. The introduction of long term pressure tele-
pressure,14 whereas as others have reported successful occlu-
metry and more sophisticated imaging techniques offer
sion of aortic aneurysms using this technique. One of these
some promise in the detection of these types of
reports has claimed to reduce the pressure to zero with coils
endoleak. It is hoped that these techniques will allow a
before tying the iliacs and doing an extra-anatomical
more rational approach in the follow-up of EVAR and
bypass for high risk aneurysms.65 This is not an original
act as early warning signs of aneurysm exclusion failure.
approach having been first performed by Moore in the
354 Endoleak complicating EVAR

6 Matsumara JS, Moore WS. Clinical consequences of

Most endoleaks can be treated successfully by periprosthetic leak after endovascular repair of abdominal aortic
endovascular methods. Open methods may be used but aneurysm. J Vasc Surg 1998; 27: 606–13.
are less favourable. A good proportion of the first and 7 Gilling-Smith GL, Martin J, Sudhindran S, et al. Freedom from
second generation grafts have developed endoleaks. endoleak after endovascular aneurysm repair does not equal
Many of these endoleaks will require secondary inter- treatment success. Eur J Vasc Endovasc Surg 2000; 19: 621–5.
ventions. To reduce endoleaks and the number of 8 Lee WA, Yehuda GW, Fogarty TJ, et al. Does complete aneurysm
secondary interventions in future there must be an exclusion ensure long-term success after endovascular repair?
improvement in EVG design, a greater understanding of J Endovasc Ther 2000; 7: 494–500.
9 Rhee RY, Eskandari MK, Zajko AB, et al. Long-term fate of the
the principles required for successful aneurysm exclu-
aneurysmal sac after endoluminal exclusion of abdominal aortic
sion and a heightened awareness of the natural history of
aneurysms. J Vasc Surg 2000; 32: 689–96.
individual endoleaks. 10 Moore WS, Rutherford RB, for the EVT investigators. Transfemoral
endovascular repair of abdominal aortic aneurysm: results of the
North American phase I trial. J Vasc Surg 1996; 23: 543–53.
11 Yusuf SW, Whitaker SC, Chuter TA, et al. Early results of
Key references endovascular aortic aneurysm surgery with aortouniiliac graft,
contralateral iliac occlusion and femoro-femoral bypass. J Vasc
Gilling-Smith GL, Brennan J, Harris P, et al. Endotension after Surg 1997; 25: 165–72.
endovascular aneurysm repair: definition, classification, and 12 Blum U, Voshage G, Lammer J, et al. Endoluminal stent-grafts for
strategies for surveillance and intervention. J Endovasc Surg infrarenal abdominal aortic aneurysms. N Engl J Med 1997; 336:
1999; 6: 305–7. 13–20.
Mohan IV, Laheij RJ, Harris PL. EUROSTAR collaborators. Risk 13 Schurink GW, Aarts NJ, Van Baalen JM, et al. Experimental study
factors for endoleak and the evidence for stent-graft of the influence of endoleak size on pressure in the aneurysm sac
oversizing when undergoing endovascular aneurysm repair. and the consequences of thrombosis. Br J Surg 2000; 87: 71–8.
Eur J Vasc Endovasc Surg 2001; 21: 344–9. 14 Marty B, Sanchez LA, Ohki T, et al. Endoleak after endovascular
Schurink GW, Aarts NJ, Van Baalen JM, et al. Experimental study of graft repair of experimental aneurysms: does coil embolization
the influence of endoleak size on pressure in the aneurysm sac with angiographic ‘seal’ lower intraaneurysmal pressure? J Vasc
and the consequences of thrombosis. Br J Surg 2000; 87: Surg 1998; 27: 454–61.
71–8. 15 White GH, May J, Waugh RC, et al. Type 1 and type 2 endoleaks:
White GH, May J, Petrasek P, et al. Endotension: an explanation for a more useful classification for reporting results of endoluminal
continued AAA growth after successful endoluminal repair. AAA repair [letter]. J Endovasc Surg. 1998; 5: 189–91.
J Endovasc Surg 1999; 6: 308–15. 16 May J, White GH, Yu W, et al. Focus on the proximal neck: the
White GH, May J, Waugh RC, et al. Type 3 and Type 4 endoleak: key to durability of endoluminal AAA repair. J Endovasc Surg
toward a complete definition of blood flow in the sac after 1997; 4: 1–27.
endoluminal AAA repair. J Endovasc Surg 1998; 5: 305–9. 17 Petrik P, Moore WS. Endoleaks following endovascular repair
of abdominal aortic aneurysm: the predictive value of
preoperative anatomic factors – a review of 100 cases. J
Vasc Surg 2001; 33: 739–44.
18 Albertini JN, Macierewicz JA, Yusuf SW, et al. Pathophysiology of
REFERENCES
proximal perigraft endoleak following endovascular repair of
abdominal aortic aneurysms: a study using a flow model. Eur J
1 Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal graft Vasc Endovasc Surg 2001; 22: 53–6.
implantation for abdominal aortic aneurysms. Ann Vasc Surg 19 Albertini J, Kalliafas S, Travis, et al. Anatomical risk factors
1991; 5: 491–6. for proximal perigraft endoleak and graft migration following
2 Volodos NL, Karpovich IP, Troyan VL, et al. Clinical experience of endovascular repair of abdominal aortic aneurysms. Eur J Vasc
the use of self-fixing prostheses for remote endoprosthetics of the Endovasc Surg 2000; 19: 308–12.
thoracic and abdominal aorta and iliac arteries through the 20 Mohan IV, Laheij RJ, Harris PL, EUROSTAR collaborators. Risk
femoral artery and as intraoperative endoprosthesis for aorta factors for endoleak and the evidence for stent-graft oversizing
reconstruction. Vasa Suppl 1991; 33: 93–5. when undergoing endovascular aneurysm repair. Eur J Vasc
3 Torsello GB, Klenck E, Kasprzak B, Umscheid T. Rupture of Endovasc Surg 2001; 21: 344–9.
abdominal aortic aneurysm previously treated by endovascular 21 McArthur C, Teodorescu V, Eisen L, et al. Histopathologic analysis
stent-graft. J Vasc Surg 1998; 28: 184. of endovascular stent grafts from patients with aortic
4 Laheij RJF, Buth J, Harris PL, et al. Need for secondary aneurysms: Does healing occur ? J Vasc Surg 2001; 33: 733–8.
interventions after endovascular repair of abdominal aortic 22 Walker SR, Macierewicz J, Elmarasy NM, et al. A prospective
aneurysms. Intermediate-term follow-up results of a European study to assess changes in proximal aortic neck dimensions after
collaborative registry (EUROSTAR). Br J Surg 2000; 87: endovascular repair of abdominal aortic aneurysms. J Vasc Surg
1666–73. 1999; 29: 625–30.
5 White GH, Yu W, May J, et al. Endoleak as a complication of 23 Matsumara JS, Chaikof EL. Continued expansion of aortic necks
endoluminal grafting of abdominal aortic aneurysms. J Endovasc after endovascular repair of abdominal aortic aneurysms. J Vasc
Surg 1997; 4: 152. Surg 1998; 28: 422–31.
 References 355

24 Illig KA, Green RM, Ouriel K, et al. Fate of the proximal aortic 42 Wever JJ, Blankensteijn JD, Eikelboom BC. Secondary endoleak or
cuff: implications for endovascular aneurysm repair. J Vasc Surg missed endoleak? Eur J Vasc Surg 1999; 18: 458–60.
1998; 28: 184–7. 43 Mehta M, Ohki T, Veith FJ, Lipsitz EC. All sealed endoleaks are
25 Jacobowitz GR, Rosen RJ, Riles TS. The significance and not the same: a treatment strategy based on analysis. Eur
management of the leaking endograft. Semin Vasc Surg 1999; J Vasc Endovasc Surg 2001; 21: 541–4 .
12: 199–206. 44 Ruurda JP, Rijbroek A, Vermeulen EGJ. Continuing expansion of
26 Walker SR, Halliday K, Yusuf SW, et al. A study on the patency of internal iliac artery aneurysms after surgical exclusion of the
the inferior mesenteric and lumbar arteries in the incidence of inflow. J Cardiovasc Surg 2001; 42: 389–92.
endoleak following endovascular repair of infrarenal aortic 45 Greenberg R, Green R. A clinical perspective on the management
aneurysms. Clin Radiol 1998; 53: 593–5. of endoleaks after abdominal aortic aneurysm repair. J Vasc Surg
27 Velazquez OC, Baum RA, Carpenter JP, et al. Relationship 2000; 31: 836–7.
between preoperative patency of the inferior mesenteric artery 46 Golzarian J, Dussaussois L, Abada HT, et al. Helical CT of the
and subsequent occurrence of type 2 endoleak in patients aorta after endoluminal stent-graft therapy: value of biphasic
undergoing endovascular repair of abdominal aortic aneurysms. acquisition. Am J Roentgenol 1998; 171: 329–31.
J Vasc Surg 2000; 32: 777–88. 47 Schurink GWH, Aarts NJM, Wilde J, et al. Endoleakage after
28 Armon MP, Yusuf SW, Whitaker SC, et al. Thrombus distribution stent-graft treatment of abdominal aortic aneurysm:
and changes in aneurysm size following endovascular aortic Implications on pressure and imaging – an in vitro study.
aneurysm repair. Eur J Vasc Endovasc Surg 1998; 16: 472–6. J Vasc Surg 1998; 28: 234.
29 Kato N, Semba CP, Dake MD. Embolization of perigraft leaks 48 McWilliams R, Martin J, Gould D, et al. Levovist-enhanced
after endovascular stent-graft treatment of aortic aneurysms. ultrasound as the primary follow-up investigation after
J Vasc Intervent Radiol 1996; 7: 805–11. endovascular aneurysm repair. J Intervent Radiol 1998; 13:
30 Lehmann JM, Macierewicz JA, Davidson IR, et al. Prevention 146–7.
of side branch endoleaks with thrombogenic sponge: one year 49 Singh-Ranger R, McArthur T, Della Corte M, et al. The abdominal
follow-up. J Endovasc Ther 2000; 7: 431–3. aortic aneurysm sac after endoluminal exclusion: a medium-term
31 Resch T, Ivancev K, Lindh M, et al. Persistent collateral perfusion morphologic follow up based on volumetric technology. J Vasc
of abdominal aortic aneurysm after endovascular repair does not Surg 2000; 31: 490–500.
lead to progressive change in aneurysm diameter. J Vasc Surg 50 Malina M, Lanne T, Ivancev K, et al. Reduced pulsatile wall
1998; 28: 242–9. motion of abdominal aortic aneurysms after endovascular repair.
32 Arko FR, Rubin GD, Johnson BL, et al. Type 2 endoleaks following J Vasc Surg 1998; 27: 624–31.
endovascular repair: preoperative predictors and long-term 51 Treharne GD, Loftus IM, Thompson MM, et al. Quality control
effects. International Congress 14 on Endovascular Interventions, during endovascular aneurysm repair: Monitoring aneurismal sac
New York, 2000. pressure and superficial femoral artery flow velocity.
33 Hinchliffe RJ, Singh-Ranger R, Davidson I, Hopkinson BR. J Endovasc Surg 1999; 6: 239–45.
Rupture of an abdominal aortic aneurysm sac secondary to type 52 Harris PL, Vallabhameni SR, Desgranges P. Incidence and risk
II endoleak. Eur J Vasc Endovasc Surg (in press). factors of late rupture, conversion, and death after endovascular
34 Resnikoff M, Darling RC 3rd, Chang BB, et al. Fate of the repair of infrarenal aortic aneurysms: the EUROSTAR experience.
excluded abdominal aortic aneurysm sac: long-term follow-up of European Collaborators on Stent-graft Techniques for aortic
831 patients. J Vasc Surg. 1996; 24: 851–5. Aneurysm Repair. J Vasc Surg 2000; 32: 739–49.
35 Baum RA, Carpenter JP, Cope C, et al. Aneurysm sac pressure 53 Lawrence-Brown MMD, Semmens JB, Hartley DE. The Zenith
measurements after endovascular repair of abdominal aortic endoluminal stent-graft system: suprarenal fixation, safety
aneurysms. J Vasc Surg 2000; 33: 32–40. features, modular components, fenestration and custom crafting.
36 White GH, May J, Waugh RC, et al. Type 3 and Type 4 Endoleak: In: Greenhalgh RM, Becquemin J-P, Davies A et al. (eds). Vascular
Toward a complete definition of blood flow in the sac after and Endovascular Surgical Techniques, 4th edn. London: WB
endoluminal AAA repair. J Endovasc Surg 1998; 5: 305–9. Saunders, 2001: 219–23.
37 Harris PL, Brennan J, Martin J, et al. Longitudinal shrinkage 54 Walker SR, Macierewicz JA, Hopkinson BR. Endovascular AAA
following endovascular aneurysm repair: a source of repair: prevention of side branch endoleaks with thrombogenic
intermediate and late complications. J Endovasc Surg sponge. J Endovasc Surg 1999; 6: 350–3.
1999; 6: 4. 55 May J, White G, Yu W, et al. Endovascular grafting for
38 Zarins CK, White RA, Schwarten DE, et al. AneuRx stent-graft abdominal aortic aneurysms: Changing incidence and indications
versus open repair of abdominal aortic aneurysms: multicentre for conversion to open operation. Cardiovasc Surg 1998; 6:
prospective clinical trial. J Vasc Surg 1999; 29: 292–308. 194–7.
39 White GH, May J, Petrasek P, et al. Endotension: an explanation 56 Harris PL. Management of endoleak and endotension. In:
for continued AAA growth after successful endoluminal repair. Greenhalgh RM, Becquemin J-P, Davies A et al. (eds). Vascular
J Endovasc Surg 1999; 6: 308–15. and Endovascular Surgical Techniques, 4th edn. London:
40 Torsello GB, Klenck E, Kasprzak B, Umscheid T. Rupture of WB Saunders, 2001: 265–70.
abdominal aortic aneurysm previously treated by endovascular 57 Kalliafas S, Albertini JN, Macierewicz J, et al. Incidence and
stent-graft. J Vasc Surg 1998; 28: 184. treatment of intraoperative technical problems during
41 Gilling-Smith GL, Brennan J, Harris P, et al. Endotension after endovascular repair of complex abdominal aortic aneurysms.
endovascular aneurysm repair: definition, classification, and J Vasc Surg 2000; 31: 1185–92.
strategies for surveillance and intervention. J Endovasc Surg 58 Marin ML, Dolmatch BL, Fry PD, et al. Treatment of type II
1999; 6: 305–7. endoleaks with onyx. J Vasc Intervent Radiol 2001; 12: 629–32.
356 Endoleak complicating EVAR

59 Edoga JK. Laparoscopic treatment for endoleak resolution 63 Sonesson B, Montgomery A, Invancev K, Lindblad B. Fixation of
following endoprosthesis exclusion of AAA. Presented to the infra-renal aortic stent-grafts using laparoscopic banding – an
International Workshop ‘New Technologies in Vascular Surgery’, experimental study. Eur J Vasc Endovasc Surg 2001; 21: 40–5.
Rome, 27 June 2001. 64 Hopkinson BR. Methods for dealing with difficult aortic necks:
60 May J, White GH, Harris JP. Techniques for surgical conversion of short, wide, conical or angulated. The 27th Global Symposium.
aortic endoprosthesis. Eur J Vasc Endovasc Surg 1999; 18: 284–9. Vascular and Endovascular Issues, Techniques and Horizons.
61 May J, White GH, Waugh R, et al. Rupture of abdominal aortic New York, 17 November 2000.
aneurysms: a concurrent comparison of outcome of those 65 Huber KL, Joseph A, Mukherjee D. Extra-anatomic arterial
occurring after endoluminal repair versus those occurring reconstruction of common iliac arteries and embolization
de novo. Eur J Vasc Endovasc Surg 1999; 18: 344–8. of the aneurysm sac for the treatment of abdominal aortic
62 Cuypers PWM, Laheij RJF, Buth J, on behalf of the EUROSTAR aneurysms in high risk patients. J Vasc Surg 2001; 33:
collaborators. Which factors increase the risk of conversion to 745–51.
open surgery following endovascular abdominal aortic aneurysm 66 Peacock JH, Brown GJ. Wiring of abdominal aortic aneurysms.
repair ? Eur J Vasc Endovasc Surg 2000; 20: 183–9. Br J Surg 1968; 55: 344–6.
 31
Graft Breakdown and Migration Complicating
EVAR

S RAO VALLABHANENI, PETER L HARRIS

Introduction 357 Management 361

Aetiology/pathophysiology 357 Endovascular secondary interventions 362
The problems and their incidence 359 References 364
Diagnosis 360

INTRODUCTION AETIOLOGY/PATHOPHYSIOLOGY

Endovascular aneurysm repair (EVAR) using stent-graft Mechanism of stent-graft function

systems that could be deployed through femoral arteries
was introduced in 1991. Although the early results were Devices which can be delivered by a transfemoral route
encouraging, late follow-up has been characterised by a high to exclude aortic aneurysms, avoiding all transabdominal
rate of secondary intervention, reaching a cumulative annual manoeuvres, call for design features that differ radically from
incidence of nearly 10 per cent for any intervention and a simple fabric tube fixed by conventional sutured anastomo-
approximately 3 per cent for late conversion.1,2 Despite close sis at open repair.4 A brief overview of the stent-graft design
surveillance and such high rates of secondary intervention, and function relevant to stent migration and disintegration is
late rupture has been observed at an annual rate of 1 per cent.2 presented here to facilitate understanding of these modes of
Although correction of an endoleak, most often type 2, has failure. Stent-grafts require not only a fabric tube to isolate the
been the commonest indication for a secondary intervention, aneurysm sac from the blood flow but also some form of
a recent EUROSTAR analysis revealed no significant associa- frame or stent to effect a ‘seal’ and fix the device in place.
tion between isolated type 2 endoleak and late rupture.2,3 The Sutures have been used to attach the stent rings to each other
important and significant risk factors leading to late rupture and to the fabric in early devices. More recent models incor-
were migration of stent-graft leading to a type I endoleak and porate the metal rings within the layers of fabric. In addition,
device disintegration leading to a type 3 endoleak, both signi- a system of radio-opaque markers is required to facilitate
fying device failure.3 Therefore the surveillance and second- visualisation and orientation of the device during deployment
ary interventions should be aimed at ensuring stent-graft and at follow-up. The entire device is crimped into a delivery
stability and integrity. Endoleak, as a consequence of device mechanism which is narrow and flexible enough to traverse
failure (secondary type 1 and type 3), is a potentially serious the iliofemoral segment. Thin Dacron has been used to facili-
finding and is a late event (see Chapter 30). tate smaller profile construction but it is more susceptible to
Although the durability of endovascular devices is wear than standard thickness fabric. Polytetrafluoroethylene
invariably subjected to rigorous testing prior to release for (PTFE), the other fabric used in stent-graft construction,
clinical use, some of the mechanisms of EVAR failure have appears to be more resistant to wear.
come to light only with long term follow-up of early recipi- Adequate apposition between stent fabric and native
ents. It is possible that some potential mechanisms for vessel wall, at the neck of aneurysm and in iliac segments,
device failure remain as yet unrecognised. This chapter is necessary to seal these junctions or ‘anastomoses’ against
addresses the clinical issues associated with device migration type 1 endoleak. This is largely dependent on radial forces
and disintegration related to endovascular aneurysm repair. exerted by the stent. Such radial forces are brought into
358 Graft breakdown and migration complicating EVAR

action by balloon expansion (e.g. Lifepath, Edwards Life Stent-graft disintegration

Sciences, Irvine, CA, USA) or spring-expansion (e.g. Zenith,
Cook, Bloomington, IN, USA) or as a mechanical function The approximately 30 000 000 pulsations a year to which the
of ‘memory’ alloys from which the stent is constructed devices are subjected increase the risk of stress fracture of
(e.g. Talent, World Medical Corp). the suture material bonding the stent frames together or to
Once deployed, the columnar strength of the device the fabric. Suture breakage may result in wire frame separa-
assists the radial forces in stabilising the stent-graft in its tion and dislocation, which, in turn, may lead to distortion.
position and resists its movement. In some devices, fixation Consequently the columnar strength is compromised and in
may be enhanced by the addition of hooks and barbs to the severe cases the basic integrity of the device may be lost.
proximal or distal stent-graft enabling it to anchor on to the Nitinol from which the wire frames of some devices are
neck of aneurysm or iliac arteries.5 Natural healing does not constructed has been shown to be susceptible to corrosion
contribute significantly to fixation and stent-grafts therefore by a host reaction leading to weakening and fracture of
rely permanently upon their mechanical properties for their stents.9 It is probable that other alloys are similarly vulner-
stability and integrity.6 A section of uncovered metal stent able. Wire frames are also susceptible to stress fracture.
on the top end of the device may assist fixation by increasing This not only compromises the strength of the device but
the frictional forces. One device (Zenith, Cook) incorp- the sharp ends of broken wire can also come into contact
orates a proximal uncovered stent with hooks intended to with the fabric. Pulsatile movement can then result in
cross the renal artery ostia and engage into the relatively erosion of fabric resulting in holes or tears with a risk of
healthy suprarenal aorta for additional fixation. sudden reperfusion of the aneurysm sac (type 3 endoleak).
The columnar strength of a device depends on its Integrity of the flow channel of the stent-graft depends on
construction. Some devices (e.g. AneuRx, Medtronic, adequate fixation between the modular components. Loss
Minneapolis, MN, USA) comprise metal rings which artic- of structural integrity and distortion may cause dislocation
ulate with each other throughout its height providing high of the contralateral limb from the main stent-graft precipi-
columnar strength. Devices with stents at fixation sites tating sudden reperfusion of the aneurysm sac.
only, with the fabric tube unsupported in the middle, have An aneurysm that has been excluded successfully shrinks
negligible columnar strength and depend entirely on radial in size not only in its anteroposterior (AP) and lateral diam-
forces and hooks for fixation. Other devices contain par- eters but also longitudinally. This may lead to kinking and
tially articulating struts or vertical wire frames to provide buckling of the stent-graft which could lead to dislocation
columnar strength of intermediate and variable degree. of limb engagement or modular separation.10 In addition,
Most devices are bifurcated but straight tube designs stent-graft distortion alters the spatial relationship between
intended for endoluminal aorto-uni-iliac use are available. the components of the device. This, in turn, may exagger-
Bifurcated design is preferred since aorto-uni-iliac repair ate movement between them causing tears in the fabric,
requires an additional femoro-femoral bypass. Some bifur- suture breakage or fracture of stent struts.7
cated models are available as a single piece. The more fre-
quently used modular construction allows customisation
of the device to utilise the anatomy fully, to extend applic- Stent migration
ability of EVAR to a wider range of aneurysms and in some
cases to facilitate accurate deployment. Modular systems The infrarenal neck of some aortic aneurysms increases in
are made up of two or three components which can be diameter at the landing zone of the stent-graft at a rate
docked and deployed to construct a bifurcated device approximating 1 mm per year following stent-graft implant-
in vivo. A major disadvantage of modular devices is the ation.11,12 It is usual to ‘oversize’ the diameter of the device
potential for disjunction of the modular parts. relative to that of aneurysm neck by at least 10 per cent or
3–4 mm. Therefore, one may expect to see compromise of
proximal fixation and seal resulting in proximal type 1
Device failure endoleaks and stent migration 3–4 years after implant-
ation.12 Aneurysmal changes in common iliac arteries result-
Materials that are known to be safe for implantation into ing in widening and shortening could result in the loss of
humans have been adopted for the fabrication of stent-grafts. distal fixation (Irfan Raza, unpublished data).
The physiological conditions within an aortic aneurysm, The physiological haemodynamic forces of the aorta gen-
however, such as the constant pulsation, the haemody- erate vectors due to flow channel geometry.13,14 The cephalo-
namic forces, host reaction, etc., have proved to be detri- caudal narrowing of the flow channel leads to a ‘wind sock
mental to the structural integrity of the device, upon which effect’, the force attempting to cause distal migration of the
the state of the aneurysm repair depends.7 Additionally, entire device. The curvature in the stent-graft elicits a ‘hose
the morphology of the aneurysm sac changes following pipe effect’ (the force that causes an open fire hose to move
implantation of an endograft, with potential adverse impact vigorously side to side like a snake), which encourages prox-
upon both the fixation and structure of the endograft.8 imal dislocation of the iliac limb engagement and distal
 The problems and their incidence 359

100 1766
872

Freedom from migration (per cent)

98
94 386
Caudad migration 96
90
Type 1a endoleak  neck dilation 91
88

Type 3a endoleak 80
Type 3b endoleak
Fabric hole
Modular 70
disconnection
Aneurysmal 60
change in iliacs
Iliac limb 50
dislocation 0 12 24 36 48 60 66
Type 1b endoleak Months

Figure 31.2 Kaplan–Meier graph of freedom from stent-graft

migration. The numbers above line are patients at risk and those
Figure 31.1 Schematic diagram of some possible modes of below the line are the percentage free from migration
stent-graft failure

Freedom from type 3 endoleak (per cent)

100 1743
874
96.3 386
93.7
migration of the proximal engagement at the neck. The 90
134
90.6
haemodynamic forces are of significant magnitude and are
84.9 36
exaggerated by curvature, angulation or kinking of the 80
device, all of which may worsen with aneurysm shrinkage 78.4
and play an important role in promoting device migration. 70
Figure 31.1 is a diagrammatic representation of some
modes of failure of endoaortic stent-grafts. The design of
60
these devices is under constant review and later gener-
ations of devices incorporate features intended to address
these modes of failure. 50
0 12 24 36 48 60 72 78
Months

THE PROBLEMS AND THEIR INCIDENCE Figure 31.3 Kaplan–Meier graph of freedom from secondary
type 3 endoleak. The numbers above line are patients at risk and
those below the line are the percentage free from migration
Stent-graft migration and disintegration are observed
throughout the follow-up. The Kaplan–Meier curve of the
incidence of migration in Figure 31.2 is derived from an
serious disintegration, shows a steady incidence affecting
analysis of 3264 patients from EUROSTAR with a mean
21.6 per cent of the cohort by 5 years (Figure 31.3).
and median follow-up of approximately 1 year. The inci-
dence of stent-graft migration can be noted to be persistent
without signs of reaching a plateau at 4 years and by then Consequences of device failure
affecting nearly 12 per cent of the cohort. The mean time
after EVAR to the diagnosis of migration is approximately Stent migration or disintegration may not show any imme-
20 months.13 diate effects and may lead to complacency in dealing with
Stent-graft disintegration is a progressive phenomenon these failures in their early stages. Haemodynamic isol-
with its effects once again noted throughout the late follow- ation of the aneurysm sac, which is the aim of stent-graft
up of all models. Suture breakage, detected on plain X-rays repair, should aim at isolating the sac not only from blood
as alteration in the alignment of stent struts, is a common flow but also from systemic blood pressure. Endotension,
and universally noted finding affecting the earliest, now with- the phenomenon of pressurisation of the stent-grafted
drawn, models. Fracture of metal struts is also a common aneurysm sac, denotes a risk of rupture. Pressure can be
finding. Secondary mid-graft endoleaks due to stent-graft transmitted through thrombus15 and through fixation sites
disintegration (fabric hole or modular component dissoci- with inadequate overlap in the absence of a detectable
ation), a finding that can be used as a surrogate measure of endoleak. In general, a 10 mm overlap is considered the
360 Graft breakdown and migration complicating EVAR

essential minimum for an adequate seal. The consequences

of ‘silent’ pressurisation of the sac may not be detectable • Anatomical factors
– Aortic neck angulation
straight away by the routine means of surveillance but may
– Angulation of iliac artery
only appear as an enlarging sac over time.16,17 An enlarging
– Higher maximal aneurysm diameter
aneurysm sac with a migrating stent-graft and inadequate
– Higher terminal aortic transverse diameter
seal is at risk of precipitous graft related endoleak and ensu-
ing rupture. Even an aneurysm that has been shown to be
shrinking, however, may be at risk if there is evidence of
increasing distortion of the stent-graft and compromise of DIAGNOSIS
the seal due to stent-graft disintegration or migration.18
Endoleak due to migration or disintegration is a late find-
ing and indicates that the ideal time for intervention has Diagnosis of device failure in its early stages is dependent
already been missed. Poor fixation, catastrophic migration upon an adequate programme of postoperative surveillance.
and disintegration of the stent-graft have been incriminated
in reports of late ruptures.19 An aneurysm sac may undergo Clinical manifestations
atrophy following successful exclusion and a sudden reper-
fusion may render it more liable to rupture than it might Migration and stent-graft disintegration occur insidiously
have been prior to exclusion. Current surveillance methods and, initially, may not be associated with any symptoms or
are limited by their inability to demonstrate directly the pres- signs until rupture ensues.21 The majority of stent-graft fail-
ence or absence of endotension. The time interval between ures are detected on routine surveillance imaging. Absence
the surveillance scans, according to current schedules, also of symptoms or signs does not signify adequacy of stent-
means that an adverse development could go undetected graft function or integrity.
for as long as nearly a year and, more significantly, may
lead to a catastrophe before it can be detected. Patients with
subtle or early changes of stent migration or disintegration Investigations
are at higher risk of such an event and call for the adoption
of flexible surveillance schedules permitting more frequent PLAIN RADIOGRAPHY
investigation in selected patients. Plain X-rays may demonstrate evidence of suture breakage
Since the mechanisms of device failure related to mater- by showing misalignment of metal struts of adjacent stent
ial fatigue and the morphological changes in the aneurysm rings. Breakages and dislocations of metal struts and frac-
sac take time to evolve, inevitably, it also takes time to eval- tures of hooks and barbs may also be seen well on plain X-
uate each new device or malfunction of that device. For this rays. The anatomical relationship of the stent-graft to bony
reason it is essential that all patients treated with aortic endo- landmarks serves as a guide in evaluating changes in stent-
grafts for repair of an aneurysm are subjected to lifelong graft position over time. Plain radiographs should be per-
surveillance. formed in four planes: AP, lateral and oblique views from
both directions. A methodical assessment of all sets of plain
radiographs taken since implantation is necessary in order to
Risk factors discern the subtle changes. Lateral views reconstructed from
computed tomography (CT) images eliminate inconsisten-
Analysis of EUROSTAR data has revealed a number of cies due to variable centring and projection. Good under-
statistically significant risk factors associated with migra- standing of the radiological appearances of stent
tion.13,20 The association may not be causal for each factor disintegration and migration is crucial to utilising the plain
and the mechanisms by which the risk factors could possi- X-rays to the fullest. Subtle findings can easily go unnoticed.
bly be associated with device failure are not discussed here.
COMPUTED TOMOGRAPHY
Risk factors for the development of
Dual-phase CT scan provides detailed information regarding
migration6,20
the morphology of the aneurysm and the presence or absence
of an endoleak and is currently the mainstay of surveillance.
• Patient factors
The position of the stent-graft can be determined accur-
– Smoking
ately in relation to anatomical landmarks such as the ostia
– Hypertension
of visceral arteries in order to diagnose migration. The
– Higher ASA (Amercan Society of
extent and adequacy of overlap at the landing zones can be
Anesthesiologists) class
assessed precisely to establish the quality of seal. Three-
• Graft related feature
dimensional reconstruction of images using a workstation
– Higher proximal diameter
provides useful information in specific circumstances.
 Management 361

Migration of subtle degree between each interval can add repair on their own, they may be harbingers of further and
up to significant levels over time. Therefore, a review of all more serious failure later. Severe disruption of stent-graft
previous imaging is necessary at each evaluation rather integrity may lead to rupture of the aneurysm by causing
than simply comparing the most recent CTs to the imme- graft related endoleak either directly or through stent-graft
diately preceding set. migration. It is important to recognise that any degree of
migration is a significant finding and points to a compro-
ANGIOGRAPHY mise in fixation even if an ‘adequate seal’ is maintained at
the time of examination. Migration may not progress at a
Angiography is not an appropriate tool for routine surveil- uniform rate, and the possibility that a minor degree of
lance. The indications for diagnostic angiography should be migration may appear as a forerunner of precipitous loss of
considered for each patient on the merits of the situation. fixation, should be considered when management plans
Examples of such indications include the enlarging aneurysm are made.
where there is plain X-ray evidence of borderline fixation The risk of unintended conversion to open repair due to
or structural deformation and deterioration of a stent-graft complications during a secondary intervention needs to be
sufficient to suspect that a fabric hole might have occurred assessed in each patient and preoperative assessment per-
but remained undetected on CT. formed accordingly.

ULTRASOUND SCAN
Selection and timing of secondary
Duplex ultrasound scan, enhanced with microbubble con-
intervention
trast media can provide useful information regarding the
presence or absence of endoleak.22 Being an operator
Late ruptures continue to occur despite a high incidence of
dependent investigation, however, it is not specific or sensi-
secondary interventions.2,19 Evidence pointing to the
tive enough to replace CT even in the diagnosis of endoleak.
importance of stent-graft migration and disintegration is
The major limitation remains the poor quality of informa-
clear and growing stronger.3,19,21 This suggests that currently
tion obtained regarding changing sac and stent-graft morph-
accepted indications and timing of secondary intervention
ology and migration.
are inappropriate or require modification. The hitherto con-
ventional focus on endoleaks should be shifted towards
MAGNETIC RESONANCE IMAGING ensuring stent-graft integrity and stability so that earlier and
Magnetic resonance imaging (MRI) is free from ionising more selective secondary interventions could be under-
radiation and is performed with or without intravenous taken before secondary type 1 or type 3 endoleaks appear.
contrast and functional measurements such as renal perfu- Such early remedy could be expected to reduce the inci-
sion can also be made. A major proportion of patients will dence of catastrophic device failure. Some factors to be
be unsuitable due to the magnetic properties of the device considered when deciding secondary interventions in the
used, the presence of metal implants such as cardiac pace- absence of an associated endoleak are listed in below and in
makers or in cases of claustrophobia. The anatomical detail the algorithm in Figure 31.4.
obtained is also limited. For these reasons, MRI is not used
as a tool of routine surveillance and has a limited role in
investigation. Considerations in endograft failure when
not associated with consequent endoleak

MANAGEMENT • Suture breakage without severe or multiple metal

ring dislocations is unlikely to compromise integrity
of the system
General principles • Multiple metal ring breaks or dislocations can
threaten the structural integrity
Complications of stent-graft migration and disintegration • ‘Landing zones’ shorter than 10 mm are prone to
may be detected on routine surveillance in apparently dislocation and consequent endoleak or
asymptomatic patients. Since the consequences of delayed development of endotension
diagnosis are potentially serious, even fatal, there is a need • Risk is worsened in all the above if associated with
for a methodical and organised surveillance system. A fail- stent-graft angulation or kinking
safe arrangement to ensure patient compliance with sur- • Loss of proximal or mid-graft seal can be more
veillance protocol and prompt review of results is therefore serious than loss of distal seal
essential. • Any migration is significant and may be followed by
Minor structural failures such as suture or strut break- precipitous loss of fixation
age are very common. While not compromising the overall
362 Graft breakdown and migration complicating EVAR

Asymptomatic migration on surveillance scan

Adequate fixation Fixation precarious Fixation lost

(10 mm overlap) (10 mm overlap) (endoleak)

Aneurysm Aneurysm Elective Admit for urgent

size ↓ size ↑ secondary intervention secondary intervention

Yes
? ‘Significant’ structural
failure or distortion
No Any migration means insecure fixation

Continue surveillance. Is the interval to Figure 31.4 Algorithm of

next scan satisfactory?
management of stent-graft migration

Late conversion has been reported to be associated with between stent-graft and native vessel may be bridged with
considerable postoperative morbidity and mortality. That, a covered stent (Figure 31.5). Similarly, fabric holes can be
combined with the fact that the majority of secondary inter- covered and ‘landing zones’ of fixation sites extended with
ventions are feasible via the transfemoral route, meant that stent extensions. Deployment of a new complete stent-
late conversion has been considered to be the last resort and graft system within a failing stent-graft has also been
undertaken only when no other means of intervention is undertaken on rare occasions. Conversion of a bifurcated
possible. In the EUROSTAR and other series, however, mor- into an aorto-uni-iliac reconstruction is also a useful
tality following late conversion was confined to patients hav- approach in some circumstances, e.g. a modular discon-
ing emergency operations for rupture2,21,23 and elective nection that could not be repaired directly. Knowledge
conversion was associated with very low risk of death. When about the device in situ and current anatomy, forward
transfemoral intervention only serves as a temporary meas- planning of access and route of deployment are essential. A
ure to ‘shore-up’ the integrity of a disintegrating stent-graft, vast array of catheters, guidewires and techniques could be
elective conversion is likely to be the preferred option. drawn from but a comprehensive account of their tech-
Stent-graft migration or disintegration may present as nical details is not within the scope of this chapter.
an acute emergency having already caused a rupture of the
aneurysm and this situation requires to be dealt with along Practical considerations
the same lines as a ruptured aortic aneurysm. Acute situ-
ations in which the aneurysm sac may be exposed to sys- The requirement for sophisticated imaging with a digital
temic blood pressure require to be treated urgently, i.e. as subtraction facility, access to a wide array of disposables
soon as possible, even if there are no symptoms. Modular used for catheter techniques, aseptic conditions and facil-
dislocation endoleaks, perigraft endoleak and graft-hole ities for anaesthesia demand a highly sophisticated environ-
endoleak are examples of this. ment for these procedures. The exact location of such
facilities depends on the hospital. It is common for short
percutaneous transfemoral procedures such as iliac limb
ENDOVASCULAR SECONDARY extensions to be performed in the interventional radiology
INTERVENTIONS suite under local anaesthetic. Longer or complex procedures,
those that require surgical exposure of access vessels, inter-
Technical considerations ventions requiring regional or general anaesthesia and those
with some risk of conversion are best undertaken in oper-
Most secondary interventions can be accomplished via the ating theatres with imaging facilities.
transfemoral route but they can be challenging and fre-
quently call for a very high degree of expertise, dexterity Conversion to conventional repair
and catheter skills. The aim is to deploy a covered stent to
restore stent-graft integrity and haemodynamic ‘seal’. A Conversion is indicated for late rupture and also when
gap between dissociated modular components or a gap other forms of secondary intervention are not feasible. All
 Endovascular secondary interventions 363

(a) (b) (c)

Figure 31.5 Modular disconnection. (a) Arrows point to the radio-opaque markers of stent-graft system well aligned initially. (b) Later
separation of these markers due to modular disconnection. (c) Treatment of the problem by a bridging stent

Assessment of fitness for such major surgery should

include objective assessment of cardiorespiratory status in
addition to baseline blood tests and electrocardiography.
Spirometry for respiratory function and echocardiography
or multiple gated acquisition scan (MUGA) scan for estima-
tion of left ventricular ejection fraction ought to be carried
out. Arterial blood gases, echocardiography, creatinine
clearance, etc., could be considered depending on individual
circumstances.
The aorta at and above the upper limit of the stent-graft
needs to be evaluated carefully by preoperative imaging to
ascertain the level of aortic cross-clamping. A suprarenal
aortic clamp is frequently required and on occasions, a
supracoeliac clamp. The risk of renal failure and the need
for dialysis is relatively high. Technical difficulties in the
removal of the stent-graft, as a result of incorporation of
the uncovered stent-graft into native vessels and the effects
of fixation barbs or hooks, should be expected. Wire cut-
ters should be available for cutting the struts.

Conclusions

Until completely durable devices become available, EVAR

patients will remain at a lifelong risk of device failure with
migration and disintegration, leading to secondary
Figure 31.6 Severe stent-graft disintegration. The metal struts
not only show misalignment but also breakage (arrows) endoleak and late rupture. With the conventional primary
focus on endoleaks, the diagnosis of subtle but incremen-
tal development of migration or deterioration of graft
integrity is often overlooked and the consequences may
fit patients with severe structural failure or migration of
be disastrous. There is a need to re-examine the indica-
stent-graft, where other forms of secondary intervention may
tions and timing of secondary interventions to eliminate
serve only as a temporary measure (Figure 31.6), should be
the risk of late rupture. Transfemoral interventions are
considered for an elective conversion to conventional repair.
364 Graft breakdown and migration complicating EVAR

aneurysms. Intermediate-term follow-up results of a European

adequate in extending the seal or in bridging defects and
collaborative registry (EUROSTAR). Br J Surg 2000; 87: 1666–73.
are usually feasible in the majority of patients. Although
2 Harris PL, Vallabhaneni SR, Desgranges P, et al. Incidence and
late conversion is technically challenging, it should be risk factors of late rupture, conversion, and death after
considered in all patients with severe stent-graft disinte- endovascular repair of infrarenal aortic aneurysms: the
gration or migration, where endoluminal interventions EUROSTAR experience. European Collaborators on Stent/graft
may only be able to offer a temporary solution. Actual or techniques for aortic aneurysm repair. J Vasc Surg 2000; 32:
imminent loss of seal requires urgent intervention. Long 739–49.
term solutions lie in alterations in the stent-graft design 3 Vallabhaneni SR, Harris PL. Lessons learnt from the EUROSTAR
and construction to impart stronger fixation and resist- registry on endovascular repair of abdominal aortic aneurysm
ance to structural deterioration. repair. Eur J Radiol 2001; 39: 34–41.
4 Veith FJ, Abbott WM, Yao JST, et al. Guidelines for development
and use of transluminally placed endovascular prosthetic grafts
in the arterial system. J Vasc Surg 1995; 21: 670–85 and J Vasc
ACKNOWLEDGEMENT Interv Radiol 1995; 6: 477–92.
5 Broeders IA, Blankensteijn JD, Wever JJ, Eikelboom BC. Mid-term
fixation stability of the EndoVascular Technologies endograft.
We wish to acknowledge Gillian Rycroft, Medical Illustrator, EVT Investigators. Eur J Vasc Endovasc Surg 1999; 18: 300–7.
for assistance with Figure 31.1. 6 Malina M, Brunkwall J, Ivancev K, et al. Endovascular healing is
EUROSTAR (European collaborators on Stent-Graft inadequate for fixation of Dacron stent-grafts in human
Techniques for Aneurysm Repair) a Europe-wide project aortoiliac vessels. Eur J Vasc Endovasc Surg 2000; 19: 5–11.
launched in 1996 to scientifically evaluate endovascular 7 Guidoin R, Marois Y, Douville Y, et al. First-generation aortic
endografts: analysis of explanted Stentor devices from the
aneurysm repair, was adopted by the European Society for
EUROSTAR Registry. J Endovasc Ther 2000; 7: 105–22.
Vascular and Endovascular Surgery in September 2001. By
8 Resch T, Ivancev K, Brunkwall J, et al. Midterm changes in
September 2001, collaborators from 107 centres had aortic aneurysm morphology after endovascular repair. J
enrolled in excess of 4000 patients into the Eurostar data- Endovasc Ther 2000; 7: 279–85.
base. Comprehensive data regarding preoperative, intra- 9 Heintz C, Riepe G, Birken L, et al. Corroded nitinol wires in
operative and follow-up details are collected prospectively explanted aortic endografts: an important mechanism of
and periodically analysed. The follow-up is structured and failure? J Endovasc Ther 2001; 8: 248–53.
uniform. See the website at www.esvs.org/start/index.asp 10 Harris P, Brennan J, Martin J, et al. Longitudinal aneurysm
(accessed 21 December 2004). shrinkage following endovascular aortic aneurysm repair: a
source of intermediate and late complications. J Endovasc Surg
1999; 6: 11–16.
11 Wever JJ, de Nie AJ, Blankensteijn JD, et al. Dilatation of the
Key references proximal neck of infrarenal aortic aneurysms after endovascular
AAA repair. Eur J Vasc Endovasc Surg 2000; 19: 197–201.
Albertini J, Kalliafas S, Travis S, et al. Anatomical risk factors for 12 Prinssen M, Wever JJ, Mali WP, et al. Concerns for the durability
proximal perigraft endoleak and graft migration following of the proximal abdominal aortic aneurysm endograft fixation
endovascular repair of abdominal aortic aneurysms. Eur J Vasc from a 2-year and 3-year longitudinal computed tomography
Endovasc Surg 2000; 19: 308–12. angiography study. J Vasc Surg 2001; 33(suppl 2): S64–9.
Beebe HG, Cronenwett JL, Katzen BT, et al. Vanguard Endograft 13 Mohan IV, van Marriewijk, Harris PL, et al. Factors and forces
Trial Investigators. Results of an aortic endograft trial: impact influencing stent-graft migration after endovascular aortic
of device failure beyond 12 months. J Vasc Surg 2001; aneurysm. J Endovasc Ther 2002; 9: 748–55.
33(suppl 2): S55–63. 14 Liffman K, Lawrence-Brown MM, Semmens JB, et al. Analytical
Broeders IA, Blankensteijn JD, Wever JJ, Eikelboom BC. Mid-term modeling and numerical simulation of forces in an endoluminal
fixation stability of the EndoVascular Technologies endograft. graft. J Endovasc Ther 2001; 8: 358–71.
EVT Investigators. Eur J Vasc Endovasc Surg 1999; 18: 300–7. 15 Marty B, Sanchez LA, Ohki T, et al. Endoleak after endovascular
Jacobowitz GR, Lee AM, Riles TS. Immediate and late explantation graft repair of experimental aortic aneurysms: does coil
of endovascular aortic grafts: the endovascular technologies embolization with angiographic “seal” lower intraaneurysmal
experience. J Vasc Surg 1999; 29: 309–16. pressure? J Vasc Surg 1998; 27: 454–61; discussion 462.
Vallabhaneni SR, Harris PL. Lessons learnt from the EUROSTAR 16 Gilling-Smith G, Brennan J, Harris P, et al. Endotension after
registry on endovascular repair of abdominal aortic aneurysm endovascular aneurysm repair: definition, classification, and
repair. Eur J Radiol 2001; 39: 34–41. strategies for surveillance and intervention [review]. J Endovasc
Surg 1999; 6: 305–7.
17 White GH, May J, Petrasek P, et al. Endotension: an explanation
REFERENCES for continued AAA growth after successful endoluminal repair.
J Endovasc Surg 1999; 6: 308–15.
18 Alimi YS, Chakfe N, Rivoal E, et al. Rupture of an abdominal
1 Laheij RJ, Buth J, Harris PL, et al. Need for secondary aortic aneurysm after endovascular graft placement and
interventions after endovascular repair of abdominal aortic aneurysm size reduction. J Vasc Surg 1998; 28: 178–83.
 References 365

19 Zarins CK, White RA, Fogarty TJ. Aneurysm rupture after of device failure beyond 12 months. J Vasc Surg 2001; 33(suppl
endovascular repair using the AneuRx stent graft. J Vasc Surg 2): S55–63.
2000; 31: 960–70. 22 McWilliams RG, Martin J, White D, et al. Detection of
20 Albertini J, Kalliafas S, Travis S, et al. Anatomical risk factors endoleak with enhanced ultrasound imaging: comparison with
for proximal perigraft endoleak and graft migration following biphasic computed tomography. J Endovasc Ther 2002; 9:
endovascular repair of abdominal aortic aneurysms. Eur J Vasc 170–9.
Endovasc Surg 2000; 19: 308–12. 23 Jacobowitz GR, Lee AM, Riles TS. Immediate and late
21 Beebe HG, Cronenwett JL, Katzen BT, et al. Vanguard Endograft explantation of endovascular aortic grafts: the endovascular
Trial Investigators. Results of an aortic endograft trial: impact technologies experience. J Vasc Surg 1999; 29: 309–16.
This page intentionally left blank
 32
Atheroembolism Complicating EVAR

JUAN C PARODI, LUIS M FERREIRA

The problem 367 Management 369

Aetiology/pathophysiology 367 Results 369
Mechanisms 368 Endovascular approach 370
Clinical manifestations 368 References 371

THE PROBLEM stages of their experience, and were no longer experiencing

them, shows that it was at least partly to do with a learning
curve or perhaps better patient selection. The aim of this
Atheroembolism results from the release of cholesterol chapter is to discuss our experience of treating patients who
rich atheromatous debris from an ulcerated aorta or an suffered atheroembolism and offer our approach in averting
abdominal aortic aneurysm (AAA) into the systemic arter- this serious complication in endovascular procedures.
ial circulation.1–2 Atheroembolism is a rare condition, but
it is becoming more common with the growing indications
for endovascular procedures and increasing numbers of
candidates undergoing them. At the same time, an ageing
AETIOLOGY/PATHOPHYSIOLOGY
population represents a rising incidence of atheromatous
‘shaggy’ aortas and aneurysms warranting treatment. The Atheroembolism consists of cholesterol crystal microem-
technical feasibility of endovascular aneurysm repair (EVAR) boli which generally occlude arterioles of 200–900 m cali-
has been conclusively established, with many centres report- bre, often leading to the development of the typical blue
ing the successful exclusion of AAAs by endoprostheses.3–5 toe syndrome (Fig. 32.1), in which arterial occlusions are
The surgical community has also enthusiastically embraced located at the level of the digital arteries of the feet.1–2 This
this new technology because it offers potentially significant atheromatous embolism syndrome is a frequently misdiag-
advantages over conventional aneurysm repair in that it nosed, and sometimes unrecognised, condition the true
avoids transperitoneal or retroperitoneal interventions, incidence and consequences of which are unknown, and
general anaesthesia and prolonged aortic clamping while also although often clinically silent, associated with significant
ensuring rapid recovery. morbidity and mortality. Clinically it may present in a
In spite of the potential advantages of EVAR, early reports variety of ways. Atheroembolism may be spontaneous but
documented a high incidence of perioperative complica- it sometimes occurs after lifting weights, coughing or
tions and mortality in excess of 5 per cent. Parodi et al.,6 tenesmus, although it is definitely more frequently observed
Gitlitz et al.7 and Thompson et al.8 reported cases of mas- during or after open aortic interventions, endovascular
sive microembolism as a significant complication of this procedures and anticoagulant or thrombolytic therapy.
technique with an incidence of 4–17 per cent, all invariably During the latter, the blue toe syndrome may occur at any
fatal cases. In contrast, however, several other representative time but, typically, it does so within 4–12 weeks of having
series failed to document evidence of peripheral microem- been started. The mechanism of embolism is presumed to
bolism which suggests that this complication may be be due to unsuccessful fibrin apposition and loosening of
related either to technique or the device used. The fact that fresh thrombus at ulcerated plaques, but it may also be pre-
even those who reported such incidents did so in the early cipitated by sudden haemorrhage into plaques, leading to
368 Atheroembolism complicating EVAR

and catheters is rapidly dispersed in retrograde fashion

upwards into the visceral and renal arteries with devastating
results. When we reviewed our cases of atheroembolism it
became clear that large and tortuous aneurysms have a
potentially higher likelihood of causing this problem. First,
that is probably because in advancing the guidewire from
the femoral artery into the aorta and then into the prox-
imal neck, the operator is negotiating the device through a
tortuous chamber often lined with friable material which
can become dislodged by those manoeuvres. In large
aneurysms especially, therefore, it is advisable to insert the
guidewire percutaneously via the brachial artery or, better
still, a floating balloon mounted on a wire, as will be dis-
cussed later. A second possible cause of atheroembolism
Figure 32.1 Embolisation to the dermal vessels with areas of applies to large aneurysms which create the necessity of
focal ischaemia, purplish discoloration of the skin and surrounding adding extensions to lengthen the device or to cover any type
petechial haemorrhages (blue toe syndrome) of endoleak. In short, the more intravascular manipulations
performed, the greater the risk for dislodgement of mate-
rial and embolism.
A new generation of endografts has evolved of greater
flexibility and of slimmer profile compared with initial
their disruption. Data suggesting that anticoagulant ther-
models. These changes in design are also probably respon-
apy causes embolism, however, are not overwhelmingly
sible for the lower incidence of atheroembolism in more
convincing.
recently reported series.
Atheroembolism can be a serious complication associated
with procedures undertaken for diagnostic or therapeutic
purposes. In a retrospective study of autopsy findings of
patients who died soon after coronary angiography, almost CLINICAL MANIFESTATIONS
30 per cent were found to have evidence of atheroembolism.1
In contrast, in 70 age-matched patients with diffuse athero-
A variety of clinical manifestations are observed in atheroem-
sclerotic disease, but no history of having undergone endo-
bolism, most commonly involving the lower limbs. The con-
luminal procedures, the incidence of atheroembolism was
dition may be heralded by hypertension, pain and a livedo
only 4.3 per cent.1
reticularis pattern on the abdomen and lower extremities,
Thromboembolic complications are relatively uncom-
cyanosis, discoloration, blue toes, gangrene, and later, ulcer-
mon after surgical AAA repair but they have been reported
ation. These features of embolism to dermal vessels gener-
in many series of endoluminal treatment. While periopera-
ally appear on the buttocks, thighs, legs and feet (see Fig.
tive mortality in patients undergoing conventional AAA
32.1). They are painful lesions caused by focal ischaemia,
repair is primarily cardiac, in those subjected to the endo-
recognisable by their purplish discoloration and surround-
luminal approach it is mainly associated with local or
ing petechial haemorrhages, and present in spite of palpable
vascular complications. One of them, namely, massive
foot pulses. If left untreated, the natural history of microem-
microembolim, is a devastating complication and proved
bolism of the extremities is one of repetitive such events;
to be fatal in over 90 per cent of affected patients. In some
more than 50 per cent of patients develop further compli-
initial series, including ours, one of the most important
cations and nearly 40 per cent experience some degree of
causes of perioperative death was related to diffuse embolism
tissue loss. Long term follow-up shows that up to 20 per cent
during endograft deployment.9–13
of these patients will die within a year of the initial event.
The most common clinical manifestation of visceral
atheroembolism is diffuse abdominal pain, distension,
paralytic ileus and bleeding. In cases of renal atheroem-
MECHANISMS
bolism the clinical picture is one of hypertension, fever,
eosinophilia and acute renal failure.
The mechanism of thromboembolism is probably related The diagnosis of atheroembolism is made having first
to intrasac manipulation by large and rigid devices dislodging anticipated that possibility, recognising the symptoms and
laminated thrombus lining the sac of the aneurysm. Most signs and confirming the source of embolism by appropri-
endovascular teams witnessed massive microembolism ate imaging. In most cases, colour duplex scans, trans-
during their early experience of inserting endografts. With oesophageal echocardiography and computed tomography
bilateral distal clamping, the debris loosened by guidewires (CT) scans are of value in defining the territory responsible.
 Results 369

Keen and Yao describe the following CT findings in patients the emboli had come from an atherosclerotic ‘shaggy’
who suffered spontaneous atheroembolism from an AAA: aorta, and in another from a limited ulcer which was acti-
irregular luminal surface, multiple lumens, heterogeneity vated by anticoagulant therapy. In addition to this cohort,
of thrombus, calcification within the thrombus, fissures three other patients who developed atheroembolism were
extending from the lumen into the thrombus and non- treated: one from poststenotic dilatation of the subclavian
contiguous areas of intraluminal thrombus.14 artery as a component of the thoracic outlet syndrome,
another from a popliteal artery aneurysm and one from
iliac stenosis.
The source of microembolism was defined by contrast-
MANAGEMENT
enhanced CT scan or angiography. Pathological studies of
skin and muscle were carried out in only two patients.
Treatment should be directed towards three goals: removal Clinical presentation included all or some of the following
of the source of atheromatous debris, symptomatic care of features: severe limb pain, livedo reticularis and lesions
the end organ wherein the emboli are located and risk factor with serpiginous edges, elevated serum levels of creatinine
modification to prevent re-embolism and progression of phosphokinase (CPK), abdominal pain, uncontrolled hyper-
the disease. tension, renal failure requiring dialysis, weight loss and
Medical management described in the literature includes paraparesis.
the use of heparin, dextran, papaverine, urokinase and Once the decision to treat the patient was made, a 3 Fr
vasodilators, among other agents. Unfortunately, the results multiperforated catheter was placed in the popliteal artery,
have not been uniformly satisfactory. Atheroembolism often most frequently inserted via the contralateral femoral artery.
produces tissue loss and even death. Since 1985, in our insti- That allowed us to combine a urokinase infusion (bolus
tution, we have treated patients with severe lower extremity 300 000 U at an infusion rate of 60 000–100 000 U per hour)
microembolism by means of intra-arterial prostaglandin with 250 g of alprostadil (Prolisina VR 0.5 mg/mL, Upjohn,
E1 (PGE1) infusion.15 The encouraging results achieved Pururs, Belgium) diluted in 250 mL of normal saline solu-
with PGE1 in Buerger’s disease (see Chapter 42) induced us tion administered as a continuous infusion over a period of
to try the drug in cases of severe atheroembolism. 2–4 hours, the rate adjusted to the patient’s tolerance of
Kurzock and Lieb discovered prostaglandins in 1930. vasodilator effects of the latter in terms of discomfort and
Von Euler in 1934 coined the term prostaglandin because even pain. The infusion was repeated depending on the
he mistook the substance for the secretory product of the response, either on the same day or on subsequent days. In
prostate. Prostaglandin E1 has several biological properties the series of patients presented here the total dose varied
which make it a suitable drug for maintaining patency of between 500 and 8000 g of PGE1.
the microcirculation. Prostanoids have been used exten-
sively in Europe for the treatment of peripheral arterial dis-
ease (prostacyclin PGI2, PGE1, and the chemically stable
RESULTS
prostacyclin analogue, iloprost). PGI2 is chemically unstable
and needs to be administered at high pH, which also has the
potential of causing vascular damage. Prostaglandin E1 is Twenty-nine patients, men of average age 71 years (58–78),
used intra-arterially because one passage through the lung were treated as follows: by AAA conventional repair
inactivates almost 70–90 per cent of the drug. (n  6), AAA endoluminal treatment (n  6), aortob-
The treatment of microembolism itself should be con- ifemoral bypass (n  9), resection of cervical rib (n  1),
sidered if the effects are severe enough to cause tissue loss resection of popliteal aneurysm (n  1), iliac stenting
or uncontrollable pain, but attention should also be given (n  1) and axillo-bifemoral bypass (n  2). The remaining
to eliminating the cause. A trial, started in 1985, using three high risk patients were treated only medically.
intra-arterial PGE1 in 29 patients admitted to our clinic Treatment with PGE1 preceded that of the source of
with the diagnosis of blue toe syndrome caused by severe embolism by several days (4–18 days).
atheroembolism, has been reported.15 All the following cri- In three of the cases of massive microembolism caused
teria had to be satisfied for inclusion in that trial: persistent by endoluminal treatment of AAAs, the intra-arterial
pain resistant to medication, severe ischaemia with risk of injection of PGE1 only provided temporary and partial
tissue loss but with distal pulses or a positive Doppler sig- improvement in skin perfusion. Two patients died of
nal present. Patients with asymptomatic or mild symptoms multiorgan failure (see Chapters 2 and 4), and one of them
of microembolism were not included in the series and suffered spinal cord injury with paraparesis. Primitive
treatment was directed solely at dealing with its source. In endovascular devices were inserted in three of the patients
19 patients the source of microembolism was an AAA, with AAA. Neither visceral nor renal embolism was recorded
occurring spontaneously in nine, during open repair in four, with the new generation of devices, all of which were of a
secondary to endoluminal treatment in five and due to slimmer profile and much more flexible and only three
diagnostic instrumentation in one. In six of these patients patients who received them developed distal embolism.
370 Atheroembolism complicating EVAR

One AAA patient with recurrent lower limb and renal that complication. While the rigidity and the broader profile
microembolism, receiving dialysis for anuria, was success- of the older devices played a significant role in atheroem-
fully treated with a stent-graft after the distal ischaemia had bolism, the most important factor determining who will or
been relieved by PGE1. Immediately after excluding the will not develop emboli is probably the severity of athero-
aneurysm, urine production recommenced, but unfortu- sclerotic disease in the aorta. Thompson et al., using an
nately he suffered massive fatal pulmonary embolism a few ultrasound based method of detecting lower limb
days later. In this group 30-day mortality was 25 per cent. The atheroembolism, demonstrated a higher incidence of par-
side effects and complications of treatment were minimal. ticle embolisation during endovascular repair compared
Recovery from renal failure occurring immediately after with conventional aneurysm surgery.16 In addition, bilat-
aneurysm exclusion was achieved in two patients, one having eral iliac artery occlusion generates turbulent flow and a
been treated endoluminally and the other by conventional net movement of those particles into the suprarenal aorta.
open repair, the inference being that microembolism is an As Lipsitz et al.17 demonstrated in an animal model, initial
important factor in renal failure. Equally, interruption of distal clamping minimises distal embolisation, but renal
the continuous showers of microemboli improves renal and/or visceral embolisation may follow. However, if out-
function. flow into one iliac artery is being maintained, embolism
All other patients showed improvement in peripheral into that extremity is more likely to occur.
ischaemia and either had minor loss or no loss of tissue. Based on these findings, we have designed our own
Long term follow-up data were available for 10 patients device (Parodi Antiembolism System or PAESreg; ArteriA,
and all had favourable outcomes. San Francisco Science, San Francisco, CA, USA) using a new
Prostaglandin E1 is a potent vasodilator which activates concept which ensures protection from distal embolisa-
fibrinolysis and inhibits leucocyte migration and activation, tion. It had been designed originally for use in the carotid
release of leucotrienes, oxygen free radicals and proteolytic territory to ensure reversal of flow during carotid angio-
enzymes. Prostaglandin E1 provokes platelet disaggrega- plasty and stenting.18 It consists of a 7–8 Fr guiding catheter
tion and inhibits platelet release of thromboxane and with an inverted pear-shaped balloon at the tip to occlude
5-hydroxytryptamine. It increases deformability of red blood the common iliac artery (CIA) through which antegrade
cells and has an antiproliferative action on vascular smooth iliac flow is permitted. Flow through the CIA is achieved by
muscle cells. In our experience of patients whose distal an iliac-to-femoral arterio-arterial shunt connecting the
pulses are present despite peripheral ischaemia and tissue guiding catheter to the distal superficial femoral artery
loss, the intra-arterial infusion of PGE1 reversed pregan- introducer at the time the endovascular device is intro-
grenous changes in the toes and relieved pain. Capillary duced and deployed.
filling and pain relief are evident almost immediately after To avoid dislodgement of fragments of thrombus, no
the injection of PGE1. Unfortunately, there have not been wires were advanced via a femoral approach retrogradely
any controlled trials demonstrating its efficacy. into the aneurysm. Instead, a catheter was inserted into
the aorta via the left brachial artery. Once the catheter had
reached the descending aorta, a Percusurge (Percusurge
Inc, Sunnyvale, CA, USA) balloon guidewire (0.014 and
ENDOVASCULAR APPROACH 0.018 inches) was inserted through the catheter and allowed
to navigate freely with aortic flow down into the aneurysm.
In most instances, direct aortic reconstruction may be the As one iliac artery had its flow interrupted by the PAES
treatment of choice in the presence of peripheral atheroem- device, the catheter was directed into the chosen iliac
bolism from an ulcerated lesion in a ‘shaggy’or dilated artery; when the wire and the catheter mounted on it were
aorta. Nevertheless, comorbidities, severe pulmonary dis- felt within the exposed common femoral artery on the lat-
ease and limited life expectancy weigh heavily against con- ter side, they were exteriorised through an arteriotomy and
ventional aortic bypass and in favour of an alternative an extra-stiff guidewire used to replace the navigator wire.
procedure such as EVAR. We believe that avoidance of The stiff wire was finally passed up to the brachial artery
general anaesthesia contributed to the lower rate of com- introducer and gently held, creating in this way a ‘through
plications in this type of patient in whom successful post- and through’ guidewire. Applying tension on the wire, we
operative weaning from mechanical ventilation might have introduced the endograft without angulation and so pre-
been very difficult. In high risk patients with non-aneurysmal vented disruption of thrombus. Further, to avoid intra-
aortic disease conventional extra-anatomical procedures aortic manipulation in relation to contralateral stump
such as axillo-bifemoral bypass are acceptable alternatives. cannulation, we prefer an aorto-uni-iliac device for this
The relation between the use of stent-grafts and application.
microembolism deserves special comment. Atheroem- In cases in which thrombus was situated at the level of
bolism is the most dreaded complication associated with the renal arteries, the Angioguard device (Angioguard Inc,
stent-grafting and, paradoxically, endoluminal exclusion Plymouth, MN, USA) was placed occluding the ostia of the
of an AAA is a rapidly emerging therapeutic tool in treating superior mesenteric and renal arteries. It consists of a low
 References 371

diagnostic and therapeutic methods when needed. We

describe two approaches for patients with distal
atheroembolism in whom endoluminal exclusion of the
aortic source of emboli is attempted using three differ-
ent devices. In each case, the stent-graft was positioned
within the infrarenal aorta preventing visceral and distal
embolisation. The dilated segment of the aorta was
thereby safely excluded from the circulation.

Key references
Kauffman JL, Shah DM, Leather RP. Atheroembolism and
microembolism syndrome. (Blue toe syndrome and
disseminated atheroembolism). In: Rutherford RB (ed).
Vascular Surgery. Philadelphia, PA: WB Saunders Company,
(a) 1995: 669–77.
Lipsitz EC, Veith FJ, Ohki T, Quintos RT. Should initial clamping for
abdominal aortic aneurysm repair be proximal or distal to
minimise embolisation? Eur J Vasc Endovasc Surg 1999; 17:
413–18.
Messina LM. Peripheral arterial embolism. In: Greenfield L J (ed).
Surgery, Scientific Principles and Practice. Philadelphia, PA:
JB Lippincott Company, 1993: 1478–92.
Parodi JC. Treatment of blue toe syndrome with intra-arterial
injection of Prostaglandin E1. In: Yao JST, Pearce WH (eds).
Aneurysms: New Findings and Treatment. Norwalk, CT:
Appleton & Lange, 1994: 325–31.
Thompson MM, Smith J, Naylor AR, et al. Microembolization during
endovascular and conventional aneurysm repair. J Vasc Surg
1997; 25: 179–86.

REFERENCES

(b) 1 Messina LM. Peripheral arterial embolism. In: Greenfield LJ (ed).

Surgery, Scientific Principles and Practice. Philadelphia, PA:
Figure 32.2 Angioguard device seen on angiography (a) JB Lippincott Company, 1993: 1478–92.
positioned within the renal arteries and (b) within both renal and 2 Kauffman JL, Shah DM, Leather RP. Atheroembolism and
superior mesenteric arteries microembolism syndrome. (Blue toe syndrome and disseminated
atheroembolism). In: Rutherford RB (ed). Vascular Surgery.
Philadelphia, PA: WB Saunders Company, 1995: 669–77.
profile guidewire based, 4 Fr filter-type device which was
3 May J, White GH, Waugh R, et al. Improved survival after
placed in the arterial ostium, capturing and removing
endoluminal repair with second-generation prostheses compared
embolic debris while maintaining distal perfusion (Fig. 32.2). with open repair in the treatment of abdominal aortic
aneurysms: a 5-year concurrent comparison using life table
method. J Vasc Surg 2001; 33: S21–6.
Conclusions 4 Beebe HG, Cronenwett JL, Katzen BT, et al. Results of an aortic
endograft trial: impact of device failure beyond 12 months.
Intra-arterial infusion of PGE1 appears to be an effective J Vasc Surg 2001; 33: S55–63.
treatment of severe atheroembolism when all other 5 Bush RL, Lumsden AB, Dodson TF, et al. Mid-term results after
treatment has failed. Surgical or endoluminal treatment endovascular repair of the abdominal aortic aneurysm J Vasc
of the primary source of atheromatous embolism is indi- Surg 2001; 33: S70–6.
6 Parodi JC, Criado FJ, Barone HD, et al. Endoluminal aortic
cated to prevent recurrence. This serious complication
aneurysm repair using a balloon-expandable stent-graft device:
can be prevented by the judicious use of endoluminal
a progress report. Ann Vasc Surg 1994; 8: 523–9.
372 Atheroembolism complicating EVAR

7 Gitlitz DB, Ramaswami G, Kaplan D, et al. Endovascular stent 13 Thompson MM, Smith JL, Bell PR. Thromboembolic complications
grafting in the presence of aortic neck filling defects: early during endovascular aneurysm repair. Semin Vasc Surg 1999; 12:
clinical experience. J Vasc Surg 2001; 33: 340–4. 215–19.
8 Thompson MM, Smith J, Naylor AR, et al. Microembolization 14 Keen RR, Yao JST. Aneurysm and embolization: detection and
during endovascular and conventional aneurysm repair. J Vasc management In: Yao JST, Pearce WH (eds). Aneurysms: New
Surg 1997; 25: 179–86. Findings and Treatment. Norwalk, CT: Appleton & Lange, 1994:
9 Jaeger HJ, Mathias KD, Gissler HM, et al. Rectum and sigmoid 305–14.
colon necrosis due to cholesterol embolization after 15 Parodi JC. Treatment of blue toe syndrome with intra-arterial
implantation of an aortic stent-graft. J Vasc Interv Radiol 1999; injection of Prostaglandin E1. In: Yao JST, Pearce WH (eds).
10: 751–5. Aneurysms: New Findings and Treatment. Norwalk, CT: Appleton
10 Sandison AJ, Edmondson RA, Panayiotopoulos YP, et al. Fatal & Lange, 1994: 325–31.
colonic ischaemia after stent graft for aortic aneurysm. Eur J 16 Thompson MM, Smith J, Naylor AR, et al. Ultrasound-based
Vasc Endovasc Surg 1997; 13: 219–20. quantification of emboli during conventional and endovascular
11 Lindholt JS, Sandermann J, Bruun-Petersen J, et al. Fatal late aneurysm repair. J Endovasc Surg 1997; 4: 33–8.
multiple emboli after endovascular treatment of abdominal 17 Lipsitz EC, Veith FJ, Ohki T, Quintos RT. Should initial clamping for
aortic aneurysm. Case report. Int Angiol 1998; 17: 241–3. abdominal aortic aneurysm repair be proximal or distal to minimise
12 Zempo N, Sakano H, Ikenaga S, et al. Fatal diffuse atheromatous embolisation? Eur J Vasc Endovasc Surg 1999; 17: 413–18.
embolization following endovascular grafting for an abdominal 18 Parodi JC, La Mura R, Ferreira LM. Initial evaluation of carotid
aortic aneurysm: report of a case. Surg Today 2001; 31: angioplasty and stenting with three different cerebral protection
269–73. devices. J Vasc Surg. 2000; 32: 1127–36.
 SECTION
7
Regional Vascular Trauma

33. Vascular injuries of the limbs 375

34. Vascular injuries of the chest 401

35. Vascular injuries of the neck 419

36. Abdominal vascular injuries 429

37. Limb replantation 443

This page intentionally left blank
 33
Vascular Injuries of the Limbs

AIRES AB BARROS D’SA, JOHN M HOOD, PAUL HB BLAIR

Introduction 375 Diagnosis and assessment 383

The problem 376 Preoperative considerations 384
Mechanisms of injury 376 Definitive surgical treatment 386
Morphology of vessel injury 379 Postoperative management 396
Pathophysiology of limb vascular injury 381 References 397
Initial management 382

INTRODUCTION During the Vietnam War this striking upturn in limb sal-
vage was maintained at 12.7 per cent.4–6 The significantly
improved long term results of vascular repair in battle
Through the millennia and even into the twentieth century casualties in Vietnam, as documented in the Vietnam Vas-
the treatment of limb vascular injuries was confined cular Registry, were attributed to evacuation by helicopter
mostly to the staunching of bleeding by cautery, styptics, within 3 hours of injury, operation by surgeons experi-
compression and ligature, in order to save life. The concept enced in vascular repair and the liberal use of autogenous
of vascular repair aimed at limb preservation was reflected vein grafts. Experimental work on the wounding capacity of
in very few anecdotal reports. In a letter William Hunter missiles has been of immense value to clinicians managing
records the occasion when Halliwell in 1759 performed the gunshot wounds.7,8
first successful vascular repair on a lacerated brachial artery
employing a farrier’s stitch. Well over a century later
Murphy in Chicago reconstructed a completely transected
femoral artery. Encouraging clinical and experimental Wartime rate of amputations
reports of vein grafting began to emerge from both sides of
the Atlantic during the early twentieth century. • World War I – 72.5 per cent
In actual practice, however, these vascular repair tech- • World War II – 35.8 per cent
niques proved to be largely impracticable when tested dur- • Korean War – 13.0 per cent
ing the bitter operational conditions of World War I, in • Vietnam War – 12.7 per cent
which high explosives and missiles accounted for an ampu-
tation rate of 72.5 per cent.1 During World War II attempts
at vascular repair were seen to be demonstrably superior to Lessons from military experience were progressively
ligation, lowering the amputation rate to 35.8 per cent.2 applied to the rising incidence of limb vascular injuries in
This was at a time when the incidence of popliteal and urban American civilian practice,9–15 a quarter of these
crural artery injury was estimated to be approximately 20 involving the upper extremity. These injuries inflicted by
per cent for each site2 and when the delay between injury knives and handguns are increasingly caused by automatic
and admission was approximately 10 hours. Despite the weapons and assault rifles in parallel with the mounting cul-
average lag time of just over 6 hours between injury and ture of gangsterism and the booming traffic in illicit drugs.
repair in the Korean War, the formal application of well Sadly, many of these assailants are juveniles and it could be
documented methods of vessel reconstruction dramat- argued that the portrayal of gratuitous violence or the glam-
ically reduced lower limb amputation rate to 13 per cent.3 orisation of brutality in the media and films, as if they were
376 Vascular injuries of the limbs

a necessary ingredient of human existence, may well play a

part in influencing the young mind. Democratic govern-
Some defining features of ‘complex’ limb
ments and caring societies deservedly place great weight on
vascular injury
free expression and libertarian values but they also have a
responsibility to ensure that those ideals are tempered by the • Caused by high energy trauma
tenets of a universal sense of morality enshrined in appro- • Concomitant artery and vein injured
priate legislation. • Transection, avulsion and damage to long vessel
segments
In some parts of the world, such as in Northern Ireland,
where terrorists, fuelled by easy access to sophisticated • Atherosclerotic artery
weaponry and poorly detected explosives, and trained in • Fracture(s) with comminution, periosteal stripping
manufacturing massive crude bombs and incendiary • Joint dislocation, closed or open
devices, have prosecuted a sustained and indiscriminate • Disruption and loss of soft tissue including collaterals
assault on a civilian population for over a quarter of a cen- • Pressure haematoma
tury and have accounted for a toll of around 3500 dead and • Nerve injury
40 000 maimed. The people of this province have endured • Contamination of open wound
physical suffering and grief with resilience and dignity,
while the very fabric of their lives was being blighted
by the systematic destruction of homes and places of Limb vascular trauma is often accompanied by substantial
work and leisure. Most of the victims resembled military blood loss and hypovolaemic hypotension, compounding
casualties and were treated predominantly in the main the complex pathophysiological events caused by ischaemic
teaching hospital the Royal Victoria Hospital, Belfast, arrest of flow. These effects, complicated further by impaired
which assumed the responsibilities of a front line evacu- drainage due to concomitant vein injury, raise compartment
ation centre.16–21 Vascular injuries treated in the hospital’s pressure and may lead to the compartment syndrome,
regional vascular surgery unit, drew on accumulated ischaemic contractures and amputation. Risk factors such as
global experience and introduced innovative approaches smoking, diabetes and peripheral atheroma, especially in the
in operative care. elderly casualty, are likely to contribute to a poor outcome.
A mounting incidence of iatrogenic limb vascular injuries These potential sequelae can be averted by resuscitative
(see Chapters 38 and 39) complicating arterial cannulation measures and vascular repair, and time is of the essence.
reflects the phenomenal proliferation of diagnostic and thera- Regardless of the aetiology and environment surround-
peutic procedures which are a feature of modern medical ing limb vessel injury, the essential principles of treatment
practice.22–26 remain the same: emergency aid at the scene, rapid trans-
port to hospital, vigorous resuscitation, accurate diagnosis
and definitive operative repair. In the complex limb vascu-
lar injury,28 especially when injuries to other systems coex-
THE PROBLEM ist, management often demands the involvement of other
disciplines. For optimum results, vascular injuries should
Recurring regional conflicts, civil wars, terrorism, violence be treated by surgeons based in hospitals with the neces-
in large conurbations and the ceaseless toll of accidents on sary facilities and resources, and with experience in vascu-
the roads and at work, inevitably, indicate that limb vascu- lar reconstruction and trauma.
lar injuries are here to stay. Vascular injuries, caused by a
host of mechanisms, will continue to challenge the clinical
acumen and operative technique of the vascular surgeon.
MECHANISMS OF INJURY
Injuries limited to vessels alone call for sound technique of
vascular repair but those that are complex also demand Penetrating
good judgement.
The defining features of limb vascular injury meriting The circumstances surrounding an incident of penetrating
the term ‘complex’, regardless of aetiology, include some vascular trauma of the limbs, namely, the type of wounding
of the following: high energy wounding agent; severe agent, the nature of the injury and the lapse of time before
injury of artery and vein; damage to soft tissues including evacuation to hospital, will strongly influence management.
muscle and collaterals possibly with a haematoma under Limbs are vulnerable to injury by glass splinters, metal
pressure; fractures, especially those with comminution and shards, bullets, shrapnel and assorted fragments from explo-
periosteal stripping; joint dislocation; nerve injury; varying sive devices. These injuries range from simple contusion and
degrees of contamination.27–30 The literature shows that puncture to complete transection of a vessel.
vascular injuries are present in 10–48 per cent of cases of The upper extremity is particularly vulnerable to stab
complex limb trauma and in this group amputations as wounds in which major vessels, and very frequently elem-
high as 85 per cent have been reported.27 ents of the brachial plexus lying in close affinity, may be
 Mechanisms of injury 377

cleanly divided but with minimal other soft tissue injury. create a degree of cavitation. The very process of cavitation
The classical self-inflicted butcher’s injury by a boning creates a suction force through the entry wound, which
knife, which may involve the external iliac artery, the com- draws in pieces of clothing, dirt and bacteria, immediately
mon femoral artery or its branches, is accompanied by tor- contaminating the wound. A relatively benign entry wound
rential and even fatal haemorrhage. conceals the severity of disruption within, but a large exit
The wounding energy of a bullet depends on its mass, its wound ought to alert the clinician.
muzzle velocity and the distance it travels before impact.4,7,8 A concentrated spread of damage by a shotgun discharged
For example, a low velocity missile (approximately 300 m/s at close range produces massive damage over a wide area of
or 1000 ft/s) will ordinarily damage structures including soft tissue, which is often underestimated during inspection
blood vessels lying directly in its path. A missile of high of the wounds. Shells, rockets, mortars and mines on the
velocity (around 750–900 m/s or 2500–3000 ft/s), dissipat- battlefield, and bombs and other explosive devices detonated
ing its energy at right angles to its trajectory, creates a tem- on a busy street by the terrorist, cause injury both by the blast
porary cavitational effect of approximately 30–40 times the wave moving faster than the speed of sound, and by metal
cross-sectional area of the bullet. The huge forces involved fragments, secondary missiles and falling masonry. The most
exceed the elastic limits of all tissues, including vessels, common iatrogenic source of penetrating vascular trauma
which are displaced, torn and obliterated well away from complicate transarterial catheterisation procedures under-
the actual path of the bullet. A humerus or tibia, for example, taken daily by cardiologists and radiologists in modem hos-
when struck by a high velocity missile will fragment into pital practice (see Chapter 38). Limb vascular trauma caused
pieces which then behave as secondary missiles, causing by high velocity missiles, bombs and shells naturally account
further soft tissue damage (Fig. 33.1). It should be remem- for much higher amputation rates than those resulting from
bered that even low velocity bullets discharged at close range stabbings and handgun injury.

Aetiology of penetrating injury

Sharp Assault Knife

object Accident Shard of glass, metal
Guns Handgun Low velocity bullet
Rifle High velocity bullet
Assault weapon High velocity bullet
Machine gun High velocity bullet
Shotgun Shot cluster or spray
Explosive Bomb Shrapnel, secondary
device missile
Mine Shrapnel
Shell, rocket, Shrapnel
mortar
Iatrogenic Cannulating Catheter, stent, milling
instrument device
Sharp instrument Scalpel, nail, pin, trocar

All penetrating limb wounds in close proximity to main

vessels must be viewed with suspicion, whether they occur
in civilian practice or as a consequence of terrorism.

Blunt

Blunt vascular trauma of the limbs usually results from

sudden deceleration in road traffic accidents and is fre-
Figure 33.1 High velocity missile injury: gross comminution of quently accompanied by other injuries. Other instances of
humerus, transection of brachial artery and ulnar nerve (reproduced such injury include falls from heights, lift and scaffolding
with permission from Barros D’Sa AAA, in Eastcott HHG (ed). Arterial crashes, as well as air, rail and mining disasters. Such
Surgery, 3rd edn. Edinburgh: Churchill Livingstone, 1992: 355–411) trauma is often of greater severity than that observed on
378 Vascular injuries of the limbs

the most severe type IIIc open fractures, extensive damage to

bone, soft tissue and long segments of artery, vein and col-
laterals as well as contamination, combine to bring about
the inevitably high amputation rates.37–39 The severity of the
associated bony injury in vascular trauma significantly influ-
ences outcome as a result of problems of bony union and of
complications such as compartment syndrome and infec-
tion, which often lead to amputation.38
Amputation rates directly related to multiple crural artery
injury are not easily found among available reports,40–42
which, in most cases, record short term results following
arterial ligation rather than repair.43,44 A low velocity missile
such as the plastic bullet has been known to cause non-pene-
trating lower limb vascular trauma.45
The high frequency of associated brachial plexus injury,
particularly in blunt avulsive trauma involving the prox-
imal upper limb, may account for a prolonged, and often
unsuccessful, functional recovery of the limb.

Aetiology of blunt injury

• Deceleration injury (high velocity)

– Road traffic accident
– Rail, air crash
– Fall
– Lift, scaffolding collapse
• Torsion, avulsion (low velocity)
– Sports injury
• Crush injury (falling masonry, etc.)
– Accident, explosions
– Mining disaster
– Natural disaster, earthquake

Figure 33.2 Angiogram shows occluded femoral artery Crushing

(arrowed) in deceleration injury of the lower limb
Crush injuries of the limbs caused by falling masonry were
sustained, notoriously, during the two world wars, but they
the battlefield and is likely to endanger the limb. Femoral
have been observed in civilian life following natural disasters
shaft fractures (Fig. 33.2) and fracture dislocations of the
such as earthquakes, bombs detonated by terrorists and in
knee, for example, carry a 10–40 per cent incidence of vas-
road traffic, farm, rail, mining and industrial accidents. The
cular injury31 and the amputation rates are correspondingly
progression of the crush syndrome, significantly causing
of the order of 32–85 per cent.32,33 The immense shearing
renal failure, can be accelerated by concomitant injury and
forces generated by the sudden violent angulation and
thrombosis of the main limb artery, especially if it happens
fracture of long bones cause disruption of adjacent vessels
to be atherosclerotic.
indirectly at points of relative fixity or directly by sharp bone
fragments.31–35 The avulsive forces in operation, particu-
larly in posterior dislocations of the knee, result in the tearing Iatrogenic
of all tissues, during which process the layers of the artery,
beginning with the intima, disrupt progressively. Notori- Iatrogenic injuries in general are dealt with in detail in
ously, and not infrequently, such a popliteal injury may not Chapters 38–40, but the following alludes to those aspects
be detected because the examining clinician fails to realise related to limb trauma.
that the dislocated knee tends to reduce spontaneously. The rising incidence of injuries of the femoral, axillary
Injuries of the popliteal trifurcation and the crural arter- and brachial arteries, as a consequence of invasive diagnos-
ies account for 10 per cent of civilian vascular trauma.36 In tic and therapeutic procedures, reflects the scale of such
 Morphology of vessel injury 379

practice in vascular, radiology and cardiology departments dislocation of the hip, subtrochanteric osteotomy and
over the past two to three decades and accounts for a signifi- osteosynthesis of an introchanteric fracture. The popliteal
cant proportion of civilian vascular trauma. Angiography vessels may be damaged during lateral meniscectomy or
and cardiac catheterisation together are responsible for knee replacement. During lumbar disc surgery lower limb
60–76 per cent of all iatrogenic vascular injuries,25,26 which flow may be compromised in the rare event of injury to the
have accelerated in proportion to the steep rise in the num- aortoiliac and iliocaval systems which may demand imme-
ber and range of interventional vascular procedures (see diate intervention.
Chapters 24, 30–32 and 38–40). Doctors in general, and In all these instances of suspected or evident iatrogenic
certainly those undertaking surgical, radiological and other arterial injury, a vascular surgeon must be sought immedi-
associated disciplines, must be cognisant of the inherent ately if serious sequelae are to be averted; these cases are
vascular risks and medicolegal implications (see Chapter potentially of medicolegal interest and therefore, must be
10) of those procedures. carefully documented.
These iatrogenic complications may be the result of poor
judgement, erroneous or unskilled technique, inaccurate
Irradiation
identification of anatomical structures or misinterpretation
of X-ray films. In contrast, difficult interventional proce-
Arteries within an irradiated field might occasionally
dures, such as percutaneous balloon angioplasty, stenting,
undergo dramatic necrosis and potentially fatal rupture of
thrombolysis and atherectomy, represent an increasingly
the wall, but this is an unusual complication of excessive
attractive alternative to surgery. The complications include
dosage and is rarely observed in current practice. More
bleeding, thromboembolism, intimal injury and dissection,
commonly, early injury to the endothelium and internal
false aneurysms and arteriovenous fistulae, and catheter
elastic lamina can be followed over the next few weeks by
retention, all of which are more common in the atheroscle-
fibrosis of the media and inflammation of the adventitia.47
rotic femoral artery.
With the passage of time the features of irradiation injury
The transaxillary approach, fortunately rarely used,
may be indistinguishable from those of atherosclerosis.
invites the risk of damage to elements of the brachial plexus.
Irradiation injury to the subclavian–axillary system may
The transbrachial approach, still popular in some cardiology
complicate treatment of breast cancer, and that of the
units, may cause problems through intimal dissection,
iliofemoral arterial system during treatment of tumours of
thrombosis and later stenosis following repair of an arteri-
the ovary, cervix and testis. Symptoms of impairment of
otomy, but an extensive collateral circulation around the
flow and even of critical limb ischaemia may ensue many
elbow protects distal flow. In cases of pre-existing superficial
years later.48
femoral artery occlusive disease, damage to the profunda
femoris artery during cannulation, and therefore to the only
remaining major source of flow to the lower extremity, may
MORPHOLOGY OF VESSEL INJURY
result in amputation. The use of the intra-aortic balloon
pump employed in the support of the failing heart, follow-
ing either acute myocardial infarction or cardiopulmonary It is important to establish precisely the nature of the arter-
bypass, may itself cause iliac and aortic dissection, throm- ial injury of an extremity for appropriate treatment and
boembolism, perforation and false aneurysm formation.46 repair.49 Cases of traumatic spasm of an artery are rare, and
Damage to the femoral vessels in the groin during sur- in most such instances at least some endothelial damage
gery for varicose veins occurs recurrently and is attribut- and contusion of the adventitia is present. More import-
able either to poor technique or to a lack of awareness of the antly, a diagnosis of traumatic arterial spasm engenders
anatomical anomalies affecting this vasculature. Arterial inactivity, which can be perilous if a limb-threatening arterial
flow to the legs is also likely to be impaired by iatrogenic injury exists. Blunt injury of a vessel, or occasionally its
injury to the major retroperitoneal vessel trunks during proximity to the path of a bullet, may cause thrombosis in
laparoscopic gynaecological interventions, cholecystectomy continuity. A frequent sequela to injury is intimal fracture,
and herniorrhaphy (see Chapter 40). which may progress to an intimal flap developing into a
A number of orthopaedic procedures, particularly in major dissection with intramural bleeding and eventual
the elderly and possibly atherosclerotic patient, can cause occlusion, especially if the tear is circumferential.
arterial injury. Chief among these is the ubiquitous total Bleeding from a laceration, whether clean or ragged, may
hip replacement, during which the external iliac and com- lead to swift exsanguination because the vessel is unable to
mon femoral arteries may be injured directly by injudi- contract circumferentially. If bleeding occurs internally,
cious retraction of the incised capsule, or indirectly by the an enlarging haematoma within fascial and bony confines
exothermic reaction caused by extrusion of the polymer will raise intracompartmental pressure to levels which can
used in preparing the acetabular base. The femoral, and occlude the artery. Alternatively, blood may continue to
in particular the deep femoral vessels, are also potentially force its way into an organising haematoma to form a false
vulnerable to injury during osteotomy for congenital aneurysm lined by endothelium within which thrombus
380 Vascular injuries of the limbs

Figure 33.3 Shrapnel injury to the lower brachial artery

resulting in a false aneurysm

Figure 33.5 Shell fragment injury of the tibioperoneal trunk

with a massive pulsatile false aneurysm in the calf; a healed tibial
fracture with a large defect is also seen (reproduced with
permission from Borros D’Sa AAB in Eastcott HHG (ed.). Arterial
Surgery, 3rd edn. Edinburgh: Churchill Livingstone, 1992: 355–411)

A false aneurysm may also develop in association with a

fistula (Fig. 33.7). If the rate of arteriovenous shunting is
marked, a high output cardiac state will manifest itself in
due course.
If a vessel is transected, either cleanly by a knife or by a
high velocity missile or shrapnel, the free ends constrict,
retract and are usually sealed by a plug of thrombus. In a
traction injury, classically observed in posterior disloca-
tions of the knee, the avulsive forces involved relentlessly
stretch the intima, which then fractures and curls back on
itself, to be followed similarly by disruption of the media
and adventitia, and as the ends retract, thrombosis ensures
occlusion; on clinical examination there may be very little
external evidence of the vascular injury.

Figure 33.4 Penetrating injury of the femoral artery with a false

aneurysm causing extrinsic pressure on the artery Morphology of arterial injury

will form (Figs 33.3–33.5). An expanding false aneurysm

• Traumatic spasm, contusion

may compress surrounding structures, particularly venous

• Intimal fracture, flap

channels, and it may eventually rupture. If the penetrating

• Dissection, intramural bleeding

injury involves both artery and adjacent vein, the resulting

• Thrombosis in continuity

arteriovenous fistula is often missed on initial clinical exam-

• Puncture, laceration

ination, although it ought to be recognised by the presence

• Occlusion by pressure haematoma

of a continuous thrill and audible bruit21 (Fig. 33.6).

• False aneurysm

Arterial flow to the limb, and equally drainage from it distal

• Arteriovenous fistula

to an arteriovenous fistula tends to be compromised.

• Transection, traction and avulsion
 Pathophysiology of limb vascular injury 381

Figure 33.6 (a) Arteriovenous fistula at the

(a) (b) popliteal trifurcation. (b) The digital subtraction
film clearly demonstrates the fistulous site

PATHOPHYSIOLOGY OF LIMB VASCULAR

INJURY

Complex biochemical and cellular pathophysiological

mechanisms provoked by limb vascular injury (see Chapter
2) bring about ischaemia-reperfusion injury (IRI) of the
muscles.50–63 Arterial injury arrests distal flow and results
in ischaemia, tissue hypoperfusion and hypoxia. Striated
muscle is vulnerable to continued warm ischaemia, and
after 6–8 hours, depending on level of injury, collateral
flow and the degree of hypovolaemic shock leads in most
cases to myonecrosis and amputation.17,56 Adenosine
triphosphate (ATP) levels deplete, xanthine dehydrogenase
is converted to xanthine oxidase causing a rise in hypoxan-
thine and xanthine. With reperfusion of ischaemic tissue
oxygen free radicals are generated51–54,57–60 liberating pro-
inflammatory mediators which cause neutrophil activation
and adhesion, and by lipid peroxidation and directly
attacking endothelial cell membranes, the permeability of
the microvasculature rises,58–63 bringing about muscle
oedema and the ‘no-reflow phenomenon’.64 Arachidonic
acid is metabolised under the action of phospholipase A2
producing potent vasoactive eicosanoids such as throm-
boxane A2, prostaglandins and leucotrienes which strongly
mediate in the pathophysiology of IRI.50 The intensity
of this process of IRI is directly proportional to the dur-
Figure 33.7 Digital subtraction angiogram showing an ation of preceding ischaemia, and will dictate the degree
arteriovenous fistula of the upper femoral vessels with associated of muscle oedema and consequent necrosis and loss of
false aneurysms function.53,54
382 Vascular injuries of the limbs

Pathophysiological sequelae of arterial

injury

• Ischaemia, hypoperfusion, hypoxia

• Hypovolaemic shock, vasoconstriction
• Ischaemia-reperfusion injury (IRI)
• Systemic inflammatory response syndrome (SIRS)
• Remote injury to gut, lung, heart, kidneys, liver, brain
• Multiple organ dysfunction syndrome (MODS)
• Multiple organ failure (MOF)

The injurious effects of IRI on gut mucosa, where xanthine

oxidase is freely available,53,54,65 produce a general release
of cytokines and bacterial translocation.66 (see Chapters 2
and 3) Portal and systemic endotoxaemia63,67,68 and the sep-
sis syndrome69,70 are not infrequent sequelae. The pul-
monary consequences of IRI resemble acute respiratory
distress syndrome (ARDS)71 manifested by raised microvas-
cular permeability and neutrophil sequestration.58 Cardiac
effects include myocardial depression, hyperkalaemic dys-
rhythmias and even cardiac arrest.50,72 Oxygen free radicals
are also implicated in IRI of cerebral tissue.57 The remote sys- Figure 33.8 Penetrating trauma causing gas gangrene
temic ramifications of IRI may bring about MODS progress-
ing to MOF and death73 (see Chapter 4).
Oedema and raised pressures within inelastic fascial of methicillin-resistant Staphylococcus aureus (MRSA) which
compartments rise more steeply in severe bone and soft plagues wound sites in vascular surgical departments around
tissue injury, and may reach levels which obstruct flow the world.75
through the main arteries. It has been proved clearly at our
centre that delay or failure to ensure outflow through an
INITIAL MANAGEMENT
injured vein, either by clamping or ligation, not only exacer-
bates the effects of IRI in leg muscle, but also causes remote
lung injury characterised by non-cardiogenic pulmonary The wounding energy dissipated on impact strongly dic-
oedema.74 The sequelae of compartment syndrome, tates the nature and extent of injury to blood vessels and
microvascular stasis and thrombosis, aseptic muscle necro- other structures and consequently the degree of risk to life,
sis, ischaemic nerve palsy, Volkmann contracture and organ and limb. In cases of major injury, survival is deter-
amputation inevitably follow49 unless timely and effective mined by speed of resuscitation at the scene, en route to
fasciotomy is undertaken. hospital, in the accident and emergency department and in
The adverse effects of IRI can be further aggravated by the operating theatre, in concert with definitive surgery.
bacterial contamination, usually by Gram-positive cocci In the exsanguinating, and especially in the multiply
and Gram-negative cocci and bacilli, some of which act syn- injured patient, standard resuscitative measures are taken
ergistically to cause cellulitis or fasciitis, leaving an under- to ensure airway, ventilation, and correction of hypo-
lying repair open to breakdown and secondary haemorrhage. volaemic shock. Control of external bleeding in the upper
The anaerobic environment of ischaemic tissue facilitates limb is easy given the superficial anatomical position of the
the regeneration of clostridial spores (Clostridium welchii, arteries. Digital compression and the use of pad and ban-
C. novyi and C. septicum) to produce gas gangrene (Fig. dage is the best option in the leg; the blind application of
33.8), with its classic features of tense oedema, crepitus, arterial forceps or clamps, particularly by the inexperi-
frothy brown watery exudate, brick-red necrotic muscle, enced, compounds vessel damage and endangers adjacent
toxaemia and cardiovascular collapse. Such an outcome is nerves. A poorly applied tourniquet will accelerate bleed-
possible in cases of prolonged ischaemia due to delay in ing and if too tight or not released in time will cause per-
exploration, suboptimal wound care in complex trauma, manent damage to vessel and nerve.
ligation of vessels and compartmental hypertension. Information should be sought from the patient or associ-
Superinfection, particularly by Pseudomonas aeruginosa, ates, for example, about the nature of the wounding agent, the
of fasciotomy incisions undertaken to relieve the effects of muzzle velocity of a gun and the distance from which it was
IRI raises the likelihood of amputation.11 That is even truer fired, the amount of blood lost, and the time interval between
 Diagnosis and assessment 383

injury and commencement of treatment. In all penetrat-

ing injuries, tetanus toxoid, prophylactic cefuroxime and
metronidazole are administered routinely and, if gas gangrene
seems likely, specific antitoxin therapy and penicillin may be
given. Analgesia is doubly useful: as well as relieving pain, nar-
cotic drugs reduce vasospasm and enhance distal flow to the
leg but caution is required in the hypotensive patient.

DIAGNOSIS AND ASSESSMENT

Clinical examination

It is vitally important that clinician should run through a

comprehensive mental checklist so as not to miss an arter-
ial injury.
1 If bleeding is continuous, is it mainly arterial or venous?
2 If a haematoma is present, is it expanding or pulsatile?
3 Is there a thrill or an audible bruit indicative of marked
stenosis or an arteriovenous fistula?
4 Are the universally recognised ‘hard’ signs of
ischaemia, such as absent distal pulses, pallor,
mottling, coolness and numbness, present?
5 If not, are there ‘soft’ signs, features often overlooked
even by the experienced observer, such as transient
ischaemia, minimal neurological deficit or a small non-
expanding haematoma?76
As the lower limb has a poorer collateral network than that
in the arm, ankle pulses are less likely to be palpable fol-
lowing vascular injury. Signs of arterial injury may well be
obscured in the multiply injured and shocked patient, but
it ought to be suspected if, following resuscitation, circula- Figure 33.9 Penetrating injury in proximity to lower superficial
femoral artery. Distal pulses were present but were reduced in
tory recovery in one limb lags behind the other.
volume, not by spasm but by mural damage that had progressed to
In assessing closed fractures caused by blunt injury, the
thrombotic occlusion. The thrombus was visible on the angiogram
index of suspicion of possible vascular injury tends to be and on the excised segment (inset) (reproduced with permission
lower than that in penetrating trauma. In most blunt from Barros D’Sa AAB. A decade of missile-induced vascular
injuries reduction, and therefore correction, of the align- trauma. Ann R Coll Surg Engl 1982; 64: 37–44)
ment of long bones, supported by general resuscitation,
restores distal flow effectively. If the slightest doubt of arterial
injury persists, angiography should be undertaken without Ultrasound
delay. The literature is replete with examples of permanent
disability or amputation following a presumed diagnosis of Routine Doppler ultrasound examination complements
spasm which, on subsequent angiographic scrutiny, invari- physical examination but calls for an intelligent interpret-
ably revealed intimal damage and thrombosis35 (Fig. 33.9). ation of the findings. For example, an audible Doppler signal
Notoriously, vascular injury accompanying dislocations of provides little reassurance that the proximal artery is intact.
the knee may be missed because of the tendency of the Much more helpful is the measurement of Doppler pressures
joint to reduce spontaneously, thereby restoring the nor- and estimation of the ankle:brachial pressure index, which
mal contour of the leg. The absence of signs of bleeding can be compared with that in the contralateral uninjured leg.
within the tissues owing to constriction of the avulsed ends An ankle:brachial pressure index below 0.90 ought to be
of the popliteal artery may further induce a false sense of viewed with suspicion. Duplex scan imaging can detect and
security. This well known pitfall in clinical examination localise injuries of the femoral and popliteal arteries, espe-
may delay diagnosis, by which time distal thrombosis may cially when these vessels lie in proximity to the path of a bullet
have progressed to a point where a distinctly unfavourable or knife. Nevertheless, non-invasive investigations take time
outcome is inevitable. and are impracticable in the presence of skin loss, fracture
384 Vascular injuries of the limbs

deformity and pain. Angiography, however, can demonstrate single-plate angiogram similar to the on-table technique is
an arterial injury expeditiously and forms a sound and reli- perfectly acceptable.
able basis for operative intervention.

PREOPERATIVE CONSIDERATIONS
Angiography
The non-salvageable limb
The competence of preoperative angiography in delineating
an arterial injury, or in excluding it, is well established, par-
With the expeditious application of the basic principles of
ticularly in penetrating leg injuries76,77 (see Figs. 33.3–33.7,
surgery it remains within the skills of the vascular surgeon
33.9). A positive angiogram is almost invariably a manda-
to salvage some of the critically injured limbs. Successes in
tory indication for exploration, except for very minor
such cases may induce a degree of overoptimism in man-
lesions. A non-operative approach in occult injuries requires
aging the mutilated and irreparable limb. Misdirected zeal
supervision, possibly further angiography and other studies,
may commit both surgeon and patient to a protracted
and sometimes delayed surgical treatment, all of which can
series of injudicious operations on an irretrievably man-
be expensive.78 A long term study, involving reasonable
gled limb, inviting complications such as infection, poten-
numbers, appraising the conservative treatment of minor
tially fatal secondary haemorrhage, poor rehabilitation and
arterial lesions defined by angiography and left unexplored,
eventually amputation of an insensate appendage.
is awaited. On the other hand, with a few exceptions, a neg-
The notion that every limb must be saved at all costs
ative angiogram gives the surgeon the necessary confidence
should be questioned: a more objective reappraisal of the
not to intervene, thus reducing the incidence of worthless
condition of a limb on admission is required, and in some
explorations.
cases it may be more prudent to proceed to primary ampu-
In penetrating wounds the yield of arterial injuries
tation and early rehabilitation with a prosthesis. An equally
detected on angiography located in close proximity to the
flawed approach would be to apply rigid guidelines for pri-
femoropopliteal system and trifurcation is low,79,80 and the
mary amputation based on the type of wounding agent, the
fact that it is an invasive and expensive procedure becomes
duration of ischaemia, the presence of injuries elsewhere or
an argument for avoiding it.81 In deciding on the need for
indeed medicolegal, social or budgetary considerations.
angiography, however, the question of the proximity of the
The introduction of various scoring systems using clinical
wound to the damaged artery is important. For example, in
criteria aimed at predicting outcome83,84 have a limited
a high velocity missile causing injury well outside its path,
role but may tip the balance when sound clinical judge-
the presence of ‘soft’ signs of arterial injury, and not least
ment is called for in a difficult case. The key reasons for
the potential medicolegal consequences of limb loss result-
early amputation are irreversible ischaemia, failed vascular
ing from a missed injury, may point to the need for angiog-
repair and sepsis, the latter two of which cannot be reliably
raphy. Where doubt exists in the stable patient, it might be
predicted, either at admission or intraoperatively.
considered negligent not to obtain an angiogram.
Dilemmas of this kind do not apply when primary
When a surgeon does not have access to angiography
amputation represents no more than the completion of a
and is compelled to rely on clinical acumen, there is no sub-
traumatic amputation or the excision of a severely crushed
stitute for meticulous and repeated physical examination.
limb. This scenario is most notoriously illustrated by the
Reliance on clinical examination alone to detect arterial
injuries resulting from the detonation of antipersonnel
injury can be most effective,82 but if clinical judgement is
mines, which are a feature of so many conflicts and which
poor, a vascular injury may well be missed and on occasion
remain a deadly threat long after the dust of war has set-
operative exploration will be fruitless. Should ischaemia
tled.85,86 In these circumstances the surgeon makes every
persist after reduction of a femoral fracture, especially in the
effort to preserve as much viable tissue and skin as possible
elderly atherosclerotic patient, timely angiography and the
for delayed closure, at which time a satisfactory stump
discovery of an injury may help to avert disaster. The high
depends on careful tailoring of the flaps.
incidence of occult and limb-threatening arterial injury
associated with dislocations of the knee, missed at clinical
examination, is a compelling argument in favour of manda- The importance of time
tory angiography in these cases.
Biplane films defining the site and type of injury assist An expeditious approach to treating a limb vascular injury is
in planning an operative approach. Fine catheter digital essential as the duration of ischaemia is pivotal to outcome.
subtraction techniques have largely displaced conventional It may be but one feature of the multiply injured patient and
angiography and permit clear definition of a vascular injury may not deserve priority during resuscitation and definitive
against a background unobscured by bone (see Fig. 33.7). surgery for life-threatening injuries of the head, chest and
In the absence of such sophisticated equipment, the percu- abdomen. Once haemodynamic stability has been restored,
taneous cannulation of the femoral artery for a one-shot, attention can be focused on the limb vascular injury. A large
 Preoperative considerations 385

survey of severe head injuries in association with concomi-

tant limb bone and vessel trauma87 has shown that the limb
injury may be treated before a head injury except in the case
of an extradural haemorrhage which constitutes an absolute
emergency.
An assortment of variables has to be taken into consid-
eration, namely the level of arterial injury, the quality
of collateral flow, which is much better in the upper limb,
the individual patient’s tolerance of ischaemia, the degree
of hypotension and the extent of associated soft tissue
and bone injuries. The relentless progression of warm
ischaemia time and its deleterious effect on striated muscle
is forewarning of an adverse result, and in some cases limb
survival is achieved at the cost of permanent neurological
impairment. Figure 33.10 Indwelling shunts: Brener in torn popliteal artery
In treating limb vascular injuries, and in particular above, and Javid in transected popliteal vein below, preparatory to
those with severe open wounds, control of bleeding, resus- vein graft replacement of each vessel (reproduced with permission
citation and definitive surgery must be regarded as over- from Barros D’Sa in Greenhalgh RM (ed.). Limb Salvage and
lapping rather than sequential stages of management, with Amputation in Vascular Disease. London: WB Saunders, 1998:
135–50)
the objective of shortening ischaemia time. Nevertheless, a
finite period of time is required for the exposure and con-
trol of vessels, identification of nerves, debridement and
wound care, attention to complex bone injuries and finally
the harvesting and preparation of donor vein to be used in
arterial and venous repair.
A heightened awareness of the consequences of pro-
longed ischaemia, and an understandable desire to press
on with surgery, may itself cause lapses in surgical prin-
ciples and in operative technique, with counterproductive
results. In this particular climate of urgency, for example,
that vascular repair might be performed before a fracture
is stabilised, a practice that tends to encourage the use of
quick but often flawed repair techniques, including lateral
suture and end-to-end anastomosis under tension, both of
which may fail. Large venous channels essential to drainage Figure 33.11 Javid shunt bridging a lengthy gap in the femoral
may be ligated simply to save time. There also remains the artery and perfusing the distal limb; another such shunt is bridging
likelihood that the eventual reduction of a fracture, which a similar gap in the adjoining femoral vein and draining the limb.
often requires the robust manipulation of bone fragments, The ends of a fractured femur (XX) are being manipulated prior to
might disrupt a delicate vascular repair. fixation (reproduced with permission from Barros D’Sa AAB,
Clearly, the merits of reducing a fracture before vascular Moorehead RJ. Combined arterial and venous intraluminal shunting
repair are obvious: damage of soft tissues and vessels by bone in major trauma of the lower limb. Eur J Vasc Surg 1989; 3: 577–81)
fragments is averted, the measurement of vein graft length to
restore arterial and venous flow will be optimal, and repair temporarily bridged by the early placement of intraluminal
can be undertaken with the confidence that it will not be dis- shunts (Fig. 33.10). With shunts in place, vascular and
turbed. This approach, however, means that stabilisation of orthopaedic surgeons no longer have to watch the clock
one or more fractures will take time. The severity of long anxiously and instead a harmonious multidisciplinary
bone damage caused by high velocity missiles and bombs approach is fostered (Fig. 33.11). The presence of shunts
demands some form of fixation so that its exact length is allows the implementation of an unhurried logical sequence
established before vascular repair. The consequent prolonga- of manoeuvres, a practice underpinned by over two decades
tion of ischaemia may place the orthopaedic surgeon under of experience and which has proved helpful in lowering the
some pressure to carry out a hurried, and perhaps technically incidence of compartment syndrome, that of fasciotomy in
imperfect, fixation bringing with it the potential immediate relieving it, and failing that the incidence of ischaemic nerve
risk of compromising a vascular repair and in the longer term palsy, contracture, sepsis, amputation and duration of stay
raising the possibility of delayed union or non-union. in hospital.28–30,49,88–93 In the few borderline cases of poten-
All these concerns about the order of arterial and bone tial primary amputation, and only when nerve continuity
repair are dispelled if the injured artery and vein are has been confirmed to be macroscopically intact, shunts
386 Vascular injuries of the limbs

Table 33.1 Comparisons between the pre-shunt (1969–1978) and post-shunt (1979–2000) periods of managing
complex limb vascular injuries, penetrating and blunt, in terms of the incidence of fasciotomy, contracture and amputation

Pre-shunt (n  34) Post-shunt (n  57)

Complex penetrating No. Per cent No. Per cent P value OR 95% CI

Fasciotomy 17/30 56.7 13/48 27.1 0.016 3.5 1.3–9.2

Contracture 13/30 43.3 8/47 17.0 0.018 3.7 1.3–10.6
Amputation 11/34 32.4 5/57 8.8 0.009 5.0 1.6–16.0

Pre-shunt (n  38) Post-shunt (n  49)

Complex blunt No. Per cent No. Per cent P value OR 95% CI

Fasciotomy 22/35 62.9 16/44 36.4 0.02 3.0 1.2–7.4

Contracture 19/34 55.9 13/45 28.9 0.02 3.1 1.2–7.9
Amputation 15/38 39.5 7/49 14.3 0.012 3.9 1.4–11.0

The incidence of all three parameters was significantly reduced following the introduction of the policy of early intraluminal shunting
of artery and vein in 1979 in both the penetrating and blunt injury groups of complex limb vascular injuries. Data were analysed using
the statistical package SPSS version 10.0 for windows (SPSS Inc., Chicago, IL, USA). Clinical characteristics of the two study groups
were compared and analysed by 2 test (Yates’ corrected), and Fisher’s exact test (when appropriate). Corresponding odds ratio (OR)
and 95% confidence interval (95% CI) values were calculated. Variables were considered statistically significant at P  0.05.

may permit more prudent assessment as to whether the limb DEFINITIVE SURGICAL TREATMENT
is salvageable with the prospect of a reasonable functional
outcome.
These shunts have made a special contribution to the Incisions and exposures
management of the very challenging subgroup of ‘com-
UPPER LIMB
plex’ limb vascular injuries, both penetrating and blunt, at
our centre. A review of the incidence of fasciotomy proced- The patient lies supine, the injured upper limb being
ures undertaken and the outcome in terms of the incidence abducted and extended palm upwards at right angles to the
of contracture and amputation are illustrated in Table body. Peripheral and central venous lines should not be
33.1. The incidence of each of these parameters prior to the inserted on this side. The draping should provide an oper-
introduction of shunts (pre-shunt period: 1969–78) was ative field with access to chest, neck and arm, as well as to
significantly higher than that after the policy of shunting one lower limb in case donor vein is required.
was put into effect (post-shunt period: 1979–2000) in both For stab wounds involving the first part of the axillary
trauma categories.28,88 artery, it is wise to secure proximal supraclavicular control
of the subclavian artery. To approach the axillary artery the
The problems in treating ‘complex’ limb incision commences below the clavicle, extending laterally
vascular injury down the deltopectoral groove and continues distally along
the course of the brachial artery. Access to the axillary artery
is possible either above or below the margins of the pec-
• Vessels are injured in 10–48 per cent of limb injuries
toralis major, or by separating the fibres of this muscle and
• 85 per cent of those limbs are amputated and fatality
dividing the pectoralis minor tendon. The intimate prox-
is not uncommon
imity of veins and nerves requires meticulous care during
• Treatment of damaged artery, vein, bone and soft
dissection. A standard longitudinal incision will expose an
tissue takes time
injured brachial artery in the upper arm or at the elbow; the
• Prolonged ischaemia and impaired venous outflow
latter is sometimes necessary in dealing with intimal frac-
aggravate IRI
ture, dissection and thrombosis following diagnostic can-
• Intensity of IRI influences outcome
nulation. Short longitudinal incisions over the radial and
• Compartment syndrome, ischaemic nerve palsy
ulnar arteries are quite adequate.
• Increased need for fasciotomy
• Climate of urgency induces lapses and flaws in
operative care LOWER LIMB
• Greater danger of myonecrosis, contracture, sepsis
For most lower limb vascular injuries the patient lies supine,
and amputation
both lower limbs being prepared and draped, permitting
 Definitive surgical treatment 387

proximal and distal extension of standard longitudinal

incisions in the injured leg, and enabling the procurement
of donor vein from the contralateral leg. Vascular injuries
of the groin are associated with torrential bleeding, for
which control of the external iliac artery has to be estab-
lished through an oblique muscle-splitting incision for
retroperitoneal access above the inguinal ligament. While
this is in progress digital pressure is maintained over the
bleeding artery and is only released after the external iliac
artery has been clamped. A longitudinal groin incision will
then enable exposure and control of the common femoral
artery, and for injuries at its bifurcation, encirclement and
control of the superficial femoral and the profunda femoris
artery is essential.
The popliteal artery can be approached medially or pos- Figure 33.12 Outlying shunt picking up flow proximal to the
teriorly, the former preserving the adjacent long saphenous injured segment and revitalising the limb distally
vein. External rotation of the thigh supported laterally by a
pack of two sterile gowns and some knee flexion provides Shunting and operative discipline
ideal medial exposure of the popliteal vessels, especially
after dividing the medial head of the gastrocnemius. The The concept of intraluminal shunting has been applied in
posterior approach with the patient being prone is rarely the past to coronary artery perfusion, is regularly used for
used: a gentle S-shaped incision is made in an oblique axis carotid endarterectomy, on occasion in replacement of the
commencing medially above and proceeding laterally thoracic aorta, and rarely, in retrohepatic caval trauma. In
downwards to expose the mid-popliteal vessels in the floor the proximal upper limb arterial injuries the protective
of the popliteal fossa, the short saphenous vein and tibial collateral network makes shunts unnecessary. In military
nerve having been identified and preserved. A medial trauma of the brachial artery, especially involving the deep
approach is used for access to the popliteal trifurcation, brachial artery6 shunting does have some value.
and separate longitudinal incisions are employed in order An indwelling shunt reconnecting the ends of a tran-
to expose segments of the crural arteries all the way down sected femoral or popliteal artery immediately arrests
to the foot. ischaemia, minimises reperfusion injury, keeps compart-
In complex injuries requiring orthopaedic intervention ment pressure at a level which reduces the need for fas-
the incisional approaches must be such as to facilitate co- ciotomy and consequently buys time for a considered
operative effort. approach to the rest of the operation. In extensive soft tissue
injuries with destruction of lengthy segments of vessels long
outlying shunts will keep the distal limb alive (Fig. 33.12). If
Control and preparation of vessels in a multiply injured patient undergoing surgery for life-
threatening injuries of the head or torso, space and condi-
Exposure of the injured vessels and digital control of bleed- tions around the operating table also allow the vascular
ing is followed by sharp dissection of the segments of artery surgeon to work unobtrusively, injured vessels can be
and vein above and below the injury before clamping. A exposed and shunted to keep the limb viable for definitive
sufficient length of artery on either side of the injury is repair later. In the case of head injuries that timely interven-
exposed and if necessary trimmed back to a point where tion minimises the local and remote impact of IRI, but if
the wall is intact. delayed, the sequelae of IRI might represent a formidable
Release of the upper clamp will usually allow a plug of further assault on the already traumatised brain.87
thrombus to be washed out and occasionally a balloon A severed femoral or popliteal vein cleared of clot is
catheter is required. If ischaemia is prolonged distal balloon shunted to re-establish drainage, particularly if damage to
exploration will often recover clot, particularly when back venous collaterals is extensive (see Figs 33.10 and 33.11). If
bleeding is poor or absent. In the latter circumstance vigor- the main venous channel is clamped rather than shunted,
ous upward milking of the limb distal to the injury may be while the distal limb continues to be perfused via an intact
helpful in expressing recent propagated clot. The young or shunted artery, a precarious rise in compartment pres-
patient does not possess an adequate collateral circulation sure and remote lung injury will supervene.74
and back bleeding is limited; by the same token, arterial One of several commercially available shunts may be
repair is imperative if the limb is to be saved. The distal employed, but if none is available, silicone elastomer or
artery is perfused with heparinised saline (20 units/mL); plastic tubing of suitable consistency, length and calibre is
systemic heparinisation is acceptable if there are no other an acceptable alternative provided that the ends are very
injuries. smoothly trimmed to avoid intimal damage. The ideal design
388 Vascular injuries of the limbs

Scoop
Syringe Snip

Shunt Staunch

Survey
Scission
Squirt

Split Stabilise

Suture Stitch
Swing
Figure 33.13 After excision of devitalised muscle, debridement
and stabilisation of fracture (XX) interposed vein grafts restore flow
through the femoral artery and vein and the deep femoral vein Figure 33.14 Aide mémoire for the Sequence of Steps in the
(reproduced with permission from Barros D’Sa AAB, Moorehead, RJ. operative management of complex limb vascular injury: staunch
Combined arterial and venous intraluminal shunting in major the bleeding, snip damaged ends of vessels, scoop out clot, syringe
trauma of the lower limb. Eur J Vasc Surg 1989; 3: 577–81) in heparinised saline, shunt both artery and vein, survey the wound
and identify nerve injury, perform scission of non-viable soft tissue,
squirt saline to irrigate wound, stabilise fractured bones, stitch
of a shunt system, purpose-built for vascular trauma,89 was vessel grafts, swing tissue for cover, suture the wound (delayed
based on experience in Northern Ireland27–30,49,74,88–93 and primary if contaminated) and, if necessary, split fascia to
led to useful experimental work on a temporary shunt.94 The decompress muscle (redrawn with permission from Barros D’Sa
Brener shunt (see Fig. 33.10) in particular has some value in AAB. Complex vascular and orthopaedic limb injuries [editorial].
that it has a side arm: placed in an artery, it forms a conveni- J Bone Joint Surg (UK) 1992; 74: 176–8)
ent portal for blood sampling, blood gas estimation and
injection of anticoagulants or contrast for on-table angiog- dividends accruing from this systematic approach, based on
raphy; placed in a vein it can act as an outlet for flushing out the use of shunts, particularly in the management of complex
stagnant blood of low pH, rich in potassium and toxic vascular trauma, are salutary28 (Table 33.1). They include
metabolites, from the distal ischaemic limb and thereby pro- improved operative technique, lower incidence of complica-
tecting the myocardium in particular. tions, better outcome and earlier discharge from hospital.
With both arterial and venous shunts in place, the vascu-
lar surgeon has ample time to survey the wound, identify and Benefits of shunting artery and vein
tag nerves, remove debris and irrigate the tissues. Restoration
of flow enables a sharp distinction to be made between viable
and dead tissue for more precise excision, and haemostasis is • Immediately restores arterial inflow and venous
outflow
more reliably achieved. Attention can then be focused on the
restoration of skeletal integrity using either internal or exter- • Minimises IRI and its consequences
nal fixation: the realigned limb is then ready for definitive • Encourages a logical sequence of operative steps
vascular repair, which on completion is certain to remain • Buys time for meticulous wound care
secure and undisturbed (Fig. 33.13). When both the injured • Buys time for ideal method of bone fixation
artery and vein have been shunted, it is immaterial whether • Buys time for correct choice of arterial reconstruction
one or other is repaired first, in this way dispelling past • Vein graft of optimum length harvested for use
debate on priority for repair. Sufficient time is also available • Compound vein graft constructed if required
to harvest vein graft and, if necessary, compound vein grafts, • Extra-anatomical bypass created if tissue
lost/contaminated
either of the panel or spiral type, can be fashioned to match
the calibre of the host vessel. The shunt offers a further • Major vein invariably reconstructed
advantage by acting as a stent, which facilitates precise sutur- • Lower incidence of failure of repair/thrombosis
ing. When an outlying shunt restores distal flow in cases of • Reduced need for fasciotomy
extensive injury of the limb, time can be taken to construct • Lowered incidence of sepsis, contracture, amputation
an extra-anatomic vein bypass through clean, unaffected • Fosters harmonious multidisciplinary cooperation
tissue at some distance from the wound.
Over two decades of experience at our centre has shown Wound care
that temporary shunting fosters a disciplined and methodical
operative routine. The key operative steps are summarised A knife wound requires minimal excision of the skin and
in a simple alliterative aide mémoire (Fig. 33.14). The clear tissues around its track. The true extent of tissue damage in
 Definitive surgical treatment 389

gunshot wounds, particularly those caused by high velocity presence of adequate collateral flow should be dismissed.
bullets, cannot always be reliably gauged from an inspec- Brachial artery injuries should not be regarded as innocuous
tion of the entry and exit wounds. That assessment becomes and left to the novice for repair because the opportunities to
even more difficult in blunt deceleration injuries, particu- obtain a good result diminish with each repeated attempt.
larly when these are closed injuries. In low velocity wounds, The Allen test should be used in determining the import-
except those caused by guns fired at very close range, only a ance of repairing injured radial and ulnar arteries.
limited amount of excision is required, whereas in high Injury to the tibial or crural arteries is unlikely to result in
velocity wounds debridement must be adequate if infection limb-threatening ischaemia unless more than one artery is
is to be averted. Devitalised muscle, easily recognised by its involved; in one report limb loss doubled when these arter-
deep purplish colour and failure to bleed or contract, must ies were ligated rather than repaired.36 Neither the literature
be completely excised (see Fig. 33.13). Attached bone frag- nor a trawl of vascular surgical opinion offers a standard
ments, debris, dirt and foreign bodies should be removed protocol for the repair of tibial arteries accompanying frac-
meticulously; this should be followed by copious irrigation, tures of the tibia.95 Although it would be correct in principle
preferably pulsatile, to remove remaining contaminants to reconstruct injured tibial arteries in order to prevent
and to lower the concentration of the bacterial inoculum. islands of ischaemia in muscle and bone, in general two are
Cursory care inevitably leads to infection, which remains sufficient, if not essential, for limb viability and good func-
one of the major causes of amputation. tion.95–97 This view is preferable to one which holds that a
A contused nerve is simply left alone, but if the transec- single Doppler pulse elicited at the ankle is sufficient,98,99 or
tion is clean and the wound is not contaminated, accurate to another that either the posterior tibial or anterior tibial
primary repair is advisable. If damage is extensive, the should be repaired, but that the peroneal artery may be
sheath of each nerve end is simply tagged for identification safely ignored.100 If any injured tibial artery is exposed, the
for delayed secondary repair as an elective procedure. opportunity should be taken to repair it unless the patient’s
general condition precludes taking time to do so. This
Management of associated fractures rationale has particular merit in situations of severe damage
to soft tissue and collaterals, as for example in type IIIc tibial
Until the management of combined bone and arterial fractures, when the chances of propagated thrombosis in a
injuries was streamlined on the basis of shunting, vascular partially damaged crural artery may convert an initially
and orthopaedic surgeons at our centre dealt with these viable leg into one that is beyond salvage. The reliance of
complex injuries in random fashion. The policy formulated fractured, and especially denuded bone, on blood supply
around the use of intraluminal shunts has established close from adjacent crural arteries cannot be overstated. In recent
cooperation between them during the perioperative period years an increasing enthusiasm for repair of these vessels has
and especially in the operating theatre.27–30 Agreed inci- proved rewarding irrespective of the poor circulatory state
sional approaches and adherence to the sequence of steps of the foot.
indicated in Fig. 33.14 has allowed each specialist to work Various morphological types of vessel injury occur, and
unhindered. Exhortations to repair the artery before bone in practice different permutations of these may be present in
fixation are attended by the real fears of suture line disrup- one damaged vessel. The optimal type of repair will depend
tion during the robust manipulations necessary to achieve on the particular circumstances of each case.
fracture reduction, and that approach is virtually obsolete
at our centre. Those who advocate vascular repair before LATERAL SUTURE
ensuring skeletal stability cannot reasonably dispute the
Closure of small puncture wounds by lateral suture is
view that it is sound practice to first stabilise bone, thereby
acceptable, especially if executed transversely using inter-
permitting optimum repair of artery and vein, confident in
rupted sutures (Fig. 33.15). Equally, transverse and short
the knowledge that they will not be disrupted (see Fig.
oblique lacerations with sharp edges can be repaired in this
33.13). In turn, particularly in complex limb trauma, the
manner. If, however, the laceration is longitudinal, and
protocol centred on the use of shunts has encouraged har-
particularly if contusion demands excision of the margins,
mony between surgeons of different disciplines.
lateral suture simply narrows the lumen (see Fig. 33.15)
and encourages thrombosis, notably in critical vessels such
Arterial repair as the brachial and popliteal, endangering the extremity.
Damaged minor arteries and veins which do not threaten
PATCH ANGIOPLASTY
the viability of the limb may be ligated but in principle all
damaged vessels should be repaired. This precept is particu- In order to preserve the calibre of the artery, an angioplasty
larly applicable to the small calibre brachial or femoral using a vein patch is usually preferable to a lateral suture
artery of a child whose limbs require sufficient blood flow (Fig. 33.16). In a closed or uncontaminated injury to a
for proper growth and development. large vessel, such as the iliac or common femoral, it would
The presumption that axillary or brachial artery ligation be quite acceptable to employ a Sauvage filamentous pros-
will not lead to upper limb amputation due to the protective thetic patch graft (see Fig. 33.16) in the knowledge that it
390 Vascular injuries of the limbs

Figure 33.15 Oblique and

transverse suture of small
lacerations and puncture
wounds. Longitudinal suture
may narrow the lumen

will heal with an endothelialised flow surface within a

matter of a few weeks.101

DIRECT ANASTOMOSIS
In cases of clean stab wounds of the artery or if there is
limited loss of length after excision of the edges, direct
end-to-end anastomosis may well be possible (Fig. 33.17).
Conversely, if there is any tension at the anastomosis (see
Fig. 33.17), thrombotic failure or actual disruption may fol-
low. Of course, a joint should not have to be flexed or major
collaterals divided, particularly in an atherosclerotic artery, to
enable approximation. These problems are of particular rele-
vance to the axillary and popliteal arteries. In such instances
excision and vein grafting is by far the better approach.

VEIN GRAFTS
101
Figure 33.16 Vein and prosthetic (Sauvage) patch The attributes of autogenous vein include durability,
angioplasty resistance to infection and an ability to draw nutrient flow
from surrounding viable tissue. The long term patency of
vein makes it a most desirable graft, a fact which has some
resonance with the predominantly younger patient who
sustains injury and stands to benefit from it.
The abundant availability of donor vein gives the sur-
geon the confidence to excise adequate, and sometimes
lengthy, segments of damaged vessel in order to leave pris-
tine ends for reconstruction (Fig. 33.18). This is especially
relevant to high velocity bullet or traction injuries involving
the popliteal vessels in which the outer appearances some-
times belie the true extent of injury to the intima. The long
saphenous vein represents the best source for donor vein,
the calibre in the upper thigh being ideal for interposition
grafting (see Fig. 33.18), while segments further down are
better suited to patch angioplasty. In circumstances in which
the deep vein of the lower limb is injured, the ipsilateral
(a) (b)
saphenous vein, which represents a precious drainage chan-
Figure 33.17 (a) Direct anastomosis without tension. nel, ought not to be disturbed; instead, the vein from the
(b) Anastomotic tension despite sacrificing branches opposite leg may be harvested. In the absence of saphenous
 Definitive surgical treatment 391

vein, the cephalic vein may be a reasonable substitute, graft drawn over it, acts conveniently as a stent and lends
although its wall is not particularly muscular. itself to disciplined suturing of the anastomoses thus avert-
In vessels such as the axillary (Fig. 33.19) and popliteal ing ‘purse stringing’. When the arterial diameter is less
(Fig. 33.20), the anastomosis is quite simply achieved by than 3–4 mm, as in the case of radial, ulnar and tibial arter-
using a continuous polypropylene everting suture com- ies, and notably the main limb arteries of children, meticu-
menced at diametrically opposite points, which then meet lous attention to technique, aided by loupes and fine
in between, access to each side being attained by rotating instruments, is essential to patency. Vasospasm of small
the vessel along its axis. An indwelling shunt, with the vein arteries is abolished by judicious inflation using a fine,
rounded catheter. The ends of the vein graft and host ves-
sel are then cut obliquely and spatulated to prevent stenosis
(Fig. 33.21). Good results are obtained by accurate coapta-
tion of intimal surfaces, eversion of the edges to prevent
inward protrusion of adventitial strands, and carefully
placed tension-free 6-0–8-0 polypropylene sutures, inter-
rupted to accommodate increasing vessel diameter with-
out stenosis as the child grows.
In the older patient damage to a diseased atherosclerotic
vessel may necessitate a lengthy vein bypass extending
well beyond the injured arterial segment. This is especially
important in situations of marked soft tissue loss and espe-
cially if severe contamination renders the wound hazardous
to vein graft. The trimmed ends of injured artery are ligated
back to healthy tissue and an extra-anatomical vein bypass
graft, perhaps unreversed but with its valves disrupted, may
take the form of a common femoral to anterior tibial bypass
tunnelled through clean tissue (Fig. 33.22).
If the calibre of available donor vein is much smaller
than that of the host vessel, the discrepancy is likely to lead
to graft failure. Such an outcome can be avoided by fash-
ioning a compound vein graft of larger diameter, which
Figure 33.18 Excision of injured segment back to pristine artery. will maintain laminar flow. This can be achieved in various
Reversed interposition vein graft ways, and shunts already in place allow ample time for this

Figure 33.19 Injury to the right

shoulder caused by a falling tree,
producing a crush injury of the axillary
artery and brachial plexus and a
humeral fracture. The preoperative
angiogram (a) shows a defect in the
midaxillary artery. The postoperative
angiogram (b) shows interposed vein
graft (arrowed) deliberately left lax
enough to permit safe adduction
(reproduced with permission from
Barros D’Sa AAB in Eastcott HHG (ed).
(a) (b)
Arterial Surgery, 3rd edn, Edinburgh:
Churchill Livingstone, 1992: 355–411)
392 Vascular injuries of the limbs

Figure 33.20 (a) A preoperative

angiogram of multiple bullet
injuries to the popliteal artery. (b) The
postoperative angiogram shows two
interposed vein grafts (between (a) (b)
arrows)

Figure 33.21 Spatulation of

vessel ends to prevent stenosis

purpose. One technique involves taking two, and if neces- a panelled compound graft (Figs 33.23 and 33.24).
sary three, equal segments of vein of appropriate length Alternatively, a length of vein, slit open longitudinally and
opened longitudinally to form panels; after excising any bereft of valves, is wrapped spirally over a bridging shunt
valves, the panels are sewn together side by side to produce of appropriate bore; the adjoining margins of vein and
 Definitive surgical treatment 393

necrosis, a prosthetic graft may have some advantage over

vein. While sepsis brings about necrosis of the body of a
vein graft, such breakdown following the implantation of a
prosthetic graft repair tends to affect the anastomoses with
the host vessel.102

FALSE ANEURYSM
False aneurysm may become apparent soon after diagnos-
tic transfemoral cardiac catheterisation because the com-
mon femoral artery lies in a superficial position. In cases of
penetrating trauma, however, a false aneurysm of a deeply
placed artery may be missed until the aneurysm manifests
itself by progressive expansion and pressure effects (see
Figs 33.3–33.5, 33.7). For obvious reasons patients in car-
diology units undergoing diagnostic cannulation may not
be fit for general anaesthesia, and sometimes regional or
local anaesthesia may have to be employed. For some years,
non-operative, ultrasound guided compression therapy
has been successfully used in selected cases.103
In those requiring surgical treatment, proximal and dis-
tal control is established, the aneurysm opened, thrombus
evacuated, the false aneurysm excised and the vessel lumen
inspected for intimal damage. Occasionally, proximal con-
trol of the external iliac artery is necessary. Back-flow from
the profunda femoris may be controlled directly or indir-
ectly by means of a balloon catheter attached to a three-
way tap. A loose atheromatous flap in the posterior wall
may have to be tacked down with a 5-0 or 6-0 polypropyl-
ene suture. The laceration is repaired transversely with
Figure 33.22 Extra-anatomical vein bypass for extensive
interrupted 5-0 double ended sutures, picking upper and
contaminated wounds lower intimal edges from within outwards. In chronic
fusiform aneurysms the opening may be rather large,
necessitating patch angioplasty or even segmental excision
host vessel are then approximated by a continuous suture and graft replacement.
to achieve a spiralled compound graft, a technique suited
to larger vessels such as the common femoral artery ARTERIOVENOUS FISTULA
(Fig. 33.25) and indeed the inferior vena cava. Arteriovenous fistulae (see Figs 33.6 and 33.7) are frequently
missed on initial examination and even on exploration,21
PROSTHETIC GRAFT most notoriously in shotgun wounds. Consequently, they
Prosthetic grafts have to be used when saphenous vein is present much later, usually accompanied by one or two false
unavailable, either because it has been removed or it is of aneurysms. The use of a sterile stethoscope intraoperatively
poor quality. Polytetrafluoroethylene (PTFE) is probably may be of value in locating fistulae if preoperative angiograms
safer and more reliable than Dacron and although the for- are not available. After control of artery and vein proximally
mer has been used with some success in closed injuries of and distally, the fistula is opened and the false aneurysms
the large vessels, its use in highly contaminated wounds excised. Reconstruction may take the form of simple lateral
should be discouraged. It ought not to be used in prefer- suture, vein patch angioplasty and occasionally excision and
ence to vein in the upper limb, in bridging vessels across graft replacement of the arterial segment involved. The vein is
joints or in replacing damaged vein. easily repaired by lateral suture or by means of a vein patch. In
The encouraging reports of the success of prosthetic order to discourage recurrence a flap of fascia may be inter-
grafts, and in particular PTFE, in stabbings and low vel- posed between adjacent suture lines.
ocity gunshot wounds21 should not lead to complacency or
indeed obscure the inherent risks of infection, secondary On-table assessment of repair
haemorrhage and a higher amputation rate when used
indiscriminately in dirty wounds. On the occasional Peroperative evaluation of the quality of repair, in particu-
instance when substantial loss of tissue leaves a repair lar of the distal anastomosis, can be achieved simply by
inadequately covered or exposed and therefore liable to assessing the quality of a simulated pulse in the distal artery
394 Vascular injuries of the limbs

Figure 33.23 Steps in the construction

of a panelled compound vein graft

technical adequacy of surgery or to identify a defect for

immediate correction. This practice provides quality con-
trol of repairs of small calibre vessels in the forearm and the
lower leg, and is of particular importance in children. In the
severely injured hypotensive patient with poor or unrecord-
able distal pulses and low core body temperature, on-table
angiography offers assurance that there is no evidence of dis-
tal thrombus. Time consuming angioscopy has been dis-
placed by intraoperative ultrasound in evaluating of the state
of the repaired vessel, but our experience so far is limited.

Vein repair

Although a patient may be fortunate not to have a limb

Figure 33.24 Popliteal artery, above, (previously shunted) compromised by ligation of a major vein, the notion that liga-
repaired using a reversed interposition vein graft (arrowed). tion is not necessarily harmful or that venous repair is of
Popliteal vein, below, (with shunt in situ) being repaired by a limited importance should be dispelled.6 Ligation is more
panelled compound vein graft
than likely to be complicated by deep vein thrombosis,
raised compartment pressure and acute oedema as well as by
produced by forcing heparinised saline from a syringe into a much higher incidence of fasciotomy than is found follow-
it using a cannula. Equally, intraoperative use of a Doppler ing vein repair. Moreover, the likelihood of venous gangrene
ultrasound probe placed directly on the vessel wall provides and amputation in the early stages is increased. In the longer
information on functional outcome. Most effectively, term, post-thrombotic changes may bring about chronic
an on-table angiogram offers quick visual proof of the symptoms of obstructive venous insufficiency.
 Definitive surgical treatment 395

Figure 33.25 Steps in the

construction of a spiralled
compound vein graft

A reconstituted major venous channel will largely prevent Compartmental hypertension resulting from oedema of
these problems and may actually enhance the patency of an muscle is largely brought about by reperfusion injury but it
adjacent arterial repair.12 Even if the vein repair fails later, an is obviously worse in severe trauma or when venous return
interim venous collateral network may have developed to the is impaired due to vein ligation or thrombosis. The tem-
extent that final occlusion of the vein repair may not be asso- porary insertion of shunts (see Figs 33.10 and 33.11) will
ciated with chronic oedema.104 Although grafts fail, there is abbreviate the period of ischaemia and consequently
evidence to show that both the patency of the repair and reduce the severity of reperfusion injury.
valve competence within the graft can be maintained in the Muscle necrosis will inevitably occur unless decompres-
long term.105 In the presence of life-threatening injuries else- sion is timely and adequate. The anterior compartment of
where, vein ligation may be necessary. Lateral suture of a the lower leg lies within rigid osseous and fascial boundaries
large diameter main venous channel is tolerated very much and is therefore the most vulnerable of the four compart-
better than that of the adjacent artery, which has a smaller ments. The muscle compartments of the forearm and palm
diameter.104 Vein graft replacement of a damaged, vein seg- are also endangered by raised pressure. Palpable distal pulses
ment is often necessary, but the larger calibre of the host ves- and reasonable capillary refill are deceptive and may delay
sel may demand the construction of a compound vein graft fasciotomy beyond the point of irreversible muscle damage.
of either the panel or spiral variety (see Figs 33.23–33.25). A number of clinical situations call for fasciotomy as a formal
This adds a little time to the operation but with a shunt in procedure: when surgical intervention is delayed beyond a
place ample time is available for precise work. period of 4–6 hours after vessel injury; concomitant injury
In situations in which both vein and artery have been of main artery and vein; significant injury to distal soft tissues
shunted, the order in which each vessel is repaired is with associated haematoma; obvious oedema and paralysis
immaterial. Alternatively, if neither vessel has been shunted, of muscle with patchy muscle necrosis; plantar flexion of the
preliminary vein repair will prevent a serious rise in venous foot after completion of vascular repair; compartment pres-
pressure when arterial flow is eventually re-established. Time sures in excess of 40 mmHg.17,18,88
permitting, an arteriovenous fistula may also be created The techniques available for fasciotomy are varied.
just distal to the grafted segment to enhance long-term Percutaneous fasciotomy is generally ineffectual and totally
patency. fails to relieve pressure in the deep compartment of the
lower leg. If fasciotomy is deemed necessary, the transcuta-
neous method is most rewarding. The lateral approach, with
Fasciotomy or without, a mid-third fibulectomy, may enable decom-
pression of the four compartments of the lower leg but this
In addition to all the other definitive procedures aimed is a rather destructive procedure risking damage to vessels
at restoring flow and drainage in the distal limb, fas- and nerves. The standard two-incision approach, the lateral
ciotomy represents an invaluable adjunctive procedure. decompressing the anterior and peroneal compartments
396 Vascular injuries of the limbs

and the medial decompressing the superficial and deep pos- replacement is essential for good flow through recon-
terior compartments, is simple and successful.106 Severe structed vessels and to ensure satisfactory perfusion of the
compression of muscles in the forearm, in both the extensor distal bed. In cases of persisting spasm, and only after
and flexor compartments, must be relieved, along with that angiographic confirmation that an arterial injury is not
in the thenar eminence of the hand. This is achieved by a present, selective intra-arterial infusion of tolazoline has
longitudinal, centrally placed incision over the extensor been advocated.107 Low dose heparin is of value in aiding
compartment and by a further curvilinear incision on the graft patency and also in preventing deep vein thrombosis,
flexor aspect, which commences in the antecubital fossa, although caution is required in the multiply injured
identifying and preserving the cutaneous nerves of the fore- patient.
arm, crosses the wrist and extends distally onto the palm In order to minimise reperfusion injury, a slow manni-
along the thenar crease to release the carpal tunnel and tol infusion over a 12–24-hour period is helpful in limiting
thenar muscles, respectively. damage. The benefits of mannitol, which were once attrib-
uted almost entirely to its osmotic properties, are now
Wound closure increasingly recognised as being achieved largely through
the accelerated inactivation of oxygen free radicals, which
The success of any vascular repair is to an extent dependent mediate in reperfusion injury and increased capillary per-
on the quality of both the vessel and the cover afforded by meability.58–63 The value of mannitol in combating the
adjacent soft tissue and muscle, which by inference effects of reperfusion injury is supported by both experi-
demands the elimination of any dead space and the pre- menta1108 and clinical109 evidence. The timing in institut-
vention of haematoma. The magnitude of soft tissue loss ing such therapy requires considered judgement and care,
(see Fig. 33.13) may be of a degree which leaves a vein graft in particular if the patient is haemodynamically unstable
exposed and vulnerable to desiccation and degeneration. on arrival. Research studies at our centre using ischaemic-
Prosthetic grafts may be immune to these influences, but if preconditioning and the use of recombinant bacterici-
the field is contaminated, suture line dehiscence and cata- dal/permeability-increasing protein may herald innovative
strophic secondary haemorrhage are very likely to occur. techniques and therapies in attenuating the systemic
In instances of soft tissue loss, the superficial muscles of inflammatory response to IRI and vitiating remote organ
the arm, or the sartorius and gracilis muscles of the leg, can injury.110–113
be freed while retaining their blood supply, and swung over Daily dressing of wounds, swab cultures and alertness to
in such a way as to ensheath a graft. Various temporary bio- early signs of gas gangrene are essential. If gas gangrene is
logical dressings, including porcine heterograft and amni- suspected, immediate surgical excision of affected tissue
otic membrane, have been used in circumstances such as under cover of penicillin and antitoxin therapy may help to
these. Ideally, however, the skills of a plastic surgeon must counteract its advance; the use of hyperbaric oxygen is
be brought to the fore in rotating muscle flaps or in the con- controversial but, if available, should be offered in serious
struction of free vascularised musculocutaneous flaps. An cases. In spite of optimal wound care and prophylactic
alternative approach, of course, is the construction of an antibiotic therapy, healing may be compromised by
extra-anatomical vein bypass (see Fig. 33.22) through clean impaired tissue perfusion of damaged tissues.
viable tissue well away from the main wound. The latter
then requires debridement followed by split-skin grafting. Failure of vascular repair
Primary suture of the wound is generally successful in
closed injuries or in clean stab wounds. In contaminated Continued maintenance of patency of vessels after repair
wounds it is wise to pursue a policy of delayed primary represents a key concern and peripheral flow must there-
suture 5–7 days later.17–19 In dirty wounds, and particularly fore be checked by observing pulses, capillary refill time
if tissue viability remains questionable, frequent inspections, and by using Doppler ultrasound. Close vigilance is neces-
on a daily basis if necessary, must be undertaken under sary and if suspicions of impaired flow and graft failure are
anaesthesia. Obviously, early attention to wound care and the aroused urgent angiography is indicated; if the vessel has
elimination of devitalised tissue will minimise the morbid- thrombosed re-exploration and fresh reconstruction are
ity associated with a necrotic and septic focus and expedite mandatory. Thrombotic occlusion is usually the product
closure. of a number of weaknesses of technique, namely, narrow-
ing caused by lateral suture, tension and constriction at the
suture line of a direct anastomosis or ‘purse stringing’ by a
POSTOPERATIVE MANAGEMENT continuous tight suture line. Other causes include inad-
equate excision of a segment with intimal damage, poor
General measures coaptation of intima at the suture line or intrusion of the
adventitia into the flow surface. When the diameter of the
In general the injured limb is nursed in the horizontal pos- vein graft is small it may fail; when a vein graft is used to
ition, but if it is swollen a little elevation is acceptable. Fluid restore flow prior to bone fixation, it may well fail if it is
 References 397

either too short, causing anastomotic tension, or too long, Mubarak SJ, Owen CA. Double incision fasciotomy of the leg for
resulting in kinking. decompression in compartment syndromes. J Bone Joint Surg
In any situation of graft failure, early reoperation will give (US) 1997; 59A: 184–7.
the limb its best second chance of uncomplicated survival, Rich NM, Baugh JH, Hughes CW. Acute arterial injuries in Vietnam:
and when revisiting the wound attention to technique is of 1000 cases. J Trauma 1970; 10: 359–69.
crucial importance. Previously sutured ends must be
trimmed back, ensuring that there is no residual evidence of
the initial injury. Clot is removed both proximally and dis-
tally, heparinised saline is infused distally, and a shunt may REFERENCES
be inserted at this stage to restore flow. Finally, the new
repair must be precise in every respect. 1 Ogilvie WH. War surgery in Africa. Br J Surg 1944; 31: 313.
2 De Bakey ME, Simeone FA. Battle injuries of the arteries in
World War II. Ann Surg 1946; 123: 534–79.
3 Hughes CW. Arterial repair during the Korean War. Ann Surg
Conclusions 1958; 147: 555–61.
4 Rich NM. Wounding power of various ammunitions. Resident
Weapons represent a global health hazard of epidemic Phys 1968; 14: 72.
proportions, particularly in the hands of irresponsible 5 Rich NM, Hughes CW. Vietnam vascular registry: preliminary
users and terrorists who target civil populations indis- report. Surgery 1969; 65: 218–26.
criminately. As long as such weapons continue to be 6 Rich NM, Baugh JH, Hughes CW. Acute arterial injuries in
used surgeons will be confronted by vascular injuries. Vietnam: 1000 cases. J Trauma 1970; 10: 359–69.
Limb vascular trauma, particularly when complex, 7 De Muth WE. Bullet velocity makes the difference. J Trauma
requires some knowledge of the wounding mechanisms 1969; 9: 642–3.
and the pathophysiological consequences thereof. The 8 Amato JJ, Billy U, Gruber RP, et al. Vascular injuries. An
experimental study of high and low velocity missile wounds.
influence of the passage of time on reperfusion injury,
Arch Surg 1970: 101: 167–74.
compartment hypertension and the potential likelihood
9 Morris GC, Creech O, DeBakey ME. Acute arterial injuries in
of limb loss must be clearly understood. Numerous fac- civilian practice. Am J Surg 1957; 93: 565–70.
tors can play a significant role in minimising the com- 10 Morris GC, Beall AC, Roof WR, et al. Surgical experience with 220
plication rate and lowering the incidence of limb loss. acute vascular injuries in civilian practice. Am J Surg 1960; 99:
The use of intraluminal shunts in both injured artery 775–81.
and vein, particularly in complex injuries, buys time and 11 Drapanas T, Hewitt RL, Weichert RC, et al. Civilian vascular
introduces a disciplined operative technique which has injuries, a critical appraisal of three decades of management.
many advantages. Unhurried and diligent care of the Ann Surg 1970; 172: 351–60.
wound, adequate debridement and precise skeletal fix- 12 Perry MO, Thal ER, Shires GT. Management of arterial injuries.
ation will create the best conditions for vascular repair of Ann Surg 1971; 173: 403–8.
13 Smith RF, Elliot JP, Hageman JH, et al. Acute penetrating arterial
both artery and vein. The avoidance of less desirable
injuries of the neck and limbs. Arch Surg 1974; 109: 198–205.
techniques of repair in favour of vein grafting and, if neces-
14 Lozman H, Beaufils AT, Rossi G, et al. Vascular trauma observed
sary, the construction of larger diameter compound vein at an urban hospital center. Surg Gynecol Obstet 1978; 146:
grafts when the situation demands will promote better 237–40.
results. When necessary, immediate and effective fas- 15 Feliciano D, Mattox KL, Graham J, Bitondo C. Five year
ciotomy will enhance tissue viability. Postoperatively, experience with PTFE grafts in vascular wounds. J Trauma 1985;
alertness to signs of graft failure and, if necessary, 25: 71–82.
re-exploration and fresh reconstruction will improve the 16 Livingston RH, Wilson RI. Gunshot wounds of the limbs. BMJ
chances of limb survival. 1975; 1: 667–9.
17 Barros D’Sa AAB, Hassard TH, LivingstonRH, Irwin JWS. Missile-
induced vascular trauma. Injury 12: 13–30.
18 Barros D’Sa AAB. Management of vascular injuries of civil strife.
Key references Injury 1982; 14: 51–7.
19 Johnston GW, Barros D’Sa AAB. Injuries of civil hostilities In:
Amato JJ, Billy U, Gruber RP, et al. Vascular Injuries. An Carter DC, Polk HC (eds). International Medical Reviews Surgery
experimental study of high and low velocity missile wounds. Vol. 1, Trauma. London: Butterworths, 1981.
Arch Surg 1970; 101: 167–74. 20 Archbold JAA, Barros D’Sa AAB, Morrison E. Genitourinary tract
Barros D’Sa AAB, Harkin DW, Blair PHB, et al. The Belfast approach injuries of civil hostilities. Br J Surg 1981; 68: 625–31.
to managing complex lower limb vascular injuries. Eur J Vasc 21 Graham ANJ, Barros D’Sa AAB. Missed arteriovenous fistulae and
Endovasc Surg (in press). false aneurysms in penetrating lower limb trauma: relearning old
Menzoian JD, Doyle JE, Cantelmo NL, et al. A comprehensive approach lessons. Injury 1991; 22: 179–82.
to extremity vascular trauma. Arch Surg 1985; 120: 801–5. 22 Shaker IJ, White JJ, Signer RD, et al. Special problems of vascular
injuries in children. J Trauma 1976; 16: 863–7.
398 Vascular injuries of the limbs

23 Boontje AH. Iatrogenic arterial injuries J Cardiovasc Surg 1978; 47 Benson EP. Radiation injury to large arteries. Radiology 1973;
19: 335–40. 106: 195–7.
24 Natali J, Benhamou AC. Iatrogenic vascular injuries. A review of 48 McCallion W. A, Barros D’Sa AAB. Management of critical upper
125 cases excluding angiographic injuries. J Cardiovasc Surg limb ischaemia long after irradiation injury of the subclavian and
1979; 20: 169–76. axillary arteries. Br J Surg 1991; 78: 1136–8.
25 Mills JL, Wideman IE, Robison JG, et al. Minimizing mortality 49 Barros D’Sa AAB. How do we manage acute limb ischaemia due
from iatrogenic arterial injuries: the need for early recognition to trauma? In: Greenhalgh RM, Jamieson CW, Nicolaides AN
and prompt repair. J Vasc Surg 1986; 4: 22–7. (eds). Limb salvage and amputation for vascular disease. London:
26 Bergqvist D, Helfer M, Jensen N, et al. Trends in civilian vascular WB Saunders, 1988: 135–50.
trauma during 30 years. Acta Chir Scand 1987; 153: 417–22. 50 Meerson FZ, Kagan VE, Kozhov YP, et al. The role of lipid
27 Barros D’Sa AAB. The value of shunting in complex vascular peroxidation in the pathogenesis of ischaemic damage and the
trauma of the lower limb In: Earnshaw JJ, Murie JA (eds). The antioxidant protection of the heart. Basic Res Cardiol 1982;
Evidence for Vascular Surgery. Shrewsbury: tfm Publishing Ltd, 77: 465–85.
1999: 189–96. 51 Halliwell B, Gutteridge SMC. Oxygen toxicity, oxygen radicals,
28 Barros D’Sa AAB, Harkin DW, Blair PHB, et al. The Belfast transition metals and disease Biochem J 1984; 219: 1.
approach to managing complex lower limb vascular injuries. 52 Korthuis RJ, Granger DN, Townsley MI, Taylor AE. The role of
Eur J Vasc Endovasc Surg (in press). oxygen-derived free radicals in ischaemia-induced increases in
29 Barros D’Sa AAB. Shunting in complex lower limb trauma. canine skeletal muscle vascular permeability Circ Res 1985; 57:
In: Greenhalgh RM, Hollier L (eds). Emergency Vascular Surgery. 599–609.
London: WB Saunders, 1992. 53 McCord 1M. Oxygen-derived free radicals in post-ischaemic
30 Barros D’Sa AAB. Complex vascular and orthopaedic limb injuries tissue injury. N Engl J Med 1985; 312: 159–63.
[editorial]. J Bone Joint Surg (UK) 1992; 74: 176–8. 54 Granger DN, Hollwarth ME, Parks DA. Ischaemia-reperfusion
31 Connolly J. Management of fractures associated with arterial injury: role of oxygen derived free radicals. Acta Physiol Scand
injuries. Am J Surg 1971; 120: 331–5. 1986; 548(suppl): 47–63.
32 Lefrak BA. Knee dislocation. Arch Surg 1976; 111: 1021–4. 55 Grisham MB, Hernandez LA, Granger DN. Xanthine oxidase and
33 Alberty RE, Goodfried G, Boyden AM. Popliteal artery injury neutrophil infiltration in intestinal ischaemia. Am J Physiol 1986;
with fracture dislocation of the knee. Am J Surg 1981; 142: 251: G567–74.
36–40. 56 Kuzon WM Jr, Walker PM, Mickle DA, et al. An isolated skeletal
34 Doty DB, Treiman RL, Rothschild PD, et al. Prevention of gangrene muscle model suitable for acute ischaemic studies. J Surg Res
due to fractures. Surg Gynecol Obstet 1967; 125: 284–5. 1986; 41: 24–32.
35 Smith RF, Szilagyi DE, Elliott JP. Fracture of the long bones with 57 Armstead WM, Mirro R, Bursija DW, Leffler CW. Post-ischaemic
arterial injury due to blunt trauma. Arch Surg 1969; 99: 315–24. generation of superoxide anion by newborn pig brain. Am J
36 Wagner WH, Calkins ER, Weaver FA, et al. Blunt popliteal artery Physiol 1988; 255: H401–3.
trauma: one hundred consecutive injuries. J Vasc Surg 1988; 58 Anner H, Kaufman RP, Vateri CR, et al. Reperfusion of ischaemic
7: 736–43. lower limbs increases pulmonary microvascular permeability.
37 Gustilo RB, Mendoza RM, Williams DN. Problems in the J Trauma 1988; 28: 607–10.
management of Type III (severe) open fractures: a new 59 Ernester L. Biochemistry of reoxygenation injury. Crit Care Med
classification of Type III open fractures. J Trauma 1984; 24: 742–6. 1988; 16: 947–53.
38 Lance RH, Bach AW, Hansen ST, et al. Open tibial fractures with 60 Smith JK, Carden DL, Korthuis RJ. Role of xanthine oxidase in
associated vascular injuries: Prognosis for limb salvage. J Trauma post-ischaemic microvascular injury in skeletal muscle.
1985; 25: 203–7. Am J Physiol 1989; 257: H1782–9.
39 Seiler JG, Richardson JD. Amputation after extremity injury Am 61 Rubin BB, Smith A, Liauw S, et al. Complement activation and
J Surg 1986; 152: 260–4. white cell sequestration in post-ischaemic skeletal muscle.
40 Kelly GL, Eiseman B. Civilian vascular injuries. J Trauma 1975; Am J Physiol 1990; 259: H525-31.
15: 507–14. 62 Beckman JS, Beckman TW, Chen J, et al. Apparent hydroxyl
41 Hartsuck 1M, Moreland HJ, Williams GR. Surgical management radical production by peroxynitrite: implications for endothelial
of vascular trauma distal to the popliteal artery. Arch Surg 1972; injury from nitric oxide and superoxide. Proc Nat Acad Sci U S A
105: 937–40. 1990; 87: 1620–4.
42 Hollemann JH, Killebrew LH. Tibial artery injuries. Am J Surg 63 Kupinski AM, Shah DM, Bell DR. Permeability changes following
1982; 144: 362–4. ischaemia-reperfusion injury in rabbit hind limb. J Cardiovasc
43 Menzoian JD, Doyle JE, Cantelmo NL, et al. A comprehensive Surg (Torino) 1992; 33: 690–4.
approach to extremity vascular trauma. Arch Surg 1985; 120: 64 Allen DM, Chen LE, Seaber AV, Urbaniak JR. Pathophysiology and
801–5. related studies of the no reflow phenomenon in skeletal muscle.
44 Pasch AR, Bishara RA, Lim LT, et al. Optimal limb salvage in Clin Orthop 1995; 122–33.
penetrating civilian vascular trauma. J Vasc Surg 1986; 3: 65 Yassin MM, Barros D’Sa AAB, Parks TG, et al. Lower limb
189–95. ischaemia-reperfusion alters gastrointestinal structure and
45 Best B, Barros D’Sa AAB. Popliteal vessel injury caused by plastic function. Br J Surg 1997; 84: 1425–9.
bullets. Injury 1987; 18: 428–9. 66 Yassin MM, Barros D’Sa AAB, Parks TG, et al. Lower limb
46 Pace PD, Tilney NL, Lesch M, et al. Peripheral arterial ischaemia-reperfusion injury causes endotoxaemia and
complications of intra-aortic balloon counter-pulsation. Surgery endogenous antiendotoxin antibody consumption but not
1977; 82: 685–8. bacterial translocation. Br J Surg 1998; 85: 785–9.
 References 399

67 Yassin MM, Barros D’Sa AAB, Parks TG, et al. Mortality following 90 Barros D’Sa AAB, Moorehead RJ. Combined arterial and venous
lower limb ischaemia-reperfusion: a systemic inflammatory intraluminal shunting in major trauma of the lower limb.
response? World J Surg 1996; 20: 961–7. Eur J Vasc Surg 1989; 3: 577–81.
68 Harkin DW, Barros D’Sa AAB, McCallion K, et al. Circulating 91 Elliott J, Templeton J, Barros D’Sa AAB. Combined bony and
neutrophil priming and systemic inflammation in limb- vascular limb trauma: a new approach to treatment. J Bone
reperfusion injury. Int Angiol 2001; 20: 78–9. Joint Surg 1984; 66-B: 281.
69 Glauser MP, Zanetti G, Baumgartner JD, Cohen J. Septic shock: 92 Barros D’Sa AAB. Upper and lower limb vascular trauma. In:
pathogenesis. Lancet 1991; 338: 732–6. Greenhalgh RM (ed). Vascular Surgical Techniques. London:
70 Bone RC. The sepsis syndrome: definition and general approach Bailliere Tindall, 1989: 47–65.
to management. Clin Chest Med 1996; 17: 175–81. 93 Barros D’Sa AAB. Leading article. Twenty-five years of vascular
71 Leeman M. Pulmonary hypertension in acute respiratory distress trauma in Northern Ireland. BMJ 1995; 310: 1–2.
syndrome Monaldi Arch Chest Dis 1999; 54: 146–9. 94 Walker AJ, Mellor SG, Cooper GJ. Experimental experience with
72 Hoffman MJ, Greenfield LJ, Sugerman HJ, et al. Unsuspected a temporary intraluminal heparin-bonded polyurethane arterial
right ventricular dysfunction in shock and sepsis. Ann Surg 1983; shunt Br J Surg 1994; 81: 195–8.
198: 307–19. 95 Brinker MR, Caines MA, Kerstein MD, Elliott MN. Tibial shaft
73 Carrico CJ, Meakins JL, Marshall JC, et al. Multiple organ failure fractures with an associated infrapopliteal arterial injury: a
syndrome. Arch Surg 1986; 121: 196–208. survey of vascular surgeons’ opinions on the need for vascular
74 Harkin DW, Barros D’Sa AAB, Yassin MM, et al. Reperfusion injury repair. J Orthop Trauma 2000; 14: 194–8.
is greater with delayed restoration of venous outflow in 96 Keeley S, Snyder WH, Weigelt JA. Arterial injuries below the
concurrent arterial and venous limb injury. Br J Surg 2000; 87: knee: fifty-one patients with 82 injuries. J Trauma 1983; 23:
734–41. 285–90.
75 Nassim A, Thompson MM, Naylor AR, et al. The impact of MRSA 97 Yeager RA, Hobson RW, Lynch TG, et al. Popliteal and
on vascular surgery Eur J Vasc Endovasc Surg 2001; 22: 211–14. infrapopliteal arterial injuries differential management and
76 Snyder WH, ThaI ER, Bridges RA, et al. The validity of normal amputation rates. Am Surg 1984; 50: 155–8.
arteriography in penetrating trauma Arch Surg 1978; 113: 424–6. 98 Segal D, Brenner M, Gorczyca J. Tibial fractures with
77 Sirinek KR, Gaskill HV, Dittman WI, et al. Exclusion angiography for infrapopliteal arterial injuries. J Orthop Trauma 1987; 1: 160–9.
patients with possible vascular injuries of the extremities – a 99 Flint LM, Richardson JD. Arterial injuries with lower extremity
better use of trauma centre resources. Surgery 1983; 94: 599–603. fracture. Surgery 1983; 93: 5–8.
78 Frykberg ER, Crump JM, Dennis JW, et al. Non-operative 100 Shah DM, Corson JD, Karmody AM, et al. Optimal management
observation of clinically occult arterial injuries: a prospective of tibial artery trauma. J Trauma 1988; 28: 228–34.
evaluation. Surgery 1991; 109: 85–96. 101 Barros D’Sa AAB, Berger K, Di Benedetto G, et al. A healable
79 McCormick TM, Burch BH. Routine angiographic evaluation of filamentous Dacron surgical fabric. Ann Surg 1980; 192:
neck and extremity trauma. J Trauma 1979; 19: 384–7. 645–57.
80 Reid JDS, Weigelt JA, Thal ER, et al. Assessment of proximity of a 102 Stone KS, Walshaw R, Sugiyama GT, et al.
wound to major vascular structures as an indication for Polytetrafluoroethylene versus autogenous vein grafts for
arteriography. Arch Surg 1988; 123: 942–6. vascular reconstruction in contaminated wounds. Am J Surg
81 Gomes GA, Kreis DJ, Ratner L, et al. Suspected vascular trauma of 1984; 147: 692–5.
the extremities: the role of arteriography in proximity injuries. 103 Feld R, Patton GM, Carabasi A, et al. Treatment of iatrogenic
J Trauma 1986; 26: 1005–8. femoral artery injuries with ultrasound-guided compression.
82 Frykberg ER, Dennis JW, Bishop K, et al. The reliability of physical J Vasc Surg 1992; 16: 832–40.
examination in the evaluation of penetrating extremity trauma: 104 Meyer J, Walsh J, Schuler J, et al. The early fate of venous repair
results at one year. J Trauma 1991; 31: 502–11. after civilian vascular trauma: a clinical, haemodynamic and
83 Gregory RT, Gould RJ, Peclet M, et al. The mangled extremity venographic assessment. Ann Surg 1987; 206: 458–64.
syndrome (MES): a severity grading system for multisystem 105 Phifer TJ, Gerlock AJ, Rich NM, McDonald JC. Long term
injury of the extremity. J Trauma 1985; 25: 1147–50. patency of venous repairs demonstrated by venography. J
84 Johansen K, Daines M, Howey T, et al. Objective criteria Trauma 1985; 25: 342–6.
accurately predict amputation following lower extremity trauma. 106 Mubarak SJ, Owen CA. Double-incision fasciotomy of the leg
J Trauma 1990; 30: 568–73. for decompression in compartment syndromes. J Bone Joint
85 Anderson N, Patha de Sousa C, Paredes S. Social cost of Surg (US) 1977; 59A: 184–7.
landmines in four countries: Afghanistan, Bosnia, Cambodia and 107 Dickerman RM, Gewertz BL, Foley DW, Fry J. Selective
Mozambique. BMJ 1995; 311: 718–21. intra-arterial tolazoline infusion in peripheral arterial trauma.
86 Coupland RM. The effect of weapons: surgical challenge and Surgery 1977; 81: 605–9.
medical dilemma. J R Coll Surg Edinb 1996; 41: 65–71. 108 Buchbinder D, Karmody AM, Leather RP, Shah DM. Hypertonic
87 Vara-Thorbeck R, Ruiz-Morales M. Severe head injury combined mannitol. Its use in the prevention of revascularization syndrome
with orthopaedic and vascular trauma of the limbs. Langenbecks after acute arterial ischaemia Arch Surg 1981; 116: 414–21.
Arch Surg 1998; 383: 252–8. 109 Shah DM, Naraynsingh Y, Leather RP, et al. Advances in the
88 Barros D’Sa AAB. A decade of missile-induced vascular trauma. management of acute popliteal vascular blunt injuries J Trauma
Ann R Coll Surg Engl 1982; 64: 37–44. 1985; 25: 793–7.
89 Barros D’Sa AAB. Leading article. The rationale for arterial and 110 Harkin DW, Barros D’Sa AAB, McCallion K, et al. Ischaemic
venous shunting in the management of limb vascular injuries Eur preconditioning before lower limb ischaemia-reperfusion protects
J Vasc Surg 1989; 3: 471–4. against acute lung injury. J Vasc Surg 2002; 35: 1264–73.
400 Vascular injuries of the limbs

111 Harkin DW, Barros D’Sa AAB, et al. Recombinant permeability-increasing protein hind limb ischaemia-
bactericidal/permeability-increasing protein attenuates the reperfusion injury. Ann Vasc Surg 2001; 15: 326–31.
systemic inflammatory response syndrome in lower limb 113 Harkin DW, Barros D’Sa AAB, McCallion K, et al. Bactericidal/
ischaemia-reperfusion injury. J Vasc Surg 2001; 33: 840–6. permeability-increasing protein attenuates systemic
112 Harkin DW, Barros D’Sa AAB, Yassin MM, et al. Gut mucosal inflammation and acute lung injury in porcine lower limb
injury is attenuated by recombinant bactericidal/ ischaemia-reperfusion injury. Ann Surg 2001; 234: 233–44.
 34
Vascular Injuries of the Chest

ALASTAIR NJ GRAHAM, KIERAN G MCMANUS, JAMES A MCGUIGAN

The problem 401 Prehospital management 408

Aetiology 401 Resuscitation 408
Pathophysiology of resuscitation 404 Operative management 408
Diagnosis 405 Postoperative care 415
Investigations 406 Results 415
General principles and pitfalls in treatment 408 References 416

THE PROBLEM experienced in civilian centres have undoubtedly modified

treatment on the battlefield. The civilian environment gen-
erates large numbers of deceleration injuries as well as cal-
The modern city and the highway contribute more major culated close-range handgun attacks.
trauma cases daily than do all the declared wars, unde- Management will vary depending on the situation. The
clared wars and terrorist atrocities in the world. The bur- patient exsanguinating a mile from a trauma centre may
den of domestic trauma also has enormous implications survive if a policy of ‘scoop and run’, emergency room tho-
for healthcare expenditure. In the USA approximately racotomy and ‘clamp and sew’ repair of the aorta prevails,
60 million injuries are recorded annually, 30 million whereas in the stable, multitrauma patient with contained
requiring professional care and 3.6 million being admitted aortic disruption a more considered approach prior to
to hospital and absorbing much of the health budget. Most transfer to a specialist centre would be reasonable.
importantly, 300 000 permanent disabilities are produced
each year and trauma is responsible for more life-years lost
than any other cause. Thoracic injuries occur at a rate of
12 per million of the population each day.1 Major chest AETIOLOGY
trauma is present in 56 per cent of trauma fatalities2 and is
directly responsible for a quarter of all deaths.3 The traditional division of thoracic injuries into penetrat-
Traumatic aortic rupture is an often missed and rapidly ing and blunt is useful in describing aetiology, diagnosis
fatal condition if not treated judiciously. In examining the and management as each category displays distinct pat-
full range of thoracic vascular injuries this chapter also terns of progress. The boundaries, however, are frequently
addresses cardiac, cervico-mediastinal, pulmonary and chest blurred as blunt injuries may also have a penetrating com-
wall vascular injuries. ponent caused by fractured ribs or extraneous objects pene-
Prior to the Vietnam War fewer than 10 cases of injuries trating the thoracic viscera (Fig. 34.1).
to thoracic great vessels had found their way into the world
literature4 whereas in the 1970s they accounted for more
than 10 per cent of reported series of vascular injuries.5 The Blunt injuries
lessons learned by these military surgeons, used effectively
during the recent urban terrorist campaign in Northern In civilian practice most great vessel injuries result from
Ireland, has been to proceed quickly to surgery without road traffic accidents. For instance, abrupt deceleration
undue attempts at resuscitation. War scenarios produce from a speed of 48 km/hour (30 mph), unmodified by seat-
proportionally fewer thoracic vascular injuries than observed belt or crumple zone side impact bars, or the anterior chest
in road traffic accidents, urban violence and terrorist wall itself, will produce a force of 172 times the force of
atrocities. The many published series of thoracic trauma gravity.6 The elastic limit of the intima and media of the
402 Vascular injuries of the chest

CARDIAC
Blunt trauma to the heart rarely has persistent conse-
quences. The heart is well protected by the sternum, one of
the stronger bones in the body. While a sudden praecordial
impact can cause dysrhythmias, even ventricular fibrilla-
tion, the commonest injury is ‘concussion’ of the heart.
Although the patient may recall a severe crushing central
chest pain it is frequently masked by the shock of the
situation and associated anterior chest wall pain. Cardiac
concussion is most often short lived, underdiagnosed
and usually inconsequential. Severe blunt trauma to the
heart, however, can result in intimal injury to the anterior
descending coronary artery causing infarction and may
Figure 34.1 A middle aged man suffered impalement by a even lead to the development of a ventricular aneurysm.
wooden fence post after crashing his car. Both penetrating and Coronary artery rupture is rarely seen but when it occurs it
blunt thoracic injuries were clearly present. A left posterolateral is rapidly fatal.8 Compression of the filled heart against the
thoracotomy allowed removal of the ‘foreign body’ followed closed mitral or tricuspid valve has been associated with
by thorough debridement and reconstruction of the chest wall. valvular injury. This may lead to cardiac instability and
He made a good recovery failure within a few days or may only come to light some
months or even years after the event.
aorta is exceeded by shearing forces as little as 80 G when
applied to the mediastinal mass of the heart and adjacent
AORTIC
structures. The forces sustained by the occupant of a vehicle
involved in a collision, however, are not simple shearing The isthmus of the thoracic aorta is immediately distal to
forces but the product of a number of complex forces the left subclavian artery. It is, as suggested by the nomen-
acting in rapid succession (see box). Similar stresses are clature, the narrowest part of the aorta situated at the junc-
placed on the mediastinal vasculature in falls from a great tion between the relatively fixed descending aorta and the
height, in aircraft crashes with the victims sitting in the mobile mediastinal structures attached to the aortic arch
‘brace’ position and in crush injury. The inner layers of the and the ligamentum arteriosum. The isthmus, therefore, is
aorta are torn and although the adventitia may contain the section of the aorta at highest risk of rupture, though it
the rupture for an undefined period of time, rupture will is not the only such site.
follow in most cases. It is rare for true intimal dissection to
occur7 and the two terms ‘rupture’ and ‘dissection’ should
not be confused. CERVICO-MEDIASTINAL
Injuries to the supra-aortic vessels occur in 5 per cent of
Component forces acting on the aorta in those investigated after severe blunt thoracic trauma and
severe anterior/posterior deceleration injury interestingly, the innominate, left common carotid and
left subclavian arteries are equally likely to be affected.9
These injuries occur most commonly in motor vehicle –
• Horizontal deceleration with ‘jerk’ deceleration as
car more frequently than motorcycle – collisions and also
the chest wall is restrained by a belt or steering wheel
leading to a horizontal component of shear stress in crush injuries, falls and similar injuries from a great
height.10 The mechanism of injury is thought to be a
• Torsion stress as the heart and its attachments are
combination of traction with violent head and arm move-
forced into the left chest and angulation stress as the
arch rotates around the isthmus ments and scalene muscle compression. Avulsion of the
supra-aortic branches can follow the simultaneous anterior
• Rotational deceleration as the upper torso rotates
impact on the chest driving the arch downwards while
anteriorly above a belt or steering wheel
accompanying hyperextension of the neck produces long
• Anteroposterior compression results when the rigid
axis traction which can exceed the internal elastic limit
posterior thoracic structures move anteriorly against
the already stationary anterior chest wall structures of the intima. Injury of the great vessels frequently results
in thrombosis and diagnosis is often delayed, but rupture
• Mediastinal structures forcibly displaced superiorly
does occur, and most commonly involves the innom-
and caudally by abdominal contents compressed by
seatbelt or other restraint inate and left subclavian arteries.11 When these injuries are
isolated the patient is usually stable and repair is pos-
• ‘Waterhammer’ effect on blood within the aorta due
sible without the use of cardiopulmonary bypass (CPB)12
to the compression of the heart and abdominal aorta
(Fig. 34.2).
 Aetiology 403

gunshot wounds or stabbings. While all penetrating mis-

siles, knives and other sharp implements puncture and
lacerate vessels and cause blood loss, the degree of damage
varies considerably. Ironically, some patients who survive
to reach hospital after injuries caused by high velocity bul-
lets, depending on velocity at impact, may have sustained
limited vessel injury and not necessarily the classic massive
cavitation injury. Nevertheless, even though the patient
may be stable, significant arterial wall contusion may have
occurred and the risk of subsequent dissection, rupture,
false aneurysm formation or thromboembolism is such
that any penetrating injury in the vicinity of the great ves-
sels demands arteriography or a similar imaging study.
High velocity injuries to the chest, however, are rarely
‘clean through and through’ injuries, the rib or other bony
structure often being hit, splintering bone and shattering
the missile. The resulting injury is more akin to a shotgun
blast at close quarters generating multiple secondary mis-
Figure 34.2 On admission after a crash, a young motorcyclist siles. Low velocity missiles have a reputation for dissecting
was found to be haemodynamically unstable. A contrast enhanced around vessels and major viscera. It may be difficult there-
computed tomography (CT) scan showed evidence of mediastinal fore to predict their track through the tissues. In addition to
haemorrhage but a discrete vascular lesion could not be seen.
laceration and bleeding, a special feature of low velocity
Digital subtraction angiography showed that the right internal
missiles is their ability to penetrate and embolise within the
mammary artery had avulsed from the subclavian artery with
acute false aneurysm formation. Surgical repair using an pulmonary or systemic vasculature.13 Shotgun injuries vary
interposition graft gave a good result. If a thoracic vascular injury depending on the distance from which they are fired and
is suspected because of hypotension, and the CT scan only shows the type of shot used. At less than 15 cm the effect is one of
a mediastinal haematoma, angiography becomes mandatory a single large, low velocity missile coring a hole in any struc-
ture with which it comes into contact. From 15 cm to 3 m,
or when clothing or bone has dispersed the shot, the body is
peppered with multiple low velocity missiles which cause
PULMONARY VESSELS
multiple lacerations. Such injuries to the ascending aorta or
Peripheral contusions of the lung involve intrapulmonary pulmonary outflow tract will make surgical repair almost
tears with alveolar haemorrhage. This low pressure system impossible. At distances greater than 3 m shotgun pellets
is easily tamponaded once the lung is reinflated by appro- tend to lodge in the subcutaneous tissues with little ill effect.
priate chest drain placement. Major central pulmonary The effects of stab injuries are usually easier to predict,
vascular injuries tend to be rapidly fatal and are usually particularly if the weapon is known. An anterior stab would
only diagnosed at thoracotomy, a procedure generally place the internal mammary arteries at risk causing signifi-
avoided in blunt trauma. cant intrapleural blood loss but usually without cardiovas-
cular collapse (see Fig. 34.2). If cardiac injury has occurred
CHEST WALL VESSELS and the patient survives to reach hospital the clinical situ-
ation is usually one of tamponade rather than exsanguination
Rib fractures caused by blunt trauma will frequently lacer-
with the right ventricle or main pulmonary artery usually
ate intercostal vessels. Bleeding from the injured internal
being the site of injury. The lungs are often injured in
mammary artery may be limited initially by vascular spasm
wounds to the lateral and posterior aspects of the chest and
but fluid resuscitation may precipitate delayed primary
with posterior wounds the descending aorta is also at risk.
haemorrhage. The likelihood of this outcome may not be
Aortic or other vascular injury should always be suspected
apparent from the findings on the initial chest X-ray and
in stab injuries and a conservative approach is acceptable
emphasises the need for a follow-up study within 6 hours.
only if accompanied by repeated clinical reassessment and
Appropriate chest drainage will usually expose such bleed-
adequate imaging to exclude vascular injury (Fig. 34.3).
ing. However, poorly inserted tube thoracostomy may itself
The distribution of major thoracic vascular injuries in one
result in iatrogenic injury to intercostal vessels.
large series from Houston, TX, USA, was as follows: descend-
ing thoracic aorta 21 per cent, subclavian artery 21 per
Penetrating injuries cent, pulmonary artery 16 per cent, subclavian vein 13 per
cent, intrathoracic vena cava 11 per cent, innominate artery
Unlike the effects of blunt injuries, all great vessels in the 9 per cent, pulmonary veins 9 per cent.14 In general, gun shot
chest are vulnerable to penetrating trauma in the form of wounds, secondary blast injuries and stab wounds of the
404 Vascular injuries of the chest

(c)
(a)

Figure 34.3 On Christmas Eve a 22-year-old man’s wife stabbed

him with a kitchen knife. (a) An entry wound was evident 10 cm to
the left of the spinous process of the eighth thoracic vertebra. He
was haemodynamically stable and the chest X-ray failed to reveal
a left pleural effusion. (b) In view of the anatomical site of the
entry wound an intravenous contrast enhanced computed
tomography scan was performed. The descending thoracic aorta
appeared normal but it was surrounded by a large haematoma
and small bilateral pleural effusions were also noted. (c) Digital
subtraction angiography revealed a false aneurysm of the
(b) descending thoracic aorta. It was approached via a left
posterolateral thoracotomy and repaired by direct suture

aorta usually lead to death within 1 hour in 90 per cent of syndrome known as ‘blast lung’ which may not be apparent
victims. Within the rather limited anatomical space of the for many hours.
thoracic inlet resides a collection of major vessels and any
penetrating trauma in this area has a high likelihood of Iatrogenic injuries
causing vascular injury. While the subclavian artery is most
at risk,15 aortic and innominate artery injuries also occur.5 One of the commonest mechanisms of iatrogenic vascular
The classic sign is an expanding haematoma. It is note- trauma is the result of inserting cannulae, either for diag-
worthy, however, that 8 per cent of patients with great vessel nostic or therapeutic purposes, presenting as a variety of
injuries presenting at one centre showed no clinical signs cervico-mediastinal arterial and venous injuries (see
of vascular injury.16 Chapters 38 and 40).

Blast injuries PATHOPHYSIOLOGY OF RESUSCITATION

Blast injuries are even more complicated than simple
penetrating or blunt trauma, because they are caused by Evidence from laboratory and clinical studies has shown
a combination of the blunt effect of the shockwave on that fluid resuscitation prior to obtaining definitive control
the torso, high and low velocity penetrating injuries from of major vessel haemorrhage results in poorer outcomes
secondary shrapnel, deceleration injuries as a result of the than those associated with a policy of ‘permissive hypoten-
body being displaced by the explosion, and finally, the sion’.17–20 Bickell et al.,17 using a swine model in which
 Diagnosis 405

haemostasis was achieved by the formation of thrombus and absent or diminished pulses in a patient with severe
after aortotomy, compared resuscitation with and without dorsal thoracic pain strongly suggests major vessel injury
80 mL/kg of isotonic crystalloid fluid. While all infused with free rupture.
animals showed an initial increase in cardiac output and In the conscious patient who has suffered disruption of
mean arterial pressure, within 30 minutes the levels of a major vessel, even of the size of the aorta, hypotension is
these parameters returned to those of untreated animals. not a constant finding. Following arterial laceration or dis-
Importantly, in the treated animals oxygen delivery was section, especially when of partial wall thickness, haemor-
lower and all of them died, whereas the untreated animals rhage may be protected by spasm or possibly tamponaded
survived. A sheep model19 consisting of a surgically lacer- by surrounding mediastinal tissues. The clinical picture is
ated pulmonary vein with chest drainage, was tested with one of a contained haematoma as opposed to free rupture
an infusion of 30 mL/kg of isotonic crystalloid: treated ani- so that clinical signs may not be evident. Suspecting and
mals were found to have a significantly increased rate, diagnosing such an injury before it progresses to free rup-
volume and duration of haemorrhage. Similar results were ture is the major duty of the trauma surgeon. The clinical
recorded in rat models of uncontrolled haemorrhage.18 triad of a grossly widened mediastinum on chest X-ray,
Clinical studies tend to support inferences from this haemothorax and transient haemodynamic instability is
experimental work. In a retrospective study of 6855 highly specific for aortic injury and a marker for impend-
patients from the San Diego County Trauma System, the ing death from free rupture.29
56 per cent given a prehospital infusion of crystalloid fluid
(620–1554 mL) had a similar mortality rate but a signifi- Aortic
cantly more negative base deficit compared with those not
receiving fluid.21 Other studies have shown that despite fluid Many patients with aortic injury report severe interscapu-
resuscitation the oxygen saturation of blood was unchanged lar pain without its significance being appreciated because
and that the lungs were more likely to suffer from fluid of the multiplicity of other injuries. Examination of the
overload.22,23 At the Ben Taub General Hospital, Houston, a chest wall is frequently unremarkable and the lung fields
prospective randomised trial of resuscitation after penetrat- may be clear. Systolic bruits are occasionally found in the
ing torso trauma was carried out, acknowledging the inher- interscapular area. The left arm pulse and arterial pressure
ent logistic problems associated with such a study. A group may be less than the right. Lower limb pulses may be diffi-
of 289 patients in whom fluid resuscitation was delayed until cult to feel despite satisfactory brachial blood pressure and
they arrival in the operating theatre was compared with 309 anuria may follow. Paraplegia may result from reduced
patients who received immediate fluid resuscitation: the lat- flow in the anterior spinal artery.
ter had a significantly higher incidence of renal failure, acute
respiratory distress syndrome (ARDS) and mortality.24
Cervico-mediastinal
Caution must be employed, however, before replacing
old inflexible guidelines with new inflexible guidelines. The
Subtle vascular signs may point to a supra-aortic vessel
results obtained in the above models of major vascular
injury. The classic sign is an expanding haematoma which
injury may not be applicable to patients with pulmonary
may mask the actual loss of pulses; either in this situation or
parenchymal injury in whom some evidence of benefit from
if a chest X-ray reveals an upper mediastinal haematoma,
a degree of volume resuscitation with hypertonic solutions is
arch aortography is recommended (see Fig. 34.2).30
observed.25 Where surgery is not immediately available26,27
or indeed appropriate,28 the withholding of resuscitation
when the cardiac index and both oxygen delivery and con- Pulmonary
sumption are low, increases morbidity and mortality.
Injuries to the venous or pulmonary vessels, which are low
pressure systems, can be effectively tamponaded by the lungs,
DIAGNOSIS especially if properly positioned chest drains allow them to
expand fully. Haemodynamic instability, rather than breath-
While penetrating vascular injuries may be solitary and the lessness, is a common feature of tension pneumothorax,
diagnosis straightforward, in blunt trauma they rarely occur particularly when accompanied by a significant haemo-
in isolation. In all multitrauma patients vascular injury must thorax. The ‘stony dullness’ of a haemothorax only becomes
be suspected until proved otherwise, and unless specific detectable when at least 1 L of blood has been lost.
vessel injuries are considered in the light of the history, they
will not be looked for and therefore will be missed. Cardiac
A major thoracic arterial injury open to the skin or
the pleural cavity usually results in rapid exsanguination. Cardiac tamponade following ventricular injury represents
Following a major deceleration injury or gunshot wound a special case of containment in which ventricular pressure
to the chest the combination of dullness, reduced air entry interferes with venous filling of the low pressure atria,
406 Vascular injuries of the chest

resulting in falling cardiac output and eventual arrest. The Once a vascular injury is suspected the next decision
patient, if conscious, is extremely anxious, gasping for is whether to employ computed tomography (CT), trans-
breath, hypotensive with peripheral shutdown, the upper oesophageal echocardiography (TOE) or angiography and
torso and face are suffused with dilated veins and the neck that usually depends on the likely associated injuries. In
is oedematous. the multiply injured patient, with potential abdominal or
intracranial injuries a CT scan is a useful initial screen-
ing investigation. Aortography is more appropriate when
Vena caval system cervico-mediastinal injuries are suspected although TOE
may be quicker to perform if life-threatening aortic injury
Injuries to the superior vena cava or the innominate
is suspected. These imaging techniques should be regarded
veins are manifested either as a tamponaded mediastinal
as being complementary rather than in competition pro-
haematoma or as free bleeding into the chest. Injury to the
vided that they can be done quickly and effectively. Surgery
intrapericardial inferior vena cava may result in cardiac
on unstable patients should not be delayed by unnecessary
tamponade, or alternatively, as in the case of an injured
extra investigations.
azygous vein, rapid exsanguination into the chest.

Contrast enhanced helical CT scan

INVESTIGATIONS
The contrast enhanced CT scan, being non-invasive, widely
available and vital in excluding intracranial, intra-abdominal
In the heat of the moment it is important to keep to a routine,
and pelvic injuries is an extremely useful initial study. False
checking airway, breathing and circulation, instituting
negative results for thoracic aortic rupture are usually due
electrocardiographic (ECG) monitoring and reviewing
to the inappropriate interpretation of poor quality scans;31,32
arterial oxygen saturation and central venous pressure.
spiral CT scans with 2 mm slices at the aortic isthmus, with
In the unstable patient, primary survey may indicate the
correctly timed injection of contrast, raise sensitivity to 100
need to perform emergency thoracotomy without further
per cent.20,33–35 The improved accuracy and speed of modern
investigation. A single penetrating cardiac injury with tam-
CT scanning has led to a trend eliminating the use of other
ponade is usually the only major cardiovascular injury
investigations in the haemodynamically unstable patient
which can be salvaged in this situation, and therefore either
and in demonstrating signs of aortic injury and periaortic
anterior thoracotomy or median sternotomy would be a
extravasation of contrast36 (Fig. 34.4b).
suitable approach. Most contained vascular injuries will be
detected on secondary survey, which should also include
the following investigations: chest and cervical X-rays, Arch aortogram
arterial blood gases, baseline haematology and biochem-
istry assessment including electrolytes and glucose, blood Aortography may be necessary on some occasions to con-
alcohol and, if applicable, a pregnancy test. firm the nature of the injury before proceeding to surgery,
Evidence of intrathoracic bleeding or pneumothorax but with the improved quality of spiral CT scans it is
may be absent on the initial chest X-ray, which therefore, required less often. Although aortography may deliver
in the stable patient, should be repeated within 6 hours. occasional false positives for thoracic aortic rupture, usu-
While lung contusion or substernal venous bleeding may ally due to the presence of atheroma or ductal diverticula,
produce a wide mediastinal shadow, progressive widening it is usually a reliable indicator of the presence of thoracic
of that shadow is strongly suggestive of an aortic rupture aortic rupture (Fig. 34.4c). A recent series comparing aor-
(box and Fig. 34.4a). tography and spiral CT showed the incidence of false posi-
tive results to be similar at 4 per cent.36
Routine angiography has been regarded as mandatory
in evaluating all penetrating neck injuries but a predictive
Radiological signs of traumatic aortic
value approaching 100 per cent for chest X-ray seen against
rupture
a negative physical examination of the neck has cast some
doubt on this dogma.30 Nevertheless, angiography remains
• Widening mediastinum
the investigation of choice when cervico-mediastinal vas-
• Loss of aortic ‘knuckle’
cular injuries are suspected.
• Apical cap
• Paratracheal stripe
• Depression of the left main bronchus
Transoesophageal echocardiography
• Deviation of the trachea/endotracheal tube to right
• Deviation of the oesophagus/nasogastric tube to
Interest has recently increased in the use of TOE in assessing
right
the patient at risk of thoracic aortic rupture (Fig. 34.4d).
 Investigations 407

This quick, portable investigation can be performed simul- would be dangerous or present considerable logistical
taneously with other procedures in the emergency room difficulties. The unstable patient with a thoracic aortic
or operating theatre. It detects concomitant valvular dis- rupture diagnosed by TOE should proceed directly to
ruption, myocardial contusion and pericardial collections surgical repair but, as false positive diagnoses are made
accurately and, unlike aortography and contrast enhanced using this test,40 in the stable patient, confirmatory angiog-
CT scans, does not involve administering potentially nephro- raphy is advisable prior to surgical exploration (see
toxic contrast material to patients already volume depleted. Fig. 34.4c).41
Transoesophageal echocardiography, however, is contraindi-
cated in patients with possible cervical spine, oropharyngeal
or maxillo-facial injury, oesophageal strictures or divertic- Magnetic resonance angiography
ula; further, ascending aortic tears may be missed due to
the intervening trachea, right main bronchus, and artefac- Although magnetic resonance angiography (MRA) has
tual acoustic shadows from cannulae.37 It should also be made significant contributions in the investigation of
borne in mind that a negative TOE does not always exclude
thoracic aortic rupture, even at the isthmus.
Despite the above advantages TOE is very operator
dependent38,39 and should be treated as a screening test
in patients whose transfer to the radiology department

(a) (c)

(b) (d)

Figure 34.4 A middle-aged man fell from a roof sustaining multiple injuries. (a) Initial chest X-ray was normal but when repeated it
showed widening of the mediastinum. (b) A computed tomography scan showed a haematoma of the mediastinum with an acute false
aneurysm at the aortic isthmus. (c) Digital subtraction angiography confirmed the diagnosis. (d) As inotropes were required to maintain
adequate blood pressure (in this injury hypertension is the norm), transoesophageal echocardiography was performed and ventricular
function was found to be impaired. Good views were obtained of the false aneurysm of the aorta, as shown
408 Vascular injuries of the chest

cervical vascular trauma it would not be regarded as a exploration where deemed appropriate with careful titra-
primary study for patients sustaining blunt or penetrating tion of fluid infusion in those whose surgery is delayed or
trauma.42 Magnetic resonance angiography may have a inappropriate.
role in excluding diaphragmatic or aortic injury when
spiral CT scan or angiography give equivocal results
and when differentiating trauma related masses from OPERATIVE MANAGEMENT
neoplasia.
The indications for surgical intervention after penetrating
thoracic trauma have evolved over some years and remain
GENERAL PRINCIPLES AND PITFALLS useful guidelines.
IN TREATMENT

The clear aim of the trauma surgeon is to ensure that poten-

Indications for surgical intervention after
tially life-threatening injuries are not missed. A majority of
penetrating thoracic trauma
patients with initially stable, but potentially fatal thoracic
injuries will have sustained either penetrating cardiac injury • Cardiac arrest
or contained rupture of the thoracic aorta. The stable • Significant or sustained haemorrhage (1 L
immediately on drain insertion, 500 mL in first
patient with penetrating cardiac trauma is at risk of cardiac
hour or 250 mL/hour thereafter)
tamponade with impaired filling of the right ventricle.
Blood loss in these patients is usually of the order of • Mediastinal traversing injury
500 mL, and therefore rapid infusion of several litres of • Major vascular injury
clear fluid will only increase central venous pressure further • Sucking chest wound
without a corresponding rise in cardiac output. Stable con- • Persistent massive air leak
tained rupture of the thoracic aorta may progress to frank • Diaphragmatic rupture
rupture if the blood pressure is suddenly raised by aggres-
sive fluid resuscitation. In unstable patients admitted with gun shot wounds to
the chest, thoracotomy is usually undertaken as an emer-
gency in the operating theatre following the steps shown in
PREHOSPITAL MANAGEMENT the algorithm in Fig. 34.5.
Thoracic aortic rupture caused by blunt deceleration
In managing patients with thoracic trauma there is no trauma is the most common thoracic vascular injury
convincing evidence of any benefit from prehospital requiring surgery and is associated with an overall mortal-
intervention beyond establishing an airway and relieving ity of up to 98 per cent.43 It is traditionally considered
tension pneumothorax. Prehospital intervention is con- to be an extremely unstable injury and its diagnosis and
traindicated for haemothorax without respiratory com- repair have taken precedence over all other injuries.44
promise. A modified version of ‘scoop and run’ with The natural history of blunt thoracic aortic injury, how-
protection of the airway and relief of tension pneumo- ever, is that those patients who actually die due to free
thorax is advisable. rupture tend to do so within 4 hours of injury45 and deaths
beyond this time are principally due to other associated
injuries. As the established mode of treatment for non-
traumatic acute aortic dissection is pharmacological con-
RESUSCITATION
trol of hypertension, principally using -blockers to
suppress myocardial contractility and nitrates to decrease
Prehospital care is continued in the accident and emergency preload and after load, some authorities are now advocat-
department with an assured airway and full oxygenation. ing a similar policy in traumatic rupture.46 If the diagnosis
Conditions such as tension pneumothorax, cardiac tam- is suspected or confirmed, medical treatment is instituted
ponade and bleeding from external wounds are attended to to control blood pressure and, after other life-threatening
urgently. A systolic blood pressure of 70–80 mmHg, usually injuries have been treated, repair of the aortic injury is car-
sufficient to maintain cerebral perfusion, is acceptable. ried out on a semi-elective basis.45 The diminution of the
There is now substantial evidence, laboratory and clinical, force of ventricular ejection by the use of -blockers is
that automatic fluid infusion prior to surgery to contend probably more effective in preventing rupture than the
with haemorrhage not only converts a contained haemor- lowering of blood pressure by intravenous nitroprusside.
rhage into a free rupture but also dilutes coagulation fac- It was first reported in 1976 that aortic injuries do not
tors and precipitates ARDS, renal failure and fluid overload. all necessarily become more complicated if left surgically
Therefore, the emphasis should be on expediting surgical uncorrected. Two patients with aortic tears proximal to the
 Operative management 409

Trauma scene Protect airway and cervical spine

Relieve pain Assess and assist breathing (oxygen)
Reassure Identify injuries
Rewarm Decompress tension pneumothorax
Run to emergency room Quickly establish intravenous access if possible
Record damage to structure and injuries to
other individuals
Communicate with major trauma centre
NO STAY AND PLAY

Emergency room Protect airway and cervical spine

Assess and assist breathing
Intubate if essential
Identify injuries
Decompress tension pneumothorax
Establish intravenous access, CVP
Massive haemothorax ECG and SaO2 (nasogastric tube if safe)
Cautious replacement of intravenous volume
Blood taken for ABG, cross-match,
Haematology and biochemistry
Stable Unstable Chest radiograph
Spiral CT Simple pneumothorax or haemothorax
TOE Chest drain unless massive haemothorax
(1.5 L estimate)

Rapidly deteriorates Widened mediastinum Normal

Emergency room Spiral CT Chest radiograph
Thoracotomy TOE
Aortogram

Operating room
Positive Negative
Thoracotomy

Observation

Figure 34.5 Algorithm of management of chest trauma with potential great vessel injury. There is no ‘golden hour’ for this scenario.
ABG, arterial blood gas analysis; CPV, central venous pressure; CT, computed tomography; ECG, electrocardiogram; SaO2 , oxygen saturation;
TOE, transoesophageal echocardiography

isthmus followed-up for up to 2 years showed no evidence Table 34.1 ‘Permissive hypotension’ in thoracic vascular trauma
of progression of those lesions.47 Expanding on this
Patient selection Diagnosed major vascular injury, or
experience, Akins et al.48 reported a series of 44 patients
Very high suspicion of major
with thoracic aortic rupture, 19 of whom received initial
vascular injury
pharmacological management without adverse compli- Anaesthesia Sedated
cations. While surgical repair was generally performed Paralysed
within 2–79 days of conservative management, in five cases Ventilated
a long term observational approach was adopted. It is Antihypertensive Esmolol
noteworthy that, of the five deaths following immediate infusion Nipride/labetalol
surgical repair, two resulting from complications of associ-
Monitoring Urinary catheter
ated injuries might have been prevented had they been Invasive arterial blood pressure
treated definitively before aortic repair. Although this Invasive central venous pressure
delayed approach in dealing with aortic rupture is now
Arterial blood pressure Systolic blood pressure
generally accepted as safe,49,50 prospective randomised 110 mmHg
trials have yet to be done. In some reports of initial conser-
Urine output 0.5 mL/kg per minute
vative management of these injuries, fatal free rupture
occurred within 24–72 hours of injury.45,51 We feel, there- ‘Danger moments’ Induction of anaesthesia
fore, that if a patient presents with a definite diagnosis Intubation
Movement to radiology or
within 5 days of injury, and even though a ‘permissive
operating table
hypotension’ regimen (Table 34.1) might have been
410 Vascular injuries of the chest

instituted, surgery should not be delayed unless there are short segment and have a discrete neck. Stenting for supra-
specific contraindications to doing so (see below). aortic injuries has been successful but, until further experi-
ence has accrued, should be regarded as experimental.

Contraindications to immediate aortic repair Surgical approaches

(after Maggisano et al.51)
The approach chosen must be based on the principle of
• Glasgow Coma Scale less than 6 within 10 days of securing proximal and distal control of the injured ves-
injury sel(s). Median sternotomy affords the best access to major
• Evidence of intracranial haemorrhage on CT scan cardiac, ascending aortic, cervico-mediastinal and main
• Acute respiratory distress syndrome pulmonary vessel injuries (Fig. 34.6a). For further access to
• Inability to tolerate one-lung anaesthesia the supra-aortic vessels distally, the median sternotomy
• Inotropic support required can be extended to an oblique cervical or supraclavicular
• Coagulopathy despite adequate treatment incision, dividing or resecting the clavicle as required, or by
• Sepsis due to other injuries, especially intra- adding a third space anterior thoracotomy to create a ‘trap-
abdominal door’ incision (see Fig. 34.6a).
In the presence of profound haemodynamic instability,
the equipment required for emergency sternotomy may
Patients diagnosed 5 days or more, but less than not be available and in these cases an anterolateral thora-
3 months, after injury may be managed with antihyperten- cotomy sited immediately inferior to the left pectoralis
sives and scheduled for repair within a few days. Patients muscle and entering the chest via the fourth intercostal
presenting after 3 months should be considered for elective space requires no more advanced equipment than a scalpel
repair based on the size and expansion rate of the false (Fig. 34.6b). If further access is required, for example to
aneurysm. institute CPB, and if conditions permit, this incision can be
easily extended across the sternum into the right chest to
create a ‘clam-shell’ approach (see Fig. 34.6b).
Endovascular approaches

There is a growing number of reports of the use of

endovascular stents in acute traumatic aortic rupture.52–58 Control and repair techniques
The left subclavian artery may have to be sacrificed as
only a short segment of aorta lies distally within which CARDIAC
a stent can be ‘landed’. In the young patient surviving The pericardium may be opened allowing relief of tam-
such a major injury, collateral perfusion of the arm is ponade and more efficient internal cardiac massage.
usually adequate, but subclavian flow can be augmented Emergency control of cardiac injury can be achieved by
either by a bypass in the neck or, via the left cubital finger pressure while resuscitation progresses. In some
fossa, placing a second stent penetrating the main aortic emergency situations the surgeon may just have the time to
stent. place the patient on CPB before executing the necessary
The procedure is usually performed under general repairs (Fig. 34.7). Definitive repair usually consists of sim-
anaesthesia, though these patients are frequently ventilator ple mattress sutures, buttressed by pericardial or prosthetic
dependent at the time of induction. Initial imaging is by pledgets. More extensive injuries are often incompatible
CT scan with three-dimensional reconstructions which with survival. Valvular injuries may become evident later
assist in planning a positioning strategy for the stent.59 and can be dealt with in a semi-urgent manner under CPB.
Magnetic resonance angiography may provide more accu-
rate and specific images of the anatomy. The femoral
SUPRA-AORTIC VESSELS
artery, usually the left, is cannulated with an 18 or 20 Fr
sheath which is passed into the proximal aortic arch where The primary objective is to secure proximal and distal con-
contrast is injected. The stent is passed over a guidewire trol of a major vessel and the inability to do so swiftly, espe-
and positioned using fluroscopy. Endoleak occurs in about cially amidst bleeding and a haematoma under pressure, is
10 per cent of cases56,57 and can be treated by further stent the most common cause of intraoperative mortality.
placement in the majority of cases (see Chapter 30). Long Immediate compression is essential while adequate expos-
term follow-up is not yet available using this new endovas- ure is being obtained for clamping. The left innominate
cular technique but it is not free from neurological or other vein may be clamped and divided with impunity to gain
side effects.60 access to life-threatening arterial bleeding beneath. The
Post-traumatic aneurysms have proven to be ideal for occasional anomalous origins of the innominate and the
endovascular stent placement as they involve a relatively right subclavian must be borne in mind. If the injury
 Operative management 411

(a) (b)

Figure 34.6 Surgical approaches. (a) A median sternotomy provides access to

the great vessels of the aortic arch. Extension upwards into supraclavicular and
oblique cervical incisions permits exposure to the subclavian and carotid
arteries, respectively. A left third space anterior thoracotomy, alone, or as a part
of the ‘book’ or ‘trapdoor’ approach (in which the clavicular attachments of the
sternocleidomastoid and strap muscles and the mid-clavicle are divided), enables
access to the left subclavian artery. (b) A left fourth space anterolateral
thoracotomy can be extended trans-sternally into the third right space to offer a
bilateral anterolateral thoracotomy or ‘clam-shell’ approach. (c) A left fourth
(c) space posterolateral thoracotomy

involves the innominate or left common carotid artery at bypass, is required. An injury at the origin of the innomi-
the origin from the aortic arch, an aortic side-biting par- nate artery is best dealt with by a 10 mm Dacron aorto-
tially occluding clamp will control bleeding (Fig. 34.8a). innominate bypass graft taken from a fresh aortotomy on
Shunting should be instituted where necessary, contribut- the proximal ascending aorta (Fig. 34.8b); the injury site is
ing favourably in ensuring maximal cerebral perfusion in closed by lateral suture reinforced with pledgets or by a
the hypotensive patient. This measure is of particular value woven Dacron patch. In order to restore flow to concur-
in simultaneous injuries to the innominate and left com- rently injured innominate and left common carotid arter-
mon carotid arteries. ies, a bifurcated Dacron graft is ideal (Fig. 34.8c), and from
Except for small limited lacerations lateral suture is not which a ‘jump’ graft may be taken to a subclavian artery if
easily achieved without tension because the ends retract necessary. Complex tears demand the institution of CPB,
and patch grafting is preferable. In practice, however, a induced hypothermic cardiac arrest and graft repair of the
Dacron softer knitted graft, either bridging the gap or as a aorta.
412 Vascular injuries of the chest

and drainage. Evacuation of a haemothorax is often pos-

sible by minimally invasive thoracoscopy.

AORTIC ISTHMUS
The standard approach is a high left posterolateral thora-
cotomy, usually at the third or fourth space, and dividing
the inferior rhomboids (see Fig. 34.6c). This permits unre-
strained access to the aortic arch and subclavian artery
for proximal control. It is important, in the heat of the
moment, not to make the incision too low as that will limit
access.

Findings
A large amount of blood may be lost into the pleural cavity
though mostly it tends to form an extrapleural haematoma
obscuring the aorta. Opening the pleura may result in free
Figure 34.7 A depressed young man managed to shoot himself rupture and digital pressure is required until clamps can be
in the chest with a crossbow and was admitted with the bolt still applied.
lodged in the centre of his sternum. The bolt had transfixed the
right and left ventricles, the descending thoracic aorta and the
ninth thoracic vertebra. Under full cardiopulmonary bypass, with Institution of bypass
cardioplegic arrest of the heart, the bolt was removed and the Where a form of shunt or bypass is to be used, preparations
injuries repaired. Cardiac injuries of this nature, requiring should be in place before making the skin incision. For
cardiopulmonary bypass for repair, are relatively uncommon partial atriofemoral bypass (Fig. 34.9a), in which the
femoral vessels are to be used, the groin should be prepared
PULMONARY and the cannulae inserted before turning the patient onto
the left lateral position. Access to the left atrium is usually
In the presence of major bleeding from the pulmonary via the atrial appendage or superior pulmonary vein.
vessels the source of loss may be obscured. Control is Alternatively, the inferior pulmonary vein may be cannu-
achieved either by clamping the pulmonary vessels within lated, thereby avoiding opening the pericardium and leav-
the pericardium or by surrounding the hilum, initially ing the cannula outside the operative field. After placement
manually, until a sloop and ‘snugger’ can be fashioned. of a purse string, the cannula is inserted and connected to
Being a low pressure system the pulmonary vessels can be a centrifugal pump ensuring arterial return to the femoral
controlled by oversewing lung tissue. Lobectomy may be artery. When a shunt, heparinised or heparin bonded, is to
necessary to establish control of more centrally placed vas- be used, purse strings are placed either in the aorta well
cular trauma or of lung tissue obliterated by a shotgun proximal to the left subclavian or in the apex of the left
blast or a disintegrated high velocity missile. In these cir- ventricle (Fig. 34.9b).
cumstances, pneumonectomy may be the only available
course, following which the chances of survival are signifi-
cantly compromised. Control
Control should precede aortic exploration. As the prox-
imal aorta may be difficult to identify it is worth dissecting
VENA CAVAL SYSTEM
the left subclavian artery and following it to the aorta. The
A trans-sternal extension of the anterolateral thoracotomy proximal clamp is usually placed proximal to and possibly
(see Fig. 34.6b) is necessary to deal with two potentially including the subclavian artery. Alternatively, the sub-
lethal and rather inaccessible injuries, namely, the poster- clavian can be controlled by clamping or by passing a
ior aspect of the intrapericardial inferior vena cava and the sloop around it twice. Distal control is usually obtained
azygous vein. with ease by clamping relatively normal aorta at the level
of the diaphragm. When the level of tear is identified it
CHEST WALL may be possible to apply the clamp closer to it thereby
While chest wall bleeding, particularly from intercostal or reducing back-bleeding from intercostal vessels.
internal mammary arteries, may result in dramatic collapse,
patients usually respond well to chest drainage and fluid Dissection of a haematoma
resuscitation. Continued bleeding to a level exceeding Once clamps are placed the mediastinal pleura and the aor-
500 mL/hour is an indication for thoracotomy. Repair will tic adventitia are opened longitudinally and the rupture is
usually mean suture ligation, removal of pleural thrombus exposed. The tear may not be immediately evident if it lies
 Operative management 413

(a) (b)

Figure 34.8 Supra-aortic vessel trunk injury. (a) A deceleration

injury of the innominate artery controlled by side-biting clamp.
(b) Aorto-innominate bypass graft and closure of the origin of the
innominate artery. (c) Aorto-innominate-left common carotid artery
bypass for penetrating injury to both great vessels, the stumps of
(c) which have been closed

in the posterior wall. The ends of the aorta will have most instances a Dacron graft is necessary (Fig. 34.9c) and
retracted and apposition to facilitate direct suture may should be matched to the size to the aorta prior to cross-
seem possible but it should not be attained at the cost of clamping. A useful tip would be to measure the size of the
tension at the point of repair. In most cases an interposition aorta on the contrast enhanced CT scan and have a suitable
woven Dacron tube graft will be necessary. graft available in theatre.

Graft and sewing techniques Issues of contention in surgery for acute

For minor tears a monofilament non-absorbable suture, thoracic aortic rupture
knotted on the outside, is commenced from each end of Paraplegia after repair of thoracic aortic rupture is a devas-
the tear. Before tying, the distal and proximal clamps are tating complication and has significant medico-legal con-
released in turn to evacuate air and thrombus. The site is sequences. It is present prior to surgery in 2.6 per cent of cases,
packed while the anaesthetists restore haemodynamic sta- having been caused by direct spinal cord injury or by injury
bility and then reinspected for leaks which are oversewn. In to the spinal arteries.61 Although the actual mechanism of
414 Vascular injuries of the chest

Centrifugal
pump

(a) (b) (c)

Figure 34.9 Deceleration injury of the isthmus of the thoracic aorta. (a) Atriofemoral pump assisted bypass in place prior to aortic repair.
(b) Ascending aorta to descending aorta shunt, which may be heparin bonded, preparatory to aortic repair. (c) Aortic ‘clamp and repair’ using
Dacron graft

postoperative paraplegia is not known it is related to clamp though mortality rates are almost identical at approxi-
time and duration of ischaemia of the anterior spinal cord. mately 15 per cent.63–65 The risk of paraplegia does rise
Despite collateral supply, relative hypoperfusion of the dis- steeply with the ‘clamp and sew’ technique once aortic
tal descending thoracic aorta during the procedure is cross-clamp times exceed 30 minutes.63,65
inevitable. Intercostal artery ligation during mobilisation Passive shunts, partial or full CPB may be employed to
of the aorta may also be a precipitating factor. increase perfusion pressure in the descending thoracic
The blood supply to the anterior two-thirds of the spinal aorta and also to prevent left ventricular distension during
cord is from the anterior spinal artery. This vessel is usually cross-clamping. Those techniques, however, are associated
well developed in the upper rather than in the lower thorax with several hazards: cardiac tamponade, ventricular fibril-
where it is small, the anterior cord being dependent on vari- lation, atrial fibrillation and phrenic nerve injury,66 all of
able segmental branches of intercostal arteries. Between T12 which may follow placement of left atrial cannulae.
and L4 vertebrae the small anterior segmental artery receives Heparin bonded shunts (see Fig. 34.9c) may not protect
flow from the artery of Adamkiewicz, a sub-branch of the against thromboembolism and cerebral infarction.67 How-
intercostals, and, in over 25 per cent of the population, ever, the degree of heparinisation required for bypass may
essential to the blood supply of the cord.62 Therefore, distal exacerbate bleeding from other injuries. Misplacement of
thoracic aortic hypotension during aortic cross-clamping the distal end of a shunt with resultant paraplegia has also
may lead to hypoperfusion of the anterior spinal cord. been reported.68
Amidst conflicting data, there does appear to be evi- Currently, widespread variation in practice is explained
dence, in uncontrolled retrospective studies, of a lower rate by the lack of prospective randomised studies which are
of paraplegia when distal circulatory support is available essential in making informed decisions.69 The experience
during repair of thoracic aortic rupture. The ‘clamp and of major referral centres accepting ‘self-selected’ stable
sew’ technique, namely, simple aortic cross-clamping patients cannot be easily extended to local institutions to
without attempts to perfuse the aorta distally (see Fig. which patients may be admitted in extremis and where
34.9c), has the advantages of simplicity and the avoidance bypass equipment may not be available. In a number of
of systemic heparinisation and its deleterious effects on series, including stable and unstable patients and where all
other injuries. This approach, however, carries a consistent techniques were used, the authors concluded that the out-
paraplegia rate of 16–20 per cent, compared with 4–6 per come depended mainly on preoperative clinical status,
cent when some form of distal circulatory support is used, associated injuries and the timing of surgery.69,70
 Results 415

A recent report detailed one case of paraplegia (1.4 per bypass using a centrifugal pump has the lowest incidence
cent) after 71 cases of thoracic aortic rupture were managed of paraplegia at 0–2.9 per cent and a mortality rate of
without distal circulatory support.71 In that institution an 11.9–14.5 per cent.63,72,74
average of 9.4 cases were treated per year, compared with Although for these reasons left heart bypass using a cen-
2.3 cases in most other large centres, and the mean cross- trifugal pump is often recommended as the technique of
clamp time was 24 minutes, with only four cases exceeding choice,75 the excellent results of the ‘clamp and sew’ technique
30 minutes. It would seem, therefore, that the ‘clamp and in expert hands71 and the use of full CPB by others73 indicates
sew’ technique is a safe option in expert hands, and sup- that the best policy has yet to be determined. Nevertheless, the
ports the early transfer of stable patients to specialised units. crucial aspects of acute management of thoracic aortic rup-
It is the only viable option in those patients who develop ture are the institution of ‘permissive hypotension’ in those in
frank rupture and are moribund; the likelihood of survival whom the diagnosis is suspected and the transfer of stable
in these cases remains low at 0–6.2 per cent.63,72 In another patients to the nearest centre wherein appropriate surgical,
large series using the ‘clamp and sew’ technique no patients anaesthetic and intensive care expertise resides.73
developed paraplegia when clamp time was less than 35
minutes; a low but significant paraplegia rate was observed
on bypass. The conclusion drawn by the authors probably
POSTOPERATIVE CARE
summarises the current best advice: first, paraplegia can be
minimised by ‘clamp and sew’ where the repair appears
straightforward, and second, that bypass should be used Rapid advances in anaesthesia, monitoring and pain relief
when prolonged cross-clamp time is anticipated.64 have contributed to improved outcomes. In selected patients,
i.e. those in whom head injury and intra-abdominal injury
Distal aortic perfusion techniques have been excluded, the current preferred method for
Considerable controversy remains regarding the optimal pain control following severe thoracic trauma is epidural
method of distal perfusion were it to be used. The tech- analgesia.76
niques employed have ranged from passive shunts, to partial The development of regional techniques of analgesia has
left heart bypass and to full CPB. The passive shunt most revolutionised the management of blunt thoracic trauma. It
commonly used is the Gott shunt which is a heparin coated has evolved from a policy of intubation and mechanical
purpose designed shunt tapered at either end, although two ventilation for all patients to one of optimising pain con-
standard CPB arterial cannulae joined by bypass tubing may trol combined with chest physiotherapy. Although many
be a satisfactory substitute. The shunt is usually inserted methods have been used, in carefully selected patients with
connecting the ascending aorta with the distal aorta (see severe thoracic trauma, epidural analgesia is the preferred
Fig. 34.9b). Access to the former site is readily achieved by a technique for pain control.76 Epidural catheters for con-
generous posterolateral thoracotomy. Less commonly, the tinuous narcotic or local anaesthetic administration are
apex of the left ventricle is chosen as the proximal cannula- both the most reliable and the most effective; once in place
tion site. In the meta-analysis by von Oppell,63 shunting they can be easily managed by nurses outside the inten-
from the ascending aorta was associated with a paraplegia sive care setting. Although the benefit in outcome using
rate of 8.2 per cent, whereas when the left ventricle was used epidural analgesia has yet to be demonstrated objectively,
it was 26.1 per cent. The high incidence for the latter tech- the parameters of pulmonary function and subjective
nique may reflect the mean cross-clamp time of 40.6 min- patient comfort have clearly improved.
utes, none being less than 39 minutes. The technique of
passive shunting was seldom employed in two recently pub-
RESULTS
lished large series;64,72 we feel, however, that it is still a useful
technique in situations when systemic heparin has to be
avoided in patients undergoing early repair of thoracic aortic The majority of treated thoracic vascular injuries have few
rupture in the presence of injuries elsewhere. long term sequelae but two specific complications of aortic
Active support of the distal circulation ranges from left trauma deserve specific mention. Late false aneurysm for-
heart bypass, i.e. either left atrial to distal aorta or femoral mation is often the result of untreated rupture caused by
artery using a centrifugal pump (see Fig. 34.9a) or femoral apparently trivial trauma, but it may also occur after suc-
vein to femoral artery bypass without an oxygenator, both cessful surgical repair. A false aneurysm may come to light
of which can be employed without systemic heparinisa- when it ruptures but in most cases it is a chance finding on
tion, to CPB with full heparinisation. Although the inci- chest X-ray taken for some other purpose. While small
dence of paraplegia for full CPB is only 2.4–4.5 per cent, asymptomatic false aneurysms can be kept under observa-
the mortality rate of 18.2–22.7 per cent is higher than usu- tion, surgical repair is usually advisable, though endovas-
ally reported63,72 for all other techniques, and therefore its cular approaches are likely to play an increasing role in
use in patients with pulmonary or intracranial injuries is management. The importance of follow-up is illustrated by
unwise.73 Recent large studies have shown that left heart the fact that false aneurysms occasionally develop into
416 Vascular injuries of the chest

aorto-oesophageal or aortobronchial fistulae, the majority REFERENCES

of which are fatal. Although definitive surgical repair of
major thoracic vascular injuries is often successful, annual
1 Besson A, Saegesser F. Color Atlas of Chest Trauma and
lifelong X-ray follow-up is mandatory.
Associated Injuries. Oradell, 1983.
2 Gebhard F, Kelbel MW, Strecker W, et al. Chest trauma and its
impact on the release of vasoactive mediators. Shock 1997; 7:
Conclusions 313–17.
3 Trunkey DD. Current therapy of trauma. In: Trunkey DD, Lewis FR.
Techniques continue to evolve in the diagnosis and (eds.) St Louis, MO: CV Mosby, 1984.
treatment, surgical and endovascular, of major thoracic 4 Rich NM, Spencer FC. Vascular Trauma. Philadelphia, PA: WB
vascular injuries. Most of the newer modalities are Saunders Co, 1978.
applicable to the stable patient presenting with con- 5 Reul GJJ, Beall ACJ, Jordan GL, Mattox KL. The early operative
tained haemorrhage. Injuries presenting with free bleed- management of injuries to the great vessels. Surgery 1973; 74: 862.
ing call for a ‘scoop and run’ approach to admission, 6 Marsh CL, Moore RC. Deceleration trauma. Am J Surg 1957;
minimal investigation and resuscitative surgery. Fluid 93: 623.
7 Mimasaka S, Yajima Y, Hashiyada M, et al. A case of aortic
resuscitation prior to surgical control of haemorrhage is
dissection caused by blunt chest trauma. Forensic Sci Int 2003;
counterproductive and increases morbidity in survivors.
132: 5–8.
The stable patient with contained haemorrhage should 8 Straub A, Beierlein W, Kuttner A, et al. Isolated coronary artery
be managed on a ‘permissive hypotension’ regimen until rupture after blunt chest trauma. Thorac Cardiovasc Surg 2003;
essential investigations are undertaken and after attend- 51: 97–8.
ing to other life-threatening injuries. 9 Chen MY, Regan JD, D’Amore MJ, et al. Role of angiography in
While CT and three-dimensional reconstruction is the detection of aortic branch vessel injury after blunt thoracic
replacing aortography as the standard investigation, trauma. J Trauma 2001; 51: 1166–71; discussion 1172.
other modalities such as TOE and MRA are playing an 10 Colour Atlas of Chest Trauma. London: Wolfe Medical
increasingly important role. The aim is to bring a stable Publications, pp. 171–6.
patient to surgery in a more controlled manner and 11 Petre R, Chilcott M, Murith N, Panos A. Blunt injury to the
supra-aortic arteries. Br J Surg 1997; 84: 606–9.
to repair or graft the injured vessels, possibly with the
12 Karmy-Jones R, DuBose R, King S. Traumatic rupture of the
aid of appropriate bypass to reduce the incidence of
innominate artery. Eur J Cardiothorac Surg 2003; 23: 782–7.
paraplegia, and to prevent renal failure or stroke. The 13 Khanna A, Drugas GT. Air gun pellet embolization to the right
urgency of the situation, however, may still demand a heart: case report and review of the literature. J Trauma 2003;
‘clamp and sew’ approach. 54: 1239–41.
Surgical repair represents just the commencement of 14 Mattox KL. Approaches to trauma involving the major vessels of
management of patients with thoracic vascular injury. the thorax. Surg Clin North Am 1989; 69: 77.
Having suffered major trauma they will generally 15 Robbs JV, Barker LW. Subclavian and axillary artery injuries. SA
require prolonged intensive care, months or years of Med J 1976; 51: 227.
rehabilitation and, given the risk of false aneurysm and 16 Meredith JW, Trunkey D. Thoracic Gunshot Wounds in
fistula formation, follow-up for life. Management of Gunshot Wounds. New York: Elsevier Science
Publishing, 1988.
17 Bickell WH, Bruttig SP, Millnamow GA, et al. The detrimental
effects of intravenous crystalloid after aortotomy in swine.
Surgery 1991; 110: 529–36.
Key references
18 Craig RL, Poole GV. Resuscitation in uncontrolled hemorrhage.
Am Surg 1994; 60: 59–62.
Bickell WH, Wall MJ Jr, Pepe PE, et al. Immediate versus delayed
19 Sakles JC, Sena MJ, Knight DA, Davis JM. Effect of immediate
fluid resuscitation for hypotensive patients with penetrating
fluid resuscitation on the rate, volume, and duration of
torso injuries. N Engl J Med 1994; 331: 1105–9.
pulmonary vascular hemorrhage in a sheep model of penetrating
Fabian TC, Richardson JD, Croce MA. Prospective study of blunt
thoracic trauma. Ann Emerg Med 1997; 29: 392–9.
aortic injury: multicenter trial of the American Association for
20 Fabian TC, Davis KA, Gavant ML, et al. Prospective study of blunt
the Surgery of Trauma. J Trauma 1997; 42: 374–83.
aortic injury: helical CT is diagnostic and antihypertensive
Mattox KL. Approaches to trauma involving the major vessels of
therapy reduces rupture. Ann Surg 1998; 227: 666–76.
the thorax. Surg Clin North Am 1989; 69: 77.
21 Kaweski SM, Sise MJ, Virgilio RW. The effect of prehospital fluids
Saletta S, Lederman E, Fein S, et al. Transesophageal
on survival in trauma patients. J Trauma 1990; 30: 1215–18;
echocardiography for the initial evaluation of the widened
discussion 1218–19.
mediastinum in trauma patients. J Trauma 1995; 39: 137–41;
22 Demling RH, Manohar M, Will JA. Response of the pulmonary
discussion 141–2.
microcirculation to fluid loading after hemorrhagic shock and
Symbas PN, Sherman AJ, Silver JM, et al. Traumatic rupture of the
resuscitation. Surgery 1980; 87: 552–9.
aorta: immediate or delayed repair? Ann Surg 2002; 235:
23 Bickell WH, Barrett SM, Romine-Jenkins M, et al. Resuscitation
796–802.
of canine hemorrhagic hypotension with large-volume isotonic
 References 417

crystalloid: impact on lung water, venous admixture, and systemic 44 Borman KR, Aurbakken CM, Weigelt JA. Treatment priorities in
arterial oxygen saturation. Am J Emerg Med 1994; 12: 36–42. combined blunt abdominal and aortic trauma. Am J Surg 1982;
24 Bickell WH, Wall MJ Jr, Pepe PE, et al. Immediate versus delayed 144: 728–32.
fluid resuscitation for hypotensive patients with penetrating 45 Pate JW, Fabian TC, Walker W. Traumatic rupture of the aortic
torso injuries. N Engl J Med 1994; 331: 1105–9. isthmus: an emergency? World J Surg 1995; 19: 119–25;
25 Matsuoka T, Hildreth J, Wisner DH. Uncontrolled hemorrhage discussion 125–6.
from parenchymal injury: is resuscitation helpful? J Trauma 46 Pate JW. Is traumatic rupture of the aorta misunderstood? Ann
1996; 40: 915–21; discussion 921–2. Thorac Surg 1994; 57: 530–1.
26 Jacobs LM. Timing of fluid resuscitation in trauma. N Engl J Med 47 Turney SZ, Attar S, Ayella R, et al. Traumatic rupture of the aorta.
1994; 331: 1153–4. A five-year experience. J Thorac Cardiovasc Surg 1976; 72:
27 Banerjee A, Jones R. Whither immediate fluid resuscitation? 727–34.
Lancet 1994; 344: 1450–1. 48 Akins CW, Buckley MJ, Daggett W, et al. Acute traumatic
28 Rady MY, Edwards JD, Nightingale P. Early cardiorespiratory disruption of the thoracic aorta: a ten-year experience. Ann
findings after severe blunt thoracic trauma and their relation to Thorac Surg 1981; 31: 305–9.
outcome. Br J Surg 1992; 79: 65–8. 49 Symbas PN, Sherman AJ, Silver JM, et al. Traumatic rupture of
29 Simon BJ, Leslie C. Factors predicting early in-hospital death in the aorta: immediate or delayed repair? Ann Surg 2002; 235:
blunt thoracic aortic injury. J Trauma 2001; 51: 906–10; 796–802.
discussion 911. 50 Kwon CC, Gill IS, Fallon WF, et al. Delayed operative intervention
30 Eddy VA. Is routine arteriography mandatory for penetrating in the management of traumatic descending thoracic aortic
injury to zone 1 of the neck? Zone 1 Penetrating Injury Study rupture. Ann Thorac Surg 2002; 74: S1892–8.
Group. J Trauma 2000; 48: 208–13. 51 Maggisano R, Nathens A, Alexandrova NA, et al. Traumatic
31 Graham AN, McManus KG, McGuigan JA, McIlrath E. Traumatic rupture of the thoracic aorta: should one always operate
rupture of the thoracic aorta. Ann R Coll Surg Engl 1995; 77: immediately? Ann Vasc Surg 1995; 9: 44–52.
154–5. 52 Ahn SH, Cutry A, Murphy TP, Slaiby JM. Traumatic thoracic aortic
32 Durham RM, Zuckerman D, Wolverson M. Computed tomography rupture: treatment with endovascular graft in the acute setting.
as a screening examination in patients with suspected blunt J Trauma 2001; 50: 949–51.
aortic injury. Ann Surg 1994; 220: 699–704. 53 Lagattolla N, Matson M, Self G, et al. Traumatic rupture of the
33 Agee CK. Computed tomographic evaluation to exclude aortic arch treated by stent grafting. Eur J Vasc Endovasc Surg
traumatic aortic disruption. J Trauma 1992; 33: 876–81. 1999; 17: 84–6.
34 Miller FB, Richardson JD, Thomas H. Role of CT in the diagnosis 54 Semba CP, Mitchell RS, Miller DC, et al. Thoracic aortic
of major arterial injury after blunt thoracic trauma. Surgery aneurysm repair with endovascular stent-grafts. Vasc Med 1997;
1989; 106: 596–602. 2: 98–103.
35 Raptopoulos V, Sheiman RG, Phillips DA, et al. Traumatic aortic 55 Orend KH, Kotsis T, Scharrer-Pamler R, et al. Endovascular repair
tear: screening with chest CT. Radiology 1992; 182: 667–73. of aortic rupture due to trauma and aneurysm. Eur J Vasc
36 Downing SW, Sperling JS, Mirvis SE, et al. Experience with spiral Endovasc Surg. 2002; 23: 61–7.
computed tomography as the sole diagnostic method for 56 Marty-Ane CH, Berthet JP, Branchereau P, et al. Endovascular
traumatic aortic rupture. Ann Thorac Surg 2001; 72: 495–501; repair for acute traumatic rupture of the thoracic aorta. Ann
discussion 501–2. Thorac Surg 2003; 75: 1803–7.
37 Lick SD, Zwischenberger JB, Mileski WJ, Ahmad M. Torn 57 Lamme B, de Jonge IC, Reekers JA, et al. Endovascular treatment
ascending aorta missed by transoesophageal echocardiography. of thoracic aortic pathology: feasibility and mid-term results.
Ann Thorac Surg 1997; 63: 1768–70. Eur J Vasc Endovasc Surg 2003; 25: 532–9.
38 Saletta S, Lederman E, Fein S, et al. Transesophageal 58 Bell RE, Taylor PR, Aukett M, et al. Results of urgent and
echocardiography for the initial evaluation of the widened emergency thoracic procedures treated by endoluminal repair.
mediastinum in trauma patients. J Trauma 1995; 39: 137–41; Eur J Vasc Endovasc Surg 2003; 25: 527–31.
discussion 141–2. 59 Horike K, Fukata Y, Kanoh M, Kurushima A. Two cases of stent graft
39 Smith DC, Bansal RC. Transesophageal echocardiography in the repair for thoracic aortic dissection: usefulness of three-dimensional
diagnosis of traumatic rupture of the aorta. N Engl J Med 1995; computed tomogram. Kyobu Geka 2001; 54: 573–6.
333: 457–8. 60 Greenberg R, Resch T, Nyman U, et al. Endovascular repair of
40 Oxorn D, Towers M. Traumatic aortic disruption: false positive descending thoracic aortic aneurysms: an early experience with
diagnosis in transoesophageal echocardiography. J Trauma 1995; intermediate-term follow-up. J Vasc Surg 2000; 31: 147–56.
39: 386–7. 61 Sturm JT, Hines JT, Perry JF. Thoracic spinal fractures and aortic
41 Vignon P, Lagrange P, Boncoeur MP, et al. Routine rupture: a significant and fatal association. Ann Thorac Surg
transoesophageal echocardiography for the diagnosis of aortic 1990; 50: 931–3.
disruption in trauma patients without enlarged mediastinum. 62 Adams HD, Von Geertruyden HH. Neurologic complications of
J Trauma 1996; 40: 422–7. aortic surgery. Ann Surg 1956; 144: 574.
42 Mirvis SE, Shanmuganathan K. MR imaging of thoracic trauma. 63 von Oppell UO, Dunne TT, De Groot MK, Zilla P. Traumatic aortic
Magn Reson Imaging Clin North Am 2000; 8: 91–104. rupture: twenty-year metaanalysis of mortality and risk of
43 Richens D, Kotidis K, Neale M, et al. Rupture of the aorta paraplegia. Ann Thorac Surg 1994; 58: 585–93.
following road traffic accidents in the United Kingdom 64 Hunt JP, Baker CC, Lentz CW, et al. Thoracic aorta injuries:
1992–1999. The results of the co-operative crash injury study. management and outcome of 144 patients. J Trauma 1996;
Eur J Cardiothorac Surg 2003; 23: 143–8. 40: 547–55; discussion 555–6.
418 Vascular injuries of the chest

65 Katz NM, Blackstone EH, Kirklin JW, Karp RB. Incremental risk 71 Sweeney MS, Young DJ, Frazier OH, et al. Traumatic aortic
factors for spinal cord injury following operation for acute transections: eight-year experience with the ‘clamp-sew’
traumatic aortic transection. J Thorac Cardiovasc Surg 1981; technique. Ann Thorac Surg 1997; 64: 384–9.
81: 669–74. 72 Fabian TC, Richardson JD, Croce MA. Prospective study of
66 Karmy-Jones R, Carter Y, Meissner M, Mulligan MS. Choice of blunt aortic injury: multicenter trial of the American
venous cannulation for bypass during repair of traumatic rupture Association for the Surgery of Trauma. J Trauma 1997; 42:
of the aorta. Ann Thorac Surg 2001; 71: 39–41; discussion 41–2. 374–83.
67 Duke BJ, Moore EE, Brega KE. Posterior circulation cerebral 73 Pate JW. Modern management of traumatic rupture of the aortic
infarcts associated with repair of thoracic aortic disruption using isthmus. Ann Thorac Surg 1998; 66: 611–12.
partial left heart bypass. J Trauma 1997; 42: 1135–9. 74 Read RA, Moore EE, Moore FA, Haenel RRT. Partial left heart
68 Verdant A, Mercier C, Page A, et al. Major mediastinal vascular bypass for thoracic aortic repair: survival without paraplegia.
injuries. Can J Surg 1983; 26: 38–42. Arch Surg 1993; 128: 747–52.
69 Tatou E, Steinmetz E, Jazayeri S, et al. Surgical outcome of traumatic 75 Von Oppell UO, Brink J, Hewitson J, et al. Acute traumatic
rupture of the thoracic aorta. Ann Thorac Surg 2000; 69: 70–3. rupture of the thoracic aorta. A comparison of techniques. S Afr J
70 DelRossi AJ, Cernaianu AC, Madden LD, et al. Traumatic Surg 1996; 34: 19–24.
disruptions of the thoracic aorta: treatment and outcome. 76 Ferguson M, Luchette FA. Management of blunt chest injury.
Surgery 1990; 108: 864–70. Respir Care Clin North Am 1996; 2: 449–66.
 35
Vascular Injuries of the Neck

S RAM KUMAR, FRED A WEAVER

The problem 419 Blunt carotid injuries 424

Historical perspective 419 Vertebral artery injuries 425
Mechanism of injury 419 References 427
Penetrating carotid injuries 420

THE PROBLEM infarction following CA repair in two neurologically

compromised patients supported Cohen’s admonition.9
Ligation was hence considered the procedure of choice in
Injury to major vascular structures of the head and neck patients who presented with ‘significant’ neurological
complicates 25 per cent of cervical injuries. Carotid artery deficits. More recent reports, however, of individual and
(CA) injuries constitute 5–10 per cent of all arterial collective experience with CA repairs in the neurologically
injuries, while injuries to the vertebral artery (VA), once compromised patient have not demonstrated a significant
thought to be uncommon, account for 30 per cent of all incidence of intracranial haemorrhage.3,10 Consequently,
cervical vascular injuries.1 Penetrating trauma is the lead- contemporary vascular surgical practice dictates repair of
ing cause of cervical vascular injury, with less than 10 per CA injuries, regardless of neurological findings.
cent of injuries following blunt trauma.2 Despite signifi- Until recently, VA injuries were considered uncom-
cant advances in the diagnosis and management of these mon, and usually only recognised at autopsy. This led to
injuries, mortality remains at 10–30 per cent, and the inci- the mistaken conclusion by Matas that isolated VA injuries
dence of permanent neurological deficit approximates were lethal.11 Few VA injuries were reported during the
40 per cent.3–5 world wars or during the Korean and Vietnam conflicts.
With the increased use of diagnostic angiography in the
management of cervical trauma, however, VA injuries have
HISTORICAL PERSPECTIVE been increasingly recognised. As with CA injuries, ligation,
first reported in 1852 by Maisonneuve,12 was the only
Ambroise Paré successfully treated a CA injury by ligation mode of therapy for VA injuries until the second half of the
in 1522.6 Ligation, which resulted in high rates of hemiple- twentieth century. In 1974, Serbinenko reported the use
gia and death, remained the only mode of management of percutaneous endovascular techniques to manage VA
until the twentieth century. The application of techniques injuries.13 Endovascular management has now replaced
of primary arterial repair, first employed during the surgical intervention as the treatment of choice for most
Korean conflict, to civilian carotid injuries, resulted in sig- VA injuries.
nificant reductions in neurological injury and mortality.
However, following a report by Wylie et al.7 of intracranial
haemorrhage after carotid revascularisation for acute
thrombotic stroke, Cohen questioned the wisdom of
MECHANISM OF INJURY
carotid repair in the neurologically compromised patient.8
He warned that reconstruction in the presence of a severe Penetrating cervical vascular injuries result from high
neurological deficit could result in a significant incidence velocity projectiles which injure vessels directly in their
of haemorrhagic infarction and death. Bradley’s subse- path (Fig. 35.1), or indirectly, by concussive forces or sec-
quent autopsy report in 1973, detailing haemorrhagic ondary missiles such as bone fragments. Injuries from low
420 Vascular injuries of the neck

velocity penetrating injuries such as stab wounds are gen- injured by bending and axial rotation of the cervical spine
erally confined to the path of penetration. than by cervical flexion or extension. The seemingly innocu-
ous activities of chiropractic neck manipulation and roller
coaster riding, where axial rotation and lateral bending may
Mechanisms of injury be accentuated, have been reported to cause VA injuries.

• Penetrating
– Knives PENETRATING CAROTID INJURIES
– Bullets
– Shrapnel
• Blunt Diagnosis
– Direct: blow to neck/Skull base fracture – CA
cervical spine fracture – VA Carotid artery injuries are classified into one of three
– Indirect: hyperextension/rotation of neck – CA anatomical zones (Fig. 35.2):
lateral flexion/rotation of neck – VA • I – injuries from the clavicle to the cricoid cartilage
• II – injuries between the cricoid and angle of mandible
Blunt injuries result from direct blows to the vessel or • III – injuries above the angle of mandible
by bony fragments from basilar or spinal fractures. Biffl A cervical bruit or thrill, a rapidly expanding cervical
et al.14 reported that 39 per cent of blunt VA injuries were haematoma and an absent CA pulse are pathognomonic
associated with a cervical spine fracture the presence of which signs of a CA injury (Fig. 35.3). Additional soft signs
was an independent clinical predictor of a VA injury. include a pulse deficit in the superficial temporal artery, signs
Significant injury may also be caused by seemingly inconse- of air embolism, active bleeding from an oropharyngeal or
quential cervical hyperextension and rotation. Excessive neck neck wound, widened mediastinum, ipsilateral Horner’s
extension or flexion stretches the CA over the transverse syndrome, or dysfunction of cranial nerves IX–XII. Con-
process of the second cervical vertebra, which may produce tralateral neurological deficits may be present, but obscured
an intimal crack or a mural contusion, typically in the distal by associated head injuries, systemic hypotension or the
internal carotid artery (ICA). The VA is more likely to be patient’s use of psychoactive substances.
The zone or location of the presumed CA injury dictates
the diagnostic work-up (Fig. 35.3). For zone II injuries,
both primary neck exploration and screening angiography
have a low positive yield.15 Consequently, many surgeons

Zone III

Common
Jugular carotid
vein artery Zone II

Zone I

Figure 35.2 Anatomical landmarks for carotid injuries by zone

(redrawn with permission from Donovan A (ed). Trauma Surgery:
Figure 35.1 Gunshot wound to the neck causing internal Techniques in Thoracic, Abdominal and Vascular Surgery.
carotid artery occlusion St Louis, MO: Mosby, 1994)
 Penetrating carotid injuries 421

have increasingly employed duplex scanning as the initial or high resistance flow suggestive of an upstream lesion in
screening assessment. Several studies have documented the non-visualised distal ICA. It should be emphasised that
the accuracy of duplex sonography in the diagnosis and the accuracy of a duplex evaluation and interpretation relies
follow-up of cervical vascular injuries when performed by heavily on a skilled technologist and an experienced reader.
a trained technologist (Table 35.1). Longitudinal and trans- If such resources are not available, angiography should be
verse grey scale and colour flow images have been shown to liberally employed despite the small yield.
correlate well with angiographic findings in over 90 per cent In contrast to the zone II injury, clinical evaluation
of injuries. Doppler waveform analysis can reveal turbulent and duplex scanning are inadequate for screening zones I

Patient with penetrating neck trauma suspicious for carotid trauma

Active cervical bleeding, expanding cervical

haematoma or airway compromise

Yes No
Immediate surgical repair Zone II injury?

Yes
Bruit, thrill, pulsatile mass
No

No Yes

Colour flow duplex Angiogram

Normal Abnormal

No injury Minimal injury Major injury Figure 35.3 Algorithmic approach to

the diagnosis and management of
Observation
penetrating carotid trauma (redrawn
Endovascular/open repair
with permission from Kumar SR,
Weaver FA, Yellin AE. Cervical vascular
injuries – carotid and jugular venous
injuries. Surg Clin North Am 2001; 81:
Follow-up
1331–44, xii–xiii)

Table 35.1 Clinical trials evaluating the role of duplex examination in the diagnosis of carotid trauma

Authors Year Findings Conclusion

Kuzniec et al.17 1998 91 per cent sensitive; 100 per cent Reproduces results of angiography
specific; 96 per cent accurate
Ginzburg et al.18 1996 100 per cent sensitive; Primary investigation of choice,
85 per cent specific patient-friendly, cheap
Fry et al.16 1994 100 per cent accuracy As effective as angiography, procedure of choice
Cogbill et al.19 1994 86 per cent accuracy Potential role in diagnosis and follow-up
20
Bynoe et al. 1991 95 per cent sensitive; 99 per cent specific; Cost effective, reliable method of diagnosis
98 per cent accurate
Meissner et al. 21 1991 100 per cent accuracy Rapid, effective screening tool
22
Greenwold et al. 1991 94 per cent sensitive; 99 per cent specific; Accurate screening tool
99 per cent accurate

Reproduced with permission from Kumar SR, Weaver FA, Yellin AE. Cervical vascular injuries – carotid and jugular venous injuries. Surg Clin North Am 2001;
81: 1331–44, xii–xiii.
422 Vascular injuries of the neck

and III CA injuries. Angiography is mandatory for injuries less than 5 mm in size or adherent, or downstream, non-
suspected in these zones (see Fig. 35.3) since bony structures obstructive intimal flaps. A high percentage of these injuries
such as the clavicle or mandible obstruct carotid insonation heal without surgical intervention,24 nevertheless, vessel
during a duplex evaluation. In addition, knowledge of the healing should be documented by follow-up duplex scans or
anatomy of the vascular tree is essential in planning the angiography. Patients with a CA occlusion on angiography
surgical approach to these injuries. Angiography also allows and a dense neurological deficit due to a large brain infarct
for endovascular management of injuries that are not on CT scan have a poor outcome regardless of treatment,4
surgically approachable and has the added benefit of whereas an occlusive ICA injury in a patient with a normal
detecting unsuspected vertebral, aortic arch, great vessel or neurological examination can be managed solely by anti-
contralateral CA injuries. coagulation and avoidance of hypotension. Anticoagula-
Brain computed tomography (CT) scans are routinely tion should be subsequently continued for a minimum of
obtained in patients with high velocity penetrating neck 3 months to prevent thrombus propagation. High zone III
injuries to evaluate associated head trauma, bony injuries injuries, not surgically accessible may best be managed by
and the state of the brain parenchyma. The presence of sig- endovascular means. All other angiographically documented
nificant parenchymal ischaemic injury soon after an occlu- CA injuries require operative repair regardless of neurologi-
sive CA injury indicates a poor neurological outcome, cal status.
regardless of therapeutic intervention. The technological
advance of helical CT scanning/angiography allows for
SURGICAL THERAPY
high resolution information concerning parenchymal injury
with the added option of reconstructing an in-continuity Zone II injuries are exposed by a neck incision overlying
image of the cervical vascular tree.23 and parallel to the anterior border of the sternocleidomas-
toid muscle. Zone I injuries frequently require a median
sternotomy for access to the more proximal portions of the
Management common carotid artery (CCA) or to deal with concomitant
arch injuries. Zone III injuries may require division of the
All injuries associated with pulsatile haemorrhage or an digastric muscle to provide exposure of the distal ICA. Access
expanding cervical haematoma with or without airway to the artery above the digastric may also be facilitated by
compromise mandate immediate surgical attention (see Fig. anterior subluxation of the mandible, which is accom-
35.3). In contrast, low velocity penetrating injuries of the plished by using archbars to fix the mandible in a subluxed
CA may result in angiographically non-occlusive ‘minimal position (Fig. 35.4). This converts a narrow triangular
injuries’, such as small intimal defects, pseudoaneurysms space between the skull and the mandibular ramus into a

(a) (b)

Figure 35.4 Technique for gaining access to

the distal internal carotid artery (redrawn with
permission from Donovan A (ed). Trauma
Surgery: Techniques in Thoracic, Abdominal and
(c) (d) Vascular Surgery. St Louis, MO: Mosby, 1994)
 Penetrating carotid injuries 423

wider, more rectangular opening, providing exposure to 3 months to prevent propagation of thrombus. As the ECA is
the CA up to the base of the skull. Osteotomy of the angle of an important collateral for cerebral flow, every effort should
mandible, instead of mandibular subluxation, is an alterna- be made to repair the isolated simple injury, but ligation is
tive method of achieving additional distal exposure. the prudent course when the injury is complex. For all CA
Vascular control is achieved in the usual fashion before repairs, routine completion angiography or duplex exami-
exposing the injury. When encountering a very distal ICA nation is mandatory to confirm a technically perfect repair
injury, a vascular balloon catheter or a balloon catheter and to assess the distal arterial tree for unsuspected injury or
shunt (Inahara–Pruitt) can be used for control by inserting thrombus.
the catheter into the vessel either through the injury itself, In the presence of associated aerodigestive tract injury or
or through a proximal arteriotomy. Once control is estab- repair, the vascular repair should be protected from saliva
lished, an intraluminal shunt accompanied by anticoagula- and bacterial contamination by the interposition of soft
tion is used to maintain antegrade flow for ICA repairs. tissue. This can be achieved by rotating the belly of the ster-
Routine use of an intravascular shunt instantly improves nocleidomastoid muscle after having divided it at its mas-
and maintains ICA cerebral perfusion. If an interposition toid insertion. In the patient with compromised airway,
graft is necessary, the shunt is passed through the lumen of early endotracheal intubation is preferred to tracheostomy,
the graft prior to anastomosis to the native artery. The graft as the latter increases the incidence of contamination of the
is then sewn in place and the shunt removed before place- vascular repair. Aerodigestive tract injury is a relative con-
ment of the last few sutures. In contrast to injuries of the traindication to the insertion of prosthetic material. When
ICA, and those of the CCA proximal to the bifurcation, a extensive soft tissue injury hinders cover of an arterial
shunt is not mandatory as cross-filling via the external repair, consideration should be given to primary ligation.
carotid artery (ECA) and circle of Willis is sufficient to
maintain cerebral perfusion.
ENDOVASCULAR THERAPY
Routine techniques of surgical repair such as lateral arte-
riorrhaphy, patch graft or excision of the injured area with The ability to treat CA injuries at the time of angiography
end-to-end or interposition graft repair are used to manage has improved with advances in catheter-based technology.
CA injuries. Autogenous grafts are preferred for zone II and Endovascular delivery of detachable coils, balloons or foam
III injuries, while prosthetic grafts are used for zone I CCA can be used to obliterate a traumatic carotid-cavernous fis-
injuries. Injuries to the carotid bulb or proximal ICA may be tula or a false aneurysm in a surgically inaccessible distal ICA
reconstructed with an interposition graft, or alternatively, segment.25 If endovascular obliteration is contemplated, the
the distal ICA may be transposed to the proximal stump patient’s neurological status must be monitored after a tem-
of the ECA (Fig. 35.5) following distal ECA ligation. Ligation porary test occlusion. If a contralateral deficit develops, a
of the ICA is recommended only when the distal vessel has bypass to the middle cerebral artery (CCA-MCA) may be
thrombosed and patency cannot be restored or if the injury required prior to obliteration. Covered stents have been
is so distal and extensive that it is not reconstructible. If ICA deployed successfully in treating traumatic false aneurysms
ligation is necessary, anticoagulation is continued for at least of the distal ICA.26

Internal carotid
artery
External
carotid
artery

Graft

Figure 35.5 (a) An injury of the proximal internal carotid artery (b) repaired by an interposition graft or (c) external carotid artery
transposition (redrawn with permission from Donovan A (ed). Trauma Surgery: Techniques in Thoracic, Abdominal and Vascular Surgery.
St Louis, MO: Mosby, 1994)
424 Vascular injuries of the neck

Table 35.2 Studies evaluating outcome following surgery for carotid trauma

Neurological deficit (per cent)

Mortality
Author Year Management At presentation At discharge (per cent)

Mittal et al.5 2000 89 per cent repair; 11 per cent ligation 0 0 17

Kuehne et al.4 1996 48 per cent repair; 8 per cent ligation; 24 12 4
44 per cent observation
Weaver et al.3 1988 68 per cent repair; 11 per cent ligation; 34 23 6
21 per cent observation
Robbs et al.27 1983 63 per cent repair; 37 per cent ligation 25 11 7
Brown et al.10 1982 80 per cent repair; 17 per cent ligation; 27 20 8
3 per cent observation

Results tool in patients with blunt neck trauma. Although the

accuracy of a duplex scan has not been evaluated prospec-
Repairs of CA injuries as described result in significant tively, experience is accruing with its application in this
improvement in absolute and relative neurological out- clinical setting. In a multi-institutional study of 60 blunt
comes.3 Even in the patient with a neurological deficit, CA injuries duplex scanning accurately demonstrated
restoration of cerebral perfusion can reverse the electrical 12 out of the 14 injuries studied.29 Other diagnostic modal-
failure surrounding the zone of irreversible cerebral injury ities include the use of helical CT scans to define zone III
(see Chapter 5), and result in dramatic neurological blunt CA injuries and gadolinium enhanced magnetic res-
improvement. In a retrospective review of 85 CA injuries, onance angiography (MRA) to rapidly image the carotid
we have shown that operative management regardless of and vertebral arteries from the aortic arch to the circle of
initial deficit improves mortality and ultimate neurological Willis.30 Until prospective studies evaluating non-invasive
status. Several other studies have corroborated these find- methods of diagnosis for blunt CA trauma become avail-
ings (Table 35.2). able, the demonstrated accuracy and universal availability
of angiography make it the primary investigative modality.
Conventional CT scans of the brain parenchyma are
BLUNT CAROTID INJURIES helpful in assessing the presence and extent of parenchy-
mal injury, bony spine injury, concomitant intracranial
Diagnosis haematomas,29 cerebral oedema, or cranial vault injuries.
A neurological deficit unexplained by CT of the brain
Blunt CA injuries can pose a significant diagnostic chal- raises the suspicion of a CA injury and is an indication for
lenge. The severity of trauma causing the carotid artery angiography.
to fracture may not be evident on physical examination.
Furthermore, a specific history or pattern of injury indica- Management
tive of CA injury cannot be elicited in the majority of
injuries. Consequently, a blunt CA injury may be missed in Maintenance of haemodynamic stability is critical when
up to two-thirds of patients. a blunt CA injury is suspected. Hypotension promotes
Although a few patients with blunt CA injury may report thrombus formation and propagation, while hypertension
loss of consciousness or present with abrupt focal neurolog- facilitates extension of the intimal dissection when present.
ical deficit at time of injury, in the majority no immediate The rare, short, symptomatic zone II dissection confined to
neurological symptoms are recorded. The classic history of the cervical CA should be repaired surgically. However,
loss of consciousness with a ‘lucid interval’ of hours to days most blunt CA injuries result in intimal disruption with
between the injury and the appearance of lateralising neu- dissection and/or thrombosis involving long segments of
rological symptoms occurs infrequently. Other less com- the CA (Fig. 35.6). Consequently, systemic anticoagula-
mon signs and symptoms include a ‘buzzing’ sound in the tion, which limits thrombus propagation and distal
ear, a cervical bruit, Horner’s syndrome, neck contusion or embolisation, is the preferred treatment for the vast major-
haematoma. Mandibular fracture is the most frequently ity of injuries. Heparin therapy is highly efficacious and is
associated injury28 and occasionally skull fractures are seen. associated independently with improved survival and
Although angiography remains the standard diagnostic better neurological outcomes.31 When instituted before
test, duplex scanning is being used increasingly as a screening the onset of symptoms, heparin therapy can pre-empt
 Vertebral artery injuries 425

VERTEBRAL ARTERY INJURIES

Diagnosis

The majority of VA injuries are clinically silent.

Neurological signs following VA trauma are rare because
of the collateral flow afforded by the contralateral VA
and the circle of Willis. In a series of 25 VA injuries,
Blickenstaff et al.33 reported only two (8 per cent) patients
with neurological symptoms of vertebrobasilar origin.
Other non-neurological findings, such as external haemor-
rhage, bruits and expanding or stable neck haematomas
are found in 10–30 per cent of VA injuries.1,33 The remain-
ing majority of injuries are an unsuspected additional find-
ing on cerebral angiography performed principally to
evaluate the carotid and brachiocephalic vessels.
In the asymptomatic patient, the possibility of a VA
injury should be suspected if the trajectory of a penetrating
injury crosses the VA or if cervical spine fractures are pre-
sent (Fig. 35.7).34 Some authors advocate angiographic
evaluation of all patients with blunt cervical spine frac-
tures.14 Cerebral angiography continues to be the gold
standard for the diagnosis, although it may be inaccurate
in characterising the true extent of the lesion.35 Although
duplex scanning has been recommended for evaluating the
vertebral arterial system, only the first part of the VA is
Figure 35.6 Carotid angiogram demonstrating long segment
accessible to insonation because the remainder of the
dissection of the internal carotid artery following blunt trauma artery is encased within the cervical spine and cranium.
Computed tomographic scans provide additional informa-
tion concerning the cervical spine and can be coupled with
neurological deterioration.31 Anticoagulation with war-
intravenous contrast enhancement and helical CT technol-
farin should then be continued for a period of at least
ogy to detect VA lesions. Gadolinium enhanced MRA per-
3 months.32
mits simultaneous evaluation of the cervical spine and
Endovascular therapy is being used more frequently for
brain parenchyma, but several anatomical factors includ-
blunt CA injuries, especially in the subgroup of patients
ing inherent arterial asymmetry and flow related venous
with contraindications to anticoagulation. The primary
enhancement limit its accuracy.36 Furthermore, the use
goal of endovascular therapy for CA dissection is restora-
of MR imaging is incompatible with the presence of
tion of the true lumen and distal blood flow. Stents and
orthopaedic fixation equipment or residual metallic debris
stent-grafts have been used successfully to tack down sepa-
from the offending injury.
rated layers and obliterate the false lumen in extracranial
CA dissections.25
Management
Results The VA contributes less than 10 per cent of total cerebral
blood flow and is anatomically unique in that the VAs join
The outcome of a blunt CA injury depends primarily on to form a single basilar artery. Consequently, a traumatic
the presence or absence of an initial neurological deficit. A unilateral VA occlusion is well tolerated except when the
high index of suspicion leading to early diagnosis facilitates contralateral VA is atretic, hypoplastic or absent, which
early anticoagulation and is associated with improved occurs in about 3.1 per cent of left and 1.8 per cent of right
neurological outcomes.32 With conservative management, VAs (Fig. 35.7). The unique anatomy of the VA permits
most arteries demonstrate uncomplicated healing in traumatic occlusion as a situation which, in most cases,
3 months. Pseudoaneurysms develop in 20–30 per cent of does not demand treatment. Non-occlusive ‘minimal’
patients treated with anticoagulation, emphasising the injuries such as intimal tears or pseudoaneurysms of less
need for careful follow-up and sequential imaging studies. than 5 cm1,24 can be simply observed.
When false aneurysms occur, they can often be managed Intervention is indicated in the following VA injuries:
by endovascular means. those associated with external haemorrhage and/or an
426 Vascular injuries of the neck

Patient with penetrating cervical trauma

Pulsatile haemorrhage/rapidly expanding neck haematoma

Yes No

Operative Cervical spine fracture, vertebrobasilar

exploration symptoms, posterior triangle violation

Yes

Four-vessel cerebral angiogram

No

Symptomatic bilateral or Yes Consider

dominant VA injury operative repair

No

Unilateral Unilateral VA occlusion,

pseudoaneurysm, AVF ‘minimal’ injury

Percutaneous transcatheter intervention

Figure 35.7 Algorithm outlining

Successful Operative ligation Observation the approach to the diagnosis and
Unsuccessful
management of suspected vertebral
Follow-up
artery (VA) injury. AVF, arteriovenous
fistulae

expanding haematoma, an angiographically documented VA, it is necessary to demonstrate angiographically that the
high flow arteriovenous fistula or a false aneurysm greater contralateral VA is of normal calibre and uninjured.
than 5 mm. With the exception of the proximal VA prior to Therapeutic anticoagulation following embolisation should
its entrance into the bony cervical canal, exposure of the be used to limit propagation of distal thrombus.
remainder of the vessel is complicated and potentially haz- Although percutaneous endovascular management of
ardous. The technical difficulties associated with operative VA injuries is possible in most patients, there are limita-
exploration of the VA and the capability of endovascular tions. The small calibre of the VA makes the use of stents or
methods in managing injuries percutaneously at the time covered stents for traumatic pseudoaneurysms difficult.37
of diagnostic angiography, render catheter-based interven- Further, extensive damage to the lumen of the VA follow-
tions preferable. The advantages of endovascular interven- ing high velocity penetrating injuries can make identifica-
tions are so significant that it is wise, when exploring the tion of the true lumen prior to endovascular manipulation
neck for other injuries, to leave a stable, known VA injury challenging.26 Occasionally, distal VA injuries are situated
undisturbed, to be managed by endovascular means in the at branch points making embolisation less attractive. For
postoperative period.33 instance, the need to sacrifice the posterior inferior
Percutaneous transcatheter embolisation with coils or cerebellar artery in order to control an injury may result in
balloons provides adequate proximal and distal control Horner’s syndrome.38
and is sufficient for the treatment of most pseudoa- On rare occasions, patients with acute haemorrhage fol-
neurysms, arteriovenous fistulae or expanding cervical lowing VA injury, possibly transection confirmed on
haematomas. On occasion, occlusion distal to the injury angiography, will require emergency surgical exploration
may necessitate a retrograde approach via the contralateral (see Fig. 35.7). Simple ligation is the preferred treatment in
VA. Even when larger pseudoaneurysms or high flow fistu- these instances. In the largest reported series of VA
lae cannot be completely occluded by embolisation alone, injuries,35 41 of 43 arteries were ligated without any neuro-
this technique does facilitate subsequent formal VA liga- logical sequelae. Both proximal and distal ligation is usu-
tion.33 Prior to percutaneous embolisation of an injured ally required to arrest haemorrhage effectively.35 Primary
 References 427

repair of the VA is almost never necessary, and given the

Key references
technical impediments, could be potentially harmful.
Although vertebral revascularisation is sometimes under- Cogbill TH, Moore EE, Meissner M, et al. The spectrum of blunt
taken in patients with non-traumatic vertebrobasilar insuf- injury to the carotid artery: a multicenter perspective.
ficiency, there are no specific reports of successful surgical J Trauma 1994; 37: 473–9.
repair of a VA injury. Nevertheless, repair may be neces- Fry WR, Dort JA, Smith RS, et al. Duplex scanning replaces
sary in the rare patient with posterior circulation neuro- arteriography and operative exploration in the diagnosis of
logical deficit and bilateral traumatic VA occlusion or potential cervical vascular injury. Am J Surg 1994; 168: 693–5.
occlusion of a dominant VA. For lesions proximal to the Kuehne JP, Weaver FA, Papanicolaou G, Yellin AE. Penetrating
bony canal direct repair may be possible; in most instances trauma of the internal carotid artery. Arch Surg 1996; 131:
however, and particularly in more distal injuries, bypass of 942–7.
Stain SC, Yellin AE, Weaver FA, Pentecost MJ. Selective
the injured segment with interval ligation will be necessary.
management of non-occlusive arterial injuries. Arch Surg
This requires familiarity with the reconstructive tech-
1989; 124: 1136–40.
niques described in respect of the VA.39,40 Weaver FA, Yellin AE, Wagner WH, et al. The role of arterial
The management of blunt vertebral injuries is less reconstruction in penetrating carotid injuries. Arch Surg 1988;
well defined. In a recent, non-randomised study of blunt 123: 1106–11.
VA trauma, Biffl et al.14 found that systemic heparinisation
resulted in improved neurological outcomes in 50 per cent
of patients in comparison with 20 per cent of the non-treated
group. Their results however, did not reach statistical REFERENCES
significance. Despite their recommendation for lower target
anticoagulation, bleeding in this population continues to 1 Demetriades D, Theodorou D, Asensio J, et al. Management
be of significant risk, with two haemorrhagic strokes among options in vertebral artery injuries. Br J Surg 1996; 83: 83–6.
heparinised patients in their study. Anecdotal reports have 2 Martin RF, Eldrup-Jorgensen J, Clark DE, Bredenberg CE. Blunt
also described successful tacking of the dissected layers of trauma to the carotid arteries. J Vasc Surg 1991; 14: 789–95.
VA intima with covered stents.37 3 Weaver FA, Yellin AE, Wagner WH, et al. The role of arterial
reconstruction in penetrating carotid injuries. Arch Surg 1988;
123: 1106–11.
4 Kuehne JP, Weaver FA, Papanicolaou G, Yellin AE. Penetrating
Results trauma of the internal carotid artery. Arch Surg 1996; 131:
942–7.
In a series of 25 penetrating VA injuries, Blickenstaff et al.33 5 Mittal VK, Paulson TJ, Colaiuta E, et al. Carotid artery injuries
reported only two (8 per cent) VA associated neurologi- and their management. J Cardiovasc Surg (Torino) 2000; 41:
cal deficits and no mortality. Demetriades et al.1 reported a 423–31.
single death due to an isolated VA injury (4.5 per cent), 6 Watson WL, Silverstone SM. Ligature of common carotid artery
and no neurological sequelae attributable to VA trauma. in cancer of the head and neck. Ann Surg 1939; 109: 1.
They also observed angiographic evidence of healing at 7 Wylie EJ, Hein MF, Adams JE. Intracranial hemorrhage following
follow-up 1 month after an intimal VA tear; several other surgical revascularization for treatment of acute strokes. J
studies have also reported spontaneous healing of most Neurosurg 1964; 21: 212–15.
small injuries. In contrast, Biffl et al.14 reported a death 8 Cohen A, Brief D, Mathewson C Jr. Carotid artery injuries. Am J
Surg 1970; 120: 210–14.
rate of 8 per cent and a 20 per cent incidence of
9 Bradley EL III. Management of penetrating carotid injuries: an
neurological deficit in patients who had sustained a blunt alternative approach. J Trauma 1973; 13: 248–55.
VA injury. 10 Brown MF, Graham JM, Feliciano DV, et al. Carotid artery
injuries. Am J Surg 1982; 144: 748–53.
11 Matas R. Traumatisms and traumatic aneurysms of the vertebral
Conclusions artery and their surgical treatment, with the report of a cured
case. Ann Surg 1893; 18: 477.
The diagnosis of cervical vascular injuries has evolved 12 Maisonnueve JGF. J Connaissances Med Chir 1852; 2: 181.
13 Serbinenko FA. Balloon catheterization and occlusion of major
considerably over the past two decades. Prompt recogni-
cerebral vessels. J Neurosurg 1974; 41: 125–45.
tion, timely diagnostic evaluation and a tailoring of surgi-
14 Biffl WL, Moore EE, Elliott JP, et al. The devastating potential of
cal and endovascular interventions to the specific blunt vertebral arterial injuries. Ann Surg 2000; 231: 672–81.
vascular injury is the key to minimising adverse events 15 Rivers SP, Patel Y, Delany HM, Veith FJ. Limited role of
and neurological injury. Further improvement in out- arteriography in penetrating neck trauma. J Vasc Surg 1988; 8:
comes will require a concerted multidisciplinary effort by 112–16.
surgeons and interventional radiologists and the appro- 16 Fry WR, Dort JA, Smith RS, et al. Duplex scanning replaces
priate application of evolving endovascular techniques. arteriography and operative exploration in the diagnosis of
potential cervical vascular injury. Am J Surg 1994; 168: 693–5.
428 Vascular injuries of the neck

17 Kuzniec S, Kauffman P, Molnar LJ, et al. Diagnosis of limbs and 29 Cogbill TH, Moore EE, Meissner M, et al. The spectrum of blunt
neck arterial trauma using duplex ultrasonography. Cardiovasc injury to the carotid artery: a multicenter perspective. J Trauma
Surg 1998; 6: 358–66. 1994; 37: 473–9.
18 Ginzburg E, Montalvo B, LeBlang S, et al. The use of duplex 30 Remonda L, Heid O, Schroth G. Carotid artery stenosis, occlusion,
ultrasonography in penetrating neck trauma. Arch Surg 1996; and pseudo-occlusion: first-pass, gadolinium-enhanced, three-
131: 691–3. dimensional MR angiography – preliminary study. Radiology
19 Cogbill TH, Moore EE, Meissner M, et al. The spectrum of blunt 1998; 209: 95–102.
injury to the carotid artery: a multicenter perspective. J Trauma 31 Fabian TC, Patton JH, Croce MA, et al. Blunt carotid injury.
1994; 37: 473–9. Importance of early diagnosis and anticoagulant therapy. Ann
20 Bynoe RP, Miles WS, Bell RM, et al. Noninvasive diagnosis of Surg 1996; 223: 513–22.
vascular trauma by duplex ultrasonography. J Vasc Surg 1991; 32 Biffl WL, Moore EE, Ryu RK, et al. The unrecognized epidemic of
14: 346–52. blunt carotid arterial injuries: early diagnosis improves
21 Meissner M, Paun M, Johansen K. Duplex scanning for arterial neurologic outcome. Ann Surg 1998; 228: 462–70.
trauma. Am J Surg 1991; 161: 552–5. 33 Blickenstaff KL, Weaver FA, Yellin AE, et al. Trends in the
22 Greenwold D, Sessions EG, Haynes JL, et al. Duplex management of traumatic vertebral artery injuries. Am J Surg
ultrasonography in vascular trauma. J Vasc Technol 1991; 15: 1989; 158: 101–15; discussion 105–6.
79–82. 34 Bear HM, Zoarski GH, Rothmann MI. Evaluation of vertebral
23 LeBlang SD, Nunez DB Jr, Rivas LA, et al. Helical computed artery injury from ballistic trauma to the neck. Emerg Radiol
tomographic angiography in penetrating neck trauma. Emerg 1997; 4: 346–8.
Radiol 1997; 4: 200–6. 35 Reid JDS, Weigelt JA. Forty-three cases of vertebral artery
24 Stain SC, Yellin AE, Weaver FA, Pentecost MJ. Selective trauma. J Trauma 1988; 28: 1007–12.
management of non-occlusive arterial injuries. Arch Surg 1989; 36 Mascalchi M, Bianchi MC, Mangiafico S, et al. MRI and MR
124: 1136–40. angiography of vertebral artery dissection. Neuroradiology 1997;
25 Hemphill JC, Gress DR, Halbach V V. Endovascular therapy of 39: 329–40.
traumatic injuries of the intracranial cerebral arteries. Crit Care 37 Waldman DL, Barquist E, Poynton FG, Numaguchi Y. Stent graft
Clin 1999; 15: 811–29. of a traumatic vertebral artery injury: case report. J Trauma
26 Gomez CR, May AK, Terry JB, Tulyapronchote R. Endovascular 1998; 44: 1094–7.
therapy of traumatic injuries of the extracranial cerebral arteries. 38 Golueke P, Sclafani S, Phillips T, et al. Vertebral artery injury –
Crit Care Clin 1999; 15: 789–809. diagnosis and management. J Trauma 1987; 27: 856–65.
27 Robbs JV, Human RR, Rajaruthnam P, et al. Neurological deficit 39 Berguer R, Flynn LM, Kline RA, Caplan L. Surgical reconstruction
and injuries involving the neck arteries. Br J Surg 1983; 70: of the extracranial vertebral artery: management and outcome.
220–2. J Vasc Surg 2000; 31: 9–18.
28 Popowich L. Blunt carotid artery trauma associated with 40 Edwards WH, Mulherin JL Jr. The surgical management of
maxillofacial injuries: report of three cases. J Oral Maxillofac proximal subclavian-vertebral artery stenosis. Contemp Surg
Surg 1984; 42: 462–5. 1980; 17: 11–19.
 36
Abdominal Vascular Injuries

JOHN V ROBBS

The problem 429 Principles of management 432

Aetiology and pathology 429 Operative management 432
Clinical presentation and diagnosis 430 Endovascular techniques 439
Investigations 431 References 440

THE PROBLEM encountered. These may be subdivided into stabs, gunshot,

shotgun and iatrogenic injuries. In general, stab wounds
cause fairly circumscribed trauma. Low velocity gunshot
Discussion on this topic revolves around injuries of the aorta wounds caused by muzzle velocities of less than 600 m/s,
and its major branches namely coeliac, mesenteric, renal generally show no more damage than that found along the
and the iliac arteries and their branches. With regard to missile track. They also tend to be relatively localised, but
venous trauma the inferior vena cava (IVC), the iliac veins the problems lie with tracing the trajectory of the missile.
and the portal system will be discussed. Injuries to the aorta High velocity missile wounds, as encountered in military
and the IVC and their major branches are relatively uncom- situations where muzzle velocities exceed 600 m/s, release
mon, certainly in the context of those admitted to hospital energy causing extensive damage around the track of the
for surgical management. In our own 15-year database of missile; when these wounds traverse the abdomen they are
patients with arterial trauma managed on the vascular serv- usually incompatible with survival. More detail on the bal-
ice, abdominal vessels were only involved in 5 per cent of the listics can be obtained by referring to an appropriate text.2
total of 3000 patients. The most frequently involved arteries Shotgun wounds at close range caused by multiple pel-
were the iliac arteries (2.5 per cent). lets deserve special consideration as they usually penetrate
The mortality for these injuries is high both in the pre- the abdominal wall and cause extensive damage. This is
hospital setting as well as following efforts at treatment. The almost invariably associated with visceral injury with or
reasons for this are difficulty in controlling haemorrhage without contamination by bowel contents, and the vascu-
prior to admission and the relative inaccessibility of the ves- lar injuries tend to be complex and may even be multiple.3
sels in the retroperitoneum making rapid access and control It should be borne in mind that as many of these vessels are
difficult. The problem is often compounded by emergency situated in the retroperitoneum a high incidence of associ-
laparotomies being carried out by relatively inexperienced ated hollow visceral injury can be expected.
surgeons and under these circumstances the situation can be Iatrogenic injuries may complicate interventional catheter
further aggravated by the release of whatever tamponade procedures. Operative misadventure is not an infrequent
exists in attempts to control bleeding with blind clamping, cause of iatrogenic vascular injury. These are beyond the
often causing further vascular damage. This is an extremely scope of this chapter but are discussed in Chapters 38–40.
challenging group of injuries from the technical point of The mechanisms encountered in penetrating trauma
view, often with little reward in terms of patient survival. may involve perforation of the vessel resulting in either
free haemorrhage or the formation of a pseudoaneurysm.
In this context, a through-and-through perforation, often
AETIOLOGY AND PATHOLOGY oblique, can be most challenging, especially when dealing
with the IVC or the aorta. Simultaneous perforation of
The mechanisms and the pathology of vascular injuries artery and adjacent vein produces an arteriovenous fistula
are varied.1 Penetrating wounds are the most frequently with or without an intervening false aneurysm; the most
430 Abdominal vascular injuries

frequently encountered in our practice have been aortocaval initially to volume resuscitation but they may gradually
and common iliac fistulae. A false aneurysm may compress become hypovolaemic once more and require ongoing
adjacent viscera or erode into the bowel or biliary system resuscitation. A more subtle manifestation of vascular injury
with resultant gastrointestinal haemorrhage. Penetrating contained by the retroperitoneum is illustrated by the
parenchymal injuries involving the liver or the kidney may patient who requires extensive volume resuscitation initially,
present with haemobilia or haematuria, respectively. stabilises, but is then found to have abdominal distension
Blunt trauma involving these vessels is relatively uncom- and a low haemoglobin level; this is highly suggestive of a
mon. One type of injury appears to involve a crushing mech- contained major vascular disruption.
anism such as a pedestrian run over by a vehicle which may
result in shearing of the major branches from the aorta. In
LOWER LIMB ISCHAEMIA
our experience we have seen distal aortic thrombosis fol-
lowing this type of injury. The fractured pelvis may damage Crush injury with iliac or aortic occlusion may present
the common iliac vessels when there is disruption of the with lower limb ischaemia. It may remain undetected at the
sacroiliac joint with cephalad displacement of the hemipelvis. time of the primary clinical survey due to hypothermia and
Fractures with disruption of the pelvic ring also frequently shock. This possibility, however, must be considered in any
cause multiple small vessel, predominantly venous, injuries. patient with pelvic fracture or significant abdominal crush
Acceleration/deceleration trauma frequently involves the injury.
renal vessels in which it would appear that the weight of the It is also likely that when intimal tears occur, thrombosis
viscera, with forward momentum, causes marked distrac- may follow within hours of the injury, and unless the patient
tion and applies shearing forces upon the renal pedicle. We is repeatedly assessed this may be missed, with unfortunate
have also seen these injuries involving the IVC at the level consequences.
of the diaphragm, the weight of the liver causing shear
stresses where the IVC traverses the diaphragm.
ANURIA
The basic pathology in all blunt trauma is that crushing
and distracting forces cause the vessel to tear from its lumi- Bilateral renal pedicle injuries are relatively uncommon.
nal surface outwards. Intima is the least elastic and there- A patient who fails to pass urine after sustaining accelera-
fore most vulnerable to tearing. The most elastic and tion/deceleration trauma, and in whom all other possible
durable element of the vessel wall is the adventitia which lower urinary tract injuries have been excluded, may have
may well remain intact although the inner layers are dis- developed bilateral renal artery thrombosis. Unilateral
rupted. Exposure of this extensive thrombogenic surface renal artery disruption may be extremely difficult to detect
may bring about thrombosis with occlusion. Partial or and often remains undiagnosed.1 If some perfusion per-
total transmural disruption may result in perforation and sists through a partially occluded vessel, the patient may
haemorrhage with false aneurysm formation. In the long well have microscopic haematuria. The only means of
term, partial disruption of the vessel wall, with an intact detecting such injuries is to be clinical aware of their possi-
adventitia is likely to lead to a localised aneurysm. These bility and to proceed with basic screening investigations
considerable forces cause a high proportion of associated such as an excretory urogram (EUG).
solid and hollow visceral injuries, further compounding
the difficulties in diagnosis and management.
INTRAOPERATIVE DIAGNOSIS
This occurs when laparotomy has been necessary, for
CLINICAL PRESENTATION AND DIAGNOSIS example for peritoneal irritation, and a retroperitoneal
haematoma is found, which may well mask a significant
vascular injury. Under these circumstances it is important
Clinically, patients can broadly be divided into two cat- to localise the haematoma anatomically, i.e. central, lateral
egories: those in whom the diagnosis is acutely apparent or or pelvic, and to assess whether it is stable, expanding or
those in whom the presentation is subtle, manifesting itself pulsatile.
days, weeks or even years later.1

Acute presentation
Acute presentation
HAEMORRHAGE
Patients may present with an obvious major intra-abdominal
• Haemorrhage

vascular catastrophe with hypovolaemic shock, poorly

• Lower limb ischaemia

responsive to resuscitation and abdominal distension. On

• Anuria

occasion patients with major venous injuries respond well

• Operative diagnosis
 Investigations 431

Delayed presentation This outcome is probably associated with an element of

traumatic pancreatitis due to activation of proteolytic
PULSATILE MASS enzymes. Massive rectal bleeding has also followed the
development of false aneurysms involving the internal iliac
A contained perforation of any of the major arteries often vessels.5
results in pseudoaneurysm formation. This pulsatile mass
compresses adjacent viscera and may manifest itself as
a duodenal obstruction. These pseudoaneurysms require HAEMATURIA
urgent intervention as their natural history is one of even- Penetrating renal parenchymal injury with the formation
tual rupture or erosion into adjacent viscera. of an intrarenal pseudoaneurysm may result in quite severe
intermittent haematuria manifesting itself several days
GASTROINTESTINAL HAEMORRHAGE after the original injury.6
Haemobilia may follow penetrating liver injuries with
partial disruption of the hepatic artery branches and the ARTERIOVENOUS FISTULA
formation of a fistula into the biliary radicles. This may Aortocaval fistula usually presents fairly acutely with the
present days to weeks after the injury and is characterised appearance of lower limb oedema and an obvious arteriove-
by right upper quadrant pain and episodes of upper gastroin- nous bruit in the abdomen. Small fistulae may remain unde-
testinal haemorrhage, usually melaena, but less frequently tected until the patient presents months, or even years, later
haematemesis.4 Injury to the peripancreatic vessels may in congestive heart failure caused by a hyperdynamic circu-
result in erosion into the duodenum of a pseudoaneurysm lation and marked lower extremity venous hypertension.
with intermittent massive gastrointestinal haemorrhage. Iliac fistulae present mainly with signs of lower extremity
venous hypertension coupled with an arteriovenous bruit.3
Figure 36.1 shows the angiogram of a 59-year man who
developed an aortocaval fistula following a gunshot injury.

Delayed presentation

• Pulsatile mass
• Gastrointestinal haemorrhage
• Haematuria
• Arteriovenous fistula

INVESTIGATIONS

Angiography

Selective retrograde femoral angiography is the gold stan-

dard of investigation. It should be reserved, however, for
stable patients in whom vascular injury is strongly sus-
pected on clinical grounds: that group comprises patients
with a pulsatile mass, arteriovenous murmur, significant
haematuria, gastrointestinal haemorrhage or an ischaemic
limb. Angiography should also be performed for patients
with stable retroperitoneal haematomas as many of these
lesions can be dealt with by interventional techniques.
There is no place for angiography in the unstable patient
who is actively bleeding.

Indications for angiography in a

haemodynamically stable patient
Figure 36.1 Retrograde femoral angiogram of a 59-year-old
man showing an aortocaval fistula following a low velocity
• Ischaemic lower limb(s)
gunshot wound to the abdomen • Pulsatile mass
432 Abdominal vascular injuries

• Arteriovenous murmur
• Suspected haemobilia
– Right upper quadrant pain
– Jaundice
– Upper gastrointestinal bleeding
• Suspected renal pedicle injury
(acceleration/deceleration)
– Anuria
– Delayed/absent function on excretory urography
– Retroperitoneal haematoma (stable)
Figure 36.2 Computed tomography (CT) angiogram showing a
– Lateral (perinephric)
pseudoaneurysm arising from the juxtarenal aorta in an 18-year-
– Pelvic old man who sustained a gunshot wound to the abdomen. At the
• Gastrointestinal bleeding time of initial wounding he underwent laparotomy at which
– Eroding false aneurysm? several small bowel perforations were repaired. He presented
1 month later with a palpable pseudoaneurysm

Excretory urography airway and insertion of at least two wide bore intravenous
cannulae (see Chapter 7B) with monitoring of progress of
The EUG is a useful screening test in patients in whom a resuscitation by means of urine output, central venous
renal injury is suspected, such as those with haematuria or pressure, systemic blood pressure, core temperature and
anuria or those with the appropriate pattern of injury in acid–base status. Those patients with obvious continuing
whom renal injury is likely. Delayed or non-excretion of haemorrhage, or in whom there are other indications for
contrast should provoke selective renal angiography. urgent exploration such as evisceration or peritonitis,
demand urgent operation. In the latter cases occult vascu-
lar injury may be present in the form of a contained retro-
Duplex scan peritoneal haematoma only discovered at operation.
In the second category, that is, those who have been sta-
This investigation has limited value in an acute presenta- bilised but in whom a vascular injury is strongly suspected
tion but may be useful in detecting the site of an iliac arte- on clinical grounds, further investigations are indicated.
riovenous fistula, or under certain circumstances and in The most useful of these is angiography which accurately
the appropriate patient, in assessing blood flow in the renal localises the injury and offers the possibility of some form
artery in someone whose abdomen is not grossly distended of endovascular intervention in the appropriate situation. An
or who is relatively thin. Published experience confirms the algorithm of management (Fig. 36.3) can be of assistance.
usefulness of this investigation in the evaluation of neck
trauma.7
OPERATIVE MANAGEMENT
Computed tomography
Access and control
Computed tomography (CT) and CT angiography are use-
ful in selected situations in which there is localised injury Once operative management has been decided upon, wide
requiring further elucidation.4 In Fig. 36.2, CT angio- exposure is absolutely of the essence in dealing with these
graphy shows a pseudoaneurysm of the juxtarenal aorta fol- injuries. The patient should be prepared and draped from
lowing a gunshot injury to the abdomen of an 18-year-old the nipple line to the knees. This gives access to the chest
man who presented 3 months after the injury. The aortic and to the groin vessels if necessary.
injury had not been detected at his initial laparotomy when The incision of choice is in the midline extending from
small bowel perforations had been repaired. the xiphisternum to the pubis (Fig. 36.4) Access may be
improved by lateral extension into the loin on either side as
required. On occasion a lower sixth or seventh intercostal
space thoracotomy or even median sternotomy may be
PRINCIPLES OF MANAGEMENT required. The surgeon should not hesitate to extend the inci-
sions to improve access; ‘keyhole’ surgery invites disaster.
The text has alluded to the spectrum of clinical presentations. A useful manoeuvre is to obtain control of the aorta at
In the unstable shocked patient the standard approach to the hiatus. The oesophagus is mobilised and retracted to
resuscitation must be followed, namely, maintenance of the patient’s left. The liver can then be retracted cephalad
 Operative management 433

Management by means of a Deaver retractor, and the aorta can be pal-

pated (Fig. 36.5a). The aorta can be compressed against the
Resuscitate spinal column either by manual pressure or with a ‘sponge
stick’, but preferably by clamping, to facilitate which the
right crus of the diaphragm overlying the aorta may be
Continuing haemorrhage Stabilised divided. The aorta is then straddled using the middle and
index finger and a straight aortic clamp can then be applied
Suspected
using the fingers as a guide (Fig. 36.5b). The aorta should
Suspected
visceral injury vascular injury
not be encircled and snared as that may risk avulsing the
lower intercostal vessels.
On opening the lesser sac, access can be obtained to the
Retroperitoneal
Laparotomy
haematoma
Angiography coeliac and superior mesenteric vessels. If more extensive
exposure of the aorta in this region is required a left vis-
ceral rotation (Fig. 36.6) can be performed by mobilising
Central Stable the colon, the spleen and the pancreas to the patient’s
Unstable (any area) (lateral) right in a plane anterior to the left kidney. This manoeuvre
(expanding) (pelvic)
(bleeding)
also provides optimal access to the left renal artery.
The infrarenal aorta is readily exposed through the root of
Laparotomy or the mesentery and the common iliac vessels by opening the
Explore endovascular procedure peritoneum over the pelvic brim.
The IVC below the liver is best exposed by right visceral
Figure 36.3 Algorithm of management in patients suspected of rotation in which the ascending colon is mobilised, along
having an abdominal vascular injury

Figure 36.6 Left and right visceral rotation procedures to gain

Figure 36.4 Incisions for access to abdominal vascular injuries access to major retroperitoneal vessels

Figure 36.5 (a) Exposure for

control of the aorta at the
diaphragmatic hiatus and
(b) control of aorta at the
(a) (b) hiatus
434 Abdominal vascular injuries

with the duodenum and pancreatic head by the Kocher blood transfusion (more than 10 units), and in whom the
manoeuvre, and reflected to the patient’s left. This expo- operation has taken more than one and a half hours.12
sure also provides optimal access to the right renal Principles of damage control involve the arrest of bleed-
vessels. ing and limiting contamination. In terms of surgical bleed-
The visceral rotation procedures are illustrated in ing non-essential vessels are ligated, and simple suture
Fig. 36.6. Access to the retrohepatic IVC can be gained by repair carried out where possible. Consideration can be
dividing the suspensory ligaments of the liver and reflecting given to balloon tamponade and the use of temporary
the liver to the patient’s left. Details are discussed under the shunts for major more essential arteries. For venous or
appropriate heading below. non-surgical bleeding packing should be entertained. All
ends may be ligated or stapled and the abdominal wounds
closed temporarily. Various strategies have been advocated
for this including the use of towel clips to approximate skin
Specific vascular scenarios edges and meshes or transparent adhesive abdominal
drapes to cover exposed bowel.13
ACTIVE ONGOING HAEMORRHAGE
In the ICU setting active re-warming is commenced using
The major problem under these circumstances is that the warmed fluids. Coagulopathy is treated, volume restored
patient is usually extremely unstable, and it is often diffi- and acidosis corrected. Bleeding not thought to be due to
cult to localise the source of haemorrhage. The initial step coagulopathy may necessitate earlier re-exploration but, if
is to ‘eviscerate’ the abdomen, taking the small bowel out planned, it should take place between 24 and 72 hours
of the abdominal cavity and allowing it to lie on the later. Consideration can be given to the operation once
abdominal wall. It is advisable to use swabs to mop up core temperature exceeds 35 °C, serum lactate is less than
blood and to restrict use of the sucker; the patient’s blood 2.5 mmol/L and the international normalised ratio (INR)
volume may rapidly disappear into the sucker reservoir and prothrombin time (PTT) are less than 1.25 times
without the surgeon fully appreciating the extent of the normal.12,14,15
loss, particularly with low pressure venous bleeding. The major consideration in the postoperative period,
It is also extremely important to keep pace with volume besides the complex metabolic derangements that may
resuscitation. If possible, tamponade the abdominal cavity continue to occur, is that of the abdominal compartment
and resuscitate before mobilising or exploring haematoma syndrome.16–18 This is the situation in which increasing
or bleeding sites. If the bleeding is obviously arterial, aortic abdominal pressures due to distended bowel, or for what-
control at the hiatus is obtained as described. This will ever reason, results in progressive oliguria, elevated airway
usually slow the bleeding down sufficiently to localise its pressure, hypoxemia and hypotension due to decreased
source and appropriate exposure can then be undertaken. venous return. It is important to be aware of this problem
Once the injury has been isolated, more localised control is after all major abdominal injuries. The optimum method
obtained if possible, for example, with side-biting clamps of measuring intra-abdominal pressure is by means of an
restoring visceral and renal perfusion while resuscitation intravesical catheter.19
is completed. Each lesion can then be dealt with on its spe- It would appear that a pressure in excess of 22 mmHg
cific merits. If the bleeding is obviously venous and welling necessitates decompression. Abdominal wounds should be
up into the operative field, it may be extremely difficult to reopened and one of the strategies of visceral containment
identify the source by virtue of its low pressure. The IVC is embarked upon.12 Detailed discussion of this subject is
best exposed by right visceral rotation, or in the case of the beyond the scope of this chapter except to note that 20 per
retrohepatic IVC by mobilisation of the right lobe of the cent or more patients in this category will develop this
liver. This will be dealt with more specifically in discussion complication, which, unless remedied early, will certainly
on IVC injuries. contribute to rapid deterioration.
The major judgement call in this type of patient is The damage control strategy is extremely labour inten-
whether or not to embark upon a ‘damage control’ strat- sive and it puts great stress on resource management.20 The
egy.8 These patients are more likely to die from ‘metabolic mortality remains in excess of 50 per cent but it is difficult
failure’ than from the actual vascular injury (see Chapter to compare various series due to the heterogeneous nature
4). The balance lies between restoring normal anatomy and of the patients and their injuries. The philosophy, in essence,
the fatal triad of hypothermia, acidosis and coagulopathy. is that some chance is better than no chance.
The strategy comprises temporising manoeuvres allowing
for active resuscitation in an intensive care unit (ICU) set-
RETROPERITONEAL HAEMATOMAS
ting with subsequent re-exploration for definitive treat-
ment.9–11 Consideration must be given to adopting this Retroperitoneal haematomas are classified into three
strategy in a patient who has a metabolic acidosis (pH less zones according to their site (Fig. 36.7), namely central
than 7.25), hypothermia (core temperature less than haematomas (zone 1), lateral haematomas (zone 2) and
34 °C), coagulopathy (non-surgical bleeding) and massive pelvic haematomas (zone 3). Central haematomas may
 Operative management 435

nephrectomy.22,26–29 When exploration is indicated, how-

ever, the renal pedicle must be controlled before opening
the haematoma. Renal artery injuries are repaired on their
Central merits.
(zone 1)
Haematomas in the medial portion of zone 2, where the
renal pelvis, renal pedicle and lateral duodenum are situ-
ated, can be problematic. Much debate has taken place as to
whether these should be explored in stable patients by virtue
Lateral Lateral of the morbidity caused by missed injuries, particularly to
(zone 2) the duodenum, and a case for routine exploration can cer-
tainly be made. These particular haematomas should prob-
ably be classified with the central zone 1 group.30
With regard to zone 3 pelvic haematomas in general,
management depends on whether the injury is actively
bleeding or contained. A contained haematoma, if non-
Pelvic
expansile and flaccid, should not be opened in the presence
(zone 3)
of a pelvic fracture, when arterial pulsation in the groin is
intact and there is no evidence of urological trauma on
urethrography and cystography. Exploration is indicated,
Figure 36.7 Retroperitoneal haematoma zones however, if a pelvic haematoma has ruptured, if it is rapidly
expanding or pulsatile in close proximity to major vessels,
if arterial pulsation in the groin is absent or if intra-
result from injuries to the great vessels of the abdomen, peritoneal rupture of the bladder has occurred.30–32
pancreas and duodenum; lateral haematomas result from An essential part of controlling haemorrhage in relation
injured kidney, renal vessels and ureter, and pelvic haemato- to pelvic fracture is urgent fixation and stabilisation of the
mas result from damage to the great vessels of the pelvis, fracture.30,31 Deep pelvic haematomas in relation to the
bladder, ureters, and the rectum. Combination haematomas internal iliac vessels and their branches even if pulsatile are
are classified as zone 4.21–23 probably best left undisturbed or treated by tamponade
All zone 1 haematomas should be explored because until the precise source of bleeding is elucidated by angiog-
of the high incidence of visceral and major vascular raphy. Most of these injuries are amenable to endovascular
injuries.23–25 The haematoma should not be entered with- procedures, embolotherapy in particular.33,34 Details of the
out attempting to resuscitate the patient as far as possible management of rectal and urological injuries are beyond
by controlling the bleeding area with pressure packs and the scope of this chapter.
volume replacement. This is often easier said than done.
If the haematoma or the bleeding site involves the
MAJOR ARTERIAL INJURIES
suprarenal area the aorta should be controlled at the
oesophageal hiatus as described. The next step will be to The options available for definitive therapy are lateral
open the lesser sac and to attempt to isolate the injury. If suture, patch angioplasty repair, end-to-end anastomosis
this does not prove possible left lateral visceral rotation or interposition grafts. Under certain circumstances liga-
should be used. If the haematoma is predominantly tion may be necessary. Repair follows well established prin-
infrarenal, or if it is situated at the base of the mesentery, the ciples35 such as debridement of the vessel ends to healthy
transverse colon should be retracted superiorly after mobil- vessel wall, removal of propagated thrombus and flushing
ising the duodenum, so that the aorta can be controlled of the distal arterial tree with heparinised saline solution.
below the left renal vein before entering the haematoma.26 Repairs must be performed with intima to intima appos-
An essential principle in managing these injuries is to avoid ition without causing luminal narrowing or any tension on
blind clamping as the IVC, left renal vein and in particular the suture line. In general, injuries compatible with sur-
the common iliac veins are fragile and very susceptible to vival can usually be repaired by simple techniques. As a
injury. Lacerations of these large vessel trunks can turn a general rule, through and through injuries are best managed
controlled situation into a hopeless blood bath. by transection of the vessel, incorporating both wounds, fol-
If a lateral zone 2 haematoma is expanding or pulsatile, lowed by reanastomosis in an end-to-end fashion.
exploration is indicated. The same would apply when there When an interposition graft or a patch closure is required
is radiological evidence of major pedicle or parenchymal autogenous material such as the long saphenous vein is prob-
injury and significant urinary extravasation. The reason for ably the best option. This applies particularly when there has
not exploring stable non-pulsatile haematomas is that been an element of contamination in the operative field.
haemorrhage adequately tamponaded by perinephric fas- Nevertheless, when there is no contamination, and in the
cia may become uncontrollable and result in unnecessary interests of expedience, prosthetic grafts may be required.
436 Abdominal vascular injuries

Mesenteric injuries is well tolerated in the presence of a patent superior mesen-

With regard to the superior mesenteric artery, definitive man- teric artery.38
agement depends largely upon the anatomical zone involved. Any mesenteric or coeliac artery disruption is poten-
Fullen and his colleagues36 classified these as follows: tially likely to result in bowel ischaemia in the postoperative
period (see Chapters 3, 4 and 29). There is some debate as
• Zone 1 – extending from the aorta to the first major to whether a second look laparotomy should be performed
branch, the inferior pancreaticoduodenal artery as routine in these patients, but no firm guidelines exist.
• Zone 2 – from the inferior pancreaticoduodenal to the
middle colic artery origin Renal artery injuries
• Zone 3 – the trunk distal to the middle colic The basic principles with regard to restoration of arterial
• Zone 4 – segmental branches to the jejunum, ileum or continuity apply. In most cases the optimum procedure is
colon. to use a saphenous graft which takes its origin from the
infrarenal aorta and is anastomosed distally end-to-side or
The chances of bowel ischaemia are maximal for zone 1, end-to-end to healthy renal artery. Unless there has been
minimal for zone 3 and extremely unlikely for zone 4 dis- partial disruption of the vessel with preservation of some
ruptions. With ligation of the proximal superior mesen- prograde flow the results of renal pedicle avulsion in terms
teric artery, collaterals may well occur through the coeliac of kidney salvage are extremely poor, mainly due to the
axis and inferior mesenteric artery branches. The patients, delay in making the diagnosis and the poor tolerance of the
however, are invariably shocked and the severe associated kidney to prolonged ischaemia.28,29,39
vasospasm and hypotension can further perpetuate bor-
derline ischaemia. Ligation of zones 1 and 2 injuries have
MAJOR VENOUS INJURIES
an extremely high mortality and it is strongly recom-
mended that repair be effected for these. It is probably also Inferior vena cava injuries
advisable to repair zone 3 injuries. For injuries close to the With simple lower IVC injuries avulsion of lumbar veins
aorta it is probably best to oversew the aortic defect and to should be avoided by resisting attempts to encircle and
restore continuity by means of a graft taken from the side clamp the IVC. It is preferable to obtain control by pres-
of the infrarenal aorta.37,38 sure above and below the laceration using sponge sticks
Clear guidelines are lacking for the treatment of the and by digital compression of the lumbar veins against
associated superior mesenteric venous injury. Such the posterior abdominal wall26,40 (Fig. 36.8a). For larger
injuries do not seem to influence overall mortality and, in injuries indwelling Foley balloon catheters may be used to
general, a simple injury can be oversewn or reanasto- control bleeding.
mosed, whereas a more complex injury with a wall defect For simple anterior puncture wounds a side-biting
would probably be best treated by ligation. atraumatic vascular clamp is applied to provide adequate
control (Fig. 36.8b). The possibility of an additional poster-
Coeliac axis injuries ior puncture wound due to a through and through injury
There is little information specifically concerning injuries should be excluded. Under these circumstances, the poster-
to this vessel. Our own experience has been to restore ior injury is best sutured through an enlarged anterior wound.
continuity of this vessel by simple suture wherever possible, Attempts to rotate the IVC should be avoided as it invari-
and if not to simply ligate it. It would appear that ligation ably results in avulsion of the lumbar veins (Fig. 36.8c).

(a) (b) (c)

Figure 36.8 Infrarenal inferior vena caval injuries: (a) control; (b) side-biting clamp control; (c) transluminal repair of posterior wall
laceration
 Operative management 437

In the presence of complex injuries which may require IVC is mobilised by right visceral rotation as well as by
interposition grafts, or if there is uncontrollable haemor- mobilising the duodenum using the Kocher manoeuvre.
rhage, ligation is the best option and is extremely well tol- The suprarenal IVC is isolated, the auricle of the atrium
erated by most patients. Grafts of the IVC have given opened and a 36 Fr chest drain tube is passed into the IVC.
uniformly poor results in terms of patency.40,41 Prior to insertion of this tube a side-hole is cut into the
proximal end of the tube through which drainage can
Juxtahepatic inferior vena cava injuries occur. The position of the tube is confirmed by palpation.
This refers to the IVC extending from above the renal veins The IVC is secured around the tube by means of snares.42
under the liver to the diaphragm, and by convention, An alternative that has been described is the use of an
includes injuries to the hepatic veins. The only tributaries endotracheal tube passed from the atrium, inflating its bal-
entering this segment of the IVC are the hepatic veins, the loon in the infrahepatic segment of the IVC so obviating
right adrenal vein and the inferior phrenic veins. Therefore, the need to mobilise it at this level.43 The technique for
by clamping the portal vein and the hepatic artery, and con- insertion of these shunts is difficult and has the potential
trolling the IVC above and below the liver, hepatic flow is for further disruption of the injured area by the tip of the
effectively arrested and there should be minimal bleeding. tube. Unless the drainage holes are correctly positioned the
Obviously, occlusion of the IVC at this level seriously shunt will have no influence whatever on bleeding. Median
compromises venous return to the heart and is doubly sternotomy is also time consuming and there is major
challenging in the hypovolaemic patient. This problem led potential for initiating further bleeding.42
to the development of the concept of the atriocaval shunt, The alternative to using the atriocaval shunt is total
in which an attempt is made to isolate the retrohepatic occlusion of the IVC and portal system.40,44,45 As already
IVC while allowing venous drainage to continue into the mentioned this severely reduces venous drainage and can
atrium by means of strategically placed apertures within cause marked hypotension. Before attempting it, therefore,
the shunt, once the IVC has been isolated42 (Fig. 36.9). the patient must be well resuscitated in terms of volume.
The first step entails performing the Pringle manoeuvre, After initiating the Pringle manoeuvre the ligamentum
i.e. cross-clamping at the porta hepatis. A total median arteriosum and the falciform ligament are divided and
sternotomy must then be carried out, followed by a pericar- the liver is gently pull in a caudad direction. The central
diotomy, isolating the intrapericardial IVC. The juxtahepatic tendon of the diaphragm is then divided to gain access to
the pericardium, which exposes the intrapericardial IVC
(Fig. 36.10a). The infrahepatic IVC is then exposed as
described earlier. The liver is then totally mobilised by
dividing the right and left triangular ligaments and the
anterior and posterior layers of the coronary ligament on
the right. The IVC can now be clamped within the peri-
cardium and below the liver. If the blood pressure drops
precipitously the aorta can be clamped at the hiatus in
order to sequestrate the upper body circulation. The liver
can now be rotated to identify the venous injury, which, in
most cases can be repaired expeditiously; again most
injuries compatible with survival can be dealt with by sim-
ple suture (Fig. 36.10b).
The mortality for these injuries is prohibitive and in
most reported series exceeds 80 per cent. No patients
requiring resuscitative thoracotomy have been reported as
survivors.42 In most experienced hands only about 50 per
cent of patients have survived shunt insertion in order to
allow repair of the injury. In our own experience, we have
attempted atriocaval shunt placement in four patients and
have had no survivors; certainly two died due to haemor-
rhage provoked by attempts to achieve this objective. We
have had 10 patients in whom total isolation was per-
formed, of whom seven survived; all these had penetrating
trauma with simple lacerations.40 All patients had some
degree of postoperative liver function derangement which
resolved within a week or so. The deaths were a direct result
Figure 36.9 Juxtahepatic inferior vena caval injuries. Use of the of blood loss and the cascade of multiple organ failure40,44
atriocaval shunt (see Chapter 4). It is extremely difficult to compare results
438 Abdominal vascular injuries

between reported series because they tend to be small and be tested. As always there is no place for blind application
represent an extremely heterogeneous group of patients. of clamps. Wherever possible one should attempt to con-
trol bleeding by compression.
Portal vein injuries As a general rule in terms of definitive repair, simple
These injuries are seldom if ever isolated and they are lateral suture, or if possible end-to-end anastomosis, is
surgically inaccessible. A high proportion of patients with the most feasible. Complex injuries with major vessel wall
these injuries exsanguinate intraoperatively.45–48 defects are best treated by ligation.49–52 Emergency portal–
Access to the portal vein is achieved by means of right systemic shunt insertion cannot be recommended as it is a
visceral rotation with mobilisation of the duodenum and complex procedure invariably performed in an unstable
head of pancreas to the right. Occasionally, in order to reach patient. In addition, the reported rate of encephalopathy
the confluence of the splenic and superior mesenteric vein, is unacceptably high53 (see Chapter 46). By the same token
it may be necessary to divide the neck of the pancreas.45 various other procedures have been reported, such as the
The major problem is to control bleeding while exposure is use of interposition grafts using either vein or prosthesis,
being established and to this end ingenuity may certainly with the occasional survivor.9
Results are uniformly poor in terms of survival and
50–80 per cent mortality is reported.45–53 The major influ-
ences on this outcome are the associated injuries and mas-
sive blood loss. In the longer term, there is little
information on the patency of attempted repairs. One
small study using duplex Doppler showed that long term
patency can be achieved following simple venorrhaphy.48
During the postoperative period there is a relatively small
risk of bowel infarction (see Chapters 3 and 29). More fre-
quently reported, however, is the massive sequestration of
fluid in the splanchnic bed, particularly after ligation,
which calls for significant fluid resuscitation. In the longer
term there are no reports related to the incidence of symp-
tomatic portal hypertension. Some authors have advocated
a second look procedure within 24 hours to exclude bowel
infarction.45,49,51,52 Our own experience is limited to 15
(a) patients for whom we have records: the mortality was 80
per cent and most of the injured vessels in survivors had
been ligated. With regard to superior mesenteric vein
injuries no specific reports can be found and the approach
advocated is that of simple repair when feasible, and failing
that, ligation.

Renal vein injuries

It is well established that the left renal vein can be ligated
provided that collaterals through the suprarenal and
gonadal vein are preserved. Whenever possible simple
repair should be attempted. The right kidney does not tol-
erate ligation of the right renal vein, as the only collaterals
are those via the renal capsule, the gonadal and adrenal
veins draining directly into the IVC.

Iliac vein injuries

Diffuse bleeding related to pelvic fractures is probably best
treated by packing followed by reoperation for removal of
the packs 24 hours later. Common or external iliac vein
injuries should be repaired wherever possible but if these
(b)
are complex, ligation is the best option, and in our experi-
Figure 36.10 Juxtahepatic inferior vena caval injuries – total ence tolerated well in the short term. Information regard-
hepatic isolation. (a) Exposure of intrapericardial inferior vena ing lower limb venous repairs in the civilian setting indicate
cava and (b) isolation of retrohepatic inferior vena cava with a high incidence of occlusion, particularly after complex
mobilisation of right lobe of liver repairs, but they did not influence limb salvage.54–57 We
 Endovascular techniques 439

have no information of results in the longer term and there ingenuity may be tested in these complex circumstances.
are few, if any, studies in the literature. Ultimately in-line reconstruction is desirable, especially in
On occasion, access to a bleeding common iliac vein young patients, but revision surgery should not be consid-
may be difficult due to the overlying arteries. Under these ered before an arbitrary period of about 6 months has
circumstances the overlying iliac artery may be transected, elapsed to ensure that all sepsis has resolved.
and once the venous injury has been attended to, reanasto-
mosed. This action is preferable to blind groping with
clamps and compounding the situation by further lacerat-
ENDOVASCULAR TECHNIQUES
ing the fragile iliac venous structures.

Interventional radiological procedures provide an attract-

Important specific vascular scenarios ive option in the stable patient with trauma involving
inaccessible vessels. The value of embolotherapy has been
• Active continuing haemorrhage demonstrated in non-essential vessels of the neck, pelvis
• Retroperitoneal haematomas and the extremities using thrombin pellets and spring coils.
• Mesenteric vessel injuries No major morbidity has been reported to date, although the
• Juxtahepatic inferior vena cava injuries literature on this subject is sparse.34,59–62 Potential hazards
• Portal vein injuries with placement of occlusive devices close to the origins of
• Concomitant arterial and colonic injuries tributaries from major parent vessels must be kept in mind,
for instance, migration and distal embolisation of the
occluding device.34 With regard to embolotherapy for arte-
riovenous fistulae an essential principle is that both the
ABDOMINAL ARTERIAL INJURY IN THE PRESENCE
arterial inflow and outflow component must be occluded.
OF CONTAMINATION
Failure to do this almost invariably results in recurrence of
This scenario is applicable, for example, to the presence of the fistula. With larger vessels there is also the danger of
sigmoid colon injury coupled with lower aortic or iliac migration and embolisation into the venous system through
artery disruption. In the presence of gross faecal contami- the fistula itself.34,59
nation it would be unwise to perform arterial repair and in A specific indication in which embolotherapy has come
particular interposition grafting (see Chapters 18 and 25). into its own is in the treatment of traumatic haemo-
Autogenous tissue tends to disintegrate once infected. bilia.4,63–67 This usually follows penetrating trauma with
Various statements regarding the ‘resistance’ to infection the formation of pseudoaneurysms, often quite sizeable,
of certain prosthetic grafts is certainly not borne out by our on the branches of the hepatic artery which form fistulae
own experience.1,58 with bile duct radicles.67,68 In a recent series,63 23 patients
In this group of patients it is advisable to ligate the with traumatic lesions were successfully treated using
major vessels within the area of contamination and to catheter occlusive techniques. There was one death not
restore circulation by means of an extra-anatomical specifically related to the embolotherapy and the long term
bypass. For example, in the presence of a contaminated results on these patients have been excellent without recur-
aortic disruption axillobifemoral bypass may be enter- rence of the problems. Liver function was not compromi-
tained. Similarly, in the presence of an iliac injury, a sed.63 In Fig. 36.11 the results of successful embolisation
femoro-femoral crossover bypass may be used. Once again for traumatic haemobilia are shown. The patient was a

Figure 36.11 Angiograms before and

after embolotherapy to hepatic artery
branches in a 29-year-old man
presenting with haemobilia in the
form of acute episodes of upper
gastrointestinal bleeding 1 week after
a gunshot wound
440 Abdominal vascular injuries

29-year-old man who had sustained gunshot trauma to the

fatality is attributable to torrential bleeding and delays in
liver and presented 2 weeks later with a gastrointestinal
securing control due to the relative inaccessibility of some
bleed. Other peripancreatic pseudoaneurysms with fistula-
of the main arterial and venous trunks. Continuing haem-
tion into the duodenum have been successfully treated by
orrhage and associated organ injury, including contami-
embolic occlusion.69
nation from perforated bowel, represent a major challenge
Renal arterial trauma is often associated with intra-
which demands timely and experienced intervention.
parenchymal pseudoaneurysms which present with signifi-
In the stable patient a CT scan, if necessary comple-
cant haematuria and lend themselves well to selective
mented by angiography, will assist in planning surgical
segmental arterial occlusion. Although segmental infarction
management and in considering the likely value of
of the kidney occurs it does not have long term sequelae in
endovascular therapies.
terms of hypertension.27 Deep-seated pulsatile haematomas
Wide exposure is crucial to gaining adequate access
associated with pelvic fractures usually arise from the dis-
both for control and in executing the necessary repairs, a
ruption of branches of internal iliac or median sacral vessels
principle of special relevance when a retroperitoneal,
and are best dealt with by selective occlusion.1,59,62
possibly pulsatile, haematoma is revealed. In cases of
The role of stent placement in trauma is not well docu-
severe trauma requiring prolonged surgery, it may be
mented and, in general, experience is limited to small
wiser to aim for ‘damage control’ as a first step followed
series.70–76 In relation to penetrating trauma, most reports
by re-exploration and definitive vascular repair when a
deal with lesions involving the subclavian and carotid ves-
degree of stability has been achieved in the ICU. Injuries
sels for which covered stents were used.73,76–78 Of particu-
to the juxtahepatic IVC, mesenteric vessels and the por-
lar interest is their successful deployment in cases of
tal venous system require specific operative manoeuvres
arteriovenous fistula.76
and solutions.
Few reports exist of blunt trauma of the abdominal ves-
sels. Self-expanding stents have been used successfully in
dissections of the abdominal aorta and iliac vessels70,77–79
(see Chapter 24). This modality of treatment, in selected
cases, represents a very attractive alternative to open sur-
Key references
gery. Medium term results with stents are good and no
Burch JM, Oritz V, Richardson RJ, et al. Abbreviated laparotomy and
major acute complications have been reported. Follow-up planned reoperation for critically ill patients. Ann Surg 1992;
ranged from 3 months to 4 years but long term evaluation is 215: 476.
essential. Ghimenton F, Thomson SR, Muckart DJJ, Burrows R. Abdominal
There are very few trauma centres in the world where- content containment: practicalities and outcome. Br J Surg
from meaningful experience with major abdominal vascu- 2000; 87: 106–9.
lar injuries can be reported. Much of this information is Hirsberg A, Walden R. Damage control for abdominal trauma. Surg
subjective and relates to experience gained retrospectively Clin North Am 1997; 77: 813–20.
or with relatively few patients. Much of what is produced Parodi JC, Schonolz C, Ferreira LM, Bergan J. Endovascular treatment
in this chapter is based on personal experience within the of traumatic arterial lesions. Ann Vasc Surg 1999: 13: 121–9.
Robbs JV, Costa M. Injuries to the great veins of the abdomen.
vascular service of the University of Natal Hospitals, South
S Afr J Surg 1984; 22: 223–8.
Africa. In general, the results remain poor for two main
reasons: massive haemorrhage that is difficult to control
and the high incidence of associated visceral injuries.
Comparative data are difficult to obtain by virtue of the REFERENCES
heterogeneous nature of the injuries in these patients,
although various trauma scoring systems may assist in 1 Robbs JV, Baker LW. Cardiovascular Trauma. Current Problems in
producing some evidence-based data.37,38 Surgery. Chicago, IL: Year Book Medical Publishers, Inc, 1984:
These injuries are probably the most challenging in XX1(4).
trauma surgery. 2 Levien LL. Ballistics of bullet injury. In: Dudley H, Carter D and
Russell RCG (eds). Rob & Smith Operative Surgery. Trauma
Surgery Part 1. 4th edn. London: Butterworths.
3 Robbs JV, Carrim AA, Kadwa AM, Mars M. Traumatic
Conclusions arteriovenous fistula: Experience with 202 patients. Br J Surg
1994; 81: 1296–9.
Blunt abdominal vascular injuries are rather uncommon 4 Moodley J, Singh B, Lalloo S, et al. Non-operative management
and largely caused by sudden deceleration and the avulsive of haemobilia. Br J Surg 2001; 88: 1073–6.
and shearing forces involved in road traffic and other acci- 5 Mokoena T, Robbs JV. Surgical management of mycotic
dents. Penetrating abdominal vessel injuries inflicted by aneurysms. S Afr J Surg 1991; 29: 103–7.
6 Angorn IB. Segmental dearterialisation in penetrating renal
guns and knives occur much more frequently and early
trauma. Br J Surg 1977; 64: 59.
 References 441

7 Corr P, Abdool Carrim ATO, Robbs JV. Colour flow ultrasound in 32 Brotman S, Soderstorm CA, Oster-Granit M, et al. Management
the detection of penetrating vascular injuries of the neck. S Afr of severe bleeding in fractures of the pelvis. Surg Gynecol Obstet
Med J 1999; 89: 644–6. 1981; 153: 823–6.
8 Burch JM, Oritz V, Richardson RJ, et al. Abbreviated laparotomy 33 Hirsberg A, Walden R. Damage control for abdominal trauma.
and planned reoperation for critically ill patients. Ann Surg 1992; Surg Clin North Am 1997; 77: 813–20.
215: 476. 34 Naidoo NM, Corr PD, Robbs JV. Angiographic embolisation in
9 Moore EE. Staged laparotomy for the hypothermia, acidosis, and arterial trauma. Eur J Vasc Endovasc Surg 2000; 19: 77–81.
coagulopathy syndrome. Am J Surg 1996; 172:405. 35 Robbs JV. Vascular trauma – general principles of surgical
10 Abramson D, Scalea T, Hitchcock R, et al. Lactate clearance and management. In: Dudley H, Carter D and Russell RCG (eds).
survival following injury. J Trauma 1993; 35: 584. Rob & Smith Operative Surgery. Trauma Surgery Part II, 4th edn.
11 Davis J, Makerise R, Holbrook T, et al. Base deficit as an indicator London: Butterworths.
of significant abdominal injury. Ann Emerg Med 1991; 20: 842. 36 Fullen WD, Hunt J, Altemeier WA. The clinical spectrum of
12 Moore E, Burch JM, et al. Staged physiological restoration and penetrating injury ton the superior mesenteric arterila
damage control Surgery. World J Surg 1998; 22: 1184. circulation. J Trauma 1972; 12: 656–64.
13 Ghimenton F, Thomson SR, Muckart DJJ, Burrows R. Abdominal 37 Asensio JA, Chahwan S, Hanpeter D, et al. Operative
content containment: practicalities and outcome. Br J Surg 2000; management and outcome of 302 abdominal vascular injuries.
87: 106–9. Am J Surg 199; 178: 235–9.
14 Martin R, Byrne M. Postoperative care and complications of 38 Asensio JA, Britt LD, Borzotta A, et al. Multiinstitutinal
damage control surgery. Surg Clin North Am 1998; 77: 929. experience with the management of superior mesenteric artery
15 Hirshberg A, Mattox K. Planned reoperation for severe trauma. injuries. J Am Coll Surg 2001; 193: 354–66.
Ann Surg 1995; 222: 3–8. 39 McAnnich JW, Caroroll PR. Renal exploration after trauma.
16 Burrows R, Edington J, Robbs JV. A wolf in wolf’s clothing – the Indications and reconstructive techniques. Urol Clin North Am
abdominal compartment syndrome. S Afr Med J 1995; 85: 46–8. 1989; 16: 203–11.
17 Meldrum DR, Moore FA, Moore EE. Selective management of 40 Robbs JV, Costa M. Injuries to the great veins of the abdomen.
the abdominal compartment syndrome: results of a prospective S Afr J Surg 1984; 22: 223–8.
analysis. Am J Surg 1997; 174: 667. 41 Degiannis E, Velmahos GC, Levy RD, et al. Penetrating injuries of
18 Ivatury RR, Diebel L, Porter JM, Simon RJ. Intra abdominal the abdominal inferior vena cava. Ann R Coll Surg Engl 1996;
hypertension and the abdominal compartment syndrome. Surg 78: 485–9.
Clin North Am 1997; 77: 783–800. 42 Schrock T, Blaisdell FW, Mathewson C Jr. Management of blunt
19 Krohn IL, Harman PK, et al. The measurement of intra abdominal trauma to the liver and hepatic veins. Arch Surg 1968; 96:
pressure as a criterion for abdominal exploration. Ann Surg 1984; 698–704.
199: 28. 43 Yellin AE, Chaffee CB, Donovan AJ. Vascular isolation in
20 Granchi TS, Liscum K. Logistics of damage control. Surg Clin treatment of juxtahepatic venous injuries. Arch Surg 1971; 102:
North Am 1997; 77: 921. 566–73.
21 Madiba TE, Muckart DJJ. Retroperitoneal haematoma and related 44 Khaneja SC, Pizzi WF, Barie PS, Ahmed N. Management of
organ injury – management approach. S Afr J Surg 2001; 39: 41–4. penetrating juxtahepatic inferior vena cava injuries under total
22 Goins WA, Rodriquez A, Lewis J, et al. Retroperitoneal vascular occlusion. J Am Coll Surg 1977; 184: 469–74.
haematoma after blunt trauma. Surg Gynecol Obstet 1992; 174: 45 Pachter HL, Spencer FC, Hofstetter RS, et al. Management of
281–90. juxtahepatic venous injuries without an atriocaval shunt. Surgery
23 Kudsk KA, Sheldon GF. Retroperitoneal haematoma. In: Blaisdale 1986; 28: 1433–8.
FW, Trunkey DD (eds). Abdominal Trauma. New York: Thieme- 46 Sheldon GF, Lim RC, Yee ES, Petersen SR. Management of injuries
Stratton, 1982: 279–93. to the porta hepatis. Ann Surg 1985; 202: 539–45.
24 Streichen FM, Dargan EL, Pearlman DM, Weil PH. The 47 McFadden D, Lawlor BJ, Ali F. Portal vein injury. Can J Surg 1987;
management of retroperitoneal haematoma secondary to 30: 91–2.
penetrating injuries. Surg Gynecol Obstet 1966; 123: 581–91. 48 Jurkovich GJ, Hoyt DB, Moore FA, et al. Portal triad injuries.
25 Costa M, Robbs JV. Management of retroperitoneal haematoma J Trauma 1995; 39: 426–34.
following penetrating trauma. Br J Surg 1985; 72: 662–4. 49 Rao R, Ivatury MD, Manohar N, et al. Portal vein injuries-
26 Weil PH. Management of retroperitoneal trauma. Curr Probl Surg noninvasive follow up of venography
1983; 20: 539–620. 50 Bostwick J, Stone HH. Trauma to the portal venous system. South
27 Angorn IB. Segmental de-arterialisation in penetrating renal Med J 1976; 68: 1369–72.
trauma. Br J Surg 1977; 64: 59–65. 51 Stone HH, Fabian TC, Turkleson ML. Wounds of the portal system.
28 Corriere JN, McAndrew JD, Bension GS. Intraoperative decision World J Surg 1982; 6: 335–41.
making in renal trauma surgery. J Trauma 1991; 31: 1390–2. 52 Pachter HL, Drager S, Godfrey N, et al. Traumatic injuries of
29 Carroll PR, Klosterman PW, McAninch JW. Surgical management the portal vein. The role of acute ligation. Ann Surg 1979; 189:
of renal trauma: analysis or risk factors, technique and outcome. 383–5.
J Trauma 1982; 22: 285–90. 53 Petersen Sr, Sheldon GF, Lim RC Jr. Management of portal vein
30 Feliciano DV. Management of traumatic retroperitoneal injuries. J Trauma 1979; 19: 616–20.
haematoma. Ann Surg 1990; 211: 108–23. 54 Fish JC. Reconstruction of the portal vein. Case reports and
31 Baumgartner F, White GH, White RA. Controversies in the literature review. Am Surg 1966; 32: 474–8.
management of retroperitoneal haemorrhage associated with 55 Nair R, Abdool Carrim ATO and Robbs JV. Gunshot injuries of the
pelvic fractures. J Nat Med Assoc 1995; 87: 33–8. popliteal artery. Br J Surg 2000; 87: 602–7.
442 Abdominal vascular injuries

56 Meyer J, Walsh J, Schuler J, et al. The early fate of venous repair 69 Baum S, Athanasoulis CA, Waltman AC, Ring EJ. Gastrointestinal
after civilian vascular trauma. A clinical haemodynamic, and haemorrhage – angiographic diagnosis and control. Adv Surg
venographic assessment. Ann Surg 1987; 206: 458–64. 1973; 7: 149.
57 Mullins RJ, Lucas CE, Ledgerwood AM. The natural history 70 Brandt MM, Kazanjian S, Wahl WL. The utility of endovascular
following venous ligation for civilian injuries. J Trauma 1980; stents in the treatment of blunt arterial injuries. J Trauma 2001;
20: 737–43. 51: 901–5.
58 Lau JM, Mattox Kl, Beall AC, et al. Use of substitute conduits in 71 Parodi JC, Schonolz C, Ferreira LM, Bergan J. Endovascular
traumatic vascular surgery. J Trauma 1977; 17: 541. treatment of traumatic arterial lesions. Ann Vasc Surg 1999: 13:
59 Grace DM, Pitt DF, Gold RE. Vascular embolisation and occlusion 121–9.
by angiographic techniques as an aid or alternative to operation. 72 Ohki T, Veith FJ, Krass K, et al. Endovascular therapy for upper
Surg Gynecol Obstet 1976; 143: 469–79. extremity injuries. Semin Vasc Surg 1998; 11: 106–15.
60 Higashida RT, Halbach VV, Fong YT, et al. International 73 Marin ML, Veith FJ, Synamon J, et al. Transluminal repair of
neurovascular treatment of traumatic carotid and vertebral penetrating vascular injury. J Vasc Intervent Radiol 1994;
artery lesions. Am J Radiol 1989; 153: 577–82. 5: 592–4.
61 Sclafani AP, Sclafani JA. Angiography and transcatheter 74 Reddy SG, Rothstein CP, Saker MB, et al. Placement of a PTFE
embolisation of vascular injuries of the face and neck. covered wall stent through a 12Fr sheath for the exclusion of a
Laryngoscope 1996; 106: 168–73. common iliac artery aneurysm. Cardiovasc Intervent Radiol 1999;
62 Robbs JV. Basic principles in surgical management of vascular 22: 152–4.
trauma. In: Greenhalgh RM (ed). Vascular and Endovascular 75 Strauss DC, Du Toit DF, Warren BL. Endovascular repair of
Techniques. Philadelphia, PA: WB Saunders, 1994. occluded subclavian arteries following penetrating trauma.
63 Green MHA, Duell RM, Johnson CD, Jamieson NV. Haemobilia. J Endovasc Ther 2001; 8: 529–33.
Br J Surg 2001; 88: 773–86. 76 Du Toit DF, Strauss DC, Blaszczyk M, et al. Endovascular
64 Fagan EA, Allison DJ, Chadwick VS, Hodgson HJ. Treatment of treatment of penetrating of thoracic outlet arterial injuries.
haemobilia by selective arterial embolisation. Gut 1980; 21: 541–4. Eur J Vasc Endovasc Surg 2000; 19: 489–95.
65 Mitchell SE, Schuman LS, Kaufman SL, et al. Biliary catheter 77 Althaus SJ, Keskey TS, Harker CP, Coldwell DM. Percutaneous
drainage complicated by haemobilia: treatment by balloon placement of self-expanding stent for acute traumatic arterial
embolotherapy. Radiology 1985; 157: 645–52. injury. J Trauma 1996; 41: 145–8.
66 Sclafani SJA, Ben-Menachem Y. Embolotherapy in abdominal 78 Vernhet H, Marty-Ane C, Lesnik A, et al. Dissection of the
trauma. In: Neal MP Jr, Tisnado J Cho S-R, (eds). Emergency abdominal aorta in blunt trauma; management by percutaneous
Interventional Radiology. Boston, MA: Little, Brown, 1989: 53–77. stent. Cardiovasc Intervent Radiol 1997; 20: 473–6.
67 Krige JEJ, Bornman PC, Terrblanche J. Liver trauma in 446 79 Sternbergh WK 3rd, Conners MS 3rd, Ojeda MA, Money SR.
patients. S Afr J Surg 1997; 35: 10–15. Acute bilateral iliac artery occlusion secondary to blunt
68 Olsen WR. Late complications of central liver Injuries. Surgery trauma – successful endovascular treatment. J Vasc Surg 2003;
1982; 92: 733–43. 38: 589–92.
 37
Limb Replantation

COLIN M MORRISON, MICHAEL D BRENNEN

The problem 443 Postoperative management 450

Definitions 443 Maintaining passive movements and encouraging active 451
Classification 443 movements
Emergency treatment 444 Complications 451
Indications for replantation 445 Secondary surgery 451
Management at the replantation centre 446 Replantation of the lower limb 451
Principles of anaesthesia 447 Clinical results 452
Recommended surgical sequence for replantation of an 447 The future 452
amputated limb References 454

THE PROBLEM DEFINITIONS

Replantation represents one of the pinnacles of reconstruc- • Replantation is the reattachment of a body part that
tive surgery, involving a wide range of tissues and surgical has been completely amputated.
techniques. It has evolved from an experimental procedure • Revascularisation is the restoration of circulation to a
in the 1950s and 1960s to a relatively common emergency devascularised but not completely amputated part.
operation performed throughout the world.1–5 The first
The distinction between replantation and revascularisa-
successful replantation of a traumatically severed limb was
tion is important when discussing the methods and results
carried out by Malt at the Massachusetts General Hospital
of reattachment of amputated parts. The survival rates
in 1962.6 Two years later the first microsurgical replanta-
following revascularisation are generally better than those
tion of a hand was performed by Chen in China.7
of replantation because adequate venous drainage often
Subsequent improvements in microvascular technique
remains intact in the former.1
have extended the indications for replantation and now
successful thumb, finger, hand, arm and leg replantations
are routinely achieved.2
CLASSIFICATION
Most major amputations occur in the workplace and
the causes vary according to local industry. Large machines,
chain saws and farming equipment are capable of trapping The condition of the amputated part and the stump can be
and amputating an entire limb. In children an alarmingly classified according to the mechanism of injury, as this
high percentage of lower limb amputations are caused by influences the management and subsequent outcome. The
lawn mowers. Studies have shown peaks of amputations mechanism of injury determines the zone of tissue damage
occurring on a Monday and Thursday at 11 o’clock in the and remains one of the most critical components in patient
morning, a time at which employees become hungry, irri- evaluation.
table and lose concentration.4 Industrial educational pro- Guillotine amputations divide the tissues sharply
grammes continue to have the greatest impact in reducing and create only a localised zone of injury. The outlook
the occurrence of this devastating injury. is favourable with a good chance of functional return
444 Limb replantation

Figure 37.1 A guillotine amputation at the mid-metacarpal Figure 37.3 An avulsion injury of the thumb. Note the tendon
level demonstrating the localised zone of injury avulsed from a more proximal level

Trauma Life Support (ATLS) protocol. The immediate pri-

orities are to establish an adequate airway, breathing and
circulation. Major life-threatening injuries take precedence
over the replantation of an amputated part (Fig. 37.4).10
A non-adherent dressing, a firm compression bandage
and elevation of the stump are usually sufficient to control
bleeding. The use of artery forceps and ligatures should be
avoided as these may cause further vessel damage. Tetanus
immunisation status should be determined and a booster
given if indicated. Broad spectrum antibiotics should be
administered if the wound is heavily contaminated or if the
patient is diabetic or immunocompromised.
X-rays are taken of the amputated part and the limb suf-
fering the amputation. X-rays of both are important in that
Figure 37.2 A crush injury with a larger zone of tissue damage they help to determine the best method of bone stabilisa-
tion. There may be missing bony fragments and this must
be taken into consideration in subsequent management.
The amputated part should be wrapped in damp, normal
(Fig. 37.1). Crushing amputations result in a larger zone of saline gauze and put into a sealed plastic bag. The sealed
tissue damage which may not be easily defined (Fig. 37.2). bag should then be placed in a container containing an ice-
The amount of tissue damage is directly proportional to saline bath to maintain a temperature of between 4 and
the size and weight of the object causing the amputation. 10 °C (Fig. 37.5). The amputated part should never be
Crushing amputations often have poor function even directly surrounded by ice or placed into hypotonic or hyper-
if they survive replantation.1 Avulsion amputations are tonic solution.
caused by a tearing force and there is usually extensive tis- All accident receiving centres should have telephone
sue injury with associated vascular damage extending well access to a replantation centre and should contact the sur-
beyond the amputation site (Fig. 37.3). A corkscrew appear- geon on call as soon as the patient is admitted. If the
ance of an artery, also known as ‘the ribbon sign’, suggests patient is a suitable candidate for replantation, then both
that an avulsion force has been applied and this segment of the patient and the amputated part should be transported
vessel should be excised and replaced by a vein graft.8,9 to the replantation centre by the most expedient means
Crushing and avulsion amputations represent relatively possible.
poor prospects for successful replantation because exten-
sive tissue resection is required to escape the zone of injury.
Emergency treatment of a patient with a
limb amputation
EMERGENCY TREATMENT
• Establish Airway, Breathing, Circulation
The patient should receive emergency treatment in accord- • Rule out other life-threatening injuries
ance with the American College of Surgeons Advanced • Control bleeding from the amputation stump
 Indications for replantation 445

(a) (b)

Figure 37.4 (a) An avulsion amputation of the upper limb in a

22-year-old man caused by a piece of farming equipment.
(b) Clinical examination showing marked bruising in the right
supraclavicular fossa. (c) A chest X-ray demonstrating widening of
the mediastinum. The subclavian artery had been avulsed,
(c) requiring operative intervention to control life-threatening
haemorrhage

replantation (see Chapter 33). The ischaemic tolerance of

• Check tetanus immunisation status/administer
an amputated extremity is inversely related to the amount
antibiotics
of muscle it contains. A proximal limb amputation can tol-
• X-ray both the amputated part and the limb
erate 4–6 hours of ischaemia before myonecrosis and the
suffering the amputation
release of toxic metabolites results in acidosis, myoglobin-
• Cool the amputated part
uria, cardiogenic shock and renal failure (see Chapters 2
• Contact the replantation centre
and 4). This is known as the replantation syndrome.
Replantation of a limb is therefore not recommended if
ischaemia time is longer than 6 hours although that period
can be prolonged by proper cooling of the amputated
INDICATIONS FOR REPLANTATION part.13,14
The smaller vessels of infants make microsurgical
Replantation is not always indicated for patients with limb repair more difficult but good nerve recovery often leads
amputations (Fig. 37.6). In considering a patient for pos- to excellent return of function. In the elderly, however,
sible replantation the surgeon must weigh the systemic nerve recovery cannot be expected with any reliability
risks against the potential loss of the amputated part.11 The and any attempt at replantation must be carefully con-
main factor in patient selection is the potential for the sidered. Successful replantation has been reported in an
replanted part to become an impediment and success must octogenerian, but the functional results were disappoint-
be measured in terms of the functional outcome rather ing.15 In patients with degenerative joint disease, post-
than limb survival alone. Careful assessment is therefore operative oedema and splinting can cause stiffness of the
essential before attempting this difficult and occasionally whole hand.
life-threatening procedure.12 Patients who have severe systemic injury or disease may
All indications for replantation must take into account not tolerate the anaesthesia and surgery well. Results with
the status of the amputated part and the patient. Amputated patients who have severe mental disease or suffer from sub-
parts that are severely crushed and those with multiple stance abuse (see Chapter 44) may also be poor. Although the
level injuries may have poor function even if they survive technical aspects of replantation in these individuals may
446 Limb replantation

Figure 37.5 The amputated part

should be wrapped in damp gauze
and put into a sealed plastic bag. The
sealed bag should then be placed
in a container containing an
ice-saline bath

• Relative contraindications to limb replantation

– Severe crush/avulsion injury
– Multiple levels of injury
– Prohibitive ischaemia time
– Significant associated injuries or medical illness
– Psychiatric instability

MANAGEMENT AT THE REPLANTATION

CENTRE

Ideally every replantation centre should have two surgical

Figure 37.6 A bomb blast injury of the upper limb. Due to the teams available at all times. One team thoroughly debrides
severe damage and multiple levels of injury replantation was not the amputated part under magnification and identifies the
appropriate various structures while the other team prepares the patient
for surgery.
not present a problem, postoperative compliance is unpre-
dictable and rehabilitation difficult (Fig. 37.7).3
The patient’s sex, cultural background, occupation and The amputated part
personality may also influence the decision to attempt
replantation. The only absolute contraindication to replanta- The amputated part is immediately taken to the operating
tion is when the patient’s associated injuries or pre-existing room, where it is cleaned using Hartmann solution. Under
medical condition preclude a prolonged surgical procedure. loupe magnification the part is carefully debrided and the
vessels and nerves identified and labelled with sutures
(Fig. 37.8). Labelling of these structures can be very helpful
in the later stages of reconstruction when fatigue influ-
Replantation strategies
ences a surgeon’s proficiency.1 The amputated part is then
kept cool until required later in the procedure.
• Indications for limb replantation
– Amputations at the level of the mid-metacarpal,
wrist or distal forearm
The patient
– Amputations at the level of the distal third of the leg
– Amputations in children
Preparation for surgery should proceed without delay.
– Bilateral amputations
Repeated preoperative examination of the injured limb
 Recommended surgical sequence for replantation of an amputated limb 447

(a) (b)

(c) (d)

Figure 37.7 (a, b) A guillotine amputation of the hand caused by a hacksaw. (c) Despite a good immediate postoperative result, the patient
failed to comply with rehabilitation. (d) The poor functional result

should be avoided, as this may increase the risk of vascular The patient should be placed on a well padded mattress,
spasm. The patient is kept warm and intravenous solutions for comfort and protection of pressure points and an
administered to maintain a hypervolaemic state. A chest indwelling urinary catheter inserted. Particular care should
X-ray and/or electrocardiogram (ECG) are taken if indicated be taken when positioning healthy limbs to avoid further
by the patient’s age and associated medical conditions. Blood injury. The patient should be kept warm and well hydrated.
is taken for a full blood count, blood type and cross-match Active warming with a warm air blanket and fluid warmer
as the patient may require a blood transfusion. helps maintain an optimum temperature of 38 °C. Good
A thorough history and examination is completed hydration is essential, allowing for blood loss at the time of
before discussing treatment options and outcomes with injury and during the operative procedure. These measures
the patient and his or her family. Consent should include reduce the incidence of thrombus formation at the vascular
permission for amputation, vein grafts, nerve grafts, skin anastomosis by avoiding vasoconstriction and hypotension.
grafts, bone grafts and possibly free flaps for soft tissue Haemoglobin levels should be monitored and maintained
coverage. at 7–8 g/dL. A low blood viscosity helps promote flow across
the site of the anastomosis. Transfusion should be avoided
unless the patient becomes symptomatic from anaemia.
PRINCIPLES OF ANAESTHESIA

Major limb replantation usually lasts many hours and

RECOMMENDED SURGICAL SEQUENCE FOR
requires general anaesthesia. This can be augmented by
regional anaesthesia using indwelling catheters. These pro-
REPLANTATION OF AN AMPUTATED LIMB
vide access for sympathetic nerve blockade aimed at
vasodilatation of the limb and for giving continuous anal- • Identification of arteries, veins, nerves and tendons
gesia, reducing unnecessary postoperative pain.16,17 • Wound debridement
448 Limb replantation

present in a distal severed forearm must be carefully evalu-

ated and excised if there is doubt about its future.
Similarly, amputations at the metacarpal level require
resection of distal avascular intrinsic muscles. The sur-
geon’s haste to re-establish blood flow must not result in
neglect of proper wound cleansing and is essential in pre-
venting myonecrosis and subsequent infection.1,13,14

Bony stabilisation

Bones may be shortened during replantation to permit bet-

ter bone contact in comminuted fractures, allow a tension
free vascular anastomosis and nerve repair and enable bet-
ter soft tissue coverage. Rigid fixation also protects the vas-
cular repair and allows early mobilisation of the replanted
limb. If possible, the periosteum is repaired to help bone
healing and minimise adhesions to flexor and extensor
tendons.
Amputations at the forearm level or higher usually
result from violent injury in which a large zone of tissue
has been affected. Consequently, the amount of bone
resected is between 2 and 4 cm. For wrist replantations,
however, as little bone as possible is resected in order to
preserve mid-carpal or radiocarpal wrist motion. For ampu-
tations at the metacarpal level, interosseous wiring through
four drill holes at 12, 3, 6 and 9 o’clock in each end
of the bone permits two strands of 24 gauge wire to be
placed perpendicular to each other. This provides stability,
allowing early motion with less chance of angulation and
Figure 37.8 The amputated part is carefully debrided. Vessels non-union.
and nerves are identified and labelled with sutures For major limb replantation, compression plating is the
preferred method of providing bone fixation (Fig. 37.9).
However this requires more time and bone exposure, pos-
• Bony stabilisation
sibly with further damage to soft tissue structures. If con-
• Vascular repair
tamination is significant, an external fixator may be used to
• Tendon repair
reduce the risk of infection (Fig. 37.10).18
• Nerve repair
• Soft tissue coverage
Vascular repair
Identification of arteries, veins, nerves
and tendons Most amputations of the upper extremity occur at the digit
or hand level. Replantation of limbs proximal to the wrist
Magnifying loupes are worn during the initial dissection involve greater muscle mass and the duration of avascular-
and structures to be repaired are identified and marked ity of the detached part is more critical. In replantations
with sutures. It is extremely important to take great care in proximal to the metacarpal level, arterial inflow must be
isolating vital structures before wound debridement. re-established after initial debridement and rapid bone
stabilisation.
If the ischaemia time is prolonged a shunt may be
Wound debridement placed between the proximal and distal arteries and veins
to perfuse the limb as quickly as possible. The Sundt and
The amputation stump is cleaned and debrided, removing Pudenz ventriculoperitoneal shunts are commercially
all contaminated, non-viable tissue. Frayed tendon ends available, but any suitable silastic tubing may be employed
and bone fragments are also removed. as long as great care is exercised during its introduction.19,20
Forearm muscles receive most of their blood supply Muscles in the hand or forearm inevitably becomes oede-
through their proximal half and, therefore, any muscle matous following reperfusion and therefore prophylactic
 Recommended surgical sequence for replantation of an amputated limb 449

Figure 37.11 Accurate suture placement is essential when

performing the end-to-end anastomosis

Figure 37.9 A compression plate used for bony stabilisation

Figure 37.12 Good arterial flow through the completed

anastomosis

Figure 37.10 An external fixator applied during replantation

repair against a yellow polyethylene background to help in
localisation of suture placement (Fig. 37.12).21 It is imperative
fasciotomy of all muscle compartments in the amputated to avoid accidentally picking up and suturing together oppos-
part should be performed prior to revascularisation. In ite walls.
addition, the carpal tunnel is released if the amputation is The field is kept moist with saline to prevent tissue des-
proximal to the wrist. iccation. If a tension-free vascular anastomosis cannot be
The limb should be re-arterialised before venous repair. performed an interposition vein graft may be used. Vein
Prior to the anastomosis, the tourniquet is released and the grafts are reversed when used for arterial repair so that any
vessels debrided back to normal intima, ensuring pulsatile valves within the graft do not impede blood flow. The vas-
proximal vessel inflow. Small vascular clamps are used to cular clamps are opened on completion of the anastomosis
hold both ends of the severed vessels in place and to con- to allow the release of toxic metabolites into the surgical
trol bleeding during repair. Loose adventitia is removed field and prevent their accumulation in the systemic
and the vessel ends are brought together by sliding the circulation. Ideally, two superficial veins should be repaired
clamps on the holding bar. Heparinised saline solution is for every arterial anastomosis to maintain a balanced
flushed into the lumen of the vessels with an irrigating circulation.
catheter to cleanse and reduce the tendency for platelets to
adhere to the intima and anastomotic site. Closed vessel-
dilating forceps are inserted into the artery or vein and Tendon repair
gently opened, stretching the vessel. This gentle dilatation
helps to prevent spasm of the muscle within the vessel wall Primary end-to-end repair gives the best opportunity for
at the site of vessel repair. tendons to regain function. This should be performed
The vascular anastomosis is performed under the oper- without undue tension, after the necrotic or damaged ten-
ating microscope or loupe magnification. An end-to-end don ends have been resected.
anastomosis is performed enabling accurate opposition of Early postoperative mobilisation promotes healing and
vessels (Fig. 37.11). Fine nylon sutures are used for vessel reduces the tendency of tendons to become adherent to the
450 Limb replantation

surrounding tissues. A four strand core suture of 4-0 nylon However, local tissue is rarely available for coverage and
or braided polyester, together with a 6-0 nylon epitendi- regional or free flaps may be needed. A split skin graft is
nous suture provides sufficient strength to allow a flexor ten- usually not appropriate.
don to be mobilised, with only a small risk of rupture at Muscle is the best choice for these wounds in terms of
the repair site. Extensor tendons only require a core suture vascularity and resistance to infection and adequate tissue
but an epitendinous suture may be added to tidy up the should be transferred to fill dead space and afford cover-
tendon ends. age of deeper structures. The muscle can be raised as a
The repaired tendon ends should lie in a healthy soft tis- myocutaneous flap or separately and covered with a split
sue bed and have good skin and subcutaneous tissue cover. skin graft taken from a suitable donor area. The pectoralis
Divided muscle bellies are repaired with 4-0 polyglactin. If major or latissimus dorsi can be very useful as pedicled
tendon damage is too great to allow primary repair a sec- flaps for coverage of the upper arm but for most forearm
ondary tendon graft or a tendon transfer will be required to injuries a free flap will be required. The surgeon can
restore movement. choose whatever muscle with which he or she is most
comfortable.

Nerve repair
POSTOPERATIVE MANAGEMENT
The immediate survival of a replanted part is determined
by the success of the vascular anastomosis, but the eventual
function of a limb depends on the quality of nerve regener- Initial dressing
ation. It is vital to identify the extent of nerve damage and
to resect the proximal and distal ends back to microscopic- The replanted limb is immobilised to protect the arterial,
ally normal tissue. In clean cut amputations, primary nerve venous and nerve repairs. A bulky dressing is applied to
repair may be possible, especially in cases where bone splint and protect the limb and to maintain it in a position
shortening has been performed. Tension must be avoided which enhances later mobility. The ideal position of the
and if the nerve cannot be sutured without tension then hand is with the metacarpophalangeal joints in 70 degree
nerve grafting is required. This can be carried out as a flexion, the interphalangeal joints in a neutral position and
primary procedure in very clean wounds but is normally the thumb in maximum palmar abduction. If this position
performed as a secondary operation, when nerve ends are compromises the vascular repair, an alternative position as
trimmed with a sharp blade and repaired with epineural close to this ideal as possible should be used. After 48
sutures under magnification. hours, the thumb and fingers are allowed free to permit
Nerve regeneration occurs at a rate of 1 mm/day. As a early active mobilisation.
result, there will be no sensory recovery in the fingertips
for 9–12 months in amputations at the wrist and for 12–18
months in forearm amputations. Coarse sensation will Monitoring the circulation
return but the quality of sensory recovery will be far from
normal on formal testing. The results are generally much Hypovolaemia and hypotension are prevented by vigorous
better in children. hydration and the continuous use of adequate analgesia.
The small muscles of the hand are responsible for fine Vasoconstriction is prevented by keeping the patient warm
manipulative movements, positioning and some power and by avoiding medication such as caffeine and nicotine.
but it is rare to get any small muscle recovery following The haemoglobin and electrolyte status are monitored and
replantation, except in children. Return of forearm func- a blood transfusion administered if required.
tion should be expected but secondary surgery, in the form Clinical monitoring is the mainstay of postoperative
of tendon grafts or tendon transfers, may be required. evaluation of a replanted part. The colour, capillary refill,
temperature and tissue turgor are carefully and frequently
recorded with observations every 30–60 minutes. Arterial
Soft tissue coverage insufficiency is manifested by a pale, cool extremity with
absent capillary refill and poor tissue turgor. Venous insuf-
The final and often most important issue in limb replanta- ficiency is manifested by a congested extremity that has
tion is management of the skin and subcutaneous tissue increased tissue turgor and extremely rapid capillary refill.
loss. Many surgeons believe that this should be achieved at Needle puncture of the skin is also helpful for assessing
the time of the first procedure, as exposed bone and neu- circulation. Bright, brisk bleeding is observed if the arterial
rovascular structures fare poorly if not covered with well and venous systems are adequate. If arterial insufficiency is
vascularised tissue. present, the wound will hardly bleed at all. If venous con-
If possible, the skin is loosely approximated to prevent gestion is present, rapid, copious bleeding of dark, deoxy-
vascular constriction when postoperative oedema occurs. genated blood is seen.
 Replantation of the lower limb 451

Arterial insufficiency of a limb is treated by lowering it

and placing it on a pillow. Venous congestion is treated by
Complications of limb replantation
elevation, loosening of the dressing and/or removal of
sutures to relieve constriction of the vascular repair. If there • Early
– Vascular thrombosis
is no early response to these measures then re-exploration
– Bleeding
and revision of the vascular anastomoses should be per-
– Infection
formed. Clinical experience has underlined the vital impor-
– Loss of the part
tance of early re-explorations when faced with circulatory
complications. • Late
– Scar contracture
– Tendon rupture
– Neuroma
MAINTAINING PASSIVE MOVEMENTS AND – Non-union
ENCOURAGING ACTIVE MOVEMENTS – Osteomyelitis
– Pain
– Cold intolerance
Once replantation has been successfully accomplished and
maintained, the rehabilitation programme can proceed.
This begins after 48 hours and the focus is directed towards
early control of oedema, passive range of motion and pro-
SECONDARY SURGERY
gressive strengthening while avoiding the common prob-
lem of joint contracture.
O’Brien24 taught that all structures should be repaired
primarily to reduce the necessity for secondary surgery.
COMPLICATIONS This is usually possible in guillotine or very minimal crush
injuries, especially if some bone shortening is performed.
Secondary surgery, however, is almost always necessary in
Early complications include vascular thrombosis leading major limb replantation.25 It may not be possible to repair
to circulatory problems, bleeding and infection.22 Careful all tendons primarily and secondary tendon repair, graft-
postoperative monitoring is necessary to detect early circu- ing or transfer may be needed to restore active movement
latory problems. If initial conservative measures do not to a particular joint.
restore the circulation rapidly then an early return to Unless the extent of nerve injury can be accurately
theatre is necessary to inspect and re-do the vascular anas- assessed it is unwise to perform primary nerve repair. A
tomosis. Bleeding may be controlled by elevation but per- delayed primary or early secondary repair may be possible
sistent bleeding may also necessitate a return to theatre to if the extent of nerve damage is limited. For larger nerve
find and control the bleeding vessels. defects cable grafting, using for example the sural nerve,
Infection is the main complication of limb replantation. may be required.
If thorough debridement has not been performed there The passive movements of joints adjacent to the site of
may be skin necrosis or necrosis of muscle. Antibiotics replantation may be restricted by scar contracture. Release
are not routinely used postoperatively but should be given of these will improve the range of movement. Release may
at the earliest signs of cellulitis and pyrexia. Deeper infec- require local, distant or free flap coverage to prevent recur-
tion may be due to unrecognised muscle necrosis and rence. Areas of non-union and bone infection are best
further radical debridement may be necessary to elimi- treated by resection and bone grafting. Arthrodesis may be
nate dead tissue and bring infection under control. Such needed for joints which are severely damaged and/or painful.
radical debridement may create a situation where flap
surgery is needed to maintain soft tissue cover for the
anastomosis.
REPLANTATION OF THE LOWER LIMB
The later complications of limb replantation include
those normally expected after trauma to the individual
tissues. Fibrosis in the skin and subcutaneous tissues may Amputation of the lower limb is usually secondary to
lead to scar contracture. The tendons may rupture or severe trauma and is commonly associated with extensive
become adherent. Neuromas in-continuity may develop local tissue destruction and injury to other sites. As a result,
and cause exquisitely tender spots. Nerve recovery may fail there have been few cases of successful outcome following
to progress. Non-union and osteomyelitis may require fur- replantation. Magee and Parker are credited with perform-
ther bone resection and perhaps bone grafting. Painful ing the first foot replantation in 196926 and sporadic
syndromes may develop and cold intolerance in the reports have followed.27–32 Restoration of function remains
replanted part is common.23 the primary goal and favourable conditions are a young,
452 Limb replantation

(a) (b) (c)

Figure 37.13 Complete amputation of the foot in a 19-year-old man following a motorcycle accident. (a) X-ray of the limb and the
amputated part. (b) Successful replantation. (c) Good return of function

healthy patient, with a sharp amputation, located in the Table 37.1 Survival rates following limb replantation
distal third of the leg (Fig. 37.13).5
The contraindications to replantation mirror those of Per cent
the upper extremity and include life-threatening associated Authors Year Level survival
injuries, a crushing or avulsion force and age or illness
which precludes a prolonged operation. Due to the larger Wang et al.39 1981 Upper and lower limb 77
mass of muscle, myonecrosis and renal failure may follow Tamai4 1982 Upper limb 94
a significant ischaemic period in this group of patients and Wood and Cooney 25
1986 Above-elbow 71
particular attention should be paid to this issue.33 In clin- Daigle and Kleinhert40 1991 Upper and lower 87
ical practice, ideal conditions are rarely encountered and limb in children
proper patient selection remains the key factor.
Axelrod and Buchler41 1991 Upper limb 93

preservation of a functioning elbow for later prosthesis

CLINICAL RESULTS fitting.37
Replantation of the lower limb remains controversial.
Revascularisation of an amputated part should be expected The available lower limb prostheses makes amputation less
in approximately 80 per cent of patients. Function, however, of a functional problem compared with the upper extrem-
is the only real measure of success and is usually deter- ity. The larger mass of leg muscle also tolerates ischaemia
mined by the quality of sensory recovery. Based on this, poorly, and without adequate sensory recovery, the skin of
success can often be predicted by the level of the amputa- the foot is at risk of recurrent breakdown.
tion (Table 37.1).3 In children, it is generally believed that replantation
A hand replanted from the mid-metacarpal level to the should be attempted with almost any part. Viability rates
distal wrist offers functional return superior to available are lower but the functional results are much better than in
prostheses (Fig. 37.14).12,13,34–36 More proximal amputa- adults.36
tions need to be considered carefully, as the risk of compli-
cations increases while the functional return decreases. At
the level of the mid-forearm, reinnervation of the small
THE FUTURE
muscles of the hand is unlikely and this may compromise
later performance. At the level of the arm or shoulder, the Composite tissue allotransplantation represents the next
results are generally poor and the main goal is often the frontier in replantation surgery. The recent hand transplants
 The future 453

(a) (b)

(c) (d)

Figure 37.14 (a) Guillotine amputation of the hand

caused by a bandsaw. (b) Replantation of the
amputated part. (c–e) Good functional outcome can
(e) be attributed to the sharp nature of the injury and the
mid-metacarpal level

performed in Lyon, France and Louisville, KY, USA, have there is uncertainty about the risk–benefit ratio of lifelong
stimulated public interest and heightened the debate about immunosuppression given the potential for organ failure,
such procedures. opportunistic infection and malignancy.
Current replantation literature confirms that significant Future successful composite tissue allotransplantation
functional return can be achieved following a hand trans- will depend on modalities to induce host tolerance such
plant, provided that the appropriate patient is selected and as major histocompatability complex matching or induc-
the procedure and rehabilitation properly implemented. tion of transplant tolerance by exposing the recipient
However, composite tissue transplant studies in animal immune system to donor marrow elements before its
models demonstrate a lack of long term survival data and maturity.38
454 Limb replantation

Conclusions REFERENCES

The primary goal of replantation is to restore function. 1 Goldner RD, Urbaniak JR. Replantation. In Green DP, Hotchkiss
The surgical team tries to produce a well perfused RN, Pederson WC (eds). Green’s Operative Hand Surgery, 4th edn.
replanted part which is supple, sensate and capable of New York: Churchill Livingstone, 1999: 1139–57.
active movement. Careful case selection is important. 2 Kleinert HE, Jablon M, Tsai TM. An overview of replantation and
Good results will be achieved with clean amputations results of 347 replants in 245 patients. J Trauma 1980; 20:
390–8.
relatively distally in the limb in young patients. Crush or
3 Pederson WC. Replantation. Plast Reconstr Surg 2001; 107: 823–41.
avulsion injuries give less favourable results. More prox- 4 Tamai S. Twenty years’ experience of limb replantation – review
imal amputations usually give poorer functional results. of 293 upper extremity replants. J Hand Surg 1982; 7: 549–56.
Recovery of sensation in elderly patients is poor and this 5 Walton RL, Rothkopf DM. Judgment and approach for
will compromise the outcome. Replantation in infants is management of severe lower extremity injuries. Clin Plast Surg
technically difficult but gives excellent results. 1991; 18: 525–3.
As well as the microvascular surgery, skeletal fixa- 6 Malt RA, McKhann CF. Replantation of severed arms. JAMA
tion, tendon repair, nerve suture and healthy skin cover 1964; 189: 716–22.
must all be given meticulous attention. Postoperative 7 Chen Z-W, Meyer VE, Kleinert HE, Beasley RW. Present
monitoring, dressings and wound care requires special- indications and contraindications for replantation as reflected by
ist nursing and medical care. Physiotherapy and occupa- long-term functional results. Orthop Clin North Am 1981; 12:
849–70.
tional therapy are essential, especially in the upper limb.
8 van Beek AL, Kutz JE, Zook EG. Importance of the ribbon sign,
The patient may also need social and even psychological indicating unsuitability of the vessel, in replanting a finger. Plast
support. Reconstr Surg 1978; 61: 32–5.
Good results can be achieved in amputations at or 9 Cooney WP III. Revascularization and replantation after upper
distal to mid-forearm level. The thumb should always be extremity trauma: experience with interposition artery and vein
considered for replantation, as should multiple finger grafts. Clin Orthop Rel Res 1978; 137: 227–34.
amputations. Single finger amputations require special 10 King C, Kuzon WM. Replantation. In: Jebson PJL, Kasdan ML (eds)
indications for replantation. The aesthetic value of replan- Hand Secrets. Philadelphia, PA: Hanley & Belfus, 1998: 229–33.
tation should not be underestimated. Many patients are 11 Sood R, Bentz ML, Shestak KC, Browne EZ. Jr. Extremity
very pleased to have the part restored even though it is of replantation. Surg Clin North Am 1991; 71: 317–29.
little functional value. 12 Hales P, Pullen D. Hypotension and bleeding diatheses following
attempted arm replantation. Anaesth Intensive Care 1982;
The indications for replantation in the lower limb,
10: 359–61.
especially with distal amputations is very limited as below- 13 Goldner RD, Urbaniak JR. Indications for replantation in the adult
knee prostheses are functionally very good. Replantation upper extremity. Occup Med 1989; 4: 525–538.
should also be considered in scalp avulsions and in ampu- 14 Goldner RD, Nunley JA. Replantation proximal to the wrist.
tations of specialised parts such as the ear, nose and penis. Hand Clin 1992; 8: 413–25.
15 Leung PC. Hand replantation in an 83-year-old woman – the
oldest replantation?. Plast Reconstr Surg 1979; 64: 416–18.
16 Berger A, Tizian C, Zenz M. Continuous plexus blockade for
improved circulation in microvascular surgery. Annals of Plastic
Key references Surgery 1985; 14: 16–19.
17 Matsuda M, Kato N, Hosoi M. Continuous brachial plexus block
Goldner RD, Urbaniak JR. Replantation. In Green DP, Hotchkiss RN, for replantation in the upper extremity. Hand 1982; 14: 129–34.
Pederson WC (eds). Green’s Operative Hand Surgery, 4th edn. 18 Tupper JW. Techniques of bone fixation and clinical experience in
New York: Churchill Livingstone, 1999: 1139–57. replanted extremities. Clin Orthop Rel Res 1978; 133: 165–8.
Kleinert HE, Jablon M, Tsai TM. An overview of replantation and 19 Barros D’Sa AAB. The rationale for arterial and venous shunting
results of 347 replants in 245 patients. J Trauma 1980; in the management of limb vascular injuries. Eur J Vasc Surg
20: 390–8. 1989; 3: 471–4.
Pederson WC. Replantation. Plast Reconstr Surg 2001; 107: 20 Barros D’Sa AAB, Moorehead RJ. Combined arterial and venous
823–41. intraluminal shunting in major trauma of the lower limb. Eur J
Tamai S. Twenty years’ experience of limb replantation – review of Vasc Surg 1989; 3: 577–81.
293 upper extremity replants. J Hand Surg 1982; 7: 549–56. 21 Rasheed T, Gordon D. Plastic innovations: background material
Walton RL, Rothkopf DM. Judgment and approach for management for microvascular anastomosis. Br J Plast Surg 1999; 52: 159.
of severe lower extremity injuries. Clin Plast Surg 1991; 22 Strauch B, Greenstein B, Goldstein R, Liebling RW. Problems and
18: 525–3. complications encountered in replantation surgery. Hand Clin
1986; 2: 389–99.
 References 455

23 Russell RC, O’Brien B McCMacLeod AM, et al. The late functional 34 Blomgren I, Blomqvist G, Ejeskar A, et al. Hand function after
results of upper limb revascularization and replantation. J Hand replantation or revascularization of upper extremity injuries:
Surg (Am) 1984; 9: 623–33. a follow-up study of 21 cases operated on 1979–1985 in
24 O’Brien BM. Replantation and reconstructive microvascular Goteborg. Scand J Plast Reconstr Surg 1988; 22:
surgery. Ann R Coll Surg Engl 1976; 58: 171–82. 93–101.
25 Wood MB, Cooney WP III. Above elbow limb replantation: 35 Meyer VE. Hand amputations proximal but close to the wrist
Functional results. J Hand Surg (Am) 1986; 11: 682–7. joint: prime candidates for reattachment (long term functional
26 Magee HR, Parker WR. Replantation of the foot: results after results). J Hand Surg (Am) 1985; 10: 989–91.
two years. Med J Aust 1972; 1: 751–5. 36 Saies AD, Urbaniak JR, Nunley JA, et al. Results after replantation
27 Chen ZW, Zeng BF. Replantation of the lower extremity. Clin and revascularization in the upper extremity in children. J Bone
Plast Surg 1983; 10: 103–13. Joint Surg (Am) 1994; 76: 1766–76.
28 Fukui A, Inada Y, Sempuku T, Tamai S. Successful replantation 37 Graham B, Adkins P, Tsai TM, et al. Major replantation versus
of a foot with satisfactory recovery: A case report. J Reconstr revision amputation and prosthetic fitting in the upper
Microsurg 1988; 4: 387–90. extremity: a late functional outcomes study. J Hand Surg (Am)
29 Lesavoy MA. Successful replantation of lower leg and foot, with 1998; 23: 783–91.
good sensibility and function. Plast Reconstr Surg 1979; 64: 760–5. 38 Lee A.WP, Mathes DW. Hand transplantation: pertinent data
30 Morrison WA, O’Brien BM, MacLeod AM. Major limb and future outlook. J Hand Surg 1999; 24: 906–13.
replantation. Orthop Clin North Am 1977; 8: 343–8. 39 Wang SH, Young KF, Wei JN. Replantation of severed limbs –
31 Tsai TM. Successful replantation of a forefoot. Clin Orthop Rel clinical analysis of 91 cases. J Hand Surg 1981; 6: 31–8.
Res 1979; 139: 182–4. 40 Daigle JP, Kleinhert JM. Major limb replantation children.
32 Usui M, Minami M, Ishii S. Successful replantation of an Microsurgery 1991; 12: 221–31.
amputated leg in a child. Plast Reconstr Surg 1979; 63: 613–17. 41 Axelrod TS, Buchler U. Severe complex injuries to the upper
33 Gayle LB, Lineaweaver WC, Buncke GM, et al. Lower extremity extremity: revascularization and replantation. J Hand Surg 1991;
replantation. Clin Plast Surg 1991; 18: 437–47. 16A: 574–84.
This page intentionally left blank
 SECTION
8
Iatrogenic Injuries

38. Injuries of arterial catheterisation 459

39. Injuries of peripheral endovascular procedures 469

40. Specialty related iatrogenic vascular injuries 487

This page intentionally left blank
 38
Injuries of Arterial Catheterisation

DAVID BERGQVIST, CHRISTER L JUNGMAN

The problem 459 Prevention of catheter injury 462

Aetiology and types of injury 459 Management of complications 463
Incidence 460 References 465
Clinical presentation 461

THE PROBLEM Inadvertent puncture

through the back wall

Arterial catheterisation or puncture may cause bleeding,

occlusion or stenosis due to thromboembolism or intimal
injury, arteriovenous fistula, pseudoaneurysm, arterial wall
dissection, and on occasion a foreign body may be left in
the arterial system. Toxic injury or allergic reaction caused
by contrast media will not be discussed here neither will we
review infectious complications.

AETIOLOGY AND TYPES OF INJURY

Haemorrhage Common Expanding retroperitoneal

femoral artery haematoma
Haemorrhage is common after arterial puncture and is usually
Figure 38.1 Inadvertent puncture through the back wall of the
localised and of minor degree. Major haemorrhage may be
common femoral artery causing retroperitoneal haemorrhage
caused by perforation of the back wall of the artery (Fig. 38.1),
which is difficult to control with manual pressure, simply
because retroperitoneal counterpressure cannot be applied.
Such haemorrhage may cause hypovolaemia and shock. aggregates and fibrin on the surface of the catheter and in the
Factors contributing to excessive blood loss are multiple holes on the catheter tip. At catheter pull-out, the fibrin
punctures, introducers of large diameter and anticoagu- sheath is washed off on the inner side of the vessel, forming
lant/antiplatelet therapy. Another factor of importance is a thrombotic plug. This thrombus may occlude the artery,
puncture through a calcified vessel wall with its lack of but it usually forms a wall-adherent, non-occlusive throm-
contractility. Haemorrhage may also be due to perforation bus. The thrombus may propagate locally or embolise dis-
by the tip of the catheter, a risk which is particularly marked tally. Heparinisation significantly decreases the amount of
in infants whose arteries are fragile. thrombotic material on the catheter.1 The amount of throm-
bus formed depends on the type of material from which the
catheter is made, the length and width of the catheter as well
Thromboembolism as the duration of catheterisation. Thrombotic material in
the holes on the tip of the catheter may embolise if the
Thromboembolism is the result of two main processes. The catheter is flushed leading to occlusion of peripheral arteries
first mechanism has to do with the formation of platelet and a clinical picture of microembolism.
460 Injuries of arterial catheterisation

The second mechanism is intimal rupture and dissection, A pragmatic way of defining major complications is to
or dislodgement of thrombotic or atherosclerotic material, include those requiring surgical intervention and those
either at the place of catheter insertion or caused by the tip of ending in fatality. The registered frequency of these
the catheter. Intimal dissection may bring about immediate complications obviously depends on the definition and
occlusion or progressive formation of thrombus, resulting in whether the study is prospective or retrospective.
occlusion or embolisation after some time. Fragmentation of Lang surveyed 11 402 procedures of retrograde percutan-
the tip of the catheter or guidewire may also occur and in eous transfemoral arteriography (Seldinger procedure),
turn these parts of the catheter may embolise. collecting data by questionnaires completed by 300 hos-
pital radiologists.17 There were 81 (0.7 per cent) complica-
Arteriovenous fistula tions, including thrombosis, requiring surgical intervention,
six of which resulted in limb loss. There were five cases of
Arteriovenous fistulae are caused by simultaneous perfor- guidewire or catheter breakage and seven cases of major
ation of an artery and adjacent vein and these have been haematoma. Massive retroperitoneal haemorrhage from
well documented.2–5 perforation of a pelvic artery occurred in three cases. These
results were compared with those from a survey of 3250
Pseudoaneurysm patients undergoing translumbar aortography by Lang,18
which resulted in only 11 major complications (0.3 per cent).
Pseudoaneurysms are found around the puncture hole in The mortality in the translumbar group, however, was higher
the artery. They are usually small and are particularly prone (0.28 per cent v 0.06 per cent).
to develop if there has been excessive bleeding, and espe- In a survey in 1981, Hessel and co-workers19 reported on
cially when patients are on anticoagulants or antiplatelet complications arising from 118 591 angiographic examin-
drugs.6,7 The risk increases with the complexity of the pro- ations from a large number of hospitals, comparing trans-
cedure.8 They may occlude spontaneously by progressive femoral, translumbar and transaxillary approaches. The
development of thrombosis,9–12 especially those with small overall complication rates was 1.7 per cent for the trans-
flow volumes.13 The longer the pseudoaneurysm neck the femoral group, 2.9 per cent for the translumbar group and
greater the chances that it will close spontaneously.14 Infected 3.3 per cent for the transaxillary group. Thirty deaths were
or mycotic aneurysms may occur when the infected tip of a reported, with the highest number in the transaxillary
catheter perforates the intima (see Chapters 18 and 44). group and the lowest in the transfemoral group. In the
transaxillary group there were also more neurological com-
Dissection plications, including seizures and hemiplegia, but also more
cases of haemorrhage, haematoma in the neurovascular
Dissection of the arterial wall occurs mainly after translum- sheath, arterial obstruction and pseudoaneurysm. In the
bar puncture or catheterisation of the aorta, when contrast translumbar group there were more cases of bleeding and of
medium is injected intramurally.15 Intestinal and renal circu- extraluminal contrast injection. The incidence of puncture-
lation may be compromised by intramural contrast medium. site thrombosis was 0.1 per cent in the transfemoral group,
compared with 0.6 per cent in the transaxillary group and
Types of catheter injury 0 per cent in the translumbar group. Emboli occurred in
0.1 per cent of the transfemoral group, 0.1 per cent of the
transaxillary group and 0 per cent of the translumbar
• Haemorrhage
group. The complication rate was more than four times
• Thrombosis – occlusive, non-occlusive, embolic
higher in hospitals doing fewer than 200 angiographies (2.7
• Arteriovenous fistula
per cent) than in those doing more than 800 (0.6 per cent).
• Pseudoaneurysm
There has been an obvious trend during the past 20
• Dissection
years of a decreasing incidence of complications reported
after arterial punctures and catheterisations, as suggested
by a series of prospective studies.20 The translumbar tech-
INCIDENCE
nique, although used less frequently today, has also been
characterised by a decreased incidence of complications.
Major complications are usually ‘complications threatening The site of catheterisation is of decisive importance in the
life, limb, or the visceral integrity of the patient, or requir- incidence and type of complication. Several reports have
ing subsequent surgical intervention for diagnosis or treat- confirmed a high frequency of complications following
ment, or significantly prolonging the hospital stay of the catheterisation of the axillary artery. Brachial plexus paralysis
patient’.16 Minor complications are ‘asymptomatic com- could be eliminated by meticulous technique and by avoid-
plication(s) of arteriography detected radiographically or a ing delay in surgical exploration once paresis has occurred.
clinically insignificant transient symptom or sign not The brachial artery has been used as an alternative to
endangering the patient or requiring further evaluation’.16 axillary artery catheterisation to avoid the risk of damage
 Clinical presentation 461

In a survey in our institution in 1994, based on the

coronary procedure registry, 49 complications occurred in
649 procedures (7.6 per cent), and of these 32 per cent
were pseudoaneurysms. Stent usage was associated with
10 per cent of pseudoaneurysms compared with 2.2 per
cent associated with percutaneous transluminal coronary
angioplasty (PTCA) only (Fig. 38.2). Surgical treatment of
haemorrhage was necessary in 10 patients (1.6 per cent). In
34 per cent of haemorrhages a transfusion was necessary.
Haemorrhagic complications presented as an obvious
expanding haematoma in all except two patients: one with
a rectus abdominis sheath haematoma and the other with a
retroperitoneal haematoma, respectively. There was no rise
in mortality related to complications at the puncture site.
Figure 38.2 Ultrasound image of pseudoaneurysm of the
common femoral artery after percutaneous transluminal coronary
angioplasty. The neck of the pseudoaneurysm and the common
femoral artery are demonstrated CLINICAL PRESENTATION

to the brachial plexus. These procedures, as with catheteri- Major haemorrhage will cause local or general symp-
sation of other peripheral arteries, can be done either by toms. The most common local symptom/sign is at the
the cut-down technique using an open arteriotomy or by a site of catheter insertion, namely, an expanding painful
percutaneous technique more commonly used in recent haematoma. In the axillary artery such a haematoma may
years. The risk of brachial artery occlusion is high, what- cause brachial plexus compression with paraesthesia, par-
ever technique is used for brachial artery catheterisation. esis, or paralysis of the arm, and in the carotid area, com-
This is probably due to the fact that the artery is narrow pression of the trachea and oedema of the larynx. Major
and that the upper limb arteries are more liable to spasm bleeding may cause general symptoms of hypovolaemic
than those of the lower limb. The exact incidence of brachial shock. A retroperitoneal haemorrhage is particularly
artery occlusion is unknown, since many patients remain treacherous, because it does not usually reveal itself by
asymptomatic. Gangrene of the fingers is rare, and is prob- local symptoms (see Fig. 38.1). Computed tomography
ably most often due to peripheral embolic occlusions. (CT) scanning is essential in diagnosing retroperitoneal
In operating rooms and intensive care units radial artery haematoma.
cannulation is often used to permit repeated arterial blood Thrombosis at the place of catheterisation may cause
sampling and pressure recording. The cannulation can be peripheral ischaemia. In common femoral artery occlusion
performed either percutaneously or via an open arteri- there is no pulsation in the groin and peripheral ischaemia
otomy. The catheter is often allowed to remain in place for is severe. In superficial femoral artery occlusion pulsation is
several days, contrary to most other peripheral artery felt in the common femoral artery, but not distally, and
catheters. Thrombosis in the radial artery, noted as a disap- ischaemic symptoms are more modest. The thrombus may
pearance of the radial pulse, occurs frequently but clinical not always be occlusive. Progression of the thrombotic
symptoms of persisting finger ischaemia are rare. In the process will alter clinical symptoms and signs within hours
majority of patients the thrombosis disappears after removal of catheterisation. Repeated investigations are therefore
of the catheter. Nevertheless, persistent finger or hand important. Symptoms are basically characterised by the five
ischaemia and even gangrene after radial artery cannulation, Ps: pain, paraesthesia, pallor, pulselessness and paralysis.
requiring amputation of the hand or lower forearm has been Emboli to the lower leg from a catheterisation site throm-
reported. It is recommended therefore, that the Allen test be bus may occlude large or medium sized arteries. Showers
carried out before the decision is made to catheterise the of microemboli may enter the small peripheral arteries and
radial artery but a negative Allen test does not exclude the cause symptoms of an apparent skin rash or the ‘blue toe
possibility of permanent damage.21 With the introduction of syndrome’.24 Catheterisation of a severely atherosclerotic
percutaneous transluminal angioplasty (PTA) the frequency aorta may cause massive embolisation or aortic occlusion,
of vascular complications seems to have increased, especially a complication which may be fatal.24
after coronary angioplasty when the haemostatic systems are Arterial occlusion after catheterisation of arteries of the
often heavily influenced by pharmacological means.8,22,23 upper extremity usually causes only mild symptoms
Pseudoaneurysm formation is the dominating complication because of the presence of a rich collateral circulation. Loss
of angiography. Messina et al.8 reported a 3.4 per cent inci- or weakening of distal pulses is the most important symp-
dence of vascular complications after interventional cardiac tom. Occasionally, pallor and some paraesthesia are evi-
catheterisation and 0.7 per cent after diagnostic procedures. dent, but paresis or pain is rare. Some patients do develop
462 Injuries of arterial catheterisation

symptoms of ‘arm claudication’ later. Gangrene of fingers pseudoaneurysm expands with the pulse wave, can be felt
or hands is a rare occurrence. by bimanual palpation and, furthermore, a machinery
Thromboembolic complications in arteries other than murmur can be heard. Ultrasound imaging will reveal the
those of the extremity may cause stroke or blindness, extent of the pseudoaneurysm (see Fig. 38.2) and the typi-
myocardial infarction, renal ischaemia with infarction, ren- cal flow profile in its neck (Fig. 38.3). Rupture is rare25 but
ovascular hypertension or intestinal infarction. Most of the the skin circulation may be compromised.26
thromboembolic complications appear while the patient is
still being catheterised or soon after catheter withdrawal.

Clinical features suggestive of catheter PREVENTION OF CATHETER INJURY

injury
Several measures can be taken to reduce the risk of arterial
• Bleeding injury. Angiography should be performed in hospitals with
• Haematoma a reasonable workload of such procedures, because the
• Ischaemia number of complications decreases with increasing experi-
• Microembolisation – ‘blue toe syndrome’, ence. Transfemoral catheterisation should be the route of
cerebral/ocular symptoms, intestinal/renal choice. The catheterisation of smaller arteries increases the
ischaemia or infarction risk for various types of injury, as does selective catheteri-
• Pulsatile swelling/thrill/bruit sation. The translumbar route is hardly ever indicated.
When arterial catheterisation is difficult or expected to be
Arteriovenous fistulae rarely have haemodynamic con- so it is best abandoned or indeed avoided altogether. In
sequences. They are diagnosed accidentally when a mur- these cases other diagnostic methods such as magnetic res-
mur or a thrill is detected. An arteriovenous fistula between onance angiography (MRA) or duplex ultrasonography
the vertebral artery and Batson’s vein plexus surrounding the should be used. The time taken for catheterisation should
artery manifests itself as a murmur, which is heard by the be short and the catheter used of small calibre. Dextran
patient and may be thought to be tinnitus. before and heparin during catheterisation together seem to
A pseudoaneurysm may form at the point of cathe- decrease the risk of thromboembolism.
terisation and on palpation is felt to be an expanding To diminish the risk of haemorrhage and later of
tumour. It may be difficult to distinguish from a small haema- pseudoaneurysm formation many angiographers use some
toma. A pseudoaneurysm develops after the haematoma form of compression device such as FemoStop (Radi Medical
has been established. In contrast to the haematoma, a Systems, Uppsala, Sweden). Another way of decreasing this

Figure 38.3 Ultrasound image of pseudoaneurysm of the common femoral artery with spectral Doppler recording. The typical ‘to and fro’
flow profile in the neck of the pseudoaneurysm is demonstrated
 Management of complications 463

risk is by the use of a tissue sealant achieved by the depos- and retained devices have been described.34,35 In a recent
ition of collagen in the extravascular puncture canal,27–29 report on the use of 3000 percutaneous suturing devices eight
although a potential complication is thrombosis when it is complications were recorded (0.03 per cent):36 thrombosis,2
accidentally injected into the vessel lumen. haemorrhage (four, with one infection) and retained devices.2
Percutaneous suturing devices to close arterial access Pathways of care are simplified in algorithms for the
puncture sites may prevent bleeding and allow the patient to detection and management of haemorrhage (Fig. 38.4) and
be mobilised more rapidly.30–33 The Food and Drug of pseudoaneurysms (Fig. 38.5).
Administration (FDA) has currently approved two suturing
devices: Prostar/Techstar and Closer. These devices may be
used with a low complication rate but bleeding, thrombosis Factors which reduce risk of catheter
injuries

• Angiographic competence
Groin arterial puncture • Preferential use of the transfemoral route
Compression • Use of small calibre catheters
The patient does not feel well, deteriorating
• Dextran and heparin to reduce thromboembolism
• Use of FemoStop®, tissue sealant, percutaneous
suture
Bleeding? No

If unstable Yes Unclear

MANAGEMENT OF COMPLICATIONS
Compression

Suspect retroperitoneal haemorrhage

Early detection of complications is the key to successful
handling. This is particularly important in cases of throm-
boembolic occlusion. The circulation in the extremities
Strong clinic or CT No should always be checked before angiography, immedi-
ately after removal of the catheter, and before the patient is
Open surgery Yes allowed to leave the hospital. Impaired circulation after
catheterisation is often ascribed to ‘spasm’, but signs of
Figure 38.4 Algorithm of the detection and management of peripheral ischaemia persisting after removal of the
haemorrhage catheter should never be ascribed to spasm until arterial

Clinical suspicion of pseudoaneurysm

after arterial puncture in the groin

With rupture
expanding haematoma Duplex ultrasonography No pseudoaneurysm
retroperitoneal

Pseudoaneurysm Repeat No further follow-up

2 cm 2 cm
diameter diameter

Local thrombin
 compression

No occlusion Occlusion

Open surgery Duplex day 1

Figure 38.5 Algorithm of the

detection and management of
No occlusion Occlusion
pseudoaneurysm
464 Injuries of arterial catheterisation

thromboembolism has been excluded by clinical means, thereby also preventing secondary formation of pseudoa-
repeat angiography or preferably duplex scan. neurysms. In case of bleeding preliminary haemostasis can
usually be obtained by upstream balloon occlusion.40
Open surgery
Non-surgical treatment of pseudoaneurysms
When exploring the femoral artery it is important to make
the skin incision proximal enough to allow for control Ultrasound guided compression has been successfully used
above the hole created by the needle or catheter.37 The hole since the early 1990s.26,42–46 One drawback is the need for
is sometimes quite large and lacerated and closure some- prolonged compression, lasting 30 minutes to more than
times requires a patch. The use of a Fogarty catheter is an hour, which may be painful for the patient and demand-
essential in clearing the distal vascular bed. The peripheral ing for the clinician.45,47–49 A modification with the
vasculature must be evaluated by angiography or duplex FemoStop compression device to occlude the pseudoa-
scan while the patient is on the operating table. neurysm has been advocated as a substitute for manual
The handling of occlusive injuries of the brachial artery compression.50,51 There is a higher failure rate in patients
after catheterisation (see Chapter 41) is more controver- who are anticoagulated, those with pseudoaneurysms
sial. Most asymptomatic patients will remain in that state, larger than 4 cm26,44,45 and in pseudoaneurysms associated
particularly if they are not active, but as a general rule these with an arteriovenous fistula.52
injuries should be repaired. Since the report by Cope and Zeit53 injecting thrombin
Stroke may be caused by carotid artery occlusion from into the aneurysmal sac, the combination of percutaneous
thrombosis, dissection or embolism. In the former case, thrombin induced thrombosis and ultrasonographically
immediate exploration and revascularisation may restore guided compression has been increasingly used54–61 (Fig.
cerebral function6 (see Chapters 5 and 11–14). Haemorrhage 38.6). Thrombin is slowly injected into the aneurysmal sac,
should be handled by early diagnosis and early exploration. preferably from the periphery and under ultrasonographic
In cases of carotid artery puncture it is important to control. The amount of thrombin used is between 100 and
observe the patient in order to be aware of the development 500 units and thrombosis occurs within seconds compared
of laryngeal and tracheal compression by a haematoma. with half an hour or more during compression alone.62 The
Early exploration and tracheotomy are advisable when rapidity of occlusion has significantly reduced the burden
there are signs of compression. Early exploration is neces- on vascular laboratory resources. Although thrombosis of
sary when bleeding is suspected after axillary artery the native artery has been described,57,63,64 this complica-
catheterisation. The patient often experiences some numb- tion seems to be extremely rare. To diminish the risk,
ness or weakness in the hand after removal of the axillary injection of small amounts of diluted thrombin solutions
catheter. Persistence of symptoms or new symptoms such has been recommended.62 Another rare complication is
as tingling, pain, decreased sensation, numbness or weak- allergy and even anaphylaxis after injection of bovine
ness indicate continued bleeding from the puncture site. thrombin.65,66 Also, if topical thrombin has been used pre-
Immediate exploration is mandatory. Haemorrhage follow- viously, i.e. during haemodialysis, IgE mediated anaphyl-
ing arterial puncture is usually easy to treat by simple axis may be induced.67 The success rate in occluding a
suture. Occasionally, a patch is necessary.
Most pseudoaneurysms are detected within one to a few
weeks after catheterisation when the haematoma has dis-
appeared. A pseudoaneurysm can thrombose and disap-
pear spontaneously, but it may also rupture. It is easy to close
the hole with one or two longitudinal stitches. Pseudoane-
urysms of the axillary artery may be more difficult to treat
surgically but early exploration is also very important. Great
care has to be taken to avoid injuring the nerve plexus
around the artery, but the neurovascular sheath must be
opened to avoid pressure on the nerves.

The use of endografts

Today, catheter or balloon induced lesions such as perfor-

ation, dissection, pseudoaneurysm and rupture or arteriove- Figure 38.6 Ultrasound image of pseudoaneurysm of the
nous fistula may be safely and effectively treated by insertion common femoral artery following injection of 1000 units of
of a covered endoprosthesis, i.e. a combination of stent and thrombin. Visible thrombus formation and the absence of colour
graft, to exclude and seal the entrance of the lesion,38–41 confirm there is no flow in the pseudoaneurysm
 References 465

pseudoaneurysm by thrombin injection is near 100 per 3 Picus D, Totty WG. Iatrogenic femoral arteriovenous fistulae:
cent. If unsuccessful another percutaneous option is the evaluation by digital vascular imaging. AJR Am J Roentgenol
delivery of coils into the aneurysmal sac.68 This method is 1984; 142: 567–70.
limited by the potential of introducing infection in the 4 Heystraten FM, Fast JH. Arteriovenous fistula as a complication
of the Seldinger procedure of the femoral artery and vein. Report
process of placing foreign material within a groin
of 2 cases. Diagn Imag 1983; 52: 197–201.
haematoma. 5 Kron J, Sutherland D, Rosch J, et al. Arteriovenous fistula: a rare
A recent approach, which has yet to be evaluated, is the complication of arterial puncture for cardiac catheterization.
local administration of recombinant activated factor VII.69 Am J Cardiol 1985; 55: 1445–6.
6 Bergentz SE, Hansson LO, Norback B. Surgical management of
complications to arterial puncture. Ann Surg 1966; 164: 1021–6.
Conclusions 7 Eriksson I, Jorulf H. Surgical complications associated with
arterial catheterization. Scand J Thorac Cardiovasc Surg 1970; 4:
Arterial catheterisation is an invasive procedure carrying 69–75.
the risk of bleeding, occlusion by various means, throm- 8 Messina LM, Brothers TE, Wakefield TW, et al. Clinical
boembolism, pseudoaneurysm, arteriovenous fistula characteristics and surgical management of vascular complications
and the added possibility that a foreign body may be left in patients undergoing cardiac catheterization: interventional
versus diagnostic procedures. J Vasc Surg 1991; 13: 593–600.
behind in the vessel. The site chosen for puncture influ-
9 Allen BT, Munn JS, Stevens SL, et al. Selective non-operative
ences the type and the incidence of complications, most
management of pseudoaneurysms and arteriovenous fistulae
of which become apparent during the procedure or soon complicating femoral artery catheterization. J Cardiovasc Surg
after catheter withdrawal. The vascular surgeon and the (Torino) 1992; 33: 440–7.
radiologist must be aware of the risks and should be pre- 10 Rivers SP, Lee ES, Lyon RT, et al. Successful conservative
pared to correct any injury inflicted, while those in other management of iatrogenic femoral arterial trauma. Ann Vasc
disciplines calling for these interventions should ensure Surg 1992; 6: 45–9.
that a vascular team is available on standby. 11 Kotval PS, Khoury A, Shah PM, Babu SC. Doppler sonographic
demonstration of the progressive spontaneous thrombosis of
pseudoaneurysms. J Ultrasound Med 1990; 9: 185–90.
12 Kent KC, McArdle CR, Kennedy B, et al. A prospective study of
Key references the clinical outcome of femoral pseudoaneurysms and
arteriovenous fistulas induced by arterial puncture. J Vasc Surg
Bergentz S-E, Bergqvist D. Iatrogenic Vascular Injuries. Berlin: 1993; 17: 125–31; discussion 131–3.
Springer Verlag, 1989. Overview of all types of iatrogenic 13 Paulson EK, Hertzberg BS, Paine SS, Carroll BA. Femoral artery
injuries but not the recent therapeutic options. pseudoaneurysms: value of color Doppler sonography in
Ettles D, Earnshaw J. Complications in interventional radiology and predicting which ones will thrombose without treatment.
thrombolysis. In: Campbell B (ed). Complications in Arterial AJR Am J Roentgenol 1992; 159: 1077–81.
Surgery. A Practical Approach to Management. London: 14 Samuels D, Orron DE, Kessler A, et al. Femoral artery
Butterworth-Heinemann, 1996. Practical guidelines. pseudoaneurysm: Doppler sonographic features predictive for
Kang S, Labropoulus N. Nonoperative treatment of femoral spontaneous thrombosis. J Clin Ultrasound 1997; 25: 497–500.
pseudoaneurysms. In: Yao J, Pearce W (eds). Practical Vascular 15 Crawford E, Beall A, Moyer J, DeBakey M. Complications of
Surgery. New York: Appleton & Lange, 1999. Details on aortography. Surg Gynecol Obstet 1957; 104: 129–41.
ultrasound-guided treatment. 16 Reiss MD, Bookstein JJ, Bleifer KH. Radiologic aspects of
Messina LM, Brothers TE, Wakefield TW, et al. Clinical renovascular hypertension. 4. Arteriographic complications.
characteristics and surgical management of vascular JAMA 1972; 221: 375–8.
complications in patients undergoing cardiac catheterization: 17 Lang E. Complications of retrograde percutaneous arteriography.
interventional versus diagnostic procedures. J Vasc Surg 1991; J Urol 1963; 90: 604–10.
13: 593–600. 18 Lang EK. Complications of direct and indirect angiography of the
Takolander R, Bergqvist D, Jonsson K, et al. Fatal thrombo-embolic brachiocephalic vessels. Acta Radiol Diagn 1966; 5: 296–307.
complications at aorto-femoral angiography. Acta Radiol 19 Hessel SJ, Adams DF, Abrams HL. Complications of angiography.
Diagn (Stockh) 1985; 26: 15–19. Radiology 1981; 138: 273–81.
20 Babu SC, Piccorelli GO, Shah PM, et al. Incidence and results of
arterial complications among 16 350 patients undergoing cardiac
catheterization. J Vasc Surg 1989; 10: 113–16.
REFERENCES 21 Wilkins RG. Radial artery cannulation and ischaemic damage: a
review. Anaesthesia 1985; 40: 896–9.
1 Antonovic R, Rosch J, Dotter CT. The value of systemic arterial 22 Oweida SW, Roubin GS, Smith RB 3rd, Salam AA.
heparinization in transfemoral angiography: a prospective study. Postcatheterization vascular complications associated with
Am J Roentgenol 1976; 127: 223–5. percutaneous transluminal coronary angioplasty. J Vasc Surg
2 Bergstrom K, Lodin H. Arteriovenous fistula as a complication 1990; 12: 310–15.
of cerebral angiography. Report of three cases. Br J Radiol 1966; 23 McCann RL, Schwartz LB, Pieper KS. Vascular complications of
39: 263–6. cardiac catheterization. J Vasc Surg 1991; 14: 375–81.
466 Injuries of arterial catheterisation

24 Takolander R, Bergqvist D, Jonsson K, et al. Fatal thrombo- 42 Fellmeth BD, Roberts AC, Bookstein JJ, et al. Postangiographic
embolic complications at aorto-femoral angiography. Acta Radiol femoral artery injuries: nonsurgical repair with US-guided
Diagn (Stockh) 1985; 26: 15–19. compression. Radiology 1991; 178: 671–5.
25 Graham AN, Wilson CM, Hood JM, Barros D’Sa AA. Risk of 43 Cox GS, Young JR, Gray BR, et al. Ultrasound-guided compression
rupture of postangiographic femoral false aneurysm. Br J Surg repair of postcatheterization pseudoaneurysms: results of
1992; 79: 1022–5. treatment in one hundred cases. J Vasc Surg 1994; 19: 683–6.
26 Feld R, Patton GM, Carabasi RA, et al. Treatment of iatrogenic 44 Hajarizadeh H, LaRosa CR, Cardullo P, et al. Ultra-sound guided
femoral artery injuries with ultrasound-guided compression. compression of iatrogenic femoral pseudoaneurysm failure,
J Vasc Surg 1992; 16: 832–40. recurrence, and long-term results. J Vasc Surg 1995; 22: 425–30;
27 Camenzind E, Grossholz M, Urban P, et al. Collagen application discussion 430–3.
versus manual compression: a prospective randomized trial for 45 Coley BD, Roberts AC, Fellmeth BD, et al. Postangiographic
arterial puncture site closure after coronary angioplasty. J Am femoral artery pseudoaneurysms: further experience with
Coll Cardiol 1994; 24: 655–62. US-guided compression repair. Radiology 1995; 194: 307–11.
28 Sanborn T, Gibbs H, Brinker J, et al. A multicenter randomized 46 Schaub F, Theiss W, Heinz M, et al. New aspects in ultrasound-
trial comparing a percutaneous collagen hemostasis device with guided compression repair of postcatheterization femoral artery
conventional manual compression after diagnostic angiography injuries. Circulation 1994; 90: 1861–5.
and angioplasty. J Am Coll Cardiol 1993; 22: 1273–9. 47 Agarwal R, Agrawal SK, Roubin GS, et al. Clinically guided
29 Henry M, Amor M, Allaoui M, Tricoche O. A new access site closure of femoral arterial pseudoaneurysms complicating
management tool: the Angio-Seal hemostatic puncture closure cardiac catheterization and coronary angioplasty. Cathet
device. J Endovasc Surg 1995; 2: 289–96. Cardiovasc Diagn 1993; 30: 96–100.
30 Aker UT, Kensey KR, Heuser RR, et al. Immediate arterial 48 Moote DJ, Hilborn MD, Harris KA, et al. Postarteriographic
hemostasis after cardiac catheterization: initial experience with femoral pseudoaneurysms: treatment with ultrasound- guided
a new puncture closure device. Cathet Cardiovasc Diagn 1994; compression. Ann Vasc Surg 1994; 8: 325–31.
31: 228–32. 49 Khoury M, Batra S, Berg R, Rama K. Duplex-guided compression
31 Gerckens U, Cattelaens N, Lampe EG, Grube E. Management of of iatrogenic femoral artery pseudoaneurysms. Am Surg 1994;
arterial puncture site after catheterization procedures: 60: 234–6; discussion 236–7.
evaluating a suture-mediated closure device. Am J Cardiol 1999; 50 Dangas G, Mehran R, Duvvuri S, et al. Use of a pneumatic
83: 1658–63. compression system (FemoStop) as a treatment option for
32 Baim DS, Knopf WD, Hinohara T, et al. Suture-mediated closure femoral artery pseudoaneurysms after percutaneous cardiac
of the femoral access site after cardiac catheterization: results of procedures. Cathet Cardiovasc Diagn 1996; 39: 138–42.
the suture to ambulate and discharge (STAND I and STAND II) 51 Trertola SO, Savader SJ, Prescott CA, Osterman FA, Jr. US-guided
trials. Am J Cardiol 2000; 85: 864–9. pseudoaneurysm repair with a compression device. Radiology
33 Carere RG, Webb JG, Ahmed T, Dodek AA. Initial experience using 1993; 189: 285–6.
Prostar: a new device for percutaneous suture-mediated closure 52 Kumins NH, Landau DS, Montalvo J, et al. Expanded
of arterial puncture sites. Cathet Cardiovasc Diagn 1996; 37: indications for the treatment of postcatheterization femoral
367–72. pseudoaneurysms with ultrasound-guided compression.
34 Eidt JF, Habibipour S, Saucedo JF, et al. Surgical complications Am J Surg 1998; 176: 131–6.
from hemostatic puncture closure devices. Am J Surg 1999; 178: 53 Cope C, Zeit R. Coagulation of aneurysms by direct
511–16. percutaneous thrombin injection. AJR Am J Roentgenol 1986;
35 Gonze MD, Sternbergh WC 3rd, Salartash K, Money SR. 147: 383–7.
Complications associated with percutaneous closure devices. 54 Walker TG, Geller SC, Brewster DC. Transcatheter occlusion
Am J Surg 1999; 178: 209–11. of a profunda femoral artery pseudoaneurysm using thrombin.
36 Nehler MR, Lawrence WA, Whitehill TA, et al. Iatrogenic vascular AJR Am J Roentgenol 1987; 149: 185–6.
injuries from percutaneous vascular suturing devices. J Vasc Surg 55 Wixon CL, Philpott JM, Bogey WM Jr, Powell CS. Duplex-directed
2001; 33: 943–7. thrombin injection as a method to treat femoral artery
37 Rutherford R, Pearce W. Acute problems following diagnostic pseudoaneurysms. J Am Coll Surg 1998; 187: 464–6.
and interventional radiological procedures. New York: Grune and 56 Liau CS, Ho FM, Chen MF, Lee YT. Treatment of iatrogenic
Stratton, 1987. femoral artery pseudoaneurysm with percutaneous thrombin
38 Nyman U, Uher P, Lindh M, et al. Stent-graft treatment of injection. J Vasc Surg 1997; 26: 18–23.
iatrogenic iliac artery perforations: report of three cases. 57 Kang SS, Labropoulos N, Mansour MA, Baker WH. Percutaneous
Eur J Vasc Endovasc Surg 1999; 17: 259–63. ultrasound guided thrombin injection: a new method for treating
39 Formichi M, Raybaud G, Benichou H, Ciosi G. Rupture of the postcatheterization femoral pseudoaneurysms. J Vasc Surg 1998;
external iliac artery during balloon angioplasty: endovascular 27: 1032–8.
treatment using a covered stent. J Endovasc Surg 1998; 5: 37–41. 58 Elford J, Burrell C, Roobottom C. Ultrasound guided percutaneous
40 Scheinert D, Ludwig J, Steinkamp HJ, et al. Treatment of thrombin injection for the treatment of iatrogenic
catheter-induced iliac artery injuries with self-expanding pseudoaneurysms. Heart 1999; 82: 526–7.
endografts. J Endovasc Ther 2000; 7: 213–20. 59 Brophy DP, Sheiman RG, Amatulle P, Akbari CM. Iatrogenic
41 Thalhammer C, Kirchherr AS, Uhlich F, et al. Postcatheterization femoral pseudoaneurysms: thrombin injection after failed
pseudoaneurysms and arteriovenous fistulas: repair with US-guided compression. Radiology 2000; 214: 278–82.
percutaneous implantation of endovascular covered stents. 60 Vermeulen EG, Umans U, Rijbroek A, Rauwerda JA. Percutaneous
Radiology 2000; 214: 127–31. duplex-guided thrombin injection for treatment of iatrogenic
 References 467

femoral artery pseudoaneurysms. Eur J Vasc Endovasc Surg 2000; 66 Sheldon PJ, Oglevie SB, Kaplan LA. Prolonged generalized
20: 302–4. urticarial reaction after percutaneous thrombin injection for
61 Tamim WZ, Arbid EJ, Andrews LS, Arous EJ. Percutaneous induced treatment of a femoral artery pseudoaneurysm. J Vasc Interven
thrombosis of iatrogenic femoral pseudoaneurysms following Radiol 2000; 11: 759–61.
catheterization. Ann Vasc Surg 2000; 14: 254–9. 67 Tadokoro K, Ohtoshi T, Takafuji S, et al. Topical thrombin-induced
62 Taylor BS, Rhee RY, Muluk S, et al. Thrombin injection versus IgE-mediated anaphylaxis: RAST analysis and skin test studies.
compression of femoral artery pseudoaneurysms. J Vasc Surg J Allergy Clin Immunol 1991; 88: 620–9.
1999; 30: 1052–9. 68 Pan M, Medina A, Suarez de Lezo J, et al. Obliteration of femoral
63 Lennox A, Griffin M, Nicolaides A, Mansfield A. Regarding pseudoaneurysm complicating coronary intervention by direct
‘Percutaneous ultrasound guided thrombin injection: a new puncture and permanent or removable coil insertion. Am J
method for treating postcatheterization femoral Cardiol 1997; 80: 786–8.
pseudoaneurysms’. J Vasc Surg 1998; 28: 1120–1. 69 Liem AK, Biesma DH, Ernst SM, Schepens AA. Recombinant
64 Forbes TL, Millward SF. Femoral artery thrombosis after activated factor VII for false aneurysms in patients with
percutaneous thrombin injection of an external iliac artery normal haemostatic mechanisms. Thromb Haemost 1999;
pseudoaneurysm. J Vasc Surg 2001; 33: 1093–6. 82: 150–1.
65 Pope M, Johnston KW. Anaphylaxis after thrombin injection
of a femoral pseudoaneurysm: recommendations for prevention.
J Vasc Surg 2000; 32: 190–1.
This page intentionally left blank
 39
Injuries of Peripheral Endovascular Procedures

AMMAN BOLIA, PETER RF BELL

The problem 469 Elastic recoil/compromised collaterals 481

Puncture related complications 469 Compromise of important branches at bifurcations 483
Management of puncture related complications 470 Accidental dissection 484
Perforation 473 Other endovascular management procedures 484
Embolic complications 478 References 486
Thrombotic complications 481

THE PROBLEM acute ischaemia and the patient should always be informed
about their incidence before consent is obtained.
Since the first description of endoluminal treatment for
peripheral vascular disease by Dotter and Judkins in 1964,1 Complications of percutaneous
there has been a substantial increase in the number of patients endovascular treatments for PVD
treated using percutaneous transluminal angioplasty (PTA).
Percutaneous transluminal angioplasty in its various forms • Puncture related complications
has been used as an alternative to reconstructive surgery in • Perforation
patients with critical limb ischaemia and, because it is minim- • Embolic complications
ally invasive, the threshold for treatment of claudicants has • Thrombotic complications
been reduced. • Compromise of important branches at bifurcations
Although the treatment is successful in a large number of • Elastic recoil and compromised collaterals
patients, it is not without complications and when these • Accidental dissection
occur, many of them require urgent attention by the oper-
ator in order to deal with the situation by endovascular or if
necessary open techniques. The incidence of complications is PUNCTURE RELATED COMPLICATIONS
dependent on the experience of the operator.2,3 The majority
of these complications can be managed percutaneously at the
time of the procedure without the need for major surgery.4–11 Main puncture sites
When percutaneous management fails, however, open sur-
gery is required. The incidence of complications requiring The commonest puncture site for peripheral vascular and
surgery ranges between 1 and 4 per cent2–12 and surgical help other interventions is the common femoral artery. Other
should always be available in centres where percutaneous sites, for example axillary, high brachial, radial and popliteal
techniques are practised. arteries may also be used for access in certain situations.
The complications of percutaneous endovascular
treatments for peripheral vascular disease (PVD) can be
categorised as puncture related, perforation, embolic, Common femoral artery
thrombotic, compromise of important branches at bifur-
cations, elastic recoil and compromised collaterals as well The possible complications arising from the common
as accidental dissection. Some of these can lead to severe femoral artery (CFA) as a puncture site include groin
470 Injuries of peripheral endovascular procedures

haematoma, which is usually a self-limiting problem, but it MANAGEMENT OF PUNCTURE RELATED

may be progressive and a substantial number may require COMPLICATIONS
transfusion or surgical evacuation. Other potential compli-
cations are listed below.
Haematoma

Haematoma at the femoral puncture site is the common-

Puncture related complications est complication with frequency ranging from 2 to 8
per cent.3,13–18 Large diameter catheters, hypertension and
• Groin haematoma: self-limiting or expanding obesity are likely to increase the chances of haematoma or
• Retroperitoneal haematoma false aneurysm formation. The majority of haematomas at
• Pseudoaneurysm the common femoral artery puncture site are self-limiting
• Arteriovenous fistula and do not require any specific treatment. Occasionally,
• Thrombosis however, a large haematoma may require blood transfu-
• Femoral nerve damage sion and/or surgical evacuation and repair.19

Retroperitoneal haematoma
Various factors contribute to the development of these
puncture related complications, namely, obesity, hyper- A high puncture in the common femoral artery may be
tension, anticoagulant treatment, over-heparinisation, a ten- required to treat either a flush occlusion or disease in
dency to bleed, low platelet count, high femoral puncture the proximal superficial femoral artery. A high puncture
causing bleeding retroperitoneally, low superficial femoral increases the risk of a retroperitoneal haematoma. The inci-
artery (SFA) puncture leading to pseudoaneurysm forma- dence of this complication is less than 0.1 per cent.8
tion, the use of large diameter catheters/sheaths and proced- When a high puncture has to be made intentionally, care
ure which is prolonged. is required to puncture that part of the common femoral
artery which can be compressed against the underlying super-
ior pubic ramus. Bleeding and haematoma formation is
Factors contributing to femoral puncture likely to occur due to inadequate compression, with pressure
related complications in the wrong place, and/or for too short a time. Prolonged
and horizontal compression on the puncture site will usually
• Obesity prevent a potentially fatal retroperitoneal haematoma.
• Hypertension Patients in whom a high puncture is anticipated should have
• Patient on anticoagulant treatment/ blood taken for ‘group and save’ so that, should transfusion
over-heparinisation be required, blood can be made available quickly. Also, the
• Generalised bleeding tendency/low platelet count nursing staff must be informed that a high puncture has been
• High puncture: retroperitoneal bleeding made so that observations are made with care and more fre-
• Low puncture: pseudoaneurysm formation quently, and action taken as soon as there is any evidence of
• Large diameter catheters/sheaths significant blood loss. Should this happen, there will be a rise
in pulse rate, a fall in blood pressure and the patient may feel
faint and demonstrate features of hypotension. As soon as it
is evident that there is blood loss, immediate transfusion has
Axillary or high brachial puncture to be started and consideration given to possible surgical
repair of the puncture site.
Access via the axillary or proximal brachial artery is used More recently, closure devices of the puncture sites have
when a groin approach is difficult due to scarring or infec- become available, methods which may be desirable in cre-
tion or when femoral pulses are impalpable and iliac occlu- ating a seal particularly in patients with hypertension or on
sive disease is suspected. The most serious complication anticoagulant treatment.
from an axillary or high brachial puncture is damage to the
brachial plexus, affecting function in the ipsilateral arm and
hand. This may lead to permanent damage or to recovery Prevention of retroperitoneal haematoma
over a period of time. Damage to the nerves can be due to and its effects
direct injury or a haematoma at the puncture site com-
pressing the brachial plexus. Direct injury usually affects • Blood for ‘group and save’
single nerves whereas compression from a haematoma may • Choose correct site for puncture, i.e. not too high
affect multiple nerves. • Maintain adequate pressure and for long enough
 Management of puncture related complications 471

immediately following the injection of thrombin helps to

• Consider using a closure device
prevent blood flowing back into the pseudoaneurysm, thus
• Lower blood pressure if hypertensive
producing more effective obliteration of the aneurysm sac.
• Inform nursing staff to keep careful observations
The use of an inflated balloon catheter across the neck of the
pseudoaneurysm, introduced from the contralateral side, has
been advocated to isolate the pseudoaneurysm and prevent
Surgical treatment of bleeding or haematoma thrombin overflowing into the systemic circulation. This
practice, however, is probably unnecessary.
Groin haematomas, once false aneurysm has been excluded
by duplex, are usually left alone and will settle spontaneously.
Patients may benefit from a short course of antibiotics to Surgical treatment of pseudoaneurysms
avoid infection. If, however, the haematoma is large then a
small incision under local anaesthesia can be used to evacu- If pressure or the injection of thrombin fails, surgical treat-
ate it and avoid risk of infection. Should the patient present ment is necessary. By the time such an operation takes place
with obvious and continuing retroperitoneal bleeding from a the groin is usually fairly indurated with thrombus and the
high puncture, blood transfusion becomes necessary and operation can be difficult. It is best to undertake the proced-
surgical intervention should be considered early rather than ure under general anaesthesia if the patient is fit. The first
late. This is a potentially fatal complication. If, in spite of move should be to gain control of the artery proximal to the
pressure and the replacement of blood, the patient remains aneurysm. This is achieved through a small incision above
haemodynamically unstable then surgical repair is indicated. the inguinal ligament and gaining access to the iliac artery
Using local or general anaesthetic, an incision is made above which can then be compressed with the finger or encircled
the inguinal ligament and the iliac artery approached with an appropriate sling for control. An incision is then
retroperitoneally through the abdominal musculature to made directly over the pseudoaneurysm, the sac opened,
expose the puncture site and arrest the haemorrhage with a the clot evacuated if possible and the artery approached.
5-0 or 6-0 Prolene suture. Bleeding can be controlled by compression or clamping of
the iliac artery and retrograde bleeding is usually dealt with
by suction as dissecting out all the branches of the artery is
False aneurysm (pseudoaneurysm) unnecessary and usually difficult. A single 5-0 suture placed
across the needle puncture site in two directions will usually
The incidence of pseudoaneurysm following common stop the bleeding. It is important to get down to the artery
femoral artery puncture ranges from 0.2 to 0.5 per cent for wall and not simply put the suture into superficial layers as
diagnostic procedures and up to 5–8 per cent for coronary this will not usually work. The wound is closed with
stenting procedures.2,21 This is likely to occur when there drainage and a course of antibiotics is given.
has been inadequate compression at the puncture site
and particularly when the puncture has been too low, for
example in the superficial femoral artery. A puncture in Thrombosis at puncture site
the superficial femoral or profunda femoris artery cannot
be adequately compressed as there is no bone immediately Thrombosis may occur at the puncture site particularly when
behind them at this level. the catheter/sheath has been passed through a stenotic seg-
A false aneurysm presents itself as a pulsatile mass at the ment and is compromising flow through it. Prolonged pro-
puncture site, which may be quite tender. A duplex scan is cedures in such a situation, despite heparinisation, may result
the easiest way of diagnosing this condition, although com- in a thrombotic complication and surgery may be required to
puted tomography (CT) and magnetic resonance imaging treat it.
(MRI) are equally good at establishing the size, shape and
extent of the pseudoaneurysm. Ultrasound compression can
be effective but it is painful and time consuming, and the Possible sequelae of untreated
operator has to apply prolonged pressure using the ultra- pseudoaneurysm
sound probe to stop the pseudoaneurysm from filling.20,22,23
Nowadays, the majority of pseudoaneurysms can be man-
aged by the injection of thrombin, in a 0.5–1 mL of 1 in
• Haemorrhage

500 units/mL concentration, through a 22 gauge needle

• Infection

directly into the pseudoaneurysm.24 This has the effect of

• Rupture

thrombosing blood within the pseudoaneurysm within sec-

• Peripheral embolisation

onds. The effect of the treatment can be assessed in real time

• Deep vein thrombosis

using duplex ultrasound. Some pressure on the probe

• Pressure on adjacent nerves
472 Injuries of peripheral endovascular procedures

Surgical treatment of thrombosis allow removal of the thrombus and assessment of the reason
for its occurrence. It may be that a plaque has dissected, in
Thrombosis occurs at the site of a puncture or catheter which case it needs to be sewn back into place and the arte-
insertion in which the limb becomes acutely ischaemic. The riotomy closed with a small patch of vein taken from the
artery should be exposed, using an appropriate incision and adjacent area. Alternatively, a small piece of Dacron may be
the vessel controlled with slings above and below the punc- used, under antibiotic cover.
ture site. A longitudinal incision into the artery will then

(a)

(b)

Figure 39.1 (a) When a perforation occurs with a catheter/guidewire, the combination can be withdrawn out of the perforation site
and the wire manipulated to find an alternative dissection. (b) An alternative method of avoiding the perforated site is to manipulate
the guidewire into a large loop so that its diameter, being larger than that of the hole at the site of perforation, will not allow it to enter
the hole, so preventing dissection along the artery
 Perforation 473

It may be better, however, to remember the original prob- either guidewire/catheter manipulations or inflation of the
lem with which the patient presented and deal with it defini- balloon catheter.
tively by an operation using a graft taken from above the area
of trauma to the first patent artery distally. If thrombosis
Perforation due to guidewire/catheter
occurs during a procedure undertaken for critical limb
manipulations
ischaemia it would be better to resort to a bypass graft rather
than try to remove the thrombus. If thrombosis has occurred
Perforation is more likely when there is calcification in the
in a healthy vessel being used for access, then local treatment
artery or when the occlusion is hard and the wire/catheter
is usually possible in the way described above. If the limb is
combination cannot penetrate the occluded segment, instead
not acutely ischaemic, it is often adequate to heparinise the
exiting through the wall of the artery (Fig. 39.1a). Crossing an
patient and wait and see what happens to the limb under
occlusion with a wire/catheter is usually a painless manoeu-
careful observation. In most cases the circulation will be ade-
vre. Therefore, if the patient experiences sudden pain during
quate for limb recovery and is dealt with electively rather
advancement of the wire/catheter, perforation is likely to
than as an emergency procedure. Any sign of deterioration,
have occurred. Every time the wire enters the perforation, the
however, should lead to an emergency bypass.
patient feels pain and this represents useful ‘feedback’ when
managing the perforation. During manipulations to find an
alternative path of dissection, this information tells the oper-
PERFORATION
ator to come out of the perforation and find an alternative
route of recanalisation (Figure 39.1b).
There are two ways in which a perforation can occur dur- When such a perforation occurs in the middle of an
ing an endovascular procedure. It may occur as a result of occlusion, particularly during subintimal angioplasty, an

Figure 39.2 A 10 cm occlusion of the popliteal artery is present. A perforation occurred during attempts at recanalisation and contrast
can be seen within the tissues outside the artery. Subsequently, an alternative dissection was found and successful recanalisation achieved
474 Injuries of peripheral endovascular procedures

(a)

Figure 39.3 (a) A 10 cm occlusion of the distal superficial femoral artery is present. During attempted recanalisation, a large perforation
occurred with substantial extravasation into the tissues. The patient was hypertensive and complained of a significant amount of discomfort.
Hence a 3 mm 1 cm embolisation coil was placed.

alternative dissection plane can be found in the majority of temporary balloon tamponade, fails, the vessel should be
the cases, and the situation retrieved so that the perforation exposed and the perforation closed with a single 5-0
becomes of no consequence (Figs 39.1b and 39.2). In a small Prolene stitch.
proportion of cases, however, the perforation is so substan-
tial that an alternative dissection plane is difficult to find and
the procedure has to be abandoned. If a significant amount Perforation due to balloon dilatation
of pain is felt due to extravasation of blood, particularly in
hypertensive patients, then one may have to consider coil The incidence of arterial rupture due to balloon dilatation is
embolisation (Fig. 39.3a). This usually involves placing a approximately 0.1 per cent.25–27 Perforation is diagnosed
5 mm 5 cm 0.035 inches coil, at the site of the perforation when, following balloon inflation, the patient feels a sharp
or just above it. After a few weeks, when the perforation has and significant amount of pain which does not recede despite
healed, another attempt can be made to recanalise the occlu- balloon deflation. This may be followed by a rise in pulse rate
sion and that usually results in a successful outcome, and drop in blood pressure. The patient may look pale, cold
whether an embolisation coil is present or not (Fig. 39.3b). and clammy, feel faint and nauseous and yawning is an add-
Perforation in a stenotic segment is extremely unlikely to be itional sign of hypotension. A quick injection of contrast will
due to catheter and wire manipulations but should this confirm the occurrence of this complication so that treat-
occur, then the management would be similar to that when ment can be carried out immediately without any significant
a perforation occurs with balloon dilatation. blood loss.
This is particularly urgent in situations where the perfor-
Surgical treatment of wire/catheter ation has occurred within a cavity, e.g. the renal or iliac arter-
perforation ies within the abdomen or the subclavian artery within the
thorax. In such cases substantial blood loss can occur very
This has never been necessary in our practice but is a theo- quickly and therefore balloon tamponade must be carried
retical possibility. If the use of coils, as described above, or out immediately. A drip must be set up and measures taken
 Perforation 475

(b)

Figure 39.3 (b) Ten months later an attempt was made at recanalisation. The guidewire found a dissection channel easily and successful
recanalisation was achieved, despite the presence of an embolisation coil

to correct blood pressure mainly with replacement fluids and external iliac artery is caused by balloon dilatation, then it
plasma substitutes while emergency cross-matched blood is may be reasonable to consider re-occluding the artery to stop
awaited. Surgical help must be sought immediately and the bleeding. This type of intervention, however, requires experi-
patient taken to theatre for emergency repair. ence and speed from the operator to minimise blood loss.
The coils to be used must be opened and ready to be deliv-
ered as quickly as possible. This is achieved by locating the tip
Prevention of perforation due to balloon of the balloon catheter at the intended site of coil placement,
dilatation the coil being introduced through a partially inflated balloon
catheter. When satisfactory placement of this coil has been
• Do not use balloon of over-sized diameter achieved, the balloon catheter is withdrawn further and the
• Do not overinflate balloon tip positioned at the distal part of the site of the perforation.
• Inflate balloon gradually Once again, another coil is delivered through the same bal-
• Raise pressure with sequential inflations in tough loon catheter. If the coils are placed satisfactorily, haemostasis
lesions will be achieved within minutes of placement and no further
blood loss will occur. It must be stressed that only an experi-
When a perforation occurs due to the use of an over-sized enced operator should undertake this manoeuvre because
balloon, the generation of high pressures rather quickly or speed is important and any delay in the placement of these
balloon rupture, it is usually substantial and may cause a lon- coils will result in substantial blood loss, with possible risk of
gitudinal split in the artery. Should such a perforation occur mortality. A less experienced operator would be wise to tam-
within what was previously an occlusion, then the occlusion ponade the perforation site with a balloon and seek the help
can be re-created by coil embolisation, ideally placed at both of a surgeon immediately.
ends of the arterial split so as to isolate the perforated seg- When a perforation occurs in a previously stenotic seg-
ment. For example, if a perforation in a previously occluded ment immediate balloon tamponade is recommended. This
476 Injuries of peripheral endovascular procedures

(a)

Figure 39.4 (a) A long tight

stenosis is present in the
polyfluorotetraethylene
(PTFE)/vein composite femoro-
popliteal graft. Initial
dilatation produced an
unsatisfactory result.
(b) Further balloon inflation
with higher pressures ruptured
the balloon, causing a
perforation in the graft. Note
contrast within the tissues and
the guidewire tip outside the
graft. The perforation was
(b) successfully managed with a
Wallgraft
 Perforation 477

Perforation

Sustained pain
Hypotension/tachycardia
Signs
Pallor/cold sweat/yawning

Set up intravenous line

Inflate balloon over Transfusion
Fluid replacement
Immediate perforation site if necessary
Blood for cross-match
action (Tamponade)

For
experienced
operators

Covered stent If perforation at site of

Treatment
if available previous occlusion,
recreate occlusion by
coil embolisation
Figure 39.5 Algorithm of the
diagnosis and management of arterial
Success To theatre perforation by balloon within a body
Failure
for repair
cavity

will have the effect of reducing blood loss and allowing the and could lead to the death of the patient. Tamponade and
operator some time to think of the best way of managing surgical intervention is the safest option and the vascular
the situation. With the availability of covered stents, an surgeons should be informed as quickly as possible. An algo-
appropriate sized stent may allow the perforation to be rithm (Fig. 39.5) offers a guide to the diagnosis and manage-
sealed, at the same time establishing flow through the artery ment of balloon perforation within a body cavity.
without the need for surgery (Fig. 39.4). It should be borne in
mind, however, that such a mode of treatment is a temporary
Surgical treatment of perforations
measure in the femoral segment where stent-grafts are
unlikely to produce long term patency. In the iliac segment,
Although balloon tamponade will control the perforation in
the management of a perforation using a stent-graft is prob-
the short term, immediate surgical treatment is advisable if a
ably the ideal solution to the problem.
solution to the problem is to be found. The approach will
Placement of a stent-graft can be a tricky and prolonged
depend upon the site of the perforation and whether the
procedure putting the patient at risk of substantial blood loss
artery was patent or occluded at that point. In addition, the
and therefore it should be undertaken with great care. The
reason for the intervention will be important. For example,
introduction of a stent-graft will require the use of an appro-
if a patient complaining of claudication is having an angio-
priately sized large sheath at the ipsilateral puncture site.
plasty, it is probably sensible to simply stop the bleeding and
Introduction of this sheath involves an exchange process
deal with the problem on another day as the leg will be no
during which the tamponade balloon needs to be removed
worse, particularly if the perforation was in an occluded seg-
and bleeding continues whilst the exchange is taking place.
ment. If, however, the leg was acutely ischaemic or becom-
The stent-graft can then be introduced through the sheath
ing worse a bypass graft should be carried out.
and placed at the site of the perforation. The entire procedure
of exchange and placement of the stent-graft can take several
minutes during which substantial haemorrhage may occur Upper limb artery perforations
through the perforation. An alternative and a safer method of
management is to introduce another balloon catheter via a These are fortunately relatively rare but if they occur it is
contralateral approach to take the place of the tamponade usually because the arm was being used for the purposes of
balloon which is then safely removed from the ipsilateral access (see Chapter 7B). It will be necessary, therefore, to
puncture site. The rest of the manoeuvres in introducing the expose the artery and control and repair the perforation with
stent-graft from the ipsilateral approach can be carried out a vein patch taken from an adjacent vein. Subclavian artery
safely, but only when the stent-graft is ready for deployment perforation is a serious and life-threatening problem as
can the contralaterally inserted tamponade balloon be access to the root of the subclavian artery, which is usually
deflated and withdrawn. where the occlusion lies, is very difficult. Balloon tamponade
It cannot be stressed enough that if the operator has little is an emergency requirement and the lesion has to be
experience, putting in coils or stent-grafts can be dangerous repaired surgically. Access can be gained by carrying out
478 Injuries of peripheral endovascular procedures

either a partial or a complete sternal split and approaching

the subclavian artery as it arises from the aortic arch. If
the lesion was a stenosis the vessel can be occluded by liga-
tures and the perforation oversewn without too much detri-
ment to the patient. Any attempt to patch the lesion at this
point is dangerous. If the arm becomes ischaemic it can be
dealt with by other means, for example a crossover graft in
the neck.

Lower limb artery perforations

In the lower limb, the commonest place for perforations is

the iliac artery, partly because it is the most frequent site for
angioplasty. If the perforation is associated with bleeding and
shock, the patient should be taken to the operating theatre
rapidly and the situation resolved. In the iliac artery a tam-
ponade balloon will usually be in place and the vessel can be
easily exposed retroperitoneally through an iliac muscle-
cutting incision leading directly to the perforated artery. The
vessels above and below are controlled with clamps and the
perforation, if it has occurred in a previously patent artery,
repaired, usually with a patch of Dacron or vein taken from
the groin. If the artery is perforated in an occluded portion, it
can be oversewn. If the vessel is occluded and provided that
the patient’s condition is satisfactory, it is usually best to
place a graft from the normal artery above, such as the com-
mon iliac or the aorta, to the groin, in order to obtain symp-
tomatic improvement. If the patient is not well, however, the
vessel can simply be ligated. If the procedure is being under-
Figure 39.6 This patient had catheter manipulations within a
taken for stenosis or occlusion no deterioration will occur significantly diseased lower abdominal aorta. It resulted in
and a definitive procedure can be carried out later. If it is cholesterol embolisation. An extensive blotchy rash appeared
impossible to construct an iliofemoral graft then a crossover soon after the procedure. The picture shows typical demarcation
graft is something which should be considered provided that of the rash at waist level extending downwards
the inflow is adequate.
For femoral lesions the same rules apply. If the perforated
vessel was already occluded and the angioplasty was being diffuse and extensive disease within the abdominal aorta,
done for claudication then simply stopping the bleeding is all extra care is necessary during catheter manipulation to avoid
that is required, with a view to definitive treatment later. A cholesterol embolisation. Sometimes, and despite all care
direct incision down to the bleeding vessel will usually suffice. taken, this complication occurs with relatively minor degrees
If, however, the procedure is being performed for rest pain or of manipulation of the catheter in the abdominal aorta.
acute ischaemia then a bypass graft, as will be described later, When it does occur, the patient experiences sudden severe
is necessary (see Chapter 15). For true crural lesions the back and abdominal pain and some pain in the legs but the
same rules apply although, generally speaking, the angio- most characteristic feature is an extensive blotchy rash which
plasty would have been undertaken for rest pain or critical is very well demarcated at waist level and extends all the way
ischaemia; under these circumstances some form of bypass down to the feet in a trouser fashion appearing either soon
will usually be necessary if the leg is to be saved. after the procedure or up to a few hours later (Fig. 39.6).
Also, characteristically, peripheral pulses are present, despite
extensive ‘trash’ embolisation. ‘Trash foot’ may eventually be
manifested as distal tissue necrosis, frequently requiring
EMBOLIC COMPLICATIONS
amputation.
Treatment is usually conservative. When cholesterol
Cholesterol embolisation (microembolisation) embolisation is higher than waist level, the renal and
mesenteric arteries may be involved, in which case there is
Excessive catheter manipulation in a diseased aorta is serious morbidity and possible mortality. Some patients
thought to cause this serious complication.28 When there is require permanent renal dialysis as cholesterol embolisation
 Embolic complications 479

Figure 39.7 Following recanalisation of the superficial femoral artery, a large embolus from the angioplasty site landed at the trifurcation
and it was successfully aspirated

into the renal circulation causes progressive and often irre- Some emboli are not haemodynamically significant,
versible loss of renal function. Apart from anticoagulation affecting one among several run-off vessels supplying the
and possibly iloprost infusion there is no treatment for this foot. In this situation the embolus may be left alone, particu-
condition and mortality is high; fortunately, this complica- larly if the limb is haemodynamically stable. Other emboli,
tion occurs rarely. particularly those larger than 3 mm, may not become
‘attached’ to the aspiration catheter, despite adequate suction
being applied with the large syringe. This may be due to the
Small and large embolisation shape of the embolus which does not allow the rim of the
(macroembolisation) catheter tip to sit uniformly over it and as such, does not
allow an adequate vacuum to be created within the catheter.
The commonest sources of such emboli are the iliac arteries, If the embolus is very large, it is difficult to maintain attach-
particularly iliac occlusions. When small emboli, up to 3 mm ment of the embolus to the catheter through suction and in
in diameter, are released into the circulation beyond an any case, even when the large embolus is attached to the
angioplastied site, they may lodge at the popliteal trifurcation catheter, it may become dislodged at the level of the sheath
or beyond (Fig. 39.7). When haemodynamically significant, when attempts are being made to aspirate it. In such a case, if
the management of such emboli is, in the first instance, more than one run-off vessel is available, then a compromise
percutaneous aspiration accomplished using a large 50 mL situation may be achieved by pushing the embolus into one
syringe attached to a non-tapered 8 Fr catheter the tip of of them, i.e. ‘push and park’32 (Fig. 39.8). This is achieved
which is brought into contact with the embolus. When aspir- by pushing the embolus either with the 8 Fr non-tapered
ating through this large syringe, a vacuum is created within catheter, already there for aspiration attempts, or with an
the catheter, which has the effect of sucking directly on the inflated balloon catheter.
embolus, which then becomes ‘attached’ to the catheter tip Thrombolysis is usually ineffective in embolic compli-
and is withdrawn along with the catheter through the sheath. cations as, in our experience, the majority of emboli appear
Frequently, the embolus becomes detached in the sheath to be of hard consistency and therefore unlikely to undergo
valve and therefore it is important to check that the valve is lysis. Moreover, the length of time it takes to set up throm-
cleared of any embolic debris before being reattached to the bolysis and to get it working means that more thrombus is
sheath. The majority of emboli of this size can be aspirated likely to form while flow remains compromised before the
percutaneously without difficulty.29–31 treatment takes full effect. It is therefore preferable to
480 Injuries of peripheral endovascular procedures

Figure 39.8 A long femoropopliteal occlusion was undergoing recanalisation but during the procedure, an embolus landed in the
tibioperoneal trunk, which could not be aspirated. This embolus was therefore pushed downwards into the peroneal artery. A three-vessel
run-off was reduced to two, but this was better than leaving the embolus in the tibioperoneal trunk, which would have meant that only
the anterior tibial artery would have been patent

resolve embolic complication in the way described and

establish flow as soon as possible. wire is inserted first of all under X-ray control using
a C-arm and passed beyond the embolus. The catheter
can then be passed over the wire and the embolus
Surgery for embolic complications withdrawn without difficulty. This can be done using
local anaesthesia of the groin or under general
If the leg becomes ischaemic in spite of attempts to remove anaesthesia. A completion angiogram is essential to
the embolus by endovascular means then surgical interven- ensure that the embolus has been retrieved.
tion becomes necessary. As the angiogram is available, the • If a groin approach is not possible or fails, then it is
operating surgeon will be aware of the location of the embo- necessary to explore the popliteal artery, using a medial
lus. The embolus is usually in the lower leg in the popliteal below-knee approach, exposing the vessel in the usual
artery or trifurcation, and if it is lower than that there is way. After the vessel has been controlled proximally
usually one crural vessel still open and the foot is usually not and distally it can be opened directly over the area
acutely ischaemic. The problem can be approached in one of where there is no pulsation and the embolus removed.
two ways. If this is done the artery should be closed with a vein
patch taken from an adjacent vein. If the embolus is
in one of the crural arteries then it can be removed
• The femoral artery can be opened in the groin and with a Fogarty catheter. Completion angiogram is
an embolectomy catheter inserted to remove the again mandatory.
thrombus. This is not usually possible if it is done
blindly with an ordinary embolectomy catheter. It is The algorithm in Fig. 39.9 offers a guide to the manage-
best done with an over the wire Fogarty catheter. The ment of embolic complications.
 Elastic recoil/compromised collaterals 481

Embolic complication conservative approach in this situation until such time as the
thrombosed segment becomes organised, for example in
3–6 months time, before repeating the angioplasty.
Cholesterol embolisation Small and large emboli
(microembolisation) (macroembolisation)
Surgical treatment for thrombosed segments
Conservative Aspiration This would be exactly the same for thrombosis occurring
management Embolectomy
in relation to catheter access as described earlier. If the
thrombosis is in a stenosed vessel then a good collateral cir-
culation may be available and an operation is not necessary
Success Failure Thrombolysis except for critical ischaemia, in which event a bypass pro-
if soft clot
cedure of appropriate type is indicated. This will depend
upon the site of the occlusion. It could be iliofemoral,
Push and park crossover, femoropopliteal or femorodistal.

ELASTIC RECOIL/COMPROMISED
Success Failure
COLLATERALS
Surgery
This complication is specific to subintimal angioplasty
Figure 39.9 Algorithm of the management of embolic where important collaterals become compromised during
complications the dissection and yet adequate flow cannot be achieved
due to elastic recoil of the occluding segment (Fig. 39.10).
This results in compromised flow to the lower limb. In the
past, emergency surgery, which includes balloon embolec-
THROMBOTIC COMPLICATIONS
tomy, thrombolysis and bypass surgery would have been
required. More recently long self-expanding stents have
Thrombotic complications can occur at the site of angio- been used to resolve the situation. Most cases with com-
plasty or remote from it, particularly when flow is being promised flow due to elastic recoil can be resolved with the
obstructed by a catheter or a sheath employed as an access use of long self-expandable stents, but in a proportion of
route to some site where angioplasty is being carried out. cases, an emergency bypass operation may still be required.
For example, during renal angioplasty, the catheter/sheath A stent in the infrainguinal arteries is unlikely to have any
may be obstructing or substantially reducing flow at a dis- significant long term patency and therefore bypass surgery
eased external iliac artery site, which is being used as the may eventually become necessary. The use of a stent in
route for the procedure. Further factors such as inadequate this situation is therefore a temporary measure in order to
heparinisation, a prolonged procedure and the use of large avoid an emergency operation (see Fig. 39.10c).
catheter/sheath sizes may play a part in causing thrombo-
sis. It can also be averted if there is prior knowledge of a
lesion on the route to a distant angioplasty site. Further, Surgery for acutely ischaemic limbs following
any significant stenosis along the route can be angioplas- endovascular intervention
tied soon after the lesion has been crossed before any fur-
ther procedures are done at the intended site. In most circumstances where endovascular techniques are
Flow can be compromised at an angioplasty site where being used for treating claudication, occlusion of the artery or
there has been a previous stenosis or an occlusion that has embolisation will not result in acute ischaemia, and watching
been dilated. Such compromise of flow may occur either the limb carefully for an hour or two may allow it to recover
due to acute thrombosis, elastic recoil or the presence of an sufficiently for a surgical or further endovascular solution to
intimal flap. In all of these cases, a stent may resolve the sit- be undertaken later. If the limb is being dealt with for critical
uation. In the case of fresh thrombosis, however, throm- ischaemia, however, some form of intervention is necessary as
bolysis may be considered or, alternatively, a stent-graft the limb usually becomes worse. This only occurs in a minor-
may resolve the complication. ity, in our experience in less than 1 per cent of cases, and it
In some situations, conservative management may be represents a surgical emergency which needs to be dealt with
appropriate. For example, when there is adequate collateral quickly. In many instances the problem can be dealt with, as
circulation around a tightly stenosed segment which then described earlier in this chapter, by removal of the thrombus
occludes during attempted angioplasty, the collateral flow and oversewing the artery with a patch. Occasionally, how-
will allow the lesion to be left alone. One may consider a ever, this will not work and further measures are required,
482 Injuries of peripheral endovascular procedures

Figure 39.10 (a) Angiogram showing a full length flush occlusion of the superficial femoral artery

Figure 39.10 (b) The occlusion was

recanalised with subintimal dissection
 Compromise of important branches at bifurcations 483

Figure 39.10 (c) Very soon after recanalisation there was a slowing of flow. Elastic recoil was noted in the upper third of the superficial
femoral artery. A long Wallstent was implanted with a successful outcome. Note the ‘spiral ribbon’ appearance of the dissection channel

particularly lower down the limb in the crural arteries where If no distal vessels can be seen on the first transfemoral
subintimal angioplasty is being performed for critical angiogram, the popliteal artery should be exposed for on-
ischaemia. In this situation, as an angiogram will have been table angiography to find a vessel which runs into the foot.
done during the procedure, the surgeon will be aware of the If no vessel can be found conforming to these requirements
condition of the proximal and distal vessels. If the leg then, based on the pre-intervention angiogram, one is
becomes painful and pale and if sensation is lost or move- chosen and exposed. Again this artery is cleared by on-table
ment is reduced then urgent treatment is needed. thrombolysis and if successful will receive the bypass graft.
First of all, the proximal vessel from which the bypass This type of acute ischaemia is fortunately rare and its
takes off is exposed. This would usually be the femoral artery occurrence is a serious situation because only half of the
in the groin or perhaps in the mid-thigh. An on-table limbs affected by distal embolisation can be salvaged. It is
angiogram is then carried out to assess the situation and see vital, however, to ensure the presence of a vascular surgeon
where a patent distal vessel lies. An appropriate vessel is then during any intervention as acute deterioration cannot be
exposed for a reversed or in situ vein bypass. Before this is forecast. The correct approach is to have a vascular team
done an angiogram of the distal run-off vessels is carried out which is multiskilled and able to deal with these situations
and if it shows good run-off into the foot then a bypass is appropriately. An algorithm offering pathways in the man-
created. If however there is a poor run-off into the foot, on- agement of compromised collaterals and elastic recoil follow-
table thrombolysis is necessary. This is achieved using strep- ing attempted subintimal angioplasty is given in Fig. 39.11.
tokinase, urokinase or tissue plasminogen activator (tPA),
made up into a solution of about 100 mL and dripped slowly
through a cannula into the distal vessel over a period of 30
minutes or so (see Chapter 16). When that has been done
COMPROMISE OF IMPORTANT BRANCHES
AT BIFURCATIONS
the angiogram should be repeated to ensure that there is
now good run-off after which the bypass graft is con-
structed. This technique will usually take care of most cases Any angioplasty at a bifurcation site is at some risk of com-
of distal embolisation unless it is of the atheromatous vari- promising the adjacent artery. The kissing balloon technique
ety, in which case nothing more can be done to save the leg. for aortoiliac segment disease has been popular for a long
484 Injuries of peripheral endovascular procedures

Subintimal superficial femoral

flaps. If blood flow is thereby compromised, the manage-
artery angioplasty ment of this complication may include stenting. It has to be
stressed, however, that the treatment of atherosclerotic
plaques using a balloon catheter does involve creating splits
Successful crossing in the intima, which are a form of dissection. In other words,
dissection is a part of the process of balloon angioplasty and
only when a dissection flap becomes haemodynamically
Compromised collaterals and
significant, should treatment with a stent be considered.
failure to establish flow Pressure readings are commonly used to assess any signifi-
cant gradients and that gradient may be enhanced with the
use of adjuvant vasodilators. Such gradients can be reduced
Attempt to establish level of
to insignificance and indeed eliminated by stent insertion.
recoil with contrast injection An alternative method of improving the haemodynam-
ics in this flow-compromised situation caused by a dissec-
tion flap may be to intentionally extend the subintimal
Long self-expanding stent at dissection until re-entry into the lumen is achieved distally.
level of elastic recoil This is a viable option when the artery distally appears to
be disease free, and is a favourable factor in the attempt to
achieve re-entry. A standard technique of subintimal angio-
plasty involves manipulating a hydrophilic guidewire into
a loop which can be engaged into the dissection flap, the
Success Failure aim being to extend the dissection, and the same loop can
be used to achieve re-entry distally.

Embolectomy
OTHER ENDOVASCULAR MANAGEMENT
Surgery Thrombolysis PROCEDURES

Bypass There is a chance that materials placed inside a patient may

be misplaced, dislodged from the desired position, released
Figure 39.11 Algorithm of the management of compromised or detached prematurely from the delivery device. Utmost
collaterals and elastic recoil following attempted subintimal
care needs to be taken to prevent such an occurrence. In
superficial femoral artery angioplasty
the unlikely event that it does, however, it is important to
consider how best to manage this problem and preferably
in a minimally invasive fashion.
time and is mainly used for disease extending from the lower
The ability to retrieve a foreign body can be an important
abdominal aorta into both common iliac arteries. The same
skill for those involved in endovascular procedures. A mis-
technique, however, applies when there is a solitary stenotic
placed stent is not a rare event and when it occurs a rapid
lesion in one of the common iliac arteries, but the lesion is
solution is required. For example, renal or subclavian artery
close enough to the origin of the other common iliac artery.
stenting may result in the stent sliding off from the balloon
Using a single balloon catheter carries a risk of embolisation
during the attempted passage of the stent to its precise destin-
into the contralateral iliac artery. It is recommended, there-
ation. When a stent has slid off the balloon, an attempt can be
fore, that even in a unilateral common iliac origin stenosis,
made to place it along the route. For example, a renal stent
the kissing balloon technique be employed.
which has come off the balloon catheter may be ‘persuaded’
At bifurcations of smaller arteries, e.g. the common iliac
to enter the common or external iliac artery where it may be
bifurcation and the common femoral bifurcation, there is
inflated and placed permanently without any significant con-
some risk of compromise to one of the branches during
sequences. Sometimes, however stents embolise distally and
attempted angioplasty of a lesion in one of these arteries. In
need to be removed by open surgery.
such situations too, the kissing balloon technique may be a
One very useful device with which any endovascular oper-
means of protection.
ator must become familiar is the Amplatz goose neck snare.
It is a clever piece of instrumentation for retrieving foreign
bodies in a tubular structure, such as an artery, a vein or even
ACCIDENTAL DISSECTION
the biliary and urinary tracts (Fig. 39.12). The goose neck
snare is designed so that it has a circular loop which is set at
Occasionally simple angioplasty of a stenotic lesion may result 90 degrees to the long axis of the catheter. When the snare is
in an extensive dissection, causing significant dissection run along a tubular structure, the loop will engage any loose
 Other endovascular management procedures 485

Figure 39.12 Complete avulsion of the

balloon from the catheter successfully
retrieved using the goose neck snare

body present in the tubular structure as long as the size of the tandem with the improved merits of materials and
snare matches with the diameter of the tubular structure. devices, the widening range of indications and the grow-
Small stents which have become dislodged may be engaged ing experience of radiologists and other operators. The
into the loop of the snare and pulled out.33 Embolisation expansion in numbers of PTAs performed correspond-
coils are occasionally misplaced or become emboli which can ingly reflects the incidence of complications that have to
be retrieved with the help of the goose neck snare. be managed by radiologists and vascular surgeons. It is
Balloon catheters are designed to burst longitudinally essential therefore that, as part of one’s training in
when the designated ‘burst pressure’ is exceeded. This endovascular procedures, the trainee is instructed in
does not pose any problems during removal. Occasionally, meticulous technique, is aware of the incidence and
however, a balloon may have a spiral or partial transverse types of complications and indeed the care necessary in
tear. The problem arises when there is a transverse tear, a avoiding them, as well as the best methods available in
rare complication due to inflation within a heavily calcified managing them when they occur. It is equally important
artery. A transversely torn balloon may be difficult to that, in the event of failure of endovascular attempts,
remove because the torn end may fold back when attempts vascular surgeons are capable of intervening successfully
are made to withdraw it. If endovascular removal of that by open surgical means.
balloon is hard then it is best achieved surgically and the
artery repaired with a patch.
Balloon catheters rarely fail but complete avulsion of the
balloon from the catheter has occurred in the past and still
remains a possibility. It is wise not to inflate the balloon
unless there is a wire through it. The presence of the wire Key references
helps with any other manoeuvres in dealing with problems
that may occur. In the above example of the avulsed balloon, Becker GJ, Katzen BT, Dake MD. Noncoronary angioplasty.
the loop of the goose neck snare can be fed over the Radiology 1989; 170: 921–40.
wire/catheter in order to engage the avulsed balloon, before Belli A, Cumberland D, Knox A, et al. The complication rate of
percutaneous peripheral balloon angioplasty. Clin Radiol
pulling it out through the sheath (see Fig. 39.12). A similar
1990; 41: 380–3.
manoeuvre can be used with dislodged/ruptured items of Matsi PJ, Manninen HJ. Complications of lower limb percutaneous
catheter, central line or avulsed wire. transluminal angioplasty: a prospective analysis of 410
procedures on 295 consecutive patients. Cardiovasc Int Radiol
1998; 21: 361–6.
Conclusions Morse M, Jeans W, Cole S, et al. Complications in percutaneous
transluminal angioplasty: relationships with patient age.
Br J Radiology 1991; 64: 757–9.
Endovascular procedures are carried out in increasing
Tegtmeyer C, Hartwell G, Selby B, et al. Results and complications
numbers each year in most hospitals and this trend of angioplasty in aortoiliac disease. Circulation 1991; 83
continues year by year. In part, this rise has occurred in (suppl I): I-53–60.
486 Injuries of peripheral endovascular procedures

REFERENCES 18 Bergquist D, Jonsson K, Weibull H. Complications after

percutaneous transluminal angioplasty of peripheral and renal
arteries. Acta Radiol 1987; 28: 3–12.
1 Dotter CT, Judkins, MP. Transluminal treatment of arteriosclerotic 19 Franco C, Goldsmith J, Veith F, et al. Management of arterial
obstructions: description of a new technic and a preliminary injuries produced by percutaneous femoral procedures. Surgery
report of its application. Circulation 1964; 30: 654–70. 1993; 113: 419–25.
2 Lewis D, Bullbulia R, Murphy P, et al. Vascular surgical 20 Chaterjee T, Do DD, Kaufmann U, et al. Ultrasound guided
intervention for complications of cardiovascular radiology: compression repair for treatment of femoral artery
13 years experience in a single centre. Ann R Coll Surg Engl pseudoaneurysm: acute and follow-up results. Catheter
1999; 81: 23–6. Cardiovasc Diagn 1996; 38: 335–40.
3 Gardiner G, Meyerovitz M, Stokes K, et al. Complications of 21 Wixon CL, Philpott JM, Bogey WM, Powell CS. Duplex directed
transluminal angioplasty. Radiology 1986; 159: 201–8. thrombin injection as a method to treat femoral artery
4 Tegtmeyer C, Hartwell G, Selby B, et al. Results and pseudoaneurysms. J Am Coll Surg 1998; 18: 464–6.
complications of angioplasty in aortoiliac disease. Circulation 22 Fellmeth BD, Roberts AC, Bookstein JJ, et al. Postangiographic
1991; 83(suppl I): I-53–60. femoral artery injuries: non-surgical repair with US-guided
5 Axisa B, Fishwick G, Bolia A, et al. Complications following compression. Radiology 1991; 178: 671–7.
peripheral angioplasty. Ann R Coll Surg Engl 2002; 84: 39–42. 23 Kazmers A, Meeker C, Kofz K, et al. Nonoperative therapy for
6 Hayes PD, Chokkalingham A, Jones R, et al. Arterial perforation postcatheterization femoral artery pseudoaneurysms. Am
during infrainguinal lower limb angioplasty does not worsen Surg 1997; 63: 199–204.
outcome: Results from 1409 patients. J Endovasc Ther 2002; 24 Owen RJT, Haslam PJ, Elliott ST, et al. Percutaneous ablation
4: 422–7. of peripheral pseudoaneurysms using thrombin: a simple and
7 Papvavassiliou VG, Walker SR, Bolia A, et al. Techniques for effective solution. CardioVasc Intervent Radiol 2000; 23:
the endovascular management of complications following 441–6.
lower limb percutaneous transluminal angioplasty. Eur J Vasc 25 O’Keeffe ST, Woods BO, Beckmann CF. Percutaneous
Endovasc Surg 2003; 25: 125–30. transluminal angioplasty of the peripheral arteries. Cardiol
8 Belli A, Cumberland D, Knox A, et al. The complication rate Clin 1991; 9: 515.
of percutaneous peripheral balloon angioplasty. Clin Radiol 26 Samson RH, Sprayregen S, Veith FJ, et al. Management of
1990; 41: 380–3. angioplasty complications, unsuccessful procedures, and early
9 Morse M, Jeans W, Cole S, et al. Complications in percutaneous and late failures. Ann Surg 1984; 199: 234.
transluminal angioplasty: relationships with patient age. Br J 27 Weibull H, Bergqist D, Jonsson K, et al. Complications after
Radiol 1991; 64: 757–9. percutaneous transluminal angioplasty in the iliac, femoral,
10 Fraedrich G, Beck F, Bonzal T, Schlosser V. Acute surgical and popliteal arteries. J Vasc Surg 1987; 5: 681.
intervention for complications of percutaneous transluminal 28 Gaines PA, Cumberland DC, Kennedy A, et al. Cholesterol
angioplasty. Eur J Vasc Surg 1987; 1: 197–203. embolisation; a lethal complication of vascular catheterisation.
11 Matsi PJ, Manninen HJ. Complications of lower limb Lancet 1988; i: 168–70.
percutaneous transluminal angioplasty: a prospective analysis 29 Murray IG, Brown AL, Wilkins RA. Percutaneous aspiration
of 410 procedures on 295 consecutive patients. Cardiovasc thromboembolectomy: a preliminary experience. Clin Radiol
Int Radiol 1998; 21: 361–6. 1994; 49: 553–8.
12 Becker GJ, Katzen BT, Dake MD. Noncoronary angioplasty. 30 Cleveland TJ, Cumberland DC, Gaines PA. Percutaneous
Radiology 1989; 170: 921–40. aspiration thromboembolectomy to manage the embolic
13 Starck EE, McDermott J, Crummy AB, Heydwolf AV. Angioplasty complications of angioplasty and as an adjunct to thrombolysis.
of the popliteal and tibial arteries. Semin Intervent Radiol Clin Radiol 1994; 49: 549–52.
1984; 1: 296–77. 31 Wagner HJ, Starck EE, Reuter P. Long term results of
14 van Andel GJ, van Erp WFM, Krepel VM, Breslau PJ. percutaneous aspiration embolectomy. Cardiovasc Intervent
Percutaneous transluminal dilatation of the iliac artery; Radiol 1994; 17: 241–6.
long-term results. Radiology 1985; 156: 321–3. 32 Higginson A, Alaeddin F, Fishwick G, Bolia A. ‘Push and park’.
15 Greenfield AJ. Femoral, popliteal, and tibial arteries: percutaneous An alternative strategy for management of embolic
transluminal angioplasty. Am J Radiol 1980; 135: 927–35. complication during balloon angioplasty. Eur J Vasc Endovasc
16 Laerum F, Castaneda-Zuniga WR, Amplatz KA. Complications of Surg 2001; 21: 279–82.
transluminal angioplasty. New York: Thieme-Stratton, 1983: 41–4. 33 Gabelmann A, Kramers SC, Tomczak R, Gorich J. Percutaneous
17 Zeitler E. Percutaneous dilatation and recanalization of iliac and techniques for managing maldeployed or migrated stents.
femoral arteries. Cardiovasc Intervent Radiol 1980; 3: 207–12. J Endovasc Ther 2001; 8: 291–302.
 40
Specialty Related Iatrogenic Vascular Injuries

LARS NORGREN

The problem 487 Gynaecological surgery 491

Orthopaedic vascular complications 488 Prevention and treatment 491
General surgery 489 References 492

THE PROBLEM These complications, which are usually treated as emer-

gencies, fall into four groups
Percutaneous interventions are increasingly used for diag- • bleeding
nostic and therapeutic procedures. In vascular surgery, the • thrombosis
Swedish Vascular Registry (Swedvasc) between 1987 and • embolism
1997 reported an increasing proportion of catheter inter- • dissection.
ventions rising from a third to three-quarters in compari- A further group of delayed problems includes false
son with open procedures on the aortoiliac segment. In aneurysms, arteriovenous fistulae and septic complications
line with this change there seems to be an increasing num- but they do not usually need emergency treatment.
ber of iatrogenic vascular injuries reported. In a Swedish Apart from catheterisation procedures there are other
study in 1995 iatrogenic causes accounted for 20 per cent specific surgical interventions which give rise to complica-
of all vascular injuries reported1 and an unpublished tions. These fall within the three main surgical specialties of
Swedvasc report valid for the year 2000 found a corres- orthopaedics, general surgery and gynaecology. Orthopaedic
ponding figure of 45 per cent. There was also a slightly operations on the lumbar disc and on the limbs for various
higher incidence of iatrogenic vascular injuries in larger conditions are at risk of producing iatrogenic vascular
rather than in smaller hospitals and this is easily explained injuries. Within the broad range of general surgical proce-
by the fact that the former treat the more complicated dures, certain categories are potentially vulnerable to vascu-
cases. lar injury and these include operations in the abdomen,
This chapter gives an overall picture of the kind of dam- laparoscopic and thoracoscopic interventions, and hernia
age that can be inflicted on arteries and veins in patients and varicose vein surgery. Gynaecological operations also
who are not primarily victims of vascular disease. They place the pelvic vasculature at some risk.
represent a small proportion of iatrogenic vascular injuries
and result from open procedures; these are discussed
under the various specialties concerned.
According to Adar et al.2 iatrogenic complications may Specialty groups and procedures at risk
be technically classified as the consequences of:

• accident • Orthopaedics
– Lumbar disc surgery
• faulty technique or routine
– Limb surgery
• errors in judgement or management
• General surgery
• failure to identify anatomical structures correctly
– Varicose vein surgery
• failure to interpret X-rays or laboratory findings
– Laparoscopic surgery
correctly.
488 Specialty related iatrogenic vascular injuries

The surgical procedure is often performed under desper-

– Thoracoscopic surgery
ate circumstances with the patient in hypovolaemic shock.
– Open abdominal surgery
On these occasions a gauze pack and manual compression
– Hernia surgery
are required while proximal control is being achieved. Even
• Gynaecology
if the bleeding emanates from a large vein, aortic clamping
gives much needed respite and a better chance of successful
resuscitation. ‘Two hands and two eyes’ are inadequate in
ORTHOPAEDIC VASCULAR COMPLICATIONS such situations. The management of these injuries depends
on the extent of the injury, and, not infrequently, graft
replacement is required for distal aortic or iliac artery injury.
Surgery for lumbar disc herniation, hip arthroplasty and
Venous injuries are certainly not easy to handle and repairs
knee operations are all procedures in which vascular prob-
by simple suture, patch, graft replacement or even a ligature
lems commonly occur.3 The overall complication rate for
may be required. Prompt diagnosis and aggressive treatment
elective orthopaedic surgery varies from country to coun-
of bleeding during lumbar disc surgery may improve the
try. Two Scandinavian studies presented figures showing
mortality, which currently stands at over 50 per cent.9
that only four out of a total of 1314 and two out of 795 vas-
Anterior approaches to the lumbar spine have become
cular injuries were caused by elective orthopaedic surgical
more popular and vascular surgeons are frequently involved
procedures. Acute orthopaedic surgery may cause bleeding
in these procedures on an emergency basis. One retrospec-
due to laceration of vessels often associated with the use of
tive study of 102 such lumbar spine procedures reported a
nails and pins. Such injuries can also cause delayed compli-
16 per cent incidence of vascular injuries requiring surgical
cations, mainly bleeding, probably as a result of infection.6
repair.10
A German study reported 43 patients with vascular injuries
Arteriovenous fistulae only rarely demand emergency
after orthopaedic surgery: 18 were seen after hip surgery, 8
treatment except in instances of heart failure when the
after knee surgery, 2 after lumbar disc surgery and 15 after
indication is obvious and the intervention urgent. This
fracture fixation.7
procedure is often difficult technically, necessitating dis-
section of both proximal and distal artery and vein. If at all
Lumbar disc surgery possible, a catheter procedure involving embolisation or
coiling, the magnitude of which is comparatively limited,
The lumbar discs are closely related to the distal aorta, the should be tried. When adequate resources become avail-
iliac arteries, the vena cava and the iliac veins. The litera- able, endovascular treatment of both arteries and veins will
ture shows that the L4/5 interspace is the most common probably be the first line of attack. Insertion of a stent-graft
site for vascular injury. This complication may result in at the site of injury is technically demanding but will stop
immediate and overt bleeding or the development of a false bleeding effectively and relatively atraumatically.
aneurysm or arteriovenous fistula, the latter commonly
presenting much later than the original injury. Such
injuries are frequently associated with intraoperative com- Hip surgery
plications. A literature survey of 68 cases with arteriove-
nous fistulae reported 17 with early signs of bleeding and Case reports on vascular injuries in conjunction with hip
16 with a peroperative or postoperative drop in blood replacement indicate an incidence of 0.004–0.25 per cent.4
pressure.3 Reoperations are frequently associated with The risk of vascular damage is considerably higher during
inflammation and a destructive process involving the revision than during primary procedures.11,12 There seems
spinal ligaments contributing to an increased risk of to be a female and left-sided dominance to the problem,
complications.8 which is more often found during hip arthroplasty than
Treatment of these complications requires a good hip fracture surgery. The deep femoral artery is more
knowledge of anatomy as well as meticulous technique. prone to injury during hip fracture surgery and the exter-
Overt bleeding from the operative field is usually a sign of nal iliac artery during arthroplasty. The severity of these
damage to large vessels. If there are signs of severe bleed- iatrogenic accidents is reflected in the incidence of func-
ing, the only effective treatment is to turn the patient and tional sequelae (19 per cent) and mortality (7 per cent).13
perform an immediate laparotomy. If, with packing, the From a technical point of view the dangerous aspects of
bleeding ceases, it is imperative to observe the patient hip replacement seem to be twofold: the use of a retractor
extremely carefully during the remaining period of surgery close to the artery carries a risk of arterial compression or
and the immediate postoperative course. Even minimal laceration, and manipulation of the femur as the femoral
signs of new bleeding should be considered serious. part of the prosthesis is being inserted causes traction of
Ultrasonography, computed tomography (CT) scanning the vessels.4 The latter may cause thrombosis within vessels
and angiography are occasionally necessary to elicit evi- already diseased.12,14 Reoperations are associated with an
dence of a retroperitoneal haematoma. increased risk of bleeding. The cement used during fixation
 General surgery 489

of the prosthesis may also cause thermal injury to the adja- compressed by the fracture and also during closed reduc-
cent vessels. Haemorrhage from the deep femoral artery tion. Careful observation during and after the procedure is
may sometimes occur at a late stage due to infection at the therefore important.
site of the prosthesis. Limb tumour surgery may also involve tackling large ves-
Prevention of such complications obviously demands sels and should be performed preferably after preoperative
sound surgical technique, and here, as in other instances, a mapping of the main arterial supply and venous drainage.
prerequisite is avoiding injury. This is especially relevant in
reoperations and in patients with atherosclerotic vascular
disease. If bleeding occurs during surgery, the application
GENERAL SURGERY
of clamps and forceps in the wound should be avoided and
compression used instead. Control is achieved through a
proximal incision over the external iliac artery. Varicose vein surgery
Tachycardia and a drop in blood pressure should be
interpreted as signs of retroperitoneal bleeding if the Injuries in varicose vein surgery are usually caused during
reasons are not otherwise obvious. ‘Distal signs’, a term dissection of the groin. Two main types of injury are seen:
coined by hand surgeons, of both ischaemia and neuro- damage to the femoral vein by ligation or resection and
logical deficit should be looked for postoperatively. Bleeding damage to arteries by ligation, resection and even stripping
may not be haemodynamically evident, but increasing of the superficial femoral artery. Even the deep femoral
swelling of the thigh, or on occasions of the gluteal region, artery has been ligated and resected. Injuries caused by the
along with motor and sensory loss, should be noted; as inadvertent injection of sclerotic agents into arteries, either
compartment pressure rises ischaemia may well follow. during surgery or at outpatient varicose vein clinics, have
With more extensive injuries to calcified vessels bypass also been reported.3 During a 20-year period, 15 injuries of
procedures are usually required, but late bleeding can often groin vessels during varicose vein surgery were reported to
be resolved by embolisation of the bleeding source. the Swedish Board of Health and Welfare; it is impossible
to estimate the true incidence of this iatrogenic injury, but
it is probably much higher.
Knee surgery In 1986 Cockett described nine cases of iatrogenic arter-
ial lesions, none of which was detected during surgery, and
As with other orthopaedic procedures there are no data
consequently repair was delayed in all cases.17 A femoral
available establishing the true incidence of vascular injuries
vein injury may be the result of an anatomical mistake dur-
during knee surgery. A report from Australia in 1987
ing dissection and frequently manifests as bleeding. This
described 12 cases after knee arthroplasty during a 10-year
type of injury is usually caused by less experienced sur-
period.15 There were three lacerations of the popliteal artery
geons. As sclerosing agents are only occasionally injected
and seven arterial thromboses. Of these seven cases, five
into the saphenous vein during a ‘high tie’ the risk of injec-
had to undergo major amputation. One false aneurysm and
tion into the superficial femoral artery appears minimal.
one arteriovenous fistula were also reported. Overall, there-
There is, however, a slight risk of percutaneous injection
fore, the incidence seems to be very low but lesions are
of sclerosant into an artery but this should be small with
found incidentally even after arthroscopy. In an American
current techniques of compression sclerotherapy. Even if
series of 118 590 arthroscopies, six penetrating injuries of
varicose vein surgery is considered to be rather simple
the popliteal artery were registered3 of which four required
technically, the need for supervised training is as essential
amputation. These injuries have also been observed after
as that for more advanced surgical procedures.
total knee replacement.16
As the anatomy of the veins in the groin is variable,
The use of a tourniquet for a bloodless field is contro-
careful dissection is mandatory. No attempt to ligate or to
versial in patients with peripheral atherosclerosis as it may
introduce a stripper should be made until the anatomy is
well cause arterial thrombosis. Furthermore, traction and
clear. Even a stripper inserted at ankle level may pass
manipulation during surgery may cause embolisation of
through a perforating vein in the thigh and end up in the
plaque from the popliteal artery.12
femoral vein. Since many of the patients undergoing vari-
cose vein surgery are rather young, dissection of an artery
Fractures and other procedures may cause spasm and its pulsation may disappear. Only
careful dissection will reveal the true anatomy, a prerequis-
It should be emphasised that closed as well as open reduc- ite to continuing with the operation.
tion of fractures, with or without osteosynthesis, may cause When deep bleeding from the femoral vein occurs unex-
compression and injury as well as thrombotic occlusion of pectedly, attempts to arrest it using artery forceps is under-
arteries, subsequently leading to bleeding or ischaemia. standable. This must always be avoided. The only acceptable
The classic example is the supracondylar fracture of the measure is to compress the bleeding vein and call for assist-
humerus in children where the brachial artery may be ance. If the injury has not been worsened by inexpert use of
490 Specialty related iatrogenic vascular injuries

instruments and clumsy clamping, the actual damage is of the large vessels and the other half of the abdominal
usually found to be quite small, and a simple lateral suture wall. These figures, however, should not be seen as the true
may stop the bleeding. If not, proximal and distal control incidence of injuries (PSR, personal communication). The
must be obtained followed by resection and reconstruction. vascular injuries are usually caused by trocar insertion and
In the majority of these cases the saphenous vein has the most commonly affected are the aorta, iliac arteries and
already been divided and may well be used for the recon- the inferior vena cava. Injuries attributable to the insuffla-
struction. If the vein has not been divided, it may assist tion needle have also been reported.22 In many countries
venous return from the leg and should be saved. In such there are now guidelines on the technique for insufflation
cases the saphenous vein from the other leg should be used. and trocar placement.23
In principle, there are two methods of repairing the Taking into consideration the size of the vessels usually
femoral vein, which is usually at least twice the diameter of damaged during laparoscopic procedures, the mortality
the saphenous vein (see Chapter 33). The first is to take risk is high and prompt treatment is required. It is likely
double the length of vein needed and split it longitudinally, that with the current degree of specialisation in surgery,
divide it into two lengths and suture these side-to-side, skilled laparoscopic surgeons are not trained to handle vas-
thereby doubling the diameter.18 The second option is to cular problems. Bleeding, revealed at laparotomy, should
split the vein longitudinally, wrap it spirally over a glass be immediately controlled by manual compression rather
staff and then suture the entire spiral into a tube graft.19 than by a ‘trial and error’ approach and, in most large
The main question, of course, is whether or not this centres, experienced vascular surgeons should be available
kind of reconstruction stays patent. An arteriovenous fis- to deal with this catastrophe.
tula can be created as an adjunctive operation. There are
no proper studies to prove the results, but even if the
Thoracoscopy
reconstruction occludes later, there may be a chance that
while it stays open collateral flow will take over. Ligature of
Thoracoscopic sympathectomy has been used increasingly
the femoral vein is rarely acceptable (see Chapter 33).
to treat palmar hyperhidrosis. Vascular injuries are not
Arterial injuries should be repaired immediately but
reported frequently and in a study of 940 operations24
delayed presentation is common. If there is the slightest
there was only one subclavian artery injury, but there have
suspicion of ischaemia, ankle blood pressures are measured
been occasional reports of mortality.
and, if damage is still suspected, angiography performed.
Reconstructions of these injuries follow general principles of
arterial surgery. If possible, a limited resection and end-to- Open abdominal surgery
end anastomosis should be carried out or, alternatively, graft
replacement, preferably using vein from the contralateral leg. Injuries to the portal vein, superior mesenteric vein and
Intra-arterial injection of sclerosing agents is particu- concomitant arteries may occur during pancreaticobiliary
larly dangerous and, if the effect is widespread, gangrene surgery. For example, vascular injuries have been reported
and amputation will follow. If the agent has been injected in 3–4 per cent of patients operated for chronic pancrea-
more locally, there is a chance that healing may occur. In titis.25 In pelvic surgery the most commonly injured vessels
either situation it is advisable to give heparin intravenously. are the iliac veins.
If for any reason, pain for example, the complication is sus- Cholecystectomy, especially as an acute procedure or
pected and the needle is still in place, heparin-saline and an after recurrent acute cholecystitis, may be extremely diffi-
-receptor blocker or reserpine should be injected through cult. There are also considerable variations in the anatomy
it. An epidural block might be of value but little scientific of the arterial supply to the liver and to the gallbladder.
proof of its value exists. When faced with serious bleeding the hepatic artery may
be ligated but even if some ischaemia does occur, the risk
of liver necrosis is limited. It is not necessary, therefore, to
Laparoscopic surgery repair the hepatic artery, a quite difficult procedure in any
event. If, however, sudden bleeding is experienced during
The tremendous increase in laparoscopic procedures dur- dissection thought to originate from the hepatic artery, the
ing the past 10 years is reflected in the rise in associated advice, as with all sudden vascular injuries, is to compress
vascular injuries, but the actual number of these cases is and not to use deep forceps blindly. The advice ‘more hands
small. In a report of 1995 laparoscopies 20 iatrogenic vas- and more eyes’ is usually sufficient to visualise a partially
cular injuries were reported,20 while in the single-centre divided artery and to permit lateral repair.
experience of 2589 laparoscopic cholecystectomies per- If a pseudoaneurysm results from the injury,26 a ligature
formed between 1990 and 1996, 0.11 per cent had major should be applied proximally and distally. Lesions of the
vascular injuries.21 In Sweden, about seven vascular injuries superior mesenteric artery and the coeliac axis are rare. The
are reported annually to the independent loss adjuster external iliac artery has on occasions been mistaken for
(PersonSkadeReglering (PSR)), half of them being injuries the ureter and opened during intended ureterolithotomy
 Prevention and treatment 491

operations. Injuries of the inferior vena cava sometimes abrupt major bleeding occurring during a surgical proced-
occur during tumour surgery and also occasionally during ure. This stressful situation may well induce the risk of
aortic surgery.27 All attempts should be made to repair the improper actions, such as the blind use of forceps and
vein, but if this proves impossible, a ligature will save the clamps. If the source of bleeding is neither absolutely clear
life of the patient. nor easily accessible for proper repair, the correct action is to
compress the bleeding region firmly with gauze packs. As
already emphasised, the assistance of another surgeon would
Hernia repair
be wise, and in many situations, a necessity. The next step is
to achieve proximal and distal control. Established principles
Iatrogenic injuries of the femoral artery or femoral vein are
of vascular repair are then followed, whether by lateral
not uncommon but are apparently not often reported.3
suture, patching or grafting. Other types of acute vascular
The risk attached to an injury is dependent on the type of
injuries resulting in thrombosis, embolism and dissection
repair and in principle, the more ‘anatomical’ the repair
are treated according to accepted principles. Complications
the lower the risk. With deep sutures the risk of damage to
appearing later, such as arteriovenous fistulae and false
the common femoral artery increases.28 The result is either
aneurysms, can usually be treated by endovascular or vascu-
bleeding or, more rarely, thrombosis with embolisation or
lar surgical techniques (see Chapters 33, 38 and 39).
even the development of a false aneurysm.
In view of the medicolegal implications (see Chapter
Bleeding from an arterial injury is usually best con-
10) detailed note-taking and good communication with
trolled by compression. Compression of the femoral vein
patient and family, and, of course, other involved profes-
and thrombosis seem to occur mainly after hernia repair
sionals, is particularly important.
using the McVay technique.29 Careful dissection and exact
repair of the fascial structures minimises the risks associ-
ated with treating these iatrogenic vascular injuries. Conclusions

Surgeons have to be aware of the potential occurrence of

GYNAECOLOGICAL SURGERY non-catheter induced vascular injuries and be prepared
to treat them. Although such injuries are relatively
Major vascular injuries are reported infrequently. In uncommon, bleeding, severe at times, can be a major
Sweden the incidence per 10 000 procedures was calculated problem. The possibility of occult bleeding, as for example
for the years 1979–83 and found to be 0.93 for laparoscopy, in lumbar disc surgery, must be considered. The value of
0.76 for laparotomy and 0.33 for vaginal operations.30 simply compressing a bleeding point rather than resort-
Laparoscopic complications most frequently involve the ing to hurried and blind clamping cannot be over-
aorta and the iliac arteries and should be detected easily by emphasised. The most worthwhile advice is not to do
the surgeon. There is, however, at least one report in the more harm and to seek assistance whenever possible; an
literature of carbon dioxide embolism following direct extra pairs of hands may be of crucial importance to
puncture of a major vessel.31 outcome. The principles of repairing iatrogenic vessel
During gynaecological laparotomy for pelvic tumour sur- trauma are the same as those for other vascular injuries.
gery vascular injuries are the most common cause of bleed- The specialists involved must be willing to improvise as
ing,30 mainly from the iliac vein or iliac artery. Caesarean none of these injuries present in a standard manner.
section has also been reported as the cause of severe vascular
problems.32 Careful handling of instruments inserted during
laparoscopy seems to be the most important factor in avoid- Key references
ing injury. If bleeding occurs, an immediate laparotomy has
Adar R, Bass A, Walden R. Iatrogenic complications in surgery. Five
to be performed, if possible with the assistance of a vascular
years’ experience in general and vascular surgery in a
surgeon. The same applies to injuries occurring during
university hospital. Ann Surg 1982; 196: 725–9.
laparotomy. It has to be stressed that firm compression and Bergentz SE, Bergqvist D. Iatrogenic vascular injuries. Berlin:
proximal control are the major life-saving measures. Springer-Verlag, 1989.
Bergqvist D, Källerö S. Reoperation for postoperative hemorrhagic
complications. Analysis of a 10-year series. Acta Chir Scand
PREVENTION AND TREATMENT 1985; 151: 17–22.
Jonung T, Pärsson H, Norgren L. Vascular injuries in Sweden
1986–1990; the result of an enquiry. Vasa 1995; 24: 130–4.
Training and experience are the alpha and omega in prevent-
Myhre HO, Sahlin Y, Saether OD, et al. Vascular complications of
ing injuries. Accidents and misinterpretations of difficult
elective orthopaedic surgery. In: Greenhalgh RM, Hollier LH
anatomy and pathology, however, cannot be totally avoided. (eds). Emergency Vascular Surgery. London: WB Saunders,
It is important, therefore, to know how to handle complica- 1992: 353–64.
tions when they occur. The most devastating complication is
492 Specialty related iatrogenic vascular injuries

REFERENCES 16 McAuley CE, Steed DL, Webster MW. Arterial complications of

total knee replacement. Arch Surg 1984; 119: 960–2.
17 Cockett FB. Arterial complications of surgery and sclerosis of
1 Jonung T, Pärsson H, Norgren L. Vascular injuries in Sweden varices. Phlebologie 1987; 40: 107–10.
1986–1990; the result of an enquiry. Vasa 1995; 24: 130–4. 18 Tera H. Emergency repair of femoral vein accidentally divided
2 Adar R, Bass A, Walden R. Iatrogenic complications in surgery. at operation for varicose veins. Acta Chir Scand 1967; 133:
Five years’ experience in general and vascular surgery in a 283–7.
university hospital. Ann Surg 1982; 196: 725–9. 19 Hobson RW, Yeager RA, Lynch TG, et al. Femoral venous trauma:
3 Bergentz SE, Bergqvist D. Iatrogenic vascular injuries. Berlin: techniques for surgical management and early results. Am J Surg
Springer-Verlag, 1989. 1983; 146: 220–4.
4 Myhre HO, Sahlin Y, Saether OD, et al. Vascular complications of 20 Nordestgaard AG, Bodily KC, Osborn RW Jr, Buttorff JD. Major
elective orthopaedic surgery. In: Greenhalgh RM, Hollier LH (eds). vascular injuries during laparoscopic procedures. Am J Surg 1995;
Emergency vascular surgery. London: WB Saunders, 1992: 353–64. 169: 543–5.
5 Støren G,Holta AL, Myhre HO, et al. Arterieskader. Tidskrift for 21 Usal H, Sayad P, Hayek N, et al. Major vascular injuries during
den Norske Laegeforening 1976; 23: 1195–7. laparoscopic cholecystectomy. An institutional review of
6 Bergqvist D, Källerö S. Reoperation for postoperative experience with 2589 procedures and literature review. Surg
hemorrhagic complications. Analysis of a 10-year series. Acta Endosc 1998; 12: 960–2.
Chir Scand 1985; 151: 17–22. 22 Dixon M, Carrillo EH. Iliac vascular injuries during elective
7 Schlosser V, Kuttler H, Kameda T. Arterial injury during orthopedic laparoscopic surgery. Surg Endosc 1999; 13: 1230–3.
and traumatological surgery. J Cardiovasc Surg 1987; 28: 46. 23 Hanney RM, Alle KM, Cregan PC. Major vascular injury and
8 Stokes JM. Vascular complications of disc surgery. J Bone Joint laparoscopy Aust N Z J Surg 1995; 65: 533–5.
Surg Am 1968; 50: 394–9. 24 Gossot D, Kabiri H, Caliandro R, et al. Early complications of
9 Sagdic K, Ozer ZG, Senkaya I, Ture M. Vascular injury during thoracic endoscopic sympathectomy: a prospective study of 940
lumbar disc surgery. Report of two cases; a review of the procedures. Ann Thorac Surg 2001; 71: 1116–19.
literature. Vasa 1996; 25: 378–81. 25 Call FP, Muhe E, Gwohardt C. Results of partial and total
10 Baker JK, Reardon PR, Reardon MJ, et al. Vascular injury in pancreaticoduodenectomy in 117 patients with chronic
anterior lumbar surgery. Spine 1993; 18: 2227–30. pancreatitis. World J Surg 1981; 5: 269–75.
11 Bergqvist D, Carlsson AS, Ericsson F. Vascular complications of 26 Thomas WEG, May RE. Hepatic artery aneurysm following
total hip arthroplasty. Acta Orthop Scand 1983; 54: 157–63. cholecystectomy. Postgrad Med J 1981; 57: 393–5.
12 Nachbur B, Meyer RP, Verkkala K, Zurcher R. The mechanisms of 27 Vollmar J. Venous trauma. Major Prob Clin Surg 1979; 23: 191–9.
severe arterial injury in surgery of the hip joint. Clin Orthop 28 Shamberger RC, Ottinger LW, Malt RA. Arterial injuries during
1979; 141: 122–33. inguinal herniorrhaphy. Ann Surg 1984; 200: 83–5.
13 Beguin L, Feugier P, Durand JM, et al. Vascular risk and total hip 29 Klausner JM, Noveck H, Skornick Y, et al. Femoral vein occlusion
arthroplasty: prevention during acetabular revision. Rev Chir following McVay repair. Postgrad Med J 1986; 62: 301–2.
Orthop Reparatrice Appar Mot 2001; 87: 489–98. 30 Bergqvist D, Bergqvist A. Vascular injuries during gynecologic
14 Matos MH, Amstutz HC, Machleder HI. Ischemia of the lower surgery. Acta Obstet Gynecol Scand 1987; 66: 9–23.
extremity after total hip replacement: report of four cases. J 31 Hanney RM, Alle KM, Cregan PC. Major vascular injury and
Bone Joint Surg Am 1979; 61: 24–7. laparoscopy. Aust N Z J Surg 1995; 65: 533–5.
15 Rush JH, Vidovich JD, Johnson MA. Arterial complications of 32 Buri P. Iatrogene Schädigung von Blutgefässen. Helv Chir Acta
total knee replacement. The Australian experience. J Bone Joint 1971; 38: 151–5.
Surg UK 1987; 3: 400–2.
 SECTION
9
Special Acute Vascular Challenges

41. Acute upper limb ischaemic states 495

42. Emergency aspects of Buerger’s disease 511

43. Acute limb vascular inflammatory conditions 523

44. Vascular emergencies caused by substance abuse 531

45. HIV/AIDS related vascular emergencies 545

46. Portal hypertension and variceal bleeding 555

47. Cold injury 573

This page intentionally left blank
 41
Acute Upper Limb Ischaemic States

KENNETH A MYERS, GREGORY W SELF

The problem 495 Conservative measures 501

Aetiology 495 Reconstructive surgery 502
Diagnosis 499 Endovascular management 505
Immediate management 501 References 507

THE PROBLEM is approximately equally due to atherosclerosis or a miscella-

neous group of non-atherosclerotic diseases.3 Other studies
suggest that patients with digital ischaemia suffer more often
Acute ischaemia is less common in the upper than in the from intrinsic small arterial disease in the hand than prox-
lower limbs, but is caused by a wider variety of disease imal major arterial disease.3 Patients occasionally develop
processes and has varied options for treatment. Ischaemia acute thrombosis in essentially normal arteries on the basis of
can affect any level from the fingertips to the forearm, and various haematological abnormalities. In the tertiary hospital
digital ischaemia is a warning that massive occlusion from setting, acute and chronic ischaemia following angiography
proximal disease may supervene. Delayed diagnosis and failed is increasingly recognised. Accordingly, a long list of possible
salvage leading to amputation is a far greater disaster than causes must be kept in mind in any individual patient.
loss of the lower limb. This chapter will attempt to provide a
plan for diagnostic evaluation and strategies for treatment of
acute ischaemia aimed at enhancing successful management Potential causes of acute upper limb
of the underlying disease before complications develop. ischaemia

• Embolism
AETIOLOGY – Cardiac diseases
– Proximal arterial diseases
– Paradoxical embolism
In contrast to the lower limb, where acute ischaemia virtu-
ally always ensues following either embolism from the • Intrinsic arterial diseases
– Occlusive atherosclerosis
heart or thrombosis on the basis of underlying athero-
– Atherosclerotic aneurysms
sclerosis, the causes in the upper limb are much more var-
– Arteritis and autoimmune diseases
ied (see Chapter 2).
– Fibromuscular hyperplasia
Less than 5 per cent of peripheral arterial reconstructions
– Cystic adventitial disease of the brachial artery
are for upper limb arterial disease.1 The incidence of lower
limb arterial reconstruction has progressively risen over the • Haematological disorders
– Hyperhomocysteinaemia
past 20 years,2 and there are now more operations for throm-
– Malignant diseases
bosis and fewer for embolism, but whether this is true for the
– Clotting factor deficiencies
upper limbs is uncertain as the upper limb series are smaller.
– Antiphospholipid antibodies
Studies of acute upper limb ischaemia have shown that
– Red cell, white cell and globulin disorders
embolism and thrombosis are almost equally respon-
– Heparin induced thrombosis syndrome
sible, that embolism is commonly due to either cardiac or
– Oral contraceptives
proximal arterial disease, and that primary occlusive disease
496 Acute upper limb ischaemic states

the course of assessment of the latter. The Joint Study of

• Repetitive external trauma
Extracranial Arterial Occlusions in more than 6000 patients7
– Thoracic outlet syndrome
showed that 17 per cent had aortic arch disease, affecting
– Athletic injuries
the left subclavian artery in approximately 15 per cent, the
– Axillary crutch injury
right subclavian artery in 10 per cent and the innominate
• Occupational injuries
artery in 5 per cent, and that stenosis was five times more
– Trauma to the hands and fingers
frequent than occlusion. The distribution of upper limb
• Iatrogenic trauma
atherosclerotic disease in a more general population has
– Arterial catheterisation
been reviewed, and it appears to affect the subclavian
– Anaesthetic arm blocks
artery in 60 per cent, the innominate in 20 per cent and
– Intra-arterial injections
more distal arteries in 20 per cent.1 The relative incidence
– Thrombosed axillofemoral bypass
of innominate or subclavian aneurysms compared with
– Irradiation arteritis
thoracic and abdominal aortic aneurysms is not clear.
– Ergotism
Occlusions most often cause mild chronic ischaemia, but
– Radial artery fistula steal
stenoses can present with multiple microemboli to cause
– Harvesting radial artery
the ‘blue finger syndrome’, while aneurysms can present
• Acute external trauma
with acute ischaemia caused by a major embolus.
– Blunt injuries
– Penetrating injuries
ARTERITIS, AUTOIMMUNE AND OTHER DISEASES
– Intravenous drug abuse
• Non-prescribed drug use A number of uncommon diseases can affect the branches
– Cannabis use of the aortic arch to cause upper limb ischaemia or cere-
brovascular insufficiency, or they may involve the small
arteries of the hands to cause digital ischaemia and necro-
Embolism sis.1,3,8 It is beyond the scope of this chapter to explore each
in detail, although they are described more extensively in
Some 10–20 per cent of all emboli are lodged in arteries to other reviews.9,10 They probably have a similar aetiology
the upper extremity, significantly more often on the right associated with an immune complex or cell mediated reac-
side. The most common site is to the brachial artery, less tion11 but the precipitating factors are unknown.
often the subclavian or axillary arteries and uncommonly
to arteries below the elbow. The larger proportion come Some causes of acute upper limb ischaemia
from the heart, particularly in patients with atrial fibrilla- associated with arteritis or autoimmune
tion or recent myocardial infarction, fewer in rheumatic disorders
heart disease and aortic valve prostheses, and rarely from
left atrial myxoma. Up to a third originate from disease in
the innominate or subclavian arteries owing to atheroscler-
• Takayasu’s disease

osis, aortitis and in particular the thoracic outlet syn-

• Buerger’s disease

drome. Paradoxical embolism can also occur.4

• Giant cell arteritis

The clinical features of acute upper limb ischaemia are

• Polyarteritis

usually quite evident although delayed in approximately

• Behçet’s syndrome

25 per cent of cases, which is significantly less common

• Scleroderma

than that in the lower limb, so that the proportion treated

• Hypersensitivity reaction

by embolectomy is higher and the outlook is better.

• Systemic lupus erythematosus

Nevertheless, several studies report amputation rates of 5–10

per cent and mortality rates of 5–15 per cent.5,6 Takayasu’s disease affects children and young adults,
with a marked female preponderance (see Chapter 43). A
panarteritis affects the major aortic branches, leading to
Intrinsic arterial diseases weakening of the wall with consequent arterial stenoses,
occlusions or aneurysms which are frequently multiple.12
ATHEROSCLEROSIS
The condition commences as a non-specific ‘rheumatic-
Proximal atherosclerosis in the upper limb is more com- like’ illness, progressing to become a chronic disease several
mon than is realised, but it may remain relatively asymptom- years later. An elevated erythrocyte sedimentation rate is
atic because of the abundant collateral circulation around usual in the acute phase, but the test returns to normal in
the shoulder. Innominate and proximal subclavian artery the ‘burned out phase’. The outlook for survival is poor
disease, which may lead to symptoms of upper limb and cere- with severe hypertension, cardiac involvement or other risk
brovascular ischaemia, is often revealed by arteriography in factors being predictors of death or major complications.13
 Aetiology 497

Buerger’s disease (thromboangiitis obliterans) usu- the popliteal artery has been reported in the brachial
ally affects young males who are smokers and is associated artery.29
with segmental thrombotic occlusions of more distal arter-
ies and veins14,15 (see Chapter 42). In the acute phase, poly- Haematological disorders
morphonuclear leucocytes infiltrate the vessel wall and
thrombus; this is followed by a subacute phase with the A wide variety of conditions can lead to a hypercoagulable
appearance of mononuclear and giant cells, and finally a state and thrombosis in small or large arteries30 (see
chronic phase of fibrosis and recanalisation. Essentially, Chapter 2). These include deficiencies of antithrombin III,
this is a clinical diagnosis. Both Japanese and American protein C or protein S, or presence of Leiden factor V,
studies showed that the larger proportion were symptom- antiphospholipid antibodies, homocysteine or antibodies
atic, of whom approximately 60 per cent had only lower to heparin causing the heparin induced thrombosis syn-
limb symptoms, 30 per cent had only upper limb symp- drome. Coagulation can be activated by oral contracep-
toms and 10 per cent had both: this was despite arterio- tives and in various malignancies. An increase in blood
graphic findings that showed changes in most lower limb viscosity due to globulin disorders, polycythaemia and vari-
and upper limb arteries, indicating that the disease is much ous lymphomas or leukaemias predisposes to thrombosis.
more extensive than can be gauged by symptoms alone.15,16
The outlook for the limbs is poor, but the prognosis for
survival is good in those who stop smoking.15,16 Regular Repetitive external trauma
cannabis smoking can cause changes similar to those of
Buerger’s disease.17 This chapter is not the place to describe acute blunt or pene-
Giant cell arteritis and polymyalgia rheumatica pre- trating trauma that happens to involve the upper extrem-
dominantly affect women, occur in most countries and are ity. There are, however, several circumstances which can
almost exclusively confined to the white population18,19 predispose to specific acute or repetitive arterial trauma
(see Chapter 43). Some 10–15 per cent of patients have clin- with severe consequences to the upper extremity.
ical involvement of arteries to the extremities10,20 but a
much larger proportion have pathological evidence of dis- THORACIC OUTLET SYNDROME
ease when studied at postmortem.20 Major arteries can be Neurovascular compression can result from impingement
affected at any level,21 but there is usually sufficient time for by either soft tissue or bony structures.31 The clinical pic-
adequate collaterals to develop. Histological examination ture is mostly one of chronic pain, neurological disturb-
reveals lymphocytic infiltration with granulomatous and ance or relatively mild vascular symptoms and only a few
giant cells, and fragmentation of the internal elastic lamina. patients have appreciable arterial insufficiency.32 Bony
Polyarteritis can be primary or secondary to conditions abnormalities are a cervical rib or its fibrous extension,
such as rheumatoid arthritis, Sjögren’s disease, cryoglobu- congenital abnormalities include the first rib, or a past
linaemia or leukaemia, 22 and the condition predominantly fractured clavicle.1
affects the endothelium of small or medium sized arteries Cervical ribs are present in less than 1 per cent of healthy
so as to cause stenoses, occlusions or small aneurysms. individuals, about 50 per cent are bilateral and it would
Behçet’s syndrome affects relatively young males from appear that less than 10 per cent of those cause symptoms33
eastern Mediterranean countries or Japan.23–25 A non-specific (see Chapter 22). The cervical rib runs along the anterior
vasculitis affects small or large vessels, obliterates vasa border of the scalenus medius muscle immediately under
vasorum and fragments elastin to cause arterial occlusions the artery (Fig. 41.1). It can rarely arise at a higher level, i.e.
or aneurysms which are frequently multiple. The aneurysms up to the fifth cervical vertebra, and occasionally the rib can
are the most frequent cause of death in some 10–25 per cent pass through the trunks of the brachial plexus. An abnor-
of patients.25 mal first rib is said to be less than one half as common as a
Scleroderma most often affects the small metacarpal cervical rib; the most common severe variant is atresia of
and digital arteries in the hands, which can lead to necrosis the first rib, leaving an exostosis on the second rib at the site
of the fingertips, but the condition is occasionally also where the scalenus anterior muscle is inserted.34
associated with occlusion of the proximal arteries to the Kinking of the artery over a bony prominence or fibrous
upper extremity secondary to severe intimal hyperplasia.26 band can cause turbulent flow, leading to poststenotic
Crohn’s disease has been described as causing axillary dilatation, then intimal disruption and aneurysm forma-
artery occlusion due to arteritis.27 tion, finally complicated by thrombosis within the aneurysm
or embolism resulting in severe acute ischaemia.
OTHER DISEASES
ATHLETIC INJURIES AROUND THE SHOULDER
Fibromuscular hyperplasia has been reported in the
brachial artery in a few patients, affecting either sex at vari- Athletic activities that require throwing occasionally cause
ous ages.28 Cystic adventitial necrosis similar to that seen in injury leading to thrombosis of the proximal arteries. This
498 Acute upper limb ischaemic states

thrombosis of the digital or ulnar arteries.36 Injuries to the

digital arteries particularly involve the index or middle fin-
Scalenus gers wherein the artery is encircled by ligaments. A condi-
medius tion referred to as the hypothenar hammer syndrome
muscle
occurs particularly in workers who use their hand as a
hammer; this includes bricklayers, printers, tilers, mechan-
ics and professional karate experts.38,39 The ulnar artery is
Cervical
rib particularly vulnerable to trauma just after it leaves the
canal under the hamate bone, where it lies superficial to the
tough transverse carpal ligament and is held tight in its
Subclavian course by its deep branch.38 Trauma can result in either
artery acute thrombosis or aneurysm formation. Ischaemia is
particularly likely if the superficial palmar arch is incom-
Scalenus
plete, but it can also be due to embolism from an aneurysm.
anterior
muscle Use of vibrating tools can cause distal small artery vasospasm
or occlusion. Digital arteries, in particular, become hyper-
Figure 41.1 A subclavian aneurysm distal to a cervical rib lying responsive causing the ‘white finger syndrome’.40
on the anterior border of scalenus medius, causing the thoracic
outlet syndrome
Iatrogenic trauma

There is a considerable capacity for medical manoeuvres to

has been particularly studied in baseball pitchers and cause damage to arteries of the upper extremity, often with
American football quarterbacks;35,36 other activities include catastrophic results. Penetrating injuries include arterial
javelin throwing or basketball. Duplex scanning shows that catheterisation for arteriography or haemodynamic mon-
the larger proportion of normal subjects have the potential itoring although the problem is far less common now
for this injury, as partial or complete occlusion of the that most investigations are performed via a transfemoral
artery does occur in the throwing position.36 Some are approach (see Chapter 38). Bleeding from the brachial or
associated with subclavian artery thrombosis due to the axillary puncture site may raise the pressure within the
thoracic outlet syndrome. Others result in axillary artery sheath of the neurovascular bundle to such a degree as to
thrombosis due to compression by the head of the humerus cause severe nerve palsies, which are frequently perman-
with the arm abducted and in external rotation, particu- ent.41 An anaesthetic arm block with accidental intramu-
larly if there is laxity of ligaments around the shoulder ral injection can cause axillary artery occlusion.42 A blunt
joint. It is possible that distal axillary artery thrombosis can needle technique is recommended, but this does not elim-
result from injury by a hypertrophied pectoralis minor inate the risk. The complication is more likely if the patient
muscle. Once cleared of clot the artery is in most cases is obese or if the artery is already diseased. Intra-arterial
found to be essentially normal. injection of therapeutic agents such as anaesthetic drugs or
Repeated external trauma to the axillary artery by the use of illicit substances by drug abusers may precipitate acute
of a crutch, though rare, is well recognised and may take severe extremity ischaemia with permanent damage43 (see
the form of acute thrombosis or the gradual development Chapter 44). The difficulties encountered in the treatment
of an axillary aneurysm with secondary thrombosis.37 of these challenging conditions is frequently compounded
by secondary venous thrombosis. Harvesting the radial
OCCUPATIONAL UPPER LIMB ISCHAEMIA artery for coronary artery bypass surgery fortunately rarely
Occupational activities occasionally lead to acute upper limb leads to hand ischaemia.44
vascular ischaemia. These are more often associated with the An increasingly common phenomenon is axillary artery
thoracic outlet syndrome causing neurological or venous thrombosis with or without embolism consequent upon
syndromes, but arterial complications can occur. Activities thrombosis of an axillofemoral prosthetic graft.45 Exertion
include mowing lawns, heavy lifting as a store assistant, may disrupt the top anastomosis to cause ‘the axillary pull-
installing overhead ducts, or playing sports as described ear- out syndrome’.46 The risk may be increased if the graft is so
lier. An occupational history consistent with repetitive exter- short that it brings about ‘tenting’ of the axillary or sub-
nal trauma should be sought if the diagnosis is not clear. clavian artery.45 Retrograde thrombosis from the upper
end of the occluded graft may extend into the subclavian
artery, or embolisation down the limb may occur from the
TRAUMA TO THE HAND AND FINGERS
cul de sac adjacent to the upper anastomosis.47 It is possi-
Athletic activities that require repetitive catching or strik- ble that the graft may shorten after some years predispos-
ing a ball with the hand can predispose to traumatic ing to thrombosis in the graft and subclavian artery, and it
 Diagnosis 499

has been suggested that this risk is reduced by making the frequent than in the lower limb because proximal muscles
anastomosis more proximal on the subclavian artery to can perform upper limb movements. The radial pulse is
reduce repetitive movements.48 The possibility of steal lost and the level of the most proximal pulse usually clearly
from the arm to a patent graft appears to be unfounded. indicates the site of occlusion. Embolism carries a consid-
Irradiation arteritis of the subclavian, axillary and erable risk of residual arm claudication, ischaemic neur-
upper brachial artery, which also compromises collaterals, opathy, ischaemic contracture or even major amputation,
presents many years after therapy for carcinoma of the usually due to delayed presentation and technical prob-
breast (see Chapter 33). Vessels occlude due to endothelial lems at operation.6 Repeated microembolism can cause
damage, disruption of the internal elastic lamina, hyaline recurrent episodes of forearm or hand pain with focal areas
thickening of the intima, obliteration of vasa vasorum, and of discoloration or skin necrosis analogous to the ‘blue toe
medial and indeed periarterial fibrosis.49 syndrome’ of the lower extremity.
Proximal artery vasospasm can result from treatment The varied presenting clinical features of chronic large
with ergot preparations and leads to severe upper or lower artery disease may include a palpable aneurysm, forearm
limb ischaemia.50,51 This complication was responsible for claudication, Raynaud’s phenomenon, digital necrosis or
withdrawal of the dihydroergotamine heparin regimen for infarction, or acute ischaemia from the occlusion. Forearm
deep venous thrombosis prophylaxis. claudication is the least common manifestation because of
Hand ischaemia can result from steal induced by an the rich collateral circulation around the shoulder and
arteriovenous fistula created for angioaccess in patients because a smaller muscle mass than that in the lower limb
requiring dialysis (see Chapter 7B) and the incidence may has to be supplied. The distal pulse status is far less predict-
be as high as 5 per cent.52 Ischaemic symptoms are not ive of chronic proximal disease than is that in the lower
uncommon early on but usually resolve rapidly as collat- limb.7 A bruit is frequently observed but is not always pres-
erals develop. The risk of persisting ischaemia is increased ent. A brachial artery systolic pressure difference between
if the ulnar artery supply is inadequate as shown by the the arms of more than 20 mmHg reliably indicates prox-
Allen test or if, as in diabetic patients, there is pre-existing imal stenotic or occlusive disease.53
distal arterial disease. Many access surgeons would now try A lump representing a cervical rib or prominent pulsa-
to avoid the otherwise standard radiocephalic (Cimino) tion of the subclavian artery pushed upwards by the rib is
fistula in type 1 diabetic patients in preference to a more usually diagnostic. The wide variety of physical man-
proximal fistula which is less likely to steal circulation. oeuvres designed to compromise the pulse in order to
Patients undergoing such access must be warned of the establish a diagnosis of thoracic outlet syndrome and the
symptoms and signs of arterial insufficiency by the surgeon high incidence of false positive results are a testimony
as the problem may not be appreciated or recognised too to the clinical challenge posed. Harris et al.54 found that
late by medical and nursing staff in other disciplines. 30 per cent of normal subjects showed apparent arterial
compression with thoracic outlet syndrome provocative
manoeuvres, and even this figure may understate the inci-
dence of false positive tests. Nevertheless, a negative test
DIAGNOSIS virtually excludes a diagnosis of vascular involvement while
a positive test provides confidence to proceed with treatment
Clinical features if a clear history of ischaemia is present. The ‘hands-up’
manoeuvre is a convenient method of observing whether
It is necessary to discuss all patients presenting with vary- the pulse disappears, the hand blanches with exercise and
ing grades of ischaemia since many conditions have a a bruit develops as the arm is then lowered (Fig. 41.2).
potential for sudden complications causing acute ischaemia. Small artery disease most commonly presents with
Diagnosis must include both the causes and consequences Raynaud’s phenomenon, digital necrosis or digital gangrene
of the diseases. In general, full clinical assessment and according to the acuteness of onset and extent of disease.
various special investigations are appropriate for most The Allen test is very useful for detecting the presence and
patients, considering the wide variety of possible causes, site of occlusion in small arteries beyond the wrist (Fig. 41.3).
not all of which are immediately obvious. Systemic features It is also used to detect inadequate ulnar artery inflow to
associated with the various causes of arteritis and auto- the hand prior to radial artery cannulation or the creation
immune disease should be sought in many patients. This of a radial artery fistula (see Chapter 7B).
requires detailed knowledge of each, which, although
beyond the scope of this chapter, is well described in sev-
eral reviews.9,10 Haematological investigations
Large artery embolism usually presents with acute
ischaemia causing pain, pallor, paralysis, paraesthesia, and A full blood examination and erythrocyte sedimentation
‘perishing coldness’, although in some cases the onset is rate would seem to be required in all patients, although the
more insidious (see Chapters 2, 15 and 16). Paralysis is less erythrocyte sedimentation rate may not be elevated even
500 Acute upper limb ischaemic states

(a) (b)

Figure 41.2 The ‘hands-up’ manoeuvre to demonstrate arterial

obstruction in the thoracic outlet syndrome, observing the hand
for pallor after exercising, feeling for loss of the brachial pulse
and return with lowering the arm, and listening for a bruit as the
arm is lowered

with florid cases of arteritis.20 A list of investigations

screening for thrombophilia in patients with an uncertain
diagnosis30 is shown below (also see Chapters 2 and 20). (c) (d)

Figure 41.3 Allen test: (a) the blanching that occurs after
Investigations performed for a thrombophilia clenching the hand a few times with the radial and ulnar arteries
screen in a patient with acute upper limb occluded, (b) prompt return of colour after releasing the radial
ischaemia with uncertain diagnosis artery, (c) failure of colour to return in the hand if the distal radial
artery is occluded or (d) failure of colour to return to a finger if
more distal disease affects the metacarpal arteries
• Protein C
• Protein S
• Antiphospholipid antibodies Radiology
• Antithrombin III
• Factor III PLAIN X-RAYS
• Factor V Leiden
• Prothrombin mutation In patients with suspected thoracic outlet syndrome, a
• Homocysteine plain X-ray of the neck will detect most cervical ribs and
first rib abnormalities. Plain X-rays of the hands and chest
may show characteristic features of scleroderma, perhaps
leading to a barium swallow.
Non-invasive investigations
ARTERIOGRAPHY
Vascular laboratory techniques can be very useful. Colour
Doppler duplex ultrasound scanning has become the pre- Most patients with a clinical diagnosis of disease poten-
liminary investigation for examining arteries to the upper tially threatening the upper limb should undergo arteriog-
extremity.55 The vessels can be visualised from the aortic raphy. This invasive investigation is frequently required to
arch to the palmar arch, but it takes a highly trained ultra- diagnose the precise site of primary disease and its conse-
sonographer to interpret the results usefully. Duplex scan- quences in the upper extremity. This approach is quite
ning has been very well evaluated for the thoracic outlet different from that applied to the lower limb where arteri-
syndrome.56 The same cautions over eliciting extrinsic ography is usually performed only if intervention is con-
arterial compression by various clinical manoeuvres must templated. Most diseases described in this review have
be expressed, but at least it can be said that a negative specific arteriographic appearances which will form the
examination by a competent ultrasonographer essentially basis for appropriate further investigations, even to the
excludes the diagnosis of vascular compression as a feature point of arterial biopsy to confirm the cause of disease. For
of the thoracic outlet syndrome (also see Chapter 22). example, temporal artery biopsy is required to establish the
 Conservative measures 501

diagnosis of giant cell arteritis, which in turn may dictate IMMEDIATE MANAGEMENT
the need for arteriography of the cerebrovascular, visceral
or lower limb arteries.
Since the precise site of disease is frequently uncertain, The patient is usually first seen by a general practitioner or at
contrast studies usually commence with arch aortography a local district hospital where facilities for diagnosis and vas-
via the groin followed by selective catheterisation of the cular surgical treatment may be limited. An arterial ultra-
innominate or left subclavian arteries if required. Digital sound may have been ordered and its findings might be the
subtraction arteriography should demonstrate all vessels reason for a vascular surgical referral. In general, eliciting the
down to the fingers. This policy is preferable to axillary or cause will require considerable experience, and almost all
brachial artery puncture, the complications of which could patients will need urgent intervention to clear the occlusion
further compromise the circulation. and to treat the cause. Thus no more than immediate sup-
Arteriography may not be warranted and would simply portive treatment should be instituted before transferring the
delay treatment in a patient with acute ischaemia due to an patient to a vascular service as an emergency. Occasionally, a
embolus which has clearly come from the heart. Arterio- patient will sustain embolism to both the arm and the brain,
graphy is required, however, if the cause is not clear and if and present with simultaneous upper limb ischaemia and a
time permits, in order to search for a possible proximal stroke. As the stroke is likely to affect the side contralateral to
arterial source. There is an increasing role for arterio- the ischaemic arm, restoration of flow to this neurologically
graphy in subacute or chronic upper limb ischaemia where intact arm becomes even more important.
time permits the application of various interventional One important immediate measure is to reduce propa-
endovascular techniques. gation of distal thrombus by anticoagulation with intra-
The arteriographic features of atherosclerosis are well venous heparin. A bolus of 20 000 units or commencement
known. The absence of atheromatous changes and large of an intravenous infusion (5000 IU bolus and 1000 IU
plaques in the large arteries, and the demonstration of seg- per hour via a mechanical intravenous pump) should suf-
ments of smooth lesions, should lead to the suspicion of fice before transferring the patient. Remember to request
other arterial diseases. Angiography may simply show occlu- intravenous cannulation in the contralateral upper limb as
sion and give no indication of the characteristic changes use of the veins of the affected limb or of either leg is inap-
diagnostic of other specific diseases. Takayasu’s disease is propriate except in critical emergencies. Even in the rare
characterised by a combination of stenotic and occlusive instance where there is concurrent embolism to the brain,
lesions with fusiform or saccular aneurysms which are anticoagulation is appropriate in most patients to limit
frequently multiple.1,57,58 Classic features for Buerger’s intracranial propagation of clot, although this must be dis-
disease are abrupt occlusion with a ‘tree root’ configur- cussed with the vascular surgeon prior to transfer.
ation of collaterals and ‘corkscrew’ tortuosity of small distal
vessels apparently due to recanalisation59 (see Chapter 42).
CONSERVATIVE MEASURES
Giant cell arteritis shows long, smoothly tapered stenoses
or occlusions1,60 (see Chapter 43). In Crohn’s disease, typ-
ically a tapering stenosis leads to occlusion.27 The features In general, conservative treatment is considered after the
of fibromuscular hyperplasia are a characteristic ‘string of underlying cause of the ischaemia has been corrected, if that
beads’ appearance. In patients with suspected thoracic out- were possible. Oral anticoagulation, permanently for throm-
let syndrome, arteriography should include views with the boembolism due to atrial fibrillation and for a period of a
arm abducted and in external rotation, but the cautions few months in cases of myocardial infarction, will reduce the
expressed as to false positive results hold just as for other risk of recurrent embolism.5 Various vasodilator drugs,
diagnostic procedures. prostaglandin or its analogues, antiplatelet agents, and rheo-
logical agents such as pentoxifylline have been used in an
attempt to improve the distal circulation in patients with
OTHER SPECIAL INVESTIGATIONS
large vessel or small vessel disease in the upper extremity.
Computed tomography (CT) has been used to evaluate the Mills and colleagues3,62 contend that there is no objective evi-
thoracic outlet syndrome.61 It was found that if a plain dence to show that any of these measures improves the out-
X-ray showed a bony abnormality, a CT scan did not come, although empirically they do use a calcium blocking
add further information, except in symptomatic patients agent and pentoxifylline. A randomised study of iloprost, a
in whom subtle bony changes were detected. Computed longer acting prostaglandin analogue given intravenously for
tomography was also found to be accurate in demonstrat- 34 weeks, showed markedly higher rates of relief of pain and
ing non-osseous compression when compared with con- limb salvage in patients with Buerger’s disease.63 Various
trols. Nevertheless, the yield from these findings in patients studies of iloprost for critical ischaemia in the lower limb
with vascular manifestations is doubtful. The value of have shown benefit,64 and these results may well be extrapo-
magnetic resonance imaging in such cases has yet to be lated to the upper extremity. The various causes of arteritis
determined. or autoimmune disease generally require treatment with
502 Acute upper limb ischaemic states

corticosteroids with or without immunosuppressive bifurcation at or just below the elbow in order to allow
drugs.9,19,20,60 Regardless of all these considerations, the selective catheterisation of each of the branches through a
patient should stop smoking. Much can be gained from good transverse arteriotomy, with closure by interrupted sutures
local care to the hand with minimal debridement, nail exci- without a patch.15,16 If the embolus is arrested more prox-
sion to drain infection, and skin protection.3 imally, a higher brachial or even axillary approach may be
A protocol has been described for treatment after intra- used, either alone or in combination with the antecubital
arterial injection of drugs by drug abusers,43 and the meas- approach. In some cases the procedure ought to be per-
ures would seem appropriate for any patient with acute or formed under local anaesthesia in view of the appreciable
severe upper limb ischaemia. Treatment includes continu- mortality due to associated heart disease.6 It is rarely
ous intravenous heparin and dextran 40 infusion, par- appropriate to manage the problem conservatively (see
enteral dexamethasone, opiates for pain control, elevation Chapters 15 and 16).
of the extremity, and early active and passive movements. A personal preference of one author (KAM) is to expose
Vasodilators such as glyceryl trinitrate (GTN) can be a valu- the brachial artery in the mid-portion of the upper arm
able addition and may be given topically or intravenously as where its calibre is larger and to perform a longitudinal arte-
the severity of the situation dictates. Heparin, dextran and riotomy, which can usually be closed by direct suture without
GTN have the advantage of easy accessibility in a variety of arterial narrowing. The tip of the Fogarty catheter is angled
medical or hospital settings. Iloprost, pentoxifylline, dex- and the balloon is pre-expanded with contrast so that it can
amethasone and other vasoactive drugs have a greater role easily be seen on fluoroscopy to ensure selective catheterisa-
in chronic ischaemic states. tion of each run-off artery (Fig. 41.4). A study in cadavers
There has been considerable recent interest in the showed that a straight catheter passed into the larger of the
ischaemia-reperfusion syndrome and its management, and two outflow arteries in 42 per cent of limbs and into one
this is just as pertinent to the upper extremity as to artery only in 96 per cent of them, whereas positioning an
ischaemia of the lower limbs65 (see Chapter 2). Mannitol angled tip catheter blindly allowed it to pass down each artery
infusion may be required to maintain renal perfusion and in 87 per cent of limbs.66 The other author (GWS) prefers a
limit swelling and raised compartment pressure in the arm, curved incision starting just above the cubital fossa and
which would further compromise circulation. extending over the bicipital aponeurosis to the tendon of
biceps. The brachial artery is well exposed and extension of
the incision allows selective catheterisation of the radial and
RECONSTRUCTIVE SURGERY ulnar artery. The affected patient is frequently elderly and
often has poorly characterised cardiac disease so the opera-
Embolectomy tion is ideally, and easily, performed using local anaesthesia
injected by the surgeon with a little supplemental intra-
Since most emboli are lodged in the brachial artery, the venous sedation given by the anaesthetist. Local anaesthesia
most favoured approach is to expose this artery at its may be given as 10–20 mL of 1 per cent lidocaine (Xylocaine)

Fogarty
catheter

Brachial
artery

Figure 41.4 Two techniques for brachial

embolectomy to clear the distal arteries (the
Radial proximal embolus is removed last): the
artery conventional technique approaches the
Ulnar brachial bifurcation to allow each branch to
artery be dissected out and selectively catheterised;
the alternative technique uses a more limited
approach to the more proximal brachial
artery and uses fluoroscopy to guide the
catheter down each branch after angling the
Conventional Alternative tip and filling the balloon with a little
technique technique contrast
 Reconstructive surgery 503

but a particularly useful solution is 10 mL of 1 per cent lido- although with slightly lower long term patency rates.1
caine mixed with 10 mL 0.5 per cent bupivacaine (Marcain) Subclavian-carotid transposition (Fig. 41.7) or carotid-
as this provides rapid onset anaesthesia and lasts for a rea- subclavian bypass (Fig. 41.8) gives good results for treating
sonable period of time postoperatively. It is important that subclavian artery occlusions.69 Axilloaxillary crossover
the attending surgeon inform the patient’s family that the bypass (Fig. 41.9) or femoroaxillary bypass (Fig. 41.10) are
prognosis is grave. The 30-day results of upper limb reserved for more difficult situations although long term
embolectomy are nearly as bad as those for the lower limb results are satisfactory.70 It is rarely necessary to use a shunt
and a reflection of underlying cardiovascular disease. if the common carotid artery is clamped. A supraclavicular
Fasciotomy of the flexor and extensor compartments approach, combined with an infraclavicular approach if
should be considered if there is marked residual swelling required, gives excellent exposure and control of the sub-
after reperfusion but is rarely needed due to the usual early clavian artery, while appropriate longitudinal incisions are
recognition of acute limb ischaemia (see Chapter 33). used to expose more distal arteries.
There is preference for synthetic grafts rather than vein
Occlusive or aneurysmal disease grafts for reconstructions involving the subclavian or

It is frequently necessary to recommend surgical reconstruc-

tion for proximal occlusive or aneurysmal disease because
of major or repetitive minor embolism, acute thrombosis
causing severe ischaemia, a major risk of such events in the
future or associated cerebrovascular symptoms. The older
patient with asymptomatic left subclavian artery stenosis,
however, is often managed conservatively.
Transthoracic reconstruction is the most direct
approach and gives the best long term patency rates of over
90 per cent67,68 but at the expense of appreciable morbidity
and mortality.1 Endarterectomy (Fig. 41.5) is only suitable
for short lesions and is technically more demanding than
bypass (Fig. 41.6), which is thus usually preferred. The
approach is through a sternal split for the innominate
artery or the left third intercostal space for the left subcla-
vian artery (also see Chapter 34). Proximal aneurysms can
be resected or an inlay bypass technique can be used.
A range of extrathoracic reconstructions is more
widely used for occlusive disease. Excellent results have
Figure 41.6 Techniques used for transthoracic bypass from the
been reported with very low morbidity and mortality rates,
aorta to the major branches

1 2

Supraclavicular
incision

Figure 41.5 The approaches for transthoracic endarterectomy Figure 41.7 A technique for extrathoracic subclavian–carotid
of the innominate or left subclavian arteries transposition
504 Acute upper limb ischaemic states

proximal axillary artery69,71 whereas most prefer autoge- rates.77,78 Asymptomatic patients in the active phase of arter-
nous long saphenous vein if the distal anastomosis is to be itis are treated conservatively, tiding them through into the
taken down into the arm (Fig. 41.11).72–74 A collective phase of being ‘burned out’ before contemplating surgical
review considers that patency rates at 23 years are greater bypass, but the latter option may be appropriate in the acute
than 80 per cent if the distal anastomosis is proximal to the phase if there are limb-threatening complications.79
brachial bifurcation, but closer to 50 per cent if it is distal to
that level.75 Good results can be obtained even with bypass
to arteries beyond the wrist.76

Takayasu’s disease

In patients with Takayasu’s disease, bypass is best com-

menced at the proximal aortic arch to avoid anastomotic
stenosis and to achieve quite acceptable late patency

Supraclavicular
incision
From
common
femoral
artery

Figure 41.8 A technique for extrathoracic carotid–subclavian Figure 41.10 Proximal anastomosis for a femoroaxillary (or
bypass femorosubclavian) bypass

Infraclavicular
incision

Figure 41.9 A technique for extrathoracic

axilloaxillary (or subclavian-subclavian) bypass
 Endovascular management 505

Arterial catheter trauma

This usually affects the brachial artery and almost invariably

results from lifting an intimal flap, so that conservative
management accepting a diagnosis of arterial spasm is
fraught with danger (see Chapter 38). Although the circu-
lation may improve via collaterals, it is usually inadequate,
and it is almost invariably better to repair the injury by direct
exposure under local anaesthesia. Preoperative arteriography
is rarely needed. It may be possible to remove the intimal
flap and close the defect with a small vein patch, but some-
times a short interposition saphenous vein graft is needed.

Postmastectomy irradiation arteritis

Critical ischaemia due to occlusion of a long segment of

artery years after the injury demands restoration of flow (see
Chapter 33). Associated fibrotic damage to the subcutaneous
Figure 41.11 Multiple incisions required for a subclavian— tissues of the anterior chest wall leaves an inelastic, poorly
brachial vein bypass graft nourished and inhospitable milieu for bypass grafting. Instead
of forcing a graft through this field, an extra-anatomic
carotid-brachial vein bypass graft can be routed well clear of
the irradiated area, with excellent long term results.83
Thoracic outlet syndrome

There is general agreement that constricting structures at Radial artery fistula steal
the thoracic outlet should be removed if vascular symp-
toms are present because progressive damage to the sub- This interesting problem can be resolved by simple ligation
clavian artery and risk of acute thromboembolic of the fistula. Alternatively, ligate the artery immediately
complications are probable. There is increasing agreement distal to the surgical fistula and construct a reversed saphe-
that, in patients with vascular manifestations, such inter- nous vein bypass from the brachial artery proximal to the
vention should routinely involve removing the first rib as fistula to the radial artery in the hand (Fig. 41.12).84
well as a cervical rib, dividing all scalene muscle attach-
ments and other fibromuscular bands, in order to avert an Hypothenar hammer syndrome
appreciable risk of recurrence.35,80–82 One of several
approaches, whether supraclavicular, infraclavicular or Those with experience of this condition advocate resection
transaxillary, will provide the necessary exposure for of an ulnar artery aneurysm with end-to-end reconstruc-
removal of the first rib. Some favour a combination of the tion by autogenous vein to avoid downstream emboli.
supraclavicular and infraclavicular80 or the supraclavicular Conservative treatment is advocated in most patients if the
and transaxillary82 approach to decompressing the outlet, ulnar artery is thrombosed, reserving bypass for the occa-
and either combination appears to be essential if arterial sional patient with ischaemia of the ulnar fingers if the pal-
reconstruction is required.80 It should be possible to mar arch is inadequate.39
achieve satisfactory exposure without ever having to divide
the clavicle.
Management of arterial complications depends on the
ENDOVASCULAR MANAGEMENT
findings at operation.80 The dilated or stenosed segment of
artery should be opened and propagated thrombus or
embolus removed using a Fogarty catheter. If the intima Balloon dilatation or stenting
appears normal, the arteriotomy can be simply closed, tak-
ing bites large enough to restore the lumen to a normal Several series report results for balloon dilatation of
diameter. If the intima is ulcerated, limited endarterec- innominate or subclavian stenosis or occlusion, with pri-
tomy and closure, primary or using a vein patch, may be mary success rates in excess of 90 per cent and acceptable
sufficient. If damage is too extensive, resection and end-to- long term results for either atherosclerosis85–88 or aortoar-
end anastomosis will not be feasible without tension, and teritis.89 Most, however, would now prefer to stent these
therefore an interposition graft should be used to replace lesions for better long term outcome, sometimes to resolve
the damaged segment. problems after balloon dilatation but most often as elective
506 Acute upper limb ischaemic states

Surprisingly few authors discuss the benefits of adding

cervical sympathectomy to thoracic outlet decompression
in patients with vascular complications. Cervical sympathe-
(a) (b) ctomy is, however, the mainstay of treatment for Buerger’s
disease in the upper limbs.101 As in the lower limb, sympa-
thectomy will be ineffective as the sole treatment of frank
tissue necrosis.

Conclusions

For most surgeons, experience with upper limb arterial

disease and acute ischaemia in particular, is limited so
that the guidelines for successful management are less
clear than are those for lower limb ischaemia. The con-
sequences from inappropriate management, however,
are likely to be far more serious. Accordingly, it is essen-
tial that each patient be carefully assessed even if the pre-
sentation appears to be only mildly symptomatic because
severe ischaemia can suddenly supervene and an opti-
mal chance for successful treatment will have been lost.
Figure 41.12 (a) An operation performed for a patient with steal Presentation with acute ischaemia almost invariably
from the hand into a radial artery fistula created for dialysis. (b) requires immediate surgical or other effective interven-
The radial artery distal to the fistula is ligated and a vein graft is tion, as the results of conservative treatment are poor.
taken from the brachial to the distal radial artery

treatment.90–93 The approach may be from the femoral

artery or retrogradely from a cut-down on the brachial Key references
artery.94 There is some early experience with covered stent-
grafts for thoracic outlet arterial injuries and aneurysms.95,96 Al-Mubarak N, Liu MW, Dean LS, et al. Immediate and late
outcomes of subclavian artery stenting. Catheter Cardiovasc
Intervent 1999; 46: 169–72. Contemporary experience and
Thrombolysis techniques for supra-aortic endovascular procedures.
Berguer R, Morasch MD, Kline RA, et al. Cervical reconstruction of
Thrombolysis, particularly using urokinase, has been exten- the supra-aortic trunks: a 16-year experience. J Vasc Surg
sively used in this context97 (see Chapter 16). Successful 1999; 29: 239–46. An experience of a large series of supra-
lysis may reveal a stenosis which can then be managed by aortic reconstructions.
Nehler MR, Taylor LM, Moneta GL, Porter JM. Upper extremity
balloon dilatation, but more extensive disease is frequently
ischemia from subclavian artery aneurysm caused by bony
present indicating the need for surgical reconstruction.
abnormalities of the thoracic outlet. Arch Surg 1997; 132:
Indeed, some highly respected authorities argue that defini- 527–32. A comprehensive analysis of arterial lesions in the
tive early embolectomy, thrombectomy or bypass surgery thoracic outlet syndrome.
is more effective.98 Thrombolysis may be effective in clear- Robbs JV, Kadwa AM, AbdoolCarrim ATO. Arterial reconstruction for
ing emboli lodged distally in the small arteries prior to Takayasu’s arteritis: medium to long term results. Eur J Vasc
reconstruction, as for example when such a complication is Surg 1994; 8: 401–7. An impressive experience of Takayasu’s
associated with the thoracic outlet syndrome,99 and has disease from South Africa.
been successfully used to treat ischaemia from Buerger’s Roddy SP, Darling RC, Chang BB, et al. Brachial artery recon-
disease100 (see Chapter 42). struction for occlusive disease: a 12-year experience. J Vasc
Surg 2001; 33: 802–5. A good overview of results to be
expected from bypass grafts in the arm, with some differences
Cervical sympathectomy from lower limb bypasses being discussed.
Sullivan TM, Gray BH, Bacharach JM, et al. Angioplasty and primary
The literature appears to reveal a growing disenchantment stenting of the subclavian, innominate, and common carotid
with the idea of adding cervical sympathectomy to a vascu- arteries in 83 patients. J Vasc Surg 1998; 28: 1059–65. An
excellent review of a large experience of endovascular treatment
lar reconstructive procedure in the neck, or of using it for
of upper limb arterial disease from a major American centre.
ischaemia caused by small vessel disease in the hands.3,62
 References 507

REFERENCES 23 O’Duffy JD. Vasculitis in Behçet’s syndrome. Rheum Dis Clin

North Am 1990; 16: 423–31.
24 Chaillou P, Patra P, Noel S, et al. Behçet’s disease revealed by double
1 Fujitani RM, Mills JL. Acute and chronic upper extremity peripheral arterial involvement. Ann Vasc Surg 1992; 6: 160–3.
ischemia. I. Large vessel arterial occlusive disease. Ann Vasc Surg 25 Tuzun H, Sayin A, Karaozbek Y, et al. Peripheral aneurysms in
1993; 7: 106–12. Behçet’s disease. Cardiovasc Surg 1993; 1: 220–1.
2 Ljungman C, Adami HO, Bergqvist D, et al. Time trends in 26 Merino J, Casanueva B, Piney E, et al. Hemiplegia and peripheral
incidence rates of acute, nontraumatic extremity ischaemia; a gangrene secondary to large and medium size vessel involvement
population based study during a 19 year period. Br J Surg 1991; in CREST syndrome. Clin Rheumatol 1982; 1: 295–9.
78: 857–60. 27 SheehanDare RA, Goodfield MJD, Wilson PD, Rowell NR. Axillary
3 Mills JL, Fujitani RM. Acute and chronic upper extremity artery occlusion as a presenting feature of Crohn’s disease.
ischemia. II. Small vessel arterial occlusive disease. Ann Vasc Surg Postgrad Med J 1989; 65: 758–60.
1993; 7: 195–9. 28 Reilly JM, McGraw DJ, Sicard GA. Bilateral brachial artery
4 AbuRahma AF, Downham L. The role of paradoxical arterial fibromuscular dysplasia. Ann Vasc Surg 1993; 7: 483–7.
emboli of the extremities. Am J Surg 1996; 172: 214–17. 29 Wali MA, Dewan M, Renno WM, Ezzeddin M. Mucoid
5 Davies MG, O’Malley K, Feeley M, et al. Upper limb embolus: a degeneration of the brachial artery: case report and a review of
timely diagnosis. Ann Vasc Surg 1991; 5: 857. literature. J R Coll Surg Edinb 1999; 44: 126–9.
6 Vohra R, Lieberman DP. Arterial emboli to the arm. J R Coll Surg 30 Kottke-Marchant K. Genetic polymorphisms associated with
Edinb 1991; 36: 835. venous and arterial thrombosis: an overview. Arch Pathol Lab
7 Fields WS, Lemak NA. Joint study of extracranial arterial Med 2002; 126: 295–304.
occlusion. VII. Subclavian steal: a review of 168 cases. JAMA 31 Nehler MR, Taylor LM, Moneta GL, Porter JM. Upper extremity
1972; 222: 1139–43. ischemia from subclavian artery aneurysm caused by bony
8 Baguneid M, Dodd D, Fulford P, et al. Management of acute abnormalities of the thoracic outlet. Arch Surg 1997; 132: 527–32.
nontraumatic upper limb ischemia. Angiology 1999; 50: 715–20. 32 Gelabert HA, Machleder HI. Diagnosis and management of
9 Scott DF, Myers KA, Lord RSA. Nonatherosclerotic diseases arterial compression at the thoracic outlet. Ann Vasc Surg 1997;
causing lower limb ischaemia. In: Myers KA, Nicolaides AN, 11: 359–66.
Summer DS (eds). Lower Limb Ischaemia. London: MedOrion 1996. 33 Pollak EW. Surgical anatomy of the thoracic outlet syndrome.
10 Hunder GG. Vasculitic syndromes. Curr Opin Rheumatol 1990; 2: Surg Gynecol Obstet 1980; 150: 97–103.
479. 34 Baumgartner F, Nelson RJ, Robertson JM. The rudimentary first
11 Lie JT, Buerger’s disease and inflammatory aspects of rib; a cause of thoracic outlet syndrome with arterial
atherosclerosis. Curr Opin Rheumatol 1990; 2: 76–80. compromise. Arch Surg 1989; 124: 109–12.
12 Sharma S, Rajani M, Kamalakar T, et al. The association between 35 Rohrer MJ, Cardullo PA, Pappas AM, et al. Axillary artery
aneurysm formation and systemic hypertension in Takayasu’s compression and thrombosis in throwing athletes. J Vasc Surg
arteritis. Clin Radiol 1990; 42: 18–27. 1990; 11: 761–9.
13 Sharma S, Rajani M, Shrivastava S, et al. Nonspecific aorto- 36 McCarthy WJ, Yao JST, Schafer MF, et al. Upper extremity arterial
arteritis (Takayasu’s disease) in children. Br J Radiol 1991; 64: injury in athletes. J Vasc Surg 1989; 9: 317–27.
690–8. 37 Lee AW, Hopkins SF, Griffen WO. Axillary artery aneurysm as an
14 Joyce JW. Buerger’s disease (thromboangiitis obliterans). Rheum occult source of emboli to the upper extremity. Am Surgeon
Dis Clin North Am 1990; 16: 463–7. 1987; 53: 485.
15 Shionoya S. Buerger’s disease: diagnosis and management. 38 Pineda CJ, Weisman WH, Bookstein JJ, et al. Hypothenar hammer
Cardiovasc Surg 1993; 1: 207–14. syndrome. Am J Med 1985; 79: 561–70.
16 Olin JW, Young JR, Graor RA, et al. The changing clinical 39 Vayssairat M, Debure C, Cormier JM, et al. Hypothenar hammer
spectrum of thromboangiitis obliterans (Buerger’s disease). syndrome: seventeen cases with long-term follow-up. J Vasc
Circulation 1990; 82: 38. Surg 1987; 5: 83–8.
17 Disdier P, Granel B, Serratrice J, et al. Cannabis arteritis 40 Bovenzi M. Vibration-induced white finger and cold response
revisited – ten new case reports. Angiology 2001; 52: 1–5. of digital arterial vessels in occupational groups with various
18 Brack A, Martinez-Taboada V, Stanson A, et al. Disease pattern in patterns of exposure to hand-transmitted vibration.
cranial and large-vessel giant cell arteritis. Arthritis Rheum 1999; Scand J Worlk Environ Health 1998; 24: 138–44.
42: 311–17. 41 McCready RA. Upper extremity vascular injuries. Surg Clin North
19 Ninet JP, Bachet P, Dumontet CH, et al. Subclavian and axillary Am 1988; 68: 7250.
involvement in temporal arteritis and polymyalgia rheumatica. 42 Ott B, Neuberger O, Frey HP. Obliteration of the axillary artery
Am J Med 1990; 88: 13–20. after axillary block. Anaesthesia 1989; 44: 773–4.
20 Paice EW. Giant cell arteritis: difficult decisions in diagnosis, 43 Treiman GS, Yellin AE, Weaver FA, et al. An effective treatment
investigation and treatment. Postgrad Med J 1989; 65: protocol for intraarterial drug injection. J Vasc Surg 1990; 12:
743–7. 456–66.
21 Greene GM, Lain D, Sherwin R, et al. Giant cell arteritis of the 44 Royse AG, Royse CF, Shah P, et al. Radial artery harvest
legs: clinical isolation of severe disease with gangrene and technique, use and functional outcome. Eur J Cardiothorac Surg
amputations. Am J Med 1986; 81: 727–33. 1999; 15: 186–93.
22 Tsokos M, Lazarou SA, Moutsopoulos M. Vasculitis in primary 45 Farina C, Schultz RD, Feldhaus RJ. Late upper limb acute
Sjögren’s syndrome: histologic classification and clinical ischemia in a patient with an occluded axillofemoral bypass
presentation. Am J Clin Pathol 1987; 88: 26–31. graft. J Cardiovasc Surg 1990; 31: 178–81.
508 Acute upper limb ischaemic states

46 White GH, Donayre CE, Williams RA, et al. Exertional disruption 68 Taha AA, Vahl AC, de Jong SC, et al. Reconstruction of the
of axillofemoral graft anastomosis: `the axillary pullout supra-aortic trunks. Eur J Surg 1999; 165: 314–18.
syndrome’. Arch Surg 1990; 125: 625–7. 69 AbuRahma AF, Robinson PA, Jennings TG. Carotid-subclavian
47 Cuschieri RJ, Vohra R, Leiberman DP. Acute ischaemia in the bypass grafting with polytetrafluoroethylene grafts for
donor limb after occlusion of axillofemoral grafts. Eur J Vasc Surg symptomatic subclavian artery stenosis or occlusion: a 20-year
1989; 3: 267–9. experience. J Vasc Surg 2000; 32: 411–18.
48 Hartman AR, Fried KS, Khalil I, Riles TS. Late axillary artery 70 Mingoli A, Sapienza P, Feldhaus RJ, et al. Long-term results and
thrombosis in patients with occluded axillary femoral bypass outcomes of crossover axilloaxillary bypass grafting: a 24-year
grafts. J Vasc Surg 1985; 2: 285–7. experience. J Vasc Surg 1999; 29: 894–901.
49 Andros G, Schneider PA, Harris RW, et al. Management of arterial 71 Ziomek S, QuinonesBaldrick WJ, Busuttil RW, et al. The
occlusive disease following radiation therapy. Cardiovasc Surg superiority of synthetic arterial grafts over autologous veins in
1996; 4: 135–42. carotidsubclavian bypass. J Vasc Surg 1986; 3: 140–5.
50 Glazer G, Myers KA, Davies ER. Ergot poisoning. Postgrad Med J 72 Katz SG, Kohl RD. Direct revascularization for the treatment of
1966; 42: 562–8. forearm and hand ischemia. Am J Surg 1993; 165: 312—16.
51 Palombo D, Mirelli M, Peinetti F, et al. Spasm of arm arteries due 73 Roddy SP, Darling RC, Chang BB, et al. Brachial artery
to ergotamine tartrate. Int Angiology 1991; 10: 513. reconstruction for occlusive disease: a 12-year experience.
52 Morsy AH, Kulbaski M, Chen C, et al. Incidence and J Vasc Surg 2001; 33: 802–5.
characteristics of patients with hand ischemia after a 74 Ristow AV, Cury JM, Costa EL, et al. Revascularization of the
hemodialysis access procedure. J Surg Res 1998; 74: 8–10. ischaemic hand using in situ veins. Cardiovasc Surg 1996;
53 Baxter BT, Blackburn D, Payne K, et al. Noninvasive evaluation 4: 466–9.
of the upper extremity. Surg Clin North Am 1990; 70: 75 McCarthy WJ, Flinn WR, Yao JST, et al. Result of bypass grafting
87–97. for upper limb ischemia. J Vasc Surg 1986; 3: 741–6.
54 Harris J, Huang W, Tyrer O, et al. Clinical and photo- 76 Nehler MR, Dalman RI, Harris EJ, et al. Upper extremity arterial
plethysmographic assessment of thoracic outlet arterial bypass distal to the wrist. J Vasc Surg 1992; 16: 633–42.
compression. J Vasc Technol 1989; 13: 203. 77 Giordano JM, Leavitt RY, Hoffman G, Fauci AS. Experience with
55 Taneja K, Jain R, Sawhney S, Rajani M. Occlusive arterial disease surgical treatment of Takayasu’s disease. Surgery 1991; 109:
of the upper extremity: colour Doppler as a screening technique 252–8.
and for assessment of distal circulation. Australasian Radiology 78 Weaver FA, Yellin AE, Campen DH, et al. Surgical procedures in
1996; 40: 226–9. the management of Takayasu’s arteritis. J Vasc Surg 1990; 12:
56 Longley DG, Yedlicka JW, Molina EJ, et al. Thoracic outlet 429–31.
syndrome: evaluation of the subclavian vessels by color duplex 79 Robbs JV, Kadwa AM, AbdoolCarrim ATO. Arterial reconstruction
sonography. Am J Radiol 1992; 158: 623–30. for Takayasu’s arteritis: medium to long term results. Eur J Vasc
57 Matsumura K, Hirano T, Takeda K, et al. Incidence of aneurysms Surg 1994; 8: 401–7.
in Takayasu’s arteritis. Angiology 1991; 42: 308–15. 80 Cormier JM, Amrane M, Ward A, et al. Arterial complications of
58 Kumar S, Subramanyan R, Ravi Mandalam K, et al. Aneurysmal the thoracic outlet syndrome: fifty-five operative cases. J Vasc
form of aortoarteritis (Takayasu’s disease): analysis of thirty Surg 1989; 9: 778–87.
cases. Clin Radiol 1990; 42: 342–7. 81 Colman PD, White GH. Complexities in the management of
59 Mills JL, Taylor LM, Porter JM. Buerger’s disease in the modern arterial compromise due to thoracic outlet syndrome. Aust
era. Am J Surg 1987; 154: 123–9. N Z J Surg 1990; 60: 100–2.
60 Perruquet JL, David DE, Harrington TM. Aortic arch arteritis in 82 Thompson JF, Webster JHH. First rib resection for vascular
the elderly: an important manifestation of giant cell arteritis. complications of thoracic outlet syndrome. Br J Surg 1990;
Arch Int Med 1986; 146: 289–91. 77: 555–7.
61 Bilbey JH, Muller NL, Connell DG, et al. Thoracic outlet syndrome: 83 McCallion WA, Barros D’Sa AAB. Management of critical limb
evaluation with CT. Radiology 1989; 171: 381–4. ischaemia long after irradiation injury of the subclavian and
62 Mills JL, Friedman EL, Taylor LM, Porter JM. Upper extremity axillary arteries. Br J Surg 78: 136–8.
ischemia caused by small artery disease. Ann Surg 1987; 206: 84 Berman SS, Gentile AT, Glickman MH, et al. Distal
521–8. revascularization-interval ligation for limb salvage and
63 Fiessinger JN, Schafer M. Trial of iloprost versus aspirin maintenance of dialysis access in ischemic steal syndrome.
treatment for critical limb ischaemia of thromboangiitis J Vasc Surg 1997; 26: 393–402.
obliterans. Lancet 1990; 335: 555–7. 85 RL, Lambert D, Chamberlain J, et al. Percutaneous transluminal
64 Second European Consensus Document on Chronic Critical Leg angioplasty of the innominate, subclavian, and axillary arteries.
Ischaemia. Eur J Vasc Surg 1992; 6(suppl A): 18–21. Eur J Vasc Surg 1990; 4: 591–5.
65 Crinnion JN, Homer-Vanniaskinkam S, Gough MJ. Skeletal muscle 86 Hebrang A, Maskovic J, Tomac B. Percutaneous transluminal
reperfusion injury: pathophysiology and clinical considerations. angioplasty of the subclavian arteries: longterm results in
Cardiovasc Surg 1993; 1: 317–24. 52 patients. Am J Radiol 1991; 156: 109–14.
66 Beckingham IJ, Roberts SNJ, Berridge DC, et al. A simple 87 Millaire A, Trinca M, Marache P, et al. Subclavian angioplasty:
technique for thromboembolectomy of the upper limb. immediate and late results in 50 patients. Catheter Cardiovasc
Eur J Vasc Surg 1990; 4: 173–7. Diagn 1993; 29: 81–7.
67 Berguer R, Morasch MD, Kline RA, et al. Cervical reconstruction 88 Selby JB, Matsumoto AH, Tegtmeyer CJ, et al. Balloon angioplasty
of the supra-aortic trunks: a 16-year experience. J Vasc Surg above the aortic arch: immediate and longterm results. Am J
1999; 29: 239–46. Radiol 1993; 160: 631–5.
 References 509

89 Tyagi S, Verma PK, Gambhir DS, et al. Early and long-term results 96 Park JH, Chung JW, Joh JH, et al. Aortic and arterial aneurysms
of subclavian angioplasty in aortoarteritis (Takayasu disease): in Behçet disease: management with stent-grafts – initial
comparison with atherosclerosis. Cardiovasc Intervent Radiol experience. Radiology 2001; 220: 745–50.
1998; 21: 219–24. 97 Johnson SP, Durham JD, Subber SW, et al. Acute arterial
90 Harris NJ, Cameron I, Beard JD, Gaines P. Percutaneous stenting occlusions of the small vessels of the hand and forearm:
of proximal subclavian artery occlusion. Eur J Vasc Endovasc Surg treatment with regional urokinase therapy. J Vasc Intervent
1995; 9: 479–80. Radiol 1999; 10: 869–76.
91 Al-Mubarak N, Liu MW, Dean LS, et al. Immediate and late 98 Pemberton M, Varty K, Nydahl S, Bell PR. The surgical
outcomes of subclavian artery stenting. Catheter Cardiovasc management of acute limb ischaemia due to native vessel
Intervent 1999; 46: 169–72. occlusion. Eur J Vasc Endovasc Surg 1999; 17: 72–6.
92 Rodriguez-Lopez JA, Werner A, Martinez R, et al. Stenting for 99 Widlus DM, Venbrux AC, Benenati JF, et al. Fibrinolytic therapy
atherosclerotic occlusive disease of the subclavian artery. Ann for upper extremity arterial occlusions. Radiology 1990; 175:
Vasc Surg 1999; 13: 254–60. 393–9.
93 Sullivan TM, Gray BH, Bacharach JM, et al. Angioplasty and 100 Lang EV, Bookstein JJ. Accelerated thrombolysis and
primary stenting of the subclavian, innominate, and common angioplasty for hand ischemia in Buerger’s disease. Cardiovasc
carotid arteries in 83 patients. J Vasc Surg 1998; 28: 1059–65. Intervent Radiol 1989; 12: 957.
94 Whitbread T, Cleveland TJ, Beard JD, Gaines PA. A combined 101 Sayin A, Bozkurt AK, Tuzun H, et al. Surgical treatment of
approach to the treatment of proximal arterial occlusions of the Buerger’s disease: experience with 216 patients. Cardiovasc
upper limb with endovascular stents. Eur J Vasc Endovasc Surg Surg 1993; 1: 377–80.
1998; 15: 29–35.
95 du Toit DF, Strauss DC, Blaszczyk M, et al. Endovascular
treatment of penetrating thoracic outlet arterial injuries. Eur
J Vasc Endovasc Surg 2000; 19: 489–95.
This page intentionally left blank
 42
Emergency Aspects of Buerger’s Disease

SEKAR NATARAJAN, DHANESH KAMERKAR

The problem 511 Laboratory tests 514

Aetiology 511 Angiography 514
Pathology 512 Natural course 515
Clinical features 513 Treatment 517
Acute presentation 513 References 520
Diagnostic criteria 513

THE PROBLEM The incidence of Buerger’s disease has been steadily

decreasing in the developed nations. At the Mayo Clinic
the incidence had regressed from 107 per 105 patients regis-
Buerger’s disease or thromboangiitis obliterans (TAO) is a tered in 1947 to 12.6 per 105 patients registered in 1985.2
non-atherosclerotic segmental inflammatory arterial dis- By contrast, the incidence of the disease in women world-
ease which mostly affects small and medium sized arteries. wide has been steadily increasing and is a reflection of the
It is most commonly encountered in young adults and is rise in numbers of women smokers.3
closely identified with tobacco usage. Von Winiwarter, in
1879, was the first to describe Buerger’s disease on discover-
ing ‘endarteritis’ and ‘endophlebitis’ in an amputated limb. AETIOLOGY
Much later, in 1908, Leo Buerger provided a detailed and
accurate description of the disease, differentiating it from
atherosclerosis, and called it ‘thromboangiitis obliterans’. The aetiology of TAO is, in short, still unknown. Although
Even though it has a worldwide distribution the condition TAO is grouped under the miscellaneous category of
is much more common in countries in Asia such as Israel, vasculitis, it differs from the more commonly encountered
India and Korea than in the USA or Europe. Cachovan in variety. Its association with smoking has been well docu-
19861 reported the incidence of Buerger’s disease as a pro- mented but the role of other aetiological factors such as
portion of peripheral arterial occlusive disease in different autoimmune mechanisms, genetic factors and hypercoag-
countries (Table 42.1). ulability states is not very clear.
There is a very strong association between tobacco
usage and TAO. Patients with TAO show a higher incidence
Table 42.1 Geographical incidence of Buerger’s disease1 of delayed hypersensitivity reaction to tobacco glycoprotein.
Papa and colleagues4 demonstrated a similar response in
Country Incidence (per cent)
healthy smokers whereas non-smokers did not respond at
all. This effect depends on individual hypersensitivity rather
Asia Israel 80.0
India 45.0–63.0 than on the amount of tobacco consumed. Progression of
Korea 16.0–66.0 disease and results of treatment are closely linked to the ces-
sation of smoking. Measurement of urinary cotinine and
Europe France 1.2–5.6 carboxyhaemoglobin in blood can be used to establish
Former West Germany 0.5–5.0
whether the patient has really stopped smoking.
Switzerland 1.0–3.0
In India Buerger’s disease is commonly observed in
UK 0.25
people who are poor and who smoke raw tobacco in the
512 Emergency aspects of Buerger’s disease

form of ‘beedies’5 but quite unusual in the richer cigarette Chronic phase
smokers who seem to develop atherosclerosis. Hence there
must be other factors such as nutritional status and infection During the chronic stage there is more extensive recanali-
which may play a role in the pathogenesis of TAO. As only a sation of the thrombus and increased fibrous tissue forma-
small number of smokers develop TAO, a lot of work has tion in the media and adventitia. The vessel becomes thin
concentrated on identifying a possible genetic predisposi- and cord-like. The elastic lamina may show minimal frag-
tion, but so far no definitive or consistent genetic cause has mentation but the most important feature is that the general
been detected. architecture of the arterial wall is well preserved. Calcification
Several studies have confirmed the presence of serum never occurs in Buerger’s disease (Figs 42.1 and 42.2).
anticollagen, antielastin, antiendothelial cell antibodies and In patients over the age of 40 years, atherosclerosis can
circulating immune complexes6–9 but a ‘cause and effect’ coexist with TAO and that has caused some confusion
relationship remains to be discovered. Chaudhry and col- about the exact diagnosis. Progression of the disease can be
leagues10 demonstrated that the level of urokinase plas- continuous and creeping or take the form of skip lesions.
minogen activator was elevated twofold and that free In the early stages the disease is segmental and starts in the
plasminogen activator inhibitor-1 was lower in patients medium and small sized muscular arteries of the extrem-
with Buerger’s disease – observations indicating some ities. It can also begin in small branches of major proximal
form of endothelial derangement. Makita et al.11 demon- arteries such as the iliac gradually extending proximally to
strated impaired endothelium dependent vasorelaxation in
the peripheral vasculature of patients with TAO, and it is
recognised that vasospasm is a prominent feature of the
disease. Attention has now been drawn to the presence of
antiphospholipid antibodies and hyperhomocysteinaemia
in patients with Buerger’s disease but the significance of
these findings is still not clear.

PATHOLOGY

Buerger’s disease is an inflammatory occlusive disease

involving both arteries and veins. It is usually segmental in
distribution with characteristic skip lesions. Pathologically,
three distinct phases in the course of the disease have been
identified.
Figure 42.1 Tibial artery with recanalised thrombus
(haematoxylin and eosin 25). Note vessel architecture is
Acute phase maintained

This is the classic presentation of Buerger’s disease and dis-

tinguishes it from atherosclerosis. The vessel is swollen
with thrombus and the arterial wall and periarterial tissue are
oedematous. Characteristically, there is evidence of trans-
mural infiltration by inflammatory cells and of microab-
scesses with many multinucleated giant cells. Cellular
thrombus is unique to Buerger’s disease. The adjacent vein
may also show similar features with thrombophlebitis.
During this stage the patient presents with features of acute
ischaemia, tenderness over the artery and inflammatory
changes in the skin over the artery.

Intermediate phase

The acute stage is followed by an intermediate phase in which

there is progressive organisation of the occlusive thrombus Figure 42.2 Close-up view of specimen in Fig. 42.1 showing
with recanalisation. The microabscesses disappear but cellular thrombus and exaggerated coiling of internal elastic
giant cell infiltration persists. lamina (haematoxylin and eosin 160)
 Diagnostic criteria 513

involve major trunks up as far as the aorta. Importantly, In a Cleveland series3 63 per cent of patients demon-
therefore, proximal artery involvement should not be strated an abnormal Allen’s test. Cold sensitivity due to
viewed as a finding which excludes a diagnosis of Buerger’s markedly increased sympathetic activity was observed in
disease. The involvement of cerebral, coronary and visceral about 40 per cent of patients but Raynaud’s phenomenon
vessels in TAO, a process extending even to vein bypass has not been noted as often in other series.16
grafts, has been documented.12–14

CLINICAL FEATURES ACUTE PRESENTATION

Buerger’s disease typically manifests in the form of acute

Buerger’s disease predominantly affects young males who
exacerbations triggered by the sudden onset of segmental
are heavy smokers. There is a strong association between
thrombosis which is usually accompanied by vague sys-
smoking and TAO but the disease is also known to occur in
temic symptoms such as low grade fever and malaise. The
non-smokers. The onset of disease is usually before the age of
patient presents with signs of acute ischaemia or a worsen-
50 and is now increasingly seen in women who use tobacco.
ing of pre-existing ischaemia. Matsushita et al.17 found that
The disease usually commences in the infrapopliteal ves-
acute exacerbation is closely associated with high circulat-
sels and is characterised by acute exacerbations and remis-
ing levels of nicotine. It is important to differentiate this
sions. As the disease progresses, more proximal arteries
picture from embolism, especially as catheter embolectomy
become involved, and the symptoms change accordingly.
for acute thrombosis in Buerger’s disease is likely to be
The classic clinical features in the early stages are foot and
unsuccessful and usually followed by a deterioration of the
instep claudication and migratory superficial throm-
ischaemia. A prior history of claudication and the presence
bophlebitis. Rubor and trophic changes are also often
of arterial occlusions in other limbs helps to differentiate it
observed. Digital ulcers are common but ulceration and
from embolism.
gangrene do not always follow the onset of claudication
but may actually precede it. Frequently, ulcer or gangrene
associated with rest pain is precipitated by trivial trauma or
DIAGNOSTIC CRITERIA
minor surgery such as the removal of a nail.
In an analysis of 112 patients Olin et al.3 documented
the typical pattern and frequency of the clinical features The diagnosis of Buerger’s disease is primarily based on the
of TAO (Table 42.2), rest pain, ischaemic ulceration and clinical picture. Shionoya15 suggested a set of classic cri-
claudication being the most prominent observed. teria which include:
Very often more than one limb is involved. In
Shionoya’s report15 the prevalence of multiple limb • onset of disease below the age of 50 years
involvement was as follows: two limbs in 16 per cent, three • a process particularly involving the infrapopliteal vessels
but frequently affecting both upper and lower limbs
limbs in 41 per cent and all four limbs in 43 per cent of
patients. Involvement of the upper limb is also quite com- • migrating thrombophlebitis
mon but is usually missed because the vessels most fre- • absence of atherogenic risk factors apart from smoking.
quently involved, namely, the ulnar and palmar arteries, Papa et al.18 suggested a point scoring system and Mills
are not always checked. Nonetheless, in most series, the and Porter19 proposed major and minor criteria for the diag-
prevalence of upper limb involvement is 45–50 per cent, nosis of Buerger’s disease. Some confusion has existed in
which is high. defining these criteria largely because many patients over
the age of 50 also display the behaviour and the angio-
graphic findings typical of TAO. Histopathologically, the
Table 42.2 Pattern and frequency of clinical features of features of atherosclerosis associated with age have only
thromboangiitis obliterans 3 added to the confusion. Sasaki et al.20 reported an inci-
dence of associated hypertension in 13.5 per cent, diabetes
Clinical features Frequency (per cent) in 4.4 per cent, and hyperlipidaemia in 7.9 per cent of their
TAO patients, and 15 per cent of them were over 50 years of
Intermittent claudication 70 (63) age at the time of diagnosis. In an analysis, Vink21 identified
Rest pain 91 (81) less than 2 per cent of patients whose symptoms started after
Ischaemic ulcer 85 (76) the age of 50 years while that was true of 29 per cent of
Upper extremity involvement 24 (28) patients in the series reported by Olin et al.3 Moreover, in
Lower extremity involvement 39 (46) addition to infrapopliteal involvement, the disease not
Both upper and lower extremities 22 (26)
uncommonly affects major proximal arteries. In one of our
Thrombophlebitis 43 (38)
series,22 26 per cent of patients presented with aortoiliac dis-
Raynaud’s 49 (44)
ease and 42 per cent with femoropopliteal disease. In India,
514 Emergency aspects of Buerger’s disease

involvement of the external iliac artery is quite often

observed.

LABORATORY TESTS

There are no specific diagnostic laboratory tests for Buerger’s

disease. Indicators of active inflammation include an ele-
vated erythrocyte sedimentation rate and raised C-reactive
protein. Immunological tests such as R factor, anti-nuclear
antibody, etc., help to rule out other systematic vasculitic
syndromes. Tests for hypercoagulability, including antiphos-
pholipid antibodies and homocysteinaemia, are also of
value. Routine electrocardiogram and echocardiogram will
help to identify any cardiac illness as well as rule out a pos-
sible source of embolism. Angiographic findings are usually
pathognomonic of Buerger’s disease.

ANGIOGRAPHY

The characteristic findings of TAO on angiography are Figure 42.4 View of distal arterial tree taken from the
seen in small and medium sized vessels such as the tibial, arteriogram in Fig. 42.3. Note the reformation of tibial vessels with
pedal, ulnar and palmar arteries, usually with apparently segmental occlusion and ‘corkscrew’ collaterals
normal large proximal arteries and the absence of athero-
matous changes. Key angiographic features are segmental
arterial occlusions, interspersed with normal sections; the
artery occludes with an abrupt cut-off, from which point
leashes of collaterals arise giving a characteristic ‘tree root’
appearance. The collaterals result in part from the recanalisa-
tion of vessels and dilated vasa vasorum which explains their
‘corkscrew’ shape (Figs 42.3–42.6), sometimes described as
‘accordion’ or ‘corrugated’ in appearance. Typically, early
venous filling also takes place. In general, the frequency
with which abrupt occlusion of arteries of the lower limb is

Figure 42.5 Arteriogram of a 30-year-old man with Buerger’s

Figure 42.3 Arteriogram of a 30-year-old man showing abrupt disease. Popliteal and femoral artery showing extensive
cut-off of superficial femoral artery. Note normal proximal artery recanalisation. Note the ‘corkscrew’ collaterals
 Natural course 515

Figure 42.6 View of arterial tree of mid-lower leg taken from

the arteriogram in Fig. 42.5 showing only collaterals

Table 42.3 Anatomical prevalence of abrupt arterial occlusion

in thromboangiitis obliterans of the lower limbs15 Figure 42.7 Arteriogram of a 40-year-old man with Buerger’s
disease. Profunda femoris artery showing early inflammatory
Artery Incidence (per cent) changes. Main trunk of profunda occluded

Superficial femoral artery 10

Popliteal artery 25
Anterior tibial artery 90
Posterior tibial artery 80
Peroneal artery 50

noted, rises as they ramify distally down the leg. Shionoya15

reported the prevalence of these angiographic findings in
his series (Table 42.3), the aortoiliac segment being
involved in 8 per cent, the femoropopliteal segment in
32 per cent, and of the crural vessels the peroneal artery is
the least affected (Figs 42.7–42.9).

Figure 42.8 Arteriogram of a 30-year-old man with bilateral

Classical angiographic findings in TAO external iliac occlusions due to Buerger’s disease. Note the extensive
collaterals and disease in branches of the internal iliac artery
• Small and medium sized vessels diseased (tibial,
pedal, ulnar, palmar) NATURAL COURSE
• Segmental occlusion
• Normal proximal arteries
• No atheromatous changes Buerger’s disease usually commences in small vessels such
• Abrupt cut-off with ‘tree root’ collaterals as the digital, pedal and palmar arteries and, expectedly,
• Corkscrew-shaped collaterals patients present initially with instep claudication and digi-
• ‘Accordion’ or ‘corrugated’ appearance of vessels tal ulceration. The disease progresses gradually to involve
affected segments of the crural and popliteal arteries, an advancing
• Early venous filling process occurring both contiguously as well as by skip
lesions. Buerger’s disease may also develop in the muscular
516 Emergency aspects of Buerger’s disease

Figure 42.11 Arteriogram of a 37-year-old smoker presenting

with sudden onset of critical ischaemia of both lower limbs
Figure 42.9 View of segments of a re-formed profunda femoris showing segmental dilatation of the right external iliac artery with
artery taken from the arteriogram in Fig. 42.8 the characteristic ‘accordion’ appearance (arrow)

Figure 42.10 Arteriogram of a 45-year-old man with Buerger’s

disease showing total occlusion of left common and external iliac
arteries with extensive ‘corkscrew’ collaterals. Note the involvement
of a lumbar artery and severe spasm of the right iliac artery

branches of the proximally situated deep femoral, internal

iliac as well as lumbar arteries (Fig. 42.10). Thus, thrombus
may form at the ostia of these vessels and progress thence
to involve the parent artery. Proximal occlusion may also Figure 42.12 View of popliteal artery and thrombotic occlusion
follow stasis thrombus secondary to more extensive distal of the left tibioperoneal trunk (arrows) taken from the arteriogram
disease. in Fig. 42.11
 Treatment 517

Buerger’s disease, involving multiple segments with nor- should be treated according to its merits with drainage,
mal skip areas, may manifest as critical ischaemia of sudden antibiotics and limited amputation. Antiplatelet drugs and
onset (Figs 42.11 and 42.12). Progress of disease with typical pentoxifylline are used in chronic cases. Vasospasm is
acute exacerbations is closely associated with heavy smoking. prominent feature of Buerger’s disease and therefore cal-
Nevertheless, in spite of stopping smoking, 10 per cent of cium channel blockers are very useful.
patients experience recurrence of ischaemic ulceration.23 As In patients who present with acute symptoms, balloon
the disease progresses proximally more collaterals occlude, catheter thromboembolectomy has been tried, but because
the ischemia worsens and the patient ultimately undergoes of the inflammatory nature of the thrombus and the vessel
major amputation. Long term survival, however, has been wall, the artery usually re-occludes. Consequently, in these
shown to be better than that reported for atherosclerosis. patients, various other treatments such as anticoagulation
and lytic therapy have been tried and the former, using
heparin, helps in preventing progression of the thrombotic
process.
TREATMENT
Low molecular weight dextran reduces blood viscosity
and enhances the microcirculation. Intra-arterial throm-
Modalities of treatment recommended bolytic therapy24 has been used in patients presenting with
for TAO gangrenous or pregangrenous lesions with an overall suc-
cess rate of 58.3 per cent (also refer to Chapter 16). Kubota
and coworkers25 have used superselective infusions of
• General
urokinase into the dorsalis pedis artery and demonstrated
– Abstinence from tobacco
– Foot care recanalisation and healing of ischaemic ulcers.
– Exercise Iloprost, a prostaglandin analogue, has been used suc-
cessfully by stimulating blood flow through its vasodilatory
• Medical
and platelet inhibitory action. Feissinger and Schafer26
– Analgesics
– Antiplatelet drugs conducted a prospective randomised double blind study in
– Calcium channel blockers patients with Buerger’s disease presenting with critical
– Pentoxifylline ischaemia, comparing a 6-hour daily infusion of iloprost
– Heparin with aspirin. On analysis at the end of the 28-day treatment
– Low molecular weight dextran period 63 per cent of patients given iloprost were entirely
– Prostaglandin relieved of ischaemic rest pain in comparison with 28 per
– Fibrinolytic agents cent treated with aspirin (P  0.05). Also, ischaemic ulcers
healed completely in 35 per cent of the study group com-
• Surgical
pared with 13 per cent in the control group. At 6 months the
– Sympathectomy – chemical; surgical (open,
minimal access) overall response rate in terms of continued pain relief and/or
– Spinal cord stimulation ulcer healing was 88 per cent in the iloprost group and 21 per
– Omentopexy cent in the aspirin group (P  0.05). The European TAO
– Ilizarov tibial corticotomy study group27 recently completed a double blind ran-
– Vascular reconstruction domised trial comparing oral iloprost with placebo in
– Amputation patients with TAO presenting with critical ischemia. Total
healing of ulcers was not significantly different between
• Miscellaneous
the treatment groups: low dose iloprost was more signifi-
– Gene therapy
– Immunosuppressants cantly effective than placebo but in high doses it failed to
show any significant beneficial effect. The conclusion
was that iloprost was more effective than placebo in the
relief of rest pain but not so in ulcer healing. It appears that
intravenous iloprost is more effective than the oral form in
Various treatment modalities have been recommended ulcer healing.
for Buerger’s disease. The mainstay of any advice is the Sympathetic ganglion blockade has been used effect-
complete abstinence from smoking and patients should be ively to relieve rest pain and healing of ulcers. As sym-
aware that smoking even one or two cigarettes can keep the pathectomy does not increase muscle blood flow it has no
disease active. In a Cleveland Clinic report,3 94 per cent of role to play in claudication. Surgical sympathectomy may
patients who stopped smoking avoided amputation whereas be reasonable, however, in patients who have severe distal
43 per cent of those who continued to smoke required it. disease where vascular reconstruction is either not possible
Regular walking and exercise helps to improve collateral or has failed. It can also be used adjunctively along with
flow, ensures better distribution and utilisation of circulat- proximal arterial revascularisation in those patients who
ing blood and assists muscle metabolism. Foot infection have extensive distal disease.
518 Emergency aspects of Buerger’s disease

Implantable spinal cord stimulators have been used healing, relief of rest pain and limb salvage are served. If the
successfully for pain relief and ulcer healing in patients patient can stay away from tobacco the period of remission
unsuitable for vascular reconstruction.28 Intramuscular is likely to last longer. Successful bypasses have been per-
gene transfer of naked plasmid DNA encoding vascular formed to isolated 10–12 cm long segments of tibial arter-
endothelial growth factor has been successfully tried in ies even when there was no continuity with the pedal arch
critically ischaemic patients.29 It has been beneficial in the (Fig. 42.13).
relief of rest pain, healing of ulcers and gives a 0.1 improve- Sasajima et al.,30 reviewing their results in 71 bypasses
ment in the ankle:brachial index. Angiogenesis has also on 61 patients over a period of 18 years, 85 per cent of
been confirmed by the appearance of new collateral vessels which were to the crural arteries or to arteries below the
on magnetic resonance imaging and contrast angiography. ankle, found that primary and secondary patency rates
Surgical revascularisation for Buerger’s disease is diffi- were 48.8 per cent and 62.5 per cent at 5 years, and
cult because of multisegmental occlusions, distal location 43 per cent and 56.3 per cent at 10 years, respectively. The
of disease and frequent presence of superficial phlebitis. patency rates were 66.8 per cent in non-smokers and 34.5
Sasaki et al.20 found that only 15.5 per cent of their patients per cent in those who continued to smoke after surgery;
were suitable for vascular reconstruction. Endarterectomy 40 per cent of the secondary failure cases underwent
is rarely possible in Buerger’s disease because of the inflam- amputation. Shionoya15 reported cumulative patency rates
matory nature of its pathology, except when the occlusion for femoropopliteal bypass of 70 per cent at 1 year and 60
is caused by stasis thrombus. It follows, therefore, that it per cent at 5 years, and for femorocrural bypass 50 per cent
can be effective in occlusions of the juxtarenal aorta and at 1 year and 44 per cent at 5 years. In our own large series
the profunda femoris artery. Bypass surgery taking the of 965 patients, 310 (32 per cent) proceeded to vascular
graft down to patent segments of vessels and even to larger bypass procedures: aortoiliac or aortofemoral 11 per cent,
collaterals has been successful and vein is the preferred iliofemoral 9 per cent, femoropopliteal 34 per cent,
graft material in these cases (see Chapter 15). Despite the femorodistal 36 per cent and upper limb bypass 10 per
short term patency of these grafts the objectives of ulcer cent.22 Lumbar sympathectomy was added as an adjuvant
procedure to tibial bypass and when the outflow was poor.
Early thrombosis was observed in 28 per cent of patients.
The cumulative 2-year and 5-year patency rates for
femoropopliteal bypass were 50 per cent and 28 per cent,
respectively, and for tibial bypass 30 per cent and 20 per
cent, respectively. The overall major amputation rate
was 22 per cent. Spasm was a major problem during dis-
section and anastomosis (Figs 42.14 and 42.15) and there-
fore gentle tissue handling and the use of vasodilators such
as papaverine and dilzem have helped to minimise this
problem.

Figure 42.13 Operative angiogram after popliteo-tibial bypass

to a 10 cm long isolated posterior tibial segment and which Figure 42.14 Operative photograph of a 35-year-old man
remained patent for 13 months. Typical ‘tree root’ and ‘corkscrew’ with Buerger’s disease presenting with proximal tibial artery
appearance of collaterals can be made out. Note the spasm in the occlusion showing a popliteal to posterior tibial reversed saphenous
posterior tibial artery just distal to the anastomosis vein graft
 Treatment 519

There have been reports of indirect revascularisation

of ischaemic limbs in Buerger’s disease by means of omen-
topexy, either as free or pedicled grafts, and also of arterial-
isation of the long saphenous vein with associated claims of
ulcer healing and relief from rest pain.31 Our treatment
policy in patients with chronic symptoms is illustrated in
Fig. 42.16.
In those patients presenting with acute symptoms,
immediate surgery is likely to fail in view of the active
inflammatory and immunological response (see Chapter 2).
It is better, therefore, to treat these patients medically. The
treatment options for acute exacerbation of Buerger’s dis-
ease are illustrated in Fig. 42.17.
Saha et al.32 recently reported a small non-randomised
trial of cyclophosphamide in the treatment of patients
with advanced TAO with modest results: improvement in
claudication, relief of rest pain and a reduction in the
influx of inflammatory cells in thrombus and vessel wall
were noted.
Figure 42.15 Operative angiogram in Fig. 42.14 showing spasm
of the posterior tibial artery. This graft remained patent for 2 years

Claudication/non-healing ulcer

Confirm diagnosis
Stop smoking

Exercise
Treatment
Foot care

Antiplatelet drugs

No relief Relief
Continue medical
treatment

(Relief)

Arteriogram Prostaglandin

(No relief)

Reconstructible Non-reconstructible

Bypass Sympathectomy

Relief Failed No relief Relief

Continue medical Continue medical
treatment treatment
Figure 42.16 Algorithm of treatment
of chronic symptoms of Buerger’s
Amputation
disease
520 Emergency aspects of Buerger’s disease

Analgesia

Narcotics Epidural

Prostaglandin infusion Lytic therapy 

Heparin  Lomodex infusion
Heparin Heparin

Pain relief  ulcer healing

Demarcation of gangrene

Local amputation No relief

Continue medical treatment
Stop smoking
Exercise Vascular repair Amputation
Antiplatelet drugs

Figure 42.17 Algorithm of treatment of

Failed
acute exacerbations of Buerger’s disease

Conclusions Shionoya S. Clinical manifestations. In: Buerger’s Disease –

Pathology diagnosis and treatment. Nagoya: The University of
Nagoya Press, 1990: 113.
Buerger’s disease or TAO is a non-atheromatous, seg-
Shionoya S. Thromboangiitis obliterans (Buerger’s disease).
mental inflammatory arterial disease affecting medium In: Rutherford RB (ed). Thromboangiitis Obliterans (Buerger’s
and small muscular arteries of both upper and lower Disease) in Vascular Surgery 4th edn. Philadelphia, PA:
limbs. It is common in young males who are heavy smok- WB Saunders, 1994: 235–45.
ers and there is a strong correlation between smoking The European Study Group. Oral iloprost in the treatment of
and progression of disease. Thromboangiitis obliterans thromboangiitis obliterans (Buerger’s disease): a double blind,
usually presents with digital ulcers, gangrene and severe randomized, placebo controlled trial. Eur J Vasc Endovasc Surg
rest pain and this clinical picture is characterised by 1998; 16: 300–7.
remissions and relapses. The angiographic findings of
segmental occlusion and ‘corkscrew’ collaterals are
pathognomonic of TAO. Complete abstinence from REFERENCES
smoking is the mainstay of treatment. In cases of
chronic disease exercise and antiplatelet drugs are worth
1 Cachovan M. Epidemiologie und geographisches
trying. In the acute case anticoagulant and thrombolytic
verteilungsmuster der Thromboangitis Obliterans. In: Heidrich H
therapy may arrest deterioration. Vascular reconstruc-
(ed). Thromboangiitis Obliterans Morbus Winiwarter – Buerger.
tion is difficult but can be successfully undertaken with Stuttgart: George Thieme, 1988: 31–6.
bypasses to patent segments of the artery. Long term 2 Lie JT. Thromboangiitis obliterans (Buerger’s disease) in women.
patency of these grafts should not be expected but they Medicine 1987; 64: 65–72.
do relieve rest pain and prevent amputation or minimise 3 Olin JW. Thromboangiitis obliterans (Buerger’s disease). In:
the extent of tissue loss, particularly if the patient gives Rutherford RB (ed). Thromboangiitis Obliterans (Buerger’s
up smoking permanently. Disease) in Vascular Surgery 5th edn. Philadelphia, PA: WB
Saunders 2000: 350–64.
4 Papa M, Bass A, Adar R, et al. Autoimmune mechanisms in
thromboangiitis obliterans (Buerger’s Disease). The role of
Key references antigen and the major histocompatibility complex. Surgery
1992; 111: 527–31.
Olin JW. Name of chapter. In: Rutherford RB (ed).Thromboangiitis 5 Jindal RM, Patel SM. Buerger’s disease in cigarette smoking in
Obliterans (Buerger’s Disease) in Vascular Surgery 5th edn. Bangladesh. Ann R Coll Surg Engl 1992; 74: 436–7.
Philadelphia, PA: WB Saunders 2000: 350–64. 6 Adar R, Papa M, Halperin Z, et al. Cellular sensitivity to collagen
Sasajima T, Kubo Y, Inaba M, et al. Role of Infrainguinal bypass in in thrombo angitis obliterans. N Engl J Med 1983; 308: 1113–16.
Buerger’s disease: an eighteen year experience. Eur J Vasc 7 Gulati SM, Saha K, Kant L, et al. Significance of circulating
Endovasc Surg 1997; 13: 186–92. immune complexes in thromboangiitis obliterans (Buerger’s
disease). Angiology 1984; 35: 276–81.
 References 521

8 Roncon A, Delgado L, Correia P, et al. Circulating immune 21 Vink M. Symposium on Buerger’s disease. J Cardiovasc Surg
complexes in Buerger’s disease. J.Cardiovasc Surg 1989; 30: 1973; 14: 1–51.
821–5. 22 Hussain SA, Sekar N. Buerger’s disease – review article. Indian
9 Eichhorn J, Sima D, Lindschau C, et al. Antiendothelial cell J Surg 2002; 64: 237–9.
antibodies in thromboangiitis obliterans. Am J Med Sci 1998; 23 Shionoya S. Thromboangiitis obliterans (Buerger’s disease). In:
315: 17–23. Rutherford RB (ed). Thromboangiitis Obliterans (Buerger’s
10 Chaudhry NA, Pietraszek NH, Hachiya T, et al. Plasminogen disease) in Vascular Surgery 4th edn. Philadelphia, PA:
activators and plasminogen activator inhibitor 1 before and WB Saunders, 1994: 235–45.
after venous occlusion of the upper limb in thromboangiitis 24 Hussein EA, el Dorri A. Intra-arterial streptokinase as adjuvant
obliterans (Buerger’s disease). Thromb Res 1992; 66: 321–9. therapy for complicated Buerger’s Disease early trials. Int Surg
11 Makita S, Nakamura M, Murakami H, et al. Impaired 1993; 78: 54–8.
endothelium dependent vasorelaxation in peripheral 25 Kubota Y, Kichikawa K, Uchida H, et al. Superselective
vasculature of patients with thromboangiitis obliterans urokinase infusion therapy for dorsalis pedis artery occlusion
(Buerger’s disease). Circulation 1996; 94(suppl II): II-211–15. in Buerger’s disease. Cardiovasc Intervent Radiol 1997; 20:
12 Donatelli F, Triggiani M, Nascinbene S, et al. Thromboangiitis 380–2.
obliterans of coronary and internal thoracic arteries in a young 26 Fiessinger JN, Schafer M. Trial of iloprost versus aspirin
woman. J.Thorac Cardiovasc Surg 1997; 113: 800–2. treatment for critical ischaemia of thromboangiitis obliterans.
13 Hassoun Z, Lacrosse M, De Ronde T. Intestinal involvement in The TAO study. Lancet 1990; 335: 555–7.
Buerger’s disease. J Clin Gastroenterol 2001; 32: 85–9. 27 The European Study Group. Oral iloprost in the treatment of
14 Lie JT. Thromboangiitis obliterans (Buerger’s Disease) in a thromboangiitis obliterans (Buerger’s disease): a double blind,
saphenous vein arterial graft. Hum Pathol 1987; 18: 402–4. randomized, placebo controlled trial. Eur J Vasc Endovasc Surg
15 Shionoya S. Clinical manifestations. In: Buerger’s Disease – 1998; 16: 300–7.
Pathology diagnosis and treatment. Nagoya: The University of 28 Chierichetti F, Mambrini S, Bagliani A, Odero A. Treatment of
Nagoya Press, 1990: 113. Buerger’s disease with electrical spinal cord stimulation –
16 Wysokinski W, Kwiatkowska W, Raczkowska BS, et al. Sustained review of three cases. Angiology 2002; 53: 341–7.
classic clinical spectrum of thromboangiitis obliterans 29 Isner JM, Baumbartner I, Rauh G, et al. Treatment of
(Buerger’s disease). Angiology 2000; 51: 141–50. thromboangiitis obliterans (Buerger’s disease) by intramuscular
17 Matsushita M, Shionoya S, Matsumoto T. Urinary nicotine gene transfer of vascular endothelial growth factor: preliminary
measurement in patients with Buerger’s Disease: effect of clinical results. J Vasc Surg 1998; 28: 964–75.
active and passive smoking on the disease process. J Vasc Surg 30 Sasajima T, Kubo Y, Inaba M, et al. Role of Infrainguinal bypass
1992; 14: 53–8. in Buerger’s disease: an eighteen year experience. Eur J Vasc
18 Papa MC, Rai I, Adar R. A point scoring system for the clinical Endovasc Surg 1997; 13: 186–92.
diagnosis of Buerger’s disease. Eur J Vasc Endovasc Surg 1996; 31 Talwar S, Jain S, Porwal R, et al. Free versus pedicled omental
11: 335–9. grafts for limb salvage in Buerger’s Disease. Aust N Z J Surg
19 Mills JL, Porter JM. Buerger’s disease: a review and update. 1998; 68: 38–40.
Semin Vasc Surg 1993; 6: 14–23. 32 Saha K, Chabra N, Gulati SM. Treatment of patients with
20 Sasaki S, Sakumo M, Yasuda K. Current status of thrombo- thromboangiitis obliterans with cyclophosphamide. Angiology
angitis obliterans (Buerger’s disease) in Japan. Int J Cardiol 2001; 52: 399–407.
2000; 75: 5175–81.
This page intentionally left blank
 43
Acute Limb Vascular Inflammatory Conditions

MATTHEW WALTHAM, KEVIN G BURNAND

Acute arterial inflammatory conditions 523 References 528

Acute venous inflammatory conditions 528

ACUTE ARTERIAL INFLAMMATORY

CONDITIONS

Vascular inflammatory conditions of the limbs may present

in the context of a wide variety of clinical syndromes,
including Buerger’s disease, scleroderma, systemic lupus
erythematosus (SLE), Takayasu’s disease and giant cell
arteritis. These diseases are characterised by inflammatory
cell invasion of arteries causing swelling, luminal obstruc-
tion and often thrombosis. Operative and endovascular
techniques are rarely indicated in their treatment and
should only be used in carefully selected patients. Surgical
biopsy of a vessel or muscle may be useful in the diagnosis
of vasculitic syndromes. Figure 43.1 Cross-section of affected small vessel in Buerger’s
disease, showing transmural inflammation and luminal thrombosis

Aetiology and pathology

Thromboangiitis obliterans (Buerger’s disease) is charac- (Fig. 43.2), a process often progressing to arterial occlusion
terised by progressive obliteration of distal vessels of the and limb ischaemia. Collateral pathways are less likely to
upper and lower limbs (see Chapter 42). The exact patho- develop with more distal occlusions, which, therefore, are
physiology is poorly understood, but patients are almost more likely to cause critical ischaemia.
invariably young cigarette-smoking males, although the Takayasu’s arteritis is an idiopathic systemic inflamma-
condition has also been described in cannabis smokers1 tory disease usually involving the aorta and its main
and women. Histological examination of affected vessels branches (see Chapter 41). It is most frequently observed
shows transmural inflammation with intimal proliferation, in young women from Far Eastern Asian countries. With
surrounded by collagen deposition (Fig. 43.1). Luminal early disease there is focal or continuous granulomatous
thrombosis occurs within the vessels, and the accompanying inflammation which leads to fibrosis of the vessel wall. This
veins may also become inflamed leading to superficial and results in multiple stenoses or occasionally aneurysms.
deep vein thrombosis. Progressive ischaemia leads first to Giant cell arteritis is the commonest form of systemic
digital gangrene and then more major amputation of the arteritis. There is focal inflammation of medium and small
limbs becomes necessary if cigarette smoking continues.2,3 arteries, usually affecting the extracranial branches of the
Systemic autoimmune diseases often cause transmural external carotid artery. The aetiology of this disease is
inflammation of small vessels, affecting individual crural unknown. Surgery plays no part in treatment but arterial
vessels, pedal arch vessels, metatarsal and digital arteries biopsy is often useful in establishing the diagnosis.
524 Acute limb vascular inflammatory conditions

Figure 43.3 Typical colour changes in the hands of a patient

with Raynaud’s disease

Figure 43.2 Angiogram showing obliteration of digital vessels

in a patient with Raynaud’s disease secondary to scleroderma

Diagnosis

CLINICAL
Critical limb ischaemia may present as pain, ulceration or
gangrene in the affected limb, which may be preceded by a
history of intermittent claudication. Buerger’s disease should
always be suspected if the patient is a young male smoker. Figure 43.4 Severe Raynaud’s disease has led to digital
Many of the vasculitides present with severe Raynaud’s amputation in this patient
phenomenon, i.e. digital pallor in response to cold
exposure followed by cyanosis and then reactive hyper-
usually extends over several years and varies in severity and
aemia (Fig 43.3). This is often intractable and may progress
speed of progression.
to digital ulceration and even frank gangrene (Fig. 43.4).
Affected limbs feel cool, distal pulses are impalpable and
INVESTIGATIONS
ischaemic paronychiae are common around the nails.
Takayasu’s disease often begins with non-specific signs Duplex ultrasound and angiography are indicated to
and symptoms indicative of a systemic inflammatory define the anatomical site of disease and to evaluate the
response. Specific features of arterial disease may then possibility of vessel reconstruction. In diseases affecting
emerge, either from limb or end organ ischaemia. The clin- distal arteries Doppler pressures in pedal vessels are
ical features affect the peripheral vascular, neurological, reduced and duplex scanning shows normal blood flow to
cardiac and pulmonary systems. The course of the disease the popliteal arteries. Duplex scanning is less valuable in
 Acute arterial inflammatory conditions 525

Figure 43.6 Angiogram showing carotid and subclavian artery

disease in a patient with Takayasu’s disease
Figure 43.5 Angiogram showing typical ‘corkscrew’ collaterals
in a patient with Buerger’s disease Active Takayasu’s disease is treated with high dose corti-
costeroids alone or together with a cytotoxic agent. Stenoses
causing critical limb, or end organ, ischaemia should be
these cases as it rarely defines disease in the distal vasculature. treated by interventional radiology or surgical bypass.
In Buerger’s disease arteriography shows a characteristic pat-
tern of normal proximal vessels and distal occlusions often MEDICAL OR CONSERVATIVE TREATMENT
with many ‘corkscrew’ collaterals (Fig. 43.5). Occluded ves-
In patients with Buerger’s disease the long term course of
sels may be biopsied to confirm the diagnosis.
the disease and the frequency and extent of amputations
In Takayasu’s disease the aorta and its main branches
depends almost exclusively on whether the patient con-
are most commonly affected, and these can be easily imaged
tinues to smoke. The emphasis of treatment must therefore
by intravenous digital subtraction angiography (Fig. 43.6).
be to stop the patient smoking.
The diagnosis is based on the clinical features and the pres-
An attempt should be made to find a cause for Raynaud’s
ence of typical configurations of affected arteries. The ery-
phenomenon, such as an underlying connective tissue dis-
throcyte sedimentation rate (ESR) and C-reactive protein
order or a cervical rib. Attacks are often precipitated by a
(CRP) level may be raised. Disease activity is sometimes
number of triggers, for example cold exposure, and avoid-
difficult to assess as the clinical features or acute phase reac-
ance of these triggers, e.g. the wearing of heated gloves or
tants do not accurately reflect blood vessel inflammation.
stopping smoking, can control the symptoms adequately.
Vasodilator drugs may also be useful particularly in pri-
mary Raynaud’s syndrome; nifedipine reduces the fre-
Management quency and severity of vasospastic attacks. Alternative
medication includes naftidrofuryl and inositol nicotinate.
GENERAL PRINCIPLES AND PITFALLS
Patients with severe vasospasm may respond to an
The surgical options for treating occlusive inflammatory infusion of prostaglandin, e.g. iloprost. The dose should
arterial disease are sympathectomy, bypass procedures and be titrated to the maximum tolerated without causing
amputation. Unfortunately, the distal nature of most vas- significant hypotension, and continued for several days.
culitides means that surgical options are limited and often Vasculitides of large muscular arteries, for example giant
have a poor outcome. In most cases surgery is only con- cell arteritis and Takayasu’s disease, should be treated pri-
sidered for intractable pain or in cases where there is ulcer- marily with corticosteroids. This is almost always success-
ation or frank gangrene. The majority of cases of arteritis ful for giant cell arteritis but 50 per cent of patients with
affecting the limbs should be treated with non-surgical Takayasu’s arteritis become refractory to steroids and sur-
methods. gery becomes necessary (see Chapter 41).
526 Acute limb vascular inflammatory conditions

SURGICAL TREATMENT Cervical sympathectomy

The sympathetic fibres to the arm synapse in the second to
Sympathectomy
fifth thoracic ganglia, with the first thoracic ganglion fusing
The sympathetic nervous system innervates the arterioles
with the inferior cervical ganglion to form the stellate gan-
and precapillary sphincters in the skin and muscle of the
glion. Inadvertent injury to this ganglion will result in a per-
limbs. The cutaneous nerves are mostly vasoconstrictor
manent Horner’s syndrome. Cervical sympathectomy may
whereas muscles have both vasoconstrictor and vasodilator
be performed by anterior cervical,9 transaxillary10 or thora-
fibres. Sympathectomy in a normal limb results in a large
coscopic11 approaches depending on surgical preference.
increase in blood flow, but its effect in an ischaemic limb is
greatly attenuated because maximal vasodilation is already Thoracoscopic cervical sympathectomy
present. Additionally, in diabetic patients peripheral neur- The thoracoscopic approach is now well established and
opathy has often already effected an ‘autosympathectomy’. has become the standard procedure although complica-
Sympathetic vasomotor blockade, however, does result tions are not uncommon. Damage to the first cervical gan-
in a redistribution of blood flow with a marked increase in glion is most easily avoided using this approach as it is not
cutaneous flow at the expense of muscular flow. Although usually possible to get any higher than the second rib from
this may worsen claudication distance the increased skin the pleural space.
blood flow will sometimes allow ischaemic nail folds and The procedure is performed under general anaesthesia,
ulceration to heal. Blockage of sudomotor activity may also preferably using a double lumen tube, with the patient
dry up ischaemic and infected ulcers.4,5 There is also evi- supine and with the arms abducted. The ipsilateral lung is
dence that dividing the sympathetic chain not only affects deflated and an artificial pneumothorax established. A
local neurotransmitter production but also reduces affer- 5 mm laparoscope is inserted through the third intercostal
ent pain fibre transmission in the spinal cord thus reducing space and advanced across the pleural cavity to identify the
or abolishing pain. sympathetic ganglia and chain passing over the necks of the
Sympathectomy cannot reverse or improve frank gan- ribs. The appropriate ganglia and interconnecting rami are
grene and if present the procedure must be combined with coagulated using a diathermy probe inserted through a sep-
ablation. Although the effects are often short lived, healing arate incision. This blunderbuss technique fails to remove
of ischaemic areas combined with cessation of smoking the sympathetic chain for histological examination and can,
may provide long term benefit. Sympathectomy is not if poorly performed, fail to interrupt the chain. For this rea-
indicated in claudication, but may help relieve symptoms son a two port technique with careful excision of the chain
in patients with rest pain who are not suitable for recon- is preferable. Care must be taken not to damage the first
structive procedures. Percutaneous sympathetic block may thoracic ganglion. The lung is reinflated at the end of the
be helpful in identifying patients with residual vasocon- procedure, the laparoscope is removed and the wounds
striction. Operative sympathectomy may be beneficial in closed. A postoperative chest X-ray is taken to identify any
those whose symptoms are relieved with perhaps some residual pneumothorax.
objective increase in skin temperature.
Transaxillary cervical sympathectomy
Sympathectomy is achieved either by surgical excision
This operation has largely been replaced by the thoraco-
of part of the sympathetic chain or by chemical ablation of
scopic procedure. The patient is placed in a lateral position
the chain. Chemical sympathectomy is performed by injec-
with the operated side uppermost and the axilla widely dis-
tion of phenol around the chain using local anaesthetic and
played by abducting the arm and flexing the forearm. An
radiographic control. Surgical excision is preferable in
8 cm oblique incision is made from latissimus dorsi, run-
patients with vasculitides because recurrence may follow
ning forwards and down across the third rib as far as the
chemical sympathectomy. Salvage surgical sympathectomy
posterior border of pectoralis major. The periosteum of
following a failed chemical sympathectomy is a very diffi-
the rib is exposed, divided with diathermy, and reflected
cult procedure.
from the superior surface to expose the costal pleura. This
The most successful application of sympathectomy is in
is divided along the upper border of the rib and a rib
the treatment of thromboangiitis obliterans or Buerger’s
retractor is inserted and opened widely. The apex of the
disease, perhaps because the condition is characterised by
lung is displaced downwards and the ganglia and intercon-
a significant component of vasospasm (see Chapter 42).
necting chain identified running beneath the costal pleura
Sympathectomy often relieves rest pain and may allow
over the necks of the ribs. The overlying pleura is opened
small areas of gangrene to heal. Sympathectomy is rarely if
and the chain and rami lifted and divided above the T2 and
ever indicated for Raynaud’s phenomenon and most
below the T5 ganglia. The wound is closed and the lung
patients can achieve satisfactory symptom control by sim-
re-expanded. A postoperative chest X-ray is taken.
ply avoiding cold, abstaining from tobacco and wearing
warm or heated gloves in the winter. Sympathectomy often Anterior cervical sympathectomy
produces dramatic early benefit6 but this is short lived and The traditional access to the sympathetic chain is through
after a year or two the benefits are marginal.7,8 a cervical approach. The patient is placed feet down supine
 Acute arterial inflammatory conditions 527

with a sandbag under the shoulders and the head turned to approach. This may be performed using either a transperi-
the opposite side. A 5 cm incision is placed 1 cm above the toneal13 or a retroperitoneal, balloon-assisted,14 approach.
clavicle with the medial end just overlying the sternomas- The benefits of this method have yet to be demonstrated
toid. The platysma, lateral fibres of sternomastoid and any but may combine the advantages of a minimally invasive
intervening veins are divided to expose the scalenus anter- approach with the certainty of surgical excision of the sym-
ior. This muscle is divided low down preserving the phrenic pathetic chain.
nerve on its surface and by reflecting it medially. That exposes
the subclavian artery which is retracted upwards or down-
wards by dividing its branches to expose the costopleural BYPASS PROCEDURES
membrane which in turn is incised to expose the apex Distal revascularisation procedures, either endovascular or
of the lung. The sympathetic chain is exposed by carefully by surgical bypass, have a high rate of failure and are rarely
stripping the lung downwards with a finger. The chain is indicated. This is largely because the secondary periarteri-
excised between the stellate and fourth cervical ganglia. tis is often very severe and makes dissection of the vessels
Lumbar sympathectomy very difficult. Unfortunately, autoimmune vasculitis also
This may be performed either by an open operation or by tends to affect distal vessels, and distal bypass procedures
injecting phenol around the lumbar chain using radio- have a much lower flow rate and patency rate than prox-
logical guidance. The latter has very low morbidity and so imal procedures. Results are therefore poor, and these pro-
has gained popularity, and may be particularly appropriate cedures should only be considered in exceptional cases.
for treating some elderly patients with ischaemic rest pain. In the presence of ulcers bypass grafting may be useful.
For younger patients surgical sympathectomy should be Even if the graft only remains patent for a short time, this
performed to ensure complete removal of the sympathetic may allow ulcers to heal. If the graft then fails, the ulcers
chain. At operation the second and third lumbar ganglia often do not recur provided that the patient stops smoking.
are removed. In males the first lumbar ganglion on at
least one side must be retained in order to preserve normal Bypass procedures in patients with Takayasu’s disease
ejaculation. Takayasu’s disease and the middle aortic syndrome are dis-
Patients with distal vessel occlusive disease, and therefore eases that are very amenable to surgical bypass (also see
poor candidates for reconstructive procedures, may be suit- Chapter 41). Takayasu’s disease most commonly causes
able for lumbar sympathectomy to relieve rest pain and stenosis of the subclavian and innominate arteries and aor-
sometimes rescue critically ischaemic tissue. The major bene- tic arch followed by the descending aorta and aortoiliac
fit is in the relief of rest pain, but it occurs in only 60 per cent region. The presenting symptoms depend upon the vessels
of cases and lasts for up to 3 years. The amputation rate is affected but transient ischaemic attacks, visual problems,
not affected.12 Surgical lumbar sympathectomy can be per- arm and leg claudication and the subclavian steal syndrome
formed either as an open procedure or laparoscopically. can all occur.15 The disease may also sometimes cause
aneurysms and aortic valve regurgitation.
Open lumbar sympathectomy
Surgical intervention should be timed to avoid acute
The lumbar sympathetic chain is approached through a
inflammatory episodes, guided by the ESR or CRP. When
transverse incision lateral to but at the level of the umbilicus
there is critical ischaemia, and immediate surgical treat-
on the side to be denervated. The oblique muscles of the
ment cannot be avoided, it should be accompanied by high
abdominal wall are divided to expose the peritoneum. This
dose steroid treatment. Arteriography must be performed
is freed by blunt dissection from the deep surface of the
to assess the disease in order to plan surgery, and it often
transversus abdominis muscle and retracted medially to
demonstrates either a smooth tapering stenosis of the
expose the retroperitoneal space. The psoas is found on the
affected vessel or total occlusion. Surgical bypass to normal
posterior wall and the groove between the medial border of
vessels beyond the limits of the disease is indicated for
this muscle, the lumbar vertebrae and the aorta on the left
patients with ischaemic symptoms, aneurysms or signifi-
and the inferior vena cava on the right is defined. The lum-
cant renal artery disease.16–18
bar sympathetic chain can usually be palpated as a firm cord
Standard operative techniques achieve good results.19–21
punctuated by a number of swellings lying in this groove on
Surgery may take the form of carotid artery reconstruction,
the front of the vertebrae. The chain is picked up with a
carotid–subclavian bypass, thoracoabdominal aortic bypass,
nerve hook and dissected up to the diaphragmatic crura and
renal artery reconstruction or aneurysm repair.22 If possi-
down to the pelvic brim. All of the rami that join the ganglia
ble the surgery should be performed in one stage to correct
are divided and the first, second and third ganglia are excised.
all affected vessels, with multiple reconstructions if neces-
The wound is closed in layers with suction drainage.
sary. Saccular aneurysmal disease has a high rate of rupture
Endoscopic lumbar sympathectomy and should be treated urgently, whereas fusiform aneurysms
The technique of laparoscopic lumbar sympathectomy has can be treated similarly to atheromatous disease. Treatment
been described as an alternative to the traditional open is by resection and replacement with a synthetic graft.
528 Acute limb vascular inflammatory conditions

AMPUTATION should be made when Trousseau’s phenomenon occurs.

The patient should be screened for autoantibodies when
When the above procedures are not indicated or fail then
there are other features of a connective tissue disease.
there is often no alternative but to amputate critically
ischaemic or gangrenous tissue. Amputation should be as
conservative as possible while ensuring that satisfactory Management
healing is achieved. Dead or devitalised tissue or bone must
be removed completely. In general, thrombophlebitis can be treated successfully
with anti-inflammatory drugs.
SURGERY FOR GIANT CELL ARTERITIS
The diagnosis of giant cell arteritis can be confirmed surgi- MEDICAL OR CONSERVATIVE TREATMENT
cally by taking a segment of the superficial temporal artery Superficial thrombophlebitis should be treated with aspirin
under local anaesthetic. The incision is placed directly over or non-steroidal anti-inflammatory drugs (NSAIDs). Venous
the vessel. Histological examination shows pronounced compression stockings may prevent recurrent attacks.
intimal thickening, round cell infiltration through all
layers of the arterial wall with destruction of the internal SURGICAL TREATMENT
elastic lamina and the presence of a few giant cells.
Surgical stripping of the long saphenous system can be
of value in preventing recurrent attacks of above-knee
ACUTE VENOUS INFLAMMATORY thrombophlebitis.
CONDITIONS
Conclusions
Acute superficial thrombophlebitis is a common problem.
It is generally considered to be a benign condition but it is Surgery has a limited role in treating patients with arterial
sometimes associated with deep vein thrombosis (DVT) or inflammatory conditions or vasculitides. Sympathectomy
other underlying pathology. is the most commonly performed operation. Bypass pro-
cedures are most effective in patients with Takayasu’s dis-
Aetiology and pathology ease and the mid-aortic syndrome. Amputation may be
necessary in patients with gangrene.
Thrombophlebitis usually affects an isolated segment of Acute superficial thrombophlebitis may complicate
superficial vein in the leg. The vein becomes acutely varicose veins but deep vein thrombosis or malignancy
inflamed, often to a greater extent than is clinically apparent. may have to be excluded before treatment.
The commonest risk factors are the presence of varicose
veins, obesity and age. It may be associated with DVT, pul-
monary embolism (PE) or autoimmune disease such as SLE Key references
or Behçet’s disease. Migrating superficial thrombophlebitis
or Trousseau’s phenomenon is associated with visceral can- Crawford ES, De Bakey ME, Morris GC, Cooley DA. Thrombo-
cer, in particular carcinoma of the pancreas. obliterative disease of the great vessels arising from the aortic
arch. J Thorac Cardiovasc Surg 1962; 43: 38–53.
Crawford ES, Snyder DM, Cho GC, Roehm Jr JO. Progress in
Diagnosis treatment of thoracoabdominal and abdominal aortic
aneurysms involving celiac superior mesenteric and renal
CLINICAL arteries. Ann Surg 1978; 188: 404.
The diagnosis of superficial thrombophlebitis is primarily Lande A. Abdominal Takayasu’s aortitis, the middle aortic syndrome
clinical. The leg is most commonly affected, presenting and atherosclerosis: a critical review. Int Angiol 1998; 17: 1–9.
Lupi-Herrera E, Sanchez-Torres G, Marcushamer J, et al. Takayasu’s
with painful inflammation over one or more veins. On
arteritis. Clinical study of 107 cases. Am Heart J 1977; 93:
examination there may be tender palpable subcutaneous 94–103.
cords along the course of the saphenous veins or their trib- Ohta T, Shionoya S. Fate of the ischaemic limb in Buerger’s disease.
utaries with associated erythema and oedema. Br J Surg 1988; 75: 259–62.

INVESTIGATIONS
Duplex ultrasound examination of the deep venous system REFERENCES
should be performed to identify deep vein thrombosis, par-
ticularly if the proximal long saphenous vein is involved. A 1 Schneider HJ, Jha S, Burnand KG. Progressive arteritis associated
search for an underlying malignancy or thrombophilia with cannabis use. Eur J Vasc Endovasc Surg 1999; 18: 366–7.
 References 529

2 McPherson JR, Juergens JL, Gifford RW. Thromboangiitis 14 Elliott TB, Royle JP. Laparoscopic extraperitoneal lumbar
obliterans and arteriosclerosis obliterans: clinical and prognostic sympathectomy: technique and early results. Aus N Z J Surg
differences. Ann Intern Med 1963; 59: 288–96. 1996; 66: 400–2.
3 Kinmonth JB. Thromboangiitis obliterans. Results of 15 Lupi-Herrera E, Sanchez-Torres G, Marcushamer J, et al.
sympathectomy and prognosis. Lancet 1948; 2: 717. Takayasu’s arteritis. Clinical study of 107 cases. Am Heart J 1977;
4 Ohta T, Shionoya S. Fate of the ischaemic limb in Buerger’s 93: 94–103.
disease. Br J Surg 1988; 75: 259–62. 16 Crawford ES, De Bakey ME, Morris GC, Cooley DA.
5 Kunlin J, Lengua F, Testart J, Pajot A. Thromboangiosis or Thrombo-obliterative disease of the great vessels arising
thromboangeitis treated by adrenalectomy and sympathectomy from the aortic arch. J Thorac Cardiovasc Surg 1962;
from 1942 to 1962. A follow-up study of 110 cases. J Cardiovasc 43: 38–53.
Surg 1973; 14: 21–7. 17 Crawford ES, Snyder DM, Cho GC, Roehm Jr JO. Progress in
6 Baddeley RM. The place of upper dorsal sympathectomy in the treatment of thoracoabdominal and abdominal aortic aneurysms
treatment of primary Raynaud’s disease. Br J Surg 1965; 52: 426. involving celiac superior mesenteric and renal arteries. Ann Surg
7 Gifford RW, Hines, EA, Craig WM. Sympathectomy for Raynaud’s 1978; 188: 404.
phenomenon Circulation 1958; 17: 5. 18 Thompson BW, Read RC, Campbell GS. Aortic arch syndrome.
8 Johnston EN, Summerly R, Birnstingl M. Prognosis in Raynaud’s Arch Surg 1969; 98: 607–11.
phenomenon after sympathectomy. BMJ 1965; 42: 962–4. 19 Robbs JV, Human RR, Rajaruthnam P. Operative treatment of
9 Telford ED. The technique of sympathectomy. Br J Surg 1935; nonspecific aortoarteritis (Takayasu’s arteritis), J Vasc Surg 1986;
23: 448. 3: 605–16.
10 Atkins HJB. Peraxillary approach to the stellate and upper 20 Fraga A, Mintz G, Valle L, et al. Takayasu’s arteritis: frequency of
thoracic ganglia. Lancet 1949; 2: 1152. systemic manifestations (study of 22 patients) and favorable
11 Hederman WP. Sympathectomy by thoracoscopy. In: response to maintenance steroid therapy with corticosteroids
Greenhalgh RM (ed). Vascular and Endovascular Surgical (12 patients). Arthritis Rheum 1972; 15: 617–24.
Techniques. London: WB Saunders 1994: 281. 21 Weaver FA, Yellin AE, Campen DH, et al. Surgical procedures in
12 Cotton LT, Cross FW. Lumbar sympathectomy for arterial disease. the management of Takayasu’s arteritis. J Vasc Surg 1990;
Br J Surg 1985; 72: 678–83. 12: 429–37.
13 Wattanasirichaigoon S, Ngaorungsri U, Wanishayathanakorn A, 22 Lande A. Abdominal Takayasu’s aortitis, the middle aortic
et al. Laparoscopic transperitoneal lumbar sympathectomy: a syndrome and atherosclerosis: a critical review. Int Angiol 1998;
new approach. J Med Assoc Thai 1997; 80: 275. 17: 1–9.
This page intentionally left blank
 44
Vascular Emergencies Caused by Substance Abuse

RONALD A KLINE, RAMON BERGUER

History 531 Amphetamines 533

The problem 531 Other agents 534
Marijuana 532 Heroin 534
Cocaine 532 Vascular system considerations 536
Angel dust, PCP 533 References 542

HISTORY white powder, is rarely obtainable in its street version with-

out being adulterated. This is notwithstanding bacterial or
viral contamination of the agent itself or the instruments of
Recreational drug use has existed for centuries and is pre- delivery. All three vascular systems, lymphatic, venous and
sent in almost all cultures, be it the opium of ancient China, arterial can be affected. The resulting injurious effects may
the coca leaves of the rainforest, or the American Indian be local or systemic.
use of peyote. Ingestion or inhalation were the usual routes The spectacular increase in drug consumption over the
of administration. Although forms of intravenous injec- past 20 years has resulted in a large number of drug related
tion began as early as 1670 it was Charles Gabriel Pravaz injuries to the vascular system. About half of the acutely
and Alexander Wood who are credited with the independ- injured patients presenting at the level 1 trauma unit of
ent yet simultaneous development in 1853 of a syringe with the Detroit Receiving Hospital, MI, USA, test positive for
a hollow needle fine enough to pierce vessels.1 This revolu- cocaine, heroin or both. In 1998 the National Institute of
tionised the administration of narcotics. The application of Drug Abuse estimated that over 21 million Americans have
this new device was in treating pain through the use of opi- tried cocaine, including three million who had done so
ate injections. Its recreational usage was not intentional. within the previous month.3 Medical examiners in
In the UK, all drugs were legal and reportedly used rou- Memphis, TN, USA, noted a 53 per cent increase in drug
tinely across society, until 1860.2 The former Prime Minister related homicides with cocaine related deaths being most
William Gladstone and Florence Nightingale used opium, responsible.4 This reflects the activities in many large cities.
while Queen Victoria used cannabis. Sir Arthur Conan Doyle Similar reports come from Salt Lake City, UT, and Tucson,
wrote a graphic description of Sherlock Holmes injecting AZ, USA, where cocaine associated violent deaths and poi-
drugs with a syringe as a normal way of relaxing during the sonings are rising.5 Substance abuse also extends to the
hiatus between cases, much to the chagrin of Dr Watson. adolescent and childhood trauma scene.6–9
The mode of consumption has not changed to date. Illicit The Drug Abuse Warning Network (DAWN) of the
substance usage is ingested, inhaled in smoking or snorting US Section of Health and Human Services appears to be
or injected. underreporting this rising usage to Congress.10 Whereas
DAWN recorded a 5 per cent incidence of cocaine use in
patients treated at the Hospital of the University of Penn-
THE PROBLEM sylvania in 1992, a prospective internal analysis showed
that 20 per cent of blunt trauma and 57 per cent of pene-
The adverse effects of substance abuse are dependent upon trating trauma victims treated during 1992 tested positive
the agent and the route of administration. Most injected for cocaine.10 In this study 38 per cent of industrial injury
agents are contaminated with diluents. Heroin, naturally a patients had cocaine ‘on board’.10
532 Vascular emergencies caused by substance abuse

Its usage is the most common cause of myocardial

Substance abuse agents infarction or stroke in young patients in America. It affects
arteries of all sizes. This cheap but potent form is smoked
• Marijuana
in a pipe while the rock is ignited from above with a butane
• Cocaine
lighter. Crack smokers have found that transpulmonary
• Angel dust, PCP
absorption is increased by the Valsalva manoeuvre, the
• Amphetamines and such drugs
smoke being blown forcefully into another person’s airway
• Heroin
as that recipient strongly inhales it. This practice may result
in air dissection into the mediastinum, pleural space and
pericardium, the sudden mediastinal emphysema causing
acute and intense pain in the mid-thorax and shortness of
MARIJUANA breath.14 Treatment is supportive until symptoms abate. If
the air dissects into the central vessels it can embolise to the
Marijuana is not usually considered a causative agent for carotid or coronary systems.
problems. Its use in conjunction with other drugs, however, The physiological effects of cocaine are local anaesthesia
cannot be ignored. While there is debate about the extent and a sympathomimetic action derived from blocking the
of harm caused by marijuana usage, research evidence presynaptic reuptake of noradrenaline and dopamine.15
indicates that more carcinogens are released in Cocaine increases coronary artery reactivity and potenti-
marijuana smoke than in cigarette smoke. Baldwin et al. ates the pressor effect of noradrenaline,16 an effect which
at University of California at Los Angeles, USA, examined is dose dependent. The arterial lesions induced by cocaine
immune cell response to smoking different drugs11 and were derive from two pathological effects. The first is an increase
able to show that alveolar macrophages from marijuana in the product of heart rate and blood pressure (p/t)
smokers had reduced bactericidal activity against which may result in dissection of the thoracic aorta or in
Staphylococcus aureus, a common pathogen in drug abusers. rupture of a berry aneurysm and subarachnoid haemor-
The same was not observed in tobacco smokers, cocaine rhage. Figure 44.1 shows the thoracic arteriogram of a
smokers, or non-smokers. Macrophages from non-smokers, 14-year-old boy who, following crack consumption, came
cocaine smokers, and tobacco smokers had similar phago- to us with an extensive thoracoabdominal dissection starting
cytic activity, whereas those from marijuana smokers at the left subclavian artery. The second effect of cocaine is
demonstrated markedly depressed phagocytic activity. spasm of medium or small sized arteries and intravascular
Marijuana smokers have been shown to have depressed thrombosis, which may be partial or complete.17
cytotoxic and immunoprotective levels of interleukin-6, Unexplained thrombi in arteries of all sizes, including
tumour necrosis factor- , and granulocyte-macrophage the aorta, have been documented.15 We have seen patchy
colony-stimulating factor.12 The combined use of marijuana thrombosis in the thoracic and abdominal aortas of cocaine
and other drugs, especially the injected forms, may account
for the frequent infections seen in this patient population.

COCAINE

Cocaine is an alkaloid extracted from the Erythroxylon coca

plant and supplied in powder form. Cocaine is snorted,
smoked and injected intravenously. Cocaine may be inhaled
in several ways. The cocaine alkaloid can be dissolved in
acetone or alcohol, purified to a more potent form, and
smoked as the so called ‘free base’.13 This has the danger of
causing injury from fire or explosion of the flammable solv-
ent. The alkaloid can be dissolved in hydrochloric acid to
form a water soluble salt, available as a crystalline granular
powder, which may be inhaled or smoked with a flam-
mable vehicle. ‘Crack’ cocaine is a neutral form, like ‘free
base’, but is made without solvent extraction. In making
crack, an alkali is added to an aqueous solution of cocaine
hydrochloride and the resulting alkaloid precipitates from Figure 44.1 Arteriogram of a 14-year-old boy with cocaine
the supernatant.13 It is from this that the ‘rocks’ are made induced thoracoabdominal aortic dissection originating just distal
and sold as such. to left subclavian
 Amphetamines 533

users resulting in distal embolisation.18 Cocaine induced other’s effects along with decreasing the dose at which
vasospasm is due not only to cocaine but also its metab- toxic and even lethal side effects can occur.
olites norcocaine and benzoylecgonine.19 Cocaine may Low doses of PCP (3–8 mg) cause mild intoxication.
also augment procoagulants by decreasing protein C and Users show impaired coordination, slurred speech and
antithrombin III levels20 and by increasing plasminogen erratic eye movement. Larger doses (8–12 mg) increase the
activator inhibitor activity, thereby inhibiting thromboly- low dose effects as well as raising heart rate and blood pres-
sis.21 Additionally, cocaine induces platelet activation and sure, and causing fever, sweating, nausea, a blank stare and
aggregation. Long term vascular changes are suggested by a shuffling, disjointed gait that some users call ‘zombie
the increase in adventitial mast cells and atherosclerosis walking’. Doses above 12 mg can unleash a range of serious
in chronic abusers.22 These myriad effects are responsible effects from profound hypotension to muscular rigidity,
for the myocardial infarctions,23 strokes,24 mesenteric convulsions, even coma and death. The titanic muscle
ischaemia,25 venous thrombosis26 and acute renal failure27,28 activity can result in rhabdomyolysis necessitating fas-
observed following cocaine intake. ciotomies. While lower dose effects may only last a few
It has been recommended that the treatment of cocaine hours, higher dose effects can continue for several days.
induced thrombosis be one of heparin anticoagulation Treatment is usually supportive until the body has been
unless terminal organ dysfunction requires urgent opera- cleared of the agent.
tive or other intervention.18 We have seen large intra-aortic
thrombi fully resolve with this treatment as long as the
patient is cocaine free. AMPHETAMINES

Methamphetamine and ‘ecstasy’ (3,4-methylenedioxy-

ANGEL DUST, PCP
methamphetamine) are two drugs now fashionable among
adolescents and young adults. Their presence at ‘rave parties’
Cocaine in powder form may be mixed with other agents. (nocturnal congregations involving dancing, sex, heavy
Although these diluents are often chemically inert, not alcohol consumption and drug usage) is almost ubiqui-
infrequently another active agent may be used, one of tous. Consumption of either of these will result in severe
which is PCP, better known as ‘angel dust’.29 The name inotropic and chronotropic effects. Like cocaine they result
PCP is the abbreviation for its chemical name 1-(1- in strokes and aortic dissections.
phencyclohexyl) piperidine. On the street, a long list of A general scheme for treating the vascular complica-
pseudonyms exists: ‘hog’, ‘squeeze’, ‘wack’ and ‘space base’ tions of non-injected substances is presented as an algo-
(when mixed with crack cocaine). Although angel dust rithm in Fig. 44.2. Identification of the specific agent is
originally referred to a combination of heroin and cocaine crucial so that substance specific treatment can be insti-
its contemporary nomenclature refers to PCP. The problem tuted. Unfortunately, that is not often forthcoming at the
of mixing PCP with cocaine is that the two potentiate each initial presentation of these patients.

Non-injected drug usage

Cocaine PCP All others

No vascular injury Vascular injury Prolonged Supportive care

supportive care

Supportive care Arterial thrombosis Venous thrombosis Haemorrhage

Routine Direct surgical

End organ at risk End organ okay
anticoagulation intervention

Open surgical intervention  Heparin and warfarin  Compression

heparin and warfarin follow-up imaging garment

Figure 44.2 Algorithm for management vascular complications of non-injected substances. PCP, 1-(1-phencyclohexyl) piperidine
(angel dust)
534 Vascular emergencies caused by substance abuse

OTHER AGENTS during the exchange from one distributor to another. The
extent of dilution is unknown unless distributors have
developed an excellent reputation for consistency. One
Anything that can be dissolved or liquefied can and has locally popular version is called ‘mixed jive’ and contains
been injected. Common oral agents that have been used the additives of strychnine, talc powder, lidocaine, quinine,
are paracetamol (acetaminophen) with codeine and penta- sugar, starch and occasionally the ‘dust’ from the interior
zocine (Talwin). Since these do not fully dissolve a com- of spent fluorescent light tubes. Local purveyors often have
mon problem with their usage is particulate embolisation. custom blends named after celebrities.
Even powdered agents such as heroin and cocaine may not The overdosed patient will be somnolent, stuporous or
fully dissolve or will precipitate in the syringe prior to in coma. The breathing pattern is shallow, respiratory rate
injection. If an intravenous route is used they are filtered slowed and minute volume diminished. Cyanosis may
by the pulmonary circulation and are directly responsible occur as heroin reduces the responsiveness of the brain
for pulmonary complications. If an inadvertent, or even stem to increases in carbon dioxide tension and depresses
intentional, intra-arterial route is used, terminal vessel occlu- the pontine and medullary centres involved in regulating
sion often leads to tissue loss,30 not uncommonly seen in respiratory rhythmicity.32,33 Deaths from opiate overdosage
the hand and frequently leading to digit or forearm ampu- are usually due to respiratory depression. Like morphine,
tations. Interdigital injections, used when an intravenous heroin induces peripheral vasodilatation and a decrease in
route is no longer possible, will often result in cannulation systemic vascular resistance augmented by a concomitant
of the digital arteries. Anticoagulant treatment with heparin, release of histamine. Therefore, opiate intoxicated patients
awaiting demarcation to occur, is followed by amputation with hypovolaemia exhibit profound hypotension.
at the level of viable tissue.31 If the carotid is used, the par- Therapy includes establishing an airway and confirming
ticles lodge in the terminal cerebral vessels causing cerebral adequate ventilation. During resuscitation, administration
infarction or seizure and carries the ignoble name of ‘huck- of naloxone hydrochloride, a specific opiate receptor
bucks’, referring to the fictional character Huckleberry antagonist, may be beneficial. Naloxone yields a dramatic
Finn. This name has arisen from the Kentucky hills where reversal of the respiratory depression and peripheral vaso-
the routine use of pentazocine had been popular. dilatation associated with opiate toxicity.32,33 Intravenous
doses of 0.4 mg, repeated every few minutes as necessary,
are effective. However, excessive naloxone may precipitate
HEROIN opiate withdrawal, thus complicating therapy. Because the
effective half-life of naloxone is shorter than that of most
Heroin derives from the poppy plant Papaver opiates, re-dosing every several hours may be necessary
somniferum.32 Morphine is another derivative from it. until the circulating opiate is excreted.
Heroin use is nearly as old as civilisation itself. The effects In heroin addiction the arterial injury is generally a con-
of heroin are mediated by the interaction with endogenous sequence of failing to cannulate a vein for injection and
opiate receptors which function as neurotransmitters, accidentally injecting the adjacent artery, the slang for this
neurohormones and modulators of neural transmission is ‘hitting pink’. The arterial damage from heroin is the
throughout the central nervous system. Heroin is delivered consequence of the chemical and bacterial contamination
to the street in powder form where it is generally diluted in of the mixture along with the bacterial contamination
water and injected intravenously and almost never in asep-
tic fashion. If venous access is no longer available then the
chronic user often resorts to subcutaneous injections.
These frequently lead to dermal ulcerations and scarring
referred to as ‘skin pop marks’ and the practice is referred
to as ‘skin popping’ (Fig. 44.3). The deleterious effects of
heroin in the abuser include overdose, vascular aneurysms
and myriad infectious complications which impair recov-
ery after injury. These infections may be at the site of injec-
tion or anywhere in the body. Endocarditis, both acute and
subacute, is frequently observed.
The problem of overdose related death or severe injury
after heroin use reflects user ignorance regarding the
strength of a particular injection.33 While heroin is distrib-
uted in various levels of purity, pure or almost pure heroin
is quite strong and may cause sudden unexpected death
when a patient thinks the injectate had been diluted. The Figure 44.3 Injection ‘pop-marks’ on the hand; note the oedema
dilution or ‘cutting’ of heroin is done to increase profits and the resulting flexion of the fingers
 Heroin 535

resulting from sharing needles and syringes. An abscess In the upper extremity, the most commonly involved
forms in the arterial wall which undergoes necrosis and arterial sites were the brachial and radial arteries. Patients
results in an infected pseudoaneurysm. As crack cocaine with upper extremity problems presented with either an
usage became popular over the past decade, the incidence infected false aneurysm or with severe hand ischaemia.
of drug induced mycotic pseudoaneurysms at our facility Those with pseudoaneurysms had severe pain, a palpable
fell precipitously. mass and cellulitis. Patients with hand ischaemia had
Twelve years ago, at the height of the heroin epidemic in mostly hand and forearm pain, neuromuscular deficits in
the USA, we reviewed our experience with heroin induced the forearm and gangrenous changes in the fingers. Exten-
arterial injuries. There were 32 upper extremity cases, 136 sive fasciotomies may be needed, as in the case of massive
lower extremity cases and four neck cases. Half of our compartment syndrome following an axillary injection of
patients had positive blood cultures on admission and of heroin (Fig. 44.6).
these, half grew methicillin-resistant Staphylococcus aureus In the neck, the subclavian or common carotid arteries
(MRSA). It should not come as a surprise that bacterial are accidentally hit while the abuser or a friend attempts
endocarditis, such as the one shown in Fig. 44.4, was a cannulation of the subclavian or internal jugular veins.
frequent finding in these patients. Infected false aneurysms The euphemism for this is ‘shooting the pocket’. Since
of the visceral arteries, such as the coeliac artery aneurysm this manoeuvre is often difficult to self-perform, a society
in Fig. 44.5, are metastatic arising from valvular endo- of fellow addicts known as ‘street doctors’ has evolved, and
carditis. for a fee, be it monetary or a portion of the injectate, will
administer the injection.

Figure 44.6 Arm fasciotomy in a drug user who injected into the
Figure 44.4 Post-mortem examination of a drug abuse patient axillary region
showing valvular vegetations from bacterial endocarditis

Figure 44.5 Arteriogram of

metastatic mycotic aneurysm of the
coeliac artery in an active drug user: (a)
(a) (b)
anteroposterior view (b) lateral view
536 Vascular emergencies caused by substance abuse

VASCULAR SYSTEM CONSIDERATIONS Although the lower extremities and groins are common
sites for injection, the upper extremities, the axilla and
neck can be involved as well. Staphylococcus aureus is again
Lymphovenous complications the most common bacterium associated with these injec-
tion sites and 40 per cent is MRSA.34 Abscesses, however,
As stated, abscess formation and cellulitis are common soft are frequently polymicrobial and include not only oral
tissue complications resulting from the use of intravenous flora but also Peptostreptococcus and Bacteroides spp.31,34
drugs obtained on the street. Repeated injections result in Although intravenous antibiotic therapy appropriate to
fibrosis and destruction of dermal and deep lymphatic tis- the organism identified is required, surgical treatment of
sues.34 With the destruction of the lymphatics the extrem- any soft tissue abscess is mandatory. This requires incision
ity is prone to additional infections in turn resulting in and drainage and sometimes excision of grossly infected
further lymphatic damage. These repeated injuries coupled and non-viable tissue. Excision may involve not only the
with the often associated ‘skin popping’ results in a chronic skin and subdermis but also the fascia, muscles and/or
woody oedematous limb. The skin develops epidermal major vessels. Such surgery should only be performed in
hypertrophy termed lichenification (Fig. 44.7), which can the operating room. Although it is tempting to aspirate
harbour fungal infection. These limbs are also prone to infected fluid percutaneously, it is not to be recommended
necrotising fasciitis (Fig. 44.8) and not infrequently emer- near vascular structures. The risk of rupturing unsuspected
gency amputation is required for sepsis control. pseudoaneurysms and causing torrential bleeding should
be borne in mind.35 If fluid is spontaneously draining from
an abscess site, any presence of blood or dark purple stain-
ing of the fluid would indicate that a major vessel is
involved and the operating surgeon must be prepared to
ligate the vessel concerned. It has been our experience that
purulent fluid not of this colour is usually not associated
with injury to a major vessel. All infected wounds should
be left open to heal by secondary intention. If a major vein
is thrombosed and infected then the entire length of that
vein and its thrombotic material must be removed and
excised completely.34
Repeated injection into a peripheral or central vein will
ultimately result in thrombosis of that vessel. Repeated
punctures and the subsequent venous endothelial damage
will often cause superficial thrombosis and phlebitis.36 As
injections are usually performed with the dominant hand
Figure 44.7 50-year-old man with a 30-year history of the contralateral side of the body is most frequently affected.
intravenous drug abuse required open ray amputation of several Treatment of superficial venous thrombophlebitis usually
toes; note hypertrophy of the skin due to chronic oedema from involves elevation of the affected extremity in order to reduce
combined venous and lymphatic obstruction oedema, anti-inflammatory medication and warm com-
presses. Because of the nature of the thrombophlebitis and
the infected needles used by the substance abuser, these
patients are at high risk of developing either local or sys-
temic sepsis. The choice of antibiotic therapy should be
driven by the results of the cultures whenever possible.
One must be cognisant of the high incidence of MRSA in
this patient population. If a local abscess forms at the site of
the superficial thrombophlebitis then the thrombosed vein
should be excised in its entirety and the wound left open
for secondary healing.
The intravenous user generally selects superficial veins
as the early targets of his or her addiction. As these vessels
thrombose the individual is forced to use larger and more
centrally placed veins, either those in the groin, axilla, sub-
Figure 44.8 A 42-year-old man, an active intravenous drug clavian, jugular and then even in the breasts, labia or the
addict, with large venous ulcers of the leg, developed necrotising dorsal penile vein. It is not uncommon, however, for some
fasciitis; he ultimately required an open knee disarticulation for drug users to selectively inject into the femoral vein
control of sepsis because it is relatively easily accessible as well as because
 Vascular system considerations 537

this site lends itself to discretion. The larger veins are no shortly prior to seeing the physician. Surgical thrombec-
less prone to thrombosis than are the superficial veins tomy also has a limited role in the care of these patients.
and over time these too will thrombose and/or become Although mycotic arterial aneurysms are common in
infected. Since many of these patients develop significant patients with repeated intravenous drug use, the entity of
oedema, erythema or pain in the affected extremity phys- mycotic venous aneurysm also exists and usually involves
ical examination cannot be relied upon to make a diagnosis the femoral vein.34,36 It is usually caused by repeated injec-
of deep vein thrombosis (DVT). It is also possible that tions into the femoral vein giving rise to septic phlebitis.
many of these patients will have chronic DVT with super- The vein may rupture because its wall is weakened by infec-
imposed acute DVT. We have relied for years on duplex tion or, alternatively, a persistent communication between
imaging of these vessels in making the differential diagno- the skin puncture site and that in the vein leads to adjacent
sis. The oedema developing in the extremities from these haematoma formation which then becomes infected.36 The
chronic repeated deep and superficial phlebitic insults will presence of a mycotic venous aneurysm should be sus-
result in massively oedematous legs extending all the way pected in patients who have a non-pulsatile painful groin
up to the femoral region. These are often lifelong compli- mass with associated signs of sepsis: pain at the site, fever,
cations. Even those who have not used drugs for decades leucocytosis. It is reported that half of all these patients
will be afflicted by chronic leg oedema as well as by very have recurrent cellulitis.31,34 Although the above would be
large venous ulcers (Fig. 44.9). These patients are also at pathognomonic for a mycotic venous aneurysm, many
high risk of experiencing complications from their DVT patients are asymptomatic. The diagnosis can be made by
which include pulmonary embolism (PE), septic PE and venous duplex imaging but in practice it is often made
bacterial endocarditis.34 unexpectedly in the operating room during groin explo-
Unless there is a specific contraindication, patients ration for a possible abscess.36
diagnosed with acute DVT should be treated with intra- Antibiotic treatment should be guided by cultures and
venous heparin following standard hospital protocols. sensitivities. Debridement of all infected tissue combined
Warfarin should be started, usually concurrently, with the with proximal and distal venous ligation is required.
heparin in anticipation of outpatient therapy. As in any Although the ideal duration of antibiotic therapy is uncer-
patient with chronic venous insufficiency and/or acute tain, a minimum of 2 weeks and up to 6 weeks, blood
DVT, oedema control should be one of the goals of treat- cultures being positive, is usually recommended. Venous
ment. All of these patients should be fitted for good med- reconstruction is rarely indicated and unlikely to be suc-
ical grade support hosiery; the degree of compression we cessful. It is critically important that when ligating the
most frequently use is between 40 and 50 mmHg, replaced infected vein one resects back to what grossly appears to be
at least every 6 months and worn for life if recurrence of normal venous wall. If residual wall infection is left at the
venous ulceration is to be prevented. site of ligation the ligatures will eventually erode through
We do not recommend the use of thrombolytic therapy and massive haemorrhage will ensue.
for DVT in these patients. They are generally unreliable in
giving a history and not infrequently the acute DVT will
have occurred superimposed on a chronic DVT, therefore Intra-arterial injection
making lytic therapy nearly useless. Thrombolytic therapy
also runs the risk of causing bleeding from arteries and/or Inadvertent intra-arterial injection often results in severe
veins which may have been punctured by the individual limb ischaemia and tissue loss. The ischaemic changes
are thought to be due not only to embolisation of non-
solubilised particulate matter but also to arterial spasm,
platelet aggregation, thromboxane release, and sympathetic
mediated vasospasm.35 As the particulate matter passes into
the capillary system vasospasm occurs in both the arterial
and venous side of the tissue bed.36 This, in addition to the
obstruction of the capillary bed by the particles, will result
in acute, severe and often unrelenting pain. The extremity
will be swollen and quite cool37 and the digits flexed or
claw like (Fig. 44.10). Sensory loss is often observed and
occurs despite palpable pulses in the major vessels leading
into the limb. Although this clinical picture is usually true
of the upper limb it is not at all uncommon to see it in the
lower extremities. The ultimate outcome is extensive loss
of tissue, and if that does not occur, permanent neuro-
Figure 44.9 Venous ulcers in a chronic intravenous drug addict logical dysfunction often ensues.38 We have found arteri-
are typically quite large ography to be of little benefit in the overall care of the patient
538 Vascular emergencies caused by substance abuse

Figure 44.10 Patient with particle embolisation of hands from

intravenous drug use

because the damage largely affects the digital vessels and

the microcirculation.
As one waits for demarcation between the viable and the Figure 44.11 Mycotic femoral aneurysm resulting from
gangrenous portions of the limb, pain management can be intravenous drug use; frequently, purulent eruptions arise from the
problematic. Although narcotic administration is neces- apex of the infected site
sary these patients have a very high tolerance to narcotics
and a very low tolerance to pain. Since there is no ceiling
on narcotic dosing, medications should be titrated up until
adequate analgesia is obtained. In the lower extremity con-
tinuous analgesia via epidural catheter can be very benefi-
cial. Intravenous heparin administration in order to achieve
a partial thromboplastin time (PTT) of two to three times
the control may limit the extent of the thrombotic event.
Once a clear line of demarcation occurs then amputation
can be planned. Elevation of the affected extremity to min-
imise oedema with will also aid in reducing the extent of
inflammation and thrombosis.31
In patients sustaining inadvertent intra-arterial drug
injections, Tait et al. have reported significant improve-
ment in distal perfusion following an infusion of prostacy-
clin iloprost.39 No major complications were associated
with the administration of this agent and they believe
it helped to control the ischaemic injury; unfortunately
they did not report the extent of tissue injury seen at initial
presentation.
Figure 44.12 Arteriogram of a drug induced mycotic femoral
If a major limb arterial segment occludes and revasculari-
aneurysm showing extravasation of contrast
sation is performed, compartment fasciotomy may be
required. Since many of these patients have chronic oedema, it
is difficult to determine by mere physical inspection whether there may be an underlying condition of renal impairment
compartment pressures are abnormal. The measurement of due to the chronic drug abuse itself. Intravenous and intra-
these pressures is advocated, avoiding fasciotomy in those arterial injection of illicit drugs will result in chronic
compartments with pressures less than 30 mmHg.39 endothelial injury and a form of drug induced, possibly
As with any major soft tissue injury involving a muscle systemic, vasculitis.
compartment, rhabdomyolysis can occur. This may be
due to the ischaemic event with subsequent reperfusion (see
Chapters 2 and 4) or it may be caused by the dissemination Arterial mycotic aneurysms
of particulate matter into the capillary bed of the muscle.
Renal failure and subsequent haemodialysis is not uncom- The classic presentation of a mycotic aneurysm is one of
mon.40 Renal failure may complicate rhabdomyolysis or a pulsatile mass over a major arterial site associated with
 Vascular system considerations 539

a puncture in the area, the presence of a bruit and intact and for the sake of efficiency, we approached the false
distal pulses.41 The typical clinical appearance of an aneurysm directly, controlling the bleeding with digital
infected femoral pseudoaneurysm is seen in Fig. 44.11. The pressure above and below the area of wall destruction.
septic patient has a painful pulsatile mass in the groin with Today, for the occasionally infected pseudoaneurysm we
some suppuration at the tip where the skin is necrosed. see, and given the very high incidence of HIV and hepatitis
The arteriogram in Fig. 44.12 shows extravasation of con- in these patients, we have gone back to proximal control to
trast through the disrupted wall of the femoral artery. The have a tidier and safer field for the surgical team. The
infection may be induced by inadvertent puncture of the femoral artery segment with inflammation or necrosis of
artery and subsequent seeding of the arterial wall or by the wall is excised. In some cases the excision and simple
adjacent spread of bacteria into the arterial wall by a local ligation is limited to the common femoral artery; in others
abscess. In the drug-using population the most common the length of artery excised also demands ligation of
site for a mycotic arterial aneurysm is the femoral artery. the superficial and deep femoral arteries, in other words,
Although any artery can be affected, the most common triple ligation. When a triple ligation is done severe limb-
sites in descending order are femoral, brachial and radial.34,31 threatening ischaemia of the leg is much more likely and
Yellin et al. report a 5 per cent incidence of unsuspected arte- one cannot predict when it will occur. Our routine is to
riovenous (AV) fistulae in these patients.34 A duplex scan will observe the patient in the recovery room and base our
help one determine the degree of tissue involvement as well decision upon the appearance of the limb and symptoms at
as identify the artery and/or vein involved and the presence that time.
of an AV fistula if one is present. It also will help in docu- At the beginning there was debate as to whether one
menting thrombosis in the venous system which may be should revascularise these legs following inflow ligation.
present. Rarely is peripheral arteriography required at this These patients do not have superficial veins for grafts, these
point. having been ruined by repeated injections and throm-
Management of these patients includes intravenous bophlebitis. One is therefore limited to the use of a pros-
antibiotic therapy, debridement of all infected tissue includ- thetic bypass which has to be routed away from the
ing the artery and/or vein and removal of the entire infected contaminated operative field (see Chapter 18). Two atopi-
segment of vessel.31,34–36,41 Although distal embolisation cal or extra-anatomical routes were used for revascularisa-
can occur from these aneurysms, rupture is the most com- tion of the lower extremity: the lateral femoral route and
mon sequela. If the artery involved is not resected back to the obturator route.35,41,43–45 In the lateral femoral route,
healthy tissue delayed rupture will occur. It is often diffi- after excising the pseudoaneurysm, the patient is prepped
cult to determine how far one needs to resect until healthy clean and using a new set of instruments, the proximal
vessel is encountered. A seasoned surgeon will have a sense external iliac artery is exposed though a retroperitoneal
of what might be termed adequate arterial wall based upon approach for inflow to the bypass tunnelled lateral to the
the appearance of the vessel and the sensation of the needle inguinal ligament into the lateral thigh; from there it is
passing through the segment. If the needle punctures very swung medially and anastomosed to a segment of distal
easily it is possible that that segment is abnormal and more superficial femoral artery.
distant resection and ligation is recommended. The obturator bypass was introduced in 1963 by Shaw
Revascularisation is not routinely required at the time and Baue46 and later expanded upon by Guida amd Moore
of ligation of mycotic pseudoaneurysms unless the extrem- in 1969.47 More contemporary modifications of this tech-
ity involved becomes ischaemic. Subsequently, almost all nique and their application, not only in drug induced
patients experience intermittent claudication, but less than femoral mycotic aneurysms but also for prosthetic graft
20 per cent ultimately require revascularisation to prevent infections, have been described.48–53 The common or
major amputation.31,36,42 In patients with infected mycotic external iliac arteries are exposed in the pelvis for the prox-
aneurysms, such procedures, when indicated, can be prob- imal anastomosis. The graft is tunnelled through the obtur-
lematic. These patients often have no usable autogenous ator foramen into the upper or middle thigh where it is
venous conduit for the repair. joined to the superficial femoral artery. The obturator
We treated 112 aneurysms and six AV fistulae of the bypass is not an easy operation in men who have a rather
lower extremities secondary to drug injection. The most narrow pelvis. More importantly, in most of these patients,
frequent presentation was a painful, pulsatile mass with the common femoral veins have thrombosed through
extensive cellulitis. Foot ischaemic complications were repeated injections, venous outflow occurring through
infrequent. In order to discriminate between a plain groin large collaterals traversing the obturator foramen; serious
abscess and an infected pseudoaneurysm we used both venous bleeding can arise from these collaterals while tun-
ultrasound and angiography. In all cases the aneurysm was nelling the graft blindly through the obturator foramen,
dealt with by a direct approach and excision of the infected even if one keeps anteromedial to the obturator vessels as
femoral artery. At the beginning of our experience we used one should. Since these grafts often lie in a subcutaneous
a separate retroperitoneal exposure of the external iliac location, it is not uncommon for many of these patients, as
artery to obtain proximal control. Further into this series, they remain addicted, to inject the easy-to-locate arterial
540 Vascular emergencies caused by substance abuse

ligation, excision and drainage of the abscess and the use

of antibiotics effective against the causative organism,
usually MRSA. Once sepsis has cleared and the wound is
clean, and if severe ischaemia causing rest pain persists,
consideration may be given to carrying out an extra-
anatomical bypass (see Chapter 18). This somewhat fatalis-
tic policy has been reaffirmed by other and more recent
series.55
The presentation of heroin induced arterial injury is dif-
ferent for the upper extremities. In the arm, the common
presentation is a compartment syndrome with neuromus-
Figure 44.13 This individual had an obturator bypass to the cular deficit and hand ischaemia. This almost certainly
mid-femoral artery but the graft became infected when it was
represents inadvertent intra-arterial injection and has been
used as a drug injection site; a perigraft abscess developed
described above.
requiring explantation of the graft followed by high thigh
amputation Because the practice of ‘shooting the pocket’ can result
in subclavian and/or carotid artery mycotic aneurysms,
treatment of these vessels is not too dissimilar from that of
the lower extremity. Although duplex imaging can be use-
graft. This of course will result in early graft thrombosis ful in confirming the diagnosis and in documenting flow
and graft infection necessitating another operation for into the affected vessel, arteriography may be more accur-
removal of the infected bypass, as seen in Fig. 44.13. A high ate in identifying the origin of the aneurysm and in plan-
thigh amputation is required frequently. ning the overall operation. Common carotid aneurysms
Some surgeons in our group performed simultaneous can be ligated and rarely need reconstruction as long as the
revascularisation if there was evidence of limb-threatening internal carotid artery is patent and can be perfused by
ischaemia or a severe drop in the ankle:brachial index while way of reversed flow from the external carotid system.34
others did not do so in septic patients. When our collective Mycotic aneurysms involving the external carotid system
experience was tabulated, including 118 femoral aneurysms can simply be ligated but those involving the internal
alone,31 we adopted a general policy of not doing simulta- carotid artery should undergo, if possible, not only exci-
neous reconstructions. Our decision was supported by the sion but also reconstruction. Again, the use of prosthetic
following facts. First, few patients need it to preserve their material in this infected site is contraindicated. If an autogen-
limbs,31,34,44 and second, among those patients who had ous venous conduit cannot be found, as is usually the
undergone revascularisation, a substantial number devel- case, the superficial femoral artery can be harvested and
oped graft infection and ended up with a higher level of used as an autogenous arterial conduit followed by replace-
amputation than would have been required for the ischaemia ment of the superficial femoral artery with an appropriate
subsequent to ligation of the pseudoaneurysm.31,46 The prosthetic graft.56
question of vascular reconstruction is generally raised after As repeatedly stated, systemic sepsis in these patients is
a triple ligation. None of the patients undergoing revas- common. Cardiac vegetations from endocarditis can cause
cularisation suffered immediate amputations. Some septic emboli to disseminate throughout the vascular tree.
patients with triple ligation who did not undergo revascu- Pulmonary abscesses and pulmonary mycotic aneurysms
larisation lost their legs. In our series, following excision do occur. The septic emboli may also affect any peripheral
of an infected pseudoaneurysm, and using either triple or central artery as well as intracerebral, adrenal, splenic,
or single ligation, 13 patients out of 102 lost their legs. mesenteric, and renal vessels.34,35 Ischaemic bowel mani-
Three months after treatment, however, the percentage of festations may occur due to embolism to the superior
limb loss experienced by patients who underwent simulta- mesenteric artery or from resulting mycotic aneurysms
neous reconstruction was slightly higher than those who which rupture. In these cases treatment involves ligation
did not, and that was because of infection of the prosthetic of the vessel and bowel resection if the segment of bowel
graft. perfused by the ligated vessel is non-viable.
If systemic sepsis is present at the time of initial pseudo-
aneurysm ligation, we do not advocate routine revascular-
isation as in these patients limb amputation may be the
safest course of treatment. This is in contrast to a report Arterial thrombosis secondary to drug use
from the New York Medical College where concomitant
restoration of blood flow was advocated, but their experi- Intra-arterial injections of drugs can directly result in
ence was limited to 15 cases.54 thrombosis of the vessel with distal embolisation. Self-
In conclusion, for infected femoral pseudoaneurysms administration of substances, and as noted above, espe-
secondary to drug injection, we recommend primary cially those of cocaine, can, due to their pharmacological
 Vascular system considerations 541

Injected drug usage

Mass Acute onset ischaemia

Adjacent to vessel Distant from vessel Arm/axilla Hand/foot Brain

Massive Intravenous heparin,

Incision and edema consider iloprost CT scan
drainage, culture
Neck Femoral region
Duplex and and sensitivity
arteriogram If demarcates, amputate Cerebral infarction
Directed
Duplex
antibiotic therapy
Fasciotomy if compartment Supportive care
Mycotic aneurysm: syndrome
Internal carotid Aneurysm of external, arterial or venous Persistent sepsis
aneurysm common carotid, subclavian Cerebral bleed
Compression garment
for edema control
Ligation back to healthy Fever of unknown
Autogenous arterial Ligation without Decompression if
vessel, wound left open origin work-up
conduit bypass with immediate appropriate and
superficial femoral artery revascularisation supportive care

Leg viable, Cardiac echo to

supportive care rule out vegetations
If need to revascularise, No aneurysm
atopical prosthetic or in-line
autogenous superficial Leg not viable Blood cultures
femoral artery bypass
Abscess
Rule out mycotic
Consider primary visceral aneurysms
Incision and drainage, amputation
culture and sensitivity
directed antibiotics Abdominal CT
Obturator or or arteriogram
lateral femoral
prosthetic bypass

Figure 44.14 Algorithm for determining the treatment of patients who have injected illicit substances. CT, computed tomography

properties, cause thrombosis of large central and/or fully resolves. It has been our experience that as long
peripheral vessels. This can occur not only in the arterial as the patient is free of the inciting agent, frequently
tree but also in the venous system. Lymphatic obstruction cocaine in any of its forms, the thrombus will resolve com-
from systemic usage of drugs other than those injected pletely and at such a time anticoagulant treatment can be
is not well documented. The aetiology of large vessel stopped.
thrombosis, including that of the thoracic and abdominal If the thrombus or thrombotic event threatens end
aorta, is unclear. It is possible that a prothrombotic state organ function then standard thrombectomy or lytic ther-
occurs associated with local spasm of the vasa vasorum apy can be used depending upon the clinical situation. In
resulting in a localised thrombotic event. This can occur those cases where this has been necessary the vessels have
despite a grossly normal endothelium seen at the time appeared grossly normal. We do not recommend long term
of operation.18 If the thrombosis occurs in the venous sys- anticoagulation in these patients, assuming that a work-up
tem then routine treatment of either superficial throm- for endogenous hypercoagulability is negative. On the
bophlebitis or DVT should be instituted. If it involves the other hand, if a demonstrable defect is detected in the
arterial tree then treatment is dictated by any complica- coagulation cascade, or if deficiencies of either protein C,
tions associated with the thrombosis. If a large thrombus is S, antithrombin III, lupus anticoagulant, anticardiolipin
identified it has been our routine not to operate to remove antibodies, factor V Leiden, PT 20210A, etc., are dis-
it unless it has embolised or end organ function is threat- covered, then these patients should be kept on lifelong war-
ened. We have successfully treated large intra-aortic farin therapy unless contraindicated by the overall medical
thrombi with intravenous heparin administration keeping condition. Patient counselling is extremely important as is
the PTT level to 2–2.5 baseline. Heparin should be con- successful antidrug detoxification if complications of anti-
tinued for at least 7–10 days in hospital. Repeat imaging, coagulant therapy are to be avoided.
be it arteriogram, spiral computed tomography or mag- A decision tree for determining the treatment of patients
netic resonance imaging, has been used to monitor the who have injected illicit substances is presented in Fig. 44.14.
regression of the thrombosis. These patients, if deemed This schema can be used as a general guide to therapy
reliable, will subsequently be started on warfarin, and but it must be tailored to the individual patient and
then discharged for follow-up imaging until the thrombus circumstance.
542 Vascular emergencies caused by substance abuse

6 Sloan EP, Zalenski RJ, Smith RF, et al. Toxicology screening in

Conclusions urban trauma patients: drug prevalence and its relationship to
trauma severity and management. J Trauma 1989;
Although the incidence of arterial complications associ- 29: 1647–53.
ated with intravenous drug use has decreased over the 7 Vogel JM, Vernberg EM. Children’s psychological responses to
past two decades, the incidence of thrombosis associated disasters. J Clin Child Psychol 1993; 22: 464–85.
with cocaine usage has increased. There has also been a 8 Fuller MG, Diamond DL, Jordan M, Walter MC. The role of a
recent rise in the recreational use of intravenous heroin substance abuse consultation team in a trauma center. J Stud
in the suburban population as a drug of choice. Whether Alcohol 1997; 56: 267–71.
9 Vogel JM, Vernberg EM. Children’s psychological responses to
there will be a concomitant rise in the complications
disasters. J Clin Child Psychol 1993; 22: 464–84.
from this usage as was noted in the urban population
10 Brookoff D, Campbell EA, Shaw LM. The under reporting of
during the 1970s and 1980s remains to be seen. Many cocaine-related trauma: drug abuse warning network
users have access to clean insulin needles, syringes and reports vs hospital toxicology tests. Am J Public Health 1993;
alcohol swabs obtainable over the counter without 83: 369.
prescription; these factors may help to reduce the histor- 11 Baldwin GC, Tashkin DP, Buckley DM, et al. Marijuana and cocaine
ical incidence of septic complications. Should that hap- impair alveolar macrophage function and cytokine production.
pen, then any complications arising from usage will be Am J Respir Crit Care Med 1997; 156: 1606–13.
dictated by the agent rather than by the route of admin- 12 Tashkin DP, Simmons MS, Sherrill DL, Coulson AH. Heavy habitual
istration. marijuana smoking does not cause an accelerated decline in FEV1
with age. Am J Respir Crit Care Med 1997; 155: 141–8.
13 Pieper B. Physical effects of heroin and cocaine; consideration for
a wound care service. J Wound Ostomy Continence Nurs 1996; 23:
248–55.
Key references 14 Sands DE, Ledgerwood AM, Lucas CE. Pneumomediastinum on a
surgical service. Am Surg 1988; 54: 434–7.
Baldwin GC, Tashkin DP, Buckley DM, et al. Marijuana and cocaine 15 Gawin FH, Ellinwood EH. Cocaine and other stimulants. Actions,
impair alveolar macrophage function and cytokine production. abuse, and treatment. N Engl J Med 1988; 318: 1173–82.
Am J Respir Crit Care Med 1997; 156: 1606–13. 16 Frishman WH, Karpenos A, Molloy TJ. Cocaine induced coronary
Berguer B, Benitz P. Surgical emergencies from intravascular artery disease. Med Clin North Am 1989; 73: 475–86.
injection of drugs. In: (eds). Vascular Surgical Emergencies. 17 Kolodgie FD, Virmani R, Cornhill JE, et al. Increase in
New York: Grune & Stratton, 1987: 309–16. atherosclerosis and adventitial mast cells in cocaine abusers;
Fuller MG, Diamond DL, Jordan M, Walter MC. The role of a an alternative mechanism of cocaine-associated coronary
substance abuse consultation team in a trauma center. vasospasm and thrombosis. J Am Coll Cardiol 1991; 17: 1553–60.
J Stud Alcohol 1997; 56: 267–71. 18 Webber J, Kline RA, Lucas CE. Aortic thrombosis associated with
Kolodgie FD, Virmani R, Cornhill JE, et al. Increase in atherosclerosis cocaine use: report of two cases. Ann Vasc Surg 1999; 13: 302–4.
and adventitial mast cells in cocaine abusers; an alternative 19 Gold MS, Washton AM, Drackis CA. Cocaine abuse;
mechanism of cocaine-associated coronary vasospasm and neurochemistry, phenomenology, and treatment. Natl Inst Drug
thrombosis. J Am Coll Cardiol 1991; 17: 1553–60. Abuse Res Monogr Ser 1985; 61: 130–50.
Webber J, Kline RA, Lucas CE. Aortic thrombosis associated with 20 Chokshi SK, Miller G, Rongione A, Isner JM. Cocaine and
cocaine use: report of two cases. Ann Vasc Surg 1999; 13: cardiovascular diseases: the leading age. Cardiology 1989; III: 1–6.
302–4. 21 Moliterno DJ, Lange Ra, Gerard RD, et al. Influence of intranasal
cocaine on plasma constituents associated with endogenous
thrombosis and thrombolysis. Am J Med 1994; 96: 492–6.
22 Kolodgie FD, Virmani R, Cornhill F, et al. Increase in atherosclerosis
REFERENCES and adventitial mast cells in cocaine abusers: an alternative
mechanism of cocaine-associated coronary vasospasm and
thrombosis. J Am Coll Cardiol 1991; 17: 1553–60.
1 Hypodermic Needle – Syringe Needle. Extract from ‘Blood and 23 Fandino J, Sherman JD, Zuccarello M, Rapoport RM. Cocaine-
Blood Transfusions’. Available at http://inventors.about.com/ induced endothelin-1-dependent spasm in rabbit basilar artery
library/inventors/blsyringe.htm (accessed 7 January 2005). in vivo. J Cardiovasc Pharmacol 2003; 41: 158–61.
2 Browne A. Drugs in Britain. The Observer. 25 March 2001, 24 Daras M, Tuchman AJ, Koppel BS, et al. Neurovascular
London. complications of cocaine. Acta Neurol Scand 1994; 90: 124–9.
3 National Household Survey on Drug Abuse: Population Estimates 25 Boutros HH, Pautler S, Chakrabarti S. Cocaine-induced ischemic
1988. Rockville, MD: National Institute on Drug Abuse, 20857. colitis with small-vessel thrombosis of colon and gallbladder.
4 Harruff RC, Francisco JT, Elkins SK, et al. Cocaine and homicide in J Clin Gastroenterol 1997; 24: 49–53.
Memphis and Shelby County: an epidemic of violence. J Forensic 26 Lisse JR, Davis CP, Thurmond-Anderle ME. Upper extremity deep
Sci 1988; 33: 1231–7. venous thrombosis: increased prevalence due to cocaine abuse.
5 Rogers JN, Henry TE, Jones AM, et al. Cocaine related deaths in Am J Med 1989; 87: 457–8.
Pima County, Arizona, 1982–1984. J Forensic Sci 1986; 31: 27 Sharff JA. Renal infarction associated with intravenous cocaine
1404–8. use. Ann Emerg Med 1984; 13: 1145–7.
 References 543

28 Wohlman RA. Renal artery thrombosis and embolization 42 Cheng SWK, Fok M, Wong J. Infected femoral pseudoaneurysm
associated with intravenous cocaine injection. South Med J 1987; in intravenous drug abusers. Br J Surg 1992; 79: 510–12.
80: 928–30. 43 Johnson JR, Ledgerwood AM, Lucas CE. Mycotic aneurysm: new
29 Turney L. Angel dust: new facts about PCP. Do It Now Foundation, concepts in surgery. Arch Surg 1983; 118: 577–82.
#123, March 2003. 44 Reddy DJ, Smith RF, Elliot JP, et al. Infected femoral artery false
30 Yeager RA, Hobson RW, Padberg FT, et al. Vascular complications aneurysms in drug addicts: evolution of selective vascular
related to drug abuse. Trauma 1987; 27: 305–8. reconstruction. J Vasc Surg 1986; 3: 718–24.
31 Berguer B, Benitz P. Surgical emergencies from intravascular 45 Fromm SH, Lucas CE. Obturator bypass for mycotic aneurysm in
injection of drugs. In: Bergan JJ, Yao ST (eds). Vascular Surgical the drug addict. Arch Surg 1970; 100: 82–3.
Emergencies. New York: Grune & Stratton, 1987: 309–16. 46 Shaw RS, Baue AE. Management of sepsis complicating arterial
32 Vilke GM, Sloane C, Smith AM, Chan TC. Assessment for deaths in reconstructive procedures. Surgery 1963; 53: 75–86.
out-of-hospital heroin overdose patients treated with naloxone 47 Guida PM, Moore WS. Obturator bypass technique. Surg Gynecol
who refuse transport. Acad Emerg Med 2003; 10: 893–6. Obstet 1969; 128: 1307–16.
33 Jaffe JH, Martin WR. Opoid analgesics and antagonists. In: 48 Tilson MD, Sweeney T, Gusberg RT, et al. Obturator canal bypass
Gilman AG, Goodman LS, Rall TW, et al. (eds). The Pharmacological grafts for sepsis lesions of the femoral artery. Arch Surg 1979;
Basis of Therapeutics, 7th edn. New York: MacMillan, 1985: 114: 1031–3.
491–531. 49 Erath HG Jr, Gale SS, Smith BM, et al. Obturator foramen grafts:
34 Yellin AE, Frankhouse JH, Weaver FA. Vascular injury secondary the preferable alternate route? Am Surg 1982; 48: 65–9.
to drug abuse. In: Ernst CB, Stanley JC (eds). Current Therapy in 50 Pearce WH, Ricco J, Yao JST, et al. Modified technique of
Vascular Surgery, 3rd edn, St Louis, MO: Mosby, 1995: 637–44. obturator bypass in failed or infected grafts. Ann Surg 1983;
35 Yeager RA, Hobson RW, Padberg FT, et al. Vascular complications 197: 344–7.
related to drug abuse. Trauma 1987; 27: 305–8. 51 Nevelsteen A, Mees U, Deleersnijder J, et al. Obturator bypass:
36 Woodburn KR, Murie JA. Vascular complications of injecting a sixteen year experience with 55 cases. Ann Vasc Surg 1987;
drug misuse. Br J Surg 1996; 83: 1329–34. 1: 558–63.
37 Silverman SH, Turner WW. Intraarterial drug abuse: new 52 Lai DTM, Huber D, Hogg J. Obturator foramen bypass in the
treatment options. J Vasc Surg 1991; 14: 111–16. management of infected prosthetic vascular grafts. Aust N Z J Surg
38 Treiman GS, Yellin AE, Weaver FA, et al. An effective treatment 1993; 63: 811–14.
protocol for intraarterial drug injection. J Vasc Surg 1990; 12: 53 Sautner T, Niederle B, Herbst F, et al. The value of obturator
456–66. canal bypass: a review. Arch Surg 1994; 129: 718–22.
39 Tait IS, Holdsworth RJ, Belch JJ, et al. Management of intra- 54 Patel KR, Semel L, Clauss RH. Routine revascularization with
arterial injection injury of Iloprost, Lancet 1994; 343: 419. resection of infected femoral pseudoaneurysms from substance
40 Dodd TJ, Scott RN, Woodburn KR, et al. Limb ischemia after abuse. J Vasc Surg 1988; 8: 321–8.
intra-arterial injection of temazepam gel: histology of nine 55 Welch GH, Reid DB, Pollock JG. Infected false aneurysms in the
cases. J Clin Pathol 1994; 47: 512–14. groin of intravenous drug abusers. Br J Surg 1990; 77: 330–3.
41 Feldman AJ, Berguer R. Management of an infected aneurysm of 56 Sessa C, Morasch MD, Berguer R, et al. Carotid resection and
the groin secondary to drug abuse. Surg Gynecol Obstet 1983; replacement with autogenous arterial graft during operation for
157: 519–22. neck malignancy. Ann Vasc Surg 1998; 12: 229–35.
This page intentionally left blank
 45
HIV/AIDS Related Vascular Emergencies

JACOBUS VAN MARLE, LYNNE TUDHOPE

The problem 545 Diagnosis 550

Aetiology 545 Management 551
Vascular pathology and pathophysiology 545 Occupational exposure and prophylaxis 552
HIV and surgery: predictors of operative outcome 546 References 553
HIV and vascular surgery 547

THE PROBLEM coat markers.5 The CD4 T cell has many important
immunological functions and destruction of these cells
causes a reduction in immune response. The disease runs a
The Joint United Nations Program on HIV/AIDS
chronic course, with approximately 10 per cent of infected
(UNAIDS) and the World Health Organization estimated
subjects progressing to AIDS within 2–3-years, and the
that there would be 39.1 million people infected with
remainder developing AIDS within a median of 10 years
human immunodeficiency virus (HIV) by the end of 2004.
from the onset of infection. The virus has been isolated
Approximately 96 per cent of people with HIV/acquired
from all body fluids and transmission occurs through oral,
immune deficiency syndrome (AIDS) live in the developing
rectal and vaginal intercourse, blood product transfusion,
world with 64.5 per cent of them in sub-Saharan Africa.1
intravenous needles, occupational acquisition and vertical
The long latent period of up to 10 years between pri-
transmission from mother to child.6
mary infection and the development of AIDS, an increased
awareness of the disease, and advances in earlier diagnosis
and treatment with an expectancy of longer survival have VASCULAR PATHOLOGY AND
together resulted in an ever increasing population who PATHOPHYSIOLOGY
may also require surgery.2 The reasons for surgical inter-
vention include pathology directly related to HIV/AIDS as
well as HIV seropositive patients who present with com- Joshi et al. were the first to report on the association
mon surgical conditions unrelated to the underlying HIV between arterial disease and HIV.7 They described a fibro-
infection. The distinction between HIV and AIDS is based proliferative occlusive disease in the coronary arteries. This
on the Centres for Disease Control classification.3 consisted of inflammation of the endothelium with lympho-
cytes and mononuclear giant cells which led to fragmen-
tation of elastin fibres and intimal fibrosis resulting in
luminal narrowing. Calabrese et al. described a systemic
AETIOLOGY necrotising vasculitis involving the small vessels in patients
with HIV infection.8
The human immune deficiency virus belongs to the family Du Pont was the first to report on aneurysmal disease
retroviridae, genus lentivirus. There is a huge variation in being associated with HIV, and since then there have been
HIV isolates with two distinct subtypes, namely, HIV-1 many publications dealing with this issue. A selection of
and HIV-2, and a high level of genetic diversity in HIV these publications is listed in the references.9–12 There is also
type1 strains.4 The virus is cytopathic and is found mainly a trend towards multiple aneurysms, atypical location and
in CD4 T cells because of the affinity of the virus for the cell a predilection for the carotid arteries.13–15 The histological
546 HIV/AIDS related vascular emergencies

appearance of HIV related aneurysms has been described HIV AND SURGERY: PREDICTORS OF
by Chetty et al.16 The principal features are involvement of OPERATIVE OUTCOME
the adventitia by a mixed acute and chronic inflammatory
cell infiltrate centred on the vasa vasorum. Occlusion of the
vasa vasorum by inflammatory oedema or cellular infiltrate There are many publications on abdominal surgery in
is a prominent feature. The inflammatory process largely HIV/AIDS patients and mortality and morbidity varies
spares the media and intima. The presence of HIV protein between 0 and 38 per cent and between 7 and 68 per cent,
within lymphocytes in the aneurysm wall has been con- respectively.25–31 Many factors interact, including the stage
firmed through immunohistochemical staining12 and poly- of immune deficiency syndrome, preoperative white cell
merase chain reaction (PCR) done on aneurysmal tissue count, serum albumin concentration, nutritional state, the
confirmed the presence of viral copies in a patient with type of operation: clean versus contaminated and emer-
multiple aneurysms.14 gency versus elective operations, and the presence of oppor-
Although it is apparent that HIV related aneurysms tunistic infections.
constitute a distinct clinical and pathological entity with
several features distinguishing them from degenerative Factors influencing outcome
and infective aneurysms, the precise pathogenesis remains
unclear. Failure to demonstrate microorganisms in the • Stage of immune deficiency syndrome
majority of aneurysms makes the hypothesis of bacter- • Preoperative white cell count
aemia, as a result of immunosuppression and secondary • Nutritional status
mycotic aneurysms, less likely.12 Weakening of the arterial • Anaemia
wall resulting in aneurysm formation may be caused by • Type of operation
direct action of the HIV itself, an immune complex mech- • Opportunistic infections
anism or ischaemia of the arterial wall resulting from
occlusion of the vasa vasorum.8,12 Although viral protein
has been demonstrated in arterial wall biopsies from
aneurysm margins, evidence to support direct viral action Stage of immune deficiency syndrome
leading to destruction of the arterial wall is still lacking.
Nair et al. were the first to report on large vessel occlu- Immunodeficiency is a more important cause of periopera-
sive disease associated with HIV infection.17 They found a tive complications than the operation itself.32 HIV infec-
histological similarity between HIV related occlusive and tion leads to reduction in the CD4 T lymphocyte count and
aneurysmal disease and suggested a leucocytoclastic vas- patients with a CD4 count of 200 cells/L are probably at
culitis of the vasa vasorum as the common pathogenesis. a greater risk of developing a postoperative infection. This
There is growing evidence that HIV influences the physi- includes worsening of preoperative infections as well as
ology of vascular endothelium.18 Certain endothelial cell newly acquired postoperative opportunistic infections. Yii
products which are considered to be markers for endothelial et al.,29 Savioz et al.30 and Consten et al.33 reported that
cell dysfunction, are significantly elevated in patients with perioperative CD4 T lymphocyte counts were significantly
HIV. These cell products include the von Willebrand factor, lower in patients with overall postoperative complications,
tissue plasminogen activator, 2-microglobulin and soluble disturbed wound healing and infections.
thrombomodulin.19 Markers of thrombin activation (frag- Tran et al. suggested that a low percentage of lympho-
ment 1 and 2) and D-dimers are used to determine whether cytes and postoperative CD4 lymphocyte counts have a
HIV infected patients have a prothrombotic state. Plasma stronger correlation with mortality.34 As expected, patients
concentration of thrombomodulin is assessed to establish with AIDS fare worse than those who are HIV positive,
whether an endothelial lesion is concurrently present or not. with poorer wound healing, more complications and a
The presence of anticardiolipin antibodies, protein S defi- higher mortality.25 It has also been found that there is no
ciency and antithrombin deficiency contribute to the hyper- significant difference in postoperative morbidity and mor-
coagulable state associated with HIV/AIDS.20–22 tality between symptom-free HIV positive patients and
HIV infection has been associated with hypertriglyceri- HIV seronegative patients and that if the CD4 T cell count
daemia which may cause endothelial damage and predispose is above 500 cells/L, the morbidity in clean surgery is
to accelerated atherosclerosis.23 There are certain class specific comparable to that of HIV negative patients.30,31,35
metabolic side effects of protease inhibitor therapy which A recent study by Mellors et al. demonstrated that
may also contribute to accelerated atherosclerotic disease. plasma viral load was the single best prognostic indicator of
These side effects include increased insulin resistance (which clinical outcome and that HIV-1 RNA concentration was
is associated with abnormalities in endothelial function, highly predictive of the rate of decline of CD4 lymphocyte
impaired nitric oxide production and diminished vasodilata- counts and progression to AIDS and death.36 Viraemia
tion) as well as abnormalities in lipid metabolism with ele- is indicative of active viral replication with continuous
vated levels of total cholesterol and serum triglycerides.24 reinfection resulting in the destruction of CD4 lymphocytes
 HIV and vascular surgery 547

and an increase in the total number of virus-producing some patients who present with complications of hyperco-
cells.36,37 agulability due to HIV/AIDS.

White cell count Vascular problems in HIV/AIDS

Several studies report significantly improved survival with • Vascular condition unrelated to underlying HIV
lower wound complication rates in patients with a higher infection
preoperative total white cell count.27,34 • Vascular disease directly related to HIV/AIDS
• Complications of hypercoagulability due to
HIV/AIDS
Nutritional status

Low serum albumin levels have been associated with poor

Due to the high incidence of HIV/AIDS in South Africa
surgical outcome in HIV patients and are independent
we follow a policy of routine testing for HIV on all admis-
prognostic factors in survival studies of AIDS patients.25,27,38
sions to the vascular unit. Patient consent is required for
Whitney et al. found a weight loss of 10 per cent to be a
HIV testing and the vast majority of patients comply. The
significant prognostic indicator. 27
incidence of HIV positivity in our unit is 15 per cent.
The majority of patients who present with vascular prob-
Anaemia lems and are found to be HIV positive, present with either
occlusive disease, aneurysms or trauma, and are otherwise
Binderow et al.38 found a trend in low versus high well, exhibiting no symptoms or signs of immunodefi-
haematocrit values as a predictor of outcome whereas Yii ciency. The spectrum of HIV/AIDS related vascular disease
et al.29 found haemoglobin concentration to be a signifi- includes fibro-obliterative disease, necrotising vasculitis,
cant prognostic indicator. aneurysms and the complications of hypercoagulability.
Spontaneous arteriovenous (AV) fistula arising from HIV
arteritis has recently been described.39
Type of operation There are some clinical characteristics specific to HIV
related vascular disease which should alert the treating
Emergency surgery has a higher mortality than elective pro-
physician to the possibility of underlying AIDS.
cedures in any cohort of patients. Savioz et al. found a signifi-
cant difference between clean and contaminated operations,
especially if the CD4 T cell count was 200 cells/L.31 They
Clinical features suggestive of HIV related
recommend that a CD4 cell count 200 cells/L should be
vascular disease
considered a threshold for contaminated surgery and it
should only be undertaken when the procedure is considered
to be unavoidable. With clean surgery, mainly orthopaedic • Younger presenting age
procedures, the probability of morbidity remains 50 per • Absence of typical risk factors for atherosclerosis
cent even with a very low CD4 count. • Multiple aneurysms
• Atypical location of aneurysms
• Features of immunodeficiency
Opportunistic infections

The presence of opportunistic infections implies an Five short case descriptions demonstrate the spectrum
immunocompromised state and results in a poorer surgical of disease treated in our unit. All the patients had a positive
outcome.25 smoking history but no other risk factors for vascular
disease.

HIV AND VASCULAR SURGERY Case report 1

Vascular problems in HIV/AIDS patients manifest them- A 25-year-old man presented with a large, tender, pulsat-
selves in three different ways. First, there are those patients ing mass involving the right axilla and infraclavicular
who present with the normal spectrum of vascular disease region (Fig. 45.1a). On angiography he was found to have
but are incidentally found to be HIV positive, and in whom a contained rupture of a right subscapular artery aneurysm
the vascular condition is not related to the underlying HIV (Fig. 45.1b) and multiple aneurysms involving the peri-
infection. Second, there are those who present with vascu- pheral arteries and abdominal aorta (Fig. 45.1c and d,
lar disease directly related to HIV/AIDS. Finally, there are respectively). The patient had AIDS with a CD4 count
548 HIV/AIDS related vascular emergencies

of 200 cells/L, but due to the fact that the aneurysm oximetry remained normal after test clamping of the com-
was severely symptomatic with pain and restriction of mon carotid artery. The aneurysm was therefore resected
movement and function in the right arm, the decision to and the vessels ligated. Postoperative recovery was unevent-
operate was made. The aneurysm was evacuated and the ful and the patient is being followed-up.
subscapular artery tied off. Postoperative recovery was
uneventful but the patient died 6 months later due to
Case report 3
advanced AIDS.
A 60-year-old man presented with a huge pulsatile mass in
Case report 2 the left popliteal fossa, venous congestion of the lower limb
and limiting claudication (Fig. 45.3a). A computed tomog-
A 29-year-old man presented with a pulsatile mass on the raphy (CT) scan and angiography confirmed the diagnosis
left side of his neck (Fig. 45.2a). Carotid angiography con- of a huge popliteal artery aneurysm (Fig. 45.3b and c,
firmed an aneurysm of the carotid bifurcation involving respectively). The patient’s CD4 count was 500 cells/L.
both the external and internal carotid arteries (Fig. 45.2b). The aneurysm was excised and a femoropopliteal bypass
His CD4 count was 200 cells/L, but he was otherwise
asymptomatic. Due to extensive involvement of the inter-
nal carotid artery, arterial reconstruction was not possible.
Carotid stump pressure was 50 mmHg and intraopera-
tive electroencephalogram (EEG) as well as transcranial

(a) (b)

(c) (d)

Figure 45.1 (a) An aneurysm in the right axilla of a 25-year-old man. (b) Angiogram reveals a contained rupture of the subscapular artery.
(c) Angiogram also shows multiple aneurysms of the peripheral arteries. (d) Angiogram additionally demonstrates an abdominal aortic
aneurysm
 HIV and vascular surgery 549

performed using the saphenous vein. The patient had an

uneventful recovery and is being followed up.

Case report 4

A 49-year-old man presented with a non-viable left lower

limb and a history of having had a right below-knee ampu-
tation three years’ previously. Angiography confirmed

(a)

(a)

(b)

(b) (c)

Figure 45.2 (a) Left carotid artery aneurysm in a 29-year-old Figure 45.3 (a) Popliteal artery aneurysm in a 60-year-old man.
man. (b) Angiogram shows the aneurysm of the carotid bifurcation (b) Computed tomography scan of the popliteal artery aneurysm.
involving both the external and internal carotid arteries (c) Angiogram also shows the popliteal artery aneurysm
550 HIV/AIDS related vascular emergencies

• CD4/CD8 ratio
• HIV-1 RNA count
• Hypercoagulability screening

The treating physician must be aware of the systemic

effects of HIV/AIDS which may influence anaesthetic
technique, operative outcome and postoperative manage-
ment. These systemic effects include cardiac, pulmonary
and central nervous system manifestations as well as
hypercoagulability and these conditions should be investi-
gated using appropriate tests. Two excellent review articles
discussing these problems in detail are recommended.40,41
The cardiac and pulmonary manifestations of HIV/AIDS
are summarised below.

Cardiac manifestations of HIV/AIDS

Figure 45.4 Angiogram showing occlusion of the infrarenal aorta • Myocarditis

• Dilated cardiomyopathy
complete occlusion of the infrarenal aorta (Fig. 45.4). The • Pericardial effusion
CD4 count was 900 cells/L. A high upper-leg amputation • Endocarditis
was performed, but due to poor wound healing he required • Coronary artery disease
disarticulation of the hip. • Pulmonary hypertension, right ventricular failure
• Drug related cardiotoxicity
• Nutritional cardiomyopathy
Case report 5 • Cardiac involvement in AIDS related tumors

A 65-year-old man presented with bilateral critical lower

limb ischaemia (Fontaine stage IV), bilateral lower limb The frequent association between tuberculosis and
swelling and empyema. He had bilateral femoropopliteal HIV/AIDS must be kept in mind.42 Spinal cord involve-
occlusive disease on angiography and a venous duplex ment, myopathy and peripheral neuropathy may occur
Doppler study confirmed bilateral deep vein thrombosis. with cytomegalovirus (CMV) or HIV infection itself.43
The CD4 count was below 200 cells/L and anticardiolipin
antibodies were positive. He refused amputation and was
managed with anticoagulation and general supportive
measures. He died within 10 days due to advanced disease. Pulmonary manifestations of HIV/AIDS41

DIAGNOSIS • Pneumocystis carinii pneumonia

– Respiratory failure
– Pneumatocoeles
In addition to the usual diagnostic work-up required for – Pneumothorax
vascular patients certain additional tests are required in the – Chronic respiratory insufficiency
patient with HIV/AIDS. These pertain mostly to predictors
of operative outcome and complications.
• Cavitatory lung disease
– Pyogenic bacterial lung abcess
– Pulmonary tuberculosis
Additional diagnostic tests in HIV/AIDS – Fungal infection
patients – Nocardia infection
• Disseminated tuberculosis
• Sexually transmitted diseases • Respiratory failure due to bacterial pneumonia
– Streptococcus pneumoniae, Moraxella catarrhalis,
• Tuberculosis
Haemophilus influenzae
• Hepatitis
– Staphylococcus aureus, Pseudomonas aeruginosa
• CD4 cell count
 Management 551

HIV status unknown Manage as if HIV ()

Emergency treatment

Adjust  to patient
HIV ()
condition,  risk, etc.

CD4 500 Manage as if HIV ()

• Conservative treatment
Elective treatment
(Consider antiretroviral CD4 200–500 • Less invasive 
treatment) • Wider application of
endovascular techniques

•  Only for severely

symptomatic patients
CD4 200 • Minimally invasive 
Figure 45.5 Algorithm for managing
• Extended prophylaxis
vascular problems in HIV/AIDS

Various factors contribute towards a procoagulant state because of HIV status or concern for subsequent complica-
in HIV/AIDS and deep vein thrombosis has been reported tions.34,41,48
to occur 10 times more frequently in these patients than in The algorithm in Fig. 45.5 provides an outline for the
the general population.44 Hypercoagulability may not only management of vascular problems in HIV/AIDS.
cause venous thromboembolism but may contribute to
bypass/graft occlusion and therefore the necessary precau-
tions should be taken. Emergency surgery

Surgery for vascular emergencies is performed regardless

Increased risk for venous thromboembolism of HIV status, because seropositivity and degree of
in HIV/AIDS immunodeficiency are often only available after comple-
tion of surgery. When confirmed HIV positive patients
• Hypercoagulability present with life- or limb-threatening vascular conditions,
– Anticardiolipin antibodies surgery may be adjusted according to the degree of
– Protein S deficiency immunodeficiency and the patient’s general condition; for
– Antithrombin deficiency example, opting for amputation rather than extensive
– Procoagulant products of endothelial cell revascularisation procedures. Our own experience sup-
dysfunction ports that of Carrillo et al. that low levels of morbidity and
• Opportunistic infections: cytomegalovirus45 mortality can be achieved with standard surgical care and
• Intravenous drug abuse techniques.49 Guth et al. found that the incidence of com-
• Malignancies plications in trauma patients with HIV was associated with
an increase in the injury severity score (ISS) rather than
HIV status.50

MANAGEMENT
Elective surgery
Concerns do exist that general anaesthesia and/or surgery
may compound immunosuppression in HIV/AIDS patients Asymptomatic HIV seropositive patients as well as those
leading to more rapid progression of the disease. The with a CD4 count above 500 cells/L have no increased sur-
immune suppression seen postoperatively in HIV seronega- gical morbidity or mortality when compared with HIV
tive patients is transient and does not affect recovery or seronegative patients and are treated as such.30,35 For those
prognosis.46,47 There is no documented evidence that sur- patients with a CD4 count of 500 cells/L, a conservative
gery or any related intervention hastens the disease process alternative to operation should be strongly considered.31
in HIV/AIDS; these interventions should not be withheld This may involve a wider application of endovascular
552 HIV/AIDS related vascular emergencies

techniques. Patients with AIDS (CD4 200 cells/L) have subclavian arteries. The cost implications of covered stent-
a median survival time of 1 year. Surgery, however, should grafts and endovascular procedures may limit their use in
still be offered to severely symptomatic patients with limb- units or in countries with budgetary constraints.
and life-threatening conditions, and in whom symptoms
can be alleviated with minimum morbidity. This type of
Antiretroviral therapy
surgery is, essentially, only palliative and should be as min-
imally invasive as possible. This includes, for example, doing
The use of highly active antiretroviral therapy (HAART)
an endarterectomy or a profundoplasty or even a lower
has resulted in dramatic declines in morbidity and mortal-
limb amputation in preference to an extensive bypass pro-
ity among HIV infected patients with advanced immune
cedure for critical lower limb ischaemia. One may also have
suppression.51 There is, however, growing concern about
to accept, for example, an extra-anatomical bypass which
patient compliance and the long term adverse effects of
may have lower long term patency than that of a more inva-
therapy. These include impact on quality of life, drug inter-
sive anatomical procedure. Ligation instead of repair of
actions, viral resistance and potential metabolic abnor-
aneurysms may also be considered in this patient group.
malities including premature cardiovascular disease. The
International AIDS Society has published updated recom-
Principles of surgery mendations for antiretroviral therapy.52 Perioperative use
at this stage is limited to patients who require elective
The standard principles of vascular surgery apply in HIV surgery that can be postponed for at least 3 months for
positive patients. Diseased arterial segments are excluded possible beneficial therapeutic effect. Tran et al., however,
or bypassed from the circulation and care is taken to found that the presence or absence of antiretroviral ther-
perform anastomoses to macroscopically normal tissue. apy and the number of antiretroviral drugs did not affect
Autogenous vein is preferred as a bypass conduit wherever surgical outcome.34
possible. Where vein cannot be used, in circumstances
where it is either unavailable or when a larger diameter
conduit is required, we will resort to polyfluorotetraethyl- OCCUPATIONAL EXPOSURE AND
ene or polyester grafts. PROPHYLAXIS
Where there is a high probability of sepsis, polyester
grafts previously soaked in rifampicin or the commercially
Healthcare workers looking after patients with HIV/AIDS
available silver-coated grafts have been used. Standard
should take the necessary precautions against accidental
antibiotic prophylaxis for vascular surgery consists of a
exposure. This includes double gloving, water resistant sur-
first generation cephalosporin. Savioz et al. found that 35
gical gowns, eye protection and careful surgical technique
per cent of infective complications were caused by oppor-
to minimise blood spillage. Guidelines for the management
tunistic infections outside the range of normal vascular
of healthcare worker exposure to HIV and recommenda-
prophylaxis.31 These complications will require therapeu-
tions for post-exposure prophylaxis and therapy have been
tic antibiotic and antifungal treatment according to culture
described in detail.53,54
and sensitivity. Prophylaxis against P. carinii pneumonia
should be given to all patients with a CD4 T cell count
of 200 cells/L.3 Similarly, a single dose of fluconazole
could probably prevent oral and oesophageal candidiasis.31
Conclusions
Endovascular management
HIV/AIDS is reaching pandemic proportions in the
developing world. Therefore, surgeons will encounter
There are no published series on endovascular management
an increasing population of HIV seropositive patients
of vascular problems in HIV/AIDS patients. Our indications
presenting with a spectrum of vascular problems.
for endovascular therapy in asymptomatic HIV seropositive
Emergency vascular surgery for trauma is performed
patients (CD4 T-cell count 500 cells/L) are the same as
regardless of HIV status. Elective surgery is undertaken
for HIV seronegative patients. In patients with a CD4 count
in patients who are not immunocompromised as it is for
of 500 cells/L, more liberal use is made of percutaneous
HIV seronegative patients. In patients, who are already
techniques, extending the application of percutaneous
immunocompromised, a more conservative approach is
transluminal angioplasty (PTA) to SCVIR (Society for
being adopted, surgery being reserved for patients who
Cardio-Vascular and Interventional Radiology) categories
are severely symptomatic. Even then, less invasive pro-
III and IV lesions. In AIDS patients, endovascular manage-
cedures are preferred. Life expectancy and the condition
ment should be considered in critical limb ischaemia,
of the patient should always be weighed against the
aneurysms and trauma in those locations which would
potential risk of the operation.
require difficult or extensive access, for example, iliac or
 References 553

10 Sinzobahamvya N, Kalangu K, Hamel-Kalinowski W. Arterial

There is a need for multidisciplinary care in managing aneurysms associated with human immunodeficiency virus (HIV)
patients with HIV/AIDS and collaboration between dif- infection. Acta Chir Belg 1989; 89: 185–8.
ferent specialists is essential. Management strategies 11 Marks C, Kuskov S. Patterns of arterial aneurysms in acquired
should be developed based on the everchanging under- immunodeficiency disease. World J Surg 1995; 19: 127–32.
standing of the pathogenesis and treatment of HIV/AIDS 12 Nair R, Abdool-Carrim ATO, Chetty R, Robbs JV. Arterial
and this will ensure optimum treatment of this rapidly aneurysms in patients infected with human immunodeficiency
expanding patient population. virus: a distinct clinical pathology entity? J Vasc Surg 1999; 29:
600–7.
13 Nair R, Robbs JV, Naidoo NG, Woolgar J. Clinical profile of
HIV-related aneurysms. Eur J Vasc Endovasc Surg 2000; 20:
235–40.
Key references 14 Tudhope LE, Van Marle J. Multiple arterial aneurysms in an HIV
infected patient: retrovirus positivity established as possible
Avidan MS, Jones N, Pozniak AL. The implications of HIV for the etiology by means of the polymerase chain reaction [abstract].
anaesthetist and the intensivist. Anaesthesia 2000; 55: Vascular Association of SA Conference, Sun City, South Africa,
344–54. 6–9 August 1999: 2.
Barbaro G. Cardiovascular manifestations of HIV infection. J R Soc 15 Veller M, Pillay J, Abdool-Carrim T, Britz R. Aneurysms in patients
Med 2001; 94: 384–90. with HIV infection: Involvement of the carotid artery bifurcation
Nair R, Robbs JV, Naidoo NG, Woolgar J. Clinical profile of HIV- [abstract]. Cardiovasc Surg 2001; 9(suppl 1): 15.
related aneurysms. Eur J Vasc Endovasc Surg 2000; 20: 16 Chetty R, Batitang S, Nair R. Large artery vasculopathy in HIV
235–40. positive patients: Another vasculitic enigma. Human Pathol
Savioz D, Chilkot M, Ludwig C, et al. Preoperative counts of CD4 2000; 31: 374–9.
T-lymphocytes and early postoperative infective complications 17 Nair R, Robbs RV, Chetty R, Naidoo NG, Woolgar J. Occlusive
in HIV-positive patients. Eur J Surg 1998; 164: 483–7. arterial disease in HIV-infected patients: a preliminary report.
Trann HS, Moncure M, Tarnoff M, et al. Predictors of operative Eur J Vasc Endovasc Surg 2000; 20: 353–7.
outcome in patients with human immunodeficiency virus 18 Chi D, Henry J, Kelly J, et al. The effects of HIV infection on
infection and acquired immunodeficiency syndrome. Am J Surg endothelial function. Endothelium 2000; 7: 233–42.
2000; 180: 228–33. 19 Blann A, Constans J, Dignat-George F, Seigneur M. The platelet
and endothelium in HIV infection. Br J Haematol 1998; 100:
613–14.
20 Bloom EJ, Abrams DI, Rogers G. Lupus anticoagulant in the
acquired immunodeficiency syndrome. JAMA 1988; 256: 491–3.
REFERENCES 21 Lafeuillade A, Alessi MC, Martin-Poizot I, et al. Protein S
deficiency in HIV infection. N Engl J Med 1991; 324: 1220.
1 World HIV and AIDS statistics. www.avert.org/wordstats.htm 22 Von Kaulla E, Von Kaulla KN. Antithrombin 111 and diseases.
2 Vipond MN, Ralph DJ, Stotter AT. Surgery in HIV-positive and Am J Clin Pathol 1967; 48: 69–80.
AIDS patients: indications and outcome. J R Coll Surg Edinb 23 Grunfeld C, Kotler DP, Hamadeh R, et al. Hypertriglyceridemia
1991; 36: 254–8. in the acquired immunodeficiency syndrome. Am J Med 1989;
3 Centers for Disease Control and Prevention. 1993 Revised 86: 27–31.
classification system for HIV infection and expanded surveillance 24 Behrens G, Dejam A, Schmidt H, et al. Impaired glucose tolerance
case definition for AIDS among adolescents and adults. MMWR and beta cell function and lipid metabolism in HIV patients
Morb Mortal Wkly Rep 1992; 41: 1–19. under treatment with protease inhibitors. AIDS 1999; 13:
4 Benn S, Rutledge R, Folks TJ, et al. Genomic heterogeneity of F63–70.
AIDS retroviral isolates from North America and Zaire. Science 25 Diettrich NA, Cacioppo JC, Kaplan G, Cohen SM. A growing
1985; 230: 949–51. spectrum of surgical disease in patients with human
5 McDouglas JS, Kennedy MS, Sligh JM. Binding of HTLV-111/LAV immunodeficiency virus/acquired immunodeficiency
to T4  T-cells by a complex of 110 K molecule. Science 1985; syndrome: experience with 120 major cases. Arch Surg 1991;
231: 382–85. 126: 860–6.
6 Quinn TC. Global burden of the HIV pandemic. Lancet 1996; 26 Binderow SR, Cavallo RJ, Freed J. Laboratory parameters as
348: 99–106. predictors of operative outcome of major abdominal surgery in
7 Joshi VV, Pawel B, Conor E, et al. Arteriopathy in children AIDS and HIV infected patients. Am Surg 1993; 59: 754–7.
with acquired immune deficiency syndrome. Pediatr Pathol 1987; 27 Whitney TM, Brunel W, Russel TR, et al. Emergent abdominal
7: 261–75. surgery in AIDS: experience in San Fransico. Am J Surg 1994;
8 Calabrese LH, Estes M, Yen-Liebermann B, et al. Systemic 168: 239–43.
vasculitis in association with human immunodeficiency virus 28 Bizer LS, Pettorino R, Ashikari A. Emergency abdominal
infection. Arthritis Rheum 1989; 32: 569–76. operations in the patient with acquired immunodeficiency
9 Du Pont JR, Bonavita JA, Di Giovanni RJ, et al. Acquired syndrome. J Am Coll Surg 1995; 180: 205–9.
immunodeficiency syndrome and mycotic abdominal 29 Yii MK, Saunder A, Scott DF. Abdominal Surgery in HIV/AIDS
aneurysms: a new challenge? Report of a case. J Vasc Surg patients: indications, operative management, pathology and
1989; 10: 254–7. outcome. Aust N Z J Surg 1995; 65: 320–6.
554 HIV/AIDS related vascular emergencies

30 Savioz D, Lironi A, Zurbuchen P, et al. Acute right iliac fossa pain 43 Price RW. Neurologic complications of HIV infection. Lancet
in acquired immunodeficiency: a comparison between patients 1996; 348: 445–52.
with and without acquired immune deficiency syndrome. Br J 44 Saber AA, Aboolian A, La Raja RD, et al. HIV/AIDS and the risk of
Surg 1996; 83: 644–6. deep vein thrombosis: a study of 45 patients with lower
31 Savioz D, Chilkot M, Ludwig C, et al. Preoperative counts of CD 4 extremity involvement. Am Surg 2001; 67: 645–7.
T-lymphocytes and early postoperative infective complications in 45 Van Dam-Mieras MC, Bruggemen CA, Muller AD, et al. Induction
HIV-positive patients. Eur J Surg 1998; 164: 483–7. of endothelial cell procoagulant activity by cytomegalovirus
32 La Raja RD, Rothenberg RE, Odom JW, Mueller SC. The incidence infection. Thromb Res 1987; 47: 69–75.
of intra-abdominal surgery in acquired immunodeficiency 46 Lennard TW, Shenton BK, Borzotta A, et al. The influence of
syndrome: a statistical revue of 904 patients. Surgery 1989; surgical operations on components of the human immune
105: 175–9. system. Br J Surg 1985; 72: 771–6.
33 Consten EC, Slors FJ, Noten HJ, et al. Anorectal surgery in human 47 Tonnensen E, Wahlgreen C. Influence of extradural and general
immunodeficiency virus-infected patients. Clinical outcome in anaesthesia on natural killer cell activity and lymphocyte
relation to immune status. Dis Colon Rectum 1995; 38: 1169–75. subpopulations in patients undergoing hysterectomy. Br J
34 Tran HS, Moncure M, Tarnoff M, et al. Predictors of operative Anaesth 1988; 60: 500–7.
outcome in patients with human immunodeficiency virus 48 Ayers J, Howton MJ, Layon AJ. Post-operative complications in
infection and acquired immunodeficiency syndrome. Am J Surg patients with human immunodeficiency virus disease. Clinical
2000; 180: 228–33. data and literature review. Chest 1993; 103: 1800–7.
35 Paiement GD, Hymes RA, LaDouceur MS, et al. Postoperative 49 Carrillo EH, Carrillo LE, Byers PM, et al. Penetrating trauma and
infections in asymptomatic HIV-seropositive orthopedic trauma emergency surgery in patients with AIDS. Am J Surg 1995; 170:
patients. J Trauma Inj Inf Crit Care 1994; 37: 545–51. 341–4.
36 Mellors JW, Munoz AM, Giorgi JV. Plasma viral load in 50 Guth AA, Hofstetter SR, Pachter HL. Human immunodeficiency
CD4 lymphocytes as prognostic markers of HIV-1 infection. virus and the trauma patient: factors influencing post-operative
Ann Intern Med 1997; 126: 946–54. infectious complications. J Trauma Inj Infect Crit Care 1996; 41:
37 Hughes MD, Daniels MJ, Fischl MA, et al. CD 4 cell count as a 251–6.
surrogate endpoint in HIV clinical trials: a meta-analyses of 51 Palella FJ, Delaney KM, Moorman AC, et al. Declining morbidity
studies of AIDS Clinical Trials Group. AIDS 1998; 12: 1823–32. and mortality among patients with advanced human immuno
38 Binderow SR, Cavallo RJ, Freed J. Laboratory parameters as deficiency virus infection. N Engl J Med 1998; 338: 853–60.
predictors of operative outcome after major abdominal surgery 52 Carpenter CC, Cooper DA, Fischl MA, et al. Antiretroviral therapy
in AIDS- and HIV-infected patients. Am Surg 1993; 59: 754–7. in adults: updated recommendations of the International AIDS
39 Nair R, Chetty R, Woolgar J, et al. Spontaneous arteriovenous Society – USA panel. JAMA 2000; 283: 381–90.
fistula resulting from HIV arteritis. J Vasc Surg 2001; 33: 186–7. 53 Centres for Disease Control and Prevention. Public Health Service
40 Barbaro G. Cardiovascular manifestations of HIV infection. J R guidelines for the management of health care worker exposure
Soc Med 2001; 94: 384–90. to HIV and recommendations for post exposure prophylaxis.
41 Avidan MS, Jones N, Pozniak AL. The implications of HIV for the MMWR Morb and Mort Wkly Rep 1998; 47(RR-7): 1–33.
anaesthetist and the intensivist. Anaesthesia 2000; 55: 344–54. 54 Department of Health. Guidelines on post-exposure prophylaxis
42 Drobniewski FA, Pozniak AL, Uttley AHC. Tuberculosis and AIDS. for healthcare workers occupationally exposed to HIV. London:
J Med Microbiol 1995; 43: 85–91. Department of Health, 1997.
 46
Portal Hypertension and Variceal Bleeding

ROWAN W PARKS, THOMAS DIAMOND

The problem 555 Emergency resuscitation and control of bleeding varices 557
Aetiology 555 Long term and definitive treatment 562
Pathophysiology 555 Prophylactic therapy 566
Clinical diagnosis 556 Results and prognosis 567
Investigations 556 References 569
Management 557

THE PROBLEM causes can be further subdivided into presinusoidal, sinus-

oidal and postsinusoidal. Prehepatic portal hypertension is
due to portal vein thrombosis, the cause of which is
Although uncomplicated portal hypertension does not unknown in the majority of cases, but may be secondary to
require treatment, urgent intervention is obviously indicated sepsis, trauma or malignant obstruction. By far the most
once bleeding occurs from oesophageal varices. Several common intrahepatic cause of portal hypertension in the
treatment options are available, but the management will USA and Europe is cirrhosis.2 The pathogenesis of this
ultimately be tailored according to the medical fitness of sinusoidal obstruction to portal flow is varied and includes
the patient, and the medical facilities and local clinical alcohol abuse, which is the most common cause, viral
expertise.1 The availability of several non-operative therapies, hepatitis, primary and secondary biliary cirrhosis and
including pharmacotherapy, balloon tamponade, endoscopic haemochromatosis. In terms of the actual incidence of
injection sclerotherapy, endoscopic variceal ligation and portal hypertension worldwide, schistosomiasis is prob-
most recently transjugular intrahepatic portal–systemic stent ably the most common cause,3 the parasite producing por-
shunt (TIPSS) has relegated emergency surgery to a sec- tal tract fibrosis responsible for the ‘pipe-stem’ appearance
ondary, yet very important, role in most institutions. Con- on macroscopic examination. The importance of this
servative therapy may be unsuccessful, and in this situation presinusoidal block is that liver function is usually nor-
surgical intervention may be the only remaining alternative. mal,4 accounting for a much improved prognosis.5 Postsinus-
Therefore, the challenge is to recognise when non- oidal or posthepatic causes of portal hypertension result
operative treatments are failing or are unlikely to be suc- from hepatic vein thrombosis either as a major vein
cessful so that surgery can be performed while the patient thrombosis (Budd–Chiari syndrome) or small vessel
still has a chance of survival. Following successful manage- veno-occlusive disease.6
ment of an acute variceal bleed, the patient should be con-
sidered for definitive long term treatment. This chapter
will outline the aetiology and pathophysiology of portal
PATHOPHYSIOLOGY
hypertension, the present methods of evaluation and the
various therapeutic approaches available.
Normal portal pressure is 5–10 mmHg with a portal flow in
the order of 1–1.5 L/min. Portal hypertension represents a
AETIOLOGY rise in portal pressure exceeding 12 mmHg. Portal hyper-
tension was previously considered to arise solely as a con-
Causes of portal hypertension can be categorised as pre- sequence of increased resistance to portal venous blood
hepatic, intrahepatic and posthepatic. The intrahepatic flow, the ‘backward flow’ theory. Recent experimental and
556 Portal hypertension and variceal bleeding

clinical studies, however, have shown both increased resist- Table 46.1 Grading of portal hypertension: Child–Pugh
ance to portal blood flow and increased portal blood flow. classification
The characteristic finding in patients with liver cirrhosis is
Number of points
systemic and splanchnic vasodilatation due to a decrease in
vascular tone. Circulating vasodilators, for example nitric 1 2 3
oxide, have been implicated in the aetiology of the vasodi-
latation associated with portal hypertension. The recogni- Bilirubin (mol/L)* 34 34–51 51
tion of vasodilatation and hyperdynamic circulation has led Albumin (g/L) 35 28–35 28
to the ‘forward flow’ theory, which proposes that increased Prothrombin time 3 3–10 10
portal venous inflow plays a central role in the pathogene- prolonged by (s)
Ascites None Slight to Moderate
sis of portal hypertension and this is independent of the
moderate to severe
resistance in the portal venous and collateral circulations. Encephalopathy None Slight to Moderate
Patients with portal hypertension characteristically moderate to severe
develop portal–systemic collateral channels at sites where
the splanchnic and systemic circulations meet, namely the Grade A 5–6 points; grade B 7–9 points; grade C 10–15 points.
gastro-oesophageal junction, the haemorrhoidal junction, * In primary biliary cirrhosis, the point scoring for bilirubin level is
adjusted as follows: 1, 68; 2, 68–170; 3, 170.
the peri-umbilical veins, the retroperitoneal veins of Retzius
and the perihepatic veins of Sappey. The collateral circula-
tion only partially decompresses the portal venous system.
In some cases blood may be shunted through the collateral
anorexia or abdominal pain. The clinical signs of chronic
circulation in sufficient quantities to cause a reversal of flow
liver failure include leuconychia, finger clubbing, palmar
in the portal vein, i.e. hepatofugal or retrograde flow.7
erythema, spider naevi, gynaecomastia (or breast atrophy
Interestingly, it is usually only the lower oesophageal
in females), testicular atrophy and loss of secondary sexual
varices that bleed. Although increased portal pressure is
hair. The existence of portal hypertension is recognised by
essential for the development of varices, the raised hydro-
splenomegaly and abdominal wall venous collaterals, caput
static pressure alone cannot be the only factor responsible
medusa being the classic manifestation.
since there is a poor correlation between the height of por-
The main complications of portal hypertension are
tal pressure and the risk of bleeding.8 The two most import-
variceal bleeding, ascites and hepatic failure. Variceal bleed-
ant risk factors that determine the risk of variceal
ing is the most dramatic of these complications, but hepatic
haemorrhage are the severity of liver disease and the size of
failure is the major determinant of survival.
the varices.9 Studies have shown that large varices are more
Disease severity is assessed by a combination of clinical
likely to rupture at lower transmural pressures than are
parameters and standard laboratory tests. The Child–Pugh
small varices.10 Endoscopic features such as ‘red spots’ and
classification is used to standardise the staging of the dis-
‘wale’ markings have been reported to be important in the
ease. The combination of ascites and encephalopathy,
prediction of variceal haemorrhage. These features represent
bilirubin concentration, albumin concentration and pro-
changes in variceal wall structure and tension associated
thrombin time are evaluated and scored (Table 46.1).2 A
with the development of microtelangiectasias.
total score from the five parameters classifies the patient as
Varices may also occur in the gastric fundus, but they
Child–Pugh grade A (score 5–6), grade B (score 7–9) or
are a source of bleeding in only 5 per cent of cases. Dilatation
grade C (score 10–15).
of the gastric submucosal veins results in portal hyperten-
sive gastropathy which affects mainly the fundus and body
of the stomach and produces an erythematous appearance
at endoscopy. An interesting, but rare, form of localised
INVESTIGATIONS
portal hypertension, known as left-sided, segmental, or
‘sinistral’ portal hypertension, occurs in those with splenic In the acute situation, blood tests for the emergency evalu-
vein thrombosis and should be suspected in patients pre- ation of haematocrit, haemoglobin, coagulation status and
senting with bleeding oesophageal or gastric varices and serum electrolytes are essential. Liver function tests and
normal liver function. serum albumin concentration give an indication of liver
function. Specific serological markers for hepatitis A, B
and C may help define the aetiology,2 while serological
assessment of autoimmune disease and primary biliary cir-
CLINICAL DIAGNOSIS
rhosis should be routine.
Although the level of activity of the underlying liver dis-
Cirrhosis results in two major events, namely hepatocellu- ease may be apparent from standard biochemical tests,
lar failure and portal hypertension. Patients with cirrhosis percutaneous needle biopsy may be used to demonstrate
may have non-specific symptoms such as general malaise, the presence of cirrhosis and to assess its activity.
 Emergency resuscitation and control of bleeding varices 557

Upper gastrointestinal endoscopy is an essential proced- management of these patients is best undertaken in a spe-
ure in the assessment of oesophageal varices. In the patient cialised centre where experienced medical and nursing staff
who is actively bleeding or has just stopped bleeding, it is have the necessary facilities.
important to establish the exact diagnosis and source of
bleeding by endoscopy but only after adequate resuscita-
tion and restoration of haemodynamic stability. Adminis- EMERGENCY RESUSCITATION AND CONTROL
tration of 10–20 mg metoclopramide intravenously prior OF BLEEDING VARICES
to endoscopy may produce a temporary cessation of bleed-
ing permitting blood and clot to be evacuated from the
Resuscitation
stomach and facilitate the visualisation of all potential
bleeding sources. In approximately 10 per cent of all patients
The initial resuscitation of a patient with bleeding
presenting with upper gastrointestinal haemorrhage in the
oesophageal varices is similar to that for any patient with
Western world, oesophageal varices are the cause of bleed-
upper gastrointestinal haemorrhage. A wide bore (16 gauge)
ing. In addition, in patients with known portal hyperten-
cannula is required for the administration of colloidal
sion and oesophageal varices, or in those in whom the
fluids, packed red cells and coagulation factors in the form
presence of portal hypertension is suspected from the his-
of fresh frozen plasma. Saline infusions should be avoided
tory and examination, non-oesophageal variceal sources
because of the risk of fluid retention and aggravation of
will be found in 10–25 per cent.
ascites. Central venous and arterial lines are necessary to
These sources include gastritis, peptic ulceration,
monitor response and to record the blood pressure.
Mallory–Weiss tears, gastric varices or portal hypertensive
Excessive transfusion or fluid administration should also
gastropathy. It is important to recognise these causes of
be avoided as this may increase the risk of further bleeding
bleeding in determining further therapeutic strategies as
by overexpansion of the circulating blood volume. Comatose
some techniques, such as balloon tamponade or scle-
patients and those with massive haematemesis may require
rotherapy, which would be indicated in patients with
endotracheal intubation to prevent aspiration.
oesophageal varices are inappropriate in those with gastric
varices or portal hypertensive gastropathy. When there is
difficulty in identifying the exact cause of bleeding, Pharmacotherapy
endoscopy should be repeated.11 In the patient being evalu-
ated in a non-acute situation, the varices should be graded Several pharmacological agents are available for the emer-
for size, tortuosity, and factors for risk of bleeding. The gency control of variceal bleeding. Most of these act by
detailed definition by some groups12,13 of cherry red spots, lowering portal venous pressure but others are thought to
varices upon varices, and the size of varices as prognostic act by constricting the lower oesophageal sphincter, hence
risk factors of further bleeding has proved useful.2 strangulating the varices and arresting haemorrhage.
Radiological studies are an integral part of evaluation. Vasoactive drugs have been widely used for the past
Ultrasound combined with Doppler study of the vessels, four decades in an attempt to control bleeding from
computed tomography (CT) and angiography may be oesophageal varices. Vasopressin was the first drug used to
indicated.14,15 The morphology of the liver can be assessed induce splanchnic vasoconstriction,16 but it is effective in
with ultrasound and CT. Ultrasound is also useful in demon- only half of the patients treated17 and its use is associated
strating the portal and hepatic veins. When ultrasound is with the development of major complications in approxi-
combined with Doppler flow measurements, directional mately 25 per cent of patients.17,18 Even the use of an
flow may be depicted in these vessels. Arteriography and analogue of vasopressin, terlipressin (triglycyl-lysine-
hepatic venography are used to demonstrate anatomy and vasopressin; Glypressin), or concomitant administration
to measure venous pressures, two mandatory investigations of the hormone with nitrates has failed to produce con-
prior to shunt surgery. If the patient’s history and labora- vincing benefits with respect to both the control of bleed-
tory findings suggest cirrhosis secondary to extrahepatic ing and reduction of side effects.19
biliary obstruction, cholangiography should be performed. The vasoactive peptide hormone somatostatin has a wide
spectrum of actions. Although in pharmacological doses
somatostatin has only a moderate vasoconstrictive effect
on the splanchnic vasculature, it reduces oesophageal
MANAGEMENT variceal pressure and has been shown to be beneficial in the
treatment of bleeding varices. The results of several ran-
The management of bleeding oesophageal varices begins domised controlled trials suggest that variceal bleeding
with emergency resuscitation and initial control of the can be controlled in 65–70 per cent of patients treated,
bleeding episode and continues with long term treatment avoiding major complications but without necessarily sig-
to eradicate the varices and reduce the risk of further nificantly reducing the mortality.20,21 Somatostatin has a
bleeding. In general, both the emergency and long term half-life of less than 3 minutes, necessitating continuous
558 Portal hypertension and variceal bleeding

intravenous infusion to achieve a therapeutic response. been made by endoscopy. If upper gastrointestinal haem-
This short half-life together with its relatively non-specific orrhage from an oesophageal or gastric neoplasm is mis-
effects and the rebound hormonal hypersecretion on cessa- taken for variceal bleeding, an inexperienced clinician may
tion of infusion, have severely limited its clinical use. inflate the gastric balloon within the oesophagus and
The more recently developed somatostatin analogue rupture the latter. With adequate resuscitation and
octreotide, however, has a much more selective action than use of pharmacological agents, particularly octreotide,
somatostatin, is longer acting and is also less expensive. insertion of a balloon tube prior to endoscopy is now rarely
Infusion of octreotide at 25 μg/hour has been shown in necessary. The one exception is possibly the patient with a
one study to be as effective as balloon tamponade of the previous history of varices or a clinical picture strongly
oesophagus in controlling variceal bleeding in the acute suggestive of portal hypertension, presenting with bleeding
situation and has less associated morbidity and mortality.22 and haemodynamic instability which does not respond to
Five randomised controlled trials have found somato- resuscitation and pharmacotherapy.
statin23,24 or octreotide25–27 to be equally effective with fewer The four-lumen (Sengstaken–Blakemore or Minnesota)
associated complications than injection sclerotherapy tube (Fig. 46.1), which is most commonly used, should be
in controlling initial variceal haemorrhage. Furthermore, stored in a refrigerator to render it less pliable, thus facili-
both somatostatin and octreotide have been shown to be tating insertion. Both balloons of a fresh tube are tested for
very effective in controlling early postinjection sclerother- leaks prior to insertion, the nasal route being preferred as it
apy bleeding, which can occur from the varices themselves, is ultimately more comfortable for the patient. Once in the
or from oesophageal ulceration or oesophagitis.28 This stomach, the gastric balloon is filled with 200 mL of air and
observation would suggest that both somatostatin and pulled back to impinge on the cardia. The oesophageal bal-
octreotide are useful as adjuvant treatment to injection loon is then inflated with air to obtain an intraluminal
sclerotherapy. Vasoactive therapy for control of variceal pressure of approximately 40 mmHg, estimated by using a
bleeding remains an attractive proposition since it is the blood pressure manometer. The tube is strapped to the
only treatment that can be initiated, without specialist nares in this position but no traction is necessary. Test
expertise, as soon as the patient enters hospital. Preliminary aspiration via the gastric lumen of the tube is undertaken
studies also suggest that octreotide is very effective in con- approximately every 15 minutes to check for further bleed-
trolling severe gastric bleeding in patients with portal hyper- ing, and this lumen may also be used for administration of
tensive gastropathy or gastric varices.29 Octreotide should neomycin and lactulose. A fourth lumen allows aspiration
be administered intravenously as either 500 μg in 500 mL of the upper oesophagus and pharynx in order to reduce the
of 5 per cent dextrose over 12 hours, or 500 μg in 60 mL of risk of bronchial aspiration. In patients who are stuporous
5 per cent dextrose over 12 hours via a syringe driver. or comatose, the airway should be protected by an endo-
Agents that act by a constrictor effect on the tracheal tube. The use of oesophageal tamponade is a tem-
lower oesophageal sphincter include pentagastrin and porary holding measure until definitive therapy is instituted.
metoclopramide.30,31 The oesophageal balloon is generally deflated after 12–24
The risk of encephalopathy is reduced by the adminis- hours to prevent oesophageal ulceration, but the position
tration of neomycin (250 mg–1 g four times daily), to
reduce the number of ammonia-producing bacteria in the
bowel, and lactulose (10–30 mL three times daily) to Gastric balloon Oesophageal balloon
(air ∼200 mL) (air ∼40 mmHg)
increase gut motility and lower faecal pH, hence reducing
ammonia absorption. Lactulose may also produce some Gastric aspiration/drugs
benefit via an antiendotoxin effect.
Pharyngo-oesophageal
aspiration
Oesophageal tamponade

Balloon tamponade is an effective method of achieving

temporary control of variceal bleeding (70–90 per cent) Oesophageal balloon
but in view of the high rate of recurrent bleeding (60 per
Gastric balloon
cent) after removal of the tube, it should be reserved for
selected situations. These include instances of continued
massive bleeding because varices cannot be sclerosed,
when surgical intervention is required because acute scle-
rotherapy has failed, and prior to transfer of a bleeding
patient to a specialised centre.
In general it is unwise, and even hazardous, to insert a Figure 46.1 Use of a four-lumen (Sengstaken–Blakemore) tube
balloon tube in a patient before a definite diagnosis has for oesophageal tamponade
 Emergency resuscitation and control of bleeding varices 559

of the gastric balloon may be maintained. If bleeding oesophageal damage and general anaesthesia is unnecessary.
recurs the oesophageal balloon may be reinflated but by The main disadvantages are the problem of removing
this stage a more definitive line of therapy should be estab- blood efficiently during active haemorrhage and the diffi-
lished. If re-bleeding does not occur within the subsequent culty in providing adequate compression of the varix after
24 hour period, the gastric balloon is deflated and the tube injection to ensure the necessary intimal damage.
removed, but these patients too will require definitive Complications of injection sclerotherapy are generally
treatment. minor and include low grade fever, retrosternal discomfort
Complications of balloon tamponade occur in about and oesophageal ulceration or mucosal sloughing. More
10–20 per cent of patients32 and include pressure damage serious complications, which are generally rare, include
to the surface of the oesophagus causing mucosal ulcer- oesophageal perforation, oesophageal stricture, pleural effu-
ation, oesophageal perforation and aspiration pneumonia. sion and empyema. With repeated injection sclerotherapy
Some patients find the tube uncomfortable and often sessions, complications become cumulative.42 If oesophageal
become distressed and agitated. ulceration becomes a significant problem either somato-
statin43 or omeprazole44 or both may be used.
In most centres, the initial sclerotherapy is performed at
Injection sclerotherapy the time of diagnostic endoscopy, and succeeds in control-
ling acute variceal bleeding in the majority (over 70 per
Crafoord and Frenckner first described injection sclero- cent) of patients,45 while in the remainder, a second scle-
therapy in 1939 in the management of acute bleeding in a rotherapy session is necessary. Following either one or two
patient with portal vein thrombosis.33 In 1973 Johnston injection sessions, acute bleeding will be controlled in
and Rodgers reported their 15 year experience with 195 90–95 per cent of patients.46,47 Failure of acute sclerother-
administrations of injection sclerotherapy for acute bleed- apy is defined as failure to control acute bleeding after two
ing in 117 patients.34 Control of bleeding was obtained in emergency injection treatments during a single hospital
93 per cent of patients, and the hospital mortality was admission.48,49 Such patients should be controlled with
18 per cent. In the following two decades, endoscopic balloon tamponade and should undergo TIPSS or be con-
injection sclerotherapy became the most frequently used sidered for one of the more major surgical options. These
form of therapy for both emergency and long term control include portal–systemic shunt, oesophageal transection/
of variceal haemorrhage.35 devascularisation or liver transplantation.
The technique relies on the obliteration of varices by a During a 37-year period in Belfast, injection sclerother-
process of sclerosis and fibrosis. Several methods are used apy using the rigid instrument was used for the control of
to achieve this objective, including intravariceal injection acute variceal bleeding. Between 1958 and 1995, 436 patients
of sclerosant, which is the most widely used method,1,36 had a total of 734 injections during 663 admissions
and paravariceal or submucosal injection,37 which aims to for acute bleeding. Bleeding was controlled in 94 per cent
produce an area of fibrosis and thickening superficial to of patients, and only 6 per cent of patients required more
the varices. Some sclerotherapists use a combination of the than one injection for control. The hospital mortality rate
intravariceal and paravariceal methods, in an attempt to was 15 per cent and, of these, 63 per cent were categorised
combine the advantages of each.38 Various sclerosing agents as Child’s grade C.
are used, including ethanolamine oleate, sodium tetradecyl The results for sclerotherapy in the treatment of acutely
sulphate, polidocanol, and more recently the tissue bleeding gastric varices have generally been poor, although
adhesive N-butyl-2-cyanoacrylate (Histoacryl).28,39–41 The reports on the use of Histoacryl glue have been very
fact that several agents are available, and are effective, prob- encouraging.50–52 Injection sclerotherapy is not indicated
ably indicates that no one in particular is vastly superior in patients with bleeding due to portal hypertensive gas-
to the others. tropathy. In these cases TIPSS or an emergency surgical
Injection sclerotherapy can be performed using either a procedure is indicated if pharmacotherapy fails.
rigid oesophagoscope or a fibreoptic endoscope. Both have
advantages and disadvantages, although the latter is more
commonly used nowadays. If one uses the rigid instru- Variceal ligation
ment, a general anaesthetic is required, but this gives the
reassurance of a protected airway. The technique for the Endoscopic variceal ligation (EVL),53 introduced in 1988,
passage of the rigid instrument requires more training and is based on the same principle as banding of haemorrhoids.
skill. The advantages are that the tip of the rigid instrument The target varix is identified and the scope advanced under
can be used to compress any bleeding point and the wide direct vision until the banding cylinder is in full contact
bore channel allows the passage of a large sucker for removal with the varix. The varix is sucked into a chamber attached
of blood or clot. Rotation of the instrument allows compres- to the tip of the endoscope resulting in a complete ‘red-
sion of the varix following injection of sclerosant. The flex- out’ and loss of endoscopic visibility. An elastic ‘O’ band,
ible instrument is technically easier to pass, carries less risk of mounted on the chamber, is then released over the varix by
560 Portal hypertension and variceal bleeding

pulling on a trip wire which runs through the working If surgery is required in the acute situation, the quickest
channel of the endoscope. The engorged varix is strangu- and simplest procedure is probably the safest. Many sur-
lated at the mucosal junction. Treatment involves ligation geons therefore favour the end-to-side portacaval shunt or
of the most distal variceal columns, commencing with the the interposition mesocaval shunt.62 Both these procedures,
bleeding varix if present. Subsequent ligations are per- however, completely divert portal flow away from the liver
formed at increasingly higher levels, proceeding in a spiral and thereby tend to cause more frequent episodes of
fashion to avoid circumferential placement of bands at the encephalopathy than do those following devascularisation
same level. A mechanism which allows repeated banding and procedures and selective shunts.63 The high incidence of
obviates the need to use an overtube for repeated removal postoperative encephalopathy, which is often chronic and
of the endoscope has been developed recently. The most incapacitating, accounts for the trend away from portal–
impressive advantage of EVL is its simplicity. In contrast to systemic shunts in recent years despite the fact that they pre-
endoscopic sclerotherapy, which requires intensive oper- vent recurrent variceal bleeding in virtually all patients.
ator training in order to obtain good results, EVL is rela- Clear indications for emergency surgical intervention in
tively easy to learn and perform competently.54 many institutions include persistent variceal bleeding
The technique has proved successful not only in the despite non-operative treatment with endoscopic tech-
treatment of acute variceal bleeding, but also in long term niques, failure of chronic sclerotherapy or EVL, and bleed-
management aimed at eradicating varices, as confirmed in ing from gastric varices or portal hypertensive gastropathy
a prospective randomised controlled clinical trial compar- which is unresponsive to acute pharmacotherapy. In our
ing this technique with sclerotherapy.55 The technique has practice, we virtually never use emergency shunt proced-
also been successfully used for patients who failed to respond ures for acute variceal bleeding.
to injection sclerotherapy.56 Reports of complications of
EVL include bleeding or perforation caused by the over-
tube,57,58 bleeding from a band induced ulcer and acute Transjugular intrahepatic portal–systemic
oesophageal obstruction caused by banded oesophageal stent shunting
varices.59
A method for producing an intrahepatic portal–systemic
stent shunt via the transjugular route was first described in
Portal–systemic shunt procedures animals by Rosch et al. in 1969.64 The first transjugular
portal–systemic shunt procedure in a patient was described
Until the use of sclerotherapy became widespread in the in 1982 when balloon catheters were inflated to produce a
early 1970s portal–systemic shunts were widely used for the tract between the portal and hepatic vein.65 The success of
treatment of acute variceal bleeding. Shunt surgery has, how- metallic expandable stents to improve patency rates since
ever, been associated with high operative mortality of 7–30 1989 has resulted in an increasing use of this technique.66
per cent and a low 5-year survival, particularly in patients The internal jugular vein, usually on the right side, is
with marked hepatic functional decompensation. For these punctured and a sheath is inserted cannulating the right or
reasons portal–systemic shunting for acute variceal bleeding middle hepatic vein under radiological screening. Then a
is not performed routinely in the vast majority of centres and transjugular liver biopsy needle or sharp stylet is directed
is reserved for the 5–10 per cent of patients in whom bleed- out of the hepatic vein, through the liver parenchyma and
ing is not controlled by pharmacotherapy or endoscopic into a large branch of the portal vein. After passing a guide
techniques. It also remains a useful treatment in countries wire into the portal vein, an angioplasty balloon catheter is
with a high incidence of non-cirrhotic portal hypertension used to expand the tract before inserting a metallic stent
that occurs during the second or third decade of life. (Palmaz or Wallstent) (Fig. 46.2).
Orloff, the main proponent of emergency shunt proced- At present the main indication for this technique is to
ures, reported an operative mortality rate of 20 per cent and control and prevent variceal haemorrhage when endo-
a 5-year survival rate of 64 per cent in a series of 94 consecu- scopic therapy has failed.67–69 It has been shown in numer-
tive unselected patients who underwent surgery within ous uncontrolled studies that TIPSS reduces re-bleeding in
8 hours of admission to hospital.60 A controlled trial by patients with recurrent variceal haemorrhage.70,71 A recent
Orloff’s group comparing routine emergency shunting with meta-analysis72 evaluated 11 randomised trials comparing
conventional medical management followed by elective TIPSS with endoscopic treatment for prevention of
shunting showed better control of haemorrhage and variceal re-bleeding and survival. Transjugular intrahep-
improved early and late survival in the emergency shunt atic portal– systemic stent shunting was compared with
group.61 Major criticisms of this trial are, first, that endo- sclerotherapy alone in five studies, with sclerotherapy plus
scopic sclerotherapy, the gold standard for treatment of acute propranolol in three studies and with EVL alone in three
variceal bleeding in most centres, was not used in the medical studies. Overall, TIPSS significantly reduced the re-bleeding
treatment arm and, second, crossover to surgical treatment rate but did not improve overall survival rate and was associ-
for patients who failed medical therapy was not allowed. ated with a significantly higher incidence of post-treatment
 Emergency resuscitation and control of bleeding varices 561

Biopsy needle first reported the use of a circular stapling gun to perform
Hepatic or stylet
Metallic stent oesophageal transection in 1974.80 It is a relatively simple
vein operation, compared with portal–systemic shunting, but
carries a similar mortality. As the portal circulation is not
disturbed, postoperative encephalopathy is generally not a
problem. The disadvantage of this procedure, however,
is that, although acute bleeding may be controlled in over
Portal Balloon 95 per cent of patients, there is, in contrast to shunt proced-
vein catheter
ures, a significant rate of recurrent variceal bleeding,
Figure 46.2 Technique to create an intrahepatic portal– approximating 30 per cent over the next 5–10 years.
systemic stent shunt Oesophageal transection (Fig. 46.3) is performed through
an upper midline incision. The left gastric venous drainage
is ligated in continuity at the upper border of the pancreas
encephalopathy compared with endoscopic treatment. in order to interrupt the main connection between the
Although some authors consider TIPSS to be the preferred hypertensive portal vein and the varices. The oesophagus is
first line treatment for uncontrolled variceal haemor- then mobilised, protecting the anterior and posterior vagi.
rhage,73,74 most support the view that TIPSS should be A linen ligature is passed around the oesophagus prior to
a rescue procedure for failures of medical/endoscopic insertion of the gun. The closed gun is introduced into the
treatments.75,76 lower third of the oesophagus via a small anterior gastro-
One advantage of TIPSS is that, being an intrahepatic tomy wound, and then slackened off to give a gap of about
procedure, the extrahepatic venous anatomy is not dis- 3 cm. The linen ligature is then tightened around the
turbed and subsequent liver transplantation can, if required, oesophagus immediately above the cardia thus invaginat-
be more easily performed.74,77 Transjugular intrahepatic ing a flange of full thickness wall between the two portions
portal–systemic stent shunting can also be employed before of the head of the gun. The gap in the head of the gun is
closure of a surgical portal–systemic shunt which is caus- tightened and the trigger pulled to complete the transec-
ing disabling encephalopathy, and it can also be used to tion (see Fig. 46.3a). It is necessary to reopen the gap on the
relieve intractable ascites. head of the gun before withdrawing it through the newly
Transjugular intrahepatic portal–systemic stent shunt- formed anastomosis. The excised portion of oesophageal
ing is successfully completed in more than 90 per cent of wall should form a complete ‘doughnut’, indicating a satis-
patients, with a procedure related mortality of less than factory transection.
1 per cent caused mainly by intraperitoneal bleeding.67 The The incidence of recurrent bleeding may be decreased
major side effect of a successfully placed TIPSS is new or by a combination of oesophageal transection and devascu-
worse encephalopathy, which develops in approximately larisation of the lower oesophagus and upper stomach (see
15–25 per cent of cases.78 In most cases the symptoms are Fig. 46.3b), but this may increase the operative risk. An
mild and are readily treated with dietary protein restriction even more extensive operation involves a thoracoabdomi-
and lactulose therapy.74 Re-bleeding occurs in 10–20 per nal approach, with oesophageal transection and extensive
cent of patients,67 and is usually associated with shunt devascularisation of the oesophagus and upper stomach,
related complications, such as intimal hyperplasia, shunt including a vagotomy, pyloroplasty and splenectomy
migration or thrombosis. Therefore, a vigorous surveil- (Sugiura operation) (see Fig. 46.3c).81 This combined pro-
lance programme is mandatory for the prompt diagnosis cedure, which was used extensively in Japan with excellent
and correction of any significant stent dysfunction and to reported results in terms of operative complications and
minimise the risk of re-bleeding. Doppler ultrasonography re-bleeding,82 is now less popular there and has never
has been used for evaluation of shunt patency, but its sen- become established in Western practice.
sitivity is inferior to angiography.79 In cases of shunt occlu-
sion, either a further shunt may be inserted or balloon
Control of active variceal bleeding
dilatation of the existing shunt can be carried out.
• Resuscitation
– Insert two 16 gauge peripheral cannulae
Oesophageal transection/devascularisation – Correct coagulation abnormalities
– Central venous access
The alternative to portal–systemic shunting for surgical
– Protect airway by elective intubation if patient
control of acutely bleeding oesophageal varices is transec-
has severe uncontrollable variceal bleeding,
tion of the oesophagus and end-to-end anastomosis using
severe encephalopathy, inability to maintain
a stapling gun.49 The aim of this procedure is to remove a
oxygen saturations above 90 per cent, or
segment of the lower 1–3 cm of the oesophagus, which is the
aspiration pneumonia
most vulnerable in terms of variceal bleeding. Vankemmel
562 Portal hypertension and variceal bleeding

(a) (b)

Figure 46.3 Oesophageal

transection/devascularisation
procedures: (a) oesophageal transection,
(b) oesophageal transection and
devascularisation and (c) Sugiura
(c) operation

LONG TERM AND DEFINITIVE TREATMENT

• Control of bleeding
– Proceed to upper gastrointestinal endoscopy as
soon as patient is haemodynamically stable During their lifetime some 70 per cent of patients will have
– Variceal band ligation is the method of first choice a further variceal bleed, the highest incidence and associ-
– If banding is difficult because of continuing ated mortality rate being during the first few months
bleeding, or this technique is not available, following the first haemorrhage, and the chance of sur-
endoscopic injection sclerotherapy should be vival returning to baseline thereafter.83 There is general
performed agreement that some form of specific long term treatment
– If endoscopy is unavailable, vasoconstrictors such as is usually indicated.
a octreotide or terlispressin should be administered, The only form of treatment that cures both the underlying
or a Sengstaken–Blakemore tube should be inserted liver disease and the portal hypertension is liver transplanta-
• Failure to control active bleeding tion. Patients presenting with bleeding oesophageal varices
– If bleeding is difficult to control, a nowadays should all be considered for liver transplanta-
Sengstaken–Blakemore tube should be inserted tion.84 Only a small percentage will ultimately receive a trans-
until further endoscopic treatment, TIPSS, or plant, but management in possible transplant candidates
surgical intervention should subsequently be directed to forms of therapy that will
– Specialist help should be sought and patient not interfere with any subsequent transplant procedure.
transferred to a specialist centre The alternatives to liver transplantation are long term
– Choice of TIPSS or other emergency surgical pharmacological management, repeated injection sclerother-
procedure should be decided depending on local apy, repeated variceal ligation, a devascularisation and tran-
expertise section operation, or a portal–systemic shunt procedure.

Over a 19-year period, a total of 139 stapled oesophageal Long term pharmacotherapy
transections with devascularisation were performed in
Belfast. Twenty-nine procedures were performed as an Although drugs have been used in the treatment of acute
emergency, with eight deaths, giving a mortality rate of variceal bleeding for many years, pharmacotherapy has only
28 per cent; this compares with a mortality rate of 14 per cent recently been investigated and used for the long term preven-
for the 110 patients who underwent elective procedures. tion of recurrent variceal haemorrhage. The mechanism of
 Long term and definitive treatment 563

action of long term pharmacotherapy is probably a reduc- earlier obliteration of varices but may be associated with a
tion in portal pressure brought about by splanchnic arteri- higher incidence of oesophageal mucosal sloughing and
olar vasoconstriction. ulceration than injections at 2- or 3-weekly intervals. Once
Nine published randomised controlled trials have com- the varices have been eradicated, follow-up endoscopy is
pared -blockers (propranolol or nadolol) to a placebo for advised every 3–6 months.
the prevention of recurrent variceal haemorrhage. A meta- The efficacy of long term sclerotherapy in eradicating
analysis of these trials shows a significant reduction in varices, to prevent recurrent bleeding and ultimately improve
re-bleeding but not in mortality.85 A number of recent stud- survival, has been the subject of several trials. Meta-analysis
ies have combined administration of isosorbide-5-mononi- of eight randomised controlled trials comparing sclerother-
trate (ISMN) with non-selective -blockade in preventing apy with a medically treated control group for the preven-
variceal rebleeding. Patch et al. randomised 102 patients sur- tion of recurrent variceal bleeding has shown that the
viving a variceal bleed to EVL or drug therapy using propra- rebleeding rate is lower in patients who underwent injection
nolol with the addition of ISMN if target reductions in portal sclerotherapy.85 However, 45–60 per cent of patients treated
pressure evaluated by the hepatic venous pressure gradient by injection sclerotherapy will re-bleed some time later.89
(HVPG), were not achieved at 3 months.86 Overall, results of Many of these patients are successfully managed with subse-
drug therapy were similar to those of EVL with no differences quent sclerotherapy sessions, but the eventual failure rate of
in survival or non-bleeding complications. Recent data sug- injection sclerotherapy is as high as 35 per cent.90,91
gest that reducing HVPG by 20 per cent or more, or reducing Failure of long term sclerotherapy is defined as massive
portal pressure to 12 mmHg is associated with a marked bleeding requiring resuscitation and blood transfusion, or
reduction in the long term risk of developing complications two or more bleeding episodes during the injection pro-
of portal hypertension and is associated with improved sur- gramme. In this situation a shunt procedure or possibly
vival.87 ‘A la carte’ treatment of portal hypertension is being a devascularisation and transection operation should be
proposed by several authors, suggesting that adaption of considered.47,92,93 In terms of improvement of overall sur-
medical therapy to the haemodynamic response is important vival it is generally perceived that sclerotherapy, while effect-
in the management strategy of patients with portal hyper- ive in treating acute variceal bleeding and reducing the risk
tension. Burroughs and McCormick treated 34 patients suffer- of recurrent bleeding, does not improve long term survival.
ing from cirrhosis and portal hypertension with propranolol This is thought to be related to the fact that it is only a pal-
and measured HVPG after a median of 4 days.88 Target HVPG liative procedure and does not alter the progression of the
reductions were achieved in 13 ‘responders’. Isosorbide-5- underlying disease process. However, a meta-analysis of the
mononitrate was added in the 21 ‘non-responders’ and controlled trails comparing chronic sclerotherapy with con-
HVPG measured again. Seven further patients achieved tar- ventional medical therapy shows a statistically significant
get HVPG reduction. Bleeding rates were lower in respon- survival advantage with the former technique.94
ders than in non-responders. The authors concluded that the
haemodynamic response to drug therapy identified patients
who were efficiently protected from variceal re-bleeding and Repeated variceal ligation
suggested that alternative treatment should be considered for
non-responders. Repeated EVL is an effective procedure for the eradication
Other agents currently being investigated for use in long of oesophageal varices. Seven published trials comparing
term pharmacotherapy include calcium channel blockers EVL with endoscopic injection sclerotherapy have been
such as verapamil, serotonin antagonists and octreotide. combined in a meta-analysis which included 547 patients.95
Drug therapy, however, has disadvantages, such as the cost This concluded that EVL carried a significantly lower rate of
of lifelong medication, problems with compliance in alco- recurrent variceal bleeding, a decreased mortality rate and
holic patients and potential side effects. reduced complications such as oesophageal stricture, while
fewer sessions were required to achieve variceal obliter-
ation. Endoscopic variceal ligation is now considered in
Repeated injection sclerotherapy many centres to be the endoscopic treatment of choice for
patients with bleeding oesophageal varices.95,96
Repeated injection sclerotherapy is currently the most widely It has been the experience of some, that the frequency of
used technique for the treatment of recurrent oesophageal recurrence of oesophageal varices is higher following EVL
variceal bleeding. If a programme of repeated sclerother- and therefore a combination of EVL and injection scle-
apy is thought to be the most appropriate long term thera- rotherapy may prove to be a useful option – using EVL to
peutic option, repeat injection sessions usually begin 1–2 treat large varices followed by a course of injection scle-
weeks after the initial emergency session and are continued rotherapy for small varices and to achieve fibrosis of the
every 1–2 weeks until the varices have been completely inner wall of the oesophagus.97 The additive effect of
eradicated. The interval between injection sessions varies sclerotherapy to patients receiving repeated EVL has been
in different centres; injections at weekly intervals produce shown in a prospective randomised trial to be safe and results
564 Portal hypertension and variceal bleeding

(a) (b) (c)

Figure 46.4 Non-selective portal-systemic

shunts: (a) end-to-side portacaval, (b) side-to-
side portacaval, (c) mesocaval prosthesis,
(d) portacaval prosthetic H graft and (e) central
(d) (e) splenorenal

in a significant reduction in the re-bleeding rate.98 Other grade C disease. Fifteen patients in the series required
authors, however, have not shown any significant benefit oesophageal dilatation because of stricture formation. Of
of combination therapy and have reported that it is associ- the 117 patients who survived to leave hospital, 46 developed
ated with a greater morbidity.99–101 recurrent haemorrhage during a follow-up period ranging
from 3 months to 19 years. In 30 patients recurrent varices
were identified and treated with post-transection injection
sclerotherapy. Often recurrent haemorrhage was of a minor
Elective oesophageal transection/ nature, and only seven of those 46 patients died as a result
devascularisation of the bleeding. The overall 5-, 10- and 15-year cumulative
survival rates for the whole series were 49 per cent, 32 per
Transection/devascularisation procedures are being used cent and 28 per cent, respectively.
less frequently for the elective or long term management of
patients with variceal bleeding. This is probably related to
the fact that, although they have the advantages of being
relatively simple to perform and not associated with Elective portal–systemic shunts
encephalopathy, they are associated with a high incidence
of recurrent variceal bleeding (approximately 30 per cent). Although many types of portal–systemic shunt have been
They may continue to have a limited role in some coun- developed, they fall into two main groups, namely non-
tries, where access to liver transplantation is not available selective and selective. Non-selective shunts decompress
and where the patient is unsuitable for a shunt procedure, the entire portal system and as a result the incidence of
namely, elderly patients, patients with advanced liver retrograde portal blood flow and the risk of postoperative
disease, diabetic patients, patients with schistosomiasis and encephalopathy are high. The diversion of portal flow away
those with previous encephalopathy. from the liver and the consequent loss of hepatotrophic
Of the 139 stapled oesophageal transections with devas- factors, is thought to cause a deterioration of hepatocellu-
cularisation in Belfast carried out between 1976 and 1995, lar metabolism and function. Examples of these non-selec-
110 were elective operations, often performed during the tive shunts include the standard end-to-side portacaval
same hospital admission but usually within a few weeks of shunt, the side-to-side portacaval shunt, the mesocaval
the onset of bleeding. There were 23 operative deaths in the shunt using a prosthetic graft, the portacaval prosthetic H
total series, 10 of these occurring in patients with Child’s graft and the central splenorenal shunt (Fig. 46.4).
 Long term and definitive treatment 565

Figure 46.5 Selective portal-systemic

shunts: (a) Warren distal splenorenal and
(a) (b) (b) Inokuchi left gastric–inferior venacaval

One of the non-selective shunts, the portacaval pros- Liver transplantation

thetic H graft, is worth further consideration. It has been
suggested that by using a small diameter prosthesis, approxi- As the results of liver transplantation have improved,
mately 10 mm in diameter, the resistance of the shunt the operation has become more widely available and
could be increased, thus decreasing the flux across the represents another treatment option for some patients with
shunt and maintaining prograde or hepatopedal portal recurrent variceal bleeding.84 In practice, economic factors
flow. The effect of the ‘partial’ shunt is not only to decom- and a restricted supply of donor organs limit its availability,
presses the portal system sufficiently to prevent recurrent and in any case a significant percentage of patients are not
variceal bleeding, but also to maintain prograde portal flow candidates for this procedure because of advanced age, active
and prevent encephalopathy and deterioration in hepato- alcohol or drug abuse, or advanced disease in other organ
cellular function;102 the results have been encouraging in a systems. Similarly, patients with extrahepatic portal vein
few centres.103–105 Other advantages are that it is a relatively thrombosis and schistosomiasis are not transplant candi-
easy shunt to construct and, as dissection in the area of the dates because hepatic functional reserve is maintained indef-
hepatic pedicle is minimal, subsequent orthoptic liver initely in these conditions. Transplantation, however, is the
transplantation is not compromised.104 treatment of choice in patients who have endstage liver dis-
Selective shunts aim to decompress only a part of the ease. Bismuth et al. have reported a 4-year survival rate of 73
portal system, namely the oesophago-gastric venous net- per cent in Child’s grade C patients undergoing liver trans-
work, to prevent variceal bleeding while still maintaining plantation,110 while survival rates among patients with com-
portal venous pressure and prograde portal flow to avert parable hepatic risk managed by primary shunt or
encephalopathy.106 Examples include the Warren distal sclerotherapy were 31 per cent and 59 per cent, respectively.
splenorenal shunt and the Inokuchi left gastric–inferior This evidence, confirmed by other investigators,84,111 sug-
vena caval shunt (Fig. 46.5). The Warren shunt is the most gests that, in the absence of any contraindications, patients
widely used and is reported to be associated with a much with end-stage liver disease should undergo liver transplanta-
lower incidence of encephalopathy than that following tion. It may also be required as the treatment of choice in
non-selective shunts.107 Nevertheless, not all patients patients with a poor quality of life and suffering from
maintain prograde portal perfusion after construction of encephalopathy, asthenia and fatigue but whose disease is
the shunt, and this is particularly true of alcoholic patients not end stage. Those patients with good hepatic function and
with cirrhosis. In addition, it is thought that, in a sig- static or slowly progressing disease are best treated in the
nificant number of patients with this shunt, prograde interim by methods other than transplantation. If future
portal perfusion may not be maintained in the long transplantation is anticipated, the technique of surgical treat-
term.108 Shunting continues to have a small but definite ment should be considered carefully: non-shunt operations
role in the management of young patients with minimal do not adversely affect subsequent transplantation and, if
hepatic impairment and who do not have a history of a shunt is required, the distal splenorenal, mesocaval or
diabetes, encephalopathy, active hepatitis or acute portacaval prosthetic H shunts, with which dissection in
bleeding.109 It may also have a place in the treatment of the area of the hepatic pedicle is minimal, are most suitable.
patients with portal hypertensive gastropathy, in whom End-to-side and side-to-side portacaval shunts should be
reduction of portal pressure may be the only successful avoided, but these techniques have been largely superseded
form of treatment. by TIPSS.
566 Portal hypertension and variceal bleeding

Initial studies of prophylactic shunt procedures during

Secondary prophylaxis of variceal bleeding the 1960s showed increased morbidity and mortality in the
in patients with cirrhosis shunt groups114–116 and they were therefore abandoned. A
more recent multicentre study from Japan, comparing
• Endoscopic variceal band ligation
prophylactic portal non-decompressive surgery with con-
– Following control of active variceal bleeding, the
ventional management, has shown significantly improved
varices should be eradicated using endoscopic
survival and reduced re-bleed rates in the shunt group at
methods. The method of first choice is band ligation
5 years.117 This study remains to be repeated in Western
– It is recommended that each varix is banded with
countries, where the aetiology of portal hypertension
a single band at weekly intervals until completely
largely differs from that in Japan, and where major
eradicated
prophylactic shunt surgery of this nature is currently
– The use of an over-tube should be avoided as it is
unjustified outside controlled trials.
associated with increased complications
Although initial controlled trials118,119 reported that pro-
– Following successful eradication of varices,
phylactic sclerotherapy reduced the risk of first variceal haem-
patients should undergo endoscopy at
orrhage and improved survival, these results were
3- to 6-monthly intervals
not borne out by a meta-analysis of 19 randomised trials
– Recurrent varices should be treated with further
assessing the effect of prophylactic sclerotherapy for
variceal eradication
oesophageal variceal haemorrhage.120 This meta-analysis
• Endoscopic injection sclerotherapy
showed a significant benefit for sclerotherapy in the preven-
– If banding is not available, injection sclerotherapy
tion of the initial variceal haemorrhage, but as there was sig-
should be used
nificant heterogeneity in the results, the authors concluded
– The sclerotherapy agent used may vary between
that the overall results were equivocal. The authors of another
institutions
meta-analysis of randomised controlled trials concluded that
• Non-selective -blocker with or without endoscopic
prophylactic sclerotherapy should not be widely applied at
therapy
present owing to insufficient documentation of benefit.121
– Non-selective -blockade may be used alone or
There is even a danger that injection treatment might precip-
in combination with endoscopic techniques
itate the first variceal bleed, which may prove fatal.122
• TIPSS
Endoscopic variceal ligation, which has an inherently
– TIPSS is more effective than endoscopic
low complication rate and brings about rapid obliteration
treatment in reducing variceal re-bleeding but
of varices, may be a better option for primary prophylaxis
does not improve survival and is associated with
of variceal haemorrhage. At least two groups have reported
more encephalopathy
a significantly reduced initial variceal bleeding rate and
• Portal–systemic shunt formation
mortality after prophylactic EVL compared with no treat-
– Continues to have a small but definite role in the
ment.123,124 The results of these two trials which have
management of young patients with minimal
demonstrated efficacy of EVL for primary prophylaxis do
hepatic impairment who do not have a history of
not yet justify a conclusion that this treatment should be
diabetes, encephalopathy, active hepatitis or
recommended for patients with varices that have not bled
acute bleeding
and further study is needed.
The most widely used form of prophylaxis is pharma-
cotherapy, particularly -blockade with propranolol. There
have been nine controlled studies on the prevention of ini-
tial variceal haemorrhage in patients with cirrhosis and
PROPHYLACTIC THERAPY
oesophageal varices. Propranolol was the active agent used
in seven studies and nadolol in two. Three meta-analy-
The aim of prophylactic therapy is to prevent a first ses,120,125,126 one based on individual data from four ran-
episode of variceal bleeding and thereby reduce subsequent domised controlled trials,126 have been published. All three
complications and mortality. As bleeding from docu- concluded that the risk of initial haemorrhage was signifi-
mented oesophageal varices will occur in only 30 per cent cantly reduced and there was a trend towards improved
of patients with cirrhosis,112,113 many authors recommend survival in cirrhotic patients with oesophageal varices.
observation until the first variceal bleed occurs. Neverthe- A recent cost-effectiveness analysis of variceal bleeding
less, the high mortality of over 50 per cent in poor-risk prophylaxis with propranolol, sclerotherapy, and shunt
Child’s C patients associated with the first bleed is reflected surgery in cirrhotic patients concluded that, at present,
in controlled trials of prophylactic treatments including propranolol is the only cost-effective modality for prophy-
pharmacotherapy, injection sclerotherapy, portal–systemic laxis of initial variceal bleeding.127 Most workers in the field
shunt procedures and extensive devascularisation and agree that patients with large varices who have not bled and
transection operations. presumably carry a high risk of haemorrhage should be
 Results and prognosis 567

offered prophylactic -blocker therapy. There are as yet cirrhosis with good liver function have good short and
insufficient data to make recommendations for small varices. long term prognosis. In patients with established cirrhosis,
Isosorbide mononitrate has been shown to be as effect- the presence of alcoholic hepatitis, hepatocellular carcin-
ive as propranolol in the prevention of first variceal bleed- oma and/or portal venous thrombosis may adversely affect
ing with no difference in survival.128 Merkel et al. have also prognosis. Operative mortality rates are also more depend-
reported that a combination of ISMN with nadolol was of ent on the status of the liver disease at the time of surgery,
greater benefit than nadolol alone;129 however, Garcia- rather than the procedure performed.
Pagan et al. reported that the addition of ISMN did not
further lower the risk of bleeding in patients already receiv-
ing propranolol.130 Long acting nitrates may be a reason-
able alternative for patients who have contraindications to,
Conclusions
or are unable to tolerate -blockers, or they could be
prescribed as an additional therapy for primary prophylaxis
The initial management of a patient with bleeding
in high risk patients.
oesophageal varices is similar to that for any patient
with upper gastrointestinal haemorrhage, and subse-
quent treatment follows a protocol with generally agreed
indications for each therapeutic option (Fig. 46.6).
Primary prophylaxis Immediate resuscitation involves administration of
fluids and blood, preferably guided by measurement of
• Methods central venous pressure, and correction of coagulation
– Pharmacological therapy with propranolol is the deficits. Emergency endoscopy is necessary to confirm
best available modality that the patient is bleeding from oesophageal varices.
– If propranolol is contraindicated or there is Pharmacotherapy, either with somatostatin or octreotide,
intolerance, variceal band ligation is the may be used if necessary. Oesophageal tamponade may
treatment of choice be necessary on rare occasions: in a patient who has
– If neither propranolol nor variceal band ligation massive variceal bleeding which disallows acute scle-
can be used, ISMN is the treatment of first choice rotherapy, when sclerotherapy fails in a patient waiting
• Diagnosis for surgery and in transporting an acutely bleeding
– All patients with cirrhosis should undergo patient to a specialised centre.
endoscopy at the time of diagnosis Once variceal bleeding has been confirmed, immedi-
– If at the time of first endoscopy no varices are ate injection sclerotherapy or band ligation should be
observed, the patient with cirrhosis should performed. Using one or sometimes two endoscopy ses-
undergo endoscopy at 3-yearly intervals sions, acute variceal bleeding will be controlled in 90–95
– If small varices are diagnosed, the patient should per cent of patients. Acute endoscopy techniques are
undergo endoscopy at yearly intervals deemed to have failed when two sessions have been unable
– If patients have grade 2 varices and Child’s B or C to control the acute bleeding episode.
disease, they should have primary prophylaxis In the 5–10 per cent of patients in whom emergency
– If grade 3 varices are diagnosed, patients should endoscopic techniques fail, an emergency surgical or
have primary prophylaxis irrespective of severity radiological procedure is indicated. The available
of liver disease options are a portal–systemic shunt (including TIPSS)
and oesophageal transection, with or without devascu-
larisation. The indications for each of these options and
their relative merits remain controversial. In general,
It is evident that further studies are required and only
emergency portal–systemic shunting produces a definite
time will tell whether pharmacotherapy, variceal ligation,
arrest of haemorrhage and prevents recurrence but
or perhaps a combination of these therapies, may be the
is associated with a high incidence of encephalopathy
most appropriate approach for prophylaxis of oesophageal
and a high operative mortality. If an emergency portal–
variceal haemorrhage.
systemic shunt is contemplated, TIPSS is now the proced-
ure of choice. Alternatively, an oesophageal transection
can be performed. This is associated with a lower inci-
RESULTS AND PROGNOSIS dence of encephalopathy but a higher incidence of
recurrent bleeding.
In general, therapeutic options for long term control of
The prognosis in an individual patient depends on the
variceal bleeding are palliative procedures designed to
severity of the bleeding episode and underlying liver func-
prevent recurrent bleeding and do not improve overall
tion.122 Patients with non-cirrhotic portal hypertension or
568 Portal hypertension and variceal bleeding

Acute upper gastrointestinal bleeding

Pharmacotherapy Resuscitation

Balloon tamponade
Urgent endoscopy
(Bleeding oesophageal varices confirmed)

Sclerotherapy/endoscopic variceal ligation (EVL)

Bleeding stops Bleeding continues Pharmacotherapy

Balloon tamponade
Long term Repeat endoscopy and
treatment options sclerotherapy/EVL
Bleeding continues

Bleeding stops
Emergency portal–systemic Emergency oesophageal
shunt (TIPSS) transection/devascularisation

Figure 46.6 Algorithm for acute management of a patient with bleeding oesophageal varices. TIPSS, transjugular intrahepatic
portal-systemic stent shunting

Control of acute variceal bleeding

Long term treatment decision

Transplant candidate Not transplant candidate

(Child–Pugh C or B with (Child–Pugh A or B with
symptomatic disease) asymptomatic disease)

Sclerotherapy, EVL Sclerotherapy Portal–systemic Oesophageal

Figure 46.7 Algorithm of or TIPSS in interim shunt (TIPSS) transection /
devascularisation
long term management of a
patient with bleeding 1/3 fail
Liver transplantation
oesophageal varices. EVL,
endoscopic variceal ligation;
TIPSS, transjugular intrahepatic Liver transplantation Shunt (TIPSS) Oesophageal transection /
devascularisation
portal–systemic stent shunting

survival. The only treatment which addresses the

Only -blockers have been clearly shown to be of ben-
underlying pathology and improves survival is liver
efit in the prevention of initial variceal bleeding. All
transplantation. When considering the appropriate thera-
patients with established cirrhosis should undergo screen-
peutic option for each patient, it is useful to first consider
ing endoscopy to assess the presence of varices. Although
whether the patient is a candidate for transplantation (Fig.
data concerning the best time intervals for repeat screen-
46.7). Patients with advanced liver disease are best treated
ing sessions are lacking, several established liver centres
by transplantation. Patients with moderate liver dysfunc-
have adopted a policy of screening patients every 2–3
tion but who are symptomatic, for example those with
years. Patients with large varices and/or endoscopic signs
encephalopathy, asthenia, fatigue or a poor quality of life,
for risk of bleeding should be placed on treatment.
are also probably best treated by transplantation. Those
Whether treating patients with small oesophageal varices
awaiting a suitable liver graft should undergo sclerother-
is cost-effective or clearly influences survival is unknown.
apy, EVL or TIPSS in the interim period.
 References 569

15 Oliver TW, Sones PJ. Hepatic angiography: Portal hypertension.

Key references In: Bernardino ME, Sones PJ (eds). Hepatic Radiology. New
York: Macmillan Publishing Co., 1984: 243–75.
D’Amico G, Pagliaro L, Bosch J. The treatment of portal 16 Kehne JH, Hughes FA, Gompertz LA. The use of surgical pituitrin
hypertension: a meta-analytic review. Hepatology 1995; 22: in the control of varix bleeding. Surgery 1956; 39: 917–25.
332–54. 17 Westaby D. The management of active variceal bleeding.
Jalan R, Hayes PC. UK guidelines on the management of variceal Intensive Care Med 1988; 14: 100–5.
haemorrhage in cirrhotic patients. Gut 2000; 46(suppl III): 18 Rector WG. Drug therapy for portal hypertension. Ann Intern
iii1–15. Med 1986; 105: 96–107.
Rosch J, Keller FS. Transjugular intrahepatic portosystemic shunt: 19 Walker S. Vasoconstrictor therapy and bleeding oesophageal
Present status, comparison with endoscopic therapy and shunt varices. Hepatogastroenterology 1990; 37: 538–43.
surgery, and future prospectives. World J Surg 2001; 25: 20 Burroughs AK, McCormick PA, Hughes MD, et al. Randomised,
337–46. double-blind, placebo-controlled trial of somatostatin for
Tait IS, Krige JEJ, Terblanche J. Endoscopic band ligation of variceal bleeding. Gastroenterology 1990; 99: 1388–95.
oesophageal varices. Br J Surg 1999; 86: 437–46. 21 Jenkins SA, Shields R. Variceal haemorrhage after failed
Teran JC, Imperiale TF, Mullen KD, et al. Primary prophylaxis of injection sclerotherapy: the role of emergency oesophageal
variceal bleeding in cirrhosis: a cost-effectiveness analysis. transection. Br J Surg 1988; 76: 49–51.
Gastroenterology 1997; 112: 473–82. 22 McKee R. A study of octreotide in oesophageal varices.
Digestion 1990; 45(suppl): 60–5.
23 Shields R, Jenkins SA, Baxter JN, et al. A prospective
REFERENCES randomised control trial comparing the efficacy of
somatostatin with injection sclerotherpay in the control of
bleeding oesophageal varices. J Hepatol 1992; 16: 128–37.
1 Johnston GW. Bleeding from oesophageal varices. In: Imrie CW, 24 Planus R, Quer JC, Boix J, et al. A prospective randomized trial
Moosa AR (eds). Gastrointestinal Emergencies. Edinburgh: comparing somatostatin and sclerotherapy in the treatment of
Churchill Livingstone, 1987: 24–45. acute variceal bleeding. Hepatology 1994; 20: 370–5.
2 Henderson JM. Portal hypertension and shunt surgery. Adv Surg 25 Sung JJ, Chung SC, Lai CW, et al. Octreotide infusion or
1993; 26: 233–57. emergency sclerotherapy for variceal haemorrhage. Lancet
3 Ryley NG, McGee JO’D. Cirrhosis and portal hypertension: 1993; 342: 637–41.
pathological aspects. In: Blumgart LH (ed). Surgery of the Liver 26 Freitas DS, Sofia C, Pontes JM, et al. Octreotide in acute
and Biliary Tract. 2nd edn. New York: Churchill Livingstone, bleeding oesophageal varices: a prospective randomized sudy.
1994: 1589–602. Hepatogastroenterology 2000; 47: 1310–14.
4 Warren KS, Reboucas G. Blood ammonia during bleeding from 27 Sivri B, Oksuzoglu G, Bayraktar Y, Kayhan B. A prospective
esophageal varices in patients with hepatosplenic randomized trial from Turkey comparing octreotide versus
schistosomiasis. N Engl J Med 1964; 271: 921–6. injection sclerotherapy in acute variceal bleeding.
5 Mohamed AE, al Karawi MA, al Otaibi R, Hanid MA. Results Hepatogastroenterology 2000; 47: 168–73.
of sclerotherapy in 100 patients comparison of the 28 Jenkins SA, Shields R, Jaser N, et al. The management of
outcome between schistosomiasis and hepatitis B. gastrointestinal haemorrhage by somatostatin after apparently
Hepatogastroenterology 1989; 36: 333–6. successful endoscopic injection sclerotherapy for bleeding
6 Tilanus HW. Budd-Chiari syndrome. Br J Surg 1995; 82: 1023–30. oesophageal ulcers. J Hepatol 1991; 12: 296–301.
7 Rector WG, Hoefs JC, Hossack KF, Everson GT. Hepatofugal 29 Jenkins SA, Baxter JN, Rennie MJ. Acute and long-term treatment
portal flow in cirrhosis; observations on hepatic hemodynamics of oesophageal varices. Br J Intensive Care 1993; 3: 65–72.
and the nature of the arterioportal communications. 30 Kleber G, Sauerbruch T, Fischer G, et al. Reduction of
Hepatology 1988; 8: 16–20. transmural oesophageal variceal pressure by metoclopramide. J
8 Lebrec D, De Fleury P, Rueff B, et al. Portal hypertension, size of Hepatol 1991; 12: 362–6.
esophageal varices and risk of gastrointestinal bleeding in 31 Taranto D, Suozzo R, de Sio I, et al. Effect of metoclopramide
alcoholic cirrhosis. Gastroenterology 1980; 79: 1139–44. on transmural oesophageal variceal pressure and portal blood
9 Jalan R, Hayes PC. UK guidelines on the management of variceal flow in cirrhotic patients. Digestion 1990; 47: 56–60.
haemorrhage in cirrhotic patients. Gut 2000; 46(suppl III): iii1–15. 32 Vlavianos P, Gimson AES, Westaby D, Williams R. Balloon
10 Mahl TC, Groszmann RJ. Pathophysiology of portal hypertension tamponade in variceal bleeding: use and misuse. BMJ 1989;
and variceal bleeding. Surg Clin North Am 1990; 70: 251–66. 298: 1158.
11 Bornman PC, Krige JEJ, Terblanche J. Management of 33 Crafoord C, Frenckner P. New surgical treatment of varicose
oesophageal varices. Lancet 1994; 343: 1079–84. veins of the oesophagus. Acta Otolaryngol 1939; 27: 422–9.
12 Beppu K, Inokuchi K, Koyanagi N, et al. Prediction of variceal 34 Johnston GW, Rodgers HW. A review of 15 years’ experience in
hemorrhage by esophageal endoscopy. Gastrointest Endosc the use of sclerotherapy in the control of acute haemorrhage
1981; 27: 213–18. from oesophageal varices. Br J Surg 1973; 60: 797–800.
13 The Italian Liver Cirrhosis Project. Reliability of endoscopy in 35 Terblanche J. Portal hypertension management. Surg Endosc
the assessment of variceal features. J Hepatol 1987; 4: 1993; 7: 472–8.
93–8. 36 Rose JDR, Crane MD, Smith PM. Factors affecting successful
14 Burns P, Taylor K, Blei AT. Doppler flowmetery and portal endoscopic sclerotherapy for oesophageal varices. Gut 1983;
hypertension. Gastroenterology 1987; 92: 824–6. 24: 946–9.
570 Portal hypertension and variceal bleeding

37 Paquet KJ, Oberhammer E. Sclerotherapy of bleeding 57 Johnson PA, Campbell DR, Antonson CW, et al. Complications
oesophageal varices by means of endoscopy. Endoscopy 1978; associated with endoscopic band ligation of esophageal varices.
10: 7–12. Gastrointest Endosc 1993; 39: 119–22.
38 Terblanche J. Has sclerotherapy altered the management of 58 Goldschmiedt M, Haber G, Kandel G, et al. A safety maneuver
patients with variceal bleeding? Am J Surg 1990; 160: 37–42. for placing overtubes during esophageal variceal ligation.
39 Gottlib JP. Endoscopic obturation of oesophageal and gastric Gastrointest Endosc 1992; 38: 399–400.
varices with a cyanoacrylic tissue adhesive. Can J Gastroenterol 59 Saltzman JR, Arora S. Complications of esophageal variceal
1990; 4: 637–8. band ligation. Gastrointest Endosc 1993; 39: 185–6.
40 Feretis C, Tabakopoulos D, Benakis P, Xenofontos M, Golematis 60 Orloff MJ, Orloff MS, Rambolt M, Girard B. Is portosystemic
B. Endoscopic hemostasis of esophageal and gastric variceal shunt worthwhile in Child’s class C cirrhosis? Ann Surg 1993;
bleeding with Histoacryl. Endoscopy 1990; 22: 282–4. 216: 256–66.
41 Kahn D, Jones B, Bornman PC, Terblanche J. Incidence and 61 Orloff MJ, Bell RH, Greenberg AG. A prospective randomized
management of complications after injection sclerotherapy: trial of emergency portacaval shunt and medical therapy in
a 10 year prospective evaluation. Surgery 1989; 105: 160–5. unselected cirrhotic patients with bleeding varices.
42 Dhiman RK, Chawla Y, Taneja S, et al. Endoscopic sclerotherapy Gastroenterology 1986; 90: 1754.
of gastric variceal bleeding with N-butyl-2-cyanoacrylate. 62 Rikkers LF, Jin G. Surgical management of acute variceal
J Clin Gastroenterol 2002; 35: 222–7. hemorrhage. World J Surg 1994; 18: 193–9.
43 Kind R, Guglielmi A, Rodella L, et al. Bucrylate treatment 63 Rikkers LF, Jin G. Variceal hemorrhage: surgical therapy.
of bleeding gastric varices: 12 years’ experience. Endoscopy Gastroenterol Clin North Am 1993; 22: 821–42.
2000; 32: 512–19. 64 Rosch J, Hanafee W, Snow H. Transjugular portal venography
44 Gimson A, Polson R, Westaby D, Williams R. Omeprazole and radiologic portacaval shunt. An experimental study.
in the management of intractable esophageal ulceration Radiology 1969; 92: 1112–14.
following injection sclerotherapy. Gastroenterology 1990; 65 Colapinto RF, Stronell RD, Birch SJ. Creation of an intrahepatic
99: 1829–31. portosystemic shunt with a Gruntzig balloon catheter. CMAJ
45 Terblanche J, Krige JEJ, Bornman PC. The treatment of 1982; 126: 267–8.
esophageal varices. Annu Rev Med 1992; 43: 69–82. 66 Richter GM, Palmaz JC, Noeldge G, Rossle M. The transjugular
46 Terblanche J, Bornman PC, Kahn D, et al. Failure of repeated intrahepatic portosystemic stent-shunt. A new nonsurgical
injection sclerotherapy to improve long-term survival after percutaneous method. Radiology 1989; 29: 406–11.
oesophageal variceal bleeding. Lancet 1983; ii: 1328. 67 Jalan R, Redhead DN, Hayes PC. Transjugular intrahepatic
47 Terblanche J, Burroughs AK, Hobbs KEF. Controversies in the portasystemic stent shunt in the treatment of variceal
management of bleeding oesophageal varices. N Engl J Med haemorrhage. Br J Surg 1995; 82: 1158–64.
1989; 320: 1398–8. 68 Rosch J, Keller FS. Transjugular intrahepatic portosystemic
48 Bornman PC, Terblanche J, Kahn D, et al. Limitations of multiple shunt: Present status, comparison with endoscopic therapy and
injection sclerotherapy sessions for acute variceal bleeding. shunt surgery, and future prospectives. World J Surg 2001; 25:
S Afr Med J 1986; 70: 34–6. 337–46.
49 Burroughs AK, Hamilton G, Phillips A, et al. A comparison of 69 Ring EJ, Lake JR, Roberts JP, et al. Using transjugular
sclerotherapy with staple transection of the esophagus for the intrahepatic portosystemic shunts to control variceal bleeding
emergency control of bleeding from esophageal varices. before liver transplantation. Ann Intern Med 1992; 116: 304–9.
N Engl J Med 1989; 321: 857–62. 70 Helton WS, Belshaw A, Althaus S, et al. Critical appraisal of the
50 Mostafa I, Omar MM, Nouh A. Endoscopic control of gastric angiographic portacaval shunt (TIPS). Am J Surg 1993; 165:
variceal bleeding with butyl cyanoacrylate. Endoscopy 566–71.
1993; 25: A11. 71 Rossle M, Haag K, Ochs A, et al. The transjugular intrahepatic
51 Pretis G, de Comberlato M, Guelmi A, et al. N-butyl- portasystemic stent-shunt procedure for variceal bleeding.
cyanoacrylate in bleeding oesophageal (EBV) and gastric (GBV) N Engl J Med 1994; 330: 165–7.
varices: experience in 135 cirrhotic patients. Endoscopy 1993; 72 Papatheodoridis GV, Goulis J, Leandro G, et al. Transjugular
25: A14. intrahepatic portosystemic shunt compared with endoscopic
52 Thakeb F, Abdel Kader S, Salama Z, et al. The value of combined treatment for prevention of variceal rebleeding: a meta-
use of N-butyl cyanoacrylate and ethanolamine oleate in the analysis. Hepatology 1999; 30: 612–22.
management of bleeding esophagogastric varices. Endoscopy 73 McCormick PA, Dick R, Panagou EB, et al. Emergency
1993; 25: A16. transjugular intrahepatic portasystemic stent shunting as salvage
53 Stiegmann G, Sun JH, Hammond WS. Results of experimental treatment for controlled variceal bleeding. Br J Surg 1994; 81:
endoscopic esophageal varix ligation. Am Surg 1988; 54: 1324–7.
104–8. 74 Jalan R, John TG, Redhead DN, et al. A comparative study of the
54 Binmoeller KF, Vadeyar HJ, Soehendra N. Treatment of transjugular intrahepatic portosystemic stent-shunt (TIPSS) and
esophageal varices. Endoscopy 1994; 26: 42–7. oesophageal transection in uncontrolled variceal haemorrhage.
55 Stiegmann GV, Goff JS, Michaletz-Onody PA, et al. Endoscopic Am J Gastroenterol 1995; 11: 1932–6.
sclerotherapy as compared with endoscopic ligation for 75 Merli M, Salerno F, Riggio O, et al. Transjugular intrahepatic
bleeding esophageal varices. N Engl J Med 1992; 326: 1527–32. portosystemic shunt versus endoscopic sclerotherapy for the
56 Saeed ZA, Michaletz PZ, Wincester CB, et al. Endoscopic prevention of variceal bleeding in cirrhosis: a randomized
variceal ligation in patients who have failed endoscopic multicenter trial. Gruppo Italiano Studio TIPS (GIST).
sclerotherapy. Gastrointest Endosc 1990; 36: 572–4. Hepatology 1998; 27: 48–53.
 References 571

76 Escorsell A, Banares R, Garcia-Pagan JC, et al. TIPS versus drug response for the prevention of bleeding. Hepatology 2002; 36:
therapy in preventing variceal rebleeding in advanced cirrhosis: 1361–6.
a randomized controlled trial. Hepatology 2002; 35: 385–92. 96 Tait IS, Krige JEJ, Terblanche J. Endoscopic band ligation of
77 Sternbeck M, Ring E, Gordon R, et al. Intrahepatic portocaval oesophageal varices. Br J Surg 1999; 86: 437–46.
shunt: a bridge to liver transplantation in patients with 97 Hashizume M, Ohta M, Ueno K, et al. Endoscopic ligation of
refractory variceal bleeding. Gastroenterology 1991; 100: 801. esophageal varices compared with injection sclerotherapy: a
78 Stanley AJ, Jalan R, Forrest EH, et al. Long-term follow-up of prospective randomized trial. Gastrointest Endosc 1993; 39:
transjugular intrahepatic portosystemic stent shunt (TIPSS) for 123–6.
the treatment of portal hypertension: results in 130 patients. 98 Lo GH, Lai KH, Cheng JS, et al. The additive effect of sclerotherapy
Gut 1996; 39: 479–85. to patients receiving repeated endoscopic variceal ligation: a
79 Ferguson J, Jalan R, Redhead DN, et al. The role of duplex prospective, randomized trial. Hepatology 1998; 28: 391–5.
Doppler in monitoring shunt function following transjugular 99 Saeed ZA, Stiegmann GV, Ramirez FC, et al. Endoscopic variceal
intrahepatic stent shunt. Br J Radiol 1995; 68: 587–9. ligation is superior to combined ligation and sclerotherapy for
80 Vankemmel M. Resection-anastomose de l’oesophage sus-cardial oesophageal varices: a multicentre prospective randomised
pour rupture de varices oesophagiennies. Nouv Presse Med 1974; trial. Hepatology 1997; 25: 71–4.
5: 1123–4. 100 Iso Y, Kawanaka H, Tomikawa M, et al. Repeated sclerotherapy
81 Sugiura M, Futagawa S. A new technique for treating is preferable to combined therapy with variceal ligation to
oesophageal varices. J Thorac Cardiovasc Surg 1973; 66: 677–85. avoid recurrence of oesophageal varices: a prospective
82 Sugiura M, Futagawa S. Esophageal transection with para- randomised trial. Hepatogastroenterology 1997; 44: 467–71.
esophagogastric devascularization (the Sugiura procedure) in the 101 Al Traif I, Fachartz FS, Al Jumah A, et al. Randomized trial of
treatment of esophageal varices. World J Surg 1984; 8: 673–9. ligation versus combined ligation and sclerotherapy for
83 Burroughs AK, Mezzanotte G, Phillips A, et al. Cirrhotics with bleeding esophageal varices. Gastrointest Endosc 1999;
variceal hemorrhage: the importance of the time interval 50: 1–6.
between admisssion and the start of analysis for survival and 102 Rypins EB, Mason RG, Conroy RM, Sarfeh IJ. Predictability and
rebleeding rates. Hepatology 1989; 9: 801–7. maintenance of portal flow patterns after small-diameter
84 Iwatsuki S, Starzl TE, Todo S, et al. Liver transplantation in the portacaval H-grafts in man. Ann Surg 1984; 200: 706–10.
treatment of bleeding esophageal varices. Surgery 1988; 104: 103 Sarfeh IJ, Rypins EB, Moussa R, et al. Serial measurement of
697–705. portal haemodynamics after partial portal decompression.
85 D’Amico G, Pagliaro L, Bosch J. The treatment of portal Surgery 1986; 100: 52–8.
hypertension: a meta-analytic review. Hepatology 1995; 104 Adam R, Diamond T, Bismuth H. Partial portacaval shunt –
22: 332–54. renaissance of an old concept. Surgery 1992; 111: 610–16.
86 Patch D, Sabin CA, Goulis J, et al. A randomized, controlled trial 105 Rosemurgy AS, Serafini FM, Zweibel BR. Transjugular intrahepatic
of medical therapy versus sendoscopic ligation for the portosystemic shunt vs. small-diameter prosthetic H-graft
prevention of variceal rebleeding in patients with cirrhosis. portacaval shunt: extended follow-up of an expanded randomized
Gastroenterology 2002; 123: 1013–19. prospective trial. J Gastrointest Surg 2000; 4: 589–97.
87 Abraldes JG, Tarantino I, Turnes J, et al. Hemodynamic response 106 Rikkers LF. Definitive therapy for variceal bleeding: a personal
to pharmacological treatment of portal hypertension and view. Am J Surg 1990; 160: 80–5.
long-term prognosis of cirrhosis. Hepatology 2003; 37: 902–8. 107 Warren WD, Millikan WJ, Henderson JM, et al. Ten years’ portal
88 Burroughs AK, McCormick PA. Prevention of variceal hypertensive surgery at Emory. Results and new perspectives.
rebleeding. Gastroenterol Clin North Am 1992; 21: 119–47. Ann Surg 1982; 195: 530–42.
89 Rikkers LF, Jin G, Burnett DA, et al. Shunt surgery versus 108 Maillard J, Flamant YM, Hay JM, Chandler JG. Selectivity of the
endoscopic sclerotherapy for variceal hemorrhage: late results distal splenorenal shunt. Surgery 1979; 86: 663–71.
of a randomized trial. Am J Surg 1993; 165: 27–33. 109 Helton WS, Maves R, Wicks K, Johansen K. Transjugular
90 Henderson JM, Kutner MH, Millikan WJ Jr, et al. Endoscopic intrahepatic portasystemic shunt vs surgical shunt in good-risk
variceal sclerosis compared with distal splenorenal shunt to cirrhotic patients: a case-control comparison. Arch Surg 2001;
prevent variceal hemorrhage in cirrhosis. A prospective, 136: 17–20.
randomized trial. Ann Intern Med 1990; 112: 262–9. 110 Bismuth H, Adam R, Mathur S, Sherlock D. Options for elective
91 Terblanche J. The surgeon’s role in the management of portal treatment of portal hypertension in cirrhotic patients in the
hypertension. Ann Surg 1989; 209: 381–95. transplantation era. Am J Surg 1990; 160: 105–10.
92 Terblanche J, Kahn D, Borman PC. Long-term injection 111 Millikan WJ, Henderson JM, Stewart MT, et al. Change in
sclerotherapy treatment for esophageal varices: a 10-year hepatic function, hemodynamics and morphology after liver
prospective evaluation. Ann Surg 1989; 210: 725–31. transplantation. Physiological effect of therapy. Ann Surg 1989;
93 Infante-Rivard C, Esnaola S, Villeneuve JP. Role of endoscopic 209: 513–25.
variceal sclerotherapy in the long-term management of 112 Grace ND. Prevention of initial variceal hemorrhage.
variceal bleeding: a meta-analysis. Gastroenterology 1989; Gastroenterol Clin North Am 1992; 21: 149–61.
96: 1087–92. 113 Kleber G, Sauerbruch T, Ansari H, Paumgartner G. Prediction of
94 Laine L. Ligation: endoscopic treatment of choice for patients variceal hemorrhage in cirrhosis: a prospective follow-up study.
with bleeding oesophageal varices. Hepatology 1995; 22: Gastroenterology 1991; 100: 1332–7.
661–5. 114 Conn HO, Lindenmuth WW, May CJ, Ramsby GR. Prophylactic
95 Bureau C, Peron JM, Alric L, et al. ‘A La Carte’ treatment of portacaval anastomosis. A tale of two studies. Medicine 1972;
portal hypertension: adapting medical therapy to hemodynamic 51: 27–40.
572 Portal hypertension and variceal bleeding

115 Jackson FC, Perrin EB, Smith AG, et al. A clinical investigation prophylaxis of variceal bleeding. Eur J Gastroenterol Hepatology
of the portacaval shunt. ii Survival analysis of the prophylactic 1996; 8: 337–42.
operation. Am J Surg 1968; 115: 22–42. 124 Lay CS, Tsai YT, Teg CY, et al. Endoscopic variceal ligation in
116 Resnick RH, Chalmers TC, Ishihara AM, et al. A controlled study prophylaxis of first variceal bleeding in cirrhotic patients with
of the prophylactic portacaval shunt. A final report. Ann Intern high-risk esophageal varices. Hepatology 1997; 25: 1346–50.
Med 1969; 70: 675–88. 125 Hayes PC, Davis JM, Lewis JA, et al. Meta-analysis of value of
117 Inokuchi K, Cooperative Study Group of Portal Hypertension propanolol in prevention of variceal haemorrhage. Lancet 1990;
in Japan. Improved survival after prophylactic portal 336: 153–6.
nondecompressive surgery for esophageal varices: a 126 Poyard T, Cales P, Pasta L, et al. and the Franco-Italian
randomized controlled trial. Hepatology 1990; 2: 1–6. Multicentre Study Group. Beta-adrenergic antagonist drugs
118 Paquet KJ. Prophylactic endoscopic sclerosing treatment of in the prevention of gastrointestinal bleeding in patients with
the oesophageal wall in varices: a prospective controlled cirrhosis and esophageal varices: an analysis of data and
randomized trial. Endoscopy 1982; 14: 4–5. prognostic factors in 589 patients from four randomised
119 Witzel L, Wolbergs E, Merki H. Prophylactic endoscopic clinical trials. N Engl J Med 1991; 324: 1532–8.
sclerotherapy of oesophageal varices: a prospective controlled 127 Teran JC, Imperiale TF, Mullen KD, et al. Primary prophylaxis
trial. Lancet 1985; i: 773–5. of variceal bleeding in cirrhosis: a cost-effectiveness analysis.
120 Pagliaro L, D’Amico G, Sorensen TIA, et al. Prevention of first Gastroenterology 1997; 112: 473–82.
bleeding in cirrhosis. A meta-analysis of randomised trials 128 Angelico M, Carli L, Piat C, et al. Isosorbide-5 mononitrate
of nonsurgical treatment. Ann Intern Med 1992; 117: versus propanolol in the prevention of first bleeding in
59–70. cirrhosis. Gastroenterology 1993; 104: 1460–5.
121 Van Ruiswyk J, Byrd JC. Efficacy of prophylactic sclerotherapy 129 Merkel C, Martin R, Enzo E, et al. Randomised controlled trial
for prevention of a first variceal hemorrhage. Gastroenterology of nadolol alone or with isosbide mononitrate for primary
1992; 102: 587–97. prophylaxis of variceal bleeding in cirrhosis. Lancet 1996; 348:
122 Santangelo WC, Dueno MI, Estes BL, Krejs GJ. Prophylactic 1677–81.
sclerotherapy of large esophageal varices. N Engl J Med 1988; 130 Garcia-Pagan JC, Morillas R, Banares R, et al. Propranolol plus
318: 814–18. placebo versus propranolol plus isosorbide-5-mononitrate in
123 Sarin SK, Guptan RK, Jain AK, et al. A randomised controlled the prevention of a first variceal bleed: a double-blind RCT.
trial of endoscopic variceal band ligation for primary Hepatology 2003; 37: 1260–6.
 47
Cold Injury

PER-OLA GRANBERG

The problem 573 Non-freezing cold injuries 579

Thermoregulation 573 Hypothermia 580
Freezing cold injuries 574 References 586

THE PROBLEM THERMOREGULATION

Humans are tropical beings but, unlike most other mam- The human body tries to maintain thermal equilibrium
mals, we lack an effective hairy covering. Our adaptation to between heat production and heat loss with a core tempera-
a cold environment was primarily an intellectual event. ture of 37  2 °C. Basal heat production under conditions
Thus, cold induced injuries are almost exclusively a result of thermal comfort is around 70–100 W. Body temperature
of our inability to protect ourselves adequately against our is regulated in the hypothalamus. Thermosensitive endor-
environment. gans in the skin, and possibly elsewhere, transmit cold per-
Serious injuries caused by cold are in most cases pre- ception via the lateral hypothalamic tracts.4
ventable, occurring only sporadically in civilian life. Various mechanisms regulate heat loss. Depending on
Despite populations of about 100 million at risk in areas both temperature and humidity, radiation accounts for
where subzero temperatures are commonplace, injuries 60 per cent of total heat loss, convection for about 18 per cent,
caused by cold are surprisingly unusual or rare.1 On the conduction for 3 per cent and evaporation for 20–30
other hand such injuries are of major significance to mili- per cent. When thermoregulation is impaired and core tem-
tary personnel operating in the cold and to populations perature starts to decline, the affected individual suffers
involved in cataclysms. For example, during World War I, cold stress, which can deteriorate into hypothermia.
115 000 British soldiers suffered cold injuries, with mor-
bidity rates ranging from 27 to 33 per cent for the years
1914 and 1915.2 The corresponding figures for the French
Mechanisms of regulating heat loss
and Italian armies were 80 000 and 38 000, respectively.
During World War II, 10 per cent of American soldiers, i.e. • Radiation: heat loss – 60 per cent
more than 90 000, suffered cold injuries.3 The great losers • Convection: heat loss – 18 per cent
in this respect, however, were the Germans; in their case • Conduction: heat loss – 3 per cent
figures exceeding 600 000 victims have been presented. The • Evaporation: heat loss – 20–30 per cent
numbers could be much higher if one included the com-
mon combination of war wounds and hypothermia. Often In a cold environment one is obliged to prevent heat loss
the cold is an enemy worse than the opposition themselves. by wearing clothing which produces a subtropical micro-
During the Fenno–Russian War of 1940, 9000 Swedish vol- climate. If, however, such protection is inadequate, physi-
unteers enlisted in the Finnish army, and of these 130 had ological adjustment to cold is accomplished by two main
to be sent back to Sweden as a result of cold injuries; this mechanisms. Heat loss by conductance is limited by
contrasts with 33 who were killed and 50 who were increased cutaneous vasoconstriction mediated by unmyeli-
wounded. nated sympathetic constrictor fibres with noradrenaline as
574 Cold injury

the transmitter.5 Furthermore, heat production can be aug- induced vasodilatation is a result of the opening of arteri-
mented by voluntary muscular activity, also by involuntary ovenous anastomoses, occurring every 5–10 minutes. This
muscular tensing, but most of all by muscular shivering. alternating constriction and dilatation, known as the
Heat production is also influenced by the secretion of ‘hunting response’, was described by Sir Thomas Lewis as
thyroxine and catecholamines, the so-called thermogenic long ago as 1930.9 Cold induced vasodilatation is a com-
hormones.6,7 promise in the human physiological plan to conserve heat
Cold injuries may be either systemic or localised. The lat- and yet intermittently preserve the function of hands and
ter injuries most often precede systemic hypothermia. The feet. It is perceived by the patient as periods of prickling
local injuries described in the next two sections constitute heat, and it becomes less pronounced as body temperature
two clinically different entities: freezing cold injuries (FCIs) decreases. Great individual variations in the degree of
and non-freezing cold injuries (NFCIs). The final section of CIVD might explain differing susceptibility to local cold
this chapter deals with systemic hypothermia per se. injury. Importantly, people indigenous to a cold climate
present a more pronounced CIVD.
In contrast to cryopreservation of living tissue when
Broad types of cold injury crystallisation occurs intracellularly and extracellularly, the
clinical congelation with a slow rate of freezing produces
• Localised only extracellular ice crystals. The process is an exothermic
– Freezing cold injuries one, liberating heat, for which reason the temperature of
– Non-freezing cold injuries the limb remains at freezing point until freezing is com-
• Systemic plete. Not until then does the limb temperature start to
– Hypothermia decline to the ambient level.10
When water in the extracellular spaces is transformed
into ice, the osmolality of this space will increase and lead
FREEZING COLD INJURIES to a passive diffusion of water from the intracellular com-
partment.11 Cell hydration alters protein structures,
membrane lipids and cellular pH, leading to destruction
Freezing cold injuries occur when the temperature drops incompatible with cell survival.12 Resistance to freezing
below freezing point and there is true freezing of the injury varies in different tissues. Skin is more resistant than
tissue. Normally, human flesh freezes at 2 °C, but super- are muscle and nerves, possibly owing to a lower water
cooling is common if the skin is dry. Skin wetness is content, both intracellularly and intercellularly, in the
conducive to frostbite, as it allows crystallisation to ter- epidermis.
minate supercooling.8 Pathogenic mechanisms thought to Indirect haemorheological factors also influence the
contribute to tissue heat loss due to FCIs include a direct pathophysiology, but their role is debatable. These factors
cryogenic insult to the cells and vascular damage with start to influence matters during the prefreeze phase, with
decreased perfusion and tissue hypoxia. sympathetic vasoconstriction, which causes transendothe-
lial plasma leakage. The increased intravascular viscosity
Pathophysiology retards flow and causes sludging. This indirect vascular effect
was earlier interpreted as similar to that found in NFCIs.
The adrenergic vasoconstriction of cutaneous vessels is of Recent studies on rabbit ears13–15 however, have shown that
great importance in the origin of frostbite. The skin’s nor- freezing causes intimal lesions in the microvasculature
mal blood flow far exceeds its nutritional requirements. In prior to any evidence of damage to other skin parenchymal
peripheral structures such as hands, feet, nose and ears, the elements. The earliest and most severe lesions were found
flow can vary considerably owing to wide arteriovenous in the intima of the arterioles, but venules and capillaries
shunts. In a warm environment the hands, for example, are subsequently demonstrated the same changes. The endothe-
superperfused and the tips of the fingers are even warmer lial cells shrank and separated from the elastic internal
than the lips. Only about 10 per cent of the flow to the hands, lamina, and this event was present in samples removed
however, is needed for tissue oxygenation; the rest creates immediately after freezing, which proves that blood reflow
warmth, thereby facilitating dexterity. Local cooling of the could not be responsible for the changes. These findings
skin occludes these shunts even in the absence of any provide evidence that the rheological part of the pathogen-
decrease in core temperature. When skin is exposed to an esis of FCI is also dependent on a cryobiological effect. Other
extreme cold environment for a longer period of time, studies confirm these results.16
even the nutrition begins to suffer. Fingernails grow more When frostbitten tissue is rewarmed, ice crystals melt as
slowly in arctic climates than in temperate zones. an endothermic reaction, absorbing the previously lost
In order to protect the viability of the peripheral parts heat of crystallisation.10 This means that the temperature
of the extremities early during cold exposure, an intermit- of the frozen area levels out at around 2 °C until all the ice
tent cold induced vasodilatation (CIVD) takes place. Cold crystals have melted. Rediffusion of water to the dehydrated
 Freezing cold injuries 575

cells then starts, leading to intracellular swelling. The thaw- risk of subsequent frostbite owing to an abnormal sympa-
ing further induces a maximal vascular dilatation, causing thetic response.
hyperaemia. The endothelial cell injury increases fluid extra- Cold metal, when grasped with the bare hand, can rapidly
vasation, leading to oedema and blister formation. The cause an FCI owing to high conductive heat loss. Most
mechanical stability of red corpuscle aggregates increases, people are aware of this but often do not realise the risk of
and showers of emboli pass along the microvessels. The handling supercooled liquids. Petrol cooled down to 30 °C
disruption of the endothelial cells exposes the basement will freeze exposed flesh almost instantly as evaporative
membrane and provides an opportunity to initiate platelet heat loss combines with conductive loss. Such rapid freezing
adhesion, which in turn starts the coagulation cascade.17 causes both extracellular and intracellular freezing, with
Blood flow then stagnates, and progressive thrombosis destruction of cell membranes, primarily on a mechanical
culminates in anoxia. basis.22 Similar FCIs have been reported following spilling
Cells that are damaged by direct freezing injury with of liquid propane directly on to the skin.23,24 This gas is
dehydration, and those that succumb due to microcircula- usually transported as a liquid under pressure at 64 °C
tory congestion and anoxia, cannot be saved. However, in and has been shown in experimental animals to cause der-
an intermediate, pivotal zone of injury located between the mal necrosis after spraying for as little as 12 seconds.25
distal frostbitten tissue and the unaffected limb, various
forms of therapy might affect the survival of tissue. Progres-
Clinical picture
sive dermal ischaemia due to the increased enzyme activity
of vasoconstricting metabolites of arachidonic acid pos-
As with a thermal injury, FCI has been classified into first,
sibly plays a role in the pathogenesis in this zone.18,19 Free
second, third and fourth degree injuries. However, in the
radicals may also act in this context. Manson et al.15 suc-
initial stages it is certainly difficult to predict the extent of a
ceeded in improving viability in frozen rat ears by admin-
frostbite according to such a scheme. Freezing cold injuries
istering superoxide dismutase (SOD) at the time of thawing,
are therefore better subdivided into superficial and deep frost-
pointing to a reperfusion injury in both FCIs and NFCIs20
bites.26 The superficial injury is limited to the skin and the
(see Chapter 2).
immediately subcutaneous tissues. A stinging, pricking
pain is often the first symptom. In a snowstorm, however,
this sensation of pain is often absent. The affected skin
Environmental factors
turns pale or waxy-white, becomes numb and will indent
when pressure is applied to it, as the underlying and sur-
The development of an FCI depends upon ambient tempera-
rounding tissues are still viable and pliable (Fig. 47.1). If,
ture and duration of exposure. However, many other factors
however, the frostbite extends into a deep injury, the skin
influence the course of events. Experiences from polar
turns white and marble-like, feels hard and adheres to the
expeditions have taught that when the ambient tempera-
adjacent tissues. One must bear in mind, however, that
ture falls below 60 °C, human flesh may freeze within a
even a superficial FCI on the extremities may appear to be
minute. It is the heat loss to which the exposed tissue is sub-
frozen solid due to the freezing alone.26
jected which determines the damage, and in this respect
windchill is an important factor. In calm weather a nor-
mothermic person is surrounded by a thin layer of warm air
next to the skin. As wind velocity increases, more heat is lost
and the danger of possible cold injury increases. Thus the
temperature feels ‘much colder’ when the wind is blowing
as is obvious from the windchill chart.21 Any movement of
air past the body has the same effect. Thus, running, skiing,
skijoring and riding in open vehicles must be considered in
addition to direct wind. During World War II the US army
suffered more than 90 000 injuries due to cold, and heavy
bomber crews suffered more FCIs than all other injuries
combined.3 However, exposed flesh will not freeze as long
as the ambient temperature is above the freezing point,
even at high wind velocities.
Clothing, wet with perspiration, rain or melted snow
increases conductive loss because of reduced insulation.
Moreover, water as a thermal conductor is superior to air Figure 47.1 A superficial local cold injury is limited to the skin,
by a factor of about 25.21 Use of alcohol and tobacco prod- which turns pale or waxy-white. The injury should be treated as
ucts, as well as poor nutrition and fatigue, are also factors early as possible in order to avoid deep frostbite. (Photo courtesy of
predisposing to frostbite. Previous cold injury increases the Anders Holmström)
576 Cold injury

Treatment

FIRST AID TREATMENT

The golden rule of treatment is to take care of the local cold
injury at once. This prevents a superficial frostbite from turn-
ing into a deep one. If the victim cannot be taken indoors,
they must be protected from the wind by the shelter of com-
panions, wind sac or similar makeshift refuge. The frost-
bitten area should be thawed by passive transmission of heat
from a warmer part of the body, for example by placing the
warm hand against the face or the cold hand into the axilla or
groin. The victim is often under cold stress, with peripheral
vasoconstriction, and therefore a warm comrade can provide
excellent therapy. Massage and rubbing the frostbitten part
with snow or a woollen muffler is contraindicated. The tissue is
filled with ice crystals and such mechanical treatment aggra-
vates the injury. Similarly, rewarming with dry heat in front of
a camp fire or a camp stove should never be considered in any Figure 47.2 If available, rapid re-warming in warm water at
circumstances as the heat does not penetrate to any depth, the 40 °C to 42 °C has been found to give the best results. Thawing
area is partly anaesthetised and treatment may even end up in is continued until the flush has extended to the tips of the toes. The
a burn injury. thawed area develops a burgundy hue and the patient needs
Frostbitten feet give warning signals of pain when the local analgesics as this rapid thawing is painful
temperature falls to around ⫹15 °C. This sensation of pain
disappears before the actual freezing takes place, as nerve
conductivity is abolished at 8 to 7 °C.27 The paradox is FCIs: first aid treatment
that the last sensation one feels is that one does not feel
anything at all. In the field frostbitten feet are best treated • Protection from the wind or trudge through snow
by rewarming them in a comrade’s armpit in the lee of nat- buffeting arms against sides
ural shelter. Under extreme conditions, when evacuation • Warm clothes, warm nourishing beverage
from cold requires travel on foot, one should avoid thaw- • Thawing of area by passive transmission of body
ing. Walking on frostbitten feet does not appear to increase heat of victim or comrade
the risk of tissue loss, whereas refreezing of a local cold • Thawing in warm water maintained at 40 °C
injury certainly does. to 42 °C
If conditions allow, the most effective treatment for a frost-
bite is thawing in warm water at 40 °C to 42 °C, a temper-
HOSPITAL TREATMENT
ature just tolerable on the back of the hand. One should not,
however, postpone thawing by time-consuming transporta- If thawing with passive warmth transfer proves unsuccess-
tion, as the severity of the injury is a product of low tempera- ful after 20–30 minutes, a deep frostbite is present and the
ture and time. The temperature of the bath should be patient should be sent, properly protected, to the nearest
checked frequently to prevent the water from falling below hospital. If delay is likely because of transportation prob-
40 °C. The thawing procedure should continue until sensa- lems, it is better to get the patient into the nearest housing
tion, colour and tissue softness return. Rapid rewarming in and thaw the injury in warm water. After complete thawing
warm water often ends up in not so much a pink, but more a the patient should be put to bed with the injured area ele-
burgundy, hue due to venous stasis (Fig. 47.2). vated (Fig. 47.3). Prompt transportation should then be
Out in the field it is important to be aware of the fact arranged to the nearest hospital, where warm water thaw-
that treatment requires more than local thawing. The ing is recommended, provided that the injury is still in the
correct treatment includes the whole individual, as frostbite frozen state. The best clinical results are obtained, paradoxi-
is often the first sign of creeping hypothermia. More cally, by rapid rewarming, raising the tissue temperature
clothes should be put on and a warm, nourishing beverage without a simultaneous rise of blood flow and oxygen
given. The victim is often apathetic and unwilling to do access. This emphasises the cryobiological aetiology of
anything. Should victims be forced to stay in the open, they frostbite as the most important one, at least in the early
should be urged to take muscular activity such as trudging stages. Frozen tissue appears to withstand hypoxia better
through the snow and buffeting their arms against their than intracellular dehydration. There is evidence to suggest
sides. Such manoeuvres open the peripheral arteriovenous that if the whole individual is immersed in a warm water
shunts in the extremities. bath, the thawing process can be accelerated.28
 Freezing cold injuries 577

Figure 47.4 Six to 12 hours after thawing, blisters appear. Every

precaution should be taken to avoid their rupture. (Photo courtesy
of Ejnar Eriksson)
Figure 47.3 The thawed extremity is placed in an elevated
position. Interdigital specimens are taken for bacterial growth.
Tetanus toxoid is given, and the extremities are kept on sterile
sheets under cradles. (Photo courtesy of Ejnar Eriksson)

Rapid rewarming gives moderate to severe pain of a

burning quality, and the patient will often require an anal-
gesic. If ibuprofen is at hand 600 mg should be given as
soon as possible and the medication repeated every 6 hours
if transportation is prolonged. The capillary damage causes
leakage of serum, with local swelling and blister formation
during the first few days. Blistering is more common on the
back of the fingers, hands and feet, where the tissues are
more lax than on the palms or the soles (Fig. 47.4). Clear
blisters are most common. Haemorrhagic blister fluid Figure 47.5 Premature surgery is malpractice and ends up in
reflects structural damage to the subdermal plexus. In infection, delay in healing and unnecessary tissue loss
order to prevent infection the blister should be kept intact
and debrided only if infected. Tetanus prophylaxis is indi-
cated. Antibiotics should be given only if there is clinical entities when dealing with surgical therapy. In a burn
evidence of infection (Fig. 47.4). injury the border between living and dead tissue is estab-
Further hospital treatment should be open and non- lished within a few days, whereas this takes months in
occlusive (Fig. 47.3). Placement of sterile cotton pledgets frostbite, in which it is difficult to predict the amount of non-
between the toes and fingers is helpful. Daily treatment in viable tissue at an early stage.
a whirlpool bath at normal skin temperature cleanses the A conservative approach is to be recommended. ‘Frostbite
wounds gently, removes bacteria, aids circulation and pro- in January, amputate in July’ and ‘Hurry up and wait
vides physiological debridement. Patients are encouraged before you ablate’ are dictums to be remembered by the
to exercise the affected digits in the whirlpool.26 Such treat- over-zealous surgeon (Fig. 47.5). The only indication for
ment frightens many surgeons, who imagine that this may early amputation is septicaemia; this is a rarity. Efforts
well transform a dry gangrene into a wet one. However, to establish the borderline between living and dead tissue
unlike that in atherosclerosis, gangrene due to frostbite is in order to decrease the duration of inpatient hospi-
essentially superficial and the necrotic tissue may extend to talisation have through the past decades been numerous.
a depth of no more than a few millimetres. Escharotomy The multiplicity of procedures recommended, e.g. plain
on the dorsum or the dorsolateral part of the fingers should X-rays, venous radioisotopic scanning, angiography, digital
be performed if joint motion is limited.26 plethysmography, laser Doppler, thermography, technetium
Being a thermogenic injury, frostbite is often compared scintigraphy, triple-phase bone scanning, magnetic reso-
with burn injuries. The inappropriateness of classifying nance imaging and angiography speak for themselves and
cold injuries as burn injuries was mentioned earlier. It is show that, so far, there is no accurate method for early
even more inappropriate, however, to compare these two detection of viability. The appropriate level of amputation
578 Cold injury

after thawing. Conversely, it has been shown in experi-

mental studies on frostbite in rabbits that there was very
little deposition of fibrin after thawing.13 Marzella
et al. have confirmed these findings and conclude that this
observation was consistent with the lack of efficacy of
heparin administration in experimental frostbite.14
Low molecular weight dextran has often been used in the
treatment of local frostbite in order to reduce sludging and
thrombosis in the small blood vessels. Clinical results are far
from encouraging, and in animal experiments, dextran, like
heparin, has to be administered in the early postoperative
period.30 Some investigators find that infusion of the rheo-
logical substance may actually increase the oedema.31
Early sympathectomy is of doubtful value. Experimental
Figure 47.6 The blackened carapace acts as a protective
evidence has shown that sympathectomy performed
covering for tissue regeneration (photo courtesy of Ejnar Eriksson)
within the first few hours of injury exacerbates oedema,
blister formation and tissue destruction.17,32 ‘Medical sym-
is not clear until the tissue declares itself. Local frostbite is pathectomy’ with intra-arterial reserpine has frequently
still a ‘wait and see injury’. Even if early recognition should been used as the first choice of treatment. The drug causes
prove possible, thereby sparing days of hospitalisation, clin- significant depletion of arterial wall noradrenaline content
ical experience has shown that surgical intervention should for up to 4 weeks.33 However, in a multiarmed experimen-
be minimal. The blackened carapace acts as a protective tal study comparing slow re-warming combined with
covering for tissue regeneration and will separate by itself intravenous low molecular weight dextran, intra-arterial
without intervention (Fig. 47.6). Premature surgery most reserpine or tolazoline, all the drug treatments were super-
often ends up in infection, unnecessary tissue loss and delay ior to slow thawing. Rapid re-warming, however, was as
in healing, rather than the reverse.1,29 If amputation must effective as any of the drug treatments.34 Randomised multi-
be performed, a modified guillotine procedure is recom- centre studies are indicated to evaluate these adjuncts in
mended at the lowest level, with later secondary closure.26 the treatment of deep FCIs.
More recently, a combination of systemic antiprosta-
glandin therapy in the form of ibuprofen and the topical
agent aloe vera has been used to inhibit localised throm-
FCIs: hospital treatment boxane production, which has been implicated as a cause
of progressive dermal ischaemia.19 Such antiprostaglandin
• Prompt transportation
treatment was reported to reduce both hospital stay and
• Rapid warm water thawing if area affected is still
morbidity in patients with frostbite. The study, however,
frozen
was not randomised and it lacked patients with severe
• Analgesia: ibuprofen 600 mg, repeated every
injuries. The breakdown products of arachidonic acid were
6 hours
first implicated in burn injuries,35 but treatment with anti-
• Tetanus prophylaxis
prostaglandin agents has now been abandoned at most
• Antibiotics only if infection present
burns centres. One must remember, however, that a burn
• Blisters kept intact and debrided only if infected
injury and frostbite differ in many respects, which is why
• Cotton pledgets placed between toes and fingers
antiprostaglandin therapy may yet have a place in the treat-
• Daily whirlpool baths at normal skin temperature
ment of frostbite. Contrary to recommended protocols19
• Dorsal or dorsolateral escharotomy if finger joint
we prefer to administer systemic antiprostaglandin when-
movement limited
ever possible, even before thawing the local injury.
• Avoid premature debridement/amputation

FCIs: adjuvant therapy

ADJUVANT THERAPY
Considerable interest has been focused on treatment with
• Anticoagulant therapy – heparin if given early
various anticoagulants believed to prevent progressive
• Low molecular weight dextran
microvascular thrombosis and tissue loss. Treatment with
• Early sympathectomy, ‘medical sympathectomy’
using intra-arterial reserpine/tolazoline
heparin was used in Korea but proved ineffective. In order
to be of value the therapy should be given early, as intravas-
• Systemic antiprostaglandin therapy as ibuprofen
plus topical aloe vera
cular cellular aggregation, stasis and thrombosis occur soon
 Non-freezing cold injuries 579

NON-FREEZING COLD INJURIES the endothelial wall. In contrast, fibrin was never observed
within the occluded vessels. Later, increased permeability
leading to tissue oedema was demonstrated by extravasation
This form of local cold injury almost exclusively affects of FITC-dextran. In the light of this Endrich et al. considered
legs and/or feet. The aetiology of NFCIs is multifactorial. that the initial event in NFCI is rheological in nature and
Prolonged exposure to cold and wet conditions is a pre- resembles that observed during ischaemia-reperfusion
requisite, combined with immobilisation causing venous injury (see Chapter 2). Iyengar et al.39 drew the same con-
stagnation due to unnatural posture or constricting foot- clusion from experiments in rabbits, where they found that
wear. Dehydration, inadequate food and stress, intercurrent free radicals (OH) generated after cooling increased signif-
illness or injury, and fatigue are also important contributory icantly upon re-warming.
factors. The recognition of this type of cold injury has Wrenn adds other interesting observations to our
remained rather unusual because it seldom occurs in civilian understanding of the pathophysiology of NFCI.36 Thus, in
life. In the military context, however, NFCI has been and addition to the duration of temperature of the medium,
seems to be a serious problem, most often due to being the susceptibility of the victim is important. The patho-
unaware of the condition and the slow and indistinct first logical hallmark of immersion foot is waterlogging of
appearance of symptoms. the thick stratum corneum on the soles of the feet. Absorp-
Non-freezing cold injury was not distinguished from tion of 1–2 g of water per hour by the foot has been
frostbite until World War I. The injury was then called reported.40,41 The amount of absorption depends on the
‘trench foot’ as it occurred in soldiers on trench duty. salinity of the medium: fresh water causes much greater
During World War II the syndrome was seen in the sur- absorption. The absorbed liquid passes into the circulation
vivors of ships sunk in arctic waters who had spent time in but some is always retained in the stratum corneum. The
water-logged boats or rafts, and this injury was therefore thicker sole is more prone to injury than the thinner skin
called ‘immersion foot’. Other names for the disorder on the dorsum of the foot.
include swamp foot, tropical jungle foot, paddy-field The pathological changes thus induced affect many tis-
foot and foxhole foot. Non-freezing cold injuries can sues. Fat degenerates early on and is replaced by fibrous tis-
occur in any condition where the environmental tempera- sue. Muscles undergo degeneration, necrosis and fibrosis
ture is lower than body temperature. The exposure time and ultimately atrophy, similar in appearance to that seen
necessary to produce NFCIs lengthens with increasing in chronic denervation.42 Osteoporosis is an early event,
temperature of the wet medium. Wrenn has suggested probably caused by vascular disturbances.
that the single term ‘immersion foot’ should be applied to Of special interest is the pathophysiology of the nerves,
this syndrome.36 In connection with the more severe as it is nerve damage which accounts for the pain, prolonged
effects of immersion at temperatures below 15 °C, dysaesthesia and hyperhidrosis. Denny-Brown et al.43
the subcategory ‘cold water immersion foot’ should be reported from experiments on cats that, in NFCI, large
used. The more benign variant occurring at temper- myelinated fibres suffered injury most likely mediated by
atures above 15 °C should be called ‘warm water immer- ischaemia. On exposing the tibial nerve in rabbit, Kennet
sion foot’. and Gilliat44 found local failure of conduction at the site of
cooling (1–5 °C for 2–4 hours), which persisted after
Pathophysiology re-warming. This was followed by distal degeneration of the
affected fibres. The fastest conducting motor and afferent
While experimental studies on FCIs have been numerous axons were particularly affected. Histology revealed pri-
over the years, there are scarcely any such studies on NFCIs mary axonal damage, particularly in large-diameter fibres.
in the literature, and the true genesis of this type of cold The changes are similar to those described after ischaemic
injury is still far from resolved. As in FCIs, sympathetic injury.45 The hypoxic theory, however, seems less probable
constrictor fibres, together with the cold itself, induce pro- in cooling nerve tissue, as the duration of cooling was too
longed vasoconstriction. The contributory effect of cold short for ischaemic injury to occur. By immersing the hind
reduces the rate of reuptake and metabolism of noradren- limb of the rabbit in a water bath at 1 °C for 10–14 hours,
aline at the nerve endings.37 Kennet and Gilliat46 reported subsequent persistent nerve
Endrich et al. presented interesting findings concerning damage to the tibial nerve, with distal degeneration of the
the microvascular ultrastructure in the NFCI.38 In an animal affected fibres. Evidence of persistent conduction block,
model, using a transparent chamber implanted into a dorsal however, was not seen. The study also showed that pure
skin fold in Syrian hamsters, they studied the integrity of cold neuropathy occurred in the absence of histological
the subcutaneous microcirculation during repeated non- abnormalities in blood vessels or muscles. This is an inter-
freezing local cold exposures. Endothelial damage was prom- esting observation as, in milder cases of NFCI, neurological
inent in the true capillaries and venous vessels, while arterioles symptoms appear to predominate, with minor involvement
remained unaffected. Smaller vessels were often completely of other tissues. The involvement of autonomic fibres in
clogged with blood cells and with leucocytes integrated into NFCI is of particular importance. The post-traumatic
580 Cold injury

hyperhidrosis and cold sensitivity are similar to the effects

of postganglionic sympathetic denervation.47
NFCIs: treatment

• Generally conservative
• Wash and dry the feet and keep them cool (no active
Clinical picture warming)
• Bed rest in warm clothes with feet elevated
Compared with an FCIS, NFCIS have an insidious onset. • Warm beverages
As the symptoms are vague victims realise the imminent • Tetanus toxoid
danger only when it is too late. The feet become cold and • Analgesic and/or NSAIDs
swollen and feel heavy, woody and numb. Clinically, the • Heparin/low molecular weight dextran
condition reveals a cool, painful, tender, white foot with • Antibiotics if indicated
wrinkled soles and most often pulseless pedal arteries on • Amputation when injury irreversible
palpation. Brawny oedema, scaling, maceration, erythema
and abrasions over pressure points from shoes or socks
may also be seen. This first, ischaemic phase can last for Prevention
hours up to a few days and is followed by a hyperaemic phase
of 2–6 weeks, during which time the feet are warm, with As the treatment of NFCI is often passive and relatively
bounding pulses and increased swelling. Anaesthesia is ineffective, prophylaxis is extremely important. Adequate
replaced by tingling and aching pain, now and then inten- foot care with time to dry the feet is crucial, as also are dry
sified by intermittent throbbing. There may be blistering socks, rest with the feet elevated and hot beverages if at all
and ulceration, which in severe cases may culminate in possible. NFCI is almost exclusively a war phenomenon.
gangrene. The British Army learned the lesson from World War I
and showed that this form of injury was preventable. For
example, the 10th British Corps, which formed an integral
Treatment part of the 5th US Army, had, during their Italian campaign,
a ratio of 1:45 between cold injuries and battle casualties;
The treatment of NFCIs is basically supportive. Out in the the corresponding ratio for the US corps was 1:4.51
field the feet should be dried carefully but, in contrast to freez- Ahle et al.51 reinvestigated Argentinian soldiers with trench
ing injuries, kept cool. The whole body, however, should be foot injuries in the Falklands war. Passive re-warming of
warmed. Plenty of warm beverages should be given. In the cold-water stressed foot was evaluated in 33 trench
hospital the patient is put to bed with the feet elevated. The foot patients who had recovered and in 15 uninjured
local injuries should not be actively warmed. Treatment subjects. Poor re-warming was noted in all injured
with 40–42 °C warm water as used for FCIs is contraindi- subjects, compared with five uninjured patients but 14
cated. Warm water is only allowed when ice crystals are normal subjects failed to re-warm normally following the
present in the tissue. cold stress, and this group was statistically indistinguish-
The feet should be carefully washed, dried and cooled able from the injured group. This could mean that a large
with a fan.48 Tetanus toxoid should be given. Non-steroid proportion of a normal population are at great risk of suf-
anti-inflammatory agents (NSAIDs) may be beneficial, but fering NFCI under appropriate conditions.
most often analgesics are required to mitigate the pain.
Heparinisation and the use of low molecular weight dex- HYPOTHERMIA
tran may be of value, but convincing reports are still awaited.
Early surgical sympathectomy should be avoided, and the
value of chemical sympathectomy is controversial. The human body strives to maintain thermal equilibrium
As a rule treatment should be conservative. Nevertheless, between heat production and heat loss, with a core tem-
fever, an elevated creatinine phosphokinase level, signs of perature of 37  2 °C. When thermoregulation is impaired
disseminated intravascular coagulation, and liquefaction and core temperature starts to decline, the individual suf-
of affected tissues indicate early surgical intervention. In fers cold stress, but not until the central temperature
most reported cases antibiotics have played a minor role in reaches 35 °C is the victim considered to be in a hypother-
recovery, even in the presence of fever and lymphadenop- mic state.
athy.49 In severe cases, however, coverage for both strepto-
coccal and staphylococcal species should be given until the Physiological reactions to decreased core
results of cultures become available.37 Patients suffering temperature
from NFCI injuries later complain of cold sensitivity,
hyperhidrosis, swollen tender feet and sometimes contrac- During cold stress an intense adrenergic vasoconstriction
tures of their toes.50 redirects blood from skin to deeper areas, i.e. from the shell
 Hypothermia 581

to the core, thereby preventing heat conduction from the moderate and below 28 °C, severe. As temperature decreases
core to the skin. This physiological reaction increases the cardiac output and blood pressure may be markedly
insulating capacity sixfold.52 The vasoconstriction is most depressed by the negative inotropic and chronotropic
pronounced in the extremities.27 The arms and legs consti- effects of hypothermia and further depressed by the con-
tute around 30 per cent of body weight and half of the body comitant hypovolaemia.
surface. Such a ‘physiological amputation’ of the extrem- At 33–32 °C systole is prolonged more than diastole.
ities means a tremendous reduction in heat loss. Adrenergic Atrial irritability in early hypothermia often induces atrial
vasoconstriction does not occur in the head and neck region. fibrillation. J waves or Osborne waves are observed at tem-
A bare headed person loses around half of their resting heat peratures lower than 32 °C.57 This electrocardiographic
production at 4 °C and this loss increases to 75 per cent abnormality is a secondary wave following the S wave. At
at 15 °C.53 lower temperatures ventricular extrasystoles are common.
Good health, good physical fitness and adaptation to Death supervenes at temperatures from 28 °C and lower,
cold improve the tolerance to the lowered temperature; most often resulting from ventricular fibrillation, but asys-
cold stress raises the respiratory rate. Immersion in cold tole may also occur.58 The lowest known adult accidental
water can increase ventilation fourfold and induce hypocap- hypothermia was a 13.7 °C core temperature,59 and the
nia. Further, cold induces shivering, which is preceded by lowest infant hypothermia was 15.2 °C.60
increased muscular tone. Shivering is believed to result Hypothermia depresses the central nervous system.
from feedback oscillations of the stretch reflex mechanism Cerebral metabolism decreases 6–7 per cent per degree
and can increase the metabolic rate four- to sixfold. Of Celsius.61 Lassitude and apathy are early signs of decreasing
importance also is good nutrition and last but not least temperature. Sluggish cerebration impairs judgement,
adequate fluid balance. Dehydration is often neglected in causes bizarre behaviour and ataxia and often ends in
this connection. lethargy and coma between 30 and 28 °C. At 28–27 °C there
A hypothermic victim has almost always been fighting are no muscle reflexes, no response to pain and no pupillary
with the cold, the snow and the wind. Cold stress elevates light reflex. The corneal reflex disappears below 25 °C.62
the catecholamines released from the sympathetic nerve The electroencephalogram flattens out at 20–19 °C.
terminals, which induces peripheral vasoconstriction, cen- Exposure to an environmental temperature colder than
tralises blood volume and increases blood pressure and one’s body temperature is the basic prerequisite for the
cardiac output.54 These events lead to diuresis. This cold occurrence of hypothermia. Extremes of age are risk fac-
induced increase of urine flow is an osmolal diuresis with tors. Neonates with an increased surface area to body mass
sodium and chloride as the main constituents independent ratio, as well as elderly persons with impaired thermoregu-
of the antidiuretic hormone. Cold diuresis decreases the latory function and reduced muscle mass and insulating fat
blood volume leading to a reduced physical working layer, run a greater risk of suffering hypothermia.
capacity.55 The kidney is only a magnifying mirror in this Victims dying of hypothermia are often found in a total
context. Increased filtration is general and fluid also moves or partial undressed state. This phenomenon has been
into the extracellular and intracellular spaces. Working in named ‘paradoxical undressing’. This peculiar behaviour
the cold furthermore induces increased water loss via res- had already been observed in 1719. When the Swedish
piration due to the low water vapour pressure of cold air King Charles XII was killed during his war against Norway,
which also disguises loss via perspiration. Most insidious, General Armfeldt, who had tried to capture Trondhjelm,
however, is the victim’s peculiar lack of thirst despite dehy- decided to beat a retreat. With 6000 men he went over the
dration56 and the difficulties of getting hold of enough border mountains but his army was caught in a horrible
water in a wintry environment. A decreased capacity for snowstorm. More than 3000 soldiers froze to death. In an
work, increased peripheral vasoconstriction and raised effort to find out whether any equipment was salvageable
blood viscosity are all factors which multiply the risk for the surviving soldiers discovered to their surprise that
local cold injuries as well as general hypothermia! many of the dead were more or less naked. Historians have
Under such circumstances leaders of groups should be later claimed that those who survived were such cruel bas-
aware of the problem and see to it that members stay well tards that they robbed their comrades of their clothing.
hydrated. The order should be: ‘Drink your fill and double Those who made such a statement had never experienced a
that volume’. In forced situations it is wise to use every snowstorm.
member’s own ‘autoanalyser’: let them pee into the snow! The reasons why a person starts to undress in the cold
Clear, pale, yellow urine indicates adequate hydration have been difficult to establish. Hallucinations and neuro-
while dark, orange and reddish-brown marks in the snow chemical changes in the brain have been suggested.63 A
tells you that somebody is in trouble. cold induced paralysis of the arterial nerves with ‘warm’
When core temperature is lower than 35 °C the human core blood perfusing the skin seems possible. Some of the
body begins to lose its ability to generate enough heat to reported victims, however, had been walking quite a dis-
maintain bodily functions. Between 35 and 32 °C, the tance while undressing, and as these individuals were hypo-
hypothermia is classified as mild, between 32 and 28 °C it is volaemic, vasodilation would promptly lead to syncope.
582 Cold injury

Recent studies have thrown a new light on this phenom- skin and airways and sequestration of water in the intercel-
enon. It is well known that when a patient gets an infection lular and intracellular compartments, all events leading to
their central temperature set-point is raised and shiver- hypovolaemia and decreased capacity for physical work.
ing begins until body temperature has reached the new The haematocrit rises by around 2 per cent per degree
set-point. In analogy the set-point is lowered in deep Celsius decline in temperature.
hypothermia and the victim experiences warmth and starts Subclinical chronic hypothermia is most often found in
to undress. The phenomenon has been named ‘anapyrexia’ elderly persons, often in association with malnutrition,
or ‘reversed fever’.64 In this context it is worth mentioning inadequate clothing and restricted mobility. Acute or
that in animal experiments alcohol has been shown to chronic alcoholism, acute intoxication and chronic meta-
induce anapyrexia,65 a finding which might explain why bolic diseases, as well as psychiatric disorders, are contribu-
drunken hypothermic victims are often lucid at tempera- tory causes in this form of hypothermia.
tures where sober individuals are unconscious. Hypothermia is a common complication in patients
It is of utmost important to recognise this phenomenon undergoing major surgery. The anaesthetic agents interfere
to avoid misinterpretation of the cause. If a person in a with thermoregulation by reducing the metabolic rate67,68
cold environment is found partly undressed or with severe and depressing the thermostatic reflexes. Muscle paralysis
disordered clothing, especially a woman or child, the police abolishes muscle tone and eliminates shivering,69 and fur-
will often interpret the situation as the result of a sexual thermore anaesthesia nullifies the sympathetic vasocon-
attack. The police officer is not competent to diagnose strictory response to cold. A fall in central core temperature
death and most often cordons the area in order to collect is common during surgery. All patients undergoing major
evidence. Not even for a medical practitioner is it always abdominal surgery and no less than two-thirds of patients
possible to determine whether such a victim is dead or not. undergoing minor abdominal surgery are found to become
The patient should, however, be brought to a hospital as hypothermic.70
soon as possible. Hypothermia is an ominous complication in patients
who have sustained major trauma. The critically injured
patient is unable to maintain body temperature in the
Different forms of hypothermia immediate postinjury period. Heat loss may be exacerbated
by infusion of unwarmed fluids and by removal of clothing.
There are many classifications of hypothermia, but from a Patients in shock who become hypothermic have a higher
practical point of view the three subdivisions given in the mortality than do normothermic victims.71,72 The critical
box below are useful. temperature seems to lie around 32–33 °C. Severely trau-
matised patients whose core temperature falls below 32 °C
have even been considered non-salvageable.73 The duration
Forms of hypothermia
of environmental exposure is a primary factor in accidental
hypothermia. A heated rescue vehicle, heated intravenous
• Accidental hypothermia
fluids and warm blankets can reduce the amount of initial
– Acute immersion hypothermia
heat loss. A heated trauma room at the accident and emer-
– Subacute exhaustion hypothermia
gency department, where the victim is promptly undressed
• Subclinical chronic hypothermia
for multiple examination, is of importance in this context.
• Hypothermia of major surgery and trauma
Evaporative heat losses can be restricted by avoiding volatile
solutions for skin cleansing. Warming and humidification
Accidental hypothermia is a major cause of death in of inspired gases is also important.
people engaged in outdoor recreational activities. Acute
immersion hypothermia occurs when a person falls into
Treatment
cold water. This medium has a specific heat 4000 times that
of air and a thermal conductivity approximately 25 times
IN THE FIELD
greater.66 The core temperature, therefore, decreases rap-
idly even if heat production is maximal. Hypothermia sets The main principle of primary care of a patient suffering
in before the victim becomes exhausted. from hypothermia is to prevent further heat loss. If the
Accidental subacute exhaustion hypothermia happens patient is conscious, they should be moved indoors or at
to skiers, climbers and walkers in the mountains. Muscular least into shelter. They should be kept in a lying position
activity maintains the body temperature as long as energy and wet clothing should be removed. Now and then there
sources are available, but when hypoglycaemia ensues, the may be difficulties in undressing a wet, stiff victim and if
victim is in a hazardous situation. Furthermore, prolonged that is the case, it would be appropriate to put that individ-
exposure to cold induces increased loss of body fluids due ual into a plastic bag, ensuring as much insulation as pos-
to cold diuresis, augmented insensible loss of water from sible and remembering to cover their head. A fire may be
 Hypothermia 583

built or a stove lit, but the patient should not be exposed the patient is in a stable environment and the procedure
close to its radiant heat which may cause a burn injury. A can be performed reasonably and continuously. In nor-
conscious patient who has been well cared for can be given mothermia, CPR induces blood flow from phasic alter-
a warm, nourishing beverage. Transportation to the near- ations75 in the intrathoracic pressure rather than from
est hospital should then be arranged. cardiac compression. Althaus et al.76 reported three cases
If the victim is comatose from hypothermia then this is of prolonged hypothermic cardiac arrest without neuro-
a ‘true medical emergency, but an emergency in slow logical sequelae after prolonged closed chest compression.
motion’.74 People die slowly in the cold. Mills has likened In one patient at 22 °C they found the heart to be hard as
the hypothermic victim to a ‘metabolic icebox’ where the stone and impossible to compress. The role of a ‘thoracic
low temperature to a large degree protects certain core pump’ with the heart as a passive conduit seems to be a
organs.26 It is mandatory to handle patients gently to min- plausible hypothesis in hypothermia as well.77
imise the risk of possible ventricular fibrillation (VF). The maxim ‘No one is dead until warm and dead’, when
Mouth-to-mouth respiratory support should be tried in dealing with a hypothermic victim, should be completed
apnoeic patients. The positive pressure often produces with ‘unless serum potassium is greater than 10 mmol/L’
gasping respiration, which indicates the presence of some because severe hyperkalaemia carries an adverse outcome.
cardiac function.48 Furthermore a core temperature of less than 12 °C must
If available, the unconscious victim should be given also be considered low enough to declare a person dead.
warm, moist oxygen and an intravenous infusion of glucose Thermometers for measuring tympanic temperature are
at 37–40 °C. Electrocardiographic (ECG) monitoring is an now available for use in the field.78,79 Resuscitation under
advantage in the prehospital setting. One must be aware, such circumstances has proved to be a waste of resources.
however, that adhesive pads for monitor leads will not stick
to the skin and needle electrodes may have to be inserted.
Shivering often obscures the interpretation of the ECG. Hypothermia: treatment in the field
The scenario out in the field generally lacks all these
facilities. The primary treatment on these occasions is to • Move patient into sheltered area and place in supine
insulate the victim and attempt to re-warm them; this position
often comes down to minimising further decreases in tem- • Handle gently to avert VF
perature. Do not try to prewarm a patient in chronic • Avoid neck pressure to detect pulse as it may trigger
hypothermia out in the field without physiological control. vagus nerve
Transportation of a hypothermic person is far less risky • Mouth-to-mouth support if apnoeic
than enthusiastic attempts to re-warm in the field where it • CPR for cardiac arrest but avoided if situation
is difficult to treat complications. threatens lives of rescuers
In contrast, when hypothermia is due to cold water • CPR contraindicated if signs of life present as chest
immersion, the cooling occurs rapidly and the extreme compression may induce VF
metabolic derangements seen in exhaustion hypothermia • Remove wet clothing, if impossible use plastic bag
are absent. This set of circumstances provides the only situ- and ensure good insulation
ation in which the victim may be treated with rapid immer- • If conscious give warm nourishing beverage
sion re-warming, even out in the open, if that is feasible. • Light fire or stove but do not expose to its radiant heat
It is often difficult, even for trained medical personnel, • If available warm moist oxygen, intravenous glucose
to determine whether or not a hypothermic person is at 37–40 °C, ECG monitoring (needle electrodes)
alive. Apparent cardiovascular collapse may actually be • Rapid immersion re-warming only if cold water
represented only by depressed cardiac output. Palpation or immersion is the cause
auscultation for at least a minute to detect spontaneous • Arrange transportation to hospital
pulses may be necessary. Pressure on the neck, which may
trigger the vagus nerve, should be avoided. The femoral
arteries in the groin should be checked instead.
HOSPITAL MANAGEMENT
In the field, the decision as to whether administer
cardiopulmonary resuscitation (CPR) should be adminis- If not previously instituted, warm, humidified oxygen
tered, is difficult. The problem is more often governed by should be administered by face mask and, after preoxy-
the situation rather than medical circumstances. If there is genation, by endotracheal tube. Continuous core tempera-
any sign of life at all, CPR is contraindicated, as prema- ture (T°) and ECG monitoring should be instituted. An
turely performed chest compressions may induce VF. intravenous infusion of isotonic dextrose in saline at
Cardiopulmonary resuscitation is not indicated when 35–45 °C should be given. Solutions containing lactate are
there is a major threat to the rescuers, but it should be ini- unsuitable as the liver cannot metabolise lactate at lower
tiated immediately after a cardiac arrest is witnessed, when temperatures. Colloids should only be administered to
584 Cold injury

patients not responding to crystalloids. A central venous compensated. Passive re-warming is physiologically sound.
pressure (CVP) catheter is often useful for registering fluid It avoids the rapid cardiovascular derangements often
overload and imminent pulmonary oedema. Laboratory encountered in patients in chronic hypotension when
evaluations should include arterial blood gases (ABGs) treated with active methods. Passive re-warming is often
uncorrected for body temperature, full blood and platelet recommended for elderly patients with intravascular vol-
counts (FBP), blood sugar, electrolyte (Es), creatinine (Cr), ume contraction.82
serum calcium, serum magnesium and serum amylase
levels, prothrombin (PT) and partial thromboplastin times
ACTIVE EXTERNAL RE-WARMING
(PTT) and fibrinogen (F) levels. A toxicology (Tox) screen,
thyroid function tests (TFTs) and cardiac isoenzymes (CIs) Patients with temperatures below 32 °C require more active
should be considered if the level of unconsciousness does re-warming by direct transfer from an exogenous source to
not correlate with the actual core temperature. Remember the surface of the body.
also that a warm skin on a cold patient is possibly due to a Examples of achieving this objective include electric
vasodilatatory drug or sepsis.80 re-warming blankets, heating pads, radiant heat sources and
Re-warming can be performed with different tech- trunk immersion. It has been stressed that the re-warming
niques, passive and active, depending on whether heat is or should be confined to truncal areas and too rapid a
is not added. Active re-warming may be delivered by exter- re-warming of the extremities with its consequent ‘after-
nal means or by core re-warming. Each has its advantages drop’ should be avoided. This afterdrop is a paradoxical
and disadvantages. The method used will often depend on decrease in core temperature of 2–3 °C occurring soon
local factors. The accessibility of resources can differ after the start of shell re-warming.82 This complication is
widely, between, for example, a hypothermic skier found believed to be initiated by peripheral vasodilatation and
in the mountains who may be many miles from a well- shunting of stagnant, cold, acidotic blood to the core, thus
equipped hospital, and in contrast, an urban, homeless further chilling the heart and possibly inducing potentially
alcoholic found in the street. fatal ventricular fibrillation. The significance of this state-
ment has been disputed. Studies on humans have shown
afterdrop, affecting both oesophagus and rectum, with-
Hypothermia: hospital management out any increased blood flow in the extremities.83 The
afterdrop phenomenon may therefore merely be due to
• Warm humidified oxygen by face mask/ the slow conductive heat transfer from the shell to the core
endotracheal intubation during aggressive re-warming.84
• Monitoring of core temperature (T°), ECG, CVP In active surface re-warming, however, the problem of
• Intravenous isotonic dextrose in saline at 35–45 °C plasma volume depletion may be exacerbated by periph-
• Blood for ABGs, FBP, sugar, Es, Cr, calcium, eral vasodilatation, resulting in hypotension. Furthermore,
magnesium, amylase, PT, PTT, F CPR is difficult to perform when the victim is immersed.
• If level of unconsciousness does not correlate with The hot bath is not recommended for the elderly.85,86 The
core T°, Tox screen, TFTs and CIs method has been advocated when hypothermia is com-
• Re-warming – passive, active (external or internal) bined with local frostbite.26
The Royal Danish Navy has adopted a field model of
re-warming in mild to moderate hypothermia proposed by
Vanggard and Gjerlof.87 It is a simple technique supplying
PASSIVE RE-WARMING
exogenous heat by immersing hands, forearms, feet and
Re-warming without adding heat takes place as a result of lower legs in water at 44–45 °C. When the distal extremities
metabolic heat created by shivering. Measures such as are warmed, the arteriovenous anastomoses in the fingers
removing wet clothes and covering the patient in layers of and toes are opened which greatly increases the venous
warm blankets are less effective at lower temperatures, as return to the heart via the superficial venous route in the
shivering ceases at around 30 °C. Passive rewarming is forearms and lower legs. This raised perfusion of the skin
the method of choice in haemodynamically stable and surface enhances the delivery of heat to the core with min-
otherwise healthy patients with a core temperature above imal countercurrent heat exchange as the superficial veins
30–32 °C.81 The rate of rewarming varies greatly with are not in close proximity to the arteries. The hypothesis
most series, reporting increases of around 0.5 °C per hour; has recently been tested on volunteers cooled in water to
the lower the body temperature, the slower the rate. If around 35 °C. The postcooling afterdrop in oesophageal
the rise in body temperature is much slower, a compli- temperature after such a treatment was decreased com-
cating disease such as hypothyroidism should be sus- pared with the shivering alone procedure.88 The extremity
pected. Generation of heat requires energy and will quickly immersion rewarming method does not require any spe-
deplete muscle glycogen stores, and this loss must be cialised equipment or expertise and can be performed in
 Hypothermia 585

the field and on board smaller ships without sophisticated consequent risk of depressed pulmonary and cardiac func-
means of heat donation. tion. Pleural lavage may also be used in combination with a
This model of extremity re-warming has been refined thoracotomy for open cardiac massage in a haemodynami-
and tested on hypothermic subjects during recovery from cally unstable, severely hypothermic patient.76
general anaesthesia.89 The re-warming was performed by a Gastric and/or colonic re-warming is used less often
water-perfused blanket (45–46 °C). In a test group the nowadays. Anatomical proximity to the heart and to the
blood vessels in the hand were distended by exposing the liver is an advantage in gastric lavage but the actual surface
distal part of the arm to subatmospheric pressure (30 area for heat exchange is small.90 Tube insertion may cause
to 40 mmHg). Compared with controls this proced- ventricular fibrillation and prolonged lavage causes severe
ure, with maximum vasodilatation, resulted in a 10-fold electrolyte imbalance. In order to prevent the latter disad-
increase in re-warming rates. This promising method has vantage, a modified Sengstaken tube has been used.92
been confirmed on cold-stressed (36 °C) healthy volun- Diathermy is a technique in which core temperature is
teers.90 The combined application of heat and subatmos- increased by using ultrasonic, short wave or microwave
pheric pressure increases local subcutaneous blood flow irradiation. The great advantage is that the re-warming takes
thereby allowing heat to be transferred subcutaneously place below the skin and subcutaneous tissues without
directly from the skin of the extremity to the critical body requiring any invasive procedure. Even though studies in
core despite the central drive for vasoconstriction.87–89 dogs have revealed a re-warming speed comparable to that
of peritoneal lavage,93 reports on its clinical use are anec-
dotal.94 There are numerous contraindications to this form
Hypothermia: active external re-warming of re-warming, for example haemorrhage, malignancies,
pregnancy, tuberculosis and implanted devices such as
arthroplasty and pacemakers, most of which are difficult to
• Re-warm trunk only (to avoid afterdrop) – electric
identify in a comatose patient.90
blankets, heating pads, radiant heat sources
Extracorporeal re-warming by means of CPB and/or
• Hot bath immersion – contraindicated in the elderly
haemodialysis is the most reliable way to maintain cardio-
and if CPR required
vascular function in severely hypothermic patients, especially
• Extremity re-warming (can be effected with water-
when cardiac activity is absent. The bypass is performed by
perfused blanket 45–46 °C)
cannulation of one or both femoral veins. The blood is oxy-
genated, warmed and returned by a mechanical pump to the
femoral artery.95,96 Cardiopulmonary bypass re-warms at
ACTIVE INTERNAL RE-WARMING
four times the rate of conventional core re-warming.97 It is
In comatose victims active internal re-warming should superior to other re-warming techniques because it possesses
take precedence over external methods. A variety of tech- the ability to oxygenate the blood and provides haemo-
niques are available for this, the most invasive type of dynamic support even in total circulatory failure. The method
treatment of hypothermia. These options include irriga- rapidly increases myocardial temperature and improves
tion of body cavities, gastric and/or colonic lavage, airway microcirculatory flow.98
re-warming, diathermy and extracorporeal blood rewarm- The drawbacks of CPB are the risk of heparinisation,
ing or cardiopulmonary bypass (CPB). haemolysis, arterial injury and air embolism. Significant
Peritoneal lavage is most widely used because it is prac- trauma constitutes the main contradiction to CPB, owing to
tical, inexpensive and easy to administer at any medical anticoagulation. A heparin-coated bypass is now available
facility.90,91 The method involves lavage using a dialysate of and seems to be preferable even in non-traumatised patients
warmed standard fluids (40–42.5 °C) which is best per- as systemic anticoagulation will exacerbate the coagulopathy
formed via two catheters and allows fluid values up to induced by hypothermia.99 Furthermore, this technique is
10–12 L/hour. This method also facilitates the manage- available only in larger hospitals. Endothelial leakage during
ment of drug intoxication and renal failure, and improves reperfusion needs massive volume supplementation during
liver function and drug detoxification by directly re-warming bypass. Patients with simultaneous deep frostbite may
the liver. Contraindications to peritoneal irrigation include develop compartmental syndromes requiring fasciotomy.100
abdominal trauma and recent abdominal surgery. In smaller units, an interesting option for extracorporeal
Pleural lavage has the advantage of re-warming the heart re-warming has been reported by Gregory et al., using a veno-
directly. Two thoracotomy tubes are inserted, one in the venous rewarmer.98 The blood is removed through a central
fifth intercostal space in the posterior axillary line and the venous catheter, warmed to 40 °C through a warming col-
other in the second or third intercostal space in the midclav- umn and returned to a second central or peripheral venous
icular line.90,91 A flow rate up to 0.5 L/min can be achieved. catheter at a flow rate of 150–400 mL per minute. The device
An adequate outlet for the fluid, however, is mandatory can be set up in about 10 minutes and without the subse-
in avoiding increasing intrathoracic pressure, with the quent use of heparin.
586 Cold injury

phaeochromocytomas and in connection with severe

Hypothermia: active internal re-warming head trauma.104,105 Pulmonary oedema is mediated via a mas-
sive sympathetic discharge, producing an intense but tran-
• Irrigation of body cavities
sient vasoconstriction with a resultant shift of blood from
– Peritoneal lavage – best in non-arrests, detoxifies,
the high resistant systemic circulation to the low resistant
improves renal and liver function
pulmonary circulation. Catecholamine levels in the uncon-
– Pleural lavage – also re-warms heart directly
scious victim are low during re-warming, but noradrenaline
– Gastric/colonic lavage – dangers are tube induced
and adrenaline levels rise rapidly, indicating some kind of
VF and electrolyte imbalance
depot effect.106,107 This kind of oedema is well controlled
• Airway re-warming – using hot, moist gas
by positive end-expiratory pressure.
– Diathermy – ultrasonic, short wave or microwave;
these have numerous contraindications
• Extracorporeal re-warming
Conclusions
– Cardiopulmonary bypass – femoral vein(s) to
femoral artery. Fast and reliable if hypothermia
severe and no cardiac activity, oxygenates blood, When dealing with cold injury it is important to distin-
gives haemodynamic support; but it has drawbacks guish between the different aetiologies. In the early
– Veno-venous – from CVP line via warming stages of cold injury physical signs are unreliable and the
column and back into a peripheral vein relevant history from the patient or witnesses is there-
fore the key to treatment. It is only when the likely the
• Amino acid infusions
aetiology has been established that specific treatment
can begin.

The use of amino acids is a new, interesting contribu-

tion to the re-warming of hypothermic patients. A bal-
anced mixture of amino acids has been proven to prevent
anaesthesia induced hypothermia and shivering. Three- Key references
quarters of the heat production occurred in extrasplanch-
Althaus V, Aeberhard P, Scheupback P, et al. Management of
nic tissue, possibly partly in the muscles. These findings are
prolonged accidental hypothermia with cardiorespiratory
of importance when dealing with accidental hypothermia. arrest. Ann Surg 1982; 195: 492–5.
Amino acid infusions ought to be included in the thera- Francis TJR, Golden FStC. Non-freezing cold injury: the
peutic arsenal.101 pathogenesis. J R Nav Med Serv 1985; 71: 3–8.
Of course, many of the techniques described can be Hamlet MP. Human cold injuries In: Pandolf KB, Sawka MN,
used synchronously. Air re-warming with inhalation of Gonzales RR (eds). Human Performance Physiology and
hot, moist gas and administration of warmed intravenous Environmental Medicine at Terrestrial Extremes. Indianapolis:
fluids should be a matter of course in every hypothermia Buchmark, 1988: 435–66.
case. For patients with a non-arrest rhythm, peritoneal irri- Jolly BT, Ghezzi KT. Accidental hypothermia. Emerg Med Clin North
gation is often preferable. It usually takes around 1 hour at Am 1992; 10: 311–27.
best to start up CPB, and during that time the abovemen- Mills WJ Jr. Out in the cold. Frostbite. A discussion of the problem
and a review of an Alaskan experience. Alaska Med 1993; 35:
tioned options should be instigated and combined with
29–66.
CPR if the patient shows haemodynamic instability.
Patients suffering from hypothermia, especially of the
chronic exhaustion type, are dehydrated. The combination
of cold diuresis, peripheral vascular leakage and increased
REFERENCES
losses from skin and airways while fighting the cold in the
cold dry air makes these losses greater than would be
expected. Intravenous infusions of warm solutions are 1 Ward M. Frostbite. BMJ 1974; 1: 67–70.
more than necessary, but one must pay attention to CVP. 2 Grattan HW. General History of Medical Services Hyiene of War.
Pulmonary oedema is a frequent complication in elderly Vol II. London: HMSO 1931.
patients at temperatures lower than 27 °C. This hazardous 3 Whayne T, DeBakey M. Cold injury ground type. Washington, DC:
US Government Printing Office, 1958.
event is usually encountered during re-warming. The aeti-
4 Spray DC. Cutaneous temperature receptors. Annu Rev Physiol
ology of this non-cardiac oedema is debatable: decreased 1986; 48: 625–9.
ciliar activity,102 destruction of surfactant from anoxic 5 Francis TJR. Non-freezing cold injury: a historical review. J R Nav
pulmonary damage103 and circulatory elements have all Med Serv 1984; 70: 134–9.
been incriminated as possible agents. Pulmonary oedema 6 Le Blanc J. Effect of reserpine on increased sensitivity to
in connection with decreased cardiac output is also noradrenaline of cold adapted rats. Am J Physiol 1966; 204:
found in hypertensive crises in patients suffering from 520–2.
 References 587

7 Jansky L. Non-shivering thermogenesis and its thermoregulatory 31 Welch GS. Frostbite. Practitioner 1974; 213: 801–4.
significance. Biol Rev 1973; 48: 85–132. 32 Golding MR, De Jong P, Sawyer PN, et al. Protection from early
8 Smith DJ Jr, Robson MC, Heggers JP. Frostbite and other cold- and late sequelae of frostbite by regional sympathectomy:
induced injuries. In: Auerbach PS, Geehr EC (eds). Management mechanism of ‘cold sensitivity’ following frostbite. Surgery 1963;
of Wilderness and Environmental Emergencies. St Louis, MO: CV 53: 303–8.
Mosby, 1988: 101–18. 33 Porter J, Lindell T, Leung B, et al. Effect of intra-arterial injection
9 Lewis T. Observation upon the reaction of vessels of the human of reserpine on vascular wall catecholamine content. Surg Forum
skin to cold. Heart 1930; 15: 177–208. 1972; 23: 183–5.
10 Vogel JE, Dellon AL. Frostbite injuries of the hand. Clin Plastic 34 Snider RL, Porter JM. Treatment of experimental frostbite with
Surg 1989; 16: 265–76. arterial sympathetic blocking drugs. Surgery 1975; 77: 557–61.
11 Merryman HT. Tissue freezing and local cold injury. Physiol Rev 35 Artursson G, Hamberg M, Johnsson CE. Prostaglandin in human
1957; 37: 233–51. burn blister fluid. Acta Physiol Scand 1973; 87: 270–6.
12 Mazur P. Causes of injury in the frozen and thawed cells. 36 Wrenn K. Immersion foot. A problem of the homeless in the
Federation Proc 1965; 24(suppl 14–15): 3. 1990s. Arch Intern Med 1991; 151: 785–8.
13 Bourne MH, Piepkorn MW, Claryton F, Leonard LG. Analysis of 37 Francis TJR, Golden FStC. Non-freezing cold injury: the
microvascular change in frostbite injury. J Surg Res 1985; 40: pathogenesis. J R Nav Med Serv 1985; 71: 3–8.
26–35. 38 Endrich B, Hammersen F, Messmer K. Microvascular
14 Marzella L, Jesudas RR, Manson PN, et al. Morphological ultrastructure in non-freezing cold injuries. Res Exp Med 1990;
characterization of acute injury to vascular endothelium of skin 190: 365–79.
after frostbite. Plastic Reconstr Surg 1989; 83: 67–75. 39 Iyengar J, George A, Russel JC, Das K. Generation of free
15 Manson PL, Jesudass R, Marzella L, et al. Evidence for an early radicals during cold injury and rewarming. Vasc Surg 1990;
free radical mediated reperfusion injury in frostbite. Free Radic 24: 467–74.
Biol Med 1991; 10: 7–11. 40 Beuttner KJK. Diffusion of liquid water through human skin.
16 Daum PS, Bowers WD Jr, Tejada J, et al. Cooling of heat of fusion J Appl Physiol 1959; 14: 261–8.
(HOF), followed by rapid rewarming, does not reduce the 41 Schenplein RJ, Blank IH. Permeability of the skin. Physiol Rev
integrity of microvascular corrosion casts. Cryobiology 1991; 28: 1971; 51: 702–47.
294–301. 42 Blackwood W. Injury from exposure to low temperature:
17 Wheatherly-White RCA, Sjöström B, Paton BC. Experimental pathology. Br Med Bull 1944; 4: 138–41.
studies in cold injury: The pathogenesis of frostbite. J Surg Res 43 Denny-Brown D, Adams RD, Brenner C, Doherty MM. The
1964; 4: 17–22. pathology of injury to nerve induced by cold. J Neuropathol Exp
18 Robson MC, Heggers JP. Evaluation of hand frostbite blister fluid Neurol 1945; 4: 305–23.
as a clue to pathogenesis. J Hand Surg 1981; 6: 43–7. 44 Kennet RB, Gilliat DM. Nerve conduction studies in experimental
19 McCauley RL, Hing DN, Robson MC, Heggers JP. Frostbite injury: non-freezing cold injury. I. Local nerve cooling. Muscle Nerve
a rational approach based on the pathophysiology. J Trauma 1991; 14: 553–62.
1983; 23: 143–7. 45 Nukada H, Dych PJ. Acute ischemia causes axonal stasis,
20 Knize DM. Cold Injury in Reconstructive Plastic Surgery, swelling, alteration and secondary demyelination. Ann Neurol
General Principles, Vol. 1, 2nd edn. Philadelphia, PA: WB 1987; 22: 311–18.
Saunders, 1977. 46 Kennet RB, Gilliat DM. Nerve conduction studies in experimental
21 Dinep M. Cold injury: a review of current theories and their non-freezing cold injury. II Generalized nerve cooling by limb
application to treatment. Medicine 1975; 39: 8–10. immersion. Muscle Nerve 1991; 14: 960–7.
22 Artursson G. The tragedy of San Juaurico – the most severe LPG 47 Emmelin N, Trendelenburg V. Degeneration activity after
disaster in history. Burns 1987; 13: 87–102. parasympathetic and sympathetic denervation. In: Reviews of
23 James NK, Moss ALH. Cold injury from liquid propane. BMJ 1989; Physiology. Berlin: Springer Verlag, 1972.
299: 950–1. 48 Hamlet MP. Human cold injuries In: Pandolf KB, Sawka MN,
24 Wegener EE, Barraza KR, Das SK. Severe frostbite caused by freon Gonzales RR (eds). Human Performance Physiology and
gas. South Med J 1991; 84: 1143–6. Environmental Medicine at Terrestrial Extremes. Indianapolis:
25 Hicks LM, Hunt JL, Baxter CL. Liquid propane cold injury: a Buchmark, 1988: 435–66.
clinicopathologic and experimental study. J Trauma 1979; 19: 49 Akers WA. Paddy foot: a warm water immersion foot syndrome
701–3. variant. I. The natural disease, epidemiology. Military Med 1974;
26 Mills WJ Jr. Out in the cold. Frostbite. A discussion of the 139: 605–12.
problem and a review of an Alaskan experience. Alaska Med 50 White JC, Warren C. Causes of pain in foot after prolonged
1993; 35: 29–66. immersion. War Med 1944; 5: 6–13.
27 Vanggard L. Physiological reactions to wet cold. Aviat Space 51 Ahle NW, Buroni JR, Mark W, et al. Infrared thermographic
Environ Med 1975; 10: 235–41. measurement of circulatory compromise in trench foot – injured
28 Renström B. Brrr … klarar kroppen kampen mot kylan. Quantum Argentine soldiers. Aviat Space Environ Med 1990; 61: 247–50.
Satis 1991; 4: 24–7. 52 Burton AC. The pattern of response to cold in animals and the
29 Edlic RF, Chang DE, Birk KA, et al. Cold injuries. Compr Ther 1989; evolution of homeothermy. In: Hardy JD (ed). Temperature: its
15: 13–21. Measurement and Control in Science and Industry, vol. 3. New
30 Webster DR, Bourn G. Low molecular weight dextran in the York: Book Corporation, 1963: 363–7.
treatment of experimental frostbite. Can J Surg 1965; 8: 53 Froese G, Burton AC. Heat loss from the human head. J Appl
423–7. Physiol 1957; 10: 235–41.
588 Cold injury

54 Hayward JS, Eckerson JD, Kemna D. Thermal and cardiovascular 78 Walpoth BH, Galdiskas J, Leupi F, et al. Assessment of
changes during three different methods of resucitation from hypothermia with a new ‘tympanic’ thermometer. J Clin Monit
mild hypothermia. Resucitation 1984; 11: 21–33. 1995; 10: 91–6.
55 Lennqvist S, Granberg PO, Wedin B. Fluid balance and physical 79 Helm M, Lampl L, Hauke J, Bock KH. Akzidentelle Hypothermie
working capacity in humans exposed to cold. Arch Environ bei Traumapatient. Anaesthesist 1995; 44: 101–7.
Health 1974; 20: 241–9. 80 Fitzgerald FT, Jessop C. Accidental hypothermia: a report of
56 Hamlet MP. Fluid shifts in hypothermia. In: Pozoz RS, Wittmers 22 cases and review of the literature. Adv Med 1982;
LE (eds). The Nature and Treatment of Hypothermia. 27: 150–3.
London/Minneapolis, MN: Croom Helm/University of Minnesota 81 Ledingham IMcA, Mone JG. Treatment of accidental
Press, 1983: 94–9. hypothermia. A prospective study. BMJ 1980; 280: 1102–5.
57 Treviko A, Razi B, Beller BM. The characteristic electrocardiogram 82 Miller JW, Danzl DS, Thomas DM. Urban accidental
of accidental hypothermia. Arch Intern Med 1971; 127: hypothermia: 135 cases. Ann Emerg Med 1980; 9: 456–61.
470–3. 83 Savard GK, Cooper KE, Veale WL, et al. Peripheral blood flow
58 Fergusson NV. Urban hypothermia. Anaesthesia 1985; 40: during rewarming from mild hypothermia in man. J Appl Physiol
651–54. 1958; 58: 4–13.
59 Gilbert M, Busund R, Skagseth A, et al. Resuscitation from 84 Webb P. Afterdrop of body temperature during rewarming: an
accidental hypothermia of 13.7°C with circulatory arrest. alternative explanation. J Appl Physiol 1986; 60: 385–90.
Lancet 2000; 355: 375–6. 85 Lloyd EL. Hypothermia and Cold Stress. London: Croom Helm,
60 Nozaki R, Ishibaski K, Adaski N, et al. Accidental proformed 1986.
hypothermia. N Engl J Med 1986; 315: 1680. 86 Moss J. Accidental severe hypothermia. Surg Gynecol Obstet
61 Ehrmanntrant WR, Ticklin HE, Faxerkras JF. Cerebral 1986; 162: 501–13.
haemodynamics and metabolism in accidental hypothermia. 87 Vangaard L, Gjerloff CC. A new simple method of rewarming in
Arch Intern Med 1957; 99: 57–61. hypothermia. Int Rev Navy Air Force Med Serv. 1979;
62 Enander A. Performance and sensory aspects of work 52: 427–30.
in cold environment – a review. Ergonomics 1984; 88 Vanggard DE, Eyolfson D, Xiaojiang X, Weseen G, Giesbrecht
27: 365–78. CG, Immersion of distal arms and legs in warm water (AVA
63 Lloyd EL. Hypothermia and Cold Stress. London: Croom Helm, Rewarming) effectively rewarms mildly hypothermic humans.
1986. Aviat Space Environ Med 1999; 70: 1081–8.
64 Cabanac M, Brinnel H. The pathology of human temperature 89 Grahn D, Brock-Utne J, Watenpaugh DE, Heller HC. Recovery
regulation: thermiatrics. Experimentia 1987; 43: 19–27. from mild hypothermia can be accelerated by mechanically
65 Granberg PO. Alcohol and cold. Arct Med Res 1991; 50(suppl 6): distending blood vessels in the hand. J Appl Physiol 1998;
43–7. 85: 1643–8.
66 Nadel ER. Energy exchanges in water. Undersea Biomed Res 90 Jolly BT, Ghezzi KT. Accidental hypothermia. Emerg Med Clin
1984; 11: 149–52. North Am 1992; 10: 311–27.
67 Burton AC, Edholm OE. Man in a Cold Environment. London: 91 Paton JF. Accidental hypothermia: a matter of turning the
Edward Arnold, 1955. scoreboard around. Med Post (Canada), 1983; 23: 4–5.
68 Mackenzie A. Hazards in the operating theatre; environmental 92 Kristensen G, Gravesen H, Benveniste D, Jordening H. An
control. Ann R Coll Surg 1973; 52: 361–5. oesophageal thermal tube for rewarming in hypothermia. Acta
69 Holdcroft A. Body Temperature Control in Anaesthesia, Surgery Anaesthesiol Scand 1985; 29: 846–8.
and Intensive Care. London: Baillière Tindall, 1981. 93 White JD, Butterfield AB, Greer KA, et al. Controlled comparison
70 Joackimsson PO. Prevention and Treatment of Intraoperative of radiowave regional hyperthermia and peritoneal lavage after
Hypothermia. Acta University Uppsalianis, no. 109. Uppsala: immersion hypothermia. J Trauma 1985; 25: 989–93.
Faculty of Medicine, 1987. 94 Huang Z, Sun QY, Shen MJ. Rewarming with microwave
71 Steineman S, Shackford SR, Davies JW. Implications of irradiation in severe cold injury syndrome. Chinese Med J 1980;
admission hypothermia in trauma patients. J Trauma 1990; 93: 119–20.
30: 200–2. 95 Philips SJ, Zeft RH, Kongathorn C, et al. Percutaneous
72 Luna GK, Maier RV, Paolin EG, et al. Incidence and effect of cardiopulmonary bypass: applications and indication for use.
hypothermia in seriously injured patients. J Trauma 1987; Ann Thorac Surg 1989; 47: 121–3.
27: 1014–18. 96 Kugelberg J, Schuller H, Bey B, Kallum B. Treatment of
73 Jurkovick GJ, Greisen WB, Luterman A, Curreri PW. accidental hypothermia. Scand J Thorac Cardiovasc Surg 1967;
Hypothermia in trauma victims: an ominous predictor of 1: 142–6.
survival. J Trauma 1987; 27: 1019–24. 97 Splittgerber FH, Talbert JG, Sweezer WP, et al. Partial
74 Fritz RL, Perrin DH. Cold exposure injuries: prevention and cardiopulmonary bypass for core rewarming in profound
treatment. Clin Sports Med 1989; 8: 111–29. accidental hypothermia. Am Surgeon 1986; 52: 407–12.
75 Neimann JT. Blood flow without cardiac compression during 98 Gregory JS, Bergsten JM, Aprehamiam C, et al. Comparison
closed chest CPR. Crit Care Med 1981; 9: 380–3. of three methods of rewarming from hypothermia:
76 Althaus V, Aeberhard P, Scheupback P, et al. Management of advantage of extracorporeal blood warming. J Trauma 1991;
prolonged accidental hypothermia with cardiorespiratory arrest. 31: 1247–51.
Ann Surg 1982; 195: 492–5. 99 von Segessen LK, Garcia E, Turina M. Perfusion without
77 Danzl DF, Pozos RS. Multicenter hypothermia survey. Ann systemic heparinisation for rewarming in accidental
Emerg Med 1987; 178: 1042–55. hypothermia. Ann Thorac Surg 1991; 52: 560–1.
 References 589

100 Hauty MG, Essig BC, Hill JG. Prognostic factors in severe 104 Baigelman W, O’Brian JC. Pulmonary effect of head trauma.
accidental hypothermia: experience from the Art. Hood tragedy. Neurosurgery 1991; 9: 729–40. Predominantly in extra-
J Trauma 1987; 27: 1107–12. splanchic tissues. Clin Sci 1996; 91: 431–9.
101 Selldén E, Bränström R, Brundin T. Augmented thermic effect of 105 Popp AJ, Shak DM, Berman RA, et al. Delayed pulmonary
amino acids under general anaesthesia occurs predominantly in dysfunction in head-injured patients. J Neurosurg 1981; 57:
extra-splanchnic tissues. Clin Sci 1996; 91: 431–9. 784–90.
102 Håkansson CH, Toremalm NG. Studies on the physiology of the 106 Wayne D. Medical afteraction conference Mount Hood. Bypass
trachea. I. Ciliar activity indirectly recorded by a new ‘light beam rewarming report T 10–88. US Army Research Institute. Natick:
reflex’ method. Ann Otol Rhinol Laryngol 1965; 74: 954–69. US Army Research Institute, 1986.
103 Dubas F. Mountain rescue and treatment of deep accidental 107 Wilkersson JE, Reven PB, Boldman NW, Horwath SM.
hypothermia. In: Cooper KE, Lomax P, Schönbaume, Veale WL Adaptations in men’s adrenal function in response to acute
(eds). Homeostasis and Thermal Stress. Basel: Karger, 1986: 49–54. cold stress. J Appl Physiol 1974; 36: 183–99.
This page intentionally left blank
Index

1-(1-phencyclohexyl)piperidine (PCP) 533 diagnosis 122 AIDS see acquired immune deficiency syndrome
Finland 116 albumin 75, 547
AAA see abdominal aortic aneurysm Sweden 116–17 albuminuria 225
abdominal aorta, imaging 98 acute leg ischaemia 107 ALI see acute limb insufficiency
abdominal aortic aneurysm (AAA) 277 acute limb insufficiency (ALI) 17 Allen test 461, 513
aortocaval fistula 281–2 acute limb ischaemia 22 allergic reactions 239
atheroembolism 367 amputations 177–8 alteplase 241
colonic ischaemia 29 anaesthesia 170–1 see also recombinant tissue plasminogen
consent 279 autologous venous grafts 191–2 activator
diagnosis 122, 277–8 combined endovascular catheter therapy alveolar haemorrhage 403
differential diagnosis 278 185–90 ambulance paramedics 62
horseshoe kidney 282 diagnosis 164–7, 182–3 amphetamines 533–4
inflammatory aneurysms 282 distal procedures 173–4 Amplatz goose neck snare 484
operative technique 279–81 endovascular options 181–94 amputation 165, 381
patient selection 278–9 fasciotomy 176–7 acute arterial inflammatory conditions 528
patient transfer 279 graft failure 204 acute limb ischaemia 177–8, 218
postoperative care 283 graft routing 176 avulsion 444
repair 67–8 inflow procedures 171–2 blunt injury 378
ruptured 39 microtibial embolectomy 176 crushing 444
screening 283 multiple outflow techniques 175 diabetes 223, 230
stenting 283 native arteries 190–1 drug abuse 536
thoracoabdominal aneurysm 282–3 occluded synthetic grafts 191 failed revascularisation 205
United Kingdom 106 portable Doppler ultrasound examination frostbite 577
abdominal arterial injury 439 166–7 guillotine 443
abdominal compartment syndrome 33, 68, 434 postoperative management 178 limb vascular injury 384
abdominal examination 277 surgery 163–78, 481–3 wartime 375
abdominal pathology 283 treatment 167–70, 183 anaemia
abdominal surgery 490–1 withholding treatment 123 chronic 78
HIV 546 wound management 177 HIV 547
abdominal vascular injuries 429–40 acute limb vascular inflammatory conditions anaesthesia
diagnosis 430–1 523–8 AAA 279
management 432 acute mesenteric arterial embolism 336, 338–9 acute limb ischaemia 170–1
pathology 429–30 acute mesenteric arterial thrombosis 336, 339 arm block 498
surgery 432–9 acute mesenteric venous thrombosis 336 carotid endarterectomy 145
abdominal X-ray (AXR) 349 acute renal failure (ARF) 45, 86 HIV/AIDS patients 551
ABG see arterial blood gas acute respiratory distress syndrome (ARDS) 42, replantation 447
ABPI see ankle: brachial pressure index 382, 405 anapyrexia 582
abscess 219, 230, 299, 536 acute thoracic aortic rupture 413–14 anastomoses 390
accidental dissection 484 acute thrombosis 18–21 failure 198
accidental hypothermia 582 acute upper limb ischaemia 495–506 type 1 endoleak 357
acetaminophen 534 aetiology 495–9 anastomotic pseudoaneurysm 301
acetylcysteine 182 diagnosis 499–501 aneurysmal disease 21, 503–4
acidosis 434, 445 endovascular management 505–6 aneurysmectomy 319
acquired immune deficiency syndrome (AIDS) management 501 aneurysms
545 reconstructive surgery 502–5 atherosclerotic 17, 19
acquired thrombophilia 20 treatment 501–2 gastric artery 330
acute aortic dissection 287–96 acute vascular insufficiency 17–24 gastroduodenal artery 330
acute arterial inflammatory conditions 523–8 acute venous inflammatory conditions 528 gastroepiploic artery 330
diagnosis 524–5 Advanced Trauma Life Support (ATLS) HIV 545
management 525–8 protocol 444 imaging 346
pathology 523 AEF see aortoenteric fistulae inflammatory 282
acute ischaemia AFBG see aortofemoral bypass graft mesenteric branch artery 330
Denmark 115–16 afterdrop phenomenon 584 mycotic 325, 537, 538–40
592 Index

aneurysms (cont’d) antithrombin III 69, 237, 497 arteriovenous fistula (AVF) 85, 90, 239, 266,
pancreatic artery 330 anuria 183, 430 380
pancreaticoduodenal artery 330 aortic aneurysm abdominal vascular injuries 431
pregnancy 331 emergency 277–84 arterial catheterisation 460
renal arteries 315–21 stent failure 358 hand ischaemia 499
rupture 352 aortic arch plaque 130 limb vascular injuries 393
thoracoabdominal 282–3 aortic dissection lumbar disc surgery 488
visceral arteries 325–32 classification 287 mycotic aneurysms 539
see also abdominal aortic aneurysm; aortic endovascular stent-graft repair 288–91 puncture related complications 470
aneurysm management 287–8 arteriovenous malformation 299
angel dust 533 pathophysiology 287 arteritis 18, 496
angiographic blush, type 4 endoleak 349 stent-graft placement 294–5 artery forceps 444
angiography aortic isthmus 412–13 arthrodesis 451
abdominal vascular injuries 431 aortic surgery ascending venography 100–1
acute arterial inflammatory conditions 524 epidural analgesia 82 ascites 89
acute limb ischaemia 168 intestinal ischaemia 29–35 ASPI see ankle systolic pressure index
blunt carotid injuries 424 United Kingdom 106 aspiration 134, 536
Buerger’s disease 514 aortic trauma 402, 405 aspirin 137, 528
carotid endarterectomy 67, 142 aortic valve prostheses 496 atherectomy, iatrogenic injuries 379
catheter injury 462 aortitis 21 atheroembolism 312, 367–71
cerebral 132, 425 aortocaval fistula 281–2 management 369
digital subtraction 95–6 aortoenteric fistulae (AEF) 303 pathophysiology 367–8
graft failure 361 aortofemoral bypass graft (AFBG) 214–15 atherosclerosis 18, 21
iatrogenic injuries 379 aortofemoral limb thrombosis 299 HIV 546
limb vascular injury 384 aortofemoral endarterectomy 172 stroke 55
lumbar disc surgery 488 aortography 406 upper limb ischaemia 496
mesenteric 336 APACHE II score 47 atherosclerotic aneurysms 17, 19
portal hypertension 557 APC resistance 238 athero-thromboembolic strokes 129
pseudoaneurysm 461 apoptosis, stroke 57 atherothrombosis 18
VAAs 325 aprotinin 81 athletic injuries 17, 497–8
Angioguard device 370 arch aortogram 406 ATLS see Advanced Trauma Life Support
angioplasty 21 ARDS see acute respiratory distress atrial fibrillation 21, 130, 414, 501
bifurcations 483–4 syndrome atrial myxoma 21
DVT 242 ARF see acute renal failure atriocaval shunt 437
graft failure 204 arm claudication 462, 499 autoimmune diseases 496, 523, 528
inflow occlusion 169 arrhythmias 90 autologous blood transfusion 80–1
mesenteric ischaemia 337 arterial allografts 218 autologous venous grafts 191–2
thrombosis 481 arterial blood gas (ABG) 254 autonomic denervation 224
ULVT 270 arterial bypass graft 163 autonomic neuropathy 224
angioscopy 232 arterial caseload 4 autosympathectomy 526
anhidrosis 224 arterial catheter 17, 21, 498 AVF see arteriovenous fistula
ankle:brachial pressure index (ABPI) 199, 200, injuries 459–65 avulsion amputation 444
228, 383 trauma 505 axillary artery 498
ankle pressure 166, 490 arterial disease, HIV 545–6 axillary ligation 389
ankle systolic pressure index (ASPI) 188 arterial exposure 171 axillary pullout syndrome 498
anterior cervical sympathectomy 526–7 arterial injury 381 axillary puncture 470
antiaggregation 136–7 angiography 384 axilloaxillary crossover bypass 503
antibiotic (rifampin)-bonded prosthetic varicose vein surgery 490 axillofemoral prosthetic graft 498
conduits 218 arterial insufficiency 450 axillo-pectoral compression 262
antibiotics 124, 212, 337 arterial mycotic aneurysms 538–40 AXR see abdominal X-ray
amputation 444, 451 arterial occlusion 461
diabetic foot disease 229–30 arterial oxygen saturation 406 ß-blockers 287, 325, 408, 563
drug abuse 537 arterial reconstruction, diabetic foot disease bacteraemia 176
prosthetic vascular graft infection 298, 302 231–2 bacterial pneumonia 550
anticoagulation arterial repair 389 bacterial translocation 42, 382
mesenteric ischaemia 337, 340 arterial strictures 340 bactericidal/permeability increasing protein
stroke 136–7 arterial surgery (BPI) 48
ULVT 261, 269 adverse events 121 bacteriology 212, 298
anticollagen 512 outcomes 105 balloon angioplasty 217, 319
anticytokine therapies 47–8 arterial thromboembolism 240 balloon catheter thromboembolectomy 517
antielastin 512 arterial thrombosis 540–1 balloon catheters 485
antiendotoxin therapy 48 arterial trauma 163 balloon dilatation 505–6
antiphospholipid antibodies 17, 69, 132, 252 arteriography perforation 474–5
hypercoagulable states 265 graft failure 199–200 balloon embolectomy 164
antiplatelet drugs 517 renal artery dissection 321 balloon tamponade 558
antiretroviral therapy 552 retrograde percutaneous transfemoral 460 Behçet’s disease 325, 497, 528
antithrombin 20 upper limb ischaemia 500–1 bifurcations, angioplasty 483–4
 Index 593

biplanar arteriography 199 calcium infusion 87 central venous catheters 265–6

bladder management 134 calf vein thrombosis 240 central venous pressure (CVP) 406, 584
blast injuries 404 cannabis 523, 532 centralisation of vascular services 3
bleeding 379 cannulation 73 cerebellar haematomas 137
arterial catheterisation 459 capillary basement membrane 225 cerebral angiography 132, 425
chest wall 412 carbon dioxide angiography 231 cerebral blood flow (CBF) 130
endoscopy 557 cardiac catheters 379 cerebral haemorrhage 142
gastrointestinal 241, 340 cardiac compression 402 cerebral hyperperfusion syndrome 158
graft-enteric erosion 299 cardiac failure 337 cerebral infarction 55, 129, 414
hip surgery 489 bypass grafts 205 pathogenesis 56–7
lumbar disc surgery 488 cardiac monitoring, central access 92 see also stroke
portal vein injuries 438 cardiac output 75 cerebral oedema 60, 132, 134
prosthetic graft infection 211 cardiac tamponade 405, 410 cerebral reperfusion 61
surgical treatment 471 cardiac trauma 402 cerebrovascular accidents
see also haemorrhage cardioembolic strokes 129–30 DVT 238
blindness 462 cardiogenic embolism 55 USA 13
blood flow, maldistribution 23 cardiogenic shock 445 cerebrovascular autoregulation 130–1
blood transfusion 77, 434 cardiomegaly 254 cervical bruit 424
high volume 78–9 cardiopulmonary bypass (CPB) 80, 402 cervical haematoma 422
in puncture related complications 471 cardiopulmonary resuscitation (CPR) cervical ribs 497, 499
blue finger syndrome 496 femoral access 89 cervical sympathectomy 506, 526
blue toe syndrome 368, 369, 461, 499 hypothermia 583 cervical vascular injuries 419
blunt injuries 17 cardiovascular system, MODS 44 cervico-mediastinal trauma 402, 405
abdominal 430 carotid angioplasty 141 CFA see common femoral artery
chest 401–2, 408 carotid artery (CA) CFD see colour-flow duplex
limb 377–8 injuries 419 Charcot’s foot 224
neck 420 lesions 141 cherry red spots 557
bone-repair 385 occlusion 464 chest pain 238
bone shortening 451 selective angiography 97 chest vascular injuries 401–16
bone washout 224 carotid endarterectomy 66–7, 141, 143–6 aetiology 401–4
bony stabilisation 448 carotid injuries diagnosis 406–8
bovine thrombin 464 blunt 424–5 investigations 406–8
bowel infarction 326 diagnosis 420–4 operative management 408–15
bowel ischaemia 33, 338, 436 management 422 postoperative care 415
bowel resection 35 surgery 422–3 prehospital management 408
bowel wall oedema 336 carotid stenosis, stroke 56 resuscitation 408
BPI see bactericidal/permeability increasing carotid surgery 106 treatment 408
protein CARS see compensatory anti-inflammatory chest wall tenderness 253
brachial artery response syndrome chest wall vessels, trauma 403, 412
embolectomy 502–3 caseloads, Denmark 8 cholecystectomy 379, 490
ligation 389 catalase (CAT) 22 cholesterol embolisation 478–9
occlusion 464 catecholamines 581 chronic anaemia 78
brachial plexus paralysis 460 catheter directed thrombolysis (CDT) 168, 239, chronic large artery disease 499
brachial puncture 470 240–2 chronic liver failure 556
brain catheter injury 319 chronic mesenteric occlusive disease 326
necrosis 57 microembolisation 478 chronic obstructive airways disease (COAD)
reperfusion 59–60 perforation 473–4 279
see also cerebral … prevention 462–3 chronic occlusions 271
breast atrophy 556 catheter therapy 185–90 chronic renal insufficiency 85–6
breast cancer 379 catheterisation 21, 73, 265 cirrhosis 336, 555–6
bruit 499 iatrogenic vascular injuries 487 CIVD see cold induced vasodilatation
cervical 424 renal artery 309 clamp and sew technique 414, 415
Budd–Chiari syndrome 267, 555 catheters 87–8 claudication 202, 226, 517, 524
Buerger’s disease 497, 501, 511–20, 523 causation 122 clawed toes 223
acute limb ischaemia 168 caval thrombus 103, 242 clindamycin 230
aetiology 511–12 CBF see cerebral blood flow clinical audit 125
diagnosis 513–14 CBP see cardiopulmonary bypass clonidine 134
pathology 512–13 CD4 cell count 550 clopidogrel 136
thrombosis 19 CDT see catheter directed thrombolysis CMV see cytomegalovirus
treatment 517–18, 525 cell death 57 COAD see chronic obstructive airways disease
bullet wounds 377 see also apoptosis coagulopathy
cell salvage 80 abdominal vascular injuries 434
C-reactive protein (CRP) 525 cellular oedema 60 blood transfusion 79
CA see carotid artery cellulitis 211, 232, 382, 451, 536 MODS 44
Caesarean section 491 central access 88 cocaine 531, 532–3, 541
calcification 473 central thrombus 103 Cockett–Thomas syndrome 242
calcium channel blockers 517, 563 central vein thrombosis 92 coeliac artery aneurysm 326, 329–30
594 Index

coeliac axis injuries 436 C-reactive protein (CRP) 525 non-invasive arterial evaluation 228–9
cold exposure 525 creatine phosphokinase (CPK) 369 physical examination 227–8
cold induced vasodilatation (CIVD) 574 creatinine clearance 46 treatment 229–31
cold injury 573–86 Crohn’s disease 497, 501 diagnosis related groups (DRGs) 11
collaterals cross-clamping 32, 414 dialysis 8, 499
compromised 481–3 crossmatching 77 diathermy in hypothermia 585
ULVT 268 CRP see C-reactive protein DIC see disseminated intravascular coagulation
colloids 75, 583 crural anastomosis 173 diffusion-weighted magnetic resonance imaging
colonic ischaemia 29, 32, 35, 43, 280 crural arteries 389 (DWI) 58, 142
colonoscopy 34 crush injury 378, 430 digital contractures 223
colostomy 35 crushing amputations 444 digital ischaemia 496
colour-flow duplex (CFD) 14 crutches 498 digital necrosis 499
combined endovascular catheter therapy cryoglobulinaemia 20, 497 digital subtraction arteriography (DSA) 95–6
185–90 crystallisation 574 acute limb ischaemia 182
common femoral artery (CFA) crystalloids 75 carotid endarterectomy 150
endarterectomy 214 CT see computed tomography endoleak 350
puncture related complications 469–70 CTA see computed tomographic angiography upper limb ischaemia 501
compartment syndrome 21, 382 CVP see central venous pressure digital ulcers 513
compartmental hypertension 395 cyanosis 165, 253, 257, 368, 524, 534 dilzem 518
compartmental pressure 237 cyclophosphamide 519 dis-incorporation of infected graft 298
compensatory anti-inflammatory response cystic adventitial necrosis 497 dissecting renal artery aneurysms 319–21
syndrome (CARS) 41 cystic medial necrosis 18 dissection
complement inhibition 48 cytokines 23, 42, 382 accidental 484
component therapy 79–80 cytomegalovirus (CMV) 550 arterial catheterisation 460
composite grafts 201 carotid artery occlusion 464
compression stockings 246 D-dimer 254 disseminated intravascular coagulation (DIC) 80
computed tomographic angiography (CTA) Dacron grafts 191 distal aortic perfusion techniques 415
349 debridement 396, 537 distal embolisation 204
computed tomography (CT) 8, 14, 96–7 deceleration injury 378 distal necrosis 165
AAA 277 decubitus ulcers 134 distal revascularisation procedures 176
abdominal vascular injuries 432 deep vein thrombosis (DVT) 237–48, 528 diuresis 46, 581
atheroembolism 368 ascending venography 100–1 diuretics 134
carotid endarterectomy 149 case study 246–8 diverticulosis 299, 336
carotid injuries 422 catheter directed thrombolysis 240–2 dobutamine echocardiography 74
central venous thrombosis 103 diagnosis 238–9 domestic trauma 401
chest vascular injuries 406 drug abuse 537 donor vein 391
graft failure 360 pathophysiology 238 dopamine 33
graft infection 298 risk factors 238 Doppler ultrasound 99
lumbar disc surgery 488 stroke 135 DVT 238
mesenteric ischaemia 336 thrombectomy 243–6 graft failure 199
portal hypertension 557 treatment 239–40 limb vascular injury 383
pseudoaneurysm 471 ultrasound 96, 101–2 upper limb ischaemia 500
pulmonary embolism 255 degenerative joint disease 445 dorsalis pedis artery 166, 227, 231, 232
renal artery aneurysms 318 delayed diagnosis, medico-legal implications drainage, infected graft 211
retroperitoneal haematoma 461 122 DRGs see diagnosis related groups
stroke 58, 130, 131, 142 Denmark drug abuse 267, 502, 531–42
thoracic outlet syndrome 501 acute ischaemia 115–16 drug induced haemorrhagic transformation 61
VAAs 325 emergency aneurysm surgery 113 DSA see digital subtraction arteriography
congenital abnormalities 267 emergency vascular procedures 111–12 duodenal tube 134
congenital webs in large veins 267 emergency vascular services 7–11 duplex graft surveillance 200–2
congestive heart failure 87, 130, 267 vascular trauma 118 duplex ultrasound 96, 142, 168
contamination, abdominal vascular surgery 439 deoxygenation see ischaemia abdominal vascular injuries 432
contrast arteriography 231 dextran 517, 578 acute arterial inflammatory conditions 524
contrast ascending venography 101 diabetes carotid injuries 421
contrast-enhanced MRA 98 acute limb ischaemia 168 diabetic foot disease 228
Cook TX2 thoracic endograft 294 amputation 223, 230 DVT 238
coralline thrombus 238 ankle pressure 166 graft failure 199, 361
coronary artery gangrene 163 mesenteric ischaemia 336
bypass surgery 498 graft failure 199, 205 PE 256
rupture 402 diabetic foot disease 223–33 ULVT 267
corticosteroids in inflammatory arterial diseases arterial perfusion 227 duty of care 122
525 arterial reconstruction 231–2 DVT see deep vein thrombosis
costo-scalene tunnel 262 history 226–7 DWI see diffusion-weighted magnetic resonance
CPK see creatine phosphokinase infection 224–5, 229–30 imaging
CPR see cardiopulmonary resuscitation ischaemia 223, 225, 230–1 dysarthria 134
crack smokers 532 limb salvageability 230 dysfibrinogenaemias 20
crackles 253 neuropathy 223–4 dyspnoea syndrome 251, 257
 Index 595

ECG see electrocardiogram endovascular repair (EVAR) 288–91 first rib resection 270
echocardiography 74, 255 atheroembolism 367–71 fistula closure 245
‘ecstasy’ 533 endoleak 345–54 fluid optimisation 75
EEG see electroencephalogram graft breakdown 357–64 fluids management 134
effort induced thrombosis 263, 264–5 intestinal ischaemia 29, 32, 33 fluorescein 338
Ehlers Danlos syndrome 278, 325 occlusion 311 Fogarty catheter 244, 464, 505
eicosanoids 41 vascular emergencies 68 foot infection 517
elastic recoil 481–3 endovascular therapy 21, 410, 423–4 foot pulses 227, 231, 383
elective surgery, HIV/AIDS patients 551–2 abdominal vascular injuries 439–40 foot ulceration 226
electrocardiogram (ECG) 254, 277, 406, 583 acutely ischaemic limbs 481–3 forearm claudication 499
electroencephalogram (EEG) 159 perforations 473–8 foreign body retrieval 484
embolectomy 502–3 end-to-end anastomosis 435, 449 fractures 389, 489
embolisation entanglement of the guidewire 90 free peritoneal rupture 281
cholesterol 478–9 Enterobacter 298 free radicals 579
graft occlusion 213 Enterobacter aerogenes 225 freezing cold injuries (FCIs) 574–8
iliac arteries 484 epidermal hypertrophy 536 pathophysiology 574–5
percutaneous 426 epidural anaesthesia 81 treatment 576–8
surgery 480 epileptic seizures 135 fresh frozen plasma transfusion 80
see also macroembolisation; equinus deformity 223 frostbite 574, 575
microembolisation ergot 499 furosemide 134
embolism 17, 238 ergotism 17
acute mesenteric arterial 336, 338–9 erythrocyte sedimentation rate (ESR) 499, 525 gadolinium 182, 327
carotid artery occlusion 464 Escherichia coli 225, 298 gangrene 163, 462
large artery 499 ESR see erythrocyte sedimentation rate acute arterial inflammatory conditions 524
paradoxical 496 EUG see excretory urogram atheroembolism 368
upper limb ischaemia 496 EVAR see endovascular repair diabetic foot 226, 229
embolotherapy 439 EVG see endovascular graft sympathectomy 526
emergency aneurysm surgery EVL see endoscopic variceal ligation gastric artery aneurysms 330
Denmark 113 excretory urogram (EUG) 430, 432 gastric varices 557
Finland 113–14 exertional compartment syndromes 21 gastritis 557
Sweden 114–15 external stenting 206 gastroduodenal artery aneurysms 330
emergency aortic aneurysm 277–84 extrathoracic reconstruction 503 gastroepiploic artery aneurysms 330
emergency medicine gastrointestinal barrier dysfunction 42–3
AAA 277–84 factor V Leiden 20, 253, 497 gastrointestinal bleeding 241, 340
amputation 444 false aneurysms 122, 380, 393, 471 gastrointestinal haemorrhage 299, 431
HIV/AIDS patients 551 see also pseudoaneurysm gastrointestinal tract, MODS 45
stroke management 142–3 fasciitis 382, 536 gene expression, upregulation 24
USA 13 fasciotomy 164, 176–7, 246, 395–6, 535 gene therapy 206
emergency vascular access 85–93 FCIs see freezing cold injuries gene transfer 205
encephalopathy 87, 558 femoral access 89–90 genetic therapies, MODS 4
end-to-end anastomosis 435, 449 femoral artery giant cell arteritis 497, 501, 523, 528
endarterectomy 503 hernia repair 491 glomerular filtration rate 87
endoleak 345–54 hip surgery 488 glutathione peroxidase 22
diagnosis 349–50 open surgery 464 glycaemic control 134
incidence 345 varicose vein surgery 489 glyceryl trinitrate (GTN) 502
management 350–2 femoral fracture 384 Gore TAG 292
pathophysiology 345–9 femoral lesions 478 graft-enteric erosion 299, 305–7
postoperative care 353 femoral nerve damage 470 graft failure 22, 197, 357–64
prevention 351 femoral pseudoaneurysm 540 diagnosis 199–202, 360–1
type 1 346–7, 351, 357 femoral shaft fractures 378 incidence 359
type 2 347–8, 352 femoral vein management 202–4, 361–2
type 3 348–9 hernia repair 491 pathophysiology 197–9, 357–8
type 4 349 varicose vein surgery 489 postoperative care 205–6
type 5 349 femoroaxillary bypass 503 risk factors 360
endoluminal prostheses 213 femoropopliteal graft 216–17 secondary interventions 362–3
endoscopic lumbar sympathectomy 527 FemoStop 462 thrombolysis 204
endoscopic variceal ligation (EVL) 559–60, 566 fibromuscular dysplasia 17, 19, 325 graft kinks 348
endoscopy 557 fibromuscular hyperplasia 497, 501 graft migration 347
endotension 349, 350, 359 fibrosis 451, 559 graft occlusion, infection 211, 213–14, 217–18
endothelial damage 237 filariasis 267 graft preservation 302
endothelial dysfunction 225 finger clubbing 556 graft routing 176
endotoxaemia 42 fingers, trauma 498 graft surveillance 199–202, 348, 360
endotoxin (LPS) 42, 43 Finland graft techniques 413
endotoxin neutralising protein 48 acute ischaemia 116 graft thrombosis 202–3
endovascular embolisation 352 emergency aneurysm surgery 113–14 groin erythema 299
endovascular graft (EVG) 345, 464 emergency vascular procedures 112 groin haematoma 470, 471
endovascular management 484–5, 505–6, 552 vascular trauma 118–19 groin infection 246, 297–307
596 Index

GTN see glyceryl trinitrate hernia repair 491 ICAM-1 see intercellular adhesion molecule-1
guidewire herniorrhaphy 379 ICU see intensive care unit
migration 90 heroin 531, 534–5 IFVT see iliofemoral vein thrombosis
perforation 473–4 hetastarch 75 IL see interleukin
guillotine amputations 443 high dependency units (HDUs) 73–4 iliac angioplasty 172
gunshot wounds 429 high volume blood transfusion 78–9 iliac artery
gut decontamination 48 high volume salvage autotransfusion 80 embolisation 484
gynaecological cancer 238 highly active antiretroviral therapy (HAART) macroembolisation 479
gynaecological surgery 491 552 perforation 478
gynaecomastia 556 hip surgery 488–9 iliac vein
hirudin 136 compression syndrome 242
HAART see highly active antiretroviral therapy histidine-rich glycoprotein 20 injuries 438–9
HAAs see hepatic artery aneurysms HIV see human immunodeficiency virus iliac vein web 267
haematemesis 299, 431 hormone replacement therapy (HRT) 238, 252 iliofemoral bypass 172
haematochezia 299 Horner’s syndrome 424 iliofemoral vein thrombosis (IFVT) 238, 241
haematocrit 78, 556 horseshoe kidney 282 iloprost 205, 479, 501, 517, 525
haematological system, MODS 44 hose pipe effect 358 IMA see inferior mesenteric artery
haematomas 204 HRT see hormone replacement therapy imaging
cervical 422 human immunodeficiency virus (HIV) 545–53 vascular emergencies 95–104
dissection 412–13 hunting response 574 see also individual techniques
groin infection 299 HVPG see hepatic venous pressure gradient immobilisation
puncture related complications 470 hydration 231 DVT 238
retroperitoneal 434–5, 470, 488 hydrostatic pressure 75 pulmonary embolism 252
surgical treatment 471 hydroxyethyl starch 75 immune complex deposition 163
thrombectomy 246 hyperabduction 263 impaired reflow phenomenon 68
haematuria 431 hypercalcaemia 183 incident reporting 125
haemobilia 326, 431 hypercapnia 132 index procedures 111
haemochromatosis 555 hypercholesterolaemia 205 infants, amputation 445
haemoconcentration 336 hypercoagulability 19, 336 infarcts with transient neurological symptoms
haemodialysis hypercoagulable states 265 (ITNS) 130
catheters 266 hypercoagulation screening 238 infection
indications 86–8 hyperglycaemia 225, 226 acute mesenteric venous thrombosis 336
vascular access 85–6 hyperhomocysteinaemia 20, 69, 199, 205 diabetic foot disease 223, 224–5, 226, 229–30
haemofiltration 312 hyperkalaemia 86, 183 diagnosis 211
haemoglobin 556 hyperkeratosis 227 graft occlusion 211, 213–14
Haemophilus influenzae 550 hyperlipidaemia prosthetic grafts 211–19
haemoptysis 238, 251, 299 diabetic foot disease 226 replantation 451
haemorrhage graft failure 199 inferior mesenteric artery (IMA) 32, 280, 347
abdominal vascular injuries 430, 434 hypernatraemia 134 inferior vena cava (IVC) 429, 436–7
arterial catheterisation 459 hyperphosphataemia 183 inferior vena caval filter (VCF) 242
gastrointestinal 299, 431 hypersensitivity angiitis 19 inflammatory aneurysms 282
intracerebral 158, 159 hypertension 87, 408 informed consent 123–4, 279
intracranial 183 atheroembolism 368 infrainguinal arterial bypass graft 231
mesenteric ischaemia 340 blunt carotid injuries 424 inherited thrombophilia 19
retroperitoneal 461 graft failure 199, 205, 360 injection scleropathy 559, 563
haemorrhagic stroke 60, 129, 137 stroke 130, 132–4, 159 inositol nicotinate 525
haemorrhagic transformation 60–1 hyperthermia 134 insulin resistance 546
haemothorax 405 hypertriglyceridaemia 546 intensive care unit (ICU)
hammer toe 223 hyperventilation 134 Denmark 8
Hampton hump 254 hypocalcaemia 183 preoptimisation 73–4
hand hypocapnia 134 intercellular adhesion molecule-1 (ICAM-1) 23,
ischaemia 499, 535 hyponatraemia 134 41
trauma 498 hypotension 134, 253 intercostal vessels 403
Hartmann solution 446 blunt carotid injuries 424 interleukin (IL) 42, 48
HDUs see high dependency units replantation 450 internal carotid artery (ICA) 420
healthcare workers 552 hypothalamus 573 internal jugular catheters 90–2
heel ulcers 231 hypothenar hammer syndrome 498, 505 interposition grafts 435
hemidiaphragm 254 hypothermia 434, 580–6 interstitial oedema 164
heparin 136, 171, 241, 266, 424, 501 hypovolaemia 336, 376, 430, 450 intestinal infarction 462
heparin cofactor II 20 arterial catheterisation 459 intestinal ischaemia 29–35
hepatic artery aneurysms (HAAs) 325, 327–9 hypothermia 581 intimal dissection 198
hepatic necrosis 329 hypoxia 132 intra-abdominal hypertension 33
hepatic reticuloendothelial dysfunction 43 intra-arterial injections 17, 490, 498, 537–8
hepatic system, MODS 45 iatrogenic injuries 378–9, 404, 429, 498–9 intracerebral haemorrhage 158, 159
hepatic venous pressure gradient (HVPG) 563 vascular 265, 487–91 intracranial haemorrhage 183
hepatitis 550 ibuprofen 577 intracranial neoplasm 241
hepatocellular failure 556 ICA see internal carotid artery intracranial pressure 134
 Index 597

intradermal ischaemia 163 large artery embolism 499 magnetic resonance imaging (MRI) 97
intrahepatic pseudoaneurysms 325, 326 lateral arteriorrhaphy 423 graft failure 361
intraoperative lysis (IOL) 192 lateral meniscectomy 379 graft infection 298, 300
intraoperative transfusion 77 lateral suture 389, 435 ischaemic penumbra 59
intraoperative volume replacement 76 LBP see lipopolysaccharide binding protein pseudoaneurysm 471
intrasac pressure 350 LDLs see low density lipoproteins SAAs 325
intravenous (IV) catheters 85 leaking aortic aneurysms 123 stroke 130, 142
intravenous (IV) drug use 536 ‘leaky bucket’ syndrome 76 malaise 211
intravenous pyelography (IVP) 320, 327 leg sepsis 298 malignancy 267, 299, 336, 379
invasive monitoring 74 leg swelling 251 Mallory–Weiss tears 557
IOL see intraoperative lysis leucocytosis 34 malnutrition 582
ipsilateral ulceration 226 leuconychia 556 mandibular fracture 424
IRI see ischaemia-reperfusion injury leucotriene (LT) B4 41 mannitol 134, 396, 502
irradiation, vessel injury 17, 379, 499 leukaemia 497 MAP see mean arterial pressure
ischaemia liability 122 Marfan’s syndrome 18, 278
bowel 33, 338, 436 lichenification 536 marijuana 523, 532
colonic 29, 32, 35, 43, 280 ligation 329 MARS see mixed antagonistic response
diabetic foot disease 223, 225, 230–1 limb haemodynamics 302 syndrome
digital 496 limb replantation 443–54 May–Thurner syndrome 242
graft occlusion 203 limb salvageability 230 mean arterial pressure (MAP) 130
hand 499 limb tumour surgery 489 mechanical thrombectomy devices 242
intradermal 163 limb vascular injuries 375–97 medial claviculectomy 270
kidney 309 diagnosis 383–4 median sternotomy 410
leg 107 management 382–3 medical history 123, 131
limb vascular injury 385 mechanisms 376–9 medical notes 124
mesenteric 335–40 morphology 379–80 medico–legal problems 121–5
peripheral 461 pathophysiology 381–2 melaena 299, 431
prosthetic graft infection 211–19 postoperative management 396–7 mesenteric branch artery aneurysms 330
thrombolysis 218 preoperative considerations 384–6 mesenteric injuries 436
VEGF 205 surgery 386–96 mesenteric ischaemia 335–40
see also acute ischaemia; acute limb see also lower limb; upper limb aetiology 335
ischaemia line infections 89 complications 340
ischaemia-reperfusion injury (IRI) 39–40, 68, lipopolysaccharide binding protein (LBP) 48 diagnosis 336–7
381, 502 litigation 5 pathophysiology 335–6
ischaemic injury 22 liver cirrhosis see cirrhosis postoperative care 339–40
ischaemic penumbra 57–8, 130 liver transplantation 562, 565 treatment 337–9
ischaemic stroke 60, 129 living wills 279 mesenteric vein thrombosis 339
acute treatment 56 LMWH see low molecular weight heparin metabolic acidosis 87
haemorrhagic transformation 60–1 lobectomy 412 metabolic failure 434
ISMN see isosorbide-5-mononitrate low density lipoproteins (LDLs) 18 metatarsals 223
isolated arm vein thrombosis 103 low molecular weight heparin (LMWH) 136, methamphetamines 533–4
isosorbide-5-mononitrate (ISMN) 563 239–40, 256, 266 methicillin resistant S. aureus (MRSA) 219, 298,
ITNS see infarcts with transient neurological lower limbs 382, 535
symptoms arterial reconstruction 68–9 metoclopramide 558
IV see intravenous artery perforations 478 MI see myocardial infarction
IVC see inferior vena cava ischaemia 430 microembolisation 478–9
IVP see intravenous pyelography replantation 451–2 microembolism 312, 368, 499
vascular injuries 386–7, 430 microtelangiectasias 556
joint dislocation 376 vessel imaging 100 microtibial embolectomy 176
jugular venous pressure (JVP) 253 LPS see endotoxin microvascular abnormalities 225
LT see leucotriene microvascular clamps 173
Kawasaki’s disease 19 lumbar disc surgery 488 migraine 159
Kayexalate 87 lumbar sympathectomy 527 missed diagnosis, medico-legal implications 122
ketamine 171 luminal narrowing 198 mitral valve prolapse 130
ketoacidosis 226 lymphangitis 232 mixed antagonistic response syndrome (MARS)
keyhole surgery 432 41
kidney ischaemia 309 macroembolisation 479–80 MODS see multiple organ dysfunction
kissing balloon technique 483 macroglobulinaemia 20 syndrome
Klebsiella 225, 298 magnetic resonance angiography (MRA) MOF see multiple organ failure
knee surgery 379, 489 acute limb ischaemia 169, 182 Moraxella catarrhalis 550
knife wounds 388 carotid endarterectomy 150 morbidity, thrombectomy 245–6
central venous thrombosis 103 morphine 545
labetalol 134 chest vascular injuries 407–8 mortality
lactate 34, 46 diabetic foot disease 231 Denmark 8, 10
lactulose 558 endoleak 350 leaking aneurysm 278
laparoscopic surgery 490 stroke 142 MODS 46
laparotomy 338, 430 ULVT 267 thrombectomy 245
598 Index

MRA see magnetic resonance antiography occlusive atherosclerosis 17 paralysis 499

MRI see magnetic resonance imaging occlusive disease 503–4 paramedics 62
MRSA see methicillin resistant S. aureus occult graft infection paraplegia 165, 405, 413
mucocutaneous lymph node syndrome 19 bacteriology 212 paraesthesia 461, 499
multiple organ dysfunction syndrome (MODS) diagnosis 211 parenteral dexamethasone 502
24, 39, 43–8, 335, 340 epidemiology 211 paroxysmal lateralising epileptiform discharges
multiple organ failure (MOF) 41 management 212–18 (PLEDs) 159
muscle occupational upper limb ischaemia 498 partial anterior circulation infarction (PACI) 55
atrophy 223 OCPs see oral contraceptives passive shunts 414
necrosis 395 octreotide 558, 563 patch angioplasty 67, 389–90, 435
paralysis 164 oedema 232, 257 patient transfers 123
soft tissue coverage 450 arterial injury 382 PCP (1-(1-phencyclohexyl)piperidine) 533
mycotic aneurysms 21, 325, 537, 538–40 compartmental hypertension 395 PDGF see platelet derived growth factor
mycronecrosis 381, 445 postoperative 445 PE see pulmonary embolism
myocardial depression 382 oesophageal tamponade 558–9 peak systolic velocity (PSV) 201
myocardial infarction (MI) 21, 130, 227, 254, oesophageal transection 561, 564 PECAM-1 see platelet-endothelial cell adhesion
267, 462, 496 oesophageal varices 555, 557 molecule-1
myoglobinuria 445 oestrogen 265 pedal anastomosis 173
myopathy 550 omeprazole 559 pedal arch 167
open abdominal surgery 490–1 pedal bypass 232
nadolol 563 open lumbar sympathectomy 527 pelvic imaging 99–100
naftidrofuryl 525 opportunistic infections 547 pelvic tumour surgery 491
naloxone 534 oral contraceptives (OCPs) 17, 238, 497 penetrating injuries 17, 376–7, 403–4, 420–4,
national databases 107 orthopaedic vascular injuries 488–9 498
nausea, dissecting renal artery aneurysms 319 osseofascial compartment hypertension 21 pentagastrin 558
neck osteoarthropathy 224 pentoxifylline 501, 517
contusion 424 osteomyelitis 226, 229, 451 peptic ulcer disease 299
vascular injuries 419–27 osteopenia 224 peptic ulceration 557
necrosis 451 osteoporosis 579 Percusage 370
acute limb ischaemia 165 osteotomy 379 percutaneous embolisation 426
brain 57 ostial atherosclerotic stenoses 309 percutaneous endovascular therapy 240–3
diabetic foot disease 224, 229 otitis media 267 percutaneous needle biopsy 556
limb vascular injuries 395 overdose 534 percutaneous suturing devices 463
negligence 122 oxidative injury 22–3 percutaneous transluminal angioplasty (PTA)
neointimal hyperplasia 198 oxidative stress 22–3 184–5, 461, 469, 552
neomycin 558 oxygen free radicals 382 perforation
nephrectomy 318, 321 balloon dilatation 474–7
nephropathy 225 pacemaker leads 266 guidewire/catheter manipulations 473–4
nerve injury 376 PACI see partial anterior circulation infarction lower limb artery 478
nerve palsies 498 PAD see peripheral arterial disease surgical treatment 477
nerve repair 450 PAF see platelet activating factor upper limb artery 477–8
neuroendocrine system, MODS 45 Paget–von Schroetter syndrome see upper limb pericarditis 87
neurogenic vasodilation 225 vein thrombosis perioperative care 73–82
neurological deficit 142 acute arterial inflammatory conditions 524 peripheral arterial disease (PAD) 8
neuromas 451 angina-like 251 peripheral endovascular procedures 469–85
neuropathy 87 arterial catheterisation 461 peripheral neuropathy 223, 550
diabetic foot disease 223–4, 225 atheroembolism 368 peripheral pulses 277
neurovascular compression 497 chest 238 peripheral vascular disease (PVD) 469–70
NFCIs see non-freezing cold injuries diabetic foot disease 223 peritoneal lavage 585
nifedipine 134, 525 dissecting renal artery aneurysms 319 permissive hypotension 404, 409
nitinol 358 HAAs 326 peroneal artery 167
nitric oxide 41, 556 large artery embolism 499 peroperative angioplasty 169–70
non-freezing cold injuries (NFCIs) 574, 579–80 mesenteric ischaemia 336 pes cavus 223
non-invasive arterial evaluation, diabetic foot pleuritic 251 PET see positron emission tomography
disease 228–9 palliative care 123 petechial haemorrhages 368
non-ketotic hyperglycaemic hyperosmolar coma pallor 461, 499 PGE1 see prostaglandin E1
226 palmar erythema 556 1-(1-phencyclohexyl)piperidine (PCP) 533
non-occlusive mesenteric ischaemia 336, 339 palmar hyperhidrosis 490 phenoxybenzamine 337
non-salvageable limb 384 Palmaz stent 352 phlebitis 239
non-steroidal anti-inflammatory drugs palpitations 251 phlebography 238
(NSAIDs) 528, 580 pancreatic artery aneurysms 330 phlegmasia caerulea dolens 20, 237, 238, 245
‘no-reflow’ phenomenom 23–4, 68 pancreaticoduodenal artery aneurysms 330 phosphokinase 183
NSAIDs see non-steroidal anti-inflammatory pancreatitis 325, 326, 336, 490 phrenic nerve injury 414
drugs papaverine 518 Physiological and Operative Severity Score for
paracetamol 534 the enUmeration of Mortality and
obstructive jaundice 326 paradoxical embolism 496 Morbidity (POSSUM) score 47, 74, 106
obturator bypass 539 paradoxical undressing 581 plasminogen 20
 Index 599

plasminogen inhibitor activator type 1 20 prosthetic grafts 393 remote vascular units 4
platelet activating factor (PAF) 23, 43 prosthetic patch angioplasty 214 renal angiography 432
platelet concentrates 79 protease inhibitor therapy 546 renal angioplasty 310, 481
platelet derived growth factor (PDGF) 206 protective sensation, in the diabetic foot 227 renal arteries
platelet-endothelial cell adhesion molecule-1 protein C 20, 69, 237, 497 aneurysms 315–21
(PECAM-1) 23 protein S 20, 69, 237, 497 embolism 312
PLEDs see paroxysmal lateralising epileptiform Proteus 298 EVAR 311
discharges Proteus mirabilis 225 imaging 98
plethysmography 263 proximal aneurysm 21 perforation 311
pleural lavage 585 proximal atherosclerosis 496 surgical problems 313–14
pleural rub 253 pseudoaneurysm 90, 204, 320, 327 thrombosis 309–11
pleuritic pain 251 abdominal surgery 490 trauma 312, 436
PMNs see polymorphonuclear leucocytes angiography 461 renal artery stenosis (RAS) 309–14
Pneumocystis carinii pneumonia 550 arterial catheterisation 460, 462 renal atheroembolism 368
pneumothorax 73, 90, 92 heroin 535 renal blood flow 309
poliomyelitis 165 non-surgical treatment 464–5 renal colic 318
polyarteritis nodosa 19, 325, 497 puncture related complications 470, 471 renal dialysis 478
polycythaemia 20 surgical treatment 471 renal failure 165
polygelines 75 pseudocoarctation 19 amputation 445
polymorphonuclear leucocytes (PMNs) 23, 41, 48 pseudocysts 325 atheroembolism 370
polymyalgia rheumatica 19, 497 Pseudomonas 298 haematocrit 78
polytetrafluoroethylene (PTFE) grafts 201, 218, Pseudomonas aeruginosa 382, 550 thrombolysis 204
292, 357 PSV see peak systolic velocity renal infarction 317
polyvinylalcohol sponge 351 PTA see percutaneous transluminal angioplasty renal insufficiency 32
popliteal artery 489 PTFE see polytetrafluoroethylene renal ischaemia 462
popliteal artery entrapment syndrome 21 PTS see post-thrombotic syndrome renal system, MODS 45
popliteal injury 378 pulmonary angiography 98, 255–6 renal vein injuries 438
portable Doppler ultrasound 166–7 pulmonary artery catheter 46 renovascular disease 39
portal hypertension 299, 325, 555–68 pulmonary dysfunction 165 renovascular hypertension 462
aetiology 555 pulmonary embolism (PE) 237, 238, 251–8, 528 re-occlusion 269
diagnosis 556 clinical presentation 251 reperfusion/reoxygenation 24
investigations 556–7 clinical signs 253–4 brain 59–60
management 557 differential diagnosis 254 injury 396
pathophysiology 555–6 investigations 254–6 replantation 443
portal–systemic shunt procedures 560, 564–5 recurrent 257 anaesthesia 447
portal vein injuries 438, 490 risk factors 251–3 complications 451
positron emission tomography (PET) 58 thrombectomy 245 indications 445–6
POSSUM see Physiological and Operative treatment 256–8 lower limb 451–2
Severity Score for the enUmeration of ULVT 267–8 management 446–7
Mortality and Morbidity pulmonary gas exchange 46 movements 451
post-carotid endarterectomy stroke 155–9 pulmonary infarction syndrome 251 postoperative management 450–1
posterior tibial arteries 167, 231 pulmonary tuberculosis 550 secondary surgery 451
postmastectomy irradiation arteritis 505 pulmonary vessels, trauma 403, 405, 412 surgical sequence 447–50
postoperative oedema 445 pulsatile haemorrhage 422 see also limb replantation
postoperative transfusion 78 pulsatile mass, abdominal injury 431 replantation syndrome 445
post-thrombotic syndrome (PTS) 237 pulsation 358 respiratory system, MODS 44–5
potassium levels and ECG findings 87 pulsion endarterectomy 172 rest pain 226
pregnancy 241 puncture related complications 469–70 restenosis 174, 270
aneurysms 331 ‘purse stringing’, avoidance by shunting 391 resuscitation 75
pulmonary embolism 252 PVD see peripheral vascular disease AAA 277
preoperative shock 32 pyrexia 134, 451 bleeding varices 557
pressure transmission 346, 349 chest vascular injuries 408
primary subclavian thrombosis 268 RAAAs see ruptured abdominal aorta aneurysms fluids 75–6
primary thrombosis 261, 263, 264 radial artery mesenteric ischaemia 337
Pringle manoeuvre 437 cannulation 461 pathophysiology 404–6
propranolol 563, 566 fistula 505 reteplase 241
prostacyclin 206 harvesting 498 retinopathy 225
prostacyclin synthase 206 radiology 8 retrograde flow 347, 556
prostaglandin E1 (PGE1) 337, 369 radionuclide imaging 300 retrograde percutaneous transfemoral
prosthetic aortic graft infection 297–307 radiotherapy 267 retroperitoneal haematomas 434–5, 470,
diagnosis 298–302 RAS see renal artery stenosis 488
incidence 297–8 Raynaud’s phenomenon 499, 524, 525 retroperitoneal haemorrhage 461
preoperative management 302 recombinant tissue plasminogen activator revascularisation 443
surgical treatment 302–5 (r-tPA) 135 anaesthesia 170–1
prosthetic bypass grafts rectal bleeding 431 Buerger’s disease 519
infection 211–19 red cell transfusion 77–8 dangers 183
surveillance programme 202 regional blocks 81–2 diabetic foot disease 227
600 Index

revascularisation (cont’d) simulated pulse 393 stroke 129–37

failed 205 single photon emission computed tomography acute limb ischaemia 165
in situ 218 (SPECT) 58 acute treatment 56
indicators 164 SIRS see systemic inflammatory response animal studies 58
mycotic aneurysms 539 syndrome anticoagulation 136–7
renal trauma 312 Sjögren’s disease 497 apoptosis 57
successful 205–6 skin breakdown, diabetic foot disease 224 arterial catheterisation 462
reversed fever 582 skin care 134 carotid artery occlusion 464
reversible ischaemic neurological deficit (RIND) skin popping 534, 536 carotid endarterectomy 141, 149–53
130, 155 SLE see systemic lupus erythematosus cerebral oedema 60
re-warming 584 slime 212 classification 129–30
rhabdomyolysis 183, 538 SMA see superior mesenteric artery cost 56
rheumatic heart disease 496 small artery disease 499 diagnosis 131–2
rheumatoid arthritis 497 small vessel disease 225 logistical problems 61–2
rib fractures 403 smoking management 132–5, 142–4
rifampicin 552 Buerger’s disease 513, 525 neuroprotective agents 137
RIND see reversible ischaemic neurological diabetic foot disease 226 pathophysiology 55–62, 130
deficit graft failure 199, 205, 360 patient studies 58–9
risk:benefit ratio 65–6 pulmonary embolism 252 post-carotid endarterectomy 155–9
risk assessment 65–9 SOD see superoxide dismutase postoperative 158
road traffic accidents 377, 401 soft tissue coverage 450 rehabilitation 135
r-tPA see recombinant tissue plasminogen soft tissue loss 396 risk factors 130
activator somatostatin 557–8, 559 thrombolysis 61, 241
rupture 325 specialisation thrombolytic therapy 135–6
ruptured abdominal aorta aneurysms (RAAAs) Denmark 7 vascular surgeons 141–6
Denmark 7, 9, 10 medico-legal problems 122 subacute bacterial endocarditis 325
Finland 112 UK 3, 5 subacute exhaustion hypothermia 582
USA 12 SPECT see single photon emission computed subclavian artery 498
tomography subclavian line placement 92–3
SAAs see splenic artery aneurysms spider naevi 556 subclinical chronic hypothermia 582
sacogram 347 spinal cord stimulators 518 subintimal angioplasty (SIA) 167, 481–3
saddle embolus 172 spiral computed tomographic angiography substance abuse see drug abuse
saline 75 (SCTA) 349 superficial femoral artery (SFA) 214
Salmonella 325 spiral computed tomography 255, 267 superinfection 382
saphenous vein 490 splenectomy 327 superior mesenteric artery (SMA) aneurysms
scar contracture 451 splenic artery aneurysms (SAAs) 325, 327 329–30, 336
scleroderma 497, 523 splenomegaly 325, 336 superoxide dismutase (SOD) 22, 575
SCM see sternocleidomastoid stab wounds 376, 403 supra-aortic vessels 410–11
scotomata 159 staffing supracondylar fracture 489
SCTA see spiral computed tomographic augmented care 74 surgery
angiography Denmark 7, 9 AAA 278, 279–81
secondary subclavian thrombosis 269 United Kingdom 105 graft occlusion 204
secondary thrombosis 261, 264, 265 Staphylococcus aureus 219, 225, 298, 550 mesenteric ischaemia 337
sedimentation rate, graft infection 211 Staphylococcus epidemidis 212, 298 see also emergency …; individual techniques
Seldinger technique 88–9, 460 stasis 237 surgical thoracic outlet decompression
selective intra-arterial vasodilator infusions 337 Stator Rotarex System 193 270–1
sensorimotor neuropathy 223, 224 stenosis, arterial catheterisation 459 suture breakage 361
sepsis stent-grafts 288, 291–4 SVR see systemic vascular resistance
graft infection 298 disintegration 358, 359 Sweden
ULVT 267 failure 358 acute ischaemia 116–17
sepsis syndrome 40, 43, 382 function 357–8 emergency aneurysm surgery 114–15
septicaemia 163 perforation 477 emergency vascular procedures 112–13
blood transfusion 79 placement 294–5 vascular trauma 118–19
graft infection 211 type 1 endoleak 346 sympathectomy 526
thrombolysis 204 stenting 141 sympathetic ganglion blockade 517
seroma 299 AAA 283 synthetic grafts 191, 503
serotonin antagonists 563 acute aortic dissection 287–96 syphilis 325
sexually transmitted diseases 550 iatrogenic injuries 379 systemic inflammatory response syndrome
SFA see superficial femoral artery ostial atherosclerotic stenoses 309 (SIRS) 24, 39, 40–3
shock, arterial catheterisation 459 ULVT 270 systemic lupus erythematosus (SLE) 247, 325,
shotgun wounds 403, 429 stents 523, 528
shunting, limb vascular injuries 387 infection 213 systemic vascular resistance (SVR) 74
SIA see subintimal angioplasty migration 358–9
sickle cell anaemia 20 sternocleidomastoid (SCM) 91 TACI see total anterior circulation infarction
side-clamping 173 street doctors 535 Takayasu’s disease 19
sigmoid colon pH monitoring 34 Streptococcus pneumoniae 550 arteritis 523, 524
sigmoidoscopy 30 streptokinase 135, 269, 483 bypass procedures 527
 Index 601

presenting symptoms and signs 325 endoleak 346 trauma centres 13

upper limb ischaemia 496, 501, 504–5 femoral lines 90 traumatic spasm 380
Talent endoprosthesis 293 mesenteric ischaemia 340 traumatic vessel injury 21
TAO see thromboangiitis obliterans peripheral endovascular procedures 481 TRISS criteria 76
TCD see transcranial Doppler sonography peripheral ischaemia 461 Trousseau’s phenomenon 528
TCE see transcatheter embolisation puncture related complications 470, 471 tuberculosis 550
temporal (giant cell) arteritis 18 radial artery 461 tumour necrosis factor (TNF) receptors, AAA
tendon repair 449–50 renal artery 309–11 repair 68
terlipressin 557 surgical treatment 472–3 TX see thromboxane
testicular atrophy 556 thromboxane (TX) A[2] 41
tetanus prophylaxis 124, 444, 577 thrombus aspiration 184 UFH see unfractionated heparin
thallium perfusion scans 74 TIAs see transient ischaemic attacks ulceration
therapeutic angiogenesis 205 tibial arteries 389 acute arterial inflammatory conditions 524
thermal injury 489 ticlopidine 136 atheroembolism 368
thermoregulation 573–4 tinnitus 462 diabetic foot disease 223, 226, 227, 229
thoracic aorta TIPSS see transjugular intrahepatic peptic 557
imaging 97–8 portal–systemic stent shunt ulcerative colitis 267
rupture 408 tissue allotransplantation 452 ulcerative plaque disease 156
thoracic injuries 401 tissue oxygenation 46 ulnar artery 498
thoracic outlet decompression 270–1 tissue plasminogen activator (tPA) 157, 206, ulnar neuropathy 204
thoracic outlet syndrome 17, 20, 262, 497, 499, 269, 483 ultrasound
505 TNF see tumour necrosis factor EVG surveillance 350
thoracoabdominal aneurysm 282–3 TNF-a binding proteins 48 lumbar disc surgery 488
thoracoabdominal (type II) aneurysm 279 tobacco 511 portal hypertension 557
thoracoscopic cervical sympathectomy 526 TOE see transoesophageal echocardiography see also duplex ultrasound
thoracoscopic sympathectomy 490 total anterior circulation infarction (TACI) 55 ultrasound venography 101
thoracotomy 408 tourniquet occlusion 232 ULVT see upper limb vein thrombosis
threshold of electric failure 57 tPA see tissue plasminogen activator unfractionated heparin (UFH) 239
threshold of membrane failure 57 training United Kingdom
thrombectomy Denmark 9 emergency vascular services 3–5, 105–9
DVT 243–6 UK 4 medico–legal problems 121–5
graft infection 214, 218 transaxillary approach 379 stroke 55, 56
morbidity 245–6 transaxillary cervical sympathectomy 526 United States
mortality 245 transbrachial approach 379 emergency vascular services 12–14
thrombin, pseudoaneurysms 464 transcatheter embolisation (TCE) 327 stroke 56
thrombin pellets 439 transcranial Doppler sonography (TCD) 150, upper gastrointestinal endoscopy 301
thromboangiitis obliterans (TAO) 19, 168, 497, 159 upper limb vein thrombosis (ULVT) 261–72
511 transcutaneous oximetry 228 anatomy 262–3
thrombocytopenia 34 transfemoral arteriography 168 classification 263–7
thromboembolectomy 157 transgraft thrombectomy 214 clinical presentation 267
thromboembolism 163 transient ischaemic attacks (TIAs) 55, 130 complications 267–9
arterial catheterisation 459–60 carotid endarterectomy 66, 141, 155 haemodynamics 263
EVAR 368 transjugular intrahepatic portal-systemic stent investigations 267
renal artery aneurysm 316 shunt (TIPSS) 555, 560–1 pathogenesis 261
thoracic aortic rupture 414 transoesophageal echocardiography (TOE) 98, treatment 269–72
thrombogenic sponge 351 368, 406–7 upper limbs
thrombolysis transthoracic reconstruction 503 acute ischaemia 495–506
acute upper limb ischaemia 506 ‘trash foot’ 176, 478 artery perforations 477–8
adverse events 204 trauma imaging 100
graft failure 204, 213, 217–18 acute mesenteric venous thrombosis 336 occupational ischaemia 498
iatrogenic injuries 379 aortic 402, 405 vascular injuries 386
lower limb arterial reconstruction 69 arterial catheter 505 venous thrombosis 103
macroembolisation 479 blunt 17, 377–8, 401–3, 420, 424–5 uraemia 87, 183
mesenteric ischaemia 337 crushing 378 urinary catheters 134
renal artery 310 fingers 498 urokinase 157, 184, 241, 269, 483
stroke 61 hand 498
ULVT 269–70 iatrogenic 378–9, 498–9 V/Q see ventilation/perfusion scans
written guidelines 124 initial management 73 VA see vertebral artery
thrombomodulin 20 intrahepatic pseudoaneurysms 325 VAAs see visceral artery aneurysms
thrombophilia 19 kidney 309, 312 valve competence 241
thrombophlebitis 253, 528 limb vascular injuries 376 valve disease 130
thrombosis pelvic 99 variceal ligation 559–60, 563–4
acute mesenteric arterial 336, 339 penetrating 17, 376–7, 420–4 varices 299
acute mesenteric venous 336 pulmonary embolism 252 bleeding 557
axillofemoral prosthetic graft 498 red cell transfusion 77 varicose veins 379
carotid artery occlusion 464 renal artery dissection 319 England 4
central vein 92 ULVT 267 surgery 489–90
602 Index

vascular access, haemodialysis 85–6 neck 419–27 vertebral artery (VA)

vascular anastomosis 449 Sweden 118–19 injuries 419, 425–7
vascular disease, HIV 547 see also vascular emergencies selective angiography 97
vascular emergencies vascular units VF see ventricular fibrillation
drug abuse 531–42 Denmark 7, 8, 9 viraemia 546
HIV/AIDS related 545–53 UK 4 visceral aortic branches, imaging 98–9
imaging 95–104 vasculitis 130, 267 visceral artery aneurysms (VAA) 325–32
medico–legal problems 121–5 vasoconstriction 450 diagnosis 326–7
prophylaxis 124 vasodilators 337 management 330–2
risk assessment 65–9 vasogenic oedema 60 treatment 327–30
Scandinavia 111–19 vasopressin 557 visceral atheroembolism 368
team approach 124 vasospasm 267, 391, 499, 526 visceral rotation 434
UK 5 VCF see inferior vena caval filter visual disturbances 159
USA 12 VEGF see vascular endothelial growth factor volume replacement, intraoperative 76
written records 124 vein bypass grafts, surveillance programme vomiting 319
see also vascular trauma 202 VTE see venous thromboembolism
vascular endothelial growth factor (VEGF) vein compressibility 101–2
205 vein graft stenosis 199 wall motion defects 74
vascular endothelium, HIV 546 vein grafts warfarin 136, 239, 257, 266
vascular inflammatory conditions 523–8 diabetic foot disease 232 warm ischaemia time 309
vascular injuries limb vascular injury 390–1 watershed phenomenon 32
iatrogenic 487–91 vein jump graft 174 wheezing 251
orthopaedic 488–9 vein patches 67, 204 white blood cell count
vascular insufficiency, clinical assessment 24 vein repair 394–5 graft infection 211
vascular laboratory 9 vena caval system 406, 412 HIV 547
vascular malformations 267 veno–venous anastomoses 174 white clot syndrome 157
vascular registries 111, 112, 119 venography 199 white finger syndrome 498
vascular repair 107, 448–9 venolysis 270 wind sock effect 358
catheter interventions 487 venous allografts 218 windchill 575
failure 396–7 venous compression 262 withholding treatment 123
vascular surgeons venous gangrene 237, 245, 246 wound care 388–9
medico–legal problems 123 venous hypertension 93 wound closure 396
stroke management 141–6 venous imaging 100–3 wound debridement 448
UK 5, 122–3 venous insufficiency 450 wound management 177
USA 14 venous thromboembolism (VTE) 239 written guidelines 124
Vascular Surgical Society 4 HIV/AIDS 551 written records 124
vascular trauma prophylaxis 135 xanthine oxidase 382
abdominal 429–40 venous thrombophlebitis 536
chest 401–16 ventilation/perfusion (V/Q) scans 255 X-rays
Denmark 118 ventilatory strategies 74 amputation 444
Finland 118–19 ventricular fibrillation (VF) 414, 581, 583 graft failure 360
limb 375–97 verapamil 563 upper limb ischaemia 500