2003 by Futura, an imprint of Blackwell Publishing Blackwell Publishing, Inc.

/Futura Division, 3 West Main Street, Elmsford, New York 10523, USA Blackwell Publishing, Inc., 350 Main Street, Maiden, Massachusetts 02148-5018, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Science Asia Pty Ltd, 550 Swanston Street, Carlton South, Victoria 3053, Australia Blackwell Verlag GmbH, Kurfurstendamm 57, 10707 Berlin, Germany All rights reserved. No part of this publication may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by a reviewer who may quote brief passages in a review.
02 03 04 05 5 4 3 2 1

ISBN: 1-4051-0387-6

Library of Congress Cataloging-in-Publication Data Vascular emergencies / edited by Alain Branchereau, Michael Jacobs. p. cm. Includes bibliographical references. ISBN 1-4051-0387-6 (alk. paper) 1. Blood vesselsWounds and injuries. 2. Blood vesselsWounds and injuriesSurgery. 3. Cardiovascular emergencies. I. Branchereau, Alain. II. Jacobs, Michael, M.D. RD598.5.V3462 2003 617.4'13044dc21 2003002267 A catalogue record for this title is available from the British Library

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

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

Edited by

ALAIN BRANCHEREAU, MD
University Hospital, Marseille, France

& MICHAEL JACOBS, MD

University Hospital, Maastricht, The Netherlands

FUTURA, AN IMPRINT OF BLACKWELL PUBLISHING

This page intentionally left blank

LIST OF CONTRIBUTORS
Marko AJDUK University Hospital Merkur Zajceva ul.9 Zagreb, Croatia Jerome ALBERTIN Departement de Chirurgie Vasculaire Hopital Adultes La Timone 264, rue Saint Pierre 13385 Marseille Cedex 05, France RaoufAYARI Departement de Chirurgie Vasculaire Hopital Adultes La Timone 264, rue Saint Pierre 13385 Marseille Cedex 05, France Joaquim BARBOSA Vascular Unit Hospital Particular de Lisboa Rua Luis Bivar, 30 1069-142 Lisboa, Portugal Xavier BARRAL Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Nord Avenue Albert Raimond 42055 Saint-Etienne Cedex 2, France Rachel BELL Department of General and Vascular Surgery Guy's & St. Thomas' Hospital Lambeth Palace Road London SE1 7EH, United Kingdom Ramon BERGUER Division of Vascular Surgery Harper Hospital, 3990 John R Detroit, Michigan 48201, USA Ricardo BOFILL Servei d'Angiologia I Cirurgia Vascular Pg Vail d'Hebron 119-129 08035 Barcelona, Spain Didier BOURRAT Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Nord Avenue Albert Raimond 42055 Saint-Etienne Cedex 2, France Bruce BRAITHWAITE Department of Vascular and Endovascular Surgery E Floor, West Block University Hospital, Derby Road Nottingham NG7 2UH, United Kingdom Alain BRANCHEREAU Departement de Chirurgie Vasculaire Hopital Adultes La Timone 264, rue Saint Pierre 13385 Marseille Cedex 05, France Jaap BUTH Department of Surgery Catharina Hospital, PO box 1350 5602 ZA Eindhoven, The Netherlands Piergiorgio CAO Unita Operativa di Chirurgia Vascolare Policlinico Monteluce, Via Brunamonti Perugia 06122, Italy Renata CASTELLANO Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Laurent CHICHE Departement de Chirurgie Vasculaire CHU Pitie-Salpetriere, 47/83, bd de 1'Hopital 75651 Paris Cedex 13, France Roberto CHIESA Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Efrem CIVILINI Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Albert CLARA Servei de Cirurgia Vascular Hospital del Mar, Paseo Maritimo, 25-29 08003 Barcelona, Spain Marc COGGIA Hopital Universitaire Ambroise Pare 9, avenue Charles de Gaulle 92104 Boulogne Cedex, France Jack COLLIN Nuffield Department of Surgery John Radcliffe Hospital Oxford, OX3 9DU, United Kingdom Andreja CRKVENAC University Hospital Merkur Zajceva ul.9 Zagreb, Croatia Philippe CUYPERS Department of Surgery Catharina Hospital, PO box 1350 5602 ZA Eindhoven, The Netherlands

VII

Lourdes DEL RIO Servicio de Cirurgia Vascular Hospital Clinico Universitario 470 HValladolid, Spain Isabelle DI CENTA Hopital Universitaire Ambroise Pare 9, avenue Charles de Gaulle 92104 Boulogne Cedex, France Lucien DUIJM Department of Vascular Surgery Catharina Hospital, PO Box 1350 5602 ZA Eindhoven, The Netherlands Bertrand EDE Departement de Chirurgie Vasculaire Hopital Adultes La Timone 264, rue Saint Pierre 13385 Marseille Cedex 05, France Ted ELENBAAS Department of Cardiac Surgery Academic Hospital Maastricht PO Box 5800 6202 AZ Maastricht, The Netherlands Lidija ERDELEZ University Hospital Merkur Zajceva ul.9 Zagreb, Croatia Jose Maria ESCRIBANO Servei d'Angiologia I Cirurgia Vascular PgValld'Hebronll9-129 08035 Barcelona, Spain Jean-Noel FABIANI Departement de Chirurgie Cardiovasculaire Hopital Europeen Georges Pompidou 20, rue Leblanc, 75015 Paris, France Jean-Pierre FAVRE Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Nord Avenue Albert Raimond 42055 Saint-Etienne Cedex 2, France Maria Jose FERREIRA Vascular Unit Hospital Particular de Lisboa Rua Luis Bivar, 30 1069-142 Lisboa, Portugal Adam FISCHER Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Natalia de la FUENTE Servei de Cirurgia Vascular Hospital del Mar, Paseo Maritime, 25-29 08003 Barcelona, Spain

Jose Maria FUENTES Servei d'Angiologia I Cirurgia Vascular PgValld'Hebronll9-129 08035 Barcelona, Spain Mauro GARGIULO Chirurgia Vascolare Universita di Modena e Reggio Emilia Policlinico Universitario, Via del Pozzo n 71 41100 Modena, Italy Philippe GERSBACH Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Olivier GOEAU-BRISSONNIERE Hopital Universitaire Ambroise Pare 9, avenue Charles de Gaulle 92104 Boulogne Cedex, France Jose GONZALEZ-FAJARDO Servicio de Cirurgia Vascular Hospital Clinico Universitario 47011 Valladolid, Spain Daniel GRANDMOUGIN Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Nord Avenue Albert Raimond 42055 Saint-Etienne Cedex 2, France George HAMILTON Department of Vascular Surgery The Royal Free Hospital NHS Trust Pond Street NWS 2QG London, United Kingdom Daniel HAYOZ Departement de Medecine Vasculaire Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Robert HINCHLIFFE Department of Vascular and Endovascular Surgery E Floor, West Block University Hospital, Derby Road Nottingham NG7 2UH, United Kingdom Brian HOPKINSON Department of Vascular and Endovascular Surgery E Floor, West Block University Hospital, Derby Road Nottingham NG7 2UH, United Kingdom Michael HORROCKS University of Bath, Room L2.27 BA2 7AYBath, United Kingdom Michael JACOBS Department of Cardiovascular Surgery Academic Hospital Maastricht PO Box 5800 6202 AZ Maastricht, The Netherlands

VIII

Isabella JAVERLIAT Hopital Universitaire Ambroise Pare 9, avenue Charles de Gaulle 92104 Boulogne Cedex, France Jean Ader JULES Centre Hospitaller Universitaire Cote de Nacre 14033 Caen Cedex, France Pierre JULIA Departement de Chirurgie Cardiovasculaire Hopital Europeen Georges Pompidou 20, rue Leblanc, 75015 Paris, France Edouard KIEFFER Departement de Chirurgie Vasculaire CHU Pitie-Salpetriere, 47/83, bd de 1'Hopital 75651 Paris Cedex 13, France Mark KOELEMAY Unit of Vascular Surgery Academic Medical Center University of Amsterdam, P.O. Box 22700 1100 DE Amsterdam, The Netherlands Brandon KRIJGSMAN Department of Vascular Surgery The Royal Free Hospital NHS Trust Pond Street NWS 2QG London, United Kingdom Dink LEGEMATE Unit of Vascular Surgery Academic Medical Center University of Amsterdam, P.O. Box 22700 1100 DE Amsterdam, The Netherlands Massimo LENTI Unita Operativa di Chirurgia Vascolare Policlinico Monteluce, Via Brunamonti Perugia 06122, Italy Marcelo LIBERATO DE MOURA Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Lars LONN Department of Radiology Sahlgrenska University Hospital SE 413 45 Goteborg, Sweden Carla LUCCI Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Dominique MAIZA Centre Hospitalier Universitaire Cote de Nacre 14033 Caen Cedex, France Miguel MARTIN-PEDROSA Servicio de Cirurgia Vascular Hospital Clinico Universitario 47011 Valladolid, Spain

Bettina MARTY Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Manuel MATAS Servei d'Angiologia I Cirurgia Vascular PgValld'Hebronll9-129 08035 Barcelona, Spain Germano MELISSANO Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Volker MICKLEY Bereich fur Gefasschirurgie Stadtklinik Baden-Baden, Balger Strasse 50 76532 Baden-Baden, Germany Bas MOCHTAR Department of Cardiac Surgery Academic Hospital Maastricht PO Box 5800 6202 AZ Maastricht, The Netherlands Jorge MOLINA Servei de Cirurgia Vascular Hospital del Mar, Paseo Maritimo, 25-29 08003 Barcelona, Spain Lars NORGREN Department of Vascular Diseases University Hospital MAS 205 02 Malmo, Sweden William PAASKE Department of Cardiothoracic & Vascular Surgery Aarhus University Hospital Skejby Sygehus 8200 Aarhus N, Denmark Philippe PACHECO Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Nord Avenue Albert Raimond 42055 Saint-Etienne Cedex 2, France Federico PAPPALARDO Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Gianbattista PARLANI Unita Operativa di Chirurgia Vascolare Policlinico Monteluce, Via Brunamonti Perugia 06122, Italy Noud PEPPELENBOSCH Department of Surgery Catharina Hospital, PO box 1350 5602 ZA Eindhoven, The Netherlands Gunnar PLATE Department of Surgery Helsingborg Hospital 251 87 Helsingborg, Sweden

IX

SalahQANADLI Departement de Radiologie Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Paola de RANGO Unita Operativa di Chirurgia Vascolare Policlinico Monteluce, Via Brunamonti Perugia 06122, Italy Jan RAUWERDA Department of Surgery Free University, Po Box 7057 1007 MB Amsterdam, The Netherlands Bo RISBERG Department of Surgery Sahlgrenska University Hospital SE 413 45 Goteborg, Sweden John ROBBS Nelson. R. Mandela School of Medicine Faculty of Health Sciences, Private Bag 7 Congella 4013, South Africa Begona ROMAN Faculdad de Filosofia, Departamento de Etica Universitat Ramon Llull, C/Claravall,l-3 08027 Barcelona, Spain Josep ROYO Servei d'Angiologia I Cirurgia Vascular Pg Vail d'Hebron 119-129 08035 Barcelona, Spain Patrick RUCHAT Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Geert Willem SCHURINK Department of Vascular Surgery Academic Hospital Maastricht PO Box 5800 6202 AZ Maastricht, The Netherlands Ludwig Karl von SEGESSER Departement de Chirurgie Cardiovasculaire Centre Hospitalier Universitaire Vaudois Rue du Bugnon 46 1011, Lausanne, Switzerland Andrija SKOPLJANAC-MACINA University Hospital Merkur Zajceva ul.9 Zagreb, Croatia Tomislav SOSA University Hospital Merkur Zajceva ul.9 Zagreb, Croatia

Andrea STELLA Chirurgia Vascolare Universita di Modena e Reggio Emilia Policlinico Universitario, Via del Pozzo n 71 41100 Modena, Italy Peter TAYLOR Department of General and Vascular Surgery Guy's & St. Thomas' Hospital Lambeth Palace Road London SE1 7EH, United Kingdom Alexander TIELBEEK Department of Radiology Catharina Hospital, PO Box 1350 5602 ZA Eindhoven, The Netherlands Ivana TONKOVIC University Hospital Merkur Zajceva ul.9 Zagreb, Croatia Yamume TSHOMBA Chirurgia Vascolare IRCCS H. San Raffaele, Via Olgettina, 60 20132 Milano, Italy Carlos VAQUERO Servicio de Cirurgia Vascular Hospital Clinico Universitario 4701 IValladolid, Spain Fabio VERZINI Unita Operativa di Chirurgia Vascolare Policlinico Monteluce, Via Brunamonti Perugia 06122, Italy Francesc VIDAL-BARRAQUER Servei de Cirurgia Vascular Hospital del Mar, Paseo Maritimo, 25-29 08003 Barcelona, Spain John WOLFE Regional Vascular Unit St. Mary's Hospital, Praed Street London W2 1NY, United Kingdom Michael YAPANIS Regional Vascular Unit St. Mary's Hospital, Praed Street London W2 1NY, United Kingdom Neval YILMAZ Department of Surgery Catharina Hospital, PO box 1350 5602 ZA Eindhoven, The Netherlands Stephane ZALINSKI Departement de Chirurgie Cardiovasculaire Hopital Europeen Georges Pompidou 20, rue Leblanc, 75015 Paris, France

FOREWORD

The subject selected for the 2003 European Vascular Course is Vascular Emergencies, and thirty one chapters in this book address the wide spectrum of urgent and emergency vascular problems. The main impetus for choosing this subject is that approximately 40% of vascular surgical practices are determined by vascular emergencies. The majority of pathologies described are applicable to every vascular surgical practice. The first chapter of the book salutes the important issue of bioethical concerns of vascular emergencies. The following three chapters focus on acute dilemmas in carotid artery disorders, including indications for emergency reconstruction. Blunt trauma of the internal carotid artery and stab wounds at the base of the neck do not occur on a daily basis in a standard vascular practice but constitute a challenging problem. Acute aortic pathology includes occlusion of the abdominal aorta as well as rupture. The latter emergency has been treated surgically for many decades and the option ofendovascular repair is appealing. While in general it is advocated that acute type B aortic dissection must be treated conservatively, new insights dictate a more surgical and endovascular attitude. Aortic emergencies also include complications of laparoscopic surgery and traumatic rupture. Acute ischemia of the upper limb is a serious problem, dictating a substantial part of our practice. Acute complications of arteriovenous fistula for hemodialysis are also addressed. Acute ischemia of the lower limb can result from embolization, thrombosis and other rare causes. Furthermore, underlying pathologies such as peripheral aneurysms and diabetes contribute to emergency situations, requiring surgical, endovascular or thrombolytic therapy. Venous emergencies are described in four chapters addressing acute thrombosis of iliocaval veins, axillary and subclavian veins, aortocaval fistula and traumatic injury of the vena cava.
XI

The last part of the book describes the subjects of acute renal artery occlusion, acute intestinal ischemia, ruptured visceral arterial aneurysms, abdominal compartment syndrome and gunshot arterial injury. We aimed for a comprehensive compilation of vascular emergencies and we could only compose this book with the crucial contribution of the authors and co-authors. Substantial editorial work has been performed by Bertrand Ede and Dirk Ubbink. We are very grateful to our secretaries Annie Barral and Claire Meertens and we appreciate the assistance of Iris Papawasiliou. The Odim team, guided by Marie-France Damia, managed once again to have both the English and French versions of this book printed in time. Blackwell Publishing/Futura contributed significantly, with editorial abetment of Joanna Levine and Jacques Strauss. The major sponsors of the biomedical industries are greatly acknowledged because the textbook and the European Vascular Course would not be possible without their continuous support and enthusiasm for this scientific assignment. Maastricht - Marseille, 2003

Michael Jacobs

Alain Branchereau

CONTENTS
Contributors Foreword

VII XI

Urgent open surgery after enaovascular AAA repair Piergiorgio Cao, Fabio Verzini, Paola De Rango Massimo Lenti, Gianbattista Parlani 71

Bioethical concerns in vascular emergencies Albert Clara, Begona Roman Jorje Molina, Natalia de la Fuente Francesc Vidal-Barraquer

Acute complications following laparoscopic surgery Marc Coggia, IsabelleDi Centa Isabellejaverliat, Olivier Goeau-Brissonniere

81

Urgent carotid surgery Alain Branchereau, RaoufAyari Jerome Albertin, Bertrand Ede

13

Acute type B aortic dissection: surgical indications and strategy Michael Jacobs, Ted Elenbaas Geert Willem Schurink, Bas Mochtar

87

Blunt injury to the carotid and vertebral arteries Ramon Berguer

Endovascular treatment of aortic type B dissection Rachel Bell, Peter Taylor XIII Traumatic rupture of the thoracic aorta Roberto Chiesa, Renata Castellano Carla Lucci, Marcelo R. Liberato de Mourn Federico Pappalardo, Germano Melissano Efrem Civilini, Yamume Tshomba 49

Penetrating injury to the blood vessels of the nect and mediastinum John Robbs

Acute abdominal aortic occlusion Pierre Julia, Stephane Zalinski Jean-Noel Fabiani

107

Has mortality rate for ruptured abdominal aortic aneurysm changed over the last 50 years? Jack Collin

Acute occlusion of the renal arteries Xavier Banal, Philippe Pacheco, Daniel Grandmougin, Didier Bourrat, Jean-Pierre Favre 125

55 Acute intestinal ischemia Brandon Krijgsman, George Hamilton 137

Ruptured AAA: should endovascular treatment be the first choice? Jaap Buth, Noud Peppelenbosch Neval Yilmaz, Philippe Cuypers Lucien Duijm, Alexander Tielbeek

61

Rupture of splanchnic artery aneurysms Joaquim Barbosa, Maria-Jose Ferreira

149

The abdominal compartment syndrome Michael Yapanis,John Wolfe

157

Endovascular treatment of blunt injury of the limbs Bo Risberg, Lars Lonn

247

Acute thrombosis of iliocaval veins Gunnar Plate, Lars Norgren

165

Rare causes of acute ischemia of the limbs Mark Koelemay, Dink Legemate

253

Acute subclavian-axillary vein thrombosis Ramon Bofill, Josep Royo, Jose Maria Fuentes Jose Maria Escribano, Manuel Matas 173

Acute arterial thrombosis of the lower limbs William Paaske

261

Aortocaval fistula Dominique Maiza, Jean Ader Jules

181

Arterial emboli of the lower limbs Michael Horrocks

275

XIV

Traumatic injury of the vena cava and its major branches Laurent Chiche, Edouard Kieffer

193

Acute thrombolysis of peripheral arterial aneurysms Ludwig Karl Von Segesser Bettina Marty, Patrick Ruchat Philippe Gersbach, Salah Quanadli Daniel Hayoz, Adam Fischer

281

Acute ischemia of the upper limb Jose Gonzalez-Fajardo Miguel Martm-Pedrosa Lourdes Del Rio, Carlos Vaquero

207

Endovascular approach to acute arterial occlusions Andrea Stella, Mauro Gargiulo

287

Acute complications of arteriovenous fistula for hemodialysis VolkerMickky

217

Thrombolysis for occlusion of bypass grafts Robert Hinchliffe Bruce Braithwaite, Brian Hopkinson

295

Gunshot and explosive projectile vascular injuries Tomislav Sosa, Ivana Tonkovic Lidija Erdelez, Andrija Skopljanac-Macina Marko Ajduk, Andreja Crkvenac

231

Acute problems of the diabetic foot JanRauwerda 301

1
BIOETHICAL CONCERNS IN VASCULAR EMERGENCIES
ALBERT CLARA, BEGONA ROMAN, JORJE MOLINA NATALIA DE LA FUENTE, FRANCESC VIDAL-BARRAQUER
The field of ethics, also called moral philosophy, involves systematizing, defending, and recommending concepts of right and wrong behavior. Although many of us would consider ourselves as trustworthy, ethical, and honest, we inevitably face choices that may hurt other people, infringe on their rights, or violate their dignity. We are always at risk of using patients as mere tools to our own ends. Ethical considerations, like diagnosis and treatment, are therefore essential features of every case of clinical care of patients. Vascular emergency patients present with problems that require quick, and sometimes immediate evaluation and intervention to save life, limb, or a serious health injury. Vascular surgeons on call have to make decisions, frequently at inconvenient hours, under circumstances of complex clinical scenarios, solitude, scarcity of hospital resources, unfamiliarity with patients, and constrained time. All these factors contribute to ethical conflicts. In the present chapter, the authors will try to provide the readers with the basic keys necessary to make a simple, reasoned, and honest analysis of ethical concerns in vascular emergencies. The reader will realize that behind the majority of our daily-practice ethical concerns, there are uncovered conflicts between moral obligations and self-interest (physician, family, or other third parties), rather than ethical dilemmas. Ethical dilemmas are infrequent and arise only if there are moral considerations for taking each of two opposing courses of action. Unfortunately, their resolution is not so easy, since determining which moral value overrides all others may reflect, at the very end, different visions of human nature.

VASCULAR

EMERGENCIES

Theories of modern biomedical ethics

Biomedical ethics (or bioethics) studies the moral issues in the fields of biological and medical sciences. It traces its roots to several early codes of ethics such as the ancient Greek Hippocratic Oath, professional codes such as the one written by the English physician Thomas Percival in the 18th century, and the Nuremberg Code for research ethics on human subjects that was established in response to the gross abuses in human experimentation performed during the Second World War. Since the end of the Second World War, a remarkable amount of attention has been given to the ethics of medical practice and research as a projection of the rapidly growing concerns generated by scientific and cultural developments during the last decades [1]. In the 1960s many technical advances occurred, such as hemodialysis, major surgical procedures including organ transplantation, and the widespread development of intensive care units and use of artificial respirators. Medically safe abortions, the contraceptive pill, prenatal diagnosis, and the first steps of genetic engineering were also developed at that time. These advances seemed to alter forever the current methods of saving, improving, and extending human lives. In the mid 1960s, the traditional moral moorings of the western medical practice came into question as a result of a series of societal changes, such as a better-educated public, the spread of participatory democracy, a decline in communally shared values, and a distrust of authority and institutions of all kinds. Therefore, the patient-physician relationship changed from a paternalistic model to one in which patient autonomy in decision-making was recognized. With the erosion of the Hippocratic synthesis and the growing complexity of bioethical concerns, many physicians sought guidance in court decisions and in legislation. Most, however, recognized the dangers of confusing law or economics with ethics, and of reducing professional ethics to nothing more than personal opinion. Some philosophers began at that time to write and speak about medical ethical issues. Many bioethical theories and decision-making models were progressively proposed (principlism, casuistry, virtue ethics, narrative ethics, feminist ethics). The history of ethics,

however, has shown humankind to be unable to reach a universally acceptable theory for guiding our actions. The wide spread of bioethical theories has also reflected this historical debate. The main questions of bioethics remain indeed among the oldest that human beings have asked themselves: the meaning of life and death, the bearing of pain and suffering, the right and power to control one's life, and our common duties to each other. PRINCIPLISM The theory of prima fade (Latin for first appearance) principles, developed by Ross, was adapted to medical ethics by Beauchamp and Childress' Principles of Biomedical Ethics [2]. W.D. Ross (The Right and the Good, 1930) believed that our moral convictions were based on duties belonging to the fundamental nature of the universe, and included the duties of fidelity, reparation, gratitude, justice, beneficence, self-improvement, and non-maleficence. The above duties are prima facie insofar as we are always under obligation unless they conflict with one another. Ross argued that there was no obvious priority among these principles, leaving our choice in the event of conflict to our own insight on a case-by-case basis. From this perspective, Beauchamp and Childress chose principles especially appropriate for medical ethics: 1 - Beneficence: duty to be of benefit to the patient, as well as to take positive steps to prevent and to remove harm from the patient. 2 - Non-maleficence: duty to not intentionally create needless harm or injury to the patient, either through acts of commission or omission. Negligence derives from not regarding this principle and includes intentionally imposing unreasonable risks as well as unintentionally imposing risks through carelessness. The debate about active euthanasia also falls within the category of non-maleficence. 3 - Respect for autonomy (self-determination): duty to leave the patient to act intentionally, with understanding, and without controlling influences that would act against a free and voluntary act. The rules of informed consent, truthfulness, privacy, and confidentiality derive from this principle. 4 - Justice: duty to provide a fair distribution of goods in society. Health resources allocation derives from this principle. These principles balance one another but often conflict. For instance, respect for autonomy can conflict with beneficence when the patient refuses

BIOETHICAL CONCERNS IN VASCULAR

a recommended therapy. Beneficence can conflict with justice in the context of resource scarcity, and so on. When principles compete, no absolute hierarchy exists for choosing to follow one principle over another. Judgments about moral precedence among competing principles are made on a caseby-case basis. Critics of principlism have claimed: 1 - the lack of a system for prioritizing principles, 2 - the lack of moral justification for the chosen principles, 3 - the underestimation of character, attitude, and motives of the person performing the action as a central factor in ethics. For these reasons, even principlism's strongest supporters admit that theories incorporating virtues, personal relationships, and other elements should be used in conjunction with the framework provided by principlism.

EMERGENCIES

CASUISTRY
Case-based reasoning, called casuistry, is another common method of bioethical reasoning. Three clinical ethicists (a philosopher - Jonsen, a physician - Siegler, and a lawyer - Winslade) identified four "topics" that are basic and intrinsic to every clinical encounter [3]. Each topic raises questions to be answered before the ethical analysis is done [4]. 1 - Medical indications: does the treatment fulfill any of the goals of medicine? With what likelihood? If not, is the proposed treatment futile? 2 - Patient preferences: what does the patient want? Does the patient have the capacity to decide? If not, who will decide for the patient? Do the patient's wishes reflect a process that is informed, understood, and voluntary? 3 - Quality of life: describe the patient's quality of life in the patient's terms: what is the patient's subjective acceptance of likely quality of life? What are the views of the care providers about the quality of life? Is quality of life less than minimal (i.e., qualitative futility) ? 4 - Contextual features: review social, legal, economic, and institutional circumstances in the case that can influence the decision and/or be influenced by the decision. Once the details of the case have been outlined according to the four topics, it is compared with a specific case (or set of similar cases) for which a moral solution has been developed in the past with professional and/or public agreement about the resolution: does the case sound like other cases you

may have encountered? Is there clear precedent (paradigm case)? How is the present case similar or different to the paradigm case? Is it similar, or different, in ethically significant ways? Thus, casuistry moves from clear past cases to more dubious ones, ordering them by paradigm analogy under some principle. The methodology is therefore similar to the practice of case law where precedents of previous trials are used for analyzing new cases that share similar circumstances. Whether casuistry is a complement or alternative to principlism is still under debate. Although casuistry works in the opposite direction of principlism, it does not eschew principles. Many bioethicists maintain that both theories share more similarities than not and that they complement each other in a system of bioethics. In addition, some critics have claimed casuistry be a product of the culture of the Middle Ages, when there was a consensus on certain principles, while no such consensus exists in today's morally heterogeneous society.

VIRTUE ETHICS
Aristotle defined virtue as "a kind of second nature" that disposes us not only to do the right thing rightly but also to gain pleasure from what we do. Virtue ethics emphasizes the character, intentions, and motives of the moral agent rather than focusing on the agent's actions or outcomes of actions. The virtuous physician naturally will do the right thing and will not likely do the wrong thing. Until the last decades, some kind of virtue ethics had been the implicit and dominant theory in traditional medical ethics since Hippocrates. The renewed interest in virtue ethics has been stimulated by the work of Alasdair MacYntre, in particular his book After Virtue (1984). MacYntre agrees that principles and rules are important for ethics, but he rejects any attempt to justify those principles or rules that abstract them from their rootedness in the historical particularities of concrete communities. The narratives that make such communities morally coherent focuses attention on the virtues correlative to those narratives. To separate ethics from its dependence on such narratives is to lose the corresponding significance of the virtues. Critics of virtue ethics may agree that having a virtuous character may incline the physician to act ethically, but they maintain that virtues alone do not give the physician sufficiently clear action guides.

VASCULAR

EMERGENCIES

ALGORITHM FOR ETHICAL ANALYSIS IN EMERGENCIES Iserson et al. [5] have developed a model specifically designed to be helpful in the emergency setting. It combines casuistry and deontological and utilitarian rules for decisions under time constraints. 1 - The first step is to ask the question: is this a type of ethics problem for which you have already worked out a rule or is this at least similar enough so that a rule could reasonably be extended to cover it? If so, then follow the rule. 2 - The second step is to ask the question: is there an option that will buy time for deliberation without excessive risk to the patient? If yes, buy time. 3 - If the first two steps do not yield a solution, then there are three rules to apply to any ethical decision. The three rules are the following ones. Impartiality: the decision-maker places in the position of the patient by saying: would you be willing to have this action performed if you were in the patient's place? Universalizability: would you be willing to use the same solution in all similar cases? Interpersonal justifiability: consider whether you would be willing to defend the decision to others, to share the decision in public.

geon, and not to be deceived by the surgeon. The patient has the positive rights to be adequately informed about the risks and benefits of surgery, to be treated by a knowledgeable, competent practitioner, to have his or her health and well-being more highly valued than the surgeon's own economic interest, and to decide whether to accept treatment under the conditions described. THE EMERGENCY SCENARIO The emergency department is not only a complex medical environment, but it presents complex clinical and ethical concerns. Ethical concerns will be discussed afterward on a subject-by-subject basis. Clinical complexity a) Unlike other diseases, vascular emergencies have been traditionally poorly protocolized. b) Patterns of disease have changed. c) New vascular technologies imply new emergency challenges. d) Surgical decision-making is often undertaken under data incompleteness. In emergency care, the database derived from history, examination, laboratory, and radiology is virtually always incomplete (up to 50% of the data may be inconclusive or frankly incorrect) [7]. A traditional assumption in surgical practice has been that decisions must be made taking into account the likely costs of under- and overtreating, promoting a challenging decision-making. e) Constrained time to make surgical decisions. Occasionally, some surgical procedures are undertaken under intense time constraints (i.e., drainage of a cervical hematoma and establishment of an airway in a postoperative carotid endarterectomy bleeding, staunching the bleeding from a major exsanguinating source). In these rare circumstances, surgical decision-making is usually straightforward. The perception of time constraint and the emotional sense of urgency are usually felt afterward. In contrast, time constraint may be more evident and may affect the surgeon's decisions when surgical need is not so immediate, such as the patient with a ruptured aortic aneurysm or an acute limb ischemia with neurologic involvement. Environment complexity a) Vascular surgeon solitude and loneliness. Vascular surgeons are scarce "goods" in the mass of health care providers. Unlike general surgery or trauma, in which there is frequently an on-call team, vascular surgeons on call are usually alone in their decisions, that is, outside from the daily-practice

Applying ethics to emergency vascular patients

THE PATIENT-SURGEON RELATIONSHIP An individual patient-surgeon relationship is formed on the basis of mutual agreement on medical or surgical care for the patient. In the absence of a pre-existing relationship, the physician is not ethically obliged to provide care to an individual person unless no other physician is available, as is the case when emergency treatment is required [6]. Once the relationship is established, the surgeon has the fiduciary duty to protect and promote the patient's interest. This primary commitment holds the surgeon's self-interest (technical, scientific, economic) in check and makes it a systematically secondary consideration. This makes the fiduciary's role morally demanding [7]. Surgical ethics is based on a recognition of the rights of patients who require the care of surgeons. The patient has the negative rights not to be killed or harmed intentionally or negligently by the sur-

BIOETHICAL CONCERNS IN VASCULAR decision-making and supervision mechanisms of many vascular departments. b) Increasing tendency to leave on-call service to junior vascular surgeons. On-call service, although at times professionally challenging, weighs more and more with increasing age and professional activity at convenient hours. In addition, many institutions have regulated a top age for on call service. Health providers should take into consideration, however, that the decision to attempt to do everything possible in all emergency circumstances, often made by inexperienced surgeons, creates sometimes logarithmically more moral problems for subsequent health professionals. c) Scarcity of hospital human and technical resources at inconvenient hours. The extraneous environment, such as the hospital laboratory, the speed of the computed tomography (CT) scan, or the availability of operative sites, etc., logistically frustrates the surgeon, creating conflicts among the critical hospital pathways and between different clinical standards of practice or practice guidelines. ETHICS AND THE LAW Surgeons are morally and legally accountable, and the two may not be concordant. Physician participation in torture, for example, may be legal in some countries but is never morally defensible. Surgeons must keep in mind the distinctions and potential conflicts between legal and ethical obligations when making clinical decisions and must seek legal counsel when they are concerned about the potential legal consequences of decisions in ethical dilemmas, when initiating policy and protocols, or when updating existing procedures. The law may vary substantially between countries. While the law is limited in its ability to provide universal guidance and direction, ethical analysis should provide a framework for determining moral duty, obligation, and conduct.

EMERGENCIES

requirement. Common law, however, recognizes that the emergency treatment of incapable persons is an exception to the requirement of consent. Otherwise, competent patients have the right to make choices regarding their health care in emergencies, just as in routine care. Respect for autonomy obligates the physician to seek for the patient the greater balance of goods over harms, as those goods and harms are understood and balanced from the patient's perspective. Consent has three components: disclosure, capacity, and voluntariness. Disclosure Mrs. EVC-1 is 80 years old and lives with her daughter in an apartment. She is fully independent and has never had a serious illness. She is admitted to the emergency department because of acute lower limb ischemia secondary to embolic disease. The vascular surgeon on call indicates prompt surgery and visits Mrs. EVC-1 to disclose benefits and risks of treatment. Before entering the emergency box, however, Mrs. EVC-1 's daughter asks the surgeon to withhold any information about risk of limb loss because her mother is very nervous. Mr. EVC-2 is a 45-year-old homeless but otherwise healthy man admitted to the emergency department because of recent-onset arm swelling. A duplex scan reveals subclavian deep venous thrombosis. The vascular surgeon on call explains the conventional anticoagulant therapy and discusses comprehensively the nature, procedure, shortand long-term benefits, complementary surgical treatment, and risks of modern fibrinolytic therapy. The patient accepts lytic treatment. Ethics and practice Disclosure refers to the provision of relevant information by the clinician and its comprehension by the patient. In many western countries, the prevailing standard of disclosure is that of the "reasonable person". The necessary elements of disclosure include clear information about the patient's diagnosis, the therapeutic alternatives to manage it, including surgical and nonsurgical treatment, the benefits and risks of each alternative, and a frank explanation of those factors about which the medical profession, and the individual surgeon in particular, are uncertain and cannot provide guarantees [7]. This disclosure may be adapted to a long (often the case) or short version according to emergency time constraints. Contrary to the common surgeon's belief, the majority of patients (more than 80%) want to know about the nature of their illness, the reason for surgery, and so on [9]. In

Ethical issues related to patient autonomy

INFORMED CONSENT Consent is the autonomous authorization of a medical intervention. The notion of consent is grounded in the ethical principles of patient autonomy and respect for people [8]. Obtaining the patient's consent to medical care is also a legal

VASCULAR

EMERGENCIES

some cultures, however, a family-centered model of decision-making is favored over one centered in the individual. "Waiver" refers to a patient's voluntary request to forego one or more elements of disclosure. In that case, the patient's reasons for waiving should be sought in order to overcome them through dialogue. If this is not possible, the patient must be informed that he can change his mind at any time or involve a family member in the decision-making process [8]. The cases EVC-1: Mrs. EVC-1 is a fully independent and capable 80-year-old woman without previous serious illnesses. Mrs. EVC-l's daughter was indeed more nervous than her mother. Withholding information during the consent process in the belief that disclosure would lead to the harm or suffering of the patient is called "therapeutic privilege" [10]. While in some cultures therapeutic privilege is widely invoked, this is not the usual case in many western countries. It is better for the surgeon to offer information and allow the patient to refuse or accept further disclosure. Accepting an inappropriate family demand to withhold information infringes on the patient's rights, violates the patient's dignity, and goes against our duty of professionalism. Conversely, there is every moral reason on the basis of confidentiality to honor requests to withhold information from family or friends if requested by the patient. EVC-2: This homeless patient is surely astonished by the surgeon's science and fine dressing. He has probably switched off his understanding after the first 100 words of the surgeon's disclosure. Mr. EVC 2 is clearly a vulnerable patient unable to understand the risks of fibrinolytic therapy in the way the surgeon has explained it. Mr. EVC-2 has probably accepted fibrinolityc therapy because of his confidence that the surgeon is looking for his best interest rather than as a consequence of personal reasoning. Although we could discuss the reasoning behind the surgeon's recommendation (patient's interest vs. surgeon-interest), there is an additional fracture in the surgeon's duty of professionalism that we must point out. Accepting consent for a risky procedure without confirming the patient's understanding of the previously cited elements of disclosure goes against respect for autonomy and is deliberately paternalistic.

Capacity Mr: EVC-3 is a 76-year-old man subjected three weeks ago to an elective aortic aneurysm resection. His postoperative course has been troublesome since the beginning (prolonged intubation, transitory renal failure, prolonged paralitic ileus). Although extubated and conscious, he still remains in the intensive care unit and has fever and some abdominal discomfort. A CT scan performed in the evening raises the concern of a bowel perforation. The vascular surgeon on call proposes prompt surgery and the patient refuses it. He claims to be too tired to fight the disease and he wants to meet his deceased wife in Heaven. A psychiatric consultation is sought to confirm the patient's competence. Mrs. EVC-4 is a 79-year-old diabetic woman admitted to the emergency department with supurative gangrene of two toes. No abscess seems to be present in the dorsal or plantar aspects of the foot. The vascular surgeon on call recommends amputation of the toes, and the patient seems to understand the surgeon's disclosure. During the dialogue, however, she refuses treatment because, "the amputation of my toes will be just the beginning of my end". Ethics and practice Capacity refers to the patient's ability to understand the information relevant to a decision and to appreciate its consequences. Capacity is specific to particular decisions and can change over time. In common law, patients are presumed capable. The surgeon develops a general impression of a patient's capacity during the clinical encounter. In some situations, however, surgeons may be unsure about a patient's capacity. Refusal of recommended treatment usually causes the surgeon to question a person's capacity, although most refusals are caused by factors other than incapacity [11]. In case of refusal, however, the greater the cost to the patient from a false-positive determination of competence, the greater the concern should be to ascertain whether the patient is truly competent [7]. When time and opportunity permit, a psychiatric consultation should be sought, if this is likely to enhance the quality of the determination of competence. Time permitting, when the patient is not competent to consent, surrogate decision makers serve to protect the best interests of the patient by choosing among reasonable options as the patient would have chosen. Since the medical team has significant input about what would be in the patient's interest medically, a decision by a surrogate that does not adhere to this standard should not be auto-

BIOETHICAL CONCERNS IN VASCULAR matically followed and may need to be reviewed by the institutional ethics committee or legal counsel. Religious beliefs: the case of Jehovah's witnesses. Patients' religious beliefs are to be respected on the basis of respect for autonomy. The case, however, is much more troublesome when such beliefs conflict against the surgeon's perceived beneficence; as in the emergency setting, the surgeon has the duty of taking care of these patients. The standard example is the Jehovah's Witnesses, who consent to all medical interventions but refuse blood and blood product transfusions. This refusal is worthy of the surgeon's respect, since these religious beliefs are as sincere as the beliefs of any other of the world's religious traditions. The surgeon, however, does have options when confronted with a patient who refuses perioperative blood product support [7]. First, the surgeon should speak to the patient in private and assure the patient of the confidentiality of the medical records. If the patient maintains the refusal, the surgeon cannot compel a competent adult patient who is not pregnant to accept the transfusion. However, the general caveat is that while competent adults are free to make martyrs of themselves, they cannot martyr their dependent children. In addition, some American hospital policies and state laws have allowed for the imposition of a surrogate decision maker to protect the interest of a minor should a parent (especially a mother) require blood products in order to prevent death and if the death of the patient would result in the child being orphaned [7]. The cases EVC-3: The case of Mr. EVC-3 was taken to an ^urgent meeting of the hospital Ethics Committee. The surgeon presenting the case was asked whether the process of disclosure had been done with empathy and care. The answer was affirmative. The Ethics Committee considered that the adequate steps had been followed (adequate disclosure and psychiatric evaluation of competence). No additional evaluation was believed necessary since the likelihood of survival (as expressed by the surgeon) was poor. Palliative care was indicated in respect of patient's autonomy. After the meeting, the surgeon had some subjective doubts about his empathy during the consent process and revisited the patient. One hour later the patient accepted surgery. The preventive ethics approach to refusal of surgery should be respectful of the patient's reasoning, on the assumption that the patient, by his or her own

EMERGENCIES

rights, has good reason for refusal but may, with additional information and/or empathy, reconsider and accept surgery, and not on the assumption that the patient capacity is in doubt. The present case exemplifies again the morally demanding fiduciary role of the surgeon (see professionalism). EVC-4: The case of Mrs. EVC-4 shows the effects of an uncovered depression in capacity. Mrs. EVC-4 had the ability to understand her problem and the proposed treatment. The unexpected reasons of Mrs. EVC-4's refusal raised doubts about her capacity and a psychiatric consultation was requested. Mrs. EVC-4 admitted to having a persistent depressed mood and several vegetative signs of depression. She accepted treatment for depression. Her foot condition stabilized with antibiotics. Some days later, the patient accepted the proposed surgery. Had prompt surgery been needed, a surrogate decisionmaker would had been sought. Voluntariness Mrs. EVC-5 is a 65-year-old diabetic woman admitted to hospital in the morning because of a toe supurative gangrene with plantar abscess. The vascular surgeon on call (surgeon A), whose service starts at 5 pm, is in the operating room treating an elective case. Mrs. EVC-5 is evaluated by another vascular surgeon (surgeon B). There will be no operating room available until the afternoon and surgeon B considers immediate surgery unnecessary. Surgeon B insinuates surgery but leaves the complete disclosure to surgeon A, who is informed by a surgical nurse that a toe amputation has been added to the surgical emergency schedule. After ending his elective case, and without delay, surgeon A goes to eat something before reentering the operating room. When he returns, Mrs. EVC-5 is already in the operating room. Surgeon A realizes that the informed consent has been insufficient and decides to complete it in the operating room. Mr. EVC-6 is a 78 year-old-man with an 8 cm aortic aneurysm. A vascular surgeon proposes elective surgery but he refuses, claiming that he has already done all he had to do in life. He signs an advanced directive refusing emergent surgery in case of rupture. A signed copy is left in the patient's chart. Six months later the aneurysm ruptures and the patient is taken, conscious, to the hospital. The vascular surgeon on call has doubts about what to do. Ethics and practice Voluntariness refers to the patient's right to come to a decision freely, without force, coercion or

VASCULAR

EMERGENCIES

manipulation. Internal and external factors can affect a patient's decision about treatment [8]. Internal factors arise from the patient's medical condition (i.e. pain). The surgeon's role is to minimize the potential controlling effect of these internal factors without jeopardizing the patient's capacity. External controlling factors may be related to the clinician, the health care setting, and the family or friends. Surgeons should take steps to minimize the potential for manipulation. Patients can be manipulated when the information they receive is incomplete or biased. For this reason, a useful strategy is to ask patients to review the information in their own words. Another source of manipulation is disclosing information just before a major procedure is to be performed. The setting (i.e., operating room) and the immediacy of the medical procedure militate against a patient being able to make a free or voluntary decision. Advance directives document the process aimed at extending the rights of competent adults to guide their medical care through periods of decisional incapacity. Advanced directives are grounded on voluntariness. Their goals are: 1 - to maximize the likelihood that medical care serves the patient's goals (promoting respect for autonomy), 2 - to minimize the likelihood of over- and undertreatment (promoting non-maleficence), 3 - to reduce the likelihood of conflicts between family members and health care providers (promoting justice), 4 - to minimize the burden of decision-making on family members or close friends (promoting respect for autonomy). In the emergency setting, however, there are practical difficulties in having such directives function [7]. Family members may or may not be aware of such directives. Emergency medical personnel do not have access to the hospital chart at the time resuscitation and other therapeutic measures are needed. By the time it is known that an advance directive exists, the patient may already have been resuscitated, be on life support, or even be in the operating room. In general, when there is unclear evidence that a patient might have refused a particular treatment, such evidence is not binding if it goes against the clear best interests of the patient needing an emergency intervention. Most people who complete advance directives are not, at that time, suffering from a terminal or fatal disease. In

completing an advance directive, most are expressing their wishes regarding the limitation of treatment when treatment will only prolong the process of dying. Therefore, the onus will fall on the surgeon to determine whether the conditions of advance directive apply. The cases EVC-5: The majority of readers whose practice includes on-call service will have occasionally met patients with cases similar to that of Mrs. EVC-5. The question of professionalism rises again in her case. Surgeon B should have personally informed surgeon A, especially about the incompleteness of the consent process, and surgeon A should have visited Mrs. EVC before entering the operating room. Neither surgeon A's physiological needs nor the existence of minor emergent surgical cases like toe amputations, debridements, or some A-V fistula revisions, justify the absence or incompleteness of the consent process before the patient enters the operating room. The reader should simply remember Iserson's question about impartiality: "would you be willing to have this action performed if you were in the patient's place?" EVC-6: Advanced directives take effect only in situations in which the patient is unable to participate direcdy in surgical decision-making. Appeals to living wills and surrogate decision-makers are ethically and legally inappropriate when individuals remain competent to guide their own care. The benefits and risks of surgical treatment together with the lethal condition of a nonsurgical attitude must be disclosed to the patient. If Mr. EVC-6 refuses surgery with an understanding of the consequences, his wishes should be honored. If he opts for surgery, then it should be performed promptly.

CONFIDENTIALITY AND TRUTH TELLING

Confidentiality is derived from the Latin confidere, to trust. Patients confide in their physicians with the understanding that what they report will not be disclosed without explicit permission. The duty to maintain confidentiality can be viewed as a prima facie obligation that may be overridden only when it conflicts with stronger moral duties. Exceptions for confidentiality are concerns for the safety of other specific persons and for public welfare (i.e., report of certain communicable/infectious diseases) . The crisis atmosphere that often attends surgical emergencies may heighten the need of family members and loved ones for information. Sur-

BIOETHICAL CONCERNS IN VASCULAR geons, however, should not allow the exigencies of an emergency situation to undermine traditional privacy safeguards. When the patient is incapacitated, the surgeon should disclose information only to the patient's surrogate, who has a legitimate "need to know" the patient's medical status. Telling the truth may seem to be a straightforward and ancient ethical principle in health care. However, the Hippocratic oath does not make any mention of truth telling to patients, and the American Medical Association's first Code of Ethics in 1847 perpetuated this attitude. This therapeutic privilege was justified by the principle of non-maleficence, and continued into this century. Today the duty of truth telling in medicine has become an ethical issue (respect for autonomy), although in many cultures it is not the norm. There are two main situations in which it is justified to withhold the truth from the patient: 1. when the surgeon has compelling evidence that disclosure will cause real and predictable harm (i.e., make a depressed patient actively suicidal), and 2. when the patient him- or herself states an informed preference not to be told the truth.

EMERGENCIES

Ethical issues related to beneficence and non-maleficence

FUTILITY Medical futility refers to interventions that are unlikely to produce any significant benefit for the patient. Two kinds of medical futility are often distinguished. A treatment is quantitatively futile when the likelihood of benefit is very poor, for example when physicians conclude that it has been useless in the last 100 cases. In addition, a treatment is qualitatively futile when the question: "What sort of life is worth preserving?" is at the case core. Surgeons have no obligation to offer or provide treatments that clearly do no benefit their patients. These therapies may increase the patient's pain and discomfort (conflicts with non-maleficence) and spend finite medical resources (conflicts with justice). However, only through dialogue can the physician understand the goals of treatment and determine futility. This approach allows for exploration of the desired outcome, acceptability of burdens, and the patient's or family's willingness to gamble with the outcome.

LIMITS TO THE AGGRESSIVE SURGICAL MANAGEMENT OF HEALTH-THREATENING EMERGENCIES A traditional assumption in surgical practice has been that emergency surgical patients should be provided rapid care at whatever level of intervention the surgeon reasonably thinks is required to preserve their lives or protect them from a serious compromise to their prior health status. Vascular surgeons on call, however, see occasional patients for whom treatment success will be very unlikely, the length of life to be secured brief, and the quality of life to be achieved marginal. These conditions create concerns as to when treatment should be characterized as futile, inappropriate, or marginally useful. In this context, surgeons look for ways of characterizing emergency patients [7]. Emergency patients for whom surgery survival is unprecedented. Vascular surgeons responding to emergency patients who are most assuredly "going," who in highest probability will very shortly die, are torn between rapid full-steamahead aggressive resuscitative measures and the recognition that interventions may succeed only in increasing the potential misery to the patient and the family. Such circumstances are represented by some patients with prolonged prehospital resuscitation following trauma and irreversible metabolic acidosis, or by patients with ruptured aortic aneurysm with certain particularities. The literature is not very prone to help to identify patients with a 100% mortality. However, the more secure surgeons are in their ability to resuscitate patients with complex surgical problems effectively, the easier it is to reach the decision not to initiate or even to terminate curative care in a patient who has an emergency surgical condition for which treatment is not reliably expected to prevent death. Vascular emergencies associated with central nervous system injury. Surgeons can be significantly frustrated when confronted with vascular surgical emergencies in patients with an associated central nervous system injury. Vascular injuries that require immediate or prompt interventions (i.e., active bleeding, acute ischemia) should be treated aggressively so long as the prognosis for the recovery of some degree of significant central nervous system function remains positive. However, especially in the postoperative period, if the most reliable prognosis is that the patient is not expected to recover to a cognitive and sapient existence, the surgeon should be willing to discontinue the trial

VASCULAR

EMERGENCIES

of intervention, given the concurrence of the surrogate decision-maker and presuming the absence of previous binding instructions. Vascular injuries not requiring immediate treatment (i.e., many thoracic aortic pseudoaneurysms) are best delayed (and controlled) if possible until the central nervous system injury prognosis can be stated with some degree of certainty. Vascular emergency patients in whom survival with severe disability may be frequent. Traumatic vascular injuries may produce some disability by themselves (i.e., paralysis after surgical treatment of a transected aorta from blunt injury) or by associated injuries (i.e., fracture malunion, bone infection, soft tissue retraction). In the face of time constraints, surgeons can usually only imperfectly engage this process so as to prepare the patient and the patient's family for the consequences of a surgical intervention. The burden of proof needed to terminate the traditional obligation to intervene can frequently not be met because of the lack of concreteness of the data in an emergency context. Because of this softness, there are grounds to qualify in favor of intervention in all but the most well founded cases of futility. Emergency cases associated with a low probability of survival but with good quality of life among survivors. This may be the case for many ruptured aortic aneurysms. At stake here is the role of costs in determining the appropriateness of therapeutic interventions. While such health policies are rarely, if ever, created, bedside surgeon decision-making must be based under beneficence and respect for autonomy principles.

text of health care, macroallocations of resources are made by politics, mesoallocations are made at the level of health institutions, and microallocations are made at the level of the patient. Resource scarcity may be due to the shortage of a finite good (i.e., an organ for transplantation), or to a shortage of economic funds. While physicians have a fiduciary duty to promote the patient's best interest, their role in resource allocation is controversial. The physician can approach resource allocation in practice by choosing tests and interventions known to be beneficial, by choosing the test or intervention with the least cost among equally beneficial options, by resolving conflictive claims for scarce resources on the basis of need and benefit, and by seeking unacceptable shortages at the level of mesoand macroallocation [8]. The physician should not approach resource allocation by subordinating the primary concern of care - his or her patient's well being- to a budgetary issue. The surgeon must also pay attention when making decisions based on "quality of life." Several studies have shown that physicians often rate the patient's quality of life much lower than the patient himself does. If the patient is able to communicate, the surgeon should engage him or her in a discussion about his or her own condition assessment. The case EVC-7: The attending surgeons should provide appropriate care for Mrs. EVC-7, since a delay in vascular surgical reconstruction could result in some neurologic sequelae and in nephropatic metabolic syndrome. Surgeons should involve the administrator on call to bring in additional skilled personnel (anesthesiologist, surgical nurses, etc.) to provide care for the patient. In this way, they clarify the responsibility of the hospital to resolve the mesoallocation problem at an administrative level. Surgeons should seek resolution of unacceptable shortages at the level of emergency care.

10

Ethical issues related to justice

Mrs. EVC-7, a 27-year-old woman, is taken to the emergency department during the night after a motor vehicle accident. She has a femur fracture and signs of acute ischemia in the limb. The attending vascular surgeon on call, together with the orthopedic surgeon, proposes immediate surgery, however no operating room will be available in the next four hours. Because of recent hospital closures in the city, no other facility is available in which to treat this patient. Ethics and practice Resource allocation is the distribution of goods and services to programs and people. In the con-

In the pursuit of professionalism

Mr. EVC-8 is a 76-year-old man with a 7 cm abdominal aortic aneurysm. He has been rejected from elective surgery because of depressed left ventricularfunction and moderate ventilatory deficit. Mr. EVC-8 comes to the emergency department with his aneurysm ruptured. The

BIOETHICAL

CONCERNS IN VASCULAR

EMERGENCIES

vascular surgeon on call wonders whether it is worth it to operate on this patient.

PROFESSION AND PROFESSIONALISM

Profession has long been recognized to encompass three essential characteristics: expert knowledge, self-regulation, and a fiduciary responsibility to place the needs of the client ahead of the selfinterest of the practitioner [12]. The dominant conception of profession is sociological. Professionalism is the basis of medicine's contract with society. It demands placing the interests of patients above those of the physician, setting and maintaining standards of competence and integrity, and providing expert advise to society in matters of health. Essential to this contract is public trust in physicians, which depends on the integrity of both individual and the whole profession [13]. In this view, ethics is an important predictor for a profession, but ethics is not its essential and indispensable defining feature. Another view of profession links it to an ethical ideal without which it cannot exit. That ideal focuses on some degree of effacement of self-interest when it is required by the good of the person seeking assistance. This conception is rooted in the etymology of the word "profession," which means "a declaration, promise, or commitment publicly announced." That promise is made in every clinical encounter when the physician offers to help those who need his or her special knowledge. That promise entails competence and putting that competence at the service of the patient, even when it means some degree of sacrifice on the part of the physician [14].

about the activity of medicine that imposes a higher standard of moral integrity on its practitioners. Each time this conflict has arisen in the past, it has resulted in a new infusion of moral sensitivity through successive professional codes. The reality of cycles of moral confusion must not obscure the specific dimensions of the present recurrence, some of which are unique, and some not [15]. What are not unique are the temptations of self-interest, power, prestige, pride, profit, and privilege that beset all humans, in all ages. In our times, however, there are two sources of unique conflict. 1 - There is a commodification of health care as a product like any other, left to the ethos of the marketplace, to competition, commercialization, and profit-making (today's moral imperatives). Physicians are not held to moral standards higher than those of the general society in which they live. 2 - There is an erosion of the foundations of professional ethics. Underlying these criticisms is a pervasive moral skepticism that denies the validity of any stable moral truth and even the capacity of reason to apprehend such truth were it to exist. Pellegrino expects a repetition of the historic cycle of deprofessionalization and reprofessionalization characteristic of periods of moral confusion [15]. Indeed, some have bet for a new code of professionalism [13]. However, it is not likely that any of these codes will change today's scenario if physicians do not have at the core of their beliefs the primacy of the welfare of their patients over any own self-interest (technical, scientific, academic). Only when medicine is a moral enterprise will that be possible. The case EVC-8: The perspective of a bad outcome frustrates any vascular surgeon faced with operating on a ruptured aortic aneurysm. The issues of futility (beneficence) and costs (justice) may rise. However, the mechanisms of rationalization of the surgeon's self-interest ("I'm going to waste some sleep hours, surely for nothing, while tomorrow I have a lot on my agenda") may often overcome any sincere ethical analysis. As we have seen previously, futility has very narrow margins and allocating resources is best managed while not at the bedside. Mr. EVC-8 must be offered surgical treatment and informed that, although it is the single curative choice, the probability of survival is very low. In respect of the patient's autonomy, his final decision will be honored.

11

CURRENT CONFUSED SCENARIO AND FUTURE PERSPECTIVES

Many individual persons, groups, and institutions play a role in and are affected by medical decisionmaking in the current practice environment. Tension and competition among the interests of clinicians, insurers, patients, and institutions for available social and health care resources unavoidably influence the patient-physician relationship [6]. All these issues have raised a deep concern about the present loss of that special dedication to competence, service, and other-than-self-interest that have been associated with the ideal physician for so long. However, in its history, medicine has witnessed recurrent cycles of moral confusion-of doubts about whatever there is something special

VASCULAR

EMERGENCIES

Conclusion
The surgeon's fiduciary duty to protect and promote the patient's interest becomes more complex and demanding in emergencies. Vascular surgeons should face all of their clinical decisions, and specially emergencies, with bioethical reasoning. This attitude is frequently not time-consuming and may help the surgeon to unmask potential conflicts between moral obligations and self-interest (sur-

geon, family, or other third parties) and to approach real ethical dilemmas with honesty, sense, and reasoning.

A CKNOWLEDGMENTS

The authors wish to acknowledge the valuable suggestions of Dr. F.Abel, President of the Institut Borja de Bioetica, Universitat Ramon Llull, Barcelona (Spain).

R E F E R E N C E S

12

1 Pellegrino ED. The metamorphosis of medical ethics: a 30-year retrospective. JAMA 1993; 269:1158-1162. 2 Beaucamp TL, Childress JF (eds). Principles of biomdical ethics, 5th edition. New York, Oxford University Press 1989: p 468. 3 Jonsen AR, Siegler M, Winslade W (eds). Clinical ethics: a practical approach to ethical decisions in clinical medicine, 4th edition. New York, McGraw-Hill 1998: p 202. 4 The University of Washington bioethics website: http:// eduserv.hscer.washington.edu/bioethics/credits.html 5 Iserson KV, Sanders AB, Mathieu DR, Buchanan AE (eds). Ethics in emergency medicine. Baltimore, Williams and Wilkins 1986. 6 Anonymous. Ethics manual. 4th edition. American College of Physicians. Ann Intern Med 1998; 128: 576-594. 7 McCullough LB, Jones JW, Brody BA (eds). Surgical ethics. New York, Oxford University Press 1998: p 416. 8 Singer PA (ed). Bioethics at the bedside: a clinician's guide. Ottawa, Canadian Medical Association 1999: p 154.

9 Dawes PJ, Davison P. Informed consent: what do patients want to know?/SocMed 1994; 87:149-152. 10 Meisel A, Roth LH, Lidz CW. Toward a model of the legal doctrine of informed consent. Am J Psychiatry 1977; 134: 285-289. 11 Appelbaum PS, Roth LH. Patients who refuse treatment in medical hospitals. JAMA 1983; 250:1296-1301. 12 Sullivan WM. What is left of professionalism after managed care? Hastings Cent Rep 1999; 29: 7-13. 13 Medical professionalism in the new millennium: a physician charter. Ann Intern Med 2002; 136: 243-246. 14 Pellegrino ED. The healing relationship: the architectonics of clinical medicine. In: Shelp E (ed). The clinical encounter: the moral fabric of the patient-physician relationship. 4th edition. Boston, Reidel 1983: pp 153-172. 15 Pellegrino ED. Medical professionalism: can it, should it survive? J Am Board FamPractmO; 13:147-149.

2
URGENT CAROTID SURGERY
ALAIN BRANCHEREAU, RAOUF AYARI JEROME ALBERTIN, BERTRAND EDE Surgery of stenoses of the internal carotid artery (ICA) is intended for lesions and chronic neurologic disorders as indicated by the NASCET, ECST, and ACAS study results. Urgent surgery for an unstable neurologic condition has given rise to a considerable amount of skepticism, because of the poor results as found in the Joint Study [1]. At present, this sentiment requires reconsideration. The combined cumulative mortality and morbidity rate of carotid surgery in the years the Joint Study was performed, was 5 % to 20% [2], whereas it currently is 1% to 5% [3], At present, the urgent diagnosis of massive ischemic and hemorrhagic strokes is possible by means of computed tomography (CT) scanning and magnetic resonance (MR) imaging. Ultrasonography, sometimes in combination with angioCT or angio-MR, now allows for a sufficiently accurate appreciation of carotid stenoses, while the use of angiography of the aortic trunk is decreasing. This wins time and reduces iatrogenic neurologic morbidity. Angiography remains important in the intra-operative check-up of the reconstruction and regarding the additional therapeutic possibilities, such as intracerebral arterial thrombolysis. The introduction of stroke centers has been an essential advancement in the emergency care of these patients through an indispensable combination of a logistic platform and a multidisciplinary approach. Unfortunately, randomized controlled trials that could clarify the indications are not present at the moment. The aim of this chapter is to provide an overview of the literature and our own experience. This chapter is dedicated to urgent carotid surgery, excluding traumatic lesions, carotid dissections and postoperative strokes.

13

VASCULAR

EMERGENCIES

Pathophysiology
Our central nervous system has the lowest resistance to ischemia. The normal cerebral blood flow is estimated at 80 mL/100 g/min. Below 20 mL/ 100 g/min, neurologic disturbances occur, which are reversible as long as the hypoperfusion is not prolonged and the threshold of 10 mL/100 g/min is not reached. These disturbances are detected by changes in the electroencephalogram (EEC) and, the more severe ones, by changes in somatosensory evoked potentials. The allowed limits before irreversible cerebral lesions occur are a flow of 0 mL/ 100 g/min during a period of 20 minutes, 10 mL/ 100 g/min for 40 minutes, and 15 mL/100 g/min for 80 minutes. Surrounding the areas of irreversible damage there is an area of nonfunctional but still viable brain tissue, which can restore its function when normal cerebral flow is re-established. Identification of this area has led to the concept of the ischemic twilight zone. This area, which has lost its autoregulatory capacity, shows a greatly unstable pressure-sensitive metabolism. A major part of the clinical manifestations of ischemic strokes is due to the dysfunctioning of this twilight zone. The aim of urgent carotid surgery is to safeguard this ischemic twilight zone. Failure of reperfusion results in a loss or deficiency of the autoregulatory system in certain areas. Clinical symptoms can be hemorrhagic events and cerebral edema. One should discern between a cerebral hemorrhage and a hemorrhage upon a stroke. The former is due to a rupture of the blood brain barrier and penetration of blood into a previously unaffected brain area. This has a poor prognosis. The latter reflects infiltration of blood into infarcted tissue. This does not necessarily have a poor prognosis. The presence of a cerebral infarction is a well-known risk factor during carotid surgery. The wall changes in the vessels of the infarcted area may lead to vascular rupture, leading to a hemorrhagic infarction upon an ischemic event. Other risk factors of reperfusion damage are multilevel lesions causing chronic hypoperfusion and severe arterial hypertension.

enables simple assessment of the evolution of the deficit before and after treatment as well as correct comparison of the results from the different studies. 0 - No deficit 1 - Minimal deficit with complete autonomy 2 - Minor deficit with incomplete autonomy not requiring assistance in daily activities 3 - Moderate deficit with walking ability 4 - Severe deficit with walking disability 5 - Disabling deficit leading to bed confinement 6 - Death CRESCENDO TRANSIENT ISCHEMIC ATTACK Crescendo transient ischemic attack (CTIA) is a recurrent, localized ischemic neurologic event characterized by spreading of the deficit, a lengthening of the duration of the event or shortening of the interval between each event. Despite the absence of prospective studies, the prognosis of not surgically treated CTIA is poor and it leads to a considerable number of strokes. PROGRESSIVE STROKE Progressive stroke is a severe neurologic deficit, showing a varying intensity but without disappearing. The natural history of these events shows a mortality rate of 14% to 36% and a morbidity rate of 54% to 69% [4]. The definition of progressive stroke is not unequivocal. Hence, various terminologies are used in the literature, including evolving stroke or fluctuating stroke. SEVERE STROKE IN THE ACUTE PHASE This is a severe neurologic deficit according to stages 4 and 5 of the modified Rankin scale. In the Oxfordshire Community Stroke study, an infarction of the complete anterior circulation is accompanied with a 30-day mortality of 39% and a risk of functional disability of 56%. Severe strokes of carotid origin also show a poor prognosis with a mortality between 16% and 55% and a risk of functional disability between 40% and 69% [5]. REGRESSIVE AND MODERATE STROKES IN THE EARLY PHASE These are moderate ischemic strokes according to Rankin stages 1 to 3 of which the neurologic state has reached a steady state. In case of a stroke due to partial obstruction of the anterior circulation, a mortality of 4% and a risk of functional disability

14

Definitions
The modified Rankin scale is used for the clinical evaluation of neurologic deficits. This scale

URGENT CAROTID
of 39% were reported in the Oxfordshire community stroke study. Surgical treatment six months after the onset of the stroke is commonly accepted as the therapy of choice for severe carotid stenoses. This expectative policy has become questionable because of the increased risk of early stroke recurrence [6] and studies showing no risk increase after early surgery [6-10]. ANATOMICAL EMERGENCIES When the neurologic state is stable, the identification of certain carotid lesions may offer an indication for urgent surgery. The most frequent possibility is a so-called subtotal stenosis. This definition is not based on the poorly known natural history of these lesions, but on hemodynamic findings representing a reduction of the flow in the ICA and the induction of a collateral circulation (Fig. 1). The second anatomical emergency is the presence of a floating thrombus at the level of the carotid bifurcation and the extracranial ICA (Fig. 2). The literature data on this subject are contradictory [11,12]. Some advocate urgent revascu-

SURGERY

larization by means of surgery or thrombolysis, others favor medical treatment followed by secondary surgery. The last type of anatomical emergency is a carotid occlusion. For this type of lesion, the clinical presentation should determine the policy. When a neurologic deficit is absent, the carotid occlusion is virtually impossible to date and does not require urgent treatment. Nicholls et al. [13] reported an incidence of 46% for severe, initially symptomatic carotid occlusions with a yearly risk of a neurologic event of 20% after a follow-up of 39 months.

Investigation of the brain

CT SCAN
Ischemic lesions are characterized by a hypodensity of the cerebral tissue, but 60% of the CT scans are falsely negative. This hypodensiry generally appears only after 36 hours. However, there are early signs during the first four hours that can be identified by scrupulous analysis. These may be a hyperdense medial cerebral artery (MCA), a disappearance of the lenticular nucleus, or indirect

15

FIG. 1 Angiography of a pre-occlusive stenosis. A - The stenosis at the origin of the ICA can be estimated at more than 90%. B - The asymmetrical intracerebral contrast distribution depicts the significant hemodynamic consequences of the stenosis, which justifies the term pre-occlusive.

VASCULAR

EMERGENCIES

signs of cerebral edema such as the disappearance of the cerebral sulci and attenuation of the cortical layer (Fig. 3). Some of these elements, like cerebral edema and a hyperdense MCA, have a poor prognosis. Hyperdensity of the MCA is an early sign, which has a sensitivity of 78% and a specificity of 93%. Disappearance of the lenticular nucleus indicates occlusion of the proximal MCA, with a sensitivity of 92% after 6 hours. Diffuse hypodensity and disappearance of the cortical sulci are correlated with a high mortality and prohibit thrombolysis.

DIFFUSION AND PERFUSION-WEIGHTED MR IMAGING

Diffusion. Diffusion imaging allows the diagnosis of ischemic lesions in the hyperacute phase of a cerebrovascular event. The diagnosis is made when an area of high intensity is found on the diffusion scan after applying diffusion gradients, whereas no signal is detectable before these gradients are applied (Fig. 4). The performance of this technique in the early detection of ischemic lesions is good, with a high sensitivity and specificity of 88% and 90%, respectively

16

FIG. 2 Two angiographic examples of a floating thrombus. Small volume thrombus distal to an ICA stenosis > 90%. A - High-volume thrombus in the ICA lumen without occlusion, distal to a stenosis of 60%. B - Despite the smaller degree of stenosis, the thrombus in B is more threatening.

FIG. 3 Early sign of stroke on CT scanning: hyperdense median cerebral artery (arrow).

URGENT CAROTID
[14,15]. False-positive results are rare. Detection limits are very small volume lesions, infratentorial localizations and TIAs. The volume of the ischemic lesion as measured with this technique has a prognostic value and is correlated with the initial clinical score and the situation after three months. By means of an apparent diffusion coefficient graph, any artifacts causing false-positive results can be eliminated. In emergency conditions, it is an essential investigation to confirm the diagnosis of a cerebrovascular event in the hyperacute phase and to appreciate the extent of irreversible damage. This image analysis takes only a few minutes. Thus, diffusion MR imaging is superior to CT scanning for the early detection of ischemic events and to quantify the extent of the ischemic cerebral area. Perfusion. Combining angiographic techniques with MR imaging enables investigation of the larger vessels of the brain (time of flight, phase contrast) . Investigation of the cerebral microcirculation by means of MR imaging can be performed using endogenous and exogenous tracers. A perfusion image is obtained by sequential scanning of the variation in signal intensity during passage of the contrast agent. Detection of diffusion asymmetries between cerebral areas allows assessment of the twi-

SURGERY

light zones around the ischemic lesion that represent poorly perfused but viable tissue areas at the time of investigation (Fig. 5). Quantitative data, such as cerebral vascular volume, cerebral flow and transit time can also be obtained, which allow quantitative analysis of the regional hemodynamic disturbances. Subtraction of the volume of the irreversible lesions, as detected by diffusion from the volume of the area with hemodynamic disturbance as shown by perfusion, allows for the assessment of the volume of the ischemic twilight area. The volume of the area with hemodynamic disturbance as measured by perfusion MR imaging is better correlated with the clinical evolution on the short and intermediate term than the volume as measured by diffusion imaging [14,15]. Thus, diffusion- and perfusion-weighted MR imaging appears to be an essential diagnostic tool in patients presenting with a cerebrovascular event in the acute phase. CT scanning merely has a poor predictive value. However, some uncertainties remain. The reversibility of the lesions as observed with diffusion MR imaging, especially in TIAs, illustrates the limitations of diffusion MR imaging in the diagnosis of irreversible cerebral lesions.

Rapid diagnosis of carotid lesions

The occurrence of a stroke requires therapeutic action as soon as possible. Noninvasive investigation should assess the etiology of the stroke. The diagnostic arsenal comprises several possibilities. Carotid duplex scanning is the primary investigation for a quick diagnosis of extracranial carotid lesions with a high sensitivity of 80%, a high specificity of 90% and an excellent correlation with angiography [16]. Transcranial doppler (TCD) is important to evaluate the cerebral collateral circulation, the cerebral vascular reserve capacity and to detect intracerebral stenoses [17]. This technique, however, is not simple and requires know-how and expertise which limits its use in emergency situations. Furthermore, it is not yet completely evaluated. Angio-CT scanning has a high sensitivity and specificity of 90% in the diagnosis of severe carotid stenoses and allows simultaneous analysis of the cerebral tissue. By means of an angio-CT, carotid stenoses may be evaluated via two-dimensional reconstructions, measuring the residual lumen in

17

FIG. 4 MR imasing diffusion image of an acute phase ischemic stroke.

VASCULAR

EMERGENCIES

18

FIG. 5 Development of a stroke in the territory of the median cerebral artery on diffusion and perfusion MR imagins. At the second hour, the diffusion imases show a small infarction (A) in the center of an ischemic penombra, even more visible on the perfusion MR imase (B) (arrow). At 24 hours, followins successful revascularization, the ischemic penombra has almost disappeared on the perfusion MR imase (D); the diffusion MR imase shows the clearly limited infarcted area (C).

T
each slice, thus showing the most stenotic area of the ICA as compared with the size in a more distal, normal area (NASCET method). Angio-MR gives exact information about the carotid stenosis and the intracerebral vascularization after a few minutes of image acquisition [14]. Appreciation of the stenosis with angio-MR may overestimate the lesion. The neurologic condition of the patient allowing cooperation during the investigation is an important limitation of the outcome of the MR and CT investigations. Subtotal stenoses are difficult to discern from carotid thromboses on MR or duplex scans because of the very low flow beyond the lesion. Despite its own neurologic morbidity [18], angiography remains the reference standard, particularly in doubtful cases, to appreciate the extracranial carotid lesion and to investigate the intracranial vessels without artifacts due to the low flow state. hemorrhagic strokes is estimated to be 15% of all strokes [22]. In general, the indications for emergency surgery are not yet clearly defined. On the one hand, surgery is able, through revascularization of the viable areas, to restore a deficit and to benefit the patient. On the other hand, surgery might also revascularize an area already lost or cause a complication due to the intervention. In other words, it may bring no benefit at all or cause deterioration. The problem is therefore to identify patients that will benefit. Today no prospective randomized trial can answer this dilemma. The data available are based on the experience of few groups and on mostly cohort studies with loose and variable inclusion criteria. Our experience is also not free from this criticism. From our medical dossiers and operation reports, we selected 15 out of 1200 cases of carotid surgery in a period of 10 years (January 1992 to December 2001), which were considered as emergency cases on the basis of an interval below 24 hours between diagnosis and intervention (Table II).

Indications and results

In 15% to 30% of the cases with cerebral ischemia, a stenosis of the cervical ICA is found [19]. The aim of early surgery after a stroke is on the one hand to restore the cerebral vascularization of the twilight zone, and on the other to exclude the emboligenic lesion in order to avoid a recurrence. This recurrence risk was found to be 3% to 5.9% in the medical arm of the NASCET study and in natural history studies. Numerous surgical studies have found discouraging results (Table I), which has led many groups to refrain from this kind of surgery. Most of these studies were performed before the introduction of CT scanning, which in part explains the observed results, considering the incidence of

Crescendo TIA
Recent series from the literature have shown encouraging results in patients with an unstable neurologic condition, selected on the basis of the clinical presentation and results of the CT scan. Table III illustrates the results of urgent carotid surgery for a crescendo TIA. The therapeutic decision is clear in this selected patient group when an accessible lesion is found. Hence, a surgical reconstruction is indicated in cases with the shortest delay, with reasonable success rates. Thus, complete healing rates of 71% to 100% can be reached. In our experience,

19

1st author [ref.]

Rob [20]

Year of ** publication
1969

Uelay

I 101 n\i

Improved N (%) 21 17 5 (28) (34) (15)

Unchanged N (%) 32 9 10 (43) (18) (31)

Worsened N (%}

Mortality N (%)
21

74 50

No No Yes

A few days < 14 days < 24 hours

(28)

Blaisdell[l] Bone [21]

1969 1990

3 8

(6) (25)

21
9

(42)
(28)

VASCULAR

EMERGENCIES

Neurologic state CTIA Progressive stroke Severe stroke

Number

Subtotal stenosis
1 6 3

Stenosis >75%

Floating thrombus

Improved neurologic state

4 6 2

Unchanged neurologic state

1 _

Death

4 8 3

2 _

i 2..
.

1 1

Subtotal stenosis + floating thrombus ** 1 subtotal stenosis + 1 stenosis >75% CTIA: crescendo transient ischemic attack

1st author [ref.]

Year of publication
1993 1994 1999

Number
12 14 21

Healing

Minor stroke

Severe stroke

Mortality

Wilson [23] Gertler [24]

100

98 71 10 19

20

Eckstein [25]

TIA: transient ischemic attack

crescendo TIAs accounted for 27% (4/15) of the indications for urgent carotid surgery. The results are in accordance with those recently reported in the literature showing 100% healing. In our opinion, crescendo TIAs form the best indication for urgent carotid surgery.

Progressive stroke
Table IV shows the results of urgent carotid surgery for progressive strokes with a normal CT scan or showing few lesions. Encouraging results compared with the natural history were reported with clinical improvement in 86% to 92% of the cases. In our experience, progressive stroke accounted for

53% (8/15) of the indications for urgent carotid surgery (Table II). The results showed clinical improvement in 75% (6/8), no change in one case (1/8), and one death (1/8). Progressive strokes form a good indication for surgery in selected patients. The aim of the selection, based on clinical and paraclinical criteria, is to identify and exclude patients in whom a revascularization is likely to have a more deleterious than beneficial effect. This may be the case in patients showing a massive lesion, a cerebral hemorrhage or a substantial impairment of their conscience. An accurate selection and preoperative work-up are necessary for a favorable outcome. For these patients, Brandl et al. [27] advise not to perform a pre-operative angiogram to reduce the neurologic morbidity and to arrange for an experienced team of carotid surgeons, with constant

URGENT CAROTID SURGERY

, r 1st author r [re/] Greenhalgh [26] Gertler [24] Brandl [27] Eckstein [25]

v / ^rr , scan xr ,_Year of ... ;. Number LI m publication %

~ , Delay > <24h A few days <24h

Improved Unchanged Worsened ,, , ,., , ,* i ,, i , , Mortality neurologic state neurologic state neurologic state A7 /0/, *% N (%) N N *(%) N *(%)

1993 1994 1998 1999

22 70 12 34

87 43
100 100

19 60 11 26

(86) (86) (92)

(76.5)

1 (4) 5 0 0 (7)

1 (4) 3 (4)

1 (4) 2 0 2 (3)

1 (8) 6
(17.5)

24 h

(5.8)

monitoring of the hemodynamics peri-operatively to ensure an optimal stability of the cerebral perfusion.

MODERATE AND REGRESSIVE STROKES IN THE ACUTE PHASE

The treatment of moderate ischemic strokes remains controversial. The initial attitude based on

the results of urgent carotid surgery from the 1960s and 1970s has led to the temporization of any intervention until after six to eight months, in which the cerebral tissue can recover (Fig. 6). In our experience, acute ischemic strokes accounted for 20% (3/15) of the indications for urgent carotid surgery. The postoperative results were an improvement in two cases (2/3) and one death

21

FIG. 6 CT scan of an acute stroke with hemorrhasic petechia and edema around the lesion (A). The same patients three months later (B).

VASCULAR

EMERGENCIES

22

(1/3). Two recent prospective studies (Table V) showed encouraging results of carotid surgery after a little disabling, ischemic stroke (Rankin scale 1 to 3). These authors concluded that early carotid surgery (within six weeks) is feasible in the acute phase of the stroke, showing a low risk of a postoperative hemorrhagic stroke and a reduction of the recurrence risk. The risk of a postoperative hemorrhagic stroke was below 1.2% in these studies. The stroke recurrence risk in the acute phase was 4.5% in patients scheduled for delayed surgery after a regressive stroke in the NASCET study. These different series showed that there was no increase in operation risk during carotid surgery in the acute phase of a stroke. The absence of a homogeneous patient selection in these series and the different study sizes do not allow for a statistical analysis of the different risk factors studied. In general, clinical and paraclinical selection of the patients in these series allowed exclusion of cerebral hemorrhages, strokes with coma and major strokes covering more than one third of the anterior vascular area on CT scan. The indication for urgent carotid surgery in severe strokes (Rankin superior to 3) in the acute phase is very controversial because of the overall poor results. Few series report results better than the natural course. Eckstein et al. [25] reported on a series of 16 patients and found a clinical improvement of 56.2% (9 cases) and a mortality of 19% (3 cases). At present it is difficult to select from these patients a group that might benefit from urgent carotid surgery. The various predictive factors studied in the literature have been disputed successively

without clear evidence. The arbitrary time limit of 6 hours between the stroke and surgical treatment was not confirmed by recent series, in which the patients were operated within 24 hours. Furthermore, the presence of emboli at the level of the MCA, which is considered as a poor prognostic factor and a centra-indication for surgery [5], has given rise in recent studies to combine thrombolysis with carotid surgery, with encouraging results. This combined approach is attractive for patients without centra-indications for thrombolysis (Fig. 7). ANATOMICAL EMERGENCIES Subtotal stenoses. In our experience, a subtotal stenosis was present in 80% (12/15) of the indications for urgent carotid surgery. The results in these patients showed clinical improvement in 75% (9/12) with a mortality of 8.3% (1/12). The natural history of these carotid lesions is poorly known because they are excluded from large prospective therapeutic trials. Therefore, this indication which is beyond discussion for the majority of surgeons, is not supported by the literature. Based on hemodynamic studies by means of TCD of high-grade stenoses [28], some centers advocate investigation of the collateral circulation in order to assess the stroke risk at the time of a carotid occlusion. This speculative attitude seems inappropriate to us, in view of the good surgical results obtained in these patients for asymptomatic carotid lesions. Floating thrombus. In our experience, a fresh carotid thrombus was present in 20% (3/15) of the patients undergoing urgent carotid surgery. Clinically one patient presented with a crescendo TIA

1st author [ref.] Piotrowski [7] Gasecki [6] Ricco [8] Ballotta [9] Eckstein [10]

Year of publication
1990 1994

Randomization
No
Yes

Number
129 100

Early surgery N/CMMR%

Delayed surgery N/CMMR%

47 / 5.3 58 / 5.2

82 / 2.5 42 / 4.8 72 / 2.8

2000 2002 2002

No
Yes

72 86
164

41/2

45/2
164 / 6.7

No

CMMR: cumulative morbidity and mortality rate

URGENT CAROTID SURGERY

and two patients with progressive strokes. Two patients improved and one died. Literature data are contradictory [11,12]. Concerning the treatment of acute cerebral ischemia, some favor urgent revascularization by means of surgery or thrombolysis, others prefer anticoagulant therapy followed by secondary surgery. The theoretical argument to postpone or refrain from surgery is the considerable risk of peroperative embolization. Pelz et al. [29] obtained results in favor of delayed surgery in a comparative study comprising 29 cases. Positive results for nondelayed surgery were also reported [11]. A consensus on the treatment of this pathology does not seem to show up. A pragmatic attitude related to the clinical situation and topographic presentation of the lesion should direct the therapeutic approach.

Principles of surgical treatment

INTRA-OPERATIVE MONITORING
Anticoagulant treatment should be initiated as soon as the diagnosis of ischemic stroke is made and centra-indications against heparin treatment are checked. The aim of this treatment is to avoid an extended thrombosis. As to normalization of the blood pressure, one should take into account the defense mechanisms of the cerebral perfusion which generate a relative hypertension. Accurate blood pressure monitoring, which is already important in carotid surgery, plays a crucial role here. Sudden and/or substantial variations in blood pressure may cause a disturbance of the autoregulation of the cerebral circulation, which can be induced

23

FIG. 7 Urgent carotid surgery for progressive stroke. Completion angiography after ICA reconstruction shows thrombosis of the median cerebral artery (A). After 20 minutes of local thrombolysis, angiography shows patency of this artery (B).

VASCULAR

EMERGENCIES

by cerebral edema, hemorrhage or hypoperfusion of the ischemic twilight zone. Intra-operative monitoring by means of EEG, TCD, evoked potentials and others is recommended by certain groups. However, the meaning of these variations observed during the acute phase of a stroke is unknown. Moreover, the delay due to their use limits their indications in emergency cases. Finally, many groups including ours recommend the systematic use of a shunt in case of urgent surgery when a neurologic deficit is present, which makes cerebral monitoring redundant.

TECHNICAL ASPECTS
The carotid reconstruction does not have particular characteristics when performed urgently. The exposition should be performed very carefully because of the risk of embolization. If the ICA is patent, it should be clamped in the first place, before finishing the dissection of the carotid bifurcation. If one suspects a carotid thrombosis of a floating thrombus at the level of the ICA, one should refrain from clamping it. Then it is important to clamp the common carotid artery first, fol-

lowed by the external carotid before completing the dissection and performing the arteriotomy of the nonclamped ICA. In case of a thrombus, opening of the ICA should be done in its proximal part in order to let the thrombus come out spontaneously, supported by the residual intracranial blood pressure. One should avoid, whenever possible, any embolectomy. In our practice a shunt is placed routinely to prevent expansion of the cerebral ischemia in the presence of a carotid stenosis. In case of carotid thrombosis, the shunt shortens the period of cerebral ischemia but also ensures a progressive cerebral reperfusion, as the flow through the shunt is similar to an ICA stenosis of 70%. A perfect carotid endarterectomy and a completion angiogram are required to minimize the risk of preoperative embolization. In the series of Ballotta et al. [9] the postoperative neurologic morbidity was due to a postoperative embolization.

POSTOPERATIVE MONITORING
One of the necessities for the postoperative period is the prevention of the reperfusion risk (Fig. 8). The therapeutic arsenal comprises symp-

24

FIG. 8 Reperfusion injury. A - Massive cerebral bleedins occurring in the immediate postoperative phase, leadins to patient's death. B - Bleeding occurring at the third postoperative day for a stroke, causing temporary aggravation with adequate final outcome.

URGENT CAROTID SURGERY

tomatic measures such as maintenance of a normal or slightly subnormal blood pressure, cerebral protection by a quick recovery after general anesthesia, and the use of edema-preventing treatments. Monitoring the cerebral circulation by means of TCD can identify the first signs of cerebral hyperperfusion, which allows for symptomatic treatment before reperfusion injury can occur [30]. Secondly, from these the surgeons select only those to investigate who may be treated surgically. The selection criteria vary among centers and in different time periods. It is therefore impossible to know what the prevalence is of the indications for urgent surgery among the total of ischemic neurologic events, or even among the total of ischemic events of carotid origin. The fundamental challenge of this kind of surgery is therefore primarily to identify and select the subgroups of patients for whom surgery can be beneficial. The second one is to organize the logistics to perform this selection and surgery under acceptable conditions and delay. MR imaging more than CT scanning offers possibilities for basic investigations, which should answer the questions raised. However, validation by means of prospective studies of the different qualitative and quantitative prognostic factors as given by MR imaging is needed. The present availability of this method is still limited and its use is far from ubiquitous. Percutaneous carotid angioplasty may play an interesting role, as it can be performed quickly. Intracerebral thrombolysis, with or without surgery or percutaneous angioplasty, offers interesting perspectives for the treatment of intracranial lesions.

THE PLACE OF THROMBOLYSIS AND ENDOVASCULAR TREATMENT

The natural history of acute ischemic strokes with intracranial carotid thrombosis and/or thrombosis of the MCA shows a mortality greater than 53% [31]. Several recent case reports showed good results of a combination of surgery and in-situ thrombolysis in patients with extra- and intracranial carotid lesions. Eckstein et al. [32] reported eight patients with a carotid occlusion or severe stenosis associated with MCA embolization treated with thrombolysis and carotid surgery. As a result, five patients suffered a minor stroke (62.5%), two were cured (25%) and one died (12.5%). However, despite the low hemorrhagic complication rates reported, the precise risk of an intracerebral hemorrhage after urgent thrombolysis remains to be shown. In the literature few data are given on the endovascular treatment in the acute phase of a stroke. In a series of 33 patients with a major disabling stroke, Endo et al. [33] reported four successful cases after intra-arterial thrombolysis, three times in combination with carotid angioplasty and once with surgery. The mortality was 58%. Finally, some heterogeneous and poorly documented publications show good results after thromboaspiration and percutaneous angioplasty of a carotid occlusion in the acute phase of a stroke.

25

Conclusion
Urgent carotid surgery is certainly surgery of the future, of which the exact indications need to be further defined. The dogma of delayed surgery after a moderate stroke appears obsolete. Prospective studies should be undertaken to address the following essential subjects: - assess the selection criteria for carotid surgery after moderate or regressive stroke, - accurately assess the optimum delay for surgery after moderate or regressive stroke, - assess the arguments in favor or against surgery for a floating thrombus and a subtotal stenosis in symptomatic and asymptomatic patients, - assess the results of angioplasty versus surgery in acute situations. All these studies should benefit by initiating stroke centers, which are essential for this purpose and for the treatment of the patients.

Comments
This general overview generates more questions than answers. The surgical series are mostly retrospective and report only on highly selected cases, of which the selection criteria are quite variable and unclear. This selection has a double effect. In the first place, the first aid department offers only few patients with a neurologic deficit with a potential carotid origin to offer to the vascular surgeons.

VASCULAR

EMERGENCIES

R E F E R E N C E S

26

1 Blaisdell WF, Clauss RH, Galbraith JG et al. Joint study of extracranial arterial occlusion. IV. A review of surgical considerations. JAMA 1969; 22; 209:1889-1895. 2 Easton JD, Sherman DG. Stroke and mortality rate in carotid endarterectomy: 228 consecutive operations. Stroke 1977; 8: 565-568. 3 Hertzer NR, O'Hara PJ, Mascha EJ et al. Early outcome assessment for 2228 consecutive carotid endarterectomy procedures: the Cleveland Clinic experience from 1989 to 1995./ Vase Surg 1997;26:1-10. 4 Toni D, Fiorelli M, Gentile M et al. Progressing neurological deficit secondary to acute ischemic stroke. A study on predictability, pathogenesis, and prognosis. Arch Neurol 1995; 52: 670-675. 5 Meyer FB, Sundt TM Jr, Piepgras DG et al. Emergency carotid endarterectomy for patients with acute carotid occlusion and profound neurological deficits. Ann Swrg-1986; 203: 82-89. 6 Gasecki AP, Ferguson GG, Eliasziw M et al. Early endarterectomy for severe carotid artery stenosis after a nondisabling stroke: results from the North American Symptomatic Carotid Endarterectomy Trial. / Vase Surg 1994; 20: 288 - 295. 7 Piotrowski JJ, Bernhard VM, Rubin JR et al. Timing of carotid endarterectomy after acute stroke. JVasc Surg\99Q; 11: 45-51. 8 Ricco JB, Illuminati G, Bouin-Pineau MH et al. Early carotid endarterectomy after a nondisabling stroke: a prospective study. Ann Vase Surg 2000; 14: 89 - 94. 9 Ballotta E, Da Giau G, Baracchini C et al. Early versus delayed carotid endarterectomy after a nondisabling ischemic stroke: a prospective randomized study. Surgery 2002 ;131: 287-293. 10 Eckstein HH, Ringleb P, Dorfler A et al. The carotid surgery for ischemic stroke trial: a prospective observational study on carotid endarterectomy in the early period after ischemic stroke. J Vase Surg 2002; 36:997 -1004. 11 Christopher M. Lofthus CM. Hyperperfusion syndrome following carotid endarterectomy. In: Lofthus CM, Kresowik TF (eds). Carotid artery surgery, Thieme, New York, 2000 : pp 321-327. 12 Combe J, Poinsard P, BesancenotJ et al. Free-floating thrombus of the extracranial internal carotid artery. Ann Vase Surg 1990; 4:558-562. 13 Nicholls SC, Bergelin R, Strandness DE. Neurologic sequelae of unilateral carotid artery occlusion: immediate and late. J Vase Surg 1989; 10: 542 -547. 14 Grunwald I, Reith W. Non-traumatic neurological emergencies: imaging of cerebral ischemia. EurRadioim?; 12:1632-1647. 15 Cosnard G, Duprez T, Grandin C. et al. Diffusion- and perfusionweighted MR imaging during the hyperacute phase of stroke. J Radial 2000; 81: 858 -869. 16 Long A, Lepoutre A, Corbillon E, Branchereau A. Critical review of non- or minimally invasive methods (duplex ultrasonography, MR- and CT-angiography) for evaluating stenosis of the proximal internal carotid artery. Eur J Vase Endovasc Surg 2002; 24:43-52.

17 Baumgartner RW, Baumgartner I, Mattle HP, Schroth G. Transcranial color-coded duplex sonography in the evaluation of collateral flow through the circle of Willis. AJNRAmJNeuroradiol 1997;18:127-133. 18 Bendszus M, Koltzenburg M, Burger R et al. Silent embolism in diagnostic cerebral angiography and neurointerventional procedures: a prospective study. Lancet 1999 6; 354:1594-1597. 19 Timsit SG, Sacco RL, Mohr JP et al. Early clinical differentiation of cerebral infarction from severe atherosclerotic stenosis and cardioembolism. Sfrofa>1992; 23: 486-491. 20 Rob CG. Operation for acute completed stroke due to thrombosis of the internal carotid artery. Surgery 1969; 65: 862-865. 21 Bone G, Ladurner G, Waldstein N, Rendl KH. Acute carotid artery occlusion - Operative or conservative management. Eur Neurol 1990; 30: 214-217. 22 Mead GE, O'Neill PA, McCollum CN. Is there a role for carotid surgery in acute stroke? Eur J Vase Endovasc Surg 1997; 13: 112-121. 23 Wilson SE, Mayberg MR, Yatsu F, Weiss DG. Crescendo transient ischemic attacks: a surgical imperative. Veterans Affairs trialists. /Vase Surg 1993; 17: 249-255. 24 Gertler JP, Blankensteijn JD, Brewster DC et al. Carotid endarterectomy for unstable and compelling neurologic conditions: do results justify an aggressive approach? JVasc Surg 1994; 19:32-40. 25 Eckstein HH, Schumacher H, Klemm K et al. Emergency carotid endarterectomy. Cerebrovasc Dis 1999; 9: 270-281. 26 Greenhalgh RM, Cuming R, Perkin GD, McCollum CN. Urgent carotid surgery for high risk patients. Eur J Vase Surg 1993; 7SupplA:25-32. 27 Brandl R, Brauer RB, Maurer PC. Urgent carotid endarterectomy for stroke in evolution. Vasa 2001; 30:115-121. 28 Stork JL, Levi CR, Chambers BR et al. Possible determinants of early microembolism after carotid endarterectomy. Stroke 2002; 33:2082-2085. 29 Pelz DM, Buchan A, Fox AJ et al. Intraluminal thrombus of the internal carotid arteries: angiographic demonstration of resolution with anticoagulant therapy alone. Radiology 1986; 160: 369-373. 30 Dalman JE, Beenakhaus 1C, Moll F et al. Transcanial doppler monitoring during carotid endarterectomy helps to identify patients at risk of postoperative hyperperfusion. Eur JVase Endovascular Surg 1999; 18: 222 - 227. 31 Jansen 0, Von Rummer R, Forsting M et al. Thrombolytic therapy in acute occlusion of the intracranial internal carotid artery bifurcation. AJNRAmJNeuroradiolWS; 16:1977-1986. 32 Eckstein HH, Schumacher H, Dorfler A et al. Carotid endarterectomy and intracranial thrombolysis: simultaneous and staged procedures in ischemic stroke./ Vase Surg 1999; 29: 459-471. 33 Endo S, Kuwayama N, Hirashima Y et al. Results of urgent thrombolysis in patients with major stroke and atherothrombotic occlusion of the cervical internal carotid artery. AJNR Am JNeuroradiolim; 19:1169-1175.

BLUNT INJURY TO THE CAROTID AND VERTEBRAL ARTERIES

RAMON BERGUER

The incidence of blunt injury to the carotid and vertebral arteries (BICV) is low but its outcome is characterized by high mortality and morbidity. BICV constitutes 0.5% to 1.0% of all blunt trauma admissions [1-3]. The incidence is much lower in the pediatric population, where it is reported to be 0.03% [4], Such a low incidence makes it inefficient to angiographically screen every patient for vascular injury arriving to a trauma center with blunt trauma of the head and neck. However, when patients who display other markers for BICV such as cervical/skull fracture or neurologic deficits are selected for diagnostic arteriography, 29% to 44% of them will be found [1,5,6] to have BICV. The outcome of blunt vascular injuries in the neck is most serious [2,3,7,8]. The incidence of stroke following demonstrated BICV ranges from 16% to 60% for carotid injuries [9,10] to 14% for vertebral injuries [7]. The death rate for BICV ranges between 25% and 31 % [2,9], being 13% to 57% [3,7,10] for carotid injuries and 4% to 67% for vertebral injuries [3,7].

3_
27

Mechanisms of injury and arterial lesions

The mechanism for disruption of the extracranial carotid artery can be a direct blow, either external (Figs. 1A and IB) or transoral, hyperextension or hyper-rotation of the neck with stretching of the artery over the transverse process of Cl (Fig. 2), direct contusion by a fragment of frac-

tured mandible, and proximity involvement of the internal carotid artery in a fracture of the temporal bone. An object striking the neck from the outside can exert a direct blow to the common or internal carotid arteries (this includes the mechanism of safety belt injury). An intra-oral blow occurs when a person falls with an object (such as a toothbrush or a lollipop) in the mouth. The object strikes the internal carotid artery through the tonsillar

VASCULAR

EMERGENCIES

FIG. 1 There is a dissection of the innominate artery extending into the subclavian artery and occluding the common carotid artery. This 24-year-old man suffered a rollover car accident, crushing his anterior chest and neck against the steering wheel. A - Arteriogram showing a luminal defect in the subclavian artery and the occlusion of the common carotid artery. B - Operative view of the dissected innominate artery.

28

fossa. The internal carotid artery traverses the temporal bone through a canal in which the adventitia of the artery and the periosteum of the bone are closely connected. A fracture and displacement of the bone entails disruption of the adjacent arterial wall (Fig. 3). Following BICV by any of the mechanisms cited above, the neurologic manifestations will be delayed for hours or days in roughly one half of cases [11-13]. The vertebral artery is rarely damaged by a direct external blow, since it is protected by bone as it ascends through the neck. In its V2 segment, the vertebral artery runs in an osteomuscular canal formed by the foramina transversaria and the intertransversaria muscles. Fractures of the lateral mass of the cervical vertebrae involving the foramen transversarium, and vertebral fractures with luxation/subluxation of the cervical spine, can result in distortion of this osteomuscular conduit containing the vertebral artery (Fig. 4) and its venae comitantes and injury to these vessels (Fig. 5). This latter mechanism is frequently seen in motor vehicle accidents and in near hanging injuries. In a motor vehicle accident or during brusque chiro-

FIG. 2 False aneurysm of the internal carotid artery from contusion of the latter against the transverse process of C1.

BLUNT INJURY TO THE CAROTID AND VERTEBRAL

ARTERIES

FIG. 3 False aneurysm of the petrosal internal carotid artery found after evaluation of a 30-year-old man who suffered a skull base fracture in a car accident.

FIG. 4 Symptomatic dissection of the V2 segment of the vertebral artery in a 27-year-old woman following a fall from a horse.

29

practic manipulation (Fig. 6), there may be sudden rotation or hyperextension of the head. The vertebral artery has a redundant loop between Cl and C2 to accommodate the wide arc of rotation that takes place between these two vertebrae. This rotation is approximately 80 degrees or about one half of the range of rotation of the entire neck. The vertebral artery is also attached to Cl and C2 by connections between its adventitia and the periosteum of the transverse foramina. During injury, particularly with extreme head rotation, the artery is stretched between Cl and C2 beyond its normal elastic range and literally snaps, the consequence being its dissection (Fig. 7), occlusion, or rupture. If it ruptures into the surrounding tissues, a false aneurysm develops; if it ruptures into a closely adjacent and also damaged vertebral vein an arteriovenous fistula ensues (Fig. 5).

BICV may result in a wide spectrum of lesions ranging from luminal irregularities (flaps, intramural hematoma, dissection), to disruption of the arterial wall (false aneurysm, fistula), to occlusion of the lumen (Fig. 8). Brain embolization may occur from an intimal flap, from the reentry tear of a dissection or from the tail of the thrombus that forms distal to an occlusion.

Diagnosis
SYMPTOMATOLOGY About one half of the patients who eventually turn out to have BICV enter the trauma unit without neurologic symptoms or signs. A substantial number (43% to 58%) of these originally asymptomatic patients will develop neurologic signs after their hospital admission [11-13].

VASCULAR

EMERGENCIES

30

FIG. 5 Ansiosram of a patient with a symptomatic arteriovenous vertebral fistula two years after blunt head and neck trauma. A - Arteriovenous fistula at V3. B - The venous end of the fistula has been obliterated by a detachable balloon. The vertebral artery intesrity has been preserved. Hypertrophy of the vertebral artery proximal to this chronic fistula is obvious.

The following specific findings markedly increase the likelihood that an individual that has suffered blunt trauma to the head and neck may harbor a BICV. 1 - A Horner's syndrome that occurs when there is disruption of the upper cervical ganglion by trauma (hyperextension or rotation) or of the peri-adventitial sympathetic fibers by dissection of the arterial wall. 2 - A neurologic deficit that may be the result of embolization (from the distal end of the dissection channel or from a thrombus beneath an intimal flap) or of ischemia (from low-flow compromise) due to the dissection or rupture of an artery resulting in its occlusion or near occlusion. In the setting of a trauma unit, the detection of neurologic deficits may be seriously hampered by high levels

of alcohol or drugs in blood, shock from other internal injuries, or concomitant head injuries. 3 - The finding of a skull base (Fig. 3), temporal bone, or cervical fracture (or luxation/subluxation) substantially increases the likelihood of concomitant injury to the internal carotid or vertebral arteries that are intimately associated to these bony structures. 4 - If sudden hyperextension or head rotation movement during the accident are suggested by the patient's description of the accident or by associated injuries to the head and neck. Hyperextension and hyper-rotation are the most common mechanisms for stretching injury of the carotid or vertebral arteries. In the series of Biffl et al. [11], setting up a screening protocol in their trauma unit for these

BLUNT INJURY TO THE CAROTID AND VERTEBRAL ARTERIES

FIG. 6 This 20-year-old man complained of neck pain after a basketball game. He underwent chiropractic treatments and twelve hours later developed symptoms of brain stem infarction. There is dissection of the distal vertebral artery (arrow). A ventriculostomy tube is in place. The patient eventually died.

1
31

FIG. 7 This 47-year-old woman developed a symptomatic vertebral artery dissection following a ski accident (arrow). B - This arteriogram, obtained six months later, shows the dilatation of the false lumen involving the V3 and V4 segments.

VASCULAR

EMERGENCIES

FIG. 8 Operative photographs of a 29-year-old worker who had his neck crashed between the posts of a hydraulic case used to repair electrical cables. A - View at operation of the hematoma in the subadventitial space of the common carotid artery. B - A cylinder of intima and media that broke off and impacted into the carotid bulb. C - Open sesment of resected common carotid artery (it was replaced with a prosthesis) showing the folded layer of intima and media and the thrombus that formed distal to it.

32
markers of BICV resulted in a 10-fold increase of this diagnosis. A seatbelt sign is not considered to be a good indicator of blunt carotid injury in retrospective studies. One review of 131 patients with seatbelt sign found only one blunt carotid injury among them (0.76%) [14]. However, if the patient has a seatbelt sign and other abnormal findings (neurologic deficit, skull or cervical fractures, etc.), then there is a good indication for angiographic screening [15] and a high probability of finding a BICV. reliable when the lesion is in the distal half of the cervical internal carotid artery, where most blunt injuries occur (skull base, infrapetrosal segment). Duplex can theoretically provide evidence of injury in the nonvisualized proximal common carotid and distal internal carotid artery by detecting abnormal flow signals. However, the noncritical stenosis created by small flaps or some dissections will go undetected while they still have a high potential for embolization and delayed occlusion. Additionally, the accuracy of duplex ultrasound of the carotid bifurcation is notoriously operator-dependent and it is unlikely that such a service will be available at all times in a trauma unit. Finally, neck swelling and associated injuries may make duplex scanning difficult and less reliable. CT angiography is not a sensitive tool to detect small defects of the arterial wall, but CT scanning of the brain is a good predictor of outcome. A large multicenter study [13] showed that those patients who showed a cerebral infarct in the admission CT of the brain had a high mortality (47%) and only

IMAGING TECHNIQUES
A few studies have analyzed the screening power of ultrasound and computed tomography (CT) angiography in the initial diagnosis of blunt vascular injury to the head and neck [1,13,16,17]. The sensitivity of the ultrasound in detecting blunt carotid injury was a remarkably high 86% in a multicenter trial [13]. Still, the study missed carotid lesions located in the distal cervical carotid and naturally the authors had no information about the intracranial carotid artery. Duplex ultrasound is not

BLUNT INJURY TO THE CAROTID AND VERTEBRAL

a poor chance (29%) of good neurologic recovery. On the contrary, of those patients with a normal CT of the brain on admission to the trauma unit, none died and 67% went on to a good neurologic recovery. Combined magnetic resonance imaging-magnetic resonance angiography has been advocated as a tool to simultaneously image the carotid and vertebral arteries and survey the brain parenchym. The experience with this modality is limited. As a triage tool for blunt injury it is probably not available in short notice in most trauma units. Even when available, placing the patient in a magnet may present problems because multiple trauma patients often require hardware to immobilize fractures and ventilatory support. Levy et al. [18] reported a low sensitivity of magnetic resonance for detecting discrete vascular blunt injuries. There is consensus among the authors of the largest series [11,12] that selection for arteriography within the large population of blunt trauma patients should be done on the basis of the clinical findings: stroke, Horner's syndrome, massive soft tissue injury of the neck, the mechanism of injury (severe hyperflexion-hyperextension and rotational injury), and concomitant bony findings (skull base fracture, fracture across the foramen transversarium, and temporal bone fracture). In patients with blunt injury to the head and neck, there is a strong correlation between those who present with a low Glasgow stroke score (equal to or less than six) and the finding of BICV.

ARTERIES

There is evidence [12] favoring anticoagulation with heparin (to be followed by Coumadin) in the treatment of BICV. The authors of the two largest series [11,12] make a good argument for systemic heparinization in the absence of centra-indications in multiple trauma patients with BCVI. Although their studies are explicidy detailed and include the largest series of BICV, they are retrospective reviews that were not analyzed using the technique of matched pairs. However, to avoid bias, they excluded patients with intracranial fistulae and those with transection or disruption of the internal carotid artery when searching by regression analyses the heparin effect on outcomes. Another retrospective review [19] of a smaller series did not find any advantage of heparin over antiplatelet therapy. It is important to weigh the advantages of anticoagulation treatment against the potential deleterious effects at the level of associated lesions. Also, cerebral infarction can be transformed in a cerebral bleeding which is most often catastrophic. Finally, some associated visceral lesions, in the case of a polytraumatic patient, mean an absolute contra-indication for anticoagulation treatment, at least in the acute phase.

SURGICAL INDICATIONS
Unfortunately there is no clearly established algorithm for surgical treatment and the indications must be discussed on a case-by-case basis. The emergency indications, always very difficult, should be distinguished from the secondary, more elective surgical indications. In the acute phase, a complete rupture of a carotid or vertebral artery or a partial rupture with evident thrombus are indications for surgical intervention. However, deep coma and severe neurologic deficit with or without extensive cerebral infarction are contra-indications for surgery. In contrast, the absence of lesions at CT scanning and a fluctuating neurologic status with preserved consciousness are elements that should speed up the decision to operate. In neurologically uninjured patients, the same arterial lesions require surgical repair, however surgery can be postponed for several hours, allowing assessment of the other polytraumatic injuries and to improve the general condition of the patient. During follow-up, angiography or CT angiography can depict specific arterial lesions that might require surgery: false aneurysms, dissection, or arteriovenous fistulae. Lesions that induce transient neurologic accidents or fistulae that cause a bruit experienced by the patient and

33

Treatment
The goals of treatment in emergency and semiemergency are to restore cerebral perfusion distal to an occlusive lesion and to prevent embolization from a thrombus that developed at the traumatized level. The goals during follow-up, if necessary, are to treat a false aneurysm, dissection, or arteriovenous fistula.

ANTICOAGULATION THERAPY
Intravenous heparin is administered in the acute phase and followed by oral anticoagulation, the duration of which depends on residual lesions identified during follow-up. The purpose of this treatment is to prevent the formation, propagation, and/or embolization of thrombus that developed at the level of the intimal rupture.

VASCULAR

EMERGENCIES

aneurysms or dissections containing thrombus despite anticoagulation, should be surgically repaired. Surgery is also indicated in case of rupture or complete thrombosis of the internal carotid artery or common carotid artery and in case of recurrent emboli resulting from arterial dissection despite anticoagulation therapy (Fig. 9). Certain aneurysms are asymptomatic. The anatomical characteristics of the infratemporal internal carotid artery and the suboccipital vertebral artery explain why false aneurysms can be surveilled because enlargement hardly occurs. As long as they behave asymptomatically, surgery is not necessary.

TECHNICAL ASPECTS
This surgery is extremely difficult and challenging, requiring unusual vascular access. Additional complicating factors include the emergency setting,

the frequently associated lesions of head and neck and extensive hematomas disturbing local anatomical landmarks. While vascular access to the common carotid artery and the bifurcation is not a problem, the majority of internal carotid artery trauma is located in the second part of the cervical region. In these circumstances it is recommended to perform nasotracheal intubation and mandible subluxation. Access to the ICA in the subparotid area requires transection of the digastric muscle and resection of the styloid apophysis. This extension allows exposure of the ICA just above the crossing of the ninth cranial nerve [20]. Access to the last centimeters of the ICA require control of the facial nerve, opening of the temporomaxillary joint, and partial ablation of the tympanal bone [21]. It must be emphasized that these complex procedures are

34

FIG. 9 A - Post-traumatic dissection of the vertebral artery that continues to be symptomatic in spite of appropriate anticoasulation. B - Exclusion bypass from the internal carotid to the vertebral artery at the level of the foramen rnasnum. The proximal vertebral artery was lisated.

BLUNT INJURY TO THE CAROTID AND VERTEBRAL ARTERIES

time consuming, which is disadvantageous in these emergency settings. Surgical repair for traumatic vertebral artery injury generally requires access to the distal vertebral artery in its suboccipital segment. This can be achieved either by a cervical-lateral approach between Cl and C2 [22] or a posterior approach between Cl and the occipital foramen [23]. The latter technique is not recommended in emergency settings since the patient must be in the ventral decubitus position. In traumatic arterial lesions, venous grafts are preferred. Clamping at the distal end is always delicate because of the fragile tissue and crushing of the intima. Furthermore, the surgical field is always narrow, allowing only limited movements. Therefore, distal arterial control can be achieved by means of an occluding balloon, causing limited intimal damage and enabling manipulation of the artery during suturing. If this technique is feasible and if 15 to 20 mm of the artery beyond the rupture are available, it is always recommended to perform an end-to-side anastomosis with closure of the transected end. This anastomotic configuration is easier to perform, allows better vision of the distal arterial segment, and provides a superior hemodynamic shape. Insertion of a shunt remains controversial. It is rarely applied because it complicates the procedure. ENDOVASCULAR TREATMENT False aneurysms can be managed with stents or stent-covered grafts. There are some initial reports of stenting as treatment for intimal flaps, proximal dissection entry points, and false aneurysms. The deployment of a stent inside a false aneurysm will not necessarily result in the thrombosis of its false lumen (Fig. 10) although such thrombosis can be

35

FIG. 10 A - This patient with BICV had two false aneurysms of the infrapetrosal internal carotid artery. B - Both false aneurysms were stented: the proximal one thrombosed; the sac of the distal false aneurysm remains patent.

VASCULAR

EMERGENCIES

induced by filling the aneurysm sac with coils through the mesh of the stent (Fig. 11). Endovascular treatment, especially in trauma patients, is specifically appealing because it is less invasive. However, this treatment comprises two drawbacks. First, short-term and long-term results are hardly known. It cannot yet be ruled out that a stent itself or the endovascular manipulations either initiate arterial occlusion or induce cerebral emboli, as already reported [24]. Second, the advantage of endovascular repair is based on the fast procedure in a trauma patient. However, this treatment requires specific expertise and infrastructure, which is not always available in every trauma center. Occlusions or neartotal occlusions result from fracture of the wall and distal embolization of a medial-intimal fragment (Fig. 8). This kind of lesion is unlikely to be successfully managed with endovascular techniques and farther distal embolization may occur during en-

dovascular manipulation. For this type of severe injury, open surgical techniques are advocated.

Summary
BICV has a low incidence but a high morbidity and mortality. The artery may be injured by a direct external blow or by adjacent bone displaced in a fracture or a luxation. Brusque stretching of the artery is a common mechanism of injury to the carotid or vertebral arteries in motor vehicle accidents. A diagnostic angiogram is indicated when a patient presents with specific features such as neurologic symptoms or craniofacial injuries. In the absence of contra-indications from associated injuries, heparin appears to be beneficial. Endovascular and direct surgical repair are recommended for specific lesions.

^
36

FIG. 11 A - This false aneurysm of the vertebral artery (C3 level) was stented but its false lumen did not thrombose. B - Coils were introduced into the false lumen through the mesh of the stent and eventually obliterated the false lumen.

BLUNT INJURY TO THE CAROTID AND VERTEBRAL ARTERIES

R E F E R E N C E S 1 Miller PR, Fabian TC, Croce MA et al. Prospective screening for blunt cerebrovascular injuries: analysis of diagnostic modalities and outcomes. Ann Swrg-2002; 236: 386-395. 2 McKevitt EC, Kirkpatrick AW, Vertesi L et al. Identifying patients at risk for intracranial and extracranial blunt carotid injuries. Am J Surg 2GQ2; 183: 566-570. 3 Berne JD, Norwood SH, McAuley CE et al. The high morbidity of blunt cerebrovascular injury in an unscreened population: more evidence of the need for mandatory screening protocols. JAm Coll Surg m\; 192: 314-321. 4 Lew SM, Frumiento C, Wald SL. Pediatric blunt carotid injury: a review of the National Pediatric Trauma Registry. Pediatr Neuroswrg 1999; 30: 239-244. 5 Kerwin AJ, Bynoe RP, Murray J et al. Liberalized screening for blunt carotid and vertebral artery injuries is justified. / Trauma 2001; 51: 308-314. 6 Biffl WL, Moore EE, Offher PJ et al. Optimizing screening for blunt cerebrovascular injuries. AmJSurg 1999; 178: 517-522. 7 Miller PR, Fabian TC, Bee TK et al. Blunt cerebrovascular injuries: diagnosis and treatment./ Trauma 2001; 51: 279-286. 8 Mclntyre WB, Ballard JL. Cervicothoracic vascular injuries. Semin Vase Surg 1998; 11: 232-242. 9 McKevitt EC, Kirkpatrick AW, Vertesi L et al. Blunt vascular neck injuries: diagnosis and outcomes of extracranial vessel injury. / Trauma 2002; 53: 472-476. 10 Biffl WL, Moore EE, Ryu RK et al. The unrecognized epidemic of blunt carotid arterial injuries: early diagnosis improves neurologic outcome. Ann Surg 1998; 228: 462-470. 11 Biffl WL, Moore EE, Offner PJ, Burch JM. Blunt carotid and vertebral arterial injuries. World J Surg mi; 25: 1036-1043. 12 Fabian TC, Patton JH Jr., Croce MA et al. Blunt carotid injury. Importance of early diagnosis and anticoagulant therapy. Ann Surg 19%; 223:513-525. 13 Cogbill TH, Moore EE, Meissner M et al. The spectrum of blunt injury to the carotid artery: a multicenter perspective. / Trauma 1994; 37: 473-479. 14 DiPerna CA, Rowe VL, Terramani TT et al. Clinical importance of the "seat belt sign" in blunt trauma to the neck. Am Surg 2002;68:441-445. 15 Rozycki GS, Tremblay L, Feliciano DV et al. A prospective study for the detection of vascular injury in adult and pediatric patients with cervicothoracic seat belt signs. / Trauma 2002; 52: 618-624. 16 Ofer A, Nitecki SS, Braun J et al. CT angiography of the carotid arteries in trauma to the neck. EurJ Vase Endovasc Surg 2001; 21:401-407. 17 Rogers FB, Baker EF, Osier TM et al. Computed tomographic angiography as a screening modality for blunt cervical arterial injuries: preliminary results. / Trauma 1999; 46: 380-385. 18 Levy C, Laissy JP, Raveau V et al. Carotid and vertebral dissections: three-dimensional time-of-flight MR angiography and MR imaging versus conventional angiography. Radiology 1994; 190: 97-103. 19 Wahl WL, Brandt MM, Thompson BG et al. Antiplatelet therapy: an alternative to heparin for blunt carotid injury. J Trauma 2002; 52: 896-901. 20 Branchereau A, Rosset E . Carotid bifurcation. In : Branchereau A, Berguer R (eds). Vascular surgical approaches. Armonk, Futura Publishing Co, 1999 : pp 1-8. 21 Thomassin JM, Branchereau A. Intrapetrosal internal carotid artery. In : Branchereau A, Berguer R (eds). Vascular surgical approaches. Armonk, Futura Publishing Co, 1999 : pp 15-20. 22 Branchereau A, Rosset E . Distal vertebral artery (C2-C1) : anatomic features and surgical approach. In : Branchereau A, Berguer R (eds). Vascular surgical approaches. Armonk, Futura Publishing Co, 1999 : pp 27-35. 23 Berguer R . Suboccipital approach to the vertebral artery. In : Branchereau A, Berguer R (eds). Vascular surgical approaches. Armonk, Futura Publishing Co, 1999 : pp 37-41. 24 MayJ, White GH, Waugh R, Brennan J. Endoluminal repair of internal carotid artery aneurysm : a feasible but hazardous procedure. / Vase Surg 1997 ; 26 : 1055-1060.

37

This page intentionally left blank

4
PENETRATING INJURY TO THE BLOOD VESSELS OF THE NECK AND MEDIASTINUM
JOHN ROBBS

Diagnosis and particularly management of penetrating injuries to the blood vessels of the neck and mediastinum may present enormous challenges to the surgical team. The anatomical areas of the neck in the context of penetrating trauma has been arbitrarily divided into three zones. Zone 1 extends from the clavicles to the cricoid cartilage, Zone 2 extends from the cricoid cartilage to the angle of the mandible, and Zone 3 extends from the angle of the mandible to the mastoid process [1] (Fig. 1). In general terms, diagnosis and management are relatively easy in Zone 2, and the difficult areas are represented by injuries in which the trajectory goes toward the base of the skull (Zone 3) or down into the superior mediastinum (Zone 1). The morbidity and mortality associated with penetrating injuries in this region are mainly related to injuries to the blood vessels, particularly the arterial system. In this chapter, discussion will be limited to the diagnosis and management of these injuries. In terms of management, the major debate revolves around whether mandatory exploration should be done for all penetrating wounds in this region or whether a more selective policy of investigation and intervention based on clinical signs should be advocated. One could argue that the most advantageous is to fully investigate all patients with penetrating wounds; however, in extremely busy trauma units, this can be unrealistic and, based on general clinical experience, it would appear that a selective policy is safe [2]. The aim of this chapter is to provide a practical guide to diagnosis and management based on our own experience with more than 4000 major vascular injuries, of which 20% to 25% involve the cervicomediastinal region.

Tr 39

VASCULAR

EMERGENCIES

FIG. 1

Diagram showing the Zones of the neck.

Etiology and pathology

In order to rationally diagnose and treat these injuries, it is essential to have a good understanding of the mechanisms of injury and pathology. Penetrating wounds may be divided into stabs, either with a knife or some other sharp implement, and missiles, which can be divided into low velocity or high velocity. High-velocity missiles are those in which the muzzle velocity exceeds 300 meters per second. This is usually associated with military weapons or hunting rifles. Low-velocity wounds such as stabs or missiles with a muzzle velocity of less than 300 meters per second, such as a standard handgun, usually produce damage that is limited to the implement or missile track. The energy transfers involved with high-velocity missiles cause a cavitation effect with tissue destruction around the actual missile track, which results in extensive associated soft tissue trauma [3]. Shotgun wounds warrant separate description. Bird shot is small caliber and densely packed within the cartridge, whereas buck shot is larger and heavier and there are fewer packed into the cartridge. The unique property of a shotgun wound is the large number of foreign bodies embedded in the tissues, particularly in the case of bird shot, with extensive associated soft tissue trauma particularly

at close range. The blood vessels involved tend to develop multiple small perforations. An additional problem occasionally encountered when the missiles enter the vessels is distal embolization into the arterial system (Fig. 2). In general, direct injury to the artery may cause either partial or complete transection. This in turn may result in pseudoaneurysm formation or, in the case of complete transection, thrombosis of both ends of the transected vessels [4,5] (Figs. 3A and 3B). Pseudoaneurysms with expansion may lead to compression of the aerodigestive tract or the brachial plexus. Small perforations, on the other hand, may temporarily seal off and then, with dissolution of the thrombus for whatever reason whether lowgrade infection or natural fibrinolysis, develop a delayed false aneurysm. This may occur gradually with progressive enlargement or may be an acute phenomenon. These patients may then present with a large pulsating hematoma, often infected, with or without compression symptoms. Brachial plexus compression under these circumstances results in major morbidity with a very guarded prognosis for

40

FIG. 2 Angiogram of a young woman who sustained a shotgun wound to the neck/upper mediastinum (Zone 1). Buck shot probably entered via the venous system and embolized to the common femoral artery (arrow).

PENETRATING

INJURY TO THE BLOOD VESSELS OF THE NECK AND

MEDIASTINUM

recovery of function [6,7]. The concept of conservative management of minimal vascular injury is not to be encouraged [8]. We believe that all perforations, even with small pseudoaneurysms that are stable, should be dealt with, as the consequences of delayed hemorrhage are highly significant in terms of morbidity. With complete transection and associated thrombosis, the thrombus tends to propagate to the first collateral. Carotid transection will propagate up to the carotid bifurcation and down to the arch of the aorta. There is also often thrombus within the lumen associated with partial lacerations (Fig. 3A). Adjacent perforations of the artery and vein will result in an arteriovenous fistula. Significant hemorrhage and the formation of a pseudoaneurysm is not an invariable feature of arteriovenous fistulae, as the arterial flow tends to take the line of least resistance into the venous system. The systemic

effects depend largely on the size of the arteriovenous shunt. In the acute setting when the mediastinal and neck vessels are involved, usually in fit young patients, the systemic hemodynamic effects are minimal. Undiagnosed fistulae may only present months or years later with the classic widened pulse pressure, systolic hypertension, and, rarely, congestive cardiac failure. We have on record patients presenting up to 14 years after the initial injury. In the acute setting, an important feature of arteriovenous fistulae is the development of thrombus at the fistula site and particularly in the downstream arterial segment, probably due to the diminished flow in that segment. When this occurs in the carotid artery, neurologic deficits occur quite frequently due to embolization of this thrombus. This is a major consideration in relation to interventional management (Fig. 3C).

41

FIG. 3 A - Diagram showing propagation of thrombus following carotid transection. B - Lateral perforation with pseudoaneurysm formation. Note intraluminal thrombus. C - Arteriovenous fistula. Note intraluminal thrombus in distal arterial component.

VASCULAR

EMERGENCIES

An injury that is becoming increasingly apparent in our practice is an intimal tear associated with a missile Shockwave. We have frequently noted at exploration an apparently intact carotid vessel with minimal bruising on the adventitial surface. On opening the vessel there is an intimal tear with superimposed thrombosis. It is postulated that the Shockwave causes acute lateral displacement of the vessel, causing distraction tears in the least elastic component of its wall, the intima, while the more elastic media and adventitia remain intact. The exposed media on the luminal surface is thrombogenic with resultant thrombosis that may be associated with thromboembolic phenomena (Fig. 4).

Clinical presentation
The entry and exit wound may give some idea of the trajectory of the missile. However, if there is a single wound it is difficult to trace the trajectory. Probing is unreliable and is certainly not recom-

42

mended, as thrombus may be dislodged with initiation of exsanguinating hemorrhage. We have also seen several patients in whom the stab wound has traversed the neck and caused a vascular injury on the opposite side. Presenting features may be grouped into emergency, subacute, and long-term. Emergency presentation is that of ongoing active hemorrhage whether to the exterior through the wound or concealed within the thoracic cavity, or the superior mediastinum. Subacute presentation occurs in patients who have stabilized hemodynamically. Pathognomonic signs of vascular injury are a pulsating hematoma, a pulse deficit in the upper limb, or a bruit whether systolic or of the arteriovenous type. It appears that on occasion the fistula may only develop later by the same mechanism already postulated for the development of delayed false aneurysm [9]. The communication may be sealed by thrombus and no bruit is audible in the acute phase. Within the next few days lysis occurs with the development of a manifest fistula. A systolic bruit may also result from compression resulting from pseudoaneurysm or partially occluding thrombus. More subtle signs of major vascular injury may be evidence of previous hemorrhage such as a history of major hemorrhage from the wound or a record of initial shock at presentation that has stabilized following fluid resuscitation, or when the patient presents with a low hemoglobin level. Long-term presentation may be with delayed onset of false aneurysm with or without compressive symptoms to the trachea or brachial plexus or an arteriovenous fistula that may present months or even years later with the consequences of systemic hemodynamic changes, a machinery bruit, or venous hypertension.

Diagnosis and management

Patients presenting in the emergency category with ongoing hemorrhage or acute stridor require urgent exploration in order to stop the bleeding and/or relieve the compression. This is a clinical decision requiring judgment as to where the site of the bleeding might be, and the appropriate exploration must be carried out. The hemodynamically stable patient without airway compression can be investigated appropriately. An erect, well-orientated chest film is extremely valuable. Widening of the mediastinum suggests

FIG. 4 Diasram illustrating postulated mechanism of arterial thrombosis following proximity (Shockwave) missile injury.

PENETRATING

INJURY TO THE BLOOD VESSELS OF THE NECK AND

MEDIASTINUM

that further investigation might be necessary. Computed tomography scans in the context of penetrating trauma are not as useful and give little more information than a high-quality antero-posterior chest radiograph. Duplex scanning is extremely useful for Zone 2 injuries [10,11]. It is accurate in this area for identifying intimal flaps, localizing arteriovenous fistulae, and detecting minor degrees of perforation (Fig. 5). This modality is of minimal value in Zone 1 or Zone 3 injuries because of anatomical consideration. Transesophageal echo has not proved particularly useful in our experience and we do not recommend it. Arch angiography using the Seldinger technique remains the gold standard for the diagnosis of cervicomediastinal vascular injuries. It has the additional advantage that interventional maneuvers can be performed when required. The indication for angiography is clinical suspicion of vascular injury in the hemodynamically stable patient, as previously

outlined. It is also recommended in our practice that routine angiography should be carried out in patients with shotgun wounds. While stab wounds can safely be observed if there is no suspicion of vascular injury on clinical grounds, our threshold for angiography is far lower with gunshot wounds, particularly those which traverse the neck, because of the possibility of a Shockwave injury. The pitfalls of interpreting angiography lie in the accurate localization of arteriovenous fistula by virtue of the rapid circulation time [7,9]. In addition, significant intraluminal thrombus may be present that is not clearly demonstrated by the radiography (Fig. 6).

Operative management
It is important that the patient be positioned and draped so that access is possible from the base of the skull to the xiphisternum. The patient should be in the supine position with a bolster between

43

FIG. 5 Duplex scan showing a common carotid to internal jugular arteriovenous fistula following a stab wound in Zone 2 of the neck.

FIG. 6 Angiogram showing a carotid to jugular fistula as a result of a gunshot wound. The young male developed an acute right hemispheric neurologic deficit while awaiting operation. At surgery there was extensive thrombus formation in relation to the fistula (not clearly seen on the angiogram).

VASCULAR

EMERGENCIES

the scapulae, the neck extended, and the head turned away from the side of the injury. It is foolish to attempt to enlarge and explore the stab wound, and the standard approach must be adapted to the blood vessels of the neck and mediastinum. The standard utility incision is placed along the anterior border of the sternomastoid muscle, which can be extended into a median sternotomy if required (Fig. 7). For Zone 2 injuries, the standard approach to the common carotid artery, the bifurcation, and the proximal internal carotid artery should be used. The aerodigestive tract can also then be explored by displacing the carotid sheath anteriorly. For lacerations of the blood vessels close to the base of the skull, access may be improved by dividing the digastric muscle. If this does not improve matters it is certainly not recommended that the temporomandibular joint be dislocated with forward displacement of the mandible, or that the ramus of the mandible be divided, as this results in unacceptable long-term morbidity.

44

FIG. 7 Incisions for exposure of cervicomediastinal blood vessels.

Excellent access to the vessels at the base of the skull can be gained by detaching the sternomastoid muscle from its insertion on the mastoid. This can be then retracted antero-inferiorly with excellent exposure of the vessels as they enter the skull. Care must be taken not to damage the tenth and twelfth cranial nerves as well as the accessory nerve, all of which can be clearly identified [12]. For Zone 1 injuries (superior mediastinum), the optimal incision is a total median sternotomy. This provides excellent access to all the mediastinal vessels including the left subclavian artery. Under certain circumstances, a limited sternotomy down to the angle of the sternum may provide sufficient exposure. We have never found it necessary to excise the clavicle or to make trap door incisions. Division or excision of the clavicle causes major longterm morbidity particularly in manual laborers. The distal subclavian artery is best approached using a supraclavicular incision. If necessary, a separate infraclavicular incision may be made in order to tunnel a graft. Division of the clavicle invariably results in non-union and a painful pseudarthrosis, which is possibly a consequence of devascularization of the bone in the process of surgical division. The arterial injuries themselves can be repaired on their merits. The usual principles of debridement of the edges of the arterial wall back to healthy tissue pertain. The vessel can then be repaired by lateral suture, patch angioplasty, or an interposition graft [4,5]. Any prospect of narrowing the vessel by lateral suture should prompt the use of a vein patch. The through-and-through wound is best totally transected, debrided, and reconstructed in an endto-end fashion. Following excision of the damaged vessel, an endto-end anastomosis should not be attempted if there is excess tension on the suture line, as this invites occlusion. The best measure of tension is whether a single tethering stitch holds the ends together; if not, it is better to reconstruct using an interposition graft, the saphenous vein being the optimum graft. While the jugular veins have been used, they are not as satisfactory because of their thin walls and wide diameter, particularly when reconstructing the internal carotid artery. In the case of injury with an intimal tear it is best to excise the damaged segment and repair by means of an interposition graft. It is extremely important, particularly with regard to the carotid vessels, to dissect meticulously in

PENETRATING

INJURY TO THE BLOOD VESSELS OF THE NECK AND

MEDIASTINUM

order to avoid dislodging thrombus. It is also important to pass an embolectomy Fogarty catheter gently, both proximally and distally, to ensure that all thrombus has been extracted prior to performing repair.

Specific problems
Penetrating wounds of the aortic arch that are compatible with survival are usually small puncture wounds that can be treated by digital occlusion and the insertion of mattress sutures deep to the occluding finger [12]. Wounds of the major branches close to the arch, particularly through-and-through wounds, are best treated by using a partially occluding clamp on the aorta, totally transecting the relevant vessel and oversewing its origin on the aorta. Continuity can then be restored by making an end-to-side anastomosis of a prosthetic graft to the intrapericardial portion of the ascending aorta and an end-to-end anastomosis to the relevant vessel. Attempts to reanastomose this type of injury invite disaster, as

it is extremely difficult to gain adequate control. It is also difficult to gain intimal apposition as the media tends to retract into the aorta. In the case of the left subclavian artery, continuity can be restored by means of a carotid to subclavian bypass if necessary (Figs. 8A and 8B). On occasion, the superior mediastinum is totally obscured by clot, which makes identification of the relevant vessels difficult. Under these circumstances, it is best to open the pericardium and expose the intrapericardiac aorta. The origins of the major vessel can then be identified by palpation and the necessary proximal control obtained prior to opening the hematoma [13]. Arch anomalies are uncommon, but they may not be identified by pre-operative angiography, and occasionally with emergency explorations they may catch the surgeon off guard (Fig. 9). In our own practice, the prevalence of some form of anomaly is 5.3%; although these anomalies are not common, this is significant [14]. The anomalies range from a common trunk from which the left common carotid and the brachiocephalic arteries arise, the so-called bouquet anomaly, to the lusorian anomaly

45

FIG. 8 Recommended sursical manasement of injuries occurring in close proximity to the aortic arch.

VASCULAR

EMERGENCIES

performed. Similarly, it is probably ill advised to restore continuity in a patient who is deeply comatose [16,17]. Most patients with localizing neurologic deficits improve, and we have not to our knowledge caused neurologic deterioration by following this practice.

ARTERIOVENOUS FISTULAE
A major problem may arise if there is failure to accurately localize the fistula site particularly in the superior mediastinum. It can be very difficult to distinguish between a fistula arising from the aortic arch from one arising from major vessels close to their points of origin. In Zone 2, duplex scanning has proved sufficiently accurate in defining the exact site of fistulation. Zone 3 presents very similar problems to those in the superior mediastinum. Duplex is not helpful and angiography can indeed be misleading as to the exact site of the fistula. The veins lie in an anterior plane to the artery, and it is relatively easy to isolate the fistula by dissecting out the vein and isolating the fistula by palpation along the course of the vein. Once found, the proximal and distal arterial components can be isolated and the artery and vein repaired [5,9]. It has been stated that it is advisable to wait for fistula to mature prior to attempting repair; however, we have not found this to be advisable, as the longer one waits the more fibrosis occurs and the more difficult the procedure becomes. In the presence of extensive fibrosis and a small deficit, it may on occasion be expedient to repair the fistula transvenously but this is rarely necessary. In general, repair constitutes simple debridement and end-to-end or lateral suture. Embolism from the clot within the arterial segment just distal to the fistula is a distinct hazard. It is important to be meticulous and gentle in the dissection of the vessels and to ensure that there is no residual thrombus in the artery prior to repair and restoration of prograde flow. Back bleeding and gentle thrombectomy using a small balloon catheter should always be performed. Air embolism through the venous component has not proved to be a problem, but it is important to clamp the veins at an early stage and to fill the venous segment with saline prior to final closure. Recurrent fistulation has never proved to be a problem, but multiple fistulae do occur and it is important, particularly in the case of shotgun wounds, to ensure that all palpable thrills have disappeared on completion of the repair. It is impor-

FIG. 9 Arch angiogram of a young male showing a typical anomaly. There is a common origin of the brachiocephalic and left common carotid arteries (arrow).
46

in which there is a short brachiocephalic trunk from which both common carotid arteries arise and the right subclavian arises from the descending aorta and traverses the posterior mediastinum behind the esophagus [13]. Under these circumstances it would be necessary to insert a shunt from the interpericardiac aorta in order to preserve cerebral flow while repair is carried out [15].

ASSOCIATED NEUROLOGIC DEFICITS

The theoretical consideration in restoring flow under these circumstances is the creation of a hemorrhagic infarct. Computed tomography scanning is notoriously unreliable within the first 24 hours of developing a neurologic deficit, and it is often not possible to obtain the necessary scan in the emergency setting. It has been our policy to restore arterial continuity in all patients with a neurologic deficit, provided that there is distal continuity, as evidenced by back bleeding following extraction of thrombus. If there is no back bleeding, ligation is

PENETRATING INJURY TO THE BLOOD VESSELS OF THE NECK AND MEDIASTINUM

tant also to attempt to trace the trajectory of the penetrating implement or missile to ensure that all injuries have been dealt with. ization through the anterior jugular system, the anterior communicating veins in the neck, and the superior intercostal veins that restored cardiac inflow via the azygos and hemi-azygos systems. Even when the azygos system was occluded, the superficial veins on the chest wall contributed to this collateralization. It must also be postulated that the vertebral plexus comes into play under these circumstances. It is well known with general surgical experience that interruption of one or even both internal jugular veins is well tolerated. Based on this evidence, it has been our policy for several years to repair simple lacerations or perform reanastomosis when feasible. Complex injuries with tissue loss cephalad to the azygos veins have been ligated [24,25]. If the vein is to be reconstructed, meticulous and gentle technique using a fine guage vascular suture (6/0) is absolutely essential. In a significant series of patients, we have not encountered major permanent morbidity as a result of this approach. Temporary facial and upper limb edema has been noted in several patients who have had ligation of the brachiocephalic veins but which resolved within a period of 4 to 5 days. Ligation of the distal subclavian vein has resulted in temporary upper limb edema but has not resulted in longterm morbidity. We have not seen large injuries of the superior vena cava caudad to the azygos vein, and this type of injury is probably incompatible with survival. Air embolism has not constituted a problem, although meticulous attention is paid to ensuring that significant amounts of air are not present in the major veins prior to restoring flow, and it is advisable to fill the segment with saline prior to inserting the final stitches.

The use of cardiopulmonary bypass

In the acute setting in most of the lesions described, cardiopulmonary bypass is unnecessary and only adds to the magnitude of the operation [18,19]. In most cases, if deemed necessary, intraluminal shunts can be during the procedure, but this is certainly not routine. Cardiopulmonary bypass may become necessary for the repair of chronic lesions presenting late in which extensive fibrosis may create major technical difficulties [20-22]. The major morbidity associated with penetrating wounds involving the extracranial cerebrovascular circulation is neurologic. In general terms, neurologic morbidity following straightforward carotid repair is negligible; however, there is a significant, approximately 5%, neurologic event rate associated with carotid arteriovenous fistulae. This, we believe, is due to thrombotic embolization. The overall mortality for mediastinal lesions is in the area of 7% and, for Zone 2 and 3 lesions, approximately 2%.

47

VENOUS INJURIES
Frequently, particularly with gunshot trauma, there is far more extensive injury to the major veins, leaving relatively large defects that do not lend themselves to simple repair. This may be the result of increased friability of the veins and their relative inelasticity. It has been stated that repair should always be attempted in this situation. However, venous repair does not enjoy the same success that one associates with arterial reconstruction. This is the result of the friability of the vessel wall and the low intraluminal pressure. There is a high occlusion rate associated with complex repair using interposition grafts, particularly with a prosthesis. In addition, as is so often the case when attempting to repair associated complex venous injuries, it significantly increases the magnitude of the operative procedure. There is a paucity of information on the collateralization of the venous system of the head and neck. Barrett performed phlebographic studies on patients with idiopathic superior mediastinal fibrosis [23]. These studies showed extensive collateral-

INTERVENTIONAL CATHETER TECHNIQUES

Embolotherapy is proving invaluable in dealing with traumatic lesions involving branches of the subclavian artery, the vertebral vessels, and branches of the external carotid [26]. In relation to the carotid system, we have had a single episode of migration of a spring coil into the internal carotid, and while the patient fortunately did not suffer any adverse sequelae, this is obviously undesirable. In relation to the vertebral vessels, particularly within the bony canal, there exists the possibility of creating a neurologic deficit with vertebral artery disruption if the posterior communicating vessels in the circle of Willis are found to be deficient [2730]. We have fortunately not encountered this problem.

VASCULAR EMERGENCIES

The use of covered stents to treat partial lacerations or arteriovenous fistulae is attractive. The currently available stents are polytetrafluoroethylene (PTFE) covered, and an increasing number of reports on the use of these stents, particularly in the subclavian artery, describe mixed success [3134]. The use of stents is probably not applicable within the aortic arch, and we strongly recommend

against their use in the carotid system in the context of trauma because of the thrombus frequently associated with these lesions. There is the ever-present possibility of infection, as all traumatic wounds are potentially contaminated. We have one patient on record with a proximal subclavian pseudoaneurysm that was stented; infection occurred and ultimately led to the patient's demise.

R E F E R E N C E S 1 RobbsJV, KeenanJ. Exploration of the neck. In: Champion HR, Robbs JV, Trunkey D (eds). Rob and Smith's operative surgery, 4th edition. London, Butterworths 1989: pp 166-172. 2 Campbell FC, Robbs JV. Penetrating injuries of the neck: a prospective study of 108 patients. flr/Swrgl980; 67: 582-586. 3 Levien LJ. Ballistics of bullet injury. In: Champion HR, RobbsJV, Trunkey D (eds). Rob and Smith's operative surgery, 4th edition. London, Butterworths 1989: pp 106-110. 4 Robbs JV. Vascular trauma. General principles of surgical management. In: Champion HR, RobbsJV, Trunkey D (eds). Rob and Smith's operative surgery, 4th edition. London, Butterworths 1989: pp 519-528. 5 Robbs JV. Basic principles in the surgical management of vascular trauma. In: Greenhalgh RM (ed). Vascular and endovascular techniques, 4th edition. London, W.B. Saunders Ltd. 2001 :pp 455-465. 6 Robbs JV, Naidoo KS. Nerve compression injuries due to traumatic false aneurysm. Ann Surg 1984; 200: 80-82. 7 Robbs JV, Baker LW. Cardiovascular trauma. Curr Probl Surg 1984; 21:1-87. 8 Frykberg ER, Crump JM, Dennis JW et al. Nonoperative observation of clinically occult arterial injuries: a prospective evaluation. Surgery 1991; 109: 85-96. 9 Robbs JV, Carrim AA, Kadwa AM, Mars M. Traumatic arteriovenous fistula: experience with 202 patients. BrJ Surg 1994; 81: 1296-1299. 10 Fry WR, Dort JA, Smith RS et al. Duplex scanning replaces arteriography and operative exploration in the diagnosis of potential cervical vascular injury. Am J Surg 1994; 168: 693-695. 11 Corr P, Abdool Carrim AT, Robbs J. Colour-flow ultrasound in the detection of penetrating vascular injuries of the neck. S Afr MedJ 1999; 80: 644 -646. 12 Robbs JV. Injuries to the vessels of the neck and superior mediastinum. In: Champion HR, RobbsJV, Trunkey D (eds). Rob and Smith's operative surgery, 4th edition. London, Butterworths 1989:pp 529-538. 13 RobbsJV. Penetrating trauma to the inlet and outlet regions of the chest. In: Westaby S, Odell J (eds). Cardiothoradc trauma. London, Arnold 1999: pp 183-195. 14 Satyapal KS, SingaramJV et al. Aortic arch branch variation case report and angiographic analysis. S Afr J Surg W02, (in press). 15 WoolgarJD, RobbsJV, Rajaruthnam P, Mohamed GS. Penetrating injuries in the innominate artery in association with abnormal aortic arch anatomy. EurJ Vase Endovasc Surg 2QQ2; 23: 462-464. 16 Perry MO, Snyder WH, Thai ER. Carotid artery injuries caused by blunt trauma. Ann Surg 1980; 192: 74-77. 17 RobbsJV, Human RR, Rajaruthnam P et al. Neurologic deficit and injuries involving the neck arteries. Br J Surg 1983; 70: 220-222. 18 Buchan K, Robbs JV. Surgical management of penetrating mediastinal arterial trauma. Eur J Cardiothorac Surg 1995; 9:90-94. 19 Graham JM, Feliciano DV, Mattox KL, Beall AC Jr. Innominate vascular injuries. / Trauma 1982; 22: 647-655. 20 Fulton JO, de Groot KM, Buckels NJ, von Oppel UO. Penetrating injuries involving the intrathoracic great vessels. S Afr J Surg 1997; 35: 82 -86. 21 Fulton JO, de Groot MK, von Oppel UO. Stab wounds of the innominate artery. Ann Thorac Surg 1996; 61: 851-853. 22 Fulton JO, Brink JG. Complex thoracic vascular injury repair using deep hypothermia and circulatory arrest. Ann Thorac Surg 1997; 63: 557-559. 23 Barret NR. Idiopathic mediastinal fibrosis. Br } Surg 1958; 46: 207. 24 Robbs JV, Reddy E. Management options for penetrating injuries to the great veins of the neck and superium mediastinum. Surg Gynaec Obstet 1987; 165: 323-326. 25 Nair R, Robbs JV, Muckart DJ. Management of penetrating cervicomediastinal venous trauma. EurJ Vase EndovascSurg 2000; 19:65-69. 26 Naidoo NM, Corr PD, Robbs JV et al. Angiographic embolisation in arterial trauma. EurJ Vase EndovascSurg 2000; 19: 77-81. 27 Thomas GI, Anderson KN, Hain RF et al. The significance of anomalous vertebro-basilar artery communications in operations on the heart and great vessels. Surgery 1956; 46: 747-757. 28 Monson DO, Saletta JD, Freeark RJ. Carotid vertebral trauma. /Trauma 1969; 9: 987 -999. 29 Schomer DF, Marks MP, Steinberg GK et al. The anatomy of the posterior communicating artery as a risk factor for ischemic cerebral infarction. NEngJMed 1994; 330: 1565-1570. 30 Jithoo R, Nadvi SS, Robbs JV. Vertebral artery embolism post subclavian artery injury with occipital lobe infarction. EurJ Vase 31 Marin ML, Veith FJ, Panetta TF et al. Transluminally placed endovascular stented graft repair for arterial trauma. J Vase Surg 1994; 20: 466-473. 32 du Toit DF, Strauss DC, Blaszczyk M et al. Endovascular treatment of penetrating thoracic outlet arterial injuries. EurJ Vase Endavasc SurgZm-, 19: 489-495. 33 Strauss DC, du Toit DF, Warren BL. Endovascular repair of occluded subclavian arteries following penetrating trauma. JEndovasc Tfcr2001; 8: 529-533. 34 Chandler TA, Fishwick G, Bell PR. Endovascular repair of a traumatic innominate artery aneurysm. EurJ Vase Endovasc Surg 1999; 18: 80 -82.

48

5
ACUTE ABDOMINAL AORTIC OCCLUSION
PIERRE JULIA, STEPHANE ZALINSKI, JEAN-NOEL FABIANI
Acute aortic occlusion is an uncommon disease with a mortality of approximately 50%, requiring urgent treatment in a specialized center. The classic differentiation between embolus and acute thrombosis of the distal aorta remains a realistic distinction. The latter has become the most frequent etiology because of the predominance of peripheral vascular pathology in an ageing population. Other causes are more rare, such as acute thrombosis of an abdominal aortic aneurysm, an occlusion due to aortic dissection or acute hemostatic disorders. If untreated, the prognosis is poor with a mortality of 75%. The initial diagnostic failures can be explained by the associated neurologic symptoms, delaying adequate treatment and therefore aggravating the prognosis. The extent of the occlusion and associated ischemia-reperfusion syndrome can induce major metabolic disorders. Despite surgical treatment, the overall mortality of this disease is approximately 50 % in the major published series [1].

49

Clinical presentation
The clinical features are mainly determined by the acute severe ischemia of the lower limbs, associated with bilateral pain and neurologic symptoms like numbness, paresthesias and, in the most severe state, even paralysis of both legs. Rest pain or severely deteriorated intermittent claudication have the same diagnostic value and suggest a superimposed thrombosis on pre-existing occlusive lesions. Additional abdominal complaints or renal insufficiency with oligo-anuria are reminiscent of intestinal or renal ischemia, indicating proximal extension of the thrombosis to the level of the visceral

arteries. The severity of symptoms depends on the quality of the collateral circulation, being poor in cases of embolization and more developed in aortoiliac occlusive disease. An occluding embolus causes an abrupt and severe bilateral pain often associated with low back and gluteal pain, and can subsequently induce major neurologic deficit with complete paraplegia. Skin mottling can often be impressive and involve both legs, proximally extending to the umbilicus. In contrast, an acute thrombosis in a diseased distal aorta with pre-existing severe claudication or rest pain will manifest either with aggravation of existing complaints or a neurologic deficit similar to

VASCULAR

EMERGENCIES

spinal cord infarction. In these nonexceptional cases, the patients are referred to a neurologic or neurosurgical department, causing significant diagnostic delays. These neurologic manifestations are not induced by spinal cord infarction but by severe ischemic neuropathy, which is reversible after revascularization [2]. Physical examination reveals absence of pulsations in both groins. The degree and extent of neurologic deficit is variable and develops in time, therefore requiring initial adequate and detailed neurologic examination, which serves as a reference for subsequent assessments. Also the location and extent of the ischemic skin should be indicated and followed. Abdominal examination is usually normal; however, meteorism, pain and decrease or disappearance of peristalsis might indicate associated intestinal ischemia [3]. Urine output should also be monitored since any alteration suggests impairment of the renal arteries. Acute onset hypertension has the same diagnostic value [4].

floating thrombus developed. This mechanism, however, is unlikely to cause a complete aortic occlusion because of the size of these thrombus masses.

ACUTE THROMBOSES
Acute thromboses represent the most common cause in several series [4,7], although Surowiec et al. [5] only encountered an incidence of 50%. They are often the end stage of obstructive aorto-iliac disease characterized by bilateral intermittent claudication. The severity of the initial symptoms greatly depends on the extent of the collateral system and suddenness of onset. If the collateral network is poorly developed and the pre-existing aortic stenosis is limited, the clinical picture is dramatic with acute neurologic deficits. In contrast, if the acute aortic occlusion is superimposed on a chronic extensive stenotic process, the clinical picture is often less impressive and predominantly manifests with ischemic rest pain. The occurrence of a thrombosis in diseased arteries is most often provoked by dehydration or cardiac decompensation, the latter being the result of myocardial infarction or severe arrhythmia [4]. Acute thrombosis of an abdominal aortic aneurysm only exceptionally occurs. The first case was described by Shumacker in 1959 [8] and the first revascularization was performed by Jannetta and Roberts [9]. Only 44 cases have recently been published, according to a review of the literature by Hirose et al. [10]. The clinical signs are similar to those of an acute aortic occlusion and only the palpation of an abdominal mass with transmitted pulsations might indicate the diagnosis. In approximately half of the reported series, the correct diagnosis was established during laparotomy. At present, contrast enhanced computed tomography (CT) allows rapid assessment. The following mechanisms of such an occlusion are: - coexisting bilateral severe iliac obstructive disease with iliac thrombosis and proximal extension in to the aortic aneurysm, - occlusion of the aneurysm neck by means of an embolus originating from the heart, - partial dislodgment of intra-aneurysmal mural thrombus causing obstruction of the lumen with secondary thrombosis, - hypotensive episode or low cardiac output with intravascular thrombosis in a pre-existing aneurysm largely filled by thrombus. The optimal treatment consists of emergency surgical repair, allowing simultaneous treatment of the

Etiology

_5
50

The most common causes of acute distal aortic occlusion are embolization and thrombosis. Other mechanisms include traumatic occlusion, aortic dissection, metabolic disorders, and pharmacologic interactions. Acute occlusions of vascular grafts leading to the signs of aortic occlusion are very uncommon and will not be addressed in this chapter.

EMBOLI
The most common source of emboli is the heart, in the past mainly from endocarditis. At present, ischemic heart disease and especially the sequelae of myocardial infarction dominate the embolic etiology [5]. Arrhythmia is often encountered, basically being atrium fibrillation, however this is rarely an isolated entity. Cardiac tumors, mainly the myxoma of the left auricle, can be the source of embolization by fragmentation of the tumor. Kao [6] reported an exceptional case of secondary acute aortic occlusion by a cardiac myxoma. Paradoxal embolization can occur in the coexistence of deep venous thrombosis, patent foramen ovale and pulmonary hypertension susceptible to cause a right-to-left shunt. Theoretically, the emboli can come from the thoracic aorta, either from an aneurysm or from an ulcerated plaque on which a

ACUTE ABDOMINAL AORTIC OCCLUSION

aneurysm and acute thrombosis. This surgical management, however, is not always feasible because of the poor physical conditions of these patients. In these cases, distal revascularization can be established by means of an axillobifemoral graft. This technique does not prevent later rupture, as observed in 15% of the patients of Schwartz et al. [11]. The prognosis of these aortic aneurysm thromboses is poor, with a mortality rate greater than 50%. Recent progress is limited and mortality rates encountered during the last years sit at 42% [10]. Acute aortic thrombosis can also be caused by aortic dissection. In the experience of Cambria et al. [12], 10 of 325 patients with acute aortic dissection required urgent aortic replacement or fenestration because of acute aortic occlusion. The mechanism of this aortic obstruction is based on the propagation of the intimal layer until the aortic bifurcation with compression of the true lumen by the false lumen (dynamic obstruction). This mechanism rarely occurs because the dissection usually extends in a unilateral fashion, or bilateral and asymmetrical, in one or two common iliac arteries. Arterial trauma is a rare cause of aortic occlusion, in which indirect localized dissection is the mechanism. Seatbelt compression is a known cause of multiple intra-abdominal lesions, however, abdominal aortic lesions rarely occur. In general, traumatic aortic injury affects the thoracic aorta in 95% and the abdominal aorta in 5% of cases [13]. The ischemic signs are immediately severe and associated intraabdominal lesions can deteriorate the prognosis, latrogenic traumatic injuries are particularly caused by intra-aortic balloon pumps [14] or complex endovascular procedures, as applied in cases of complicated aortic dissections [15]. Aortic thromboses can also be induced by hypercoagulability syndromes. These syndromes mainly cause venous thromboses, but if arterial complications occur the prognosis is poor [4]. The clinical manifestations of aortic thrombosis in patients with nephrotic syndrome are dramatic, as reported by Imamura et al. [16]. Szychta et al. described ulcerative colitis as a result of acute aortic thrombosis [17]. Finally, lupus anticoagulants are associated with aorto-iliac occlusions and smaller caliber arteries [18,19]. Recently, DiCenta et al. [20] described a case of infrarenal aortic occlusion associated with inferior vena cava occlusion in a patient with circulating anticoagulants. Several pathologic states can be identified, including heparin-induced thrombopenia, antithrombin III, protein C, and protein S deficits. It is therefore necessary, in case of acute aortic occlusion, to fully investigate the coagulation parameters because coagulation disorders require immediate treatment in order to prevent recurrences.

Additional investigations
Duplex scanning of the lower limbs does not provide additional information to that found during physical examination. However, the technique might identify involvement of one or more visceral arteries, realizing that obesity or abdominal meteorism can limit its sensitivity. Furthermore, duplex scanning is technician dependant. Arteriography comprising aortography and arteriography of the lower limbs is debatable. In the older publications arteriography was recommended as a standard technique. There are indeed several arguments to justify this technique. It provides a perfect diagnosis and allows accurate delineation of the proximal extension of the occlusion and the status of the renal arteries, as well as the distal outflow [3]. Noteworthy to mention is the advantage of an injection in the ascending aorta which often allows better and earlier visualization of the lower limb arteries via the internal mammary and epigastric arteries. In case of embolization, arteriography might depict emboli in other regions like the superior mesenteric and renal arteries in particular but also emboli resulting from thrombus fragmentation at an arterial bifurcation, leading to a secondary shower of emboli affecting the femoral, popliteal and crural arteries. This is a relative advantage because in these ill patients the femoral bifurcation can be visualized on the initial angiography and the distal outflow can be assessed by intra-operative angiography. The disadvantages of arteriography include deterioration of renal function, delay of revascularization and the fact that it hardly modifies the surgical strategy, except in visceral involvement and aortic dissection [1,21]. CT scanning is hardly cited in the literature, probably because the majority of publications are outdated. At present, the spiral CT scanners will systematically be used, especially in the acute setting of the disease. In fact, this technique allows visualization of the total thoraco-abdominal aorta and its main visceral branches. It might also depict the cardiac cavities in the search for thrombus formation. Magnetic resonance imaging has not been applied in acute aortic occlusion

_
51

VASCULAR

EMERGENCIES

because of its limited availability and less maneuverability in acute settings. If a cardiac embolic source is suspected, echocardiography should be performed to identify valvular disease or dyskinetic ventricular areas due to ischemic cardiopathy. Residual intracardiac thrombus mass is rarely diagnosed but is an argument for adequate and prolonged anticoagulation treatment in order to limit the risk of embolic recurrences [22].

Treatment
The medical treatment is similar to the management of severe acute ischemia of the lower limbs. Intravenous heparin starts with a dose of 2 000 to 3 000 IU, followed by continuous infusion of 300 to 400 lU/kg per 24 hours, frequently controlled with blood samples. The role of thrombolysis is poorly defined, but it is not applicable in acute cases in which immediate surgical revascularization is required. However, it might be considered as a first attempt in patients presenting with a subacute clinical picture. Thrombolysis can re-open at least one iliac axis, followed by an additional endovascular procedure such as thrombectomy and angioplasty [23]. At present, however, the experience with this method remains anecdotal [24]. This technique will be most effective in hypercoagulability disorders but, in severe ischemia associated with absence of collaterals and thromboses in the microcirculation, its applicability will be limited by the duration of the procedure. Surgery remains the predominant treatment to guarantee a revascularization as fast as possible. The urgency is even more obvious if neurologic deficits are present, indicating severe ischemia. If renal or mesenteric artery involvement is suspected, laparotomy is required in order to restore patency by means of thrombectomy or direct embolectomy. Bypass surgery can be performed if proximal atherosclerotic lesions are present in these arteries. In general, a large surgical field is prepared, including the abdomen, both groins and at least one axilla. The first strategy is generally a retrograde bilateral thrombectomy, which might be successful in embolization of the aortic bifurcation. Completion angiography of both legs delineates patency of the distal arteries and might indicate the need for additional embolectomy. In thromboses of diseased arteries this technique is not effective and revascularization by means of

52

bypass grafting is required. A classic procedure to perform in case of failed thrombectomy is an axillobifemoral bypass, especially in order to reduce the risks associated with laparotomy [25]. Actually, the choice between a direct revascularization via abdominal access or indirect extra-anatomical bypass basically depends on age and general condition of the patient [4,26]. Emergency direct revascularization seems to offer remarkable results [27]. Extraanatomical grafts are indicated in older patients with or without cardiac insufficiency. Intra-arterial blood pressure measurement and insertion of a Swann-Ganz catheter are recommended. Acute aortic occlusion associated with severe ischemia can induce significant metabolic disorders during and after revascularization due to reperfusion of substantial tissue areas. These disorders can cause major complications in one or two legs: neurologic sequelae, compartment syndrome, extensive muscular necrosis leading to amputation or reperfusion injury with multiple organ failure and subsequent death. Controlled limb reperfusion [28,29] and systematic fasciotomies are recommended. It seems logical that, in the near future, these extremely ill patients will benefit from these controlled reperfusion methods to reduce the associated morbidity and mortality [30]. Babu et al. [4] have analyzed the main factors that determine clinical outcome. Bad left ventricular function is a major risk factor since mortality in these patients was 85% as compared to 23% in patients with an adequate left ventricular. Hypercoagulability disorders are also associated with an impressive mortality of 83%. Other risk factors for poor prognosis include suprarenal thrombosis, distal obstructive disease and direct onset extremely severe ischemia. However, neurologic deficit due to spinal cord ischemia is not a poor prognostic risk factor because all cases normalized after revascularization. Therefore, neurologic deficit is not a contra-indication, but rather an indication for prompt revascularization.

Conclusion
Acute abdominal aortic occlusion is a rare phenomenon and is associated with an overall mortality of approximately 50%, mainly due to the comorbidity in the majority of patients such as cardiac pathology and extensive vascular disease of

ACUTE ABDOMINAL AORTIC OCCLUSION

which the aortic occlusion is the final event. Misleading neurologic symptoms can occur in 30%, in which absence of femoral pulses should indicate cardiovascular etiology instead of lower limb paralysis. Medical treatment is still based on anticoagulation. Therapy of choice is surgical revascularization, either by embolectomy, direct grafting or extra-anatomical bypass. Fasciotomy with or without controlled reperfusion should be implemented in the surgical protocol in order to limit compartment syndromes and the general consequences of revascularization. By means of such an integrated strategy, the morbidity and mortality of this severe disease can be reduced.

R E F E R E N C E S

1 Verrier C, Bertrand P, Mercier C, Piquet P. Occlusion aigue du carrefour aortique. In: Kieffer E (ed). Urgences vasculaires non tmumatiques. Paris, AERCV, 1998: pp 373-381. 2 Mozingo J, Denton 1C Jr. The neurological deficit associated with sudden occlusion of the abdominal aorta due to blunt trauma. Surgery 1975; 77: 118-125. 3 Webb KH, Jacocks MA. Acute aortic occlusion. Am] Surg 1988; 155: 405-407. 4 Babu SC, Shah PM, NitaharaJ. Acute aortic occlusion-factors that influence outcome. / Vase Surg 1995; 21: 567-575. 5 Surowiec SM, Isiklar H, Sreeram S et al. Acute occlusion of the abdominal aorta. Am J Surg 1998; 176: 193-197. 6 Kao CL, Chang JP. Abdominal aortic occlusion: a rare complication of cardiac myxoma. Tex Heart /ntf/2001; 28: 324-325. 7 Favre JP, Gay JL, Gournier JP, Barral X. Acute occlusions of the aorta. / Chir 1995; 132: 7-12. 8 Shumacker H. Surgical treatment of aortic aneurysms. Postgrad AM1959; 25:535-548. 9 Jannetta P, Roberts B. Sudden complete thrombosis of an aneurysm of the abdominal aorta. NEnglJMed 1961; 264: 434-436. 10 Hirose H, Takagi M, Hashiyada H et al. Acute occlusion of an abdominal aortic aneurysm-case report and review of the literature. Angiology 2000; 51: 515-523. 11 Schwartz RA, Nichols WK, Silver D. Is thrombosis of the infrarenal abdominal aortic aneurysm an acceptable alternative? J Vase Surg 1986; 3: 448-455. 12 Cambria RP, Brewster DC, Gertler J et al. Vascular complications associated with spontaneous aortic dissection. / Vase Surg 1988; 7: 199-209. 13 Dajee H, Richardson IW, type MO. Seat belt aorta: acute dissection and thrombosis of the abdominal aorta. Surgery 1979; 85: 263-267. 14 Sakakibara Y, Sasaki A, Nakata H et al. Acute aortic thrombosis after intra-aortic balloon pumping. Jpn J Thorac Cardiovasc Swg2000;48: 123-125. 15 Lookstein RA, Mitty H, Falk A et al. Aortic intimal dehiscence: a complication of percutaneous balloon fenestration for aortic dissection. / Vase Interv Radiol 2001; 12: 1347-1350. 16 Imamura H, Asaka M, Saito A et al. Thrombosis of the abdo-

minal aorta in a patient with nephrotic syndrome. Nippon Jinzo Gakkat Shi 2001; 43: 608-612. 17 Szychta P, Reix T, Sevestre MA et al. Aortic thrombosis and ulcerative colitis. Ann Vase Swrg-2001; 15: 402-404. 18 Setoguchi M, Fujishima Y, Abe I et al. Aorto-iliac occlusion associated with the lupus anticoagulant. Report of two cases. Angiology 1997; 48: 359-364. 19 Komori K, Okadome K, Onohara T et al. High aortic occlusion associated with lupus anti-coagulant. EurJ Vase Surg 1992; 6: 302-306. 20 DiCenta I, Fadel E, Mussot S et al. Occlusion of the aorta and inferior vena cava in a patient with circulating anticoagulants. Ann Vase Swrg2002; 16: 380-383. 21 Dossa CD, Shepard AD, Reddy DJ et al. Acute aortic occlusion. A fourty-year experience. Arch Surg 1994; 129: 603-608. 22 Busuttil RW, Keehn G, Milliken J et al. Aortic saddle embolus. A twenty-year experience. Ann Surg 1983; 197: 698-706. 23 Buth J, Cuypers P. The diagnosis and treatment of acute aortic occlusions. J Mai Vase 1996; 21: 133-135. 24 Cunningham M, May S, Tucker W, Gerlock A. Response of an abdominal aortic thrombosic occlusion to local low-dose streptokinase therapy. Surgery 1983; 93: 541-544. 25 Drager SB, Riles TS, Imparato AM. Management of acute aortic occlusion. Aw/Swig 1979; 138: 293-295. 26 Meagher AP, Lord RS, Graham AR, Hill DA. Acute aortic occlusion presenting with lower limb paralysis. J Cardiovasc Surg 1991; 32: 643-647. 27 Bradbury AW, Stonebridge PA, John TG et al. Acute thrombosis of the non-aneurysmal abdominal aorta. EurJ Vase Surg 1993; 7: 320-323. 28 Schlensak C, Doenst T, Bitu-Moreno J, Beyersdorf F. Controlled limb reperfusion with a simplified perfusion system. Thorac Cardiovasc Surg 2000; 48: 274-278. 29 Vogt PR, von Segesser LK, Fagotto E et al. Simplified, controlled limb reperfusion and simultaneous revascularization for acute aortic occlusion. / Vase Surg 1996; 23: 730-733. 30 Julia P, Fabiani JN. Ischemia-reperfusion and compartment syndrome. In: Branchereau A.Jacobs M (eds). Complications in vascular and endovascular surgery (part II). Armonk, Futura Publishing Company, 2002: pp 11-21.

5_
53

This page intentionally left blank

6
HAS MORTALITY RATE FOR RUPTURED ABDOMINAL AORTIC ANEURYSM CHANGED OVER THE LAST 50 YEARS?
JACK COLLIN

Annual rates of abdominal aortic aneurysm (AAA) rupture continue to increase with the ageing of populations throughout the developed world (Figure). Deaths from AAA represent the rates of tobacco smoking over the previous 60 years. Only a third of patients with ruptured AAA have emergency surgery, and operative mortality is around 50%. Intra-operative mortality rates have not changed for a generation but postoperative mortality rates have decreased by 3.5% per decade with improved management of organ failure. Most patients with AAA rupture die without surgery. Consequently, small improvements in operative mortality have negligible impact on overall AAA mortality. In the short term early detection of AAA by population screening and elective repair of aneurysms greater than 55 mm diameter offer the best chance to reduce AAA mortality. In the long term current low rates of tobacco smoking among the higher socio-economic groups of North America and Europe presage the progressive disappearance of AAA as a major cause of premature death.

55

National statistical data

AAA is a disease of the elderly. Rupture of an AAA is uncommon in men under age 55 years or women younger than 60 years. It becomes increasingly common with advancing age. The incidence of death from AAA as a percentage of all deaths peaks in men aged between 70 and 75 years. There

is, however, no peak in the incidence of deaths from AAA per thousand at risk with increasing age. The older the cohort studied, the more deaths from AAA are seen to occur per thousand at risk [1]. The relative risk of death from AAA is 10 times greater in men than in women at age of 60 years. Among those who survive into their late 80s, men are only three times more likely to die from AAA

VASCULAR

EMERGENCIES

56

than women of the same age. Since among the very elderly, women outnumber men by more than three to one, the false perception has sometimes arisen that the disease is more common in elderly women than men. Throughout the developed world the proportion of elderly men and women in the population has been increasing since the agricultural revolution. In the United Kingdom, data from the 2001 census show that for the first time in history those over 65 years of age are more numerous than those under 16 years. In addition to being old and male, in order to be at high risk of developing an AAA it is necessary to have smoked a substantial quantity of tobacco over a number of years. To be at high risk of AAA rupture, a patient with a known AAA should be a current smoker with poorly controlled hypertension [2]. In Europe, cigarette smoking became increasingly common in men during and after the First World War of 1914 to 1918. European women began to smoke in large numbers from the Second World War (1939 to 1945) onward. In Northern Europe and North America, smoking rates among adult males have been slowly declining for the last 30 years, particularly in the higher socio-economic groups, but smoking rates among women have increased. With all the above information in mind, it comes as no surprise to discover that over the last 20 years deaths from AAA in the United Kingdom have continued to increase [3]. Unless cigarette smoking significantly declines as a European addiction, there are sound reasons to predict that the number of patients presenting with AAA will continue to increase.

Has elective AAA surgery reduced overall AAA mortality?

Data from the United Kingdom Multicentre Abdominal Aortic Aneurysm Screening Study (MASS) [4] have shown that for patients with AAA diameters greater than 55 mm measured by ultrasonography, the number needed to treat (NNT) with elective AAA repair to prevent one death from AAA over the following four years is five. For patients with AAA diameters of 50 to 55 mm, the NNT is unknown but likely to be at best very substantially larger. At worst it may be replaced by a number needed to harm (NNH). The United Kingdom Small Aneurysm Trial [5] showed that patients with AAA antero-posterior diameters of 40 to 54 mm measured by ultrasonography randomized to elective surgical treatment were more likely to die from an AAA-related cause than those randomized to best medical treatment. Not surprisingly, after elective surgery fewer deaths occurred from AAA rupture but the number of rupture deaths prevented was exceeded by the number of operative deaths from elective AAA repair. It is clearly inadequate to use death rates from ruptured AAA alone as the measure of successful management of AAA. Scrutiny of death certification in the MASS study revealed that for many patients who died within 30 days of elective surgery, AAA was not mentioned as a cause of death and was not therefore recorded in national statistics.

HAS MORTALITY RATE FOR RUPTURED AAA

Present evidence supports the conclusion that for patients with AAA diameter less than 55 mm, elective AAA repair is likely to do more harm than good. Until the publication of data from the Small Aneurysm Trial, it had become common to perform elective repair on AAAs of modest size. The current fashion for transluminal graft insertion has ensured that many patients at low risk of AAA rupture continue to have surgical treatment for small AAAs. The advocates for these internal fashion accessories justify their enthusiasm by claims that the risks of insertion are low and that they protect the wearer from subsequent AAA rupture. Neither claim currently withstands evidence-based analysis.

CHANGED OVER THE LAST 50 YEARS?

tality is less than 85%. It is likely therefore that some reduction in overall mortality from ruptured AAA could be achieved by better access to specialist vascular surgical services. Most published data on AAA mortality are difficult to interpret. Clarity is obscured by misuse and inappropriate interchange of the distinctly different terms peri-operative, peroperative (intra-operative), and postoperative. Additional fog is created by often-deliberate confusion of emergency and even urgent AAA surgery with repair of ruptured AAA. Bown et al. [7] have made a valiant attempt to penetrate this obfuscation with a recent metaanalysis of 171 articles covering 50 years of ruptured AAA surgery. They have shown that after allowance for publication bias, the intra-operative mortality for ruptured AAA surgery has remained constant at 20%. They deduce that when analysis of retrospective data from individual institutions reveal a high intra-operative mortality rate, authors tend not to report it but instead disclose only the overall 30-day operative or in-hospital mortality. Operative mortality for urgent or emergency operations undertaken in the belief that an AAA has ruptured or is about to rupture (acute AAA surgery) is around a third of that for ruptured AAA surgery and three times that for elective AAA repair. Inclusion of acute AAA surgical data with that for ruptured AAA has the remarkable property of improving the operative mortality figures for both elective and ruptured AAA surgery. In the author's experience, local audit of data often reveals large differences in intra-operative mortality rates between individual surgeons. Part of the difference can be explained by the relative willingness or reluctance of different surgeons to operate on moribund patients with little or no chance of survival. Operative technique and skill play an additional part. Surgery for ruptured AAA is difficult and unforgiving of inexperience or carelessness. The most important principle underlying success is to lose as little blood as possible during the operation additional to that which has already and inevitably been lost to the circulation by pre-operative bleeding. The common causes of unnecessary pre-operative blood loss are listed in the Table. Most intra-operative deaths tend to be attributed to a cardiac cause, which has the advantage of making the surgeon feel better. The reality is different. In almost all cases, the proximate cause of the myocardial ischemia and resultant arrhythmia, asystole, or

What treatment do patients with ruptured AAA receive?

For practical purposes without surgical repair, rupture of an AAA has a mortality of 100%. Whether death occurs within minutes, hours, or days is determined by the site of rupture and the strength of the peri-aortic retroperitoneal connective tissue. Anterior free rupture into the peritoneal cavity results in exsanguination within minutes. Posterior or lateral contained rupture permits transportation of the patient to a hospital. If the patient is fortunate, the hospital will have a vascular surgeon available who can perform emergency AAA repair. If the patient is unfortunate, interhospital transfer will have to be arranged if judged to be appropriate. Finally, both the patient and the vascular surgeon need to agree that an attempt to repair the AAA rupture is sensible and worthwhile. Fewer than 50% of patients with AAA rupture have sufficient initial containment of hemorrhage to survive long enough to reach the hospital alive. Those who do are further reduced in number by delayed diagnosis, unnecessary diagnostic imaging, lack of a vascular surgeon, dilatory transfer arrangements, patient choice, and clinical selection based on age, coexistent disease, and estimated probability of operative survival. Overall, around one third of patients who have AAA rupture currently undergo an operation for its attempted repair. In individual hospitals, operation rates on those admitted range from less than 50% to almost 100%. Patel et al. [6] have shown that ruptured AAA repair is cost-effective provided the operative mor-

57

VASCULAR

EMERGENCIES

Postoperative mortality after ruptured AAA repair

1 Failure to occlude the neck of the AAA before incising the posterior peritoneum. Injury to renal, gonadal or adrenal veins by blind clamping of the AAA neck. Injury to iliac veins by unnecessary dissection behind the common iliac arteries. Delayed recognition of aorto-caval fistula. Inadequate control of inferior mesenteric, lumbar and median sacral arteries. Heparin administration to a patient already fully anticoagulated by massive blood loss and hypothermia. Reperfusion back-bleeding through already oversewn common iliac artery origins. Closure of the abdomen before coagulation has been restored by administration of fresh frozen plasma, platelets, cryoprecipitate and rewarming.

4 5

58

pump failure is blood loss and inadequate blood replacement. The learning curve for successful repair of ruptured AAA is long for this most difficult of vascular surgical operations. In major tertiary referral centers in the US, ruptured AAA may represent as little as 5% of all AAA operations performed [8,9]. In the United Kingdom in some regional vascular centers [10], they account for more than 50% of all AAA operations. Most surgeons currently operating on patients with ruptured AAA perform fewer than five such operations per annum. Even in major referral centers, few individual surgeons operate on more than 10 patients with ruptured AAA each year. In such circumstances it is no surprise that intraoperative mortality remains around 20% as it has done since the early days of the operation fifty years ago. It is improbable that intra-operative mortality rates will improve in the foreseeable future.

In contrast to intra-operative surgical skill-based mortality there is evidence that postoperative 30-day mortality has progressively fallen by approximately 3.5% per decade [7]. This steady fall in postoperative death rates is attributable to the development of the specialty of intensive care medicine and improvements in the management of organ failure. Numerous authors have attempted to devise predictive scoring systems for postoperative death based on both pre-operative and early postoperative evaluation of the function of different organ systems [11-13]. In order of relative importance, survival is determined by control of cardiac, respiratory, renal, and hepatic failure. To date, most success has been achieved by expert management of renal and cardiac failure and least impact has been made on deaths from hepatic and respiratory failure. It is now unusual for a patient to succumb from renal failure alone without coexistent failure of at least one other organ system. The cynical view is sometimes expressed that, with profligate use of intensive care unit facilities, it is now possible to keep many patients alive for 30 days postoperatively who have little if any prospect of independent survival. It is certainly the case that beyond five days intensive care the rule of diminishing returns begins to apply. Few patients who are not fit for discharge from an intensive care unit five days after surgery will ultimately leave the hospital alive. With healthcare systems around the world in financial crisis, the cost effectiveness of long-term intensive care is increasingly being challenged.

Overall mortality for ruptured AAA

At present only a third of patients who rupture an AAA will have an operation. Thirty-day operative mortality rates are around 50%, with wide variations in reported rates from different centers. Two fifths of all operative deaths occur during the operation and intra-operative mortality has remained unchanged for decades. It is reassuring to learn that while today's surgeons are no better than their

HAS MORTALITY RATE FOR RUPTURED AAA CHANGED OVER THE LAST 50 YEARS?
mentors, at least they are no worse. The overall mortality from ruptured AAA remains stubbornly between 80% and 85%. The apparent 3.5% reduction per decade in postoperative mortality attributable to the growth of the expensive and laborintensive specialty of intensive care medicine has had a negligible and undetectable effect on AAA mortality statistics. The most hopeful sign that tobacco smoking in the developed world may be in terminal decline is that it has now been largely abandoned by the trend-setting intellectual elite of Northern Europe and North America. Target growth areas for tobacco sales are now largely third world countries. Secondary prevention is directed at reducing the risk of AAA rupture in those who already have the disease. The proven interventions are smoking cessation and control of hypertension. Specific therapies with drugs to reduce growth rates of AAA at present are unproven. Recently the United Kingdom MASS study [4] established the value of a screening program for AAA detection in men aged 65 to 74 years. Elective surgical repair of AAAs with a diameter greater than 55 mm reduces the incidence of death from AAA-related causes, although all cause mortality is not affected. It is probable that current AAA-specific mortality could be reduced by restricting elective AAA surgery to those patients with AAA antero-posterior diameters measured by ultrasonography of at least 55 mm.

Can AAA mortality be reduced?

Beyond any shadow of doubt, the most important factor underlying the current high incidence of death from AAA was the 20th century phenomenon of cigarette smoking by at its peak up to 75% of the adult male population. Primary prevention of this disease is likely to be much more effective than any imaginable cure. Fashions in human habits and addictions change and the emotional props of one century are prone to disappear in the next as quickly as they appeared in the last.

R E F E R E N C E S 1 Anonymous. Office for National Statistics. Mortality by cause (24). Series DH2, London. The Stationery Office 1999. 2 Brown LC, Powell JT. Risk factors for aneurysm rupture in patients kept under ultrasound surveillance. United Kingdom Small Aneurysm Trial Participants. Ann Surg 1999; 230: 289-297. 3 Anonymous. Office for National Statistics. Mortality by cause (5-24). Series DH2, London. The Stationery Office (1981 -1999). 4 Anonymous. Multicentre Aneurysm Screening Study Group. The Multicentre Aneurysm Screening Study (MASS) into the effect of abdominal aortic aneurysm screening on mortality in men: a randomised controlled trial. Lancet m% 360:1531-1539. 5 Anonymous. The United Kingdom Small Aneurysm Trial Participants. Mortality results for randomised controlled trial of early elective surgery or ultrasonographic surveillance for small abdominal aortic aneurysms. Lancet 1998; 352:1649-1655. 6 Patel ST, Korn P, Haser PB et al. The cost-effectiveness of repairing ruptured abdominal aortic aneurysms. / Vase Surg 2000:32;247-257. 7 Bown MJ, Sutton AJ, Bell PRF, Savers RD. A meta-analysis of 50 years of ruptured abdominal aortic aneurysm repair. Br J Surg 2002; 89: 71 4 -730. 8 Cooley DA, Carmichael MJ. Abdominal aortic aneurysm. 9 Lawrie GM, Morris GC Jr, Crawford ES et al. Improved results of operation for ruptured abdominal aortic aneurysms. Surgery 1979: 85; 483 -488. 10 Collin J, Murie J, Morris PJ. Two year prospective analysis of the Oxford experience with surgical treatment of abdominal aortic aneurysm. Surg Gymcol Obstet 1989: 169; 527-531. 11 Davies MJ, Murphy WG, Murie JA et al. Pre-operative coagulopathy in ruptured abdominal aortic aneurysm predicts poor outcome, fir/ Swrg 1993; 80: 974-976. 12 Meesters RC, van der Graaf Y, Vos A, Eikelboom BC. Ruptured aortic aneurysm: early postoperative prediction of mortality using an organ system failure score. 5r/Swrgl994; 81: 512-516. 13 Brady AR, Fowkes FG, Greenhalgh RM et al. Risk factors for postoperative death following elective surgical repair of abdominal aortic aneurysm: results from the United Kingdom Small Aneurysm Trial. On behalf of the United Kingdom Small Aneurysm Trial participants, fir/ Surg 2000; 87: 742-749.

59

This page intentionally left blank

7
RUPTURED AAA: SHOULD ENDOVASCULAR TREATMENT BE THE FIRST CHOICE?
JAAP BUTH, NOUD PEPPELENBOSCH, NEVAL YILMAZ PHILIPPE CUYPERS, LUCIEN DUIJM, ALEXANDER TIELBEEK

Rupture of an abdominal aortic aneurysm (AAA) remains lethal despite rapid prehospital transport, early diagnosis and resuscitations, expeditious surgical repair and progress in anesthesia and intensive care. Mortality rates remain between 32% and 70% with significant associated morbidity [1-5]. Most centers quote rates near 50 % [6-8]. These high operative mortality rates reflect the magnitude of the physiologic stress of patients following rupture. Hemorrhage, prolonged hypotension, laparotomy and prolonged lower limb ischemia because of aortic clamping all contribute to the risk of cardiac complications, multiple organ failure and death. In addition, the patients are usually elderly and often have pre-existing comorbidities. Hypotension following rupture is often controlled at first by tamponade within the retroperitoneum, but the relaxation of the abdominal tone at induction of general anesthesia often precipitates cardiovascular collapse. Exposure of the neck of the aneurysm together with the dissection through the hematoma causes disruption of retroperitoneal veins and small arteries, resulting in further hemorrhage that is often difficult to control in coagulopathic patients. In the presence of large retroperitoneal hematoma, the aorta is frequently clamped at the supraceliac segment. This renders the viscera and lower extremities ischemic, which contributes to the establishment of a fibrinolytic state [9,10] and has a dramatic effect on cardiac afterload and lactic acid production. Subsequent reperfusion of the lower limbs adds further physiologic injury. Secondary bleeding episodes and other complications such as renal failure, adult respiratory distress syndrome, and colonic and gallbladder ischemia are ultimately responsible for most of the deaths.

VASCULAR

EMERGENCIES

The excessive operative mortality also has important resource implications since most patients will spend many days in the intensive care unit before finally succumbing to the complications of rupture and emergency surgery [11]. In general,, patients who undergo open surgery represent a relatively favorable subset, as a considerable number are not operated at all because of significant comorbid factors. These patients inevitably will die without a surgical option [12].

Endovascular approach
Endovascular repair of ruptured AAA (rAAA) offers the possibility of a significant reduction in operative mortality. This approach relies on the intravascular deployment of an aortic stent graft, introduced via the femoral arteries to exclude the aneurysm from the circulation [13,14]. Laparotomy is avoided and the procedure can be performed under local anesthesia. It is likely that the risk of turning a contained rupture into intraperitoneal hemorrhage by the induction of general anesthesia would be significantly reduced. Additional blood loss due to opening of the retroperitoneal hematoma is avoided, while prolonged infra- or suprarenal aortic clamping is not necessary. Additionally, cardiac stress and duration of lower limb ischemia will be minimized. There is no ample evidence that endovascular aortic repair (EVAR) is technically feasible and safe in patients scheduled for elective AAA repair [1518]. In this chapter we will describe the results of a consecutive cohort study, comparing the impact of a protocol of preferential management by EVAR. The study group includes patients who were mostly, but not all, treated by EVAR. The study group will be compared with a historical control group consisting of patients treated routinely by open surgery for symptomatic or ruptured AAA (rAAA).

62

Methods
From May 2001 onward, patients with symptomatic nonruptured AAA (snrAAA) and rAAA of the abdominal aorta presenting at the Catharina Hospital in Eindhoven were treated according to a well-defined management protocol involving intentto-treat by emergency (e)-EVAR. Patients were considered symptomatic nonruptured AAA if there were

no signs of hemorrhage outside the wall of the aneurysm on computed tomography (CT), but they had acute pain in the abdomen and an abdominal aneurysm that was painful at palpation. Aneurysms were defined ruptured AAA if there was extravasation of blood surrounding the aneurysm at CT examination. In patients who did not undergo CT examination, a retroperitoneal hematoma at open surgery was the criterion for rupture of the aneurysm. Within the study period, all patients who were referred to our hospital with a symptomatic aneurysm of the abdominal aorta were prospectively analyzed and included in this study. On arrival in the emergency ward, the intravenous fluid infusion rate was minimized. The protocol dictated that patients were taken to the radiology department for emergency CT examination with intravenous contrast infusion to opacity the aorta. An exception was made for patients in profound shock or those who had a cardiac arrest during transportation to the hospital. Diameter and length of the infrarenal neck of the aneurysm were measured and the decision whether endovascular repair was feasible was taken and communicated with the operating room staff. Exclusion criteria for e-EVAR were a short neck (less than 10 mm length), a wide neck (more than 30 mm in diameter), and inaccessible iliac arteries. Following CT examination, patients were quickly transported to the operating room for the selected emergency procedure, or taken to the intensive care unit (ICU) for further optimization (only in snrAAA). Patients with rupture of their aneurysm were preferentially treated with an aorto-uni-iliac (AUI) endograft (Fig. 1) combined with a crossover bypass. The study group described above was compared with a control group of patients with symptomatic aneurysms, who were treated by open procedure, between January 1999 and May 2001. This historical control group was retrospectively analyzed by hospital chart review. Primary outcome events that

RUPTURED AAA: SHOULD ENDOVASCULAR TREATMENT BE THE FIRST CHOICE?

63

FIG. 1 A - Intra-operative angiogram demonstrating rAAA B - Intra-operative angiogram demonstrating deployment of proximal component of AUI device C - Intra-operative angiogram demonstrating bilateral iliac arteries D - Postoperative CT examination demonstrating functioning AUI device with complete exclusion of the aneurysm.

VASCULAR

EMERGENCIES

were compared included: 30-day or in-hospital mortality, morbidity, length of hospital and ICU stay, intra-operative blood loss, requirement of blood products and overall fluid infusion during operation. Statistical analysis was performed using SPSS for Windows version 9.0. Chi-square and Fisher tests were used for the comparison of discrete variables and the Mann-Whitney test was used for continuous variables. Continuous variables are presented as the mean range. A p value of smaller than 0.05 was considered significant.

Results
PATIENTS
From May 2001 until June 2002, 40 consecutive patients in the study group were admitted and treated in our hospital because of a ruptured or symptomatic infrarenal abdominal aneurysm (group I, Table I). Fourteen patients had snrAAA and 26 rAAA. Twenty-six patients received endovascular repair (EVAR subgroup) and 14 patients conventional open surgery (COS [conventional open surgery] subgroup, Table II). While there was a trend that

a/ , // Male/female; 'J _ Group I Study group (40 patients)

Mean age , v (range) Years

SnrAAA/ AAA rAAA N

Mean ^ ., , 0 AAA (range) cm

Systolic -.^ u <100mmHg N (%)

r ,. n , Laraiac events Pulmonary events %} N i%} N (/c) ^ (/o) ^

34/6

73.0 (56.8-90.0)

14/26

7.0 (3.6-10.0)

16

(40)

12

(30)

(18)

Group n Control group

22/6

64

(28 patients)

73.2 (58.1-86.7)

6/22

7.5 (4.0-10.5)

14

(56)

(32)

6 (21)

Of Male/female N Subgroup with EVAR (26 patients) Subgroup with

COS

IN SnrAAA/ rAAA N Mean 0AAA (range) cm

I (40 PATMS) Systolic -JOOmrnHg N (%} Cardiac events Pulmonary events N (%) N (%)

Mean age (range) Years

23/3

74.1 (56.8 - 90.0)

10/16

6.7 (3.6-8.8)

(31)

(31)

(23)

11/3

71.0 (58.2 - 80.9)

4/10

7.7 (5.0-0.0)

(57)

4 (29)

(7)

(14 patients)
COS: conventional open surgery EVAR: endovascular abdominal aortic aneurysm repair rAAA: ruptured abdominal aortic aneurysm SnrAAA: symptomatic nonruptured abdominal aortic aneurysm

RUPTURED AAA: SHOULD ENDOVASCULAR

patients in the conventional group were hemodynamically less stable and had larger aneurysms, these group differences in patient characteristics were not significant (Table II). The control group with routine open repair consisted of 28 patients treated between January 1999 and April 2001 (Group II). Six patients in this group had an snrAAA and 22 an rAAA (Table I). There were no significant differences with regard to patient characteristics, presence of pre-operative shock, or previous cardiac and/or pulmonary events between groups I and II. In the study group, 33 (83%) patients underwent emergency CT scanning. Seven patients did not undergo CT scanning, three in the EVAR and four in the COS subgroup. Reasons for not performing a CT scan were profound hypovolemic shock (in two patients) and logistical reasons in five patients. The mean neck length was significantly longer in

TREATMENT

BE THE FIRST

CHOICE?

EVAR compared with COS, 18.0 (6 - 36) mm and 7.5 (0-15) mm, respectively (p = 0.004). The neck diameter in the two subgroups was not statistically different, at 23.8 (17-33) and 27.8 (20-34) mm in EVAR and COS, respectively.

APPLICABILITY OF E-EVAR
EVAR was performed in 26 of the 40 patients in the study group. Reasons for COS were unavailable endovascular specialists in six patients and unsuitable anatomical (dimensions of the infrarenal neck) or technical (profound hypovolemia) reasons in eight patients. Thus, the feasibility of EVAR based on an acceptable aortoiliac anatomy and hemodynamicswas80% (32/40).

PROCEDURAL DETAILS
The intra-operative and hospital aspects in groups I and II are summarized in Table III. In the EVAR

Study group (40 patients)

Total (40 patients)

Mean operation time (range) - Minutes Anesthesia - Number Local Regional General Technique performed - Number Tube Bifurcated AUI + crossover Mean blood loss (range) - Ml Mean fluid infusion (range) - Ml ICU stay (range) - Hours Hospital stay (range) - Days

EVAR (26 patients)

154 (80 - 270) 15

COS (14 patients)

155 (90 - 240)

Control group (28 patients)

155 0 - 270) 15 8 17

176 (100-240)

65

14
21 7
3900 (300-12000) 9000 (5 000 -13 700)

9 2 5 10 5 19 19 2600 1100 1800 (100 - 2 500) (100-6000)* (500 - 6 000) 6600 4700 104 000 (1 500 - 20 500) (1 500 - 12 500) (3 400 - 20 500)* 154 81 46 (0 - 408) (0 - 220) (12 - 408) 12.3 7.2 22.1 (1-60) (0 - 60) (0-21)

11

122 (11-864)

13.0 (0 - 59)

*p<0.01 **p<0.004 number indicates patients unless stated otherwise AUI: aorta-uni-iliac device COS: conventional open surgery EVAR: endovascular abdominal aortic aneurysm repair

VASCULAR

EMERGENCIES

subgroup, no conversions to open surgery were required. Most often an AUI device was used in combination with a femorofemoral cross-over bypass (19 patients). Two patients received a tube endograft and five received a bifurcated stent graft. Of the patients who received a bifurcated stent graft, two patients had an snrAAA. In 88% of the cases in the EVAR subgroup, either local (in 15 patients) or regional (in 8 patients) anesthesia was used. Two patients with an snrAAA and one with an rAAA received general anesthesia. The patient with ruptured aneurysm was in deep hypovolemic shock at arrival in the hospital. The responsible anesthesiologists felt that in this patient a more rapid control of the hemodynamic situation might be obtained by general anesthesia. Mean operation time was 155 (80 to 270) minutes in group I, and 176 (100 to 240) in group II, which was not a significant difference. Mean blood loss in group I was 1 800 (100 to 6 000) mL, and in group II 3 900 (300 to 12 000). This difference was statistically significant (p = 0.01). In addition, there was a significant difference be-

tween groups I and II with regard to total fluid infusion (blood components, fresh frozen plasma, and crystalloids combined; p = 0.0040). Administration of blood components was comparable in the two groups.

PERI-OPERATIVE MORBIDITY AND MORTALITY

The hospital and ICU stay in the study group and control group was not statistically different (Table III). The peri-operative mortality rate in the study group with preferential EVAR was significantly lower than in group II: 20% and 43%, respectively (p = 0.043) (Table IV). If only patients with rupture of their aneurysms were considered, the mortality rate in group I was 31% and in group II 50%. This difference did not reach the level of statistical significance (p = 0.10). Causes of death included continued bleeding, cardiac failure, multiorgan failure, respiratory insufficiency and bowel ischemia. The latter constituted 50% (two patients) of the causes of death in the EVAR subgroup. The rate of

66

Total (40 patients) Mortality (%) Cardiac Pulmonary Continued bleeding Multiorgan failure Bowel ischemia Postoperative morbidity (% of survivors) Cardiac Pulmonary Multiorgan failure Coagulation disorder Cerebral vascular accident (CVA) Reoperation Wound infection/hematoma
p = 0.04 = 0.3

Study gn,p (40 patients) EVAR COS (26 patients) (14 patients)

Control group ? * { Paiens)

8 (20)*

2 2 2 2
14 (44)** 2 3 1 1 3 4

2 1 1 2 1 1

12 (43)* 2 1 2 4 3
4 (25)**

2 1

2
1

1
4

RUPTURED AAA: SHOULD ENDOVASCULAR TREATMENT BE THE FIRST CHOICE?

postoperative morbidity was higher in the study group than in the control group (35% versus 14%, respectively) (Table IV). However, this difference did not reach statistical significance.

General considerations
Conventional open surgery has been the gold standard for the treatment of acute symptomatic aneurysms for five decades. During this period, the peri-operative mortality and morbidity have improved only modestly [18]. After the successful introduction of elective stent graft repair of asymptomatic abdominal aneurysms [16-18], this technique now also becomes appealing for the treatment of acute symptomatic aneurysms [13,14,19]. Mortality rates in these studies range from 10% to 45%, which is promising considering the prohibitive mortality in open surgery. None of these studies, however, was based on an intention to treat by EVAR protocol or on a consecutive patient group. Rather, patients were selected on the bases of availability of experienced staff and other practical aspects. In a previous report from our group, Yilmaz et al. assessed the outcome in a consecutive series; however patients with open procedures during the same period were excluded from the analysis [20]. The present study is the first in which all treated patients were included, irrespective of whether the method of aneurysm repair was by stent graft or by open technique.

FOLLOW-UP OF PATIENTS
In the EVAR subgroup, three patients demonstrated an endoleak during follow-up from 30 days to 14 months: two patients had a type I endoleak and one patient a type II. The latter patient is still under survey and an intervention for coiling has been planned. Both patients with type I endoleak refused further intervention because of their ages of respectively 90 and 80 years. The 90-year-old patient has a follow-up time of 14 months and he remains without symptoms. The 80-year-old patient was discharged from further follow-up at his own request. Follow-up was achieved in 90% of the patients in the study group and 86% in the control group. The patients, with recorded follow-up data were followed in the hospitals from where they originally were referred. The 6-month survival in group I was 74% and in group II 52% (Fig. 2). The difference of 22% already existed after the first postoperative month and there was no further change of this difference in the subsequent follow-up period.

67

FIG. 2 Life table graph of patients in the study group and in the control group. Initially existing difference sustains during the first months of follow-up.

VASCULAR

EMERGENCIES

68

The present study group, consisting of patients with EVAR and COS, understandably demonstrates less differences in pre-operative characteristics compared to controls undergoing surgical repair than in our previous study. No differences in operating time and ICU admission time were noted between groups I and II. Patient selection criteria for EVAR involve in the first place the presence of a suitable infrarenal neck. This was apparent in the study group from a significantly shorter neck in the subgroup with COS compared to the subgroup with EVAR. The use of local rather than general anesthesia may be one of the important factors determining the outcome of treatment [21]. In our study group, 58% received local or regional anesthesia. These types of anesthesia do not influence the tone of the abdominal wall. Relaxation of the abdominal musculature during general anesthesia may change a contained rupture into an intraperitoneal or free rupture, reducing the chance of survival significantly [3]. Local anesthesia has the additional advantage of leaving the sympathetic tone of the arterioles unchanged. Patients with rAAA usually are in a state of compensated shock with maximal vasoconstriction. Release of the sympathetic tone at induction of general anesthesia may cause complete cardiovascular collapse, as most of the surgeons know from practical experience. Reduction of blood loss and fluid administered during operation in the EVAR subgroup are likely related to the avoidance of general anesthesia as much as from avoiding open surgery. Pre-operative CT scanning on the one hand appears quite useful for ascertaining whether endovascular treatment is feasible and for measuring anatomical dimensions. A drawback may be the time delay until the treatment commences. In the present series, only two patients had blood pressures too low to allow an additional delay of 10 to 15 minutes, which is the usual time an emergency CT scan takes in our institution. During CT examination, the operating room is prepared for the operative procedure, making the actual time delay even less. The use of AUI rather than bifurcated endografts is a matter of debate. During our institutional experience, we have developed a strong preference for AUI endografts. The advantages of this device type include a quick introduction and deployment, which rapidly lowers intra-aneurysmal blood pressure and provides control on the intra-abdominal

bleeding [22]. Only one groin needs to be explored under local anesthesia, which seems more easily tolerated by the patient in emergency circumstances than bilateral groin exploitation. An additional advantage seems to be that patients with complex iliac artery anatomy can more be frequently treated by AUI devices since only one suitable side is needed. Nevertheless, Orend et al. and Lachat et al. used bifurcated stent grafts in their selected patient population, with comparable operating times and excellent results [21-23]. To demonstrate the impact of endograft treatment on the first-month mortality of acute AAA requires a carefully analysis. Simple assessment of EVAR-treated patients will lead to skewed outcome, as patients with short necks or profound shock will have the highest risk. It is of note that the present study is also the first to demonstrate a significant difference in first-month mortality in favor of e-EVAR compared to conventional surgery in a combined group of patients with ruptured and symptomatic AAAs. The advantage of e-EVAR continues during the first postoperative months. Apparently there is no catch-up of mortality by delayed events and the favorable effect on survival seems durable. Admittedly, our present follow-up periods were still rather short. It is of note that the incidence of nonlethal postoperative complications in group I was somewhat higher than in group II. A plausible explanation may be that the occurrence of complications in the first group was assessed prospectively and in the latter retrospectively. Nevertheless, this finding signifies the fact that an rAAA remains a serious condition irrespective of the type of treatment and, secondly, that open surgery may increase the overall insult to a patient's condition, resulting in a higher mortality. The present study has several flaws. First, a relatively small number of patients were included and the follow-up period was short. Second, despite the intent to treat by EVAR protocol, six patients (15%) received conventional open repair because of an unavailable endovascular specialist at the time of admission. When these patients also would have been treated by EVAR, the difference in early mortality might have been even greater. Third and perhaps most importantly, the control group in this study was retrospectively analyzed, which may influence the comparability with the study group as well as accuracy of recording of events. A large-scale multicenter study is needed to confirm that emergency EVAR for acute symptomatic and ruptured

RUPTURED AAA: SHOULD ENDOVASCULAR TREATMENT BE THE FIRST CHOICE?

aneurysms is associated with improved survival. Once our findings are confirmed by such a study, additional evidence will be required from a randomized, controlled trial comparing EVAR and open surgery for well-defined indications, distinguishing patients with truly ruptured and symptomatic nonruptured AAA. The importance of a trial monitoring committee that terminates the study as soon as a statistical significant difference in operative survival is obtained at regular interim analyses is obvious in such a randomized, controlled trial. Late complications associated with e-EVAR included the occurrence of endoleaks. In the present study, three endoleaks (12%) of survivors were present after one month. Two patients had a type I endoleak but refused further treatment; they remain without symptoms so far. Nevertheless, in our opinion, these endoleaks should be treated either by the use of a giant Palmaz stent, an aortic extension endograft, or by laparoscopic banding of the aorta. The policy for type II endoleaks may be similar to that for elective EVAR, and intervention should be dependent of eventual increase in size of the aneurysm. In conclusion, emergency endovascular repair of ruptured and acute symptomatic abdominal aortic aneurysms is justified when the patient has a suitable anatomy, most importantly an adequate infrarenal neck. The first-month mortality, in the present study, was significantly lower than in a control group receiving surgical repair. Further study of a larger study population to confirm our findings is needed.

R E F E R E N C E S

1 Johansson G, Swedenborg J. Ruptured abdominal aortic aneurysms: a study of incidence and mortality. BrJ Surg 1986; 73: 101-103. 2 Johansen K, Kohler TR, Nicholls SC et al. Ruptured abdominal aortic aneurysm: the Harborview experience. / Vase Surg 1991;13:240-247. 3 Gloviczki P, Pairolero PC, Mucha P Jr et al. Ruptured abdominal aortic aneurysms: repair should not be denied. / Vase Surg 1992; 15:851-859. 4 Kniemeyer HW, Kessler T, Reber PU et al. Treatment of ruptured abdominal aortic aneurysm, a permanent challenge or waste of resources? Prediction of outcome using a multi-organ dysfunction score. EurJ Vase Endovasc Surg 2000; 19: 190-196. 5 Noel AA, Gloviczki P, Cherry KJ Jr et al. Ruptured abdominal aortic aneurysms: the excessive mortality rate of conventional repair. / Vase Surg 2001; 34: 41-46. 6 Katz DJ, Stanley JC, Zelenock GB. Operative mortality rates for intact and ruptured abdominal aortic aneurysms in Michigan : an eleven-year statewide experience. J Vase Surg 1994; 19: 804-817. 7 Kantonen I, Lepantalo M, Brommels M et al. Mortality in ruptured abdominal aortic aneurysms. The Finnvasc Study Group. EurJ Vase Endovasc Surg 1999; 17: 208-212. 8 Prance SE, Wilson YG, Cosgrove CM et al. Ruptured abdominal aortic aneurysms: selecting patients for surgery. EurJ Vase Endovasc Surg 1999; 17: 129-132. 9 Green RM, Ricotta JJ, Ouriel K, DeWeese JA. Results of supraceliac clamping in the difficult elective resection of infrarenal aortic aneurysms. / Vase Surg 1989; 9: 124-134. 10 Illig KA, Green RM, Ouriel K et al. Primary fibrinolysis during supraceliac aortic clamping./ Vase Surg 1997; 25: 244-254. 11 van Ramshorst B, van der Griend R, Eikelboom BC. Survival and life quality after surgery for ruptured abdominal aneurysm. In: Greenhalgh RM, Mannick JA (eds). The cause and management of aneurysms. London, W.B. Saunders, 1990: pp 433-440. 12 Bradbury AW, Makhdoomi KR, Adam DJ et al. Twelve-year experience of the management of ruptured abdominal aortic aneurysm. BrJSurglWl; 84: 1705-1707.

13 Ohki T, Veith FJ, Sanchez LA et al. Endovascular graft repair of ruptured aortoiliac aneurysms. J Am Coll Surg 1999; 189: 102-113. 14 Greenberg RK, Srivastava SD, Ouriel K et al. An endoluminal method of hemorrhage control and repair of ruptured abdominal aortic aneurysms. / Endovasc 77zer2000; 7: 1-7. 15 Matsumura JS, Pearce WH. Early clinical results and studies of aortic aneurysm morphology after endovascular repair. Surg Clin North Am 1999; 79: 529-540. 16 Zarins CK, White RA, Schwarten D et al. AneuRx stent graft versus open surgical repair of abdominal aortic aneurysms: multicenter prospective clinical trial. J Vase Surg 1999; 29: 292-308. 17 Becquemin JP, Lapie V, Favre JP, Rousseau H. Mid-term results of a second generation bifurcated endovascular graft for abdominal aortic aneurysm repair: the French Vanguard trial. / Vase Swr#1999;30: 209-218. 18 Cuypers P, Buth J, Harris PL et al. Realistic expectations for patients with stent-graft treatment of abdominal aortic aneurysms. Results of a European multicenter registry. Em J Vase Endovasc Surg 1999; 17: 507-516. 19 Yusuf SW, Whitaker SC, Chuter TA et al. Emergency endovascular repair of leaking aortic aneurysm. Lancet 1994; 344:1645. 20 Yilmaz N, Peppelenbosch N, Cuypers PW et al. Emergency treatment of symptomatic or ruptured abdominal aortic aneurysms : the role of endovascular repair. J Endovasc Ther 2002; 9: 449-457. 21 Lachat ML, Pfammatter Th, Witzke HJ et al. Endovascular repair with bifurcated stent-grafts under local anaesthesia to improve outcome of ruptured aortoiliac aneurysms. EurJ Vase Endovasc Surg 2002; 23: 528 -536. 22 Gawenda M, Heckenkamp J, Zaehringer M, Brunkwall J. Intraaneurysm sac pressure. The holy grail of endoluminal grafting of AAA. EurJ Vase Endovasc Swig 2002; 24: 139-145. 23 Orend KH, Kotsis T, Scharrer-Pamler R et al. Endovascular repair of aortic rupture due to trauma and aneurysm. Eur J Vase Endovasc Surg 2002; 23: 61 - 67.

69

This page intentionally left blank

8
URGENT OPEN SURGERY AFTER ENDOVASCULAR AAA REPAIR
PIERGIORGIO CAO, FABIO VERZINI PAOLA DE RANGO, MASSIMO LENTI, GIANBATTISTA PARLANI

Urgent open surgery after endovascular aneurysm repair (EVAR) is infrequently required but can carry a significant mortality in case of aneurysm rupture. Urgent treatment is usually indicated in cases ofischemic or hemorrhagic complications of EVAR and, rarely, for endograft infection. This repair may involve urgent conversion with graft removal at the time of original operation (primary conversion) or at a later stage (secondary conversion), as well as additional procedures with the endograft left in place (open surgery without conversion). Conversion to open repair after EVAR is usually more troublesome than a standard elective operation because it is usually subsequent to the endovascular procedure in patients who are often excluded to open treatment due to severe comorbidities. The best method to avoid the high morbidity and mortality of urgent open surgery after EVAR is prevention. Pre-operative assessment with appropriate patient selection and infra-operative monitoring with possibility of open conversion are essential. Strict surveillance for all patients undergoing EVAR is crucial.

1
71

Indications
Rapidly evolving technology has resulted in a substantial improvement in success rates of EVAR since the earliest published reports [1-4]. Nevertheless, although minimally invasive, EVAR can be associated with an adverse early outcome. Recent studies have addressed serious complication rates occurring in the mid and long term after EVAR, prima-

rily the risk of rupture [5-12]. For some of these adverse events, an open repair can be urgently required. This chapter addresses surgical indications, techniques and reported experiences on urgent open surgery after EVAR. Urgent open surgery after EVAR is usually indicated in cases of ischemic or hemorrhagic complications and rarely for graft infection.

VASCULAR

EMERGENCIES

Surgical treatment may involve urgent conversion with graft removal at the time of the original operation (primary conversion) or at a later stage (secondary conversion), as well as additional procedures with the endograft left in place (open surgery without conversion). PRIMARY CONVERSION Despite adequate pre-operative imaging, correct sizing of prostheses and appropriate case selection, EVAR may fail and require immediate conversion to open repair. Primary conversion usually occurs for vessel injury or endograft misplacement during attempted EVAR in patients who have undergone hours of intervention and may have received a large quantity of contrast agent. Common causes of emergent conversion can be summarized as follows. 1 - Rupture of the aorta or other vessels due to instrumentation during the stent graft implantation procedure. Although this is no longer an absolute indication for immediate conversion, open repair probably remains the safest treatment in these cases. 2 - Inability to deploy the graft because of inadequate size of femoral or iliac arteries. 3 - Graft deployment failure due to mechanical dysfunction of the delivery system (e.g., catheter sheath rupture or elongation during retrieval, wire breakage with partially deployed endografts, impossibility to retrieve the catheter after deployment). 4 - Inaccurate deployment of the endograft (e.g., irreversible twisting of the endograft leading to inability to complete deployment or to limb occlusion) . 5 - Misplacement of the endograft that can lead to aortic, renal or iliac obstruction. 6 - Inadequate exclusion of the abdominal aortic aneurysm (AAA) after apparent proper deployment (primary endoleak). 7 - Graft occlusion. Improvements in technology and interventional techniques, as well as increasing experience, have resulted in a reduction in primary conversion rates. May et al. found that primary conversion rates reduced from 20% to 2%, accurate case selection being crucial to obtain success of deployment [13]. Access problems can be largely overcome with appropriate case selection, the use of smaller diameter devices, and planning preventive open technique as performing iliac conduits. The risk of aortic rupture can be minimized by limiting the oversizing of balloons in balloon-expandable devices depending

on the severity of the atherosclerotic disease and by avoiding dilatation beyond the graft-covered vessel. Immediate device migration can be avoided by proper aortic neck selection: the presence of mural thrombus, inverted funnel shape and heavy circumferential calcification, in addition to a length of less than 15 mm and a diameter larger than 30 mm, are relative contra-indications for the use of the endoluminal repair. SECONDARY CONVERSION One of the main indications for urgent secondary conversion after EVAR is aneurysm rupture. Prevention of AAA rupture with its related morbidity and mortality is the primary driving force behind the concept of aneurysm repair, as the natural history of an untreated AAA is to enlarge and rupture. In spite of the immediate successful exclusion of aortic aneurysm at the time of implantation of the endograft, reports on rupture of AAA after EVAR continue to accumulate, showing a risk as high as 1% per year [14]. Migration, modular component disconnection and other mechanical failures like fabric perforation with acute onset of endoleak represent the most frequent causes of urgent secondary conversion after EVAR. Another indication for urgent open repair after EVAR is represented by graft infection, a rare complication reported in the literature. More prone to this complication seem to be those grafts which are not premounted into a delivery catheter at the time of their insertion and therefore are more exposed to operator manipulations, and patients subjected to multiple re-interventions for any cause [15]. Aortoduodenal fistulae have been reported as rare sequelae of EVAR secondary to stent graft disruption in a preexisting inflammatory process causing adhesions between the bowel and the aortic wall [16]. OPEN SURGERY WITHOUT CONVERSION Some of the indications mentioned above can be the cause of urgent open surgery with the graft left in place, either immediately or after a primarily successful procedure. The more frequent causes of immediate urgent open surgery without conversion are hypogastric occlusion requiring revascularization and iliac artery rupture not amenable to endovascular repair. During follow-up the most frequent indication is represented by graft limb occlusion due to severe kinking of the graft limb.

8_
72

URGENT OPEN SURGERY AFTER ENDOVASCULAR AAA

REPAIR

Surgical techniques
PRIMARY AND SECONDARY CONVERSION
Several differences exist between conversion without deployment, immediate explantation and late explantation. In the case of aortic rupture, the difficulties encountered in obtaining vascular control and in dealing with an aorta that has lost much of its integrity during the process of removing the endograft renders surgical repair much more difficult than repairing a "de novo" ruptured AAA. On the other hand, it has been postulated that the presence of an endograft may somehow limit the quantity and rate of blood loss outside the aneurysmal sac, reducing the hemodynamic changes of the AAA rupture [7]. In conversions occurring during primary graft placement, the possibility of using suprarenal aortic control through balloon inflation must be considered. This technique may be applied quickly, particularly when a guide wire is already in place in the thoracic aorta. However, it may be difficult to maintain the inflated balloon in place because of the pulsatile pressure of the blood flow; balloon migration can be prevented by advancing the balloon over an extra stiff guide wire and fixing both to the field drapes with a hemostat. Alternatively, the balloon can be sustained with a long introducer fixed to the patient's skin. Balloon control is not free from significant complications such as visceral vessel embolization and visceral ischemia, and can render open surgery even more difficult in an operative field already complicated by the presence of the endograft. For this reason, it may be preferable to remove the balloon as early as possible after gaining access to the aortic neck with a standard cross-clamp. Aortic and iliac cross-clamping technique may vary on the basis of the position and degree of deployment of the endograft. A standard open AAA technique may be used in patients in whom the endoluminal approach has been abandoned due to access problems (primary conversion). In such cases, infrarenal aortic clamping is usually possible [17,18]. In cases of secondary conversion for patients with an endograft with infrarenal fixation deployed immediately below the renal arteries, clamps can be gently applied to the aorta and common iliac arteries as in standard open AAA repair, including the underlying graft. After initial clamping and

opening of the aneurysm sac, the proximal end of the endograft may be removed by cautious traction, eventually after opening the aortic clamp and crossclamping the main trunk of the endograft to reduce blood loss. Following this, the iliac clamps will be opened and closed in sequence in a similar fashion to allow removal of the underlying endograft within the jaws. Supraceliac control of the aorta has been recommended by many authors [5,10-11]. This approach is usually required in the following circumstances: 4 when the endograft has been incorrectly deployed over the renal arteries, * in patients who experienced rupture at the level the proximal neck, e.g., for balloon inflation during deployment of the endograft, 4- in patients with serious hemorrhage until control of the infrarenal neck is obtained, * in cases of grafts with suprarenal fixation, * in cases of late conversion when the long standing endograft presence is associated with a periaortic inflammatory response with tissue fibrosis, and removal of the endograft may result in a weak and thinned-out aortic neck for proximal anastomosis, 4 during removal of an infected endoprosthesis. The proximal anastomosis is performed in a standard fashion, usually including a dacron felt or pledgets in the suture line to reinforce the weakened aortic wall. Although the procedure of choice in all instances of explantation is complete removal of the endograft, in the case of endografts anchored by attachment systems at the level of the renal arteries or above, it may be impossible to remove the complete graft, especially during a secondary conversion. In such cases the amputation of the infrarenal portion of the endoprosthesis, by cutting the metal frames of suprarenal attachment system, may be indicated. When complete removal of the proximal portion of the covered graft is impossible, the technique used by Lawrence-Brown can be applied. The proximal transected endograft is incorporated in the proximal suture line, which anastomoses the new graft to the neck of the aneurysm. This technique might be less traumatic than attempting to remove the entire prosthesis. However, some authors speculate that the risk of pseudoaneurysm formation would be increased with the inclusion technique,

8_
73

VASCULAR

EMERGENCIES

particularly if there had been a proximal endoleak [18,19]. Distal anastomoses are usually carried out at the level of the iliac bifurcation in case of iliac aneurysmal involvement. In the case of nonectatic iliac arteries, gentle removal of the iliac limbs of the endograft can prevent intimal damage and an aortic anastomosis can be performed (Fig. 1). In case of firm adhesion of the endograft limbs in the iliac arteries, as in endografts with hooks for distal fixation system, the inclusion technique can be applied also for the distal anastomosis. Alternatively, complete iliac artery transection below the iliac endograft ends should be performed.

OPEN SURGERY WITHOUT CONVERSION

At the time of the initial EVAR operation, some complications can lead to the subsequent need for additional open surgery without the obligation to remove the total graft. The most common indication for open surgery without conversion is bilateral hypogastric occlusion due to incorrect graft planning with errors in length measurement, low graft deployment, or dis-

74

FIG. 1 Gentle iliac limbs removed and clamped prior to distal aortic anastomosis durins secondary conversion for acute type III endoleak that occurred two years after EVAR.

tal migration of the graft. In such cases, risk of intestinal ischemia is high and can be lowered by internal iliac revascularization, at least at one side. The side to be revascularized must be chosen on the basis of specific anatomical characteristics (e.g., common trunk length, presence of an iliac aneurysm, calcium extent and possibility of mobilizing the artery). In general, it may be preferable to choose the left side over the right because of the larger collateral pathways between the right internal iliac and the mesenteric circulation. Surgical approach to the iliac bifurcation is usually best obtained through an eight centimeter oblique incision on the right or left lower quadrant with a retroperitoneal approach. In some patients with a prior vertical inguinal incision, it may be preferable to extend the skin incision cranially with subsequent oblique fascial incision. After mobilizing the hypogastric artery, vascular clamps are applied and the internal iliac is divided. A direct re-implantation of the distal segment over the external iliac is usually difficult because of the endograft inside the proximal external iliac, so the best option is to perform an end-to-end anastomosis with a short segment of PTFE or dacron graft and subsequently suture end-to-side to the distal external iliac. Another possible immediate ischemic complication after EVAR is inadvertent renal occlusion during deployment. An endovascular rescue treatment by stent placement is usually successful in case of partial coverage of the renal ostium, whereas endograft removal is usually indicated in case of total bilateral renal flow obstruction. In the rare event of unilateral renal impairment, the endograft may be left in place and an extra-anatomical renal revascularization can be performed: the hepatic, splenic or iliac arteries can be used as inflow arteries without the need for aortic clamping and graft removal (Fig. 2). Limb ischemia is another complication usually treated by open surgery without endograft removal. Thrombolysis is usually indicated for late obstruction and in cases of viable, nonthreatened ischemic limbs. In our experience, most of the cases of limb ischemia have occurred early after EVAR (within three months) and were related to undiagnosed iliac dissection, graft kink, or compression. In such cases, an open approach with thrombectomy, eventually followed by balloon and stent angioplasty, is an option reserving the femorofemoral extraanatomical bypass in case of failure of mechanical clot removal.

URGENT OPEN SURGERY AFTER ENDOVASCULAR AAA

REPAIR

Review of published cases and our experience

True incidence of AAA rupture after EVAR is difficult to assess because many cases have been reported as isolated anecdotes or from subgroups of patients treated with specific endografts that are no longer in use. Bernhard et al. reported 5 ruptures (4.9%) among 103 patients treated with a firstgeneration Endovascular Technologies (EVT) graft, but none in the subsequent 583 patients who received a second-generation graft [12]. Zarins et al., reporting on 1067 patients treated with the AneuRx endograft (Medtronic AVE), found that the one-year risk of rupture by life table analysis was 0.4%, and the two-year risk of rupture was 2.6% [10]. Vallabhaneni and Harris, reporting on patients from EUROSTAR registry, calculated a rupture rate of about 1% per year that significantly increased after the fifth year up to 2.9% by life table analysis. They also identified type III endoleak (RR 7.47), migration (RR 5.35), and aneurysm diameter at last measurement (RR 1.057) as three significant independent predictors of late rupture in the EUROSTAR population [20].

FIG. 2 Left renal revascularization with splenic artery transposition in a patient with renal occlusion due to endograft coverage of renal ostium: 6-month computed tomography reconstruction showing the proximal part of the aortic endograft, regular patency of the right renal artery and of the spleno-renal transposition.

8_
75
Colon ischemia is one of the most feared complications after EVAR. Fortunately, it occurs rarely and most of the time the mild degree of a mucosal ischemia can be treated medically without sequelae. However, in the rare case of transmural colonic infarction, an urgent bowel resection is needed. Hemorrhagic complications of EVAR can rarely be treated with endovascular methods. Iliac ruptures can seldom be repaired by endografting and urgent open surgery is more often required, especially when rupture occurs near the iliac bifurcation, usually determined by too aggressive balloon dilatation outside the stent graft distal ends. These maneuvers can actually tear off the hypogastric origin, and positioning of a covered stent would not arrest backbleeding from the internal iliac circulation. In such cases, the same approach described for hypogastric revascularization can be used to suture the damaged artery with the aid of balloon hemostatic control inside the stent graft. Eventually, an iliac bending can be applied as shown in Figure 3.

FIG. 3 Left iliac artery repair with bending after rupture secondary to iliac graft ballooning. A retroperitoneal approach with abdominal left lower quadrant incision was performed.

VASCULAR

EMERGENCIES

76

At least 47 cases of aortic aneurysm ruptures after EVAR have been reported in referenced journals and these were recently reviewed by Bernhard et al. [12]. Bifurcated configurations represented the majority of the used grafts (30 patients). Different models of endografts had been used, including 8 Medtronic-AneuRx, 5 EVT-EGS, 3 GuidantAncure, 10 homemade grafts, 5 MinTec Stentor, 2 Medtronic/World Medical-Talent and 11 Boston Scientific Vanguard. Four manufacturers were unreported. The time between implant and rupture ranged from 3 days to 85 months, with a mean of 16.4 (16.8) months (median 16 months). Berhard et al. found that among the causes of rupture, all but one were associated with endoleaks, but only 18 of them were evident before rupture. With respect to the type of endoleak, 12 were type I, 2 were type II, and 4 were undetermined. Primary type I endoleaks (i.e., recognized at implantation) had a tendency to rupture early (within one to six months), which supports the recommendation to treat them without delay. The same authors found that an increase in aneurysm diameter was recorded in 13 of the 31 ruptured cases with available information (42%). The majority of patients had evidence of other abnormalities noted on prerupture imaging studies, some of which were recognized only in retrospection: besides aneurysm enlargement, migration and loss of device integrity were the most common underlying problems contributing to the onset of an acute endoleak at the time of rupture. Some of these problems appeared to be

related to early generation designs that are no longer manufactured, whereas newer endografts have been followed for too short a time to determine their durability. Therefore, it is mandatory that rigorous surveillance is applied to all patients until sufficient long-term evidence exists to define the most reliable follow-up regimen for any given device. In the Bernhard review of ruptured aneurysm after EVAR, open surgical repair was performed in 40 of 47 patients with an overall peri-operative mortality rate of 41% [12].

Personal experience
Between April 1997 and October 2002, 451 consecutive patients underwent elective EVAR for AAA at our unit. Median follow-up was 25 months (range 1 to 66 months). Early mortality rate in our cohort was 1.1% (5/451), while immediate conversion rate was 1.5% (7/451). At a median follow-up of 24 months, overall late mortality occurred in 45 patients (10.2%). Causes of late mortality included 3 AAA ruptures, 8 cardiac failures, 1 bowel occlusion, 1 pulmonary embolism, 10 myocardial infarctions, 11 cancers, 2 traumas, 1 renal failure, 1 suicide, 2 strokes, 3 respiratory insufficiencies, 1 intracranial hemorrhage and 1 bleeding duodenal ulcer. Aneurysm-related death rate (i.e., peri-operative mortality and late mortality due to aneurysm rupture or any aortic re-intervention) is shown in Table I and in Figure 4.

Fottow-up interval (month)

Number at risk

Number of terminal events

5 0 0 1 1 0 0 0 0

% of patients free from aneurysm-related death 98.75 98.75 98.75 98.26 97.71 97.71 97.71 97.71 97.71

Cumulative standard error 0.0056 0.0056 0.0056 0.0074 0.0092 0.0092 0.0092 0.0092 0.0092

0-6 6-12 12-18 18-24 24-30 30-36 36-42 42-48 48-54

451 350 298 253 220 150 118 59 32

URGENT OPEN SURGERY AFTER ENDOVASCULAR AAA

REPAIR

FIG. 4 Freedom from aneurysm-related death in 451 patients after endovascular AAA repair (life table analysis).

Urgent open surgery during or immediately after EVAR was required in 28 cases (6.2%) (see Table II). Seven immediate urgent conversions were performed: six patients required immediate conversion because of impossibility to deploy the endograft as

planned, in all cases related to narrow, tortuous and calcified vessels. In the remaining patient, aortic neck rupture occurred at the time of graft deployment and required emergent conversion to open repair; a supraceliac aortic control was necessary

Indication far urgent open surgery

Primary conversion Impossibility to progress into diseased iliac arteries Aortic rupture Secondary conversion Symptomatic AAA expansion for acute type I endoleak secondary to migration AAA rupture due to component disconnection Open surgery without conversion Iliac rupture Graft limb or iliac occlusion Unintentional bilateral hypogastric occlusion Renal infarction and parenchymal hemorrhage secondary to renal angioplasty Left colon infarction, without hypogastric occlusion Renal ischemia due to renal artery occlusion * Peri-operative: within 30 days AAA: abdominal aortic aneurysm

Peri-operative* N
7 6 1 -

Late N

2 1 1

14 2 6 3 2 1

5 4

VASCULAR

EMERGENCIES

8_
78

and a tube graft was implanted. Unfortunately, the patient died two days after surgery because of multiple organ failure due to massive blood losses. This case represents the only death related to urgent open surgery after EVAR in our cohort of patients (5.2%). Two patients underwent secondary urgent conversion (at 18 and 40 months after EVAR, respectively) for acute proximal type I endoleak due to graft migration and for AAA rupture due to proximal cuff-aortic body disconnection, respectively. Three other patients experienced rupture during follow-up to our knowledge; they were admitted to other hospitals and died without conversion to open repair. Urgent open surgery without conversion was performed in 19 patients: 14 early and 5 late. Early open surgery included: * two nephrectomies: one for intraparenchymal hemorrhage after simultaneous EVAR and renal stenting for an AAA associated with renal stenosis, the other for renal infarction due to inadvertent left renal artery coverage; * one left colectomy for intestinal ischemia in a patient with bilateral patency of internal iliac arteries; * four femorofemoral bypass grafts and two thrombectomies for limb ischemia; * two iliac artery repairs for intraprocedural rupture: one patient was treated with arterial suture and bending and the other with external iliac-tofemoral bypass; * three external-to-internal iliac bypasses in four patients after unintentional bilateral hypogastric coverage. The fourth patient remained without

events although hypogastric revascularization has not been attempted because of the presence of extensive iliac calcification. Late open surgery included: * one spleno-renal bypass for rapidly evolving renal insufficiency and hypertension secondary to renal artery occlusion; * four femorofemoral bypass grafts for limb ischemia due to graft occlusion.

Conclusion
We can conclude that urgent open surgery after EVAR is required infrequently but, in the case of AAA rupture, carries a significant mortality, similar to that expected for patients without prior endograft surgery. The potential extremely dangerous consequences of complications after EVAR impose a high level of attention 1 - pre-operatively, with appropriate patient selection, 2 - in the operating room, with fully equipped patient monitoring and possibility of open conversion, 3 - during follow-up. The best way to avoid the high morbidity and mortality associated with urgent open surgery after EVAR is prevention. All EVAR. patients should therefore be indefinitely monitored at six-month intervals, and abnormalities should be identified and carefully considered for elective endovascular or open correction.

R E F E R E N C E S 1 ParodiJC, PalmazJC, Barone HD. Transfemoral intraluminal graft implantation for abdominal aortic aneurysms. Ann Vase S7gl991; 5: 491-499. 2 Coppi G, Moratto R, Silingardi R et al. The Italian trial of endovascular AAA exclusion using the Parodi endograft. JEndovasc Surg 1997; 4: 299-306. 3 Moore WS. The EVT tube and bifurcated endograft systems: technical considerations and clinical summary. EVT investigators. JEndovasc Surg 1997; 4: 182-194. 4 May], White GH, Waugh R et al. Comparison of first- and second- generation prostheses for endoluminal repair of abdominal aortic aneurysms: a six-year study with life table analysis. J Vase Surg 2000; 32: 124-129. 5 Lumsden AB, Allen RC, Chaikof EL et al. Delayed rupture of aortic aneurysms following endovascular stent grafting. Am J Swrgl995; 170: 174-178. 6 Torsello GB, Klenk E, Kasprzak B, Umscheid T. Rupture of abdominal aortic aneurysm previously treated by endovascular stentgraft. / Vase Surg 1998; 28: 184-187. 7 May J, White GH, Waugh R et al. Rupture of abdominal aortic aneurysms: a concurrent comparison of outcome of those occurring after endoluminal repair versus those occurring de novo. EurJ Vase Endovasc Surg 1999; 18: 344-348. 8 Krohg-Sorensen K, Brekke M, Drolsum A, Kvernebo K. Periprosthetic leak and rupture after endovascular repair of abdominal aortic aneurysm: the significance of device design for long-term results. / Vase Surg 1999; 29: 1152-1158.

URGENT OPEN SURGERY AFTER ENDOVASCULAR AAA

9 Breek JC, Hamming JF, Lohle PN et al. Spontaneous perforation of an aortic endoprosthesis. EurJ VascEndovasc Surg 1999; 18: 174-175. 10 Zarins C, White RA, Fogarty TJ. Aneurysm rupture after endovascular repair using the AneuRx stent graft. J Vase Surg 2000; 31: 960-970. 11 Politz JK, Newman VS, Stewart MT. Late abdominal aortic aneurysm rupture after AneuRx repair: a report of three cases. /Vase Surg 2000; 31: 599-606. 12 Bernhard VM, Mitchell SM, Matsumura JS et al. Ruptured abdominal aortic aneurysm after endovascular repair. / Vase Swrg2002;35: 1155-1162. 13 May J, White GH, Yu W et al. Endovascular grafting for abdominal aortic aneurysms: changing incidence and indication for conversion to open operation. Cardicrvasc Sing 1998; 6: 194-197. 14 Harris PL, Vallabhaneni SR, Desgranges P et al. Incidence and risk factors of late rupture, conversion, and death after endovascular repair of infrarenal aortic aneurysms: the EUROSTAR experience. European Collaborators on Stent/graft Techniques for Aortic Aneurysm Repair. / Vase Surg 2000; 32: 739-749.

REPAIR

15 Matsumura JS, Katzen BT, Hollier LH, Dake MD. Update on the bifurcated Excluder endoprosthesis: phase I results. J Vase Surg 2001; 33 (2 suppl): S150-153. 16 Norgren L, Jernby B, Engellau L. Aortoenteric fistula caused by a ruptured stent-graft: a case report. J Endovasc Surg 1998; 5: 269-272. 17 Schlensak C, Doenst T, Hauer M et al. Serious complications that require surgical interventions after endoluminal stent-graft placement for the treatment of infrarenal aortic aneurysms. /fcSwrg2001; 34:198-203. 18 May J, White GH, Harris JP. Techniques for surgical conversion of aortic endoprosthesis. EurJ Vase Endovasc Surg 1999; 18: 284-289. 19 Jacobowitz GR, Lee AM, Riles TS. Immediate and late explantation of endovascular grafts: the endovascular technologies experience. / Vase Surg 1999; 29: 309-316. 20 Vallabhaneni R, Harris P. Overview of the complications following endovascular AAA repair. In: Branchereau A, Jacobs M (eds). Complications in vascular and endovascular surgery (Part II). Armonk, Futura Publishing Co, 2002: pp 129-136.

79

8_

This page intentionally left blank

9
ACUTE COMPLICATIONS FOLLOWING LAPAROSCOPIC SURGERY
MARC COGGIA, ISABELLE DI CENTA ISABELLE JAVERLIAT, OLIVIER GOEAU-BRISSONNIERE

Laparoscopy is now considered the technique of choice for many surgical procedures, especially in general surgery, gynecology and urology. It offers the advantages of minimally invasive surgery compared with conventional surgery: reduced surgical trauma, minimal postoperative pain, faster recovery and shortened postoperative hospital stay. With its overwhelming success, increasing numbers of patients are undergoing laparoscopy for a variety of procedures. Despite its popularity, a number of complications specifically related to the laparoscopic approach have been described. Injuries to the great retroperitoneal vessels are the most severe complications. These complications, which occur with Veress needles or trocars, are completely unknown in conventional surgery. More recently, aortic surgery has entered the field of laparoscopic surgery. Even though no specific vascular complications have been reported during totally laparoscopic or laparoscopic-assisted aortic procedures, such complications are possible and will probably occur with the growing experience in these new approaches.

_ J 81

Major vascular Complications in I " . laparoscopic surgery Major vascular complications can occur during the set-up phase of laparoscopy (76.5%), related to the early maneuvers to enter the peritoneal cavity, or

can be secondary to the surgical dissection required fr sPec^lc laparoscopic procedures (23.5%) [1]. The incidence of major vascular complications during laparoscopic surgery is between 0.05% and 0.25% [1-6]. This incidence might seem insignificant in view of the considerable number of laparoscopic procedures, but it is probably underestimated because many surgeons do not publish their experience [3].

VASCULAR

EMERGENCIES

THE SET-UP PHASE OF LAPAROSCOPY

Major vascular injuries. Major vascular injuries occurring during laparoscopic surgery are dependent on the technique used to enter the peritoneal cavity. The favored method of establishing a pneumoperitoneum in laparoscopic surgery is the blind (or closed) technique. This involves insufflation of the peritoneal cavity through a needle (Veress), which is blindly placed intraperitoneally. Subsequently, a trocar is inserted blindly to allow introduction of a laparoscope. Predisposing factors for visceral and vascular injury due to perforation by the Veress needle or the first trocar are adhesion between viscera and the abdominal wall resulting from previous inflammation or surgery, patients with a thin abdominal wall and poor technique or surgical inexperience. Using the blind technique to enter the peritoneal cavity, it is important to know the relationship of the umbilicus to the aortic bifurcation. Kurd et al. [4] showed that the location of the umbilicus was more caudal in heavier women and negatively correlated with body mass index (BMI). In nonobese women (BMI less than 25 kg/m2), the mean location of the umbilicus was 0.4 cm caudal to the aortic bifurcation, and was at or cephalad to the aortic bifurcation in 53%. In overweight women (25<BMI<30 kg/m2), the mean umbilical location was 2.4 cm caudal to the aortic bifurcation. In obese women (BMI higher than 30 kg/m2), the mean umbilical location was 2.9 cm caudal to the aortic bifurcation and was located at the level of the umbilicus in 30%. In all cases, the umbilicus was located cephalad to where the common iliac vein crossed the midline. The open technique is an alternative method of creating the pneumoperitoneum. In this technique the peritoneal cavity is opened under direct vision and a blunt-tipped trocar is introduced. Although the open technique appears to be safer than the closed technique, most surgeons prefer to use the closed technique because it is faster, requires a smaller incision and is not associated with leakage of carbon dioxide. However, operative time of open laparoscopy is significantly shorter than that of closed laparoscopy. Moreover, insufflation of the peritoneal cavity through the Veress needle is done with a low-flow setting. In open laparoscopy, insufflation through the trocar can be done with high flow because of its diameter. Gas leakage can be solved by anchoring the trocar to the abdominal fascia and by replacing additional sutures around the trocar.

82

Major vascular injuries related to the early maneuvers to enter the peritoneal cavity occur with the blind insertion technique of the Veress needle and primary trocar [1-6]. In a review of the literature, Bonjer et al. [6] collected six retrospective studies concerning 489 335 patients operated with the closed technique. The total incidence of vascular injuries was 0.075%. Of the specified injuries, 39.8% were caused by insertion of the Veress needle. The location of vascular injury due to insertion of a trocar was reported as aorta (13/38), inferior vena cava (5/38), iliac artery (12/38), iliac vein (7/38) and mesenteric artery (1/38). For Chapron et al. [1], major vascular injury occurred in the following locations: vena cava (23.8%), external iliac vessels (23.8%), aorta (19%), common iliac vessels (19%), mesenteric vessels (9.6%) and unspecified (4.8%). In six other retrospective studies concerning 12 444 patients operated with the open technique, vascular injuries were not reported [6], Prospective randomized studies of open versus closed laparoscopy have not been performed. Indeed, since the incidence of vascular injuries during laparoscopy is low, a prospective study comparing open and closed technique would require excessive numbers of patients to allow statistically significant statements to be made [3]. However, the comparative study performed by Bonjer et al. [6] is in accordance with the literature review and shows that the incidence of vascular injury associated with closed laparoscopy is significantly greater than that of open laparoscopy. Vascular injuries secondary to Veress needle insertion usually result in small two to three millimeters puncture lacerations, which can be repaired primarily with placement of a few interrupted vascular sutures to obtain adequate hemostasis. The indications for exploration include blood via Veress needle aspiration, hemodynamic instability, active intra-abdominal hemorrhage or expanding retroperitoneal hematoma [2,3]. When there is a return of blood via the Veress needle, the sheath must be left in place in order to tamponade the bleeding and help locate the site of injury [3]. Usal et al. [3] underlined the importance of inserting the insufflation needle at a 45 degree angle in order to prevent injuries. Major vascular traumas secondary to the primary trocar are caused by the same mechanism as the Veress needle. The indications for exploration include hemodynamic instability, active intra-abdominal hemorrhage or expanding retroperitoneal hematoma [2,3]. Vascular injury is recognized imme-

ACUTE COMPLICATIONS FOLLOWING LAPAROSCOPIC

diately in most cases and prompts emergency laparotomy to repair the lacerated vessel [1,6]. Regardless of the mechanism of vascular injury, it is essential that the posterior wall of the injured vessel is explored to assure that a through and through injury has not occurred [2], To prevent perforation of viscera or blood vessels, disposable trocars with safety shields were developed [6]. The safety shield is supposed to shoot forward after the peritoneum has been penetrated. However, the shield can be held back for a considerable distance, particularly in cases of insufficient pneumoperitoneum. This leaves the stylet of the trocar unprotected in the peritoneal cavity. Then, in thin patients or when the abdominal cavity is not sufficiently elevated, the unprotected stylet can easily cause vascular injury [6]. Mortality of major vascular injury is about 10% [1,3] and represents the second most common cause of death for laparoscopic procedures after anesthesia related causes. Delay in diagnosis is probably the most significant contributor to associated morbidity and mortality [2]. Other vascular injuries. The use of multiple trocar sites increases the risk of injury to the epigastric vessels [3]. These injuries can be recognized by the dripping of blood down the trocar sleeve into the abdomen or by hematoma formation around the trocar insertion site. It is important to avoid these injuries by placing the trocar after lighting the abdominal wall videoscopically and identifying the epigastric vessels, especially in thin patients. Gas embolism. Gas embolism is a rare but often fatal complication of laparoscopy. It is caused by direct insufflation of carbon dioxide into a blood vessel, especially vena cava or its branches. While the incidence of gas embolism is 0.001% in the closed technique, open laparoscopy is highly unlikely to cause gas embolism because the trocar is inserted under direct vision. Delay in diagnosis can be due to a number of factors, the most important being the absence of blood in the peritoneal cavity or the presence of a small retroperitoneal hematoma which can easily remain undetected,

SURGERY

depends on the severity of the hemorrhagic syndrome and on the surgeon's experience. Carbon dioxide pneumoperitoneum may cause a significant decrease in splanchnic vessel blood flow [7,8]. Decrease in visceral blood flow results from drop in cardiac output (-30%), direct mechanical compression of intra-abdominal blood vessels or via humoral mechanisms and mesenteric vasoconstriction due to hypercapnia (transperitoneal absorption of carbon dioxide). Moreover, carbon dioxide pneumoperitoneum at a pressure above 15 mmHg may cause visceral vasoconstriction as a result of intra-operative release of vasopressin, increasing portal vein pressure due to retained C02- Use of halothane may also contribute to the reduction in visceral vessel blood flow [8], Such a decrease in splanchnic vessel blood flow may be catastrophic in patients with certain preexisting conditions (impairment of splanchnic vessels, hypercoagulable states). Then it is important that intra-abdominal pressure be kept as low as possible throughout the procedure and should always be maintained below 15 mmHg. Minute volume insufflation should be maintained at an adequate level to avoid hypercapnia (less than 8 L/min).

Major vascular complications in laparoscopic aortic surgery

Vascular surgery recently entered the field of laparoscopy, and specific vascular complications of laparoscopic aortic surgery have not yet been reported. Since November 2000 we developed a new approach for totally laparoscopic aortic procedures [9]. Initially we used this technique to treat severe aorto-iliac occlusive disease. Since January 2002 we have used this totally laparoscopic approach for the surgical treatment of infrarenal aortic aneurysms. Vascular injuries are possible during the laparoscopy procedures or during the postoperative period.

83

LAPAROSCOPY PROCEDURES
Major vascular complications encountered during laparoscopy procedures were hemorrhagic complications and clamping problems. Hemorrhagic complications. Intraprocedural hemorrhagic complications were noted in two patients and necessitated surgical conversions. One patient operated for severe aorto-iliac occlusive disease

THE LAPAROSCOPY PROCEDURE

Major vascular injury during the laparoscopy procedure is possible, as it can be observed during conventional surgery. The most common cause is the use of monopolar electrode (iliac artery and vein) and sharp dissection (vena cava) [1]. It requires surgical treatment, if possible by laparoscopy but it

VASCULAR

EMERGENCIES

84

had an acute bleeding from the left prosthetic limb after completion of the aorto-prosthetic anastomosis. This was due to a leakage of the clamp that was applied on the prosthetic limb before tunneling. Actually, this complication is avoided by a binding previously placed around the left prosthetic limb. The second patient was operated at the beginning of our experience in totally laparoscopic procedures for infrarenal aortic aneurysm. He had a back-bleeding from a lumbar artery after a laparoscopic endo-aneurysmorrhaphy with aorto-bi-iliac bypass. Because of the poor visibility, it was not possible to locate the origin of the bleeding and a surgical conversion was made. Actually, with the use of a stapler device designed for laparoscopic hernia repair, an internal control of the lumbar arteries is always possible. Moreover, it is remarkable to note that with the intraperitoneal pressure, the black-bleeding from the lumbar arteries is less than usually observed in conventional surgery. Clamping problems. Aortic clamping is a challenging problem during laparoscopic aortic surgery, especially when the infrarenal aorta is extensively calicified. Pre-operative computed tomography scan is then essential to assess the location and extent of aortic calcifications before a laparoscopic procedure is considered. Surgeons consider that aortic calcifications are contra-indications for totally laparoscopic aortic procedures [10-14]. However, with our laparoscopic approach of the abdominal aorta, a simple suprarenal aortic clamping is possible. The approach of the suprarenal aorta is obtained with the use of a partial mediovisceral rotation. This technical point is important regarding the technical difficulties encountered with calcified vessels. Indeed, we performed, if necessary, grasping forceps endarterectomies of the aortic wall and then moved the clamp distal to the renal arteries before performing the aorto-prosthetic anastomosis. If back-bleeding is observed from the distal aorta, we perform a rapid endarterectomy to reposition the aortic clamp and suture the aortic stump. A balloon occlusion of the distal aorta could also be used. THE POSTOPERATIVE PERIOD Major vascular complications encountered in the postoperative period were not specific but their mechanisms were related to the aortic laparoscopic technique.

Hemorrhagic complications. In our experience, two patients had major hemorrhagic complications. The third patient of our series was re-operated eight hours after the end of the laparoscopic procedure for an aorto-prosthetic suture line bleeding. At re-operation, bleeding from large needle holes was observed on the posterior wall of the aorta. In one point, the suture had torn out of the aorta. We believe that these anastomotic problems were due to needle size (3/0 poplypropylene sutures) and brittleness of the aortic wall which was endarterectomized during the initial procedure. Actually, when we perform an aortic endarterectomy before the anastomosis, the suture line is carried out with a thin polypropylene suture in order to reduce the size of needle holes. Prosthetic pledgets are also used to reinforce some stitches. Another patient had a retroperitoneal hematoma, which required a surgical exploration at day 21. We did not observe any anastomosis leakage or other causes for this retroperitoneal hematoma. Four other patients had retroperitoneal hematoma, which did not necessitate surgical exploration but required blood transfusions. We think that these hematomas resulted from coagulopathy. This coagulopathy was induced by hemodilution and the use of heparinized saline for irrigation of the operative field. Actually, these two factors have been corrected and, in the latest patients, we did not observe such retroperitoneal hematomas. However, it is important to know that with laparoscopy, tamponade of the operative field is not as effective as in conventional aortic surgery. Therefore it is important to control the surgical hemostasis, especially at the level of the suture line(s) and the dissection planes. Thrombotic complications. Prosthetic limb thrombosis is a complication of aortic and peripheral vascular surgery. However, it remains rare after aorto-iliac or aortofemoral surgery. Two factors are predisposing for such complications after aortic laparoscopic procedures: 1 - the aortic anastomosis time can be long and blood can then stagnate in the limbs of the prosthesis, 2 - twist of prosthetic limbs is possible because with laparoscopy it is more difficult to position the limb during the tunneling. If these technical problems are not recognized during the procedure, they can lead to postoperative acute thrombosis of the prosthesis.

ACUTE COMPLICATIONS FOLLOWING LAPAROSCOPIC SURGERY

Conclusion
Laparoscopic surgery is technically demanding, especially in vascular surgery. Knowledge of surgical anatomy is very important to avoid lesions. The

surgeon's experience is also an important and even crucial factor for preventing vascular injuries. We believe that learning curve and laparo-training are essential and allow for a decrease in the number of technical errors.

R E F E R E N C E S 1 Chapron CM, Pierre F, Lacroix S et al. Major vascular injuries during gynecologic laparoscopy. / Am Coll Surg 1997 ; 185 : 461-465. 2 Saville LE, Woods MS, Laparoscopy and major retroperitoneal vascular injuries. SurgEndosc 1995 ; 9 : 1096-1100, 3 Usal H, Sayad P, Hayek N et al. Major vascular injuries during laparoscopic cholecystectomy. An institutional review of experience with 2589 procedures and literature review, Surg Endosc 1998 ; 12 : 960-962, 4 Hurd WW, Bude RO, DeLancey JO, Pearl ML. The relationship of the umbilicus to the aortic bifurcation: implications for laparoscopic technique. Ofetel Qjmcd 1992 ; 80 : 48-51. 5 Catarci M, Carlini M, Gentileschi P, Santoro E. Major and minor injuries during the creation of pneumoperitoneum. A multicenter study on 12,919 cases. SurgEndosc 2001 ; 15 : 566-569. 6 Bonjer HJ, Hazebroek EJ, Kazemier G et al. Open versus closed establishment of pneumoperitoneum in laparoscopic surgery, Br JSurg 1997; 84: 599 -602, 7 Sternberg A, Alfici R, Bronek S, Kimmel B. Laparoscopic surgery and splanchnic vessel thrombosis. / Laparoendosc Adv Surg Tech 1998 ; 8 : 65-68. 8 Richmond BK, Lucente FC, Boland JP. Laparoscopy-associated mesenteric vascular events: description of an evolving clinical syndrome. / Laparoendosc Adv Surg Tech 1997 ; 7 : 363-367. 9 Coggia M, Bourriez A, Javerliat I, Goeau-Brissonniere 0. Totally laparoscopic aortobifemoral bypass: a new and simplified approach. EurJ VascEndovasc Swrg2002 ; 24 : 274-275. 10 Dion YM, Gracia CR. A new technique for laparoscopic aortobifemoral grafting in occlusive aorto-iliac disease. J Vase Surg 1997 ; 26 : 685-692, 11 Barbera L, Ludemann R, Grossefeid M et al. Newly designed retraction devices for intestine control during laparoscopic aortic surgery: a comparative study in an animal model, SurgEndosc 2000 ; 14 : 63-66, 12 Said S, Mall J, Peter F, Muller J. Laparoscopic aortofemoral bypass grafting: human cadaveric and initial clinical experiences. / Vase Surg 1999 ; 29 : 639-648. 13 Alimi Y, Hartung 0, Orsoni P, Juhan C. Abdominal aortic laparoscopic surgery: retroperitoneal or transperitoneal approach ? EurJ Vase Endovasc Surg 2000 ; 19 : 21-26. 14 Ahn SS, Hiyama DT, Rudkin GH et al. Laparoscopic aortobifemoral bypass. / Vase Surg 1997 ; 26 : 128-132.

85

This page intentionally left blank

10
ACUTE TYPE B AORTIC DISSECTION: SURGICAL INDICATIONS AND STRATEGY
MICHAEL JACOBS, TED ELENBAAS GEERT WILLEM SCHURINK, BAS MOCHTAR

The clinical spectrum of acute type B aortic dissection is broad, varying from rupture with subsequent death to asymptomatic, uncomplicated presentation. Patient population also comprises different ages and etiology, ranging from the young Marfan patient to the older hypertensive, atherosclerotic patient. This heterogenic etiology, symptomatology and patient population obviously requires a differentiated approach. The indications for emergency operation for acute type A aortic dissection are established, but the management of acute type B dissection remains controversial. In general it is advocated that patients who have type B acute dissection without complications be treated conservatively during the acute phase and surgical treatment be selected if the aortic diameter becomes enlarged during the chronic phase. However, recent publications question this medical treatment as the most optimal approach in acute type B dissections and propose a more surgical attitude. This chapter summarizes the main issues determining the choice for surgical treatment in patients with acute type B dissection, also addressing the outcome of conservative treatment with subsequent related indications for surgical intervention. Surgical strategies are described whereas endovascular techniques are discussed in another chapter of this book.

._ L(J 87

Non-surgical treatment Since most patients with acute type B dissection suffer from (severe) hypertension, medical treatment is focused on hypotensive therapy. In general,

patients are admitted to a critical care unit and EGG and blood pressure are closely monitored. During the acute phase of dissection, nitrate, calcium channel antagonist and B-adrenergic receptor blocker medications are administered intravenously. If no

VASCULAR

EMERGENCIES

10

complications occur, medical treatment is continued until the blood pressure reaches acceptable levels and can be managed by oral medication. Following discharge from the hospital the intensive follow-up programme starts, including control of blood pressure measurements and CT/MRA imaging. This conservative policy generally offers acceptable outcome in the majority of cases with limited aortic enlargement during follow-up, assuming that antihypertensive treatment is successful [1]. According to a definition with a debatable period of fourteen days, acute type B dissections become chronic type B dissections and retrospective analysis of these patients might identify indications for surgical treatment of acute patients in order to prevent serious complications in the chronic phase. Kato et al. [2] performed univariate and multivariate factor analyses in patients with type B dissection who had been treated medically during the acute phase to determine the predictors for chronic phase enlargement (greater than 6 cm) of the dissected aorta. The predominant predictors for aortic enlargement in the chronic phase were the existence of a maximum aortic diameter of or greater than 40 mm during the acute phase and a patent primary entry site in the thoracic aorta. The values of actuarial freedom from aortic enlargement for these patients at 1, 3 and 5 years were 70%, 29% and 22%, respectively. No aortic enlargement was observed in the other patients throughout the entire follow-up period. The authors recommended surgical treatment during the acute phase in patients with a large aortic diameter (superior or equal to 40 mm) and a patent primary entry site in the thorax. The clinical value of these recommendations were completely confirmed by Kozai et al. [3] who described poor prognosis in these patients with open false lumen and large aortic diameter. Noteworthy is the analysis of Sueyoshi et al. [4] concerning type B aortic intramural hematoma (IMH). This IMH, initially described as dissection without intimal tear, is believed to be rupture of the vasa vasorum and represents a continuum of aortic dissection in which progression of an IMH with rupture into the lumen may be one mechanism for aortic dissection. In the experience of Sueyoshi et al. [4] maximum aortic diameter and maximum aortic wall thickness on initial CT image were predictive for progression of the affected aorta in patients with type B IMH. They recommended careful surveillance studies in patients with a maximum aortic diameter of 40 mm or more or a maximum

aortic wall thickness of 10 mm or more because these patients are prone for complications directly related to aortic rupture. Marui et al. [5] also addressed the question whether acute-phase factors could be identified, predictive for chronic-phase aortic enlargement, by studying chronic-phase enlargement of dissections in patients without complications during the acute phase. In 101 patients with uncomplicated type B acute dissection, univariate and multivariate factor analyses were performed similar to the results of Kato et al. [2], a patent false lumen and a maximum aortic diameter greater than 40 mm were independent predominant predictors for chronic phase enlargement (larger than 60 mm). Actuarial freedom from aortic enlargement for the patients with a maximum aortic diameter of 40 mm and a patent false lumen at 1.5 and 10 years were 43%, 33% and 22% respectively, whereas in patients with a maximum aortic diameter of less than 40 mm and a closed false lumen, the values were 97%, 94% and 84%, respectively. The authors suggest that patients with a patent false lumen and a maximum aortic diameter of larger than 40 mm should undergo surgery during the sub acute or early chronic phase, whereas patients with a closed false lumen and a maximum aortic diameter of less than 40 mm should continue to receive hypotensive treatment. The main reason for this more aggressive approach is the fact that mortality rate of surgical treatment for patients with acute type B aortic dissection without complications has recently decreased. Other arguments put forward by Marui et al. are that the surgical results for late chronic phase cases of enlarged aortas are not better than the results for acute phase surgery and the strength of the dissected aorta during the early chronic phase is satisfactory. The results of the above mentioned studies clearly justify a more selective strategy for patients with acute type B dissection. Before identifying such a selective approach it would be important to analyze the complications of conservative, medical treatment.

Complications in acute type B aortic dissection

Several retrospective studies have been performed to determine whether there are initial findings during the acute phase of type B dissection which can

TT
predict the long-term outcome of the disease. Genoni et al. [6] reviewed case records of 130 patients treated for type B aortic dissection over a tenyear period. Thirty one percent of the patients were operated on in the acute phase (less than 14 days), 29% were operated on in the chronic phase and 45% were treated medically. The most frequent indications for emergency surgery were malperfusion (34%), potential rupture (27%) and aortic rupture (19%). Indications for urgent surgery (before discharge from the first hospitalization) were malperfusion, left pleural effusion, increasing aortic diameter and persistent pain. Overall mortality rate in the acute phase was 10.8%, 5.6% of whom who had until then only medical treatment and 22% of patients operated on in this phase. In-hospital mortality rate was 9% for the medically treated patients compared with 9.1% for patients undergoing urgent surgery and 27% for those undergoing emergency surgery. Age, persistent pain and malperfusion were significant independent predictors of the need for surgery. Paraplegia, leg ischemia, pleural effusion and aortic diameter larger than 4.5 cm were significant predictors of poor survival. Patients without malperfusion, pleural effusion, rupture or shock had a significantly better event-free survival. The actuarial survival rate for high-risk patients (malperfusion, rupture, shock) was 62% at one year and 40% at 5 years. The corresponding values for low-risk patients were 94% and 84%, respectively. Based on their experience the authors advocate medical treatment in uncomplicated dissections, mainly consisting of long-term beta-blocker treatment. Emergency surgery should be performed for malperfusion such as leg ischemia and visceral/ renal ischemia, potential rupture and rupture. Surgical mortality in the acutely operated patients mainly occurs in patients with rupture and shock. It should be emphasized that surgical patients are at very high risk and can therefore not be compared with uncomplicated, medically treated patients. Carrel et al. [7] assessed early and long-term outcome of acute B aortic dissection after initial conservative treatment. Seventeen per cent of 225 patients underwent replacement of the descending aorta within the first week after hospitalization, indicated by rupture, extensive dilatation, malperfusion or pseudocoarctation with uncontrollable hypertension. All other patients underwent primary conservative treatment. Hospital mortality during and after initial conservative treatment was 17.6%, due to rupture, intestinal ischemia or cardiac failure. Hospital mortality after early surgery was 21% for the overall time period. After hospital discharge from the initial acute dissection, surgery for chronic dissection was performed in 47 patients because of expanding descending aortic aneurysm mainly (mortality 4/47). The authors recommend conservative treatment in uncomplicated type B aortic dissections and urgent surgery for rupturing disease, distal malperfusion, uncontrollable hypertension or pain. Surgery should also be considered in younger patients or Marfan Syndrome with 5 cm diameter of the aorta at initial evaluation, in patients with limited false aneurysm or retrograde dissection into the aortic arch, and those with poor medical compliance or uncontrollable proximal hypertension.

Indications for surgery

Based on the above-mentioned clinical outcome of conservative treatment and the complications in acute type B aortic dissections, indications for surgery can be distilled. Furthermore, this analysis also suggests that the timing of surgical intervention should be differentiated, as summarized in Table I. Emergency surgery is indicated in patients presenting with rupture, shock and hemodynamic instability. Malperfusion of visceral organs and kidneys, as well as spinal cord and lower limbs, requires immediate treatment. An uncomplicated type B dissection

89

Emergency surgery (open/endovascular) Rupture, shock, malperfusion, retrograde A Urgent surgery Unremitting pain, uncontrollable hypertension, rapid aortic expansion Early chronic-phase surgery Patent false lumen with initial aortic diameter > 4 cm Elective surgery Diameter enlargement during chronic phase > 6 cm (Marfan > 5 cm)

VASCULAR

EMERGENCIES

which extends retrogradely into the aortic arch necessitates emergency surgery via sternotomy. Urgent surgery is indicated in case of unremitting pain, uncontrollable hypertension and rapid aortic expansion. In the early chronic phase, patients with a patent false lumen and initial aortic diameter larger than 4 cm should be considered for surgical repair in order to prevent long-term complications. Elective surgery during the chronic phase is limited to patients who show enlargement greater than 6 centimeters. In patients with Marfan or EhlersDanlos this threshold diameter is 5 centimeters. It should be emphasized, however, that all these recommendations are based on clinical experiences and not on prospectively proven evidence. However, the complex and inhomogeneous etiology and population will never allow these kind of studies.

Selective and surgical management of acute type B aortic dissection

w
90

In an attempt to describe a selective management strategy for acute type B aortic dissection, Schor et al. [8] retrospectively analyzed the outcome of their management of 68 consecutive patients. Medical therapy consisted of aggressive antihypertensive therapy. Patients with unremitting pain or uncontrollable hypertension despite antihypertensive treatment underwent early operation. Emergency operation was also performed for rupture or significant aortic dilatation. Furthermore, if CT-scans showed increasing peri-aortic or intrapleural fluid extravasation, rapid aortic expansion or significant aortic dilatation, urgent operation was also recommended. Three patients died soon after admission, 17 underwent operation and 48 were treated conservatively. In the surgical group, mortality did not occur and major complications were encountered in 59% of patients. In the conservative group, only one patient (2%) died of rupture. However, 12 patients (25%) required aorta-related interventions. Actuarial survival at one and five years were not different between both groups (approximately 90% and 75%, respectively). When interpreting these results it is important to emphasize that the author's hospital serves as a tertiary referral center, indicating that many of the patients are self selected, which means that those who died or were unfit for referral, never reached the hospital. Importantly,

survival of both modalities improved significantly during the last decade. Rapid diagnosis, intensive monitoring and aggressive modern pharmacologic management increased survival rates of non-operative management. Taking all these issues into account, the authors support a selective management of acute type B aortic dissection. It should be emphasized, however, that the reader is left without a clear algorithm because no decision tree was proposed. Elefteriades et al. [9] supported a complicationspecific approach: medical management for uncomplicated acute descending dissections and surgical intervention for complicated dissections. Nine of 100 consecutive patients with acute dissection died before any treatment was started. Sixty of the 91 surviving patients had a benign course and 31 patients had a course complicated by rupture (8), vascular occlusion (17), early expansion or extension (12) and continued pain (4). Forty-two patients were operated (21 early, 21 late): direct aortic replacement (32), fenestration (6) and thrombo-exclusion (4). Six patients died after surgery and six developed paraplegia. The authors conclude that the vast majority of patients will survive with medical management for acute type B dissection. Surgical intervention is only indicated in complicated dissection and includes direct aortic replacement for rupture, fenestration for vascular occlusion, and aortic replacement with thrombo-exclusion as an option for patients with acute expansion. Recently, the management and outcome of acute aortic dissection (A and B) was evaluated in the experience of 12 international referral centers [10]. A total of 465 patients were treated: 62% type A and 38% type B dissections. In type A patients, surgery was performed in 72%. In type B patients, 20% underwent surgical therapy and 4% percutaneous fenestration. Medically treated type B patients had 10.7% in-hospital mortality, whereas postoperative mortality was 31.4%. This high surgical mortality rate could not be supported by Lansman et al. [11]. They described their results of surgery (within 14 days) in 34 patients with acute type B dissection. Indications for surgery were pain, rupture, malperfusion or uncontrolled hypertension. Thirty patients underwent left thoracotomy, 3 had thoraco-abdominal incisions, and one underwent thrombo-exclusion. Surgery was performed under hypothermic circulatory arrest (50%) or partial bypass (50%). There were no hospital mortalities, however, significant complications

ACUTE TYPE B AORTIC DISSECTION: SURGICAL INDICATIONS AND STRATEGY

occurred in 47% of patients including respiratory problems, renal failure, myocardial infarction and chylothorax. Neurologic deficit occurred in 6% of cases. Sasaki et al. [12] operated upon 22 patients with acute type B dissection and compared sub total prosthetic replacement of the thoraco-abdominal dissected aorta (n = 7) with partial replacement of the descending aorta at the intimal tear (n = 15). All procedures were performed with femorofemoral partial cardiopulmonary bypass. There was one early and one late death in the partial replacement group and no mortality in the other group, resulting in an overall mortality rate of 9%. Umana et al. [13] recently addressed the question what the best treatment for acute type B dissection is: medical, surgical or endovascular stentgrafting. Over a 36-year period they treated 189 patients with acute type B dissection. Sixty-seven patients received early surgical treatment. The two main determinants of death were shock (hazard ratio [HR] = 14.5) and visceral ischemia (HR = 10.9). Arch involvement, rupture, stroke, previous sternotomy and coronary or lung disease roughly doubled the hazard. Surgical mortality decreased over the decades from 57% (1963 to 1969) to 27% (1990 to 1999). In general, their results of medical and surgical therapy were comparable. With improving surgical outcome they defend the tactic of early operation, especially in younger individuals. Table II summarizes mortality rates of emergency surgery in acute type B aortic dissection.

Surgical techniques
The surgical spectrum ranges from a limited fenestration to a complete thoraco-abdominal aortic replacement. This wide range obviously indicates that the results of different series in the literature cannot be compared since detailed information about the applied techniques is most often lacking. A limited left thoracotomy with aortic fenestration carries a significant lower mortality and morbidity as compared to a thoraco-laparatomy with complete aortic replacement. The extend of the disease, aneurysmatic involvement and associated malperfusion determine the surgical strategy. Furthermore, surgeon's preference and experience finally decide which approach and technique will be applied.

ADJUNCTIVE PROCEDURES
Operations on the descending and thoraco-abdominal aorta have been notorious for their detrimental effects on organs supplied from these portions. Paraplegia, renal failure and visceral infarction are the most feared complications of extensive aortic repair. Because of the fragile aortic quality with subsequent longer cross-clamp times and the lack of collateral networks, these complications occur even more frequently in acute dissections as compared to degenerative aneurysmatic disease. Spinal cord protection can be achieved by means of retrograde aortic perfusion or profound hypothermia and circulatory arrest. The latter, however,

1 0

91

First author [ref.] Hagan [10] Lansman [11] Sasaki [12] Umana [13] *

Surgical technique

Number of patients

Mortality

Miscellaneous Aortic replacement Aortic replacement Miscellaneous

35 34

31
0 9
57 (1963 - 1969) 27(1990-1999) 14

22 67
22

Safi [14]

Aortic replacement

* Analysis of the entire experience over a 36-year period demonstrates a trend towards lower early surgical mortality risk over the decades from 57% to 27%.

VASCULAR EMERGENCIES

10
92

requires full heparinization carrying the disadvantage of bleeding complications. Especially in acutely dissected aortic tissue, hemostasis is an immense burden. Therefore, we prefer left heart bypass with cannulation of the left atrium or pulmonary vein and the femoral artery, only requiring minimal heparinization (0.5 mg/kg). The additional technique of cerebrospinal fluid drainage has extensively been evaluated and recently proven to be an important asset in the prevention of paraplegia [15]. Our surgical protocol in elective and emergency surgery includes assessment of motor evoked potentials. This technique is an accurate method for detecting cord ischemia, guiding surgical tactics to reduce neurologic deficit [16]. Unlike atherosclerotic aneurysmatic thoraco-abdominal aortas, acutely dissected aortas contain many patent intercostal and lumbar arteries. Monitoring motor evoked potentials helps to identify those arteries which are crucial for spinal cord integrity, requiring reattachment or any technique to revascularize. Distal aortic perfusion also has the advantages of visceral and renal perfusion during proximal crossclamping. However, as soon as the abdominal part of the aorta is excluded by clamps, the visceral and renal arteries are no longer perfused. In our experience the technique of selective organ perfusion prevents visceral and renal ischemia, even in patients with pre-existing renal insufficiency [17]. Four perfusion catheters are connected to the left heart bypass-system and inserted in the origins of the celiac axis, superior mesenteric and both renal arteries. In patients with acute aortic dissection this technique allows continuous organ perfusion, offering the surgeon the opportunity to perform an optimal vascular reconstruction of the visceral and renal arteries. These arteries are almost always involved in the dissecting process and originate from the false or true lumen. Furthermore, side branches might be compressed or even contain extension of the dissection, threatening distal outflow to the target organ. All these dramatic anatomic changes require optimal surgical repair, which is obviously time consuming. In addition, the fragility of the arterial and/or aortic wall often requires re-inforcement by means of teflon strips, adding even more time to perform the reconstruction. Selective organ perfusion allows these time consuming procedures. Additional advantages of distal aortic perfusion include unloading the heart, prevention of ischemiainduced metabolites and cooling/rewarming the patient.

GENERAL CONSIDERATIONS
Surgery for acute dissection has two primary objectives: replacement of the aortic segment at risk of imminent or manifest rupture and relief of distal organ malperfusion. Because rupture is frequently located in the proximal half of the descending thoracic of infrarenal aorta, replacement can be limited to these portions. It is rather uncommon that the entire thoraco-abdominal aorta has to be replaced, however, pre-existing aneurysmatic disease and extensive malperfusion might require such an extensive procedure. Several techniques have been described during the last decades, including thrombo-exclusion, tailored aortoplasty and local glue aortoplasty. The most recent and frequently reported techniques are prosthetic replacement of the thoracic aorta and aortic fenestration [18]. In recent years improvements have been made in suture material, vascular clamps, impermeable grafts, tissue adhesives and antihemorrhagic agents. Acutely dissected aortas can preferably be anastomosed with 4-0 polypropylene sutures. Cross-clamping a friable, dissected aorta should be performed with rubber-shod instruments because they are the least traumatic. The use of biologic glue has greatly contributed to the performance of blood-tight anastomoses. These glues, however, are mainly used in type A aortic dissections. In acute type B dissections the proximal anastomosis can be re-inforced but distal application is not very popular because conjoining the aortic layers might induce exclusion of important side branches.

AORTIC REPLACEMENT
The intimal tear is most frequently located in the proximal descending thoracic aorta and eventual rupture is most often situated in this segment. The standard approach, with the patient in the right lateral decubitus position and the left pelvis rotated posteriorly for access to the left femoral vessels, is via the fourth or fifth intercostal space. In dissections affecting the thoraco-abdominal aorta the left thoracotomy is performed in the sixth intercostal space and extended with a laparotomy. We only transect the anterior 5-10 cm of the diaphragm in order to limit postoperative pulmonary complications. The detailed surgical techniques are extensively described [18]. In summary, limited heparinization is administered if distal aortic perfusion is performed. Following cross-clamping, a longitudinal aortotomy

ACUTE TYPE B AORTIC DISSECTION:

visualizes the intimal tear, and false and true lumen. The dissected membrane is transversely removed at the level of the targeted proximal anastomosis and the complete aortic wall is then transected. In a fragile aorta we perform a sandwich-technique, including a circular teflon-felt inside and outside the aortic layer with subsequent end-to-end anastomosis to this re-inforced circumference (Fig. 1). The level of the distal anastomosis can vary considerably and depends on many parameters. The most limited scenario comprises attachment a few centimeters distal to the proximal anastomosis. In acute dissection this anastomosis is often performed to the true lumen, in chronic dissection to the outer layer. Attachment to the true lumen, however, carries the potential risk of malperfusion of side bran-

SURGICAL INDICATIONS AND STRATEGY

ches originating from the false lumen. We therefore prefer distal fenestration with a length of several centimeters and re-inforce the outer layer with inner- and outer circumferential teflon strips. The elephant trunk principle might also be applied. In case the total descending thoracic aorta is replaced, the dissected membrane is completely resected, leaving a longitudinal rim at the non-dissected edges. Three French (F) occlusion catheters are inserted in the back bleeding intercostal arteries and based on our motor evoked potential information we reattach the important segmental vessels. Also this anastomosis is re-inforced with an exernal teflon strip (Fig. 2). It is not uncommon that an acute type B dissection with the tear in the proximal descending aorta

93

FIG. 1 Proximal anastomosis, just cephalad to the dissection tear. In fragile aorta the anastomosis can be re-inforced with a teflon layer outside, inside or both.

VASCULAR

EMERGENCIES

does not cause ischemic problems in the thoracic segment (paraplegia) but leads to visceral or renal ischemia, necessitating either fenestration or open repair. The latter is performed through a laparotomy, extended with a limited, left anterior thoracotomy via the eighth or ninth intercostal space, allowing adequate cross-clamping of the distal thoracic aorta. Since adequate and safe reconstruction demands time, we prefer to use selective perfusion in this situation as well. Obviously, the left heart cannot be reached via this low thoracotomy and partial extracorporeal bypass is installed (femorofemoral cannulation). Following cross-clamping, aortotomy and opening of the dissected membrane, the selective catheters are inserted in the visceral and renal arteries, and intercostal and lumbar arteries are temporarily occluded with 3F balloon catheters. With a dry surgical field and continuous organ perfusion the surgeon now has all the time to analyze the damage. At this point several surgical techniques are possible depending on the extension of the dissection

into the side branches. Visceral or renal arteries which still have a conjoined connection with all layers of the aorta and their origin offer the easiest reconstruction. In case the aortic wall is strong enough, resection of the dissected membrane with subsequent closure of the aorta (teflon re-inforced) can solve the problem. This local fenestration is often performed to relief a static obstruction of the visceral arteries. It often occurs, however, that the origin of the visceral and/or renal arteries are disconnected from the inner aortic layer and compressed by the false lumen or even occluded if no re-entry exists. Also, the aortic dissection can extend into the visceral and/or renal arteries (Fig. 3). In all these circumstances we prefer vascular repair by means of separate polyester grafts (Fig. 4). The affected artery is dissected as far as the dissection reaches and transversely transected. Depending on its diameter, a 6 or 8 mm polyester graft is anastomosed in an endto-end fashion with 5-0 or 6-0 prolene sutures, with or without a small, circumferential teflon strip. This

10
94

FIG. 2 Reattachment of three intercostal arteries, temporarily occluded with 3F balloon catheters. A circular teflon strip re-inforces the anastomosis.

ACUTE TYPE B AORTIC DISSECTION:

SURGICAL INDICATIONS AND STRATEGY

FIG. 3

Acute type B dissection with extension into the visceral (A) and renal (B) arteries.

10. 95

FIG. 4 Diagram showing a dissected aorta with involvement of visceral and renal arteries. Selective organ perfusion is continuously provided, allowing optimal vascular reconstruction of the affected arteries.

peripheral reconstruction is performed under continuous perfusion. After all arteries are repaired the proximal anastomoses of the aortic tube graft to the dissected distal thoracic aorta is performed according to the above described technique, fol-

lowed by the distal anastomosis and connection of the selective grafts to the tube graft. Infrarenal fenestration is in most cases effective to relief a dynamic obstruction. The goal is to reestablish the same arterial pressures in the false and

VASCULAR

EMERGENCIES

true lumen by creating a large re-entry in the infrarenal aorta. It should be noted that primary closure of the diseased aorta is not always feasible, requiring a small interposition graft with re-inforced anastomoses. Lower limb ischemia, in the absence of proximal ischemic problems, is most often relieved by means of catheter fenestration or extra-anatomic bypass (femoro-femoral bypass, axillo-femoral bypass).

CONNECTIVE TISSUE DISEASE

Acute type B dissection in patients with EhlersDanlos or Marfan disease who require emergency surgery is a surgeon's nightmare. Acute dissection in a pre-existing fragile aorta indeed diminishes successful outcome of open aortic repair, especially if the entire thoraco-abdominal aorta has to be replaced in case of a dissecting thoraco-abdominal aneurysm. The only advantage is that these patients are most often young and have an adequate cardiopulmonary function. The same surgical principles are applicable in these patients with even more emphasis on re-inforcement and reconstruction with selective grafts.
INTRA-OPERATIVE COMPLICATIONS

10
96

It is evident that emergency surgery in acutely dissected aortas is associated with a high mortality and morbidity. It is also clear that morbidity and mortality are depending on the extent of the surgical procedure. The main dilemma for the surgeon is to perform a palliative, limited procedure to get the patient alive from the operating table or increase the short-term risk by performing a more extensive and curative procedure in order to improve long-term outcome. It is not really possible to design an algorhythm because too many factors and parameters determine the final outcome. However, it depends very much on intra-operative complications which require immediate action. The first problem which can arise after laparotomy is the appearance of (severely) ischemic bowel. In general, the vascular reconstruction is performed first, followed by judgement of the viability of the bowel. Most often bowel resection is required and it is not uncommon that the visceral ischemia is already beyond repair, leaving minimal changes to survive. Distal aortic perfusion by means of left heart bypass has the potential risk of aggravating malperfusion of the visceral, renal or spinal cord arteries.

The structure and composition of the dissection is unpredictable and it can occur that reversal of flow excludes side branches which were initially patent. Intra-operative monitoring of urine output and motor evoked potentials can assist in the assessment of acute malperfusion. Fortunately this is an uncommon complication, however we have observed this phenomenon several times. Immediate arrest of the left heart bypass is obviously necessary and the next strategy depends on the extend of the procedure. If replacement of the total abdominal or thoracoabdominal aorta is planned we reverse the aortic reconstruction and start with implantation of a bifurcation graft to both iliac of femoral arteries. The body of the graft is then clamped and a cannula is inserted in the prosthesis, allowing extracorporeal circulation and selective organ perfusion. Following, the reconstruction is performed from the distal into the proximal direction. Intra-operative bleeding problems comprise a major difficulty. Meticulous surgical techniques are a prerequisite, but major bleedings also arise from the dissected layers and coagulation disorders. Intensive assessment of the complete coagulation spectrum and the administration of coagulation products are of imminent importance. It might be necessary to pack the complete surgical field and remove the gauzes one or two days later.

Conclusion
Acute type B dissection usually exists without acute complications and is generally treated conservatively. Emergency surgery is indicated in patients presenting with rupture, shock and hemodynamic instability. Immediate surgery is required in case of malperfusion of the visceral organs and kidneys, spinal cord and lower limbs. Urgent surgery is recommended in patients with unremitting pain, uncontrollable hypertension and rapid aortic expansion. In the early chronic phase, patients with a patent false lumen and initial aortic diameter larger than four centimeters should be considered for surgical repair. Improved surgical techniques and the application of adjunctive procedures have substantially contributed to augmented surgical outcome. Future hope and ambition are focused on hybrid vascular procedures, in which minimal invasive techniques play a major role.

ACUTE TYPE B AORTIC DISSECTION: SURGICAL INDICATIONS AND STRATEGY

R E F E R E N C E S 1 Iguchi A, Tabayashi K Outcome of medically treated Stanford B aortic dissection. Jpn arc/1998; 62: 102-105. 2 Kato M, Bai H, Sato K et al. Determining surgical indications for acute type B dissection based on enlargement of aortic diameter during the chronic phase. Circulation 1995; 92 (SupplII): 107-112. 3 Kozai Y, Watanabe S, Yonezawa M et al. Long-term prognosis of acute aortic dissection with medical treatment: a survey of 263 unoperated patients./pn Qrr/2001; 65: 359-363. 4 Sueyoshi E, Imada T, Sakamoto I et al. Analysis of predictive factors for progression of type B aortic intramural hematoma with computed tomography. / Vase Swrg2002; 35: 1179-1185. 5 Marui A, Mochizuki T, Mitsui N, et al. Towards the best treatment for uncomplicated patients with type B acute aortic dissection. Circulation 1999; 100 (Suppl II): 275-280. 6 Genoni M, Paul M, Tavakoli R, et al. Predictors of complications in acute type B dissection. Eur J Cardio Thor Surg 2002; 22: 59-63. 7 Carrel T, Nguyen T, Gysi J, et al. Acute type B aortic dissection: prognosis after initial conservative treatment and predictive factors for a complicated course. Schweiz Med Wochenschr 1997; 127: 1467-1473. 8 Schor JS, Yerlioglu ME, Galla JD, et al. Selective management of acute type B aortic dissection: long-term follow-up. Ann Thor Swrgl996; 61: 1339-1341. 9 Elefteriades JA, Lovoulos CJ, Coady MA, et al. Management of descending aortic dissection. Ann Thorac Surg 1999; 67: 2002-2005. 10 Hagan PG, Nienaber CA, Isselbacher EM, et al. The International Registry of Acute Aortic Dissection (IRAD). New insights into an old disease. JAMA 2000; 283: 897-903. 11 Lansman SL, Hagl C, Fink D et al. Acute type B aortic dissection: surgical therapy. Ann Thor Surg 2002; 74: SI833-1835. 12 Sasaki S, Yasuda K, Kunihara T et al. Surgical results of Stanford type B aortic dissection. Comparison between partial and subtotal replacement of the dissected aorta. / Cardiovasc Surg 2000; 41: 227-232. 13 UmanaJP, Miller DC, Mitchell RS. What is the best treatment for patients with acute type B aortic dissections. Medical surgical or endovascular stent-grafting? Ann Thor Surg 2002; 74: S1840-1843. 14 Safi HJ, Miller CC, Reardon MJ et al. Operation for acute and chronic aortic dissection: recent outcome with regard to neurologic deficit and early death. Ann Thor Surg 1998; 66: 402-411. 15 Coselli JS, Lemaire SA, Koksoy C et al. Cerebrospinal fluid drainage reduces paraplegia after thoracoabdominal aortic aneurysm repair: results of a randomized clinical trial. / Vase Swig 2002; 35: 631-639. 16 Jacobs MJ, Elenbaas TW, Schurink GW et al. Assessment of spinal cord integrity during thoracoabdominal aortic aneurysm repair. Ann Thor Swrg2002; 74: S1864-1866. 17 Jacobs MJ, Eijsman L, Meylaerts SA et al. Reduced renal failure following thoracoabdominal aortic aneurysm repair by selective perfusion. Eur] Cardiothorac Surg 1998; 14: 201-205. 18 Borst HG, Heinemann MK, Stone CD. Surgical treatment of aortic dissection. Churchill Livingstone Inc., 1996, 357 p.

10
97

This page intentionally left blank

11
ENDOVASCULAR TREATMENT OF AORTIC TYPE B DISSECTION
RACHEL BELL, PETER TAYLOR
Acute aortic dissection is the most common aortic emergency,, with an incidence double that of ruptured abdominal aortic aneurysms [I]. It affects 10 to 20 people per million population per annum [2]. Without treatment, 36% to 72% of patients will die within 48 hours of diagnosis [3]. The Stanford classification is based on the presence or absence of involvement of the ascending aorta. Type A dissection involves the ascending aorta. Type B dissection does not involve the ascending aorta and the primary intimal tear is usually just distal to the origin of the left subclavian artery in the descending thoracic aorta. The distinction between acute and chronic dissection is arbitrarily based on time since the onset of symptoms: less than 14 days for acute dissection and greater than 14 days for chronic dissection.

1 1

99

Treatment
The preferred treatment for acute type B aortic dissection is medical management with antihypertensive drugs and B-blockers. This is based on reports of survival rates of only 50% for open surgery compared with 80% for medical treatment alone [4]. Emergency surgery or endovascular intervention is reserved for ongoing pain, refractory hypertension, localized false aneurysm, end-organ ischemia, and rupture [5,6]. These complications occur in 30% to 40% of patients. Recently, the International Registry for Acute Aortic Dissection (IRAD)

reported that mortality in patients with type B dissection treated medically was 11% compared with 31% in those who required surgery [7]. Pain usually resolves with adequate treatment for hypertension. The majority of patients have some fluid in the left chest that is usually a transudate secondary to the dissection, which must be differentiated from rupture by the appearance of extravasated contrast medium. Renal and mesenteric ischemia carry a surgical mortality of 50% and 88%, respectively [5]. In the long term, those treated medically have a 30% to 40% risk of aneurysm formation, which requires treatment if they are symptomatic or if the

VASCULAR

EMERGENCIES

diameter is more than 5.5 to 6 cm in diameter [8,9]. Aneurysms related to previous dissection have a high risk of rupture and appear to rupture at a smaller diameter than degenerative aneurysms [8]. Currendy there is no level I evidence regarding the diameter at which aneurysms related to previous dissection should be repaired. Endovascular treatment of complicated type B dissection includes covering the primary entry tear with a stent graft, percutaneous fenestration of the intimal flap, and stenting of obstructed aortic side branches. Endovascular treatment has also been recommended for penetrating ulcers and intramural hematomas of the descending aorta, which have been identified as precursors of dissection [7]. Technical advances have had an enormous impact on diagnosis and treatment of aortic dissection, and endoluminal repair is now considered the first line treatment for complicated type B dissection.

Endovascular treatment
The endovascular techniques used to treat type B dissection are: 1 - placing a stent graft over the primary entry tear to obliterate the flow into the false lumen, 2 - percutaneous fenestration of the intimal flap, 3 - stenting the aortic side branches. Factors that influence the type and timing of intervention are the type of dissection, the mechanism of obstruction, and the aortic branches involved. The endovascular techniques are complimentary and a combination of procedures may be required depending on the clinical status of the patient and the aortic branches affected. The primary endovascular technique is placement of a stent graft across the primary entry tear,

Mechanisms involved in aortic dissection

n
100

Aortic dissection is thought to occur when there is a tear in the aortic intima, allowing blood to track between the intima and the media. The association between intramural hematoma and aortic dissection suggests that a dissection may also start as a bleed from the vasa vasorum into the aortic media. A dissection can extend a variable distance, in a retrograde or antegrade fashion. The blood-filled space between the intima and the media becomes the false lumen. The dissection is commonly spiral in shape and often extends into the abdominal aorta. The right renal artery and the visceral arteries are usually perfused from the true lumen and the left renal artery from the false lumen (Fig. 1). This, however, is not always the case and depends on the relationship of the intimal flap to the origins of these vessels (Fig. 2). End-organ ischemia can be caused by static obstruction resulting from extension of the dissection into aortic side branches, or by dynamic obstruction due to distension of the false lumen causing the intimal flap to bow into the true lumen resulting in true lumen collapse (Fig. 3). Patency of the false lumen in the thorax correlates closely with the late complications of aortic dissection such as aneurysmal dilatation and death from rupture [10]. In chronic aortic dissection, a high flow rate in the false lumen is associated with an increased risk of rupture [11].

FIG. 1 Arteriogram showing "floating viscera" with contrast filling the celiac axis seemingly unconnected to the aorta.

ENDOVASCULAR TREATMENT
as this can relieve both static and dynamic obstruction of aortic branches. Experiments have shown that decreasing the false lumen inflow by placing a stent across the primary entry tear is the most effective treatment for true lumen collapse [12,13]. Fenestration and aortic branch stenting are usually limited to those patients who have residual ischemia after redirection of inflow into the true lumen. However, fenestration can be used as a first line treatment for acute limb ischemia, which has become the major indication for this technique.

OF AORTIC TYPE B DISSECTION

surization of the false lumen and subsequent rupture. It may be necessary to design a stent graft specifically for the treatment of aortic dissection. Some authorities have recommended stenting all patients presenting with acute type B dissection, to reduce the incidence of late aneurysm formation. However, this should be assessed against best medical treatment in a randomized controlled trial. Computerized tomography (CT) and calibration angiography are used to determine the dimensions of the stent graft. The advent of multislice CT and magnetic resonance angiography has supplanted the need for conventional angiography. However, angiography may still be used to delineate the primary tear and to show which arterial branches come off the true and false lumens. In cases of dissection it is difficult to know the original diameter of the aorta, and the width of the non-dissected aorta proximal to the entry tear is used to size the diameter of the stent. It is recommended that the devices be only oversized by 10%, as there is a risk of rupture with devices that are too big associated with perforation of the intimal flap. Intravascular ultrasound (IVUS) can be used to identify the location of the primary entry tear and to demonstrate the relation of the true and false lumens and the aortic branches. It is essential that

Imaging and devices

The commercially manufactured devices currently available in Europe are the Talent thoracic stent (Medtronic AVE), which consists of a nitinol frame covered with polyethylene, and Endofit (Endomed, Inc), which is a nitinol frame with a polytetrafluoroethylene (PTFE) cover. The Thoracic Excluder (W.L. Gore & Associates), manufactured from nitinol and PTFE, was recently withdrawn voluntarily from the market for redesigning following reports of fractures in the nitinol frame in 10% of patients in the United States. There are problems with the rigid devices in this clinical setting, as they can erode through the intimal flap, causing pres-

101

FIG. 2 CT scan showing an acute aortic dissection with the left renal artery originating from the true lumen.

FIG. 3 CT scan showing "true lumen collapse" associated with partial thrombosis of the false lumen.

VASCULAR

EMERGENCIES

all patients have radiological follow-up with CT scan and plain chest X-ray at 3 months, 6 months, 12 months, and annually thereafter. The ideal outcome is complete thrombosis of the false lumen. In patients with partial thrombosis of the false lumen, there is still a risk of aneurysmal dilatation. The long-term durability of stent grafts is unknown but there is the potential for development of stent fractures, endoleaks, and migration.

PERCUTANEOUS FENESTRATION
This technique can be used in cases of true lumen collapse in order to direct blood flow into the true lumen, decrease the pressure in the false lumen, and allow reperfusion of aortic branches that have been isolated by the dissection. Arterial access is from the femoral artery and ideally biluminal

Endovascular techniques
COVERED STENTS
The procedures can be carried out in theater or in the angiography suite. High-quality imaging is essential to ensure accurate graft placement. The operation can be performed under general, regional, or local anesthesia. Regional or local blocks are increasingly performed, as they allow assessment of distal neurologic function throughout the procedure. Devices are usually inserted via the common femoral artery and 5 000 units of unfractionated heparin are given intravenously prior to insertion of the introducer sheath. The stent graft is introduced over a stiff guide wire into the true lumen and deployed to cover the primary entry tear. The proximal neck length must be at least two centimeters to ensure adequate fixation (Figs. 4A and 4B). It is possible to cover the left subclavian artery without pre-emptive revascularization. Occasionally the proximal neck requires balloon dilation after deployment to exclude a type I endoleak. However, it is not advisable to perform balloon dilation in the area of dissection because of the risk of rupture and of tearing the fragile intimal flap. Deployment of stent grafts distally within the dissection that do not cover the primary tear may cause retrograde dissection and turn a type B dissection into a type A with disastrous consequences. Early devices required systemic hypotension for deployment to prevent migration, but with the newer stent grafts this is not necessary. Following deployment, further angiography is performed to confirm the position of the stent graft and exclusion of the primary entry tear. The arteriotomy is repaired and the wound is closed with dissolvable sutures. The patients are observed in recovery for several hours after which they return to the surgical ward. A CT scan is performed prior to discharge if there is any concern about the exclusion of the false lumen.

n
102

FIG. 4 A - Arch aortography showing the false lumen filling from a primary entry tear just distal to the left subclavian artery. B - After deployment of a stent graft across the primary entry tear resulting in exclusion of the false lumen.

ENDOVASCULAR TREATMENT OF AORTIC TYPE B DISSECTION

manometry is performed before and after intervention. IVUS can be used to help direct balloon fenestration of the intimal flap. Single or multiple tears are made in the intimal flap using a biopsy needle, a guide wire is positioned at the puncture hole and a 12 to 14 mm angioplasty balloon is used to complete the fenestration. Alternatively, fenestrations can be created by dilating re-entry tears identified using IVUS. Fenestration should be performed in the abdominal aorta as more proximal fenestrations fail to provide adequate redirection of blood flow in the distal aorta. Sometimes fenestration alone does not provide adequate reperfusion and adjunctive procedures may be required [14]. In acute dissection the intimal flap is fragile, tears easily and the fenestrations remain patent. For chronic dissection the intimal flap is more rigid and deployment of a bridging stent across the fenestration has been used to ensure adequate distal flow. In cases of true lumen collapse in the iliac arteries, placing guide wires in the true and false lumens and advancing a single introducer sheath over both wires causes a longitudinal tear in the intimal flap leading to reperfusion of the ischemic limb. However, the technique is difficult and not widely practiced. Fenestration is contraindicated in the presence of partial or complete thrombosis of either lumen because of the associated risk of distal embolization. plegia and this has been born out by the clinical results. Care must be taken with patients with Marfan syndrome, as there are often multiple tears in the intima and just covering the primary tear may not be effective. In these patients, the need to exclude the false lumen must be weighed against the increased risk of paraplegia. The Stanford group was the first to report their experiences with custom-made grafts (Table I). They reported that revascularization of blocked aortic branches occurred in 76% of patients. Longterm follow-up has shown 79% (15/19) of patients having complete thrombosis of the false lumen [16]. Nienaber et al. reported zero mortality and morbidity in a small series of 12 patients treated by stenting across the primary tear [17]. They have since reported excellent results in an additional 82 patients (Table I) [18]. However, it is difficult to interpret this data, as it does not specify the distribution of acute and chronic cases or the number with rupture or end-organ ischemia. In comparison with open surgery, this endoluminal technique has low morbidity and mortality rates. Importantly, the risk of paraplegia appears to be significantly lower than for conventional surgery. Endoluminal repair avoids thoracotomy, aortic crossclamping, and cardiac bypass. As a consequence, blood loss is minimal and admission to intensive care is rarely required. Endoluminal exclusion promotes thrombosis of the false lumen, which has been shown to decrease the risk of aneurysm formation [10] (Figs. 5A and 5B).

11
103

UNCOVERED STENTS
Residual static obstruction of aortic branches after stenting across the primary entry tear can occur. In this circumstance, an uncovered stent can be placed in the true lumen of the aortic branch to relieve the obstruction. Stents can also be placed from the false lumen of the aorta to restore flow to the true lumen of the branch vessel. Deployment of an infrarenal stent graft in the true lumen of the abdominal aorta can be used to treat limb ischemia.

PERCUTANEOUS FENESTRATION AND AORTIC BRANCH STENTING

These adjunctive techniques can be technically complex and time consuming, and are only required in a small number of patients. Slonim et al. reported successful revascularization in 93% (37/40) of patients treated with a combination of fenestration and stenting. However, the 30-day mortality rate for the group was 25% (10/40). The causes of death were multiorgan failure, rupture of the false lumen, and right heart failure (Table II) [14]. Williams et al. reported similar results with successful revascularization in 88% (21/24) and a mortality rate of 25% (6/24) [21]. The severity and duration of visceral ischemia prior to intervention have an important impact on outcomes and patient survival. One of the disadvantages of endovascular repair is that the degree of mesenteric ischemia prior to revascularization cannot be visualized,

Results
COVERED STENTS
Kato et al. showed experimentally that the false lumen could be excluded by simply covering the primary entry tear with a covered stent graft. Reducing inflow into the false lumen promotes complete thrombosis [15]. Covering only a short length of descending thoracic aorta reduces the risk of para-

VASCULAR

EMERGENCIES

, , 1st author r ,, ret 1 r

,, Year

, . , Number Type technical Jr , , of of success ,q} patients dissection

( /O)

, , . , Revasculathrombosis of . . , ,, . . . J ,, , nzation of Complications , , ,,T , , Jfalse lumen ,q. (Tranches (Number ojf cases)
( /OJ : (ff ,

, , 30-da} /. mortality /&\

( /C)

Follow-up (months)
4-28

Dake [16] 1999

19

A 4 B 15

100

C* 15/19 (79) P** 4/19 (21)

76

Colonic infarction(l) Limb ischemia (1) Pneumonia (1) Renal failure(l)

3/19 (16)

Nienaber [17] Rehders [18]

1999 2001

12 94*

A 0 B 12

100

C 12/15 (80) P 3/15 (20)

12 12

100

98

Stroke (1) TIA(l) Bleeding (3) Stroke (2) Paraplegia(l) Type 1 endoleak (1) Intimal tears Saccular aneurysm

Lonn[19] 2002

16

A 2 B 14

100

C 15/16(94) P 1/16 (6)

12

Lee [20]

2002

41

87.8

1-91

11

104

Complete thrombosis * Partial thrombosis Contains the 12 patients from 1999 series TIA: transient ischemic attack

FIG. 5 A - CT scan showing an acute type B dissection associated with a small, contained leak into the left hemithorax. B - One-year follow-up CT scan showing thrombosis and shrinkage of the false lumen.

ENDOVASCULAR TREATMENT

OF AORTIC TYPE B DISSECTION

1st author [ref.]

Year

Number of patients 1997

Type of dissection

Fenestration

Successful Complications revascularisation (Number of cases) (%) 21/24 Sepsis (1)

30-day mortality (%) Total: 6 (25) 3 multi-organ failure 1 ruptured false lumen 2 complications of surgery

Williams [21]

24

13 Type A 8 TypeB 3 Atypical

6 fenestrations 2 fenestrations + PTA 12 fenestrations + stents 4 stents alone

Slonim [14]

1999

40

10 Type A 30 TypeB

2 fenestrations 0 fenestrations + PTA 14 fenestrations + stents 24 stents alone

37/40 (93)

Thrombosed Total: 10 (25) renal stent 5 multi-organ(1) failure Distal 2 ruptured false embolization lumen (1) 1 right heart failure 1 complication of surgery
1 NA

Gaxotte [22]

2002

41

39/41 (95)

Total: 7 (17)

11
105

NA: not available PTA: percutaneus transluminal angioplasty

which can have catastrophic consequences. Unfortunately, exploratory surgery in these patients is often delayed and, as a result, multiorgan failure and death are common. Cases of mesenteric and renal ischemia require prompt diagnosis and rapid intervention to prevent complications and death.

Conclusion
The mid-term results for endoluminal treatment of type B aortic dissection are encouraging but this may be a highly selected group. These techniques

are relatively new compared with open surgery and are still under evaluation. It appears that stent graft placement in the acute situation can prevent late aneurysm formation by facilitating complete thrombosis of the thoracic aortic false lumen. However, there are still problems with rigid stent grafts, which can erode through the intimal flap and cause pressurization of the false lumen leading to rupture. There is a need for a stent graft that is specifically designed for the treatment of aortic dissection that is blunt with a tapered end. Despite these problems, endoluminal repair has a role in the treatment of acute and chronic type B dissection, penetrating ulcers, and intramural hematomas.

VASCULAR

EMERGENCIES

Future trials
Despite increasing experience in the endoluminal treatment of acute and chronic aortic dissection, there are still areas that require further investigation. Ideally this research should take place under the auspices of randomized controlled trials. 4> The role of endoluminal repair in the treatment of uncomplicated type B dissection versus best medical treatment.

The role of endoluminal repair in the treatment of complicated type B dissection versus surgical treatment. The indications for further procedures to ensure complete thrombosis of the thoracic and abdominal false lumens, such as the use of uncovered stents distally. The precise diameter at which asymptomatic aneurysms related to chronic dissection should be repaired. The long-term durability of the stent grafts.

R E F E R E N C E S

- -g j j .-* * 106

1 Kouchoukos NT, Dougenis D. Surgery of the thoracic aorta. N EngJMed 1997; 336:1876-1888. 2 Pate JW, Richardson RL, Eastridge CE. Acute aortic dissections. Ann Surg 1976; 42: 395 -404. 3 Anagnostopoulos CE, Prabhakar MJ, Kittle CF. Aortic dissections and dissecting aneurysms. Am J Cardiol 1972; 30: 263-273. 4 Wheat MW Jr. Current status of medical therapy of acute dissecting aneurysms of the aorta. World J Surg 1980; 4: 563-569. 5 Cambria RP, Brewster DC, Gertler J et al. Vascular complications associated with spontaneous aortic dissection./tocSwrgl988; 7: 199-209. 6 Fann JI, Sarris GE, Mitchell RS et al. Treatment of patients with aortic dissection presenting with peripheral vascular complications. Ann Swrgl990; 212: 705-713. 7 Ha an PG ' Nienaber CA, Isselbacher EM et al. The International Registry of Acute Aortic Dissection (IRAD): new insights into an old disease. JAMA 2000; 283: 897-903. 8 Griepp RB, Ergin MA, Galla JD et al. Natural history of descending thoracic and thoracoabdominal aneurysms. Ann Thorac Surg 1999; 67:1927-1930, discussion 1953-1958. 9 Juvonen T, Ergin MA, Galla JD et al. Risk factors for rupture of chronic type B dissections./ Thorac Cardiovasc Surg 1999; 117: 776-786. 10 Bernard Y, Zimmermann H, Chocron S et al. False lumen patency as a predictor of late outcome in aortic dissection. Am] Cardiol 2001; 87:1378 -1382. 11 Erbel R, Oelert H, Meyer J et al. Effect of medical and surgical therapy on aortic dissection evaluated by transoesophageal echocardiography. Implications for prognosis and therapy. The European Cooperative Study Group on Echocardiography. Circulation 1993; 87: 1604-1615. 12 Chung JW, Elkins C, Sakai T et al. True-lumen collapse in aortic dissection: part I. Evaluation of causative factors in phantoms with pulsatile flow. Radiology 2000; 214: 87-98.

13 Chung JW, Elkins C, Sakai T et al. True-lumen collapse in aortic dissection: part II. Evaluation of treatment methods in phantoms with pulsatile flow. Radiology 2000; 214: 99-106. 14 Slonim SM, Miller DC, Mitchell RS et al. Percutaneous balloon fenestration and stenting for life-threatening ischemic complications in patients with acute aortic dissection. / Thorac Cardiovasc Surg 1999; 117: 1118-1126. 15 Kato N, Hirano T, Takeda K et al. Treatment of aortic dissections with percutaneous intravascular endoprosthesis: comparison of covered and bare stents./ Vase Men Radiol 1994; 5: 805-812. 16 Dake MD, Kato N, Mitchell RS et al. Endovascular stent-graft placement for the treatment of acute aortic dissection. NEngJ Med 1999; 340: 1546-1552. 17 Nienaber CA, Fattori R, Lund G et al. Nonsurgical reconstruction of thoracic aortic dissection by stent-graft placement. NEng JMed 1999; 340:1539-1545. 18 Rehders TC, Nienaber CA. Complications of stent graft placement in the thoracic aorta. In: Branchereau A, Jacobs M (eds). Complications in vascular and endovascular surgery - Part I. Armonk, Futura Publishing Co 2001: pp 185-192. 19 Lonn L, Delle M, Lepore V et al. Endograft therapy of the thoracic aorta in aortic dissections. Cardiovasc Mervent Radiol 2002;25(supp2):S158. 20 Lee DY, Choi DH, Shim WH et al. Elective endovascular treatment of aortic dissections with stent grafts. Cardiovasc Mervent Radiol 2002; 25 (supp 2): S158. 21 Williams DM, Lee DY, Hamilton BH et al. The dissected aorta: percutaneous treatment of ischemic complications - principles and results. / Vase Men Radiol 1997; 8: 605 - 625. 22 Gaxotte VD, Haulon S, Willoteaux S et al. Endovascular treatment in complications of aortic dissection: retrospective study on 52 patients. Cardiovasc Mervent Radiol 2002; 25 (supp2):S157.

12
TRAUMATIC RUPTURE OF THE THORACIC AORTA
ROBERTO CHIESA, RENATA CASTELLANO, CARLA LUCCI MARCELO R. LIBERATO DE MOURA, FEDERICO PAPPALARDO GERMANO MELISSANO, EFREM CIVILINI, YAMUME TSHOMBA
Aortic injury may be secondary to several mechanisms; the more frequent are penetrating, iatrogenic, and blunt trauma (Table I) [1-4]. In a recent large autopsy study, all penetrating trauma victims died before reaching the hospital, whereas 5.5% of the blunt trauma victims were admitted to the hospital alive [5], explaining why the surgeon will generally face blunt thoracic aortic trauma. Thus, in this chapter mainly blunt lesions will be discussed. Blunt aortic injury occurs as the result of motor vehicle accidents, falls, and crush injuries, and it may account for 10% to 15% of deaths caused by motor vehicle accidents [6]. Thoracic aorta injuries have been implicated as the cause of death in as many as 17% of all motor vehicle crash fatalities. It is estimated that between 70% and 90% of patients sustaining this injury die at the scene from free aortic rupture. The 10% to 20% of patients with thoracic aorta injuries who survive long enough to reach the hospital have a dismal prognosis: it has been estimated that approximately 30% of them will succumb within 6 hours, 40% to 50% within 24 hours, and 90 % within 4 months, unless expedient and proper diagnostic and therapeutic measures are undertaken [7]. Pate et al. found that associated injuries were present in more than 90 % of patients with aortic transection, and 24% of them required a major operation before aortic repair [8]. The major complications of survivors diagnosed and treated have been related to associated injuries and spinal cord injury related to the surgical management. The characteristics of thoracic aorta injuries have made it difficult for single centers to accumulate large series of patients. Most studies have been performed on relatively small populations [9-11], done over long time periods [12,13], or concentrated on one particular method of treatment [14-17]. Thus, many questions and controversies persist with regard to the optimal methods of diagnosis and treatment of this pathology.

VASCULAR

EMERGENCIES

Penetrating
Gunshot wounds Stab wounds Other: ice picks, shrapnel, spearing Motor-related injuries: motor vehicular, motorcycle and auto-pedestrian accidents Fall Crush injury Work/recreational-related: industrial accidents, bull goring, equestrian accidents

latrogenic
Central line insertion Cardiac catheterization Pacemaker insertion Intra-aortic balloon counterpulsation Expandable stents (tracheal, bronchial or vascular)

Pathogenesis
MECHANISMS OF TRAUMA
More than 90% of thoracic great vessel injuries are caused by penetrating external or iatrogenic trauma [1]. Penetrating injuries include complete or partial transections and arteriovenous fistulae. Because the vessels involved are generally large, the mechanism of muscular retraction usually fails to control hemorrhage, which can result in rapid blood loss [18]. Most penetrating lesions are located at the ascending aorta and arch branches, with a very poor prognosis, which renders them rare in the clinical setting [5]. Although penetrating mechanisms predominate, the number of patients with aortic disruption due to blunt trauma has continued to increase [19]. Aortic rupture in blunt trauma results most commonly from sudden high-speed deceleration or, less frequently from chest compression. Other mechanisms involved in blunt aortic injuries might include compression of the vessels between bony structures such as sternum and spine, and profound intraluminal hypertension during a severe traumatic event [20-22]. There is much debate as to the type of motor vehicle collision that causes aortic disruption. Although most authors agree that a direct frontal deceleration is the most common mechanism that can cause traumatic injury to the aorta (Fig. 1), recent studies suggest that as many as 40% of the thoracic aortic injuries may be a consequence of side impact collisions [23-26].

LESION TOPOGRAPHY
The typical point of injury is located in the most proximal descending aorta, at the site of insertion of the ligamentum arteriosum, just distal to the origin of the left subclavian artery. At this point, a highly mobile region of the aorta is placed between two fixed segments: the aortic arch is anchored with the neck vessels including the left subclavian artery, and the descending thoracic aorta is fixed to the thorax by the ligamentum arteriosum and by the

12
108

FIG. 1 Common mechanism of aortic injury: a high-speed frontal deceleration collision.

TRAUMATIC RUPTURE OF THE THORACIC AORTA

intercostal vessels. The mobile part of the aorta, the distal part of the arch and the most proximal part of the descending are only loosely fixed to the chest wall by the parietal pleura. With abrupt thorax deceleration, the fixed portions decelerate with the chest, but the loosely fixed parts continue to move forward until they finally decelerate: aortic rupture occurs at the interface between these two parts [27]. Figure 2 provides some information regarding the site of thoracic aortic lesions in 144 patients studied in a trauma registry [7]. To clarify the terminology, total rupture or transection applies to the full thickness separation of the aorta, which can be partial or circumferential (Fig. 3). On the other hand, injury or partial rupture refers to lesions without a complete wall disruption. True traumatic dissection, which involves a longitudinal separation of the media along the length of the aorta, has rarely been reported [4].

Diagnosis
Each clinical situation, depending on the patient's general status, must be individually judged searching for the best diagnostic tools.

MEDICAL HISTORY
Although an accurate history and physical examination are necessary, clinical signs and symptoms
FIG. 2 Principal location of thoracic aortic lesions.

109

FIG. 3 Types of aortic rupture: (A) partial, (B) circumferential, and (C) double.

VASCULAR

EMERGENCIES

are often lacking in these patients. A high index of suspicion is the cornerstone for timely diagnosis and repair of thoracic aorta injuries or rupture. Aortic injury must be suspected in any patient with high-energy trauma to the chest. The violence and magnitude of the injury forces must be inferred from historical and clinical findings. Patients with a history of motor vehicle crash, with victims in the same accident, or those having fallen from a great height, must be suspected of having an aortic trauma.

tive predictive value, the supine chest X-ray is a valuable screening tool for mediastinal hemorrhage, but is worth little as far as definitive diagnosis is concerned [30]. Other important signs are irregularity or blurring of the aortic knob contour, presence of a left apical cap and a tracheal displacement [31]. It is important to obtain serial follow-up chest films in patients with suspected mechanism of injury, because radiographic abnormalities may be absent on initial evaluation. Spiral computed tomography (SCT). SCT is currently considered not only as a screening method to select patients for thoracic aortography, but also as a definitive diagnostic procedure recognizing aortic injuries and rupture (Fig. 5). Compared to aortography, it is less invasive, faster to obtain, more readily available and less expensive [32]. Some direct signs of aortic lesions such as intimal flap, intramural hematoma or dissection, aortic wall or contour irregularity, pseudoaneurysm, and pseudo-coarctation are seen on SCT [33]. The presence of a hemomediastinum is well characterized by a SCT and represents an important indirect sign of aortic trauma. The use of two- and threedimensional reconstructions on SCT aortography creates images very close to those obtained by con-

CLINICAL FINDINGS
Aortic trauma is often obscured by the presence of other serious injuries [27]. Nevertheless, if the diagnosis is made, it can overshadow the presence of other severe and more lethal lesions. Head trauma, massive abdominal hemorrhage, extensive burns and severe respiratory insufficiency can delay the diagnosis and surgical repair of traumatic aortic injuries. The radiological finding of an expanding mediastinal area, increasing hemothorax and/ or prolonged anuria conversely confers priority to the treatment of a thoracic aortic lesion. The clinical signs can be as unspecific as a deep or interscapular thoracic pain, which can happen with aortic adventitia distension. However, patients can rarely present more specific syndromes like an aortic pseudo-coarctation, coursing with upper limb hypertension and reduced or absent femoral pulses [28]. Intimal flaps or dissections can cause ischemic complications, and finally if a total rupture is present, bleeding to mediastinum and pleura will cause hemodynamic instability, worsening pain and death in the majority of cases.

110

IMAGING STUDIES
The diagnostic practice depends on the patient's conditions on hospital admission. The kind of aortic and associated lesions influence the outcome and diagnostic studies, according to the patient's hemodynamic stability and general status. Chest X-ray. The first step is to obtain a front chest X-ray, as initial radiological evaluation, in every high-speed trauma patient with suspected blunt traumatic aortic injury. There are several radiological findings in cases of traumatic aortic rupture. Many studies have shown that the widened mediastinum on chest X-ray is associated with more than 90% of thoracic aortic injuries (Fig. 4) [29] . With a 90% sensitivity, a 25% specificity and a 95% nega-

FIG. 4 Chest X-ray showing a widening mediastinum and left lung contusion.

TRAUMATIC RUPTURE OF THE THORACIC AORTA

FIG. 5 SCT demonstrates: (A) descending aortic rupture at the isthmus level causing an intimal flap and left hemothorax; (B) thoracic aortic injury at the descending part with vessel wall irregularity, left hemothorax, and lung contusion.

ventional aortography [34], providing the surgeon with important anatomical information. The negative predictive value of 99.9% for aortic lesions has been considered with a normal mediastinum and a regular aorta seen on SCT [35]. A total body study can also provide important information regarding associated lesions on stable patients with craniofacial, thoracic and abdominal injuries. Angiography. Aortography is traditionally considered the gold standard imaging study to detect aortic injury, to define its location and extent (Fig. 6). It also provides important information about vascular anomalies and other factors that influence operative strategy. The demonstration of an irregular or discontinued contour of the aortic lumen represents the aortographic diagnosis of blunt traumatic aortic injuries. Intimal flap, aortic dissection, post-traumatic coarctation or luminal outpouching relating to a pseudoaneurysm are current aortographic patterns caused by blunt traumatic aortic lesions. Thoracic aortography can detect blunt traumatic aortic injuries with a sensitivity and specificity of 95% to 99% and 94% to 100%, respectively [36-38]. False-negative examinations relate to incomplete series, inadequate injections or projections. False positives often relate to prominent ductus diverticulum or from ulcerated atheromas.

Transesophageal echocardiography (TEE). TEE combined with color doppler flow mapping can accurately demonstrate blunt traumatic isthmic aortic lesions [39,40] (Fig. 7). This diagnostic method can be rapidly and simultaneously realized with other procedures like mechanical ventilation or laparotomy. It is not indicated in patients with a difficult airway or suspected spinal cord injury. Moreover, atheromatous disease and pneumomediastinum may interfere with the accuracy of TEE [36]. TEE may be very useful in unstable patients, as in such cases it is not possible to perform other time-consuming studies. Magnetic resonance angiography (MRA). MRA results are comparable to those of SCT and conventional aortography. Nevertheless MR has several limitations such as the time spent completing the exam and the inaccessibility of the patient during examination, which excludes its routine use in urgent cases [30]. Intravascular ultrasound. Intravascular ultrasound imaging has been described as a complementary method clarifying slight focal aortic abnormalities not visible by thoracic aortography. Its use is still limited and it is mainly associated with endovascular procedures [41]. Diagnostic practice for traumatic aortic injury is based on mechanism of

111

VASCULAR

EMERGENCIES

11
112
FIG. 6 Distal thoracic aortic pseudoaneurysm evidenced by luminal outpouchins durins aortosraphy in (A) frontal and (B) lateral views.

FIG. 7 TEE showins a postisthmic aortic pseudoaneurysm with a clear larse neck and hish-velocity flow inside (TL: true lumen; PA: pseudoaneurysm).

TRAUMATIC RUPTURE OF THE THORACIC AORTA injury, chest X-ray and SCT scan or aortography; all of these modalities have limitations and thus must be considered in concert. the aorta is more difficult and tedious. Current indications for delaying the aortic repair in the hemodynamically stable patient include trauma to the central nervous system with coma, respiratory failure from lung contusion, body surface burns, blunt cardiac injury, tears of solid organs that will undergo nonoperative management, retroperitoneal hematoma, contaminated wounds, age 50 years or older and medical comorbidities [4] (Fig. 8). SURGICAL APPROACH (Fig. 9) As with any operative procedure, patient positioning and skin incision are important, as adequate exposure is mandatory for proximal and distal control of the great vessels. General exposure and skin preparation should include the anterior neck, thorax, abdomen and a lower extremity. When a subclavian injury is suspected, the ipsilateral arm should be prepared and draped in a fashion that maintains free mobility of the shoulder. The posterolateral thoracotomy provides excellent exposure to virtually all portions of the hemithorax [44]. The incision extends from behind the medial border of the scapula, below its tip, and then forward to the anterior axillary line. The chest can be entered through the fourth to the seventh intercostal space, usually in the bed of the fifth rib. This incision generally offers an adequate access for replacing the upstream two thirds of the descending aorta from the distal aortic arch down to the eighth intercostal space. If required, the fifth or sixth ribs may be removed. Particular care must be exercised when positioning hypotensive patients for this incision, because the right decubitus position interferes with venous return and can aggravate the condition. This approach may allow a transverse sternotomy in case of difficult control of the proximal aorta. Associated phreno-celiotomy may be performed to treat previously undiagnosed abdominal vascular and visceral injuries. For hypotensive patients with undiagnosed injury, the mainstay of thoracic trauma surgery is the left anterior thoracotomy through the fourth intercostal space, with the patient in the supine position. This is also the incision of choice for rapid access to the mediastinum for open cardiac massage. The incision should start 2 cm lateral to the sternum to avoid injury to the internal mammary artery. It then follows a gentle curve laterally to the axilla. When additional exposure is needed, an anterior thoracotomy may be extended across the sternum in the midline, as previously described; it is also possible

Treatment
TIMING The treatment of patients who reach the hospital alive remains controversial. The original strategy of immediate aortic repair has been challenged by recent reports of successful delayed repair [42,43] Although traumatic rupture of the thoracic aorta has traditionally been considered a surgical emergency, there exists a patient population for whom nonoperative management may be appropriate. Indications for urgent operative repair include hemodynamic instability, increasing hemorrhage from the chest tubes and radiographic evidence of an expanding hematoma. Delayed repair may be considered in selected hemodynamically stable patients, who may not necessarily benefit from immediate repair, including patients with severe head injuries, risk factors for infections (major burns, sepsis, heavily contaminated wounds) and severe multisystem trauma with poor physiologic reserve. A basis for nonoperative management is that maintaining the systolic blood pressure below 120 mmHg or mean arterial pressure less than 80 mmHg significantly reduces the risk of rupture. The risk of fatal rupture of the peri-aortic hematoma in hemodynamically stable patients has been estimated to be 4.5% within the first 72 hours, but it does not increase if conservative treatment is further continued [43]. This might be related to the maintenance of aortic adventitia continuity in patients who survive. In those, hemorrhage is contained by the surrounding mediastinal structures. In fact, some patients may develop chronic pseudoaneurysms. However, given that progression and rupture were documented also in patients with small injuries, the reason for delaying operation should not be based solely on the size of the lesion, but also on high-risk physiologic criteria. Although a small risk of free rupture still remains, data support the concept that nonoperative management of aortic lesions can be utilized safely in selected cases. In some cases of smaller aortic tears, the lesion may heal on its own. The downside of the delayed repair is that, due to the extensive scarring at the site and around the injury, the surgical dissection of

113

VASCULAR

EMERGENCIES

FIG. 8

Schematic diasnosis and manasement flow chart of traumatic aortic lesions.

114
to extend the incision posteriorly in order to have a better control of proximal aorta. A median sternotomy is preferred for injuries of the ascending aorta and of the innominate and proximal carotid arteries. Extension into the neck, along the anterior border of the right sternocleidomastoid muscle, allows access to the proximal right subclavian artery and to the right vertebral artery. A thoraco-sternotomy [45] associates the antero-lateral thoracotomy through the third intercostal space to a median sternotomy; this access permits a rapid access to the mediastinum, to have the control of the proximal aorta and, if is necessary, to perform cardiopulmonary bypass. during thoracic and thoraco-abdominal aortic surgery. In emergency situations, however, the operation must be performed as expeditiously as possible without these techniques. The technique of choice is indeed dictated by the urgency of the patient's condition, availability of technical personnel and surgeon preference at each hospital. CLAMP-AND-SEW The single clamp-and-sew method of repair has many strong advocates, who point to the technique's simplicity, the low paraplegia rate if crossclamp times are short, and a low mortality rate as compared with approaches that use heparin [17,29, 46,47]. Sweeney et al. reported a mortality rate of 12% and a permanent paraplegia rate of 1.3% with a mean cross- clamp time of 24 minutes [47]. However, a cross-clamp time this brief cannot be guaranteed [10] and most surgical groups do not meet this mark [48,49]. Experimental data and clinical series have demonstrated that the occlusion of the descending thoracic aorta longer than 30 minutes

TECHNIQUES OF ORGAN PERFUSION/REPAIR

Discussion continues regarding the optimal management of injury to the thoracic aorta after trauma, because of concerns about spinal cord ischemia and subsequent neurologic deficit. Several methods for indirect assessment of perfusion are available

TRAUMATIC RUPTURE OF THE THORACIC AORTA

is associated with the development of postoperative paraplegia [49-51]. The average international crossclamp time, as reported by von Oppell et al., is 41 minutes [51]. Even those who advocate the clamp-and-sew technique have patients with paraplegia because of unexpectedly long cross-clamp times. Most recently, groups using this technique reported paraplegia rates in the 2% to 24% range (Table II). This is in contrast to the 0% to 7% of those who use active distal support [8,49,54]. Cross-clamping the thoracic aorta renders ischemia of all organs distal to the clamp, including the liver, intestines, kidneys and skeletal muscle. Van Norman et al. demonstrated that aortic unclamping after emergency surgery for traumatic thoracic aorta tears was associated with a decrease in P02, an elevated serum potassium level, and a marked metabolic acidosis that persisted for up to six hours after declamping [55]. Roberts et al. demonstrated that after release of the cross-clamp in unshunted animals, there was a reduction in renal function to 50% to 85% of baseline values [56]. Occlusion of the descending thoracic aorta is also associated with a multitude of physiologic derangements including the development of proximal hypertension, which increases the afterload and places strain on the left ventricle. A reduction of the cardiac index by 29% has been demonstrated together with an increase in left ventricular wall stress with the application of a proximal thoracic cross-clamp [57]. This is extremely important, considering that

i l
115

FIG. 9 Most common approaches to the thoracic aorta: (A) postero-lateral thoracotomy, with possible abdominal extension; (B) anterior thoracotomy, with possible contralateral extension and/or sternotomy.

1st author [ref.] Schmidt [15] Maggisano [43] Sweeney [47] Fabian [49] Attar [52] Jahromi [53]

Year

Number of patients

Mortality (%)

Paraplegia/ paraparesis* (%)

1 1 1

1992 1995 1997 1997 1999 2001

32 36 71 73 54 21

5 3 9 11 12 2

(16)

(3.7)

(8)
(13) (15) (22) (10)

(3) (2)
(16.4)
(24) (16)

12 10 3

* Related to live patients

VASCULAR

EMERGENCIES

many patients with blunt aortic injuries have an associated cardiac contusion; this places the patients at increased risk for peri-operative cardiac arrest, postoperative myocardial infarction and the development of adult respiratory distress syndrome after surgery [58]. Furthermore, the use of vasodilators such as nitroprusside to control proximal hypertension after the application of the proximal crossclamp reduces arterial perfusion pressure to the spinal cord, while the application of the cross-clamp increases cerebrospinal fluid pressure [59].

loading, which may be helpful in elderly patients and in those with myocardial contusions [60]. The most common methods of distal circulatory support are complete and partial cardiopulmonary bypass via an external pump and left atrial to aortic or femoral bypass (left heart bypass, LHBP).

CARDIOPULMONARY BYPASS
Cardiopulmonary bypass has the ability to oxygenate, scavenge shed blood and heat and cool as desired [8,48]. However, the use of full anticoagulation in a multiply injured patient may increase the risk of bleeding and death. For this reason complete cardiopulmonary bypass has largely fallen into disfavor. Patients with thoracic aortic injuries at multiple levels who require extensive repair are notable exceptions [18]. Partial cardiopulmonary bypass with heparinbonded circuits has been validated as an attractive option in the setting of traumatic ruptured aorta as a means of avoiding the use of systemic heparinization. Cannulation of the right atrium via the femoral vein is simple and provides a clear and unobstructed working field [61]. It can provide adequate distal circulatory support and safely heat, cool, oxygenate, and transfuse as required as opposed to LHBP; with partial cardiopulmonary bypass, improved oxygenation may also be attained in the presence of lung contusions [8]. Some literature results are summarized in Table III.

ACTIVE DISTAL CIRCULATORY SUPPORT

Active distal circulatory support is very effective in reducing the risk of paraplegia, particularly when long cross-clamp times are needed. Simple aortic clamping is known to raise cerebrospinal fluid pressures, which can exacerbate intracranial injuries. Unloading of the proximal aorta with an active distal support system may minimize that rise. The adjuncts for distal aortic perfusion reduce but do not eliminate the risk of neurological deficits [53]. The origin of the artery of Adamkiewicz arises between T5 and T8 in 12% to 15% of patients [8]. A low placement of a distal aortic clamp may result in obstruction of the artery of Adamkiewicz [8]. Alternatively, a clamp placed proximally to the left subclavian artery may transiently reduce the spinal cord blood flow through the vertebrobasilar system. Distal support has other theoretical advantages over simple clamping: it provides proximal cardiac un-

12
116

ler author [ref.] Soyer [54] Pate [8] Fabian [49] Fabian (full) [49] Gammie [48] Attar [52] Jamieson [62]

Year
1992 1995 1997 1997 1998 1999
2002

Number of patients

Mortality (%)

Paraplegia/ paraparesis * (%)

43 88 39 22 10 43 42

3 6 5 5 1

7
5

(7) (7) (12.8) (22.7) (10) (16)

0 2 3 1 0 0 0

(2) (7.7) (4.5)

(12)

* Related to live patients

TRAUMATIC RUPTURE OF THE THORACIC AORTA LEFT HEART BYPASS

LHBP, which is connected between the left atrium and distal aorta or a femoral artery (Fig. 10), can be used with little or no heparin because an oxygenator is not required. It also has some limitations; most LHBP systems do not incorporate a heat exchanger and are dependent on adequate pulmonary function for oxygenation. Cannulation of the left atrial appendage or pulmonary vein can sometimes be difficult in the presence of an extensive mediastinal hematoma. Additionally, because of the risk of air embolization in these closed systems, physicians are reluctant to rapidly infuse volume through them. Some authors also described clamp-related embolic events due to heparineless circuit [63]. Injuries to the aortic arch, innominate artery or ascending aorta, which represent a minority of aortic trauma cases, cannot be repaired with LHBP. Full cardiopulmonary bypass or even profound hypothermic circulatory arrest may be necessary. Furthermore, in patients who present in extremis, necessitating immediate control of hemorrhage and restoration of blood volume before repair, partial bypass may offer no advantage. However, the majority of aortic transections occur at the isthmus; the patients who survive to surgery are usually hemodynamically stable, with the peri-aortic hematoma contained in the mediastinum. These injuries may be complex or close to the proximal cross-clamp, thus complicating and potentially prolonging the repair. The adjuvant use of LHBP may help in extending the critical window, so the aortic crossclamp can be applied safely in such situations (Table IV).

1 1
117

FIG. 10 Illustration of LHBP7 an active method of distal perfusion.

1st author [ref.] Read [64] Kipfer [65] Contino [66] Fabian [49] Gammie [48] Symbas [67]

Year
1993 1994 1994 1997 1998

Number of patients
16 10 24 69 14 19

Mortality (%)
2 0 5 10 1 5

Paraplegia/ paraparesis (%)

0 0 0 2 0 0

(13)

(20.8) (14.5)
(7) (26)

(2.9)

2002

VASCULAR

EMERGENCIES

SURGICAL TECHNIQUES
Injury usually originates at the medial side of the aorta, at the level of the ligamentum arteriosum. The goal of initial operative approach is to obtain proximal and distal control of the descending thoracic aorta. Vascular clamps are applied to three locations: proximal aorta, distal aorta and subclavian artery. The hematoma is entered and backbleeding from intercostal arteries is controlled. Care is taken to avoid indiscriminate ligation of intercostal vessels: only those required for adequate repair of the aorta should be ligated. During aortic reconstruction clamps should be moved as close as possible to the site of injury, in order to reduce spinal cord ischemia, especially when using the clamp-sew technique. Primary suture and graft interposition (Fig. 11) are the main strategies of aortic repair after a traumatic lesion. The first is usually the preferred choice, being simple and fast. This technique is adequate in cases of partial laceration, but also in disruptions when aortic stumps are not too distant and severely injured. Frequently in aortic trauma, adventitial layers are retracted and must be carefully included in the aortic suture. During this phase, the esophagus must be accurately separated from the posterior aortic wall. Teflon-felt pledgets can be useful to reinforce a friable suture line. Finally, the absence of a pros-

thetic graft decreases risk of infection. Graft interposition has been used in more than 85% of the reported cases [11], and is advisable when more than 2 cm of the vessel is injured. The aorta is grafted with a straight polyester graft which is kept as short as possible.

COMPLICATIONS
Significant postoperative complications related to the thoracic injury can develop in these patients. Cardiac, renal, pulmonary and neurologic morbidities have been reported in the literature with rates up to 50% [68]. Pate et al. registered significant postoperative complications in 72.7% of the 44 patients operated after 1984. The most common complications in their experience were adult respiratory distress syndrome and pneumonia (29.5%), severe systemic hypertension (20.5%), coagulopathy, renal failure, serious cardiac arrhythmias, late tension pneumothorax and thoracic wound infection [8]. There were additional problems stemming from orthopedic, abdominal, cranial and burn injuries. Paraplegia is the most devastating complication after repair of descending aortic transection since most of the patients who suffer blunt aortic injury are young. The personal loss caused by paraplegia and the economic impact on society are enormous.

12
118

FIG. 11 Methods of aortic repair: (A) simple suture, and (B) graft interposition.

TRAUMATIC RUPTURE OF THE THORACIC AORTA

The two central issues associated with prevention of paraplegia are duration of cross-clamp time and use of distal aortic perfusion. Patients with acute aortic rupture are at a greater risk for paraplegia during aortic repair than patients with chronic aneurysms or coarctation. This increased risk is likely due to the lack of preformed collaterals along with the additional complications unique to trauma patients including pulmonary and cardiac contusions, shock, hypoxia and hypo-osmolality from fluid overload [8]. Late deaths, in fact, are most often due to multiple organ failure [69]. endovascular treatment. Nevertheless, the placement of vascular stent grafts has not yet been shown to increase the risk of paraplegia compared to conventional surgical intervention. The largest series of thoracic aortic diseases electively treated with endoluminal grafts reported a paraplegia rate of 3.6% [70]. Up to 70% of traumatic aortic injuries affect the isthmus segment excluding only a few branches to the spinal cord if an endograft is placed (Fig. 12). The length of the covered aorta may be limited to a few centimeters near the diseased segment, which could lessen the risk of medullar ischemia. Lachat et al. have reported significant lower rates of morbidity and mortality for stent grafting of acute aortic lesions demonstrating the benefits and advantages offered by this minimally invasive technique [71]. In this report patients were hemodynamically stable to undergo contrast enhanced CT and angiographic evaluation to determine the suitability for stent grafting. TEE and intravascular ultrasound have been described as helpful associated diagnostic studies [72]. Early endovascular treatment was considered in the treatment of stable nonbleeding lesions, after recovery of associated life-threatening injuries. The rupture had to be contained, with a proximal neck of normal

ENDOVASCULAR INTERVENTIONS
Despite advances in surgical and peri-operative care, conventional surgery for acute aortic rupture still carries a significant morbidity and mortality. Thus, endovascular procedures may be an attractive option for treating this kind of lesions. Theoretical advantages of stent grafting are multiple; the absence of aortic cross-clamping prevents rising of intracranial pressure in some situations like severe head injury. Patients with pulmonary contusions do not need single-lung ventilation as used for conventional surgical repair. Critical statements have pronounced a high risk of spinal cord ischemia due to the exclusion of intercostal arteries during

119

FIG. 12 A less invasive technique: endovascular exclusion of an isthmic injury.

VASCULAR

EMERGENCIES

12
120

appearance and length greater than 5 mm and a diameter less than 36 mm. Depending on the associated lesions and the bleeding risk, no heparin at all or a maximum dose of 5000 IU was administrated intravenously and completely reversed after stent graft delivery. The immediate technical success rate was 100%, as the aortic lesion could be excluded in all the cases; there was one early death due to hemorrhagic shock in a patient with a semicircumferential rupture who probably had an undetected incomplete proximal sealing. Because most injuries occur at the aortic isthmus, much concern has been raised regarding the placement of rigid devices in an angulated aortic arch. As previously described, this can cause an inadequate seal of the thoracic aorta. However, this has been largely overcome with newer and more flexible devices. This new therapeutic strategy has few drawbacks worth describing. The delivery system calibers are actually of large diameter (18Fr to 24Fr), being difficult to introduce through small and spastic arteries of young people or tortuous and calcified arteries of older people. A further concern is the stock inventory of devices, which must be available for emergency cases in different sizes and lengths. Another topic of concern is the relatively young population suffering from aortic trauma, which could be treated with this method. Stent grafts are actually produced mainly for treatment of chronic aortic pathologies, especially aneurysms, and do not come in adequate diameters for the small vessels of young people. Moreover, the mean age (38.7 years) of this population raises many questions about the long-term follow-up after endoluminal treatment [49]. Additional study of the endovascular intervention role in aortic injury is required to obtain a stronger support for its use in this scenario.

ment, five of these arrived in extremely critical conditions. The other two died during urgent thoracotomy dictated by hemodynamic deterioration, before the bleeding could be controlled. An early chest X-ray was obtained in nine patients. SCT and/or aortography studies were performed in seven patients with stable hemodynamic conditions. Two patients also underwent TEE. Associated lesions were found in every patient of our series (Table V), and were present mainly in those patients who died in the emergency department. The most common lesions were other thoracic injuries in 12 patients and abdominal visceral injuries in 7 patients. Table VI summarizes the clinical and surgical features of the patients. Hospital mortality for patients undergoing surgical repair was 3/8 (38%). One death occurred in a 56-year-old woman involved in a motor vehicle accident. Thoracic aortic rupture repair was performed using the clamp-and-sew technique, and during wound closure she suffered from an irreversible cardiac arrest. The second was a 43-year-old man who died of uncontrolled bleeding during repair of a penetrating lesion of median aortic arch. The third death was in a 48-year-old man who underwent repair of an isthmic pseudoaneurysm using LHBP, and died on the 12th postoperative day of multiple organ failure. One patient developed paraplegia following a

Injury

Number of patients

Personal experience
Between January 1988 and September 2002,15 patients with acute injury or rupture of the thoracic aorta were admitted to the emergency department of our institution. There were 11 (73%) males and 4 (27%) females aged 19 through 68 years (mean age of 39.8 years). Ten patients had history of motor vehicle crash, two suffered from a motorcycle crash and three had lesions secondary to penetrating traumas. Seven patients died in the emergency depart-

Other thoracic traumas (including fractures and great vessels) Abdominal visceral Closed-head injury Peripheral fractures Head fractures Abdominal vascular Neck fracture

12 7 6 6 4 2 1

TRAUMATIC RUPTURE OF THE THORACIC AORTA

Case Sex
1 2 M M

Age
28 19

Lesion characteristics
Isthmic rupture (MVC) Descending thoracic aorta perforation (Gunshot) Isthmic rupture (MVC) Descending thoracic aortic dissection and rupture (MVC) Descending thoracic aortic perforation (Metallic bar for building construction) Descending thoracic aortic injury -pseudoaneurysm (MC) Isthmic injury pseudoaneurysm (MVC) Ascending aorta rupture (MVC) Isthmic injury pseudoaneurysm (MVC) Ascending aorta perforation (Stab wound) Isthmic and diaphragmatic aortic rupture (MC) Ascending rupture (MVC) Isthmic rupture (MVC) Isthmic rupture (MVC) Isthmic rupture

Procedure

In-hospital outcome

Follow-up

Urgent thoracotomy Death in emergency room (hemorrhagic shock) Urgent thoracotomy Death in emergency room (hemorrhagic shock) Graft interposition (clamp and sew) Graft interposition (clamp and sew) Urgent thoracophreno-laparotomy Death in immediate postoperative period (cardiac arrest) Discharged Alive after 7 years

56

31

43

Death in operating room (uncontrolled bleeding)

26

Graft interposition (LHBP) Graft interposition (LHBP) Graft interposition (CPB) Graft interposition (LHBP) Sternotomy (simple suture) Autopsy findings

Discharged

Alive after 1.5 year

34

Discharged

Alive after 6 months

8 9

F M

63 48

Discharged Death in postoperative period (MOF) Discharged

Alive and paraplegic after 2 months

121

10

38

Alive after 5 years

11

22

12 13 14 15

M M M F

56 31 68 35

Autopsy findings Autopsy findings Autopsy findings Autopsy findings

(MVC)
CPB: cardiopulmonary bypass LHBP: left heart bypass MC: motorcycle crash MOF: multiple organ failure MVC: motor vehicle crash

VASCULAR

EMERGENCIES

graft interposition in the ascending aorta under cardiopulmonary bypass. The other four patients suffered no major complications and were discharged in good condition.

Conclusion
The major difficulty in the evaluation of data concerning blunt aortic injury is that retrospective reviews often collect patients with all aortic injuries together, comparing outcomes for injuries in dif-

ferent locations of the aorta with different methods of repair and different surgeons at different institutions. Concomitant severe lesions are usually present in patients suffering from aortic trauma, and thus diagnostic and therapeutic timing must be rapidly established. The polytrauma patient should be managed under a multidisciplinary approach and treated in a specialized center. Organizing trauma centers is a very important issue. In these places, patients can be better treated, and prospective studies will be helpful for creating guidelines and treatment algorithms.

R E F E R E N C E S

~j [/ *-~ 1 2

1 Mattox KL, Feliciano DV, Burch J et al. Five thousand seven hundred sixty cardiovascular injuries in 4459 patients. Epidemiologic evolution 1958 to!987. Ann Surg 1989; 209: 698-707. 2 Tsai FC, Chang YS, Lin PJ, Chang CH. Blunt trauma with flail chest and penetrating aortic injury. EurJ Cardiothorac Surg 1999; 16: 374-377. 3 Siersema PD, Tan TG, Sutorius FF et al. Massive hemorrhage caused by a perforating Gianturco-Z stent resulting in an aortoesophageal fistula. Endoscopy 1997; 29: 416-420. 4 Esterra A, Mattox KL, Wall MJ. Thoracic aortic injury. Semin Vase Surg 2000; 4: 345 -352. 5 Dosios TJ, Salemis N, Angouras D, Nonas E. Blunt and penetrating trauma of the thoracic aorta and aortic arch branches: an autopsy study. / Trauma 2000; 49: 696 - 703. 6 Greendyke R. Traumatic rupture of aorta: special reference to automobile accidents. JAMA 1966; 195: 527-530. 7 Hunt JP, Baker CC, Lentz CW et al. Thoracic aorta injuries: management and outcome of 144 patients. / Trauma 1996; 40: 547-556. 8 Pate JW, Fabian TC, Walker WA. Acute traumatic rupture of the aortic isthmus: repair with cardiopulmonary bypass. Ann Thome Surg 1995; 59: 90 -99. 9 Higgins RS, Sanchez JA, DeGuidis L et al. Mechanical circulatory support decreases neurologic complications in the treatment of traumatic injuries of the aorta. Arch Surg 1992; 127: 516-519. 10 DelRossi AJ, Cernaianu AC, Madden LD et al. Traumatic disruptions of the thoracic aorta: treatment and outcome. Surgery 1990; 108: 864-870. 11 Lee RB, Stahlman GC, Sharp KW. Treatment priorities in patients with traumatic rupture of the thoracic aorta. Am Surg 1992; 58: 37-43. 12 Duhaylongsod FG, Glower DD, Wolfe WG. Acute traumatic aortic aneurysm: the Duke experience from 1970 to 1990. / Vase Surg 1992; 15: 331 -343. 13 Cowley RA, Turney SZ, Hankins JR et al. Rupture of thoracic aorta caused by blunt trauma. A fifteen-year experience./ Thorac CardiovascSurgim; 100: 652-661. 14 Verdant A, Cossette R, Dontigny L et al. Acute and chronic traumatic aneurysms of the descending thoracic aorta: a 10 year experience with a single method of aortic shunting. / Trauma 1985; 25: 601-607. 15 Schmidt CA, Wood MN, Razzouk AJ et al. Primary repair of traumatic aortic rupture: a preferred approach./ Trauma 1992; 32: 588-592.

16 Verdant A, Page A, Cossette R et al. Surgery of the descending thoracic aorta: spinal cord protection with the Gott shunt Ann Thorac Surg 1988; 46: 147-154. 17 Mattox KL, Holzman M, Pickard LR et al. Clamp/repair: a safe technique for treatment of blunt injury to the descending thoracic aorta. Ann Thorac Surg 1985; 40: 456-463. 18 Bongard F. Thoracic and abdominal vascular trauma. In: Rutherford RB (ed). Vascular surgery. 5th edition. Philadelphia, W.B. Saunders Co, 2000: pp 872-892. 19 Feliciano DV, Bitondo CG, Mattox KL, et al. Chilian trauma in the 1980's. A 1-year experience with 456 vascular and cardiac injuries. Ann Surg-1984; 199: 717-724. 20 Williams JS, Graff JA, UkuJM, SteinigJP. Aortic injury in vehicular trauma. Ann Thorac Surg 1994; 57: 726-730. 21 May E. Clinical evaluation of the critically injured. Springfield, Charles Thomas, 1975. 22 Zehnder M. Delayed post-traumatic rupture of the aorta in a young, healthy individual after closed head injury. Angiology 1956;7:252-2'56. 23 Ben-Menachem Y Rupture of the thoracic aorta by broadside impacts in road traffic and other collisions: further angiographic observations and preliminary autopsy findings. / Trauma 1993; 35:363-367. 24 Katyal D, McLellan BA, Brenneman FD et al. Lateral impact motor vehicle collisions: significant cause for blunt traumatic rupture of the thoracic aorta. J Trauma 1997; 42: 769-772. 25 Feczko JD, Lynch L, Pless JE et al. An autopsy case review of 142 nonpenetrating (blunt) injuries of the aorta. J Trauma 1992; 33: 846-849. 26 Verdant A. Major mediastinal vessel injury: an underestimated lesion. Can J Surg 1987; 30: 402-404. 27 Pasic M, Ewert R, Engel M et al. Aortic rupture and concomitant transection of the left bronchus after blunt chest trauma. Chest 2000; 117:1508-1510. 28 Symbas PN, Tyras DH, Ware RE, Hatcher CR Jr. Rupture of the aorta. A diagnostic triad. Ann Thorac Surg 1973; 15: 405-410. 29 Hilgenberg AD, Logan DL, Akins CW et al. Blunt injuries of the thoracic aorta. Ann Thorac Surg1992; 53: 233-239. 30 Patel NH, Stephens KEJr, Minis SE et al. Imaging of acute thoracic aortic injury due to blunt trauma: a review. Radiology 1998; 209:335-348. 31 Tisnado J, Tsai FY, Als A, Roach JF. A new radiographic sign of acute traumatic rupture of the thoracic aorta: displacement of the nasogastric tube to the right. Radiology 1977; 125: 603-608.

TRAUMATIC RUPTURE OF THE THORACIC AORTA

32 Novelline R, RheaJT, Rao PM, StukJL. Helical CT in emergency radiology. Radiology 1999; 213: 321 - 339. 33 Gavant ML. Helical CT grading of traumatic aortic injuries. Impact on clinical guidelines for medical and surgical management. Radial Clin North Am 1999; 37: 553-574. 34 Reardon MJ, Hedrick TD, Letsou GV et al. CT reconstruction of an unusual chronic posttraumatic aneurysm of the thoracic aorta. Ann Thome Surg 1997; 64: 1480-1482. 35 Wicky S, Wintermark M. Schnyder P et al. Imaging of blunt chest trauma. EurRadioimQ; 10:1524-1538. 36 Ben-Menachem Y. Assessment of blunt aortic-brachiocephalic trauma: should angiography be supplanted by transesophageal echocardiography?/Trauma 1997; 42: 969-972. 37 Gavant ML, Menke PG, Fabian T et al. Blunt traumatic aortic rupture: detection with helical CT of the chest. Radiology 1995; 197:125-133. 38 Mirvis SE, Shanmuganathan K, Miller BH et al. Traumatic aortic injury: diagnosis with contrast-enhanced thoracic CT: five year experience at a major trauma center. Radiology 1996; 200: 413-422. 39 Brooks SW, Young JC, Cmolik B et al. The use of transesophageal echocardiography in the evaluation of chest trauma. / Trauma 1992; 32: 761-768. 40 Buckmaster MJ, Kearney PA, Johnson SB et al. Further experience with transesophageal echocardiography in the evaluation of thoracic aortic injury. J Trauma 1994; 37: 989-995. 41 Lobato AC, Quick RC, Phillips B et al. Immediate endovascular repair for descending thoracic aortic transection secondary to blunt trauma. JEndovasc 77^2000; 7:16-20. 42 Parmley LF, Mattingly TW, Manion WC et al. Nonpenetrating traumatic injury of the aorta. Circulation 1958; 17: 1086-1101. 43 Maggisano R, Nathens A, Alexandrova NA et al. Traumatic rupture of the thoracic aorta: should one always operate immediately? Ann Vase Surg 1995; 9: 44-52. 44 Hurley EJ, Blaisdell FW. Aortic injuries. In: Blaisdell FW, Trunkey DD (eds). Cervicothoradc Trauma. New York, Thieme, 1986: pp 223-245. 45 Villard J, Vial P, Dureau G et al. Thoraco-bisternotomy in cardiovascular surgery. Nouv Presse Med 1982; 11: 3647-3649. 46 Tatou E, Steinmetz E, Jazayeri S et al. Surgical outcome of traumatic rupture of the thoracic aorta. Ann Thorac Surg 2000; 69: S ^ 70-73. F 47 Sweeney MS, Young DJ, Frazier OH et al. Traumatic aortic transections: eight-year experience with the "clamp-sew" technique. Ann Thorac Surg 1997; 64: 384-389. 48 Gammie JS, Shah AS, Hattler BG et al. Traumatic aortic rupture: diagnosis and management. Ann Thorac Surg 1998; 66: 1295-1300. 49 Fabian TC, Richardson JD, Croce MA et al. Prospective study of blunt aortic injury: Multicenter Trial of the American Association for the Surgery of Trauma. / Trauma 1997; 42: 374-383. 50 Katz NM, Blackstone EH, Kirklin JW, Karp RB. Incremental risk factors for spinal cord injury following operation for acute traumatic aortic transection. J Thorac Cardiovascular Surg 1981; 81: 669-674. 51 von Oppell UO, Dunne TT, De Groot MR, Zilla P. Traumatic aortic rupture: twenty-year metaanalysis of mortality and risk of paraplegia. Ann Thorac Surg 1994; 58: 585-593. 52 Attar S, Cardarelli MG, Downing SW et al. Traumatic aortic rupture: recent outcome with regard to neurologic deficit. Ann 77wracSwrgl999;67:959-965. 53 Jahromi AS, Kazemi K, Safar HA et al. Traumatic rupture of the thoracic aorta: cohort study and systematic review. J Vase Surg 2001; 34:1029-1034. 54 Soyer R, Bessou JP, Bouchart F et al. Acute traumatic isthmic aortic rupture. Long-term results in 49 patients. EurJ Cardiothorac Swrgl992;6:431-437. 55 van Norman GA, Pavlin EG, Eddy AC, Pavlin DJ. Hemodynamic and metabolic effects of aortic unclamping following emergency surgery for traumatic thoracic aortic tear in shunted and unshunted patients./Trauma 1991; 31:1007-1016. 56 Roberts AJ, Nora JD, Hughes WA et al. Cardiac and renal responses to cross-clamping of the descending thoracic aorta. J Thorac Cardiovasc Surg 1983; 86: 732 - 741. 57 Kouchoukos NT, Lell WA, Karp RB, Samuelson PN. Hemodynamic effects of aortic clamping and decompression with a temporary shunt for resection of the descending thoracic aorta. Surgery 19'79; 85: 25-30. 58 Kram HB, Appel PL, Shoemaker WC. Increased incidence of cardiac contusion in patients with traumatic thoracic aortic rupture. Ann Surg 1988; 208: 615-618. 59 Szwerc MF, Benckart DH, Lin JC et al. Recent clinical experience with left heart bypass using a centrifugal pump for repair of traumatic aortic transection. Ann Swrgl999; 230: 484-492. 60 Hug HR, Taber RE. Bypass flow requirements during thoracic aneurysmectomy with particular attention to the prevention of left heart failure. J Thorac Cardiovasc Surg-1969; 57: 203-213. 61 von Segesser LK, Weiss BM, Garcia E et al. Reduction and elimination of systemic heparinization during cardiopulmonary bypass. / Thorac Cardiovasc Surg 1992; 103: 790 - 799. 62 Jamieson WR, Janusz MT, Gudas VM et al. Traumatic rupture of the thoracic aorta: third decade of experience. Am J Surg 2002; 183:571-575. 63 Duke BJ, Moore EE, Brega KE. Posterior circulation cerebral infarcts associated with repair of thoracic aortic disruption using partial left heart bypass. / Trauma 1997; 42:1135-1139. 64 Read RA, Moore EE, Moore FA, Haenel JB. Partial left heart bypass for thoracic aorta repair. Survival without paraplegia. Arch Surg 1993; 128: 746 -752. 65 Kipfer B, Leupi F, Schuepbach P et al. Acute traumatic rupture of the thoracic aorta: immediate or delayed surgical repair? Eur J Cardiothorac Surg 1994; 8: 30 - 33. 66 Contino JP, Follette DM, Berkoff HA et al. Use of Carmedacoated femoral-femoral bypass during repair of traumatic aortic pseudoaneurysms. Arch Swrgl994; 129: 933-939. 67 Symbas PN, Sherman AJ, Silver JM et al. Traumatic rupture of the aorta: immediate or delayed repair? Ann Surg 2002; 235: 796-802. 68 Butler KL, Moore EE, Harken AH. Traumatic rupture of the descending thoracic aorta. AORV/1996; 63: 917-925. 69 Chiesa R, Lucci C, Castellano R et al. Multiple organ failure after surgical arterial reconstruction. In: Branchereau A, Jacobs M (eds). Complications in vascular and endovascular surgerj (Part II). Armonk, Futura Publishing Company, 2002: pp 23-32. 70 Mitchell RS, Miller DC, Dake MD et al. Thoracic aortic aneurysm repair with an endovascular stent graft: the "first generation". Ann Thorac Surg 1999; 67:1971-1974, discussion 1979-1980. 71 Lachat M, Pfammatter T, Witzke H et al. Acute traumatic aortic rupture: early stent-graft repair. EurJ Cardiothorac Swig 2002; 21: 959-963. 72 Thompson CS, Rodriguez JA, Ramaiah VG et al. Acute traumatic rupture of the thoracic aorta treated with endoluminal stent grafts. J Trauma 2002; 52:1173-1177.

12. 123

This page intentionally left blank

ACUTE OCCLUSION OF THE RENAL ARTERIES

XAVIER BARRAL, PHILIPPE PACHECO DANIEL GRANDMOUGIN, DIDIER BOURRAT, JEAN-PIERRE FAVRE

Acute renal ischemia rarely occurs. During the last ten years we have only encountered some twenty publications in the literature addressing this issue. This limited scientific attention, associated with our personal data analysis, invites for three comments. - Occlusion of atherosclerotic stenoses of the renal artery silently develops and is only diagnosed in the chronic stage, explaining the exceptional observation of acute failure despite the high incidence of renal artery disease. - Acute renal ischemia occurs in the setting of varying and uncommon pathologies without mutual causal relations. This multifactorial etiology delays the diagnostic process and does not allow a general treatment. - In the literature, acute renal ischemia is evaluated in clinical cases which are pooled on the basis of a specific etiology or treatment. Only one publication comprises more than 30 patients thanks to a compilation of 20 years of experience [1]. Therefore, the treatment of this seldom acute ischemia is most often based on personal belief rather than scientific evidence.

1 1
125

Incidence
The exact incidence of acute renal ischemia is difficult to assess. Only few authors report on their personal data. The publication of Hoxie and Coggin is often referred to, describing 205 recent renal infarctions in 14 471 autopsies carried out in a nine-year period, accounting for 1.4% renal infarctions in a deceased population [2]. This publication dates

from an era (1940) in which effective treatment like anticoagulation did not exist. A more recent and interesting study is described by Domanovits et al. [3] who, in a period of 45 months, found 17 cases of renal infarction among 248 882 patients admitted to an emergency center (0.007%). In our personal experience, excluding postoperative and early post-transplantation occlusions, we have encountered 18 cases of acute renal occlusion in

VASCULAR

EMERGENCIES

a twenty-year period. These figures represent approximately 1.2% of our transplant and renal artery surgery and 0.13% of our overall peripheral vascular surgical procedures. The incidence of unilateral or bilateral acute occlusion in a single functioning kidney is even more exceptional and is found in 5% to 20% of acute renal artery obstruction [4]. Noteworthy is the fact that the majority of acute infarctions occur in the left renal artery (more than 60%). The more oblique origin of the left renal artery in relation to the aorta favors preferential passage of emboli as compared to the right side [1,5].

suprarenal ulcerating plaques. Finally, some cases of paradoxical embolism have been described [13].

ACUTE ARTERIAL THROMBOSIS

The second important cause of acute renal ischemia is thrombosis, encountered in 32.6% of cases (Table I). The mechanisms of these thromboses are extremely variable. Progression of atherosclerotic or dysplastic stenosis, spontaneous or accelerated by angiotensin-converting enzyme inhibitors, can lead to complete occlusion. Furthermore, local thrombosis can be associated with diseases like arteritis [14], Behcet's disease [15], oral contraceptives [16], marijuana [17] or cocaine abuse [18], coagulation disorders [19,20], ergotamine intoxication [21], Chagas disease [22], umbilical artery catheterization [23] and heparin allergy (Fig. 1). Finally, acute renal ischemia can be due to secondary thrombosis in an aortic dissection, extending to the renal arteries [24,25] or in an isolated, spontaneous renal artery dissection [26].

Etiology
The causes of acute renal artery occlusion can be distinguished in three main groups: emboli, thrombosis and trauma (Table I).

EMBOLI
In the literature only eight series included more than ten patients and the rate of embolization was 43.5%. These renal emboli correspond to 2.3% of all embolic cases [11]. In 70% the source of emboli is the heart, either caused by arrhythmia, valve disease or myocardial infarction [12]. Atherosclerotic emboli are more uncommon but can be dislodged during aortic cross-clamping, originating from

TRAUMATIC CAUSES
The third group of acute occlusions consists of post-traumatic reasons and was identified in 23.8% of patients. Dorsolumbar or thoraco-abdominal trauma is most often the consequence of a car accident or other acceleration/deceleration impact. Two mechanisms can cause a subintimal rupture and subsequent thrombosis:

13
126

1st author [ref.] Stables [6]

Ouriel [1] Lacombe [7] Blum [8] Salam [9] Bouttier [10] Domanovitz [3] Personal experience Total

Year of publication

Number of patients

Emboli N(%)

Thrombosis N(%)

Trauma N(%)

1976 1986 1992 1993 1993 1997 1999

2002
8 series

21 35 20 14 10 12 17 18
147

0
13 5 14 2 12 14
4
64 (43.5)

16 15 0 4
0 3

21 6 0 0 4 0
0 4
35 (23.8)

10 48 (32.6)

ACUTE OCCLUSION OF THE RENAL ARTERIES

arrhythmia or emboligenic cardiopathy, the diagnosis of acute renal occlusion should be kept in mind. The likelihood of a renal artery embolus can reach 80% [3] and only immediate diagnosis and treatment offer a possibility for a functional recovery of the affected kidney. Therefore, emergency contrast enhanced computed tomography (CT) scanning should be performed, especially if the general and hemodynamic condition of the patient are acceptable. Again, any further time delay threatens the affected kidney (Fig. 2). If the clinical situation is less urgent, assessment of serum lactate dehydrogenase is useful. In case of renal artery occlusion, this enzyme is generally elevated [32]. Serum ureum and creatinine are routinely investigated but they have limited diagnostic value. In case of oliguria secondary to renal artery occlusion, it is worthwhile to assess the levels of sodium in serum and urine, which are equivalent in most cases. The excreted sodium fraction is actually 100% and the equivalent sodium levels in serum and urine should indicate immediate CT scanning [33]. CT angiography is currently the method of choice to confirm the suspicion of renal artery occlusion. Performed before and after the injection, it allows accurate assessment of the renal artery and the parenchymatous damage [34]. Contrast enhanced CT allows for differentiation between recent acute occlusions or pre-existing occlusions, the latter being characterized by renal atrophy, thin cortex and irregular capsule. Magnetic resonance angiography could also be an useful method; however, its availability and accessibility are inferior to those of CT. Duplex scanning of the kidney is an easy technique but it depends on the technician and degree of meteorism and offers less information on the renal parenchym [35]. At present it is a second-choice technique and is only applied in the absence of CT. The goal should be not to duplicate diagnostic tests and not to lose important time and start treatment as soon as possible. Even though arteriography remains the gold standard in the diagnosis of renal artery occlusions, the technique is currently replaced by the latest generation CT scanners. Arteriography allows us to identify the type of occlusion and the level of contrast reinjection in the distal renal artery. If selective angiography of the origin of the renal artery does not delineate the distal reinjection, selective angiography of renal artery side branches, lumbar or diaphragmatic arteries can be performed (Fig. 3).

FIG. 1 Acute occlusion of the left renal artery, discovered three days after heparine allersy. Medical treatment.

- crush injury of the artery against the spine, - tearing-off or overstretching the renal artery pedicel. These traumatic lesions of the renal artery [27,28] are observed in approximately 4% of abdominal trauma. The injury can be the result of an iatrogenic trauma like selective renal artery arteriography [29] or any endovascular procedure (angioplasty, stenting, endograft). It should be noted that the latter causes decreased significantly because in the majority of cases these complications can be solved immediately [30].

_
127

Diagnostics
Because of the low incidence and the extremely variable clinical features, acute renal artery occlusions often pass unnoticed. The clinical diagnosis is established in one of five patients [31], with a substantial delay of additional investigations confirming renal artery involvement. Typically, the clinical picture is characterized by an intense pain in the flank with possible referred pain to the external genital organs, nausea, vomiting, paralytic ileus, and micro- or macroscopic hematuria. Arterial hypertension might also be present and light fever can develop after a few hours. In patients presenting with nephretic colic, previous

VASCULAR

EMERGENCIES

FIG. 2 A - Post-traumatic acute occlusion of the left renal artery. CT scan four hours after the accident. B - Angiography immediately performed after the CT scan. C - Completion angiography showing filiform intrarenal arteries and inhomogeneous parenchyma. No functional outcome.

FIG. 3 Catheterization of lumbar arteries, allowing vizualization of the right renal artery distal to the proximal occlusion.

ACUTE OCCLUSION OF THE RENAL ARTERIES

Value of renal function

Before any treatment is started, renal function should be analyzed. In fact, all the costs and potential iatrogenic complications cannot be justified if an infarcted kidney is revascularized without a functional outcome. Analysis of the literature shows that in many cases a successful anatomical reconstruction of an occluded renal artery leads to a poor functional outcome (Table II). In contrast, certain investigations are required in order to assess whether a kidney is still functional. The associated time delay can obviously progress the renal infarction and this dilemma contributes to a blind treatment without adequate idea of prognosis in many patients. This prognosis depends on the distal perfusion pressure, the collateral network, and the level and duration of the renal artery occlusion. In 1956 Morris et al. [36] showed that with a distal perfusion pressure of 25 mmHg the kidney can survive two hours of arterial occlusion. At this pressure level, the kidney is anuric due to the insufficient filtration pressure at the level of the afferent glomerular arteries. Following reconstruction, however, an adequate arterial pressure above 60 mmHg allows recovery of the kidney. In this way, as observed in ischemic myocardium, the hibernating glomeruli can regain their function after revascularization. This also explains why severe hypertension discovered during acute renal artery occlusion carries a good prognosis be-

cause it illustrates a functional juxtaglomerular function, secreting renin [1]. In contrast, the absence of hypertension often indicates massive renal infarction without hope for recovery. This minimal residual pressure is related to the quality of the collateral network at the time of arterial occlusion. In case of acute thrombosis in a pre-existing stenosis, the developed collateral vessels in the ureter, diaphragm, and lumbar and suprarenal areas can all contribute to a residual pressure with which the kidney can be viable. Conversely, in traumatic or embolic cases, the collateral network is not developed and the antegradely stasis-induced thrombosis will rapidly reach the renal hilus, not allowing parenchymatous perfusion. This circulatory arrest immediately induces a massive, irreversible infarction. In these complete occlusions without collaterals, the delay of intervention to achieve a functional recovery can only be extremely short. In a healthy animal kidney, a two-hour ischemia is tolerated. Functional recovery occurs in two to three weeks, indicating the time to regenerate the urothelium. Beyond four hours there will be no recovery anymore [37]. In humans, renal survival beyond two hours of ischemia is very unlikely [8,38]. This fundamental role of the collateral network explains why the duration of occlusion prior to revascularization has no prognostic value [1]. It only explains, in general terms, that an acute thrombosis in a preexistent stenotic renal artery carries a better prognosis than a post-traumatic or embolic occlusion.

1 1 129

1st author [ref.] Bouttier [10] Ouriel [1] Salam [9]

Year of publication
1987 1987 1992

Number of patients

Treatment Surgery Surgery Urokinese Streptokinese Urokinese Urokinese

Succesfull angiography

Functional succes

10 13 10 14
1

8 4 3 8 0

10

7
13
1

Blum [8] Carey [13]

1993 1999

An angiographical success was obtained in 39 of 48 cases with only 23/48 functional successes

VASCULAR

EMERGENCIES

Strategy
In practice, the clinical context guides the therapeutic strategy. In case of an embolus or traumatic occlusion and if the CT angiography or arteriography show an arterial occlusion with absence of distal renal perfusion, the main decision criterion is the time interval between the occlusion and the diagnosis. If the interval exceeds three hours, as observed in the majority of series (Fig. 4), revascularization is not justified. Additional to this interval, the time of desobstruction must be added: 2 to 3 hours for surgery and 6 to 12 hours for thrombolysis. In both circumstances, the changes of success are minimal. This abstention from therapy has no major impact on the renal function if the contralateral is normal. The only exception for a desperate salvage revascularization would be in case of bilateral acute occlusion or occlusion of a single kidney, even if the time interval seems to be excessive. Only few successes have been reported after six hours [39]. In some cases, the clinical and anatomical picture is less clear. It can occur that distal to an arterial occlusion CT scanning shows a motley image with perfused and hypodense areas in the renal parenchyma. In this situation, active anticoagulation protects the kidney against further ischemia. Emergency arteriography (Fig. 5) will depict the

130

renal arterial tree and, according to the localization of the obstacle either in the common trunk or side branches, the decision is made to continue the anticoagulation treatment or perform a desobstruction. In case of acute thrombosis or even in embolization or trauma and if CT angiography or arteriography visualize kidney perfusion distal to the obstacle, the situation is completely different. Adequate anticoagulation therapy should be started and revascularization can be postponed. Some authors [40] even recommend to prepare the patients during several days in order to reduce the surgical risks resulting from pulmonary edema, hypertension or acute renal failure with oligo-anuria, occurring in 15% of patients in the experience of Lacombe [7]. Dialysis or ultrafiltration is started and the cardiac condition is optimized. This time period also allows assessment of the actual and remaining renal function. Isotope renography is the most accurate technique, assessing the overall renal function and percentage of function of the occluded kidney. Combined with the inuline clearance, it estimates the persistent glomerular filtration of the affected kidney. This isotopic analysis should be combined with duplex scanning because the size and morphology of the occluded kidney are two parameters that will determine subsequent therapy. In case of a small kidney with a lowered corticomedullar index and increased arterial resistance, the success

FIG. 4 Delay between the assumed moment of occlusion and the diasnosis, before the start of treatment ( personal experience and data from the literature). Only 7.3% of patients are diasnosed before the third hour, in the majority of cases after iatrosenic occlusion.

ACUTE OCCLUSION OF THE RENAL ARTERIES

false route or bleeding. Furthermore, extensive occlusions involving the hilus of the kidney as well as the majority of noniatrogenic traumatic cases require surgery.

SURGICAL TECHNIQUES
Embolectomy. Surgical access for renal artery embolectomy is via a subcostal laparotomy or lumbotomy. The renal artery dissection should extend to the level of the bifurcating branches in order to direct the embolectomy catheters in each branch. A transverse arteriotomy distal to the aortic ostium allows renal artery clamping once the inflow is restored. Distally, the handling of a N 2 balloon catheter should be very cautious to prevent a intrarenal arterial rupture. Once the desobstruction has been performed, it is recommended to rinse the kidney and eliminate all residual micro thrombi with 500 cc of Ringer's lactate under a pressure of 150 to 200 mmHg. A continuous flow of fluid at the end of this rinsing process is a good prognostic sign. In contrast, a minimal drop-by-drop flow suggests intraparenchymatous occlusions with subsequent secondary failure. Following local heparinization, the arteriotomy is closed with a 7/0 monofilament suture. During renal embolectomy we prefer to avoid aortotomy because distal desobstruction and rinsing is much more complicated. Furthermore, it requires suprarenal cross-clamping, compromising the contralateral kidney. Renal artery trauma. Kidney injury is often part of a severe polytraumatic framework which prohibits any thrombolytic treatment. Especially if the contralateral kidney is not affected, emergency surgical repair is not recommended in these multi-injured patients. Hemostatic difficulties after opening of the renal compartment as well as the incertainty of the result of the revascularization determine the temporizing attitude of the majority of surgeons. However, if the patient becomes hypertensive due to renal hypoperfusion by the collateral network, nephrectomy can be considered. In case of traumatic injury of an anatomical and functional monokidney, the prognosis is obviously different, requiring emergency salvage surgery. In order to achieve realistic changes of success the procedure has to be performed ex vivo [42] because of the three following major reasons. 1 - In case of renal artery subintimal rupture, antegrade extension of the thrombus can involve the different branches of the artery. Distal thrombectomy

FIG. 5 artery.

Example of distal embolization of the right renal

rate of a revascularization is minimal, especially if the individual function is less than 15% of the overall function [41]. These patients benefit most from anticoagulation therapy. In contrast, if the volume of the kidney corresponds to the posture of the patient and the corticomedullar structure is preserved, a surgical or endovascular procedure should be considered, depending on the aspect of the angiographic lesions.

131

Treatment
The desobstruction of acute renal artery occlusions has benefited from the progress in noninvasive techniques: in-situ thrombolysis with or without angioplasty, angioplasty with or without a stent, percutaneous thrombosuction and fenestration. The choice of method depends on the clinical conditions, preference and experience of the team and the type of lesion. In general, surgery should be performed in cases in which recanalization might induce complications like progress of dissection,

VASCULAR

EMERGENCIES

is very difficult to perform acurately in a contused, bleeding area. 2 - Even if the scan does not reveal any transections, a contusion or capsular fissure might induce severe bleeding after reperfusion, carrying the risk of secondary nephrectomy. 3 - Renal transections require repair by means of gluing, which must be performed in a dry environment. Furthermore, the glue must be in free air during ten minutes before any blood contact. Following cooling and rinsing with Euro-Collins, the ruptured side branches are repaired with pladget-reinforced sutures and the transections are filled with glue. The kidney is then surrounded by a hemostatic gauze in order to limit the potential bleeding after reperfusion. After the kidney has been repaired, the arterial reconstruction is performed according to the ex vivo techniques. During the whole procedure the renal area is tamponated. Finally, the kidney is autotransplanted in the lower abdominal aorta and caval vein or common iliac vessels. Acute thrombosis. If the angiography shows reinjection in the renal trunk, revascularization can be performed by means of thrombectomy, thromboendarterectomy or aortorenal bypass. The surgical access is also via laparotomy or lumbotomy. The applied material is preferentially synthetic because of the frequently necessary long-term balloon dilatations of venous grafts. In children, the use of an arterial autograft is recommended because it will grow with time, offering a permanent reconstruction. If the selective angiography shows a renogram with a poorly visible intrahilar re-injection [6], we and others believe that ex vivo repair associated with autotransplantation offers the best changes to save the kidney. The performance of several distal anastomoses in small caliber arteries is delicate and implies clamping of renal artery branches to avoid uncomfortable back-bleeding, which even further deteriorates renal ischemia in a poorly protected kidney. Ex vivo repair avoids all these disadvantages [43] and allows high-quality completion angiography [44].

THROMBOLYSIS
Selective catheterization of an occluded renal artery and local infusion of thrombolytic agents is currently the most frequently applied technique, according to numerous recently published case reports. The agents as well as the prescribed doses vary considerably: Pilmore et al. [45] apply streptokinase at 30 000 units/hour (U/h), Sternbergh et al. [46] use urokinase at 60 000 U/h, Salam et al. [9] follow a protocol with a loading dose of 250 000 units urokinase and infusion at 60 000 U/h or loading dose of 90 000 units streptokinase with continuation at 5 000 U/h. We apply a local dose of 250 000 units urokinase and continue with 2 000 U/h/kg. We perform a coagulation check-up every four hours and temporarily stop the fibrinolysis if serum fibrinogen is less than one gramme. The average duration of treatment reported in the literature differs considerably. It varies from 2 hours in the shortest treatment of partial occlusion to an average of 23 hours with extremes of 50 hours. In general, there is no consensus on the type of agent, the doses and the duration after which the treatment becomes useless. Only one single animal study compared the effect of thrombolytic agents in acute renal artery occlusions: rt/PA was more effective than urokinase; streptokinase was not evaluated.

11
132

PERCUTANEOUS THROMBOSUCTION
Percutaneous thrombosuction techniques are applied mainly in peripheral and arteriovenous fistulae desobstructions [48], whereas renal artery occlusions treated by this method are rarely reported. Most often, the attempts were technical failures or partial successes, necessitating completion by local thrombolysis. Different devices have been used. - Fogarty catheter, introduced in the renal artery via a guide wire, - thromboaspiration catheter based on the vortex effects like the Amplatz Clotbuster catheter [49], - thromboaspiration catheter based on the venturi effect like the Rheolitic Angio-jet catheter [50].

ANGIOPLASTY AND STENTING

The series in the literature addressing acute and chronic renal artery occlusions do not always differentiate between angioplasty with or without a stent. A recent publication reported on five emergency desobstructions for acute renal artery thrombosis by means of angioplasty [51]. Angioplasty is obviously completed with stent placement if the results of thrombolysis or throm-

Endovascular techniques
The advantage of endovascular techniques is the rapid availability as soon as the patient enters the radiology suite.

ACUTE OCCLUSION OF THE RENAL ARTERIES

bosuctionaire are only partially successful. Primary stenting is recommended in short thromboses with truncal visualization and secondary occlusions after previous stenting. The latter indication is rather debatable for us because of the high incidence of recurrences. In general, the applied techniques are rather similar to the methods used in percutaneous treatment of renal artery stenoses. Finally, it would be interesting to compare the techniques of desobstruction as a function of the delay of treatment between the moment of obstruction and the end of the revascularization. In practice, the published data are too vague. Most often the exact moment of occlusion is not known and the time intervals vary considerably (from 24 hours to 2 weeks), which is probably caused by the chronic ischemic process. In an attempt to assess the results, we have collected the published cases and related them with each technique, obviously being a debatable methodology. We have identified 149 renal artery emergency desobstructions in the literature between 1972 and 2002. Our personal experience comprises 18 cases, adding up to 167 kidneys in 155 patients, 12 of whom who suffered from bilateral occlusion. In 98 kidneys the technique was a surgical procedure and in 69 kidneys an endovascular approach was chosen. In the surgical series, 51 of 98 kidneys were classified as functional [52], which was similar to the endovascular results (36/69; 52%). Mortality was 9.2% in the surgical group and 1.4% in the endovascular one. Our personal experience depicts comparable results with a functional kidney in 47% of patients. If the surgical outcome is assessed as a function of etiology, the emboli seem to carry the best prognosis (68% of salvaged kidneys), followed by acute thrombosis (63%). Traumatic occlusions have a poor prognosis (10%). Our personal experience slightly differs, showing a better prognosis in cases with acute thrombosis (Table I). Thrombolysis followed by angioplasty with stenting offers the best endovascular results, with a success rate of 55%. Thrombolysis as a monotherapy is effective in 50% of cases; this is similar to the results of angioplasty. Thrombosuction has only been described in six cases: three failures, two partial successes, one complete success.

FENESTRATION OF AORTIC DISSECTIONS

Some thoracic aortic dissections can cause acute occlusion of one or both renal arteries by extension of the false lumen, the mechanism of which is variable. The intimal flap can behave as an obstructing valve, occluding the renal artery. Otherwise, thrombosis of the false lumen might extend to the ostium and obliterate the artery. In any cases, the goal of treatment is to re-install communication between the renal arterial trunk and the circulating lumen. The most common applied technique consists of perforating the intimal flap by means of a balloon catheter and bursting the flap by inflating the balloon, re-establishing communication between the true and false lumen and reperfusing the kidney (Fig. 6). If the reperfusion is inadequate, a stent can be deployed in the renal artery. The proximal part of the stent is left unattached in the aortic circulating lumen. This technique generally guarantees adequate revascularization [52]. These techniques have recently been developed and longterm outcome of these stents positioned between the aorta and renal artery is unknown.

133

For several reasons, the reported results in the literature only have a limited value. The majority of publications comprise one, two or three patients and most often only report on therapeutic success, not indicating whether failures occurred as well. The smaller series report on different etiologies with variable diagnostic and therapeutic measures. It is rather difficult to compare the outcome of an injured kidney in a young adult with renal artery embolization in an old person due to arrhythmia. The most extensive series date back to the 1980s and 1990s with a compilation of data spread over 15 to 20 years. This implies that several patients have been treated in the 1970s with limited means, before arrival of CT scanning, duplex, scintigraphy, and endovascular and thrombolytic therapies.

Conclusion
Acute renal artery occlusions have a worse prognosis than chronic occlusions. Improvement of the prognosis is only possible if the delay between the moment of occlusion and the final diagnosis is reduced to a minimum. If feasible, the endovascular techniques show similar results as compared to surgery, with a distinctly lower morbidity and mortality.

VASCULAR

EMERGENCIES

II

134

FIG. 6 A - Type B dissection with extension in the abdominal aorta, with occlusion of the right renal artery. B - Catheterization of the right renal artery (via femoral access). Note that the catheter is not in the true lumen. C - Restored patency of the right renal artery after fenestration and placing a stent in the right renal arterial trunc and the true lumen.

ACUTE OCCLUSION OF THE RENAL ARTERIES

R E F E R E N C E S

1 Ouriel R, Andrus CH, Ricotta JJ et al. Acute renal artery occlusion: when is revascularization justified? / Vase Surg 1987; 5:348-355. 2 Hoxie H, Coggin C. Renal Infarction. Arch Intern Med 1940; 65: 587-594. 3 Domanovits H, Paulis M, Nikfardjam M et al. Acute renal infarction. Clinical characteristics of 17 patients. Medicine 1999; 78:386-394. 4 Fu GY, Candela RJ, Mishkind M et al. Bilateral renal artery occlusion: an unusual presentation of atrial fibrillation and hypertrophic cardiomyopathy. Clin Cardzo/1994; 17: 631-633. 5 Lessman RK, Johnson SF, Coburn JW, Kaufman JJ. Renal artery embolism: clinical features and long-term follow-up of 17 cases. Ann Intern Med 1978; 89: 477-482. 6 Stables DP, Fouche RF, de Villiers van Niekerk JP et al. Traumatic renal artery occlusion: 21 cases./ Urol 1976; 115: 229-233. 7 Lacombe M. Acute non-traumatic obstructions of the renal artery. J Cardiovasc Swrgl992; 33:163-168. 8 Blum U, Billmann P, Krause T et al. Effect of local low-dose thrombolysis on clinical outcome in acute embolic renal artery occlusion. Radiology 1993; 189:549-554. ' 9 Salam TA, Lumsden AB, Martin LG. Local infusion of fibrinolytic agents for acute renal artery thromboembolism: report of ten cases. Ann Vase Surg 1993; 7: 21 - 26. 10 Bouttier S, Valverde JP, Lacombe M et al. Renal artery emboli: the role of surgical treatment. Ann Vase Surg 1988; 2:161 -168. 11 Fogarty TJ, Buch WS. The management of embolic and thrombotic arterial occlusions. In: Rutherford RB (ed). Vascular surgery. Philadelphia, W.B. Saunders Compagny, 1977: pp 423-431. 12 Jones RE, Tribble CG, Tegtmeyer CJ et al. Bilateral renal artery embolism: a diagnostic and therapeutic problem. J Vase Surg 1987; 5: 479-482. 13 Carey HB, Boltax R, Dickey KW, Finkelstein FO. Bilateral renal infarction secondary to paradoxical embolism. Am J Kidney Dis 1999; 34: 752-755. 14 Hoover LA, Hall-Craggs M, Dagher FJ. Polyarteritis nodosa involving only the main renal arteries. Am J Kidney Dis 1988; 11:66-69. 15 Fukuda T, Hayashi K, Sakamoto I, Mori M. Acute renal infarction caused by Behcet's disease. Abdom Imaging 1995; 20: 264-266. 16 Saint F, Quintela R, Salomon L et al. Acute renal artery occlusion in a 15-year-old girl taking oral contraceptives. BJUInt 2002; 89: 787-788. 17 Lambrecht GL, Malbrain ML, Coremans P et al. Acute renal infarction and heavy marijuana smoking. Nephron 1995; 70: 494-496. 18 Goodman PE, Rennie WP. Renal infarction secondary to nasal insufflation of cocaine. AmJEmergMed 1995; 13: 421 -423. 19 Peddi VR, Kant KS. Catastrophic secondary antiphospholipid syndrome with concomitant antithrombin III deficiency. J Am Soc Nephrol 1995; 5: 1882 -1887. 20 Mysiak A, Kobusiak-Prokopowicz M, Kaiser T, Lipinska M. Thrombophilia complicated by multiple thrombo-embolic lesions in the arterial system - case report. Kardiol Pol 1997; 47:492-495. 21 Janssen van Doom K, Van der Niepen P, van Tussenbroeck F, Verbeelen D. Acute tubulo-interstitial nephritis and renal infarction secondary to ergotamine therapy. Nephrol Dial Transplant 2000; 15:1877-1879. 22 Mohallem SV, Ramos SG, dos Reis MA et al. Prevalence of renal infarcts in autopsies of chronic Chagas disease patients. Rev Soc Bras Med Trop 1996; 29: 571 - 574. 23 Greenberg R, Waldman D, Brooks C et al. Endovascular treatment of renal artery thrombosis caused by umbilical artery catheterization. / Vase Surg 1998; 28: 949 - 953. 24 Frank RG, Peyser D, Herasme V. Global renal infarction secondary to a dissecting thoracic aneurysm. Urology 1996; 48: 930-931.

25 Strichartz SD, Gelabert HA, Moore WS. Retrograde aortic dissection with bilateral renal artery occlusion after repair of infrarenal aortic aneurysm. JVasc Swrgl990; 12 : 56-59. 26 Lok SY, Chalvardjian A, Common AA. Primary renal artery dissection. CanAssocRadiolJ\9%; 46: 54-56. 27 Haas CA, SpirnakJP. Traumatic renal artery occlusion: a review of the literature. Tech Urol 1998; 4:1 -11. 28 Brown MF, Graham JM, Mattox KL et al. Renovascular trauma. AmJ Surg 1980; 140: 802 -805. 29 Gallucci M, Alpi G, Cassanelli A et al. Renal infarction secondary to selective arteriography. Rays 1985; 10: 95-97. 30 Rees CR, Palmaz JC, Becker GJ et al. Palmaz stent in atherosclerotic stenoses involving the ostia of the renal arteries: preliminary report of a multicenter study. Radiology 1991; 181: 507-514. 31 Gasparini M, Hofmann R, Stoller M. Renal artery embolism: clinical features and therapeutic options./ Urol 1992; 147: 567-572. 32 Winzelberg GG, Hull JD, Agar JW7 et al. Elevation of serum lactate dehydrogenase levels in renal infarction. JAMA 1979; 242:268-269. 33 Liano F, Gamez C, Pascual J et al. Use of urinary parameters in the diagnosis of total acute renal artery occlusion. Nephron 1994; 66:170-175. 34 Amilineni V, Lackner DF, Morse WS, Srinivas N. Contrastenhanced CT for acute flank pain caused by acute renal artery occlusion. AJR Am J Roentgenol 2000; 174:105 -106. 35 Martin KW, McAlister WH, Shackelford GD. Acute renal infarction: diagnosis by Doppler ultrasound. Pediatr Radial 1988; 18: 373-376. 36 Morris GC, Heider CF, Moyer JH. The protective effect of subfiltration arterial pressure on the kidney. Surg Forum 1956; 6:623-627. 37 Hamilton PB, Phillips RA, Hiller A. The duration of ischemia required to produce uremia. AmJ Physiology 1948; 152: 517-521. 38 Schefft P, Novick AC, Stewart BH, Straffon RA. Renal revascularization in patients with total occlusion of the renal artery. JUrol 1980; 124:184-186. 39 Barral X, Farcot M, Favre JP et al. Complications vasculaires de la transplantation renale. EncylMed Chir (Editions Scientifiques et Medicales Elsevier SAS, Paris) Techniques chirurgicales Chirurgie vasculaire, 43.310,1989: 21 p. 40 Higgins RM, Goldsmith DJ, Charlesworth D et al. Elective rather than emergency intervention for acute renal artery occlusion with anuria. Nephron 1994; 68: 265-267. 41 Barral X, Gournier JP, Favre JP. Resultats tardifs de la revascularisation des occlusions chroniques de 1'artere renale. In: Cormier JM, Cormier F, FichelleJM, MarzelleJ (eds). Techniques et strategies en chirurgie vasculaire. Paris, Pharmapost, 1995: pp 265-272. 42 Barral X, Lorin S, Grandmougin D, Favre JP. Chirurgie de 1'artere renale. Encyl Med Chir (Editions Scientifiques et Medicales Elsevier SAS, Paris) Techniques chirurgicales Chirurgie vasculaire, 43.110.C, 2002:18 p. 43 Barral X, Gournier JP, Favre JP. Chirurgie ex-vivo des arteres renales. In: Kieffer E (ed). Chirurgie des arteres renales. Paris, AERCV, 1993: pp 205-214. 44 Barral X, Favre JP, Gournier JP. Le controle arteriographique peroperatoire des arteres renales. In: Branchereau A, Magnan PE. Methodes de controle peroperatoire des restaurations vasculaires. Marseille, CVN, 1994: pp 33-38. 45 Pilmore HL, Walker RJ, Solomon C et al. Acute bilateral renal artery occlusion: successful revascularization with streptokinase. AmJ Nephrol IMS; 15: 90-91. 46 Sternbergh WC 3rd, Ramee SR, DeVun DA, Money SR. Endovascular treatment of multiple visceral artery paradoxical emboli with mechanical and pharmacological thrombolysis. JEndovasc 77^2000; 7:155-160.

135

VASCULAR

EMERGENCIES

47 Haberstroh J, Wagner G, Kiefer T et al. Renal artery occlusion model in dogs for the evaluation of thrombolytic agents. J Invest Surg 1997; 10:183-188. 48 Rousseau H, Otal P, Colombier D et al. Percutaneous arterial thrombectomy. Endovasular Impact 1997; 2: 46-50. 49 Coleman CC, Krenzel C, Dietz G\ et al. Mechanical thrombectomy: results of early experience. Radiology 1993; 189: 803-805. 50 Van Omrnen VG, van der Veen FH, Geskes GG et al. comparison of arterial wall reaction after passage of the Hydrolyser device

versus a thrombectomy balloon in an animal model./ Vase Interv ' 51 Dwyer KM, Vrazas JI, Lodge RS et al. Treatment of acute renal failure caused by renal artery occlusion with renal artery angioplasty. Am J Kidney Dis 2002; 40: 189-194. 52 Beregi JP, Cocheteux B, Koussa M et al. Traitement endovasculaire des malperfusions au cours des dissections aortiques. In: Kieffer E, Fabiani JN. Chirurgie des dissections aortiques. Paris, AERCV, 2002: pp 231 -239.

136

14
ACUTE INTESTINAL ISCHEMIA
BRANDON KRIJGSMAN, GEORGE HAMILTON
Intestinal ischemia is an uncommon condition presenting particular problems of diagnosis and management. The prevalence of the disease is difficult to establish. In the United Kingdom, approximately 2000 deaths a year are attributable to intestinal vascular insufficiency, with 1883 deaths in 2000 [1]. At least 833 (44%) were classified as acute (834 being unspecified as either acute or chronic). Women are more often affected than men by a ratio of 2:1. The elderly are more commonly affected, the incidence being rare below fourty five years of age, and with the majority of deaths occurring after seventy five years of age. Most cases are caused by emboli (64%), followed by arterial thrombosis (27%), venous thrombosis (3.5%), and nonocclusive mesenteric ischemia (4.8%) [2]. Mortality is high and has changed little since the 1970s, despite interventional advances (Table I).

11
137

Pathophysiology
The arterial circulation to the gut has extensive collaterals and arcades providing multiple sources of blood inflow (Fig. 1). This explains why vascular occlusion is well tolerated as evidenced by the relative lack of clinical intestinal ischemia despite the high prevalence of atherosclerotic disease of the aorta and visceral arteries. Certain collateral patterns are recognized, depending on which artery is blocked. When either the celiac or superior mesenteric artery (SMA) is compromised, the main collateral circulation is by the gastroduodenal and pancreaticoduodenal arteries. The main collateral channels between the SMA and inferior mesenteric

artery (IMA) occur in the region of the splenic flexure between the middle and left colic arteries. In the presence of either SMA or IMA occlusion, the marginal artery of Drummond and the arch of Riolan (an ascending branch of the left colic artery anastomosing with branches of the SMA) enlarge significantly. In the presence of an IMA occlusion, another important collateral circulation is between the internal iliac artery and the left colic artery via the superior hemorrhoidal arteries. The SMA is the critically important vessel in maintaining visceral perfusion, as demonstrated by increased blood flow after eating. This is not seen in the celiac artery. In chronic ischemia, all patients have SMA stenosis or occlusion, in addition to celiac artery and/or IMA involvement (Table II) [27-29].

VASCULAR

EMERGENCIES

1st author [ref.] Czerny [2]

Cho [3]

Year

Number of patients
145

Mortality rate %
26 52 32 48 24 81 60 35 59 66 80 72 96 45 70 68 40 51 40 69 71 92 27 82 77 58

Arteries
SMA, CA, IMA SMA, CA SMA, IMA

% of patients involved

1997

57 25 14
4

2002 2002
2001

48 58
170

Park [4] Endean [5] Foley [6] Mamode [7] Newman [8] Urayama [9] Duron [10] Klempnauer [11] Bronner [12]

SMA only SMA occlusion

CA: celiac artery IMA: inferior mesenteric artery SMA: superior mesenteric artery

2000
1999 1998 1998 1998 1997 1997 1996 1996 1995 1995 1992 1990 1990 1990 1989 1988 1987 1986 1984 1982

21 57 98 34
492

43

90 20 47 28 34 43
100

14
138

Voltolini [13] Konturek [14] Ward [15] Deehan [16] Inderbitzi [17] Levy [18] Batellier [19] Bapat [20] Finucaine [21] Sitges-Serra [22] Wilson [23] Lazaro [24] Andersson [25] Sachs [26] Total

62 65 20 32 83
102

About 10% to 20% of the cardiac output flows through the visceral circulation (500 to 1200 mL/ min.). Although intestinal mucosa forms one half of the intestinal tissue mass, it receives a disproportionately large 75% of resting blood flow. All causes of acute intestinal ischemia result in prolonged hypoxia with persistent severe pain out of proportion to the clinical findings. This is associated with early development of acidosis, hyperamylasemia and leucocytosis. The ischemic damage to the mucosa results in the loss of the mucosal barrier. This allows translocation of bacteria, endotoxins and cytokines with major systemic effects such as septicemia and multiorgan failure. Successful reperfusion releases free radicals causing an inflammatory response with release of many cytokines, activated leucocytes and inflammatory mediators. This results in local mucosal and hepatic inflammation and systemic effects leading to multiorgan failure [27].

23 60 30

Etiology
OCCLUSIVE DISEASE
Emboli. The SMA is the most common site of embolic occlusion although the celiac artery can be affected. There is classically an underlying cardiac problem giving rise to the organized thrombus that

1962

ACUTE INTESTINAL

ISCHEMIA

If
139

FIG. 1 A - Schematic representation of the collateral circulation of the intestine. B - Angiosraphic appearance of arch of Riolan from superior mesenteric artery (stented at its origin). C - Angiographic appearance of marginal artery of Drummond. D - Initial angiogram demonstrates occluded IMA. The delayed film shows the colonic supply.

VASCULAR

EMERGENCIES

14 140

embolizes. This is usually atrial fibrillation or less commonly a mural thrombus from an acute myocardial infarction. A history of previous embolic events is not uncommon. Other causes of emboli include iatrogenic intra-aortic manipulations, paradoxical emboli through a septal defect, atrial myxoma or primary aortic tumors [30]. The history is of constant severe epigastric or periumbilical pain of sudden onset. It is frequently followed by copious vomiting and explosive diarrhea. Typically the patient has been previously well and asymptomatic. The abdominal signs are often lacking or nonspecific, with distension in association with absent or normal bowel sounds without any signs of peritonism. This combination of severe abdominal pain out of proportion to the clinical findings is typical. Peritonism or blood in the stool or vomitus indicates severe advanced intestinal ischemia with likely infarction and is generally a late clinical feature. The presence of proximal SMA pulsation and the distribution of intestinal ischemia are intra-operative clues for an embolus. The occlusion in embolism is usually distal to the origin of the pancreaticoduodenal and middle colic branches, which allows some blood flow to the small intestine to be maintained. The stomach, duodenum, and proximal jejunum are normal with ischemia extending to the mid transverse colon. Thrombosis. Superior mesenteric arterial thrombosis may occur as the result of progression of SMA stenosis that had not previously been diagnosed or treated. There is often a history of intestinal or food fear with severe weight loss, the hallmark of chronic intestinal ischemia in about 65% of patients [4]. The typical patient is female and a heavy smoker, often with evidence of widespread arterial disease including previous myocardial infarction or daudication. As with embolic occlusion, the combination of severe abdominal pain out of proportion to the clinical findings is typical. The thrombosis of the SMA occurs at the origin of the artery. In contrast to embolic disease, the proximal SMA pulse is absent and the distribution of intestinal ischemia is more extensive. Only the stomach, duodenum and distal colon are spared. In the young patients, fibromuscular dysplasia can cause mesenteric arterial thrombosis with equally devastating results [31]. Intravenous cocaine abuse is another increasing problem accounting for intestinal ischemia in the young patients. The extent of intestinal ischemia and infarction tends to be focal

and less than that seen with atherosclerotic thrombosis [32]. The mechanism of ischemia appears to be occlusive rather than due to vasospasm. Mesenteric ischemia should be considered in the differential diagnosis when evaluating a young patient with a history of cocaine abuse presenting with an acute abdomen. Some prothrombotic states such as hyperhomocysteinemia or the 20210 A prothrombin gene mutation have resulted in primary arterial thrombosis [33,34]. Mesenteric venous thrombosis (MVT) is rare and accounts for 5% to 15% of all acute mesenteric ischemia. It is classified as primary (where no cause is recognized) or secondary. Secondary MVT may follow hypercoagulable states, portal venous stasis and hypertension, intra-abdominal infection and inflammation or malignancy, use of oral contraceptives and splenectomy (see Table III). Long-term anticoagulation is required for MVT, because of the high recurrence rates. The clinical presentation is usually less acute than that of arterial occlusion. Severe but vague abdominal pain that tends to be colicky and slowly progressive is usually present. Few abdominal signs are present except tenderness, distension and decreased bowel sounds. The pain is out of proportion to the physical findings. Fecal occult blood is present in the majority of patients. There is a pyrexia of greater than 38 C in 25% to 50% of patients, and 20% have a tachycardia [35, 36]. Leucocytosis ranges from 12000 to 29000 [37]. Frank peritonitis is seen only when transmural infarction or perforation has occurred. Surgical findings include blood-stained free peritoneal fluid at laparotomy. The affected bowel is cyanotic and edematous with a rubbery texture. Mesenteric arterial pulsations are present but the veins contain fresh thrombus that extrude when the veins are cut. Infarction is most common in the mid small bowel. Other occlusive causes. Other causes for mesenteric occlusion include aortic dissection or isolated SMA dissection [38]. Vascular toxicity following chemotherapy with 5-FU, cisplatin and vincristine has been reported [39], as has been occlusion by Schistosoma mansoni infection [40]. NONOCCLUSIVE MESENTERIC ISCHEMIA Mesenteric ischemia can occur without arterial occlusion. It can develop in severe systemic illness in association with cardiorespiratory shock and multiorgan failure. Typically these patients are on critical

ACUTE INTESTINAL
care units. Shock, either cardiogenic or secondary to septicemia, is often present. The physiologic response of mesenteric vasoconstriction in response to both normovolemic and hypovolemic shock appears to be mediated by the renin-angiotensin axis [41]. It is usually aggravated by the use of inotropic agents. It is a distinct complication of open heart surgery with an incidence of 0.2% to 0.4 % although its mortality rate is 70% to 100% [42]. Long-term dialysis patients are another group of susceptible patients who may suffer from nonocclusive mesenteric ischemia (NOMI). A major charac-

ISCHEMIA

teristic finding is an episode of severe hypotension or hypovolemia immediately before the onset of abdominal symptoms [43,44]. Abdominal pain is the most common symptom, although in the unconscious patient abdominal distension, gastrointestinal bleeding, leucocytosis or fever may be the presenting features. The mortality rate is approximately 65% [45].

IATROGENIC INTESTINAL ISCHEMIA

Intra-aortic manipulations such as interventional radiological procedures or intra-aortic balloon pumps post cardiac surgery can result in this condition. Embolization can occur following pulsatile cardiopulmonary bypass [46]. Embolization is usually widespread involving the kidneys, pelvis and lower limbs, as well as the viscera, and carries a very high mortality. The correct treatment is not known but should probably include full intravenous anticoagulation, appropriate resuscitation and infusion of prostacyclin or its analogues. Embolization due to cholesterol may be suspected by an eosinophilia in the blood film. This may occur in patients undergoing systemic thrombolysis or anticoagulation with aortic manipulation. Anticoagulation is relatively contraindicated and the use of statins would seem beneficial. Left colonic ischemia can occur following aortic reconstruction after interruption of the direct or collateral blood supply. It is more common following aneurysm repair after ligation of a patent IMA. The accepted incidence is 2% (ranging from 0.2% to 10%). Mortality in this condition is 40% to 50%, and approaching 90% if there is full bowel wall thickness ischemia [47]. Diagnosis can be difficult in the postoperative period and depends on raised clinical awareness. Watery diarrhea, bloody or not, should prompt urgent bedside colonoscopy. A leucocytosis greater them 20000, fever and shock, and severe metabolic acidosis should alert the surgeon to the possibility of severe colonic ischemia.

Hypercoagulable states Neoplasms Deep venous thrombosis Pregnancy Factor V Leiden Protein C deficiency Protein S deficiency Homocysteine Antithrombin III deficiency Anticardiolipin antibodies 20210 A prothrombin gene mutation Hyperfibrinogenemia Myeloproliferative disorders Sickle cell disease Abdominal pathology Malignancy Splenectomy Splenomegaly Estrogen containing oral contraceptives Inflammation/infection Pancreatitis Inflammatory bowel disease Intra-abdominal abscess Mechanical venous occlusion Portal hypertension Vohoilus Intussusception

141

Diagnosis
Acute intestinal ischemia is a life-threatening surgical emergency, yet can be a difficult diagnosis to make, with delay contributing directly to infarction. The majority of cases are diagnosed more than 12 hours after the onset of symptoms [16]. Delayed diagnosis accounts for the majority of malpractice claims involving acute mesenteric ischemia in the

MVT: mesenteric venous thrombosis

VASCULAR

EMERGENCIES

14

142

United States [48]. Diagnosis depends on a high index of suspicion. The main presenting feature is the combination of severe abdominal pain out of proportion to the clinical findings, as discussed above. Serum levels of lactate and leucocytes are elevated in the majority (65% to 90%) of patients to greater than 50 U/L and 15000/mL, respectively [2,49-51]. Hyperamylasemia is seen in just under half the patients with acute mesenteric ischemia [50]. Elevation of serum inorganic phosphate levels have been proposed as a marker of mesenteric ischemia, as it is extensively found in gut, but this only occurs in 15% to 33% of such patients [50,51]. However, in those patients who did have elevated phosphate levels, it predicted extensive injury and poor prognosis [52]. The fibrinolytic marker D-dimer is elevated in thrombo-embolic occlusion of the SMA, although levels are also raised in other conditions of acute bowel ischemia such as strangulation or ruptured aortic aneurysm [53]. Animal studies have suggested intestinal fatty acid binding protein (I-FABP) as a serum marker reflecting bowel ischemia. Early human studies show promise, as patients with ischemic bowel disease demonstrate significantly higher I-FABP levels than either healthy subjects or patients with acute abdominal pain. Patients with mesenteric infarction had the highest serum I-FABP levels [54]. Plain radiographs of the abdomen may reveal nonspecific bowel dilatation or, in MVT, wall edema (thumbprinting); or gas in the bowel wall or portal vein. Unfortunately they are not helpful in most cases. Mesenteric angiography will confirm the diagnosis of arterial occlusion but at the cost of delay in treatment. If there are clear abdominal signs of peritonitism, urgent laparotomy without angiography is the best course of action. In the remainder of patients suspected of acute intestinal ischemia without abdominal signs, angiography is indicated with lateral views of the visceral aorta and its branches. In acute SMA thrombosis, there is usually no visualization of the entire artery because of the ostial nature of the disease, although delayed views may show slow filling of the distal SMA. SMA embolization usually allows visualization of the proximal artery to just beyond the level of the middle colic artery. Although noninvasive imaging modalities such as computed tomography (CT), magnetic resonance

imaging (MRI) and ultrasound can evaluate the aorta and the origins of splanchnic arteries, selective angiography of mesenteric arteries is still the gold standard in diagnosing peripheral splanchnic vessel disease in NOMI [55]. Angiography will show intestinal arterial vasospasm and, most importantly, exclude a significant arterial lesion. Abdominal CT scanning has led to the correct diagnosis in 80% of patients presenting with acute intestinal ischemia in some centers [2]. Findings supporting acute mesenteric ischemia include arterial or venous thrombosis, intramural gas, portal venous gas, a focal lack of bowel wall enhancement and liver or splenic infarcts. The newer generation of multislice CT scanners allow a more detailed study of the small bowel and mesenteric vessels which has, in some authors' experience, eliminated the need for additional imaging studies such as angiography [56]. MVT can be diagnosed on contrast enhanced CT scanning by demonstration of thrombus within the superior mesenteric vein [57]. Venous infarcted bowel is clearly demonstrated by CT when other signs for MTV such as ascites, bowel wall thickening, bowel dilatation and pneumatosis intestinalis are present. It would seem that this is the imaging modality of choice in MVT, as it seems to be a more sensitive test than angiography [58]. While it has good results in the hands of its advocates [59], duplex scanning is impaired by the increased intestinal gas that is frequently present in this condition. Magnetic resonance angiography will reliably demonstrate the proximal mesenteric vessels, but presently imaging of the more distal branches and occlusions is not good.

Treatment
NONSURGICAL

In all cases, the patient should be initially resuscitated, given broad-spectrum intravenous antibiotics and fully heparinized. As yet, the twin goals of mesenteric revascularization and resection of nonviable bowel can only be achieved by surgical means. Surgery is indicated in all patients with peritonitis. Angiography in patients without peritonitis may demonstrate NOMI or MVT. In NOMI, treatment is nonoperative and depends on optimizing cardiac output and treating underlying conditions such as sepsis. Intramesenteric arterial infusion of papaver-

ACUTE INTESTINAL ine at a dose of 30 to 60 mglr1 may be beneficial. Up to 65% of patients who have undergone cardiac surgery have had symptomatic improvement within hours when diagnosed early [60]. If MVT is diagnosed at angiography, intra-arterial thrombolytic therapy has been given successfully [61]. Nonoperative management by full anticoagulation for acute MVT is feasible when the initial diagnosis is certain and when the bowel infarction has not led to transmural necrosis and bowel perforation. The morbidity, mortality, and survival rates are similar in cases of surgical and nonoperative management [62]. Other reported endovascular procedures for acute intestinal ischemia include fenestration and stent placement in aortic dissection [63,64], angioplasty and stenting in an acute occlusion in a patient with chronic mesenteric insufficiency [65,66], and angioplasty alone [67]. This contrasts with an increasing use of angioplasty for chronic mesenteric ischemia [28].

ISCHEMIA

proximal flow cannot be established, SMA thrombosis is likely and reconstructive surgery will be required. SMA reconstruction. Revascularization can be performed either using bypass grafting from the aorta to the patent SMA or by re-implantation of the healthy portion of the SMA into the aorta. In the presence of perforated bowel or infarction requiring bowel resection, prosthetic grafts must not be used. Reversed saphenous vein aortomesenteric grafting or direct SMA re-implantation is the procedure of choice in this situation (Fig. 2). Bypass with vein or prosthesis is prone to kinking due to its configuration, and great care must be taken to align the grafts to avoid this complication. Single vessel revascularization is usually adequate in the emergency [6] and probably also in the nonemergency situation [4,68]. Assessment of intestinal viability. Determination of which portions of bowel are nonviable can be difficult especially when there is extensive infarction. The decision about how much to resect can be crucial to the long-term outcome. Clinical assessment by detecting pulsation in the arcades, color of the bowel, peristalsis and bleeding from cut edges is most commonly used. This is often complemented by the use of the doppler probe to detect flow in the intestinal wall in addition to flow in the arcade vessels. Other techniques include the use of fluorescein and inspection under Wood's lamp, pulse oximetry and laser doppler flowmetry. Ischemic bowel is resected with the aim of preserving as much bowel as possible, particularly where a short gut syndrome may be created. Thus, several segmental resections with multiple anastomoses may be necessary. Contemporary practice is to leave all marginally viable intestine after revascularization, providing there is re-inspection of the bowel after 24 to 48 hours. In cases of MVT, resection of infracted bowel with liberal margins should be performed with a primary anastomosis if perfusion is adequate. Venous thrombectomy has poor results with a high recurrence rate and is rarely indicated. Post aneurysm repair colon ischemia requiring surgery usually necessitates a Hartmann's procedure, a primary anastomosis being contraindicated. Consideration must be made to protect the aortic graft and its limbs from infection or contamination by using antiseptic soaked swabs. Exposed grafts at the end of the procedure should be covered by a wellperfused omental pedicle flap.

SURGICAL
Laparotomy is indicated in patients with peritonitis after rapid resuscitation. The first step is to assess the degree and extent of bowel viability. Free, foul smelling peritoneal fluid is a sign of advanced necrosis even if perforation has not occurred. Ischemic bowel has a characteristic appearance with loss of its normal sheen. It is dull, gray in color and flabby in tone without any peristalsis. Infarcted bowel is purplish black in color, often friable and perforated. In many cases the bowel ischemia will be so extensive and advanced that no further surgical treatment is undertaken and palliative care given. Where there is hope of sufficient bowel viability, revascularization should be performed before any bowel resection is considered. After successful revascularization, previously precarious segments of intestine may recover and resection of clearly ischemic bowel can then take place. SMA embolectomy. The proximal portion of the SMA is dissected free from the surrounding fat and lymphatic tissue just as it emerges from the pancreatic neck into the base of the mesentery. Approximately 3 to 4 centimeters of artery is cleared, with care taken not to damage the branches. Heparin (5000 units) is given intravenously. A transverse arteriotomy is made and a 3F or 4F embolectomy catheter is passed proximally and distally to clear the embolus and reestablish vigorous pulsatile flow. If

143

VASCULAR

EMERGENCIES

144

FIG. 2 A - Schematic representation of revascularization of the SMA with: bypass taking care to avoid kinking and obstruction B - Or re-implantation of SMA into the aorta. C - Angiographic appearance of aorto-SMA bypass with vein graft. There is co-existing left common iliac occlusion. D - Angiographic appearance of re-implanted SMA into aorta, which has a small saccular aneurysm at the site of occluded vein graft (aortoceliac bypass).

ACUTE INTESTINAL
Postoperative management. This is especially important in those patients who have undergone extensive bowel resection. Fluid losses (especially potassium and magnesium) must be carefully monitored and replaced. Total parenteral nutrition may be started early in the postoperative period, and may need to be continued for months in cases with short gut syndromes. Abdominal compartment syndrome has been reported following revascularization of chronic mesenteric ischemia [69] and should be considered in acute revascularization, especially if there has been limited or no resection. A second look laparotomy 24 to 48 hours after the initial procedure is mandatory to assess the viability of the marginally perfused parts of the bowel and to check the intestinal anastomoses. Further laparotomies are indicated until the viability of the bowel is established. In cases where major intestinal resection with primary anastomosis has taken place, there is an increasing vogue to performing a second look laparoscopy under local anesthesia. This is estimated to reduce the number of unnecessary laparotomies by two thirds [70]. Continued intensive care is required with optimization of cardiac and respiratory status. This is in part a response to intestinal reperfusion and may lead to multiorgan failure, and patients undergoing endovascular revascularization should be cared for similarly. There is an associated deterioration in hepatic function with transaminases rising 90- to 100-fold [27]. The hepatic impairment and associated coagulopathy is usually transient, returning to baseline within seven to ten days. In cases of MVT, long-term anticoagulation is required because of the high recurrence rates. Complications such as ascites and portal hypertension may require specific treatment [37].

ISCHEMIA

Prognosis
Regardless of etiology, prognosis depends crucially on rapid diagnosis and institution of treatment to prevent or at least minimize bowel infarction. In one study, mortality was 86% if the diagnosis was made more than 24 hours after onset of symptoms compared to 50% if diagnosis was within 24 hours [71]. The presence of peritoneal signs results in a worse mortality of 82% compared to 33% when

abdominal signs are absent. Extensive intestinal infarction in this series had a 100% mortality rate whereas early revascularization in limited bowel infarction to less than one meter only had a mortality of 18%. Circulatory collapse was another prognostic indicator with a 100% mortality compared with 50% in patients with a systolic blood pressure greater than 100 mmHg [71]. Studies, generally retrospective, demonstrate the worse prognosis of acute compared to chronic mesenteric ischemia. Thirty-day postoperative mortality is approximately 52% to 70% (compared to less than 5%), the majority due to bowel infarction [3,16]. Long-term patency and symptom-free survival can be expected after successful mesenteric arterial reconstruction for acute mesenteric ischemia. Late survival rates have been reported as 54% to 73% at 5 years and 20% at 10 years. Excluding perioperative deaths, the probability of long-term survival does not appear to differ between acute and chronic mesenteric ischemia [2,3,5]. A significant decrease in the overall mortality rate, from 77% to 59%, has been observed in France over the past decade [10]. This was a retrospective multicenter study by the French associations of surgical research comparing the periods from 1980 to 1985 and 1990 to 1995. The main improvement was seen in the patients' pre-operative condition. Mortality rates decreased from 83% to 63% in thrombotic acute mesenteric ischemia and from 51% to 19% in MVT. There were no significant increases in frequencies of angiography, vascular or second look procedures. Long-term follow-up of patients with chronic mesenteric ischemia has shown that recurrent symptoms at three years were more common if revascularized by endovascular means compared with surgery (34%, 95% CI 14%-54% versus 13%, 95% CI 6%-21%) [72] (Fig. 3). Most of these symptoms occurred in the first year (28% recurrence at 1 year and 34% at 3 years). Mortality was similar with both treatments in this group of patients with chronic symptoms. The literature suggests that there is a mean technical success rate of 91% (8%) with an immediate pain relief rate of 79% (9%). This is associated with a complication rate of 18% (15%) and a mortality rate of 4%. There is no reason to believe that endovascular methods would confer any symptom or survival advantage in patients with acute symptoms unless there has been no progression to intestinal gangrene.

if
145

VASCULAR

EMERGENCIES

Image Not Available

FIG. 3 Cumulative freedom from recurrent symptoms in chronic mesenteric ischemia. At 3 years, recurrent symptoms were more common in the sroup treated by endovascular means (34%) compared to open sursery (13%)[72], Reproduced with permission from the Journal of Vascular Surgery, Mosby, Inc.

Conclusion
Acute intestinal ischemia remains a surgical emergency with a high mortality and the key to its management is prompt diagnosis. Current treatment is by laparotomy, revascularization and excision of dead bowel. Some acute occlusions can be treated endovascularly with assessment of bowel viability by minimally invasive means.

14
146

R E F E R E N C E S 1 Anonymous. Office for National Statistics. Mortality by cause (2000'- 27, 46). London, The Stationery Office 2001. 2 Czerny M, Trubel W, Claeys L et al. Acute mesenteric ischemia. Zentmlbl Chir 1997; 122: 538-544. 3 Cho JS, Carr JA, Jacobsen G et al. Long-term outcome after mesenteric artery reconstruction: a thirty seven-year experience. / Vase Surg 2002; 35: 453 - 460. 4 Park WM, Gloviczki P, Cherry KJ Jr et al. Contemporary management of acute mesenteric ischemia: factors associated with survival. / Vase Surg 2002; 35: 445-452. 5 Endean ED, Barnes SL, Kwolek CJ et al. Surgical management of thrombotic acute intestinal ischemia. Ann Surg 2001; 233: 801-808. 6 Foley MI, Moneta GL, Abou-Zamzam AM Jr et al. Revascularization of the superior mesenteric artery alone for treatment of intestinal ischemia. J Vase Srg2000; 32: 37-47. 7 Mamode N, Pickford I, Leiberman P. Failure to improve outcome in acute mesenteric ischaemia: seven-year review. Eur J Surgim; 165: 203-208. 8 Newman TS, Magnuson TH, Ahrendt SA et al. The changing face of mesenteric infarction. Am Surg 1998; 64: 611-616.

9 Urayama H, Ohtake H, Kawakami T et al. Acute mesenteric vascular occlusion: analysis of 39 patients. Eur J Surg 1998; 164: 195-200. 10 Duron JJ, Peyrard P, Boukhtouche S et al. Acute mesenteric ischemia: changes in 1985-1995. Surgical Research Associations. Chimr^e 1998; 123: 335-342. 11 Klempnauer J, Grothues F, Bektas H, Wahlers T. Acute mesenteric ischemia following cardiac surgery. / Cardiovasc Surg 1997; 38: 639-643. 12 Bronner JF, Boissel P. Acute ischemia and arterial mesenteric infarction in patients aged over 75. Apropos of a comparative series of 38 cases. / Chir 1997; 134: 109-113. 13 Voltolini F, Pricolo R, Naldini G, Parziale A. Acute mesenteric ischemia. Analysis of 47 cases. Minerva Chir 1996; 51: 285-292. 14 Konturek A, Cichon S, Gucwa J, Rogula T. Acute intestinal ischemia in material of the III Clinic of General Surgery, Collegium Medicum at the Jagellonian University. Przegl Lek 1996; 53: 719-721. 15 Ward D, Vernava AM, Kaminski DL et al. Improved outcome by identification of high-risk nonocclusive mesenteric ischemia, aggressive reexploration, and delayed anastomosis. Am J Surg 1995; 170: 577-581. 16 Deehan DJ, Heys SD, Brittenden J, Eremin 0. Mesenteric ischaemia: prognostic factors and influence of delay upon outcome. JR Coll SurgEdinb 1995; 40: 112-115. 17 Inderbitzi R, Wagner HE, Seiler C et al. Acute mesenteric ischaemia. Eur] Surg 1992; 158: 123-126. 18 Lew PJ, Krausz MM, Manny J. Acute mesenteric ischemia: improved results. A retrospective analysis of ninety-two patients. Surgery 1990; 107: 372-380. 19 Batellier J, Kieny R. Superior mesenteric artery embolism: eighty-two cases. Ann Vase Surg 1990; 4: 112-116. 20 Bapat RD, Aiyer PM, Relekar RG et al. Ischemic bowel disease. Indian] Gastroenterol 1990; 9: 19-22. 21 Finucane PM, Arunachalam T, O'Dowd J, Pathy MS. Acute mesenteric infarction in elderly patients. J Am Geriatr Soc 1989; 37: 355-358. 22 Sitges-Serra A, Mas X, Roqueta F et al. Mesenteric infarction: an analysis of 83 patients with prognostic studies in 44 cases undergoing a massive small-bowel resection. Br] Surg 1988; 75: 544-548.

ACUTE INTESTINAL
23 Wilson C, Gupta R, Gilmour DG, Imrie CW. Acute superior mesenteric ischaemia. Br JI Sure: 1987; 74: 279-281. o ' 24 Lazaro T, Sierra L, Gesto R et al. Embolization of the mesenteric arteries: surgical treatment in twenty-three consecutive cases. Ann Vase Surg 1986; 1: 311-315. 25 Andersson R, Parsson H, Isaksson B, Norgren L. Acute intestinal ischemia. A fourteen-year retrospective investigation. Ada CMrScandim; 150: 217-221. 26 Sachs SM, Morton JH, Schwartz SI. Acute mesenteric ischemia. Surgery 1982; 92: 646-653. 27 Harward TR, Brooks DL, Flynn TC, Seeger JM. Multiple organ dysfunction after mesenteric artery revascularization./ Vase Surg 1993; 18: 459-469. 28 Hallisey MJ, Deschaine J, Illescas FF et al. Angioplasty for the treatment of visceral ischemia. / Vase Interv Radiol 1995; 6: 785-791. 29 Johnston KW, Lindsay TF, Walker PM, Kalman PG. Mesenteric arterial bypass grafts: early and late results and suggested surgical approach for chronic and acute mesenteric ischemia. Surgery 1995; 118: 1-7. 30 Higgins R, Posner MC, Moosa HH et al. Mesenteric infarction secondary to tumor emboli from primary aortic sarcoma. Guidelines for diagnosis and management. Cancer 1991; 68:16221627. 31 Hamed RM, Ghandour K. Abdominal angina and intestinal gangrene. A catastrophic presentation of arterial fibromuscular dysplasia: case report and review of the literature. J Pediatr Surg mi; 32: 1379-1380. 32 Hoang MP, Lee EL, Anand A. Histologic spectrum of arterial and arteriolar lesions in acute and chronic cocaine-induced mesenteric ischemia: report of three cases and literature review. AmJSurgPatholim; 22: 1404-1410. 33 Gradman WS, Daniel], Miller B, Haji-Aghaii M. Homocysteineassociated acute mesenteric artery occlusion treated with thrombectomy and bowel resection. Ann Vase Surg 2001; 15: 247-250. 34 Seeburger JL, Stepak M, Fukuchi SG et al. Multiple arterial thrombo-embolisms in a patient with the 20210 A prothrombin gene mutation. Arch Surg 2000; 135: 721-722. 35 Kumar S, Sarr MG, Kamath PS. Mesenteric venous thrombosis. NEnglJMedmi; 345: 1683-1688. 36 Hassan HA, Raufman JP. Mesenteric venous thrombosis. South Md/1999;92: 558-562. 37 Divino CM, Park IS, Angel LP et al. A retrospective study of diagnosis and management of mesenteric vein thrombosis. Am /Surg2001; 181: 20-23. 38 Vignati PV, Welch JP, Ellison L, Cohen JL. Acute mesenteric ischemia caused by isolated superior mesenteric artery dissection. / Vase Surg 1992; 16: 109-112. 39 Allerton R. Acute mesenteric ischaemia associated with 5-FU, cisplatin and vincristine chemotherapy. Clin Oncol (R Coll Radiol) 1996; 8: 116-117. 40 Anayi S, Al-Nasiri N. Acute mesenteric ischaemia caused by Schistosoma mansoni infection. Br Med J (Clin Res Ed) 1987; 294: 1197. 41 Bailey RW, Bulkley GB, Hamilton SR et al. Protection of the small intestine from nonocdusive mesenteric ischemic injury due to cardiogenic shock. Am J Surg 1%^; 153: 108-116. 42 Schutz A, Eichinger WT, Breuer M et al. Acute mesenteric ischemia after open heart surgery. Angiology 1998; 49: 267-273. 43 Dahlberg PJ, Risken WA, Newcomer KL, Yutuc WR. Mesenteric ischemia in chronic dialysis patients. AmJNephrol 1985; 5: 327-332. 44 Gusmao L, Santana A, Riocarvalho I et al. Mesenteric ischemia in hemodialysis. Acta Med Port 1992; 5: 169-171. 45 Allen KB, Salam AA, Lumsden AB. Acute mesenteric ischemia after cardiopulmonary bypass. / Vase Surg 1992; 16: 391-396. 46 Ogino H, Miki S, Ueda Y, Tahata T. A case of bowel necrosis due to acute mesenteric ischemia following pulsatile cardio-

ISCHEMIA

pulmonary bypass. Ann Thorae Cardiovase Surg 1998; 4: 34-36. 47 Farkas JC, Calvo-Verjat N, Laurian C et al. Acute colorectal ischemia after aortic surgery: pathophysiology and prognostic criteria. Ann Vase Surg 1992; 6: 111-118. 48 Fink S, Chaudhuri TK, Davis HH. Acute mesenteric ischemia and malpractice claims. South AM/2000; 93: 210-214. 49 Meyer T, Klein P, Schweiger H, Lang W. How can the prognosis of acute mesenteric artery ischemia be improved? Results of a retrospective analysis. ZentraW CAir 1998; 123: 230-234. 50 Tsai CJ, Kuo YC, Chen PC, Wu CS. The spectrum of acute intestinal vascular failure: a collective review of 43 cases in Taiwan. BrJ Clin Praet 1990; 44: 603-608. 51 Gorey TF, O'Sullivan M. Prognostic factors in extensive mesenteric'ischaemia. Ann R Coll Surg Engl 1988; 70: 191-194. 52 May LD, Berenson MM. Value of serum inorganic phosphate in the diagnosis of ischemic bowel disease. Am J Surg 1983; 146: 266-268. 53 Acosta S, Nilsson TK, Bjorck M. Preliminary study of D-dimer as a possible marker of acute bowel ischaemia. 5r/Swrg2001; 88: 385-388. 54 Kanda T, Fujii H, Tani T et al. Intestinal fatty acid-binding protein is a useful diagnostic marker for mesenteric infarction in humans. Gastroenterology 1996; 110: 339-343. 55 Trompeter M, Brazda T, Remy CT et al. Nonocdusive mesenteric ischemia: etiology, diagnosis, and interventional therapy. Eur Radiol 2002; 12: 1179-1187. 56 Horton KM, Fishman EK, Multi-detector row CT of mesenteric ischemia: can it be done? Radiographics 2001; 21: 1463-1473. 57 Choudhary AM, Grayer D, Nelson A, Roberts I. Mesenteric venous thrombosis: a diagnosis not to be missed! / Clin Gastroenteronm; 31: 179-182. 58 Morasch MD, Ebaugh JL, Chiou AC et al. Mesenteric venous thrombosis: a changing clinical entity. / Vase Surg 2001; 34: 680-684. 59 Danse EM, Laterre PF, Van Beers BE et al. Early diagnosis of acute intestinal ischaemia: contribution of colour doppler sonography. Ada Chir Belg 1997; 97: 173-176. 60 Klotz S, Vestring T, Rotker J et al. Diagnosis and treatment of nonocdusive mesenteric ischemia after open heart surgery. Ann Thorae Swig 2001; 72: 1583-1586. 61 Train JS, Ross H, Weiss JD et al. Mesenteric venous thrombosis: successful treatment by intraarterial lytic therapy. / Vase Interv Radiol W8; 9: m-m. 62 Brunaud L, Antunes L, Collinet-Adler S et al. Acute mesenteric venous thrombosis: case for nonoperative management. / Vase Surg mi; 34: 673-679. 63 Leung DA, Schneider E, Kubik-Huch R et al. Acute mesenteric ischemia caused by spontaneous isolated dissection of the superior mesenteric artery: treatment by percutaneous stent placement. Eur Radiol %m\ 10: 1916-1919. 64 Slonim SM, Miller DC, Mitchell RS et al. Percutaneous balloon fenestration and stenting for life-threatening ischemic complications in patients with acute aortic dissection. / Thorae CardiowMcSurg 1999; 117: 1118-1126. 65 Brountzos EN, Critselis A, Magoulas D et al. Emergency endovascular treatment of a superior mesenteric artery occlusion. Cardiovase Intervent Radiol 2001; 24: 57 - 60. 66 Loomer DC, Johnson SP, Diffin DC, DeMaioribus CA. Superior mesenteric artery stent placement in a patient with acute mesenteric ischemia. / Vase Interv Radiol 1999; 10 : 29-32. 67 Rundback JH, Rozenblat GN, Poplausky M et al. Re: jejunal artery angioplasty and coronary stent placement for acute mesenteric ischemia. Cardiovase Intervent Radiol 2000; 23: 410-412. 68 Bjorck M, Acosta S, Lindberg F et al. Revascularization of the superior mesenteric artery after acute thrombo-embolic occlusion. BrJ Surg m% 89: 923-927. 69 Sullivan KM, Battey PM, Miller JS et al. Abdominal compartment syndrome after mesenteric revascularization. / Vase Surg 2001; 34: 559-561.

147

VASCULAR

EMERGENCIES

70 Seshadri PA, Poulin EC, Mamazza J, Schlachta CM. Simplified laparoscopic approach to "second-look" laparotomy: a review. Surg Laparosc Endosc Percutan Tech 1999; 9: 286-289. 71 Giulini S, Bonardelli S, Cangiotti L et al. Factors affecting prog-

nosis in acute intestinal ischemia. Int Angiol 1987; 6: 415-420. 73 Kasirajan K, O'Hara PJ, Gray BH et al. Chronic mesenteric ischemia: open surgery versus percutaneous angioplasty and stenting. / Vase Surg 2001; 33: 63-71.

14

148

F I .)
RUPTURE OF SPLANCHNIC ARTERY ANEURYSMS
JOAQUIM BARBOSA, MARIA-JOSE FERREIRA

Splanchnic artery aneurysms represent an uncommon but important form of mesenteric vascular disease since nearly 22 % present as clinical emergencies, including 8.5% that result in death [1]. In routine autopsies, they appear in 0.01% to 0.2%, being diagnosed more often in the aged population [2]. Two necropsy studies in the elderly found a 10% prevalence of visceral aneurysm, supporting that its incidence is dependent on the age of the studied population [3]. All visceral arteries can be affected by aneurysms. However, the most frequently affected are the splenic, hepatic, superior mesenteric and celiac arteries being responsible for almost 90% of the visceral aneurysms.

15
149

History
The first report of a visceral artery aneurysm (VAA) dates from 1770 by Beaussier, who described a splenic artery aneurysm in an anatomic demonstration. In 1881, the American president James A. Garfield died from rupture of a splenic artery aneurysm, two months after a gunshot in the abdomen. DeBakey and Cooley reported, in 1953, the first successful resection of a superior mesenteric artery aneurysm. Before 1960 only a few reports of VAA were published with high mortality rates. In 1970, Stanley et al. reviewed the reports in the literature concerning 1098 cases, but since then the number of cases increased dramatically [1].

Because splanchnic artery aneurysms are rare, the reports in the literature mainly comprise small series or particular forms of presentation. Accounting to the small number there is a trend to enclose in the series different etiologies as well as true and false aneurysms. All these factors contribute to the incomplete knowledge and poor description of its natural history. At present, advances in imaging technology and endovascular procedures are modifying the diagnostic and therapeutic strategies and partly explain the growing number of reported aneurysms. But despite all these advances, a significant number of these lesions are not discovered until rupture, which often results in the death of the patient. Therefore,

VASCULAR

EMERGENCIES

an aggressive approach to the diagnosis and management of these aneurysms is required. However, if there is no doubt about the surgical management of ruptured visceral aneurysms, controversy remains regarding the optimal management of the asymptomatic cases because the risk-benefit is difficult to access in the absence of a comprehensive natural history. New therapeutic and less aggressive techniques offer new options for some patients, but there is still no sufficient long-term follow-up for evaluation of the results.

Incidence
The introduction of more sophisticated imaging techniques, like high-resolution computed tomography, high-resolution ultrasound and magnetic resonance angiography permitted to discover asymptomatic aneurysms. More than 3 000 splanchnic artery aneurysms have been published in the international literature. Traditionally, the reported incidence of the aneurysms related to their anatomic location is: splenic artery 60%, hepatic artery 20%, superior mesenteric artery 5.5%, celiac artery 4%, gastroduodenal and gastro-epiploic artery 4%, jejunal and ileal-colic artery 3%, pancreatoduodenal and pancreatic 2%, gastroduodenal 1.5% and mesenteric inferior artery less than 1% [4]. However, in the last decade the most common visceral aneurysm reported in the literature occurred in the hepatic artery. This fact can be associated with the increased use of percutaneous diagnostic and therapeutic biliary procedures, as well as the non-operative management of blunt hepatic trauma and the increased use of CT scan for diagnosis. As a result, there is an increased detection of post-traumatic false aneurysms of the hepatic arterial branches, particularly intrahepatic aneurysms. More than 80% of hepatic artery aneurysms were traditionally extra hepatic but now, these represent only 66%. The great majority of the hepatic artery aneurysms are solitary (91%) and most commonly located in the common hepatic or the right hepatic artery. They occur mainly in males (66%), with an average age of 56 years (range 18-86 years). The true incidence of splenic artery aneurysms is unknown, but has been reported to be between 0.02% and 10.4% in the general population and 8.8% to 50% in cirrhotic patients. The mean age of presentation is 52 years (range 2-93 years) and

78.8% of the splenic artery aneurysms occur in women. Most of them are small (smaller than 2 cm), saccular and almost 80% are located in the mid-or distal splenic artery [5]. Aneurysms of the splenic artery occur in association with fibrodisplasia of the renal arteries. About 4% of patients with renal fibrodisplasia present with concomitant splenic artery aneurysms. Both intracranial and splenic aneurysms are found in patients with medial displasia [6]. Aneurysms of the celiac artery constitute 4% of the visceral artery aneurysms and are characteristically associated with aneurysms at other locations: abdominal aortic aneurysms are present in 20% of patients and other visceral artery aneurysm in 40% [6]. Aneurysms of the superior mesenteric artery may be saccular or fusiform and are almost always located in the proximal 5 centimeters of the artery. In the past, most of these aneurysms were discovered in women, but now the incidence in men is 66%.

Etiology
Most aneurysms are caused by atherosclerosis or medial degeneration, however, some authors defend that the atherosclerotic changes are a secondary process to the appearance of the aneurysm. Other, less common causes include trauma, infection and iatrogenic injury. The etiology of splenic artery aneurysms remains unclear, although several conditions have been associated like fibrodisplasia, pancreatitis, portal hypertension, hemodynamic and endocrine changes in pregnancy, atherosclerosis, inflammatory and infectious disorders, arteritis and collagen vascular diseases. The most common pathologic findings are defects of the media characterized by fragmentation of elastic fibers and loss of smooth muscle. The first responsible factor for splenic artery aneurysms is the presence of systemic arterial fibrodisplasia. Patients with arterial fibrodisplasia exhibit splenic artery aneurysms with a frequency six times greater than the normal population. They are more common in women with multiple pregnancies, probably due to the presence of estrogens and progesterone receptors in the arterial wall. Hormonal shifts during pregnancy and relaxin (a gestational hormone responsible for the dilatation) can induce degenerative changes in the splenic artery like disruption of the internal elastic lamina, fragmentation of elastic fibers and fibrodisplasia of the media.

15
150

RUPTURE OF SPLANCHNIC ARTERY

The development of splenic artery aneurysms in patients with portal hypertension and splenomegaly can be due to a hyperkinetic state in the spleen. These hemodynamic factors also explain the same situation in patients after liver transplantation and represent an additional risk factor in pregnancy. Another mechanism described is infection secondary to sub acute bacterial endocarditis in intravenous drug abusers. Enzymatic injury to pancreatic and peripancreatic arteries (splenic, pancreatoduodenal and gastroduodenal) or from erosion of a pseudocyst into adjacent visceral arteries explain aneurysm formation in the course of a pancreatic inflammatory process. These aneurysms are closely related to the intensity and duration of the disease and the incidence varies between 10% to 17% in chronic pancreatitis. In celiac artery aneurysms, the most common pathologic finding is medial degeneration and atherosclerosis. Other findings include poststenotic dilatation and re-entry from aortic dissection. The majority of cases are true aneurysms (55%) but false aneurysm secondary to penetrating trauma and infection have also been reported. Historically, mycotic aneurysms of the hepatic artery had the highest incidence but nowadays they account for only 3% and usually occur in intravenous drug abusers. False aneurysms now account for nearly 50% of reported hepatic aneurysms, due to iatrogenic maneuvers during percutaneous biliary procedures or trauma. The etiology of superior mesenteric aneurysms is diverse, but a significant number is caused by infection secondary to subacute bacterial endocarditis by nonhemolytic streptococcus. The incidence of infected aneurysms was 58% to 63%, but during the last decade it decreased to 33%. This was accompanied by the increase of true atherosclerotic aneurysms (25%) as well as those related to inflammatory processes like pancreatitis (12%). Nonetheless, the superior mesenteric artery remains the most common site of infection of a peripheral muscular artery. Dissection also affects this artery more than any other visceral artery.

ANEURYSMS

artery aneurysms is discovered in the asymptomatic state. However, in the study of Carr et al., 63% of the VAA were symptomatic at the time of presentation and 23.9% presented with rupture [7]. A classic presentation of an asymptomatic aneurysm is the appearance of a circular calcification on an abdominal radiography (Fig. 1). Today they are depicted by ultrasound or CT scan. If aneurysms are symptomatic, the most common symptom is abdominal pain. At physical examination a palpable mass or abdominal bruit can be found. Rupture of an aneurysm causes severe abdominal pain and hypovolemic shock. In case of a ruptured splenic artery aneurysm, the hemorrhage is initially confined to the lesser sac, resulting in a period of hemodynamic stability. Following, the hemorrhage ruptures into the peritoneal cavity, which is the so called double-rupture phenomenon first described by Brockman, in 1930 [6]. Splenic artery aneurysm rupture into the retroperitoneum (11%), stomach (16%), colon (8%), pancreas (6%) and 13% develop gastro-intestinal bleeding secondary to erosion of the aneurysm into adjacent viscera. In splenic artery aneurysms, particularly those associated with inflammatory lesions, bleeding is usually brisk but varies from short, repeated and self-limited episodes to massive hemorrhage. In

75
151

Clinical presentation and diagnosis

Nowadays, with the increased use of abdominal imaging techniques a great number of splanchnic
FIG. 1 Plain X-ray of the abdomen showing the calcifications of an aneurysm of the splenic artery.

VASCULAR

EMERGENCIES

11
152

patients with pancreatitis, the presence of a false aneurysm can be suspected by persisting or abrupt increasing abdominal pain or hemodynamic instability and/or gastro-intestinal bleeding with no other obvious cause. In up to 80% of patients with a hepatic artery aneurysm, rupture can occur in equal frequency into the peritoneal cavity and biliary tract or duodenum, gallbladder, portal vein or stomach. Abdominal pain is present in 55% of patients and gastrointestinal bleeding or hemobilia occurs in 46% of patients. Jaundice can be present up to 10% of patients due to extrinsic compression of the bile duct by the aneurysm. Aneurysms of the celiac and superior mesenteric artery are most often symptomatic at presentation and are suspected because of intermittent abdominal pain. This can result from repetitive embolization of mural thrombus with subsequent mesenteric ischemia. Between 38% and 50% of mesenteric artery aneurysms and 13% of celiac aneurysms present with rupture, manifested by acute abdominal pain, hypotension or sudden death. Aneurysms of the gastric and gastro-epiploic artery most often present with rupture (90%), two thirds manifested by gastro-intestinal bleeding and one third with intraperitoneal bleeding. Aneurysms of the inferior mesenteric artery are very rare and presentation is characterized by abdominal pain, gastro-intestinal bleeding or shock. The majority is discovered during exploratory laparotomy for gastro-intestinal hemorrhage. Despite advances in endoscopic diagnosis, the source of gastro-intestinal bleeding is not found in 5% of patients [8]. This is complicated by the fact that 70% to 80% of such bleeding stops spontaneously. Although rare, splanchnic artery aneurysms can cause recurrent gastro-intestinal bleeding due to erosion into adjacent organs. The rarity of this pathology contributes to the delay in the diagnosis and higher mortality. Early angiography is recommended in most cases of gastro-intestinal bleeding in patients with no endoscopic diagnosis. CT-scan can also demonstrate an aneurysm and is very useful, particularly in elderly patients because it offers a lesser risk. It's also useful in defining the dimensions of the aneurysm and to delineate its relations with the surrounding organs [8] (Fig. 2). Clinical assessment of visceral aneurysms is very difficult but selective angiography is the most valuable exam for the diagnosis and therapeutic planning (Fig. 3).

A selective biplanar arteriography is mandatory for the study of the celiac and superior mesenteric arterial aneurysms. Spiral CT angiography with three-dimensional reconstructions can give detailed anatomic information and may identify the vessel origins. CT scanning is also useful in the diagnosis of ruptured aneurysm, identifying a contained hematoma within the lesser sac or retroperitoneum [9].

FIG. 2 Giant aneurysm of the splenic artery. Note the calcifications within the splenic parenchyma due to previous infarctions secondary to embolism from the splenic artery.

FIG. 3 Selective ansiosram of the celiac trunk showins an aneurysm located at the second third of the splenic artery.

RUPTURE OF SPLANCHNIC ARTERY

ANEURYSMS

Risk of rupture
The risk of rupture of visceral aneurysms is overestimated by collective reviews because there is a bias due to including symptomatic cases as well. Furthermore, the true incidence of visceral artery aneurysms is unknown. The reported risk of rupture of splenic artery aneurysms varies from 3.0% to 9.6%. In the study of the Mayo Clinic the rate of rupture was 2.9% for women and 10.9% for the male population with an overall risk of 4.6% [10]. There appears to be a bimodal age distribution with the majority of aneurysm rupture presenting either in young adults or in the elderly. For instance, in splenic artery aneurysms the increased incidence of rupture in young adults is associated with pregnancy or portal hypertension and carries a poor prognosis. On the other hand, the risk of rupture for aneurysms in elderly patients is probably less than 2% with a much lower mortality risk. The risk of rupture of a splenic artery aneurysm is also linked to the size of the aneurysm (more than 2 centimeters of diameter), pregnancy, cirrhosis and liver transplantation. In these two last groups the risk is particularly high in patients with oc-1-antitripsin deficiency due to the excessive proteolitic activity. Calcification of the aneurysm was classically associated to higher risk of rupture but this fact has never been proven. B-blockade may be protective against rupture. Growth rates are slow and growth is infrequent. Stanley suggests that the true risk of rupture of splenic artery aneurysm is less than 2%. Data of a single series suggests a risk of rupture between 3% to 10% [4]. The mortality after rupture remains high and was recently reported to be 36% [11]. Rupture of splenic artery aneurysms in pregnant women carries a high maternal and fetal mortality (70% and 95% respectively). This rupture is more common in the third trimester, labor or post-partum and in multipares. On the other hand, mortality related to aneurysms associated with pancreatitis varies between 15% and 43%, depending on the patient clinical state, site and nature of the bleeding, lesion and the type of management. The exact risk of rupture of hepatic aneurysm is not known, but nearly 2/3 of the recent reports present with rupture into either the peritoneal cavity or an adjacent viscera, most likely the biliary tract. The mortality of these selected patients was 21% [5] but it can be assumed that the mor-

tality of free rupture is much higher and an aggressive approach is clearly indicated. Blumenberg et al., in 1974, consider the risk of rupture of mesenteric artery aneurysms to be quite small [12], but Stone et al., in 2002, reported a rupture rate of 38% [13]. Probably, this variability of the reported risk of rupture can be related to the evolution of referred etiologies for this pathology. In this serie, male patients were particularly prone for rupture with a risk of 50%. The mortality rate of ruptured superior mesenteric aneurysms is 30%. The natural history of celiac artery aneurysms appears to be one of expansion and rupture, with some authors reporting a risk of rupture of approximately 13% [11]. The mortality rate with rupture is very high, near 100%, and is often complicated with intestinal infarction. The global mortality rate after rupture of a visceral artery aneurysm ranges between 35% and 100% as a result of the catastrophic hemorrhage that may ensue, where as mortality of elective treatment is less than 10% [14]. Therefore aggressive surgical policy of all aneurysms is recommended, especially because rupture cannot be predicted.

Indications for the surgery

At present there is a trend towards observation of most small aneurysms in asymptomatic, high risk, elderly patients. Indications for surgery remain controversial in some groups of patients. Repair is mandatory for splenic artery aneurysms in the following situations: symptomatic aneurysms, pregnancy or women in childbearing age, liver transplant patients, and diameter larger than two centimeters. Furthermore, patients should have a reasonable operative risk and life expectancy of more than two years. Operative intervention should be considered in all but the most high-risk patients with hepatic, celiac artery and superior mesenteric artery aneurysms, considering the high mortality with rupture. Aggressive surgical management of pancreatic and pancreaticduodenal artery aneurysms is also mandatory, because of the very high risk of rupture at the time of clinical presentation (56% for the gastroduodenal and 38% for the pancreaticduodenal artery).

1L 153

VASCULAR

EMERGENCIES

Treatment
The current options for management of a VAA include endovascular and conventional surgery. Due to the high morbidity associated with major operations, various minimally invasive techniques have been developed. Transcatheter embolization has been used successfully in the treatment of splenic and hepatic artery aneurysms, although end-organ ischemia, painful splenic infarction and late vessel recanalization are potential problems [15]. Various laparoscopic techniques have also been applied to the treatment of this pathology as well as the use of endovascular techniques with the use of stent grafts. The option for each treatment modality depends on the presentation, location of the aneurysm, size, and general condition of the patient as well as the experience of the vascular center. Conventional surgical treatment of the visceral artery aneurysms usually involves ligation or resection of the aneurysm with or without vascular reconstruction. Surgical strategies differ somewhat for each type of aneurysm.

SURGERY

15
154

Rupture of visceral artery aneurysms requires urgent surgical repair. In this emergency setting, ligation of the aneurysm without reconstruction can be the best alternative whenever there is adequate collateral circulation, particularly in unstable or high surgical risk patients. Ligation of the hepatic artery proximal to the gastroduodenal, right gastric artery, gastro-epiploic and pancreatoduodenal is possible without major ischemic complications. Ligation of the splenic artery is not always followed by spleen necrosis and ligation of the superior mesenteric artery can be performed, followed by segmental bowel resection in a setting of emergency. In large aneurysms or for those involved in inflammatory tissue, the best option is to ligate the proximal and distal artery from inside the aneurysm. Celiac or hepatic artery aneurysms involving the gastroduodenal artery and superior mesenteric artery require resection of the aneurysm and revascularization by interposition or bypass graft. For splenic artery aneurysms, we propose a golden rule, that consists of dividing the artery in three main segments. For the proximal aneurysms we advise ligation of the splenic artery, normally without splenectomy. Those who are located to the median third of the artery (Fig. 3), an intra-aneurys-

matic ligation with splenectomy, seems to be the best solution. Considering the distal segment, ligation of the splenic artery and splenectomy is the preferential therapeutic option, in our opinion. It is mandatory to perform the drainage of the pancreatic area, in all suspected pancreatic injury, to avoid the complications of a pancreatic fistula. Surgical options for repair of celiac artery aneurysms include aneurysmectomy with or without revascularization, aneurysmorraphy, reimplantation and ligation. The latter option is not always accompanied by hepatic necrosis, but this complication is more frequent in the presence of liver disease. Superior mesenteric aneurysms pose unique challenges for surgical repair. They extend well beyond the region usually encountered with atherosclerotic process. In addition, inflammation and infection are common features of these aneurysms, making the dissection difficult. Therefore one of the most attractive techniques for the treatment of such aneurysms is ligation. There is no assurance that the collateral flow is sufficient to prevent ischemia, thus a careful assessment of adequate mesenteric blood flow is necessary. Stone et al. described their experience with ruptured aneurysms and ligation was the most used technique employed in 75% of the patients [13]. However, 38% of the patients with rupture needed concomitant bowel resection. No patient that underwent elective intervention required bowel resection. Revascularization was recommended only if bowel ischemia was present in patients undergoing operation for rupture or if pre-operative symptoms or evaluation, mainly angiography, suggested mesenteric ischemia. Most series have also found that adequate collateral blood flow exists and ligation without revascularization is the procedure of choice. Other techniques for repair of the superior mesenteric artery aneurysm comprise aneurysmectomy used in 35% of the reported cases and aneurysmorraphy in 21%. The first technique is used after proximal and distal ligation and to resect the infected aneurysm. Aneurysmorraphy is used when the aneurysm is saccular and a portion of the arterial wall is free of disease. In only a minority of cases the revascularization was undertaken, accounting for less than 15% in the literature [11]. Revascularization can be made either by interposition or, in larger or more distal lesions, by aorto-mesenteric bypass. When the aneurysm is not infected and there is no intestinal

RUPTURE OF SPLANCHNIC ARTERY

ischemia, a synthetic graft can be used. In the other cases it is necessary to use vein. Intrahepatic large and proximal aneurysms may require partial liver resection. Alternatively, proximal and distal ligation can be undertaken, accepting the risk of liver necrosis. Currently, approximately 37% of hepatic artery aneurysms are treated with percutaneous embolization. [5].

ANEURYSMS

MINIMAL INVASIVE TECHNIQUES

Laparoscopic ligation, resection or clipping of the aneurysm constitute other therapeutic alternatives. These approaches can be difficult in patients with previous abdominal operations, obese patients or for aneurysms located in pancreatic parenchyma or deep in the splenic hilus. During the past ten years, percutaneously directed embolization became the most employed technique for the treatment of intrahepatic aneurysms and situations like those associated with pancreatitis. With respect to treatment of intrahepatic aneurysms, good immediate outcome with success rates between 67%100% have been described [16,17]. Complications of this technique include hepatic necrosis, abscess formation or sepsis. Early failure was related to the inability to catheterize selectively the vessel due to arterial spasm, rupture of the aneurysm during the procedure or arterial perforation by the catheter [16]. Embolization is a very attractive therapeutic modality in high-risk patients or for lesions difficult to access surgically, like the aneurysms associated with pancreatitis. For some authors, embolization is considered a temporary process until definite surgery is performed. Surgery is reserved for instable patients, failed embolization or for complications like infection or extrinsic compression. As these complications are most likely to occur in large aneurysms, excision or drainage should be considered after successful embolization. Several reports of patients treated with embolization as a sole therapy described a high technical success rates (75%-100%), low morbidity (14%-25%) and low death rates (048%). However, most reports include a small number of patients and short followup [8,17]. Carr et al. conclude that embolization alone is not adequate because of the high primary failure rate with necessity to convert to open surgery [18]. All patients should be followed by CT-scan after embolization because there is a risk of recurrent hemorrhage (12%) and recanalization (37%). There are some reports describing ultrasound guided thrombin occlusion of visceral aneurysms at

the time of the surgery [19]. The use of thrombin injection, was initially used in the treatment of femoral artery pseudoaneuryms. Endovascular treatment by means of stent grafts is another option, with the advantage of the lower procedural risk and because it excludes the aneurysm while preserving end-organ perfusion [15]. The arterial anatomy and the location of the aneurysm have a large impact on the technical ability to place a stent graft. Normal caliber artery on either side of the aneurysm is required to allow safe sealing. The introduction of more flexible stent grafts and smaller introducer systems, also allow treatment in tortuous vessels. The risk of vessel rupture or thrombosis is also of major concern. In conclusion, despite advances in the diagnosis and therapy, the heterogeneity of the VAA indicates that management should be individualized and tailored. Prognosis of this disorder depends on the anatomic location of the aneurysm, primary pathology and general condition of the patient.

Summary
The splanchnic artery aneurysms are an uncommon pathology with a high morbidity and mortality rate. Rupture is the most frequent and dangerous complication. The incomplete knowledge of the natural history of the disease contributes to the complexity of these situations. The etiology of the process seems to be very different according the involved vessel. Age, general condition of the patient, size and location of the aneurysm play an important role in the therapeutic decision. In case of rupture, all elective conventional surgical techniques can be applied, including simple ligation of the vessel. The main objective is to stop the bleeding and save the patients life and preserve the viability of the organ. Endovascular procedures are a good alternative, especially in high-risk patients. However, long-term follow-up is required to assess durability and safety of these techniques.

155

R E F E R E N C E S 1 Stanley JC, Thompson NW, Fry WJ. Splanchnic artery aneurysms. Arch Surg 1970; 101: 689-697. 2 Carr SC, Mahvi DM, Hoch JR et al. Visceral artery aneurysm rupture. / Vase Surg 2001; 33: 806-811.

VASCULAR

EMERGENCIES

3 Busuttil RW, Brin BJ. The diagnosis and management of visceral artery aneurysms. Surgery 1980; 88: 619-624. 4 Zelenock GB, Stanley JC. Splanchnic artery aneurysms. In: Rutherford RB (ed). Vascular surgery, 5th ed. Philadelphia, W. B. Saunders Co., 2000: pp 1369-1382. 5 Shanley CJ, Shah NL, Messina LM. Common splanchnic artery aneurysms: splenic, hepatic, and celiac. Ann Vase Surg 19%; 10: 315-322. 6 Carr SC, Pearce WH. Management of visceral artery aneurysm. In: Yao JST, Pearce WH (eds). Practical vascular surgery. Stamford, Appleton & Lange, 1999: pp 241-258. 7 Carr SC, Pearce WH, Vogelzang RL et al. Current management of visceral artery aneurysms. Surgery 1996; 120: 627-634. 8 Wagner WH, Cossman DV, Treiman RLV et al. Hemosuccus pancreaticus from intraductal rupture of a primary splenic artery aneurysm. / Vase Surg 1994; 19: 158-164. 9 Brunei WG, Greenberg HM. CT demonstration of a ruptured splenic artery aneurysm. J Comput Assist TomogrlQQl; 15: 177178. 10 Abbas MA, Stone WM, Fowl RJ et al. Splenic artery aneurysms: two decades experience at Mayo clinic. Ann Vase Swrg-2002; 16: 442-449. 11 Messina LM, Shanley CJ. Visceral artery aneurysms. Surg Clin North Am 1997; 77:425-442.

12 Blumenberg RM, David D, Slovak]. Abdominal apoplexy due to rupture of a superior mesenteric artery aneurysm: clip aneurysmorrhaphy with survival. Arch Surg 1974; 108: 223-226. 13 Stone WM, Abbas M, Cherry Kf et al. Superior mesenteric artery aneurysms: is presence an indication for intervention? / Vase Surg 2002; 36: 234-237. 14 Settembrini PG, Jausseran JM, Roveri S et al. Aneurysms of anomalous splenomesenteric trunk: clinical features and surgical management in two cases. / Vase Surg 1996; 24: 687-692. 15 Larson RA, Solomon J, Carpenter JP. Stent graft repair of visceral artery aneurysms. / Vase Surg 2002; 36: 1260-1263. 16 De Perrot M, Berney T, Buhler L et al. Management of bleeding pseudoaneurysms in patients with pancreatitis. Br J Surg 1999; 86: 29-32. 17 Gambiez LP, Ernst OJ, Merlier OA et al. Arterial embolization for bleeding pseudocysts complicating chronic pancreatitis. Arch Surg 1997; 132: 1016-1021. 18 Carr JA, Cho JS, Shepard AD et al. Visceral pseudoaneurysms due to pancreatic pseudocysts: rare but lethal complications of pancreatitis. / Vase Surg 2000; 32: 722-730. 19 Mclntyre TP, Simone ST, Stahlfeld KR. Intraoperative thrombin occlusion of a visceral artery aneurysm. J Vase Surg 2002; 36: 393-395.

11
156

16
THE ABDOMINAL COMPARTMENT SYNDROME
MICHAEL YAPANIS, JOHN WOLFE

Intra-abdominal pressure (IAP) has been studied since the late 19th century. Indeed, the effects of raised IAP in cats were documented by Heinricius, who found that they could be fatal due to inhibition of respiration. Later, seminal studies by Emerson, Coombs, and Overholt looked at IAP, its variation, and factors that could influence it. Although the clinical consequences of raised IAP have been known for many decades, the "abdominal compartment syndrome" (ACS) was only popularized as a concept in the early 1980s [1], when multi-organ dysfunction was identified, in conjunction with a distended abdomen and intra-abdominal hypertension (IAH).

157

Definition
Broadly, compartment syndrome can be defined as a rise in pressure in a body compartment, causing reduction in tissue perfusion thereby jeopardizing tissue function and viability. While this concept has been applied to ACS, there is no widely accepted definition. It is thought to occur when there is a sustained rise in IAP in conjunction with a distended abdomen, increased airway pressures, hypoxia, hypercarbia, a fall in cardiac output, and oliguria despite adequate fluid resuscitation. Char-

acteristically, the abnormal parameters show marked improvement on abdominal decompression.

NORMAL IAP AND IAH

The normal range for LAP is thought to be marginally above atmospheric pressure (0 to 5 mmHg) [2]. However, marked variation has been found with diaphragmatic, abdominal wall, and pelvic muscular contraction, body position, sex, and body mass index [3]. The exact point at which this raised IAP becomes significant is to some extent arbitrary, and definitions may quote values ranging from

VASCULAR

EMERGENCIES

more than 10 mmHg to more than 25 mmHg. Most authors suggest a pressure of higher than 20 mmHg, as indicating IAH [4].

Epidemiology
The majority of studies are based in trauma center intensive care units [5] on a select group of high-risk patients. Furthermore, they are frequently retrospective or small prospective series with variable definitions of IAH or ACS, and subjects may have undergone prophylactic decompression prior to LAP measurement. Therefore, the true incidence of ACS is difficult to ascertain. Ivatury et al. [6] showed that 32.9% of 70 patients admitted to surgical intensive care with life-threatening penetrating abdominal trauma developed IAH (IAP superior to 25 cmHsO). In a prospective series by Meldrum et al. [5], 14% of 145 acutely injured patients, with an injury severity score higher than 15, requiring emergent laparotomy, developed ACS. Recently, Hong et al. [7] prospectively studied 706 patients of all risks, with or without prior laparotomy, admitted to a trauma intensive care unit. Only 2% of the patients developed IAH (IAP> 20 mmHg), while only 1 % went on to develop fullblown ACS. This study implies that the true incidence of ACS is very low. Nevertheless, in a survey of the current opinions of American trauma surgeons, 85% had experienced ACS at least once in the previous year, suggesting that it is a condition of which every surgeon exposed to vascular trauma should be aware [8].

phylactic mesh closure has been undertaken, the incidence of ACS has been higher. Ruptured aortic aneurysm can result in ACS through massive intra- and retroperitoneal hemorrhage and the large volume of fluid used during resuscitation. In the experience of Fietsam et al. [10], 4 of 104 patients operated on for ruptured aortic aneurysm developed ACS. All received more than 25 L of fluid resuscitation. Maxwell et al. [11] studied the role of fluid resuscitation in the development of secondary ACS. One half percent of 1216 trauma admissions (in the absence of abdominal injury) underwent abdominal decompressions for visceral edema. Fluid resuscitation averaged 19 5 L of crystalloid and 29 10 units of blood before laparotomy. Consequently, they recommended that IAP should be monitored when resuscitation volumes approached 10 L of crystalloid or 10 units of packed red cells.

16
158

Retroperitoneal hemorrhage Ruptured aneurysm Acute pancreatitis Spontaneous Mesenteric ischemia Intestinal obstruction Colonic pseudo-obstruction rogemc Postoperative Trauma Burns Massivefluid Resuscitation Chronic Obesity Ascites Ovarian tumor Pneumoperitoneum Ileus Acute gastric dilatation Severe abdominal

Etiology (Table I)
Essentially, any condition associated with visceral expansion or which gives rise to an increase in intraor retroperitoneal fluid, usually blood, can cause ACS. Causes are divided into acute or chronic. Of the acute causes, abdominal trauma is the most commonly reported. Probable risk factors include severe trauma, massive fluid resuscitation, a high injury severity score, and the requirement of a damage control laparotomy. In a study by Ertel et al. [9], 17 of 311 patients with severe abdominal or pelvic injury, requiring damage control laparotomy, and admitted to the intensive care unit, developed ACS. As many as 47% of these patients required packing. Where primary fascial as opposed to pro-

THE ABDOMINAL COMPARTMENT

SYNDROME

Pathophysiology
The elucidation of the mechanisms of pathogenesis in IAH has relied to a large extent on animal models. The advent of laparoscopy, especially in the last decade, has demonstrated that some of the organ changes seen in laparoscopy can mirror those with increased IAP as IAH begins to develop. The rise in IAP seen in ACS is often caused by the accumulation of large volumes of intra- and extraperitoneal blood and fluid. This may relate partly to the underlying pathology, but also to a bleeding diathesis induced by consumption of coagulation factors. Intra-abdominal packing and massive tissue edema from the liberal use of crystalloids and the exudation of fluid due to enhanced capillary permeability can precipitate the development of ACS. Organ dysfunction in ACS is usually multifactorial. Damage not only occurs as a result of the pressure effects of IAH, but also because of ischemic damage from hypovolemic shock and then reperfusion from resuscitation (ischemia-reperfusion injury). It has been shown that hemorrhage followed by resuscitation produces greater cardiorespiratory damage at a lower IAP than IAH alone [12]. More recently, in an animal study by Oda et al. [13], sequential hemorrhagic shock, resuscitation, and ACS were associated with significantly increased portal and central venous cytokine levels and more severe lung injury than hemorrhagic shock or ACS alone. In addition, there is often associated comorbidity in an elderly population, which enhances any pressure induced damage. This includes ischemic heart, respiratory, and renovascular disease. However, the importance of raised IAP is demonstrated by the often rapid reversal of organ dysfunction following abdominal decompression [6]. The response of the systems to a raised IAP is graded. At pressures of up to 10 mmHg, there are no effects. As the IAP exceeds 15 mmHg, cardiovascular changes take place, which become more profound as the IAP exceeds 20 mmHg. Oliguria can arise at lower pressures, but it becomes more evident when they exceed 20 mmHg. As IAP approaches 40 mmHg, anuria ensues.

lower pressure (intrathoracic), the anatomical obstruction of the IVC by diaphragmatic distortion caused by IAH, and a drop in flow from the IVC and retroperitoneal veins. In turn, this leads to a reduction in preload. Peripheral vascular resistance increases due to compression of vascular beds, resulting in an elevated afterload. Owing to a rise in intrathoracic pressure, cardiac compliance falls, increasing ventricular filling pressures. The combination of the changes in preload, afterload, and cardiac compliance cause a fall in cardiac output. Consequently, in order to maintain blood pressure, there is a reflex tachycardia. Mean arterial pressure does not change significantly. The presence of underlying ischemic heart disease renders the myocardium susceptible to the effects of an increased afterload and tachycardia, and to the development of cardiac failure due to volume overload.

THE KIDNEYS
In 1923, Thorington and Schmidt documented the effects of raised IAP on urine output. Total anuria was found to occur with an IAP of 30 mmHg. Sugrue studied the changes in renal function in IAH after laparotomy and found that the odds ratio for developing renal impairment with increased IAP was 12.4 [4]. The mechanism for this relates to a rise in retroperitoneal pressure causing renal vein compression, and an increase in renal vascular resistance. These changes manifest in a fall in renal blood flow and glomerular filtration rate. Various humoral factors may play a role in oliguria, including rises in renin, aldosterone, and antidiuretic hormone production with increased IAP. Ureteric obstruction is not thought to occur.

THE RESPIRATORY SYSTEM

IAH forces the diaphragm upward thereby reducing chest wall and lung compliance. This necessitates a rise in peak airway pressure to maintain ventilation. Since movement of the diaphragm is impaired, the upper zones of the lungs, which normally have a higher ventilation-perfusion ratio, are ventilated even more [15]. This enhances any ventilation-perfusion mismatch. Hypoxia and hypercarbia ensue.

THE CARDIOVASCULAR SYSTEM

A rise in LAP leads to a reduction in venous return [14]. This is due to functional obstruction of the inferior vena cava (IVC) as it moves from an area of high pressure (intra-abdominal) to an area of

INTRACRANIAL PRESSURE (ICP)

An increase in ICP is caused by restriction in lumbar venous return. In addition, a raised ICP and a

VASCULAR

EMERGENCIES

reduced cerebral perfusion pressure have been shown to be caused by rises in intrathoracic and pleural pressures causing functional venous outflow obstruction to the cerebral veins via the jugular venous system [16].

42 mmHg [7]. Meldrum (see below) devised a grading system for management, based on the value of IAP [5].

CLINICAL FEATURES
The clinical features of a distended and tense abdomen, the requirement of high atrial pressures to maintain cardiac output, increasing peak airway pressures with increasing hypoxia and hypercarbia, and oliguria are consistent with ACS.

GASTROINTESTINAL SYSTEM
Effects of increased IAP have been identified in hepatic, gastric, and intestinal perfusion. Hepatic blood flow may become significantly reduced with even mild IAH, and this is independent of cardiac output. A fall in gastric mucosal pH with IAH may precede the development of ACS. In an animal model, Diebel et al. [17] demonstrated a reduction in ileal mucosal blood flow and evidence of bacterial translocation with a raised LAP. They proposed that this might explain septic complications and organ failure in ACS. However, the part that bacterial translocation plays in ACS is not proven. Friedlander et al. [18] demonstrated a reduction in superior mesenteric arterial (SMA) flow following hemorrhage and resuscitation before increasing IAP. The reduction in SMA flow appeared greater than would be expected from the fall in cardiac output. It was concluded that optimization of hemodynamic status was not sufficient to maintain SMA flow but that early abdominal decompression was important.

MEASUREMENT OF IAP (TABLE II)

In a recent prospective blinded study, the sensitivity and accuracy of abdominal examination in assessing IAP were evaluated. When IAP was 15 mmHg, sensitivity and accuracy were 56% and 84%, respectively [19]. This implies that clinical examination is insufficient on its own at determining whether LAP is increased to such levels. However, it is not clear whether the accuracy would be sufficient at the level of pressure normally encountered in ACS. The methods of measurement of IAP can be divided into direct and indirect techniques. Direct techniques have usually been employed in animal experiments, although laparoscopy has permitted the direct assessment of IAP via a pressure transducer connected to the laparoscope. Indirect methods have largely been used to measure IAP by assuming that visceral luminal pressure reflects the IAP generally. The most popular technique involves utilizing a bladder catheter. It was first described by Kron et al. [1] in 1984 and was later validated. It is thought that the bladder acts as a passive diaphragm when 50 to 100 mL is instilled into an

16
160

ABDOMINAL WALL
The abdominal cavity can accommodate small volumes with relatively little increase in IAP. However, the pressure-volume curve for the abdominal wall is exponential, and as IAH ensues, relatively small volume changes create large rises in IAP. IAH has been shown to reduce abdominal wall muscle perfusion, which increases the likelihood of wound healing problems including dehiscence or later herniation. Furthermore, abdominal wall ischemia may predispose to wound infection.

Diagnosis
This relies on the accurate measurement of IAP, having established that clinical findings are consistent with ACS. The difficulty arises where IAP is not very high, since organ dysfunction is frequently multifactorial. In Hong's cohort of patients, those with IAH that did not require decompressive laparotomy for ACS had a mean IAP of 26 mmHg, while those that went on to develop ACS had a mean of

Direct

Metal cannula Intra-peritoneal catheter Laparoscopy

Indirect

Inferior vena cava Nasogastric Urinary Catheter

THE ABDOMINAL COMPARTMENT empty bladder. When the volume of sterile saline has been introduced into the bladder via the Foley catheter, the sterile tubing of a catheter bag is crossclamped just distal to the culture aspiration port. The end of the tubing is connected to the catheter and the clamp partly released such that fluid from the bladder can fill the tube proximal to the clamp. After reclamping the tube, a 16-guage needle is placed through the culture aspiration port and connected to a pressure transducer. The top of the symphisis pubis is used as the zero point with the patient supine. The technique may be unreliable where the bladder is small and contracted or where there is a large compressive pelvic hematoma. Once measurements have begun, they should be repeated approximately every four to six hours to detect adverse changes in IAP. OTHER INVESTIGATIONS Since ACS is a clinical diagnosis, computed tomography scanning should not be used routinely as an investigation. Recently, Pickhard et al. [20] identified findings common to four patients with proven ACS. These included tense infiltration of the

SYNDROME

retroperitoneum, extrinsic IVC compression by hematoma or exudates, and an increase of antero-posterior to transverse diameter (positive round belly sign). Gastric tonometry has been shown to identify patients with IAH likely to go on to ACS, although it is not widely used.

Management (Figure)
SUPPORTIVE TREATMENT The aim of supportive treatment should be to maximize cardiac filling pressure by fluid resuscitation, and inotropic support as necessary. An accurate assessment will be required by the use of a central line and Swann-Ganz catheter to determine pulmonary artery wedge pressures and cardiac output. In addition, attempts should be made to treat coagulopathy using clotting factors, and hypothermia by warming. Cheatham et al. [21] have recently described the use of abdominal perfusion pressure. The abdominal perfusion pressure is to equal the difference between

ii
161

FIGURE

Treatment diasram for abdominal compartment syndrome related to intra-abdominal pressures.

VASCULAR

EMERGENCIES

the mean arterial pressure and the ZAP and was found to be a better predictor of survival and a better target for resuscitation. A grading system has been proposed for the management of IAH (Table III). The authors of this study felt that purely volume resuscitation may be appropriate for an LAP of up to 25 mmHg. As the pressure exceeded 25 mm Hg, tissue perfusion pressure was thought to be impaired, requiring decompression. When greater than 35 mmHg, the cause was felt to be an arterial bleed and it was believed that decompression followed by a further reexploration was necessary [5].

repaired several months later. A rectus sheath myocutaneous advancement flap or an absorbable mesh and skin graft can be used for definitive closure. An alternative to decompressive laparotomy has been described in an animal model, where continuous external negative abdominal pressure was evaluated for the reversal of some of the effects of IAH. Continuous external negative abdominal pressure appeared to reduce IAP, central venous pressure, and intracranial pressure significantly, and there was a trend for a lower peak inspiratory pres-

ABDOMINAL DECOMPRESSION AND CLOSURE

The mainstay of treatment of ACS is abdominal decompression. It may be undertaken prophylactically, following a laparotomy where the wound would otherwise have to be closed under extreme tension or where the risks of ACS are thought to be high, such as massive hemorrhage and fluid resuscitation and the use of abdominal packs. Despite the use of pre-emptive measures, ACS can still occur, although with a lower incidence [9]. In a 10-year case control study at the Mayo Clinic, patients who underwent primary mesh closure were compared with those who underwent standard closure but then required a mesh for IAH. The two groups of patients had similar demographic and clinical profiles, but multi-organ failure scores, endoscopically detected colonic ischemia rates, and mortality from multi-organ failure were lower in the prophylactic mesh group. Various pre-operative and intra-operative risk factors were identified in those requiring a mesh postoperatively who would benefit in a prophylactic mesh. These included pre-operative anemia, prolonged shock and cardiac arrest, intra-operative massive resuscitation, profound hypothermia, and severe acidosis [22]. The type of closure can be either temporary or definitive (Table IV). Temporary closure may be achieved with a variety of methods. For example, a large fluid bag or polypropylene mesh can be sutured in a tension-free manner to the abdominal wound. It is also important that any closure keep fluid and viscera reasonably contained. After several days, when the edema has subsided, an attempt can be made at closure. Where this is not possible, the mesh can be reduced in size as the edema improves or the abdomen can be closed in stages, where temporary bag or mesh can be removed and replaced with a skin graft. The large ventral hernia can be Table ffl
^ , Grade

GRADING OF THE ABDOMINAL

COMPARTMENT SYNDROME [5]

Bladder pressure \T mmHg

Recommendation

10-15 16-25
III

Maintain normovolemia Hypervolemic resuscitation Decompression Decompression and re-exploration

16
162

26-35
>35

IV

Table IV Temporary

METHODS OF ABDOMINAL WALL CLOSURE

Skin clips PTFE Polypropylene mesh Fluid bag, e.g., "Bagota" Zipper

Definitive

Myocutaneous advancement flap Skin grafting of an absorbable mesh

THE ABDOMINAL COMPARTMENT SYNDROME

sure [23]. It is suggested that such a device could be used in humans when IAH occurs but where a decompressive laparotomy is not required. Despite the immediate improvements noted, the mortality of this subgroup of patients remains high and in excess of 50% [7,15].

Outcome of treatment
Abdominal decompression has been shown to improve cardiorespiratory and renal function. A sudden reduction in IAP leads to a fall in blood pressure as a result of the sudden reduction in peripheral vascular resistance. Patients must therefore be volume resuscitated and inotropes may be required.

Conclusion
Although ACS is relatively rare, it remains common enough for any vascular surgeon to recognize its existence. Prophylactic measures such as abdominal wall mesh closure or prompt decompressive laparotomy can reduce mortality of this otherwise universally fatal condition.

R E F E R E N C E S 1 Kron IL, Harman PK, Nolan SP. The measurement of intraabdominal pressure a criterion for abdominal re-exploration. Ann Surg 1984; 199: 28 -30. 2 Sugrue M. Intra-abdominal pressure. Clinical Intensive Care 1995; 6:76-79. 3 Sanchez NC, Tenofsky PL, Dort JM et al. What is normal intraabdominal pressure? Am Surg mi; 67: 243-248. 4 Sugrue M, Buist MD, Hourihan F et al. Prospective study of intra-abdominal hypertension and renal function after laparotomy. BrJSurgWb; 82: 235-238. 5 Meldrum DR, Moore FA, Moore EE et al. Prospective characterization and selective management of the abdominal compartment syndrome. AmJSurglWT, 174: 667-673. 6 Ivatury PvR, Porter JM, Simon RJ et al. Intra-abdominal hypertension after life-threatening penetrating abdominal trauma: prophylaxis, incidence and clinical relevance to gastric mucosal pH and abdominal compartment syndrome. / Trauma 1998; 44: 1016-1023. 7 Hong JJ, Cohn SM, Perez JM et al. Prospective study of the incidence and outcome of intra-abdominal hypertension and the abdominal compartment syndrome. 5r/Surg 2002; 89: 591-596. 8 Mayberry JC, Goldman RK, Mullins RJ et al. Surveyed opinion of American trauma surgeons on the prevention of the abdominal compartment syndrome. JTrauma 1999; 47: 509-514, 9 Ertel W, Oberholzer A, Platz A et al. Incidence and clinical pattern of the abdominal compartment syndrome after "damagecontrol" laparotomy in 311 patients with severe abdominal and/ or pelvic trauma. Crit Care Med 2000; 28:1747-1753. 10 Fietsam RJr. ( Villalba M, Glover JL, Clark K. Intra-abdominal compartment syndrome as a complication of ruptured abdominal aortic aneurysm repair. Am Swrgl989; 55: 396-402. 11 Maxwell RA, Fabian TC, Croce MA, Davis KA. Secondary abdominal compartment syndrome: an underappreciated manifestation of severe hemorrhagic shock. J Trauma 1999; 47: 995-999. 12 Simon RJ, Friedlander MH, Ivatury RR et al. Hemorrhage lowers the threshold for intra-abdominal hypertension-induced pulmonary dysfunction./Trauma 1997; 42: 398-405. 13 OdaJ, Ivatury RR, Blocher CR et al. Amplified cytokine response and lung injury by sequential hemorrhagic shock and abdominal compartment syndrome in a laboratory model of ischaemiareperfusion./Trauma 2002; 52: 625-632. Schein M, Wittman DH, Aprahamian CC, Condon RE. The abdominal compartment syndrome: the physiological and clinical consequences of elevated intra-abdominal pressure. / Am Coll Surg 1995; 180: 745 -753. Cullen DJ, Coyle JP, Teplick R, Long MC. Cardiovascular, pulmonary and renal effects of massively increased intraabdominal pressure in critically ill patients. Crit Care Med 1989; 17:118-121. Bloomfield GL, Ridings PC, Blocher CR et al. A proposed relationship between increased intra-abdominal, intrathoracic and intracranial pressure. Crit Care Med 1997; 25: 496-503. Diebel LN, Dulchavsky SA, Brown WJ. Splanchnic ischemia and bacterial translocation in the abdominal compartment syndrome. J Trauma 1997; 43: 852 - 855. Friedlander MH, Simon RJ, Ivatury R et al. Effect of hemorrhage on superior mesenteric artery flow during increased intraabdominal pressures. J Trauma 1998; 45: 433-439. Kirkpatrick AW, Brenneman FD, McLean RF et al. Is clinical examination an accurate indicator of raised intra-abdominal pressure in critically injured patients? Can J Surg 2000; 43: 207-211 Pickhardt PJ, ShimonyJS, Heiken JP et al. The abdominal compartment syndrome: CT findings. AJRAmJRoentgenol 1999; 173: 575-579. Cheatham ML, White MW, Sagraves SG et al. Abdominal perfusion pressure: a superior parameter in the assessment of intraabdominal hypertension. / Trauma 2000; 49: 621-627. Rasmussen TE, Hallett JW Jr., Noel AA et al. Early abdominal closure with mesh reduces multiple organ failure after ruptured abdominal aortic aneurysm repair: guidelines from a 10-year case control study./ Vase SMrg2002; 35: 246-253. Bloomfield G, Saggi B, Blocher C, Sugerman H. Physiologic effects of externally applied continuous negative abdominal pressure for intra-abdominal hypertension. J Trauma 1999; 46: 1009-1016.

14

15

16 17 18 19

163

20 21 22

23

17
ACUTE THROMBOSIS OF ILIOCAVAL VEINS
GUNNAR PLATE, LARS NORGREN

Pulmonary embolism and post-thrombotic sequelae were earlier common and frequently deleterious complications of deep venous thrombosis (DVT). An important landmark in the treatment of DVT was the development ofheparin and coumarin derivatives during the 1930s and 1940s. Systemic thrombolysis was attempted during the 1960s, and has recently been followed by catheter-directed thrombolysis. Although surgical removal of DVT was initially attempted during the early 20th century [1], it was not until the late 1970s that venous thrombectomy gained a more widespread use. At present, DVT of the infrapopliteal, popliteal, and femoral veins is usually treated with heparin, unfractionated heparin or low molecular weight heparin, oral anticoagulation, and compression. A more invasive treatment may be considered for more proximal thrombi in the iliofemoral or iliocaval segment.

17
165

Pathophysiology
Proximal DVT, iliofemoral, or iliocaval thromboses may have the same etiology as any DVT, e.g., thrombophilia, malignancy, trauma, and surgery, or may be of idiopathic origin. There are some specific features with pelvic vein obstruction, however. Compression of the left iliac vein by an overriding iliac artery, the May-Thurner syndrome [2] is a common cause of iliofemoral thromboses and a caval vein thrombosis is sometimes due to malformations of the venous anatomy. Blue phlegmasia

(phlegmasia cerulea dokns) induced by a rapid occlusion of proximal veins is frequently associated with malignant disease [3]. The major hazards associated with proximal DVT are occurrence of pulmonary embolism and later development of post-thrombotic sequelae. Scintigraphic evidence of pulmonary embolism is present in 40% to 50% of patients with iliofemoral thrombosis [4]. The most severe consequence of the post-thrombotic syndrome, leg ulceration, is caused mainly by destruction of the femoropopliteal vein valves [5] sometimes in combination with

VASCULAR

EMERGENCIES

venous outflow obstruction. An impaired outflow may also cause venous claudication.

Treatment
CONSERVATIVE TREATMENT
Patients of old age and/or with extensive malignancy and without signs of quickly developing blue phlegmasia should be treated with anticoagulation only. Infusion of unfractionated heparin or subcutaneous low molecular weight heparin injections followed by oral anticoagulation combined with compression treatment of the leg should be initiated. Such treatment alone is always acceptable and other treatment modalities are always combined with anticoagulation.

Clinical presentation
Although some proximal thromboses may develop slowly and give rise to limited symptoms, the most frequent presentation is a rapid swelling of the leg with pain in the leg or groin. The pain may be caused by an inflammatory process in the veins, phlebitis, but can also be caused by an increased compartmental pressure, giving rise to an impaired tissue perfusion, which is the underlying mechanism in phkgmasia ceruka dolens. Clinical examination usually reveals a swelling of the entire limb with discoloration and dilated superficial veins. An inferior vena cava thrombosis may cause symptoms from both legs simultaneously. Not infrequently, patients experience symptoms of pulmonary embolism even before leg swelling is evident.

VENOUS THROMBECTOMY
During the early experience with venous thrombectomy, it was learned that, while pulmonary embolism was prevented [6], the risk of post-thrombotic sequelae and vein valve incompetence was not abolished [7]. Since the late 1970s, there has been a new interest in venous thrombectomy [8]. The main objectives of venous thrombectomy are to re-establish and preserve venous patency and to preserve vein valve function, thereby avoiding major complications like venous gangrene, pulmonary embolism, and late post-thrombotic sequelae. The operative mortality has been extremely low in several series [8]. Since thrombectomy is usually an effective method to remove thrombi from the deep venous system, development of venous gangrene is usually prevented. In addition, venous thrombectomy is often successful in restoring iliac vein patency even when systemic thrombolysis has failed [9]. Early results have been excellent [8] (TableI). The long-term benefit with surgery has not been convincingly proven, however (Table II). At 5- and 10-year re-examinations in the only prospective randomized study for iliofemoral thrombosis, the clinical and physiologic results were slightly better following venous thrombectomy combined with an arteriovenous fistula compared to anticoagulation treatment alone [13]. Iliofemoral venous patency and venous outflow were better following surgical than medical treatment, but valvular competence was not clearly preserved by this surgical procedure. The long-term results suffer because of several missing follow-up examinations, which might obscure a possibly beneficial effect. Still, venous hypertension was significantly less prevalent following surgery than with anticoagulation treatment alone [13]. As with thrombolytic treatment, the question still re-

Diagnosis

17
166

A carefully obtained history is crucial for the selection of therapy. Previous episodes of DVT and the duration and extent of the current episode influence the choice of treatment. Diagnostic principles also depend on which treatment could be reasonable. It is therefore important to determine the etiology of the thrombosis, especially whether a neoplasm may have caused the present situation. It is also important to check for bleeding disorders. Duplex ultrasonography is most useful in the investigation of DVT of the leg veins, although the pelvic veins may be difficult to demonstrate. Computed tomography scanning may assist in the evaluation of the iliocaval veins and could further reveal malignancies in this region. The same may be true for magnetic resonance imaging. There is also a place for venography. In order to delineate the proximal extension of an iliofemoral or iliocaval thrombosis, puncture of both femoral veins and catheterization of an arm vein to reach the upper extension of a thrombus in the inferior vena cava is sometimes required. This kind of extensive investigation is only performed in those cases where an interventional treatment is planned. When a catheter-directed thrombolytic treatment is contemplated (see below), puncture of the ipsilateral popliteal vein is the preferred approach.

ACUTE THROMBOSIS mains whether the post-thrombotic sequelae including leg ulceration are prevented by this surgical procedure.

OF ILIOCAVAL

VEINS

Indication for surgery

Both our and other's experience show that a careful patient selection is of utmost importance in achieving good results. The main indications for venous thrombectomy are phlegmasia cerulea dolens and other iliofemoral thrombosis if the thrombus is fresh (less than five to seven days) and the patient is young and healthy. The success of surgery depends greatly on the age of the thrombus. After five to seven days, organization of the thrombosis

and vein valve destruction makes complete venous clearance impossible, and rethrombosis and valvular incompetence are likely to develop after thrombectomy. A careful analysis of the presenting symptoms and the history are therefore crucial, since no better means of establishing the age of the thrombus are available. Occasionally, venography may demonstrate recanalized veins without valves indicating previous thrombosis, but other signs, e.g., the appearance of the thrombus and presence of collaterals, are less reliable indicators of thrombus age. The extension of the thrombosis has implications on the operative technique but not as much on selection of treatment. Thromboses limited to the iliac veins are usually easy to remove surgically with good long-term results, but these patients also do

Table I

POSTOPERATIVE RESULTS OF ILIOCAVAL VENOUS THROMBECTOMY

Year of publication Period Number of thrombectomies Mortality Pulmonary embolism Primary patency Secondary patency

[ref.]

Capdevilla [10] Juhan [11]

1993 1997

1974-1991 NP

177 77

3.3

3.9 0

NA 87

II 167

NA: not available

, ., t author r f [rej.Jl

, f Ar Number of ., i . thrombectomies

, r bollow-up '

No , sequelae v

Moderate , sequelae V

Severe , sequelae ot

Patency at 5 years %
79 84

Patency at 10 years %
71 84

Capdevilla [10] Juhan [11]

177

71 22 months 8.5 years (5-13 years) 10 years

63.2
80

30.9 13.7

5.9 6.3

77

Plate [12]

13

54

NA

42*

* Occlusion 17%, stenosis 41%, normal patency 42% NA: not available

VASCULAR

EMERGENCIES

11
168

fairly well with conservative treatment alone. Complete clearance of the deep venous system is possible in more than a third of patients with a thrombosis involving the entire extremity, with excellent long-term outcome. Even if the late results following thrombectomy of extensive thrombosis are not as good as with proximal thrombosis, these patients also do worse following conservative treatment [14]. Therefore, the extension of the thrombus per se does not affect the decision for or against surgery. There are no comparative therapeutic results of vena cava thromboses. Deep venous thromboses limited to the femoropopliteal segment have occasionally been operated upon, but the usefulness of surgery has not been established in these cases. Iliofemoral thromboses developing during pregnancy must be managed in close cooperation with an obstetrician. Venographic examination should be minimized to avoid undue radiation, but venograms necessary to decide the proper treatment and to perform a safe procedure must be obtained. During early pregnancy, thrombectomy combined with an arteriovenous fistula may be performed safely. In these cases, the fistula is not closed until several weeks after delivery. During late pregnancy, venous thrombectomy may be combined with simultaneous caesarean section. Since these patients are usually young, they are likely to develop late sequelae if a venous outflow obstruction persists and if the peripheral vein valves are destroyed. Pregnant women and women taking contraceptives are presently the two groups most commonly subject to venous thrombectomy [15].

Operative technique
Routine laboratory tests including anticoagulation studies, blood typing, and cross-matching should be administered at admission. Full dose intravenous anticoagulation treatment should be initiated in cooperation with the radiologist and the anesthetist to decrease the risk of thrombus propagation or embolization during any time required for diagnostic procedures and preparation for surgery. Prophylactic antibiotics are administered to reduce the risk of wound infection, which is otherwise common in this edematous tissue in proximity to the groin. It must be remembered that these patients have accumulated a significant amount of fluid in

their swollen leg and are often hypovolemic. This volume deficit must be adjusted by intravenous administration of crystalloid and/or colloid solutions prior to surgery. The operation may be performed under local or epidural anesthesia, but general intubation anesthesia is preferred. This allows application of a positive end-expiratory pressure of 10 to 15 cm water during manipulation of the thrombus in the iliac or caval veins, which probably reduces the risk for pulmonary embolism. Arrangements for autotransfusion are optimal, since these patients may lose a significant amount of blood during clearance of thrombus material from the deep veins. The long saphenous vein, the common femoral vein, and the superficial femoral artery are exposed using a vertical or transverse groin incision with minimal perivascular dissection and meticulous hemostasis. A transverse venotomy is performed in the common femoral vein, controlling bleeding with digital compression or application of soft Fogarty clamps. A venous balloon catheter (Fogarty, n8-10) is introduced into the vena cava, the balloon is inflated with contrast material and intra-operative fluoroscopy is utilized to verify the position of the catheter in the vena cava. Repeated thrombectomy is performed until no further thrombus material is retrieved from the iliac veins. Occasionally, the catheter may pass into the ascending lumbar vein, which may rupture if the balloon is inflated. There is also a risk that thrombus material at the caval bifurcation may become dislodged into the vena cava during return of an inflated balloon from the ascending lumbar vein. Occasionally, it may be difficult to pass the catheter from the left iliac vein into the vena cava due to compression from the overlying artery and intraluminal fibrosis (venous spur). Sometimes this problem can be overcome by bending the tip of the catheter. A guide wire may also be helpful. The peripheral veins are cleared with simultaneous obstruction of the proximal common femoral vein to prevent embolization. If the thrombosis is fairly fresh, the peripheral thrombi are easily milked out of the deep venous system by vigorous manual compression from the foot toward the groin. This must be repeated several times and may cause significant blood loss, which must be noted and compensated. Some distal thrombus material may be retrieved by careful retrograde introduction of a balloon catheter into the deep and superficial femoral veins. Competent vein valves sometimes compromise this. Passage of these valves may be

ACUTE THROMBOSIS accomplished after careful inflation of the balloon to dilate the valvular region. When clearance of the peripheral veins has been completed, patency of the iliac vein is secured by additional thrombectomy to remove thrombi that may have become dislodged from the internal iliac vein or the peripheral veins. Clearance of the iliac segment is checked by intra-operative venography or angioscopy. The venotomy is closed with interrupted or continuous 6/0 polypropylene sutures. An arteriovenous fistula is constructed by performing an end-to-side anastomosis between the divided long saphenous vein and the superficial femoral artery. An infant-feeding catheter may be introduced through a (long) tributary of the saphenous vein and left in place for postoperative venographic control. A suction drainage is brought out through a separate stab incision and the wound is closed in layers avoiding kinking of the fistula.

OF ILIOCAVAL

VEINS

Technical problems
In the presence of thrombosis in the inferior vena cava, there is an increased risk of creating pulmonary emboli. Some surgeons have used additional balloon catheters introduced into the same venotomy or through the contralateral groin veins with the balloon inflated cephalad of the thrombus as protection against embolism during the procedure. This has been proven unnecessary if the thrombus is confined to the iliofemoral veins. A more secure control of the vena cava itself is preferable if this vein is also involved. In elderly patients with proven pulmonary embolism, it is often sufficient to insert a permanent filter into the inferior vena cava cephalad of the thrombus. This is easily performed percutaneously or by surgical exposure of the right internal jugular vein. Various devices are currently available for intraluminal control of the vena cava [16]. The Greenfield filter is the most widely used filter, and has proven effective and safe. In most young patients, it is better to remove the caval thrombus in order to preserve venous outflow from both lower limbs. The inferior vena cava is exposed using a right-sided transverse infracostal incision with extraperitoneal dissection. The vein is controlled by application of soft Fogarty clamps. A longitudinal venotomy is performed and the vein edges are separated with stay stitches during careful removal of

the thrombus by milking, suction and use of venous Fogarty catheters. The venotomy is closed with continuous or interrupted 5/0 polypropylene stitches. It is advantageous to perform exploration of the vena cava simultaneously with thrombus removal from the groin in order to secure venous patency before the cavotomy is closed. With minimal protrusion of the thrombus into the vena cava, it is not always necessary to perform a cavotomy. Left-sided iliofemoral thromboses are often caused by compression from the overlying right common iliac artery with secondary intraluminal fibrosis, a so-called venous spur [2]. In these cases, the risk for rethrombosis is greatly increased and the long-term results are more dubious. Some surgeons use angioscopy and ring catheters to overcome such venous spurs [17]. Others suggest direct exposure and reconstruction of the caval inflow. The value of such aggressive measures has not been proven. Therefore, we have previously relied solely upon iliofemoral thrombectomy in combination with an arteriovenous fistula to keep as much as possible of the vein patent and to promote development of collateral flow. Recently, we have utilized intra-operative or postoperative balloon angioplasty with stent application with some success. In ascending thrombosis where the peripheral thrombus is of an older age, clearance of the peripheral veins may be difficult. Gruss [18] has suggested exploration of the popliteal and/or tibial veins with construction of a peripheral fistula to allow clearance and maintenance of venous patency in such cases. Intra-operative administration of local streptokinase has also been used. The benefit of these additional procedures is not obvious, since it is likely that the valvular function has already been destroyed in cases with old thrombosis. Hence, ligation of the superficial femoral vein has been proposed to prevent pulmonary embolism and late valvular incompetence. Although still controversial, it seems that this procedure is fairly well tolerated if the deep femoral vein is patent and competent. In severe cases of phlegmasia ceruka dolens, increased pressures are often present in all muscle compartments of the calf. It is of utmost importance to achieve a rapid decompression of these compartments. Except for immediate and successful thrombectomy, this is best accomplished by fasciotomy of all muscle compartments of the calf. In less severe cases, the need for fasciotomy may be assessed by measurement of intracompartmental pressures, the critical level being around 30 mmHg [19].

17
169

VASCULAR

EMERGENCIES

Postoperative management
Repeated thrombectomy combined with a venous reconstruction may be considered in cases of early rethrombosis with massive leg swelling, although the long-term patency of such procedures has not been convincing [20]. The preferential procedure would be a cross-femoral or iliocaval bypass using a large autogenous vein or ringed polytetrafluoroethylene prosthesis with a protective arteriovenous fistula. Percutaneous transluminal balloon dilation combined with stent implantation has occasionally been successful in treating residual stenoses of the iliac vein in combination with repeated thrombectomy. Full anticoagulation should be resumed as soon as possible (meticulous operative hemostasis is therefore crucial). Oral anticoagulation is started on the first postoperative day and continued for at least six months. The patient is allowed to ambulate on the first postoperative day. Compressive stocking support should be used for several months and as long as leg swelling persists. In most cases the patient may be discharged from the hospital one week after the operation. The arteriovenous fistula can be closed after six to eight weeks, when healing of the venous endothelium has occurred. With a superficially placed fistula, this is easily accomplished using local anesthesia and a short skin incision over the maximum thrill created by the fistula. An intra-operative venogram may be obtained to check iliofemoral venous patency. The fistula is then divided and ligated. This procedure is performed as an outpatient procedure. Other techniques using interventional radiology have also been used.

with DVT did not present centra-indications to systemic thrombolytic treatment. In a recent meta-analysis [22], it was again stressed that despite early vein patency after systemic thrombolysis, there is no proven reduction in the incidence of post-thrombotic sequelae. One investigation showed that, out of 35 patients treated with streptokinase, only 2 patients were free from symptoms and had normal venous function after a mean of 29 months followup [23].

CATHETER-DIRECTED THROMBOLYSIS
To reduce the risk of bleeding during systemic thrombolysis, the thrombolytic agent may be infused directly into the thrombus using intravenous catheters. This catheter-directed thrombolysis may also ameliorate the lysis and decrease the time required. The presently preferred approach has been through the ipsilateral popliteal vein using ultrasound guidance to avoid additional venous or arterial punctures. An alternative approach would be via the internal jugular vein. The lytic agents most often used in this setting have been urokinase and recombinant tissue plasminogen activator administered using pulse-spray technique or a slow infusion. The effect of the thrombolysis checked frequently with venography and the position of the catheter is corrected accordingly. Initially, most centers applied a temporary vena cava filter to prevent pulmonary embolism during the thrombolysis. This has now been abandoned at most places. Following completed thrombolysis, an underlying stenotic process, e.g., the May-Thurner syndrome, may be dealt with. Balloon dilation of such lesions usually combined with placement of venous stents increases the chance of restoring patency thereby reducing the risk for recurrent thrombosis [24]. A few larger short-term studies of catheter-directed thrombolysis have been published. An excellent technical success rate was demonstrated in a prospective study from 1997 [25]. A national multicenter registry report was presented in 1999 claiming that the treatment is safe and effective [26]. Major bleeding complications in 11% and a primary iliofemoral patency rate of 60% at one year were reported. In these centers, urokinase was used as the thrombolytic agent and 71% of the thromboses treated were confined to the iliofemoral segment; the remaining thrombi were localized more distally. Recently, a randomized controlled trial comparing catheter-directed thrombolysis with anticoa-

17
170

INDUCED THROMBOLYSIS
Systemic thrombolysis with either streptokinase or urokinase may dissolve thrombotic material and restore venous patency. In a pooled analysis f:~om six randomized studies evaluating streptokinase treatment versus heparin, it was noted that thrombolysis occurred almost four times more often after streptokinase treatment. However, bleeding complications are common and recurrent thrombosis giving rise to post-thrombotic sequelae has been seen frequently. In three studies that allowed comparison of bleeding complications, thrombolysis induced a three times greater rate of bleeding than after heparin treatment [21]. In addition, various reports have shown that only 7% to 20% of patients

ACUTE THROMBOSIS OF I LI oc AVAL VEINS

gulation treatment of iliofemoral venous thrombosis was presented [27]. In this study, the iliofemoral patency rate at 6 months was 72% following thrombolytic treatment compared to 12% with anticoagulation. In one series of 10 patients, catheter-directed thrombolysis and stent placement were used for treatment of the May-Thurner syndrome [28]. After a mean of 15 months, 9 of 12 patients were asymptomatic. It has also been shown that this mode of treatment improves health-related quality of life [29]. Neither of the series presented specifically studied patients with iliocaval thrombosis. With thrombotic engagement of the vena cava, one has to consider whether the risk of pulmonary embolism due to fragmentation of the thrombus during thrombolysis requires caval protection. In such cases, applications of a permanent filter as the sole solution or placement of a temporary filter during thrombolytic treatment are probably safer solutions. the age of the patient, extension of the thrombosis and the etiology (cancer, obesity, trauma, MayTurner syndrome). Anticoagulation combined with elastic stocking is a generally accepted treatment especially in older patients or patients with cancer (Figure). A more aggressive attitude is justified in younger patients in order to avoid the late post-thrombotic sequelae. Non-randomized studies show the beneficial effects of thrombolysis or surgical thrombectomy compared to anticoagulation treatment (Table III), but there are no studies comparing the results of thrombolysis and thrombectomy. Thrombolysis has gained a predominant role in the treatment of iliocaval thromboses. Some authors consider all deep venous thromboses an indication for fibrinolytic teratment, irrespective location or delay. Mewissen et al. [26] are more conservative and do not apply fibrinolysis for recurrent thromboses or femoropopliteal thromboses older than 10 days. Obviously, the centra-indications for thrombolysis are directly determined by the potential complications of the treatment. At present, iliocaval surgical thrombectomy should be considered for recent iliocaval thromboses in young patients inwhom a contra-indication or failure of thrombolytic treatment can be expected.

Choice of treatment
The choice between anticoagulation treatment, thrombolysis and surgical thrombectomy depends on many factors, the most important of which are

17
171

60 - 70 years
(Or impending gangrene)

> 60 - 70 years
(Or disseminated malignancy, or contra-indications)

Centra-indication for thrombolysis?

Anticoagulation

NO

YES

Catheter-directed thrombolysis (or anticoagulation)

Surgical thrombectomy (or anticoagulation)

FIGURE. Treatment algorhythm for iliofemoral and iliocaval thrombosis. Our conclusion is that anticoagulation is usually an accepted method, althoush a more agsressive approach should be attempted in young patients to avoid late sequelae.

VASCULAR

EMERGENCIES

1st author [ref.]

Plate [12]

Year of publication

Treatment
Anticoagulation (n=l7)

Patency

No sequelae

1997 2001
2002

10 years

Thrombectomy
AbuRahma [24]

(n=13)

29 42 18 69 12 72

24 42 78 30

Anticoagulation (n=33)

5 years

Endovascular
Elsharawy [27]
Anticoagulation

(n=18) (n=17) (n=18) 6 months

Thrombolysis
Endovascular: thrombolysis angioplasty stent

R E F E R E N C E S 1 Leriche R. Traitement chirurgical des suites eloignees des phlebites et des grands cedemes non medicaux des membres inferieurs. Bull Mem Soc Nat Chir 1927; 53: 187-195. 2 May R, Turner]. The cause of the predominantly sinistral occurrence of thrombosis of pelvic veins. Angeiology 1957; 8: 419-425. 3 Perkins JM, Magee TR, Galland RB. Phlegraasia caerulea dolens and venous gangrene. BrJ Surg 1996; 83: 19-23. 4 Plate G, Ohlin P, Eklof B. Pulmonary embolism in acute iliofemoral venous thrombosis. BrJ Surg 1985; 72: 912-915. 5 Shull KG, Nicolaides AN, Fernandes e Fernandes J et al. Significance of popliteal reflux in relation to ambulatory venous pressure and ulceration. Arch Surg 1979; 114: 1304-1306. 6 Fontaine R, Tuchmann L. The role of thrombectomy in deep venous thromboses./ Cardiovasc Surg 1964; 5: 298-312. 7 Lansing AM, Davis WM. Five-year follow-up study of iliofemoral venous thrombectomy. Ann Surg 1968; 168: 620-628. 8 Plate G, Einarsson E, Ohlin P et al. Thrombectomy with temporary arteriovenous fistula: the treatment of choice in acute iliofemoral venous thrombosis. / Vase Surg 1984; 1: 867-876. 9 Stiegler H, Hiller E, Arbogast H et al. Thrombectomy after unsuccessful thrombolytic therapy of deep leg vein thromboses: an effective procedure? Vasa 1992; 21: 280-288. 10 Capdevilla JM, Estallo L, Ballon H, RancanoJ. Resultats de la thrombectomie veineuse. In: Branchereau A, Jausseran JM. Chirurgie des veines profondes. Marseille, CVN, 1993 : pp 32-36. 11 Juhan CM, Alimi YS, Barthelemy PJ et al. Late results of iliofemoral venous thrombectomy. / Vase Surg 1997; 25: 417-422. 12 Plate G, Eklof B, Norgren L et al. Venous thrombectomy for iliofemoral vein thrombosis. Ten-year results of a prospective randomised study. EurJ VaseEndovasc Surg 1997; 14: 367-374. 13 Plate G, Akesson H, Einarsson E et al. Long-term results of venous thrombectomy combined with a temporary arteriovenous fistula. EurJ Vase Surg 1990; 4: 483-489. 14 Browse NL, Clemenson G, Thomas ML. Is the postphlebitic leg always postphlebitic? Relation between phlebographic appearances of deep-vein thrombosis and late sequelae. BrMedJIQSO; 281: 1167-1170. 15 Torngren S, Bremme K, Hjertberg R, SwedenborgJ. Late results of thrombectomy for ilio-femoral venous thrombosis. Phlebology 1991; 6: 249-254. 16 Greenfield LJ, DeLucia A 3rd. Endovascular therapy of venous thromboembolic disease. Surg Clin North Am 1992; 72: 964-989. 17 Jakob H, Maass D, Schmiedt W et al. Treatment of major venous obstruction with an expandable endoluminal spiral prosthesis. J Cardiovasc Surg 1989; 30: 112-117. 18 Gruss JD. Venous reconstruction. Part 1. Phlebology 1988; 3: 7-18. 19 Qyarfordt P, Eklof B, Ohlin P. Intramuscular pressure in the lower leg in deep vein thrombosis and phlegmasia cerulea dolens. Ann Surg 1983; 197: 450-453. 20 Plate G, Einarsson E, Eklof B et al. Iliac vein obstruction associated with acute iliofemoral venous thrombosis. Results of early reconstruction using polytetrafluoroethylene grafts. Ada Chir Stand 1985; 151:607-611. 21 Goldhaber SZ, Buring JE, Lipnick RJ, Hennekens CH. Pooled analyses of randomized trials of streptokinase and heparin in phlebographically documented acute deep venous thrombosis. An JMed 1984; 76: 393 -397. 22 Wells PS, Forster AJ. Thrombolysis in deep vein thrombosis: is there still an indication? Thromb Haemost 2001; 86: 499-508. 23 Albrechtsson U, Anderson J, Einarsson E et al. Streptokinase treatment of deep venous thrombosis and the post-thrombotic syndrome. Follow-up evaluation of venous function. Arch Surg 1981; 116:33-37. 24 AbuRahma AF, Perkins SE, Wulu JT, Ng HK. Iliofemoral deep vein thrombosis: conventional therapy versus lysis and percutaneous transluminal angioplasty and stenting. Ann Surg 2001; 233: 752-760. 25 Bjarnason H, Kruse JR, Asinger DA et al. Iliofemoral deep venous thrombosis: safety and efficacy outcome during five years of catheter-directed thrombolytic therapy. / Vase Interv Radial 1997; 8: 405 -418. 26 Mewissen MW, Seabrook GR, Meissner MH et al. Catheterdirected thrombolysis for lower extremity deep venous thrombosis: report of a national multicenter registry. Radiology 1999; 211: 39-49. 27 Elsharawy M, Elzayat E. Early results of thrombolysis vs anticoagulation in iliofemoral venous thrombosis. A randomised clinical trial. EurJ Vase Endovasc Surg 2002; 24: 209-214. 28 Patel NH, Stookey KR, Ketcham DB, Cragg AH. Endovascular management of acute extensive iliofemoral deep venous thrombosis caused by May Turner syndrome. J Vase Interv Radial 2000; 11: 1297-1302. 29 Comerota AJ, Throm RC, Mathias SD et al. Catheter-directed thrombolysis for iliofemoral deep venous thrombosis improves health-related quality of life. / Vase Surg 2000; 32: 130-137.

17
172

18
ACUTE SUBCLAVIAN-AXILLARY VEIN THROMBOSIS
RAMON BOFILL, JOSEP ROYO, JOSE MARIA FUENTES, JOSE MARIA ESCRIBANO, MANUEL MATAS

Acute subclavian-axillary vein thrombosis (SAVT) is uncommon compared to deep vein thromboses in other vascular territories. While the risk of developing fatal pulmonary thromboembolism in this condition is relatively small, it can occur even with anticoagulant treatment. Furthermore, clinical manifestations of pain, cyanosis, swelling and functional limitations can be disabling. The development of secondary SAVT can be a consequence of general diseases, a state of hypercoagulation, or tumors, but the main cause is the increasing use of intravenous devices for diagnostic or therapeutic purposes. The primary form of SAVT, first described more than a century ago and known as Paget-Schroetter syndrome, is usually the result of repeated effort by the upper extremity and scapular girdle. This condition presents most frequently in young, otherwise healthy individuals performing strenuous occupational or recreational activities. The literature shows particular interest in this type of primary venous thrombosis, with reference to diagnostic and therapeutic methods directed toward recuperating patency of the affected venous vasculature and maintaining good long-term clinical and hemodynamic results. The main points of controversy regarding treatment reside in the following: timing of initiation of fibrinolytic therapy and anticoagulation; assessment of the need, ideal timing and choice of approach for surgical decompression of the thoracic outlet; and the suitability of other interventional techniques aimed toward recovering venous patency or correcting residual occlusive lesions after surgery. Nevertheless, experience documented in the literature regarding this condition shows relative agreement on some aspects. Since primary SAVT generally occurs in young patients at the height of physical activity, rehabilitation should be as quick and effective as possible. Thus, the process is considered to require other measures in

VASCULAR EMERGENCIES

addition to anticoagulant therapy. The current use of thrombolytic agents together with surgical or endovascular procedures, applied after careful assessment of each individual case, has resulted in a notable improvement in the outcome ofSAVT.

Etiology
SAVT can be divided into primary and secondary forms and each form requires a different type of therapeutic management. Primary SAVT was described 1875 and published in 1884 by J. Paget and Von Schroetter, respectively. In 1949 Hughes conducted a study involving more than 300 patients with acute spontaneous venous thrombosis of the upper extremities, and first used the term PagetSchroetter syndrome to refer to this condition. The causes of primary SAVT are mainly related to strenuous upper body activity, venous compression at the thoracic outlet, and upper limb immobility or anatomical abnormality causing intimal damage to the vein at the thoracic outlet. The anterior scalene muscle, the subclavian muscle and the clavicle against the first rib are three key points of narrowing of the superior thoracic outlet at which the axillary and subclavian veins can be compressed, particularly with movements of abduction. Anatomical anomalies can produce other susceptible points [1,2]. Repeated compression has a traumatic effect on the vein wall that leads to thickening, fibrosis and intimal proliferation. Thrombosis develops when heightened physical activity of the upper limb increases this traumatic effect. Secondary SAVT can be related to systemic diseases such as cardiac failure, an hypercoagulation state often associated with neoplasm, amyloidosis, sarcoidosis, or oral contraception, or as a consequence of a local aggression in the venous territory due to radiotherapy, instrumentation, indwelling catheters, pacemaker wires, parenteral nutrition, dialysis fistulae, intravenous drug abuse, shoulder trauma or tumor mass compression.

devices and catheters. It has been estimated that more than 18% of patients with central venous catheters develop partial or total thrombosis. Certain factors, such as the duration of catheter implantation and the type of material used in their manufacture, can influence the development of venous occlusion [3]. The possibility of presenting venous thrombosis is almost null before the sixth day after catheter placement and is relatively remote before the fifthteenth. The new materials used in catheter manufacture such as polyurethane, silicone and silicone elastomer, offer better results regarding maintenance of catheter patency and a smaller risk of pulmonary thrombo-embolism, as compared to polyvinyl or polyethylene [4].

Clinical presentation
Swelling, pain or discomfort, venous engorgement and mild cyanosis are the classic symptoms of SAVT. These symptoms become enhanced with increased activity of the affected extremity. Swelling of the hand and fingers is especially disabling and less tolerable than the swelling of the foot that occurs in post-thrombotic syndrome in the lower extremity, which is usually mild. The great majority of patients (more than 80%) develop these symptoms in the first 24 hours. Nonetheless, a delay in their presentation is not infrequent and may extend to months if the patient has had episodes of similar symptoms previously as a result of intermittent extrinsic venous compression without actual thrombosis [5]. SAVT occurs in both sexes and at all ages, but the characteristic patient with this condition is a young or middle-aged male, often presenting a history of vigorous use of the affected upper limb. Immobilization or forced positioning can be the cause of SAVT in an elderly person. The upper right limb is affected significantly more often, a fact explained by the preponderance of right-sided dominance in the general population. Secondary SAVT usually has a more benign expression, probably owing to the anatomical characteristics of the affected location and a more limited

18 174

Incidence
The primary form of SAVT accounts for less than 2% of all cases of deep venous thrombosis. The secondary form is becoming notably more prevalent due to the increasing use of central venous system

ACUTE SUBCLAVIAN-AXILLARY
extension of the thrombotic lesion. For this reason the secondary form of SAVT is usually more susceptible to the favorable effect of collateral circulation. From the hemodynamic standpoint, it is very demonstrative that dialysis patients with a thrombosed temporary venous access in the subclavian territory manifest symptoms of proximal venous occlusion only after establishment of an arteriovenous fistula at the level of the elbow. This is because the clinical effects of the thrombosis associated with the initial venous access are expressed only when the collateral circulation becomes overloaded and cannot absorb the increase in venous supply. There have been cases of SAVT secondary to paraneoplastic origin in which significant extension of the thrombus has produced symptoms of ischemia resulting from the difficulty in providing arterial supply due to compromised venous return.

VEIN

THROMBOSIS

Therapeutic management
The important etiological and clinical differences between the primary and secondary forms of SAVT are reflected in the therapy used for their management. Treatment for primary SAVT has changed considerably since Paget and Schroetter's description of this condition more than a century ago, but it is still controversial. Moreover the development of thrombolytic and endovascular treatment has provided a new arsenal to sustain this controversy. Treatment for secondary SAVT is mainly limited to postural measures directed toward favoring venous return, elastic contention and anticoagulant therapy, when there are no centra-indications. Surgical resection of tumors responsible for SAVT often provides good results [6].

ANTICOAGULANT THERAPY

Diagnostic methods
After recording the clinical history and performing a complete physical examination, doppler ultrasound is the initial technique used for establishing the diagnosis of SAVT. This noninvasive method is limited by bony structures that can produce acoustic shadowing and the occasional presence of highly developed collateral vessels that may induce diagnostic errors. The diagnostic reliability of doppler ultrasound is highest when results are positive in patients with clinical symptoms clearly suggestive of SAVT. Being a noninvasive, reliable technique that is comfortable for the patient, doppler ultrasound is also especially useful for repeated follow-up studies of any treatment applied. Nonetheless, the gold standard for the diagnosis of SAVT is still contrast enhanced venography, preferably through the basilica vein (Fig. 1). Despite its more invasive nature, it allows proper static and dynamic evaluation of the deep venous system and the collateral circulation, and visualization of extrinsic compression. Venography is also excellent for the application and control of thrombolytic treatment and any associated surgical or endovascular technique. The basic examination of any patient with suspected SAVT should also include general laboratory analyses with special attention to coagulation alterations, as well as radiography and computed tomography scanning directed toward the detection of thoracic outlet anomalies.

In the current therapeutic management of primary SAVT, anticoagulation has a co-adjuvant function with other treatments. Administration of heparin-Coumadin was advocated in the past because of its favorable effect on the stabilization and recanalization of primary thrombi and the prevention of pulmonary embolism [7,8]. Nowadays this monotherapy has lost favor because of its deficient long-term control of the disabling consequences of primary SAVT. The same can be said regarding elastic contention and postural measures. Moreover, in the use of anticoagulant treatment, as well as in the

175

FIG. 1

Thrombosis of right subclavian-axillary vein.

VASCULAR

EMERGENCIES

other therapeutic measure described below, there is still relevant controversy regarding factors related to timing [7-9]. In contrast, anticoagulation has a fundamental role in the treatment of secondary SAVT, even when the thrombosis is more extensive than is usually seen in these cases and the symptoms more intense. In general the behavior and course of the secondary form is similar to that of deep venous thrombosis in other locations. For this reason the use of anticoagulant treatment in secondary SAVT adheres to more standardized criteria, and the management of contra-indications and complications associated with heparin-Coumadin administration follows the same pattern as in other thromboses. Thus, the indication for vena cava filters for the prevention of pulmonary thrombo-embolism is the same in secondary SAVT as in cases where the sources of the emboli are the lower extremities.

FIG. 2 Venogram with acute thrombosis recanalized with lytic therapy.

THROMBECTOMY
Thrombectomy, initially introduced as an isolated surgical technique and later associated with thoracic oudet decompression, showed notably good outcome in some of the published series, though the number of patients included was not extensive. In at least one of these reports there was evidence that the results were favorable in the long term. Nevertheless, this is an aggressive approach that requires extensive dissection of the subclavian vein to insure proper control of lumen repair. Since the advent of thrombolytic therapy, thrombectomy has been displaced from its prominent position as the technique of choice for the treatment of SAVT [10]. and its application should be performed under protocolized analytic control. When angiography demonstrates no significant changes at 24 hours after the initiation of thrombolytic treatment, interruption of the treatment is recommended. At this time, there is no established protocol for dosage of thrombolytic agents. Table I shows an example of a possible therapeutic model from the various regimens used in the published experience of several authors. Most groups applying thrombolytic therapy use urokinase rather than streptokinase or RPT because of its relative safety and efficacy. Nonetheless, the major drawback to all these agents is still a risk of hemorrhage.

18
176

THROMBOLYTIC THERAPY (Fig. 2)

Catheter-directed thrombolysis is, in the opinion of many experts, the first therapeutic option to be used in well-selected cases of primary SAVT [11]. This variant of thrombolytic treatment presents several advantages over systemic thrombolysis. One of the main features is the possibility for real-time angiographic control of the results obtained in lysis of the thrombus. Moreover, the presence of extrinsic compression can be detected in functional examinations and corrected later. The maximum recommended interval between the initiation of the thrombotic episode and application of effective treatment is notably longer for thrombolysis than for thrombectomy, with delays of 5 to 10 days considered acceptable. Naturally, thrombolytic treatment is contra-indicated in patients at risk for hemorrhage

THORACIC OUTLET DECOMPRESSION

After restoring venous patency with thrombolytic treatment, persistence of the musculoskeletal mechanisms producing thoracic outlet compression in the axillary-subclavian venous territory will, of course, continue to act unfavorably (Fig. 3). This circumstance is clearly reflected in post-thrombolysis venography controls, which show filling defects when the patient is at rest or in positions that exacerbate the problem. Most authors agree that definitive correction of this factor is essential to avoid new episodes of thrombosis and to obtain long-term patency. Surgery, rather than endovascular therapy, proposed more than a decade ago, is generally re-

ACUTE SUBCLAVIAN-AXILLARY

VEIN

THROMBOSIS

Thrombolytic agent Urokinase Initial bolus First variable infusion rate Second variable infusion rate

> 200 000 U

150 000 - 250 000 U/h for 6 hours 60 000 -100 000 U/h for no more than 72 hours

Under monitored surgical intensive care

Heparin Na administered concomitanly () Heparin Na infusion concomitanly (*) () 5 000 U intravenous bolus

() Not in all groups (**) Infusion rate must be adapted to maintain PTT ratio at 2.0 - 2.5 times normal

commended as the best method for eliminating these compression mechanisms [5,12-15]. The poor outcome recorded with the use of endovascular treatment, specifically stent placement, is attributed to the compressive force of the musculoskeletal structures, which in these circumstances is superior to the capacity of the stent to maintain the lumen

of the vessel or the caliber of the vein sufficiently open. Deformity or breakage of the stent has occurred in some cases. Table II lists the various surgical techniques used for thoracic outlet decompression and the timing criteria applied by each author.

ENDOVASCULAR THERAPY
After successful thrombolytic treatment and proper thoracic outlet decompression, angiographic examination may still demonstrate vessel lumen irregularities due to residual thrombotic material or inflammatory alterations of the venous intima. Several opinions regarding therapy for these cases have been voiced in the literature, depending on the extension of the lesion and the patient's symptoms. In these circumstances endovascular therapy would seem to be a reasonable option. It is interesting to note that the introduction of endovascular thrombolysis has led to changes in the concept of conventional surgery for this condition, making it much less aggressive. The classic transclavicular approach with clavicle resection and direct repair of the subclavian vein has been substituted by the current use of the transaxillary, supraclavicular or infraclavicular approaches plus complementary endovascular therapy, particularly the use of stents [12,16,19]. In cases of contralateral stenoses detected during bilateral venography in patients with symptoms in only one upper limb, the benefit of prophylactic correction has not been demonstrated.

177

FIG. 3 Residual compressive abnormality at costoclavicular space after successful thrombolytic therapy during upper extremity abduccion.

VASCULAR

EMERGENCIES

1st author [ref.] Machleder [11]

Year

Number of patients
50

Immediate thrombolysis 43

Surgical decompression

Surgical approach

Average follow-up 3.1 years

1993

12 PTA pre-operatively 12 transaxillary 9 PTA postoperatively 1 transclavicular 6 first rib resection

Azakie [12]

1998

33

2 transaxillary 31 months 32 supraclavicular 21 plus infraclavicular 5 years

Heron [8]

1999

54

0 thrombolysis 48 anticoagulation alone 18

3 first rib resection

Lee [15]

2000

22 18

13 scalenectomy 1 first rib resection 9 early surgical decompression 9 staged surgical decompression 8 first rib resection + venolysis 10 first rib resection 9 scalenectomy 3 bypass 16 23 2 14 first rib resection scalenectomy exostosis resection stent

13 supraclavicular

22.3 months

Angle [16]

2000

18

18 transaxillary

3 months

Arko [17]

2001

12 10

12

3.9 months

18 178

Feugier [14]

2001

45 months

Kreienberg [18]

2001

23

4 years

Lokanathan [13] 2001

25

1 claviculectomy 12 PTA

1 transaxillary

2.9 years

PTA: percutaneous transluminal angioplasty

MANAGEMENT OF IMPOSSIBLE OR INADEQUATE THROMBOLYSIS

In certain cases, thrombolysis may not be possible, it may be contra-indicated, or it may be completely or partially ineffective. In these situations, the therapeutic alternatives are directed toward correcting symptomatic residual obstruction. Opinions on the suitability of using one or another of these

techniques depends on the assessment of several parameters including length of the lesion [20], severity of the symptoms and the age and activity requirements of the patient at work or in sports [14,17]. In general terms, it is accepted that surgical treatment has a favorable outcome in carefully selected patients with relatively short obstructive lesions and certain disability in their daily activities.

ACUTE SUBCLAVIAN-AXILLARY
Jugular vein turndown seems to be favored over claviculectomy [2], bypass or venous interposition.

VEIN

THROMBOSIS

Possible scenarios in the near future

In a comprehensive review of the literature carried out by Rutherford in 1998 and published in Seminars in Vascular Surgery, the term possible scenarios was used to refer to the sequence in which existing therapeutic options are applied after the appearance of a new treatment such as, for example, thrombolysis [10]. The conclusion formulated after

an overview of the most recent publications on SAVT is that the configuration of these therapeutic scenarios is beginning to coincide, but always within the context of the experience of individual groups. The series may be more or less extensive, but the situation is far from a collective effort in which consensus is achieved after obtaining data from large prospective, multicenter studies. Such a collective effort is needed to resolve doubts around the use of low molecular weight heparin in secondary SAVT, for example, and to establish specific criteria on the use of thrombolytic, surgical and endovascular treatment at the best time and in the best way possible for patients with either form of this condition.

R E F E R E N C E S 1 Palerme LP, Chan AM, Hsiang YN. Axillary vein thrombosis secondary to congenital stricture in a left-sided superior vena cava. Ann Vase Surg 2000; 14: 648-651. 2 Hurlbert SN, Rutherford RB. Primary subclavian-axillary vein thrombosis. Ann Vase Surg 1995; 9: 217-223. 3 Martin C, Viviand X, Saux P, Gouin F. Upper-extremity deep vein thrombosis after central venous catheterization via the axilary vein. Crit Care Med 1999; 27: 2626-2629. 4 Monreal M, Raventos A, Lerma R et al. Pulmonary embolism in patients with upper extremity DVT associated to venous central lines. A prospective study. Thromb Haemost 1994; 72: 548-550. 5 Hicken GJ, Ameli FM. Management of subclavian-axillary vein thrombosis: a review. Can J Surg 1998; 41: 13-25. 6 Branchereau P, Alric P, Berthet JP et al. Surgical exposure of superior sulcus lung tumors with vascular involvement. Ann Vase Surg mi; 15: 206-211. 7 Tilney ML, Griffiths HJ, Edwards EA. Natural history of major venous thrombosis of the upper extremity. Arch Surg 1970; 101: 792-796. 8 Heron E, Lozinguez 0, Emmerich J et al. Long-term sequelae of spontaneous axillary-subclavian venous thrombosis. Ann Intern Med 1999; 131: 510-513. 9 Donayre CE, White GH, Mehringer SM, Wilson SE. Pathogenesis determines late morbidity of axillosubdavian vein thrombosis. AmJSurgim; 152: 179-184. 10 Rutherford RB. Primary subclavian-axillary vein thrombosis: the relative roles of thrombolysis, percutaneous angioplasty, stents, and surgery. Semin Vase Surg 1998; 11: 91-95. 11 Machleder HI. Evaluation of a new treatment strategy for PagetSchroetter syndrome: spontaneous thrombosis of the axillarysubclavian vein. J Vase Surg 1993; 17: 305-317. 12 Azakie A, McElhinney DB, Thompson RW et al. Surgical management of subclavian-vein effort thrombosis as a result of thoracic outlet compression./ Vase Surg 1998; 28: 777-786. 13 Lokanathan R, Salvian AJ, Chen JC et al. Outcome after thrombolysis and selective thoracic outlet decompression for primary axillary vein thrombosis. / Vase Surg 2001; 33: 783-788. 14 Feugier P, Aleksic I, Salari R et al. Long-term results of venous revascularization for Paget-Schroetter syndrome in athletes. Ann Vase Sing 2001; 15: 212-218. 15 Lee WA, Hill BB, Harris EJ Jr et al. Surgical intervention is not required for all patients with subdavian vein thrombosis. J Vase Swrg2000; 32: 57-67. 16 Angle N, Gelabert HA, Farooq MM et al. Safety and efficacy of early surgical decompression of the thoracic outlet for PagetSchroetter syndrome. Ann Vase Swrg2001; 15 : 37-42. 17 Arko FR, Harris EJ, Zarins CK, Olcott C 4th. Vascular complications in high-performance athletes. / Vase Surg 2001; 33: 935-942. 18 Kreienberg PB, Chang BB, Darling RC 3rd et al. Long-term results in patients treated with thrombolysis, thoracic inlet decompression, and subclavian vein stenting for Paget-Schroetter syndrome. / Vase Surg 2001; 33 (2 Suppl): S100-105. 19 Kunkel JM, Machleder HI. Treatment of Paget-Schroetter syndrome. A staged, multidisciplinary approach. Arch Surg 1989; 124: 1153-1158. 20 Molina JE. Need for emergency treatment in subclavian vein effort thrombosis. J Am Coll Surg 1995; 181: 414-420.

19
AORTOCAVAL FISTULA
DOMINIQUE MAIZA, JEAN ADER JULES
Spontaneous rupture of an abdominal aortic aneurysm (AAA) into the inferior vena cava (IVC) system is the most common cause of an aortocaval fistula (ACF). It is a seldom but formidable complication because the symptoms are protean and can mislead the diagnosis. If not treated, it can lead to right heart failure resistant to medical treatment and finally ending in the patient's death. Discovery during the surgical cure of such an aneurysm is associated with significant blood loss and a high peri-operative mortality. The pre-operative diagnosis can improve the prognosis by allowing the surgical team to anticipate and adapt the treatment. In the aim to clarify the symptomatology and the results of surgical treatment, the Association Universitaire de Recherche en Chirurgie (AURCFrance) performed a survey which led to a retrospective study concerning 39 cases of active ACFs operated between 1973 and 1989.

19. 181

Retrospective study of the AURC

Retrospective review identified a total of 39 patients who underwent operation in 11 of the 28 AURC centers for a documented arteriovenous fistula (AW) between a ruptured abdominal aortic or iliac aneurysm into the IVC or iliac vein (Table I). Patients with another cause of ACF were not included in the study group. The patients (36 men, 3 women) were aged between 42 years and 87 years (mean 69 years). The records of these patients were

reviewed in detail to specify the etiology and the location of the fistula, the clinical symptoms at the time of the initial presentation, the management and the results of the surgical treatment. All 39 aneurysms were of atherosclerotic origin. In 37 of these, aneurysm involved the infrarenal aorta, extending into the iliac vessels in 30 cases, whereas two patients had isolated common iliac artery aneurysms. The average length of the fistula was 30 21 mm (3 to 80 mm). In 34 patients, the fistula was eroded into the vena cava, in 4 patients into the right common iliac vein and in 1 patient into the left medial iliac vein. In 4 patients, a concomitant retroperitoneal aneurysm rupture was present.

VASCULAR

EMERGENCIES

n , Center

c burgeon 6

Number t of , .. . patients
4 2 7 4 1 1 10 2 1 5 2

Clinical manifestations
Abdominal

Angers CHU Bordeaux CHU Caen CHU Clermont-Ferrand CHU Creteil CHU Grenoble CHU Lille CHU Nancy CHU Paris, Hopital Ambroise Pare Rouen CHU Toulouse CHU Purpan

JM Chevallier JM Serise D Maiza JP Ribal JP Becquemin H Guidicelli JL Magne C Stankowiak (Y) G Fieve F Bacourt J Testart J Watelet A Barret

Pain Bruit Mass Thrill Cardiac Tachycardia Dyspnea Global heart failure Right heart failure Shock Angina Renal Oliguria Anuria Venous engorgement Lower limb edema Lower limb cyanosis Abdominal cyanosis Lower limb pulsatile varicose veins Lower limb subacute ischemia Priapism

31 27 21 13

79 69 54 33

24 15 10 3 9 1

61.5 38.5 25.6 7.7 23.1 2.6

12 12

30.8 30.8

19 12 3 3 2 2 1 1

48.7 30.8 7.7 7.7 5.1 5.1 2.6 2.6

19
182

The clinical manifestations are summarized in Table II. The continuous abdominal bruit with systolic intensification observed 27 times, was looked for only in 29 patients. The pathognomonic triad associating abdominal pain, abdominal bruit and abdominal mass was present in 13 cases. The diagnosis of an ACF was recognized before surgery in 29 patients (74.4%) and in 23 patients based only on the physical findings. The observation of an abdominal bruit or abdominal bruit associated with abdominal pain and an abdominal mass allowed the diagnosis in most of the cases (Table III) . The mean interval between the fistula formation, determined by the appearance of symptoms and the diagnosis, was 7 12 days (some hours to 60 days). The / \ j / angiographic or echographic examinations were realized only in a fickle way. A pre-operative aortography was performed in 16 patients. The 13 aortographies performed by arterial way always allowed to assert or to confirm the diagnosis, and in
7

Rectal bleeding Abdominal pulsatile varicose veins

k T
1

Neurologic Paraplegia Hemiplegia 1 1 2.6 2.6

rr , t /7777/Y7/ t . ,. manifestations

Number
27

Aortocaval J fistula recognized before surgery 1\ /o/\ ( /o)

22 (81)

Abdominal bruit Abdominal bruit + abdominal pain + abdominal mass

13

11

(85)

AORTOCAVAL

FlSTULA

Number
Aortic control Upstream to the fistula Suprarenal aortic clamping Infrarenal aortic clamping Downstream to the fistula Aortic or iliac clamping Endovascular occlusion Not clarified Venous control
30 5 34

12 patients to clarify the location of the fistula. On the other hand, three aortographies performed by venous way allowed only one time to assert the diagnosis. A computed tomography (CT) scanning with contrast performed in five patients allowed to assert or to confirm the diagnosis in three patients, twice to clarify the origin of the fistula and twice to observe an associated retroperitoneal hematoma. An abdominal echography, performed in 11 patients, always confirmed the diagnosis of an AAA but suspected the presence of the fistula only in one patient because of the existence of a lower extension of the vena cava.

7 2

Treatment
The surgical approach to the vessels has been by way of laparotomy in 38 patients (32 xyphoid-topubis midline incisions and 6 supra-umbilical transverse incisions). Because of the history of abdominal surgery, a retroperitoneal abdominal route was chosen in one patient. In four patients, a circulatory arrest occurred at the time of aortic clamping. Three patients could be resuscitated by closed-chest heart massage. The modalities of arterial and venous control are summarized in Table IV. The chronology of the venous control with regard to the aortotomy is clarified only in 35 patients: in 30 patients the venous control was realized only after the aortotomy. The techniques of restoring arterial and venous flow are summarized in Table V. They concern only 37 patients because two patients died before the repair: one of a cardiac arrest at the time of the aortic clamping; the other one, to whom the diagnosis of ACF was not established before surgery, died of an uncontrolled venous bleeding during the aortotomy. Mean blood loss was 4 600 3 500 mL (1 000 to 15 000 mL). An auto transfusion system was used in only two patients. Operative mortality, defined as death occurring during the hospitalization, was 28.2% (11 patients). There were three intraoperative deaths (7.7%): one patient died of cardiac arrest at the time of aortic clamping, the two others of hemorrhage. There were eight postoperative deaths (20.5%) occurring between 0 and 39 days (mean 1014 days). The causes and incidence of peri-operative deaths are summarized in Table VI. Pulmonary and cardiac complications are responsible for 54.5% of deaths. Twelve patients (30.7%) had one or several nonlethal complications as summarized in Table VII.

Upstream to the fistula

Clamping

Endovascular occlusion Downstream to the fistula

Clamping Endovascular occlusion Digital compression through opened aneurysm Not clarified

6 13
11

11 10

19. 183

Techniques
Arterial repair Aorto-aortic bypass Aorto-bi-iliac bypass Aortobifemoral bypass Aorto-iliac bypass Ilio-iliac bypass Venous repair Lateral suture Patch Ligation
7 18 8 2

19 49 22 5 5 76 11 13

VASCULAR

EMERGENCIES

Number ofdeaths
Pulmonary complications Cardiac failure Coronary disease Hemorrhage Renal failure Biliary peritonitis Staphylococcus aureus septicemia

Patients %

Deaths %

3 2 1

7.7 5.1 2.6 5.1 2.6 2.6 2.6

27.2 18.2
9.1

2 1 1 1

18.2
9.1 9.1 9.1

chronic renal failures) within 6 5 days (1 to 15 days). Thirteen patients presented a pre-operative congestive cardiac failure. Four of them died during the hospitalization. Among the survivors, the time of recovery could only be assessed in seven patients: in six cases, the cardiac function normalized in 3 3 days; one patient kept signs of left cardiac failure during two months. Two patients presented with pre-operative neurologic signs. A right hemiplegia due to the shock in a patient who had a left carotid stenosis recovered quickly without aftereffects in the postoperative period. One pre-operative paraplegia has only partially declined and improved to a partial paraparesis allowing walking. In seven patients, the vena cava patency was assessed by a cavography realized at distance of the repair: twice the vena cava was thrombosed and in five patients it was patent.

General considerations
Spontaneous rupture into the adjacent venous system is an unusual complication of ruptured AAAs. According to the literature, the incidence of this event is quite low, occurring between 1% and 7% of all ruptured aneurysms [1-8] (Table VIII). Only two patients (1.4%) suffered from such a rupture in the AAA prospective study realized by the AURC from January until December 1989 [4]. Ruptured arteriosclerotic aneurysms are the most common cause of ACF [6,9-12]. This etiology represents 92.8% of the ACF observed during the same period by the AURC members having participated in our study.

Complications

Number
6

19
184

Deep venous thrombosis Digestive hemorrhage Wound infection Chronic renal failure Colorectal ischemia Paraplegia Serratia septicemia

2 2 1 1 1 1

UNUSUAL CASES OF ACF

Very rare cases of ACF in association with inflammatory ruptured AAA have been reported [13,14]. ACF resulting from rupture of syphilitic or mycotic aneurysms, as well as aneurysmal lesions seen in Marfan's syndrome, Ehlers-Danlos syndrome [15] or Takayasu's arteritis [16] are exceptional. Very rarely, neoplasms may also cause a major AVF by erosion of adjacent arterial and venous structures [17]. Traumatic ACF due to a simultaneous penetrating injury of the aorta or iliac artery and the IVC or one of the iliac veins represent 10% to 20% of current cases. The most frequently reported traumatic lesions are those caused by gunshot abdominal wounds and iatrogenic complications of lumbar spine surgery [11,18-20]. Davidovic et al. reported

Pulmonary embolism due to deep venous thrombosis was not encountered. No gastrointestinal complications required surgery. The postoperative paraplegia observed in one patient recovered spontaneously within a few days. Nine of 21 patients who had a pre-operative creatinine greater than 150 pmol/L died during the hospitalization. One patient suffered from chronic renal failure. Eleven patients regained a normal renal function (n = 9) or identical to the previous state (two pre-existent

AORTOCAVAL

FISTULA

1st author [ref.]

Year of publication

Type of study

Period

Number of operated AAA

2152 1034

Number of operated ruptured AAA

373 146

Aortovenous fistulas (%)

26 2

Ghilardi [5] Koskas [4]

1994 1997

Retrospective Single-center retrospective Multicenter retrospective Single-center retrospective Single-center prospective

1965 -1992 1989

(6.9) (1.4)

Bednarkiewicz [6

1997

1980 -1994

580

111

(6.3)

Tsolakis [7]

1999

1990 -1997

112

52

(7.7)

Alonso Perez [8]

2000

1996 -1997

NP

144

(2)

AAA: abdominal aortic aneurysm

one case of ACF caused by blunt abdominal trauma after a traffic accident and one fistula developed after cardiac catheterization [21]. Although ACFs resulting from penetrating wounds frequently have a dramatic presentation, iatrogenic fistula produces less acute symptoms than traumatic and spontaneous aortic or iliac fistula. Pagni et al. reported in 1996 a case of ACF secondary to a retro-aortocaval false aneurysm which had ruptured into the IVC, in an 81-year-old woman, 7 years after primary infrarenal aortic aneurysm repair [22]. In 2002, Tuma et al. reported a case of contained rupture of an AAA and tear of the IVC 15 months after placement of an aortic endograft [23]. The aortic rupture was probably caused by poor proximal fixation of the graft, and an angulated right iliac limb of the stent graft penetrated into the IVC just above the common iliac junction and caused sealed perforation. Erosion of the aneurysm into the venous system occurs most often in the infrarenal IVC. More rarely, it occurs in an iliac vein [3,11] or in left renal vein [3,24,25]. In a review, Mansour et al. reported 16 cases of spontaneous aorto-left renal vein fistula, and 94% had a retro-aortic left renal vein [24].

CLINICAL ASPECTS
The fistulization of the aneurysm may be asymptomatic and discovered at operation. In general, total or partial obstruction of the fistula by the intraaneurysmal mural thrombus exists [3,11]. In the majority of the cases, the clinical symptomatology is highly variable associating, in different degrees, local symptoms due to the aneurysmal rupture and regional and general signs due to the high-flow AW. These protean clinical manifestations and the rarity of such lesions may lead to delay in the diagnosis (Table IX). When the diagnosis of aneurysm is overlooked, the untoward hemodynamic consequences of the AW may lead to misdiagnosis such as phlegmasia cerulea, sciatic pain with motor deficits and congestive heart failure [10]. Conversely, the palpation of an aneurysm associated with abdominal pain and shock can lead to the diagnosis of retroperitoneal rupture.

185

PATHOPHYSIOLOGY
Hemodynamic consequences of the fistula are responsible for the protean clinical picture which includes cardiac, renal, venous or even neurologic

VASCULAR

EMERGENCIES

19

AORTOCAVAL

FISTULA

findings. Cardiac manifestations are secondary to the shunting of blood through the fistula into the low-resistance and low-pressure venous system, increasing venous return, and consequently cardiac output and blood volume [28]. Systolic blood pressure decreases transiently and then returns to the normal level while diastolic pressure remains low. The flow rate through the fistula and the myocardial reserve determine whether the arteriovenous communication is tolerated. Usually, tolerance is initially good. Gradually, however, right and then overall heart failure develops and does not respond to medical treatment. Of traumatic fistulas only 30% of the patients present with signs of cardiac failure because these generally have a small size and arise in young people having a good cardiac reserve [9-11]. Several pathologic mechanisms explain renal failure [29-31]. Hypertension of the renal veins is a hemodynamic obstacle, initiating a drop in renal blood flow and thereby in glomerular filtration [31]. The decreased mean arterial pressure and the heart failure add to the renal low-flow state. Reactional hypersecretion of aldosterone by the juxtaglomerular apparatus induces water and salt retention, which then leads to water overload. The clinical manifestations of renal failure are oliguria or anuria and azotemia. Regional venous hypertension is a constant finding and involves mainly the lower limbs. Symptoms are generally bilateral but they can be unilateral when the fistula is located in one iliac vein [10]. Heart failure and venous compression secondary to the aneurysmal mass increase venous hypertension. Clinically edema and cyanosis are frequent. Superficial veins may be distended or pulsatile at palpation. The arterial inflow to the leg reduced by the fistula explains the frequent low-grade ischemia observed distally. These findings, especially in the one-sided forms, may lead to the incorrect diagnosis of phlegmasia cerulea [10,11]. Venous hypertension in the pelvic veins may lead to hematuria [32-34], rectal hemorrhage [6,10, 11] or priapism [12,33], which may cause diagnostic confusion. Neurologic manifestations have occasionally been reported in the literature. Brain disorders, ranging from obtundation to syncope, result from the low cerebral blood flow [9,10]. Full-scale paraplegia is rare and can result from functional arterial insufficiency due to the increased spinal venous stasis induced by venous hypertension in the epidural space associated with arterial lesions in atherosclerotic patients [10,35].

DIAGNOSIS
The diagnosis of an aortocaval or iliac AVF is dependent on recognition of its clinical features. Whatever the mode of clinical expression, a simple complete clinical examination should lead to the diagnosis: a continuous bruit with systolic accentuation is considered pathognomonic for AW. In the presence of a painful abdominal mass or signs of pelvic or lower limb venous hypertension or progressive cardiac failure, the diagnosis becomes more obvious. Common features in patients with aorto to left renal vein fistula include abdominal pain, hematuria, impaired renal function and nonvisualization of the left kidney. Less common but also present in the majority of cases are a left-sided bruit, pulsatile abdominal mass and proteinuria. Overt signs of high-output cardiac failure are less likely because of the smaller size of the renal vein compared with the IVC and iliac vein [3,24]. If the clinical presentation is not acute, there is time to perform pre-operative imaging studies. Various diagnostic techniques like B-mode ultrasonography, doppler ultrasonography, aortography, CT scan and magnetic resonance (MR) angiography will confirm the diagnosis in most patients, unless the fistula is occluded by a thrombus. Ultrasonography associated with doppler analysis can easily confirm the diagnosis [36]. Because of its reliability and noninvasiveness, this method can be superior to aortography. Although angiography remains the best method to document the presence of ACF for many authors [3,11], the diagnosis may not be apparent if thrombus material occludes the fistula. Aortography is not really mandatory, mostly confirms only what is clinically evident and should not delay surgical treatment in emergent situations. In contrast, arteriography can be very valuable if the clinical findings are unclear, and delineation of anatomy may be particularly important in the management of traumatic fistulas. The role of CT scan during the AURC study is probably underestimated due to the time period of the survey (1973-1989). CT scan has been demonstrated to be similarly useful in establishing the diagnosis of ACF with its characteristic findings of caval effacement, loss of fat plain between the aorta and IVC, and rapid flow of contrast from the aorta into a dilated IVC [37,38]. In case of retro-aortic left renal vein fistula, CT scan can demonstrate not only the anomalous location of the left renal vein but also the AAA and poor enhancement of the left kidney. Early and intense enhancement of the FVC is

187

VASCULAR

EMERGENCIES

underlined with gadolinium-enhanced MR angiography. The exact location of the fistula and a complete pre-operative assessment of AAA can be performed with this noninvasive imaging technique [39,40]. Moreover, the absence of iodinated contrast media makes it particularly suitable for patients with renal insufficiency, but the long time for imaging acquisition requires the patient to be stable. SURGICAL STRATEGY If the diagnosis is not established before surgery, the surgeon may still recognize the entity at operation before opening the aneurysm, by palpating a thrill over the major vascular structures and observing engorgement of the vena cava and adjacent retroperitoneal or pelvic veins. Classic surgical repair consisting of suture of the fistula from within the sac in patients with aneurysm or lateral venography in traumatic injuries, must be performed as soon as possible. No attempts to improve pre-operative cardiac or renal failure are worthwhile because medical therapy is unavailing. All nonoperated patients will die in some days or some weeks after the appearance of symptoms [12]. Treatment is currently well established. Intraoperative hemodynamic monitoring requires placement of at least one arterial and one right atrial pressure line. If there is a history of heart failure or coronary artery disease, intraoperative monitoring of left ventricular filling pressure using a SwannGanz catheter is recommended [10]. Continuous hemodynamic monitoring is necessary to prevent fluid overload at the beginning of surgery, which may worsen congestive cardiac failure. At crossclamping, an increase in peripheral resistance and reduction in venous return may cause ventricular fibrillation and cardiac arrest. Furthermore, hypovolemic shock caused by bleeding during fistula repair or by sudden peripheral resistance decrease after aortic clamp removal, should be avoided [41]. As intraoperative blood loss is often significant, autotransfusion equipment is helpful to decrease the heterologous blood transfusion [11,42]. It was only used twice during the AURC study, which is explained by the time period of the survey. Minimal aneurysm manipulation is important to prevent paradoxical embolization of intraluminal debris [43]. The aortic control is similar to standard aortic aneurysm surgery without AVF, even if supraceliac control is necessary in the presence of large aneurysms. The proximal aortic clamping must be performed before the distal arterial clamping to avoid

19
188

a sudden increase of the arteriovenous shunt [11]. The IVC control is determined by specific anatomical conditions [10]. When the reno-caval junction can be approached easily, the vena cava may be clamped distal to the renal vein before opening the aneurysm. When venous clamping or compression is performed in this manner, blood loss is limited and the risk of air or atheromatous embolism is reduced. The strategy of compression to control massive venous hemorrhage from the fistula should be chosen before the aneurysm is opened. In case of an aorto-renal vein fistula, proximal and distal control of a retro-aortic left renal vein is difficult and dangerous and should not be attempted. The dissection of the infrarenal aorta should be performed under direct vision to avoid injury to a retro-aortic left renal vein. In this situation, direct compression of the left renal vein from inside the aneurysm lumen should be carried out. After aortotomy, the fistula defect should be oversewn inside the aneurysm cavity after all thrombus and debris are removed. Venous bleeding from the fistula can be controlled by direct finger compression or sponge stick if the fistula is small or, in others cases, by balloon occlusion catheters inflated in the vena cava on both sides of the fistula [1,11,27]. Transvenous positioning of balloon catheters in the vena cava before aortic opening was found to be helpful in reducing hemorrhage by Ingoldby et al. [44] and Naito et al. [45]. Closure of the venous tear should be rapid and done from inside the aneurysm by a running suture, taking wide bites on the edges of the fistula. A patch is rarely necessary and it increases operative time and blood loss but can be useful in some cases of traumatic ACF. In case of inadequate venous bleeding control, large fistula or iliac fistula, it is possible to ligate the vein at both sides of the fistula [10,11]. Although IVC ligation is generally well tolerated, many complications (recurrent deep vein thrombosis, leg edema, varicose veins) may occur [7]. In case of friable vena cava wall, the replacement of the destroyed part of the IVC with graft interposition has been proposed but with an increased postoperative thrombosis and pulmonary embolization risk [5, 46]. Woolley et al. described an aortic exclusion with proximal and distal ligation of the aorta and bypass grafting in one patient [47]. In this situation, the fistula remains undisturbed but only communicates with the excluded aneurysm sac. This operation could cause deep venous thrombosis and continued aneurysm growth. Routine use of IVC

AORTOCAVAL

FISTULA

interruption by clips should not be performed, since embolization can occur after technical failure [11,27]. Aortic repair is performed according to standard prosthetic graft implantation.

SURGICAL RESULTS
During the postoperative period, cardiovascular and renal symptoms induced by the high-flow AVF generally decrease rapidly in surviving patients. Postoperative complications are due to the general consequences of the blood transfusion, pre-existing comorbidities or general infectious problems. They are explained by the necessity of an urgent surgical treatment, without wasting the time necessary for an adequate pre-operative evaluation in older arteriosclerotic patients with coronary artery disease and severe physiologic disturbances. Reported operative mortality after surgical repair of ACF caused by aneurysmal rupture is high, ranging from 6% to 36% (Table X). In the studies of Brewster et al. and Davidovic et al., all deaths occurred in patients in whom the correct diagnosis was not

established before surgery [11,21]. In a review of 184 aortovenous fistulas (159 aorto-caval, 8 aortoiliac, 17 aorto-renal), Calligaro et al. reported a survival of 72% of patients with ACFs and 88% of patients with aortorenal vein fistulas after surgery [3]. The lower mortality rate found with aortorenal vein fistulas was explained by an associated lower incidence of ruptured aneurysms, because only 12% of patients with an aortorenal vein fistula had a separate retroperitoneal rupture compared with 22% of patients with an ACE In the experience of Brewster et al., iliac AW had a better prognosis than ACF because most of these patients were less hemodynamically compromised [11]. Results are much better in cases of traumatic or iatrogenic ACFs, because of younger age, absence of heart disease and better hemodynamic status of patients [11,20,21].

ENDOVASCULAR TREATMENT
Endovascular stent graft repair may offer an attractive and minimally invasive treatment alternative

1st author [ref.] Harrington [26] Calligaro [3]

Year of publication
1989 1990

Time period

Number of Number of AW in ruptured AAA traumatic AW

Overall postoperative mortality %

Postoperative mortality traumatic AW

189

1975 - 1989

10
159 ACF 8AIF 17ARF

0 0 0 0 0 6 0 2 2 0

20 28 29 12 36 10
36.4

Review

Salo [32] Brewster [11] Ghilardi [5] Davis [27] Davidovic [21] AURC

1990 1991 1994 1998

1976 - 1988

11 14 22 16 14 39

30 years
1965 - 1992 1970 - 1997

0 0 0
.

2002 2002

NA
1973 - 1989

25
28.2

ACF: aortocaval fistula AIF: aorto-iliac fistula ARF: aorto-renal fistula AW: arteriovenous fistula NA: not available

VASCULAR

EMERGENCIES

19
190

to open repair management especially in cases in which a pre-operative diagnosis can be made. It offers the theoretical advantage of reduced intraoperative blood loss and can subsequently reduce the mortality and the morbidity rates. Boudghene et al. [48] reported successful treatment of ACF with percutaneous stent grafts in an experimental study in which the ACF was created percutaneously in eight sheep. The use of an endovascular stent graft in the repair of a spontaneous aortovenous fistula has only been reported four times for ruptured AAA (three ACF and one aorto-left renal vein fistula) , one time for an ACF that recurred after conventional AAA repair and one time for an ACF due to retroperitoneal sarcoma [17,49-53]. Such application was not without technical problems during introduction and deployment of the stent graft. In their case, Beveridge et al. [49] described the technical problems during introduction and deployment of the contralateral limb of the bifurcated stent graft because of the displacement of the common iliac artery by the aneurysm. Umscheid and Stelter [50] used a bifurcated stent graft in their case because the distal aortic neck was very short and the iliac vessels had a normal caliber. However, postoperative CT showed that one of the iliac limbs of the bifurcated graft was compressed by the narrow distal cuff of the aneurysm, potentially leading to limb occlusion or claudication in the future. Lau et al. [51] used an aorto-uni-iliac device with a femorofemoral cross-over graft from the right to left side, and the left common iliac artery was embolized with metallic coils after successful deployment of the stent grafts to excluded inflow into

the aneurysm sac. The intraoperative angiogram showed complete exclusion of the aneurysm with absence of endoleak and closure of the ACF, but a repeat spiral CT showed a persistent fistula with a small inflow from the left common iliac artery. The coils had clearly not occluded the left common iliac artery. Two weeks later, the patient was readmitted with acute ischemia of both legs due to occlusion of the aortic stent graft. Post-thrombolysis arteriography showed a patent graft with good run off and no evidence of any endoleaks or persistence of ACF, but with a persistent stenosis at the junction between the aorto-uni-iliac graft and the iliac extension. Sultan et al. [52] reported a case of a contained rupture of the aorta communicating with the left renal vein. An aorto-aortic stent graft was used and the postoperative CT showed no evidence of an endoleak. The left renal vein, however, continued to fill the saccular para-aortic structure. The endovascular repair was only partially successful, as the defect in the left renal vein had to be repaired at open laparotomy few days later. Duxbury et al. [17] treated a right common iliac AW resulting from a retroperitoneal sarcoma with a straight stent graft, but ten days after stenting the patient had clinical deterioration and angiography showed partial prolapse of the proximal end of the stent graft into the IVC, enlarging the fistula. A bifurcated stent graft was placed without complication. Unlike a rupture or contained rupture of an AAA, spontaneous ACF may not always require emergency treatment, and time may be available for obtaining the suitable endovascular stent graft.

R E F E R E N C E S 1 Baker WH, Sharzer LA, Ehrenhaft JL. Aortocaval fistula as a complication of abdominal aortic aneurysms. Surgery 1972; 72: 933-938. 2 Duppler DW, Herbert WE, Dillihunt RC, Ray FS. Primary arteriovenous fistulas of the abdomen. Their occurrence secondary to aneurysmal disease of the aorta and iliac arteries. Arch Suigim-, 120: 786-790. 3 Calligaro KD, Savarese RP, DeLaurentis DA. Unusual aspects of aortovenous fistulas associated with ruptured abdominal aortic aneurysms. / Vase Surg 1990; 12: 586-590. 4 Koskas F, Kieffer E. Surgery for ruptured abdominal aortic aneurysm: early and late results of a prospective study by the AURC in 1989. Ann Vase Surg 1997; 11: 90-99. 5 Ghilardi G, Scorza R, Bortolani E et al. Primary aortocaval fistula. Cardiovasc Surg 1994; 2: 495-497. 6 Bednarkiewicz M, Pretre R, Kalangos A et al. Aortocaval fistula associated with abdominal aortic aneurysm: a diagnostic challenge. Ann Vase Surg 1997; 11: 464-466. 7 Tsolakis JA, Papadoulas S, Kakkos SK et al. Aortocaval fistula in ruptured aneurysms. Eur J Vase Endovasc Surg 1999; 17: 390-393. 8 Alonso-Perez M, Segura RJ. Surgical risks of emergency AAA repair. In: Branchereau A, Jacobs M (eds). Surgical and endovascular treatment of aortic aneurysms. Armonk, Futura Publishing Co, 2000 : p p 271-280. 9 Reckless JP, McColl I, Taylor GW. Aortocaval fistulae: an uncommon complication of abdominal aortic aneurysms. Br J Surg 1972;59:461-462. 10 Petetin L, Pelouze GA, Mercier V et al. Rupture of abdominal aortic aneurysm into the inferior vena cava: a study of seven cases. Ann Vase Surg 1987; 1: 572-577.

AORTOCAVAL

FISTULA

11 Brewster DC, Cambria RP, Moncure AC et al. Aortocaval and iliac arteriovenous fistulas: recognition and treatment. J Vase Swig 1991; 13: 253-265. 12 Gilling-Smith GL, Mansfield AO. Spontaneous abdominal arteriovenous fistulae: report of eight cases and review of the literature. BrJSurgim; 78: 421-425. 13 Ferrari M, Bonanomi G, Fossati N et al. Surgical management of inflammatory abdominal aortic aneurysm associated with occult aortocaval fistula. Surgery 2000; 127: 234-236. 14 Farid A, Sullivan TM. Aortocaval fistula in ruptured inflammatory abdominal aortic aneurysm. A report of two cases and literature review. / Cardiovasc Surg 1996; 37: 561-565. 15 Taniyasu N, Tokunaga H. Multiple aortocaval fistulas associated with a ruptured abdominal aneurysm in a patient with EhlersDanlos syndrome. Jpn CircJ 1999; 63: 564-566. 16 Gronemeyer PS, deMello DE. Takayasu's disease with aneurysm of right common iliac artery and iliocaval fistula in a young infant: case report and review of the literature. Pediatrics 1982; 69: 626-631. 17 Duxbury MS, Wells IP, Roobottom C et al. Endovascular repair of spontaneous non-aneurysmal aortocaval fistula. Eur J Vase Endovasc Swrg2002; 24: 276-278. 18 Krishnasastry KV, Friedman SG, Deckoff SL, Doscher W. Traumatic juxtarenal aortocaval fistula and pseudoaneurysm. Ann Vase Surg 1990; 4: 378-380. 19 Pincu M. Traumatic aortocaval fistulas of late diagnosis. / Vase Swgl994; 19: 1097-1098. 20 Papadoulas S, Konstantinou D, Kourea HP et al. Vascular injury complicating lumbar disc surgery. A systematic review7. Eur J Vase Endovasc 5wrg2002; 24: 189 -'195. 21 Davidovic LB, Kostic DM, Cvetkovic SD et al. Aorto-caval fistulas. Cardiovasc Surg 2002; 10: 555-560. 22 Pagni S, Halene S, Kwass W, Khachane V. Ruptured aortic pseudoaneurysm: a rare presentation as aortocaval fistula. ) Cardiovasc Surg 1997; 38: 165-168. 23 Tuma MA, Hans SS. Rupture of abdominal aortic aneurysm with tear of inferior vena cava in a patient with prior endograft. / Vase Surg 2002; 35: 798-800. 24 Mansour MA, Rutherford RB, Metcalf RK, Pearce WH. Spontaneous aorto-left renal vein fistula: the "abdominal pain, hematuria, silent left kidney" syndrome. Surgery 1991; 109: 101-106. 25 Masood T, Naylor AR, Edge JM et al. Double aortovenous fistula: a unique presentation of a ruptured abdominal aortic aneurysm. Eur] Vase Surg 1994; 8: 107-109. 26 Harrington EB, Schwartz M, Haimov M et al. Aorto-caval fistula: a clinical spectrum. / Cardiovasc Surg 1989; 30: 579-583. 27 Davis PM, Gloviczki P, Cherry KJ Jr et al. Aorto-caval and ilioiliac arteriovenous fistulae. Am J Surg 1998; 176: 115-118. 28 Houben PF, Bollen EC, Nuyens CM. "Asymptomatic" ruptured aneurysm: a report of two cases of aortocaval fistula presenting with cardiac failure. Eur] Vase Surg 1993; 7: 352-354. 29 Albalate M, Gomez Octavio J, Llobregat R, Fuster JM. Acute renal failure due to aortocaval fistula. Nephrol Dial Transplan. 1998; 13: 1268-1270. 30 Gregoric ID, Jacobs MJ, Reul GJ, Rochelle DG. Spontaneous common iliac arteriovenous fistula manifested by acute renal failure: a case report. J Vase Surg 1991; 14: 92-97. 31 Brunkwall J, Lanne T, Bergentz SE. Acute renal impairment due to a primary aortocaval fistula is normalised after a successful operation. Eur] Vase Endovasc Surg 1999; 17: 191-196. 32 Salo JA, Verkkala KA, Ala-Kulju KV et al. Hematuria is an indication of rupture of an abdominal aortic aneurysm into the vena cava. / Vase Surg 1990; 12 : 41-44.

33 Saxon SR, Glover WM, Youkey JR. Aortocaval fistula and contained rupture of an abdominal aortic aneurysm presenting with pelvic venous congestion. Ann Vase Surg 1990; 4: 381-383. 34 Steinke TM, Reber PU, Hakki H, Kniemeyer HW. Haematuria and an abdominal aortic aneurysm- -warning of an aortocaval fistula. Eur] Vase Endovasc Surg 1999; 18: 530-531. 35 Jauslin PA, Muller AF, Myers P, Velebit V. Cauda equina syndrome associated with an aortocaval fistula. Eur] Vase Surg 1991; 5: 471-473. 36 Daxini BV, Desai AG, Sharma S. Echodoppler diagnosis of aortocaval fistula following blunt trauma to abdomen. Am Heart J 1989; 118: 843-844. 37 Quiroga S, Alvarez-Castells A, Hidalgo A et al. Spontaneous aortocaval fistula: CT findings with pathologic correlation. Abdom Imaging 1995; 20: 466-469. 38 Rosenthal D, Atkins CP, Jerrius HS et al. Diagnosis of aortocaval fistula by computed tomography. Ann Vase Surg 1998; 12: 86-87. 39 Walter F, Blum A, Quirin-Cosmidis I et al. An aortocaval fistula diagnosed with 1.5-T magnetic resonance angiography. / Cardiovasc Magn Reson 2000; 2: 213-216. 40 Gaa J, Bohm C, Richter A et al. Aortocaval fistula complicating abdominal aortic aneurysm: diagnosis with gadoliniumenhanced three-dimensional MR angiography. Eur Radiol 1999; 9: 1438-1440. 41 Miani S, Giorgetti PL, Arpesani A et al. Spontaneous aortocaval fistulas from ruptured abdominal aortic aneurysms. Eur] Vase Surg 1994; 8 : 36-40. 42 Doty DB, W'right CB, Lamberth WC et al. Aortocaval fistula associated with aneurysm of the abdominal aorta: current management using autotransfusion techniques. Surgerj 1978; 84: 250-252. 43 Cooperman M, Deal KF, Wooley CF, Evans WE. Spontaneous aortocaval fistula with paradoxical pulmonary embolization. Am JSurg 1977; 134: 647-649. 44 Ingoldby CJ, Case WG, Primrose JN. Aortocaval fistulas and the use of transvenous balloon tamponade. Ann R Coll Surg Engl 1990; 72: 335-339. 45 Naito K, Sakai M, Natsuaki M, Itoh T. A new approach for aortocaval fistula from ruptured abdominal aortic aneurysm. Balloon occlusion technique under echogram guidance. Thorac Cardiovasc Surg 1994; 42: 55-57. 46 Kiskinis DA, Saratzis N, Megalopoulos A et al. Primary aortocaval fistula in association with ruptured aneurysms. Ann Vase Surg 1994; 8: 496-499. 47 Woolley DS, Spence RK. Aortocaval fistula treated by aortic exclusion. / Vase Surg 1995; 22: 639-642. 48 Boudghene F, Sapoval M, Bonneau M, Bigot JM. Aortocaval fistulae: a percutaneous model and treatment with stent grafts in sheep. Circulation 1996; 94: 108-112, 49 Beveridge CJ, Pleass HC, Chamberlain J et al. Aorto-iliac aneurysm with arteriocaval fistula treated by a bifurcated endovascular stent-graft. Cardiovasc Intervent Radiol 1998; 21: 244-246. 50 Umscheid T, Stelter WJ. Endovascular treatment of an aortic aneurysm ruptured into the inferior vena cava. J Endovasc Ther 2000; 7: 31-35. 51 Lau LL, O'reilly MJ, Johnston LC, Lee B. Endovascular stentgraft repair of primary aortocaval fistula with an abdominal aorto-iliac aneurysm. / Vase Surg 2001; 33: 425-428. 52 Sultan S, Madhavan P, Colgan MP et al. Aorto-left renal vein fistula: is there a place for endovascular management? J Endovasc Srgl999;6:375-377. 53 Gandini R, Ippoliti A, Pampana E et al. Emergency endograft placement for recurrent aortocaval fistula after conventional AAA repair. J Endovasc Tfor 2002; 9: 208-211.

191

TRAUMATIC INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

LAURENT CHICHE, EDOUARD KIEFFER

No vascular lesion is more appalling to manage than that affecting the caval veins and their branches. Once limited to cases from military conflicts, these traumas have become, since nearly thirty years ago, the legacy of specialized trauma centers, particularly in North America from which most of the publications originate. The impact of the initial shock, the nearly always concomitant vascular or extravascular lesions, exposure difficulties and the control over and repair of damaged veins all explain the very high mortality associated with these lesions, despite the use of modern techniques. After a short historic overview, this chapter will address the trauma mechanisms, the local consequences and the general therapeutic approach to these lesions. We will focus on the description of damage to the inferior vena cava (IVC) because this occurs with the highest incidence, and we will not discuss trauma to the suprahepatic veins (SHVs) because this is usually associated with parenchymal damage to the liver, which is beyond the scope of this chapter.

20_
193

History
The modern era of vascular anastomoses started in 1912 with the work of Alexis Carrel [1]. In 1894, this French physician was frustrated by the technical impossibility to repair a lethal injury to the portal vein of President Sadi Carnot resulting from a dagger stab. Nevertheless, Eck had described earlier, in 1877, an anastomosis between the IVC and the portal vein in the dog, and in 1882 Schede closed an injury of the femoral vein laterally in a

human being. The first experimental end-to-end anastomosis is ascribed to Clermont in 1901. The first attempts of venous repair were reported during the Balkan war in 1911 and subsequently during the two World Wars. In 1916, Taylor [2] reported on the first patient to survive an IVC injury. Since the Korean conflict (1950 to 1953), the interest in venous repairs has been growing as illustrated by the series of 37 IVC injuries reported by Ochsner et al. in 1961 [3]. At that time, only 19 survivors of this type of injury had been published.

VASCULAR

EMERGENCIES

In 1981, the Vietnam Vascular Registry [4] reported 82 traumatic lesions of the caval vein, comprising 78 IVCs and four superior vena cavae (SVC). In 1978, the group in Houston had already reported a series of 301 cases, observed in 30 years of civil practice [5]. It is estimated that 10% to 15% of the penetrating traumas of the abdomen are accompanied by major venous injury and that one in 50 shot wounds damages the IVC [6]. In 50% of the cases, the patients die from hemorrhage before reaching the hospital. These data illustrate the place traumatic lesions of the IVC currently occupy outside the battlefield.

lateral position, is spared most of the time during frontal attacks on the left part of the body. All kinds of trauma, varying between a localized puncture and injuries with substantial tissue loss, can occur. Bullet injuries are the most destructive. Except for the trauma to the caval vein, they are often responsible for other serious lesions to neighboring structures that induce a risk of sepsis, which in part influences the prognosis.

IATROGENIC TRAUMA
This may concern vessel wall perforations due to the introduction of an endoluminal catheter or caval filter, of which the prognosis is usually good when untreated, or injuries occurring during surgery. Trauma occurring during thoracoscopy or laparoscopy may either pass unnoticed or be revealed secondarily, or may be diagnosed directly because of excessive blood loss during the procedure. As this type of trauma is becoming more and more frequent due to the increasing use of these techniques, it forms the topic of another chapter of this book. Finally, direct iatrogenic damage to the left innominate vein or the intrapericardial caval vein can complicate certain cardiac re-interventions.

Trauma mechanisms
In general, the superior thoracic, intrapericardial and retrohepatic segments of the caval vein may sustain either a penetrating or a closed trauma. The IVC is usually involved in single penetrating injuries.

CLOSED TRAUMA

20
194

Closed injuries usually result from a severe deceleration trauma in the horizontal direction such as in traffic accidents, or in the vertical direction, like in a fall from high altitude. The resulting shear forces cause a partial or total avulsion of the venous segment, as can be observed at the atriocaval or hepatocaval junctions, or at the level of the azygos, renal, or superior mesenteric veins. The common and internal iliac veins are damaged where they pass bony structures, especially after pelvic fractures. In the literature, 10 cases of post-traumatic thrombosis of the IVC were found in 1999 [7], occurring 3 days to 3 years after a high-energetic trauma. This thrombosis could have been secondary to organization of an initially localized thrombus, which allowed spontaneous hemostasis after an injury to the vessel wall. As an anecdote, a laceration of the IVC was reported following a nonpenetrating trauma caused by the high-pressure water jet of an industrial cleaner [8].

CONCOMITANT LESIONS
Nearly all patients who have suffered a trauma of the caval vein have at least one concomitant arterial or venous and/or neighboring visceral organ injury [9]. The most frequent arterial lesions concern the abdominal aorta. In several series [9-13], this association has significantly influenced the mortality risk. In rare cases, direct identification of an aortocaval fistula resulting from a penetrating trauma could protect the patient from a fatal hemorrhage. Some chronic aortocaval fistulae have been described [14], as have some cavorenal fistulae [15]. The venous lesions most frequently associated involve the portal, hepatic, splenic and superior mesenteric veins. Although all organs may be damaged, the most common ones are the liver, the kidneys, the duodenum, the small intestines and the colon. Recently, a traumatic cause of duodenum fistulae was found in 9 out of 37 (24%) cases in the literature [16]. The trauma was either related to a penetrating trauma of the abdomen, ingestion of a foreign body or early or late migration (7 days to 11 years) of a caval filter.

PENETRATING TRAUMA
The most frequent penetrating traumas are those from gunshots rather than stabbing weapons. The statistical preponderance of dexterity among the attackers explains why the caval vein, in its right

TRAUMATIC

INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

The fluid suppletion, initially consisting of macromolecules, also requires large quantities of blood products, red blood cell concentrates, plasma concentrates and platelets. Restoration of the blood volume is indispensable during this kind of surgery. In the majority of series, if specified, the mean number of blood product units supplied varied between 15 and 30, having a statistically significant impact on the mortality. It is important to prevent hypothermia using external warming devices and by warming the administered products.

Local consequences of caval vein trauma

Half of the patients who suffered from a caval vein trauma, and particularly those with an arterial lesion, present with a severe hemorrhagic shock. As opposed to the arteries, the veins show a poor vasoconstrictor response and cannot generate an effective hemostasis through their own physiologic properties. Because of the absence of valves, IVC injuries are not only characterized by bleeding from the iliac, but also by reflux from the atrial side. These veins, however, do have some features that can lead to a spontaneous hemostasis, at least temporarily. The circulation, including the caval veins, is maintained at low pressure. In case of an IVC injury, the neighboring tissues can cause an effective plugging that limits the hemorrhage and sometimes leads to thrombus formation and healing of the vessel wall. The hemorrhage may thus be contained by the retroperitoneum, the pancreas, the duodenum or the posterior side of the liver. This phenomenon is particularly seen in simple lesions caused by stabbing weapons or low-velocity shot wounds, but is exceptional in case of a perforating trauma of the IVC or damage secondary to highvelocity fire weapons.

General intervention principles

Generally speaking, no additional investigation should interfere with the initial reanimation of a patient with an injured abdomen or thorax. Not until the condition has been stabilized can a transthoracic cardiac or abdominal ultrasound exam be performed. When the clinical suspicion of a major vascular trauma has risen, a surgical exploration is justified. In several North American series, patients in whom reanimation remained insufficient underwent an urgent clamping of the thoracic descending aorta through a left anterolateral thoracotomy. Although meant to restore arterial pressure in these dying patients, this heroic maneuver yielded disappointing results. In 2001 Carr et al. [9] reported only 14 survivors out of 151 cases in the literature treated in this manner. In another series [11] reporting on the outcome of 302 abdominal vascular injuries, thoracotomy was performed in 131 cases (43%), of which 43 times (14%) were in the shockroom and 88 times (29%) in the operating room. The survival rates in these two groups were only 2% and 10%, respectively. In another recent series on 136 IVC injuries, one single patient survived out of 25 who underwent a thoracotomy for clamping [12]. These results can be explained by the fact that the most severely injured patients would obviously have died already on the site of the accident, were it not for the modern means of transport and prehospital reanimation.

20
195

Primary reanimation measures

In case of a penetrating trauma, the localization of one or more openings and the estimation of the route of the damaging agent can help to suspect a trauma of the caval vein. Sudden profound hemodynamic shock just after induction of anesthesia or after installation of mechanical ventilation is a very suggestive sign. A rapid response to massive fluid suppletion confirms this suspicion. If not, it indicates that the structures usually capable of internal compression are damaged and cannot prevent the caval bleeding, which is a poor sign. In order to ensure a quick filling, the reanimation team should have two large venous accesses at their disposal. It is possible to use the saphenous vein at the ankle, even in patients with IVC lesions [17]. This has been proven experimentally [18]. The rich venous supplies, together with the internal plugging capacity of a caval vein that has regained approximately its normal size, explain this controversial fact.

COMMON PRINCIPLES OF MAJOR VENOUS INJURIES

Patients who make it to the operating room should be treated by a surgical team of sufficient number. In order to respond to a possible decompensation

VASCULAR

EMERGENCIES

20
196

during anesthetic induction or surgical incision, they are usually anesthetized after preparation of the skin and operation field. In most cases the patient is positioned in the dorsal decubitus position, and the operation field runs down from the thoracic and abdominal areas to the knees. The unanimously acknowledged surgical access in case of an abdominal venous injury is a median laparotomy from xyphoid to pubic area, to which a median sternotomy is added. After the intestines are moved aside, the peritoneal cavity is emptied of a variable amount of blood by means of aspiration into a reservoir. A nonpulsating hematoma of black blood directly identifies a venous injury. Continuous aspiration of it carries the risk of exsanguination and should be prohibited. Preferably, gauzes should be packed into all abdominal quadrants, as it is difficult to localize the origin of the venous bleeding, which expands like a flat surface rather than a jet. This hemostasis enables to optimize the hemodynamic situation by correcting the volume. A temporary clampage of the infrarenal aorta only, performed without difficulty and visibly, or even simple compression, can help to reach a stable situation. In contrast with arterial injury, for which clamping of the inflow to obtain adequate hemostasis usually suffices, hemostasis of a venous injury requires checking of both the inflow and the outflow tract. Use of conventional clamps is dissuaded, especially when they have to be applied half blindly. Moreover, the posterior side of the IVC should be checked even more carefully to avoid additional damage. Whenever possible, hemostasis should be achieved through direct digital compression, fixed gauzes, covered atraumatic clamps placed under visual control or by using tourniquets. Most of the traumatic lesions of the caval vein and its branches may then be treated by means of direct closure or simple ligature in certain localizations. The use of prosthetic material in the form of patches or tubular grafts is rarely necessary. Autologous venous jugular or saphenous grafts, sometimes composed and organized in order to obtain sufficient material for the reconstruction (helical graft), are preferred.

rected, the search for venous lesions can begin. The first maneuver to perform is to tilt the patient's head down to minimize the risk of a massive air embolization. On the basis of the points of impact, the route of the damaging agent and the localization of the hematoma, it is often possible to predict the traumatized vein. Schematically, the median supramesocolic hematomas suggest a lesion of the superior mesenteric vein. Lateral, retroperitoneal peri-renal hematomas point to an injury of the renal veins. Lateral pelvic hematomas indicate a lesion of the iliac vein. Finally, hematomas in the portal region, in the middle of the hepatic pedicle, designate a portal vein injury, and those in the retrohepatic area point toward a trauma of the retrohepatic IVC and/or SHV. The decision to explore contained hematomas for which hemostasis could be achieved is controversial in view of the technical difficulties to be anticipated after decompression of venous hematomas. This is not so much the case when an infrarenal IVC injury is suspected, because of the frequently associated visceral or vascular lesions and the relatively simple reconstruction techniques. In contrast, it is more controversial in case of a retrohepatic IVC lesion, which shows no spontaneous hemorrhagic problems, but may lead to a fatal shock after mobilization of the liver [6]. A re-bleeding is indeed rare, whereas it is estimated that 10% to 40% of the patients die from a direct exposure during the attempt to repair the lesion [6].

TRAUMA OF THE IVC

The IVC is more frequently subjected to trauma than the SVC. The abdominal vascular structures are mostly injured, mainly through a penetrating trauma [11], The problems of each segment are described separately. Infrarenal IVC. The infrarenal IVC is involved in 25% to 50% of the injuries to the IVC [4,5,13,1924]. Its exposure via a right mediovisceral approach is preferred over a median retroperitoneal approach, which offers the right access to the abdominal aorta and the left renal vein but less to the IVC. The coloparietal exposure is pursued while retaining the duodenal area and the head and body of the pancreas to the left (Fig. 1). This allows exposure of the right kidney and its ureter. After rotation of the viscera, the IVC can be exposed nicely from the junction of the iliac veins to its juxtahepatic segment.

Specific aspects for each area

When the hemorrhage is controlled by means of packing gauzes and the hemodynamics are cor-

TRAUMATIC INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

A serious hemorrhage occurring during exposure of the IVC should be compressed by an assistant to allow the surgeon to continue the exposure. Subsequently, various procedures can be chosen for the hemostasis and reconstruction. The simplest one is to apply gauze over the lesion. However, maintaining sufficient compression during the whole duration of the venous repair is difficult. Total clamping of the IVC at a distance from the lesion by means of conventional clamps or tourniquets is risky due to the vulnerability of the vessel wall and the possibility of permanent reflux supplied by the collateral circulation from the azygos and lumbar veins. A smaller vessel wall trauma can be treated by means of bipolar endoluminal clamping using occlusive balloon catheters (Fogarty), or urinary catheters (Foley). Lesions of the anterior or lateral sides of the IVC can best be handled using a large Satinsky clamp (Fig. 2), which has the advantage that it partially clamps the lumen and thereby does not induce venous hypertension. Direct closure is the technique of choice to repair lesions of the infrarenal IVC. In urgent cases, a quick closure using thicker stitches (3/0 or 4/0) is preferable to trying to do a more esthetical but more time-consuming reconstruction using thinner stitches. This is underlined by the finding that there is no relation between the outcome of the reconstruction and the degree of residual stenosis [6]. Rarely, a secondary enlargement of the sutured area is required. Lesions of the posterior side can be closed, either via direct exposure by rotating the IVC (Fig. 3) after resection of one or more lumbar veins, or via a transcaval approach (Fig. 4) after having enlarged the entry opening in the anterior side, which is to be repaired afterward. In case of major destruction of the vessel wall, particularly in combination with shock and other vascular lesions, ligation of the infrarenal IVC or iliocaval junction together with ligation of all lumbar veins is generally advocated.

197

FIG. 1 Approach of the IVC, from its orisin to the subhepatic segment, by mediovisceral rotation after mobilizing the right colon, duodenum, and pancreas.

VASCULAR

EMERGENCIES

FIG. 2

Lateral clamping of the vena cava by means of a Satinsky clamp after digital control of the bleeding.

20
198

FIG. 3 Exposure and closure of a lesion in the posterior side of the IVC after transection of lumbar veins and axial rotation.

TRAUMATIC INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

FIG. 4 Closure of a perforating lesion of the IVC. The posterior wall is sutured first via the interior of the vein after having enlarged the lesion in the front. This is closed secondarily.

199
Juxtarenal and juxtahepatic IVC. The juxtarenal IVC extends two centimeters above and below the junction with the renal veins. It runs upward through a short juxtahepatic IVC, accessible via the lower border of the liver. Involvement of these segments comprises 20% to 50% of the traumatic lesions of the IVC [13,19,21-23]. Access to the Juxtarenal segment is obtained by a right mediovisceral rotation (see above). Exposure of the posterior side by means of an axial rotation requires either a right retrorenal release, mobilizing the kidney to the median line, or a complete resection between two clamps of the right renal vein, which is to be reconstructed after repositioning the IVC. The juxtahepatic segment can be approached similarly, by means of a complete right mediovisceral rotation, or more selectively by mobilizing only the duodenum and pancreas. The narrow relation between the anterior side of the IVC and the posterior side of the portal vein should be kept in mind, to avoid any damage to the latter during the dissection. The clamping techniques do not differ much from those used for the infrarenal segment. To obtain a dry operation field at the Juxtarenal level requires the simultaneous clampage of both renal veins, sometimes including the renal arteries. Control over the inflow and outflow of the Juxtarenal segment usually does not cause any problems. However, the outflow of the juxtahepatic IVC may be difficult because it is a short segment. Exerting anteroposterior pressure on the liver and thereby compressing the end of the juxtahepatic JVC leads to the same result. Reflux from the suprarenal vein, if any, is usually minimal. The reconstruction options for the Juxtarenal and juxtahepatic segments of the IVC can be superposed on those for the infrarenal segment in case of simple lesions. They basically consist of direct closure. Although a stenosis with a 75% caliber reduction due to this suture may be acceptable [6], we find it important to preserve a residual lumen of at least 30% of the initial diameter. Correction of remaining stenoses regarded excessive is the only indication

VASCULAR

EMERGENCIES

200

for enlargement procedures by means of a venous patch rather than prosthetic material. Complete replacement of this caval segment using a venous or prosthetic graft is exceptional and restricted to cases with extensive vessel wall damage in patients whose hemodynamic condition and hemostasis allow for a complex reconstruction. Retrohepatic IVC. Traumatic lesions of the retrohepatic IVC are the ones most feared. They comprise 10% to 40% of the IVC injuries and are associated with the worst prognosis. They are frequently associated with severe damage to the liver parenchyma and/or SHV. They should be suspected when single clampage of the hepatic pedicle does not lead to a reduced hemorrhage. The retrohepatic caval segment is the least surgically accessible. The risks of serious hemorrhage and massive air embolization during its exposure are unpredictable and there are extremely high mortality rates during the operation. Moreover, the most recent reports [6,12] advise to refrain from this procedure, as far as the structures surrounding the IVC are intact and ensure sufficient hemostasis, either spontaneously or after field compression. Re-inforcement of these structures is sometimes even better than exposing maneuvers, which may be fatal. In case of active hemorrhage despite tamponade, access to the retrohepatic segment may be obtained by rotating the hepatic lobe to the right, after transection of the right triangular ligament, to reach the lesions that appear from the right lateral side, or by rotating the left lobe, after transection of the left triangular ligament, for those appearing from the left lateral side. These maneuvers greatly benefit from an extension of the median laparotomy by means of a right thoracophrenotomy and, even better, by a vertical median sternotomy which offers an additional exposure of the intrapericardial IVC and the heart. The anterior side of the IVC, united with caudate lobe through multiple veins is inaccessible without dissecting the hepatic parenchyma. The context of these maneuvers limits the indications. In case of active bleeding from the anterior side of the liver, it may be better to "finger fraction" the liver parenchyma after clamping the hepatic pedicle (Pringle maneuver), which limits the bleeding. The nonviable tissues are removed progressively to reach the IVC and to repair the lesion across the hepatic defect [25]. Numerous procedures have been proposed in an attempt to dry the operation field and to enable the repair of a lesion of the retrohepatic IVC. They all have

their pros and cons and deserve a more detailed description. 1) Vascular exclusion of the liver. Complete vascular exclusion of the liver, as proposed in traumatology by Heaney et al. [26] in 1966, gives a dry operation field. It consists of a quadruple clampage (Fig. 5), which comprises clamping of the vena cava, just above the renal veins and at the intrapericardial level, after a previous total clamping of the aorta where it emerges from the diaphragm, and followed by clamping of the complete hepatic pedicle. Exclusion of the hepatic vasculature is limited by the severity of the shock and the initial hypovolemia. Without substantial filling beforehand, the main consequence of the quadruple clampage is an abolished venous return, which may cause arrhythmias and sudden cardiac arrest. 2) Atriocaval shunt. The blossoming of transplantation techniques has facilitated the concept of shunting in traumatic lesions of the retrohepatic IVC. In 1968 Buckberg et al. [27] described the placement of a completely internal shunt. This shunt, placed between the juxtarenal IVC and the right atrium, was combined with selective perfusion of the hepatic pedicle. The flow was maintained by means of a cannula inserted in the retrohepatic IVC isolated between two tourniquets. In the same year, Schrock et al. [28] reported on the use of a shunt, similar to the one of Buckberg, in a case with a trauma to the liver and SHV, introduced via a right thoracoabdominal approach, from the right atrium to the juxtarenal IVC. The subsequent modification made by Fullen et al. [29] in 1974, is the systematic approach of traumatic lesions of the retrohepatic IVC through a median sternolaparotomy. This approach enables the easy introduction of the atriocaval shunt across a pocket made onto the right atrium, and facilitates hepatic mobilization maneuvers. The most frequently used shunt is a simple thoracic drain with a large caliber (higher than 32F) in which an additional lateral orifice should be cut to allow for the blood to return to the right atrium. Finally, the retrohepatic IVC is excluded by closing the infrahepatic and intrapericardial tourniquets. The upper end of the drain exteriorized from the right atrium is of course clamped, but may also be used to administer substantial intracardial transfusions. The use of a large-caliber endotracheal tube (superior to 9 mm), in which a lateral opening is cut, seems in our view to offer additional advantages (Fig. 6). Its curvature allows its introduction while pushing on the posterior side of

TRAUMATIC INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

20
201
FIG. 5 Vascular exclusion of the liver through quadruple clamping. First the aorta is clamped, followed by clamping of the hepatic pedicle, the suprarenal vena cava, and the intrapericardial vena cava.

the FVC without inserting it into a SHV or exteriorizing it via the opening of the caval injury. In addition, the distal balloon which ensures the watertightness at the level of the juxtahepatic FVC, renders unnecessary the direct inspection of it. In all cases the administration of heparin is not necessary. Despite its apparent simplicity, the atriocaval shunt has not really brought the desired benefice in caval traumatology. In various series, very high mortality rates (80% to 100%) were found despite the use of a shunt [30]. On the basis of the North American literature between 1971 and 1996, Carr et al. [9] showed a survival of only 29 patients out of 90 who were treated with a shunt, i.e., a mortality rate of 68%. The majority of these failures

FIG. 6 The usage of an endotracheal balloon tube as atriocaval shunt to treat traumatic lesions of the retrohepatic inferior vena cava.

VASCULAR

EMERGENCIES

20
202

were due to massive bleeding when the shunt was positioned. The other failures were mainly caused by a late decision to place a shunt in an already exsanguinated patient, and the occurrence of a massive gas embolization, a perforation of the IVC or, more rarely, a thrombosis of the shunt. 3) Cavo-atrial balloon shunt. The disappointing results of the atriocaval shunt have led to the introduction of balloon shunts, inserted via a saphenofemoral route, which has a number of lateral orifices that drain the blood from the juxtarenal IVC to a single distal axial opening [31] (Fig. 7). Inflation of a circumferential balloon, localized a few centimeters away from the distal end of the shunt, enables internal tamponade of the caval lesion. Placement of this shunt can be performed before the laparotomy, although the positioning of the balloon is rather uncertain. The main advantage is that it does not require a sternotomy. It is essential to fill the balloon with water to avoid any risk of embolization in case of rupture. The risk of intrashunt thrombosis is not averted as heparinization is not usually done in this kind of injury. 4) Hypothermic circulatory arrest and active venovenous shunting. The place of extracorporeal circulation (ECC) techniques in retrohepatic IVC injuries is limited. Initially proposed in cases with major hepatic trauma, ECC allows for a hypothermic circulatory arrest in order to obtain a completely clear operation field [32,33]. High-dose heparinization, which is a prerequisite for this technique, is often contra-indicated in case of multiple other lesions. For this purpose the use of active venovenous shunting has been proposed according to what was already in use for certain hepatic transplantations [34]. Together with clamping of the caval vein on both sides of the liver and clamping of the hepatic pedicle, this kind of shunting via a saphenofemoral approach allows drainage of the caval territory and, for some, the portal area distal the inferior mesenteric vein. Venous return to the heart is accomplished via the internal jugular or axillary vein. Powered by a pump, this shunt requires only small dosages of heparin (50 lU/kg), but needs some installation time, like the conventional ECC, for access, placement of the cannulas and installation of the circuit. These two techniques do not yield control over the initial bleeding and are rarely used in the end. They do not make initial hemostasis redundant by means of compression. Their application requires a stabilized hemodynamic situation. The true aim is to enable a direct repair of

the retrohepatic IVC in a dryer operation field. Similar to the procedure for the infra or juxtarenal IVC, direct closure is preferred. Complex reconstructions at this level should remain an exception. The reconstruction of an IVC lesion at a juncture point of an SHV may require a venous or pericardial patch. In these cases downward retraction of the liver is often sufficient to expose the lesion, especially when it is located at the end of the left SHV. It should be replaced by standard rotatory maneuvers.

FIG. 7 Scheme of a cavo-atrial balloon shunt introduced via the saphenofemoral route. The balloon, inflated at the level of the retrohepatic caval lesion, ensures the hemostasis.

TRAUMATIC INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

JUXTADIAPHRAGMATIC

AND INTRAPERICARDIAL IVC

The short juxtadiaphragmatic and intrapericardial segments of the IVC form the end of it. Its anatomical situation makes it subject to penetrating trauma and closed deceleration injuries. In the latter, the heart is violently pushed forward and sometimes undergoes a rotation around its venous axis, which induces an avulsion of the IVC at the diaphragmatic orifice. The sudden occurrence of a cardiac tamponade due to a massive hemopericardium rapidly leads to death and explains why these lesions are so rarely seen in practice. A global review in 1990 [35] reported on 15 of these cases (7%) of 219 IVC traumas of which 8 (53%) were lethal. Access to the terminal segment of the IVC is easily made through a median sternolaparotomy. A longitudinal median opening of the pericardium along the right atrium exposes the IVC at its cavo-atrial junction. Digital control of the lesion is common and may also be obtained by means of a Satinsky clamp. Vascular exclusion of the liver, an atriocaval shunt or ECC may be performed. Direct closure is the ideal mode of venous repair. Interposition of a short prosthetic tube or reconstruction by means of venous or prosthetic patch is rarely used. A complete transection of the IVC may be repaired via a transatrial route using ECC under the condition that the stump of the juxtadiaphragmatic IVC has a sufficient length and solidity. In one case [36] where the proximal stump was retracted into the hepatic parenchyma, the reconstruction was performed in two sessions, consisting of a double ligation of the IVC, followed by an interposition of a re-inforced prosthetic graft (22 mm) between the juxtarenal IVC and the right atrium. The 24-hour interval between the sessions was used to stabilize the hemodynamic situation and correct the disturbed coagulation and hypothermia, while the abdomen was only partially closed by means of a mesh attached to the borders of the laparotomy incision to minimize the intra-abdominal pressure while the juxtadiaphragmatic IVC was ligated. In another case [37] an avulsion of the IVC occurred after a cardiac reintervention using ECC, which was treated by means of a triangular pericardial tissue transplant stitched at a distance from the borders of the lesion.

genic perforation of the SVC caused by an endoluminal procedure is usually less severe. A median sternotomy is the preferred approach in these injuries. The presence of a right hemothorax suggests an avulsion of the azygos vein, which is better exposed via a right thoracotomy. The main difficulty in the exposure is related to the amount of bleeding from a low-pressure circulation with a very high flow. Lateral closure, performed under digital control or after placing a lateral Satinsky clamp, is the technique of choice. The urgency of the situation does not justify the use of composed venous grafts. Replacement of an extremely damaged SVC may require an autologous internal jugular vein or a guarded prosthetic graft.

INJURIES OF THE MAJOR BRANCHES

Iliac veins. Apart from the osseous trauma to the pelvis, iatrogenic lesions from laparoscopic procedures or maneuvers to obtain control at the common iliac artery are the most common causes of this type of injury. Access to the iliocaval junction and the distal ends of the iliac veins becomes difficult due to the bleeding that quickly fills the pelvic cavity. The best route is obtained via mobilization of the right colon, retaining it to the left of the aorta and the right common iliac artery, transected between two clamps. The ureter is the only critical structure in this region. Hemostasis is obtained by compressing the IVC with gauzes against the spine and the common iliac veins against the promontory. Hemostasis of iliac vein injuries extending into the internal iliac veins is more difficult, because of their rich collateral circulation. Direct closure is standard. A too extensive venous damage should be treated with ligation of an iliac vein or iliocaval junction. The large number of side branches between the iliac veins and the reno-azygolumbar network explains why these ligations are usually well tolerated. Renal veins. Injuries to the renal veins account for 8% to 12% of abdominal vascular injuries [5, 23,39,40]. They occur in 6% to 13% of the IVC injuries [10,23,38,41,42]. These veins are injured by penetrating as well as by closed traumas. Iatrogenic trauma of the left retro-aortic renal vein during abdominal aortic surgery, should be avoided by careful analysis of the pre-operative computed tomography (CT) scan. Surgical exploration of retroperitoneal peri-renal hematomas is not justified in the case of a closed trauma, until an intravenous

203

SVC INJURIES
Traumatic lesions of the SVC are mainly caused by penetrating trauma. The surviving patients, mainly described as case reports, are few [38]. latro-

VASCULAR

EMERGENCIES

204

pyelography, a renal angiogram or a CT scan in a stable patient have shown a normal renal excretion. In contrast, surgical exploration of peri-renal penetrating injuries is standard. The right renal vein is accessed by means of mobilization of the duodenum and colon. The left renal vein is approached via a median longitudinal incision of the retroperitoneum. To reduce the bleeding it may be useful to control and clamp the renal arteries. A lateral closure, performed under digital control, either after clamping the injured vein or after lateral clamping of the IVC at its ostium, can most often be realized. If ligation of the right renal vein is necessary, one should choose between a nephrectomy and a renal autotransplantation if the hemodynamics allow for it. Ligation of the left renal vein is well tolerated as long as the capsular and genital veins are spared. Azygos vein. The azygos vein can be injured through penetrating and closed traumas. A massive hemothorax is common and indicates an approach via a right thoracotomy. The bleeding can be enhanced while spreading the ribs. There is a considerable risk of air embolization, which is why one should opt for an initial compression maneuver rather than excessive aspiration of the operation field. Final treatment consists of ligation of the azygos vein with lateral closure of the SVC when it is controlled by a Satinsky clamp. Venous brachiocephalic trunk. The right and left venous brachiocephalic trunks join to form the SVC. A lesion at this level may extend into the SVC. In practice, the left trunk is the most affected because of its length and its tract, which crosses the median line and is exposed particularly during sternotomies for cardiac procedures. Ligation is well tolerated without any sequelae.

Postoperative measures
Whether the lesion of the IVC is repaired, ligated or simply compressed, certain measures must be taken to prevent thrombo-embolic complications resulting from stasis of blood in the lower part of the body. These comprise elevation of the legs, elastic compression and antithrombotic drugs. There is no consensus on the anticoagulation regimen. Late sequelae, such as deep venous insufficiency, are rare regardless of the treatment chosen.

FOLLOW-UP AND PROGNOSTIC FACTORS In nearly 50% of cases, patients suffering from a traumatic lesion of the IVC die before any treatment is given. The results of 20 series in the literature, comprising 2032 operated patients, are presented in the Table [4,5,9-13,19-24,40-46], which shows that the mortality rate for all the different locations is 44% on average, ranging from 21% to 75%. No improvement is clear with time, despite a substantial progression in treatment options. Nevertheless it is plausible that the improvement of collection and transport of the wounded has led to the operation of more and more diseased patients. Death occurs mostly within 24 hours. In more than 90% of the cases this is due to massive bleeding, mainly during attempts to explore and repair the damage. During the early follow-up period, the bleeding induces an irreversible disseminated intravascular coagulopathy. Multi-organ failure or a sepsis, often related to intestinal lesions, are the other major causes of death after this period. The presence of shock on arrival, the existence of concomitant lesions, especially of the aorta, the closed character of the trauma [19,20,38], absent spontaneous hemostasis of the lesion and the retrohepatic orjuxtadiaphragmatic localization are significantly associated with a poor outcome. We observed a mean mortality rate of 69% (42% and 100%) in 17 series comprising a total of 344 patients operated for a retrohepatic lesion of the IVC (Table). In comparison, the mean mortality rate in case of an infrarenal lesion of the IVC is estimated at 24.5% [46]. The mortality rates of traumatic lesions of the SVC and the azygos vein surpass 50% [47]. That of the iliac veins is estimated at a mean of 25%, ranging from 10% to more than 40%, depending on whether accompanying vascular lesions are present. Finally, renal vein lesions are mortal in 12% to 30% of the cases and up to 50% when other vascular lesions are present [11,40,46]. The most frequently reported complications, whichever localization of the trauma, are respiratory complications (pulmonary embolism, pneumonia, respiratory insufficiency), abdominal infections and collections (subphrenic and parietal abscesses, peritonitis, pancreatitis, etc.), thoracic infections (mediastinitis), renal complications (renal insufficiency, venous thrombosis) and intestinal complications (gastroduodenal ulcers, ileus). The longterm prognosis of the patients surviving the acute period is, however, excellent.

TRAUMATIC INJURY OF THE VENA CAVA AND ITS MAJOR BRANCHES

1st author [ref.] Turpin [24] Graham [5] * Byrne [43] Wright [4] Sirinek [41] Millikan [44] Berguer [22] Stewart [23] Wiencek [45] Burch [10] * Klein [21] Leppaniemi [13] Feliciano [46] * Ombrellaro [20] Rosengart [19] Kuehne [12] Hansen [42] Asensio [11] Carr [9] Davies [40] Total - Mean (range)
* Series from the same center IVC: inferior vena cava NS: not specified

Overall Year Number Deaths (%)

53 37 38 23 65 38 59 30 57 37 21 39

Retrohepatic IVC Number Deaths (%)

9

1977 1978 1980 1981 1983 1983 1985 1986 1986 1988 1994 1994 1995 1997 1999 1999
2000

34
301

77 60
100

69
2

26 82 51 58 54 77 67
276

NS NS 12 11
6

NS NS 83 73 67 60 64 42 66 42 67 78 83 66
100

15 36 12
3

38 23
495

29.5
48 51 52 55 75 28

88 12
9

27 37
136

36
9

20
205

47 77 38 88

2001 2001 2001

12
3

43 NS 69 (42 - 100)

43.5
44 (21 - 75)

NS
344

2032

R E F E R E N C E S 1 Friedman SG. A history of vascular surgery. Mount Kisco, Futura Publishing Company, 1989: 212 p. 2 Taylor DC. Two cases of penetrating wounds of the abdomen involving the inferior vena cava. Lancet 1916; 2: 60-67. 3 Ochsner JL, Crawford ES, DeBakey ME. Injuries of the vena cava caused by external trauma. Surgery 1961; 49: 397-406. 4 Wright CB, Hiratzka LF, Hobson RW 2nd et al. Management of vena caval injuries: the Vietnam vascular registry review. JCardiovasc Surgl%\; 22: 203-212. 5 Graham JM, Mattox KL, Beall AC Jr, DeBakey ME. Traumade injuries of the inferior vena cava. Arch Surg 1978; 113: 413-418. 6 Buckman RF, Pathak AS, Badellino MM, Bradley KM. Injuries of the inferior vena cava. Surg Clin North Am 2001; 81:1431-1447. 7 Cellarier G, Carli P, Laurent P et al. Thrombose cave posttraumatique. Presse Med 1999; 28:1575 -1578. 8 EstreraAL, AucarJA, Wall MJJretal. Hydroblast injuries to the small bowel and inferior vena cava./ Trauma 1999; 47:979-981.

VASCULAR

EMERGENCIES

20
206

9 Carr JA, Kralovich KA, Patton JH, Horst HM. Primary venorrhaphy for traumatic inferior vena cava injuries. Am SurgWQl; 67:207-214. 10 Burch JM, Feliciano DV, Mattox KL, Edelman M. Injuries of the inferior vena cava. AmJSurg 1988; 156: 548-552. 11 Asensio JA, Chahwan S, Hanpeter D et al. Operative management and outcome of 302 abdominal vascular injuries. AmJSurg2000; 180:528-534. 12 Kuehne J, FrankhouseJ, Modrall G et al. Determinants of survival after inferior vena cava trauma. Am Surg 1999; 65: 976-981. 13 Leppaniemi AK, Savolainen HO, Salo JA. Traumatic inferior vena caval injuries. Scand J Thome Cardiovasc Surg 1994; 28: 103-108. 14 Pincu M. Traumatic aortocaval fistulas of late diagnosis. / Vase Surg 1994; 19: 1097-1098. 15 Kavic SM, Atweh N, Ivy ME et al. Renal artery to inferior vena cava fistula following gunshot wound to the abdomen. Ann Vase Sjg 2002; 16: 666-670. 16 Guillem PG, Binot D, Dupuy-CunyJ et al. Duodenocaval fistula: a life-threatening condition of various origins./ Vase SMrg-2001; 33:643-645. 17 Blaisdell FW. Management of injuries to the vena cava and portal veins. In: Bergan JJ, Yao JST (eds). Vascular surgical emergencies. Orlando, Grune & Stratton, 1987: pp 263-274. 18 Posner MC, Moore EE, Greenholz SK et al. Natural history of untreated inferior vena cava injury and assessment of venous access.}Trauma 1986; 26: 698-701. 19 Rosengart MR, Smith DR, Melton SM et al. Prognostic factors in patients with inferior vena cava injuries. Am Surg 1999; 65: 849-856. 20 Ombrellaro MP, Freeman MB, Stevens SL et al. Predictors of survival after inferior vena cava injuries. Am Surg 1997; 63:178-183. 21 Klein SR, Baumgartner FJ, Bongard FS. Contemporary management strategy for major inferior vena caval injuries. / Trauma 1994;37:35-42. 22 Berguer R, Wilson RF, Wiencek RG Jr. Traumatismes aigus de la veine cave inferieure. In Kieffer E (ed). Chirurgie de la veine cave inferieure et de ses branches. Paris, Expansion Sdentifique Francaise, 1985: pp 101-104. 23 Stewart MT, Stone HH. Injuries of the inferior vena cava. Am Suigl986;52:9-13. 24 Turpin I, State D, Schwartz A. Injuries to the inferior vena cava and their management. Am] Surg Wl; 134: 25-32. 25 Pachter HL, Spencer FC, Hofstetter SR et al. The management of juxtahepatic venous injuries without an atriocaval shunt: preliminary clinical observations. Surgery 1986; 99: 569-575. 26 HeaneyJP, Stanton WK, Albert DS et al. An improved technique for vascular isolation of the liver. Ann Surgl9Q6; 163: 237-241. 27 Buckberg GD, Ono H, Joseph WL. Hypotension following revascularization of the anoxic liver: factors influencing its occurrence and prevention. Surgery 1968; 63: 446-458.

28 Schrock T, Blaisdell WF, Mathewson C. Management of blunt trauma to the liver and hepatic veins. Arch Surg 1968; 96: 698-704. 29 Fullen WT), McDonough JJ, Popp MJ, Altemeier WA. Sternal splitting approach for major hepatic or retrohepatic vena cava injury. / Trauma 1974; 14: 903-911. 30 Burch JM, Feliciano DV, Mattox KL. The atriocaval shunt: facts and fiction. Ann Surg 1988; 207: 555-568. 31 Pilcher DB, Harman PK, Moore EE. Retrohepatic vena cava balloon shunt introduced via the sapheno-femoral junction. /Trauma 1977; 17: 837-841. 32 Launois B, de Chateaubriant P, Rosat P, Kiroff GK. Repair of suprahepatic caval lesions under extracorporeal circulation in major liver trauma./ Trauma 1989; 29: 127-128. 33 Hartman AR, YunisJ, Frei LW, Pinard BE. Profound hypothermic circulatory arrest for the management of a penetrating retrohepatic venous injury: case report. J Trauma 1991; 31:1310-1311. 34 Biffl WL, Moore EE, Franciose RJ. Venovenous bypass and hepatic vascular isolation as adjuncts in the repair of destructive wounds to the retrohepatic inferior vena cava. / Trauma 1998; 45:400-403. 35 van de Wai HJ, Draaisma JM, Vincent JG, Goris RJ. Rupture of the supradiaphragmatic inferior vena cava by blunt decelerating trauma: case report J Trauma 1990; 30: 111-113. 36 Frezza EE, Valenziano CP. Blunt traumatic avulsion of the inferior vena cava: case report. J Trauma 1997; 42:141 -143. 37 Demetriades D. Penetrating injuries to the thoracic great vessels.} Card SurgWl; 12 (2 Suppl): 173-180. 38 Kudsk KA, Bongard F, Lim RC Jr. Determinants of survival after vena caval injury. Analysis of a fourteen-year experience. Arch Swrg 1984; 119:1009-1012. 39 Jackson MR, Olson DW Beckett WC Jr et al. Abdominal vascular trauma: a review of 106 injuries. Am Surg 1992; 58: 622-626. 40 Davis TP, Feliciano DV, Rozycki GS et al. Results with abdominal vascular trauma in the modern era. Am SurgWQl', 67: 565-571. 41 Sirinek KR, Gaskill HV 3rd, Root HD, Levine BA. Truncal vascular injury: factors influencing survival./ Trauma 1983; 23: 372-377. 42 Hansen CJ, Bernadas C, West MA et al. Abdominal vena caval injuries: outcomes remain dismal. Surgery 2000; 128: 572-578. 43 Byrne DE, Pass HI, Crawford FA Jr. Traumatic vena caval injuries. Am}Surg 1980; 140: 600-602. 44 Millikan JS, Moore EE, Cogbill TH, KashukJL. Inferior vena cava injuries: a continuing challenge. JTrauma 1983; 23: 207-212. 45 Wiencek RG Jr, Wilson RF. Inferior vena cava injuries: the challenge continues. Am Surg 1988; 54: 423-428. 46 Feliciano DV. Truncal vascular trauma. In: Callow AD, Ernst CB (eds). Vascular surgery. Theory and practice. Stamford, Appleton & Lange, 1995: pp 1059-1085. 47 Mattox KL. Thoracic vascular trauma.} Vase Surg 1988; 7: 725729.

ACUTE ISCHEMIA OF THE UPPER LIMB

JOSE GONZALEZ-FAJARDO, MIGUEL MARTIN-PEDROSA LOURDES DEL RIO, CARLOS VAQUERO

Acute ischemia of the upper limb is seen infrequently compared with events in the leg. In general, occlusion of a major artery in the upper limb is better tolerated than in the lower limb, possibly because the potential for the development of an extensive collateral blood flow, but also because the bulk and work rate of the muscles of the arm are considerably less than those of the leg. Failure to recognize a critical arterial injury, however, can result in disabling ischemic symptoms including cold intolerance, claudication, and pain. Acute arm ischemia can be caused mainly by embolism, thrombosis, or trauma, although other vascular disorders (e.g., drug abuse, arteritis, thoracic outlet syndrome, hypercoagulable states) may contribute to the development of upper extremity arterial insufficiency. Today, iatrogenic brachial arterial catheter injury secondary to coronary and peripheral vascular procedures is the most common cause of direct arterial injury requiring surgery. Although there is some controversy about how aggressively acute arm ischemia should be treated, particularly when the arm appears viable, current management is aimed at restoring arterial inflow to the limb. Functional integrity of the upper limb, and more specifically the hand, is essential for daily activity, and the consequences of impaired function or amputation of an arm are disastrous, with loss of independence and/or livelihood. This chapter reviews the evidence on acute upper limb ischemia in the literature, and describes the management strategies for the treatment of traumatic and atraumatic acute occlusions of the upper limb extremity.

ZiL 207

Acute ischemia of the upper limb may occur secondary to a variety of causes, including emboli,

thrombosis, and trauma [1]. The incidence accounts for a mean of 16.6% (range 7% to 31.5%) of cases of acute ischemia of the limbs [2]. These patients usually tend to be older than those with

VASCULAR

EMERGENCIES

leg ischemia, and females present a slight preponderance in both observational and operative series.

EMBOLISM
Embolism is considered the most common cause of acute arm ischemia (74%) [2]. The emboli are attributed to a variety of sources. Cardiac embolism is the most frequently reported cause of acute arm ischemia (58% to 93%) [25] and atrial fibrillation is the usual etiology. Over the years the incidence of atrial fibrillation has remained fairly constant although the cause of fibrillation has changed from valvular heart disease as a result of rheumatic fever to ischemic heart disease and myocardial infarction. Rare causes include endocarditis, atrial myxoma, ventricular aneurysm, cardiac failure, and paradoxical embolism [2]. Non-cardiac embolism determines 1% to 32% of the acute arm emboli [2-5]. Proximal upper limb stenosis caused by atherosclerotic plaque or external compression (cervical ribs) can result in thrombo-embolism or atheroembolism, which may cause large vessel occlusion or acute digital ischemia. Other causes include atheroma in the aortic arch, primary subclavian aneurysm or aneurysm secondary to extrinsic compression from thoracic outlet syndrome, old fracture, and chronic trauma such as that from the use of crutches. Rarer sources are the proximal end of an occluded axillofemoral graft, arteritis, malignant emboli, and fibromuscular dysplasias [2]. Despite a classic embolic presentation and operative findings, an embolic source may not be found in at least 12% of patients.

sis or by producing emboli that occlude the distal circulation (Raynaud's phenomenon) [5]. Less common causes include arteritis from connective tissue disorders (scleroderma), radiation arteritis, hyperthrombotic conditions and thrombosis associated with malignancy or steroid use [5].

TRAUMA
Trauma is responsible for 15% to 45% of cases of acute arm ischemia [5,7], This ischemia may result from blunt or penetrating injuries. In penetrating injuries, especially from high-velocity missiles, the artery can be damaged either by direct penetration or by shattering of bone. These injuries usually lead to complete transection of the artery with retraction of the damaged ends of the vessel and subsequent thrombosis, thus limiting blood loss but producing severe distal ischemia. Although penetrating injuries can be caused by bullets and knives, current brachial artery injury secondary to cardiac or peripheral vascular catheterization is probably the most common cause of direct arterial injury requiring surgery. In blunt trauma, vascular injury can be mediated by joint dislocation or fracture. These injuries may produce severe longitudinal vessel traction by compression or laceration leading to intimal disruption, subintimal hematoma, and subsequent thrombosis. Because of the possibility of intimal damage causing delayed arterial occlusion, arteriography may be indicated in select instances of this type of injury. latrogenic arterial trauma. With the increasing performance of percutaneous transluminal angioplasty and insertion of intravascular devices, the risk of acute upper ischemia has increased in recent years [8], occurring in less than 1% of all brachial catheterizations. The acute arterial occlusion is the result of intimal flap or dissection from wire or catheter manipulation. The risk of thrombosis is increased if the procedure is therapeutic or if the catheter is left in place for long time. Surgical treatment is frequently indicated, and unless the ischemia is mild, observation for improvement in adults is inappropriate. Early repair under local anesthesia followed by simple thrombectomy and closure is satisfactory in most cases. Occasionally, patch angioplasty or segmental resection and anastomosis can be necessary. On the other hand, cannulation of the radial artery for pressure monitoring causes occlusive thrombosis in 10% to 20% of patients, most of which do not result in distal is-

21
208

THROMBOSIS
Reports suggest that 5% of cases in population studies and 9% to 35% in surgical series are due to thrombosis [2-5]. Jivegard et al. [6] estimated that in patients who had embolectomy, the chance of thrombosis being the true cause was 5.5% in the arm. Most of the proximal arterial lesions that can cause emboli can also result in thrombosis, including atherosclerotic plaques, aneurysm, acute aortic dissection, and arteritis (Takayasu's disease) [2]. Atherosclerosis in the upper extremity appears especially prominent in older men. The disease may be at the origin of the great vessels or distally in the axillary or brachial arteries. Aneurysms of the subclavian or axillary arteries may also result in upper extremity ischemia through two mechanisms. They may directly cause ischemic symptoms by thrombo-

ACUTE ISCHEMIA OF THE UPPER LIMB chemia because of the rich collateral network in the hand [9]. Vascular trauma associated with fractures and dislocations. Associated vascular injury in the context of general orthopedic trauma is quite infrequent (3.8% to 6.5%). However, the incidence does vary remarkably with the type of orthopedic injury encountered, being extremely rare in midshaft long-bone fractures but much more frequent with joint dislocations [10]. Given that the closer an artery lies to a bone, the more likely it is to be injured, the majority of vascular injuries associated with blunt trauma affect the brachial artery just above and below the antecubital fossa in the upper extremity. Blunt arterial injuries of the shoulder. Although vascular injury associated with blunt trauma about the shoulder is uncommon, it can be devastating. The mobility of the shoulder allows traction and avulsion of the underlying vascular and neurological structures with sometimes relatively minor bony injury. Typically, arterial injury in the shoulder area presents as a pulse or blood pressure deficit with associated shoulder or supraclavicular hematoma. The degree of ischemia is often not severe because of good collateral circulation, but there is often a distal neurological deficit caused by brachial plexus injury (Fig. 1). Furthermore, when brachial plexus injury occurs, the incidence of vascular injury is high (approximately 35%), and associated arterial injury should be carefully investigated [10]. Arteriography is important in these injuries to determine the presence and, more importantly, the level of the injury. Often, proximal control can be obtained through a supraclavicular approach. Blunt arterial injuries of the elbow. Supracondylar fractures and elbow dislocations are the most common injuries associated with blunt trauma in the upper limb. A review by Bunt et al. [11] showed a 3% incidence of vascular injury with supracondylar fractures and a 10% incidence with open elbow dislocations. Dislocation of the joint or direct supracondylar fracture essentially bow-strings the artery, and the resultant traction leads to intimal disruption or even transection. Clinical examination including documentation of pulse by palpation and doppler will identify most vascular injuries. However, because of extensive collateral flow around the elbow, up to 10% of patients with upper limb arterial injuries have preserved distal pulses [7,9,10]. When these fractures or dislocations present with obvious distal ischemia, emergent reduction is required. If perfusion is immediately restored with full pulses equal to the contralateral limb, the limb can be observed. Occasionally, reduction will improve perfusion but a pulse will not return. In these cases, angiographic assessment should be performed. The major morbidity with brachial artery injuries is attributable to the associated nerve injuries that occur twice as frequently in the traumatized upper limb as in the lower limb. Thus, although viability is maintained by vascular repair, functional outcome depends mainly on the nerve injury.

1L
209

FIG. 1 Twenty-five-year-old man with blunt impact of the left shoulder. The patient had a pulseless and cool left arm with brachial plexus deficit secondary to the joint dislocation. Angiogram shows occlusion of the proximal axillary artery and reconstitution of the proximal brachial artery. At operation, the axillary artery was badly contused and was repaired with saphenous vein bypass graft.

VASCULAR

EMERGENCIES

21
210

MISCELLANEOUS CAUSES latrogenic intra-arterial injection. Acute ischemia of the upper extremity after intra-arterial drug injection is rare. Its treatment is frequently unsuccessful and its prognosis is usually poor, often leading to gangrene. This complication is rarely seen iatrogenically [12] and is most frequently seen in intravenous drug abusers. Digital artery thrombosis is the final common pathway in the generation of critical ischemia and necrosis, which may be immediate or delayed. Different mechanisms have been suggested for this, including vasoconstriction, crystal formation of the drug in small vessels, mechanical obstruction by constituents or impurities, direct cytotoxicity, intimal damage, stasis, and thrombosis. Other factors that affect the amount of tissue damage are the dose and the nature of substances being injected. Arterial injury secondary to drug abuse. Vascular injury associated with drug abuse is a challenging problem of relatively recent origin. Its manifestations in the acute setting are frequently catastrophic [13]. The most common site of inadvertent arterial injection in drug abuse has been the antecubital fossa. Diagnostic studies should include color-flow duplex scanning. Arteriography is not advisable because it may exacerbate vasospasm and intimal damage without changing the course of therapy. Because clotting and obstruction begin in the small arterioles and can progress to involve more proximal vessels, systemic anticoagulation is indicated in an effort to retard the process and prevent further tissue loss. Lytic therapy is not advisable because this may exacerbate extravasation, which has been well documented in experimental models, and contributes to the development of a compartment syndrome. Intravenous papaverine, prostaglandin Eg, or iloprost may also be helpful. Prophylactic antibiotics are recommended to prevent infection of compromised soft tissues.

pends on both the acuteness of the episode and the localization of the occlusion. Subsequent development of ischemic changes depends on the efficiency of the collateral circulation around the shoulder and the elbow. The most common sites of upper limb occlusion are the axillary and brachial arteries, consistent with the likely sites for embolic occlusion. If acute changes are confined solely to the hand, it is likely that the distal vessels are affected or that there is a microvascular phenomenon.

Diagnosis
A careful history and physical examination often suggests a diagnosis or at least eliminates a number of disease categories [1]. A history of ischemic heart disease or recent myocardial infarction is of paramount importance. Thrombosis in an unusual site such as the axillary artery is more likely to give rise to acute symptoms and in these cases an identifiable factor may be present in the history, e.g., radiation for breast cancer or a thoracic outlet syndrome. The differential diagnosis with thrombosis is important because treatment of the latter frequently requires bypass surgery [3,4]. In general, the more proximal the lesion, the more likely thrombosis is the cause of the acute ischemia, especially if there is evidence of previous Raynaud's-like symptoms due to distal embolization [2]. This information allows for selection of the best non-invasive diagnostic work-up. Initial evaluation consists of pulse palpation and segmental doppler pressure measurement on the involved as well as the uninvolved extremity. The most reliable physical sign of acute ischemia is the absence of pulses in the affected limb in the radial and ulnar arteries at the wrist or the brachial artery in the antecubital fossa [1]. The arterial pulse should be palpable in the vessels proximal to the occlusion and in the opposite unaffected limb. The distribution of pulses will provide information about the level of the occlusion. Arterial occlusion caused by blunt trauma, fractured bones, joint dislocations, or prolonged extrinsic pressure can be recognized by a reduction in arterial pressure distal to the site of the injury (Fig. 2). In these circumstances, it must be emphasized that the presence of an audible doppler signal or even a palpable pulse does not exclude an injury, because collateral development may continue to supply some blood flow to the peripheral

Clinical presentation
The symptoms and signs of acute ischemia in the upper limb are the same as those for the lower limb and are traditionally called the six P's: pain, pallor, paralysis, paresthesia, pulselessness, and perishing cold [I]. The patient usually complains that the limb has suddenly become lifeless and numb with a variable degree of pain. The severity of the symptoms de-

ACUTE ISCHEMIA
tissues [7,9,10]. Careful pressure measurements are necessary to avoid overlooking a potentially disastrous injury. A duplex scan should be obtained whenever there is doubt. Depending on the clinical presentation, the nature of the trauma, and the uncertain duplex diagnosis, arteriography can be considered.

OF THE UPPER LIMB

of 6% [1]. The only centra-indications to surgery are inoperatibility (serious medical condition that precludes surgical intervention even under local anesthesia) and distal emboli extended to side branches that are inaccessible for balloon embolectomy or thrombectomy. These patients are better treated with anticoagulation and vasodilators. The danger of significant disability from an embolus lodging in the radial or ulnar artery is minimal and operation is not warranted for an embolus in this area [1].

Management strategies
Surgical intervention remains the most reliable method of restoring arterial inflow to the acute ischemic limb. Conservative treatment will leave one third of patients with an unsatisfactory outcome, and even accounts for an amputation rate

NON-TRAUMATIC ACUTE ISCHEMIA OF THE UPPER LIMB

Pre-operative evaluation. The difficulty in the typical case of acute arterial ischemia relates to distinguishing an embolic from a thrombotic occlusion. A pre-operative arteriogram is not indicated in a clear-cut case of embolism but may be useful in making the distinction between thrombosis and embolism and in planning an operative strategy. An inappropriate embolectomy or thrombectomy is associated with high morbidity and mortality rates [2]. However, the delay in getting to the operating room must be balanced against the possible benefit of this test. Pre-operatively, the patient should be given adequate analgesia and 5 000 units heparin intravenously. Successful embolectomies can be performed even after several days of symptoms; there is no rigid time criterion for surgical intervention. The time of ischemia is not as critical as the status of the extremity. An arm that remains viable for several days after embolization usually has an intact distal circulation, enabling restoration of flow by embolectomy at this later stage. Embolectomy. The patient is placed supine on an operating table and the affected limb, axilla, and upper chest are prepared with antiseptic solution and draped. Local anesthesia with an anesthesiologist monitoring is the best and safest way to begin an embolectomy. Mild sedation can be used if necessary. The brachial artery is exposed through an S-shaped incision over the antecubital fossa. Further dissection will expose the biceps tendon, the median nerve, and the brachial artery. The artery should be dissected out to reveal its division into radial and ulnar branches and all three vessels should be controlled with rubber slings or arterial clamps. This is supported by Beckingham et al. [14], who were able to cannulate only one artery

21
211

FIG. 2 Supracondylar fractures and elbow dislocations are the most common injuries associated with acute upper arm ischemia.

VASCULAR

EMERGENCIES

of the forearm in 42% of occasions using a brachial arteriotomy and conventional embolectomy catheter. A transverse incision in the brachial artery is preferred to prevent narrowing when the arteriotomy is repaired. A Fogarty catheter should be passed proximally to restore inflow and then distally down both the radial and ulnar arteries until back-bleeding is obtained. Following this, the vessels should be flushed with heparinized saline solution and the arteriotomy closed with 6/0 prolene. After release of the clamps, the radial pulse should be palpable at the wrist. Postoperatively the patient should be fully heparinized. Long-term anticoagulation should be routine after embolectomy unless there are medical contra-indications, particularly when the heart is the source as there is a high risk of recurrent embolism. Anticoagulation not only reduces the rate of recurrence, but also reduces postoperative mortality.

MEDICAL THERAPY
There is no universally accepted treatment protocol. Rest, analgesia, and heparinization are recommended. Some patients improve on heparin alone. The mainstays of treatment are aimed at improving the microcirculation. Regional nerve blocks, stellate ganglion blocks, and sympathectomies have been performed in the past for pain relief but do not improve the eventual outcome. The use of thrombolytic agents in the treatment of patients with acute upper arm ischemia seems to be an effective means of removing thrombus from occluded vessels. However, the potential value of this technique has been documented in small series with limited long-term follow-up data [17]. Heparinization and vasodilators have been widely used to reverse the thrombotic process and maximize the surviving circulation. Recently, iloprost (prostacycline, a prostaglandin inhibitor) and prostaglandin Eg have shown be a useful adjunct medical therapy [12].

OPERATIVE ADJUNCTS
Failure to restore the radial pulse is a string indicator of poor outcome. In this situation, treatment options include repeat embolectomy, use of intra-operative thrombolysis [15], or peri-operative angiography to identify any residual occlusion or arterial lesion. When the source of emboli is a proximal arterial lesion, definitive surgery or angioplasty may be required, either at the same time as embolectomy or later. If adequate inflow to the brachial artery is not achieved, we prefer to delay the operation to a second stage with general anesthesia, because the technical procedure should be tailored to the particular cause. Acute ischemia resulting from thrombosis has a poor outcome if treated by simple catheter thrombo-embolectomy and is more likely to require reconstructive surgery [2-4]. Today, an atherosclerotic stenosis of the subclavian or axillary artery is managed by angioplasty, with or without stenting. Occasionally, surgical bypass may be required if angioplasty is unsuccessful. Proximal lesions can be dealt with by anatomical or extra-anatomical grafting, while saphenous vein bypass to distal arm arteries is effective for selected distal occlusions [2-4]. Vascular throracic outlet compression is a clear indication for relieving surgery, since the extrinsic compression can be the cause of occlusion or distal embolism [16]. In these cases, poststenotic subclavian aneurysms require resection and grafting to prevent recurrent embolism (Fig. 3).

TRAUMATIC ACUTE ISCHEMIA OF THE UPPER LIMB

Pre-operative evaluation. Given that vascular injuries associated with limb fractures and dislocations are usually caused by high-energy trauma, the surgical dictum of life before limb prevails in the treatment of these injuries. When significant upper extremity ischemia coexists with a life-threatening cervical, thoracic, or abdominal injury, such lesions must be initially treated before initiating definitive treatment for any extremity. These patients demand a well-organized multidisciplinary approach for appropriate care [18]. In this regard, we have found that it is best to have orthopedic and vascular surgeons present in the room during the entire operation. This frequently allows the orthopedic team to choose their fixation devices in such a way that they will not hinder or complicate the subsequent vascular repair. Precise and prompt diagnosis of arterial injury is paramount in order to ensure limb viability. In most cases of penetrating or blunt arterial trauma, preoperative diagnostic arteriography is not necessary. However, surgical exploration may not be sufficient for precise location of the extension of the arterial damage in all cases. An arteriogram, for example, is especially useful in defining shotgun blast injuries in which there is a wide distribution of pellets or in patients with extensive hematoma or compart-

21
212

ACUTE ISCHEMIA
ment syndrome associated with fractures and dislocations [7,10]. Operative preparation. The operative preparation of the patient for combined and orthopedic repairs is conducted in a standard fashion regardless of whether arteriography has been performed pre-operatively or will be performed during the operation. In these cases of upper extremity injuries, at least one lower extremity is prepared and draped to allow vein harvesting, if necessary. Sequence of repair. Although the sequence of vascular repair and fracture stabilization remains

OF THE UPPER LIMB

unclear, orthopedic stabilization and immobilization should precede vascular repair in almost all patients. This is desirable for two reasons [7]: 1 - if orthopedic reduction and stabilization are performed after vascular repair, the necessary manipulation and traction may place the previously performed vascular repair under excessive tension leading to disruption; 2 - it allows the vascular surgeon to debride soft tissue that has been injured and route the bypass graft through uninvolved tissue planes without fear that subsequent orthopedic repair will expose or injure the vascular repair. Operative repair. After administration of intravenous antibiotics, the injured vessel is approached through a longitudinal incision. Proximal and distal control is established away from the focus of the lesion, using vessel loops, and the traumatized segment is then isolated and exposed. A survey of associated venous and nerve injuries as well as the extent of soft tissue injury is also undertaken. The vessel wall adjacent to the fracture site should be carefully evaluated for any sign of contusion. Afterward, a balloon embolectomy catheter should be used routinely for clearing all possible distal thrombotic material, and loco-regional anticoagulation with saline heparinized solution should be administered. There are no reports of systemic heparinization having an effect on outcome [19]. Although direct end-to-end anastomoses are preferred by some

2J_
213

FIG. 3 Severe acute ischemia of the upper limb in a youns patient with a cervical rib. Angiography shows thrombosis of the axillary artery and reconstitution of the brachial artery through collateral flow (A).

This patient was treated during several years for Raynaud's-like symptoms and upper limb digital ischemia. Repeated trauma contributes to the intimal damage and subclavian artery aneurysm (B).

VASCULAR

EMERGENCIES

21
214

authors, our experience has been that the ends of the brachial artery retract readily and, hence, this anastomosis tends to be under tension. It is our preference to use an interposition autogenous vein graft. The preferred conduit for repair of most arterial injuries is autologous greater saphenous vein [3,4]. The greater saphenous vein has an adequate diameter for most repairs of injured upper extremity arteries. Associated venous injuries. Management of concomitant venous injuries in the upper extremity is less controversial than in the lower limb, because acute or chronic edema is unusual after ligation of venous injuries in conjunction with arterial repair, as confirmed by the experience of Timberlake et al. [20]. All venous injuries were handled by ligation in conjunction with the appropriate arterial reconstruction, and neither fasciotomies nor subsequent amputations were reported. Based on this experience, we recommend repair of concomitant venous injuries in the upper extremity when technically simple and ligation of extensive injuries that would otherwise require interposition grafting. Soft tissue coverage. Vascular repairs absolutely require viable soft tissue coverage [10]. Failure to do so leads to graft infection and potentially lifethreatening hemorrhage. Many combined arterial and bone injuries have associated extensive soft tissue destruction that, following debridement, leaves extensive defects. It is not advisable to perform arterial or venous repairs in the depth of such defects because many of these wounds will require frequent daily debridement of devitalized and infected tissue. This debridement may result in exposed grafts in the depths of the wounds and place them at risk of infection and graft blow-outs. When such wounds are encountered, it is preferable to debride and ligate the injured vessels and perform arterial reconstruction using a saphenous vein bypass graft

originating from an uninvolved artery proximal to the injury, routing this bypass through intermuscular or subcutaneous tissue planes remote from the injured area, and performing the distal anastomosis distal to the injured area. This approach frequently precludes the repair of associated venous injuries, which in such cases should be ligated. Ligation. Arterial ligation is more likely to be considered when these injuries occur very distally in the limb and are associated with good collateral circulation. This may be the case in isolated injuries of single arteries of the forearm. Studies focusing on small vessel trauma have shown that outcome is limited not by the ligation or repair of the injured vessel but by the frequent concomitant nerve injury. Evaluation of patency rates after repair of singlevessel lacerations reveals that 50% to 68% of these vessels remain open at a minimum follow-up of five months. It has been demonstrated that the remaining vessel in a single vessel forearm injury experiences a compensatory flow volume increase of 46% to 65% [9]. Fasciotomy. Distal pulse palpation or doppler registry must be obtained to ensure that an adequate functional result has been achieved. Immediate postoperative follow-up consists of neuromotor, neurosensory, and vascular evaluations with special attention to the development of early signs of compartment syndrome. When any significant degree of increased compartment pressure is detected on physical examination, full-length compartment fasciotomy is performed immediately. Adherence to this policy may lead to an occasional unnecessary fasciotomy, but delay in performance of fasciotomy until the signs of compartmental hypertension are clearly evident frequently leads to irreversible muscle necrosis, nerve damage, and significantly decreased function in an anatomically salvaged extremity.

R E F E R E N C E S 1 Williams N, Bell PR. Acute ischaemia of the upper limb. BrJ 2 Eyers P, EarnshawJJ. Acute non-traumatic arm ischaemia. BrJ Surg 1998; 85: 1340- 1346. 3 Katz SG, Kohl RD. Direct revascularization for the treatment of forearm and hand ischemia. Am/Swrg-1993; 165: 312-316. 4 Pentti J, Salenius JP, Kuukasjarvi P, Tarkka M. Outcome of surgical treatment in acute upper limb ischaemia. Ann Chir Gpa>n995;84:25-28. 5 Edwards JM, Porter JM. Upper extremity arterial disease: etiologic considerations and differential diagnosis. Semin Vase 5u^ 1998; 11: 60-66.

ACUTE ISCHEMIA
6 Jivegard L, Holm J, Schersten T. The outcome in arterial thrombosis misdiagnosed as arterial embolism. Ada Chir Scand 1986; 152: 251-256. 7 Schuler JJ, Meyer JP. Arterial injuries associated with extensive soft tissue trauma. In: Ernst CB, Stanley JC (eds). Current therapy in vascular surgery, 2d edition. Philadelphia, B.C. Decker Inc 1991: pp 636-643. 8 Nehler MR, Taylor LM Jr, Porter JM. latrogenic vascular trauma. Semin Vase Surg 1998; 11: 283-293. 9 Hammond DC, Gould JS, Hanel DP. Management of acute and chronic vascular injuries to the arm and forearm. Indications and technique. Hand Clin 1992; 8: 453-463. 10 Winkelaar GB, Taylor DC. Vascular trauma associated with fractures and dislocations. Semin Vase Surg 1998; 11: 261-273. 11 Bunt TJ, Malone JM, Moody M et al. Frequency of vascular injury with blunt trauma-induced extremity injury. Am J Surg 1990-160:226-228. 12 Samuel I, Bishop CC, Jamieson CW. Accidental intra-arterial drug injection successfully treated with Iloprost. EurJ Vase Surg 1993; 7: 93-94. 13 Cooper JC, Griffiths AB, Jones RB, Raftery AT. Accidental intraarterial injection in drug addicts. EurJ Vase Surg 1992; 6: 430433.

OF THE UPPER LIMB

14 Beckingham IJ, Roberts SN, Berridge DC et al. A simple technique for thrombo-embolectomy of the upper limb. EurJ VflscSigl990;4:173-177. 15 Gonzalez-Fajardo JA, Perez-Burkhardt JL, Mateo AM. intraoperative fibrinolytic therapy for salvage of limbs with acute arterial ischemia: an adjunct to thrombo-embolectomy. Ann Vase Surg 1995; 9: 179-186. 16 Gelabert HA, Machleder HI. Diagnosis and management of arterial compression at the thoracic outlet. Ann Vase Surg 1997; 11:359-366. 17 Widlus DM, Venbrux AC, Benenati JF et al. Fibrinolytic therapy for upper extremity arterial occlusions. Radiology 1990; 175: 393-399. 18 Katsamouris AN, Steriopoulos K, Katonis P et al. Limb arterial injuries associated with limb fractures: clinical presentation, assessment and management. EurJ Vase Endovasc Surg 1995; 9: 64-70. 19 Pillai L, Luchette FA, Romano KS, Ricotta JJ. Upper extremity arterial injury. Am Swrgl997; 63: 224-227. 20 Timberlake GA, O'Connell RC, Kerstein MD. Venous injury: to repair or ligate, the dilemma. / Vase Swrgl986; 4: 553-558.

2L 215

22
ACUTE COMPLICATIONS OF ARTERIOVENOUS FISTULA FOR HEMODIALYSIS
VOLKER MICKLEY

In the immunocompromised and multimorbid dialysis patient, access complications can rapidly become life- or at least limb threatening. Access thrombosis should be treated urgently to prevent implantation of a central venous catheter for hemodialysis. In autologous fistulae, pre-operative clinical examination and color-coded duplex-sonography help to identify the causative stenosis, whereas in grafts on-table angiography immediately after surgical or interventional declotting is mandatory. Correction of the stenosis is an integral part of any declotting procedure. The endovascular surgeon would be the ideal partner for patients with thrombosed access, because he is free to choose the best interventional or surgical treatment option depending on the site and the extent of the obstruction. Access infection must be taken very seriously because it can result in major bleeding and septic complications. Immediate hospitalization, adequate antibiotic treatment, and consequent surgical intervention often resulting in access abandonment are the only means to reduce the otherwise high mortality rate. In the majority of patients with (pseudo-)aneurysm or peripheral steal syndrome, adequate pre-operative imaging and careful planning of the procedure together with skillful surgical technique will allow for correction of the vascular pathology and at the same time result in preservation of access function.

ZiZ/ 217

Renal failure End-stage renal failure (ESRF) is one of the major problems in public health care with an estimated

prevalence of 200 000 patients in Europe and growing at a rate of about 8% per year. Due to the limited availability of donor kidneys and to logistic and medical limitations of peritoneal dialysis, life-long

VASCULAR

EMERGENCIES

hemodialysis remains the only treatment option for the majority of ESRF patients. The native arteriovenous fistula (AVF) has been shown to provide the most reliable and durable access to the patient's bloodstream. Complication rates are lower than those of arteriovenous grafts (AVG) and central venous catheters (CVC), and patient survival is significantly longer, even when corrected for comorbidity [1,2]. However, when hypoplastic or exhausted peripheral veins or severely diseased peripheral arteries make construction of an autologous access impossible, a graft or even a catheter will be needed for starting or maintaining renal replacement therapy. Complications associated with vascular access are a major cause of morbidity and mortality in hemodialysis patients. In particular, those acute complications leading to failure of the access site (thrombosis, infection) represent a direct threat to the patient's life. Peripheral circulatory disturbances (aneurysm, steal syndrome) will at least endanger the afflicted extremity. Timely and effective treatment is necessary, aiming at restoration and preservation of access function whenever possible.

advantages and potential dangers of CVC for hemodialysis should be the main motivation for immediate declotting of a thrombosed access with correction of any underlying stenosis in such a way that the access can be used again for the next planned hemodialysis session. Pre-operative CVC implantation should only be considered in patients with severe electrolyte disturbances or hyperhydration, when immediate hemodialysis is necessary. The various surgical and interventional alternatives for the treatment of stenosed and thrombosed AVF and AVG are discussed extensively in an earlier EVC textbook [6]. This chapter therefore addresses only their application in acute thrombosis. AVF THROMBOSIS Diagnosis of fistula thrombosis is easily made from clinical findings. There is no bruit or murmur on auscultation and the typical thrill will be missing on palpation, although there may be pulsation of a vein segment close to the anastomosis. Identification and treatment of the underlying stenosis are integral elements of any access declotting procedure. As only superficial (or superficialized) veins are used for AVF construction, in most patients inspection and thorough palpation will reveal a fibrous and stenotic vein segment as the cause of thrombosis. Additional color-coded duplex-ultrasound may be helpful in adipose patients or otherwise questionable cases to identify the site of stenosis and the extent of thrombosis. Due to their lack of side branches, cephalic or transposed basilic veins will thrombose up to their junctions with the brachial or axillary vein, whereas forearm AW often thrombose only up to the next patent side branch. Treatment of AVF thrombosis must aim at preservation of the puncture site. Only then can CVC implantation be avoided. Surgical and interventional methods are equally effective as far as clot removal is concerned. Location and extent of the underlying stenosis, however, must be taken into account in every individual patient in order to choose the best treatment modality. The following classification of AW stenosis is intended to help to identify the adequate corrective procedure (Fig. 1). Type I - stenosis (anastomotic venous stenosis). About 80% of stenoses leading to occlusion of AW are found at or close to the arteriovenous anastomosis. Surgical dissection and mobilization of the vein on the one hand, and the extremely turbulent flow within a functioning arteriovenous

22
218

Thrombosis
The most common complication of arteriovenous access for hemodialysis is thrombosis. One might argue that acute occlusion of the access site is not a life-threatening complication, because it can be bridged with a CVC for hemodialysis until a planned intervention the next day or even some days later. However, one must not forget that implantation of a CVC is associated with a high rate of acute and late complications. Depending on the site of catheter insertion, arterial puncture, hemothorax, pneumothorax, cardiac tamponade, and retroperitoneal hematoma have been reported to occur at frequencies varying between 1% and 10% [3]. Even more important is the 40% to 50% risk of central venous obstruction after multiple or prolonged catheterization of mediastinal veins causing disabling arm swelling when located centrally to a functioning access and often requiring complex radiological or extensive surgical corrective measures [4,5]. And, of course, there is a significant risk of catheter-associated bacteremia with high rates of metastatic complications and death in the immunocompromised ESRF patient [3]. The well-known dis-

ACUTE COMPLICATIONS OF ARTERIOVENOUS FISTULA FOR HEMODIALYSIS

22_
219

TYPE III

FIG. 1

Classification of AVF stenoses.

VASCULAR

EMERGENCIES

22
220

connection on the other, have been implicated as the causes. After interventional declotting, a guide wire can be passed through the stenosis, and a percutaneous transluminal angioplasty (PTA) can be attempted. In most cases, these fibrotic stenoses afford highpressure balloons and prolonged dilatation times. Nevertheless, long-term results are so disappointing that even dedicated interventionalists recommend surgical revision [7]. After surgical dissection, the access vein is transversely opened proximal to the stenosis without affecting the puncture site. Central clot can be removed with a Fogarty catheter, assisted by digital massage in tortuous or aneurysmatic veins. On-table completion phlebography is mandatory to rule out residual clot and additional stenoses. In peripheral radiocephalic AW, distal ligation of the stenotic vein segment and proximal re-anastomosis of the cephalic vein to the radial artery is the easiest and most durable method of reconstruction [8]. Proximal re-anastomosis is often difficult in type Istenoses of brachiocephalic fistulae, when the vein must be extensively mobilized to bridge the greater distance to the artery. In these cases, a short interposition graft substituting the stenosed segment is the better alternative, because the needling segment of the access vein remains untouched [9]. Type II - stenosis (stenosis of the needling segment). Short stenoses within the needling segment or between two needling areas of the access vein may be late consequences of venous cannulations before AW construction. Multiple or long stenoses may reflect fibrotic reactions of the vein wall to the repeat cannulations for hemodialysis. PTA is the only means to treat a puncture site stenosis and at the same time completely preserve the site for immediate hemodialysis access; it should therefore be attempted first. As stent implantation is strictly contra-indicated in the needling segment, stenosis recoil, early failure, or repeated failures within short time intervals are indications for surgery [7]. In order to preserve as much of the access site as possible, short stenoses should be patched. In tortuous veins, resection of the stenosis and endto-end anastomosis of the vein may be possible. The skin incision should be as short as possible to leave some untouched access vein for the next hemodialysis sessions. When the wound is completely healed, the patched or re-anastomosed segment can be punctured again. Unfortunately, there are no studies

comparing the patency rates of autologous and synthetic patches. The latter have at least the theoretical advantage that the venous capital of the ESRF patient is not further damaged by harvesting a segment of superficial arm or saphenous vein. Multiple or long type II-stenoses should be bypassed. The author prefers a greater saphenous vein interposition graft when the caliber of the access vein upstream the stenosis is less than 5 mm, and ePTFE in larger veins. Comparative studies, however, are lacking. Only the diseased vein segment should be replaced, because 1 or 2 cm of untouched vein proximally or distally to the interposition graft may be sufficient for immediate postoperative hemodialysis access. Type III - stenosis (functional stenosis). In brachiocephalic AW, a stenosis of the vein at its junction with the axillary vein can cause thrombosis because the cephalic vein in most individuals lacks sufficient collateral side branches. Following interventional declotting, PTA of the stenosis must be performed very carefully in order to avoid rupture. Once a rupture occurs, a stent must be implanted reaching into the subclavian vein possibly causing axillary vein thrombosis and thus making later access grafts to the brachial or axillary vein impossible [7]. The surgical alternative is to dissect the central part of the cephalic vein distal to its stenosis and transpose it to the central basilic or brachial vein. Of course, late stenosis of this anastomosis can occur. Comparative studies of PTA versus surgery for this rather infrequent problem are lacking. AVG THROMBOSIS Like in AW, clinical diagnosis of AVG thrombosis is based on absence of thrill and murmur. Preoperative clinical detection of stenosis in a clotted AVG is not that easy. Arterial and venous graft anastomoses lie deeply under scar tissue, so that palpation is unlikely to identify stenotic vessel segments. Once the graft is clotted, neither color-coded duplexultrasound nor arteriography or phlebography can help to detect anastomotic stenoses. Consequently, after surgical or interventional declotting, on-table angiography of the AVG delineating both anastomoses together with the feeding artery and the draining vein is mandatory to define and immediately treat the cause of thrombosis. By definition, AVG patients are those with exhausted peripheral veins. Treatment of AVG ste-

ACUTE COMPLICATIONS OF ARTERIOVENOUS FISTULA FOR HEMODIALYSIS

nosis therefore must aim not only at preserving the graft as the puncture site but also at preserving the patient's already reduced venous capital. In contrast to surgery, interventional procedures do not require (venous) graft material or more proximal (venous) anastomotic sites. These advantages make them an intriguing alternative for the treatment of the majority of AVG stenoses. A simple classification of AVG stenosis based on clinical considerations is proposed to help identify the adequate surgical or interventional procedure for every individual finding (Fig. 2).

22_
221

FIG. 2

Classification of AVG stenoses.

VASCULAR

EMERGENCIES

Type I - stenosis (arterial anastomotic stenosis). The arterial anastomosis of an AVG is almost always a side-to-end anastomosis. Arterial anastomotic stenosis resulting from intimal hyperplasia will therefore have a complex three-dimensional configuration involving the artery immediately upstream and downstream from the anastomosis, and the anastomotic graft segment. PTA of such a stenosis can be very difficult. Depending on the access chosen for interventional graft declotting and depending on the angle in which the graft was sutured to the artery, one of the three areas of stenosis will be difficult or even impossible to traverse with a guide wire and balloon catheter. Surgical revision with resection of the anastomosis, patch or graft reconstruction of the artery, and re-anastomosis of the AVG is also demanding in small and diseased peripheral arm vessels but allows for complete correction of the stenosis. Type II - stenosis (midgraft stenosis). In some individuals with longstanding AVG access, midgraft stenosis occurs due to excessive ingrowth of fibrous tissue through multiple puncture holes. These stenoses can be dilated or curetted [10], thus leaving in place the cause of the problem. The more straightforward therapeutic option is to bypass the stenosed graft segment with a new prosthesis. If only a part of the access site must be replaced, the remaining old puncture site can be cannulated for hemodialysis. If the access site is completely stenotic, half should be curetted and the rest bypassed, just to avoid the need for CVC implantation. When restenosis develops in the curetted graft segment, it can be bypassed at a later date, again without the need for a dialysis catheter. Type III - stenosis (venous anastomotic stenosis). The great majority of grafts occlude because of progressive stenosis of the venous anastomosis. The combination of surgical trauma to vein wall and endothelium during graft implantation, compliance mismatch between graft and vein, and flow disturbances in the anastomotic area are considered the main causes. Although the graft-to-vein anastomosis is frequently sutured in an end-to-side fashion, hemodynamically it is an end-to-end anastomosis (provided the vein valves distal to the anastomosis are competent). Therefore, PTA of the stenosis is simple and should be attempted first, because it is the most vein-preserving treatment modality, although surgical thrombectomy and graft revision has been

shown to be more effective in most controlled trials [11]. In short stenoses, the surgical alternative is patch angioplasty. Stenoses longer than 5 cm or complete occlusions of the draining vein afford graft extension to a more proximal vein segment [12,13]. Graft extension should also be considered in early or repeat restenoses.

Infection
Hemodialysis patients are at permanent risk of arteriovenous access infection. Uremia itself has a strong immunosuppressive effect. Again and again the natural skin barrier is broken by repeated access punctures allowing for eventual bacterial invasion. Arm swelling caused by central venous obstruction or hematoma caused by a false puncture or compression technique are additional risk factors. The same is true for pyoderma induced by uremic pruritus and dermatitis. While the most frequently grown organism is Staphylococcus aureus, a wide spectrum of gram-positive and gram-negative bacteria has been isolated in access-related bacteremia. In ESRF, bacteremia is always a serious and lifethreatening problem. About 20% to 25% of bacteremic episodes are followed by metastatic complications, and 5% to 7% of the patients die [14,15]. Therefore, even localized puncture-induced access infections without signs and symptoms of bacteremia must be treated consequently and adequately. Patients should be hospitalized, and systemic antibiotic therapy should be given, according to sensitivity testing. In cases accompanied by chills and fever or positive blood cultures, early abandonment of the access site must be considered.

22
222

AVF INFECTION
Life-threatening infections of native AVF are rare. Postoperative deep wound infections are always anastomotic infections (Fig. 3A). If untreated they will result in formation of a false aneurysm (Fig. 3B) or in anastomotic rupture and profuse bleeding. Immediate surgical revision is mandatory. In cases with severe local inflammation, ligation of the fistula vein and reconstruction of the artery with a vein patch to the arteriotomy or a venous interposition graft (Fig. 3C) are necessary. Suturing a more proximal fistula through a clean operative field can provide early use of the revised fistula. Only in selected patients with minor signs of local infection, repair of the disrupted suture line may be performed.

ACUTE COMPLICATIONS OF ARTERIOVENOUS FISTULA FOR

HEMODIALYSIS

223

FIG. 3 PATIENT 1. A - Purulent discharge from broken skin incision three weeks after surgical revision of a formerly stenosed brachiocephalic fistula anastomosis. B - Pre-operative angiogram showing false aneurysm in extensively infected and partially necrotic access vein. C - Postoperative angiogram after complete resection of the infected access vein and arterial reconstruction with a short saphenous vein interposition graft. An end-to-end anastomosis of the cubital artery was not possible because the wall of its elongated proximal portion was heavily inflamed and partially necrotic.

VASCULAR

EMERGENCIES

In largely dilated access veins, parietal will develop, especially when there is (type II- or type III-) stenosis leading to a in flow velocity (Figs. 4A, 4B). A parietal

thrombus a central reduction thrombus

can cause sterile thrombophlebitis like in superficial varicose veins in the leg. After eventual inoculation of bacteria, however, septic endophlebitis will develop, often complicated by endocarditis and septic pulmonary metastases. Immediate extirpation of the whole access vein is mandatory together with long-term antibiotic treatment.

AVG INFECTION
Reported infection rates for prosthetic grafts range from 11% to 35% [16]. The rate of infection associated with graft implantation may be as high as 15%. Needling-induced late infection can be treated with segmental graft resection and bypass [6], together with long-term antibiotics and close monitoring for recurring infection. Early infection following graft implantation always involves the complete graft with both anastomoses. To prevent or treat anastomotic rupture and bleeding, the graft must be completely excised. At the arterial anastomosis a small cuff may be left for easier closure of the arteriotomy [17]. The draining vein can be ligated proximally and distally to the anastomosis unless it is the main draining vein of the extremity.

22
224

Aneurysm
Whereas a graft has a defined caliber and will not, or will only minimally, dilate over time, successive enlargement of the vein lumen is a prerequisite of successful long-term AW function and efficient

FIG. 4 PATIENT 2. A - Aneurysmatic degeneration of the right upper arm cephalic vein nine years after creation of an elbow fistula. A proximal type Ill-stenosis of the access vein could be assumed from inspection of the right infraclavicular region. Reddening of the skin with local pain, chills, and fever indicated bacterial endophlebitis. Blood cultures were positive for growth of Staphylococcus aureus. Immediate and complete extirpation of the access vein was performed. B - Pre-operative angiogram showing aneurysmatic degeneration of the right cephalic vein due to type Ill-stenosis.

ACUTE COMPLICATIONS OF ARTERIOVENOUS FISTULA FOR

hemodialysis therapy. An arterial aneurysm has been defined as a vessel segment having a maximal diameter larger than 1.5 times that of the normal artery. This definition can easily be adapted to a graft with a known diameter, but it seems problematic in an access vein, which can and will change its "normal" diameter over time. A circumscript ve-

HEMODIALYSIS

nous dilation should be called an aneurysm when its diameter is more than 1.5 times larger than the maximal diameter of the "normal" access vein upstream and downstream (Fig. 5). Excessive dilation/ elongation of the access vein without obvious "normal" segments should rather be called aneurysmatic degeneration (Fig. 4).

22
225

FIG. 5 PATIENT 3. A - A rapidly expandins anastomotic aneurysm 4.5 years after creation of a peripheral Brescia-Cimino-fistula on the left forearm. Pre-operative angiogram demonstrates additional postaneurysmatic stenosis. B - Postoperative angiogram after resection of the aneurysm with end-to-end anastomosis of the radial artery and re-anastomosis of the cephalic vein to the proximal radial artery.

VASCULAR

EMERGENCIES

AVF ANEURYSM
Aneurysm formation is a common phenomenon in native AW and in general can be treated as routine. A postaneurysmatic stenosis must be investigated as a possible causative factor in all cases (Figs. 4B, 5A). Needling of only a small area of the vein can also lead to aneurysmatic dilation due to the permanently repeated impairment of the wall integrity of a short vessel segment. An aneurysm should be treated when the postaneurysmatic stenosis causes low access flow with inefficient hemodialysis and the risk of thrombosis. Mural thrombi detected by color-coded duplex-ultrasound or clinically apparent thrombophlebitis are indicators for the hemodynamic relevance of the stenosis and for impending complete access thrombosis. Likewise, mural thrombi are a risk factor for septic complications (Fig. 4A). There are two indications for emergency surgery in access vein aneurysms: septic phlebitis (see above) and rapid expansion with impending skin necrosis or active bleeding. Rapidly expanding aneurysms are mainly located adjacent to the arterial anastomosis, and are associated with a postaneurysmatic stenosis (Fig. 5A). When the arteriovenous anastomosis is involved in the aneurysmatic dilation, resection of the aneurysm (and the postaneurysmatic stenosis) must be combined with reconstruction of the artery. A simple end-to-end anastomosis is often possible if the artery was sufficiently elongated during the time of aneurysm growth (Fig. 5B). Otherwise, a short interposition vein graft will be necessary to restore arterial continuity. The access vein is then re-anastomosed to the proximal artery, again if necessary with an interposition graft. When the arterial anastomosis is not involved, and when there is a short segment of "normal" vein before the aneurysm, resection of the dilated vein segment and reconstruction with a vein-to-vein graft will suffice.

in contra-indicated, because this would compromise the needling segment. Education of patient and nurses is necessary to avoid further area puncture in favor of the so-called rope ladder technique.

Steal syndrome
In arteriovenous access, the steal phenomenon is defined as retrograde diastolic flow from the distal artery into the access. This is a normal finding in functioning access, because resistance in the central venous system is significantly lower than in peripheral muscle arteries. Steal syndrome (i.e., steal phenomenon with signs and symptoms of peripheral ischemia) does not normally occur after arteriovenous access construction, because relative ischemia will induce vasodilation. This results in enhanced net arterial inflow into the respective extremity and increased collateral flow to the hand. The sensitive balance between enhanced arterial inflow and arteriovenous blood drainage can be disturbed when access flow exceeds the compensatory capacity of the arterial tree of the arm, when severe mediosclerosis or arterial obstructions prohibit sufficient vasodilatation, or in a combination of both mechanisms. Like peripheral arterial obstructive disease in the lower extremities, steal syndrome can be classified according to the Fontaine classification. Stage I-steal syndrome is synonymous with steal phenomenon, as there is retrograde flow in the distal artery (demonstrated by color-coded duplex-sonography) but no symptoms of ischemia. Stage II-steal syndrome connotes forearm or hand pain during exertion/elevation or during hemodialysis, stage Ill-steal syndrome is characterized by rest pain, and stage IV by necrosis.

22
226

STAGE II - STEAL SYNDROME AVG PSEUDO-ANEURYSM

In grafts, aneurysms are always associated with local destruction of the graft wall (Fig. 6). This will occur some months after graft implantation, when only one or two areas are needled. By definition, these aneurysms represent pseudo-aneurysms, as the normal graft wall is missing in a certain area and the wall of the pseudo-aneurysm consists of fibrous tissue. Resection and replacement of the destroyed graft segment(s) is the treatment of choice. Percutaneous implantation of a stent graft As a result of fistula maturation, blood flow tends to rise in AVF over time. Once stage II-disease has developed, there is a significant risk of deterioration. Patients with this condition must be monitored very carefully so that they are treated when symptoms become disabling and before rest pain or necrosis occur. In grafts, stage II-disease is likely to disappear when venous anastomotic stenosis successively reduces access flow. After correction of such a stenosis, there is a high risk of recurrence and even deterioration of symptoms.

ACUTE COMPLICATIONS OF ARTERIOVENOUS FISTULA FOR HEMODIALYSIS STAGE III - AND STAGE IV - STEAL SYNDROME
Once rest pain or acral necrosis has occurred, urgent diagnostic work-up and therapy are indicated. Color-coded duplex-ultrasound will demonstrate diastolic retrograde flow in the distal artery and should be used to measure flow volume in the proximal brachial artery. Arteriography including the aortic arch and the proximal arm arteries is mandatory to exclude proximal stenosis. Visualization of the forearm and hand arteries will be possible only after digital compression of the access. With an estimated frequency of 0.2% to 2.0%, the incidence of severe ischemic complications in radiocephalic AW is rather low when compared to the 2.7% to 7% in grafts and the 10% to 25% in brachiocephalic or brachiobasilic basilic fistulae [18]. Surgical ligation or interventional embolization of the radial artery distal to the anastomosis will block retrograde inflow and, in most cases, cure steal syndrome. In elbow fistulae, steal is associated with high flow volume in one third of the patients. Several banding procedures have been described for flow reduction. They are only effective when they result in a high-grade stenosis, and therefore are associated with the risk of persisting or recurring steal syndrome on the one hand and insufficient access flow or access thrombosis on the other. In the majority of patients, in whom steal syndrome occurs due to severe mediosclerosis and at low access flow (600 mL/min or less), sufficient banding would

22
227

FIG. 6 PATIENT 4. A - Rapidly expanding pseudo-aneurysm 1.5 year following implantation of an upper arm curved AVG. The punctured graft areas can easily be identified from the needle scars. B - Intraoperative angiogram demonstrates a large graft pseudo-aneurysm with partially floating mural thrombus and a smaller one without thrombus beneath the distal puncture site.

VASCULAR

EMERGENCIES

possibly result in a reduction of flow below the quantity needed for efficient hemodialysis therapy. For this situation the distal revascularization-interval ligation (DRIL) procedure (Fig. 7) was developed [19]: ligation of the cubital artery will block retrograde arterial flow without significant reduction of access flow. A bypass from the brachial artery more than 5 cm proximal to the access anastomosis to the cubital artery distal to the ligation will further enhance hand perfusion without deteriorating access function. In AVG, the risk of thrombosis is inversely correlated with flow volume. Efficient graft banding will therefore lead to reduced patency. DRIL is likely to be the better alternative because it has virtually no influence on graft flow.

Conclusion
The rapidly growing number of hemodialysis patients and the growing number of access-related complications demand increasing attention from vascular surgeons and interventional radiologists. In particular, those potentially limb- (steal syndrome, aneurysm) and life-threatening (thrombosis, infection) complications pose specific problems. A dedicated endovascular surgeon would be the ideal partner for access patients because he is free to choose or combine conventional surgical and endovascular procedures according to the nature and location of the access problem.

(J (J j~j~ 228

FIG. 7 DRIL technique (Distal Revascularization Interval Ligation) : ligation of the cubital artery will block retrograde arterial inflow without significant reduction of access flow.

R E F E R E N C E S

1 Dhingra RK, Young EW, Hulbert-Shearon TE et al. Type of vascular access and mortality in U.S. hemodialysis patients. Kidney Int 2001; 60: 1443 -1451. 2 Pastan S, Soucie JM, McClellan WM. Vascular access and increased risk of death among hemodialysis patients. Kidney Int 62; 2002: 620-626. 3 Mickley V. Central venous catheters: many questions, few answers. NephrolDial Transplant 2002; 17:1368-1373. 4 Mickley V, Gorich J, Rilinger N et al. Stenting of central venous stenoses in hemodialysis patients: long-term results. Kidney Int 1997; 51: 277-280. 5 Mickley V. Subclavian artery to right atrium haemodialysis bridge graft for superior vena caval occlusion. Nephrol Dial Transplant 1996; 11: 1361 -1362. 6 Tordoir JHM, van de Sande F, Leunissen KML. Complications of vascular access for hemodialysis. In: Branchereau A, Jacobs M

8 9 10

(eds). Complications in vascular and endovascular surgery. Parti. Armonk, Futura Publishing Company 2001: pp 225-235. Turmel-Rodrigues L, Raynaud A, Bourquelot P. Percutaneous treatment of arteriovenous access dysfunction. In: Conlon PJ, Schwab SJ, Nicholson ML (eds). Hemodialysis vascular access. Practice and problems. New York, Oxford University Press 2000: pp 183-202. Oakes DD, SherckJP, Cobb LF. Surgical salvage of failed radiocephalic arteriovenous fistulae: techniques and results in 29 patients. Kidney Int 1998; 53: 480-487. Polo JR, Vazquez R, Polo J et al. Brachiocephalic jump graft fistula: an alternative for dialysis use of elbow crease veins. Am J Kidney Dis 1999; 33: 904-909. Puckett JW, Lindsay SF. Midgraft curettage as a routine adjunct to salvage operations for thrombosed polytetrafluoroethylene hemodialysis access grafts. Am]Surg 1988; 156:139-143.

ACUTE COMPLICATIONS

OF ARTERIOVENOUS FISTULA FOR HEMODIALYSIS

15 Nielsen J, Ladevoged SD, Kolmos HJ. Dialysis catheter-related septicaemia - Focus on Staphylococcus aureus septicaemia. NephrolDial Transplant 1998; 13: 2847-2852. 16 Padberg FTJr, Lee BC, Curl GR. Hemoaccess site infection. Surg Gynecol Obstet 1992; 174:103-108. 17 Deneuville M. Infection of PTFE grafts used to create arteriovenous fistulas for hemodialysis access. Ann Vase Surg 2000; 14: 473-479. 18 Morsy AH, Kulbaski M, Chen C et al. Incidence and characteristics of patients with hand ischemia after a hemodialysis access procedure. J Surg Res 1998; 74: 8-10. 19 Berman SS, Gentile AT, Glickman MH et al. Distal revascularization-interval ligation for limb salvage and maintenance of dialysis access in ischemic steal syndrome. J Vase Surg 1997; 26: 393-404.

11 Green LD, Lee DS, Kucey DS. A metaanalysis comparing surgical thrombectomy, mechanical thrombectomy, and pharmaco mechanical thrombolysis for thrombosed dialysis grafts. J Vase S M rg2002;36:l-7. 12 Marston WA, Criado E, Jaques PF et al. Prospective randomized comparison of surgical versus endovascular management of thrombosed dialysis access grafts. / Vase Surg 1997; 26: 373 - 381. 13 PolakJF, Berger MF, Pagan-Marin H et al. Comparative efficacy of pulse-spray thrombolysis and angioplasty versus surgical salvage procedures for treatment of recurrent occlusion of PTFE dialysis access grafts. Cardiovasc Intervent Radiol 1998; 21: 314-318. 14 Marr KA, Sexton DJ, Conlon PJ et al. Catheter-related bacteriemia and outcome of attempted catheter salvage in patients undergoing hemodialysis. Ann Intern Med 1997; 127: 275-280.

22
229

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR INJURIES

TOMISLAV SOSA, IVANA TONKOVlC, LIDIJA ERDELEZ ANDRIJA SKOPLJANAC-MACINA, MARKO AJDUK, ANDREJA CRKVENAC
The number of military and civilian gunshot and explosive projectile vascular injuries has increased in the past decades. In most of the conflicts, non-Geneva weapons are in use, with the majority of wounds made by high-velocity fragments and high-velocity soft-point bullets. The initial mine or mortar fragment velocity has reached 2000 m/s [1,2]. The doubling of missile velocity quadruples its kinetic energy, and also the increase in mass augments the kinetic energy. Quadrupling of the projectile mass can have an eightfold increase in kinetic energy at the same velocity. Acting this way, the missile disperses a high quantity of the kinetic energy in the tissues, provoking the grinding effect. In addition to its direct effect, the bullet can damage the blood vessels with the high-velocity cavitational effect, resulting in stretching, disrupting, or thrombosis. The formation of a temporary cavity, however, is not a new phenomenon associated with modern high-velocity weapons.

23_
231

Missile passage
Some investigators maintain that a larger exit than entry wound is evidence of the devastating potential of an increase in velocity. While exit wounds are larger than entry wounds in 60% of cases, the difference in the size of these wounds is not per se directly attributable to the velocity, since the velocity is greater at the smaller entry wound and lesser at the greater exit wound. The larger exit wound is caused by projectile yaw or by pro-

jectile fragmentation or is a result of multiple secondary bone fragment projectiles. The point to be born in mind is that while the high-velocity projectile has the potential for higher energy transfer with subsequent greater tissue disruption, this may not always be the case. Missiles that penetrate the human body disrupt, destroy or contuse tissue, invariably resulting in a contaminated wound. The penetrating missile or fragment destroys tissue by crushing it as it punches the hole through the tissue. This hole is the permanent cavity. After

VASCULAR

EMERGENCIES

passage of the projectile, the walls of the permanent cavity are stretched radially outward. The maximum lateral tissue displacement delineates the temporary cavity. Any damage in this area is a result of the tissue stretching. The sonic shock wave precedes the projectile's passage through the tissue. Although the magnitude of the sonic wave may range up to pressures of 100 atm, its duration is so brief, about 2 us, that it does not displace tissues or have detectable harmful effect on tissues (Fig. 1).

23
232

FIG. 1 Sonic shock wave and temporary cavity made by outward stretch of the permanent cavity (modified from [2]).

Types of projectiles
+ .45 Automatic. The full metal jacketed military bullet is one of few that do not yaw significantly in soft tissue. Lack of yaw results with deep penetration. The crush tissue disruption remains nearly constant throughout the bullet path, and the temporary cavity is too small to show a stretch wounding effect, being maximal at the point of entry and gradually diminishing with penetration (Fig. 2). + .22 Long Rifle. The solid lead round-nosed bullet yaws through 90 degrees and travels base-forward for the last half of tissue path. The crush effect reaches its maximum when the bullet is traveling sideways, but it is too small to add a detectable stretch wounding effect (Fig. 3). + .38 Special. This has a special lead, roundnosed bullet. Seventy percent of these bullets yaw

through 90 degrees and travel base-forward for the latter part of their tissue path, while 30% after yawing straighten and travel point-forward for the remainder of their paths. The temporary cavity is 20% greater than that made by the Long Rifle. ^ 9 mm Parabellum. This is widely used both in pistols and submachine guns. It produces the profile that resembles that of the .38 Special but the maximum temporary cavity is about two centimeters larger in diameter and will show some stretch effects (radial splits) in less elastic, more susceptible tissues such as those of the liver. * 7,62 Nato FMC-FMC. Full metal cased, it is still used in sniper's rifles and machine guns. It shows the characteristic behavior in tissue observed in nondeforming pointed bullets. It yaws through 90 degrees and, after reaching the base-forward position, continues the rest of its path with little or no yaw. However, the rotation impaired to the bullet by the rifled gun barrel is sufficient to cause point-forward travel in the air but not in tissue, where such bullet shape and location of center of mass outweigh rotation effects. The tissue disruption in the first 15 to 18 cm of bullet penetration is minimal. At 20 to 35 cm, a large temporary cavity is produced, after marked bullet yaw. If this temporary cavity occurs in the liver, the amount of tissue disruption is likely to make survival improbable (Fig. 4). + AK-74. This smaller caliber Russian assault rifle uses the full metal cased bullet with the copper-plated steel jacket as its predecessor, the AK-47. A unique design feature of AK-74 is an air space inside the jacket at the bullet's tip. The air space serves to shift the bullet's center of mass toward the rear. This bullet yaws after only 7 cm of tissue penetration, assuring an increased temporary cavity stretch disruption, even in extremity hits. The typical exit thigh wound is stellate, with skin split measuring 9 to 13 cm across. The underlying muscle split is about half that size. The bilobed yaw pattern results from initial bullet yaw returning to zero yaw (first lobe), but then yawing a second time (second lobe) to 180 degrees, where the center of mass stabilizes the projectile in base-forward travel (Figs. 5 and 6). 4 .357 Magnum JSP. The jacketed pistol softpoint bullet and the jacketed hollow-point bullet flatten their tips on impact. This expansion or mushrooming results in doubling of effective bullet diameter in tissue, enabling the bullet to crush four times as much tissue. This conversion of the bullet to a

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR INJURIES

FIG. 2 .45 AUTOMATIC. The short, round-nosed, full metal cased bullet penetrates deeply but does not deform or yaw significantly (modified from [2]).

a23_
233
FIG. 3 LONG RIFLE. This solid bullet yaws through 90 degrees, similar to the .38 Special bullet (modified from [2]).

FIG. 4 7,62 NATO CARTRIDGE. After about 16 cm of penetration, the bullet yaws through 90 degrees, forming a large temporary cavity at the point of maximum yaw (modified from [2]).

VASCULAR

EMERGENCIES

FIG. 5 AK 47. Used very widely throuShout the world. Marked yaw besins about 25 cm after penetration (modified from [2]).

23
234
nonaerodynamic shape causes increased temporary cavity tissue stretch, as does the yawing of the bullet. The maximum temporary cavity occurs at a shallower penetration depth. This soft-point pistol bullet is typical of the type most commonly used by law enforcement agencies in the United States. It decreases the penetration depth and seldom perforates the injured (Fig. 7). + 7,62 SP. Expansion occurs on impact, but after this the bullet flattens, its pieces break off and make their own separate paths of crushed tissue. These bullet fragments penetrate up to 9 cm radially from the bullet path. The following temporary captations stretch the muscle already weakened by multiple perforations. The fragment paths act to concentrate the force of the stretch, increasing its effect and causing pieces of muscle to be detached. This synergistic effect is seen only with bullets that fragment. Very few victims with torso shots survive (Fig. 8). ^ .22 CAL FMC. The large permanent cavity made by the M-193 bullet was observed by surgeons

FIG. 6 AK 74. The bullet yaws early without deformation. As this bullet strikes soft tissue, lead flows forward filling the air space inside the bullet's tip. This produces an asymmetrical bullet, explaining the unusual curve made by the bullet path in the latter part of its penetration (modified from [2]).

who served in Vietnam. Its point-forward travel is about 12 cm, after which it yaws to 90 degrees, flattens and breaks around the groove around the bullet mid section. The bullet point remains a flattened triangular piece, and the rear portion breaks into many fragments that penetrate up to seven centimeters radially from the bullet path causing a much enlarged permanent cavity. The bullet breaks in half at the shooting distance of 80 m. At a range of 180 m, the projectile does not break and the wounding mechanism is the same as that of the AK74 (Fig. 9). + M-855 .22 CAL FMC. This is a heavier M-855 bullet used as the standard bullet for the US Armed Forces. The wound profile is similar to that produced by the M-193, although the tip does not remain in one piece. + The smaller M-193, AK-74, and M-855 are susceptible to deflection and disturbance of their point-forward flight. This can result in large yaw angles at impact and a shallower location in the body, but with maximum tissue disruption.

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR

INJURIES

This is an expanding bullet that makes a larse temporary cavity just after its penetration (modified from [2]).

23
235
FIG. 8 7,62 SOFT-POINT BULLET. The fragmentation of this bullet is responsible for the massive tissue disruption (modified from [2]).

FIG. 9 .22 CALIBER FULL METAL CASED BULLET. This was the standard weapon of the US Armed Forces for a long period, but lately has been replaced by a new rifle of the same caliber but with a heavier bullet (modified from [2]).

VASCULAR

EMERGENCIES

+ 224 Soft-point. This is designed for maximum deformation and fragmentation. The amount and type of damage is similar to that caused by the M-193 bullet, but the location of the maximum disruption is at a shallower penetration depth (Fig. 10). + 12-Gauge shotgun, #4 buckshot. This gun loaded with 27 pellets of #4 buckshot at a close range causes massive crush-type tissue disruption. At this range, soft tissue impact deforms the individual pellets increasing its section from the original 6 mm to 10 mm. The 27 perforations result in severe disruption of blood supply to tissue between

the multiple wound channels. Vascular reconstruction is impossible in this case. Maximum soft tissue disruption is expected at ranges under 25 m. Striking velocity of the projectile decreases with distance. When bone is struck, the penetration depth will be less but the degree of soft tissue disruption will be greater due to increased projectile deformation and the creation of secondary bone fragment missiles (Fig. 11). ^ Fragments from explosive devices (mines, grenades). The great majority of these projectiles are of blunt and irregular shape, not aerodynamic

23
236
FIG. 10 224 SOFT-POINT BULLET. This is a typical .22 caliber center-fire hunting bullet (modified from [2]).

FIG. 11 12-GAUGE SHOTGUN with #4 size buckshot. This weapon is used by the military and law enforcement groups for special situations (modified from [2]).

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR

and made of soft material. This causes them to loose velocity rapidly in the air. Initial fragment velocity is greater than 1800 m/s, but the wounds of survivors indicate that striking velocities were about 600 m/s. In cases where the victim is close, such as stepping on a landmine, the injury pattern is similar to that produced by # 4 buckshot at close range. In this situation, the anatomical integrity is totally destroyed. If the survivor is at some distance, the projectile track made by the fragment is consistent with its size and generally remains constant throughout its part. The temporary cavity stretches a little, while the projectile rotates, flattens and fragments. Rotation produces the grinding effect on soft tissues and implicates the vascular repair with run-in and runoff far beyond the wounding area.

INJURIES

Impact of projectile type on surgical procedure

A shot through the soft tissue of an average human thigh by a 7,62 Nato soft-point bullet can result in exit wounds up to 13 cm in diameter with massive tissue loss. If the wound is located in the proximal femoral area, vascular reconstruction is nearly always impossible. If the wound is located more distally on superficial femoral vessels, vascular reconstruction should be performed, but with special attention to the possible rupture or severe distension of the sciatic nerve, the disruption of which produces the afunctional, neurotrophic extremity. With such wounds located in the popliteal fossa, amputation is the first choice. If a shot is located cruraly with a viable foot, debridement and conservative treatment is recommended. The vascular surgeon must realize that the same destructive potential is obtained with the Nato 7,62 FMC bullet, but the exit wound does not exceed 2 cm in its largest dimension. If a patient presents after leg or arm gunshot with a .22 Long Rifle, .38 Special, or .45 automatic bullet with entrance and exit holes less than one centimeter in diameter, accompanied by vessel injury, vascular reconstruction should always be attempted. The surgical wound treatment rendered should be minimal. M-193 bullets do not cause any more tissue disruption than the .22 Long Rifle in the first 12 cm of penetration. Reconstruction of neurovascular bundles can therefore be performed, often followed

by muscular flap covering. In the torso wound, if the victim survives, the aortic and iliac damage can be substantial and requires emergent hemostasis, debridement and often extra-anatomical bypasses. AK-47 and especially AK-74 wounds might be extremely dangerous in the upper thigh or thoracic or abdominal cavity. Due to its directional change in the soft tissues, the bullet can enter the abdominal or thoracic cavity through the entry upper thigh wound. Aorto-iliac and thoracic aorta injury can be missed in such patients. Plain X-ray of the entire torso and neck is therefore recommended. The .357 Magnum JSP bullet is particularly dangerous if it enters the vascular areas. As the maximum of its temporary cavity lies within the first 15 centimeters after the point of entry, vascular injury can be produced not only by destruction, but also with arterial and venous stretching and distension. Vascular injury can be missed during the first exploration by a surgeon with limited experience in vascular surgery. A 12-gauge shotgun with #4 buckshot produces nonrepairable vascular and soft tissue injury when fired at a range up to 3 m (close range). Short ranges, if most of the pellets are slowed with bone contact, can allow a vascular reconstruction accompanied by external bone fixation. A high incidence of associated deep venous injury (82%), nerve injury (37%), fracture (33%), massive tissue loss (43%), and compartmental hypertension (39%) is observed [3]. After vascular injury with fragments from explosive devices, in victims with preserved extremity shape, vascular reconstruction should always be attempted. A substantial number of distant mine or mortar vascular injuries present in the late stage, i.e., several months after injury, causing troublesome access to the vascular structures because of the severe fibrosis that occurs as the result of tissue grinding [2,4,5].

23_
237

Epidemiology of vascular trauma

Analysis is limited to hospital trauma victims and results in a skewed picture of the true epidemiology of vascular-injury-related morbidity and mortality. In the recent past, accurate and comprehensive identification of gunshot and explosive projectile

VASCULAR

EMERGENCIES

vascular injury cases was difficult and many patients had complex hospital courses. The creation of computerized trauma registries during the past decade has partially facilitated epidemiological studies of those vascular injuries. In the majority of studies, vascular trauma is defined on the basis of patients who had blood vessels repaired. Thus, the true epidemiology of vascular injury is poorly characterized because such definition excludes individuals who die before arriving at the hospital. Furthermore, only a few studies included vascular trauma in persons who die in hospitals before blood vessel repair can be accomplished, or diagnosed blood vessel injuries that are not repaired. Because most epidemiological reports on noniatrogenic vascular injury have not been population based, direct calculation of incidence rates is not possible [6].

extent of the lesion, operative strategy and patient prognosis. The outcome is also representatively engraved in patients with gunshot and explosive vascular injuries when compared with other trauma patients. Injury severity score is 40% higher, hospital stay is prolonged by 30%, intensive care unit stay prolonged by 20%, and mortality and hospital charges doubled [13].

Trends in anatomical distribution of vascular injuries

23
238

The reverse proportion of anatomical distribution of vascular injuries can be noticed when military conflicts are compared with large urban, civilian series. In war circumstances, the majority of vascular injuries are located in the extremities (91% to 97%) and only 1% to 5% and 2% to 4% are sited in the neck and trunk, respectively. In larger civilian series, however, the neck injuries are represented in 12% and trunk injuries in 54%. Injuries to the blood vessels in the extremities are represented in 34% of individuals [6-10].

Vascular repair of gunshot and explosive projectile vascular injuries can generally be divided into 1 - primary: performed with the time gap needed for the transport, 2 - secondary: performed several months or longer after injury. In general, primary vascular reconstruction is burdened by the emergency setting, shock and its complications, primary infection, injury of the adjacent tissues, concomitant injuries, compartment syndrome and postischemic syndrome. Secondary reconstructive procedures are difficult because of severe fibrosis that, in some instances, make vascular structures inaccessible, cause trouble with hemostasis, graft stenosis and a longer rehabilitation period.

Clinical signs of vascular injury

The early, immediate signs of vascular injury can be dominated by arterial, venous or combined trauma. Arterial signs include the 5 Ps (pain, pallor, pulselessness, paresthesia, and palsy), pulse deficit distal to the gunshot wound, decreased capillary refill, diminished distal doppler segmental pressures and arterial hemorrhage. Venous injury is characterized by edema, numbness, distal hyperemia, stasis and venous hemorrhage. Late signs of vascular injury include arteriovenous fistulae and false aneurysms. Duplex scanning can locate the defect in the arterial wall and identify fistulae or false aneurysms [4,5,14,15]. Magnetic resonance angiography or angiography can complete the diagnostic process.

Mechanism of vascular injury

Penetrating trauma is the dominant cause of noniatrogenic blood vessel injuries, accounting for 50% to 90% of all such injuries. In urban settings, gunshot wounds that cause nonfatal vascular injury are most likely to involve the abdominal vessels, followed by the lower extremities [10-12]. The ballistic variable such as missile caliber and velocity together with the missile properties influence substantially the mechanism of the blood vessel lesion,

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR

INJURIES

Treatment of gunshot and explosive vascular injuries

Prior to the vascular reconstruction, general precautions must be performed, such as adequate treatment of hypovolemic shock and anesthesiologic measures to combat respiratory distress syndrome and renal failure. Time is the most important factor determining the outcome of the immediate vascular reconstruction [4,16-21]. Ischemia-reperfusion injury can result in compartment syndrome. If the total ischemic time is longer than 6 hours in patients older than 60 years or longer than 4 hours in young individuals, fasciotomy should be considered [4,17,22-25]. In general, intra-operative doppler can be performed [26, 27] but completion angiography is recommended. Because of the high incidence of stenosis, arteriovenous fistulae, or false aneurysms, postoperative surveillance by means of duplex scanning should be performed [28]. In venous injury, especially if venous stasis is present, deep venous reconstruction is the preferred method [23,25,29-33]. After surgery, limb elevation and compressive stockings decrease edema formation and anticoagulation is started. Small and limited arteriovenous fistulae can be treated by intraluminal occlusion (stent, coils), whereas large fistulae require surgical occlusion and anatomical reconstruction if possible [4,5,14,15]. In surgery, severe fibrosis aggravates the exposure and adjacent structures can be incorporated in the fibrosis. Completion angiography is recommended because multifragment lesions might have caused more distally located smaller fistulas. False arterial aneurysms also require resection and anatomical reconstruction.

According to recent reports, color-flow duplex has a 50% to 95% sensitivity, a 99% specificity and accuracy of 98% in detecting vascular injuries in the extremities. The accuracy in the thoracic outlet, the axilla and the calf is considered questionable. It is especially useful in detecting occult venous injuries that would not be detected arteriographically. However, it is highly operator dependent and should be done exclusively by an interpreter who has experience in vascular sonography [27,35,36]. If the penetrating injury lies distal to the deltoideopectoral groove or inguinal ligament, due to the algorithm for the arterial trauma with maximum sensitivity while minimizing patient morbidity and costs, we do not consider diagnostic arteriography necessary [37-40]. If time allows, arteriography is useful in proximal injuries, active hemorrhage or expanding hematoma. It is therefore used liberally in abdominal and thoracic gunshot vascular injuries in stable patients. Arteriography is mandatory when there is uncertain suspicion of vascular injury. Furthermore, when arteriovenous fistulae and/or false aneurysm are suspected, the diagnosis and endovascular occlusion of these lesions can be obtained during the same diagnostic procedure. Vertebral lesions can be detected exclusively by arteriography, which reveals occult arterial injury in 9% of patients [41]. A selective policy in use of arteriography should be followed [42,43]. Intra-operative arteriography can demonstrate distal sequential injuries [44].

25
239

Gunshot or projectile injuries of supra-aortic arteries

In another chapter of this book, J. Robbs addresses the treatment of arterial injuries to the neck arteries. In general, gunshot or explosive projectile carotid injuries are often associated with injuries of the trachea, esophagus, spine and thorax. Associated injuries are not pathognomonic for carotid injury. Shock at presentation predicts mortality (40% with shock, 8% without), but it does not predict neurologic outcome. Coma at presentation predicts mortality (higher than 70%) but is not a good predictor of neurologic outcome. Approximately 50% of patients who are initially neurologically intact after partial gunshot injury develop neurologic deficit or die. Early diagnosis

Diagnostic considerations for gunshot and explosive vascular injuries

Immediately after the injury, assessment of the ankle-brachial index is useful. A pressure index less than 0.90 predicted arterial injury with 87% sensitivity and 97% specificity. Physical examination in conjunction with doppler pressure at a threshold value of 0.90 yielded a sensitivity of 72.5% and a specificity of 100% [34].

VASCULAR

EMERGENCIES

23
240

and treatment are imperative for reducing mortality and morbidity. Vertebral arterial injury requires selective management, depending on the type of injury. Narrowing or compression is best managed by systemic anticoagulation and/or endovascular stenting. Vertebral occlusion only requires exploration in case of large hematoma. Arteriovenous fistulae require embolization or, if not successful, surgical ligation. False aneurysms can be managed by endovascular stenting or surgical repair. Innominate artery injury occurs less frequently (5.5%) than subclavian artery injury (26%). Major, life-threatening bleedings are not only caused by these arterial lesions but mainly determined by internal jugular, subclavian or innominate vein injuries [45]. Temporary shunts to maintain cerebral blood flow, while the innominate or common carotid artery are being prepared, are used infrequently. None of the several reports describe the indications for use or document the results. No valid study has been conducted to clarify this problem but many trauma surgeons suggest that if there is a scanty backflow from the artery, or stump pressures less than 70 mmHg, a temporary shunt may be needed. A variety of shunting techniques have been used including inlaying shunts, combined external and internal shunts and external ones. The role of each technique has not yet been defined. Primary ligation may be chosen in some unstable patients in whom survival depends on prompt completion of the procedure. To ensure normal function, repair of all major arteries is recommended whenever possible. In the presence of heavy contamination, remote bypass techniques may be employed in selected patients. Subclavian-carotid, axillary-axillary and carotid-carotid bypasses have been used successfully to avoid placing a graft directly into a heavily contaminated wound, such as may be encountered after erosion of a tracheostomy tube [46-49].

brink defect of the aortic wall, while bleeding is stopped by hypotension and temporary clotting. Hypovolemic shock limits the pre-operative diagnostic work-up and occasionally a computed tomography scan can be performed in stable patients. Side defects of the aortic wall are associated with late complications such as pseudo-aneurysms [50]. In emergencies, direct repair is the only choice. When direct pledged sutures are impossible, the best alternative is partial resection of the thoracic aorta followed by graft replacement. Patching in this area has not shown encouraging long-term results. During the resection, the use of a Gott shunt can be beneficial [51].

Abdominal vascular gunshot and projectile injuries

Initial management includes primary and secondary surveys, resuscitation and diagnostic studies. Many patients with abdominal vascular gunshot injuries present in hypovolemic shock. Hospital mortality is up to 75% for the aortic injuries, and 66.7% for the inferior vena cava (FVC) injuries [52]. Treatment algorithms differ for these patients. When suprarenal aortic injury is suspected, emergency thoracotomy for descending thoracic aortic control can be life saving in patients with extreme hypotension [53]. It can be done quickly in a bloodless field. However, this maneuver carries higher mortality rates and is still in the area of significant debate. Variables such as recent loss of vital signs, anticipated delay in transfer to the operating room and requirement for open cardiac massage still are the indications for emergency thoracotomy [54]. After laparotomy, supraceliac aortic control should always be obtained. Clamping in this area improves the hypotension and facilitates further exploration [55]. In injuries of the suprarenal area, distal control can be obtained with second, infrarenal clamping. Then, wide exposure can be obtained by way of medial visceral rotation of the descending colon, spleen, pancreas, stomach and left kidney. After the division of the left diaphragmatic crus, the new clamp is positioned more proximally (supraceliac) and the previous one removed. The same maneuver should be made with the infrarenal clamp. Suprarenal gunshot aortic injury carries high mortality rates particularly because it is often associated with additional vascular or visceral injuries.

Gunshot or explosive projectile injuries of the descending thoracic aorta

Patients with direct gunshot injuries of this region usually die before reaching the hospital. Among survivors with acute injury, we can often find the

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR After debridement and rinsing, simple defects can be repaired primarily. Larger defects require patch angioplasty. If the circumferential aortic segment has been shot or debrided, interposition grafting will be required, preferably with antibiotic impregnated material with re-implantation of the visceral and renal arteries. Gunshot injury of the infrarenal aorta is approached anteriorly after supraceliac control. After approaching the defect, reclamping below the renal arteries should be performed whenever possible. In a heavily contaminated field, ligation and extraanatomical reconstruction is a reasonable alternative, like axillobifemoral or bilateral axillo-unifemoral bypass. Iliac gunshot/projectile injuries are burdened with a high percentage (up to 97%) of associated injuries [56]. Intestinal lesions occur in 88% and genitourinary in 38%, yielding an average of 2.8 associated injuries per patient. Overall mortality is 23%, mostly attributed to exsanguination. If arterial ligation in a contaminated field becomes necessary, early extra-anatomical revascularization should be considered because of its improved outcome in comparison with delayed revascularization. Yet, based on the experience of prosthetic repairs in the presence of intestinal injuries where no infection was noted, the repair can be performed even with antibacterial prosthesis. Significant postoperative limb swelling occurs in all patients who undergo venous ligation and venous reconstruction should therefore be attempted. Portal region gunshot/explosive projectile injuries are very rare and highly lethal. They are associated with liver and sometimes spinal trauma. Portal vein ligation has an associated 90% mortality, whereas repair has a 42% mortality [57]. Extrahepatic arterial ligation after the Kocher maneuver is better tolerated than portal vein ligation, with a mortality rate of 58% (the repair carries 86% mortality). The surgeon must observe the early development of hepatic ischemia, which may necessitate delayed revascularization or further parenchyma resection. A repair of at least one vessel must be seriously considered. The first choice is portal vein repair. Mesenteric vascular injuries like celiac axis injuries can usually be ligated. The presence of adequate back-bleeding is a positive predictive sign for the uneventful outcome. Surgical repair is preferable, performed with autologous venous material. Repair of the superior mesenteric artery has provided better outcomes and should be done whenever possible [58]. Associated intestinal

INJURIES

injuries should be treated with contemporary stomas. Precise control of JVC injuries may be difficult because of the diffuse nature of the bleeding. Usually, the surgeon is faced with the patient whose abdominal cavity is filled with blood, and the only guidelines to IVC injury are venous blood at the puncture site and the position of the entry and exit wounds. Digital or sponge stick compression and, at times, balloon tamponade are better methods of control than the vessel loops or clamps. If the bullet has perforated the infrarenal IVC, the posterior defect can be repaired from the surgically extended anterior one. At times, ligation of the infrarenal IVC can be the only alternative. It is followed by lower extremity edema and leg elevation and compression stockings should be employed early in the postoperative period. Ligation of the suprarenal-infrahepatic IVC is less tolerated and repair should be attempted. The degree of contamination and the physiologic status of the patient dictate the appropriate method of repair. If the contamination is not severe, the best results are achieved with externally supported, large-diameter prosthetic conduits [59,60]. Retrohepatic venous injuries are some of the most challenging injuries to treat. The complete vascular isolation of the liver is mandatory, supported with the packing of the injured veins. During the repair, cavo-atrial shunt has been advised, whereas others employ veno-venous bypass [60]. If the hemorrhage is supported by coagulopathy, one may be forced to pack the area and address the injury at a later time [61]. Because mortality rates remain overwhelmingly high despite attempts of repair, stable hematomas in the retrohepatic area, confirmed by computed tomography scan, should be left unexplored.

241

Gunshot and explosive projectile vascular injuries in the extremities

Many patients with extremity gunshot/projectile arterial injuries have associated venous, nerve and orthopedic injuries due to the bullet, pellet or fragment mechanism of wounding [4,18,24,62,63]. The preferred diagnostic method is still matter of debate. Most experienced surgeons, together with the physical examination and plain X-ray, use duplex scanning for the assessment of the occult arterial

VASCULAR

EMERGENCIES

23
242

trauma because of the high sensitivity (95% to 100%), specificity (99% to 100%), and predictive accuracy (98%) when compared to that of arteriography in the setting of extremity and cervical vascular injuries in zone II [59,60]. It also saves time, the most precious constant prerequisite for a successful vascular repair. It can be used to determine hemodynamic significance of arterial injuries by providing the real-time assessment of blood flow velocities and doppler waveforms. Examples of the ultrasound findings include luminal narrowing or widening, luminal defects, vessel occlusion, blunted peak velocities, spectral broadening, arterialization of the venous blood flow, false aneurysms and deep venous thrombosis. Furthermore, adequate studies can be obtained in patients with open wounds using the sterile probe covers, and in unstable patients in the operating room. The major limitation to this technique is the availability of trained vascular technicians. Respecting the prognosis, in the presence of concomitant injuries and when the time gap is longer than six hours, vascular repair should be done first [4]. Bone stability is maintained mainly with external fixation. Venous repair has been debated extensively [64]. Repair of major venous injuries is advocated unless there is an overwhelming medical reason not to do this. The best results are achieved when nerve repair is done simultaneously.

ments in order to avoid overly frequent use of fasciotomy [73,74]. Physical findings are pain disproportionate to the injury, pain with passive stretching, weakness, hypo-esthesia or paraesthesia, paralysis, tenseness and tenderness to palpation. Allopurinol, steroids, superoxide dismutase, and mannitol can be of help in preventing and diminishing compartment syndrome with their suppression of xanthine oxidase, inflammatory response, free oxygen and with scavenge of the free radicals, respectively [75]. FASCIOTOMY Fasciotomies should be performed according to the recommended techniques, preferably with double incision (on both sides of the extremity) [69,76, 77]. Open fasciotomy should always be performed after gunshot injury. Wounds are left open with wetto-dry dressings. Dressing changes are accomplished at the bedside at a minimum of twice a day. Postfasciotomy wound management ought not to be anatomical, following the rule: less and tensionless is better [78]. The complications of fasciotomy arise mainly from its delay rather than the procedure itself [20,65-67,79,80]. Fasciotomy wound infection is primarily related to extent of tissue injury, duration of ischemia and time gap of fasciotomy. Nerve injury (saphenous, posterior tibial,

Compartment syndrome
Compartment syndrome as a condition that exists when increased pressure within a limited space compromises the circulation and function of the tissues within that space most commonly affects the lower extremities, although its occurrence in the upper extremities should not be ignored [43,6567]. It occurs primarily in response to direct musculoskeletal trauma (23% to 50% of cases) [68-70] or vascular injury with resultant ischemia and reperfusion [68,71]. Fasciotomy is more often required for vascular trauma (28% in a series of extremity vascular injuries) than for nontraumatic acute arterial occlusions (8% to 14% of patients) [70,72]. The table summarizes the different mechanisms of compartment syndrome in the extremities. Diagnosis is based on a high index of suspicion and frequent meticulous clinical examination supplemented by compartmental pressure measure-

Decreased compartment size (Extrinsic) Surgical closure of fascial defects Large scars Increased compartment contents (Intrinsic) Bleeding Vascular injuries Fractures Contusions Muscular tears Ischemia/reperfusion edema from arterial injuries Venous thrombosis Shock

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR INJURIES

peroneal) may occur during fasciotomy, most commonly with the subdermal approach. Inadequate debridement of nonviable muscles may lead to wound sepsis and necessitate amputation. concomitant abdominal part is intact, which is rarely the case after gunshot/explosive injuries. If there is a suspicion that collaterals are damaged or contused, iliac venous repair is mandatory. There is often the question which injury should be fixed first. It is most important to repair the injured artery as soon as possible to minimize the anoxia in the distal extremity. In cases when blood flow can be temporarily established by an internal shunt, it is better to repair the vein first. In many clinical situations it may be more expeditious to repair the venous injury first, at least by lateral suture, to establish hemostasis and permit better arterial exposure. However, the best way is to do both repairs as fast as possible.

Concomitant venous repair in the management of the gunshot arterial injuries in the extremities
The management of concomitant venous injuries has been controversial for decades [25,81-89]. One of the strongest arguments for the immediate repair and preservation of deep venous drainage of the extremity is to facilitate blood flow and prevent secondary complications of the edema that can aggravate the condition of the injured limb [80,86,90,91]. Barcia et al. [92] found an immediate 50% to 75% decrease in femoral arterial inflow associated with acute venous occlusion. However, Hobson et al. reported a return to baseline arterial blood flow after approximately 72 hours [93]. Injuries to the femoral, popliteal, iliac and axillar veins should be repaired. In a significantly injured extremity with arterial, venous and extensive soft tissue trauma, venous occlusion that is not corrected can significantly affect limb perfusion [94]. After the missile injury, tissue is not only destroyed in its direct path but also in a much wider surrounding zone of injury. These cavitation effects are especially notable when high-velocity weapons are used. Such types of wounds may be seen in military conflicts as well as in civilian gunshot wounds. These types of wounds with the blast effect in the surrounding tissue do not easily lend themselves to primary venous repairs. Most gunshot/explosive projectile venous injuries must be managed by interposition graft techniques or ligation. A duplicated femoral vein exists in 20%, and if a preserved duplicated vein exists, the other injured vein may be ligated. Most of the tibial vein injuries may be ligated due to the six outflow veins in this region; this is not the case with the popliteal vein, which always should be reconstructed because of the significant edema after simple ligation. After gunshot injury, reconstruction of cubital, brachial and axillary veins is recommended to improve the function of the hand. Repair of the iliac vein may not be an obligation when there is absence of inguinal injury and the subcutaneous tissue of the

Personal experience
During a 25-year period we have treated 33 civilian gunshot vascular injuries. Wartime and explosive projectile vascular injuries were much more frequent. Between 1991 and 1995 we treated 220 patients with 347 vascular injuries. There were 207 arterial and 140 venous injuries as well as associated fractures and nerve injuries. The mean prehospital time for this group was 7 hours (range 3 to 18 h) [24]. Two hundred and seven (59.6%) arterial and 140 (40.4%) vein lesions were treated. This is compatible with the previous reports [24,95]. Six percent of these vascular injuries were caused by high-velocity fragments from exploding devices and 16.4% by high-velocity softpoint bullets. The mean age of patients was 25 years (range 16 to 47 years). The most common vascular injury was to the popliteal artery and vein. Injuries to the superficial femoral and brachial vessels were slightly less frequent. Ten of the patients had isolated venous injuries. The fractures were stabilized by an external fixator in 90.4% of cases. Fractures appear to be the concomitant injuries that aggravate the vascular repair because they occur in up to 34% of patients making them candidates for the complex treatment. Associated nerve lesions seriously compromise the outcome. In our series, they presented in 40.4% of casualties. The best results were obtained after simultaneous nerve reconstruction during the one procedure. Seventy-four fasciotomies (48% on upper and 62% on the lower limbs) were performed because

23_
243

VASCULAR

EMERGENCIES

of compartment syndrome. This represents 44.5% among all gunshot injuries of the extremity vessels. Vessel injuries repaired within four hours after the incident did not require fasciotomy. The muscle viability can be estimated clinically (color, contractility, bleeding). However, in 16 patients muscle biopsy was of help in the decision for fasciotomy. Arteriovenous fistulae presented in 5.4% vascular injuries, and false aneurysms were encountered in 9.5%. Acute arteriovenous fistulae were present in only 7 gunshot injuries (2.0%) and acute false aneurysms in 6 (1.7%). The rest presented as the late sequelae of arterial and venous injury. As stated before, peripheral arterial injury was repaired by means of autologous venous interposition or bypass grafting in most patients. Eighteen patients who had blood vessel injury underwent amputation (8.1%). The indications for amputation were sepsis, deep venous thrombosis and extensive myonecrosis. Concomitant vein, nerve, soft tissue and bone necrosis were present in all amputees. Sixteen patients died. Death was caused

by associated craniocerebral, pulmonary and bowel injuries, as well as with sepsis. During follow-up the vascular status in all examined patients was entirely satisfactory and none of these patients suffered from ischemic symptoms. One late amputation was done because of the neurotrophic limb, not because of ischemia.

Conclusion
Gunshot and projectile vascular injuries are complex and require extensive knowledge about the different types of bullets and missiles and subsequent surgical management. Arterial repair is mandatory in most cases in order to preserve viability. Venous repair deserves a selective approach, but permanent edema is a threat for healing processes. Associated surgical procedures like fasciotomy, bone, organ and nerve repair are extremely important, especially for the functioning and quality of life on the long term.

23
244
R E F E R E N C E S 1 Rich NM, Spencer FC. Vascular Trauma. Philadelphia, W.B. Saunders Co., 1978 : 610 p. 2 Bowen TE, Bellamy RF (eds). Emergency war surgery, Second United States revision of the Emergency war surgery NATO handbook. Washington, DC, US Department of Defense, 1988 : 446 p. 3 Meyer JP, Lim LT, Schuler Jj et al. Peripheral vascular trauma from close-range shotgun injuries. Arch Surg 1985; 120: 1126-1131. 4 Luetic V, Sosa T, Tonkovic I et al. Military vascular injuries in Croatia. Cardiovasc Surg 1993; 1 : 3-6. 5 Feliciano DV, Cruse PA, Burch JM, Bitondo CG. Delayed diagnosis of arterial injuries. Am J Surg 1987; 154: 579-584. 6 Caps MT. The epidemiology of vascular trauma. In: Rutherford RB (ed). Vascular surgery, 5th ed. Philadelphia, W.B. Saunders Co., 2000 : pp 857-862. 7 Bole PV, Purdy RT, Munda RT et al. Civilian arterial injuries. Ann Surg 1976; 183: 13-23. 8 Bongard F, Dubrow T, Klein S. Vascular injuries in the urban battleground: experience at a metropolitan trauma center. Ann Vase Surg im-4: 415-418. 9 Rich NM, Baugh JH, Hughes CW. Acute arterial injuries in Vietnam: 1000 cases. J Trauma 1970; 10: 359-369. 10 Smith ATJr, Kuller LH, PerperJA et al. Epidemiology of homicide in Allegheny County, Pennsylvania, between 1966-1974 and 1984-1993. Prev Med 1998; 27: 452-460. 11 Spigner C. Race, class and violence: research and policy implications. Int] Health Serv 1998; 28: 349-372. 12 Wintemute GJ, Wright MA, Parham CA et al. Criminal activity and assault-type handguns: a study of young adults. Ann Emerg Med 1998; 32: 44-50. 13 Mattox RL, Feliciano DV, Burch J et al. Five thousand seven hundred sixty cardiovascular injuries in 4559 patients. Epidemiologic evolution 1958 to 1987. Ann Surg 1989; 209: 698-705. 14 Khoury G, Sfeir R, Nabbout G et al. Traumatic arteriovenous fistulae: the Lebanese war experience. EurJ Vase Surg 1994; 8: 171-173. 15 Porcellini M, Bernardo B, Selvetella L et al. Late arteriovenous fistulas and pseudo-aneurysms of the extremity secondary to civilian firearms. G Chir 1997; 18: 91-96. 16 Hussain ST, Aslam S, Khan RA et al. An observational study of 256 cases of vascular trauma in the northwestern province of Pakistan. Ann R Coll SurgEngl 2001; 83: 388-391. 17 Ozisik K, Erturk M. Management of military vascular injuries. / Cardiovasc Surg 2001; 42: 799-803. 18 Sfeir RE, Khoury GS, Kenaan MK. Vascular trauma to the lower extremity: the Lebanese war experience. Cardiovasc Surg 1995; 3: 653-657. 19 Fasol R, Irvine S, Zilla P. Vascular injuries caused by anti-personnel mines. / Cardiovasc Surg 1989; 30: 467-472. 20 Nair R, Abdool-Carrim AT, Robbs JV. Gunshot injuries of the popliteal artery. BrJSurgmO; 87: 602-607. 21 Fletcher JP. Vascular trauma. Ann Acad Med Singapore 1992; 21: 294-296. 22 Armstrong K, Sfeir R, Rice J, Kerstein M. Popliteal vascular injuries and war: are Beirut and New Orleans similar IJ Trauma 1988; 28: 836-839. 23 Sharma PV, Ivatury RR, Simon RJ, Vinzons AT. Central and regional hemodynamics determine optimal management of major venous injuries. / Vase Surg 1992; 16: 887-894.

GUNSHOT AND EXPLOSIVE PROJECTILE VASCULAR

24 Tonkovic I, Petrunic M, Fiolic Z et al. War injuries of the arteries. Lijec Vjesn 1997; 119: 316-319. 25 Timberlake GA, Kerstein MD. Venous injury: to repair or ligate, the dilemma revisited. Am Surg 1995; 61: 139-145. 26 Johansen KH. Vascular trauma. In: Rutherford RB (ed). Vascular surgery, 5th ed. Philadelphia, W.B. Saunders Co., 2000 : pp 857-916. 27 Hood DB, Weaver FA, Yellin AE. Nonarteriographic evaluation of penetrating vascular injuries. In: Ernst CB, StanleyJC (eds). Current therapy in vascular surgery, 4th ed. St. Louis, Mosby 2001 : pp 580-583. 28 Wicky B, Largiader J, Pouliadis G, Zollikhofer C. Clinical and angiographic follow-up of vascular injuries of the major arteries of the extremities. Chirurg 1984; 55: 244-248. 29 Zamir G, Berlatzky Y, Rivkind A et al. Results of reconstruction in major pelvic and extremity venous injuries. / Vase Surg 1998; 28:901-908. 30 Yelon JA, Scalea TM. Venous injuries of the lower extremities and pelvis: repair versus ligation. J Trauma 1992; 33: 532-538. 31 Meyer J, Walsh J, Schuler J et al. The early fate of venous repair after civilian vascular trauma. A clinical, hemodynamic, and venographic assessment. Ann Surg 1987; 206: 458-464. 32 Feliciano DV, Mattox KL, Graham JM, Bitondo CG. Five-year experience with PTFE grafts in vascular wounds. / Trauma 1985; 25: 71-82. 33 Timberlake GA, O'Connell RC, Kerstein MD. Venous injury: to repair or ligate, the dilemma. / Vase Surg 1986; 4: 553-558. 34 Johansen K, Lynch K, Paun M, Copass M. Non-invasive vascular tests reliably exclude occult arterial trauma in injured extremities. / Trauma 1991; 31: 515-522. 35 Weaver FA, Yellin AE, Bauer M et al. Is arterial proximity a valid indication for arteriography in penetrating extremity trauma? A prospective analysis. Arch Surg 1990; 125: 1256-1260. 36 Schwartz MR, Weaver FA, Bauer M et al. Refining the indications for arteriography in penetrating extremity trauma: a prospective analysis. / Vase Surg 1993; 17: 116-124. 37 Bynoe RP, Miles WS, Bell RM et al. Noninvasive diagnosis of vascular trauma by duplex ultrasonography. / Vase Surg 1991; 14: 346-352. 38 Fry WR, Smith RS, Sayers DV et al. The success of duplex ultrasonographic scanning in diagnosis of extremity vascular proximity trauma. Arch Srgl993; 128: 1368-1372. 39 Bergstein JM, Blair J, Edwards J et al. Pitfalls in the use of colorflow duplex ultrasound for screening of suspected arterial injuries in penetrated extremities. / Trauma 1992; 33: 395-402. 40 Gagne PJ, Cone JB, McFarland D et al. Proximity penetrating extremity trauma: the role of duplex ultrasound in the detection of occult venous injuries./ Trauma 1995; 39: 1157-1163. 41 Pasch AR, Bishara RA, Lim LT et al. Optimal limb salvage in penetrating civilian vascular trauma. J Vase Surg 1986; 3: 189-195. 42 Degiannis E, Levy RD, Sofianos C et al. Arterial gunshot injuries of the extremities: a South African experience. J Trauma 1995; 39: 570-575. 43 Pillai L, Luchette FA, Romano KS, Ricotta JJ. Upper-extremity arterial injury. Am Surg 1997; 63: 224-227. 44 Roberts RM, String ST. Arterial injuries in extremity shotgun wounds: requisite factors for successful management. Surgery 1984; 96: 902-908. 45 Perry MO. The management of acute vascular injuries. Baltimore, Williams & Wilkins, 1981 : pp 55-65. 46 Flint LM, Snyder WH, Perry MO, Shires GT. Management of major vascular injuries in the base of the neck. An eleven-year experience with 146 cases. Arch Surg 1973; 106: 407-413. 47 Reul GJ Jr, Beall AC Jr, Jordan GL, Mattox KL. The early operative management of injuries to the great vessels. Surgery 1973; 74: 862-868. 48 Calhoon JH, Grover FL* Trinkle JK Chest trauma. Approach and management. Clin Chest Med 1992; 13 : 55-67.

INJURIES

49 Thai ER. Aortic arch and proximal brachiocephalic penetrating injuries. In: Ernst CB, Stanley JC (eds). Current therapy in vascular surgery, 2nd ed. Philadelphia, B.C. Decker, 1991 : pp 617-622. 50 Hood JM, Blair PHB. Vascular trauma. In: Beard JD, Gaines PA (eds). Vascular and endovascular surgery, 2nd ed. London, W.B. Saunders Co, 2001 : pp 169-197. 51 Kram HB, Appel PL, Wohlmuth DA Shoemaker WC. Diagnosis of traumatic aortic rupture: a ten-years retrospective analysis. Ann Thorac Surg 1989; 47: 282-286. 52 Sriussadaporn S. Abdominopelvic vascular injuries. J Med Assoc Thai 2000; 83: 13-20. 53 Feliciano DV. Abdominal vascular injuries. Surg Clin North Am 1988; 68: 741-755. 54 Wiencek RG Jr, Wilson RF. Injuries to the abdominal vascular system: how much does aggressive resuscitation and prelaparotomy thoracotomy really help? Surgery 1987; 102: 731-736. 55 Rotondo MF, Schwab CW, McGonigal MD et al. Damage control: approach for improved survival in exsanguinating penetrating abdominal injury./ Trauma 1993; 35: 375-383. 56 Carrillo EH, Spain DA, Wilson MA et al. Alternatives in the management of penetrating injuries to the iliac vessels. / Trauma 1998; 44: 1024-1030. 57 Jurkovich GJ, Hoyt DB, Moore FA et al. Portal triad injuries. /Trauma 1995; 39: 426-434. 58 Landau DS, Schuler JJ. Blunt and penetrating abdominal vascular injuries. In: Ernst CB, Stanley JC (eds). Current therapy in vascular surgery, 4th ed. St. Louis, Mosby, 2001 : pp 605-609. 59 Salam AA, Stewart MT. New approach to wounds of the aortic bifurcation and inferior vena cava. Surgery 1985; 98: 105-108. 60 Rogers FB, Reese J, Shackford SR, Osier TM . The use of venovenous bypass and total vascular isolation of the liver in the surgical management of juxtahepatic venous injuries in blunt hepatic trauma. / Trauma 1997; 43: 530-533. 61 Denton JR, Moore EE, Coldwell DM. Multimodality treatment for grade V hepatic injuries: perihepatic packing, arterial embolization, and venous stenting./ Trauma 1997; 42: 964-968. 62 Radonic V, Baric D, Petricevic A et al. War injuries of the crural arteries. Br/Swrg 1995; 82: 777-783. 63 Dajani OM, Haddad FF, Hajj HA et al. Injury to the femoral vessels - the Lebanese War experience. EurJ Vase Surg 1988; 2: 293-296. 64 lerardi RP, Rich N, Kerstein MD. Peripheral venous injuries: a collective review. J Am Coll Surg 19%; 183: 531-540. 65 Matsen FA 3rd, Winquist RA, Krugmire RB Jr. Diagnosis and management of compartmental syndromes. / Bone Joint Surg Am 1980; 62: 286-291. 66 Myers SI, Harward TR, Maher DP et al. Complex upper extremity vascular trauma in an urban population. / Vase Surg 1990; 12: 305-309. 67 Sriussadaporn S. Vascular injuries of the upper arm. JMed Assoc 7M 1997; 80: 160-168. 68 Vitale GC, Richardson DJ, George SM Jr, Miller FB. Fasciotomy for severe blunt and penetrating trauma of the extremity. Surg Gynecol Obstet 1988; 166: 397-401. 69 Wrilliams AB, Luchette FA, Papaconstantinou HT et al. The effect of early versus late fasciotomy in the management of extremity trauma. Surgery 1997; 122: 861-866. 70 Abouezzi Z, Nassoura Z, Ivatury RR et al. A critical appraisal of indications for fasciotomy after extremity vascular trauma. Arch Surg 1998; 133:547-551. 71 Perry MO. Compartment syndromes and reperfusion injury. Surg Clin North Am 1988; 68: 853-864. 72 Nypaver TJ, Whyte BR, Endean ED et al. Nontraumatic lowerextremity acute arterial ischemia. AmJSurg 1998; 176:147-152. 73 Whitesides TE, Haney TC, Morimoto K, Harada H. Tissue pressure measurements as a determinant for the need for fasciotomy. Clin Orthop 1975; 113: 43-51. 74 Field CK, Senkowsky J, Hollier LH et al. Fasciotomy in vascular trauma: is it too much, too often? Am Surg 1994; 60: 409-411.

25
245

VASCULAR

EMERGENCIES

75 Fan tone JJ. Pathogenesis of ischemia-reperfusion injury: an overview. In: Zelenock G (ed). Clinical ischemic syndromes: mechanisms and consequences of tissue injury. St. Louis, Mosby 1990. 76 Mubarak SJ, Owen C. Double-incision fasciotomy of the leg for decompression in compartment syndromes. / Bone Joint Surg ylwl977;59: 184-187. 77 Nghiem DD, Boland JP. Four-compartment fasciotomy of the lower extremity without fibulectomy: a new approach. Am Surg 1980; 46: 414-417. 78 Cuschieri J, Anagnostopoulos P, Kralovich KA et al. Fasciotomy wound management: less is better. J Trauma (in press). 79 Ernst CB, Kaufer H. Fibulectomy-fasciotomy. An important adjunct in the management of lower extremity arterial trauma. J Trauma Wl; 11: 365-380. 80 Conkle DM, Richie RE, Sawyers JL, Scott HWJr. Surgical treatment of popliteal artery injuries. Arch Surg 1975; 110:1351 -1354. 81 Borman KR, Jones GH, Snyder WH 3rd. A decade of lower extremity venous trauma: patency and outcome. Am J Surg 1987; 154: 608-612. 82 Menzoian JO, Doyle JE, Cantelmo NL et al. A comprehensive approach to extremity vascular trauma. Arch Surg 1985; 120: 801-805. 83 Meyer J, Walsh J, Schuler J. The early fate of venous repair following civilian vascular trauma. A clinical, hemodynamic and venographic assessment. Ann Swrg-1987; 206: 458-464. 84 Mullins RJ, Lucas CE, Ledgerwood AM. The natural history following venous ligation for civilian injuries. / Trauma 1980; 20: 737-743. 85 Nypaver TJ, Schuler JJ, McDonnell P et al. Long-term results of venous reconstruction after vascular trauma in civilian practice./ Vase Surg 1992; 16:762-768.

86 Pappas PJ, Haser PB, Teehan EP et al. Outcome of complex venous reconstructions in patients with trauma. / Vase Surg 1997; 25: 398-404. 87 Phifer TJ, Gerlock AJ Jr, Rich NM, McDonald JC. Long-term patency of venous repairs demonstrated by venography./Trawmfl 1985; 25: 342-346. 88 Rich NM, Gomez ER, Coffey JA et al. Long-term follow-up of venous reconstruction following trauma. In: BerganJJ, YaoJST (eds). Venous disorders. Philadelphia, W.B. Saunders Co., 1991 : pp 471-481. 89 Wells JK, Hagino RT, Bargmann KM et al. Venous morbidity after superficial femoral-popliteal vein harvest. / Vase Surg 1999; 29: 282-291. 90 Keen RR, Meyer JP, Durham JR et al. Autogenous vein graft repair of injured extremity arteries: early and late results with 134 consecutive patients. / Vase Surg 1991; 13: 664-668. 91 Richardson JB Jr., Jurkovich GJ, Walker GT et al. A temporary arteriovenous shunt (Scribner) in the management of traumatic venous injuries of the lower extremity. J Trauma 1986; 26: 503-509. 92 Barcia PJ, Nelson TG, Whelan TJ Jr. Importance of venous occlusion in arterial repair failure: an experimental study. Ann SurglW; 175: 223-227. 93 Hobson RW 2nd, Howard EW, Wright CB et al. Hemodynamics of canine femoral venous ligation: significance of combined arterial and venous injuries. Surgery 1973; 74: 824-829. 94 Menzoian JO, Mendez MV. Penetrating arterial injuries in the extremities. In: Ernst CB, Stanley JC. Current therapy in vascular surgery, 4th ed. St. Louis, Mosby, 2001 : pp 604-605. 95 Radonic V, Baric D, Giunio L et al. War injuries of the femoral artery and vein: a report on 67 cases. Cardiovase Surg 1997; 5: 641-647.

23
246

24
ENDOVASCULAR TREATMENT OF BLUNT INJURY OF THE LIMBS
BO RISBERG, LARS LONN

Endovascular treatment of various vascular disorders has been established as part of clinical routine. Preferentially, endovascular techniques are used in elective cases with stenoses, fistulae, or aneurysms. In recent years, the technique has been applied more frequently in emergency cases. It has turned out to be lifesaving as well as an attractive alternative to large open exposures. To date, endovascular repair has been performed mostly for injuries of central vessels, and reports of its application to limb injuries are sparse [1-4].

2f
247

Arterial injury
Traumatic lesions of extremity arteries result mainly from penetrating injuries and motor vehicle accidents. Vascular blunt injuries to the extremities are rare. The literature contains only case reports often related to youth and sports [5,6]. Isolated intimal injuries with spontaneous healing have been reported after blunt trauma [7]. Blunt vascular lesions occur from external direct trauma to the vessels or from secondary trauma following, for example, fractures or joint dislocations. The symptoms usually develop as a consequence of the induced ischemia. Bleeding problems following blunt injuries are extremely rare. The vessel pathology after blunt trauma includes wall fractures, hematomas, intimal fractures, dissections, and thrombosis. There can be a wide range of structural

damage and symptoms and there is not necessarily a direct correlation between extent of the injury and symptoms. Angiography is the gold standard examination for patients with documented or suspected vascular limb injuries. It has been considered important to document the presence or absence of such injuries, since even small injuries have been thought to progress to occlusion. [8-11]. Among patients with blunt popliteal injury, angiography was considered unnecessary in those with normal neurovascular examination [12]. Standard management at most institutions, however, includes mandatory repair of detected lesions. The indication for intervention must be set by the clinical symptoms and not judged on the angiographic images, i.e., do not treat the images. It has been clearly demonstrated in previous studies that

VASCULAR

EMERGENCIES

post-traumatic changes as discovered on angiography can undergo spontaneous resolution [13-15].

Diagnosis
Significant clinical findings in extremity trauma include pulse deficit, bruit, expanding hematoma, a history of hypotension, or signs of bleeding and eventually neurologic deficit. The clinical examination must include pulse palpation and, in doubtful cases, ultrasound-based pressure measurement. Calculating the ankle-brachial index (ABI) may be helpful in lower extremity injuries. Diagnosis is established by duplex-ultrasound, standard arteriography, computed tomography angiography (CTA), or magnetic resonance angiography (MRA), isolated or in combination. The decision of which technique to use must be based on local organization and routine. In most departments, standard angiography is routine. It has the advantage that catheters and guide wires will already be in place if an endovascular procedure is to follow.

Patients with multitrauma are usually examined with a spiral CT. Multislice CTA in the evaluation of patients with suspected injuries in the arteries is routine [16]. Bleeding can often be identified on CT before hemodynamic instability develops, and this time window can be an opportunity for endovascular treatment. If nonsurgical therapy is the first option, CTA is unnecessary and it is better to move forward to angiography as soon as possible. Color-flow duplex-ultrasound mapping with a linear-array transducer is an option for detection of vascular injuries. The imaging can be carried out in the emergency room. The fourth method for diagnosis (Table: C), MRA, has not gained widespread use for acute situations due to logistic problems and availability of machines. The two latter techniques are advantageous in that neither involves the use of ionizing radiation or iodinated contrast material.

24
248

Facilities
Modern vascular centers are (or should be) equipped with facilities for a variety of urgent endovascular procedures. Blunt vascular limb injuries would seldom require access to embolization facilities. Large stent graft devices such as used for ruptured aortic aneurysms would never seem to be needed. The shelf in the intervention suite should contain various guide wires, catheters, stents, and various embolization materials. The situation may not be different from penetrating injuries, where also a variety of stent-grafts should be available.

Types of lesions
A
Rupture, wall fracture Focal narrowing Occlusion, thrombosis Intimal flap Intimal dissection Endothelial irregularities Intramural hematoma AV fistulas Pseudoaneurysm

Endovascular techniques
B
Stent Stent grafts

Imaging techniques
C
Angiography (DSA) Ultrasound

Occlusive materials such as: - coils - (detachable) balloons

Computed tomography angiography (DSA) Magnetic resonance angiography (MRA) Contrast enhanced MRA (CE-MRA)

Management
Arterial integrity is often threatened in patients with blunt trauma lesions of the extremities. Digital subtraction angiography (DSA) quickly confirms sites of hemorrhage and other lesions (Table: A). Despite the options available for endovascular interventions, acute bleeding is best managed by embolotherapy (Table: B).

-PVA
- gelfoam - thrombin, bucrylate Thrombolysis

ENDOVASCULAR TREATMENT
Vascular injuries must be recognized, diagnosed, and treated quickly and accurately at the emergency ward. Occasionally, suspected vascular trauma results in rapid deterioration of the clinical condition. A surgical approach to control bleeding is then necessary. In the future, with increasing availability of skills in vascular interventional techniques, endovascular treatment should be possible also in these severe cases and not only as a guide to surgical treatment. There is a controversy regarding whether and to what extent angiography should be carried out. As indicated above, the clinical symptoms must determine if angiography is going to be performed or not. In asymptomatic cases, many angiograms will be negative and conservative management with observation during 24 hours may be sufficient. DSA is therefore indicated when: 1 - vascular injury is suspected due to clinical symptoms 2 - multiple lesions are suspected 3 - location of injury is uncertain 4 - endovascular therapy is feasible 5 - abnormal pulse waves on doppler with decreased ankle-brachial index are at hand.

OF BLUNT INJURY OF THE LIMBS

and only in special occasions is the procedure continued with direct antegrade puncture. A flush DSA of the aortic arch in left anterior oblique position may be useful to detect any other underlying disease before selective DSA is performed.

Endovascular treatment options

Embolotherapy is a fundamental aspect of interventional radiology. Treatment of hemorrhage following trauma focuses mainly on embolization (Table: B). The word embolus means plug, and embolization is a method for plugging vessels in a controlled manner via the endovascular route. Embolic agents function by inducing direct mechanical obstruction of the vessel and provide a framework for thrombus formation. This results in occlusion, and certain agents also incite an inflammatory reaction. The material can be divided into permanent or temporary occlusive agents based on its properties. The level of destination from a large vessel down to capillary levels is determined by the size of the product. Treatment depends on appropriate planning and selection of material, where angiography is the prerequisite. The technique depends on the location of the lesion. The collateral network in extremities is rich and therefore it is important to embolize both proximal and distal to the injury site to prevent retrograde filling in certain cases. Intravascular administration of various agents and materials to induce occlusion is administrated percutaneously. Thus, ongoing hemorrhage will arrest or eliminate vascular supply in order to simplify if a surgical treatment is needed. Particulate material includes coils, gelatin sponge, and polyvinyl alcohol foam (PVA). Coils are made of steel fragments surrounded by a thrombogenic material (i.e., wool strands) or, as in the case of detachable coils, a positively charged electrode (i.e., platinum coils) promotes clot formation by attracting negatively charged cells. The concept of hemorrhage control by transluminal balloon catheter tamponade, i.e., detachable transcatheter balloons, is used primarily in the cerebrovascular regions. An additional role for balloons is the achievement of temporary hemostasis until surgical intervention is allowed in the more global arteries of the limbs.

Technique of angiography
Conventional DSA with selective catheterization, serial imaging, and at least two orthogonal planes should be obtained on every patient with blunt trauma. These oblique images routinely acquired will reduce the risk of missing lesions on single projection images. Superselective angiography is almost always needed to confirm diagnosis.

249

LOWER LIMB ANGIOGRAPHY

A routine contralateral femoral puncture (Seldinger technique) under local anesthesia is standard approach for lower extremity lesions. Brachial access can be used when femoral pulses are absent. A catheter is positioned in the distal aorta and a flush DSA used to visualize the global arteries in cases of lower limb traumas before selective DSA is performed.

UPPER LIMB ANGIOGRAPHY

Patients with upper limb symptoms should have the brachiocephalic or left subdavian artery and the more distal arm vessels investigated. For upper limb injuries, the femoral technique is likewise used

VASCULAR

EMERGENCIES

Gelfoam is a resorbable material used mainly to temporarily control bleeding because of its tendency for recanalization. A liquid tissue adhesive such as bucrylate has low viscosity and is a very aggressive embolizing agent causing necrotizing vasculitis. The material can be infused percutaneously to obliterate traumatic lesions but this requires skill and experience and certain precautions. It is mainly used in portal veins for obliteration before liver surgery. In cases of pseudoaneurysm formation with narrow necks, thrombin injection is an option, but it should be administered percutaneously directly into the sac [17,18]. There is a potential risk of overspill of thrombin into the native artery if administered through the angiography catheter. Combination of endoluminal stents and stent grafts can also be applied for minimal arterial repair. The concept is to bridge the defect in the vessel wall with endoluminal graft supported by a bare stent material, usually self expandable. The risk for device infection following stent graft repair of trauma cases is a concern but is not definite.

stents placed. Two patients had injuries to the abdominal aorta, two to the carotid artery, and two to the subclavian artery. The early and even longterm follow-up was satisfactory [23].

Own case
A 51-year-old man had his right upper arm caught in the sliding door of a van. At arrival to the hospital he had numbness with loss of sensibility on the medial part of the lower right arm and the hand was cooler. The motor function was not affected. Blood pressure on the left arm was 120 mmHg and on the right 60 mmHg. The radial artery was not palpable. Clear doppler insonation was achieved in the right radial and ulnar arteries. The patient was taken to angiography because of suspected intimal tear with occlusion. The angiography was performed by the femoral route. The right subclavian artery was catheterized. Contrast injection revealed an intimal tear in the brachial artery with a roll-up for a distance of 2 to 4 cm approximately 10 cm above the elbow causing complete occlusion (Fig. 1). The lesion was easily passed

Survey of the literature

24
250

Only few case reports deal with endovascular treatment of blunt injuries, and none with limb injuries. Endovascular repair of vascular injuries to the extremities as reported in the literature dominates by penetrating injuries [3]. Still, open repair is the standard of care at most large institutions. In a recent report on lower extremity injury on 550 patients during a 10-year period, therapy was based on open surgical reconstruction [19]. In that study, penetrating injuries dominated and 20% were blunt trauma. Open repair is the gold standard for blunt injuries to the extremities. This is particularly true for popliteal injuries. Shortcomings in diagnosis and therapy may add to subopdmal results [20,21]. Our survey of the literature has not revealed any report on endovascular repair of blunt vascular injuries to the extremities. One single case from our own institution is presented below. The first report of stent treatment of blunt arterial injury was by Althaus et al. in 1996. They reported successful stent reconstruction of an iliac lesion [22]. In a report by Brandt et al. in 2001, six patients over a 6-year period had endovascular repair following blunt injuries. Each patient had one or more

FIG. 1 Selective digital subtraction angiography (DSA) of the right brachial artery: a 2 cm occlusion resulting from an intimal tear is shown (corresponding to the blunt trauma region).

TTTTTT
by a guide wire and percutaneous transluminal angioplasty (PTA) catheter. The defect was treated by balloon occlusion for 3 x 2 minutes to seal the intima to the vessel wall. After the first ballooning, a defect was still noticed (Fig. 2). A second ballooning almost completely restored vessel integrity. A small defect was still seen on contrast injection, but there was free and unhindered blood flow across this area (Fig. 3). The patient was given glycerolnitrat (0.2 mg) four times during the procedure. Since there was a clear clinical improvement, no stenting was performed. The introducer was left in place for a possible re-intervention. Local heparin infusion through the introducer as well as low molecular heparin subcutaneously was given. The patient was observed over night with adequate heparin administration. The clinical picture was normalized on the following day, with palpable radial pulse. Blood pressure was identical bilaterally. A duplex-doppler examination was performed and demonstrated that the brachial artery was open but a small defect (1 mm) was still seen in the area of injury. One-month follow-up with duplex-doppler demonstrated totally normalized conditions. There were no remaining clinical symptoms.

Discussion
Endovascular techniques have not been used commonly for blunt extremity trauma as reported in the literature. In fact, we have not encountered any published case. Since blunt injuries are not rare, this is a little bit surprising. The applicability of endovascular techniques would seem to be similar for both penetrating and blunt injuries. By large, many vascular injuries are still treated by open surgery, but endovascular techniques seem to be taking over. This is likely to occur also with blunt injuries. Blunt trauma to large vessels, as denoted above, has turned out to be a more or less established indication for endovascular therapy. Endovascular treatment of extremity vessels such as brachial and femoral arteries is next to come. The dismal results from treatment of atherosclerotic patients with long PTA or stents in the femoral/popliteal/distal arteries may, however, dampen the enthusiasm. The situation may be different in the traumatized patient. These patients most certainly belong to a much younger age group, in which immediate and long-term results

24
251

FIG. 2 After prolonged "gluing" of the dissection into the arterial wall by a 6 mm balloon occlusion, the artery is opened up. A substantial intimal lesion is visualized.

FIG. 3 Final result: restored vessel integrity with minimal residual defect,

VASCULAR

EMERGENCIES

may be better. There may still be concern about introducing foreign material into young people with long life expectancies.

Conclusion
Endovascular techniques are being increasingly utilized for emergency vascular cases. There is no doubt that the immediate success rate is high. Since

this is a novel technology, long-term data are lacking. Most experience hitherto is from penetrating injuries, but as more experience is gained more blunt injuries will be treated. Of particular concern is the applicability of the technique in growing individuals and people with a long life expectancy. This is important since the trauma population differs vastly in age distribution from the aging atherosclerotic patients in whom endovascular technology is mostly being applied today.

R E F E R E N C E S

JA & ' "059

1 Brunkwall J, Lindblad B, Ivancev K et al. latrogenic AV-fistula treated by a graft-covered self-expandable stent. Eur J Vase EndovascSurgl9%; 12: 243-245. 2 Chuter TA, Ivancev K, Lindblad B et al. Endovascular stent-graft exclusion of an aortobronchial fistula./ Vase Interv Radiol 1996; 7:357-359. 3 Marin ML, Veith FJ, Panetta TF et al. Transluminally placed endovascular stented graft repair for arterial trauma. JVasc Surg 1994; 20: 466-473. 4 Scharrer-Pamler R, Gorich J, Orend KH et al. Emergent endoluminal repair of delayed abdominal aortic rupture after blunt trauma. JEndovasc Swrg-1998; 5:134-137. 5 Sarfati MR, Gait SW, Treiman GS, Kraiss LW. Common femoral artery injury secondary to bicycle handlebar trauma./ Vase Surg 2002; 35: 589-591. 6 Sotta RP. Vascular problems in the proximal upper extremity. Clin Sports Med 1990; 9: 379-388. 7 Kestenberg WL. Review of intimal arterial injuries. Surgery versus conservative management. Am Swrgl990; 56: 504-506. 8 Geuder JW, Hobson RW 2nd, Padberg FT Jr et al. The role of contrast arteriography in suspected arterial injuries of the extremities. Am Swig 1985; 51: 89-93. 9 Kelly GL, Eiseman B. Civilian vascular injuries. / Trauma 1975; 15:507-514. 10 King TA, Perse JA, Marmen C, Darvin HI. Utility of arteriography in penetrating extremity injuries. Am J Surg 1991; 162:163-165. 11 Perry MO, Thai ER, Shires GT. Management of arterial injuries. Ann Surg 1971; 173: 403-408. 12 Abou-Sayed H, Berger DL. Blunt lower-extremity trauma and popliteal injuries: revisiting the case for selective arteriography. Arch Surg 2002; 137: 585 -589. 13 Dennis JW, Frykberg ER, Veldenz HC et al. Validation of nonoperative management of occult vascular injuries and

accuracy of physical examination alone in penetrating extemity trauma: 5 to 10 year follow-up./ Trauma 1998; 44: 243-253. 14 Frykberg ER, Crump JM, Dennis JW et al. Nonoperative observation of clinically occult arterial injuries: a prospective evaluation. Surgery 1991; 109: 85-96. 15 Frykberg ER, Vines FS, Alexander RH. The natural history of clinically occult arterial injuries: a prospective evalutation. /Trauma 1989; 29: 577-583. 16 Soto JA, Munera F, Cardoso N et al. Diagnostic performance of helical CT angiography in trauma to large arteries of the extremities. J Computer Assisted Tomography 1999; 23:188-196. 17 Lonn L, Olmarker A, Geterud K et al. Treatment of femoral pseudoaneurysms. Percutaneous US-guided thrombin injection versus US-guided compression. Ada Radiol 2002; 43: 396-400. 18 Kang SS, Labropoulos N, Mansour MA, Baker WH. Percutaneous ultrasound guided thrombin injection: a new method for treating postcatherization femoral pseudoaneuryms. / Vase Surg 1998; 27:1032-1038. 19 Hafez HM, Woolgar J, Robbs JV. Lower extremity arterial injury: results of 550 cases and review of risk factors associated with limb loss./ Vase Surg2001; 33:1212-1219. 20 Gupta R, Quinn P, Rao S, Sleunarine K Popliteal artery trauma. A critical appraisal of an uncommon injury. Injury 2001; 32: 357-361. 21 Kwolek CJ, Sundaram S, Schwarcz TH et al. Popliteal artery thrombosis associated with trampoline injuries and anterior knee dislocations in children. Am Swrgl998; 64:1183-1187. 22 Althaus SJ, Keskey TS, Harker CP, Coldwell DM. Percutaneous placement of self-expanding stent for acute traumatic arterial injury./Trawmfl 1996; 41: 145-148. 23 Brandt MM, Kazanjian S, Wahl WL. The utility of endovascular stents in the treatment of blunt arterial injuries./ Trauma 2001; 51:901-905.

25
RARE CAUSES OF ACUTE ISCHEMIA OF THE LIMBS
MARK KOELEMAY, DINK LEGEMATE

Acute upper or lower limb ischemia is an urgent condition that requires prompt restoration of blood flow to save the threatened limb. Acute limb ischemia most frequently occurs unilaterally, but patients may also present with bilateral symptoms. With regard to the etiology, a distinction can be made between arterial embolism, arterial thrombosis, and dissection. It is estimated that some 60% of acute upper extremity ischemia and 70% percent of acute lower extremity ischemia is caused by arterial embolism. Some 80% to 90 % of peripheral arterial embolisms are of cardiac origin., such as arrhythmias, cardiac aneurysm, and intracardial thrombus formation after myocardial infarction, dilatory cardiomyopathy, cardiac valve vegetations as in endocarditis or other (innate) cardiac valve anomalies, and atrial myxoma. Acute ischemia due to arterial thrombosis is the result of low blood flow, increased blood viscosity, or a procoagulative state (Virchow's triad) superimposed on preexistent atherosclerotic changes in the vascular endothelium. This chapter describes some rare causes of acute limb ischemia, which are listed in the Table. Data were collected by a systematic literature search of the PUBMED database using keywords arterial, embolism, and thrombosis with references to related articles, and using cross-references to complete the search. It was definitely not our aim to present a complete overview of all reported cases on a specific cause of acute limb ischemia.

^ Zi5 253

Anatomic CARDIAC While the most common cardiac disorders are noted briefly in the introduction, a few rare caus-

es of peripheral arterial embolism of cardiac origin must be mentioned. Young or middle-aged patients presenting with acute limb ischemia and concomitant venous thrombosis (VTE) or pulmonary embolism (PE) should raise the suspicion

VASCULAR

EMERGENCIES

Anatomic Cardiac Paradoxical arterial embolism Atrial myxoma Metastatic pulmonary tumors Metastatic germ cell tumor Peripartum cardiomyopathy Aorta Aortic dissection Aortic (aneurysm) thrombosis Mural aortic thrombus Primary malignancy of the aorta Infectious aortitis Peripheral arteries Popliteal entrapment Primary peripheral artery dissection Osteochondroma Persistent sciatic artery External compression Drugs Cocaine Human chorionic gonadotropin Anabolic steroids Heparin Chemotherapy Thrombin injection Prothrombotic state Malignancy related (Trousseau's syndrome) Anticardiolipin antibodies Lupus anticoagulant antibodies Hyperhomocysteinemia Nephrotic syndrome Inflammatory bowel disease Vasculitis Takayasu's disease Giant cell arteritis Cogan syndrome Kawasaki disease Trauma Blunt abdominal trauma Bullet embolism latrogenic Orthopedic surgery Surgery in high lithotomy position Aortic cannula embolism Catheterization Puncture sealing devices Intra-aortic balloon pump Aortic bifurcation endoprosthesis

25
254

of paradoxical embolism. This is caused by the passage of a right-sided venous or cardiac thrombus into the arterial circulation by a patent foramen ovale (PFO) or another intracardiac defect, generally at the atrial level. Autopsy studies have revealed that up to 35% of the population has a PFO, which has been suggested to be a risk factor for ischemic stroke. Little is known about the incidence of acute limb ischemia in patients with a PFO. Travis et al. found paradoxical embolism to be the cause of acute limb ischemia in 13 of 2764 (0.4%) patients presenting in a ten-year period [1]. The definitive diagnosis is established by the presence of VTE or PE, arterial emboli, and a right-to-left shunt. Transesophageal echocardiography is the investigation of choice if transthoracic echocardiography does not identify an intracardiac thrombus or PFO. Treatment modalities for the prevention of further emboli vary and include observation, antiplatelet therapy, anticoagulants, placement of vena cava filters, and closure of the PFO by percutaneous or operative procedures [1,2]. Other sources of cardiac embolism include neoplasms such as atrial myxoma or primary pulmonary malignancies invading the left atrium or the pulmonary veins. Occasionally, other malignancies cause tumor emboli, as in a young patient with bilateral acute leg ischemia in whom a germ cell tumor invaded the right lung and left atrium [3]; this patient finally died of massive pulmonary embolus after repeated peripheral embolectomies and tumor resection. Dilatory cardiomyopathy is a known risk factor for peripheral embolism. It is estimated that one of every 3000 to 15000 pregnancies is complicated by peripartum cardiomyopathy, with symptoms of dyspnea, peripheral edema, and fatigue. Whereas such patients are at increased risk of venous thromboembolism, we found one patient with popliteal and tibial artery embolism that resolved during intravenous heparin administration [4].

AORTA
Aortic dissection and aortic thrombosis are wellknown causes of acute extremity ischemia. In addition, peripheral emboli can originate from a diseased aorta. The prevalence of nonaneurysmal mural aortic thrombus has been reported to be 0.45% in a series of 10671 consecutive autopsies [5]. Eight of these patients had clinical signs of distal embolism. To date, 80 cases of distal embolization from mural aortic thrombus have been

RARE CAUSES OF ACUTE ISCHEMIA OF THE LIMBS documented in the literature [6]. One explanation for mural thrombus formation is a hypercoagulative state, whereas local aortic inflammation due to pancreatitis, and steroid therapy for Crohn's disease have also been postulated to cause aortic thrombi [7]. Depending on local availability, transesophageal echocardiography, magnetic resonance angiography (MRA), or computed tomography scanning can be used to establish the diagnosis. The optimal treatment modality for mural aortic thrombus has yet to be established. In the literature several management strategies have been suggested including anticoagulation therapy, aortic thrombo-endarterectomy, and graft replacement of the diseased aortic segment [6-8]. Primary malignancies of the aorta are extremely rare. From 1873 to 1997, only 87 cases have been reported in the literature [9]. In two thirds of these cases, the diagnosis was made post mortem, whereas the first manifestation in the other patients most often was peripheral tumor embolism. The male: female ratio was 2:1, the mean age of the patients 60 years. Angiosarcoma, sarcoma, leiomyosarcoma, and malignant fibrous histiocytoma were the most frequently found types of aortic tumors. If pathologic examination of an embolus reveals an aortic malignancy, MRA of the aorta can be helpful to determine whether the tumor can be resected. The prognosis, however, remains poor, because some 70% of the patients already have distant metastases. Bacteremia can cause infections of the normal nonaneurysmal aorta. In the pre-antibiotic era, cardiac valve vegetations frequently were the cause of infectious aortitis, but now the role of endocarditis has been diminished. At present, patients with a compromised immune system who are 50 to 70 years old and have an aortic aneurysm are at risk for infectious aortitis, typically caused by Staphylococcus aureus, Salmonella, or fungal infections. The most serious complication of infectious aortitis is rupture of the aorta, but infectious aortitis should also be kept in mind as a cause for peripheral embolization. We found case reports on lower limb embolization due to infections with S. Pneumoniae and Paracoccidiodes brasiliensis [10,11]. PERIPHERAL ARTERIES An abnormal anatomical relation between the popliteal artery and the medial head of the gastrocnemius muscle causes medial displacement and compression of the artery, known as popliteal entrapment [12]. It is a rare condition, predominantly found in young men, who may complain of claudication, paresthesia, or numbness of the foot. A rare presentation of popliteal entrapment is embolization from the diseased popliteal artery to the foot arch and digital vessels [13]. Several classifications for popliteal entrapment have been suggested. In the most common type, the popliteal artery is compressed between the medial head of the gastrocnemius muscle and the medial femur condyle. In the other variants, a slip of gastrocnemius muscle, a fibrous band, or the popliteus muscle may cause constriction of the artery [12]. Finally, physiologic popliteal entrapment may exist due to gastrocnemius, soleus, or plantaris muscle hypertrophy in athletes [12,14]. The definitive management depends on the local situation and may consist of simple release of the popliteal artery by division of the medial head of the gastrocnemius muscle or a short venous bypass graft in case of an occlusion. Primary dissection of the peripheral arteries is a rare condition that should be considered in young patients with acute onset limb ischemia. In a literature review published in 1969, 110 case reports were identified [15]. Only nine cases concerned the lower extremity arteries. More recently, additional dissections of the femoral and popliteal arteries have been described, with symptoms ranging from claudication to acute limb ischemia. Primary dissection occurs predominantly in patients under 50 years of age, more often in men, who are hypertensive in 90% of the cases. Underlying diseases associated with spontaneous dissection include Marfan syndrome, cystic media degeneration, and fibromuscular dysplasia. Osteochondroma are benign tumors of the metaphysis of long bones that can displace adjacent blood vessels such as the superficial femoral and popliteal artery. The majority of patients are younger than 25 years and have no manifestations of atherosclerotic disease [16,17]. The cartilaginous part of the tumor matures into bone, which continuously damages the artery. This may finally result in pseudoaneurysm formation, arterial thrombosis, luminal stenosis due to extrinsic compression, or arteriovenous fistula. Some authors therefore advise prophylactic resection of osteochondroma in the vicinity of a vessel. A similar mechanism caused embolism of the lower limb in a patient with an osteosynthesis screw that was used to repair a comminuted femoral fracture. The screw and surrounding callus caused superficial femoral artery occlusion and distal

255

VASCULAR

EMERGENCIES

25 256

embolization that did not respond to thrombolytic therapy and resolved only after thrombo-endarterectomy and removal of the screw [18]. The sciatic artery is the earliest fetal artery to supply blood to the lower extremity by connecting the internal iliac and popliteal artery. The sciatic artery involutes with the development of the superficial femoral artery. Its proximal part becomes the superior gluteal artery, whereas the distal part forms a plexus on the surface of the great adductor muscle. Based on angiographic studies, the prevalence of a persistent sciatic artery is estimated to be between 0.025% and 0.04% [19]. Like other arteries, the sciatic artery is prone to atherosclerosis and aneurysmal disease in particular, which may present as a pulsatile mass in the gluteal region, as chronic lower limb ischemia, or as acute ischemia due to distal embolization from a sciatic artery aneurysm. Management depends on the nature of the lesion and comprises surgery in most cases. In a recently published report, two patients were successfully treated with placement of a covered stent for exclusion of an aneurysm, and dilation and stent placement for claudication [20]. Finally, external compression of peripheral arteries may lead to acute extremity ischemia, such as in a patient with a proven occlusion by duplex scanning of the right common iliac artery due to massive fecal impaction of the sigmoid colon and rectum; the symptoms fully subsided after removal of the fecaloma [21].

Drugs
Several drugs have been reported to be associated with acute limb ischemia. Cocaine is a potent vasoconstrictor and is known to cause myocardial infarction. The use of cocaine, whether taken intranasally or smoked as crack, has also been reported to cause aortic, common iliac, and popliteal artery thrombosis [22,23]. It is hypothesized that cocaine induces platelet aggregation and that it may augment procoagulants by decreasing protein C and antithrombin III (AT-III) levels, resulting in arterial thrombosis. An additional risk of drug abuse is distal embolization caused by accidental intra-arterial drug injection. In a review, Stewart et al. [24] addressed the risk of thrombo-embolic events after ovarian stimulation, in vitro fertilization and gamete intrafallopian trans-

fer, and the use of human chorionic gonadotrophin (HCG). The majority of events were deep VTE and PE, with 15 cases of arterial thrombosis of the carotid, cerebral, subclavian, and iliac arteries. Possible explanations for arterial thrombosis include decreased AT-III levels and increased fibrinogen levels induced by exogenous HCG or the presence of an unknown thrombophilia in combination with HCG. Danazol is an androgenic steroid that is used primarily for endometriosis and has both prothrombotic and antithrombotic effects. It influences the clotting cascade by increasing platelet counts and elevating both anticoagulant and procoagulant protein levels. Alvarado et al. described two patients using danazol, for endometriosis and idiopathic thrombocytopenic purpura, respectively, with acute lower extremity arterial thrombosis and who had no risk factors for arterial disease [25]. Both patients were treated successfully with thrombolytic therapy and withdrawal from danazol. Other anabolic steroids such as testosterone, nandrolone, sustanon, dianabol, and testovarin have also been reported to cause diffuse arterial thrombosis or emboli from left-sided intraventricular cardiac thrombus [26-28], Suggested mechanisms for arterial thrombosis include platelet activation, increased levels of procoagulant factors and a decrease in fibrinolytic activity. One of the side effects of unfractionated heparin administration is the occurrence of heparin-induced thrombocytopenia (HIT) in 2% of patients. The diagnosis of HIT is established by a decrease in platelet count by 50 000/mm3 or an absolute decrease below 100 000/mm3 and a positive heparin-dependent platelet antibody test. It is estimated that 30% to 75% of patients with HIT develop paradoxical venous or arterial thrombosis (heparin-induced thrombocytopenic thrombotic syndrome [HITTS]) or white clot disease [29]. The mechanism of clot formation is not fully understood, and it is believed to be the result of immune activated platelets combining with fibrin. A loss of local endothelial integrity and heparin-induced immune complex interactions may cause thrombosis. A few case reports describe arterial thrombosis of the lower extremity arteries. The optimal treatment of HITTS has not yet been established and comprises, apart from heparin withdrawal, warfarin, dextran, aspirin, low molecular weight heparin, prostacyclin analogues or thrombolytic therapy with urokinase or streptokinase. The administration of chemotherapeutic agents such as etoposide, cisplatin, vinblastine, methotrex-

RARE CAUSES OF ACUTE ISCHEMIA

ate and bleomycin for treatment of pulmonary and germ cell cancer has been reported to cause peripheral arterial thrombosis [30-32]. In addition, premenopausal patients receiving adjuvant therapy for breast cancer were at a 1.6% risk of arterial thrombosis when tamoxifen was added to chemotherapy compared to 0% in patients receiving chemotherapy alone [33]. The reason for increased risk of both venous and arterial thrombosis is not clear. As mentioned in the next section, the presence of a malignant tumor carries an increased risk of arterial thrombosis due to a hypercoagulable state. Suggested mechanisms for chemotherapy-induced thrombosis include a decrease in proteins C and S and AT-III levels. Thrombin injection is an effective and safe treatment of pseudoaneurysms, e.g., of the femoral artery, and is associated with a low complication rate [34]. However, arterial thrombosis can result after accidental injection of thrombin in the femoral artery [35].

OF THE LIMBS

Prothrombotic state
Deficiencies in proteins C and S, AT-III and factor VIII, and factor V Leiden mutation are thrombophilia that are well known risk factors for spontaneous venous and arterial thrombosis. A less common cause of arterial thrombosis is malignancy-related hypercoagulability, also known as Trousseau's syndrome [36]. It has been described predominantly in association with malignancies of the lung, pancreas, and intestinal tract, but may be caused by all cancers. The exact etiology of malignancy-related arterial thrombosis is unclear. There is evidence that tumors can produce procoagulant factors, cause deficiencies of proteins C and S and AT-III, and induce platelet activation. The treatment of arterial thrombosis in such cases is not different from usual. Some authors advocate prompt treatment of the underlying malignancy, but this is not supported by scientific evidence. A special group of malignancies are the acute leukemias that cause thrombosis by means of leukemic cell sedimentation or disseminated intravascular coagulation [37]. Antiphospholipid antibodies are autoantibodies that may cause procoagulant states. The most commonly detected subgroups are anticardiolipin antibodies, lupus anticoagulant antibodies, and anti-B2glycoprotein-1 antibodies. The prevalence of these

antibodies ranges between 1% and 5% in a young healthy population and increases with age. Most patients with antibodies will never suffer from a thrombo-embolic event. However, 50% to 70% of patients with systemic lupus erythematosus and antiphospholipid antibodies may develop the antiphospholipid syndrome, which is characterized by venous or arterial thrombosis. The majority of thromboses are venous (55% of cases). Arterial thrombosis affects the brain (50%), the coronary arteries (27%), and the peripheral arteries (23%). The increased incidence of mitral valve vegetations in these patients causes arterial embolism. The best prophylaxis and treatment seems to be warfarin medication, probably lifelong, as withdrawal from warfarin leads to a high recurrence of thromboembolic events [38]. Hyperhomocysteinemia is an independent, moderate risk factor of arterial occlusive disease. Although the relation between hyperhomocysteinemia and venous and arterial thrombosis is less clear, it must also be kept in mind as a source of peripheral arterial thrombosis, as described in a patient with femoral artery thrombus causing acute ischemia of the lower limb; correction of blood homocysteine with vitamin suppletion may prevent further arterial sclerosis and thrombosis. The nephrotic syndrome has its highest prevalence in young patients and has been reported to cause arterial thrombosis, mainly of the superficial femoral arteries. Although rare, there are a few reports on arterial thrombosis in adult patients, located in the cerebral, mesenteric, renal, and aortic and femoral arteries [40,41]. Arterial thrombosis in the nephrotic syndrome is thought to be a result of decreased AT-III levels, platelet dysfunction, alteration in levels of clotting factors, increased blood viscosity, hypoalbuminemia, and administration of steroids. Inflammatory bowel disease is also associated with thrombo-embolic events with an estimated prevalence between 1.3% and 6.4%, with two thirds being venous thromboses. Patients with Crohn's disease and those with ulcerative colitis are at risk of developing aortic or lower extremity arterial thrombosis [42,43]. The origin of arterial thrombosis in inflammatory bowel disease is as yet unclear because of its low prevalence. It is hypothesized that a primary AT-III deficit causes the hypercoagulable state and subsequent thrombosis. Among other explanations are increased platelet activation, a decrease in tissue plasminogen activator and elevated anticardiolipin

257

VASCULAR

EMERGENCIES

antibodies. There is no uniform therapy; the patients described in several case reports and in a survey among members of the American Society of Colon and Rectal Surgeons were all treated with surgical thrombo-embolectomy.

Trauma
The association between vascular injuries and fractures or luxations is well known. Other less common causes of vascular injuries must not be overlooked. Blunt abdominal trauma by seat belt injury with subadventitial rupture of the common or external iliac arteries can result in acute leg ischemia [51,52]. It is assumed that shear forces cause a disruption of the intima and subsequent thrombosis. Blunt abdominal trauma can also dislodge thrombus from an abdominal aortic aneurysm. One patient known to have an abdominal aortic aneurysm developed an acute ischemic leg from a seat belt injury [53], while four other patients developed bilateral leg ischemia after the Heimlich maneuver, as a result of aortic aneurysm thrombosis and thrombo-embolism to both legs [54,55]. Bullet fragments must also be considered as sources of arterial embolisms because of the increasing possession and use of guns. Many papers have reported delayed onset of peripheral arterial ischemia due to late migration of bullet fragments from large caliber arteries to peripheral arteries, including paradoxical bullet embolism in a patient withaPFO [56,57].

Vasculitis
Takayasu's disease is a nonspecific inflammatory process of the aorta and its major branches and the pulmonary arteries that causes arterial stenosis or aneurysm formation. The origin of the inflammation that causes changes in all vascular wall layers is unknown. While the majority of symptoms are caused by alterations in the thoracic aorta and subclavian arteries, patients have also presented with acute leg ischemia due to emboli originating from the infrarenal aorta [44] and aorto-iliac bifurcation [45]. Therapy comprised thrombolysis in the former case and aorto-iliac bifurcation prosthesis in the latter, in combination with steroids. Giant cell arteritis is another nonspecific vasculitis of the larger arteries that may cause acute upper or lower extremity ischemia [46,47]. Like in Takayasu's disease, steroid therapy is the treatment of choice. Cogan syndrome is a rare condition, characterized by nonsyphilitic interstitial keratitis and vestibulo-auditory symptoms which progress to deafness. In 10% of these patients, vasculitis develops that may manifest itself in occlusion or aneurysmal dilatation of the coronary arteries, mesenteric ischemia or neurologic symptoms. One case was described with acute onset calf claudication due to a thrombus in the popliteal artery that was successfully treated with thrombolytic therapy [48]. Kawasaki disease is a mucocutaneous lymph node syndrome characterized by high, persistent fever, prominence of conjunctival vessels, changes in the oropharyngeal mucous membranes, rash and lymphadenopathy. It occurs most frequently in children under four years, but it has been described in adults too. A week after the initial symptoms, vasculitis develops and aneurysms start forming, predominantly in the coronary arteries and other, similar locations as in Cogan syndrome. One report describes a middle-aged patient with a history of Kawasaki disease who presented with acute ischemia as a result of a thrombosed popliteal aneurysm [50].

25
258

latrogenic
Patients undergoing orthopedic surgery are at risk for acute peripheral ischemia. Total knee arthroplasty is associated with a 0.03% to 0.17% risk of vascular complications such as thrombosis, popliteal artery transection and aneurysm formation [58,59]. It is unclear whether the application of a tourniquet increases the risk factor for arterial thrombosis. Additional case reports were found on iliac artery occlusion and aortic thrombosis after retroperitoneal exposure of the spine. Recognition of this complication could prove difficult because sensorimotor disturbances may also be caused by transient nerve damage. In addition, total hip replacement has been noted to cause bilateral thrombosis of the iliac arteries. The ankle-brachial index decreases significantly by placing patients in the high lithotomy position [60]. Geeraerts et al. described a patient who developed acute external iliac thrombosis superimposed on preexisting atherosclerotic lesions in the

RARE CAUSES OF ACUTE ISCHEMIA OF THE LIMBS

lower extremity after transurethral bladder tumor resection [61]. Further embolic complications were related to femoral artery catheterization for diagnostic angiographies, percutaneous transluminal (coronary) angioplasties, sealing of the femoral artery punctures with closure devices [62] and embolism by a buffer of an aortic cannula that was used during coronary bypass surgery in a patient with acute leg pain [63]. The use of an intra-aortic balloon pump was also reported to cause acute aortic thrombosis. In the near future, patients can be expected to have acute leg ischemia due to thrombosis of one of the limbs of an endovascular device for abdominal aortic aneurysm repair [64].

R E F E R E N C E S 1 Travis JA, Fuller SB, Ligush J Jr. et al. Diagnosis and treatment of paradoxical embolus./ Vase Surg 2001; 34: 860-865. 2 AbuRahma AF, Downham L. The role of paradoxical arterial emboli of the extremities. AmJSurgl9%; 172: 214-217. 3 Singh A, Jenkins DP, Dahdal M et al. Recurrent arterial embolization from a metastatic germ cell tumor invading the left atrium. Ann Thome Surg 2000; 70: 2155-2156. 4 Carlson KM, Browning JE, Eggleston MK, Gherman RB. Peripartum cardiomyopathy presenting as lower extremity arterial thrombo-embolism. A case report. JReprod Med 2000; 45: 351-353. 5 Machleder HI, Takiff H, Lois JF, Holburt E. Aortic mural thrombus: an occult source of arterial thrombo-embolism. J Vase Swrgl986;4:473-478. 6 Rossi PJ, Desai TR, Skelly CL et al. Paravisceral aortic thrombus as a source of peripheral embolization. Report of three cases and review of the literature. / Vase Surg 2002; 36: 839 - 843. 7 Hahn TL, Dalsing MC, Sawchuk AP et al. Primary aortic mural thrombus: presentation and treatment. Ann Vase Surg 1999; 13: 52-59. 8 Bowdish ME, Weaver FA, Liebman HA et al. Anticoagulation is an effective treatment for aortic mural thrombi. J Vase Surg 2002; 36:713-719. 9 Seelig MH, Klingler PJ, Oldenburg WA, Blackshear JL. Angiosarcoma of the aorta: report of a case and review of the literature. / Vase Surg 1998; 28: 732 - 737. 10 Maclennan AC, Doyle DL, Sacks SL. Infectious aortitis due to penicillin-resistant Streptococcus pneumoniae. Ann Vase Surg 1997; 11: 533-535. 11 Cherri J, Freitas MA, Llorach-Velludo MA, Piccinato CE. Paracoccidioidomycotic aortitis with embolization to the lower limbs. Report of a case and review of the literature. J Cardiovasc Surg 1998; 39: 573 -576. 12 Lambert AW, Wilkins DC. Popliteal artery entrapment syndrome. BrJSurg 1999; 86:1365-1370. 13 Fong H, Downs AR. Popliteal entrapment syndrome with distal embolization. A report of two cases. / Cardiovasc Surg 1989; 30: 85-88. 14 Lepori L, Perren A, Gallino A. The popliteal artery entrapment syndrome in a patient using anabolic steroids. NEnglJMed 2002; 346:1254-1255. 15 Rabkin DG, Goldstein DJ, Flores RM, Benvenisty Al. Spontaneous popliteal artery dissection: a case report and review of the literature. J Vase Surg 1999; 29: 737-740. 16 Eschelman DJ, Gardiner GAJr., Deely DM. Osteochondroma: an unusual cause of vascular disease in young adults. / Vase Interv &Mfio/1995;4:605-613. 17 Vasseur MA, Fabre 0. Vascular complications of osteochondromas. / Vase Surg 2000; 31: 532 - 538. 18 Gronefeld G, Zegelmann M, Schulte-Huermann D, Landgraf H. An unusual cause of arterial vascular occlusion with peripheral embolism. VASA 1993; 22: 260-263. 19 Ikezawa T, Naiki K, Moriura S et al. Aneurysm of bilateral persistent sciatic arteries with ischemic complications: case report and review of the world literature. J Vase Surg 1994; 20: 96-103. 20 Gabelmann A, Kramer SC, Wisianowski C et al. Endovascular interventions on persistent sciatic arteries. JEndovasc 77^2001; 8:622-628. 21 Hoballah JJ, Chalmers RT, Sharp WJ et al. Fecal impaction as a cause of acute lower limb ischemia. AmJ Gastroenterol 1995; 90: 2055-2057. 22 Mirzayan R, Hanks SE, WTeaver FA. Cocaine-induced thrombosis of common iliac and popliteal arteries. Ann Vase Surg 1998; 12: 476-485. 23 Webber J, Kline RA, Lucas CE. Aortic thrombosis associated with cocaine use: report of two cases. Ann Vase Surg 1999; 13: 302304. 24 Stewart JA, Hamilton PJ, Murdoch AP. Thrombo-embolic disease associated with ovarian stimulation and assisted conception techniques. Hum Reprod 1997; 12: 2167-2173. 25 Alvarado RG, Liu JY, Zwolak RM. Danazol and limb-threatening arterial thrombosis: two case reports. J Vase Surg 2001; 34:11231126. 26 Falkenberg M, Karlsson J, Ortenwall P. Peripheral arterial thrombosis in two young men using anabolic steroids. EurJ Vase Endovasc Surgml; 13: 223-226. 27 McCarthy K, Tang AT, Dalrymple-Hay MJ, Haw MP. Ventricular thrombosis and systemic embolism in bodybuilders: etiology and management. Ann Thorac Swrg2000; 70: 658-660. 28 Nieminen MS, Ramo MP, Viitasalo M et al. Serious cardiovascular side effects of large doses of anabolic steroids in weight lifters. Eur Heart Jim; 17:1576-1583. 29 Murphy KD, McCrohan G, DeMarta DA et al. The heparininduced thrombocytopenia and thrombosis syndrome: treatment with intra-arterial urokinase and systemic platelet aggregation inhibitors. CardiovascInterventRadiol 1996; 19:123-127. 30 Cool RM, Herrington JD, W7ong L. Recurrent peripheral arterial thrombosis induced by cisplatin and etoposide. Pharmacotherapy 2002; 22:1200-1204.' 31 Molloy RG, Welch GC, Drury JK, Abel BJ. Arterial thrombosis after chemotherapy with cisplatin, vinblastine and methotrexate. BrJ Clin Pract 1995; 49: 50 - 51. 32 Vos AH, Splinter TA, van der Heul C. Arterial occlusive events during chemotherapy for germ cell cancer. NethJMed 2001; 59: 295-299. 33 Saphner T, Tormey DC, Gray R. Venous and arterial thrombosis in patients who received adjuvant chemotherapy for breast cancer./Qm Onco/1991; 9: 286-294. 34 Friedman SG, Pellerito JS, Scher L et al. Ultrasound-guided thrombin injection is the treatment of choice for femoral artery pseudonaeurysms. Arch Swrg-2002; 137: 462-464. 35 Sadiq S, Ibrahim W. Thrombo-embolism complicating thrombin injection of femoral artery pseudoaneurysm: management with intra-arterial thrombolysis./Fast Infers fladio/2001; 12: 633-636.

259

VASCULAR

EMERGENCIES

25 260

36 Rigdon EE. Trousseau's syndrome and acute arterial thrombosis. Cardiovasc Surg WOO; 8: 214-218. 37 DiGiovanni RJ, Crilley P, Kerstein MD. Peripheral arterial occlusion in acute promyelocytic leukemia. Cardiovasc Surg 1999;7:258-260. 38 Levine JS, Branch DW, Rauch J. The antiphospholipid syndrome. NEnglJMed20Q2; 346: 752-763. 39 Tan YM, Chia KH. Acute thrombosis associated with hyperhomocysteinemia. EurJ Vase EndovascSurg 2002; 24: 279-280. 40 Nakamura M, Ohnishi T, Okamoto S et al. Abdominal aortic thrombosis in a patient with nephrotic syndrome. AmJNephrol 1998; 18: 64-66. 41 Lee CH, Chen KS, Tsai FC et al. Concurrent thrombosis of cerebral and femoral arteries in a patient with nephrotic syndrome. AmJNephrol 2000; 20: 483-486. 42 Szychta P, Reix T, Sevestre MA et al. Aortic thrombosis and ulcerative colitis. Ann Vase Surg2001; 15: 402-404. 43 Novotny DA, Rubin RJ, Slezak FA, Porter JA. Arterial thromboembolic complications of inflammatory bowel disease. Report of three cases. Dis Colon Rectum 1992; 35:193-196. 44 Buchner N, Sanner B, Tepel M et al. Acute ischemia of the legs and rapidly progressing renal failure in a 39-year-old patient. /7iferoirfl999;40:555-560. 45 Pistorius MA, Jego P, Sagan C et al. Arterial embolic manifestations in the legs revealing isolated aorto-iliac Takayasu's disease. JMd Vase 1993; 18: 331-335. 46 Rivers SP, Baur GM, Inahara T, Porter JM. Arm ischemia secondary to giant cell arteritis. AmJSurg 1982; 143: 554-558. 47 Greene GM, Lain D, Sherwin RM et al. Giant cell arteritis of the legs. Clinical isolation of severe disease with gangrene and amputations. AmJMed 1986; 81: 727-733. 48 Amatucci G, Del Mastro G, landoli R. Horton giant cell arteritis of the legs. Report of a case. / Cardiovasc Surg 1997; 38: 309 - 312. 49 Bastug DE, Dominic A, Ortiz 0 et al. Popliteal artery thrombosis in a patient with Cogan syndrome: treatment with thrombolysis and percutaneous transluminal angioplasty. Cardiovasc Interv /tofcon997;20:57-59. 50 Bradway MW, Drezner AD. Popliteal aneurysm presenting as acute thrombosis and ischemia in a middle-aged man with a history of Kawasaki's disease. / Vase Surg 1997; 26: 884 - 887.

51 Shindo S, Okamoto H, Nagai M et al. Acute ischemia of the lower legs from blunt abdominal trauma: an unusual cause of atheroembolism. Case report. / Trauma 1994; 36: 451-453. 52 Gupta N, Auer A, Troop B. Seat belt-related injury to the common iliac artery: case report and review of the literature. /Trauma 1998; 45: 419-421. 53 Jiao LR, Ramanathan A, Ackroyd J. An acute lower limb ischaemia with an unusual cause. EurJ Vase Endovasc Surg 1998; 15:547-549. 54 Mack L, Forbes TL, Harris KA. Acute aortic thrombosis following incorrect application of the Heimlich maneuver. Ann Vase Surg 2002; 16:130-133. 55 Ayerdi J, Gupta SK, Sampson LN, Deshmukh N. Acute abdominal aortic thrombosis following the Heimlich maneuver. Cardiovasc Surg2002: 10; 154-156. 56 Adegboyega PA, Sustento-Reodica N, Adesokan A. Arterial bullet embolism resulting in delayed vascular insufficiency: a rationale for mandatory extraction. JTrauma 1996; 41:539-541. 57 Schurr M, McCord S, Croce M. Paradoxical bullet embolism: case report and literature review. J Trauma 1996; 40:1034-1036. 58 Bellemans J, Stockx L, Peerlinck K et al. Arterial occlusion and thrombus aspiration after total knee arthroplasty. Clin Orthop 1999; 366:164-168. 59 Ohira T, Fujimoto T, Taniwaki K. Acute popliteal artery occlusion after total knee arthroplasty. Arch Orthop Trauma Surg 1997; 116: 429-430. 60 Canterbury TD, Wheeler WE, Scott-Conner CE. Effects of the lithotomy position on arterial blood flow in the lower extremities WVmjlWZ; 88:100-101. 61 Geeraerts T, Albaladejo P, Droupy S et al. Acute thrombosis of external iliac artery after short procedure in the high lithotomy position. Anaesthesiology 2000; 93:1353-1354. 62 Tomlinson MA, Beese R, Banwell M et al. Sequential retroperitoneal venous hemorrhage and embolism of an angio-seal puncture closure device complicating iliac artery angioplasty. J Endovasc Surg im; 3: 264-269. 63 Wijesinghe LD, Coughlin PA, Gill K. An unusual cause of femoral embolus. Cardiovasc Surg 2000; 8: 287-288. 64 Parent FN 3rd, Godziachvili V, Meier GH 3rd et al. Endograft limb occlusion and stenosis after ANCURE endovascular abdominal aneurysm repair. J Vase Surg 2002; 35: 686-690.

26
ACUTE ARTERIAL THROMBOSIS OF THE LOWER LIMBS
WILLIAM PAASKE

An arterial thrombus is a clot formed from the constituents of the blood, and arterial thrombosis signifies the formation, or the presence, of a thrombus within the artery indicating the dynamic processes leading to partial or complete cessation of blood flow through the vessel and the events thereafter to the static, functional endpoint of that progression. The term acute, in clinically detectable situations, relates to the immediate symptoms and/or signs becoming apparent as a consequence of the thrombosis; the transition from acute to chronic thrombosis is not always clear in the temporal domain, but for all practical purposes it would seem reasonable to restrict the term acute to clinical situations in which a sudden deterioration of the peripheral arterial circulation is anamnestically and/or clinically evident. The dynamic nature of the thrombotic process often extends beyond the time of immediate symptoms or signs; this is particularly evident in some but not all cases of thrombosis resulting from atherosclerotic occlusive disease. In other situations, the thrombus-generated symptoms are stable after they have become clinically recognizable, but the clinical course is often characterized with alleviation of symptoms and mitigation of signs in the early hours after presentation.

^ \) 261

Incidence OI thrombotic events The exact frequency of admissions for acute thrombotic events is not known; the Danish vascular surgical registry of procedures, Karbase, covers the entire activity of the country in the services of

vascular surgery. From 1996 to 2000, the frequency of admissions for acute ischemia was 11 patients per 100 000 inhabitants of all ages per year, but the reporting system does not differentiate between embolus and thrombosis, and the figure also ineludes events on the upper extremities. It is interesting to observe that the male/female ratio was

VASCULAR

EMERGENCIES

26
262

0.98 in this Danish material. The Swedish vascular registry, Swedvasc, includes 913 open procedures from acute ischemia in year 2001 (corresponding to 10 per 100 000 inhabitants of all ages per year, remarkably close to the Danish figure in spite of pronounced regional variations) with half of these procedures given as interventions for embolus, the other half for thrombosis. This distribution is surprising and must be considered problematic, with a gross overestimation of the embolic etiology. The percentage of procedures on the lower extremities was 81.7% in the series. Overall, in the Swedish year 2001 material, 10.3% of the patients admitted for acute ischemia died within 30 days (average 1995 to 2001 around 12% to 13%); 8.1% were amputated within 30 days, and only 58.2% of the patients had an uncomplicated course within the first month after admission followed by some procedure (open operation, balloon angioplasty). Of the cases recorded in year 2001 where the procedure was indicated as thrombectomy/embolectomy of the leg, 12.7% died within 30 days after the procedure (and 23.5% occluded). For bypass procedures, the 30day mortality was 15.3%, and the occlusion rate was 21.6%; the cases treated with balloon or catheter in the leg resulted in a 5.1% 30-day mortality and a 30-day occlusion frequency of 26.1%. Thrombolysis for lower limb ischemia was performed in 208 primary and supplementary cases in Denmark in the year 2 000; this corresponds to 3.6 lyses per 100 000 inhabitants per year (lysis of grafts is included in this overall figure). Common to the Danish and the Swedish materials was a massive presence of risk factors and comorbidities in these patients. In Denmark, 47.2% were noted in the national material from 1996 to 2000 as having cardiac disease, 35.2% hypertension, 25.1% pulmonary disease, and 17% previous cerebrovascular disease; 15.4% had diabetes mellitus, and 77.6% were frank enough to report that they were, or had been, smokers. Only 3.7% had been amputated previously. The purpose of this chapter is to give a review of the mechanisms, diagnosis, and treatment of acute arterial thrombosis in the native arteries of the lower extremities. Considerations on arterial emboli are discussed in another chapter of this book [1]. Thrombosis of grafts and of other previously performed arterial reconstructions have been discussed previously [2] and shall not be repeated here. Since an interesting clinical problem is the determination of an embolic or thrombotic cause of the disease

manifestations, some considerations will be included with reference to thrombo-embolism.

Mechanisms
ATHEROSCLEROTIC OCCLUSIVE ARTERIAL DISEASE The most frequent etiological factor for thrombus formation in the legs is atherosclerotic occlusive arterial disease where complex pathogenetic processes lead to changes of the morphology, the biochemistry, the physiology, and the biomechanics of the vessel wall together with local disturbances of the blood to vessel wall interface interactions, and generalized changes in the blood constituents [3]. It has been demonstrated that different atherosclerotic plaques contain varying proportions of the plaque components. From several studies, mainly on coronary and carotid arteries, it has become clear that the main plaque constituents are the connective tissue extracellular matrix (collagen, proteoglycans, fibronectin, elastic fibers), crystalline cholesterol, cholesteryl esters and phospholipids, and cells (monocyte-derived macrophages, T lymphocytes, smooth muscle cells) [4]. The precise nature of the specific details of the atherosclerotic processes in the lower extremity arteries is less well known, but there is evidence to suggest that the practical classification of the plaques based on their histological composition and structure as suggested by the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association [5] can be generalized and applied to lower extremity arteries as well. Although this classification has been criticized [6] and updated [7], the concept of the vulnerable arteriosclerotic plaque [8, 9] seems suitable for interpretation of infrainguinal arterial pathologies too, since the lesions now [7] denoted types VI (surface defect, hematoma, thrombosis), VII (calcification predominates [10]), and VIII (fibrous changes predominate) appear applicable for the purposes of peripheral vascular surgery. The term vulnerable implies vulnerability to plaque disruption with the probability of subsequent acute thrombosis. The question of stable versus unstable atherosclerosis has, for obvious reasons, attracted enormous attention for coronary and carotid arteries [11], and the notion of a high-risk or vulnerable atherosclerotic plaque with superpositioned thrombus formation [12] must be investigated in detail in the lower extremities to decide whether it will

ACUTE ARTERIAL THROMBOSIS OF THE LOWER LIMBS

be clinically meaningful to extend this interesting conceptual framework to the patients with vascular disease in the legs. However, there can be no doubt that plaque disruption is often the crucial, final common process that converts a nonocclusive, often clinically silent atherosclerotic lesion to a potentially limb-threatening, invalidating, or fatal condition [13]. The recent developments in clinical imaging, especially high-resolution multicontrast magnetic resonance (MR) technology and electron beam computed tomography (for assessment of calcification), should be applied for plaque imaging and characterization in extremity arteries as a next step to the investigations performed on the aorta, coronary, and carotid arteries [4]. The effect of early detection of peripheral lesions as markers of localized and possibly generalized atherosclerotic disease followed by risk reduction and modification together with targeted medical therapy, for the time being notably statins [14], is an apparent and urgent subject for concerted international multicenter trials comprising large-scale cohorts. The presently available experimental evidence obtained on autopsy-based models with doppler anemometry [15] as well as with advanced noninvasive MR techniques in vivo [16,17] has definitely shown an association between atherosclerosis development and low and oscillating wall shear stress on the interface between the blood and the endothelium. This means that there are now firm observations to support the theory of a relation between the physical forces acting on the arterial wall and the development of atherosclerosis in certain predisposed segments of the arterial tree. Extrapolating from these data, one must assume that the clinical correlates of the atherosclerotic process to the thrombotic mechanisms, at least to some unspecified degree, reflect the uneven distribution of the frictional forces acting on the arterial wall in the lower extremity as well as in the aorta [17] and carotid arteries [18,19]. Arterial remodeling also in connection with atherosclerotic stenosis development is well known. The anatomical changes induced by the remodeling process may add risk for thrombosis with hypertension, unstable shape of plaque, thickness of plaque, and the extent of the stenosis as risk factors for outward remodeling at least of carotid arteries [20]. This is probably of clinical importance in restenosis following angioplasty in leg arteries too, and whether new, experimental drugs such as sirolimus [21] will be useful for extremity arterial disease is open to speculation. In a recent study, pairs of atherosclerotic femoral arteries showed concordance in plaque size, expansive remodeling, and occurrence of plaques containing a lipid-rich core, but no concordance in plaque inflammation between right and left arteries. These findings strongly suggest that not only the amount of atherosclerosis but also arterial remodeling and plaque lipid deposition is influenced by systemic factors [22]. Thrombosis leads to endothelial dysfunction [23]. These phenomena are evidently secondary and obligatory to the atherosclerotic degeneration of the vessel wall, but the pathophysiologic consequences have not been clarified in details specifically with regard to the leg arteries [24]. It is of high theoretical interest that modest hyperinsulinemia, mimicking fasting hyperinsulinemia of insulin-resistant states, abrogates endothelium-dependent vasodilatation in large conduit arteries, probably by increasing oxidant stress [25]. Since nitrous oxide, which is formed in normal endothelial cells, is of great importance for the modulation of the peripheral vascular resistance, especially in diabetics [26], there is evidence for the fact that vasodilatation (by NO-induced reduction of intracellular calcium) together with autoregulation of blood flow and the veno-arteriolar sympathetic axon reflex are disturbed in patients with atherosclerosis in the legs [27]. Oxidation of lipids consumes NO in itself, and the overall effect of decreased NO in atherosclerosis is then increased risk of platelet aggregation, increased platelet adhesion to the endothelial surface, increased invasion of monocytes, higher arterial muscular tonus, and hypertrophy of the smooth muscle cells in the vessel wall. Finally, it has been shown that patients with intermittent claudication have disturbances in various rheological variables (red blood cell aggregation, hematocrit, viscosity; leukocyte and platelet activation in atherosclerotic regions) [28-30] that will add to the thrombogenic potential. Also, the amount of exercise-provoked micro-albuminuria in claudicants is associated with the degree of ischemia [31,32], an interesting observation, since micro-albuminuria is a marker of endothelial dysfunction [33]. Local arterial thrombosis secondary to atherosclerosis is the crucial stage of multiple, extremely complex, local, and generalized processes leading to blood flow interruption and, possibly, to acute

263

VASCULAR

EMERGENCIES

clinical presentation; it is, however, not the end point of the pathological process as is well known [5].

ARTERIAL THROMBOEMBOLISM AND EMBOLI

In the view of this author, the concept of arterial thrombo-embolism is dangerous. In the classical literature, it is repeatedly envisaged that a thrombus can cause embolization in the more distally located vascular tree by partial or complete loosening of the thrombus at the site of formation with subsequent symptoms or signs arising de novo or being accentuated by a deterioration of the pre-embolic clinical condition. This is certainly of high clinical importance in the carotid artery territory [34], often resulting in focal neurologic deficits such as amaurosis fugax, transitory ischemic attacks, and stroke. In patients with atherosclerosis causing thrombosis of the major arteries of the leg, there are frequently synchronous atherosclerotic lesions in the crural arteries as well. The development of thrombosis in the major leg arteries may lead to perfusion reduction in the more distal arteries and decrease of the linear blood flow velocities, so this multifaceted condition, often accompanied by congestive heart disease, hypotension, and dehydration, causes a stagnant, intra-arterial blood pool causing favorable conditions for distal thrombosis. It seems to me that an embolic etiology can rarely be established with confidence in patients with suspected or symptomatic occlusive arterial disease in the legs, whereas secondary crural artery thrombosis arising as described rather than caused by thrombo-embolism gives a more meaningful interpretation of the clinical situation. In other cases, of course, peripheral thrombosis may be secondary to embolization. In well-established situations such as in the presence of an ulcerated aortic plaque or a shaggy aorta, thrombo-embolism may be a relevant underlying mechanism behind for example blue toe syndrome (most frequently cholesterol athero-embolism); it is often obvious after percutaneous transluminal angioplasty [35], and occasionally after open reconstructions [36]. Overall, the incidence of primary embolization of the leg arteries is rapidly decreasing because of the improved medical treatment of rheumatic heart disease, atrial fibrillation, and myocardial infarction.

26
264

ANEURYSM
The development of nonspecific abdominal aortic aneurysms can no longer be attributed to simple atherosclerotic wall degeneration because mul-

tiple studies have shown complex disturbances in the normal molecular biology of the aortic wall in these aneurysm patients [37]. The debate on a possibly genetic etiology has not been settled as yet [38] although generalized aneurysmal disease (and systolic blood pressure) are associated with polymorphic variation in the fibrillin-1 gene [39]. Also, the expression of angiogenic factors such as vascular endothelial growth factors [40] and the oxidative stress [41] seem to be involved in the pathogenesis of aneurysms. Autoimmune [42,43] or even infectious pathogeneses (notably Chlamydia pneumoniae) have also been suggested, the latter being highly controversial [44]. Since there is some degree of concordance between abdominal aortic and popliteal aneurysms (around 10% of patients with abdominal aneurysm also have popliteal aneurysm) , it seems reasonable to suggest common etiology and pathogenesis for development of some aneurysmal types. Recently, this notion has found experimental substantiation, since the architectural changes in the vessel wall (primarily fragmentation of the elastic lamellae), the loss of vascular smooth muscle cells, and the increased amount of inflammatory infiltrate seem to be analogous; besides a number of biochemical markers, including those of apoptosis, support a unifying theory for arterial aneurysm development [45]. Experimental evidence has also pointed toward generalized involvement of the arterioles as well as the larger arteries [46]. There is also reduction in tensile strength and stiffness in venous tissue from patients with abdominal aortic aneurysm that is associated with disruption and reduction of the elastin contents of the vein wall; these changes are analogous to those observed in the arterial aneurysmal wall and confirm the systemic nature of this disorder [47]. Further, it has been shown that aortic aneurysms appear to be an important source of circulating interleukin-6 where the concentration is influenced by genotype [48], but it remains to be examined whether this holds true for popliteal aneurysms as well. It has been demonstrated that there is a surprisingly high occurrence of popliteal aneurysm in patients with the popliteal entrapment syndrome [49]. These aneurysms are seen in young patients, predominantly below 30 years of age, and it seems reasonable to infer that the cause of the development of aneurysm in this situation is the hemodynamic disturbances induced by the stenotic compression of the artery leading to poststenotic aneurysm formation, a remodeling response.

ACUTE ARTERIAL THROMBOSIS OF THE LOWER LIMBS

As with other aneurysms, those in the popliteal artery may thrombose; while the frequency is not known, older materials suggest that it is around 40% [50] with one quarter presenting with signs of acute and one quarter with chronic ischemia. Thrombosis must be conceived as being due to the pathologic blood flow patterns within and below the aneurysmal sac, but it cannot be excluded that emboli loosened from the aneurysm may impede run off in a progressive manner and result in a variant of stagnant flow thrombosis. Cystic adventitial disease of the popliteal artery rarely leads to thrombosis whereas claudication is common [51]. At least in the United States, lesions of the popliteal artery in high-performance athletes seem rather frequent [52]. companied by pulsating hematomas, local spasms, and intramural bleeding. Under unfavorable circumstances, hematomas in surrounding soft tissues can compress the artery with local occlusion and thrombosis. After stretching of the artery, and, of course, transection, thrombosis may occur. A special case occurs in drug addicts who inject various substances into the arteries. These arterial traumas have their highest prevalence in younger patients without notable collateral vessel formation which give rise to deep, distal ischemia that often necessitates urgent attempts of revascularization.

PHLEGMASIA CERULEA DOLENS

This is a rare but important cause of acute arterial thrombosis of the lower limbs. Phkgmasia cerulea dolens is the term for the clinical picture seen in massive, acute venous ilio-femoral thrombosis with life-threatening obstruction of the venous return, included the microvascular collaterals; occasionally compartment syndrome accompanies the serious state. The disease is most often seen on the left side in critically ill patients with medical problems such as hypercoagulation, serious infections (streptococcal fasciitis, Staphillococus aureus sepsis), incompensated heart disease, cancer or other chronic conditions that interfere with the clotting of the blood or the patient's mobility (postoperative state), but it can also occur rapidly and spontaneously in previously healthy individuals, such as pregnant or post partum women, and in children [57]. The reason for the accompanying secondary arterial thrombosis is stagnant flow due to the primary outflow obstruction.

ARTERIAL TRAUMA
The mechanisms of arterial trauma with subsequent acute thrombosis and the patterns of complications of vascular injury management are well documented in the literature [53]. In civilian practice in Denmark and Sweden, only 0.5 and 1.0 procedures are performed per 100 000 inhabitants per year for vascular trauma [54], respectively. Around two thirds affect the lower limb arteries [55]. The exact frequency of interventions performed by vascular surgeons for iatrogenic trauma is not known but it probably exceeds 40% of all vascular injuries in European countries without armed conflicts [56]. The increasing volume of invasive examinations and treatments (cardiac catheterizations, angiographic procedures, percutaneous transluminal angioplasty, orthopedic and neurosurgical procedures) has led to an increasing frequency of symptomatic lesions that need open, vascular surgical interventions. The mechanisms behind the acute thrombosis are multifactorial, and spasm must always be kept in mind. Local thrombosis can occur after contusion damages or in the vicinity of the path of a projectile through the vessels. Even with minimal endothelial damage, thrombosis can result as a consequence of local platelet aggregation and fibrin accumulation. The conditions for local thrombosis also occur after fracture of the intima, either partial or complete (blunt violence or penetrating trauma). Thrombosis is likely to occur when a local dissection and coiling of the intima reduces blood flow, and even a small, localized, intramural hematoma changes the properties of the overlying endothelium to make it highly thrombogenic. Local lacerations are frequently ac-

26
265

Miscellaneous
AORTIC DISSECTION [58]
Aortic dissection occasionally or often (12% to 55%, materials vary) affects inflow to the legs, exclusively or in combination with ischemia of other end organs [59]. The ischemia is produced by obliteration of the peripheral vessel origin (s) into the dissection or obliteration of the true lumen by an expanding false lumen. Marfan's syndrome sometimes causes lower limb ischemia [60-62], whereas leg ischemia due to Ehlers-Danlos syndrome seems more often to be caused by rupture of the often aneurysmatic aorta or iliacs [63,64]. Spontaneous dissection of the infrarenal aorta is very rare but it

VASCULAR

EMERGENCIES

should be known that 50% extend into the iliac or femoral arteries [65].
FlBROMUSCULAR DYSPLASIA

Fibromuscular dysplasia of the lower limbs is rare [66] and can be unilateral as well as bilateral and give rise to hypoplasia and anisomelia in children [67]. The literature contains a few reports on isolated crural artery affection [68]; the profunda femoris artery can also be diseased [69], whereas iliac artery disease seems to be a bit more frequent. It has not been investigated whether the endothelial properties are deranged in this disease, but due to the very nature of the wall changes, local vasomotor control must be assumed to be severely compromised.

teritis nodosa with aneurysms and those with systemic lupus erythematosus are seldom seen by vascular surgeons, but one should be aware that these diseases occasionally include secondary thrombosis in the clinical pattern. Cannabis as a cause of arteritis in young smokers seems to be forgotten as a cause of thrombosis and distal necrosis but should be kept in mind [84].

COAGULATION DISORDERS, TOBACCO, DRUGS

It is beyond the scope of this chapter to discuss these states; long, systematic lists can be found in relevant textbooks and in recent reviews [85]. Spasms must, as always, be kept in mind (ergotism, cocaine, heroin, temazepam, digitalis).

CYSTIC ADVENTITIAL DISEASE

Cystic adventitial disease most often affects the popliteal artery [70] but can also be found in veins; it is characteristic that the lesions appear near an articulation (knee, ankle, elbow, wrist) [71], but lesions in the common femoral artery have also been described [72]. Ultrasound will provide the diagnosis, although MR scanning will probably be indicated for precise visualization.

Diagnosis
Acute arterial thrombosis of the lower limbs is a clinical diagnosis that can be made with a high degree of certainty if based on a careful, analytical synthesis of anamnesis, symptoms, and signs. Only in selected cases, it should be considered to order supplementary examinations. Besides, it is essential to decide whether treatment should be initiated instantly, or whether it is advisable to postpone treatment as a consequence of one or more delaying diagnostic procedures. In other cases, waiting with treatment can be indicated after a balanced judgment (watchful waiting). The overriding consideration must always be that the risk to limbs and life for each and every patient must be minimized. A judgment of the safety of the system in which the surgeon is compelled to act must, as usual, be balanced toward the aggregated risk to which the patient is exposed. The clinical picture is well known: it can vary from the asymptomatic to the catastrophic. The development of symptoms is dependent on the anatomical localization and extent of the thrombosis, the degree of a possible collateral circulation, and the general state of the patient. One extreme is the sudden appearance of intense Pain with Pallor, Paresthesia, Paralysis, Pulselessness, and Poikilothermia (the six P); in other cases, a sudden occlusion of for example the superficial femoral artery can occur asymptomatically in an elderly patient with slowly progressing atherosclerosis with gradual development of compensatory collat-

26
266

EXTERNAL COMPRESSION
External compression of the arteries with subsequent thrombosis in the lower limbs can be seen in the popliteal entrapment syndrome [73-75], the compartment syndrome [76], in pregnancy with aortic dissection [77], and in combination with trauma.

ARTERITIS
Takayasu's disease, which has recently been described with simultaneous occurrence of Marfan's syndrome [78], Buerger's disease (thromboangiitis obliterans) [79], and Horton giant cell arteritis [80] are unusual and rare causes of thrombosis in western countries. Adamantiades-Behcet's disease in the angio variant is characterized by venous thrombosis and, on the arterial side, aneurysms (often multiple) are frequently encountered together with localized or more widespread arterial thromboses [81,82]. Due to the huge number of immigrants to northwestern Europe from the Middle East and Asia, the prevalence and incidence will increase here to approach that of the Mediterranean regions. Raynaud's disease occasionally causes thrombosis in distal arteries [83]. Patients with periar-

ACUTE ARTERIAL THROMBOSIS eral circulation. The conventional sign of pallor is misleading, cyanosis may be elicited or supercede pallor. An estimation of the venous filling or of the capillary return may be helpful. In many cases the symptoms will be relieved within hours, but other patients will have to be revascularized within a time span of four to six hours, less in children and young adults. Usually, the muscles will exhibit histological signs of warm ischemia after some four hours with edema (macroscopic fish flesh appearance) whereas the skin commonly can tolerate a more extended time span before necrosis develops, except in phlegmasia cerulea dolens where bullae as a sign of skin necrosis can occur rapidly. In a pathophysiologic context, it must be remembered that thrombosis in the major arteries, where adequate collaterals are not present or not adequate, will be followed by thrombosis in the microcirculation. This means that there is a possibility for the postoperative occurrence of the no reflow phenomenon, i.e., the capillary perfusion will not show signs of improvement in spite of even perfect blood delivery through the revascularized arterial system. Pulse palpation is a difficult art and meaningful interpretation is difficult. In light cases there will be moderate pain and pale skin with no effect on motor or sensory nervous function. The next step is more intense pain and development of paresthesia with numbness progressing to loss of sensation. The sense of light touch will be diminished and proprioception may disappear, whereas decreased temperature sense and two-point discrimination will disappear later (but before the deep pain and pressure senses are lost) since these sensory modalities are conducted through less hypoxia-sensitive fibers. The most serious case is a livid, anesthetic foot without capillary perfusion and venous filling together with rigidity of the crural muscles (rigor mortis) causing lack of dorsiflexion and plantar flexion of the toes alone or in combination with paresis of the ankle joint. In these patients, amputation may be an immediate option. A special variant is present in diabetics with preexisting sensory deficits. The Joint Council of The Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery have approved a set of recommended standards for reports dealing with lower extremity ischemia [86]. These are, however, problematic since presence or absence of arterial and venous doppler signals are included in the algorithm for classification into the clinical cate-

OF THE LOWER LIMBS

gories. The purely clinical descriptors for categorization of this type of ischemia are appropriate and may help to decide the plan of action. In some eminent textbooks (for example the standard work edited by Rutherford [87]), emphasis is laid on documenting the doppler segmental pressures. This can also be questioned since the methodological uncertainties [88] in combination with the erroneous measurements in certain patient groups (notably diabetics [89,90]) make it dangerous to interpret these recordings in a meaningful way. Even standardized ankle and toe blood pressure measurements performed in specialist laboratories with strain gauge equipment cannot be relied upon for diagnosis of acute thrombosis, especially in atherosclerotic patients, because of the wide standard variations [91]. The point is that acute thrombosis is a clinical diagnosis and that reliance on seemingly "objective" vascular laboratory measurements is dangerous. To supplement the clinical examination, ultrasound duplex scanning can be performed with reasonable confidence as a pre-operative procedure in at least chronic atherosclerotic occlusive disease [92,93], but in the acute situation its use is more problematic for arterial affection. In highly specialized laboratories, duplex may obviate the need for pre-operative angiography [94], but some systematic series are not so optimistic [95-97] with regard of getting precise, and usable, information as to the run off in distal, crural and pedal arteries for potential bypass procedures. Angiography, as duplex, should only in rare cases be considered as diagnostic but rather as a necessary pre-operative, or preinterventional, tool for detailed planning of the therapeutic procedure. Ultrasound duplex is the diagnostic procedure of choice for phlegmasia cerulea dolens, with very high sensitivity and specificity (around 95%) [98, 99]. It should be performed absolutely at once without delay as soon as clinical suspicion has arisen, and on liberal indications. The standard imaging technique for arterial thrombosis remains digital subtraction angiography (DSA) with intra-arterial injection of a contrast agent. The newer digital subtraction techniques [100] combined with contrast (signal to noise) enhancing agents have made MR angiography an option [101-104], but where thrombolysis or other endovascular procedures (for example suction/aspiration [105]) are possibilities, the first choice remains DSA, except in diabetics and patients with

26. 267

VASCULAR

EMERGENCIES

26
268

impaired kidney function. The new spiral or multislice computed tomography scanners can provide good three-dimensional multiplanar reconstructions of the major arteries by bench postprocessing; experience with crural and pedal arteries is limited but promising [102,106,107]. In selected difficult cases, we have occasionally found it helpful to visualize the degree of ischemia in the crural skeletal muscles by plain MR imaging (estimates of muscle edema) or by MR spectroscopy. Advanced indicator kinetic methods such as the single injection, residue detection method [108, 109] have been used experimentally in humans, and animals, to determine the capillary permeability of selected indicators in the peripheral tissues and the heart [110,111]. We recently attempted to combine these methods (which are based on the principles of irreversible thermodynamics and stochastic black box analysis) with the investigative options presented by positron electron tomographical scanning and apply them to patients with intermittent claudication and critical ischemia. The work is in progress in our institution and has not as yet been published. These technologies, which are very expensive and time consuming, open new windows of opportunity to assess objectively the degree of acute ischemia.

Treatment
The decision to treat acute thrombosis of the leg arteries is dependent on a complex set of factors of which not all can be subjected to a strict scientific analysis. The available literature is, also in this case, often not clear with respect to essential data [112,113]. The starting point is to decide whether treatment is indeed indicated. In an unknown proportion of cases, acute thrombosis of the superficial femoral artery is entirely asymptomatic, presumably often with absence or very discrete signs of clinical manifestations, as already noted by Fontaine [114]. The gray zone, where slight clinical symptoms and signs decline into minimal or absent problems after a period of observation, presents itself as an opportunity for the vascular surgeon to contemplate restraint of therapeutic enthusiasm: It can be done, but should it? For example, acute onset of claudication will frequently, if not always, merit a phase of observation and conservative treatment (stop

smoking and keep walking [115]). Often, strict scientific information for an informed clinical risk assessment is simply not available, and the strategy must be left to best clinical judgment, which may later be the subject of forensic controversy, complaints or litigation. Of the 424 claims recorded against vascular surgeons in the United Kingdom, 36 cases regarded alleged failure to recognize or treat ischemia [116]. Acute limb ischemia was defined by the TransAtlantic Inter-Society Consensus (TASC) as any sudden decrease or worsening in limb perfusion causing a potential threat to extremity viability [117], and acute arterial thrombosis was identified as one of the two major causes of acute limb ischemia, the other one being embolus. If the acute ischemia is, or potentially or manifestly becomes, a threat for limb viability, vascular surgical treatment is often, but not always, contraindicated. If the general state of the patient makes vascular surgical intervention problematic, or if the result of a considered analysis of the comorbidities is destructive, primary amputation might be indicated; this is generally the case in situations where the leg shows signs of motor paralysis as a result of extreme ischemia. Since a thrombosis may show signs of progression, some authors [85] advocate the immediate routine administration of intravenous heparin. If time permits, the general state of the patient should be optimized after consultation with experienced medical colleagues, notably anesthesiologists and cardiologists. In the acute phase, the overall strategy is to intervene in order to convert the acute situation into that immediately prior to this acute episode; this means that definite lege artis arterial reconstruction should be postponed and undertaken as an elective procedure, if at all possible. At the preliminary stage, the intra-arterial catheter-directed thrombolytic therapies are presumed helpful according to vascular surgical mainstream thinking to alleviate the acute symptoms, occasionally to dissolve the thrombus, to increase or restore distal perfusion, and to make visualization of the run off possible, either immediately or after a period of drug infusion where the window of therapeutic expectancy thus becomes widened [118]. Translated, this means that thrombolysis can be indicated in the acute patient coming in the evening and going on overnight to convert the patient's treatment from a possibly risky emergency operation to a more safer, considered, and elective daylight case after proper visualization of the run off options. This has

ACUTE ARTERIAL THROMBOSIS OF THE LOWER LIMBS the additional advantage of giving more time to alleviate or remove associated generalized risk factors. It must be stressed, though, that a recent Cochrane review addressed the question of whether the preferred initial treatment for acute limb ischemia was surgery or thrombolysis; embolic, thrombotic and other cases were included in the analysis [119]. All in all, five trials were identified for inclusion according to the Cochrane entry criteria, and the aggregated material comprised 1 283 patients, notably the STILE 1994 [120-123] and TOPAS 1996 [120, 124-126] series; also included was the small (n=20) Swedish material [127]. The conclusion was as follows: "Universal initial treatment with either surgery or thrombolysis cannot be advocated on the available evidence. There is no overall difference in limb salvage or death at one year between initial surgery and initial thrombolysis. Thrombolysis may be associated with a higher risk of ongoing limb ischemia, and of hemorrhagic complications, including stroke. The higher risk of complications must be balanced against risks of surgery in each patient." This attempt at a structured meta-analysis from the presently available literature must be considered most problematic. As the Cochrane reviewers themselves note, there were major differences in patient demographics, including severity of ischemia, site of occlusion, prosthetic or native vessel, and lytic regime and agent, and caution is required in interpreting the results of the meta-analysis of the data. It is necessary to point out in this connection that the data from the STILE study have indicated that for patients with subacute limb ischemia due to native vessel occlusion, surgery is both more effective and more durable than thrombolysis, as pointed out by the Leicester group [128]. Recently reported studies continue to report heterogeneous materials [129,130], although smaller but more focused series [131] seem to corroborate the view that thrombolysis has its place in the acute phases. Whether advancing age is a risk factor for lysis is controversial [132,133]. A Markov decision tree analysis based on the TOPAS series information and expenditures in a New York environment concluded that initial surgery provided the most efficient and economical use of resources for acute lower extremity arterial occlusion; besides, the high cost of thrombolysis was related to the expense of the lytic agents, the need for subsequent interventions in patients treated with initial lysis, and the long-term costs of amputation in patients who fail lytic therapy [134]. In conclusion, and in line with the Thrombolysis Consensus Document of 1998 [135], the scientific evidence for thrombolysis simply is not yet available but the data must nevertheless be considered to be sufficiently cogent for approval of agents. Finally, it should be noted that systemic intravenous infusion of thrombolytic agents for native artery occlusion must be considered obsolete [135]. The problem of lysis of thrombosed popliteal aneurysms, and the distal vasculature, has been very much debated, but there seems to be a growing consensus for lysis followed by permanent reconstruction [136-138] and also for elective operation of nonthrombosed aneurysms at that location [139]. Recent European series maintain the importance of an aggressive surgical approach to acute thrombosis [128]. The most common site of thrombosis in atherosclerotic patients is the superficial femoral artery. In these cases, the proper strategy is revascularization with an in-situ or reversed saphenous vein graft [113]; the second choice is implantation of a synthetic graft. There is still controversy as to whether PTFE (polytetrafluoroethylene) is preferable to dacron (polyester; polyethyleneterephthalate) [140, 142], and a recent Cochrane review [143] concluded that there is no clear evidence which type of graft is best for femoropopliteal grafting, that in terms of autologous graft patency in-situ and reversed vein grafts are equally successful while human umbilical vein performs better than PTFE, and that a distal vein cuff may improve primary patency for below-knee PTFE femoropopliteal grafting. In a very recent study, obviously not included in the Cochrane review, the two-year results of the Pop-Up multicentered (Denmark, Norway, Finland) and randomized study comprising 427 patients were presented [144]: 413 patients with 208 dacron (Uni-Graft, B-Braun Aesculap) and 205 PTFE (Goretex; W.L. Gore Inc.) grafts were available for analysis. The two-year primary patency rates were 68% for Dacron and 57% for PTFE (p=0.05), and the secondary patency rates were 76% and 63% respectively (p=0.01); amputations, mortality, and 30-day complications occurred at identical rates. However, this series did not focus on surgery for acute ischemia due to thrombosis, but it seems reasonable to extrapolate the results. Isolated thromboses of the common femoral artery and the deep femoral artery are rarer and can usually be managed by local thrombo-endarterectomy through groin incisions; they are not suitable for endovascular procedures, whereas the newer stent

26
269

VASCULAR

EMERGENCIES

26
270

graft endoprostheses [145-148] might prove their efficiency for recanalization of the superficial femoral artery, but their place in the treatment of acute thromboses has not been settled. The Amplatz thrombectomy device has been used for recanalization of acute occlusions of both the superficial and the deep femoral arteries, but the reports are anecdotal [149]. The Rotarex rotational thrombectomy catheter has been deployed in acute and subacute thromboses with occlusions of the middle or distal third of the superficial femoral artery or the popliteal artery; the results seem to warrant further studies [150]. Generally speaking, percutaneous thrombectomy may prove to become an option for treatment of acute thromboses [151], also for more distal pathology [152]. There is still controversy regarding whether short or long reversed saphenous vein grafts should be applied for treatment of a localized popliteal thrombosis. Some surgeons prefer to use the distal superficial femoral artery or the proximal popliteal artery as the inlet site for the bypass whereas others advocate mounting of a conventional saphenous vein in-situ graft from the groin to the infragenicular popliteal artery. The case for the long graft is corroborated by the fact that later development of atherosclerosis in the superficial femoral artery may necessitate further procedures in the long term. The short-term results seem to be identical. The endovascular techniques with endografts and/or stents are not yet established procedures [153]. The in-situ long saphenous vein graft technique (with or without the new endoscopic techniques [154] or blind valvulotomy [155]) is eminently suitable for thromboses necessitating a distal anastomosis below the knee joint, even for reconstruction of the foot arteries, as documented by a wealth of literature reports. The strict scientific evidence, however, as to the superiority of the in-situ technique

in preference to the reversed method, is not available [113], and presently the choice of procedure must be left to the discretion of the surgeon [156, 157]. Implantation of synthetic prostheses with distal anastomosis below the knee joint should be restricted to limb salvage cases after consideration of other options, for example the use of, spliced, arm veins [158] or other venous conduits. The efficacy of subintimal angioplasty of isolated infragenicular lesions is well documented [159, 160], but the place of this sophisticated treatment is not settled for acute cases. Finally, one should always exert a high level of suspicion for the development of revascularization compartment syndrome whether after thrombolysis or open surgery for acute thromboses [161].

Conclusion
Surgical treatment of acute lower limb thromboses is one of the core areas in clinical vascular surgery accounting for around 10 open procedures per 100 000 inhabitants of all ages per year in the Scandinavian countries corresponding to 10% to 15% of the total workload. Thrombolysis accounts for 3.6 cases per 100 000 per year (graft lysis included) . Typical patients are high-risk old age pensioners with complex histories, previous interventions, together with presence of multiple risk factors and comorbidities. The diagnosis is made on clinical grounds, and whereas the available diagnostic procedures can establish the precise location of the thrombotic processes and clarify the treatment options, the clinical decision of whether, when, and how to treat is highly complex and dependent on a multitude of factors, of which only a few have solid scientific evidence.

R E F E R E N C E S

1 Horrocks M. Arterial emboli of the lower limbs. In: Branchereau A, Jacobs M (eds). Vascular emergencies. Elmsford, Blackwell Publishing, Inc./Futura Division, 2003: pp 275-280. 2 Paaske WP. Overview of graft occlusion after lower limb revascularization: methodological aspects. In: Branchereau A, Jacobs M (eds). Complications in vascular and endovascular surgery (Part II). Armonk, Futura Publishing Co, 2002: pp 245-252. 3 Libby P. Changing concepts of atherogenesis.j7Kto"n Med 2000; 247:349.358. 4 Fayad ZA, Fuster V. Clinical imaging of the high-risk or vulne-

rable atherosclerotic plaque. CircResmi; 89: 305-316. 5 Stary HC, Chandler AB, Dinsmore RE et al. A definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis. A report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. Circulation 1995; 92:1355-1374. 6 Virmani R, Kolodgie FD, Burke AP et al. Lessons from sudden coronary death: a comprehensive morphological classification scheme for atherosclerotic lesions. Arterioscler Thromb Vase Biol 2000; 20:1262-1275.

ACUTE ARTERIAL THROMBOSIS OF THE LOWER LIMBS

7 Stary HC. Natural history and histological classification of atherosclerotic lesions: an update. Arteriosckr Thromb Vase Biol20QQ; 20: 1177.1178. 8 Fuster V. Lewis A. Conner Memorial Lecture. Mechanisms leading to myocardial infarction: insights from studies of vascular biology. Circulation 1994; 90: 2126-2146. 9 Falk E, Shah PK, Fuster V. Coronary plaque disruption. Circulation 1995; 92: 657-671. 10 Wexler L, Brundage B, Grouse J et al. Coronary artery calcification: pathophysiology, epidemiology, imaging methods, and clinical implications. A statement for health professionals from the American Heart Association. Writing Group. Circulation 1996; 94: 1175-1192. 11 Falk E. Stable versus unstable atherosclerosis: clinical aspects. Am Heart/1999; 138 (Pt 2): S421-425. 12 Rauch U, Osende JI, Fuster V et al. Thrombus formation on atherosclerotic plaques: pathogenesis and clinical consequences. Ann Intern Med 2001; 134: 224-238. 13 Worthley SG, Osende JI, Helft G et al. Coronary artery disease: pathogenesis and acute coronary syndromes. Mt Sinai J Med 2001; 68: 167-181. 14 Anonymous. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo- controlled trial. Lancet 2002; 360: 7-22. 15 Pedersen EM, Agerbaek M, Kristensen IB, Yoganathan AP. Wall shear stress and early atherosclerotic lesions in the abdominal aorta in young adults. Eur J Vase Endovasc Surg 1997; 13: 443-451. 16 Oyre S, Pedersen EM, Ringgaard S et al. In vivo wall shear stress measured by magnetic resonance velocity mapping in the normal human abdominal aorta. EurJ Vase Endovasc Surg 1997; 13: 263-271. 17 Pedersen EM, Oyre S, Agerbaek M et al. Distribution of early atherosclerotic lesions in the human abdominal aorta correlates with wall shear stresses measured in vivo. Eur J Vase Endovasc Surg 1999; 18: 328-333. 18 Oyre S, Ringgaard S, Kozerke S et al. Accurate noninvasive quantitation of blood flow, cross-sectional lumen vessel area and wall shear stress by three-dimensional paraboloid modeling of magnetic resonance imaging velocity data. / Am Coll Cardiol 1998; 32: 128-134. 19 Stokholm R, Oyre S, Ringgaard S et al. Determination of wall shear rate in the human carotid artery by magnetic resonance techniques. EurJ Vase Endovasc Surg 2000; 20: 427-433. 20 Saito D, Oka T, Kajiyama A et al. Factors predicting compensatory vascular remodelling of the carotid artery affected by atherosclerosis. Heart 2002; 87: 136-139. 21 Poon M, Badimon JJ, Fuster V. Overcoming restenosis with sirolimus: from alphabet soup to clinical reality. Lancet 2002; 359: 619-622. 22 Vink A, Schoneveld AH, Richard W et al. Plaque burden, arterial remodeling and plaque vulnerability: determined by systemic factors? / Am Coll Cardiol 2001; 38: 718.723. 23 Reil TD, Moore WS, Kashyap VS et al. The effects of thrombus, thrombectomy and thrombolysis on endothelial function. EurJ Vase Endovasc Surg 2000; 19:'162-168. 24 Rossi M, Cupisti A, Perrone L et al. Acute effect of exerciseinduced leg ischemia on cutaneous vasoreactivity in patients with stage II peripheral artery disease. Microvasc Res 2002; 64: 14-20. 25 Arcaro G, Cretti A, Balzano S et al. Insulin causes endothelial dysfunction in humans: sites and mechanisms. Circulation 2002; 105: 576.582. 26 Baron AD. Vascular reactivity. Am] Cardiol 1999; 84: 25J-27J. 27 Midttun M, Sejrsen P, Paaske WT. Peripheral blood flow rates and microvascular responses to orthostatic pressure changes in claudicants before and after revascularisation. Eur J Vase Endovasc Surg 1999; 17: 225-229. 28 Woodburn KR, Lowe GD, Rumley A et al. Relation of haemostatic, fibrinolytic, and rheological variables to the angiographic extent of peripheral arterial occlusive disease. Int Angiol 1995; 14: 346.352. 29 Dupuy-Fons C, Brun JF, Quere I et al. Rheology and occlusive arterial disease of the legs. J Mai Vase 1996; 21: 165-170. 30 Lee AJ, Fowkes FG, Rattray A et al. Haemostatic and rheological factors in intermittent claudication: the influence of smoking and extent of arterial disease. BrJHaematol 1996; 92: 226-230. 31 Hickey NC, Shearman CP, Gosling P, Simms MH. Assessment of intermittent claudication by quantitation of exerciseinduced micro-albuminuria. EurJ Vase Surg 1990; 4: 603-606. 32 Khaira HS, Maxwell SR, Shearman CP. Antioxidant consumption during exercise in intermittent claudication. BrJSurg 1995; 82: 1660.1662. 33 Seifalian AM, Atwal A, White S et al. A role for near infrared spectroscopy in the assessment of intermittent claudication. Int Angiol 2001; 20: 301 -306. 34 Gaunt ME, Brown L, Hartshorne T et al. Unstable carotid plaques: pre-operative identification and association with intraoperative embolisation detected by transcranial doppler. EurJ Vase Endovasc Surg 1996; 11: 78-82. 35 Al Hamali S, Baskerville P, Fraser S et al. Detection of distal emboli in patients with peripheral arterial stenosis before and after iliac angioplasty: a prospective study. / Vase Surg 1999; 29: 345.351. 36 Sharma PV, Babu SC, Shah PM, Nassoura ZE. Changing patterns of atheroembolism. Cardiovasc Surg 1996; 4: 573.579. 37 Paaske WP. Abdominal aortic aneurysm: an update of basic research. In: European Association of Surgical Sciences (ed). Up-date in vascular surgery. London, Foxwell & Davies Ltd, 2001: pp 55-64. 38 Powell JT, MacSweeney ST, Greenhalgh RM et al. Interaction between fibrillin genotype and blood pressure and the development of aneurysmal disease. Ann N Y Acad Sci 1996; 800: 198-207. 39 MacSweeney ST, Skidmore C, Turner RJ et al. Unravelling the familial tendency to aneurysmal disease: popliteal aneurysm, hypertension and fibrillin genotype. Eur J Vase Endovasc Surg 1996; 12: 162-166. 40 Kobayashi M, Matsubara J, Matsushita M et al. Expression of angiogenesis and angiogenic factors in human aortic vascular disease. / Surg Res 2002; 106: 239-245. 41 Miller FJ Jr, Sharp AVJ, Fang X et al. Oxidative stress in human abdominal aortic aneurysms: a potential mediator of aneurysmal remodeling. Arterioscler Thromb Vase Biol 2002; 22: 560-565. 42 Hirose H, Tilson MD. Abdominal aortic aneurysm as an autoimmune disease. Ann N Y Acad Sci 2001; 947: 416-418. 43 Rasmussen TE, Hallett JWr Jr, Schulte S et al. Genetic similarity in inflammatory and degenerative abdominal aortic aneurysms: a study of human leukocyte antigen class II disease risk genes./ Vase Sug2001; 34: 84-89. 44 Vammen S, Vorum H, Ostergaard L et al. Immunoblotting analysis of abdominal aortic aneurysms using antibodies against Chlamydia pneumoniae recombinant MOMP. Eur J Vase Endovasc Surg 2002; 24: 81-85. 45 Jacob T, Ascher E, Hingorani A et al. Initial steps in the unifying theory of the pathogenesis of artery aneurysms. / Surg Res 2001; 101: 37-43. 46 Midttun M, Sejrsen P, Paaske WT. Is non-specific aneurysmal disease of the infrarenal aorta also a peripheral microvascular disease? EurJ Vase Endovasc Surg 2000; 19: 625-629. 47 Goodall S, Crowther M, Bell PR et al. The association between venous structural alterations and biomechanical weakness in patients with abdominal aortic aneurysms. / Vase Surg 2002; 35: 937-942. 48 Jones KG, Brull DJ, Brown LC et al. Interleukin-6 (IL-6) and the prognosis of abdominal aortic aneurysms. Circulation 2001; 103: 2260-2265. 49 Gyftokostas D, Koutsoumbelis C, Mattheou T, Bouhoutsos J. Post stenotic aneurysm in popliteal artery entrapment syndrome. / Cardiovasc Surg 1991; 32: 350-352.'

26
271

VASCULAR

EMERGENCIES

-272

50 Inahara T, Toledo AC. Complications and treatment of popliteal aneurysms. Surgery 1978; 84: 775-783. 51 Scobie TK, Curry RH. Cystic adventitial disease of the popliteal artery. Can J Surg 1975; 18 : 46-50. 52 Arko FR, Harris EJ, Zarins CK et al. Vascular complications in high-performance athletes. / Vase Surg 2001; 33: 935-942. 53 Rich NM. Complications of vascular injury management. Surg Clin North Am 2002; 82: 143-174. 54 Paaske WP. EUROVASC report 1999. EurJ Vase Endovasc Surg 2001; 22: 282. 55 Paaske WP, Lorentzen JE, Hansen HJ. Peripheral arterial injuries. Acta Chir Scand Suppl 1980; 502: 176-180. 56 Fingerhut A, Leppaniemi AK, Androulakis GA et al. The European experience with vascular injuries. Surg Clin North Am 2002; 82: 175-188. 57 Hood DB, Weaver FA, Modrall JG, Yellin AE. Advances in the treatment of phlegmatia cerulea dolens. Am J Surg 1993; 166: 206-210. 58 Erbel R, Alfonso F, Boileau C et al. Diagnosis and management of aortic dissection. Eur Heart J mi; 22: 1642-1681. 59 Slonim SM, Miller DC, Mitchell RS et al. Percutaneous balloon fenestration and stenting for life-threatening ischemic complications in patients with acute aortic dissection. / Thome Cardiovasc Surg im; 117: 1118-1126. 60 Babu SC, Shah PM, Clauss RH. Unappreciated causes of ischemia in the leg. Am J Surg 1982; 144: 225-227. 61 Millar AJ, Gilbert RD, Brown RA et al. Abdominal aortic aneurysms in children. J Pediatr Surg 1996; 31: 1624-1628. 62 Shah PM, Clauss RH. Dissecting hematoma presents as acute lower limb ischemia: diagnostic patient profile and management. / Cardiovasc Surg 1983; 24: 649-653. 63 Cikrit DF, Glover JR, Dalsing MC, Silver D. The Ehlers-Danlos specter revisited. Vase Endovascular Surg 2002; 36: 213-217. 64 Germain DP. Clinical and genetic features of vascular EhlersDanlos syndrome. Ann Vase Sw/g2002; 16: 391-397. 65 Mozes G, Gloviczki P, Park WM et al. Spontaneous dissection of the infrarenal abdominal aorta. Semin Vase Surg 2002; 15: 128.136. 66 Sauer L, Reilly LM, Goldstone J et al. Clinical spectrum of symptomatic external iliac fibromuscular dysplasia. / Vase Surg 1990; 12: 488-496. 67 Flowers MJ, Eyres KS, Bell MJ, Gaines PA. Unilateral lower limb hypoplasia in arterial fibromuscular dysplasia. Clin Orthop 1996; 324: 217.221. 68 Patra P, De Lajartre AY, Chaillou P et al. Fibromuscular dysplasia of the arteries of the leg. A propos of one case. / Mai fee 1987; 12: 185-188. 69 Schneider PA, LaBerge JM, Cunningham CG, Ehrenfeld WK. Isolated thigh claudication as a result of fibromuscular dysplasia of the deep femoral artery. / Vase Surg 1992; 15: 657-660. 70 Levien LJ, Benn CA. Adventitial cystic disease: a unifying hypothesis. / Vase Surg 1998; 28: 193-205. 71 Chakfe N, Beaufigeau M, Geny B et al. Extra-popliteal localizations of adventitial cysts. Review of the literature. J Mai Vase 1997; 22: 79-85. 72 Carlsson S, Sandermann J, Hansborg N. Adventitial cystic disease in the common femoral artery. Ann Chir Gynaecol 2001; 90: 63.64. 73 Lambert AW, Wilkins DC. Popliteal artery entrapment syndrome. BrJSurgim; 86: 1365-1370. 74 Stager A, Clement D. Popliteal artery entrapment syndrome. Sports Med im; 28: 61-70. 75 Turnipseed WD. Popliteal entrapment syndrome. / Vase Surg 2002; 35: 910-915. 76 Ulmer T. The clinical diagnosis of compartment syndrome of the lower leg: are clinical findings predictive of the disorder? J Orthop Trauma 2002; 16: 572-577. 77 Pakarian F, Ray SA, Bewley S. Relief of leg ischaemia by caesarean section: a case report of aortic dissection during pregnancy. EurJ Vase Endovasc Surg 2000; 20: 209-210.

78 Back HJ, Shin KC, Lee YJ et al. Takayasu's arteritis concurrent with Marfan syndrome. A case report. Angiology 2000; 51: 435.439. 79 Olin JW. Thromboangiitis obliterans (Buerger's disease). N EnglJMed 2000; 343: 864.869. 80 Le Hello C, Levesque H, Jeanton M et al. Lower limb giant cell arteritis and temporal arteritis: follow-up of eight cases. JRheumatoimi;28: 1407-1412. 81 Bradbury AW, Milne AA, Muria JA. Surgical aspects of Behcet's disease. BrJSurgim; 81: 1712-1721. 82 Ozeren M, Mavioglu I, Dogan 0V, Yucel E. Reoperation results of arterial involvement in Behcet's disease. EurJ Vase Endovasc Su2000; 20:512-519. 83 Belch JJ, Ho M. Pharmacotherapy of Raynaud's phenomenon. Drugs 1996; 52: 682-695. 84 Disdier P, Granel B, Serratrice J et al. Cannabis arteritis revisited. Ten new case reports. Angiology 2001; 52: 1-5. 85 Liapis C, KakisisJ. Acute arterial occlusion of the extremities. In: European Association of Surgical Sciences (ed). Up-date in vascular surgery. London, Foxwell & Davies Ltd, 2001: pp 111.123. 86 Rutherford RB, Baker JD, Ernst C et al. Recommended standards for reports dealing with lower extremity ischemia: revised version. / Vase Surg 1997; 26: 517-538. 87 Zierler RE, Sumner DS. Physiological assessment of peripheral arterial occlusive disease. In: Rutherford RB (ed). Vascular surgery. 5th edition. Philadelphia, W.B.Saunders Co 2000: pp 140.165. 88 Jeelani NU, Braithwaite BD, Tomlin C, MacSweeney ST. Variation of method for measurement of brachial artery pressure significantly affects ankle-brachial pressure index values. EurJ Vase Endovasc Surg 2000; 20 : 25-28. 89 Osmundson PJ, Chesebro JH, O'Fallon WM et al. A prospective study of peripheral occlusive arterial disease in diabetes. II. Vascular laboratory assessment. Mayo Clin Proc 1981; 56: 223-232. 90 Vincent DG, Salles-Cunha SX, Bernhard VM, Towne JB. Noninvasive assessment of toe systolic pressures with special reference to diabetes mellitus. / Cardiovasc Surg 1983; 24: 22-28. 91 Tonnesen KH, Noer I, Paaske W, Sager P. Classification of peripheral occlusive arterial diseases based on symptoms, signs and distal blood pressure measurements. Acta Chir Scand 1980; 146: 101.104. 92 Aly S, Jenkins MP, Zaidi FH et al. Duplex scanning and effect of multisegmental arterial disease on its accuracy in lower limb arteries. EurJ Vase Endovasc Surg 1998; 16: 345-349. 93 Eiberg J, Madycki G, Hansen M et al. Ultrasound imaging of infrainguinal arterial disease has a high interobserver agreement. EurJ Vase Endovasc Surg 2002; 24: 293-299. 94 McCarthy MJ, Nydahl S, Hartshorne T et al. Colour-coded duplex imaging and dependent Doppler ultrasonography in the assessment of cruropedal vessels. BrJ Surg 1999; 86: 33-37. 95 Larch E, Minar E, Ahmadi R et al. Value of color duplex sonography for evaluation of tibioperoneal arteries in patients with femoropopliteal obstruction: a prospective comparison with anterograde intra-arterial digital subtraction angiography. / Vase Surg 1997; 25: 629.636. 96 Wain RA, Berdejo GL, Delvalle WN et al. Can duplex scan arterial mapping replace contrast arteriography as the test of choice before infrainguinal revascularization? / Vase Surg 1999; 29: 100-109. 97 Avenarius JK, Breek JC, Lampmann LE et al. The additional value of angiography after colour-coded duplex on decision making in patients with critical limb ischaemia. A prospective study. EurJ Vase Endovasc Surg 2002; 23: 393-397. 98 Aitken AG, Godden DJ. Real-time ultrasound diagnosis of deep vein thrombosis: a comparison with venography. Clin Radiol 1987; 38: 309-313.

ACUTE ARTERIAL THROMBOSIS OF THE LOWER LIMBS

99 Tan SS, Chong BK, Thoo FL et al. Diagnosis of deep venous thrombosis: accuracy of colour doppler ultrasound compared with venography. Singapore Afef/1995; 36: 362-366. lOORuehm SG, Nanz D, Baumann A et al. 3D contrast-enhanced MR angiography of the run-off vessels: value of image subtraction. / Magn Reson Imaging 2001; 13: 402-411. 101 Goyen M, Ruehm SG, Barkhausen J et al. Improved multi-station peripheral MR angiography with a dedicated vascular coil. JMagn Reson Imaging 2001; 13: 475-480. 102 Katz DS, Hon M. CT angiography of the lower extremities and aortoiliac system with a multi-detector row helical CT scanner: promise of new opportunities fulfilled. Radiology 2001; 221: 7.10. 103Svensson J, Leander P, Maki JH et al. Separation of arteries and veins using flow-induced phase effects in contrast-enhanced MRA of the lower extremities. Magn Reson Imaging 2002; 20 : 49-57. 104 Wang Y, Winchester PA, Khilnani NM et al. Contrast-enhanced peripheral MR angiography from the abdominal aorta to the pedal arteries: combined dynamic two-dimensional and boluschase three-dimensional acquisitions. Invest Radiol 2001; 36: 170.177. 105Zehnder T, Birrer M, Do DD et al. Percutaneous catheter thrombus aspiration for acute or subacute arterial occlusion of the legs: how much thrombolysis is needed? EurJ Vase Endovasc Surg 2000; 20: 41.46. 106Puls R, Knollmann F, Werk M et al. Multi-slice spiral CT: 3D CT angiography for evaluating therapeutically relevant stenosis in peripheral arterial occlusive disease. Rontgenpraxis 2001; 54: 141-147. 107 Rubin GD, Schmidt AJ, Logan LJ, Sofilos MC. Multi-detector row CT angiography of lower extremity arterial inflow and runoff: initial experience. Radiology 2001; 221: 146-158. lOSPaaske WP, Sejrsen P. Permeability of continuous capillaries. Dan Med Bull 1989; 36: 570.590. 109 Paaske WP, Sejrsen P. Microvascular function in the peripheral vascular bed during ischaemia and oxygen-free perfusion. Eur J Vase Endovasc Surg 1995; 9: 29-37. 110 Haunso S, Paaske WP, Sejrsen P, Amtorp 0. Capillary permeability in canine myocardium as determined by bolus injection, residue detection. Ada Physiol Scand 1980; 108: 389-397. 111 Svendsen JH, Paaske WP, Sejrsen P, Haunso S. Capillary permeability of 1311-albumin in canine myocardium as determined by bolus injection, residue detection. Microvasc Res 1989; 37: 352-356. 112 Wolfe JHN, Cheshire NJ. Does distal revascularization for limb salvage work? In: Greenhalgh RM, Fowkes FGR (eds). Trials and tribulations of vascular surgery. Evidence based vascular surgery. London, WB Saunders, 1996: pp 353.364. 113 Paaske WP. Femoropopliteal reconstruction. In: Branchereau A, Jacobs M (eds). Critical limb ischemia. Armonk, Futura Publishing Co, 1999: pp 147.164. 114 Fontaine R, Kim M, Kieny R. Die chirurgische Behandlung der peripheren Durchblutungsstorungen. Helvetica Chirurgica Ada 1954; 5/6: 499-533. USHousley E. Treating claudication in five words. Br Medjl%8; 296: 1483-1484. 116 Campbell WB, France F, Goodwin HM. Medicolegal claims in vascular surgery. Ann R Coll SurgEngl 2002; 84: 181-184. 117 Anonymous. Management of peripheral arterial disease (PAD). TransAtlantic Inter- Society Consensus (TASC). Section C: acute limb ischaemia. EurJ Vase Endovasc Surg 2000; 19 Suppl A: SI 15S143. 118 Hall TB, Matson M, Belli AM. Thrombolysis in the peripheral vascular system. Eur Radiol 2001; 11: 439-445. 119 Berridge DC, Kessel D, Robertson I. Surgery versus thrombolysis for acute limb ischaemia: initial management. Cochrane Database Syst Rev 2002; 3: CD002784. 1200uriel K, Shortell CK, DeWeese JA et al. A comparison of thrombolytic therapy with operative revascularization in the initial treatment of acute peripheral arterial ischemia. J Vase Surg 1994; 19: 1021-1030. 121 Anonymous. Results of a prospective randomized trial evaluating surgery versus thrombolysis for ischemia of the lower extremity. The STILE trial. Ann Surg 1994; 220: 251-268. 122 Weaver FA, Comerota AJ, Youngblood M et al. Surgical revascularization versus thrombolysis for nonembolic lower extremity native artery occlusions: results of a prospective randomized trial. The STILE investigators. Surgery versus thrombolysis for ischemia of the lower extremity. J Vase Surg 1996; 24: 513.523. 123 Comerota AJ, Weaver FA, Hosking JD et al. Results of a prospective, randomized trial of surgery versus thrombolysis for occluded lower extremity bypass grafts. Am J Surg 1996; 172: 105.112. 1240uriel K, Veith FJ, Sasahara AA. Thrombolysis or peripheral arterial surgery: phase I results. TOPAS investigators. JVasc Surg 1996; 23: 64-75. 125 Ouriel K, Veith FJ. Acute lower limb ischemia: determinants of outcome. Surgery 1998; 124: 336-342. 126 Ouriel K, Veith FJ, Sasahara AA. A comparison of recombinant urokinase with vascular surgery as initial treatment for acute arterial occlusion of the legs. Thrombolysis or Peripheral Arterial Surgery (TOPAS) investigators. NEnglJMed 1998; 338: 1105-1111. 127Nilsson L, Albrechtsson U, Jonung T et al. Surgical treatment versus thrombolysis in acute arterial occlusion: a randomised controlled study. EurJ Vase Surg 1992; 6: 189-193. 128Pemberton M, Varty K, Nydahl S, Bell PR. The surgical management of acute limb ischaemia due to native vessel occlusion. EurJ Vase Endovasc Surg 1999; 17: 72-76. 129Korn P, Khilnani NM, Fellers JC et al. Thrombolysis for native arterial occlusions of the lower extremities: clinical outcome and cost. / Vase Surg 2001; 33: 1148-1157. 130 Swischuk JL, Fox PF, Young K et al. Transcatheter intra-arterial infusion of rt-PA for acute lower limb ischemia: results and complications. / Vase Interv Radiol 2001; 12: 423-430. 131 Suggs WD, Cynamon J, Martin B et al. When is urokinase treatment an effective sole or adjunctive treatment for acute limb ischemia secondary to native artery occlusion? Am J Surg 1999; 178: 103.106. 132 Braithwaite BD, Davies B, Birch PA et al. Management of acute leg ischaemia in the elderly. BrJSurg 1998; 85: 217-220. 133 Lambert AW, Trkulja D, Fox AD et al. Age-related outcome for peripheral thrombolysis. EurJ Vase Endovasc Surg 1999; 17: 144.148. 134Patel ST, Haser PB, Bush HL Jr, Kent KG. Is thrombolysis of lower extremity acute arterial occlusion cost-effective? J Surg Res 1999; 83: 106-112. 135 Anonymous. Thrombolysis in the management of lower limb peripheral arterial occlusion. A consensus document. Working party on thrombolysis in the management of limb ischemia. Am/Carrfw/1998;81: 207-218. 136Dorigo W, Pulli R, Turini F et al. Acute leg ischaemia from thrombosed popliteal artery aneurysms: role of pre-operative thrombolysis. EurJ Vase Endovasc Surg 2002; 23: 251-254. 137 Steinmetz E, Bouchot 0, Faroy F et al. Pre-operative intra-arterial thrombolysis before surgical revascularization for popliteal artery aneurysm with acute ischemia. Ann Vase Surg 2000; 14: 360.364. 138Greenberg R, Wellander E, Nyman U et al. Aggressive treatment of acute limb ischemia due to thrombosed popliteal aneurysms. EurJ Radiol 1998; 28: 211-218. 139 Michaels JA, Galland RB. Management of asymptomatic popliteal aneurysms: the use of a Markov decision tree to determine the criteria for a conservative approach. EurJ Vase Surg 1993; 7: 136-143. 140 Robinson BI, Fletcher JP, Tomlinson P et al. A prospective randomized multicentre comparison of expanded polytetrafluoroethylene and gelatin-sealed knitted dacron grafts for femoropopliteal bypass. Cardiovasc Surg 1999; 7: 214-218.

273

VASCULAR

EMERGENCIES

26
274

141 Green RM, Abbott WM, Matsumoto T et al. Prosthetic aboveknee femoropopliteal bypass grafting: five-year results of a randomized trial. / Vase Surg 2000; 31: 417.425. 142 Post S, Kraus T, Muller-Reinartz U et al. Dacron vs. polytetrafluoroethylene grafts for femoropopliteal bypass: a prospective randomised multicentre trial. EurJ Vase Endovasc Surg 2001; 22: 226.231. 143Mamode N, Scott RN. Graft type for femoro-popliteal bypass surgery. Cochrane Database Syst Rev 2000; 2: CD001487. 144Jensen LP, Schroeder TV, on behalf of the Pop-Up Study Group. Dacron is better than PTFE for above-knee femoropopliteal bypass. A randomized clinical trial. Book of Abstracts, XVI Annual Meeting of the European Sodety for Vascular Surgery, 2002: pp 141-142. 145LammerJ, Dake MD, Bleyn J et al. Peripheral arterial obstruction: prospective study of treatment with a transluminally placed self-expanding stent-graft. International Trial Study Group. Radiology 2000; 217:95.104. 146Bauermeister G. Endovascular stent-grafting in the treatment of superficial femoral artery occlusive disease. / Endovasc Ther 2001; 8: 315-320. 147Deutschmann HA, Schedlbauer P, Berczi V et al. Placement of Hemobahn stent-grafts in femoropopliteal arteries: early experience and midterm results in 18 patients. J Vase Interv Radial 2001; 12: 943-950. 148 Rubin BG, Sicard GA. The Hemobahn endoprosthesis: a selfexpanding polytetrafluoroethylene-covered endoprosthesis for the treatment of peripheral arterial occlusive disease after balloon angioplasty. / Vase Surg ZOO}; 33 (2 Suppl): S124-S128. 149Gorich J, Rilinger N, Sokiranski R et al. Mechanical thrombolysis of acute occlusion of both the superficial and the deep femoral arteries using a thrombectomy device. AJR Am J Roentgenol 1998; 170: 1177-1180. 150 Berczi V, Deutschmann HA, Schedlbauer P et al. Early experience and midterm follow-up results with a new, rotational

thrombectomy catheter. Cardiovasc Intervent Radiol 2002; 25: 275-281. 151 Morgan R, Belli AM. Percutaneous thrombectomy: a review. EurRadiol2002;\2:205.2\7. 152 Desgranges P, Kobeiter K, d'Audiffret A et al. Acute occlusion of popliteal and/or tibial arteries: the value of percutaneous treatment. EurJ Vase Endovasc Surg 2000; 20: 138-145. 153 Rosenthal D, Matsuura JH, Clark MD et al. Popliteal artery aneurysms: is endovascular reconstruction durable? J Endovasc Ther 2000; 7: 394-398. 154 Suggs WD, Sanchez LA, Woo D et al. Endoscopically assisted insitu lower extremity bypass graft: a preliminary report of a new minimally invasive technique. / Vase Surg 2001; 34: 668-672. 155 Gangadharan SP, Reed AB, Chew DK et al. Initial experience with minimally invasive in-situ bypass procedure with blind valvulotomy. / Vase Surg 2002; 35: 1100-1106. 156Lawson JA, Tangelder MJ, Algra A, Eikelboom BC. The myth of the in-situ graft: superiority in infrainguinal bypass surgery? EurJ Vase Endovasc Surg 1999; 18: 149.157. 157 Eugster T, Stierli P, Aeberhard P. Infrainguinal arterial reconstruction with autologous vein grafts: are the results for the insitu technique better than those of non-reversed bypass? A longterm follow-up study. / Cardiovasc Surg 2001; 42: 221-226. 158 Browning N, Zammit M, Rodriguez D et al. Use of arm veins for lower extremity arterial bypass. Results, anatomical features and technical considerations. S AfrJ Surg 2000; 38: 36-41. 159 Ingle H, Nasim A, Bolia A et al. Subintimal angioplasty of isolated infragenicular vessels in lower limb ischemia: long-term results. J Endovasc Ther 2002; 9: 411.416. 160 Varty K, Nydahl S, Nasim A et al. Results of surgery and angioplasty for the treatment of chronic severe lower limb ischaemia. Eurj Vase Endovasc Surg 1998; 16: 159-163. 161 Defraigne JO, Pincemail J. Local and systemic consequences of severe ischemia and reperfusion of the skeletal muscle. Physiopathology and prevention. Ada Chir Belg 1998; 98: 176-186.

ARTERIAL EMBOLI OF THE LOWER LIMBS

MICHAEL HORROCKS

It is now well recognized that acute leg ischemia is a dangerous condition leading to loss of limb and loss of life. In a major review of acute leg ischemia reported by the Vascular Surgical Society of Great Britain and Ireland, limb salvage rates were only 70 % and there was a 22% mortality rate during the same hospital stay [1], With the declining incidence of rheumatic heart disease and increased evidence of acute on chronic ischemia, an accurate and speedy diagnosis is essential so that the appropriate treatment can be given. Acute arterial ischemia is caused either by an embolusfrom a proximal source or by primary thrombosis of the artery itself. It is important to try and distinguish between these two principal causes, as the outcome may be quite different and the treatment required may be selected according to etiology.

27
275

Background
In a review of intra-arterial thrombolysis in the United Kingdom, it was shown that there were approximately 5 000 patients presenting annually with acute lower limb ischemia [2]. The etiology was found to be thrombo-embolic disease in the vast majority of cases, and thrombosis had now replaced embolism as the principal cause of acute ischemia. Thrombosis accounted for approximately 60% of all cases with embolism accounting for less than 30%. The review also showed that intra-arterial thrombolysis was being used increasingly as a first line

treatment for acute lower limb ischemia, even when the evidence suggested that the cause was embolic. In a review of acute and acute on chronic leg ischemia, Kauhanen et al. emphasized the importance of distinction between acute embolic ischemia and acute on chronic ischemia caused by thrombus superimposed on an atherosclerotic stenosis [3]. In a series of 194 acutely ischemic lower limbs of which 189 underwent balloon catheter thrombo-embolectomy, even in retrospect, the diagnosis of acute embolic ischemia was only made in 48% of patients, the remainder being acute on chronic ischemia. In 22% of the patients, the emergency surgeon was

VASCULAR

EMERGENCIES

unable to distinguish between acute embolic ischemia and acute on chronic ischemia. The results of these studies emphasize the difficulties of clinical assessment and suggest that routine diagnostic angiography would be an important evaluation tool in planning optimum management, particularly as many patients can be treated by catheterbased techniques.

it should be remembered that there are less common causes giving rise to arterial emboli. In a report by Pereira et al., infective endocarditis was found in 24% of patients causing arterial emboli secondary to heart disease, and a further 28% secondary to dilating cardiomyopathy [4].

Investigations Presentation
The classic presentation of acute embolic arterial ischemia is the six P's: Pain, Paresthesia, Pulselessness, Pallor, Paralysis and Perishing cold. There is often an obvious source of an embolus such as cardiac arrhythmia (usually atrial fibrillation), recent history of myocardial infarction, or an alternative source such as a large abdominal aortic aneurysm. There is usually evidence that the rest of the peripheral vascular tree is normal with no history of previous claudication and a full compliment of pulses in the other limbs. When clot or atherosclerotic plaque from a proximal source embolizes into the arterial tree, the embolus usually lodges at the bifurcation of an artery. The common sites for lodging are the bifurcations of the aorta, the common iliac artery, the common femoral artery and the popliteal artery. This usually produces sudden and very severe ischemia of the limb below the embolus and, as there has been little opportunity for the development of collateral circulation, the ischemia is usually profound. This clear-cut pattern of presentation is easy to recognize and should be treated with great urgency. The classic picture of acute embolic ischemia is becoming increasingly rare with the decline in rheumatic heart disease and the better treatment of atrial fibrillation. More commonly, patients present with a more complicated picture. Patients with atrial fibrillation may also have peripheral vascular disease giving rise to a less acute picture because of preexisting collaterals around the site of the embolus. In the presence of such peripheral vascular disease, even if an embolus has caused the sudden onset of ischemia, the results of embolectomy are much less certain. In these cases, making the diagnosis is more difficult and treatment is associated with increased risk of postoperative complications. While atrial fibrillation associated with ischemic heart disease or rheumatic heart disease remains the most common source of peripheral embolism, In the presence of clear-cut criteria for an embolus like sudden onset of leg pain, clinical evidence of heart disease like myocardial infarction or atrial fibrillation, and no evidence of peripheral vascular disease (normal pulses in the other leg), it is usually considered safe to proceed straight to surgical embolectomy. If there is doubt as to the diagnosis, urgent arteriography is mandatory, with a provision to treat by catheter techniques if appropriate.

Techniques of embolectomy
It has now been almost fourty years since the first embolectomy technique was recorded by Fogarty et al. [5]. More recently, new percutaneous thrombectomy techniques have been used for removing emboli including aspiration thrombo-embolectomy, angioscopic thrombo-embolectomy and mechanical thrombectomy with or without fibrinolytic infusions. BALLOON THROMBO-EMBOLECTOMY Although balloon embolectomy is a well-tried and proven technique, problems remain with its use. It is well recognized that after removal of an embolus from an arterial wall by embolectomy catheter, residual thrombus may well be left behind. Furthermore, secondary thrombosis of runoff vessels may be inadequately cleared, giving rise to poor flow and the possibility of secondary thrombosis. Thrombo-embolectomy using a Fogarty catheter may also cause arterial perforation and pseudo-aneurysm formation, and may at a later stage give rise to intimal hyperplasia [6,7]. The degree of endothelial damage that leads to intimal hyperplasia appears to be independent of the type of balloon catheter used. It increases with increasing sheer force (pressure) on the vessel wall and with multiple passages of the balloon [8]. It has been suggested that the passage of an embolectomy catheter over a guide wire under X-ray control might reduce the incidence of arterial

276

ARTERIAL EMBOLI OF THE LOWER LIMBS

perforation, although there are no randomized data to support this. There is little evidence that the type of embolectomy catheter used has any influence on complication rate or outcome [9]. Recently, mechanical thrombectomy has been promoted as an effective way of clearing thromboembolic material in an occluded artery. These mechanical devices either combine clot maceration with suction removal (Angiojet Hydrolysor, Oasis) or use clot maceration alone (Amplatz thrombectomy device). Most of this experience appears to relate to acute on chronic ischemia rather than embolic disease and there are no separate data for emboli as opposed to thrombosis. There are many reports of successful removal of the embolized material following mechanical disruption from the distal vessel. popliteal artery with control of the origin of each of the calf vessels is essential. This allows direct control of the embolectomy catheter with passage into each of the calf vessels. Following an arteriotomy of the popliteal artery, it is wise to repair the artery with a vein patch to prevent narrowing and secondary thrombosis.

Quality control
On completion of what appears to be a satisfactory embolectomy, it is essential to do an on-table angiogram or similar investigation to ensure that the artery is adequately cleared. It is well recognized that up to 70% of embolectomies are incomplete after the surgeon feels he or she has a satisfactory result, angiography showing residual thrombus or further thrombus in a more distal part of the limb. In the author's experience, it is a good idea to include intra-arterial thrombolysis inserted distally after successful embolectomy. This can remain in the distal circulation while the upstream part of the embolectomy is performed and during the period of arteriotomy closure. Completion angiography should be completed prior to closure of the groin wound.

SURGICAL EMBOLECTOMY
In cases of a saddle embolus, there is sudden occlusion of the aorta by a large embolus sitting at the aortic bifurcation. This is best approached by a bilateral femoral route. It is important not to try and remove the embolus through one single femoral artery, as there is a real danger of dislodging clot into the contralateral leg. A bilateral femoral approach can be done perfectly adequately under local anesthetic although a regional or general anesthetic may be preferable in a very distressed patient. For a patient who has an embolus in either the iliac or common femoral arteries, a femoral approach is satisfactory. It would be normal practice to give heparin 5 000 international units intravenously once the decision has been made to do an embolectomy, but this may be delayed if the patient is to have a spinal or epidural anesthetic. Local anesthesia by direct infiltration is adequate, allowing full mobilization of the common femoral artery and the origin of the superficial and profunda femoris artery. Following arteriotomy, an appropriately sized Fogarty catheter should be passed upward to ensure that the inflow is clear, and then distally down both the profunda femoris and the superficial femoral artery. Pressure on the balloon should be kept to a minimum compatible with removal of the clot. For emboli that have gone to the popliteal artery or more distally, it may be advantageous to explore the popliteal artery rather than the femoral artery. This is best done by a medial approach. If there is evidence that there is thrombus going into the individual calf vessels, then exploration of the

277

Results
In a review of lower limb emboli in a single vascular unit, Burgess et al. found a mean delay before diagnosis of 29 hours in 71 lower limb emboli (range 1 to 264 hours) [10]. Two thirds of the causes were atrial fibrillation and one third was other sources. Thirty-day mortality was 45% with an amputation rate of 15% and an overall complication rate of 62%. There was no evidence of improving figures over a ten-year period. Factors that increased mortality were a delay before diagnosis, grade of surgeon performing the operation, and inadequate inflow or outflow at operation. Factors found to affect limb salvage rate adversely were intermittent claudication, lack of immediate pre-operative heparinization and juniority of operating surgeon. In a study of 397 patients, Becquemin and Kovarsky reviewed their experience of lower limb arterial embolism. The in-hospital mortality rate was 15%. Risk factors for amputation were comorbidity, severity of

VASCULAR

EMERGENCIES

27
278

ischemia, distal embolus and delay in treatment [11]. The role of thrombolysis in the treatment of embolic occlusion of the lower extremity was reported from the STAR Registry by Huettl and Soulen [12] in which 45 of 306 consecutive cases of lower limb arterial occlusion were treated with urokinase and registered in the Society of Cardiovascular Radiology Transluminal Angioplasty and Revascularization Registry. The diagnosis of embolus was based on the combination of clinical and angiographic data. Fifty of the patients had atrial fibrillation, 40% previous myocardial infarction and 35% had a prior cerebrovascular event. At the time of presentation 71% of the limbs were deemed to be viable, 27% were threatened and one had irreversible ischemia. The mean duration of symptoms to treatment was 8.6 days and the average length of occlusion was 17 centimeters. The distribution of the emboli was 4% in the aorto-iliac segment, 65% in the femoropopliteal segment, 24% in the tibial segment and 7% were into an arterial graft. There was a technical success rate of 69% with a one-year primary patency rate of 79%. These results are similar to those reported for surgical embolectomy and appear to confirm no advantage over surgical procedure. Much of the data related to the management of acute critical ischemia with thrombolysis do not clearly differentiate between embolic and primary thrombotic ischemia. In their review of surgery versus thrombolysis for acute limb ischemia for the Cochrane database, Berridge et al. reviewed 1283 patients in five trials [13]. They found no significant difference in limb salvage or death at 30 days, 6 months, or one year between initial surgery and thrombolysis. Thrombolysis was thought to be associated with a higher risk of ongoing limb ischemia and hemorrhagic complications including stroke. They concluded that the higher risk of complications must be balanced against the relative risk of surgery in each patient. These patients were mainly acute on chronic ischemia rather than embolic, and no differentiation is made between the results in those who were found in retrospect to have an embolus and those with a primary thrombosis. A further study of 204 patients (107 female, 97 male) with 224 episodes of lower limb ischemia treated surgically between 1993 and 1997 is reported by Ilic et al. [14]. Admission and operation were within 6 hours in 9% of the patients, 6 hours to 24 hours in 33% of the patients and greater than 24 hours in 56% of the patients. More than half

the patients had motor and sensory deficit at the time of admission. The majority of their emboli were discovered in the popliteal artery; a transfemoral approach was used in 60% of cases and a transpopliteal approach in 40%. In addition to doing an embolectomy, they had to perform a bypass operation in 14 cases, fasciotomy in 43 cases, and administer intra-operative streptokinase in two cases. Quality control was performed using continuous wave doppler and occasionally intra-operative angiography. Early amputation rate was 10% and subsequently limb salvage was recorded at 77% at one year. Complete recovery was recorded in 90% of the limbs salvaged, but there was a peroneal nerve palsy in 5% of cases. The one-month mortality rate was 12%. Factors that indicated a worse prognosis were location of the embolus in the abdominal aorta or the popliteal artery, delay in presentation or presentation with neurologic deficit. There was a small incidence (2%) of further emboli occurring requiring treatment within one year despite apparently adequate anticoagulation.

Trash foot
One form of embolism into the foot that is related to operative procedures is trash foot, which usually occurs following repair of abdominal aortic aneurysm. In a review, Kuhan and Raptis studied 1601 aortic reconstructions performed between 1976 and 1995 [15]. Trash foot occurred in 32patients (44 limbs, 23 cases following aortic aneurysm repair and 9 cases following an aortofemoral bypass for occlusive disease): 13.6% underwent an early amputation and 20% required a delayed amputation. Thirty-day mortality was 25%. There was no suggestion that surgical embolectomy improved the outcome. COMMENTS In a review of thrombolysis in lower extremity acute arterial occlusion, Patel et al. looked at the relative costs of surgery versus thrombolysis [16]. Although this study was aimed at acute on chronic thrombosis rather than embolic thrombosis, their conclusions were that surgery was more efficient and more cost effective than thrombolysis in treating acute lower limb ischemia. In a review of percutaneous aspiration embolectomy in acute embolic occlusion of the infrainguinal arteries, Wagner and Starck [17] reviewed 102 patients, 62% of whom had limb-threatening ischemia.

Eighty-six percent had cardiac disease as the source of their embolus. The initial clinical success rate of percutaneous aspiration embolectomy was 87% with major complications occurring in 9% of cases. The thirty-day mortality was 4%, comparing well with similar results with Fogarty catheter embolectomy. In their hands percutaneous aspiration embolectomy had a higher success rate and a lower mortality. They advocated percutaneous aspiration embolectomy because it was a simpler technique combining diagnosis with therapeutic procedures and enabling the treatment of tibial and pedal vessels as well as femoral and popliteal arteries. In a review of 159 cases of arterial embolism between 1991 and 1993, Wolosker et al. reported that the patients were almost equally split between male and female [18]. In most cases, the cause of the embolus was well established, with the most common cause (78%) being atrial fibrillation. The site of occlusion was most frequently the femoral artery, and although all patients presented with severe lower limb ischemia, none had gangrene on admission. Seventy percent of patients presented to the hospital within 24 hours, and all patients were submitted to lower limb embolectomy with a Fogarty catheter, 70% of which were done through the femoral artery. Fasciotomy was performed on 48 patients because of compartment syndrome. Nineteen patients died immediately after operation, most of these due to heart failure. Twenty-three (16%) of the 140 surviving patients were submitted to amputation after the occlusion of artery branches that had undergone embolectomy. One hundred twenty-seven limbs were preserved in the 117 patients (83%). Those patients with muscle tenderness, paralysis or ischemia lasting longer than 24 hours, had the worst results in relation to limb preservation. It was concluded that patients who presented with lower limb embolism who were in a good general condition and who did not have any necrosis of the limb or neurologic deficit, had a good outcome from the point of view of both their limb and their life. A very low complication rate was recorded from the surgical Fogarty catheter.

of sudden onset of acute ischemia with no evidence of cardiac disease and no evidence of peripheral vascular disease, surgical embolectomy with a Fogarty embolectomy catheter seems the most appropriate treatment. In many cases, however, the differential diagnosis between a surgical embolus and thrombosis of a previously diseased artery may be difficult to tell. If there is doubt, the patient should be submitted for urgent arteriography with a view to either thrombolysis, catheter embolectomy or surgical embolectomy. If the embolus is localized and small with relatively normal vessels, then surgical embolectomy may seem the most attractive way of removing the embolus. Completion angiography is mandatory before wound closure. If there is more extensive thrombus, and other catheter techniques may be more helpful, then a radiological solution is more appropriate. Care must be taken to pass an embolectomy catheter as few times as possible, minimizing the damage to the endothelium. In those institutions in which there is experience and radiological expertise in catheter techniques for removing emboli, the results are largely similar to those of surgical embolectomy. If these techniques fail, however, it is imperative that there is vascular surgical expertise to perform a surgical embolectomy or to do some form of reconstruction to revascularize the limb. Factors that affect limb salvage and survival are age of patient, comorbidity, delay in presentation, seniority and experience of the surgical staff, and the presence of neurosensory deficit at the time of presentation. The management of surgical embolus remains a high-risk and challenging procedure best dealt with by vascular surgeons.

279

R E F E R E N C E S 1 Campbell WB, Ridler BMP, Szymanska TH. Current management of acute leg ischemia: results of an audit by the Vascular Surgical Society of Great Britain and Ireland. Br J Surg 1998; 85: 1498-1503. 2 Golledge J, Galland RB. Lower limb intra-arterial thrombolysis. Postgrad AW/1995; 71: 146-150. 3 Kauhanen P, Perakyla T, Lepantalo M. Clinical distinction of acute and acute on chronic leg ischaemia. Ann Chir Gjnaecol 1995; 84: 335-338. 4 Pereira Barretto AC, Nobre MR, Mansur AJ et al. Peripheral arterial embolism. Report of hospitalised cases. Arq Bras Cardiol 2000; 74: 324-328. 5 Fogarty TJ, Cranley JS, Krause RJ et al. A method for extraction of arterial emboli and thrombi. Surg Gynecol Obstet 1963; 116: 241-244.

Conclusion
Embolization of the lower limb is a relatively uncommon but limb- and life-threatening problem requiring urgent surgical and radiological management. In those patients who have a clear-cut history

VASCULAR

EMERGENCIES

6 Nevelsteen A, Suy R. Arterial rupture and pseudoaneurysm formation secondary to the use of the Fogarty balloon catheter. Ada ChirBelg 1987; 87: 300-303. 7 Cronenwett JL, Walsh DB, Garrett H. Tibial artery pseudoaneurysms: delayed complication of balloon catheter embolectomy. J Vase Surg 1988; 8: 483-488. 8 Gloor B, Schopke C, Largiader J. Damage to the vessel wall by the Fogarty balloon catheter. Helv ChirActa 1994; 60: 749-752. 9 Schwarcz TH, Dobrin PB, Mrhdcka R et al. Balloon embolectomy catheter-induced arterial injury: a comparison of four catheters. / Vase Surg 1990; 11: 382-388. 10 Burgess NA, Scriven MWT, Lewis MH. An 11-year experience of arterial embolectomy in a district general hospital./,/? Coll Surg Edinb 1994; 39: 93-96. 11 Becquemin JP, Kovarsky S. Arterial emboli of the lower limbs: analysis of risk factors for mortality and amputation. Association Universitaire de Recherche en Chirurgie. Ann Vase Surg 1995; 9 (Suppl): S32-S38.

12 Huettl EA, Soulen MC. Thrombolysis of lower extremity embolic occlusions: a study of the results of the STAR Registry. Radiology 1995; 197: 141-145. 13 Berridge DC, Kessel D, Robertson I. Surgery versus thrombolysis for acute limb ischaemia: initial management. Cochrane Database Sjst Rev 2002; CD002784. 14 Ilic M, Davidovic L, Lotina S et al. Arterial embolisms of the lower limb extremities. Srp Arh Celok Lek 2000; 128: 276-280. 15 Kuhan G, Raptis S. "Trash foot" following operations involving the abdominal aorta. Aust N ZJ Surg 1997; 67: 21-24. 16 Patel ST, Haser PB, Bush HL, Kent KG. Is thrombolysis of lower extremity acute arterial occlusion cost-effective? J Surg Res 1999; 83: 106-112. 17 Wagner HJ, Starck EE. Acute embolic occlusions of the infrainguinal arteries: percutaneous aspiration embolectomy in 102 patients. Radiology 1992; 182: 403-407. 18 Wolosker N, Kuzniec S, Gaudencia A et al. Arterial embolectomy in lower limbs. Rev Paul Med 1996; 114; 1226-1230.

27
280

28
ACUTE THROMBOLYSIS OF PERIPHERAL ARTERIAL ANEURYSMS
LUDWIG KARL VON SEGESSER, BETTINA MARTY, PATRICK RUCHAT PHILIPPE GERSBACH, SALAH QUANADLI, DANIEL HAYOZ, ADAM FISCHER

Acute thrombosis of peripheral arterial aneurysms is a serious problem. In contrast to still patent peripheral arterial aneurysms, where rupture and consecutive bleeding can be an issue, the main problem after thrombotic aneurysm occlusion is, in general, the lack of distal runoff due to previous micro-embolization with progressive occlusion of the more distal arterial bed. Hence, standard aneurysm repair, i.e., resection in combination with interposition of an autologous or other graft, often results in recurrent occlusion. In order to increase the success rate under such circumstances, the therapeutic approach must be more specifically tailored to the problem of distal malperfusion and not only the issues related to the peripheral aneurysm per se. Pre-operative thrombolysis has been recommended [1] under such circumstances because it has the potential to not only recanalize a peripheral aneurysm, but also to re-open at least a part of the distal vascular bed, thus allowing for an improved runoff after surgical aneurysm repair. In this chapter we will discuss our experience with pre-operative thrombolysis for acute popliteal aneurysm thrombosis and its background.

^_ ZiO 281

for lower limb peripheral artery nrrliiinn in geuei ai ucciubioii in o-ptipral The Working Party on Thrombolysis in the Management of Limb Ischemia has produced an excellent document [2] in which the relevant issues for thrombo-

lysis in the management of lower limb peripheral arterial occlusion are well defined. These include the end oints of the studies the clinical P ' Presenta' tion of the patients ^ ischemia, the definition of terms used in conjunction with fibrinolytic therapies, the description of the catheter systems, the available thrombolytic agents, the recommended choice and dose of lytic agents (appendix A of this

VASCULAR

EMERGENCIES

document includes over 40 reported dosage schemes in catheter thrombolysis), the indications for thrombolysis, the centra-indications, adjunctive treatments, monitoring and complications.

Thrombolysis for acute ischemia after popliteal aneurysm occlusion (personal experience)
In our hands, routine work-up in patients with acute ischemia of the lower limbs includes, in addition to the clinical assessment, an ultrasonographic study prior to angiography. As a result, the presence of a peripheral arterial aneurysm is often known prior to the initial angiography and the latter can be planned accordingly. In this setting, angiography is usually started through a contralateral femoral approach and a cross-over maneuver is performed to reach the superficial femoral artery of the affected ischemic limb. As previously reported [3], the standard technique was to embed the catheter first in the distal part of the aneurysm and to inject an initial bolus of 200 000 international units (IU) of urokinase in order to achieve some degree of thrombus lacing for accelerated thrombolysis. After positioning the tip of the catheter in the proximal part of the thrombus, this maneuver is followed by a continuous infusion of 75 000 to 100 000 lU/h of urokinase. Simultaneously, intravenous heparin (standard dosage is 15 000 IU/24h for a 70 kg adult patient) is given systemically. In addition to clinical followup, angiographic assessment with distal catheter tip advancement in accordance to the progression of thrombolysis is performed every 6 to 24 hours. Throughout the fibrinolytic therapy, the patients are monitored in our intermediate care unit and the protocol includes systematic assessment of consciousness, noninvasive blood pressure measurement, quantification of diuresis, clinical status of the lower limb, the puncture site, as well as lab work with a focus on coagulation parameters and indicators of myolysis and renal function.

The most frequent risk factors in this group of patients were arterial hypertension (42%), active smoking (24%), hypercholesterolemia (22%), overweight (19%), pre-operative renal failure (14%), former smoking (7%; together with active smoking: 31%) and others. Additional aneurysms were detected at the level of the abdominal aorta in 26 patients (44%), iliac artery (20%), femoral artery (19%), thoracic aorta (7%) and elsewhere. Previous surgical procedures related to the cardiovascular system were documented in 61% of patients, including previous femoral artery procedures in 22%, aneurysm repair in 17% and coronary artery bypass grafting in 17%.

PATIENTS UNDERGOING THROMBOLYSIS

Thirteen patients (mean age 7610 years) with documented popliteal aneurysm by means of duplex ultrasound underwent thrombolysis prior to surgical aneurysm repair as described previously [3]. Duration of ischemia prior to hospital admission was less than 6 hours in 6 of 13 patients (46%) and more than 6 hours in 7 of 13 patients (54%). Twelve patients had ischemia of category Ha [4] with slight sensomotoric disturbance. One patient had severe ischemia category lib with an immediately threatened limb. A typical image of an initial angiography is shown in Fig. 1.

28
282

RESULTS
Mean duration of lysis was 3215 hours with infusion of 2 400 000 800 000 IU of urokinase. Thrombolysis resulted in a patent popliteal artery with one- or two-vessel runoff in 10 of 13 patients (77%). A typical angiographic control with repermeabilization of the popliteal axis after 22 hours of intra-arterial lysis is shown in Fig. 2. Fig. 3 shows the result after 44 hours of intra-arterial thrombolysis in the same patient. Clean inflow and outflow segments have been restored prior to surgical aneurysm repair. In contrast, Fig. 4 shows an angiographic control examination after 60 hours of intra-arterial thrombolysis without restoration of distal outflow. This type of persistent lytic failure was documented in three of 13 patients (23%). Ten patients underwent subsequent bypass grafting. A reversed saphenous vein was used in eight patients and a prosthetic graft was implanted in two patients. The distal anastomosis was performed at the infragenicular popliteal artery in seven patients and at the tibial artery in three patients. In seven

PATIENTS
All patients hospitalized in our institution are systematically registered in our patient analysis and in a tracking system. By retrospective analysis we identified 59 admissions to our unit for popliteal aneurysms (age 36 to 91 years).

ACUTE THROMBOLYSIS

OF PERIPHERAL ARTERIAL

ANEURYSMS

patients the outcome was successful. There were four bypass graft occlusions in three patients because of insufficient runoff. One patient with a venous conduit and one with a prosthetic conduit had lytic failure with absence of runoff. The third patient underwent venous bypass grafting twice but both failed. An overview of the outcome in these patients is given in Table I. For this small series of patients who underwent intra-arterial thrombolysis prior to surgical popliteal aneurysm repair, patency rate at one month was 68% and limb salvage rate was 83%, whereas early mortality was 15%. Mean follow-up was 15 9 months

and the 12-month patency rate was 46%. Three patients have completed the 24-month follow-up and had a patent femoropopliteal graft (two veins and one endoprosthesis).

General considerations
Popliteal aneurysm occlusion is a serious event. In the past, direct surgical aneurysm repair with intra-operative exploration of the distal vascular bed was routine. However, distal tibial artery occlusion

283

FIG. 1 Preprocedural angiosraphy showing distal superficial femoral artery occlusion.

FIG. 2 Angiographic control after 22 hours of intra-arterial thrombolysis showing repermabilization of the femoropopliteal axis and residual intraluminal thrombotic material (same patient as Fig. 1).

VASCULAR

EMERGENCIES

28
284
FIG. 3 Angiographic control after 44 hours of intra-arterial thrombolysis: clean inflow and outflow prior to popliteal aneurysm repair (same patient as Figs. 1 and 2). FIG. 4 Angiographic control examination after 60 hours of intra-arterial thrombolysis documenting failure of distal vessel repermeabilization: aneurysm repair is not indicated.

Success Successful lysis Failed lysis Aneurysm repair Occlusion Amputation Death
10/13

Failure

9/10 1/10 0/10 1/10

3/13 2/3 2/3 2/3 1/3

due to previous embolization from the popliteal aneurysm is not uncommon. In our opinion, intraprocedural attempts to re-open the distal vascular bed using balloon catheters and/or intra-operative lysis are not a reliable approach in patients with thrombosed popliteal aneurysms. The experience presented here shows that the mean duration of thrombolysis required for re-opening the arterial bed and achieving a clean inflow and outflow is usually more than 24 hours (see also Figs.l through 4) and requires considerable doses of intra-arterial urokinase application (superior to 2 000 000 IU). In contrast, intra-operative thrombolysis allowing for a few doses of 100 000 IU of urokinase over a

ACUTE THROMBOLYSIS

OF PERIPHERAL ARTERIAL ANEURYSMS

the limb salvage rate varied between 70% and 100% (Table II). Unfortunately, many parameters necessary for thorough analyses are not available in a substantial number of studies cited and therefore the conclusions made must be considered with caution. There are a number of arguments in favor of the concept proposed by Marty et al. [3] to use the success of thrombolysis as predictor for outcome of surgical repair of thrombosed popliteal aneurysm. The images reproduced here speak for themselves. Figs. 1 through 3 document successful thrombolysis prior to successful surgery and are in sharp contrast to Fig. 4, in which no reliable peripheral perfusion was obtained after 60 hours despite increased dosage of thrombolytic agents. For the latter case, the final outcome could be expected to be negative, and finally was. Although further validation of this approach is recommended for the future, for the time being, based on our experience and common sense, we have adopted the described approach for our current practice.

short period is quite limited in its potential and can therefore not reach the same degree of repermeabilization. As a matter of fact, Fig. 2 shows that the 22 hours of thrombolysis allows for repermeabilization of the distal vascular bed. However, continued thrombolysis over more than 40 hours is necessary to clean out the proximal femoral inflow (Fig. 3). The limb salvage rate of 83% reported here for acute occlusion of popliteal aneurysms must be compared with the results reported in the literature for similar patient populations. Bowrey et al. [5] report a limb loss rate of 24% in their group (pre-operative thrombolysis was used in 9 of 17 patients), whereas Dawson et al. [6] describe a median limb loss rate of 25% (range 9% to 36%). Marty et al. [3] compiled the results of 12 series from the literature in which pre-operative thrombolysis was performed prior to surgical popliteal aneurysm repair. In the cited reports, the number of patients included varied between 1 and 18. Early graft patency was between 73% and 100%, whereas

First author [ref.] Schwarz [7] Ferguson [8] Bowyer [9] Thompson [10] Ramesh [11] Carpenter [12] Varga [13] Garramone [14] Hoelting [15] Debing [16] Taurino [17] Greenberg [18] Personal experience [3]

Year
1984 1986 1990 1993 1993 1994 1994 1994 1994 1997 1997 1998 2001

Number of patients
1

Failure %
0

Mortality %
0 0

Early graft patency %

100

Limb salvage %

285

100

10
9

NA 33
0

NA
100 100

70
100
100 at 30 days 83 at 18 months 100 at 62 months 73 at 30 days
100

11
0 0 0 0 0

6* 12

17
0

NA
100

7 including 1* 18 including 9*
3 9 2 8

11
0 0 0

73
100 100 100

NA
0

100 at 5 years 100 at 5 years

87

13
0

NA
0

NA 83 68

6 including 1*
13

100 at 2 years 83 at 30 days

23

15

VASCULAR

EMERGENCIES

R E F E R E N C E S 1 Callum K, Bradbury A. ABC of arterial and venous disease. Acute limb ischemia, fir Afed/2000; 320: 764-767. 2 Anonymous. Thrombolysis in the management of lower limb peripheral arterial occlusion. A consensus document. Working Party on Thrombolysis in the Management of Limb Ischemia. Ajn/Carrfio/1998;81: 207-218. 3 Marty B, Wicky S, Ris HB et al. Success of thrombolysis as a predictor of outcome in acute thrombosis of popliteal aneurysms. / Vase Surg 2002; 35: 487-493. 4 Anonymous. Transatlantic Inter-Society Consensus (TASC) Working Group. Management of peripheral arterial disease. EurJ Vase Endovasc Surg 2000; 19: S115-S143. 5 Bowrey DJ, Osman H, Gibbons CP, Blackett RL. Atherosclerotic popliteal aneurysms: management and outcome in forty-six patients. EurJ Vase Endwasc Swrg2003; 25: 79-81. 6 Dawson I, Sie RB, van Bockel JH. Atherosclerotic popliteal aneurysm. BrJSurg 1997; 84: 293-299. 7 Schwarz W, Berkowitz H, Taormina V, Gatti J. The pre-operative use of intraarterial thrombolysis for a thrombosed popliteal artery aneurysm. J Cardiovasc Surg 1984; 25: 465-468. 8 Ferguson L, Paris I, Robertson A et al. Intra-arterial streptokinase therapy to relieve acute limb ischemia. / Vase Surg 1986; 4: 205-210. 9 Bowyer R, Cawthorn S, Walker W, Giddings A. Conservative management of asymptomatic popliteal aneurysm. Br J Surg 1990; 77: 1132-1135. Thompson J, Beard J, Scott D, Earnshaw J. Intra-operative thrombolysis in the management of thrombosed popliteal aneurysm. BrJ Surg 1993; 80: 858-859. Ramesh S, Michaels J, Galland R. Popliteal aneurysm: morphology and management. BrJ Surg 1993; 80: 1531-1533. Carpenter], Baker C, Roberts B et al. Popliteal artery aneurysms: current management and outcome. / Vase Surg 1994; 19: 65 - 72. Varga Z, Locke-Edmunds J, Baird R. A multicenter study of popliteal aneurysms. / Vase Surg 1994; 20: 171-177. Garramone R, Gallagher J, Drezner D. Intra-arterial thrombolytic therapy in the initial management of thrombosed popliteal artery aneurysms. Ann Vase Surg 1994; 8: 363-366. Hoelting T, Paetz B, Richter G, Allenberg J. The value of preoperative lytic therapy in limb-threatening acute ischemia from popliteal artery aneurysm. Am] Surg 1994; 168: 227-231. Debing E, Brande PVD, Tussenbroek FV et al. Intra-arterial thrombolysis followed by elective surgery for thrombo-embolic popliteal aneurysms. Acta Chir Belg 1997; 97: 137-140. Taurino M, Calisti A, Grossi R et al. Outcome after early treatment of popliteal artery aneurysms. IntAngio 1998; 17: 28-33. Greenberg R, Wellander E, Nyman U et al. Aggressive treatment of acute limb ischemia due to thrombosed popliteal aneurysms. EurJ Radial 1998; 28: 211-218.

10 11 12 13 14 15 16 17 18

28
286

29
ENDOVASCULAR APPROACH TO ACUTE ARTERIAL OCCLUSIONS
ANDREA STELLA, MAURO GARGIULO

Peripheral artery occlusion preventing blood flow to a limb is caused by embolism or the natural progression of peripheral arterial disease, usually atherosclerosis. Progression of arterial disease is characterized by the gradual extension of occlusion and is manifested by concomitant change in the clinical symptoms. The progression from claudication to rest pain, ischemic ulcers or gangrene may be the result of a single acute event exacerbating existing chronic ischemia. However, even if progression may be correlated with acute events, this clinical condition cannot be defined as true acute ischemia as it is induced by segmentary thrombosis. True acute limb ischemia (ALI) is the sudden and unheralded occlusion of a vessel. It is a limb- and often life-threatening condition. The thrombosis causing ALI does not remain segmentary but extends rapidly especially to the limb extremity. Mortality can be as high as 20 % and if treatment is delayed, amputation may be required in some 30 % of cases. The factors triggering ALI are not so much insufficient collateral pathways as the development of thrombosis, the severity of the underlying disease and, hence, the extension of peripheral atheromatous lesions, the occlusion site, the speed of onset and the presence of venous thrombosis. If these factors are accompanied by spasm and a fall in blood pressure, the clinical condition may rapidly become one of irreversible ischemia.

_^ &J 287

.treatment Options The various disease management procedures proposed such as direct surgery, thrombectomy and endovascular treatment have all shown fairly simi-

lar results, variations depending largely on the type of presenting thrombosis and time to treatment. After the introduction of the Fogarty catheter, many thromboses were treated with simple thrombectomy, a procedure that gave acceptable results in

VASCULAR

EMERGENCIES

29
288

the short term but high thrombus recurrence rates due to persistence of the primary cause of occlusion. Endovascular treatment, especially fibrinolytic drug therapy, achieves artery revascularization in most cases and successfully resolves most peripheral thromboses. Vessel revascularization usually allows identification of the cause of thrombosis and so enables a decision as to possible definitive treatment with endovascular procedures or open surgery. Despite the numerous studies conducted in the 1990s to establish the best treatment of acute peripheral ischemia, management regimens still vary considerably. Although partly due to the initial treatment carried out by the receiving hospital, these discrepancies are also due to the varying professional skills of the surgeon charged with the case. Moreover, not all hospitals have a vascular radiologist with the equipment necessary to perform endovascular procedures. These variables have been evidenced by a study carried out among members of the Vascular Surgical Society of Great Britain and Ireland (VSSGBI) [1]. Questionnaires were returned describing 539 events in 474 patients: 55% of patients admitted to the hospital were placed in the care of a vascular surgeon, 25% under a general surgeon with no vascular specialization and 20% under other medical staff. Forty percent of delays in presentation were attributable to the patient, the general practitioner or the transfer. Delayed referrals were the result of a late decision on the part of the admitting team (35%), or delay by the vascular team because no vascular surgeon or radiologist was present in the hospital. Because the most effective means of dissolving clot is activation of plasminogen bound to fibrin within the matrix of the clot, it is natural that current therapy approaches to vessel thrombolysis for acute ischemia are based on catheter delivery of the plasminogen activator directly into the thrombosis (Table I). If delivered into the thrombus, the plasminogen activator is protected from neutralization by circulating plasminogen activator inhibitors. Despite the local or regional delivery, ongoing intrathrombus infusion of lytic agents often results in systemic activation of plasminogen with breakdown of fibrinogen, clotting factors and other plasma protein [2]. Therefore lysis of the thrombus is obtained with direct infusion of thrombolytic agents or with percutaneous mechanical thrombectomy. Streptokinase was the most frequently used agent until the landmark article of McNamara and Fischer in 1985 documented improved results with locally administrated high-dose urokinase (UK) [3].

Appropriate medical treatment can achieve revascularization of an occluded artery in a procedure that can constitute a scheduled treatment of choice. Today lesions causing occlusion, stenosis or aneurysm can be successfully treated with endovascular procedures, with reduced trauma and overall risk of complications. Important additional goals of thrombolytic therapy are to provide a complete diagnosis of the disease, restore patency of branch vessels inaccessible to mechanical thrombectomy or direct surgery, and avert major vascular reconstruction surgery in favor of a limited, less extensive procedure or endovascular therapy.

Arterial thechniques for peripheral thrombolysis

After Dotter's report on the use of the coaxial catheter to restore flow to the extremity of the limb [4], Gruentzig in 1974 introduced a double-lumen balloon catheter which revolutionized endovascular

Thrombolytic therapy Streptokinase Urokinase Recombinant tissue-type plasminogen activator (rt-PA) Single-chain urokinase-type plasminogen activator (scu-PA) Acylated plasminogen-streptokinase activator complex (APSAC) Percutaneous mechanical thrombectomy Percutaneous aspiration thrombectomy Fullback thrombectomy and trapping Rotational and hydraulic recirculation thrombectomy Non-recirculating mechanical thrombectomy

ENDOVASCULAR APPROACH TO ACUTE ARTERIAL therapy [5]. Introduced in the late 1980s, percutaneous transluminal angioplasty can now be converted to a successful procedure with the placement of a stent. In 1974 Dotter reported the use of streptokinase for selective clot lysis [6]. When treating patients with ALI, most interventional radiologists choose to access the vascular system through the common femoral artery contralateral to the symptomatic limb. This approach facilitates thrombolysis while not precluding subsequent procedures on the affected limb. The inferior epigastric artery and deep circumflex iliac artery are the landmarks to locate the inguinal ligament. Access is through the common femoral artery below the inguinal ligament. A puncture site that is too proximal may lead to retroperitoneal hematomas, dangerous complications of thrombolytic therapy. To avoid this, a basic arteriogram should be performed from the renal level up to both femoral arteries. Once the arterial occlusion has been identified, it is important to determine if the clot is lysable. McNamara described the guide wire transversal test in which an attempt is made to pass a floppy-tipped guide wire into the clot. If the guide wire cannot be easily advanced into the clot, there is only a 10% likelihood of success [7]. Smith et al. then showed that an initial 2- to 4-hour lytic infusion at the proximal edge of the clot will soften it up sufficiently to allow the passage of a guide wire through and subsequently and increase the likelihood of successful thrombolytic therapy [8]. Depending on the occlusion site, antegrade or retrograde puncture may be performed: antegrade access is more suitable for femoropopliteal thrombosis, whereas retrograde access is indicated for iliac and aortic artery occlusions. Acute iliac occlusions are treated from the original puncture site. Superficial artery and popliteal artery occlusions can be treated from either the initial puncture point or from the affected limb. Using the original puncture site eliminates the risk of bleeding from an additional puncture point, which would render procedures such as angioplasty and stenting in the popliteal or tibial area more difficult. For occlusions below the common femoral artery, an antegrade puncture can be performed. This technique can be challenging, however, especially in obese patients, as the guide wire preferentially advances into the profunda femoral artery. Several techniques have been developed to conduce the guide wire into the superficial femoral artery. After documenting that

OCCLUSIONS

the common femoral artery has been effectively entered, the needle should be redirected to the contralateral wall and the wire re-advanced. Alternatively, the floppy tip of a moveable core wire can be advanced into the profunda and allowed to herniate into the superficial femoral artery. A third method is to exchange the needle for a directional catheter, retract it under fluoroscopy and redirect a wire into the superficial femoral artery. Once the wire is placed in the superficial femoral artery, a vascular sheath is introduced to facilitate the procedure and reduce local complications. Infusion of 5000IU heparin is recommended to prevent thrombus formation induced by partially obstructed flow and catheter manipulations. After assessing the clot with the guide wire transversal test, the next step is to advance a catheter with multiple side holes into the occluded artery. The catheter should be positioned so that the proximal side hole is at or close to the top of the clot and the distal side hole is just proximal to the end of the clot. Follow-up arteriography should be performed as necessary and when practical, i.e., in the afternoon if the therapy was started in the morning, or the next morning if commenced in the afternoon. Patients often may experience increased pain during lysis. This is due to the embolization of small clots into the peripheral vascular field. Symptoms are short lived, however, and resolve with prolonged infusion. When the angiographic follow-up identifies emboli down the leg, the catheter should be advanced into the tibial arteries and infusion continued until the emboli dissolve. Inability to pass the guide wire through the occlusion is indicative of either atherosclerotic disease, neointimal fibroplasias or a highly organized and calcified thrombus that will not respond to a lytic agent. Therefore primary operative reconstruction is recommended in these cases. Certain safety guidelines and important technical aspects are fundamental during intra-arterial thrombolytic therapy. Patients should be carefully selected and include only those who are cooperative and able to endure the prolonged bed rest involved (24 to 36 hours) with minimal movement. All invasive procedures and needle punctures should be kept to a minimum. When thrombolytic therapy is selected, it is preferable, when technically feasible, to puncture and cannulate through the contralateral groin to minimize bleeding around the catheter insertion site. In the event of an endovascular protocol being adopted, many factors are predictive of the success

29_
289

VASCULAR

EMERGENCIES

of intra-arterial thrombolysis: the guide wire passing into the clot, short duration of the occlusion, localization of the occlusion (iliac, femoral, or popliteal) and visualization of distal vessels. After successful infusion, an arteriogram should be performed and blood studies repeated. If the cause of occlusion is identified it can be managed immediately with standard interventional techniques. If percutaneous correction is not possible, systemic anticoagulation should be continued until the definitive operative procedure is completed. The technique selected to repair the lesion may have substantial impact on long-term results.

Complications of thrombolytic therapy

Complications of thrombolytic therapy include hemorrhage, allergic reactions, embolism, stroke, reperfusion syndrome and arteriography complications. As numerous complications result from the invasive nature of intra-arterial thrombolytic therapy, there are several absolute and relative contraindications that must be complied, most importantly severe ischemia with neurologic changes and devitalized limb with irreversible ischemic damage (Table II). Patients with profound ischemia characterized by pallor, paresis and decreased sensation cannot sustain the length of time required for thrombolytic therapy to have an effect and are best treated with emergency surgical intervention. Bleeding is the most common and feared complication of thrombolytic therapy; it occurs as the result of the active lytic state, but can also result from the induced coagulopathy. The systemic effect of lytic agent and heparin therapy used to reduce the risk of pericatheter thrombosis is monitored by the fibrinogen level, prothrombin time, partial thromboplastin time, and fibrin degradation products during infusion. Although bleeding can occur in any patient receiving lytic therapy, patients with significant hypofibrinogenemia appear to be at greatest risk of a serious hemorrhagic complication. If the fibrinogen level drops below 100 mg/dL, infusion should be slowed or stopped to allow the restoration of circulating fibrinogen and clotting factors. However, the efficacy of catheter thrombolysis is evaluated angiographically and clinically and does not depend on laboratory markers of systemic fibrinolysis. In general, laboratory values do

not correlate with hemorrhagic complications or the efficacy of lysis. Some bleeding occurs because of the trauma during invasive diagnostic or therapeutic procedures. Since catheter-directed thrombolysis is by definition an invasive procedure, an albeit minimal number of bleeding complications is inevitable. The National Audit of Thrombolysis for Acute Limb Ischemia (NATALI) database reports an 8.1% incidence of major hemorrhage compared to 5.1% reported for the series reviewed by Berridge et al. and 12.5% reported by TOPAS [9,10]. Although intra-arterial urokinase has been demonstrated to increase the salvage of acutely ischemic skeletal muscle in an experimental model, thrombolytic therapy in the setting of acute ischemia introduces certain potential complications. Myonephropathic syndrome (acidosis, hyperkalemia and myoglobinuria), compartment syndrome and

29
290

Absolute Active bleeding source Cerebrovascular accident or other intracranial process (< 2 months) Relative Recent major surgery, eye surgery, organ biopsy, postpartum status Recent trauma Severe and uncontrolled hypertension Coagulation deficits Pregnancy Hemorrhagic retinopathy Hepatic failure Endocarditis Absolute contra-indications for intra-arterial thrombolytic therapy Severe ischemia with neurologic changes Devitalized limb with irreversible ischemic damage

ENDOVASCULAR APPROACH TO ACUTE ARTERIAL hemorrhagic necrosis may result when profoundly ischemic tissue is reperfused after thrombolysis. Occasional complications of arteriography and arterial cannulations, such as arterial wall injury, mural dissection, or thrombosis, must also be taken into account. As the intra-arterial thrombus is lysed and blood flow through the vessel restored, embolization of the clot into the distal vasculature may occur. The numerous reports in the literature show a wide variation of distal embolization incidence rates ranging from 16% to 65% [11,12]. Belkin and O'Donnell [13] have characterized the resulting clinical picture as "the storm before the calm" since these patients usually suffer an acute worsening of their ischemic symptoms that almost always resolves with continued thrombolytic infusion. Thus, when the patient suffers sudden clinical deterioration with distal embolization, the reflex to discontinue therapy should be resisted. Pericatheter thrombosis remains a significant clinical problem when the patient does not receive concomitant heparinization. Before heparin use, pericatheter thrombus occurred in 36% of cases. Today the routine administration of heparin has reduced this to less than 5% [11].

OCCLUSIONS

Results of thrombolytic therapy

Some patients develop acute leg ischemia as a preterminal event. Recognition of these patients and thoughtful conservative treatment are part of good medical practice. In these cases, the association between nonviable limbs and mortality counsels appropriate management that does not include amputation. Conversely, other patients presenting with unsalvageable legs are best managed with early amputation. Publications on the management of acute leg ischemia tend to reflect the practices of specialist units. The management of acute ischemia has changed radically in recent years. Many general surgeons now rightly consider themselves insufficiently experienced to manage the condition. The results of the current treatment reflect technical advances such as the development of multiside-hole catheters ensuring infusion of high-dose UK directly into the thrombus. Assessing the suitability of thrombolytic management of an ischemic

leg must be based on a clear definition of the severity of the presenting ischemia and the understanding that thrombolytic therapy alone is seldom sufficient. The ischemia scoring system generally used in the literature refers to Rutherford's threecategory classification: viable (not immediately threatened), threatened (salvageable if promptly treated) and irreversible (major tissue loss, amputation regardless of treatment) [14]. In 1998 an intercontinental consensus paper was published on the use of thrombolytic therapy in occlusive peripheral arterial disease affecting lower limbs. These authors broadened category II into a) salvageable if threatened, and b) salvageable if threatened as emergency [15]. Although many randomized trials compare surgery with lysis, the results are unfortunately inconclusive. In one severely ischemic group of patients, while mortality was higher in patients undergoing initial surgery, no difference was seen between the two groups in regard to limb salvage. In the single-center Rochester study, the rate of limb salvage was identical at 80%. More cardiopulmonary complications were reported in patients taken directly to surgery, which explains the higher long-term mortality rate in the surgical group: at 12 months follow-up, 84% of patients randomized to UK were alive compared to only 58% of patients randomized to primary surgery. Mortality was comparable in patients without cardiopulmonary complications, averaging approximately 10% at 12 months irrespective of treatment. In conclusion, thrombolytic therapy was associated with a reduction in cardiopulmonary complications, improved survival and limb salvage [16]. There were no differences in mortality and amputation rates when lysis and surgery were compared in the 392 patients included in the Surgery versus Thrombolysis for Ischemia of the Lower Extremity (STILE; three treatment groups: rt-PA, UK or primary operation) study. However, this study contained many patients with late presentation [17]. Post hoc analysis of surrogate end points, such as death and amputation, hemorrhage and peri-operative complications favored surgery. The rate of major amputation was higher in native arterial occlusions treated with thrombolysis: 10% thrombolysis versus 0% surgery at one year. Recurrent ischemia was also more common in the lysis group. When recent occlusions (below 14 days) were considered separately, however, fewer amputations were reported in the patients who underwent lysis. This study suggests that thrombolysis provides greatest benefit to patients

291

VASCULAR

EMERGENCIES

29
292

with acute occlusion of bypass graft if treated before day 14. In this study, UK and rt-PA are equivalent. ALI (within 14 days) was best treated with thrombolytic therapy, as amputation rates were lower when infusion was instituted promptly. Primary surgical intervention was seen to be more effective in patients with recent ischemia (above 14 days) and native arterial occlusions which respond poorly to thrombolytic therapy. The Thrombolysis or Peripheral Arterial Surgery (TOPAS) study, conducted at 113 centers in the United States, Canada and Europe and including patients with lower limb arterial occlusion (within 14 days), reports similar amputation and mortality rates at discharge and at one year in 544 patients, (68.2% and 68.8% for UK and surgical patients, respectively) [18]. At 6 months follow-up, 31.5% of patients undergoing a thrombolytic percutaneous procedure were amputation-free. This figure fell to 25.7% at one year. In conclusion, 4000 lU/min recombinant UK (rUK) for 4 hours followed by 2 000 lU/min would appear to be a safe treatment regimen, achieving limb salvage and patient survival rates similar to peripheral arterial surgery. Generally, concomitant heparin therapy increases the risk of hemorrhagic complications: major bleeding occurred more frequently with thrombolysis (12.5% versus 5.5%), with intracranial hemorrhage developing in 1.6% of the thrombolytic group. Nevertheless the TOPAS findings confirm that ALI may be treated with catheter-directed thrombolysis, achieving similar results while avoiding the need for open surgery in many patients. Although the STILE and TOPAS studies both claimed better results for graft occlusions compared to native vessel occlusions, adjuvant procedures were usually required to maintain patency.

Mechanical thrombectomy
Percutaneous mechanical thrombectomy (PMT) avails itself of a heterogeneous group of devices and techniques designed to clear intravascular thrombotic occlusion using a combination of mechanical dissolution, fragmentation and aspiration. While thrombolytic therapy is the standard treatment for occlusion, it is associated with a definite risk of hemorrhage and embolic complications. Consequently PMT may be required in a number of patients. PMT devices are of variable complexity, design, mechanism of action, efficacy and cost. Several ones are

now gaining favor among European practitioners and have been approved by the Food and Drug Administration. Many PMT devices work with recirculation, trapping and lysing the thrombus. The first device to use hydrodynamic recirculation was the Kensey (Trac-Wright) athero-ablative catheter. Dr Kurt Amplatz of the University of Minnesota has designed several recirculation prototypes using different techniques. There are now two categories of recirculation mechanism: rotational recirculation in which the vortex is produced by high-speed rotation, and hydraulic recirculation in which highspeed fluid jets create a stagnation pressure gradient (venturi effect), trapping and dissolving the evacuated thrombus. Another class of PMT techniques consists of devices whose primary function is direct thrombus fragmentation and aspiration. However, these simple devices with negligible recirculation carry high risks of embolization and vascular injury. A further class of PMT instrumentation uses ultrasound energy for selective thrombus ablation. The efficiency of mechanical thrombectomy could be defined as the percentage weight or volume of thrombus cleared or fragmented by the treatment. This effect depends on technique, thrombus volume, composition, organization, wall adherence, the extent of atherosclerotic disease and the coagulation activity of the patient. Another problem is vessel wall damage: there is no standardized system for acute vascular injury. The chronic manifestations of vascular injury from transcatheter applications include intimal hyperplasia, atherosclerosis, pseudoaneurysm and fistula. These complications seem to occur less as compared to the Fogarty balloon catheter. Catheter clot lysis is now gaining increasing acceptance as it becomes evident that not all cases of critical ischemia are best treated surgically or with thrombolysis alone. Thromboaspiration was first developed by Starck et al. in 1985 to treat infrainguinal embolic occlusions [19]. A 5F to 8F catheter is introduced into an 8F sheath and advanced to the thrombus, where aspiration is performed manually with a syringe. While aspiration is certainly more rapid (less than one hour), it does not carry the contra-indications of thrombolysis and is also suitable for long occlusion. It can only be used on recently formed thrombus (within a day). Desgranges et al. [20] report a success rate of 43% using thrombus aspiration alone and 81% when thrombus aspiration is associated

ENDOVASCULAR APPROACH TO ACUTE ARTERIAL

with thrombolysis and lesion repair using angioplasty or surgery. Castaneda et al. [21] report promising results in their study comparing the Fogarty embolectomy catheter, the Arrow-Trerotola peripheral thrombectomy device and the MTI-Castaneda over-the-wire brush. While all devices proved effective in removing the thrombus from the iliac arteries treated, they all caused lesions extending into the media. Other multicenter clinical studies and case reports have shown factors such as lateral wall pressure, catheter size, catheter tip sharpness, balloon eccentricity, fluid- versus gas-filled balloons, syringe size, velocity of catheter motion and presence of blood in the vascular lumen to be regularly underestimated as a cause of vascular injuries that are incriminated for failure. In 1997 Henry et al. [22] reported their results with the Hydrolyser catheter: technical success was obtained in 79% of all the primary arterial thromboses treated. While this catheter has the advantage that it can be used during the same endovascular procedure, it does not allow complete debridement of the artery and must be associated with complementary interventional and/or fibrinolytic procedures. Hopfner et al. [23] reported their results in 51 patients with the shredding embolectomy thrombectomy catheter (S.E.T. catheter, now OASIS thrombectomy system), performing hydromechanical thrombectomy (venturi effect). The procedure entails the introduction of an ipsilateral antegrade 8F sheath and monitoring of systemic blood pressure and heart rate. The S.E.T. catheter proved completely successful in 12% of the patients, requiring a mean time of 6.4 minutes. Rapid reduction of thrombus mass was achieved in the other patients. Revision procedures were required in 88% and peripheral embolization in 10%, compared with the 5% reported in the literature for thrombolysis. The above results justify the still limited use of these devices, even though removing part of the thrombus remains an interesting theoretical proposal. Further technical improvements to these devices should make them suitable for regular use, thereby cutting the time and cost of treating acute thrombosis.

OCCLUSIONS

General considerations
Many studies show that lytic therapy offers an effective therapeutic option for the treatment of ALL The principal objective of thrombolytic therapy is

to remove the occlusion, thus restoring the vascular patency and to make the diagnosis of the disease. There are many potential attractions of thrombolytic therapy over an initial surgical approach because these findings confirmed that ALI could be treated with catheter-directed thrombolysis, achieving similar results but avoiding the need for open surgery in many patients. The distal arterial bed can be cleared of small thrombi more efficiently with a balloon preparing the artery for final treatment. Intra-arterial UK infusions have been proposed as the first-choice treatment for acute arterial and graft thrombosis [24]. Immediate success rates with intrathrombus UK can be as high as 88% in native arteries, and meta-analysis of the English literature reveals an average success rate of 85% in the most recent series with a complication rate of 20% [25]. Nevertheless, acceptance of thrombolytic therapy as the procedure of choice is not yet universal. One of the major potential advantages of thrombolysis is to limit the extent of subsequent major surgical therapy and, if possible, to conduce the patient into the endovascular treatment. In the experience of Suggs et al. [26], in 23% of their patients it proved to be the sole therapy required for limb salvage, and in a number of other cases successful lysis uncovered stenotic lesions that were amenable to angioplasty and stenting. Long-term follow-up revealed that successful thrombolytic therapy with or without a required subsequent endovascular procedure resulted in durable arterial patency. Similar results have been reported by Wholey et al. [27], with a one-year patency rate of 87% for native vessels following successful lysis and angioplasty in patients with ischemia of less than 14 days duration. The STILE and TOPAS trials did not analyze the long-term patency of successfully lysed native arteries. Moreover Faggioli et al. [25] clearly reported data supporting the need of adjunctive therapy to improve long-term patency, confirming the failure of thrombolysis as sole therapy. Clinical factors such as duration of ischemia and severity of limb ischemia have been suggested as important aspects of lytic results. A post hoc analysis of the STILE trial (entered only patients with ischemia within 14 days) which stratified all patients by duration of ischemia, showed the significant benefit of thrombolysis for the acutely ischemic limbs. Nevertheless, amputation rate was similar in patients with ischemia longer than 14 days and patients who had a shorter period of ischemia: 14% versus 10% [17].

293

VASCULAR

EMERGENCIES

When evaluating the thrombolysis, the length of the occlusion has an important impact on the results: a multifactorial analysis of the TOPAS data showed that lesions with a length greater than thirty centimeters predicted better amputation-free survival than surgery.

Conclusion
In conclusion, patients with significant motor or sensory deficits, femoropopliteal occlusion and dia-

betes must be treated by emergent angiography and surgery if possible. Unfortunately the results for these patients tend to be worse than for those with less profound ischemia. Thrombolysis should have a role as a treatment option for ALI secondary to native artery occlusion, for patients with extensive thrombus rapidly occluding the distal bed, and for young patients with minimal atherosclerotic disease. In addiction, short focal lesions are best treated with an endovascular approach reducing the impact of the surgery.

R E F E R E N C E S

\/ O *** *^ 004

1 Campbell WB, Ridler BM, Szymanska TH. Current management of acute leg ischaemia: results of an audit by the Vascular Surgical Society of Great Britain and Ireland. BrJ Surg 1998; 85: 1498-1503. 2 Comerota AJ, Malone MD. Simplified approach to thrombolytic therapy of arterial and graft occlusion. In: YaoJST, Pearce WH (eds). Practical vascular surgery. Stamford, Appleton & Lange, 1999: pp 321-334. 3 McNamara TO, Fischer JR. Thrombolysis of peripheral arterial and graft occlusions: improved results using high-dose urokinase. AJR Am J Roentgenol 1985; 144: 769 - 775. 4 Dotter CT, Judkins MR Transluminal treatment of arteriosclerode obstruction: description of a new technique and preliminary report of its application. Circulation 1964; 30: 654. 5 Gruntzig A, Hopff H. Percutaneous recanalization after chronic arterial occlusion with a new dilator-catheter (modification of the Dotter technique) Dtsch Med Woehenschr 1974; 99: 2502-2511. 6 Dotter CT, Rosch J, Seaman AJ. Selective clot lysis with lowdose streptokinase. Radiology 1974; 111: 31-37. 7 McNamara TO, Bomberger RA. Factors affecting initial and sixmonth patency rates after intra-arterial thrombolysis with highdose urokinase. Am J Surg 1986; 152: 709-712. 8 Smith DC, McCormick MJ, Jensen DA, Westengard JC. Guide wire traversal test: retrospective study of results with fibrinolytic therapy. / Vase Interv Radial 1991; 2: 339-342. 9 Campbell B, Kinsella D. Complications of thrombolytic therapy and their avoidance. In: Greenhalgh RM, Powell JT, Mitchell AW (eds). Vascular and endovascular opportunities. London, W.B. Saunders, 2000: pp 505-516. 10 Berridge DC, Makin GS, Hopkinson BR. Local low dose intraarterial thrombolytic therapy: the risk of stroke or major haemorrhage. BrJ Surg 1989; 76: 1230-1233. 11 Belkin M, Belkin B, Bucknam CA et al. Intra-arterial fibrinolytic therapy. Efficacy of streptokinase vs urokinase. Arch Surg 1986; 121: 769-773. ' 12 O'Donnell TFJr, Coleman JC, Sentissi J et al. Comparison of direct intra-arterial streptokinase to urokinase infusion in the management of failed infrainguinal ePTFE grafts. In: Veith FJ (ed). Current critical problems in vascular surgery. St. Louis, Quality Medical Publishing, 1989: pp 80-88. 13 Belkin M, O'Donnell TFJr. Complications of thrombolytic therapy. Bernhard VM, TowneJB (eds). Complications in vascular surgery. St. Louis, Quality Medical Publishing, 1991: pp 433-441. 14 Rutherford RB, Flanigan DP, Gupta SK et al. Suggested stan-

15

16

17 18

19 20

21 22 23 24 25 26

27

dards for reports dealing with lower limb ischemia. / Vase Surg 1986; 4: 80-94. Anonymous. Thrombolysis in the management of lower limb peripheral arterial occlusion. A consensus document. Working party on thrombolysis in the management of lower limb ischemia. Am] Cardiol 1998; 81: 207-218. Ouriel K, Shortell CK, DeWeese JA et al. A comparison of thrombolytic therapy with operative revascularization in the initial treatment of acute peripheral arterial ischemia. J Vase Surg 1994; 19: 1021-1030. Anonymous. Results of a prospective randomised trial evaluating surgery versus thrombolysis for ischemia of the lower extremity, the STILE trial. Ann Surg 1994; 220: 251-268. Ouriel K, Veith FJ, Sasahara AA. A comparison of recombinant urokinase with vascular surgery as initial treatment for acute arterial occlusion of the legs. Thrombolysis or Peripheral Arterial Surgery (TOPAS) Investigators. NEngJMed 1998; 33 8: 1105-1111. ' Starck EE, McDermottJC, Crummy AB et al. Percutaneous aspiration thromboembolectomy. Radiology 1985; 156: 61-66. Desgranges P, Ernenwein D, Kobeiter H et al. Ischemies aigues spontanees des membres inferieurs. In: Kieffer E (ed). Urgences vasculaires non traumatiques. Paris, AERCV, 1998: pp 163-179. Castaneda F, Li R, Patel J et al. Comparison of three mechanical thrombus removal devices in thrombosed canine iliac arteries. Radiology mi; 219: 153-156. Henry M, Amor M, Henry I, Alloaoui M. Thrombectomy with the hydrolysing catheter. A propos of 50 cases. Arch Mai Coeur Vaiss 1997;' 90: 797-804. Hopfner W, Vicol C, Bohndorf K, Loeprecht H. Shredding embolectomy thrombectomy catheter for treatment of acute lower-limb ischemia. Ann Vase Surg 1999; 13: 426-435. McNamara TO, Bomberger RA, Merchant RF. Intra-arterial urokinase as the initial therapy for acutely ischemic lower limbs. Circulation 1991; 83 (2 suppl): 1106-119. Faggioli GL, Ricotta JJ. Thrombolysis therapy for lower extremity arterial occlusion. Ann Vase Surg 1993; 7: 297-302. Suggs WD, Cynamon J, Martin B et al. When is urokinase treatment an effective sole or adjunctive treatment for acute limb ischemia secondary to native artery occlusion? Am J Surg 1999; 178: 103-106. Wholey MH, Maynar MA, WTioley MH. Comparison of thrombolytic therapy of lower-extremity acute, subacute, and chronic arterial occlusions. Cathet Cardiovasc Diagn 1998; 44: 159-169.

30
THROMBOLYSIS FOR OCCLUSION OF BYPASS GRAFTS
ROBERT HINCHLIFFE, BRUCE BRAITHWAITE, BRIAN HOPKINSON

The standard surgical management of bypass graft occlusion has been thrombo-embolectomy with additional procedures performed as necessary to treat any underlying occlusive lesion. This approach yields rather disappointing results. Alternatively, some authors advocate graft removal and replacement in an attempt to improve the outcome. The surgical management of bypass graft occlusion is associated with significant morbidity and mortality. Many patients are simply too frail to undergo further major arterial bypass surgery. With an aggressive surgical approach only 12% of patients will be alive at five years HI Thrombolysis was popularized in the 1970s by Dotter as an endovascular method of restoring patency in native artery occlusions. The minimally invasive nature of thrombolysis and its potential to reveal occlusive lesions amenable to endovascular therapy and to restore flow in thrombosed distal vessels became established. Later, reports emerged regarding its use in the management of bypass graft occlusion.

The evidence
Despite the widespread use of thrombolysis in lower limb ischemia, a recent systematic review of published data found only ten reports of randomized, controlled trials [2]. The majority concern thrombolysis of native arterial occlusions. Only one of these publications, a subgroup of patients identified from the Surgery versus Thrombolysis for Ischemia

of the Lower Extremity (STILE) trial, specifically reported the results of surgery or thrombolysis in patients with occluded bypass grafts of less than six months duration [3]. Other studies of thrombolysis for bypass graft occlusions are retrospective and consist of small groups of heterogeneous patients. In addition to these studies, a large national database of thrombolysis, the National Audit of Thrombolysis in Acute Leg Ischaemia (NATALI), was maintained in the United Kingdom.

VASCULAR

EMERGENCIES

Delivery and type of lytic agent

The delivery of the thrombolytic agent to its target site may be achieved via a systemic route or directly into the occluded bypass graft. The former technique has been abandoned in peripheral arterial (both native and bypass) occlusions due to the high number of bleeding problems and the more favorable results achieved by direct intrathrombus delivery. The ability to catheterize the graft and cross the occlusion with a wire has a significant effect on outcome. Failure to catheterize an occluded graft occurred in 39% of patients in the STILE trial [3]. Sullivan et al. found that difficulties catheterizing an occluded graft could be negotiated by placing the tip of the thrombolysis catheter at the origin of the graft thus creating a stump through which a wire could be manipulated in to the occlusion [4]. An alternative technique to facilitate delivery of intrathrombus lysis is to perform direct graft puncture. This technique is particularly useful in patients with extra-anatomical prosthetic bypass grafts. Prophylactic antibiotics are required where direct puncture of prosthetic grafts is undertaken. A number of thrombolytic dosing regimens have been reported. In Braithwaite's report of native arterial thrombolysis, there was no difference with respect to limb salvage or complications following lowor high-dose therapy, however, more secondary procedures were performed in those patients receiving the high-dose regimen [5]. There is no evidence to suggest which may be the preferred technique for bypass graft occlusion. The pulse-spray technique achieves thrombolysis through the delivery of lytic agent under pressure via a multi-side-hole catheter. The pressure created contributes to the mechanical lysis of thrombus. In the study by Hye et al. using pulse-spray for bypass graft occlusion, only 27% of patients required an infusion of thrombolytic agent outside the radiology suite [6]. Lysis using pulse-spray alone took 118 minutes, whereas requirement for an infusion increased the time to 13.5 hours. In the Thrombolysis or Peripheral Arterial Surgery (TOPAS) trial, all native and bypass graft occlusions were pooled together but higher doses of lytic agent were required for patients with bypass graft rather than native arterial occlusions. There is no clear evidence to suggest which thrombolytic agent should be used in graft occlusion. In one study streptokinase was successful in

clearing thrombus in only 57% of cases, whereas in other studies, almost 90% of grafts have been successfully opened with tissue plasminogen activator (tPA) [7,8]. In the STILE trial there was no difference between the efficacy of urokinase and rt-PA with respect to the composite outcome measure. Heparin is a useful adjunct during thrombolysis although it may increase the hemorrhagic complications. In our institution it is infused alongside the indwelling sheath to prevent peri-catheter thrombus during infusion of lytic agent.

Outcome of graft thrombolysis and prognostic factors

There is a significant shortfall of patients who fail to undergo graft catheterization or in whom thrombolysis is prematurely terminated due to complications. The United Kingdom NATALI database demonstrated that successful lysis was achieved in 70% of attempts. Complete lysis was achieved in 48% and partial lysis in 22% of patients. The one-year patency of lysed grafts was 33%. On an intentionto-treat basis, this figure dropped to 20% [9]. A 30% one-year patency following initial successful graft thrombolysis is remarkably similar among disparate studies. The role of thrombolysis is primarily to restore flow and reveal the underlying lesion that contributed to the occlusion. Lesions may be amenable to endovascular treatment such as angioplasty and are multiple in a third of cases [6]. Gardiner et al. found that one-year patency rates were better in those patients who had a correctable underlying lesion (86% versus 37%) [10]. Alternatively, lysis may direct traditional surgical therapy and allow a minimal access approach. Indeed, some studies have shown that intimal hyperplasia or vein graft stenosis may be better managed by open surgery rather than endovascular techniques [11]. One study found a tendency for a higher initial recanalization rate in patients with prosthetic bypass grafts, although these did not reach statistical significance [4]. Immediate success may be greater with suprainguinal grafts than their infrainguinal counterparts [12]. Arterial bypass grafts that occlude in the first postoperative year are usually a result of either technical error or an incorrect choice of operation.

296

THROMBOLYSIS

FOR OCCLUSION OF BYPASS

GRAFTS

Consequently, it may be expected that thrombolysis will have less favorable results in patients with grafts that are less than one year old. Nackman et al. were able to demonstrate by multivariate analysis that grafts less than 12 months old were a significant predictive factor in the failure of thrombolysis [13]. If lysis was undertaken successfully in grafts less than one year old, 44% required an amputation, rising to 69% where lysis has failed. The duration of ischemia has a major impact on the outcome of thrombolysis. Thrombolysis is more effective in acute graft occlusion. Comerota et al. demonstrated a clear reduction in major amputation at one year (48% versus 20%) in patients undergoing thrombolysis compared to surgery where the duration of ischemia was less than 14 days [3]. Conversely, the same data from the STILE trial was unable to detect any difference if the ischemia had been present for greater than two weeks. At one year, there was little difference in composite clinical outcome whether the patient was undergoing thrombolysis of a graft above (80%) or below (81%) the inguinal ligament. A review of the available results in the literature by Browse et al. appeared to demonstrate an improved outcome in patients undergoing thrombolysis for suprainguinal bypass occlusion compared with those obtained from infrainguinal lysis [14]. Galland's analysis of the NATALI database was unable to support this finding. There was no difference in patency between the two sites [9]. Thrombolysis may restore the patency of occluded collateral vessels and improve lower limb circulation without fully opening the graft. This may be sufficient to permit limb salvage. In one study, in excess of 60% of limbs were salvaged at one year even though only 30% of grafts were patent [13]. However, long-term patency is significantly greater if lysis was complete. Galland et al. [9] demonstrated a 12-month patency of 39% where graft patency had been restored with at least one vessel runoff. Where lysis was incomplete, the patency was significantly less (17%).

predictor of outcome. The length of occlusion appeared to be a more important factor. Patients with longer occlusions appear to fare better with thrombolysis, which may be related to the difficulty of choosing an outflow site for a bypass graft [15]. Diabetic patients appear to respond poorly to attempts at graft thrombolysis and do badly following graft lysis. They were less likely to undergo successful initial thrombolysis and had inferior long-term patency if recanalized. Although the presence of diabetic distal disease was suspected as the cause, it could not be proven. Runoff was similar in both groups. Hye et al. [6] suggested that venous bypass occlusions may have a worse outcome compared to their synthetic counterparts because of vein wall ischemia. This observation appears to be born out in other author's experience [16,17]. One study, however, found that venous bypass occlusions respond well to thrombolysis [4]. The results of the latter study may be misleading because of confounding factors such as the state of the runoff and the presence of underlying occlusive lesions. No firm conclusions can be drawn about the use of warfarin following the successful restoration of graft patency by thrombolysis. Both Nackman et al. [13] and Galland et al. [9] demonstrated that longterm patency was not affected by the use of warfarin, although the number of patients was small and anticoagulation was prescribed at the discretion of the surgeon. Our own feeling is that aspirin should be used for the prevention of other cardiovascular events but the mainstay of graft preservation remains mechanical. If a bypass graft reoccludes following thrombolysis, further attempts at lysis are associated with a poor outcome. An analysis of the NATALI database found that five of eight such grafts reoccluded at three months. Subsequent surgical attempts at limb salvage following failed thrombolysis similarly have a low chance of success. In Sullivan's series, of the 11 thrombolytic failures only three were successfully treated by revision or graft replacement [4].

Life table analyses for long-term patency

Following the STILE trial, a multifactor analysis of patients undergoing thrombolysis found that the nature of the occlusion was not an independent

Periprocedural mortality
The reported peri-operative mortality for the surgical management of bypass occlusion is surprisingly low (no deaths in Comerota's series). It is probable that a significant number of patients are

VASCULAR

EMERGENCIES

not offered revascularization due to their comorbidity, although results from the STILE trial suggested 92% of designated procedures were carried out on an intent-to-treat basis. The accepted overall mortality from thrombolysis in the periprocedural period is in the area of 3% to 5%. The majority of patients will perish from stroke, myocardial infarction, or hemorrhagic complications. The long-term survival following lysis of occluded grafts is encouraging. In Nackman's paper [13], 89% of patients were still alive at two years and 84% of patients in another study were alive at 42 months [6].

Logistics
Thrombolysis is a labor-intensive technique. Patients may require multiple visits to the interventional radiology suite. In a study by Browse et al. [14], an average of four angiograms were required in every patient. Furthermore, graft occlusions appear to take longer to successfully lyse than their native arterial counterparts. Pulse-spray techniques may reduce the time to achieve patency and the labor intensity of lysis by means of its mechanical action on thrombus. However, even when pulsespray techniques are employed, graft occlusions appear to take longer to lyse than their native arterial counterparts (93 minutes versus 65 minutes in one study) [18]. The incidence of complications means that medical and nursing staff must be familiar with the technique. The patient is usually nursed on a high dependency unit. A protocol for the treatment of thrombolysis complications may reduce their severity and simplify patient care. The general impression might be that there is little cost difference between bypass surgery and thrombolysis in native vessel occlusion. However, Rickard et al. found that on average they were spending $ 2440 Australian more per patient with graft occlusion (1997) [19].

ciated with significant pain. Thrombolysis should not be considered in patients whose grafts have occluded less than four weeks following insertion. The porosity of prosthetic grafts poses risk of major hemorrhage through the graft wall and anastomosis. The complications of arterial thrombolysis are not insignificant. In a study from Nottingham, major hemorrhage occurred in 5% of patients, minor in 15%, and stroke in 1% [20]. In Hye's study of pulsespray thrombolysis of occluded bypass grafts [6], 27% of patients had a complication directly attributable to thrombolysis. The axilla is a more frequent site of bleeding complications than the groin because of the lack of supporting connective tissue [14]Direct graft puncture offers an alternative approach. In experienced hands it is associated with low rates of local complication. It is especially useful in patients with prosthetic extra-anatomical femorofemoral bypass grafts in whom local complications may be high [21]. Evidence from the UK NATALI database suggests that bleeding complications were higher in patients undergoing graft thrombolysis [9]. Prosthetic grafts in particular seem to have more local complications [3]. In another series, more complications were experienced with prosthetic graft lysis (53%) compared to vein graft (41%) and native artery bypass (29%), although these figures did not reach statistical significance. Furthermore, acute worsening of ischemia due to embolization appeared to be almost twice as common in graft thrombolysis (2.7% versus 1.5%) [22]. In that study of 19 patients, 11 patients required amputation and two patients died despite further lysis or surgery. Paradoxically, lower rates of hemorrhagic complications occurred in patients with bypass graft occlusions in the TOPAS trial and no difference could be demonstrated with regard to major morbidity at one year in the STILE trial.

Alternative indications
Occlusion of the body and iliac limbs of endovascular stent grafts is not an uncommon complication. The conventional management of endovascular stent graft occlusion is a femorofemoral bypass, an axillobifemoral graft, or conversion to open repair. Conversion to open repair carries a mortality of at least 20%. Unfortunately, thrombolysis may cause significant complications. A case from

Complications of thrombolysis for occluded grafts

Thrombolysis requires a great deal of patient cooperation. The procedure requires that the patient lies still for a number of hours and it is asso-

THROMBOLYSIS FOR OCCLUSION OF BYPASS GRAFTS

the Malmo group demonstrated that thrombolysis caused loss of graft seal with the aortic wall, aneurysm expansion and subsequent rupture [23]. In addition, thrombolysis has been used to good effect in patients with occluded hemodialysis fistula. The most important predictor of successful restoration of patency with lysis in these cases appears to be the adequacy of venous outflow. replacement with a venous conduit, whereas others have generally found limb salvage rates of 40% at three years [28]. Gardiner et al. [10] suggested that graft replacement with a venous conduit offered substantially better patency than prosthetics. Seventy-seven percent of vein grafts were patent at 16 months compared with 45% of prosthetic grafts at 14 months. However, the results may be explained in part by the large numbers of underlying correctable lesions that were found in those undergoing replacement with vein (85% versus 39%) [9].

Conventional surgical management

The diagnosis of bypass graft occlusion does not mandate revascularization. In the absence of critical ischemia (30% of occluded grafts in the STILE trial), nonoperative management may be appropriate. Alternatively, primary amputation should be considered if the graft has occluded due to global progression of atherosclerotic disease in the runoff vessels or if the patient is too frail to survive major arterial surgery. Standard surgical practice for the revascularization of critically ischemic limbs following bypass involves one of two approaches. First is graft thrombectomy (with or without an adjunctive procedure), and second is graft removal and replacement. Robinson et al. [24], however, reported a cumulative secondary patency rate of 47% at one year in patients undergoing surgical thrombectomy and an adjunctive procedure (in those patients treated for early graft failure). Twenty-six percent of patients required amputation at 1 month and 41% at one year in that study. Graft thrombo-embolectomy alone is akin to thrombolysis. Both involve removal of the thrombus without treatment of the underlying cause. One study reported a 29% patency at one year in 40 occluded above-knee prosthetic grafts undergoing revision surgery [25], The majority had graft thrombectomy with or without patch angioplasty. Others have demonstrated a two-year patency of less than 40% with thrombectomy and graft revision, decreasing to less than 30% at five years [26]. Where the population of patients is matched with respect to patient demographics, graft type and occlusion, short-term patency rates were better with thrombolysis than thrombectomy alone (86% versus 42%) [27]. Some authors have attained good longterm outcomes using graft replacement surgery. Edwards et al. [1] achieved a 5-year secondary patency of 71% and limb salvage of 90% using graft

Conclusion
The results from repair of failing bypass grafts are superior to those of grafts that are occluded. Graft replacement provides more durable results than thrombectomy alone or with angioplasty, especially if performed using autogenous vein. Patients who appear to benefit from thrombolysis are those with occluded grafts where the duration of ischemia has been short. Grafts greater than one year old and in which an underlying cause of the occlusion can be treated also appear to do well. Those patients without distal runoff and in whom graft replacement cannot be performed with an autogenous venous conduit may also be considered in a favorable light for thrombolysis. Diabetics and patients with venous bypass graft occlusions do poorly following successful thrombolysis. More evidence is required before absolute conclusions can be drawn regarding the indications for thrombolysis in bypass graft occlusion. The decision to perform thrombolysis for bypass graft occlusion will depend, in part, on the experience and distribution of resources in individual centers.

R E F E R E N C E S 1 Edwards JE, Taylor LM Jr, Porter JM. Treatment of failed lower extremity bypass grafts with new autogenous vein bypass grafting. / Vase Surg 1990; 11: 136-145. 2 Palfreyman SJ, Booth A, Michaels JA. A systematic review of intra-arterial thrombolytic therapy for lower limb ischaemia. EurJ Vase Endavasc Surg 2000; 19: 143-157. 3 Comerota AJ, Weaver FA, Hosking JD et al. Results of a prospective, randomised trial of surgery versus thrombolysis for occluded lower extremity bypass grafts. Am J Surg 1996; 172: 105-112.

VASCULAR

EMERGENCIES

30
300

4 Sullivan KL, Gardiner GA Jr, Kandarpa K et al. Efficacy of thrombolysis in infrainguinal bypass grafts. Circulation 1991; 83 (2suppl):I99-1105. 5 Braithwaite BD, Buckenham TM, Galland RB et al. Prospective randomized trial of high-dose bolus versus low-dose tissue plasminogen activator infusion in the management of acute limb ischaemia. Thrombolysis Study Group. BrJ Surg 1997; 84: 646-650. 6 Hye RJ, Turner C, Valji K et al. Is thrombolysis of occluded popliteal and tibial bypass grafts worthwhile? J Vase Surg 1994; 20: 588-597. 7 Seabrook GR, Mewissen MW, Schmitt DD et al. Percutaneous intra-arterial thrombolysis in the treatment of thrombosis of lower extremity arterial reconstructions. / Vase Surg 1991; 13: 646-651. 8 van Breda A, Robison JC, Feldman L et al. Local thrombolysis in the treatment of arterial graft occlusions. / Vase Surg 1984; 1:103-112. 9 Galland RB, Magee TR, Whitman B et al. Patency following successful thrombolysis of occluded vascular grafts. EurJ Vase Endovasc Surg m\; 22: 157-160. 10 Gardiner GA Jr, Harrington DP, Koltun W et al. Salvage of occluded arterial bypass grafts by means of thrombolysis. J Vase Swrgl989;9: 426-431. 11 Berkowitz HD, Fox AD, Deaton DH. Reversed vein graft stenosis: early diagnosis and management. / Vase Surg 1992; 15: 130-142. 12 Berridge DC, al-Kutoubi A, Mansfield AO et al. Thrombolysis in arterial graft thrombosis. EurJ Vase Endovasc Surg 1995; 9: 129-132. 13 Nackman GB, Walsh DB, Fillinger MF et al. Thrombolysis of occluded infrainguinal vein grafts: predictors of outcome. / Vase Surg 1997; 25: 1023 -1032. 14 Browse DJ, Torrie EP, Galland RB. Low-dose intra-arterial thrombolysis in the treatment of occluded vascular grafts. BrJ Sttrg-1992;79:86-88. 15 Ouriel K, Veith FJ. Acute lower limb ischemia: determinants of outcome. Surgery 1998; 124: 336-342.

16 Durham JD, Geller SC, Abbott WM et al. Regional infusion of urokinase into occluded lower-extremity bypass grafts: longterm clinical results. Radiology 1989; 172: 83-87. 17 Parent FN 3rd, Piotrowski JJ, Bernhard VM et al. Outcome of intraarterial urokinase for acute vascular occlusion. J Cardiovasc Surg 1991; 32: 680 -689. 18 Valji K, Roberts AC, Davis GB, Bookstein JJ. Pulsed-spray thrombolysis of arterial and bypass graft occlusions. AJR Am J Romtgmol 1W1; 156: 617-621. 19 Rickard MJ, Fisher CM, Soong CV et al. Limitations of intraarterial thrombolysis. Cardiovasc Surg 1997; 5: 634-640. 20 Berridge DC, Makin GS, Hopkinson BR. Local low dose intraarterial thrombolytic therapy: the risk of stroke or major haemorrhage. BrJ Surg \m; 76: 1230-1233. 21 Guest P, Buckenham T. Thrombolysis of the occluded prosthetic graft with tissue-type plasminogen activator. Technique, results and problems in 23 patients. Clin Radiol 1992; 46: 381-386. 22 Galland RB, Earnshaw JJ, Baird RN et al. Acute limb deterioration during intra-arterial thrombolysis. Br J Surg 1993; 80: 1118-1120. 23 Resch T, Lindblad B, Lindh M et al. Aneurysm expansion and retroperitoneal hematoma after thrombolysis for stent-graft limb occlusion caused by distal endograft migration. J Endovasc ThermO; 7: 446-450. 24 Robinson KD, Sato DT, Gregory RT et al. Long-term outcome after early infrainguinal graft failure. J Vase Surg 1997; 26: 425-438. 25 Anonymous. Results of a prospective randomized trial evaluating surgery versus thrombolysis for ischaemia of the lower extremity. The STILE trial. Ann Surg 1994; 220: 251-268. 26 Whittemore AD, Clowes AW, Couch NP, Mannick JA. Secondary femoropopliteal reconstruction. Ann Surg 1981; 193: 35-42. 27 Graor RA, Risius B, Young JR et al. Thrombolysis of peripheral arterial bypass grafts: surgical thrombectomy compared with thrombolysis. A preliminary report. J Vase Surg 1988; 7: 347-355. 28 Bartlett ST, Olinde AJ, Flinn WR et al. The reoperative potential of infrainguinal bypass: long-term limb and patient survival. / Vase Swrg 1987; 5: 170-179.

ACUTE PROBLEMS OF THE DIABETIC FOOT

JAN RAUWERDA

Diabetes mellitus (DM) is a metabolic disease caused by hereditary as well as environmental factors. There are two different types of DM. DM type 1 (10% to 20%), also known as insulin-dependent diabetes, is caused by a destruction ofthefl-cells of the pancreas, due to an autoimmune response. It leads to insulin deficiency and hyperglycemia. It affects primarily children and adolescents. The so-called type 2 DM (80% to 90%) is characterized by a gradual defect in insulin production in the f-cells as well as an insulin resistance caused by a diminished response of the receptors in the target organs and an increase of glucose production by the liver. Before the diagnosis of DM type 2 is made, the disease is already present for four to seven years. It was calculated that in 2000 more than 755 million people worldwide would suffer from DM. This number will increase to more than 250 million in 2010, with immense consequences for vascular surgeons [1].

301

DM and foot problems

The annual incidence of foot ulcers is 2% to 3%. Cross-sectional studies showed that the prevalence of a foot ulcer and/or amputation is 3% to 5%. The local recurrence rate of foot ulcers is high, varying from 34% after one year to 70% after 5 years. This means that, of all the individuals with DM, 15% to 25% will develop a foot ulcer [2]. There is a country-based variation in the relative risk of diabetes-related lower extremity amputation

(Table I). It is not surprising that 40% to 60% of all amputations are performed in patients with DM [3]. Within 3 years of amputation, 30% to 50% of the patients undergo an amputation of the contralateral leg. Even the life expectancy is influenced by the amputation. The 5-year survival rate of a patient with DM and a healed ulcer is 58%. After an amputation due to ischemic lesions it is only 27%, reflecting the end stage of the disease. It has been calculated that almost 20% of all the hospitalizations and costs of patients with diabetes are related to foot ulceration and/or amputation.

VASCULAR

EMERGENCIES

1st author [ref.] Gujral [4] Trautner [5] Siitonen [6] Lawee [7] Anonymous [8] Deerochanawong [9] Most [10] Miller [11]
Lee [12]

Location Leicestershire, United Kingdom Leverkusen, Germany Eastern Finland Ontario, Canada Washington State, USA Newcastle upon Tyne United Kingdom Six states in the USA New Jersey, USA Oklahoma Indians, USA Pima Indians, USA Dutch population

Year 1980-1985 1990-1991 1978-1984 1987-1988

Incidence per year

14/10 000 21/10 000

Men: Women: 35/10000 24/10000 44/10000 51/10000 57/10000 60/10000 77/10000 180/10000 241/10000

1988
1989-1991 1976-1978 1979-1981 1972-1980 1972-1984

Nelson [13] Van Houtum [14]

25/10 000

The International Working Group of the Diabetic Foot showed that the primary healing costs varies from 7 000 to 10 000 dollars. The direct costs related to amputation, in association with the diabetic foot problem, varies from 30 000 to 60 000 dollars. The indirect costs, which result from the loss of jobs, reduced productivity, the individual patient costs and not even calculating the loss of quality of life, for the diabetic foot in the United States has been estimated as 4 billion dollars per year. So the diabetic foot is not only a specific patient problem but it is also of significant economic interest [15]. The costs associated with diabetic foot ulceration and lower extremity amputations are shown in Table II. To reduce the incidence of amputations, the St. Vincent Declaration has been adopted. The goal of this declaration is to reduce the amputation rate of patients with DM by 50%. In this respect, a coordinated multidisciplinary screening or even the start of so-called Diabetic Foot Clinics has been shown to be successful. Neuropathic ulcers can be healed in 68% of patients, ischemic ulcers in 72%, and an amputation rate reduction of 50% has been achieved. Compared to the individual costs of primary healing,

the multidisciplinary approach to prevent amputation is cost effective [24-26].

Pathogenesis of foot problems

Diabetic foot ulcers are the result of al least two risk factors. This pathway is shown in Table III. Diabetic peripheral neuropathy involves mostly all fibers of the nervous system (sensory, autonomic and motor fibers). Sensory neuropathy leads to diminished sensation of pain, pressure awareness and propriocepsis. Motor neuropathy results in atrophy of the intrinsic foot muscles leading to flexion deformity of toes and even a disturbed walking pattern. All those factors lead to high-pressure zones, mostly on the plantar surface of the metatarsal heads I and V and dorsal side of toes. The autonomic neuropathy leads to diminished sweat secretion, which is responsible for vulnerable dry skin with cracks and fissures. Angiopathy also plays a role. The so-called microangiopathy is responsible for a hampered circulation by opening the arteriovenous shunts. This

ACUTE PROBLEMS OF THE DIABETIC FOOT

1st author [ref.] A. Primary healing Bouter [16] Apelqvist [17] B. Healing with amputation Connor [18] Bouter [16] Bild [19] Reiber [20] Thompson [21] Apelqvist [2] Van Houtum [22] C. Long-term costs (3-year period) Apelqvist [23]

Country

Costs

1988 1994 1987 1988 1989 1992 1993 1993 1995 1995

The Netherlands1 Sweden3 United Kingdom1 The Netherlands1 United States1 United States2 New Zealand1 Sweden3 The Netherlands1 Sweden3

10000 7000 14000 15000 8000-12000 20000-25000 11000 11000 14500 Primary healing 16 1006-26 7007 Healing with an amputation 43 1004-63 1005

1 In-hospital costs 2 In-hospital costs, rehabilitation included 3 Total direct costs until healing 4 Minor amputation 5 Major amputation 6 Without ischemia 7 With ischemia

31
303
use the standard noninvasive vascular laboratory tests because of the noncompressibility of the crural arteries. Beside neuropathy and angiopathy, little foot trauma as well as infections can lead to severe limb-threatening complications.

leads to edema noticed as a warm foot with distended veins, but also a diminished capillary flow. It is not certain whether this microcirculatory disturbance can be influenced. Another specific pattern is the so-called macroangiopathy. In comparison with the distribution of atherosclerotic lesions in non-diabetics, it is more common, affects younger individuals without sex difference, is multisegmental, but most strikingly it is a pattern of multisegmental crural vessel occlusions with surprisingly often intact foot arteries. In 40% of cases, the popliteal artery is palpable. The natural history of these atherosclerotic lesions shows much faster progress than in non-diabetics. Besides these atherosclerotic lesions, the media calcinosis (Monckenberg disease) is found more often in diabetics, with a frequency up to 30 times. This calcinosis of the muscular layer of the arterial wall, probably due to neuropathy, has no influence on the peripheral circulation, but makes it difficult to

The diabetic foot and associated complications

The World Health Organization's definition of the diabetic foot refers to foot problems in patients with DM, associated with ulceration and/or infection, with or without destruction of deep tissues, related to neurological abnormalities and varying degrees of peripheral vascular disease in the lower limb. Diabetic foot problems can be limb and life threatening, caused by infection, ischemia, or a combination of these complications and most often require urgent surgical intervention.

VASCULAR EMERGENCIES

OKimDmMncFQ&r 1999 [IS]

DIABETES MELLITUS

NEUROPATHY
1 \ 1 1
Automatic

Postural and coordination deviation

Decreased pain sensation and propriocepsis

Diminished sweating

ir

Foot deformities, stress and shear pressures

Inadequate foot wear, non-compliance, neglect, unawareness, lack of patient and staff education

ACUTE PROBLEMS OF THE DIABETIC FOOT

Although there is not much written about emergent foot surgery in patients with DM, Taylor et al. described the principles of treatment [27]. These principles are: 1 - debridement with or without phalangeal and/or ray amputation, 2 - broad-spectrum antibiotics (intravenous), 3 - noninvasive vascular testing and angiography, 4 - revascularization procedures. There are three groups of diabetic foot problems that require urgent intervention: 1 - septic diabetic foot without ischemia, 2 - septic diabetic foot with ischemia, 3 - tissue loss with ischemia. This is mostly caused by the special foot anatomy as well as metabolic disturbances aggravated by infection.

FOOT ANATOMY
The purpose of the human feet is to support the entire body weight, to absorb forces, to accommodate propulsion, and to facilitate rotational movements. Our feet are well adapted for this monumental task. These adaptations are reflected in special anatomical features such as a thickened epidermis and subcutaneous tissue consisting of fatty pads, which provide protection of the underlying soft tissues and bony structures and effectively cushion the high-stress forces. Moreover, the foot muscles are adapted to perform their task, by means of arrangement in separate anatomical compartments.

SEPTIC DIABETIC FOOT WITHOUT ISCHEMIA

Patients with a septic foot show an infected ulcer, arthritis, osteomyelitis or even a subplantar abscess combined with general symptoms of infection (fever, deregulated glucose metabolism, etc.). Even in patients with an uninterrupted arterial circulation (palpable pedal pulses, biphasic arterial signals on doppler examination, ankle-brachial index [ABI] higher than 0.9), this infection can be limb threatening.

FOOT COMPARTMENTS
Four separate foot muscles are recognized: a medial plantar, a central plantar, a lateral plantar compartment, and an interosseus compartment (Figs. 1 and 2).

305
1 - Lateral compartment 2 - Central compartment 3 - Medial compartment it - Interosseus compartment 5 - Lateral compartment 6 - Central compartment 7 - Medial compartment

FIG. 1 The plantar aspect of the foot.

FIG. 2

Transverse section of the foot.

VASCULAR

EMERGENCIES

306

The borders of the foot compartments are dorsally formed by the metatarsal bones and the interosseus fascia and, on the plantar side, by the fascia plantaris. Medially, the intermuscular septum extends from the calcaneus to the head of the first metatarsal bone, and on the lateral side, the intermuscular septum ranges from the calcaneus to the fifth metatarsal. Finally, on the central side, individual muscles are divided by the intermuscular septum. Proximally, there is a fibrous attachment between the navicular bone, the posterior tibial muscles and the medial cuneiform bone. The lateral neurovascular bundle penetrates this septum. The lateral plantar compartment houses the intrinsic foot muscles of the fifth toe, such as the flexor digiti minimi brevis and abductor muscle. The medial plantar compartment contains the intrinsic foot muscles of the big toe, including the flexor hallucis brevis and abductor hallucis muscles as well as the flexor hallucis longus muscle. The plantar compartment contains all the intrinsic foot muscles of the other toes: the adductor hallucis muscle, the quadratus plantae muscles, the flexor digitorum brevis muscle, as well as the tendons and muscles of all the long flexor muscles. The interosseus compartment contains only the interosseus muscles. Walking on a deformed foot leads to high-pressure zones. This can result in a foot ulcer. Preferential ulcer locations are the planter side of metatarsophalanges I and V. If this ulcer becomes infected,

edema increases the pressure in the adjacent compartment leading to a compartment syndrome.

Metabolic changes
Another predisposing factor for elevated compartment pressure are metabolic changes. In diabetic patients, glucose is preferentially metabolized by means of the so-called sorbitol pathway. This may lead to an excess conversion of glucose to sorbitol, which in turn is responsible for increased tissue concentration of several split products. Together, these products and the hydropholic molecules, sorbitol in particular, are responsible for edema within the compartments, leading to a higher compartment pressure. Another metabolic change in patients with diabetes is caused by greater affinity of HbAlC for oxygen than normal for hemoglobin. This phenomenon, which lowers the tissue oxygen concentration, is known as pseudohypoxia. All these factors are responsible for an increased capillary permeability [28]. A foot infection results in edema and, even in case of a normal arterial circulation, can result in digital arterial thrombosis causing wet gangrene of a toe or even a compartment syndrome of the foot with necrosis with the whole content of that of a specific compartment (Fig. 3).

FIG. 3 Necrosis example of the internal compartment of the foot.

ACUTE PROBLEMS OF THE DIABETIC FOOT

This explains why an infected foot ulcer can cause extensive tissue necrosis, even in a normal vascularized foot. Fig. 4 shows the relation between probability of wound healing and noninvasive measurements. These tests allow for the probability of wound healing to be estimated and also indicate additional angiography and revascularization procedures [15]. Absent foot pulses, ABI less than 0.6, abnormal toe pressure (ATI) less than 0.4, or noncompressible crural vessels are indications for angiography. This strategy appears not to be routine in all hospitals. A retrospective analysis of 283 diabetics from 18 hospitals in the Netherlands in 1994 showed that optimal vascular evaluation concerning palpation of foot pulses, noninvasive testing, duplex scanning of the vessels, angiography, etc., were only performed in 53% of patients undergoing a major amputation and in 45.3% undergoing a minor amputation. Only the university hospitals showed adequate vascular investigations and revascularization procedures [30].

TREATMENT
The treatment consists of broad-spectrum antibiotics (intravenous) and a drainage procedure. For adequate debridement, knowledge of the anatomy of the foot is mandatory. When the infection is over, healing by secondary intervention is to be expected. Recently some good results with use of vacuum therapy have been described [29].

The septic diabetic foot with ischemia

This group of patients shows an acute diabetic foot problem related to ischemia and infection. Also in this category, aggressive debridement and broad-spectrum antibiotic treatment is required. Noninvasive tests, if not performed pre-operatively than shortly after the primary operation, can provide information of the arterial circulation. Although there is lack of prospective trials, there is scarce evidence that transcutaneous oxygen pressure measurements as well as ABI/or toe pressure measurements can proclaim wound healing.

Angiography
In case of impaired renal function with good femoral pulses and an uninterrupted blood flow in the aortofemoral segment on duplex examination, antegrade puncture of the ipsilateral femoral artery can be performed to make an angiography of the leg and foot to diminish the total amount of the

307

TcPo2 Toe pressure Ankle pressure

mmHg
0 20 40 60 80
100 120 140

FIG. 4

Wound healing versus different noninvasive measurements.

VASCULAR

EMERGENCIES

308

nephrotoxic contrast medium to avoid renal failure. In obese patients, a Seldinger technique from the contralateral groin is preferable. Characteristically, angiography shows multiple segmental crural occlusions, but surprisingly often the pedal arteries are patent and suitable for revascularization. If angiography shows no pedal arteries, duplex and/or magnetic resonance angiography can be helpful [31]. An angiographic scoring system of the foot arteries, advised by the Joint Vascular Society Councils, is associated with bypass graft patency and limb salvage. In a prospective series, a foot score greater than 7 correlated with a bypass failure rate of 2%, and a score less than 7 correlated with a bypass failure rate of 30%. However, this scoring system proved not to be able to discriminate between candidates for a pedal bypass procedure or a primary amputation [32]. Not every patient with an acute foot problem and ischemia should be a candidate for revascularization. In this respect, the prehospital state of mobility is a very helpful scoring system to discriminate between primary amputation and revascularization (Table IV). Patients who are pre-operatively scored in categories 4 and 5 are probably better off with a primary major amputation [33]. Treatment also consists of aggressive debridement and antibiotics. In the other groups, revascularization should be performed soon after the debridement in order to prevent a major amputation. Because there is often multilevel disease, a combination of percutaneous transluminal angioplasty (PTA) and/or stent implantation for inflow improvement combined with outflow procedures (femorocrural, popliteal crural, or pedal bypasses)

must be performed. A combination of infra-inguinal revascularization surgery and PTA of aortoiliac or even femoropopliteal segment shows to be effective and there is no evidence that diabetes per se is a limiting factor [34].

Tissue loss without ischemia

The third category is patients presenting with acute ischemia and/or gangrene without infection. Noninvasive testing with angiography and revascularization are the best ways to prevent a major amputation. From this category, excellent results are described in the literature (3-year patency higher than 70%, limb salvage superior to 80%) [35,36]. In case of large soft tissue defects, even the use of a free flap transfer combined with crural or pedal bypass is effective, although reconstructions on the plantar side without restoration of sensibility have shown to be of less value [37]. However, the five-year life expectancy of patients with neuro-ischemic foot defects in diabetic patients is less than 25%. From the literature it is known that there is no evidence that crural pedal reconstruction in diabetics have a worse prognosis than in non-diabetics [38]. In case of crural stenotic lesions, PTA is recommended. Although there are some reports of good results of durability of PTA of multiple crural lesions, others report only a patency of 55% after one month and 32% after one year [39,40]. Nevertheless, the improved arterial inflow by means of PTA of crural lesions can be sufficient diabetic foot healing.

Personal results of urgent foot surgery in diabetic patients

In a retrospective analysis of all consecutive admissions for urgent foot surgery in diabetic patients in our tertiary referral center during a five-year period, 198 patients were analyzed. The patient characteristics are shown in Table V. In 127 (64%) patients, the indication for urgent foot surgery was ischemia with or without infection. One hundred and seventeen patients underwent 175 vascular procedures. Peri-operatively, 14 patients died, eight patients as a result of sepsis and/or respiratory insufficiency, two patients because of myocardial events, and four

fabtel?
1 2 3 4 5

DEGREES OF MOBILITY

Free mobility in- and outdoors Free mobility outdoors with support Free mobility indoors with support Wheelchair dependency Bed: not able to move around

ACUTE PROBLEMS OF THE DIABETIC FOOT

TdileV

PATIENT CHARACTERISTICS OF ?$irt&i

1996-2001 198 199

Period Number of patients Number of involved legs

Septic foot Septic ischemic foot Ischemic foot

72 78 49

Duration of Diabetes millitus - Years Age (range) - Years Vascular involvement - N Primary amputation - N Vascular procedures - N

17.1 67.9(41-94) 127 10 117

Inflow Outflow Combined Total Primary assisted patency - % Peri-operative mortality - % Two-year limb salvage rate - c , Two-year survival - %

45 100 30 175

62 12.6 81.1 52

patients, mostly young ones, refused further treatment because they had end-stage diabetic disease with blindness, dialysis dependency and an impending major amputation. The two-year primary assisted patency of crural/ pedal reconstruction was 62%, with a limb salvage of 81.1%, comparable to the results in the literature.

Routine postoperative control

After secondary wound healing with or without revascularization, the plantar surface of the foot is deformed. This deformity again leads to high-pressure zones. These high-pressure zones are a potential risk for development of recurrent foot ulcers and, consequently, a risk for amputation. This amputation risk is expressed in the risk profile for the diabetic foot, also known as the Simms classification, which determines the frequency of postoperative control (Table VI). In a two-year follow-up study, the amputation risk in patients within category 3 was 25% [41]. Also in a prospective controlled trial, this risk classification showed to be effective.

In a series of 225 diabetic patients during a threeyear follow-up, ulceration occurred in 5.1% in category 0, in 14.3% in category 1, in 18.8% in category 2, and 54.8% in category 3 (bilinear association p < 0.001). All amputations took place in group 2 and 3 (respectively, 3.1% and 20.9%; p < 0.001) [42]. In general, a multidisciplinary approach is effective in lowering the amputation rate in diabetic patients and the frequency of control depends on the risk profile [43].

309

fttteYT

SIMMS CMSSIHCATI0N QREKEP1OHLE

Category

Neuropathy

Foot deformity

Foot surgery

VASCULAR

EMERGENCIES

R E F E R E N C E S 1 Boulton AJ. The pathogenesis of diabetic foot problems: an overview. Diabet Med 1996; 13 (Suppl 1): S12-16. 2 Apelqvist J, Larsson J, Agardh CD. Long-term prognosis for diabetic patients with foot ulcers. / Intern Med 1993; 233: 485-491. 3 van Houtum WH, Lavery LA. Regional variation in the incidence of diabetes-related amputations in the Netherlands. Diabetes Res Cfoz/tow* 1996; 31: 125-132. 4 Gujral JS, McNally PG, O'Malley BP, Burden AC. Ethnic differences in the incidence of lower extremity amputation secondary to diabetes mellitus. Diabet Med 1993; 10: 271-274. 5 Trautner C, Haastert B, Giani G, Berger M. Incidence of lower limb amputations and diabetes. Diabetes Care 1996; 19: 1006-1009. 6 Siitonen 01, Niskanen LK, Laakso M et al. Lower-extremity amputations in diabetic and nondiabetic patients. A populationbased study in eastern Finland. Diabetes Care 1993; 16: 16-20. 7 Lawee D, Csima A. Diabetes-related lower extremity amputations in Ontario: 1987-1988 experience. Can J Public Health 1992; 83: 298-302. 8 Anonymous. Lower extremity amputations among persons with diabetes mellitus. Washington, 1988. MMWR Morb Mortal Wkly Rep 1991; 40: 737 -739. 9 Deerochanawong C, Home PD, Alberti KG. A survey of lower limb amputation in diabetic patients. Diabet Med 1992; 9: 942-946. 10 Most RS, Sinnock P. The epidemiology of lower extremity amputations in diabetic individuals. Diabetes Care 1983; 6: 87-91. 11 Miller AD, Van Buskirk A, Verhoek-Oftedahl W, Miller ER. Diabetes-related lower extremity amputations in New Jersey, 1979 to 1981. JMed SocNjmb; 82: 723-726. 12 Lee JS, Lu M, Lee VS et al. Lower-extremity amputation. Incidence, risk factors, and mortality in the Oklahoma Indian diabetes study. Diabetes 1993; 42: 876-882. 13 Nelson RG, Gohdes DM, Everhart JE et al. Lower-extremity amputations in NIDDM. 12-year follow-up study in Pima Indians. Diabetes Care 1988; 11: 8-16. 14 van Houtum WH. Diabetes related lower extremity amputations. Thesis, Amsterdam, 1998. 15 Anonymous. International Working Group on the Diabetic Foot. International consensus on the diabetic foot. Bruxelles, International Diabetes Federation, 1999: 96 p. 16 Bouter KP, Storm AJ, de Groot RR et al. The diabetic foot in Dutch hospitals: epidemiological features and clinical outcome. Eur/Mn993; 2: 215-218. 17 Apelqvist J, Ragnarson-Tennvall G, Persson U, Larsson J. Diabetic foot ulcers in a multidisciplinary setting. An economic analysis of primary healing and healing with amputation. J Intern Med 1994; 235: 463-471. 18 Connor H, The economic impact of diabetic foot disease. In: Connor H, Boulton AJM, Ward JD (eds). The foot in diabetes. Chichester, John Wiley & Sons, 1987: pp 145-149. 19 Bild DE, Selby JV, Sinnock P et al. Lower-extremity amputation in people with diabetes. Epidemiology and prevention. Diabetes Care 1989; 12:24-31. 20 Reiber GE, Pecoraro RE, Koepsell TD. Risk factors for amputation in patients with diabetes mellitus. A case-control study. Ann Intern Med 1992; 117: 97-105. 21 Thomson FJ, Masson EA, Boulton AJ. The clinical diagnosis of sensory neuropathy in elderly people. Diabet Med 1993; 10: 843-846. 22 van Houtum WH, Lavery LA, Harkless LB. The costs of diabetes-related lower extremity amputations in the Netherlands. Diabet Med 1995; 12: 777-781. 23 Apelqvist J, Ragnarson-Tennvall G, Larsson J, Persson U. Longterm costs for foot ulcers in diabetic patients in a multidisciplinary setting. Foot Ankle Int 1995; 16: 388-394. 24 Apelqvist J, Larsson J. What is the most effective way to reduce incidence of amputation in the diabetic foot? Diabetes Metab Res RevmQ; 16 (Suppl 1): S75-83. 25 Holstein PE, Sorensen S. Limb salvage experience in a multidisciplinary diabetic foot unit. Diabetes Care 1999; 22 (Suppl 2) : S97-103. 26 Meltzer DD, Pels S, Payne WrG et al. Decreasing amputation rates in patients with diabetes mellitus. An outcome study. JAm Podiatr Med Assoc 2002; 92: 425-428. 27 Taylor LM Jr, Porter JM. The clinical course of diabetics who require emergent foot surgery because of infection or ischemia. J Vase Surg 1987; 6: 454-459. 28 Rauwerda JA. Foot debridement: anatomic knowledge is mandatory. Diabetes Metab Res Rev 2000; 16 (Suppl 1): S23-26. 29 McCallon SK, Knight CA, Valiulus JP et al. Vacuum-assisted closure versus saline-moistened gauze in the healing of postoperative diabetic foot wounds. Ostomy Wound Manage 2000; 46: 28-32, 34. 30 van Houtum WH, Bakker K, Rauwerda JA, Heine RJ. Vascular assessment before lower extremity amputation. The diabetic foot 2001; 4: 185-193. 31 Hofmann W, Forstner R, Kofler B et al. Pedal artery imaging. A comparison of selective digital subtraction angiography, contrast enhanced magnetic resonance angiography and duplex ultrasound. EurJ Vase Endovasc Surg 2002; 24: 287-292. 32 Toursarkissian B, D'Ayala M, Stefanidis D et al. Angiographic scoring of vascular occlusive disease in the diabetic foot: relevance to bypass graft patency and limb salvage./ Vase Srg2002; 35: 494-500. 33 Luther M. Surgical treatment of chronic critical leg ischaemia. A five-year follow-up of survival, mobility and treatment level. EurJ Surg 1998; 164: 35-43. 34 Faries PL, Brophy D, LoGerfo FW et al. Combined iliac angioplasty and infrainguinal revascularization surgery are effective in diabetic patients with multilevel arterial disease. Ann Vase Swig 2001; 15:67-72. 35 Pomposelli FB Jr, Arora S, Gibbons GWr et al. Lower extremity arterial reconstruction in the very elderly: successful outcome preserves not only the limb but also residential status and ambulatory function. / Vase Surg 1998; 28: 215-225. 36 Dorweiler B, Neufang A, Schmiedt W, Oelert H. Pedal artery bypass for limb salvage in patients with diabetes mellitus. Eur J Vase Endovasc Surg 2002; 24: 309-313. 37 Vermassen FE, van Landuyt K, Combined vascular reconstruction and free flap transfer in diabetic arterial disease. Diabetes Metab Res Rev 2000; 16 (Suppl 1): S33-36. 38 Hamdan AD, Saltzberg SS, Sheahan M et al. Lack of association of diabetes with increased postoperative mortality and cardiac morbidity: results of 6565 major vascular operations. Arch Surgmt; 137: 417-421. 39 Danielsson G, Albrechtsson U, Norgren L et al. Percutaneous transluminal angioplasty of crural arteries: diabetes and other factors influencing outcome. EurJ Vase Surg 2001; 21: 432-436. 40 Neufang A Kraus 0, Dorweiler B et al. Pedal bypass surgery in diabetic foot syndrome: indications, technique and outcome. Med Klin 2002; 97: 256-262. 41 Rith-Najarian SJ, Stolusky T, Gohdes DM. Identifying diabetic patients at high risk for lower-extremity amputation in a primary health care setting. A prospective evaluation of simple screening criteria. Diabetes Care 1992; 15: 1386-1389. 42 Peters EJ, Lavery LA. Effectiveness of the diabetic foot risk classification system of the International Working Group on the Diabetic Foot. Diabetes Care 2001; 24: 1442-1447. 43 Van Gils CC WTieeler LA, Mellstrom M et al. Amputation prevention by vascular surgery and podiatry collaboration in highrisk diabetic and nondiabetic patients. The Operation Desert Foot experience. Diabetes Care 1999; 22: 678-683.

11
310

Conception et realisation ODIM, Z.A. La Carretiere 04130VOLX Acheve d'imprimer sur ses presses en fevrier 2003 N d'imprimeur: 92.775 Depot legal: ler trimestre 2003 Imprime en FRANCE