Textbook of Interventional Cardiology, 5th Edition

1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899

TEXTBOOK OF INTERVENTIONAL CARDIOLOGY

ISBN: 978-1-4160-4835-0

Copyright © 2008, 2003, 1999, 1994, 1990 by Saunders, an imprint of Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permissions may be sought directly from Elsevier’s Rights Department: phone: (+1) 215 239 3804 or (+44) 1865 843830 (UK); fax: (+44) 1865 853333; e-mail: [email protected]. You may also complete your request on-line via the Elsevier website at http://www.elsevier.com/permissions.

Notice Knowledge and best practice in this field are constantly changing. As new research and experience broaden our knowledge, changes in practice, treatment, and drug therapy may become necessary or appropriate. Readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of the practitioner, relying on his or her experience and knowledge of the patient, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the Editor assumes any liability for any injury and/or damage to persons or property arising out of or related to any use of the material contained in this book. The Publisher Library of Congress Cataloging-in-Publication Data Textbook of interventional cardiology / [edited by] Eric J. Topol.—5th ed. p. ; cm Includes bibliographical references and index. ISBN 978-1-4160-4835-0 1. Heart—Interventional radiology. 2. Angioplasty. 3. Cardiovascular system—Diseases— Treatment. I. Topol, Eric J., 1954[DNLM: 1. Cardiovascular Diseases—surgery. 2. Cardiovascular Diseases—drug therapy. 3. Cardiovascular Surgical Procedures—methods. WG 168 T355 2008] RD598.35.I55T49 2008 617.4′12059—dc22

2007018033

Executive Publisher: Natasha Andjelkovic Developmental Editor: Agnes Hunt Byrne Publishing Services Manager: Frank Polizzano Project Manager: Michael H. Goldberg Design Direction: Steven Stave

Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org

Printed in China Last digit is the print number:

9

8

7

6

5

4

3

2

1

This book is dedicated to my family—my wife, Susan, and our kids, Sarah and Evan, who have been with me every step along the way.

Contributors

Jorge R. Alegria, MD

Matthew C. Becker,

Fellow, Interventional Cardiology, Mayo Clinic College of Medicine, Rochester, Minnesota Balloon Angioplasty: Is It Still a Viable Intervention?

Fellow, Cardiovascular Disease, Cleveland Clinic, Cleveland, Ohio Hypertrophic Cardiomyopathy

Alexandra Almonacid, MD

Robert H. Beekman III,

Research Associate, Harvard Medical School; Assistant Director, Angiographic Core Laboratory, Brigham and Women’s Hospital, Boston, Massachusetts Qualitative and Quantitative Coronary Angiography

R. David Anderson,

MD, MS

Peter B. Berger,

Paolo Angelini,

Farzin Beygui,

MD

Clinical Professor of Medicine, Baylor College of Medicine; Interventional Cardiologist, Texas Heart Institute, St. Luke’s Episcopal Hospital, Houston, Texas Surgical Standby: State of the Art MD

Assistant Clinical Professor of Medicine, Medical University of Ohio, Toledo; Director, Peripheral Vascular Intervention, Riverside Methodist Hospital; Investigator, MidWest Cardiology Research Foundation, Columbus, Ohio Venous Interventions

Saif Anwaruddin,

MD

Professor of Pediatric Cardiology, University of Cincinnati College of Medicine; Professor, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Transcatheter Therapies for Congenital Heart Disease

Associate Professor of Medicine and Director of Interventional Cardiology, Division of Cardiovascular Medicine, University of Florida College of Medicine, Gainesville, Florida Elective Intervention for Chronic Coronary Syndromes

Gary M. Ansel,

MD

MD

Associate Chief Research Officer and Director of the Center for Clinical Studies, Geisinger Health Systems, Danville, Pennsylvania Intervention in Complex Lesions and Multivessel Disease MD, PhD

Permanent University Hospital Staff, Faculté de Médicine Pitié-Salpêtrière, Université Paris VI; Senior Consultant, Pitié-Salpêtrière University Hospital, Paris, France Transradial Percutaneous Coronary Intervention for a Major Reduction of Bleeding Complications

John A. Bittl,

MD

Interventional Cardiologist, Munroe Regional Medical Center, Ocala Heart Institute, Ocala, Florida Role of Adjunct Devices: Cutting Balloon, Thrombectomy, Laser, Ultrasound, and Atherectomy

MD

Fellow, Cardiology, Cleveland Clinic, Cleveland, Ohio Inflammation Status

Christopher Bajzer,

MD

Associate Director, Peripheral Intervention, Heart and Vascular Institute, Cleveland Clinic, Cleveland, Ohio Renal Artery Stenosis

Anthony A. Bavry,

MD, MPH

Fellow, Interventional Cardiology, Cleveland Clinic, Cleveland, Ohio Late Stent Thrombosis

Ashley B. Boam,

MS

Chief, Interventional Cardiology Devices Branch, Center for Devices and Radiological Health, U.S. Food and Drug Administration, Rockville, Maryland Regulatory Issues

Philipp Bonhoeffer,

MD

Professor of Cardiology, Institute of Child Health; Chief of Cardiology and Director of the Cardiac Catheterisation Laboratory, Great Ormond Street Hospital, London, United Kingdom Pulmonary and Tricuspid Valve Interventions

vii

viii

Contributors Michael Brandle,

Ryan D. Christofferson,

MD, MS

Associate Professor of Endocrinology, Zurich School of Medicine, Zurich; Division Chief, Division of Endocrinology and Diabetes, Department of Internal Medicine, Kantonsspital St. Gallen, St. Gallen, Switzerland Diabetes

Ralph G. Brindis,

MD, MPH

Clinical Professor of Medicine, University of California, San Francisco, School of Medicine, San Francisco; Senior Advisor for Cardiovascular Diseases, Northern California Kaiser Permanente; Physician, Department of Cardiology, Kaiser Permanente Oakland Medical Center, Oakland, California Quality of Care in Interventional Cardiology

David Buckles,

PhD

Louise Coats,

MBBS, MRCP

Clinical Research Fellow, Institute of Child Health and Great Ormond Street Hospital, London, United Kingdom Pulmonary and Tricuspid Valve Interventions

Antonio Colombo,

MD

Faculty of Medicine and Surgery, Vita-Salute San Raffaele University; Director of Invasive Cardiology, San Raffaele Scientific Institute and Columbus Hospital, Milan, Italy Ostial and Bifurcation Lesions

Chief, Peripheral Vascular Devices Branch, Center for Devices and Radiological Health, U.S. Food and Drug Administration, Rockville, Maryland Regulatory Issues

Bertrand Cormier, MD

Heinz Joachim Büttner,

Marco A. Costa,

MD

Director, Interventional Cardiology, Herz-Zentrum Bad Krozingen, Bad Krozingen, Germany Evidence-Based Interventional Practice

Christopher P. Cannon,

MD

Associate Professor of Medicine, Harvard Medical School; Associate Physician, Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts Lipid Lowering in Coronary Artery Disease

Ivan P. Casserly,

BSc, MB BCh

Assistant Professor of Medicine, University of Colorado School of Medicine and University of Colorado Hospital; Director of Interventional Cardiology, Denver Veterans Affairs Medical Center, Denver, Colorado Carotid and Cerebrovascular Interventions

Matthews Chacko,

MD

Assistant Professor of Medicine, Johns Hopkins University School of Medicine; Director of Peripheral Interventions, and Faculty, Interventional Cardiology, Coronary Care Unit, and the Thayer Firm, Division of Cardiology, Johns Hopkins Hospital, Baltimore, Maryland Thrombolytic Intervention

Derek P. Chew,

MBBS, MPH

Associate Professor of Medicine, Flinders University School of Medicine; Director of Cardiology, Flinders Medical Centre, Adelaide, South Australia, Australia Anticoagulation in Percutaneous Coronary Intervention

Leslie Cho,

MD

Director, Women’s Cardiovascular Center; Medical Director, Preventive Cardiology and Rehabilitation, Cleveland Clinic, Cleveland, Ohio Gender and Ethnicity Issues in Percutaneous Coronary Intervention

MD

Fellow, Interventional Cardiology, Cleveland Clinic, Cleveland, Ohio Percutaneous Mitral Valve Repair

Hospital Doctor, Service Médecine-Cardiologie, Institut Hospitalier Jacques Cartier, Massy, France Mitral Valvuloplasty MD, PhD

Associate Professor of Medicine and Director of Research, Division of Cardiology, University of Florida College of Medicine Jacksonville, Jacksonville, Florida Restenosis

Alain Cribier, MD Professor of Medicine, University of Rouen; Chief, Department of Cardiology, Hôpital Charles Nicolle, Rouen, France Percutaneous Aortic Valvular Approaches: Balloon Aortic Valvuloplasty and Percutaneous Valve Replacement with the Cribier-Edwards Bioprosthesis

Fernando Cura,

MD, PhD

Vice Director, Interventional Cardiology and Endovascular Therapies, Instituto Cardiovascular de Buenos Aires, Buenos Aires, Argentina Access Management and Closure Devices

Pranab Das,

MD

Fellow, Interventional Cardiology, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois Bioabsorbable Stents

P. J. de Feyter,

MD, PhD

Erasmus Medical Center, Rotterdam, The Netherlands Percutaneous Intervention for Non-ST Segment Elevation Acute Coronary Syndromes

Robert S. Dieter,

MD, RVT

Assistant Professor of Medicine, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois Upper Extremities and Aortic Arch

John S. Douglas, Jr., MD Professor of Medicine, Emory University School of Medicine; Director of Interventional Cardiology and Cardiac Catheterization Laboratories, Emory University Hospital, Atlanta, Georgia Percutaneous Intervention in Patients with Prior Coronary Bypass Surgery

Contributors Stephen G. Ellis,

Hidehiko Hara,

MD

MD

Director, F. Mason Sones Cardiac Catheterization Laboratory, Cleveland Clinic, Cleveland, Ohio Drug-Eluting and Bare Metal Stents

Fellow, Preclinical Research, Minneapolis Heart Institute and Foundation, Minneapolis, Minnesota The Left Atrial Appendage

Helene Eltchaninoff,

Motoya Hayase, MD

MD

Professor of Medicine, University of Rouen; Chief, Cardiac Catheterization Laboratory, Department of Cardiology, Hôpital Charles Nicolle, Rouen, France Percutaneous Aortic Valvular Approaches: Balloon Aortic Valvuloplasty and Percutaneous Valve Replacement with the Cribier-Edwards Bioprosthesis

Associate Director, Interventional Cardiovascular Therapy, The Skirball Center for Cardiovascular Research, Cardiovascular Research Foundation, Orangeburg, New York Percutaneous Revascularization Procedures

Nezar Falluji,

Professor of Medicine, University of Pennsylvania School of Medicine; Director, Interventional Cardiology and Cardiac Catheterization Laboratories, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Support Devices for High Risk Percutaneous Coronary Intervention

MD

Assistant Professor, Division of Cardiology, Linda and Jack Gill Heart Institute, University of Kentucky, Lexington, Kentucky Lower Extremity Interventions

Andrew Farb,

MD

Medical Officer, Center for Devices and Radiological Health, U.S. Food and Drug Administration, Rockville, Maryland Regulatory Issues

Peter J. Fitzgerald,

MD, PhD

Professor of Medicine (Cardiology) and Engineering, Stanford University School of Medicine; Director, Center for Cardiovascular Technology, Stanford University, Stanford, California Intravascular Ultrasound

Shmuel Fuchs,

MD

Howard C. Herrmann,

Russel Hirsch,

MD

MBChB

Associate Professor of Pediatric Cardiology, University of Cincinnati College of Medicine; Director, Cardiac Catheterization Laboratory, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Transcatheter Therapies for Congenital Heart Disease

David R. Holmes, Jr., MD Professor of Medicine, Mayo Clinic College of Medicine; Consultant, Mayo Clinic, Rochester, Minnesota Balloon Angioplasty: Is It Still a Viable Intervention?; The Left Atrial Appendage

Associate Professor of Cardiology, Sackler School of Medicine, Tel Aviv University, Tel Aviv; Director, Catheterization Laboratory Service, Golda-Hasharon Hospital, Rabin Medical Center, Petah Tikva, Israel Percutaneous Myocardial Revascularization: Lasers and Biologic Compounds

Yasuhiro Honda,

Valentin Fuster,

Professor of Medicine, Mount Sinai School of Medicine; Director, Cardiovascular Institute, Mount Sinai Hospital, New York, New York Atherothrombosis and the High-Risk Plaque

Lecturer in Cardiology, University of Rostock School of Medicine; Deputy Head, Division of Cardiology and Angiology, University Hospital Rostock, Rostock, Germany Aortic Vascular Interventions (Thoracic and Abdominal)

Mario J. Garcia,

Eduardo Infante de Oliveira,

MD, PhD

MD

Professor of Medicine and Radiology, Mount Sinai School of Medicine; Director of Cardiac Imaging, Mount Sinai Hospital, New York, New York Functional Testing and Multidetector Computed Tomography

Lowell Gerber,

MD

Chief of Cardiovascular Services, Florida Hospital Heartland, Sebring, Florida Percutaneous Aortic Valvular Approaches: Balloon Aortic Valvuloplasty and Percutaneous Valve Replacement with the Cribier-Edwards Bioprosthesis

Hussam Hamdalla,

MD

Assistant Professor and Associate Program Director, Interventional Cardiology Fellowship, University of Kentucky College of Medicine, Lexington, Kentucky Role of Platelet Inhibitor Agents in Percutaneous Coronary Intervention

MD

Co-Director, Cardiovascular Core Analysis Laboratory, Center for Cardiovascular Technology, Stanford University, Stanford, California Intravascular Ultrasound

Hüseyin Ince,

MD

MD

Staff Cardiologist, Hospital de Santa Maria, Faculdade de Medicina de Lisboa, Lisbon, Portugal Renal Artery Stenosis

Bernard Iung,

MD

Professor of Cardiology, University of Paris VII; Hospital Doctor, Service de Cardiologie, Hôpital Bichat, Paris, France Mitral Valvuloplasty

Alice K. Jacobs,

MD

Professor of Medicine, Boston University School of Medicine; Director, Cardiac Catheterization Laboratories and Interventional Cardiology, Boston Medical Center, Boston, Massachusetts Regional Centers of Excellence for the Care of Patients with Acute Ischemic Heart Disease

ix

x

Contributors Hani Jneid,

Alexandra J. Lansky,

MD

Division of Cardiology, University of Louisville, Louisville, Kentucky Percutaneous Balloon Pericardiotomy for Patients with Pericardial Effusion and Tamponade

Samuel L. Johnston,

MD

Fellow, Cardiology, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois Upper Extremities and Aortic Arch

Samir R. Kapadia, MD Associate Professor of Medicine, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University; Staff, Cleveland Clinic, Cleveland, Ohio Imaging for Intracardiac Interventions; Mitral Valve Repair; Hypertrophic Cardiomyopathy

Adnan Kastrati,

MD

Professor of Cardiology, Technische Universität; Director, Catheterization Laboratory, Department of Cardiology, Deutsches Herzzentrum, Munich, Germany Percutaneous Coronary Interventions in Acute ST-Segment Elevation Myocardial Infarction

Dean J. Kereiakes,

MD

Professor of Clinical Medicine, The Ohio State University College of Medicine, Columbus; Medical Director, Christ Cincinnati Heart and Vascular Center; Medical Director, Lindner Center for Research and Education, Cincinnati, Ohio Regional Centers of Excellence for the Care of Patients with Acute Ischemic Heart Disease

Morton J. Kern,

MD

Professor of Medicine, University of California, Irvine, School of Medicine, Irvine; Associate Chief of Cardiology and Director of the Cardiac Care Unit, UC Irvine Medical Center, Orange, California Intracoronary Pressure and Flow Measurements

Matheen A. Khuddus,

MD

Fellow, Cardiology, Division of Cardiology, University of Florida College of Medicine, Gainesville, Florida Elective Intervention for Chronic Coronary Syndromes

Young-Hak Kim,

MD

Assistant Professor of Medicine, Ulsan University; Attending Physician, Asan Medical Center, Seoul, South Korea Percutaneous Intervention for Left Main Coronary Artery Stenosis

Ran Kornowski,

MD

Associate Professor of Cardiovascular Medicine, Sackler School of Medicine, Tel Aviv University, Tel Aviv; Director, Interventional Cardiology and Cardiac Catheterization Laboratories, Beilinson and GoldaHasharon Hospitals, Rabin Medical Center, Petah Tikva, Israel Percutaneous Myocardial Revascularization: Lasers and Biologic Compounds

MD

Associate Professor of Clinical Medicine, Columbia University College of Physicians and Surgeons; Director of Clinical Services, Center for Interventional Vascular Therapy, New York–Presbyterian Hospital/Columbia University Medical Center, New York, New York Qualitative and Quantitative Coronary Angiography

John M. Lasala,

MD, PhD

Professor of Medicine, Washington University School of Medicine; Director, Interventional Cardiology, and Medical Director, Cardiac Catheterization Laboratory, Barnes-Jewish Hospital, St. Louis, Missouri Percutaneous Closure of Patent Foramen Ovale and Atrial Septal Defect

Robert J. Lederman,

MD

Investigator, Cardiovascular Branch, Division of Intramural Research, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland Cardiovascular Interventional Magnetic Resonance Imaging

Michael J. Lim,

MD

Assistant Professor of Medicine and Director, Interventional Cardiology Fellowship Training Program, Saint Louis University School of Medicine; Director, Cardiac Catheterization Laboratory, Saint Louis University Hospital, St. Louis, Missouri Intracoronary Pressure and Flow Measurements

A. Michael Lincoff,

MD

Professor of Medicine, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University; Vice Chairman for Research, Department of Cardiovascular Medicine, and Director, Cleveland Clinic Cardiovascular Coordinating Center, Cleveland Clinic, Cleveland, Ohio Abrupt Vessel Closure

Thomas R. Lloyd,

MD

Professor of Pediatrics, University of Michigan Medical School; Director, Cardiac Catheterization Laboratory, C. S. Mott Children’s Hospital, Ann Arbor, Michigan Transcatheter Therapies for Congenital Heart Disease

Daniel B. Mark, MD, MPH Professor of Medicine, Duke University School of Medicine; Co-Director, Coronary Care Unit, and Attending Physician, Duke University Medical Center, Durham, North Carolina Medical Economics in Interventional Cardiology

Bernhard Meier,

MD

Professor of Cardiology, Faculty of Medicine, University of Bern; Director of Cardiology, University Hospital, Bern, Switzerland Chronic Total Occlusion

Victor M. Mejia,

MD

Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Support Devices for High Risk Percutaneous Coronary Intervention

Contributors Gilles Montalescot,

Masakiyo Nobuyoshi,

MD, PhD

MD, PhD

Professor of Cardiology, Institut de Cardiologie, PitiéSalpétrière Hospital, Paris, France Transradial Percutaneous Coronary Intervention for a Major Reduction of Bleeding Complications

Clinical Professor, Kyoto University Faculty of Medicine, Kyoto; Chairperson, Kokura Memorial Hospital, Kitakyushu, Fukuoka Prefecture, Japan Small-Vessel and Diffuse Disease

Pedro R. Moreno,

Igor F. Palacios,

MD

Associate Professor of Medicine, Mount Sinai School of Medicine; Director, Interventional Cardiology Research, Mount Sinai Hospital, New York, New York Atherothrombosis and the High-Risk Plaque

Douglass A. Morrison,

MD, PhD

Interventional Cardiologist, Yakima Heart Center, Yakima, Washington Extent of Atherosclerotic Disease and Left Ventricular Function

Debabrata Mukherjee,

MD, MS

Gill Foundation Professor of Interventional Cardiology, University of Kentucky College of Medicine; Director, Cardiac Catheterization Laboratories, University of Kentucky Medical Center, Lexington, Kentucky Periprocedural Myocardial Infarction and EmbolismProtection Devices; Bioabsorbable Stents; Lower Extremity Interventions

Srihari S. Naidu,

MD

Director, Cardiac Catheterization Laboratory, WinthropUniversity Hospital, Mineola, New York Support Devices for High Risk Percutaneous Coronary Intervention

Brahmajee K. Nallamothu,

MD, MPH

Assistant Professor of Internal Medicine, Division of Cardiology, University of Michigan Medical School, Ann Arbor, Michigan Renal Dysfunction

Craig R. Narins,

MD

Assistant Professor of Medicine and Cardiology and Assistant Professor of Vascular Surgery, University of Rochester School of Medicine and Dentistry, Rochester, New York Preoperative Coronary Intervention

Gjin Ndrepepa,

MD

Associate Professor of Cardiology, Technische Universität and Deutsches Herzzentrum, Munich, Germany Percutaneous Coronary Interventions in Acute ST-Segment Elevation Myocardial Infarction

Franz-Josef Neumann,

MD

Honorary Professor of Cardiology, Albert-LudwgsUniversität, Frieburg; Medical Director and Chairman, Herz-Zentrum Bad Krozingen, Bad Krozingen, Germany Evidence-Based Interventional Practice

Christoph A. Nienaber,

MD, PhD

Professor of Internal Medicine and Cardiology, University of Rostock School of Medicine; Head, Division of Cardiology and Angiology, University Hospital Rostock, Rostock, Germany Aortic Vascular Interventions (Thoracic and Abdominal)

MD

Associate Professor of Medicine, Harvard Medical School; Director, Knight Catheterization Laboratory, Massachusetts General Hospital, Boston, Massachusetts Percutaneous Balloon Pericardiotomy for Patients with Pericardial Effusion and Tamponade

Seung-Jung Park,

MD, PhD

Professor of Medicine, Ulsan University; Director, Asan Medical Center, Seoul, South Korea Percutaneous Intervention for Left Main Coronary Artery Stenosis

Uptal D. Patel,

MD

Assistant Professor of Medicine and Pediatrics, Divisions of Nephrology and Pediatric Nephrology, Duke University School of Medicine, Durham, North Carolina Renal Dysfunction

Marc S. Penn,

MD, PhD

Director, Bakken Heart-Brain Institute, Cleveland Clinic, Cleveland, Ohio Stem Cell Therapy for Ischemic Heart Disease

Carl J. Pepine,

MD

Eminent Scholar, American Heart Association–Suncoast Chapter Chair, Professor of Medicine, and Chief, Division of Cardiovascular Medicine, University of Florida College of Medicine, Gainesville, Florida Elective Intervention for Chronic Coronary Syndromes

Marc A. Pfeffer,

MD, PhD

Dzau Professor of Medicine, Harvard Medical School; Senior Physician, Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts Angiotensin-Axis Inhibition

Jeffrey J. Popma,

MD

Research Associate, Harvard Medical School; Director, Invasive Cardiovascular Services, St. Elizabeth’s Medical Center; Director, Angiographic Core Laboratory, Brigham and Women’s Hospital, Boston, Massachusetts Qualitative and Quantitative Coronary Angiography

Mark J. Post,

MD, PhD

Professor of Vascular Physiology and Chair of the Department of Physiology, University of Maastricht, Maastricht, The Netherlands Angiogenesis and Arteriogenesis

Vivek Rajagopal,

MD

Staff Cardiologist, Cardiac Disease Specialists, Piedmont Hospital, Atlanta, Georgia Other Adjunctive Drugs for Coronary Intervention: bBlockers and Calcium Channel Blockers

Stephen R. Ramee,

MD

Section Head, Invasive/Interventional Cardiology, Ochsner Clinic Foundation, New Orleans, Louisiana Chronic Mesenteric Ischemia: Diagnosis and Intervention; Acute Stroke Intervention

xi

xii

Contributors Kausik K. Ray,

MD, MRCP

Senior Clinical Research Associate, University of Cambridge; Honorary Consultant Cardiologist, Addenbrookes Hospital, Cambridge, United Kingdom Lipid Lowering in Coronary Artery Disease

Marco Roffi,

MD

Associate Professor of Medicine, University of Geneva; Director, Interventional Cardiology Unit, University Hospital, Geneva, Switzerland Diabetes

Javier Sanz,

MD

Assistant Professor of Medicine, Mount Sinai School of Medicine; Staff Cardiologist and Associate Director of CT and MRI in Cardiology, Mount Sinai Hospital, New York, New York Atherothrombosis and the High-Risk Plaque

Wolf Sapirstein,

MD

Medical Officer, Center for Devices and Radiological Health, U.S. Food and Drug Administration, Rockville, Maryland Regulatory Issues

Albert Schömig,

MD

Professor of Medicine, Technische Universität; Chief, Department of Cardiology, Deutsches Herzzentrum, Munich, Germany Percutaneous Coronary Interventions in Acute ST-Segment Elevation Myocardial Infarction

Daniel G. Schultz,

MD

Director, Center for Devices and Radiological Health, U.S. Food and Drug Administration, Rockville, Maryland Regulatory Issues

Robert S. Schwartz,

MD

Medical Director of Pre-Clinical Research, Minneapolis Heart Institute and Foundation, Minneapolis, Minnesota The Left Atrial Appendage

Mehdi H. Shishehbor,

DO, MPH

Interventional Fellow, Department of Cardiovascular Medicine, and National Institutes of Health K12 Scholar, Cleveland Clinic, Cleveland, Ohio Imaging for Intracardiac Interventions

Mitchell J. Silver,

DO

Associate Professor of Cardiology, Ohio University College of Medicine, Athens; Staff Cardiologist/Vascular Medicine, Riverside Methodist Hospital, Columbus, Ohio Venous Interventions

Daniel I. Simon,

MD

Herman K. Hellerstein Professor of Cardiovascular Research, and Director, Case Cardiovascular Center, Case Western Reserve University School of Medicine; Chief, Cardiovascular Medicine, and Director, Heart and Vascular Institute, University Hospitals Case Medical Center, Cleveland, Ohio Restenosis

Michael Simons,

MD

A.G. Huber Professor of Medicine and Director, Angiogenesis Research Center, Dartmouth Medical School, Hanover; Chief of Cardiology, DartmouthHitchcock Medical Center, Lebanon, New Hampshire Angiogenesis and Arteriogenesis

B. Clay Sizemore,

MD

Cardiology Fellow, Division of Cardiovascular Medicine, University of Florida College of Medicine, Gainesville, Florida Elective Intervention for Chronic Coronary Syndromes

Goran Stankovic,

MD

Assistant Professor of Medicine, University of Belgrade School of Medicine; Interventional Cardiologist, University Institute for Cardiovascular Diseases, Clinical Center of Serbia, Belgrade, Serbia Ostial and Bifurcation Lesions

Steven R. Steinhubl, MD Associate Professor, University of Kentucky College of Medicine, Lexington, Kentucky Role of Platelet Inhibitor Agents in Percutaneous Coronary Intervention

Srihari Thanigaraj,

MD

Associate Professor of Medicine, Cardiology Division, Washington University School of Medicine and BarnesJewish Hospital, St. Louis, Missouri Percutaneous Closure of Patent Foramen Ovale and Atrial Septal Defect

Eric J. Topol,

MD

Director, Scripps Translational Science Institute; Chief Academic Officer, Scripps Health; Professor of Translational Genomics, The Scripps Research Institute; Senior Consultant, Division of Cardiovascular Diseases, Scripps Clinic, La Jolla, California Inflammation Status; Thrombolytic Intervention

Christophe Tron,

MD

Chief, Intensive Care Unit, Department of Cardiology, Hôpital Charles Nicolle, Rouen, France Percutaneous Aortic Valvular Approaches: Balloon Aortic Valvuloplasty and Percutaneous Valve Replacement with the Cribier-Edwards Bioprosthesis

Alec Vahanian,

MD

Professor of Cardiology, University of Paris VII; Head of Cardiology Department, Hôpital Bichat, Paris, France Mitral Valvuloplasty

Robert A. Van Tassel,

MD

Senior Consultant, Cardiology, Minneapolis Heart Institute and Foundation, Minneapolis, Minnesota The Left Atrial Appendage

Ron Waksman,

MD

Professor of Medicine, Georgetown University School of Medicine; Associate Chief of Cardiology and Director of Experimental Angioplasty and New Technologies, Washington Hospital Center, Washington, DC

Vascular Brachytherapy for Restenosis

Contributors Christopher J. White,

MD

Chair, Department of Cardiology, Ochsner Clinic Foundation, New Orleans, Louisiana Chronic Mesenteric Ischemia: Diagnosis and Intervention; Acute Stroke Intervention

Paul G. Yock,

MD

Martha Meier Weiland Professor of Medicine and Bioengineering, Stanford University School of Medicine; Director, Program in Biodesign, Stanford University, Stanford, California

Intravascular Ultrasound Hiroyoshi Yokoi,

MD

Director of Clinical Section, Department of Cardiology, Kokura Memorial Hospital, Kitakyusu, Fukuoka Prefecture, Japan

Small-Vessel and Diffuse Disease Alan Zajarias,

MD

Assistant Professor of Medicine, Cardiology Division, Washington University School of Medicine; Interventional Cardiologist, Barnes-Jewish Hospital, St. Louis, Missouri Percutaneous Closure of Patent Foramen Ovale and Atrial Septal Defect

Khaled M. Ziada,

MD

Assistant Professor of Medicine, Division of Cardiovascular Medicine, and Associate Director, Interventional Cardiology Fellowship Program, University of Kentucky College of Medicine, Lexington, Kentucky Periprocedural Myocardial Infarction and EmbolismProtection Devices

Andrew A. Ziskind,

MD, MBA

Professor of Medicine, Washington University School of Medicine; President, Barnes-Jewish Hospital, St. Louis, Missouri Percutaneous Balloon Pericardiotomy for Patients with Pericardial Effusion and Tamponade

xiii

Preface

Radical, maximally invasive surgery was performed for the preparation of this fifth edition of Textbook of Interventional Cardiology. There are over 30 new chapters and 70 new authors, for a complete revamping of the coverage of the ever-burgeoning field of interventional cardiology. Section 1, Patient Selection, which is now considered highly important, is new to this edition. As percutaneous interventions have supplanted surgical ap-proaches for many types of patients and anatomical subsets, the risk and benefit assessment is critical. New chapters dedicated to arterial inflammation at baseline, which may be quite pivotal for long-term prognosis; functional testing, especially with multidetector CT angiography; and demographics, such as gender, ancestry, diabetes, and renal disease, have been added to help guide cardiologists in patient selection. An overview of evidence-based practice in interventional cardiology helps pull much of this together. The complexity of coronary interventions has drastically changed, with approaches to left mainstem lesions that are unprotected, complex bifurcations, and diffuse disease now more common. Chapters on these topics, as is the case throughout the book, are written by international authorities. New chapters on transradial intervention and peri-access site management, which may be an important segue to facilitate outpatient intervention, are especially pragmatic. The past year has been checkered in the field, with marked public attention given to late thrombosis of drugcoated stents and the results of the COURAGE trial. Late stent thrombosis, to which a new chapter is dedicated, is certainly a lingering concern that has led to more prolonged dual antiplatelet therapy and a shift in practice in the United States toward more bare metal stents. While the incidence of late stent thrombosis is quite low, we clearly need more information in order to prevent this dreaded complication. The COURAGE trial sparked debate as to whether percutaneous coronary intervention procedures were even warranted as compared with a pharmacologic-only strategy. The trial had major shortcomings, but the most important was the selection of the endpoint of death or myocardial infarction. No prior trial had shown benefit for this endpoint in the history of interventional cardiology, so to anticipate that this could be possible defies any Bayesian or a priori knowledge of the field. While interventional cardiology has been under

fire for these two issues, the hope is that this will settle with the realization that there has been truly remarkable and relentless progress in the field. One of the most exciting frontiers is the transformation of select hospitals into interventional centers of excellence. Two new chapters address this opportunity. One tackles acute myocardial infarction and acute coronary syndromes. The other chapter deals with the concept of stroke centers, performing acute intervention on patients with evolving stroke. What are the other new frontiers for this field? The book delves much more deeply into each type of “big artery,” noncoronary intervention with the lower and upper extremities, mesenteric, renal, carotid, and cerebrovascular arterial beds, along with venous interventions. This is a major difference from the last edition—the practice of interventional cardiology now extends to virtually all of the major artery beds. Certainly intracardiac intervention is a promising new dimension, with intracardiac echo, left atrial appendage closure, and percutaneous repair of the mitral valve or aortic valve. Using catheter-based intervention for stem cell therapy, regeneration therapy, or angiogenesis are particularly topical and important research paths. And the same applies for detection of vulnerable plaque and the controversy of whether nonobstructive inflamed segments of arteries should undergo intervention to preempt plaque fissure, erosion, or rupture. All of these topics are covered in newly added chapters. The chapters on quality of care and regulatory issues also are new and present salient perspectives on the practice and regulatory aspects of the field. Cumulatively, this book not only has hopefully tracked the progress in the field but also has provided a futuristic perspective. Compared with the field when the first edition of this textbook was published in the 1980s, when all there was to work with were relatively primitive balloon angioplasty catheters and a bit of roulette as to whether a major coronary dissection would be induced, the practice of interventional cardiology today is unrecognizable. Rarely is just a balloon used, the procedure is almost invariably calm and predicatable, and now the real interventional cardiologist is “pan-vascular” and evolving to practice an “intracardiac” genre, facile in all of the noncoronary vasculature procedures including the ability to close a patent foarmen ovale or left atrial appendage, or perform transcatheter valve repair.

xv

xvi

Preface Of the five editions of this book, I believe this one has captured and anticipated the field better than any other. I am especially grateful to the 125 authors from all over the world who shared their expertise and have put together an unprecedented reference source for our field. Michael Goldberg and his book production team at Elsevier have been formidable supporters, providing an exceptional layout; Natasha Andjelkovic, executive publisher, and Agnes Byrne, developmental editor, also at Elsevier, were most helpful in getting this project off the ground, along with my prior editorial assistant, Donna Wasiewicz-Bressan. I also want to express my deepest

thanks to my friend and colleague Dr. Paul Teirstein, who has shown me a whole new level of interventional cardiology since my arrival in La Jolla. We all hope that the interventional cardiologist will find this a particularly useful reference source for what still remains the most remarkable discipline in medicine—one in which immediate gratification for patients can be achieved, and longterm imaginative solutions to complex challenges just keep accruing at a breakneck pace.

Eric J. Topol

CHAPTER

1

Inflammation Status Saif Anwaruddin and Eric J. Topol

KEY POINTS 䊏 Inflammation should be considered a risk factor for coronary atherosclerosis and acute coronary syndromes (ACS). 䊏 Molecular biomarkers may help to define the inflammatory state and to describe characteristics such as plaque morphology and acute thrombosis. 䊏 Within interventional cardiology, despite technologic advances, inflammation still remains a limiting factor in terms of restenosis, stent thrombosis, microembolization, and so on. 䊏 Adjuvant medical therapy has improved outcomes in interventional cardiology, and this may, in part, be

Although atherosclerosis and acute coronary syndromes (ACS) are related, they remain distinct entities from both a pathophysiologic and a clinical standpoint. Many patients develop severe atherosclerotic disease of coronary vessels but never experience an ACS, whereas others die of an ACS without symptomatic evidence of significant antecedent coronary atherosclerosis. This heterogeneity probably reflects differences in genetic heritability and environmental factors that, in turn, contribute to differences in both the predisposition and the response to injury.1 As our understanding of these entities improves, evidence supporting the role of inflammation as a central component of both of these processes continues to accumulate. Inflammation appears to be integral in the induction and propagation of atherogenesis and atherothrombosis. The perpetuation of atherosclerosis by inflammation is also a concept that has important implications, both for the identification of patients at risk and in the treatment of clinically apparent disease. The presence of clinically detectable levels of inflammation in otherwise asymptomatic patients should be considered a harbinger of potentially adverse outcomes. Current emphasis on risk factor modification focuses primarily on the traditional coronary artery disease risk factors, including smoking, dyslipidemia, hypertension, and the presence of diabetes. Despite affecting the response of patients with atherosclerosis and ACS to various therapies and ultimately dictating their clinical course, inflammation has emerged as a risk factor that needs to be addressed and modi-

due to potent anti-inflammatory properties of these drugs. 䊏 The optimal timing of pretreatment with adjuvant medical therapy in both elective and emergent percutaneous coronary intervention (PCI) has been shown to be of significance and may be related to controlling the inflammatory response to injury from PCI. 䊏 Future directions in interventional cardiology will focus on device-based therapies, but there will also be a need to discover newer drug therapies that allow for more effective and efficient methods of modulating inflammation.

fied. How to accomplish this task is a question is of tremendous value, not only in terms of preventive strategies, but also for currently available percutaneous coronary interventions (PCI). Controversy exists regarding the best way to define inflammation as a risk factor in otherwise healthy patients and in those with preexisting coronary artery disease. Although C-reactive protein (CRP) has been extensively studied, both as an independent marker of risk and as an active participant in the process, some researchers have questioned the clinical value of this marker. The focus on a single vulnerable plaque is only the tip of the proverbial iceberg. The ACSs seem to be driven by inflammation, a process that more globally affects the entire coronary tree. Although treating individual “unstable” plaques remains enticing, a more complete approach to interventional therapy must be employed, with an emphasis on treating the vulnerable lesion in the context of the broader inflammatory component. As details of the underlying molecular and genetic mechanisms become more apparent, treatment of atherosclerosis and ACS may ultimately become more individually tailored. Understanding the connection between inflammation and thrombosis is vital to achieving an appreciation of the pathophysiology behind ACS and coronary artery disease. To better serve our patients with this information, viable methods of quantifying arterial inflammation as a modifiable risk factor need to be developed and utilized. Ultimately, this information can help rationalize treatment strategies in order to overcome current obstacles within interventional

3

4

Patient Selection Markers of inflammatory status in coronary disease MPO

Marker of plaque vulnerability

IL-18

Plaque morphology

MMPs

Development of plaque instability

HGF

Evidence of acute thrombosis

Th1/Th2 balance CRP

Proinflammatory vs anti-inflammatory balance Systemic inflammatory state

Figure 1-1. Markers of inflammatory status in coronary disease. CRP, C-reactive protein; HGF, hepatocyte growth factor; IL-18, interleukin-18; MMPs, matrix metalloproteinases; MPO, myeloperoxidase; Th, helper T cells.

cardiology. With the use of drug-eluting stents (DES) and adjuvant medical therapy, this process is already underway, but it continues to evolve. The challenge is to overcome the limitations, including restenosis and thrombosis, that have important inflammatory underpinnings. This chapter provides an overview of the complex inflammatory components that contribute to atherothrombosis and how percutaneous strategies induce or are influenced by arterial inflammation.

INFLAMMATION, ATHEROSCLEROSIS, AND ATHEROTHROMBOSIS Quantifying Inflammatory Status The evaluation and modification of clinical predictors in patients with known coronary atherosclerosis, or in those at risk for developing atherosclerosis or atherothrombosis, is well established. The limitation of these traditional clinical predictors lies in their inability to adequately incorporate other elements, such as inflammation. Attempts to quantify the degree of inflammation and its significance require understanding of the underlying molecular factors involved in the inflammatory and thrombotic processes (Fig. 1-1). Molecular biomarkers of inflammation can be used to predict the future risk of clinical events or to evaluate an appropriate response to therapy. Their use may facilitate targeted therapeutic strategies based on a comprehensive molecular risk profile rather than simply on clinical characteristics. The challenge remains in being able to accurately define and measure the inflammatory state. Although many candidates have been considered, only a select number are supported by the available clinical data. Even fewer have been rigorously evaluated in large-

scale clinical studies to ensure their utility and to confirm their value. Candidate markers need to undergo a meticulous process of evaluation to examine their worth in the clinical context, including an assessment of their practicality, their costeffectiveness, and whether they add information beyond that which is already known. Furthermore, whether specifically targeting these markers with medical therapy affects clinical outcomes remains to be seen. C-reactive Protein Traditionally defined as an acute phase reactant, CRP has achieved recognition as a marker of inflammation. The value of high-sensitivity CRP (hsCRP) as a marker of systemic inflammation is in its ability to predict cardiovascular risk. Extensive large-scale epidemiologic data exist that support the ability of CRP to predict the risk of future cardiovascular events in otherwise healthy individuals,2,3 in those with unstable angina, and in patients who have undergone PCI procedures. The association between CRP and cardiovascular events only strengthens the importance of inflammation in ACS and atherosclerotic disease. Although CRP is produced in the liver, there is ample evidence to suggest that it is actively involved in atherosclerotic disease. Its role in upregulation of adhesion molecule expression in endothelial cells and in controlling macrophage recruitment lend support to its involvement in atherosclerosis and ACS. In addition, autopsy studies have demonstrated CRP immunoreactivity in plaques with vulnerable morphology.4 Therefore, CRP represents an attractive target for medical therapy, in both primary and secondary prevention strategies. However, studies have not addressed CRP as a treatable risk factor per se, but have focused instead on the secondary effects of treatment on CRP levels. An ongoing study, the Justification for the Use of Statins in Primary Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER), is attempting to address this question.5 Platelets: Mediators Of Inflammation and Thrombosis Platelets are central to the processes of atherosclerosis and ACS, because they provide a link between inflammation and thrombosis. In the context of ACS and PCI, the effect of platelet inhibition probably extends beyond the ability to inhibit thrombosis, involving regulation of platelet-mediated inflammation (Fig. 1-2). The interactions among platelets, endothelium, and leukocytes facilitate the process of plateletmediated inflammation and thrombosis. Antiplatelet therapy continues to be an effective method of preventing thrombosis; however, a significant benefit may occur through modulation of the inflammatory properties inherent to platelet function. Considerable interest has been generated for the platelet-derived CD40 ligand (CD40L) and soluble CD40 ligand (sCD40L) in the context of atherosclerosis and ACS. Although it was originally thought to

Inflammation Status vWF

}

Platelet activation GP-1b


lllbB3

sCD40

Figure 1-2. Platelets as mediators of inflammation. αIIbβ3, a glycoprotein receptor; CD40 R, CD40 receptor; GP1b, glycoprotein-Ib; IL-8, interleukin-8; MCP-1, monocyte chemoattractant protein-1; PSGL, P-selectin glycoprotein ligand; sCD40, soluble CD40; vWF, von Willebrand factor. (From Anwaruddin S, Askari A, Topol EJ: Redefining risk in acute coronary syndromes using molecular medicine. J Am Coll Cardiol 2007;49:279-289.)

Leukocyte P-Selectin Peroxynitrates

PSGL IL-8, MCP-1 Platelet-endothelial cell interaction

• Inhibits endothelial repair • Increased thromboxane levels • Prevents endothelial cell migration • Prevents angiogenesis

be involved in cellular development within the context of humoral immunity, CD40L has been found on other cell types, including eosinophils, T cells, basophils, and monocytes, in both bound and soluble forms. Commensurate with its widespread distribution is the myriad of functions CD40L participates in related to atherogenesis and ACS. By facilitating direct interaction with endothelial cells, platelet-bound CD40L is pro-inflammatory and has been shown to upregulate cellular adhesion molecules, increase secretion of chemokines,6 and increase tissue factor production. Facilitating the activation of monocytes,7 in combination with its other roles, places CD40L in the center of the atherosclerotic process. In addition to potent pro-inflammatory effects, CD40L regulates the development and stability of thrombus in ACS. Stability is maintained by the interaction between the lysine-arginine-glutamic acid domain of the CD40L and the platelet αIIbβ3 receptor.8 Whereas inhibition of CD40L results in more stable plaque morphology, CD40L left unchecked engages in destabilizing activities. When anti-CD40L antibody was administered to apoE −/− mice treated with anti-CD40L, a reduction in plaque lipid and inflammatory content was noted, without any effect on the size of the lesion.9 CD40L is also influential in the production and release of matrix metalloproteinases (MMPs), which are thought to be responsible for degradation of the fibrous cap of the atheromatous plaque. These effects of CD40L support its role in plaque instability and atherothrombosis. Clinically, the ratio of sCD40L to CD40L provides important prognostic value. In patients presenting with ACS, elevated sCD40L was an independent predictor of death and recurrent myocardial infarction (MI).10 Data from the Dallas Heart Study noted that sCD40L is not a marker of clinically silent atherosclerotic disease, nor is it associated with traditional risk factors for coronary atherosclerosis, suggesting a separate inflammatory process involved in the genesis

CD40 R

Adhesion molecules

of an ACS.11 A large, placebo-controlled randomized trial of platelet glycoprotein inhibitors in PCI for ACS showed particular benefit among those patients with elevated sCD40L.12 The interaction and communication among platelets and leukocytes is essential to the process of inflammation and its sequelae. Inasmuch as the interaction between platelets is emphasized in this scenario, leukocyte and platelet interactions are also vital to the development of thrombosis in ACS. One of the key intermediaries between platelets and leukocytes is P-selectin and its interaction with the Pselectin glycoprotein ligand (PSGL). In experimental models, P-selectin was shown to be important to the processes of thrombosis and thrombus stability.13 P-selectin potentially represents an important target for therapy, given its presence in thrombosis. Clinically, in patients presenting with chest pain, elevated levels of P-selectin were predictive of future troponin I positivity.14 In apparently healthy women, P-selectin levels were predictive of future cardiovascular events.15 However, direct measurement of platelet-monocyte aggregates may represent a more sensitive marker than P-selectin, and it remains to be seen whether P-selectin will be of value as a true connection between inflammation and thrombosis. Leukocytes and Inflammation in Acute Coronary Syndromes and Coronary Atherosclerosis The inflammatory responses leading to the disruption of plaque in ACS and subsequent events is characterized by a varied cellular presence. The relationship between monocyte-derived macrophages and the pathogenesis of atherosclerotic coronary artery disease has been well studied. The importance of neutrophils, lymphocytes, and mast cells in plaque disruption and thrombosis has become apparent. The value of leukocytosis in acute myocardial infarction may extend beyond simple prognosis and may predict patient response to revascularization strategies.

5

6

Patient Selection

Mast cell degranulation HGF

MMPs

IFN-

T Cell

VCAMs

IL-18

MPO: • Oxidized LDL • Breakdown NO • “Marker of vulnerable plaque”

MCP-1: Summons macrophages/monocytes to sites of injury

Several studies have assessed the prognostic value of leukocytosis in the setting of ACS, supporting a relationship between leukocytosis and adverse cardiac events during hospitalization for acute MI.16 In ACS, the presence of neutrophils is being recognized as an important component of acute plaque rupture. Furthermore, elevated neutrophil counts in those with acute MI are associated with suboptimal angiographic results after fibrinolysis.17 The monocyte-macrophage is central to the events leading to formation of the atherosclerotic plaque and to promoting ongoing inflammation, which may trigger an ACS via an array of leukocyte secretory products. These leukocyte secretory products serve to provide a potential mechanistic link between inflammation and the pathogenesis of atherothrombosis and atherosclerosis (Fig. 1-3). Myeloperoxidase Production and release of myeloperoxidase (MPO) from granules is known to occur in both neutrophils and monocytes. MPO is involved in several processes

Figure 1-3. Leukocyte secretory products. HGF, hepatocyte growth factor; IL-18, interleukin-18; INF-γ, interferon-γ; LDL, low-density lipoprotein cholesterol; MCP-1, monocyte chemoattractant protein-1; MMPs, matrix metalloproteinases; MPO, myeloperoxidase; NO, nitric oxide; VCAMs, vascular cell adhesion molecules. (From Anwaruddin S, Askari A, Topol EJ: Redefining risk in acute coronary syndromes using molecular medicine. J Am Coll Cardiol 2007;49:279-289.)

that modulate atherogenesis and coronary inflammation. For example, MPO has the ability to oxidize low-density lipoprotein (LDL) cholesterol, to break down nitric oxide (NO), to regulate endothelial homeostasis, and to modulate NO function in inflammatory processes.18 Elevated levels of MPO have been demonstrated in patients with coronary artery disease and have been implicated as having a role in plaque destabilization.19 An elevated level of MPO at the time of presentation with an ACS was associated with a worse prognosis irrespective of troponin T level,20 emphasizing the importance of the underlying inflammatory state. It is plausible that the presence of persistent inflammation may contribute to future events. Therefore, the role of MPO as a “marker of the vulnerable plaque” even in the troponin-negative ACS patient has been suggested.21 MPO serves to highlight the role of neutrophils in initiation of the events leading to ACS and events occurring immediately after MI. In addition to prognostic value, it provides a direct indication of underlying plaque instability in those with suspected ACS.

Inflammation Status Interleukin-18 Interleukin-18 (IL-18) is a cytokine that is capable of inducing production of interferon-γ (IFN-γ) in T lymphocytes and of supporting differentiation of the TH1 subset of helper T lymphocytes. IL-18 production is increased by stimulation from IL-1β. The resultant increase in IL-18 and the binding to the IL-18 receptor on macrophages and vascular smooth muscle cells results in upregulation of IFN-γ, MMPs, various cytokines, and vascular cell adhesion molecules (VCAMs). Furthermore, IL-18 has been found in atherosclerotic plaques. Increased IL-18 expression results in a “vulnerable plaque morphology,” defined by thin cap atheroma and more intraplaque hemorrhage.22 However, the exact mechanism by which IL-18 contributes to atherosclerotic disease remains controversial. Although increased IL-18 expression has been noted in patients with ACS as well as in those with known coronary artery disease who are at increased risk of death from cardiovascular causes, the role of IL-18 in ACS is still undetermined and needs to be understood in relation to other factors. The ratio of IL-18, a pro-inflammatory cytokine, to that of IL-10, an anti-inflammatory cytokine, may be more important than the actual quantity of IL-18 itself.23 Nevertheless, future investigation is necessary before the utility of this cytokine is fully understood. Matrix Metalloproteinases MMPs serve to regulate the extracellular environment through breakdown and proteolysis of matrix components, so as to facilitate a favorable environment for cellular development. MMPs are produced in a propeptide form and undergo eventual cleavage in the extracellular environment. MMPs have been implicated in processes such as neointimal hyperplasia, left ventricular remodeling, and formation of vascular aneurysms. MMP-9, or gelatinase B, and MMP-2, or gelatinase A, are thought to be involved in the development of plaque instability. Local release of these factors is believed to degrade the fibrous cap of the atherosclerotic plaque. Evidence to support the involvement of MMPs in ACS has come from small clinical studies. Higher blood levels of MMP-2 and MMP-9 have been noted in ACS patients compared with healthy controls. Although elevated MMP-9 concentrations were associated with an increased hazard ratio of cardiovascular death after adjusting for clinical confounders,24 what was not entirely evident was whether MMP-9 provided prognostic information beyond that conveyed by other biomarkers of inflammation typically examined in ACS. Although preliminary work with MMPs is suggestive of their participation in ACS, future translational and clinical investigations need to examine not only MMPs but also their relationship with tissue inhibitors of metalloproteinases (TIMPs). This regulatory association is important, and an improved understanding of it may

provide more accurate and more valuable insight as to the significance of MMPs/TIMPs in coronary atherosclerosis and ACS. Pregnancy-Associated Plasma Protein A Pregnancy-associated plasma protein A (PAPP-A), a zinc-binding metalloproteinase that is secreted by activated macrophages, fibroblasts, vascular smooth muscle cells, osteoblasts, and placental syncytiotrophoblasts, functions to activate insulin-like growth factor-1 (IGF-1) through actions on IGF-binding protein (IGF-BP). Although the role for PAPP-A as a biomarker for Down syndrome during pregnancy is well established, its potential role with respect to coronary atherosclerosis and ACS has only recently been recognized. PAPP-A has been found in higher concentrations in unstable plaques from patients dying as a result of ACS. In addition, circulating levels of PAPP-A have been shown to be significantly higher in patients with unstable coronary syndromes versus stable angina.25 Furthermore, an elevated PAPP-A level was an independent predictor of a 6-month combined primary end point of cardiovascular events, including mortality, in a study of troponin I–negative patients presenting with ACS.26 Although the findings are quite preliminary, both PAPP-A and the ratio of PAPP-A to proMBP27 have correlated with extent of coronary atherosclerosis in stable angina. Although PAPP-A may be present in the unstable plaque in ACS and in peripheral blood of patients with coronary atherosclerosis, its role remains undefined. Some have postulated that, through proteolytic breakdown of IGF-BP activating IGF-1, PAPP-A is able to mediate pro-atherogenic functions and may participate in local inflammatory processes.28 Conversely, IGF-1 may be protective in coronary and systemic vascular disease, and lower IGF-1 levels may actually predict adverse cardiac events.29 PAPP-A may simply be a marker of atherosclerotic disease and not directly involved in the pathogenesis.30 Although these possibilities remain intriguing, it is premature to advocate for the clinical use of PAPP-A in the absence of a proven, pertinent mechanism. Hepatocyte Growth Factor Hepatocyte growth factor (HGF), a growth factor originally thought to be important in cellular growth and development, possesses characteristics that underscore a potential purpose in ACS. In the context of acute plaque rupture and thrombosis, there appears to be a relationship between thrombus formation and release of HGF. This appears to be dependent on the presence of mast cells as an intermediary. Perhaps, through activation of the thrombin receptor on mast cells, release of heparin leads to elevated HGF release from the extracellular matrix. Clinically, significantly higher levels of HGF were demonstrated in patients with chest pain and evidence of acute thrombosis (from ACS, aortic

7

8

Patient Selection dissection, or acute pulmonary emboli) compared with those without evidence of thrombosis.31 Similar findings were noted in patients who presented with cerebral infarction and unstable angina. It has been suggested that elevated levels of HGF may be protective in the ACS setting. Although HGF release may be a direct response to acute thrombosis and subsequent myocardial injury, larger studies are needed to validate these findings and to determine whether any relationship exists between levels of HGF and clinical prognosis. T Cells and Interferon-γ Supporting the inflammatory basis for atherosclerosis and ACS is the involvement of specific subsets of T lymphocytes. Helper T (TH) lymphocytes are central mediators of inflammation. TH1 subsets are proinflammatory and express cytokines such as IFN-γ, IL-2, and tumor necrosis factor-β (TNF-β). In contrast, TH2 cells are responsible for regulating humoral immunity and attenuating inflammation. That T cells are involved in the progression of atherosclerosis has been shown by transfer of CD4+ T cells into B cell– and T cell–deficient apoE−/−/scid/scid mice, which resulted in worsening of atherosclerotic disease and upregulation of IFN-γ secretion.32 In ACS, higher levels of T-cell activation have been noted and are thought to be independent of ischemic injury. Significantly higher levels of TH1 CD4+ cells were observed in patients with unstable angina, compared with controls. Similar findings were noted in patients with acute MIs.33 The functional significance of the presence of TH1 cells in unstable coronary syndromes may be related to expression and release of IFN-γ. Downstream activation of monocyte-macrophages may result from TH1-mediated IFN-γ release. Upregulation of IFN-γ– inducible genes in monocytes occurs in unstable angina, further supporting TH1-mediated activation of macrophages in ACS.34

A GENETIC BASIS FOR INFLAMMATION The ability to quantify and eventually treat inflammation as a risk factor for cardiovascular disease may hinge on a proper understanding of the determinants of inflammation. A complex phenotype such as inflammation is strongly influenced by genetics, which may help explain the variation in inflammatory status among those with coronary disease. The discovery of myocyte enhancer factor 2A (MEF2A) and its relationship to MI illustrated the intricate relationship between genetics and clinical cardiology.35 In addition, specific haplotypes in the gene encoding for the 5-lipoxygenase activating protein (FLAP) have been shown to confer an increased risk of MI and stroke in selected populations.36 FLAP appears to modulate the production of pro-inflammatory intermediaries such as leukotriene B4 that are believed to be of significance in coronary disease. Furthermore, upregulation of the leukotriene pathway

is also associated with a haplotype of the leukotriene A4 hydrolase (LTA4H) gene, which also confers a risk of MI.37 Targeting these pathways based on genetic risk could represent a novel method of identifying and treating inflammation as it relates to MI and coronary atherosclerosis. FLAP inhibitors have been used in a prospective, randomized fashion to demonstrate a reduction in inflammatory biomarkers among those carrying haplotypes associated with increased risk.38 As the tools used to facilitate discovery of such markers improve, so will both our understanding of the complex nature of inflammation as it relates to coronary disease and our ability to modulate it.

INFLAMMATION AND PERCUTANEOUS CORONARY INTERVENTION The treatment of stable yet symptomatic coronary atherosclerosis and unstable coronary atherothrombosis has progressed from balloon angioplasty to the modern era of DES and will continue to evolve. Adjunctive medical therapy in the form of statins, glycoprotein inhibitors, and thienopyridines and antithrombin agents has also been a mainstay of treatment aimed at controlling risk factors and progression of disease, both at the time of intervention and afterward. The ultimate success or failure of PCI and management of coronary atherosclerosis varies among individuals. Beyond technical considerations, revascularization has been and will continue to be limited in efficacy by one major variable—inflammation. The focus of treatment must address inflammation to improve outcomes in PCI. Many of the challenges faced by interventional cardiologists are rooted in problems related to inflammation, a process mediating the response to injury after insult to the endothelial integrity imparted by PCI. The resultant neointima formation has been noted to be a significant problem and occurs as a result of cell death and inflammation. Although DES have been able to temper this process, they are neither a definitive nor a perfect solution.39,40 Inflammation is a more prominent force, not only in the initiation and propagation of disease, such as in ACS, but also in response to injury after PCI. Inflammation as a Response to Injury: Pathobiology and Clinical Significance The pathobiology of the arterial response to injury in the form of PCI with stent deployment has been extensively examined. Stenting leads to an acute inflammatory response, followed by chronic inflammation. Within minutes after stent deployment, an intense reaction to injury consists of platelet activation and accumulation, expression of adhesion molecules, leukocyte recruitment, and thrombus formation. The degree of injury likely determines the resultant inflammatory response and eventual restenosis (Figs. 1-4 and 1-5). In autopsy studies, pathologic data suggest that stent deployment leading to

Inflammation Status

Figure 1-4. Angiogram demonstrating in-stent restenosis of a sirolimus-eluting stent in the middle left anterior descending artery (arrow).

assault, as does leukocyte adhesion to endothelial cells and to platelets. The expression of neutrophil adhesion molecules, particularly the integrins CD11b/ CD18 (now ITGAM/ITGB2) known as the membrane attack complex (MAC-1), increases after PCI with bare metal stents (BMS) in patients with single-vessel coronary atherosclerosis and is strongly correlated with the risk of restenosis. The correlation between MAC-1 expression, as a surrogate marker of leukocyte activation, and neointimal hyperplasia has been examined in MAC-1 −/− mice. After endothelial denudation, a limited leukocyte presence and a reduction in the degree of neointimal hyperplasia were noted.43 The response-to-injury hypothesis appears to involve a complex interplay among platelets, leukocytes, fibrin, and other components. The release of a multitude of cytokine mediators, such as MCP-1, IL6, IL-1, and TNF-α, appears to coordinate this effort. The end result is neointimal hyperplasia leading to restenosis. The potential exists to establish molecular targets, such as MAC-1, with the aim of developing specific therapies to combat inflammation. The significance of these findings is highlighted by the influence of inflammation on clinical outcomes. Leukocytosis, a nonspecific surrogate marker of inflammatory response across the spectrum of ACS, has been demonstrated to be an ominous sign. In the setting of PCI, peak circulating monocyte count has been shown to be associated with angiographic restenosis at 6 months.44 Biomarkers of Inflammation and Percutaneous Coronary Intervention

Figure 1-5. Intravascular ultrasound (IVUS) demonstrating severe in-stent restenosis.

arterial medial fracture, particularly deep into lipidrich plaque, is associated with a higher degree of inflammatory infiltrate, increased neointimal thickness, and neoangiogenesis.41 Clinical factors associated with restenosis, particularly in diabetics, include longer stent length, active tobacco use, smaller arterial reference diameter, and inflammatory state as determined by CRP level.42 Leukocyte infiltration as a response to injury remains an important feature of this inflammatory

Inflammatory biomarkers have also been used to quantify systemic inflammation to assess the relation between inflammation and restenosis in PCI. Preprocedural levels of sCD40L were examined prospectively and found to be predictors of restenosis at 6 months in patients undergoing PCI for stable angina.45 CRP elevation is more commonly used as a measure of inflammatory status after coronary stent implantation. Levels of hsCRP were shown to rise across a translesional gradient, both in patients with angina and in those who had undergone PCI, suggesting local production of CRP or increased local release of CRP-rich thrombus.46 In patients with stable coronary disease who underwent PCI, elevated postprocedural levels of CRP were noted and supported a robust inflammatory response. The rationale of predicting outcomes using preprocedural measures of inflammation remains controversial, because pre-PCI levels of CRP and IL-6 have not been shown to correlate with in-stent restenosis after PCI.47 In 483 patients with stable or unstable angina who underwent coronary intervention with BMS, elevated CRP and lipoprotein(a) predicted adverse cardiac events at 1 year, but the association did not hold for in-stent restenosis.48 Prospective investigation of 276 patients who had undergone PCI with BMS for both stable angina and unstable

9

Patient Selection 32

P = .002

P = .01 CRP tertiles Upper tertile Middle tertile Lower tertile

24 Incidence, %

10

16

8

0 Angiographic restenosis (diameter stenosis 50%)

Clinical restenosis (target vessel revascularization)

Figure 1-6. Incidence of angiographic and clinical restenosis in three groups defined by the change in C-reactive protein (∆ CRP) after percutaneous coronary interventions. (Redrawn from Dibra A, Mehilli J, Braun S, et al: Inflammatory response after intervention assessed by serial C-reactive protein measurements correlates with restenosis in patients treated with coronary stenting. Am Heart J 2005;150:344-350.)

coronary syndromes demonstrated that preprocedural CRP levels were predictive of increased rates of restenosis and worse clinical outcomes after adjusting for the presence of unstable coronary disease.49 Given the available data, it is difficult to make definitive conclusions regarding the utility of the preprocedural inflammatory state, particularly in the population of patients with stable angina. In ACS, however, an assessment of the baseline inflammatory state may provide valuable prognostic information. Clinically, a rise in postprocedural CRP has been shown to correlate with in-stent restenosis at 6 months after PCI. In 1800 patients with either stable or unstable angina undergoing PCI, peak postprocedure CRP level strongly correlated with both angiographic and clinical restenosis (Fig. 1-6). Those patients in the highest tertile of postprocedure increase in CRP level also had higher 30-day rates of stent thrombosis, death, MI, or target vessel revascularization.50 Alternatively, a return of post-PCI CRP levels to baseline 72 hours after intervention was highly predictive of event-free survival over a 1-year period in a prospective study of 81 consecutive patients with one-vessel stable angina. Therefore, the postprocedural rise in CRP appears to provide a consistent correlation with future risk of restenosis.

GENETICS, INFLAMMATION, AND RESTENOSIS Inflammation as a response to injury in the post-PCI population is not a universal phenomenon and begs the question of individual genetic susceptibility. The

presence of genetic polymorphisms may help define susceptibility and may affect the selection of therapy for certain patients undergoing PCI. The inflammatory response may, in part, be determined by underlying genetic predisposition. The value of such information, beyond traditional risk factors of restenosis (e.g., diabetes), will be determined in studies of large populations. The Genetic Determinants of Restenosis (GENDER) project is one of the studies that have set out to examine possible genetic risk factors. Although the information generated from these studies is only preliminary, it accomplishes two tasks—the first is to generate hypotheses regarding possible novel mechanisms for restenosis, and the second is to encourage larger-scale studies to validate these ideas. The data from most of these studies is obtained in smaller populations, so the possibility exists that these associations between clinical events and polymorphisms may not exist. Furthermore, most of these studies are of selected populations, and generalization to other ethnic groups or races may not be appropriate. Even where the associations have been proven to be robust in large numbers of subjects, genetic epidemiologic studies do not elucidate pathophysiologic mechanisms. Polymorphisms have been examined in many different components of the complex array of factors involved in restenosis (Table 1-1). Inflammatory markers such as interleukins, selectins, MMPs, and proteins involved in platelet aggregation and the renin-angiotensin system have been examined, among others. Whether the same polymorphisms are responsible for restenosis in those with DES remains to be seen; however, the idea of a genetic basis for this inflammatory response to injury may allow for both selectively directed therapy and the development of newer therapies directed specifically at known genetic targets.

INFLAMMATION AND DRUG-ELUTING STENTS Clinical restenosis occurs as a result of both injury to the vessel and the underlying atherosclerotic and inflammatory burden. With the introduction of the sirolimus-eluting (Cypher [Cordis, Johnson and Johnson[) stents and the paclitaxel-eluting (Taxus [Boston Scientific]) stents, potent anti-inflammatory drugs have been applied with the aim of reducing the inflammatory response to injury. Large clinical trials support a reduction in the rate of clinical restenosis with DES compared with BMS.39,40 The relationship between the inflammatory milieu and the DES is a complex one, at best. Attempting to understand this dynamic may help to define the limitations of these stents and how best to use them. Defining the inflammatory state through the use of baseline levels of biomarkers or changes in levels is perhaps a simplistic estimation of the events that constitute a complex process. An appreciation of the relationship between DES and inflammation first requires an understanding that the inflammatory

Inflammation Status Table 1-1. Genetic Polymorphisms and the Risk of Restenosis Polymorphism

Study

Follow-up

N and Procedure

Clinical Outcome

Results

Comments

LPL (8p22) polymorphisms: −93/T/G Ser447Ter Asp9Asn Asn291Ser Angiotensinogen: 235Met/Thr, T174M, A(−6)G, AT1R: 1166A/C, T810A AT2R: 1675G/A, 2123A HO-1: Polymorphism in promoter region

Monraats et al86

9 mo

3104 (PCI)

TVR by either PCI or CABG

Ser447Ter associated with decreased risk of TVR after PCI

BMS used. Ser447Ter SNP codes for a stop codon

Wijpkema et al87

9.6 mo

2987 (PCI)

1° end point: TVR 2° end point: clinical restenosis

Relationship still significant after adjustment for use of ACE inhibitors

48 different polymorphisms in 34 genes associated with inflammatory mediators such as adrenergic receptor, various interleukins, CSF, complement, NOS, SDF-1

Monraats et al88

9.6 mo

3104 (PCI)

TVR as defined by PCI or CABG 1 mo after procedure

Fractalkine (FKN) receptor polymorphisms: V249I and T280M

Niessner et al89

365 (PCI)

Toll-like receptor SNPs: TLR-2 Arg753Gln TLR-4 Asp299Gly

Hamann et al90

1° End point: 6 mo 2° End point: 2 yr 6 mo

1° end point: Clinical ISR 2° end point: Clinical re-ISR Angiographic restenosis at 6 mo

Interferon-γ: IFNG T874A IFN-γ receptor: IFNGR1 C–56T IFNGR2 A839G

Tiroch et al91

Variable depending on end point

2591 (stent)

287 with TVR, 327 with 2° endpoint; AT1R 1166CC associated with 1° (P = .007) and 2° (P = .002) end points β2-adrenergic receptor (ADRB2) Gly16Gly variant associated with increased risk of TVR, even after adjusting for diabetes, age, gender, and smoking status Restenosis in 25% of patients; I249 allele associated with restenosis TLR-2 Arg754Gln SNP associated with restenosis after PTCA and stent IFN and IFN receptor genotypes not found to have an association with ISR

SNPs of APOE gene: −219G/T 113G/C, 334T/C, 472T/C

Koch et al92

6-mo angiographic follow-up, 1 yr clinical

1850 (stent)

206 (PTCA) and 182 (stent)

1° end point: Clinical (1 yr) and angiographic (6 mo) ISR 2° end point: Death and nonfatal MI (1 yr) 1° end point: angiographic restenosis at 6 mo and TVR at 1 yr 2° end point: death and nonfatal MI (1 yr)

CD14 (−260T/T), eotaxin gene (−1328A/A), and CSF2 gene (117Thr/Thr) variants were associated with lower risk of TVR Risk is associated with both ISR and recurrent ISR Risk of this SNP and restenosis was higher in females than in males No association found; however, authors attributed lack of observed relationship to complexity of molecular processes

No relationship found between APOE and either end point

ACE, angiotensin-converting enzyme; APOE, apolipoprotein E; BMS, bare metal stent; CABG, coronary artery bypass grafting; CSF, colony-stimulating factor; DES, drug-eluting stent; IFN, interferon; ISR, in-stent restenosis; MI, myocardial infarction; NOS, nitric oxide synthase; PCI, percutaneous coronary intervention; PTCA, percutaneous transluminal coronary angioplasty; SDF-1, stromal-derived factor-1; SNP, single nucleotide polymorphism; TVR, target vessel revascularization.

state exists at multiple levels. Deployment of DES may have different effects on inflammation in the coronary versus the systemic setting. Clinical studies have examined these relationships (Table 1-2). Most support a local inflammatory milieu that is separate from the systemic inflammatory state and suggest that the use of DES may protect the local environment from outside inflammatory forces. A postprocedural rise in markers of systemic inflammation with the use of BMS, although paralleled with the use of DES, predicts outcome with regard to angiographic restenosis only when using BMS (Fig. 1-7). Other data sets suggest a relationship between the systemic

inflammatory response and the type of stent used. The differences between these data sets may be explained by many factors, including differences in timing of serum measurements, use of adjunctive therapies (e.g., heparin, glycoprotein IIb/IIIa inhibitors), the small size of these studies, and the lack of randomization in some. A substudy of the Randomized Trial to Evaluate Relative PROTECTion Against Post-PCI Microvascular Dysfunction and Post-PCI Ischemia Among Anti-Platelet and AntiThrombotic Agents (PROTECT-TIMI 30) noted a significant reduction in CRP and in troponin I rise with DES versus BMS after adjustment for variables

11

Patient Selection Table 1-2. Relationship of Stent Type to Inflammation and Restenosis Study

N

Design

End Points

Results

Comments

Gaspardone et al93

160

Nonrandomized; BMS, DES, or DEXs; low risk

Angiographic restenosis at 12.9 mo, CRP rise after procedure

No differences noted in CRP levels after procedure, but lower rate of restenosis in DES group

Dibra et al94

301

Randomized; SES vs. BMS; low risk

Change in CRP; 6-mo angiographic follow-up for restenosis

Higher restenosis rate in BMS vs. SES

de la TorreHernandez et al95

300

Nonrandomized; SES vs. BMS; low risk

Gogo et al96

75

Nonrandomized; SES vs. BMS; ACS included but no STEMI

Change in CRP; 4-6 mo and 12 mo clinical follow-up assessing for MACE (death, MI, TVR) Change in IL-6, IL-1Ra, CRP 24 hr after PCI

CRP elevation similar in both groups, but the relation to restenosis was noted in BMS No difference in rise of inflammatory markers between DES and BMS

Kim et al97

67

Nonrandomized; DES vs. BMS; lowrisk population

CRP levels at various time points after PCI; 6 mo clinical followup, angiogram at 6 mo

Gibson et al51

665 DES, 139 BMS

Nonrandomized to stent type; NSTEMI patients undergoing PCI

CRP, IL-6, RANTES, sCD40L, TnI, PT F1.2, CK-MB at three time points

Lower CRP levels with DES; greater-diameter stenosis and late lumen loss in BMS, although angiographic follow-up was incomplete No differences noted with regard to sCD40L, RANTES, CK-MB, PT F1.2; rise in CRP and TnI was higher among those with BMS

Although CRP elevation was predictive of restenosis, DES had a much lower rate of restenosis. Suggests that systemic inflammatory state may not be reflective of local environment. Elevated CRP levels in BMS group predicted restenosis, but restenosis in SES group was independent of rise in CRP The relationship between systemic inflammation (CRP) and restenosis holds only for BMS Evidence to support the idea that the systemic inflammatory response does not influence restenosis in DES population Suggests that the use of DES can alter systemic inflammatory conditions

In high-risk population, the benefit of DES may be in its effects on the microcirculation more so than on systemic inflammation

ACS, acute coronary syndrome; CK-MB, creatine kinase MB fraction; CRP, C-reactive protein; DES, drug-eluting stents; DEXs, dexamethasone eluting stents; IL, interleukin; MACE, major adverse cardiac event; NSTEMI, non-ST-segment elevation myocardial infarction; PCI, percutaneous coronary intervention; PT F1.2, prothrombin fragment 1.2; RANTES, regulated upon activation, normal T cell expressed and secreted; sCD40L, soluble CD40 ligand; SES, sirolimus-eluting stent; STEMI, ST-segment elevation myocardial infarction; TnI, troponin I; TVR, target vessel revascularization.

P = .04 30

Angiographic restenosis, %

12

20 P = .37 10

0 Sirolimus-eluting stent group

Bare stent group

Figure 1-7. Angiographic restenosis (percentage) among patients whose postprocedure change in C-reactive protein levels were higher than the median (purple bars) and those whose values were no higher than the median (blue bars) in bare metal stent and sirolimus-eluting stent groups. (Redrawn from Dibra A, Ndrepepa G, Mehilli J, et al: Comparison of C-reactive protein levels before and after coronary stenting and restenosis among patients treated with sirolimus-eluting versus bare metal stents. Am J Cardiol 2005;95:1238-1240.)

such as diabetes, randomized treatment assignment to eptifibatide or bivalirudin, myocardial perfusion, epicardial artery perfusion, and duration of clopidogrel pretreatment. No significant differences were noted between the groups in regard to other markers such as sCD40L, RANTES, creatine kinase MB fraction (CK-MB), and prothrombin fragment F1.2.51 Although this difference is intriguing, it may be more the result of an improved post-PCI microcirculatory status with the use of DES, as well as a reduction in myocardial injury (lower troponin I), and subsequent response to this (lower CRP). Whereas these studies presented varied conclusions with regard to the effects of DES on systemic inflammation, the populations sampled were also quite different in risk and ACS status. The enthusiasm for the reduction in clinical and angiographic restenosis noted with the advent of DES has been tempered by the real concern for stent thrombosis in patients who have received DES (Fig. 1-8). In a large, prospective, observational study of 2239 patients, stent thrombosis complicated the course of 1.3% of patients receiving either sirolimuseluting or paclitaxel-eluting stents. Fourteen patients had subacute stent thrombosis (≤30 days), and 15

Inflammation Status

Figure 1-8. Angiogram demonstrating subacute stent thrombosis of a paclitaxel-eluting stent in the proximal left anterior descending artery (left anterior oblique caudal projection).

patients had late stent thrombosis (>30 days); 13 of these 29 patients died.52 The factors associated with stent thrombosis from this study included premature cessation of antiplatelet therapy, diabetes mellitus, reduced ejection fraction, bifurcation lesions, and renal failure. The “real world” experience has been replicated in other studies and continues to be slightly higher than noted in larger studies. Although the incidence of stent thrombosis at 30 days appears to be similar between DES and BMS, there is concern about the occurrence of late stent thrombosis in the DES population. The duration of antiplatelet therapy in these patients has come into question as a result. It remains difficult to accurately define the true incidence of stent thrombosis, given the various definitions (both clinical and angiographic) used. The pathobiology of DES placement is important to understand, because stent thrombosis may limit widespread use of DES. Postmortem studies in late stent thrombosis of BMS have elucidated possible mechanisms, including plaque disruption at stent borders, and stenting across ostia of branch vessels, among others.53 The drug-release characteristics of the sirolimus-eluting stent and the paclitaxel-eluting stent differ as to the time required for release of the drug, potentially contributing to their safety profiles. Sirolimus is completely released by 6 weeks, whereas paclitaxel-eluting stents are thought to retain drug over a longer period. Although paclitaxel is a useful drug to temper the inflammatory response to injury in the form of restenosis, it has been shown in animal models to delay healing after stent placement, particularly in higher doses.54 Specifically, fibrin deposition, increased vessel wall inflammation, and

intraintimal hemorrhage were noted with paclitaxeleluting stents. Delayed endothelial healing and persistent fibrin deposition, as seen with DES use, leave the segment of the artery vulnerable to thrombosis and highlight the need for potent antiplatelet therapy. Autopsy findings of incomplete endothelialization and fibrin deposition in late stent thrombosis with DES and the presence of eosinophil infiltrates suggest that, while preventing neointimal hyperplasia, DES also affect the normal healing process.55 The concept of delayed healing is seen again in the context of overlapping placement of stents for longer angiographic lesions. Histologic study of overlapping DES has revealed delayed arterial healing, increased inflammatory cellular infiltrate (notably eosinophils), and fibrin deposition. These observations were made in both types of DES, but were greater in paclitaxeleluting stents.56 Delayed healing secondary to higher local content of anti-inflammatory drugs, a hypersensitivity reaction, or both is offered as a possible mechanistic explanation. The decision to cease antiplatelet therapy after DES implantation is problematic, because premature discontinuation may have disastrous consequences. Inflammation complicates the use of stents in the treatment of atherosclerotic coronary disease, as a response to injury in the healing process with the use of BMS and in the context of delayed endothelialization and stent thrombosis with DES. Finding a balance in this complex situation remains a challenge.

ADJUVANT MEDICAL THERAPY: TARGETING INFLAMMATION The pervasive nature of inflammation in the context of ACS and PCI is hard to dispute. Many of the complications and obstacles faced in treating patients with ACS and those undergoing PCI are undoubtedly related to inflammation. Although current therapies employed to treat atherosclerosis and ACS are directed at reducing platelet aggregation, preventing thrombosis, and controlling lipid levels, their ultimate benefit may lie in their ability to modulate inflammation (Fig. 1-9). Quantification of inflammatory risk may help improve the ability to determine prognosis and to treat many patients with ACS and with coronary atherosclerosis. Much of the evidence for this statement has come from clinical and translational experimental work with currently available medical therapy. Although these various adjuvant therapies are not specifically directed at treating inflammation, their secondary effects on established markers of inflammation have been assessed in numerous studies. Statin Therapy (HMG-CoA Reductase Inhibitors) The use of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins) in ACS and coronary atherosclerosis has illuminated the role of the potential nonlipid effects of these drugs. Statin

13

14

Patient Selection • Ability to affect platelet↑ NO synthesis, inhibit isoprenoids, ↑ Th2Th1 leukocyte interactions Regulate inflammatory cyotkine expression, ↓ ICAM expression • Affect degree of circulating markers of inflammation llb/llla Inhibitors Isoprenoid synthesis Statin Nitric oxide IL-18 CD40L IL-6 TNF-a Th2

ICAMs

• Ability to affect platelet-leukocyte aggregation • Reduction of P-selectin expression • Inhibition of TRAP stimulation of protease activated receptors Clopidogrel

therapy is thought to affect inflammatory processes contributing to atherosclerosis. Specifically, statins produce “pleiotropic” effects by actions on isoprenoids. By inhibiting the production of isoprenoids, statins interfere with G protein–mediated upregulation of transcription factors responsible for inflammatory signaling.57 Statins can affect numerous pathways that contribute to inflammation, including adhesion molecules, cytokines, and a variety of inflammatory cell types. A reduction in the levels of ICAM-1/CD54 and CD18/ CD11a by statins has been shown to occur through reduction of messenger RNA (mRNA) transcript.58 Statins have the ability to downregulate leukocyte mRNA for various cytokines such as IL-6, IL-8, and MCP-1. Statins also affect the balance of helper T lymphocytes by favoring a higher proportion TH2 versus TH1 cells.59 By affecting these pathways, statins may hinder progression of atherosclerotic plaque or avert the onset of an ACS. Clinically, statins such as atorvastatin and simvastatin reduce recurrent ischemia after ACS60 and improve mortality in patients with coronary artery disease.61 Benefits attributed to statin use include lipid-independent effects such as reductions in CRP and serum amyloid A (SAA) with intensive statin therapy over a 16-week period.62,63 Although intensive statin therapy has been found to improve clinical outcomes, patients with lower CRP levels as a result of intensive treatment have fewer clinical events, independent of LDL levels, supporting the premise that control of inflammation in ACS leads to a reduction in clinical events.

Adjuvant therapy: Anti-inflammatory properties

Figure 1-9. Anti-inflammatory effects of adjuvant medical therapy in percutaneous coronary interventions. ICAM, intravascular cell adhesion molecule; IL, interleukin; NO, nitric oxide; Th, helper T cells; TNF-a, tumor necrosis factor-α; TRAP, thrombin receptor agonist peptide.

The Reversing Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) study randomly assigned 657 patients to receive intensive statin therapy with either atorvastatin 80 mg or pravastatin 40 mg. Intensive therapy with atorvastatin resulted in a greater reduction in CRP and a corresponding reduction in the progression of atherosclerotic disease as assessed by intravascular ultrasound (IVUS).64 In the context of PCI, statin therapy has been proven to reduce periprocedural complications and to improve short- and long-term outcomes (Table 1-3) (Fig. 1-10). Conceptually, the anti-inflammatory properties of statin therapy may reduce the response to injury induced by PCI and periprocedural complications, including microembolization.65 In contrast to most secondary prevention studies, statin use in the context of PCI has demonstrated a mortality benefit (Fig. 1-11). Multiple studies have also demonstrated a reduction in the periprocedural CK and CK-MB elevations with statin therapy, which have been shown to predict adverse long-term outcomes. A reduction of myonecrosis resulting from pretreatment with statins before PCI may help to explain the reduction in short- and long-term mortality. In the context of PCI, the benefit of statin therapy appears to be distinct from that observed in the secondary prevention trials. The mechanism of statin-induced reduction in myonecrosis may involve a decrease in the postprocedural inflammatory response, as supported by the Atorvastatin for Reduction of MYocardial Damage during Angioplasty—Cell Adhesion Molecules (ARMYDA-CAMs) substudy.66

Inflammation Status Table 1-3. Effect of Statin Therapy on Percutaneous Coronary Intervention Study

N 98

Chan et al99

Chan et al

Study Design

Population

Outcomes

Results

Comments

1552

Prospective, nonrandomized

Elective or urgent PCI

MI or death at 1 yr

Statins were especially effective in those with hsCRP in highest quartile

5052

Prospective, nonrandomized; propensity analysis used

Excluded MI or cardiogenic shock

Death at 30 days and at 6 mo after PCI

Pretreatment with statins was associated with reduced mortality at 1 yr and less periprocedural MI Statin therapy remained independent predictor of survival at 30 days and 6 mo Reduction in 1-yr mortality and restenosis; improvement in LV function

Hong et al100

202

Prospective, randomized

Low EF, PCI for AMI

Death, AMI, CVA, angiographic restenosis, repeat PCI/CABG, LV function

Gaspardone et al101

223

Prospective, observational, nonrandomized; three groups (pretreatment, post-treatment, none)

Stable angina, normal LV function, baseline normal CRP level

CRP level at 24 and 48 hr and primary combined end point of death, MI, and TVR at 6 mo

Reduction in primary end point with pretreatment and post-treatment groups driven by TVR

Prospective, observational

Excluded patients >80 yr, cardiogenic shock, inhospital mortality, and life expectancy 15 hr) placebo at 3 hr was associated (minimum) to 24 hr with a significant (maximum) before reduction in the PCI primary end point 600 mg of clopidogrel No benefit of given at 2-3 hr, 3-6 hr, treatment noted 6-12 hr, or >12 hr beyond 2-3 hr of before PCI pretreatment Pretreatment with For those undergoing 600 mg of PCI, no significant clopidogrel at differences in intervals from 1 to MACE if pretreatment 6 hr before PCI was 30 kg/m2 ≥20 µg/min or albumincreatinine ratio ≥30 mg/g

Men: 88 cm —

Men: 35% Multivessel CAD; angina or ischemia Multivessel CAD; angina or ischemia; LVEF >30%

Medically refractory unstable angina; high CABG risk†

Treatment Groups (N)

Repeat Revasc. (%)

Mortality %

P Value

CABG (33) PTCA (29)

24.2 at 6.5 yr 6.9 at 6.5 yr

.09

32% had single-vessel CAD; stents not used

CABG (30) PTCA (29) CABG PTCA CABG (60) PTCA (64)

10.0 at 3 yr 6.9 at 3 yr 24.5 at 8 yr 39.9 at 8 yr 12.5 at 4 yr 22.6 at 4 yr

NA

Single-center study; stents not used

NA

Stent use rare

19.4 at 5 yr 34.5 at 5 yr 25.6 at 7 yr 44.3 at 7 yr 3.1 at 1 yr 6.3 at 1 yr 4.2 at 3 yr 7.1 at 3 yr 8.3 at 5 yr 13.4 at 5 yr 19 at 1 yr 14 at 1 yr 34 at 5 yr 26 at 5 yr

.003

81% IMA use; stents not used

CABG (180) PTCA (173) CABG PTCA CABG (96) Stent (112) CABG Stent CABG Stent CABG (79) PCI (65) CABG PCI

11.1 at 7 yr 69.9 at 7 yr 3.1 at 1 yr* 22.3 at 1 yr* 8.4 at 3 yr* 41.1 at 3 yr* 27.5 at 5 yr* 42.9 at 5 yr* 35 at 1 yr‡ 49 at 1 yr‡ 46 at 5 yr‡ 51 at 5 yr‡

Comments

.23

.001 .294

89% IMA use; 3.5% glycoprotein IIb/IIIa inhibitor use

.39 .27 .27

54% stent use; 11% glycoprotein IIb/IIIa inhibitor use

.27

CABG, coronary artery bypass graft surgery; CAD, coronary artery disease; IMA, internal mammary artery; LVEF, left ventricular ejection fraction; NA, not available; PCI, percutaneous coronary interventions; PTCA, percutaneous transluminal coronary angioplasty; revasc., revascularization. *Combines absolute rates of repeat CABG and PCI. † Prior heart surgery, myocardial infarction within 7 days, LVEF 70 yr, or balloon pump use. ‡ Includes revascularization or unstable angina. Adapted from Flaherty JD, Davidson CJ: Diabetes and coronary revascularization. JAMA 2005;293:1501-1508.

rates of death, stroke, or MI were 25.0% with PCI and 19.8% with CABG. Diabetic patients treated with stenting had a lower event-free survival rate at 5 years (54.5%) compared with those undergoing CABG (25.0%) owing to the difference in the rates of repeat revascularization (42.9% versus 10.4%, respectively). The mortality rates did not differ (see Table 2-6). A comparison of the 1-year mortality rate among diabetic patients enrolled in the BARI (1996) and ARTS (2001) trials suggests an improvement in outcomes over time. The 1-year mortality rates in the two studies were 6.4% and 3.1%, respectively, with CABG and 11.2% and 6.3% with PCI. Such findings may reflect differences in patient selection or may indeed express an improvement in the revascularization and medical management of diabetic patients. An indirect comparison between PCI and CABG results is also possible using registries. As an example, PCI outcomes of 857 BARI-eligible patients (23% with diabetes) treated within the NHLBI Dynamic Registry were compared with those of 904 patients randomized to PTCA in the BARI trial.78 Stents and GP IIb/IIIa antagonists were used in 76% and 24% of cases, respectively. A dramatic decrease in both abrupt vessel closure (1.5% versus 9.5%) and need for inhospital CABG (1.9% versus 10.2%) was observed in the more contemporary patient group. No difference

in in-hospital mortality was observed. Among diabetic patients, the survival at 1 year within the group of BARI-eligible NHLBI Dynamic Registry patients was similar to that observed in the BARI-CABG group (92.1% versus 93.6%). However, such comparisons must be interpreted with caution, because the favorable outcomes of the registry patients may also be the result of improved medical management. A registry conducted by the Northern New England Cardiovascular Disease Study Group evaluated 5-year mortality rates among patients undergoing revascularization procedures in a large regional database linked to the national death index.79 A subset of 7159 patients with diabetes treated between 1992 and 1996 was examined. Of those, 2766 patients (736 PCI and 2030 CABG) had similar profiles to diabetic patients randomized in the BARI trial. After adjustment for differences in baseline characteristics, patients treated with PCI had significantly higher mortality than those undergoing CABG (HR = 1.5). When stratified for severity of disease, the difference in mortality remained significantly higher for PCI in the setting of three-vessel but not two-vessel disease.79 Similarly, a single-center experience of 2319 consecutive diabetic patients (265 PCI, 2054 CABG) undergoing coronary revascularization in the late 1990s detected a significantly higher 5-year mortality rate

35

36

Patient Selection Table 2-7. Putative Explanations for the Survival Benefit of CABG over PCI in Diabetic Patients More complete revascularization Less myocardium at risk on follow-up Diabetes predicts restenosis after PTCA but not graft failure on follow-up Less disease progression in untreated segments CABG may convey a survival benefit in the setting of subsequent Q-wave MI Risk associated with repeat revascularization due to restenosis after PCI may negatively impact long-term survival CABG, coronary artery bypass grafting; MI, myocardial infarction; PCI, percutaneous coronary intervention; PTCA, percutaneous transluminal coronary angioplasty.

with PCI versus CABG (adjusted HR = 1.7 for noninsulin-treated patients and 2.6 for insulin-treated patients).80 Further comparative analyses of the outcomes of diabetic patients undergoing CABG or multivessel PCI relied on databases of several New York cardiac registries. A total of 37,212 patients (33% with diabetes) undergoing CABG and 22,102 patients (25% with diabetes) undergoing stent-based PCI between 1997 and 2000 were indentified.81 At 3 years, a significant mortality reduction was observed among patients undergoing CABG, with the adjusted HR ranging from 0.59 to 0.71, according to the extension of atherosclerotic involvement. The observed reduction missed statistical significance only in the subgroup of diabetic patients with two-vessel CAD and no involvement of the left anterior descending coronary artery. Explaining the Mortality Benefit of CABG Several hypotheses have been formulated to explain the mortality benefit associated with CABG over PTCA suggested from the randomized trials (Table 2-7).74 The survival advantage of CABG among diabetic patients in BARI was limited to those who received at least one IMA graft. In addition, although diabetic patients in the CABG and PTCA groups had a similar mean number of significant lesions (3.5 versus 3.4), 87% of all intended vessels were successfully bypassed with CABG, but only 76% of vessels with significant lesions were successfully revascularized with PTCA.82 As expression of a less complete revascularization, in the BARI trial diabetic patients had more jeopardized myocardium after PTCA than after CABG.83 In addition, within the PTCA group, diabetic patients a had significantly higher increase in jeopardized myocardium at 1 year compared with nondiabetic patients. These findings are an expression of both restenosis and disease progression in untreated segments. In contrast, among CABG patients, diabetes was not associated with a percentage increase in jeopardized myocardium at angiographic follow-up.83 Furthermore, an analysis of all BARI-eligible diabetic patients (n = 641) revealed that the rate of Q-

wave MI in the first 5 years after revascularization was similar after PCI or CABG (approximately 8% to 9%), but at the same time the associated risk of death was substantially reduced in patients who underwent CABG (adjusted RR = 0.09).84 These results suggest that CABG provided greater protection from death after ischemic events in diabetic patients. Finally, whereas diabetic patients in the BARI trial had markedly greater restenosis after PTCA than nondiabetic individuals, graft patency in the CABG group was not influenced by diabetic status.85 The long-term survival advantage of CABG over PCI in diabetic patients may therefore, in part, result from having a more durable restoration of flow conveyed by CABG without the risk of a repeat revascularization procedure, as was frequently the case in the PTCA group. CABG in the Era of Drug-Eluting Stents So far, no study comparing DES implantation and surgery has been completed. Indirect information on the potential for DES to compete with CABG can be derived from the ARTS II study, a prospective multicenter registry of patients undergoing multivessel PCI with SES implantation, matched to the randomized patients included in the ARTS I trial of CABG versus stenting. In the subgroup of 367 diabetic patients, the 1-year MACE rate in ARTS II was 15.7%, similar to the rate in the CABG group of ARTS I (14.6%).86 There were no statistically significant differences in the rates of death (2.5% versus 2.1%), cerebrovascular accident (0% versus 5.2%), or MI (0.6% versus 2.1%), but a higher repeat revascularization rate was observed in ARTS II (12.6% versus 4.2%). Because occlusive restenosis occurs more frequently in diabetic patients than in nondiabetics and has been associated with increased long-term mortality among diabetics, DES treatment has theoretically the potential to improve survival in this patient population.87 It is encouraging that the mortality rate among diabetic patients at 1 year was 11.2% in the PTCA arm of BARI, 6.3% in the stent arm of ARTS I, and only 2.5% in ARTS II.86 A study sponsored by the NHLBI, the Future Revascularization Evaluation in patiEnts with Diabetes mellitus: Optimal Management of multivessel disease (FREEDOM) trial, will compare DES-based PCI and CABG in 2400 diabetic patients with multivessel disease. The primary end point will be all-cause mortality, nonfatal MI, and stroke. The study will have a parallel registry of approximately 2000 patients, and the overall study duration will be 5 years. In addition, the Coronary Artery Revascularization in Diabetes (CARDia) trial is currently randomizing 600 diabetic patients in the United Kingdom and Ireland to CABG or PCI (with either BMS or DES).88 The primary end point will be a composite of death, nonfatal MI, and cerebrovascular accident at 1 year. Additional data on costs, quality of life, and cognitive function are being collected, and follow-up will extend for 3 to 5 years.

Diabetes Early Invasive Versus Conservative Strategy

The high prevalence of abnormal glucose metabolism in patients with CAD, and in particular among those with acute manifestations of the disease, was recently confirmed in large-scale surveys in both the United States and Europe. Within the U.S. CRUSADE (Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the ACC/AHA guidelines) registry, among 46,410 patients with non-ST-elevation ACS, the prevalence of diabetes was 33%.89 Within the National Registry of Myocardial Infarction (NRMI), the prevalence of diabetes among patients presenting with ST-elevation MI (STEMI) and non-STelevation MI (NSTEMI) was 27% and 34%, respectively.90 In the Euro Heart Survey, glucose metabolism was addressed among 2854 patients with stable CAD and 2107 patients with unstable CAD.91 The overall prevalence of diabetes was approximately 30% in both groups. Among unstable CAD patients without known diabetes, an OGTT detected IGT in 36% and diabetes in 22% of cases. In the stable CAD group those proportions were 37% and 14%, respectively (Fig. 2-8).91 Diabetic patients, compared with nondiabetics, more frequently have characteristics and comorbidities that may negatively affect outcomes in the setting of ACS.92 However, several studies have shown that diabetes remains an independent predictor of shortterm morbidity and mortality after accounting for imbalances in baseline characteristics, a notion recently reinforced by an analysis of the CRUSADE registry (Table 2-8).92 Also, in the lung run, diabetic patients presenting with non-ST-elevation ACS have significantly higher rates of mortality and morbidity, recurrent MI, stroke, and heart failure compared with nondiabetic counterparts.28 Recent data from the Euro Heart Survey suggest that in this setting the mortality rate is particularly high among diabetic women.4 Importantly, in ACS, even prediabetes (i.e., fasting glucose levels between 100 and 126 mg/dL) is associated with increased CV risk.93

In diabetic patients with non-ST-segment elevation ACS, the positive impact of an early invasive strategy can be derived from subgroup analyses of large-scale randomized studies. The Fragmin and Fast Revascularisation during InStability in Coronary artery disease (FRISC II) study randomized 2457 patients to an invasive or conservative strategy and detected a significant survival benefit associated with the invasive strategy at 1 year.94 The reduction in 1-year death or MI associated with early coronary angiography followed by revascularization (if needed) was marked among diabetic patients (n = 299), in terms of relative and particularly of absolute risk reduction (39% and 100 90 80 Prevalence (%)

NON-ST-ELEVATION ACUTE CORONARY SYNDROMES

70 60 50 40 30 20 10 0

OGTT

FPG

Acute admission = = = =

OGTT

FPG

Elective consultation

Normal Impaired fasting glucose Impaired glucose tolerance Newly detected diabetes

Figure 2-8. Prevalence of abnormal glucose regulation in patients without known diabetes mellitus in the Euro Heart Survey assessed by oral glucose tolerance test (OGTT) or fasting plasma glucose (FPG). (From Bartnik M, Ryden L, Ferrari R, et al: The prevalence of abnormal glucose regulation in patients with coronary artery disease across Europe. Eur Heart J 2004;25:1880-1890.)

Table 2-8. In-Hospital Clinical Outcomes in Diabetic Patients with Non-ST-Elevation Acute Coronary Syndromes in the CRUSADE Registry Adjusted Odds Ratio (95% CI) Clinical Outcome Death (%) Reinfarction(%) Congestive heart failure (%) Shock (%) Red blood cell transfusion (%)

Nondiabetic (N = 31,049)

NIDDM (N = 9,773)

IDDM (N = 5,588)

NIDDM*

IDDM†

4.4 3.2 8.0

5.4 3.5 12.4

6.8 3.8 13.7

1.14 (1.02-1.29) 1.07 (0.96-1.19) 1.25 (1.16-1.34)

1.29 (1.12-1.49) 1.07 (0.93-1.24) 1.19 (1.09-1.31)

2.5 12.9

3.2 17.4

3.5 20.8

1.22 (1.05-1.41) 1.31 (1.23-1.40)

1.18 (0.97-1.44) 1.51 (1.40-1.63)

ACS, acute coronary syndromes; CI, confidence interval; IDDM, insulin-dependent diabetes mellitus; NIDDM, non-insulin-dependent diabetes mellitus. *Nondiabetic vs. type 2 diabetic patients. † Nondiabetic vs. IDDM patients. From Brogan GX, Peterson ED, Mulgund J, et al: Treatment disparities in the care of patients with and without diabetes presenting with non-ST-segment elevation acute coronary syndromes. Diabetes Care 2006;29:9-14.

37

38

Patient Selection FRISC II: 1-year death or MI Noninvasive, diabetes mellitus 30% Invasive, diabetes mellitus 20% Noninvasive, no diabetes 10% Invasive, no diabetes 0% 0

60

120

A

180

240

300

360

Follow up (days)

TACTICS: 6-month death, MI, rehospitalization for ACS 30

27.7%

Invasive Conservative

25 20.1%

20

16.4% 14.2%

15 10 5 0

B

Diabetes

No diabetes

Figure 2-9. Outcomes according to diabetic status in the FRISC II (A) and TACTICS (B) trials of invasive versus conservative strategy in acute coronary syndromes (ACS). MI, myocardial infarction. (A, From Norhammar A, Malmberg K, Diderholm E, et al: Diabetes mellitus: The major risk factor in unstable coronary artery disease even after consideration of the extent of coronary artery disease and benefits of revascularization. J Am Coll Cardiol 2004;43:585-591; B, Data from Cannon CP, Weintraub WS, Demopoulos LA, et al: Comparison of early invasive and conservative strategies in patients with unstable coronary syndromes treated with the glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J Med 2001;344:1879-1887.)

9.3%, respectively) (Fig. 2-9). Among nondiabetics, the effect was less pronounced (28% and 3.1%, respectively). Because of differences in sample size, the benefit observed barely missed statistical significance in diabetic patients but achieved it in nondiabetic individuals. In addition, diabetic patients undergoing early invasive therapy had a 38% reduction in the relative risk of 1-year death (7.7% versus 12.5%), again not reaching statistical significance owing to the small sample size.94 In the Treat Angina with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy (TACTICS)-TIMI 18 trial, an early invasive strategy was associated with a significant 22% reduc-

tion in the relative risk of death, MI, or rehospitalization for ACS at 6 months, compared with an early conservative strategy.95 All patients were treated with aspirin, clopidogrel, and tirofiban. Diabetic patients derived a greater benefit than nondiabetics from an early invasive strategy, in terms of both absolute (7.6% versus 1.8%) and relative (27% versus 13%) event reduction at 6 months (see Fig. 2-9). According to the 2002 Guidelines of the European Society of Cardiology (ESC), diabetes patients with ACS are to be classified automatically as high risk and therefore qualify for an early invasive strategy and for GP IIb/IIIa receptor inhibitors on top of standard treatment.96 Within the CRUSADE registry, however, diabetic patients had a statistically significant lesser chance to get early coronary angiography compared with nondiabetic individuals.89 Coronary Artery Bypass Surgery The only randomized trial that has compared CABG with PCI in patients with ACS was the AWESOME trial.97 This study compared the two revascularization strategies in patients who had medically refractory unstable angina and were at high risk for adverse outcomes with CABG. Among 2431 patients identified, 454 were considered acceptable for both PCI and CABG; 1650 patients were not deemed to be candidate for both therapies and entered a physiciandirected registry, and the 327 who were considered candidates for both treatment but refused randomization entered a patient-choice registry. Overall, diabetes prevalence was 31%. The respective CABG and PCI 3-year survival rates for diabetic patients were 72% and 81% for those randomized, 85% and 89% for those in the patient-choice registry, and 73% and 71% for those in the physician-directed registry.97 None of these differences was statistically significant. These results must be interpreted with caution because, from both a surgical perspective (left IMA used in 70% of cases) and an interventional perspective (stents and GP IIb/IIIa antagonists used in 54% and 11% of patients respectively), the way patients were revascularized may not comply with current standards. Nevertheless, CABG and PCI appear to be comparable options for high-risk diabetic patients with ACS, and the choice of revascularization should be made individually based on coronary anatomy, ventricular function, age, and comorbidities.

ST-ELEVATION MYOCARDIAL INFARCTION Paralleling the observations for non-ST-elevation ACS, diabetes is also an independent predictor of morbidity and mortality in STEMI. A retrospective study evaluating admission glucose of 141,680 patients presenting with acute MI demonstrated a linear correlation between glucose level and mortality (Fig. 2-10).98 Compared with individuals with admission glucose levels of 110 mg/dL or less, the hazard ratios for mortality for those with glucose

Diabetes 100%

1-Year mortality

90%

30-Day mortality

80% 70% 60%

Figure 2-10. Relationship between admission plasma glucose values and 30-day and 1-year mortality rates among patients presenting with acute myocardial infarction. (From Kosiborod M, Rathore SS, Inzucchi SE, et al: Admission glucose and mortality in elderly patients hospitalized with acute myocardial infarction: Implications for patients with and without recognized diabetes. Circulation 2005;111: 3078-3086.)

50% 40% 30% 20% 10% 0%

ⱕ100

⬎170 to 240

⬎140 to 170

⬎110 to 140

⬎240

Plasma glucose (mg/dL)

levels of more than 110 to 140, more than 140 to 170, more than 170 to 240, and more than 240 mg/ dL were 1.1, 1.3, 1.5, and 1.8 at 30 days and 1.1, 1.2, 1.3, and 1.5 at 1 year, respectively. The impact of diabetes on outcomes after the acute MI phase was addressed in a contemporary largescale study, the VALsartan In Acute myocardial iNfarcTion (VALIANT) trial.99 The study enrolled 3400 patients with known diabetes, 580 patients with newly diagnosed diabetes, and 10,719 patients with no diabetes. At 1 year, patients with previously known and newly diagnosed diabetes had similar increased risks of mortality (adjusted HR = 1.4 and 1.5, respectively) and of CV events (adjusted HR = 1.4 and 1.3, respectively), compared with nondiabetics. Similarly to what is observed in the setting of non-ST-elevation ACS, diabetic patients are exposed less frequently to evidence-based therapy in the management of acute MI. According to the Swedish Register of Information and Knowledge about Swedish Heart Intensive care Admissions (RIKS-HIA), after adjustments for differences in baseline characteristics, patients with diabetes were significantly less likely than nondiabetics to be treated with reperfusion therapy, heparins, statins, or revascularization but more likely to receive ACE inhibitors100 (Fig. 2-11). Importantly, the same analysis documented a mortality benefit associated with the administration of several of these therapies in the diabetic population (Fig. 2-12).

and 1.5 lives saved per 100 patients treated, respectively. Whereas CABG in the setting of STEMI is typically reserved for failed PCI and for MI-related mechanical complications, primary PCI may be preferred over thrombolytic therapy in diabetic patients. However, the data to support this notion are limited. In a pooled analysis on a total of 367 diabetic patients enrolled in 11 randomized trials, allocation to

Reperfusion Therapy

Figure 2-11. Likelihood of receiving various treatments in diabetic and nondiabetic patients with acute myocardial infarction in the RIKS-HIA registry, after adjustment for baseline characteristics. Horizontal lines indicate odds ratio (OR) ±95% confidence interval. ACE, angiotensin-converting enzyme; LMWH, low-molecular-weight heparin; revasc, revascularization. (From Norhammar A, Malmberg K, Ryden L, et al: Underutilisation of evidence-based treatment partially explains for the unfavourable prognosis in diabetic patients with acute myocardial infarction. Eur Heart J 2003;24:838-844.)

With respect to fibrinolytic therapy, the meta-analysis of the Fibrinolytic Therapy Trialists’ Collaborative Group involving all of the large randomized trials of fibrinolytic therapy versus placebo in STEMI demonstrated a greater than twofold survival benefit at 35 days among diabetic patients (n = 2236), compared with nondiabetics (n = 19,423), corresponding to 3.7

OR Reperfusion

0.83

Heparin/LWMH

0.88

Aspirin

0.97

β-blockade

0.97

Statin

0.88

ACE inhibition

1.45

Revasc ⬍14 days

0.86 0.5

1 No

1.5 Yes

2

Diabetes mellitus

39

40

Patient Selection OR Reperfusion

0.67

Heparin/LWMH

0.69

Aspirin

0.50

β-blockade

0.65

Statin

0.70

ACE inhibition

0.90

Revasc ⬍14 days

0.53 0

0.5

1

1.5

1-Year mortality diabetic patients

Figure 2-12. Effects of various treatments on 1-year mortality rate in patients with diabetes mellitus in the RIKS-HIA registry. Horizontal lines indicate odds ratio (OR) ±95% confidence interval. ACE, angiotensin-converting enzyme; LMWH, lowmolecular-weight heparin; revasc, revascularization. (From Norhammar A, Malmberg K, Ryden L, et al: Underutilisation of evidence-based treatment partially explains for the unfavourable prognosis in diabetic patients with acute myocardial infarction. Eur Heart J 2003;24:838-844.)

primary PCI led to a reduction in death or nonfatal MI at 30 days compared with fibrinolytic therapy (9.2% versus 19.3%; P < .05).101 Overall, the benefit of primary PCI over thrombolytic therapy was greater in diabetic compared with nondiabetic patients (number needed to treat to save one life [NNT] = 10 and 16, respectively). These data were generated before the availability of stents or GP IIb/IIIa inhibitors. Within the Comparison of Angioplasty and Prehospital Thrombolysis In acute Myocardial infarction (CAPTIM) trial, a small group of diabetic patients with acute MI (n = 103) were randomized to prehospital thrombolysis or a more contemporary primary PCI (stents were used in 83% of cases, and GP IIb/IIIa inhibitors in 27%).102 The 30-day incidence of death, recurrent MI, or stroke tended to be higher in diabetic individuals receiving fibrinolysis than in those undergoing mechanical reperfusion (21.0% versus 8.8%; P = .09). The difference was driven by the higher, although not statistically significant, mortality rate in the fibrinolysis group (13.0% versus 5.3%). A single-center retrospective analysis including a limited number of diabetic patients (n = 202) treated with reperfusion therapy for STEMI detected a significantly lower 1-year incidence of death or reinfarction in patients treated with primary PCI (n = 103), compared with those undergoing fibrinolysis (19.4% versus 36.4%, respectively).103 Stents (92%) and GP IIb/IIIa antagonists (63%) were broadly used in the setting of primary PCI.

The Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial aimed to determine the benefits of stent implantation over PTCA and abciximab over placebo in patients with STEMI, using a 2 × 2 factorial design. The study demonstrated that even in the primary PCI era diabetic patients (n = 346) had worse outcomes than nondiabetic individuals (n = 1736).104 Accordingly, the incidence of death, disabling stroke, reinfarction, or ischemic TVR at 1 year was 21.9% in diabetics and 16.8% in nondiabetics (P < .02). The difference was driven by increased rates of death (6.1% versus 3.9%; P = .04) and TVR (16.4% versus 12.7%; P = .07) among diabetics. The rates of restenosis and TVR at 1 year were significantly reduced in diabetic patients who underwent routine stenting compared with balloon angioplasty (21.1% versus 47.6% and 10.3% versus 22.4%, respectively). Aggressive Glucose-Lowering Therapy The Diabetes mellitus, Insulin Glucose infusion in Acute Myocardial Infarction (DIGAMI) study was designed to test the hypothesis that intensive glucoselowering therapy in patients with diabetes and acute MI would improve outcomes. A total of 620 patients were randomized to either standard treatment (controls) or standard treatment plus insulin-glucose infusion titrated according to glucose levels for at least 24 hours, followed by subcutaneous insulin treatment for 3 months after discharge. Active treatment was associated with a statistically significant mortality reduction at 3.5 years (33% versus 44%; RRR = 0.72) (Fig. 2-13). This translated into an impressive NNT of 9. Nevertheless, mortality remained elevated, underscoring the high risk of this patient population. In addition, insulin infusion was associated with a reduction in recurrent MI and heart failure rates at follow-up. In the DIGAMI 2 study, three glucose-lowering strategies were compared in 1253 diabetic patients with suspected acute MI: group 1 received an acute insulin-glucose infusion titrated to glucose levels for 24 hours, followed by insulinbased long-term glucose control; group 2 received insulin-glucose infusion for 24 hours followed by standard glucose control; and group 3 received routine metabolic management according to local practice.105 At 2 years, the mortality rates in the three groups were comparable (23.4%, 22.6%, and 19.3%, respectively) (Fig. 2-14), and no significant differences in nonfatal MI or stroke were detected. Against the expectations, the achieved blood glucose levels during the study period were identical in the three groups. The trial was stopped prematurely because of slow enrollment and lack of funding. Taken together, the results of the DIGAMI trials may be reconciled by stating that, in diabetic patients with acute MI, aggressive glucose lowering appears to be critical, irrespective of how this goal is achieved. In addition, it is not the metabolic effect on the myocardium of a glucose-insulin-potassium infusion per se that improves outcomes, but the associated

Diabetes 30 Control 25

Insulin

Figure 2-13. One-year mortality curves in diabetic patients with acute myocardial infarction randomized in the DIGAMI trial to either insulin infusion or control therapy. (From Malmberg K, Ryden L, Efendic S, et al: Randomized trial of insulin-glucose infusion followed by subcutaneous insulin treatment in diabetic patients with acute myocardial infarction [DIGAMI study]: Effects on mortality at 1 year. J Am Coll Cardiol 1995;26:57-65.)

Mortality (%)

20

15

10

5

0 0

Mortality (%)

25 20 Group 1 Group 2 Group 3

10 5 0 0

0.5

1.0

1.5

2.0

100

150

200

250

300

350

400

Days

30

15

50

2.5

3.0

Years

Figure 2-14. Three-year mortality curves in diabetic patients with acute myocardial infarction randomized in the DIGAMI 2 trial according to three different glucose-lowering strategies (for details see the text). (From Malmberg K, Ryden L, Wedel H, et al: Intense metabolic control by means of insulin in patients with diabetes mellitus and acute myocardial infarction [DIGAMI 2]: Effects on mortality and morbidity. Eur Heart J 2005;26:650-661.)

glucose-lowering effect. Accordingly, in a randomized study enrolling 20,201 patients with STEMI primarily treated with thrombolytic therapy, glucoseinsulin-potassium infusion for 24 hours did not influence 30-day CV mortality or morbidity among either nondiabetic or diabetic individuals.106

ANTITHROMBOTIC THERAPY IN DIABETES Aspirin and Clopidogrel Data on the efficacy of antiplatelet therapy for primary prevention in patients with diabetes are limited. The only prospective randomized study has been the Early Treatment Diabetic Retinopathy Study (ETDRS), which enrolled 3711 diabetic patients in the 1980s and randomized them to aspirin 650 mg/ day or placebo.107 The administration of aspirin over

5 years was associated with a nonsignificant reduction in all-cause mortality and in fatal or nonfatal MI (RR = 0.91 and 0.83, respectively). In the secondary prevention setting, the Antiplatelet Trialist Collaboration demonstrated that prolonged use of an antiplatelet agent (mainly aspirin) among 5126 diabetic patients was associated with only a modest, nonsignificant benefit over placebo (RRR = 7%).108 Information on which oral antiplatelet agent may be best suited for diabetic patients in the prevention setting can be derived from a subgroup analysis of the only large-scale head-to-head comparison, the Clopidogrel versus Aspirin in Patients at Risk of Ischemic Events (CAPRIE) trial. Among 3866 diabetic patients, the adenosine diphosphatase (ADP) (P2Y12) receptor antagonist clopidogrel (75 mg/day) was found to be superior to aspirin (325 mg/day) in the composite of ischemic and bleeding events over 2 years (RRR = 14.5%).109 Accordingly, the number of ischemic or bleeding events prevented with clopidogrel per year among 1000 treated patients was 9 in the nondiabetic group, 21 in the diabetic group overall, and 38 in the insulin-treated group. These results were not considered to be strong enough, and aspirin remains the first-line antiplatelet agent for CV prevention, even among diabetic patients. The ADA recommends aspirin (72 to 162 mg/day) indefinitely for all diabetic patients with evidence of CVD and in the primary prevention setting for individuals older than 40 years of age with one or more CV risk factors or albuminuria.110 The Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trial investigated the safety and efficacy of long-term administration of aspirin (75 to 162 mg/day) and clopidogrel (75 mg/day), compared with aspirin alone, in patients with established atherosclerotic disease or with multiple CV risk factors.111 In the large diabetic population enrolled (n = 6556), no benefit of the combination therapy was observed after a median follow-up of 28 months, whereas the

41

42

Patient Selection bleeding rate increased. With respect to patients undergoing PCI, the Clopidogrel for the Reduction of Events During Observation (CREDO) study randomized patients either to a 300-mg loading dose followed by 12 months of clopidogrel therapy or to no loading dose and clopidogrel treatment for 1 month on top of aspirin. Among 560 diabetic patients, the benefit of pretreatment and prolonged clopidogrel therapy was modest (RRR = 11.2%) compared with the benefit of this regimen observed among 1556 patients without diabetes (RRR = 32.8%).112 In the setting of non-ST-elevation ACS, aspirin remains a cornerstone of therapy, although specific data for diabetic patients are lacking. The Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial randomized patients with ACS primarily medically managed to aspirin or aspirin plus clopidogrel for 9 to 12 months. Diabetic patients (n = 2840) derived only a modest, nonsignificant benefit from the combined treatment (death, MI, or stroke rate = 14.2% versus 16.7%). Among patients undergoing PCI in the trial, the benefit of the combined antiplatelet therapy was less marked (RR = 0.77) among diabetic patients compared with nondiabetic ones (RR = 0.66).113 With respect to STEMI, the ClOpidogrel and Metoprolol in Myocardial Infarction Trial (COMMIT) in China randomized 45,852 patients with suspected acute MI to clopidogrel on top of aspirin or aspirin alone for a mean of 15 days.114 Allocation to clopidogrel led to a modest but statistically significant relative risk reduction (9%) in death, reinfarction, or stroke during the treatment period. Limitations of the study included the lack of reperfusion therapy (approximately 50% received fibrinolysis, but primary PCI was performed only in isolated cases). No information on diabetic patients has so far been reported. The CLopidogrel as Adjunctive Reperfusion TherapY (CLARITY)-TIMI 28 trial randomized patients receiving fibrinolytic therapy for acute MI to clopidogrel (300-mg loading dose followed by 75 mg/day) or placebo. At 30 days, the incidence of CV death, recurrent MI, or recurrent ischemia leading to urgent revascularization was reduced by 20% in those receiving clopidogrel therapy.115 No subgroup analysis addressing the diabetic patients enrolled in the main trial (n = 575) is currently available. Nevertheless, among the 282 diabetic patients who underwent PCI during their index hospitalization, pretreatment with clopidogrel resulted in a 39% reduction in 30-day events, although this difference was not statistically significant owing to the small sample size.116 Because of the results of the COMMIT and CLARITY-TIMI 28 trials, the U.S. Food and Drug Administration expanded the indications of clopidogrel for STEMI in August 2006. Overall, clopidogrel treatment for up to 1 year is indicated in diabetic patients presenting with ACS or undergoing PCI. Conversely, its role in the long-term prevention setting still needs to be defined. Although resistance to both aspirin and clopidogrel has been described in diabetic patients, the clinical relevance of these findings remains unde-

termined. Newer P2Y12 antagonists that have reversible action or are suitable for intravenous administration are currently being tested. Data on the diabetic population are lacking. Glycoprotein IIb/IIIa Receptor Antagonists The use of intravenous platelet GP IIb/IIIa receptor inhibitors and stents has markedly reduced the early hazard in diabetic patients undergoing PCI. In the Evaluation of Platelet IIb/IIIa Inhibitor for Stenting (EPISTENT) trial, abciximab halved the risk of death, MI, or urgent revascularization at 30 days among diabetic patients undergoing stenting compared with placebo (12.1% versus 5.6%, respectively). The observed event rate was comparable to that of abciximab-treated nondiabetic patients (5.2%). A pooled analysis of three early abciximab trials demonstrated a significant 1-year mortality rate reduction among diabetic patients randomized to the drug compared with placebo (2.5% versus 4.5%).117 The Intracoronary Stenting and Antithrombotic Regimen: is abciximab a Superior Way to Eliminate Elevated Thrombotic risk in diabetics (ISAR-SWEET) study demonstrated that, among 701 low-risk diabetic patients, abciximab did not confer additional benefit on top of aspirin and a high clopidogrel loading dose (i.e., 600 mg > 2 hours before PCI).118 However, the study excluded ACS and insulin-treated diabetic patients. The question whether one GP IIb/IIIa inhibitor rather than another may be preferable in diabetic patients was addressed in a subgroup analysis of the TARGET trial, the only head-to-head comparison thus far performed. Among the 1117 diabetic patients enrolled, randomization to abciximab or tirofiban at the time of PCI led to comparable outcomes for up to 1 year.46 In particular, no difference was observed in terms of TVR or late mortality, suggesting that the non-GP IIb/IIIa properties of abciximab (such as vitronectin and αMβ2 [Mac-1] receptor inhibition) do not translate into a long-term clinical benefit among diabetic patients. In the setting of non-ST-segment elevation ACS, although the overall impact of GP IIb/IIIa receptor inhibitors used in a conservative setting has been modest,119 a mortality benefit was detected among diabetic patients. Accordingly, the meta-analysis of the diabetic populations (n = 6458) enrolled in the six large-scale trials of GP IIb/IIIa inhibitors in ACS detected a highly significant 26% mortality reduction associated with the use of these agents at 30 days, compared with placebo (Fig. 2-15).120 These findings were reinforced by a statistically significant interaction between treatment and diabetic status. The use of these potent platelet inhibitors was associated with a similar proportionate reduction in mortality for patients treated with insulin and for those treated with diet or with oral hypoglycemic drugs. Even more striking was the mortality reduction (70%) associated with the use of GP IIb/IIIa inhibitors among the diabetic patients who underwent PCI (see Fig. 2-15).

Diabetes Diabetic patients with ACS: 30 day-mortality

Placebo

IIb/IIIa

P = .33

6.1%

5.1%

687

P = .07

4.2%

1.8%

362

P = .17

6.7%

3.6%

1677

P = .022 P = .51

7.8%

5.0%

6.2%

4.6%

4.8%

4.9%

6.2%

4.6%

Trial

N

PURSUIT

2163

PRISM PRISM-PLUS GUSTO IV PARAGON A PARAGON B Pooled

Odds ratio and 95% CI

412

P = .93

1157 6458

OR 0.74 0

Figure 2-15. Meta-analysis of six randomized, placebo-controlled trials demonstrating the effect of platelet glycoprotein IIb/IIIa inhibitors (IIb/IIIa) on 30-day mortality among diabetic patients with acute coronary syndromes (ACS): A, overall benefit; B, efficacy among patients who underwent inhospital percutaneous coronary intervention (PCI). The data are reported as odds ratios with 95% confidence intervals (CI) and corresponding probability (P) values. Values lower than 1.0 indicate a survival benefit of IIb/IIIa. (From Roffi M, Chew DP, Mukherjee D, et al: Platelet glycoprotein IIb/IIIa inhibitors reduce mortality in diabetic patients with non-ST-segmentelevation acute coronary syndromes. Circulation 2001;104:2767-2771.)

P = .007

A

0.5

1

IIb/IIIa better

1.5

2

Placebo better

Diabetic patients with ACS undergoing PCI: 30 day-mortality Trial

N

PURSUIT

457

Odds ratio and 95% CI

Placebo

IIb/IIIa

3.3%

2.4%

2.5%

0.0%

PRISM

147

P = .57 P = .50

PRISM-PLUS

107

P = 1.00

1.8%

0.0%

GUSTO IV

239

P = .037

6.5%

1.2%

PARAGON A

45

P = .31

7.1%

0.0%

PARAGON B

284

P = .06

4.3%

0.7%

1279

P = .002

4.0%

1.2%

Pooled

OR 0.30 0

B

With respect to putative mechanisms underlying the preferential benefit of GP IIb/IIIa inhibitors observed among diabetic patients, an hypothesis was generated by an in vitro study in which the blood of diabetic (n = 35) and nondiabetic (n = 38) individuals was exposed to pharmacologic concentrations of abciximab, tirofiban, and eptifibatide.121 The assessment of fibrinogen-binding capacity with flow cytometry after exposure to 1 µmol/L ADP showed that GP IIb-IIIa antagonists inhibited platelet activation to a greater extent in blood from the diabetic patients. The decreased rate of fibrinogen binding after platelet activation was thought to be a consequence of glycation of the GP IIb/IIIa receptor, which may subsequently enhance the inhibitory function of GP IIbIIIa antagonists.121 Despite the preferential benefit from GP IIb/IIIa antagonists in the setting of non-STelevation ACS, data from the US NRMI registry including more than 60,000 patients with NSTEMI showed that diabetic patients had a significantly lesser chance to be treated with these potent platelet inhibitors than did nondiabetic individuals.122 The value of GP IIb/IIIa inhibitors for diabetic patients at the time of mechanical revascularization for STEMI cannot be adequately assessed, because few data are available. In a small, placebo-controlled, randomized trial with abciximab for stent-based primary PCI, the use of the GP IIb/IIIa antagonists

0.5 IIb/IIIa better

1

1.5

2

Placebo better

among diabetic patients (n = 53) led to a significantly lower mortality rate at 6 months (0% versus 16.7%), as well as a reduced rate of reinfarction.123 Within the previously mentioned CADILLAC trial, no benefit of abciximab in terms of morbidity or mortality was observed among 346 lower-risk diabetic patients treated with either PTCA or stents for acute MI.104 Anticoagulants The Superior Yield of the New strategy of Enoxaparin, Revascularization, and Glycoprotein IIb/IIIa inhibitors (SYNERGY) trial compared the low-molecular-weight heparin (LMWH) enoxaparin with unfractionated heparin (UFH) in 9978 ACS patients undergoing early invasive strategy and found no difference in outcomes at 30 days and 6 months in the overall study population or in the diabetic cohort (n = 2926).124 The Aggrastat-to-Zocor (A-to-Z) trial randomized 3987 ACS patients to enoxaparin or UFH on top of aspirin and tirofiban and found no benefit of enoxaparin. Among diabetic patients (n = 751), the composite of death, MI, or refractory ischemia at 30 days was nonsignificantly lowered with enoxaparin (8.4% versus 10.7%).125 Heparin and LMWH should be seen as equivalent alternatives for diabetic patients in the setting of ACS and PCI.

43

44

Patient Selection The value of a bivalirudin-based antithrombotic strategy for PCI was studied in the Randomized Evaluation in PCI Linking Angiomax to reduced Clinical Events 2 (REPLACE-2) trial. The study showed the noninferiority of bivalirudin plus provisional GP IIb/ IIIa inhibition compared with routine GP IIb/IIIa inhibition on top of aspirin and clopidogrel in terms of 30-day death, MI, urgent revascularization, or inhospital major bleeding. The outcomes up to 1 year in the two groups were also comparable among the 1624 diabetic patients enrolled, suggesting that bivalirudin may be seen as an alternative antithrombotic agent in the PCI setting.126 The interest for this compound has also been reinforced by a recent ex vivo human study showing that both bivalirudin and the combination of the GP IIb/IIIa antagonist eptifibatide plus heparin achieved marked reductions in total thrombus formation and fibrin deposition in diabetic patients undergoing PCI.127

studies have documented that higher levels of habitual aerobic fitness and/or physical activity are associated with significantly lower cardiovascular and overall mortality rates among diabetic individuals. To achieve and maintain effective lifestyle modifications, diabetic subjects should receive multidisciplinary counseling by dietitians, diabetes educators, exercise trainers, and physicians. Smoking doubles the CV morbidity and mortality risk among diabetic individuals.7 In addition, smoking has been associated with premature diabetes-related microvascular complications. Nicotine replacement or bupropion therapy is an effective element to include for smoking cessation in combination with behavioral interventions. Persistent abstinence of tobacco use remains one of the major goals in prevention of CVD in nondiabetic as well in diabetic subjects.7 Hyperglycemia

PREVALENCE AND MANAGEMENT OF CARDIOVASCULAR RISK FACTORS AND TREATMENT GOALS Aggressive CV risk factor modification, including optimal glycemic control, cigarette smoking cessation, control of blood pressure and cholesterol levels, and weight reduction and exercise, is an essential part of diabetes care. In fact, CV morbidity and mortality rates increase more steeply in diabetic subjects than in nondiabetic ones in the presence of additional risk factors. Table 2-9 summarizes the recommended treatment goals according to the ADA.7 Based on the human and financial costs associated with diabetes-related complications, lifestyle and pharmacologic interventions to prevent diabetes are a logical and cost-effective strategy.128 Dietary intervention, increased physical activity, and moderate weight loss not only improve glycemic control but also lower blood pressure and favorably affect lipid metabolism. Regular physical activity may reduce HbA1c levels by 10% to 20%, both systolic and diastolic blood pressure by 5 to 12 mm Hg, and triglyceride levels by 20% and may increase high-density lipoprotein (HDL)-cholesterol levels. Large cohort

Table 2-9. Treatment Goals for Diabetic Patients According to the American Diabetes Association Glycemic control Hemoglobin A1c Preprandial capillary plasma glucose Peak postprandial capillary plasma glucose Blood pressure Lipids Low-density lipoprotein (LDL) Triglycerides High-density lipoprotein (HDL)

3 g/dL fall in hemoglobin, or a >4 g/dL fall in hemoglobin in the absence of overt bleeding, or any red cell transfusion of 2 or more units* Intracerebral hemorrhage or bleeding the causes hemodynamic compromise or requires intervention

Myocardial infarction (after CABG) Myocardial infarction (not periprocedural) TIMI major bleeding TIMI minor bleeding Major bleeding (REPLACE-2 definition) GUSTO severe or life-threatening bleeding GUSTO minor bleeding

Bleeding that requires transfusion but does not case hemodynamic compromise

*All calculations of falls in hemoglobin are adjusted for any transfusion by the Landefeld index. CABG, coronary artery bypass grafting; CK-MB, creatine kinase MB fraction; GUSTO, Global Use of Strategies to Open Occluded Arteries in Acute Coronary Syndromes trial; PCI, percutaneous coronary intervention; REPLACE-2, Randomized Evaluation in PCI Linking Angiomax to reduced Clinical Events 2; TIMI, Thrombolysis in Myocardial Infarction trial.

Exosite 1

Catalytic site Exosite 2 Fibrinogen binding

Fibrinogen

Figure 10-3. Thrombin binding sites.

substrate recognition sites are involved in the binding of heparin, fibrinogen, and thrombomodulin, and the catalytic site is responsible for the serine protease activity and is blocked by the direct thrombin inhibitors.3 Adverse Events after Percutaneous Coronary Intervention Improvements in interventional techniques and refinements in antithrombotic therapies have led to a decline in the incidence of ischemic complications after PCI. Further refinement of antithrombotic strategies can be considered a two-edged sword, with improved prevention of ischemic complications potentially leading to an increase in bleeding complications (Table 10-1). Although the relationship between periprocedural myocardial infarction (MI)

has been widely debated, several studies using data from large-scale clinical trials demonstrate an excess risk of mortality with creatine kinase MB fraction (CK-MB) elevations of greater than three times the upper limit of normal. However, other studies have observed an excess in mortality only at higher degrees of myonecrosis, such as more than five times the upper limit of normal or when there is the development of Q waves. In an analysis of patients enrolled in the Randomized Evaluation in PCI Linking Angiomax to reduced Clinical Events 2 (REPLACE-2) study, a CK-MB elevation greater than or equal to three times the upper limit of normal was associated with a 3.5-fold excess risk of mortality at 12 months and accounted for 13.2% of all mortality seen by 12 months (Fig. 10-4).4 This forms the threshold definition for periprocedural (within 48 hours) MI within many PCI trials of adjunctive pharmacotherapy. In contrast, weighing the clinical significance of bleeding events reductions in hemoglobin has been less robustly examined. Factors contributing to this uncertainty include a nonstandardized approach to the recognition and reporting of clinical events in clinical trials and the lack of routine assessment of blood loss after PCI. Nevertheless, several studies demonstrate a substantial increase in early and late mortality associated with Thrombolysis In Myocardial Infarction (TIMI) major and minor bleeding after PCI. In an analysis by Kinnaird and colleagues examining 10,974 patients over a 10-year period, major bleeding events were associated with an approximately 10-fold excess in mortality and an approximately threefold excess in non-Q-wave MI.5 Urgent revascularization and Q-wave MI were also increased. In the analysis of REPLACE-2, TIMI minor or greater bleeding accounted for 3.9% of all the mortality

167

168

Pharmacologic Intervention Monitoring of Anticoagulation

20

20

15

13.7%

13.4%

15 13.2% 11.6% 10

10 7.6 5.3 5

4.6%

3.5

Odds ratio

Percent attributable fraction

Percent attributable fraction Odds ratio

5

2.8 2.0 1 0
1xULN
2xULN
3xULN
5xULN
10xULN

20

20

15

15 12.0%

10

10

6.1 5

5 4.0%

3.9% 3.5%

1.6

2.2

2.3

1

0 Protocol major/ Protocol minor bleed major bleed

TIMI major/ minor bleed

TIMI major bleed

Figure 10-4. Relationships among ischemic events, bleeding events, and late mortality.

observed with a 2.3-fold relative risk, whereas a bleeding event meeting the TIMI major criteria was associated with a 6.1-fold excess mortality risk (see Fig. 10-4). These observations raise the question of how to weigh bleeding and ischemia relative to each other. From a statistical perspective, combining these end points may increase the likelihood of observing no difference between therapies, and therefore assessment of noninferiority may be better conducted on ischemic and bleeding end points separately. However, from a clinical perspective, bleeding and ischemic events are associated with substantial adverse outcomes and the combined consideration remains a clinical imperative.

Odds ratio

Percent attributable fraction

Percent attributable fraction Odds ratio

Various assays, including the activated clotting time (ACT), ecarin clotting time (ECT), and Factor Xa levels, have been used to monitor the therapeutic effect of anticoagulants during PCI. However, the correlation between the levels achieved with these assays with the various agents and clinical events have been studied only retrospectively. The relationship between the assay level achieved and clinical events is also influenced by concomitant antiplatelet therapy. In the context of unfractionated heparin therapy, increasing levels of ACT are associated with a modest reduction in periprocedural ischemic events, but a moderate excess in bleeding events.6 In contrast, when heparin is given with abciximab, ischemic events are lower, with little further reduction in events at higher ACT levels but a substantial increase in bleeding events (Fig. 10-5). The ACT assay is not as useful for monitoring the efficacy of enoxaparin and the other low-molecularweight heparins (LMWHs), with lesser degrees of prolongation in this assay observed.7 Traditional laboratory-based factor Xa assays also remain impractical for catheter laboratory use. The ENOX assay (Rapidpoint) is a whole-blood, point-of-care assay that correlates with laboratory enoxaparin-induced anti-Xa levels.8 The Evaluating Enoxaparin Clotting times (ELECT) study explored the relationship between the ENOX assay results and clinical outcomes among 445 patients receiving subcutaneous or intravenous enoxaparin before PCI. There was a nonsignificant and nonlinear association between the ENOX times and ischemic complications, whereas bleeding events increased with greater ENOX times. ENOX times of between 250 and 450 seconds (correlating with anti-FXa levels of between 0.8 and 1.8 IU/mL) for intraprocedural anticoagulation and levels of less than 200 to 250 seconds for sheath removal when enoxaparin is used have been recommended. In contradistinction to both heparin and LMWH, bivalirudin is generally associated with greater prolongation of ACT. This effect appears to occur in a dose-dependent manner, although no gradient of benefit with respect to ischemic or bleeding events has been observed across the range of ACT values recorded at the doses studied within clinical trials.9 Despite the higher ACT levels, lower rates of bleeding have been consistently observed, potentially highlighting the limited value of ACT in predicting clinical events with this agent. The ACT provides qualitative but not quantitative information about bivalirudin and is of value only in determining if this agent was effectively administered. As a possible clinical alternative, the monitoring of these agents with the ECT may be more appropriate. Measurements based on this test appear to better correlate with plasma bivalirudin and hirudin levels.10 Whether levels based on this assay evolve to recommended targets for therapy remains to be established.

Anticoagulation in Percutaneous Coronary Intervention 15%

7 day, death/MI/TVR

n  5216

10% 41% P  .026

5%

10.1%

11.1%

8.6%

8.9%

6.6%

7.7%

7.5%

9.8%

0% 250

300

400

350

450

Minimum ACT

TIMI major or minor bleeding

20%

15%

10%

5%

16.9%

12.4% 8.6%

250

300

9.9% 12.4% 13.7% 12.4% 16.9%

0%

Figure 10-5. Relationship between activated clotting time (ACT) and outcome with heparin.

Unfractionated Heparin Heparin Pharmacology Unfractionated heparin is a heterogeneous group of glycosaminoglycans of various lengths (5000 to 30,000 daltons; mean, 15,000 daltons) that exhibits a high-affinity for antithrombin. This binding augments antithrombin’s enzymatic inactivation of thrombin, factor Xa, and factor IXa, with its effects on thrombin being the most pronounced.11 Given heparin’s reliance on antithrombin for a therapeutic effect, it is considered as an indirect antithrombin. The antithrombin effect of heparin requires the simultaneous binding of heparin, antithrombin, and thrombin. Consequently, molecules smaller than 18 saccharides lack sufficient length to simultaneously span antithrombin and thrombin and do not exhibit antithrombin activity. These smaller molecules account for up to two thirds of unfractionated heparin preparations. Thrombin inactivation by heparin also oc-curs by heparin-cofactor II, an enzyme with specific activity for thrombin, but it requires much higher heparin levels than the heparin-antithrombin

350

400

450

Maximum ACT

pathway. However, the anti-factor Xa effects of heparin do not depend on simultaneous binding of antithrombin and factor Xa, and antithrombin effects therefore are observed across a wider range of saccharide chain lengths. Pharmacokinetic heterogeneity is also observed, because larger heparin molecules are cleared more rapidly, and attenuation of heparin’s antithrombin effect is faster relative to its anti-factor Xa effect. The activated partial thromboplastin time (aPTT) and in vivo anticoagulant effect have an imperfect correlation. The elimination of unfractionated heparin is initially through rapid but saturable metabolism within endothelial cells and macrophages (zero-order kinetics), followed by slower renal clearance (firstorder kinetics). The plasma half-life depends on the dose administered and is approximately 1 hour at doses of 100 IU/kg. In the context of excessive dosing or perforation/excessive bleeding, unfractionated heparin can be reversed by the administration of protamine. However, the clinical efficacy safety and efficacy of this strategy is not well established. Increasingly, the limitations of heparin have been appreciated. These limitations include the activation

169

170

Pharmacologic Intervention of platelets; a dependence on antithrombin levels; nonspecific binding to plasma protein; an inability to inhibit clot bound thrombin; and direct binding to platelet factor-4 contributing to heparin induced thrombocytopenia in 1% to 3% of treated patients. Platelet activation by heparin is evidenced by an increase in the expression of platelet surface adhesion molecules.12 Nonspecific binding to plasma proteins secreted by platelets and endothelial cells in the setting of inflammation, and thrombosis may contribute to reduced bioavailability.13 The heparinantithrombin complex results in a large molecular structure that has limited capacity to access thrombin and FXa bound within thrombus.14 Clinical Data for Unfractionated Heparin Worldwide, unfractionated heparin remains the mainstay anticoagulant for patients undergoing PCI. Despite this fact, there are no prospective, randomized data to demonstrate the relative efficacy of this agent over placebo, and current dosing recommendations are empirical. Nevertheless, clinical experience and anecdotal evidence demonstrate the need for some degree of anticoagulation in the setting of balloon- and stent-induced vascular injury. In the absence of prospective, randomized data, several studies point toward the benefits and risks associated with greater degrees of anticoagulation with heparin in PCI. Early case-control studies in the era of percutaneous transluminal coronary angioplasty (PTCA) suggest that patients experiencing acute closure and death or urgent revascularization had lower ACT levels than those not experiencing these complications.15,16 Similarly, among 403 patients randomized to heparin (5000 units IV or 20,000 units IV) before balloon angioplasty, those receiving the higher dose experienced a nonsignificant reduction in the rate of death, MI, acute vessel closure, and repeat intervention (8.0% versus 12.5%, P = NS) but an increased rate of bleeding complications (20% versus 6%, P < .001).17 Weight-adjusted dosing has been studied as a strategy to reduce the variability in dose response. In a 400-patient, randomized trial assessing weightadjusted dosing compared with higher fixed dosing, the weight-adjusted strategy was not associated with superior efficacy or safety, although earlier sheath removal was afforded. Nevertheless, pooled analysis of heparin-only– treated patients enrolled in several randomized clinical trials suggests that there is a gradient of benefit associated with increasing degrees of anticoagulation, with commensurate risk of bleeding events. This analysis suggests that ACT levels in excess of 350 seconds are associated with fewer ischemic events, although bleeding rates also increase at these levels.6 Such levels of anticoagulation are not required when concomitant glycoprotein (GP) IIb/IIIa inhibition is used, and the relevance of these data in the context of pretreatment with thienopyridines is not known.18 These observations have also been difficult to dem-

onstrate in smaller studies in which the initial heparin doses, and therefore the ACT levels, achieved were lower.19 In contrast, the available data do not support the use of prolonged heparin infusions after PCI for the prevention of subacute ischemic events, in which no significant reduction in ischemic events is observed, but there is a clear excess in bleeding events and increased length of stay.20 This is especially true for patients receiving GP IIb/IIIa inhibition. Low-Molecular-Weight Heparin Pharmacology The LMWHs are produced by chemical or enzymatic depolymerization of unfractionated heparin, resulting in heparin fragments with a mean molecular weight that is approximately 30% of most unfractionated heparin preparations. However, the molecular size of the heparin molecules still varies, and anticoagulant characteristics remain heterogeneous, although more predictable, compared with those of heparin.21 Although between 25% and 50% of the heparin molecules retain antithrombin activity, the principal effect of the LMWHs is the inhibition of anti-FXa by antithrombin. Compared with unfractionated heparin, the LMWHs demonstrate a more consistent dose response as a result of less nonspecific plasma protein binding, as well less platelet activation and platelet factor-4 interactions leading to less heparin-induced thrombocytopenia. The longer half-life of this agent provides a more convenient means of prolonged anticoagulation before PCI among patients with acute coronary syndrome (ACS). Clearance is achieved by renal excretion, and the biologic half-life is increased in those with renal failure (Table 10-2). Several small studies have explored the various dosing strategies for the use of enoxaparin in PCI. Adequate levels of anti-FXa were observed in patients 2 to 8 hours after subcutaneous dosing of enoxaparin (1 mg/kg bid), and in those receiving an additional 0.3-mg/kg intravenous dose 8 to 12 hours after subcutaneous dosing at 1.0 mg/kg.22 Other investigators have suggested that doses as low as 0.5 mg/kg of intravenous enoxaparin may be safe and efficacious, while enabling easier sheath management, although a fourth of the patients in this study also received a GP IIb/IIIa inhibitor.23 Some evidence suggests that enoxaparin may be reversed by the intravenous administration of protamine, but these data are limited. Clinical Data on Enoxaparin For the available LMWHs, most data support the use of enoxaparin in PCI. Interest in the use of enoxaparin in patients undergoing PCI has emerged from its use in patients with ACS. In general, these data suggest at least equal efficacy, if not modest superiority, with respect to ischemic complications compared

Anticoagulation in Percutaneous Coronary Intervention Table 10-2. Features of Low-Molecular Weight Heparins Property

Unfractionated Heparin

Mean molecular mass (daltons) Dependence on antithrombin Anti-Xa : anti-IIa activity Half-life (minutes) Bioavailability Subcutaneous absorption Binding to plasma proteins Binding to platelets or macrophages Antigenicity/HITS Clearance Protamine neutralization

Enoxaparin

15,000 Yes 1 ≈60 + to + + + ++ +++ ++ ++ Renal ++++

Bivalirudin

5,000 Yes 2-4 ≈240 ++++ ++++ + + + Renal ++

2,180 No No anti-Xa activity 25 ++++ — — — — Renal/proteolysis —

HITS, heparin-induced thrombocytopenia syndrome; plus signs represent relative strength.

with heparin, with a modest excess in bleeding events in the context of invasively and conservatively managed patients. The initial reported experience with enoxaparin specifically in PCI includes a series of studies performed by the National Investigators Collaborating on Enoxaparin (NICE) study group. These studies explored enoxaparin without abciximab (NICE-1) and with abciximab (NICE-4) in patients undergoing PCI and compared these historically with the arms of the Evaluation in Percutaneous Transluminal Coronary Angioplasty to Improve Long-Term Outcome with Abciximab GP IIb/IIIa Blockade (EPILOG) and Evaluation of Platelet IIb/IIIa Inhibitor for Stenting (EPISTENT) trials, respectively (Fig. 10-6).24 The NICE3 registry addressed outcomes among ACS patients receiving the various intravenous GP IIb/IIIa inhibitors, with PCI being left to the discretion of the investigator. The NICE-1 registry assessed enoxaparin (1.0 mg/ kg IV) without GP IIb/IIIa inhibition before coronary intervention in 828 patients undergoing elective and urgent PCI. The primary study end point was in-hospital and 30-day major hemorrhage. Minor bleeding,

the need for any transfusion, and the composite ischemic end point of death, MI, and urgent revascularization were also examined. Key exclusion criteria were acute MI within 24 hours, recent fibrinolysis (3 days), prior LMWH within 12 hours, thrombocytopenia less than 100,000/mL, and serum creatinine level higher than 2.5 mg/dL. In this group, at least one stent was placed in 85% of patients, aspirin was administered to all patients, and clopidogrel pretreatment was at the discretion of the treating interventionalist. Arteriotomy closure devices were not permitted and the protocol was prescriptive with respect to the time for sheath removal (4 to 6 hours). In the study without concomitant GP IIb/IIIa inhibition, major hemorrhage occurred in 1.1% of patients, with minor hemorrhage and transfusions occurring in 6.2% and 2.7% of patients, respectively. The composite ischemic end point of death, MI, and urgent revascularization at 30 days was observed in 7.7% of patients, with MI occurring in 5.4% of cases.24 In the very similar NICE-4 protocol, 818 patients received enoxaparin (0.75 mg/kg) and abciximab (0.25-mg/kg bolus and 0.125-µg/kg/min infusion).24 In this study, 88% of patients received a bare-metal 30 Day Outcomes

Death, MI, UVR

Minor hemorrhage

Major hemorrhage

15.5%

16%

13% 10.8%

10.3%

10% 7.7% 7%

7.0%

6.8%

6.2%

5.3%

Figure 10-6. Observational studies with low-molecular-weight heparins in percutaneous coronary intervention contrasted with events in the Evaluation of Platelet IIb/IIIa Inhibitor for Stenting (EPISTENT) trial.

4% 1.1%

2.8%

1.5% 0.4%

1% 0% NICE-1 (n  828)

2.9%

2.2% 1.7%

EPISTENT Hep (n  809)

NICE-4 (n  818)

EPISTENT Abx (n  794)

Dalteparin (n  103)

171

172

Pharmacologic Intervention stent. The use of closure devices was not permitted. Inclusion and exclusion criteria and clinical end point definitions were similar to those used in the NICE-1 study. In NICE-4, major hemorrhage and minor hemorrhage were reported in 0.4% and 7.0% of patients, respectively, with transfusions required in 1.8% of cases. The composite ischemic event of death, MI, or urgent revascularization by 30 days occurred in 6.8% of patients, suggesting that enoxaparin may confer a similar level of efficacy and safety as observed with unfractionated heparin in the context of abciximab therapy.24 Employing a noncontrolled observational design, the NICE-3 study reported bleeding and ischemic events among 671 patients presenting with ACS and treated with tirofiban, eptifibatide, or abciximab.25 Within this population, 43% underwent PCI. By 30 days, death, MI, and urgent revascularization were observed in 1.6%, 5.1%, and 6.8% of patients, respectively. The primary end point of non–CABG-related major bleeding was reported in 1.9% of patients by 30 days. Although numerically higher than the rates observed in other studies, interpretation of these results is hampered by the uncontrolled nature of the study design. Other observational data in the setting of ACS also suggest that enoxaparin is safe and efficacious among ACS patients undergoing PCI. Subgroup analysis of 4676 patients undergoing PCI in the ExTRACT-TIMI 25 study suggests that the incidence of death or MI may be reduced with enoxaparin compared with heparin among patients who have received fibrinolysis for ST-segment elevation myocardial infarction (STEMI) (10.7% for enoxaparin versus 13.8% for heparin, P = .001), with no significant excess in bleeding complications. Similarly, the larger subgroup analysis of 4687 unstable angina and non-ST-segment elevation ACS patients undergoing PCI in the Superior Yield of the New Strategy of Enoxaparin, Revascularization and Glycoprotein IIb/IIIa Inhibitors (SYNERGY) study observed comparable rate of 30-day death or MI rates, with a slight excess in bleeding events.26 Two randomized studies have been more optimally designed to examine the relative clinical risks and benefits of enoxaparin among patients undergoing PCI. The Coronary Revascularization Using Integrillin and Single bolus Enoxaparin (CRUISE) study randomized 261 elective or urgent PCI patients to enoxaparin (1 mg/kg IV) or heparin, with all patients receiving eptifibatide.27 This small study reported no difference in the rate of bleeding complications or angiographic complications (6.3% versus 6.2%, P = NS) during the procedure. Similarly, there were no differences in ischemic end points at 48 hours or 30 days. Several other randomized studies have been too small to demonstrate clear benefits with enoxaparin over heparin, with a meta-analysis of these studies demonstrating no difference in bleeding or ischemia.28 The largest study directly addressing enoxaparin use among patients undergoing PCI was the Safety and Efficacy of Intravenous Enoxaparin in Elective Percutaneous Coronary Intervention: an International

Randomized Evaluation (STEEPLE) trial.29 This study randomized 3528 patients to intravenous enoxaparin (0.5 mg/kg; n = 1070), intravenous enoxaparin (0.75 mg/kg; n = 1228), or ACT-adjusted unfractionated heparin (n = 1230). The primary end point was non–CABG-related, protocol-defined bleeding by 48 hours (but not using the TIMI or Global Use of Strategies to Open Occluded Arteries in Acute Coronary Syndromes [GUSTO] scales), with the ischemic end points at 30 days also reported as secondary end points. In this study, GP IIb/IIIa inhibition and thienopyridines were used in about 40% and 95% of patients, respectively, with drug-eluting stents deployed in 57% of patients; 16% of cases involved multivessel intervention. At 48 hours, enoxaparin was associated with a lower rate of protocol-defined major and minor bleeding (6.0% for enoxaparin at 0.5 mg/kg versus 6.6% for enoxaparin at 0.75 mg/kg versus 8.7% for heparin; P = .0014), with most of the benefit driven by reductions in major bleeding (1.2% for enoxaparin at 0.5 mg/kg versus 1.2% for enoxaparin at 0.75 mg/kg versus 2.8% for heparin; P = .004 and P = .007). However, no difference was seen when the TIMI or GUSTO definitions of bleeding were applied, and there was no difference in the rate of transfusion. The composite end point of death, MI, or urgent revascularization at 30 days favored the unfractionated heparin arm, although these differences did not reach statistical significance and met a broad noninferiority boundary (Fig. 10-7). These results suggest that enoxaparin is a viable alternative to heparin, producing modest reductions in bleeding risk. However, the nonblinded nature of this study is of concern and confirmation of this result with other trials will be required. Clinical Data for Dalteparin Data for dalteparin are limited, with disappointing results suggesting that further clinical development of this agent for catheter laboratory use is unlikely.30 In a dose-ranging study of 107 patients, 4 patients received dalteparin (120 U/kg) less than 8 hours before PCI and received an additional 40 U/kg (1 patient) or no further LMWH (3 patients). The remaining patients were randomized to 40 U/kg given intravenously (27 patients) or 60 U/kg given intravenously (76 patients) at the beginning of the procedure, with all patients receiving aspirin and abciximab. However, three early thrombotic events led to the decision to unblind the study and terminate the 40-U/kg arm. In this trial, death, MI (CK > three times the upper limit of normal), and urgent revascularization were observed in 15.5% of patients overall, whereas major hemorrhage and transfusion each occurred in 2.8% of patients. Although inadequately powered to fully evaluate the clinical utility of this agent among patients undergoing PCI, these event rates are higher than commonly seen in modern PCI trials, and further adequately controlled studies with this agent have not been performed.

Anticoagulation in Percutaneous Coronary Intervention

0.051

9.0%

8.5% NS

0.01

6.8%

6.5% 6.0%

6.2%

5.9%

5.8%

NS

Figure 10-7. Protocol-defined major bleeding and ischemic events in the Safety and Efficacy of Intravenous Enoxaparin in Elective Percutaneous Coronary Intervention: An International Randomized Evaluation (STEEPLE) trial. CABG, coronary artery bypass grafting; GUSTO, Global Use of Strategies To Open Occluded Arteries in Acute Coronary Syndromes trial; MI, myocardial infarction; TIMI, Thrombolysis In Myocardial Infarction trial; URV, urgent revascularization.

NS

3.0% 2.2%

2.0% 1.6%

1.5% 0.6% 0.8%

0.0% Non-CABG related bleeding

Enoxaparin 0.5 mg/kg (n  1070)

Pentasaccharide and Hexadecasaccharide Two novel indirect thrombin inhibitors have entered phase IIb and phase III clinical development. Fondaparinux, a pentasaccharide, and a hexadecasaccharide are both synthetic molecules that mimic the biologically active sequence of heparin in its interaction with antithrombin. Given that these molecules are short, their principal effect is the inactivation of FXa, as seen with the LMWHs in contradistinction to unfractionated heparin.31 Similarly, these agents have relatively long half-lives, offering once-daily dosing regimens. These agents are not reversed by protamine and require the administration of factor VII concentrates. Given the pharmacokinetic characteristics, the initial interest with these agents has been for the treatment of patients presenting with acute coronary syndromes. The only large-scale trial evaluating a substantial number of patients undergoing coronary intervention is the Organization to Assess Strategies in Acute Ischemic Syndromes (OASIS)-5 trial. In this trial of 20,078 ACS patients randomized to enoxaparin or fondaparinux, 6207 patients underwent coronary intervention. Among these patients, no differences in ischemic complications were observed, though a benefit with fondaparinux was evident with respect to bleeding events compared with enoxaparin (8.8% for enoxaparin versus 3.3% for fondaparinux; P < .001). However, a substantial number of patients undergoing PCI received unfractionated heparin in

TIMI major/minor bleeding

GUSTO moderate or severe bleeding

Enoxaparin 0.75 mg/kg (n  1230)

Death/MI/URV

Unfractionated heparin (n  1228)

both arms of the study. The protocol was modified during the study to ensure the use of heparin in the fondaparinux arm owing to a higher rate of catheterrelated thrombosis (0.5% for enoxaparin versus 1.3% for fondaparinux; P = .001). Dedicated randomized trials of these agents, as stand-alone antithrombotic strategies or in combination with other antithrombins and antiplatelet agents in PCI, are awaited. Direct Thrombin Inhibitors Pharmacology The direct thrombin inhibitor hirudin, found in the saliva of the medicinal leech Hirudo medicinalis, is the prototypical molecule of this class. Hirudin is a 65– amino acid protein that forms a stable noncovalent complex with thrombin.31 With two domains, the NH2 terminal core domain and the COOH terminal tail, the hirudin molecule inhibits the catalytic site and the anion binding exosite in a two-step process. An initial ionic interaction leads to a rearrangement of the thrombin-hirudin complex and the subsequent formation of a tighter, irreversible 1 : 1 bond.3 This complex and tight binding of hirudin to thrombin helps account for the highly specific effect of hirudin on thrombin. Generally, the direct thrombin inhibitor molecules are smaller than the indirect thrombin inhibitors and consequently demonstrate greater efficacy for the inhibition of clot-bound

173

174

Pharmacologic Intervention thrombin, in addition to their effects on fluid-phase thrombin.31,32 Two forms of recombinant hirudin (rhirudin) have been developed, one with a sulfated Tyr63 and the other without this change. The nonsulfated tyrosine molecule appears to have a 10-fold lower affinity for thrombin compared with the naturally occurring compound. The hirudin-thrombin interaction offers a method for categorizing the other direct thrombin inhibitors, which have been divided into univalent and bivalent molecules. The univalent molecules, dabigatran, argatroban, melagatran, and the oral prodrug, ximelagatran, inhibit only the catalytic site and inactivate only fibrin-bound thrombin. The thrombin inhibition provided by these agents is less robust than that observed with hirudin, because dissociation leads to some residual thrombin activity. Argatroban, the only one of these agents approved for use in PCI, binds to the apolar-binding site adjacent to the catalytic site and provides competitive inhibition. The bivalent molecules, recombinant hirudin and bivalirudin, bind to the catalytic site and at least one of the exosites. Although the interaction between hirudin and thrombin is irreversible, the inhibition provided by bivalirudin is more transient. Bivalirudin is a synthetic 20–amino acid molecule with two domains. These are targeted toward the anionbinding exosite and catalytic sites, which are linked by four glycine spacers. Given the shorter amino acid chain length compared with hirudin, bivalirudin exhibits less avid ionic binding. Cleavage of the bivalirudin molecule at the Arg-Pro bond of the amino-terminal extension by thrombin itself enables the release of the thrombin active site for further thrombotic activity. This in part accounts for the shorter half-life of bivalirudin compared with hirudin and may account for some of the reduced bleeding risk seen with this agent. Several other direct thrombin inhibitors have been developed in addition to those discussed, but they have not yet found a clinical role in the catheterization laboratory.33 All available agents approved for use in PCI require parenteral administration. With the exception of argatroban, these agents are cleared renally, and clearance is attenuated in the setting of reduced renal function. In the setting of excessive dosing or bleeding, these agents can be removed by hemofiltration. Argatroban is primarily eliminated through hepatic metabolism, and dose reduction in the setting of hepatic dysfunction is required. However, renal function also influences dosing.34 Bivalirudin also undergoes proteolysis within the plasma, contributing to its shorter half-life and relatively constant elimination characteristics even among patients with mild to moderate renal impairment (see Table 10-2). Nevertheless, dose attenuation is required for patients with creatinine clearance less than 30 mL/min. These agents are not reversed by protamine. Nonspecific measures such as transfusion of blood products, including fresh frozen plasma, and local measures are recommended in the context of active bleeding.

Clinical Evidence of Direct Thrombin Inhibition This class of agents, particularly bivalirudin, has emerged as a useful alternative to heparin as an anticoagulant among patients undergoing PCI as an adjunct and alternative to GP IIb/IIIa inhibition. Early trials with hirudin focused on the prevention of restenosis in the setting of balloon angioplasty. Although no anti-restenotic effect was evident, reductions in early ischemic events were observed. These agents have found a role in the management of patients with heparin-induced thrombocytopenia (i.e., argatroban and bivalirudin), whereas most recent data suggest that improved thrombin inhibition with bivalirudin enables sparing of GP IIb/IIIa inhibition in most patients undergoing PCI. Clinical Evidence for Hirudin The first large-scale, randomized trial of direct thrombin inhibition in PCI was the Hirudin in a European Trial Versus Heparin In the Prevention of Restenosis after PTCA (HELVETICA) trial.35 In this study, 1141 unstable angina patients undergoing balloon angioplasty received either of two dose regimens of hirudin or unfractionated heparin. Patients receiving intravenous hirudin experienced a reduction in early cardiac events within 96 hours (hirudin arms combined, RR = 0.61; 95% CI: 0.41 to 0.90; P = .023). However, in this study, the primary end point was event-free survival at 7 months, and for this end point, there were no differences between the three treatment arms, whereas similar rates of restenosis were observed. In the angioplasty substudy of STEMI patients of the Global Utilization of Strategies to Open Occluded Coronary Arteries IIb (GUSTO IIb) trial, 503 patients undergoing PTCA were randomized to hirudin or heparin.36 Hirudin resulted in a 23% (P = NS) reduction in death, MI, or stroke at 30 days. A benefit with hirudin in the setting of PCI is also evident from other observational studies. Among all patients undergoing PCI in the GUSTO IIb (ST elevation, [randomized] and non-ST elevation ACS [physician discretion]), a reduction in 30-day MI among the hirudin group (n = 672) compared with those receiving heparin (n = 738) was seen (4.9% versus 7.6%, P = .04), and a nonsignificant excess in bleeding was observed.37 Likewise, analysis of the OASIS-2 trial of unstable angina patients randomized to heparin or hirudin, assessing the outcomes in 172 patients undergoing PCI within 72 hours of randomization, provides similar conclusions.38 Although the study was observational and relatively small, the rate of death or MI at 96 hours was lower among hirudintreated patients compared with those receiving heparin (6.4% versus 21.4%; OR = 0.30; 95% CI: 0.10 to 0.88) and 35 days (6.4% versus 22.9%; OR = 0.25; 95% CI: 0.07 to 0.86). Caution should be exercised when interpreting this nonrandomized comparison. Nevertheless, these observations led to a metaanalysis of direct thrombin inhibition drawn from

Anticoagulation in Percutaneous Coronary Intervention 16

OR 0.66 (0.48–0.91)

UFH 14

DTI

Event rate (%)

12 10

OR 0.48 (0.29–0.73)

8 6 4 2

OR 0.94 (0.86–1.03)

OR 0.55 (0.57–1.33)

OR 0.55 (0.21–1.48)

0

Figure 10-8. The relative impact of direct thrombin inhibition in invasive and conservative management of acute coronary syndromes. DTI, direct thrombin inhibitor; MI, myocardial infarction; PCI, percutaneous coronary intervention; UFH, unfractionated heparin.

two PCI and nine ACS trials (N = 35,970), including data for bivalirudin and the univalent direct thrombin inhibitors, that reported a beneficial effect linked to the timing of PCI.39 Among patients undergoing PCI within 72 hours of randomization, the direct thrombin inhibitors were associated with lower rates of death or MI (OR = 0.66; 95% CI: 0.48 to 0.91) compared with heparin. In this analysis, a reduction in bleeding was driven by the benefits observed in the PCI trials (Fig. 10-8). In contrast, a more modest effect was documented when PCI was delayed after 72 hours. No benefit with these agents over heparin was observed in the context of conservative management. Clinical Evidence for Argatroban For the widespread application to patients undergoing PCI, argatroban has not been studied in largescale, randomized clinical trials. However, as an alternative to heparin among patients with heparininduced thrombocytopenia syndrome (HITS), results of a small case series totaling 151 patients suggest that this agent is safe.40,41 Similarly, a small, nonblinded, uncontrolled study of argatroban administered to patients treated with abciximab (n = 150) and eptifibatide (n = 2) suggests that the combinations of these agents is at least feasible, but evidence defining the absolute benefits and risk associated with argatroban in the context of modern intervention practice is still lacking.42 Clinical Evidence for Bivalirudin Clinical studies with bivalirudin constitute the bulk of evidence supporting the role of direct thrombin inhibition in the context of PCI. The first large-scale study with bivalirudin in the context of balloon angioplasty was with the Bivalirudin Angioplasty Study (BAT).43,44 Initially published in 1995, the trial was conducted in the era before coronary stenting,

MI Pre-PCI

MI Death or MI On day of PCI post-PCI Early PCI

Death or MI early PCI

Death or MI

Conservative treatment

thienopyridine use, and intravenous GP IIb/IIIa inhibition. In context of urgent or elective angioplasty, 4312 patients were randomized to bivalirudin (1 mg/ kg bolus and 2.5 mg/kg/hr infusion) or high-dose unfractionated heparin. A subgroup of 741 post-MI patients underwent stratified randomization to the same treatment arms. Randomization to bivalirudin provided a 22% reduction (6.2% versus 7.9%, P = .039) in death, myocardial infarction, or urgent revascularization and a 62% reduction (3.9% versus 9.7%, P < .001) in major bleeding events at 7 days. Among this stratified post-MI subgroup, the triple ischemic end point was reduced by 46% by 90 days (OR = 0.54; 95% CI: 36 to 0.81; P = .009). The advent of GP IIb/IIIa inhibition required two smaller pilot studies exploring the incremental benefits of bivalirudin among PCI patients receiving modern antiplatelet therapies. The Comparison of Abciximab Complications with Hirulog for Ischemic Events Trial (CACHET) A/B/C studies explored the role of bivalirudin with routine or provisional use of abciximab use in 208 patients undergoing coronary angioplasty and stenting. Within this small study, a promising reduction in bleeding events without excess ischemic events was observed.45 The Randomized Evaluation of PCI Linking Angiomax to reduced Clinical Events (REPLACE)-1 study employed a less prescriptive design, randomizing 1056 PCI patients to 0.75 mg/kg and 1.75 mg/kg/hr of bivalirudin or 60 to 70 U/kg of heparin with GP IIb/IIIa inhibition (i.e., abciximab, eptifibatide, or tirofiban) provisionally, routinely, or not at all at the discretion of the interventional cardiologist. Stents and GP IIb/IIIa inhibition were used in approximately 85% and 76% of patients, respectively. A nonsignificant benefit favoring the use of bivalirudin was observed at 48 hours in terms of ischemic and bleeding complications, despite the liberal use of GP IIb/IIIa inhibition.46 In the largest trial of antithrombotic therapy in PCI performed, the REPLACE-2 study enrolled 6010 patients undergoing elective or urgent coronary intervention. Randomization was to bivalirudin

175

176

Pharmacologic Intervention (0.75 mg/kg and 1.75 mg/kg/hr IV) and provisional abciximab or eptifibatide versus the planned use of these GP IIb/IIIa inhibitors and heparin (65 mg/kg IV), conducted in a double-blind, double-dummy manner.47 The commonly used “triple ischemic end point” of death, MI or urgent revascularization by 30 days was assessed with a noninferiority design. The major exclusions to this study were patients presenting with STEMI undergoing PCI for reperfusion, patients at significant risk for bleeding, or those requiring dialysis. As a result, approximately 50% of patients underwent PCI for an ACS, multivessel intervention was undertaken in about 15% of cases, and saphenous vein graft intervention occurred in 6% of patients. Provisional use of a GP IIb/IIIa inhibitor was permitted for coronary dissection, thrombus formation, unplanned stenting, slow flow, distal embolization, and ongoing clinical instability in the bivalirudin arm, whereas provisional placebo was used in the arm of patients already receiving GP IIb/IIIa inhibition. Among bivalirudin treated patients, GP IIb/IIIa inhibition was used in 7.5% of procedures. In contrast, 5.2% of heparin and GP IIb/IIIa inhibitiontreated patients received provisional placebo (P = .002). Pretreatment with a thienopyridine, mostly clopidogrel, was administered in 86% of patients. Bivalirudin (plus provisional GP IIb/IIIa inhibition) was associated with a nonsignificant excess in ischemic events (7.6% for heparin and GP IIb/IIIa inhibition versus 7.9% for bivalirudin; OR = 1.09; 95% CI: 0.90 to 1.32; P = .40) but met the boundary for noninferiority. In contrast, bleeding events were significantly reduced when evaluated by the TIMI criteria or the slightly broader protocol definition that included blood transfusion (4.1% for heparin and GP IIb/IIIa inhibition versus 2.4% for bivalirudin, P < .001). Reduced vascular access site events accounted for a large proportion of this bleeding benefit. Despite inadequate power to assess the benefit with respect to mortality, assessment of 12-month events demonstrated a lower point estimate for mortality with bivalirudin (1.6% versus 2.5%, P = .16).48 The nonsignificant excess in early MI was not associated with an excess in late mortality. These data are further supported by the results of the Acute Catheterization and Urgent Intervention Triage Strategy (ACUITY), a trial of antithrombotic therapy among ACS patients undergoing early invasive management. Among this high-risk patient population, the strategy of bivalirudin with bailout use of GP IIb/IIIa inhibition was associated with a slight but nonsignificant excess in ischemic events, with significant reduction of bleeding events (Fig. 10-9). Overall, when considered in a combined end point of ischemia and bleeding, the use of bivalirudin was “not inferior” to heparin or LMWH and a GP IIb/IIIa inhibitor. However, this was a complex trial, in which the timing of the GP IIb/IIIa inhibitor was also randomized to “upstream use” or “in cath-lab” initiation. As with the bivalirudin versus heparin or LMWH comparison, the in-catheter laboratory initiation of GP IIb/IIIa inhibition was associated with slightly

more ischemic events and fewer bleeding events. Although the noninferiority margin for these comparisons was met, results of longer follow-up of these patients will be needed to evaluate the relative efficacy and safety of these strategies. Pooled analysis of the randomized clinical trial experience with bivalirudin in PCI, including 11,638 patients (bivalirudin, 5861; heparin, 5777) demonstrates a reduction in the incidence of death, MI, revascularization, and major bleeding (7.8% versus 10.8%, P < .001) with this direct thrombin inhibitor at 48 hours. This large clinical trial experience also suggests a benefit with respect to mortality alone, despite the very low rate of events (0.01% versus 0.02%, P = .049). Consistent with the individual trial data, reductions for major bleeding were substantial (2.7% versus 5.8%, P < .001).49 The impact of bivalirudin-based strategies compared with eptifibatidebased strategies on measures of coronary flow after intervention have shown mixed results that are difficult to interpret in the context of large-scale clinical data that show no difference between these strategies.50 Limited, uncontrolled series also report experiences with bivalirudin with other interventional technologies, including drug-eluting stents, brachytherapy, and peripheral intervention. These results suggest an ischemic and bleeding profile that is consistent with the large-scale clinical trials despite their observational nature.51-53

TREATMENT OF SPECIAL GROUPS With the broad array of therapies available, weighing the limitations and benefits of each approach is often difficult. In many patients, the use of unfractionated heparin remains a safe and efficacious choice, especially in the context of pretreatment with a thienopyridine and the planned used of GP IIb/IIIa inhibition. However, in specific high-risk populations, the decision to use an alternative antithrombotic strategy may be considered. ST-Segment Elevation Myocardial Infarction Although the efficacy of hirudin has been explored in the context of primary PCI with balloon angioplasty in the GUSTO IIb study (discussed earlier), there are no randomized studies to optimally evaluate the risks and benefits of using enoxaparin or bivalirudin in the context of primary or rescue PCI. Nevertheless, a small, observational series suggests that bivalirudin is a feasible anticoagulant in primary PCI.36,54 Similar small series have been described for the use of enoxaparin in the context of liberal GP IIb/IIIa use. Transitioning from Upstream Management to the Catheter Laboratory Extrapolation of the clinical experience with unfractionated heparin suggests that the degree of anticoagulation required during PCI is greater than that

Anticoagulation in Percutaneous Coronary Intervention UFH/Enox  GPI (N  4603)

Bivalirudin alone (N  4612)

Bivalirudin  GPI (N  4604)

P  .32

30-day events (%)

8

7.3%

P  .35

7.7% 7.8%

6

P  .34

P  .78

5.0% 5.4% 4.9%

4 2.3% 2

2.7%

2.4%

1.3%1.5%1.6%

0 Ischemic composite

Death

Myocardial infarction

Unplanned revascularization for ischemia

A

Heparin  GPI (N  4603)

Bivalirudin  GPI (N  4604)

Bivalirudin alone (N  4612)

P  .31 12

11.8%

Figure 10-9. Ischemic and bleeding outcomes with bivalirudin versus bivalirudin/glycoprotein IIb/IIIa inhibition versus heparin (A) or lowmolecular-weight/glycoprotein IIb/IIIa inhibition (B) in the Acute Catheterization and Urgent Intervention Triage Strategy (ACUITY) trial. CABG, coronary artery bypass grafting; UFH, unfractionated heparin.

30-day events (%)

10

11.1%

P .001 9.1% P  .38

8 5.7%

6 4

P .0001 5.3% 3.0%

2 0 All major bleeding

B

required during the medical management of patients presenting with ACS. As a result, strategies have evolved to optimize the antithrombin therapies for ACS patients proceeding to PCI while already receiving one of these agents. Among patients being treated with heparin, an ACT-guided approach is recommended, with an additional 20 to 50 U/kg administered intravenously to achieve an ACT longer than 200 to 250 seconds when concomitant GP IIb/IIIa inhibition is planned and more than 300 to 350 seconds when heparin is the sole agent. Data on enoxaparin suggest that PCI can proceed without additional dosing when the procedure is occurring within 8 hours of the subcutaneous dose, and an additional intravenous bolus of 0.3 mg/kg is recommended when the delay is 8 to 12 hours. Outside this window, a dose of 0.75 mg/kg given intravenously should be administered regardless of GP IIb/IIIa inhibition use based on results of the SYNERGY study. Among patients receiving infusions of bivalirudin, an additional bolus of 0.5 mg/kg and an increase in the

Non-CABG major bleeding (primary end point)

infusion rate to 1.75 mg/kg were shown to be safe and efficacious in the ACUITY study, regardless of GP IIb/IIIa use (Table 10-3). Whether switching between agents when transitioning to invasive management is safe has not been prospectively evaluated. Decreased Renal Function Increased ischemic and bleeding events are observed among patients with renal dysfunction. Analyses of the randomized clinical trial experience with bivalirudin suggests that the relative benefits of this agent in terms of bleeding complications and ischemic complications is preserved.55,56 In absolute terms, among patients with at least moderate renal dysfunction (creatinine clearance 300-350 seconds without concomitant glycoprotein IIb/IIIa inhibition. ACS, acute coronary syndrome; ACT, activated clotting time; PCI, percutaneous coronary intervention.

Heparin/GPIIb/IIIa inhibition

Bivalirudin

15.0%

13.6% 10.6%

10.0% 7.2% 5.5% 5.0% 3.0%

1.9%

1.2%

0%

0% Normal eGFR
90 mL/min

Mild eGFR 60–90 mL/min

Moderate eGFR 30–59 mL/min

Severe eGFR  30 mL/min

Protocol defined bleeding Heparin/GPIIb/IIIa inhibition

Bivalirudin

15.0% 11.8% 9.1%

10.0% 7.7% 6.7%

6.1% 5.0%

4.7%

4.9%

5.7%

0% Normal eGFR
90 mL/min

Mild eGFR 60–90 mL/min

Moderate eGFR 30–59 mL/min

Severe eGFR 30 mL/min

30-day death/MI/urgent revascularisation

with renal impairment. However, given the reliance on renal elimination for these agents, their use is not expected to be an optimal choice. Diabetes Subgroup analysis of randomized clinical trials appears to indicate that abciximab provides substantial benefits in terms of reduced repeat revascularization and mortality among diabetic patients, with comparable effects observed with tirofiban. Clinical trial evidence with bivalirudin supports similar con-

Figure 10-10. Relationship between renal function and outcomes with bivalirudin. eGFR, estimated glomerular filtration rate.

clusions. In the REPLACE-2 study of bivalirudin and provisional GP IIb/IIIa inhibition compared with heparin and GP IIb/IIIa inhibition, bivalirudintreated diabetic patients experienced a lower, but nonsignificant rate of mortality at 12 months (2.3% versus 3.9%, P = NS). No difference in the rate of 30day bleeding and ischemic outcomes was observed.57 Although the long-term effects of enoxaparin-based strategies in diabetic patients have not been reported, a substantial rate of concomitant GP IIb/IIIa use in these studies will limit the interpretation of these data.

Anticoagulation in Percutaneous Coronary Intervention Heparin-Induced Thrombocytopenia HITS precludes the use of unfractionated heparin during PCI. Although the rate of HITS is less frequent with the LMWHs, cross-reactivity with these agents is observed and may be associated with increased rates of ischemic and bleeding complications. Whether pentasaccharides and hexadecasaccharides are safe and efficacious in this context has yet to be determined. The direct thrombin inhibitors are well suited to the management of HITS patients requiring PCI. Observational data for argatroban suggest that this agent can be safely used as an alternative to heparin in these patients. Case reports with recombinant hirudin (lepirudin) suggest that the use of this agent is also feasible.58 Similarly, a registry of 52 HITS patients receiving bivalirudin before PCI reported a 96% rate of freedom from death, Q-wave MI, and emergent CABG. Thrombocytopenia (platelet counts 40 to 60, ≤40 mg/dL) and correlating this with risk of subsequent adverse events.37 Among almost 2000 subjects with 4-month LDL-C data available, about 90% had LDL-C levels less than 100 mg/dL (2.59 mmol/L). Compared with the reference group (LDL-C level of 80 to 100 mg/dL), the hazard of death, MI, stroke, recurrent ischemia, and revascularization was lower among patients with LDL-C levels between more than 40 and 60 (HR = 0.76) and lowest among those with LDL-C levels less than or equal to 40 mg/dL (HR = 0.61). There was no excess risk of adverse events at these low levels of LDL-C. It is not necessary to reduce the dose of a statin if the resultant LDL-C levels fall well below guideline recommendations. These results suggest the possibility that further LDL-C lowering beyond the new guideline optimal goal of less than

2.0

Figure 11-10. Updated Cholesterol Treatment Trial (CTT) meta-analysis, including the Treating to New Targets (TNT), Individualized Dosing Efficacy versus Flat Dosing to Assess Optimal Pegylated Interferon Therapy (IDEAL), and Management of Elevated Cholesterol in the Primary Prevention Group of Adult Japanese (MEGA) trials, showing the linear relationship between low-density lipoprotein cholesterol (LDL-C) reduction and clinical risk reduction. (Modified from Baigent C, Keech A, Kearney PM, et al: Efficacy and safety of cholesterol-lowering treatment: Prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005;366:1267-1278.)

70 mg/dL (1.8 mmol/L) may translate into additional clinical benefit. Reduction in C-Reactive Protein Levels Statins possess pleiotropic effects that are mediated by HMG-CoA reductase but are not dependent on lowering of LDL-C levels (Fig. 11-11). All statins lower CRP levels, in part related to the statin dose. In PROVE IT, the median levels of CRP were similar in the 80-mg atorvastatin and 40-mg pravastatin groups (12.2 and 11.9 mg/L, respectively; P = .60) at study entry, but they were significantly lower in the atorvastatin group than in the pravastatin group at 30 days (1.6 versus 2.3 mg/L, P < .001), 4 months (1.3 versus 2.1 mg/L, P < .001), and the end of the study (1.3 versus 2.1 mg/L, P < .001). Although the levels of LDL-C and CRP were reduced by statin therapy at 30 days, the correlation between the achieved values was weak. (r = .16; P = .001) (i.e., less than 3% of the variance in achieved CRP levels was explained by the variance in achieved LDL). Because there are other known correlates of CRP in statin-naïve subjects, further assessment from PROVE IT performed a crosssectional analysis of the relationship between ontreatment uncontrolled risk factors and CRP levels.38 These factors were defined as body mass index (BMI) greater than 25 kg/m2 (i.e., World Health Organization’s cutoff for overweight), blood pressure higher than 130/85 mm Hg, glucose concentration higher than 110 mg/dL, triglyceride level greater than 150 mg/dL (i.e., Adult Treatment Panel III of the National Cholesterol Education Program’s cutoffs), HDL level less than 50 mg/dL, LDL-C level equal to or greater than 70 mg/dL, and smoking. An increase

Lipid Lowering in Coronary Artery Disease Within liver and vascular walls Acetyl CoA  Acetoacetal CoA Dose dependent on inhibition by statins

Dose dependent on inhibition by statins

HMG CoA

Mevalonate

Rapid vascular effect

Isopentanyl PP Hepatic inhibition

Geranyl PP

Endothelial, inflammatory and smooth muscle cells Translocates to the cell membrane Prenylation

Farnesyl PP

Liver Hepatocytes

Figure 11-11. Inhibition of HMG-CoA reductase leads to low-density lipoprotein cholesterol (LDL-C)– mediated effects through the liver and nonlipid-related effects in the vessel wall. (From Ray KK, Cannon CP: The potential relevance of the multiple lipid-independent (pleiotropic) effects of statins in the management of acute coronary syndromes. J Am Coll Cardiol 2005;46:1425-1433.)

Squalene

RhO Geranyl g PP

Cholesterol Activation of transcription factors

in incremental risk factor burden (i.e., number of uncontrolled risk factors present) was associated with an increase in CRP values (Fig. 11-12). Among patients allocated to standard therapy, the CRP level was 3.8 mg/L (interquartile range [IQR]: 1.9, 7.8) when seven uncontrolled risk factors were present and 1.0 mg/L (IQR: 0.7, 2.1) when none were present (P < .0001 for trend). However, among patients allocated intensive therapy, the corresponding CRP levels were lower and ranged from 2.4 mg/L (IQR: 1.7, 5.7) to 0.8 mg/L (IQR: 0.4, 1.2) (P < .0001 for trend). In this population, in which everyone received a statin, prior randomization to 80 mg of atorvastatin was associated with a 27% lower CRP compared with 40 mg of pravastatin (P < .0001) independently of LDL-C, triglyceride, and HDL levels and other correlates of CRP, such as age, gender, glycemia, blood pressure, smoking, and BMI.38

At 30 days, the median LDL concentration was approximately 70 mg/L, and the median CRP level was approximately 2 mg/L. At 30 days, patients in whom statin therapy resulted in LDL-C levels less than 70 mg/dL had lower age-adjusted rates of recurrent MI or CHD death compared with those who did not achieve this goal (2.7 versus. 4.0 events per 100 person-years, P = .008). Despite the minimal correlation between LDL-C and CRP levels, an identical difference in the age-adjusted rates of events was also observed among patients in whom statin therapy resulted in CRP levels of less than 2 mg/L compared with those in whom statin therapy resulted in higher CRP values (2.8 versus 3.9 events per 100 personyears, P = .006). Patients who had achieved LDLC levels less than 70 mg/dL and CRP levels less than 2 mg/L had the lowest risk of recurrent events, whereas those with LDL-C levels more than

Standard therapy (Prava 40) Intensive therapy (Atorva 80)

4

Risk factors 1) BMI  25 kg/m2 2) Current smoker 3) HDL  50 mg/dL 4) TG  150 mg/dL 5) Glucose  125 mg/dL 6) BP  130/85 mm Hg 7) LDL  70 mg/dL

3.5 CRP mg/L

3

Figure 11-12. Relationships among uncontrolled risk factors, statin therapy, and achieved C-reactive protein (CRP) levels. (From Ray KK, Cannon CP: The potential relevance of the multiple lipid-independent (pleiotropic) effects of statins in the management of acute coronary syndromes. J Am Coll Cardiol 2005;46:1425-1433.)

Cellular activation Adhesion molecules Thrombosis Cytokine production Vasodilation

2.5 2 1.5 1

Ptrend  .0001 for each

0.5 0 0

1

2

3

4

5

Number of uncontrolled risk factors

6–7

193

Pharmacologic Intervention 70 mg/dL and CRP levels more than 2 mg/L had the highest risk. Hazard ratios for recurrent events among patients whose values were more than 70 mg/dL for LDL-C and below 2 mg/L for CRP, those whose values were less than 70 mg/dL for LDL and more than 2 mg/L for CRP, and those whose values were more than 70 mg/dL for LDL-C and more than 2 mg/L for CRP, compared with those whose values of achieved LDL-C were less than 70 mg/dL and CRP less than 2 mg/L (i.e., reference group), were 1.3, 1.4, and 1.9, respectively (for trend across groups, P < .001) (Fig. 11-13A). Similar data have emerged from the A to Z Trial, showing that subjects who achieve a low CRP level with high-dose statin therapy and those who achieve the dual goals of LDL-C levels less than 70 mg/dL and CRP levels less than 2 mg/L are at lower risk for recurrent events (see Fig. 11-13B).39 Meta-analysis of achieved CRP levels in the PROVE IT and A to Z trials demonstrated that the adjusted risk of death or recurrent MI of a CRP value greater than 2 mg/L is 1.43 (95% CI: 1.2 to 1.7). These secondary prevention data demonstrate that using statin therapy to achieve target levels of both LDL-C and CRP decreases the risk of recurrent MI and CHD death among patients with ACS. Whether CRP is causally related to risk or is a marker remains unclear, but several lines of evidence suggest that it is an important player in mediating cardiovascular risk (Fig. 11-14). These data support the hypothesis that therapies designed to reduce inflammation after ACS may improve cardiovascular outcomes.

Cumulative rate of recurrent MI or death from coronary causes

194

0.10

LDL cholesterol  70 mg/dL CRP  2 mg/liter

0.08

LDL cholesterol  70 mg/dL CRP  2 mg/liter

0.06

LDL cholesterol  70 mg/dL CRP  2 mg/liter

0.04

LDL cholesterol  70 mg/dL CRP  2 mg/liter

0.02

0.00 0.0

A

0.5

1.0

1.5

2.0

2.5

Follow-up (years)

Cumulative probability of death or MI (%) CRP  2 and LDL  70 N 1244

8 7

CRP  2 and LDL  70 N  500

6

CRP  2 and LDL  70 N  1140

5

CRP  2 and LDL  70 N  659

4

Apolipoprotein B

3

Although the clinical benefits of statins appear to be predominantly related to LDL-C–mediated effects, the LDL-C level incompletely measures atherogenic lipoproteins, and measurement of the concentration of apolipoprotein B (apoB), which is a direct measurement of the concentration of proatherogenic particles (e.g., LDL-C, VLDL, IDL), and measurement of non-HDL-C, which reflects the cholesterol concentration of atherogenic lipoproteins or the total cholesterol to HDL ratio, provide alternative approaches. The debate regarding the choice of the best lipid parameter has further intensified with apparently conflicting evidence between prospective studies.40-42 In statin trials, conflicting data exists about whether on-treatment lipid values alone explain the totality of the benefits of statin therapy. In the Air Force/ Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) trial, on-treatment apoB values appeared to be a superior marker of on-treatment efficacy compared with LDL-C levels, and the ontreatment apoB/A-I ratio appeared to explain the entire benefit of statin therapy in this trial.43 In contrast, the much larger LIPID trial suggested that the proportion of the treatment effect explained by reductions in LDL-C was 52%, compared with 67% for apoB,44 suggesting that nonlipid-related effects may also contribute to the long-term benefit of

2 1 0 0

B

120

240

360

480

600

Follow-up after month 4 (days)

Figure 11-13. Clinical benefit of achieving the dual goals of lowdensity lipoprotein (LDL) less than 70 mg/dL and C-reactive protein (CRP) levels of less than 2 mg/L with statin therapy in the Pravastatin or Atorvastatin Evaluation and Infection Therapy– Thrombolysis In Myocardial Infarction 22 (PROVE IT-TIMI 22) (A) and Aggrastat to Zocor (A to Z) trials (B), showing the benefit of achieving dual goals. (A, from Ridker PM, Cannon CP, Morrow D, et al, for the Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 (PROVE IT-TIMI 22) Investigators: C-reactive protein levels and outcomes after statin therapy. N Engl J Med 2005;352:20-28; B, from Morrow DA, de Lemos JA, Sabatine MS, et al: Clinical relevance of C-reactive protein during follow-up of patients with acute coronary syndromes in the Aggrastat-to-Zocor Trial. Circulation 2006;114:281-288.)

Lipid Lowering in Coronary Artery Disease

statins. There is little conclusive evidence to suggest that we should measure apoB as an alternative to LDL-C in routine clinical practice. Oxidized Low-Density Lipoprotein Oxidized phospholipids (OxPLs) exist within atherosclerotic plaques and are bound by lipoprotein(a) in plasma. Circulating levels of oxidized LDL are strongly associated with angiographically documented CAD and therefore may contribute to the pathogenesis of atherosclerosis.45 In the MIRACL trial, high-dose atorvastatin reduced the total apoB-containing OxPLs by 29.7%, as well as reducing apoB levels by 30%. When normalized per apoB-100, compared with placebo, atorvastatin increased the OxPL/apoB level (9.5% versus −3.9%, P < .0001). These data suggest that atorvastatin treatment results in enrichment of OxPLs on a smaller pool of apoB particles, which may contribute to the reduction in ischemic events after ACS observed in MIRACL.46 HDL as a carrier of excess cellular cholesterol in the reverse cholesterol transport pathway is believed to provide protection against atherosclerosis. In reverse cholesterol transport, peripheral tissues (e.g., vessel-wall macrophages) remove their excess cholesterol through the ATP-binding cassette transporter A1 (ABCA1) to poorly lipidated apolipoprotein A-I, forming pre-HDL. HDL consists of a heterogeneous class of lipoproteins containing approximately equal amounts of lipid and protein (Fig. 11-15). The various HDL subclasses vary in quantitative and qualitative content of lipids, apolipoproteins, enzymes, and lipid transfer proteins, resulting in differences in shape, density, size, charge, and antigenicity. Assessment of HDL-C measures the cholesterol content of all these HDL subclasses and is therefore a crude marker of reverse cholesterol transport. A large number of prospective, observational studies have generally reported inverse associations between HDL-C concentrations and the risk of CHD,47-52 with the largest study reporting that a 1 mg/dL higher HDL-C concentration is associated

CRP

}

Figure 11-14. Schematic of putative mechanisms by which C-reactive protein (CRP) may mediate cardiovascular (CV) risk. BMI, body mass index; BP, blood pressure; HDL, high-density lipoprotein; IL-6, interleukin-6; LDL, low-density lipoprotein; NO, nitric oxide; TF, tissue factor; TG, triglycerides.

CV risk factors 1) BMI 2) Smoking 3) HDL 4) TG 5) Glucose 6) BP 7) LDL

Molecular mechanisms ↑ Rho kinase ↑ NFKB ↑ RAGE ↑ IL-6 ? Others

Intermediate biological mechanisms ↑ expression adhesion molecules on endothelium Reduces NO bioavailability Induces TF expression Activates complement

}

Clinical consequences

Plaque rupture

Thrombosis

with a 2% lower risk of CHD in men and 3% lower risk in women.47 The association for HDL-C is proportionally about 50% stronger in women than in men.47 The American National Cholesterol Education Program considers HDL-C to be an optional secondary target of lipid treatment,23 whereas the European Consensus Panel recommend a minimum target for HDL of 40 mg/dL (1.03 mmol/L) in certain patients, such as diabetics,53 but the relevance of the latter recommendation is unclear in light of limited data from clinical trials. Raising High-Density Lipoprotein Cholesterol Levels Fibrates, which are agonists of the PPAR-α receptor, increase the hepatic production of apolipoprotein A-I, which is the principal lipoprotein contained in the HDL particle and that raises HDL levels by approximately 10% to 15% (although others have shown almost no effect). Niacin reduces the hepatic clearance of the mature HDL particle, raising circulating HDL-C levels by 20% to 30%. Cholesterol ester– transfer protein (CETP) is a plasma glycoprotein that facilitates the transfer of cholesteryl esters from HDLC to apoB-containing lipoproteins and triglycerides from apoB-containing lipoproteins to HDL. Increasing the triglyceride content of HDL increases its clearance, thereby reducing HDL-C levels. Cholesterol Ester Transfer Protein Inhibition Humans with CETP deficiency due to molecular defects in the CETP gene have markedly elevated plasma levels of HDL-C and apolipoprotein A-I, suggesting that CETP inhibition may increase HDL-C levels. A phase II trial assessed the relative efficacy of torcetrapib in addition to different doses of atorvastatin at raising HDL.54 The addition of 60 mg of torcetrapib was associated with an increase in HDL of about 55% when added to 80 mg of atorvastatin, compared with about a 40% increase in HDL when used alone. The addition of 60 mg of torcetrapib to atorvastatin reduced the LDL/HDL ratio significantly,

195

196

Pharmacologic Intervention B

C-II

Reverse cholesterol transport

LPL

VLDL

LDL E

B

Oxidation

Liver CD36 SR-A

Hepatocyte LDL receptor

Bile adds FC and PL

Cholesterol pool

Cholesterol pool

Arterial-wall macrophage

SR-B1

ApoA-1 LRP

Cholesteryl ester transfer protein inhibitor

Gastrointestinal tract

LPL

A-1

-HDL

B

FC

C-II Chylomicrons and remnants

ABCA1

ABCA1

Peptide

CE

LCAT

E A-1

Nascent pre- -HDL

Infusion of ap-oA-1 Milano and phospholipids

A-1

Lipid-poor apoA-1

Figure 11-15. The reverse cholesterol transport pathway of high-density lipoprotein (HDL). Lipid-poor pre-β-HDL cholesterol, rich in apolipoprotein A-I (apoA-I), is synthesized by the liver or intestinal mucosa and released into the circulation, where it promotes the transfer of excess cellular-free cholesterol (FC) from macrophages to apoA-I by interacting with the ATP-binding cassette transporter A1 (ABCA1) in arterial wall macrophages. Plasma lecithin-cholesterol acyltransferase (LCAT) converts free cholesterol in pre-β-HDL cholesterol to cholesteryl ester (CE), resulting in the maturation of pre-β-HDL cholesterol to mature α-HDL cholesterol. The α-HDL cholesterol is transported to the liver by a direct or indirect pathway. In the direct pathway, selective uptake of cholesteryl ester by hepatocytes occurs with the scavenger receptor, class B, type 1 (SR-B1). In the indirect pathway, HDL cholesterol cholesteryl ester is exchanged for triglycerides in apolipoprotein B–rich particles (B), LDL cholesterol, and very-low-density lipoprotein (VLDL) cholesterol through cholesteryl ester-transfer protein (CETP), with the uptake of cholesteryl ester by the liver through the low-density lipoprotein (LDL) receptor (LDLR). Cholesterol that is returned to the liver is secreted as bile acids and cholesterol. Acquired triglycerides in the modified HDL cholesterol particle are subjected to hydrolysis by hepatic lipase (HL), thereby regenerating small HDL cholesterol particles and pre-β-HDL cholesterol for participation in reverse cholesterol transport. E, apolipoprotein-E–rich particles; PL, plasma lecithin. (Modified from Brewer HB Jr: Increasing HDL cholesterol levels. N Engl J Med 2004;350:1491-1494. Copyright 2004 Massachusetts Medical Society. All rights reserved.)

and when added to 80 mg of atorvastatin, it resulted in an LDL/HDL ratio of less than 1 (Fig. 11-16). However, in December 2006, Pfizer halted the development of torcetrapib due to a 60% observed increase in deaths in a phase III trial. The relevance of these biologic changes and the role of CETP inhibition remains unproved. Fibrates and Niacin A variety of studies have assessed the relative merit of raising HDL-C levels using fibrates or niacin (discussed previously). A meta-analysis of these data suggests that there is a 2.5% reduction in CHD events for every 1% rise in HDL-C levels with fibrates and a 1.7% reduction per 1% rise in HDL-C levels with

niacin.55 Niacin raises HDL-C by levels of approximately 28%, but its use has been limited by side effects, notably flushing mediated by prostaglandins. Using ultrasound and measurements of carotid intima–media thickness, niacin has been shown to attenuate and may reverse atherosclerosis.56 This concept is being tested in the large-scale 20,000 patient Heart Protection-2 study, combining niacin with an antiflushing agent in a Chinese population with CHD or high-risk equivalent. Overall, the effects of fibrates on HDL-C levels are modest. This was most recently demonstrated in the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) trial, in which long-term treatment with fenofibrate to raise HDL-C concentrations and lower triglyceride levels was assessed in subjects

Lipid Lowering in Coronary Artery Disease Baseline LDL/HDL

Treatment LDL/HDL

4.5

Figure 11-16. Effect of torcetrapib (60 mg) in addition to atorvastatin on the low-density lipoprotein/highdensity lipoprotein (LDL/HDL) ratio in subjects with hyperlipidemia compared with earlier statin-only trials. (Data from references 24, 43, 54, 63.)

Mean LDL-C to HDL-C ratio

4 3.5 3 2.5 2 1.5 1 0.5 0 AF/Tex

4S

TNT

T/A

T/A

T/A

T/A

20–40 mg lova

20–40 mg simva

80 mg atorva

60/10 mg tor/atorva

60/20 mg tor/atorva

60/40 mg tor/atorva

60/80 mg tor/atorva

with type 2 diabetes and total blood cholesterol concentrations of less than 251 mg/dL (6.5 mmol/ L).57 At the end of the trial, LDL-C levels fell from 120 mg/dL (3.07 mmol/L) to 95 mg/dL (2.43 mmol/ L) in subjects allocated fenofibrate and to 101 mg/dL (2.60 mmol/L) in the placebo group. Triglyceride levels were also lower in the fenofibrate group— 131 mg/dL (1.47 mmol/L) versus 166 mg/dL (1.87 mmol/L)—but HDL-C levels were not significantly different—44 mg/dL (1.13 mmol/L) versus 44 mg/dL (1.12 mmol/L)—compared with placebo. The differences in lipid levels between treatment groups decreased during the trial, particularly among patients receiving additional lipid-lowering therapy. Overall, 5.9% of patients on placebo and 5.2% of those on fenofibrate experienced the primary end point (P = .16). There was a significant 24% reduction in nonfatal MI but a nonsignificant increase in CAD mortality of 19%. Total cardiovascular events were significantly reduced by 11%, predominantly reflecting a 21% reduction in coronary revascularization. Overall, the FIELD trial failed to demonstrate a significant benefit of an agent that predominantly reduces triglyceride levels and raises HDL levels in a high-risk population. This finding might have resulted from an unequal, excess use of statin therapy in subjects allocated to placebo. Given the benefits of statin therapy among diabetics in the Collaborative Atorvastatin Diabetes Study (CARDS) trial58 and in subgroups of other trials,17 there is little evidence to suggest that fibrates are an alternative first-line treatment to statins for the management of dyslipidemia in diabetics. It remains to be seen whether adding a fibrate to statin therapy will reduce cardiovascular risk among diabetics, as is being assessed in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial.

Glitazones Glitazones are a novel class of agents that stimulate the PPAR-γ receptor, improve glycemic control, and have favorable effects on dyslipidemia, in particular reducing triglycerides and raising HDL-C levels. The Prospective Pioglitazone Clinical Trial in Macrovascular Events (PROACTIVE) trial was a placebocontrolled trial that assessed the benefit of a PPAR-γ agonist pioglitazone in stable type 2 diabetic subjects with evidence of macrovascular disease. Patients with evidence of coronary, cerebral, or peripheral arterial disease and an HgbA1c level greater than 6.5% despite treatment were considered eligible for the study (N = 5238). Compared with placebo, pioglitazone was associated with a 0.8% versus 0.3% reduction in HgbA1c from baseline (P < .0001) and an 11.4% versus 1.8% reduction in triglycerides (P < .0001). Significantly, pioglitazone raised HDL-C levels by 19%, compared with 10.1% in the placebo group (P < .0001), and reduced the LDL/HDL ratio by 9.5%, compared with 4.2% (P < .0001). The primary end point of mortality, nonfatal MI, stroke, ACS, coronary or peripheral arterial revascularization, and above-knee amputation tended to be lower in patients allocated pioglitazone (HR = 0.9; P = .095). The secondary end point of mortality, nonfatal MI, and stroke was reduced by 16% among patients allocated pioglitazone (P = .027). The overall findings of PROACTIVE, which included nonacute end points such as limb amputation, were neutral. However, a significant reduction in the secondary end point of death, nonfatal MI, and stroke was observed, raising the possibility that PPAR-γ agonists may be of value as adjunctive therapy to statins among patients with diabetes and macrovascular disease. However, these findings require validation in

197

198

Pharmacologic Intervention further clinical trials. The differences in glycemic control between treatments were modest, but large differences in HDL-C levels were observed with pioglitazone. It is possible that the beneficial effects of pioglitazone may not be due just to effects on glycemia but may also be related to beneficial effects on atherogenic dyslipidemia or systemic inflammation.55 Triglycerides Hypertriglyceridemia is a strong predictor of CHD. Prospective studies such as the Multiple Risk Factor Intervention Trial (MRFIT) show that the adjusted risk of a fatal or nonfatal CHD event is greater among subjects with triglyceride levels of 200 mg/dL or higher. This was the result regardless of whether the subjects were in a fasting or nonfasting state. The Whitehall II study showed the potential relevance of combining triglycerides and cholesterol in risk prediction. There is also an inverse relationship between serum levels of HDL-C and triglycerides, with low serum HDL-C levels representing an independent risk factor for cardiovascular disease and the so-called atherogenic lipid triad, consisting of high serum triglyceride levels, low serum HDL-C levels, and a preponderance of small, dense, LDL-C particles. This feature is particularly common in obesity and insulinresistant states. More than 2.5 million deaths each year worldwide are weight related, with cardiovascular disease the leading cause. Although modification of nutrition and physical activity is the cornerstone of therapy for obesity, pharmacotherapy focusing on improvement of the metabolic risk profile in patients who are at high risk for diabetes and cardiovascular disease may be required. Endocannabinoid Blockers The endocannabinoid system contributes to the physiologic regulation of energy balance, food intake, and lipid and glucose metabolism through central and peripheral effects. This system consists of endogenous ligands and two types of G protein–coupled cannabinoid receptors. The CB1 receptor is located in several brain areas and in a variety of peripheral tissues, including adipose tissue. Compared with

wild-type animals, CB1-knockout mice have leaner body composition, but this lean phenotype is not fully explained by changes in food intake. Stimulation of the CB1 receptors in fat cells promotes lipogenesis and inhibits the production of adiponectin, a cytokine derived from adipose tissue that has potentially important antidiabetic and antiatherosclerotic properties.59 The system may provide a possible treatment target for high-risk overweight or obese patients. The RIO Trials The RIO trials assessed the efficacy and safety of rimonabant (doses of 5 and 20 mg/day) in reducing body weight and improving cardiovascular risk factors in overweight patients (Table 11-2).60-62 The primary efficacy measure was weight loss, and secondary efficacy measures included changes in metabolic factors. Consistently across these trials, 5 mg of rimonabant had little effect, but 20 mg of rimonabant reduced body weight by about 6 kg (see Table 11-2), reduced triglycerides by approximately 5% to 14%, and increased HDL-C levels by approximately 10% to 21% compared with placebo (Fig. 11-17). Taken together, the trials of rimonabant suggest that CB1 receptor blockade in patients with adverse cardiovascular risk factors or obesity ameliorates metabolic abnormalities. In view of the residual cardiovascular risk observed among patients despite risk factor modification with statins and control of blood pressure, this novel strategy may be a useful adjunct to current therapeutic regimens. However, the potential cardioprotective effects of rimonabant and similar agents require assessment in clinical trials before definitive conclusions can be made. These potential cardioprotective effects may be offset by a significant excess of side effects, such as nausea and depression, which have led to a high discontinuation rate in the clinical outcome trials conducted. Because these trials are of relatively short duration (1 to 2 years) and it appears that weight gain and adverse cardiometabolic risk recurs after drug cessation, the long-term tolerability of these agents needs to be further assessed. These side effects may limit the widespread use of these agents.

Table 11-2. Trials of Rimonabant in Subjects with Obesity Showing Effects on Weight Loss and Lipids Trial

Population

RIO Europe

BMI >30 or BMI >27 with untreated hypertension or dyslipidemia BMI 27-40 with TG 150-703 mg/dL or TC/HDL ratio >4.5 BMI >30 or BMI >27 with untreated hypertension or dyslipidemia

RIO Lipids RIO North America

N

Study Duration

1507

1 yr

1036 3045

BMI, body mass index; HDL, high-density lipoprotein; TG, triglycerides.

Change in HDL

Change in Weight with Rimonabant (20 mg)

−14% (P < .001)

+21% (P < .001)

−6.6 kg (P < .001)

1 yr

−13% (P < .001)

+10% (P < .001)

−6.7 kg (P < .001)

2 yr

−5% (P < .001)

+11% (P < .001)

−6.3 kg (P < .001)

Change in TG

Lipid Lowering in Coronary Artery Disease HDL-C

Triglycerides 30

10 Rimonabant at 5 mg

0 Placebo –5

–10 P  .001

Rimonabant at 20 mg

–15

20

Rimonabant at 5 mg P  .017

15

P  .001

10 Placebo 5

0

–20 0

A

Rimonabant at 20 mg

25 Change in HDL cholesterol (%)

Change in triglyceride level (%)

5

12

24

36

52

Week

0

12

B

24

36

52

Week

Figure 11-17. Changes in triglyceride (A) and high-density lipoprotein cholesterol (HDL-C) (B) levels with rimonabant in the Rimonabant In Obesity Lipids (RIO-Lipids) trial. (Modified from Despres JP, Golay A, Sjostrom L, for the Rimonabant in Obesity-Lipids Study Group: Effects of rimonabant on metabolic risk factors in overweight patients with dyslipidemia. N Engl J Med 2005;353:2121-2134.)

Acyl-coenzyme A : Cholesterol Acyltransferase inhibitors

demonstrate any beneficial effect of pactimibe on adverse cardiovascular outcomes.

Acyl-coenzyme A : cholesterol acyltransferase (ACAT) is an enzyme that esterifies cholesterol in a variety of tissues. In some animal models, ACAT inhibitors have antiatherosclerotic effects. In an IVUS study of 408 patients with angiographically documented coronary disease, patients assigned the ACAT inhibitor pactimibe (100 mg/day) did not show any significant difference in the progression of atherosclerosis. The change in percent of atheroma volume was similar in the pactimibe and placebo groups (0.69% and 0.59%, respectively; P = .77). However, both secondary efficacy variables assessed by means of IVUS showed unfavorable effects of pactimibe treatment. Compared with baseline values, the normalized total atheroma volume showed significant regression in the placebo group (−5.6 mm3, P = .001) but not in the pactimibe group (−1.3 mm3, P = .39; P = .03 for the comparison between groups). The atheroma volume in the most diseased 10-mm subsegment regressed by 3.2 mm3 in the placebo group, compared with a decrease of 1.3 mm3 in the pactimibe group (P = .01). After the preliminary results of the ACAT IntraVascular Atherosclerosis Treatment Evaluation (ACTIVATE) were revealed in October 2005, clinical trials of pactimibe ceased worldwide. In patients with CAD, pactimibe failed to reduce atherosclerosis progression compared with usual care and had some proatherogenic effects. Although the study was not powered for clinical outcomes, ACTIVATE failed to

CONCLUSIONS All patients with CAD benefit from statin therapy, with no apparent threshold below which benefit is absent. Intensive statin therapy reduces cardiovascular events and atherosclerotic disease progression compared with standard therapy and therefore should be considered the standard of care for patients with CAD. In addition to important reduction of LDL-C levels, intensive statin therapy reduces inflammation, which appears to be particularly important in the early benefits observed in ACS patients, and has important contributions thereafter to long-term risk reduction. Beyond statin therapy, the data for other agents that favorably alter lipid profiles are unclear, but potential benefits of agents that raise HDL-C levels or significantly lower triglyceride levels are being investigated. REFERENCES 1. Kannel WB, Dawber TR, Friedman GD, et al: Risk factors in coronary heart disease. An evaluation of several serum lipids as predictors of coronary heart disease; the Framingham Study. Ann Intern Med 1964;61:888-899. 2. Kannel WB, Castelli WP, Gordon T, McNamara PM: Serum cholesterol, lipoproteins, and the risk of coronary heart disease. The Framingham Study. Ann Intern Med 1971;74: 1-12.

199

200

Pharmacologic Intervention 3. Neaton JD,Wentworth D: Serum cholesterol, blood pressure, cigarette smoking, and death from coronary heart disease. Overall findings and differences by age for 316,099 white men. Multiple Risk Factor Intervention Trial Research Group. Arch Intern Med 1992;152:56-64. 4. Neaton JD, Blackburn H, Jacobs D, et al: Serum cholesterol level and mortality findings for men screened in the Multiple Risk Factor Intervention Trial. Multiple Risk Factor Intervention Trial Research Group. Arch Intern Med 1992;152: 1490-500. 5. Law MR, Wald NJ: Risk factor thresholds: Their existence under scrutiny. BMJ 2002;324:1570-1576. 6. Lowering blood cholesterol to prevent heart disease. National Institutes of Health Consensus Development Conference Statement. Natl Inst Health Consens Dev Conf Consens Statement 1985;5:27. 7. The Lipid Research Clinics Coronary Primary Prevention Trial results. I. Reduction in incidence of coronary heart disease. JAMA 1984;251:351-364. 8. Stamler J: The coronary drug project—Findings with regard to estrogen, dextrothyroxine, clofibrate and niacin. Adv Exp Med Biol 1977;82:52-75. 9. Canner P, Berge K, Wenger N, et al: Fifteen year mortality in Coronary Drug Project patients: Long-term benefit with niacin. J Am Coll Cardiol 1986;8:1245-1255. 10. Carlson L, Danielson M, Ekberg I, et al: Reduction of myocardial reinfarction by the combined treatment with clofibrate and nicotinic acid. Atherosclerosis 1977; 28:81-86. 11. Frick M, Heinonen O, Huttunen J, et al: Efficacy of gemfibrozil in dyslipidaemic subjects with suspected heart disease. An ancillary study in the Helsinki Heart Study frame population. Ann Med 1993;25:41-45. 12. Rubins HB, Robins SJ, Collins D, et al: Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high-density lipoprotein cholesterol. Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial Study Group. N Engl J Med 1999;341:410-418. 13. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: The Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-1389. 14. Sacks FM, Pfeffer MA, Moye LA, et al: The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events Trial Investigators. N Engl J Med 1996;335:10011009. 15. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. N Engl J Med 1998;339:1349-1357. 16. 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. 17. Collins R, Armitage J, Parish S, et al: MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5963 people with diabetes: A randomised placebo-controlled trial. Lancet 2003;361:2005-2016. 18. Baigent CA. Keech PM, Kearney L, et al: Efficacy and safety of cholesterol-lowering treatment: Prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet 2005;366:1267-1278. 19. Schwartz GG, Olsson AG, Ezekowitz MD, et al: Effects of atorvastatin on early recurrent ischemic events in acute coronary syndromes: The MIRACL study: A randomized controlled trial. JAMA 2001;285:1711-1718. 20. Cannon CP, Braunwald E, McCabe CH, et al: Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004;350:1495-1504. 21. De Lemos JA, Blazing MA, Wiviott SD, et al: Early Intensive vs a delayed conservative simvastatin strategy in patients with acute coronary syndromes: Phase Z of the A to Z Trial. JAMA 2004;292:1307-1316. 22. Wiviott SD, de Lemos JA, Cannon CP, et al: A tale of two trials: A comparison of the post-acute coronary syndrome lipidlowering trials A to Z and PROVE IT-TIMI 22. Circulation 2006;113:1406-1414.

23. Grundy SM, Cleeman JI, Merz CN, et al: Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110: 227-239. 24. LaRosa JC, Grundy SM, Waters DD, et al: Intensive lipid lowering with atorvastatin in patients with stable coronary disease. N Engl J Med 2005;352:1425-1435. 25. Pedersen TR, Faergeman O, Kastelein JJ, et al: High-dose atorvastatin vs usual-dose simvastatin for secondary prevention after myocardial infarction: The IDEAL study: A randomized controlled trial. JAMA 2005;294:2437-2445. 26. Cannon CP, Steinberg BA, Murphy SA, et al: Meta-analysis of cardiovascular outcome trials comparing intensive versus moderate statin therapy. J Am Coll Cardiol 2006;82:438-445. 27. Scirica BM, Morrow DA, Cannon CP, et al: Intensive statin therapy and the risk of hospitalization for heart failure after an acute coronary syndrome in the PROVE IT-TIMI 22 study. J Am Coll Cardiol 2006;47:2326-2331. 28. Horwich TB, Hamilton MA, Maclellan WR, Fonarow GC: Low serum total cholesterol is associated with marked increase in mortality in advanced heart failure. J Card Fail 2002;8: 216-224. 29. Rauchhaus M, Clark AL, Doehner W, et al: The relationship between cholesterol and survival in patients with chronic heart failure. J Am Coll Cardiol 2003;42:1933-1940. 30. Ray KK, Cannon CP: The potential relevance of the multiple lipid-independent (pleiotropic) effects of statins in the management of acute coronary syndromes. J Am Coll Cardiol 2005;46:1425-1433. 31. Ray KK, Cannon CP, McCabe CH, et al: Early and late benefits of high-dose atorvastatin in patients with acute coronary syndromes: Results from the PROVE IT-TIMI 22 trial. J Am Coll Cardiol, 2005;46:1405-1410. 32. Cannon CP, Ray KK, Braunwald E: Reply. J Am Coll Cardiol 2006;48:852-853. 33. Hulten E, Jackson JL, Douglas K, et al: The effect of early, intensive statin therapy on acute coronary syndrome: A metaanalysis of randomized controlled trials. Arch Intern Med 2006;166:1814-1821. 34. Ballantyne CM: Clinical trial endpoints: Angiograms, events, and plaque instability. Am J Cardiol 1998;82(6A): 5M-11M. 35. Okazaki S, Yokoyama T, Miyauchi K, et al: Early statin treatment in patients with acute coronary syndrome: Demonstration of the beneficial effect on atherosclerotic lesions by serial volumetric intravascular ultrasound analysis during half a year after coronary event. The ESTABLISH Study. Circulation 2004; 110:1061-1068. 36. Hong MK, Lee CW, Kim YK, et al: Usefulness of follow-up low-density lipoprotein cholesterol level as an independent predictor of changes of coronary atherosclerotic plaque size as determined by intravascular ultrasound analysis after statin (atorvastatin or simvastatin) therapy. Am J Cardiol 2006;98:866-870. 37. Wiviott SD, Cannon CP, Morrow DA, et al: Can low-density lipoprotein be too low? The safety and efficacy of achieving very low low-density lipoprotein with intensive statin therapy: A PROVE IT-TIMI 22 substudy. J Am Coll Cardiol 2005; 46:1411-1416. 38. Ray KK, Cannon CP, Cairns R, et al: Relationship between uncontrolled risk factors and C-reactive protein levels in patients receiving standard or intensive statin therapy for acute coronary syndromes in the PROVE IT-TIMI 22 trial. J Am Coll Cardiol, 2005;46:1417-1424. 39. Morrow DA, de Lemos JA, Sabatine MS, et al: Clinical relevance of C-reactive protein during follow-up of patients with acute coronary syndromes in the Aggrastat-to-Zocor Trial. Circulation 2006;114:281-288. 40. Ridker PM, Rifai N, Cook NR, et al: Non-HDL cholesterol, apolipoproteins A-I and B100, standard lipid measures, lipid ratios, and CRP as risk factors for cardiovascular disease in women. JAMA 2005;294:326-333. 41. Pischon T, Girman CJ, Sacks FM, et al: Non-high-density lipoprotein cholesterol and apolipoprotein B in the prediction of coronary heart disease in men. Circulation 2005;112: 3375-3383.

Lipid Lowering in Coronary Artery Disease 42. Denke MA: Weighing in before the fight: Low-density lipoprotein cholesterol and non-high-density lipoprotein cholesterol versus apolipoprotein B as the best predictor for coronary heart disease and the best measure of therapy. Circulation 2005;112:3368-3370. 43. Gotto AM Jr, Whitney E, Stein EA, et al: Relation between baseline and on-treatment lipid parameters and first acute major coronary events in the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS). Circulation 2000;101:477-484. 44. Simes RJ, Marschner IC, Hunt D, et al: Relationship between lipid levels and clinical outcomes in the Long-term Intervention with Pravastatin in Ischemic Disease (LIPID) trial: To what extent is the reduction in coronary events with pravastatin explained by on-study lipid levels? Circulation 2002; 105:1162-1169. 45. Tsimikas S, Brilakis ES, Miller ER, et al: Oxidized phospholipids, Lp(a) lipoprotein, and coronary artery disease. N Engl J Med 2005;353:46-57. 46. Tsimikas S, Witztum JL, Miller ER, et al: High-dose atorvastatin reduces total plasma levels of oxidized phospholipids and immune complexes present on apolipoprotein B-100 in patients with acute coronary syndromes in the MIRACL trial. Circulation 2004;110:1406-1412. 47. Gordon DJ, Probstfield JL, Garrison RJ, et al: High-density lipoprotein cholesterol and cardiovascular disease. Four prospective American studies. Circulation 1989;79:8-15. 48. Gordon T, Castelli WP, Hjortland MC, et al: High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Am J Med 1977;62:707-714. 49. Multiple risk factor intervention trial. Risk factor changes and mortality results. Multiple Risk Factor Intervention Trial Research Group. JAMA 1982;248:1465-1477. 50. Assmann G, Schulte H, von Eckardstein A, Huang Y: Highdensity lipoprotein cholesterol as a predictor of coronary heart disease risk. The PROCAM experience and pathophysiological implications for reverse cholesterol transport. Atherosclerosis 1996;124(Suppl):S11-S20. 51. Jacobs DR Jr, Mebane LL, Bangdiwala SI, et al: High density lipoprotein cholesterol as a predictor of cardiovascular disease mortality in men and women: The follow-up study of the Lipid Research Clinics Prevalence Study. Am J Epidemiol 1990;131:32-47. 52. Gordon DJ, Knoke J, Probstfield JL, et al: High-density lipoprotein cholesterol and coronary heart disease in hypercholesterolemic men: The Lipid Research Clinics Coronary Primary Prevention Trial. Circulation 1986;74:1217-1225.

53. Chapman MJ, Assmann G, Fruchart JC, et al: Raising highdensity lipoprotein cholesterol with reduction of cardiovascular risk: The role of nicotinic acid—A position paper developed by the European Consensus Panel on HDL-C. Curr Med Res Opin 2004;20:1253-1268. 54. Thuren T: Torcetrapib Phase 2 Factorial Study A3071026. American Heart Association Scientic Sessions, 2005. 55. Birjmohun RS, Hutten BA, Kastelein JJ, Stroes ES: Efficacy and safety of high-density lipoprotein cholesterol-increasing compounds: A meta-analysis of randomized controlled trials. J Am Coll Cardiol 2005;45:185-197. 56. Taylor AJ, Sullenberger LW, Lee HJ, et al: Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol (ARBITER) 2: A double-blind, placebo-controlled study of extended-release niacin on atherosclerosis progression in secondary prevention patients treated with statins. Circulation 2004;110:3512-3517. 57. Keech A, Simes RJ, Barter P, et al: Effects of long-term fenofibrate therapy on cardiovascular events in 9795 people with type 2 diabetes mellitus (the FIELD study): Randomised controlled trial. Lancet 2005;366:1849-1861. 58. Colhoun HM, Betteridge DJ, Durrington PN, et al: Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes Study (CARDS): Multicentre randomised placebo-controlled trial. Lancet 2004;364:685-696. 59. Okamoto Y, Kihara S, Ouchi N, et al: Adiponectin reduces atherosclerosis in apolipoprotein E-deficient mice. Circulation 2002;106:2767-2770. 60. Van Gaal LF, Rissanen AM, Scheen AJ, et al: Effects of the cannabinoid-1 receptor blocker rimonabant on weight reduction and cardiovascular risk factors in overweight patients: 1-year experience from the RIO-Europe study. Lancet 2005;365:1389-1397. 61. Despres JP, Golay A, Sjostrom L, for the Rimonabant in Obesity-Lipids Study Group: Effects of rimonabant on metabolic risk factors in overweight patients with dyslipidemia. N Engl J Med 2005;353:2121-2134. 62. Pi-Sunyer FX, Aronne LJ, Heshmati HM, et al: Effect of rimonabant, a cannabinoid-1 receptor blocker, on weight and cardiometabolic risk factors in overweight or obese patients: RIO-North America: A randomized controlled trial. JAMA 2006;295:761-775. 63. Pedersen TR, Olasson AG, Faergerman O, et al: Lipoprotein changes and reduction in the incidence of major coronary heart disease events in the Scandinavian Simvastatin Survival Study (4S). 1998;97:1453-1460.

201

CHAPTER

12 Angiotensin-Axis Inhibition Marc A. Pfeffer KEY POINTS 䊏 Pharmacologic inhibition of the renin-angiotensinaldosterone system has provided important mechanistic insights into cardiovascular disease progression and offers a means to reduce morbidity and mortality in a variety of cardiovascular conditions. 䊏 Angiotensin-converting enzyme (ACE) inhibitors are one of the extensively studied classes of pharmacologic compounds. 䊏 ACE inhibitors and angiotensin receptor blockers (ARBs) reduce the risk of progression to end-stage renal disease. 䊏 ACE inhibitors and ARBs are effective in reducing clinical events in patients with symptomatic heart failure and reduced left ventricular ejection fraction. Concomitant use in one study resulted in incremental benefits.

Clinical use of inhibitors of the renin-angiotensin system (RAS) had relatively modest early expectations attributed to high plasma renin activity with angiotensin-converting enzyme (ACE) inhibitors specifically developed for the treatment of hypertension. From this relatively humble beginning, RAS inhibitors have emerged as major pharmacologic advances based on definitive clinical trial documentation of their benefits in prolonging survival and reducing cardiovascular and renal morbidity (Fig. 12-1). This rise to prominence of ACE inhibitors and angiotensin receptor blockers (ARBs) over the past 3 decades has been based on primary observations from quality mechanistic studies in animals, such as the attenuation of adverse left ventricular remodeling after myocardial infarction (MI) and the slowing of progression of renal dysfunction, leading to major clinical outcome trials and on secondary benefits first observed in the clinical trials stimulating basic investigations to provide the mechanistic underpinnings for the newly discovered clinical benefits. This progress can be attributed to cross-fertilization between data from basic laboratory research and randomized clinical trials, which support each other in the advancement of clinical care methods and in under-

䊏 Several ACE inhibitors and one ARB are effective in reducing clinical events in high-risk acute myocardial infarction patients, but their combination did not offer incremental improvements. 䊏 Aldosterone inhibition has been shown to be effective in treating severe heart failure and in patients with acute heart failure complicating myocardial infarction. 䊏 In patients with stable coronary artery disease, the addition of an ACE inhibitor to conventional risk factor modification lowers the risk of cardiovascular events. 䊏 All inhibitors of the renin-angiotensin-aldosterone system, when used in clinically effective doses, can result in significant hyperkalemia.

standing the mechanisms that interrupt disease progression.1

APPLICATIONS OF ANGIOTENSIN-CONVERTING ENZYME INHIBITORS AND ANGIOTENSIN RECEPTOR BLOCKERS Systemic Hypertension On the basis of its prevalence in the aging population and the direct pathophysiologic links of elevated blood pressure to atherosclerosis, stroke, MI, sudden death, and heart failure, hypertension is the most important population-attributable and modifiable risk factor for cardiovascular events.2 Sustained lowering of arterial pressure is the major and critical mechanism by which antihypertensive therapies reduce the incidence of adverse cardiovascular events.3,4 With the establishment of the importance of blood pressure reduction, the field has matured to the point that placebo-controlled trials in hypertension are considered unethical. To attempt to demonstrate the importance of a specific compound, contemporary trials usually compare different classes of antihypertensive agents.

203

204

Pharmacologic Intervention Non-ACE pathways (e.g., chymase)

Vasoconstriction Cell growth Na/H2O retention Sympathetic activation

Angiotensinogen 3

renin

AT1

Angiotensin I

2

Angiotensin II 1

ACE Aldosterone 4

Cough, angioedema benefits?

Inactive fragments

↑ Bradykinin

AT2

Vasodilation Antiproliferation (kinins)

Figure 12-1. Four pharmacologic sites of inhibition of the renin-angiotensin-aldosterone system. 1, Angiotensin-converting enzyme (ACE) inhibitor reduces the conversion of angiotensin I to angiotensin II and also limits the breakdown of bradykinin; 2, angiotensin type 1 receptor blocker or angiotensin receptor blocker (ARB) blocks the action of angiotensin II at the AT1 receptor; 3, renin inhibition blocks the cleavage of angiotensinogen into angiotensin I; 4, aldosterone inhibitor reduces the influence of aldosterone on renal tubules and other tissues. (From McMurray JV, Pfeffer MA, Swedberg K, Dzau V: Which inhibitor of the renin-angiotensin system should be used in chronic heart failure and acute myocardial infarction? Circulation 2004;110:3281-3288.)

Major cardiovascular events ACEi vs D/
B 6 2581/20631 CCB vs D/
B 9 2998/31031 ACEi vs CCB 5 1953/12562

∆BP 3450/26799  2/0 3839/37418  1/0 2011/12541  1/ 1

1.02 (0.98–1.07) 0.31 1.04 (1.00,1.09) 0.92 0.97 (0.92–1.03) 0.22

Cardiovascular death ACEi vs D/
B CCBvs D/
B ACEi vs CCB

6 1061/20631 9 1237/31031 5 870/12562

1440/26799  2/0 1584/37418  1/0 840/12541  1/ 1

1.03 (0.95–1.11) 0.36 1.05 (0.97–1.13) 0.33 1.03 (0.94–1.13) 0.56

6 2176/20631 9 2527/31031 6 1763/12998

3067/26799  2/0 3437/37418  1/0 1683/12758  1/ 1

1.00 (0.95–1.05) 0.76 0.99 (0.95–1.04) 0.71 1.04 (0.98–1.10) 0.68

Total mortality ACEi vs D/
B CCB vs D/
B ACEi vs CCB

0.5

1.0

2.0

Relative risk Favors 1st listed

In the context of comparator trials of antihypertensive compounds, the clinical mandate of obtaining adequate blood pressure control frequently requires use of multiple agents, which confounds interpretation of the results. Overall, ACE inhibitors have not consistently distinguished themselves as clearly superior to other effective blood pressure–lowering therapies in reducing cardiovascular risk in patients with hypertension (Fig. 12-2).4,5 The Captopril Prevention Project (CAPPP) reported similar clinical outcomes comparing an ACE inhibitor with a β-blocker–based antihypertensive regimen in more

Favors 2nd listed

Figure 12-2. Meta-analysis of outcomes of hypertensive patients treated with different antihypertensive agents. Blood pressure (BP)–lowering regimens are based on different drug classes. Mean BP difference between the first and second treatment regimens are shown. ACEi, angiotensin-converting enzyme inhibitor; CCB, calcium channel blocker; D/βB, diureticand/or β-blocker–based regimens. (From Williams B: Recent hypertension trials: Implications and controversies. J Am Coll Cardiol 2005;45:813-827.)

than 10,000 patients with essential hypertension.6 Similarly, in the smaller U.K. Prospective Diabetes Study, the level of blood pressure control in diabetic patients had more influence on clinical outcomes than whether an ACE inhibitor or β-blocker was used.7 The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) compared the cardiovascular event rates of more than 40,000 patients with hypertension plus an additional risk factor, who were randomized to a strategy of an initial antihypertensive therapy based on a diuretic (i.e., chlorthalidone), calcium antagonist

Angiotensin-Axis Inhibition (i.e., amlodipine), α-adrenergic blocker (i.e., doxazosin), or ACE inhibitor (i.e., lisinopril).8 The α-blocker arm was discontinued early by the Data Safety Committee after it became apparent that more cardiovascular events, particularly heart failure, were occurring with this therapy compared with the diuretic group.9 The other three arms went to completion, and for the primary end point of fatal coronary heart disease or nonfatal MI, no differences were apparent between these three classes of antihypertensive agents when used as initial therapy. However, for several clinically important secondary outcomes, the diuretic-based regimen was thought to be superior.10 In contrast, in the Second Australian National Blood Pressure Study (ANBP2), fewer cardiovascular events were observed in elderly hypertensive patients randomized to the ACE inhibitor (i.e., enalapril) regimen compared with a diuretic (i.e., hydrochlorothiazide) as initial therapy.11 As an even newer class of antihypertensive agents, several ARBs have accepted the challenge of addressing whether their use favorably influences morbidity and mortality in hypertension. In the Losartan Intervention of Endpoint (LIFE) trial, the clinical outcomes of blood pressure control with losartan were compared with those of the β-blocker atenolol in more than 9000 hypertensive patients with electrocardiographic evidence of left ventricular hypertrophy.12 In this trial, the ARB-based regimen was shown to be even more effective in reducing the primary end point of death, MI, or stroke than the β-blocker. This difference in clinical events occurred predominantly because of lower rates of stroke, despite similar achieved blood pressures and comparable use of diuretics as the second-line agents.13 Of the many major antihypertensive clinical outcome trials, LIFE is unique in showing a difference in clinical event rates between classes of agents that could not be accounted for by differential blood pressures. Although smaller and not as adequately statistically powered, the Study on Cognition and Prognosis in the Elderly (SCOPE) had similar trends for clinical benefits with the ARB candesartan.14 In the Valsartan Antihypertensive Long-term Use Evaluation (VALUE), valsartan was compared with amlodipine in more than 15,000 patients with hypertension plus another risk factor.15 As opposed to the LIFE trial, in which the ARB was shown to be superior to the β-blocker in reducing clinical events, in VALUE, the picture was mixed for ARB compared with a calcium channel blocker. During the early drug titration phase, blood pressure lowering was initially better with the calcium channel blocker, and overall, this was associated with fewer strokes compared with the ARB arm. However, the primary end point of cardiac events was not different between the two antihypertensive regimens. The results of the Morbidity and Mortality After Stroke (MOSES) study, which compared the effects of the ARB eprosartan with effects of the calcium channel blocker nitrendipine in 1400 stroke survivors requiring antihypertensive therapy, contraindicated the VALUE findings.

Although achieved systolic arterial pressures were comparable (average, 133 mm Hg), the stroke patients randomized to the ARB had fewer cardiovascular events than those randomized to the calcium channel blocker.16 Because a high proportion of patients with hypertension require multiple antihypertensive agents for adequate blood pressure control, the traditional focus of clinical trials with administration of a single first agent becomes less clinically relevant than determination of the optimal combination of antihypertensive agents to most safely and effectively lower an individual’s cardiovascular risk profile. The AngloScandinavian Cardiac Outcomes Trial (ASCOT) compared a strategy of blood pressure control using a calcium channel blocker (i.e., amlodipine) and adding an ACE inhibitor (i.e., perindopril) as needed for blood pressure control versus a β-blocker (i.e., atenolol) and adding a diuretic (i.e., bendroflumethiazide) as needed for blood pressure control in almost 20,000 patients with hypertension and multiple cardiovascular risks.17 Randomization to the calcium channel blocker–ACE inhibitor strategy was found to be associated with a lower risk of major cardiovascular events compared with the beta-blocker and diuretic as first- and second-line therapies. Although the blood pressure control was greater in the calcium channel blocker–ACE inhibitor group, the observed difference in cardiovascular events (16% reduction) was greater than what would be anticipated from the blood pressure differential (2.7-mm Hg systolic pressure). Despite only suggestive and by no means definitive data that indicated unique clinical benefits beyond blood pressure control across classes of antihypertensive therapies, ACE inhibitors and ARBs are an important part of the therapeutic armamentarium for treatment of hypertension to reduce blood pressure and, more importantly, cardiovascular risks. The cardiovascular benefits of inhibiting the RAS that have been demonstrated in certain higher-risk populations are probably the key to their appeal to frontline clinicians for use in individuals with even less overt disease and risks. Renal Protection ACE inhibitors and ARBs of the RAS have earned their designation as renal protective. Clinical trials have definitely demonstrated that the use of an ACE inhibitor in patients with diabetic nephropathy and nondiabetic nephropathy can slow the progression of renal disease and forestall the need for dialysis. In patients with juvenile diabetes and nephropathy, randomization to captopril reduced the proportion of patients who had a doubling of serum creatinine levels or needed renal transplantation.18 The AfricanAmerican Study in Kidney Disease (AASK), comparing an ACE inhibitor (i.e., ramipril) and calcium channel blocker (i.e., amlodipine), was prematurely terminated when it became clear that patients randomized to ramipril had a lower incidence of

205

206

Pharmacologic Intervention progression to dialysis.19 The development of microalbuminuria was also decreased by the ACE inhibitor trandolapril.20 In patients with type 2 diabetes and proteinuria, two separate studies with ARBs demonstrated improved clinical outcomes compared with other antihypertensive strategies. In the Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan (RENAAL) study, losartan was associated with fewer patients progressing to a doubling of creatinine levels or needing renal replacement therapy.21 In the Irbesartan Diabetic Nephropathy Trial (IDNT), despite the similar blood pressure lowering with a regimen based on irbesartan compared with a regimen based on amlodipine or other antihypertensives, fewer renal end points were experienced by those on the ARB than in the two other antihypertensive groups (Fig. 12-3).22 Together, IDNT and RENAAL provide convincing evidence that an ARB is of particular value in reducing renal failure in patients with diabetic nephropathy. However, it must be acknowledged that the overall balance of cardiovascular events did not favor either class of antihypertensive agents because a reduction in heart failure admissions with the ARBs was offset by fewer atherosclerotic end points in the calcium channel blocker groups. These studies also underscored the difficulties and the importance of blood pressure lowering. In IDNT, achieving systolic blood pressure control approaching 120 mm Hg was associated with the best protection against cardiovascular events.23 In patients with type 2 diabetes with only microalbuminuria, the ARB irbesartan resulted in a dose-dependent, not blood pressure–dependent, reduction in the risk of developing proteinuria,24 and valsartan was more effective than amlodipine in reducing albumin excretion.25 Taken together, the ACE inhibitors and the ARBs have earned their preferential role in the management of diabetics and others at risk for renal failure. Because there has not been a major head-to-head comparison of ACE inhibitors and ARBs in renal disease, results can only be inferred, but both inhibitors of the RAS can be considered effective in slowing the progression of renal disease and should be used in treating appropriate patients.26 Although a meta-analysis indicated that there was no specific additional renal protective effect of ACE inhibitors or ARBs compared with other antihypertensive agents,27 this interpretation is not uniformly accepted.28 National Kidney Foundation guidelines strongly endorse the use of an ACE inhibitor or ARB to treat diabetics who have hypertension and chronic kidney disease.29 Their recommendation of a target systolic blood pressure of 130 mm Hg also serves to underscore the practical need for combinations of classes of antihypertensive agents to best limit progression of kidney disease. Although the combination of an ACE inhibitor and an ARB can be effective in reducing proteinuria30 definitive clinical outcome trials of this combination have not been reported.31

Congestive Heart Failure Angiotensin-Converting Enzyme Inhibitors in Heart Failure The early clinical trials of ACE inhibitors in heart failure, Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS),32 Studies of Left Ventricular Dysfunction (SOLVD) Treatment,33 and Vasodilators in Heart Failure-II (V-HeFT II),34 have so convincingly demonstrated the importance of inhibiting the RAS with an ACE inhibitor in patients with symptomatic heart failure and depressed ejection fraction that each of the major international society guidelines designate this therapy with its highest endorsement for use in this setting.35-37 The more contemporary and clinically relevant question is whether the use of an ARB can provide incremental value alone or in combination with more extensively used ACE inhibitors (Fig. 12-4). Angiotensin Receptor Blockers in Heart Failure With the proven benefits of ACE inhibitors in the treatment and prevention of heart failure, the clinical value of ARBs had to be demonstrated relative to an ACE inhibitor.38 In an early head-to-head evaluation in elderly patients with symptomatic heart failure, there was an apparent survival benefit with use of the ARB in a relatively small trial (fewer than 50 deaths) designed to evaluate tolerability.39 Because this was not a definitive finding and mortality was not the primary prespecified outcome of this study, a much larger trial was undertaken. With more than 3000 patients and more than 500 deaths in the Losartan Heart Failure Survival Study II (ELITE II), losartan (50 mg/day) was not found to be superior to captopril (150 mg/day) because the trend was toward a survival benefit with the ACE inhibitor.40 Another approach in ascertaining the clinical value of an ARB in heart failure employed by the Valsartan in Heart Failure Trial (Val-HeFT) tested whether the ARB valsartan provided supplemental benefit in modernly managed patients.41 Although a survival benefit with the addition of the ARB was not demonstrated, the prespecified co-primary outcome of death, hospitalization for heart failure, treatment for worsening heart failure, or resuscitated sudden death was reduced by 13% in the group randomized to receive valsartan (Fig. 12-5).41 Small but significant and consistent improvements in left ventricular ejection fraction and functional status supported these observations of reductions in heart failure admissions with the use of valsartan.42 Because the concurrent medication used by the Val-HeFT cohort was not uniformly distributed, subgroup analyses were performed with respect to other proven therapies for heart failure. In the 7% of the population that at baseline were not receiving an ACE inhibitor, relatively large morbidity and mortality benefits of the ARB were apparent. Although potentially important,

Angiotensin-Axis Inhibition RR 23% P = .0006

Irbesartan Amlodipine Placebo

0.6

RRR 20% P  .024

RRR –4% P = NS

Proportion with primary end point

0.5 0.4 0.3 0.2 0.1 0.0 0

6

12

18

24

30

36

42

48

146 128 122

65 46 53

54

Months of follow-up No. at risk Irbesartan Amlodipine Placebo

579 565 568

555 542 551

528 508 512

496 474 471

304 287 280

400 385 401

216 187 190

A 50

Placebo Risk reduction, 16%

Primary composite end point (%)

P=.02 40 Losartan

30

20

10

0 0

12

24

36

48

295 329

36 52

Months of study No. at risk Placebo 762 Losartan 751

689 692

554 583

B Figure 12-3. Primary end points in the Irbesartan Diabetic Nephropathy Trial (IDNT) (A) and Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan (RENAAL) study (B). (A, from Lewis EJ, Hunsicker LG, Clarke WR, et al: Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med 2001;345:851-860; B, from Brenner BM, Cooper ME, de Zeeuw D, et al: Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med 2001;345:861-869.)

207

208

Pharmacologic Intervention

HBP CAPPP ALLHAT ANZ2

CAD EUROPA PEACE IMAGINE

MI Consensus II ISIS-4 GISSI-3 SMILE SAVE AIRE TRACE

HF CONSENSUS I SOLVD V-HeFT II PEP-CHF

Vascular HOPE

DM Prevention DREAM

DM Renal Collab Study ABCD REIN AASK

this observation was from a small subgroup, and the robustness of the finding was questioned. Another subgroup analysis from this study generated an even greater clinical quagmire because it indicated that patients receiving a β-blocker who were randomized to the ARB appeared to have a higher risk of death.41 With both β-blockers and ACE inhibitors firmly established as lifesaving and current international recommendations for the treatment of symptomatic heart failure with depressed ejection fraction strongly endorsing concurrent use of both, the result of this “triple-therapy” subgroup analysis was of particular concern.36,37 The Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity (CHARM) program with more than 7000 patients and 3.7 years of follow-up has amassed the largest patient exposure of any randomized, controlled clinical trial in heart failure.43 The CHARM program is a combination of three integrated protocols evaluating the use of candesartan in heart failure patients with depressed and preserved ejection fractions (see Fig. 12-5). In the depressed ejection fraction (EF) protocols (EF 15 mm) stents of various designs have been approved for use in the United States. Randomized clinical trials are needed to document the superiority of this untested approach to traditional long-balloon angioplasty.

EMERGING APPROACHES Despite the various mechanical approaches discussed earlier, percutaneous revascularization of long lesions

Small-Vessel and Diffuse Disease continues to be associated with increased risks of periprocedural complications and late recurrence. It stands to reason that potential benefits afforded by a variety of novel and emerging strategies aimed at increasing the safety and long-term efficacy of angioplasty may be especially evident in the setting of more complex lesion morphologies such as the long lesion. A series of prospective clinical trials has demonstrated the efficacy of potent antagonists of platelet IIb/IIIa receptor for patients undergoing percutaneous revascularization. Administration of abciximab has significantly reduced occurrence of acute ischemic events in the setting of high- and low-risk PTCA, and benefits have persisted for 3 years of follow-up.84-87 Results of the IIb/IIIa Platelet Receptor Antagonist 7E3 in Preventing Ischemic Complications (EPIC) trial have been analyzed partially. The benefits of abciximab during angioplasty are not diminished by the presence of adverse lesion characteristics, including lesion length. Because of the propensity for increased late luminal loss, long lesions may be especially responsive to locally delivered ionizing radiation therapy. According to preliminary clinical data, it appears to be especially effective in reducing neointimal hyperplasia after arterial wall injury.88 Gene-based therapy, currently in the preclinical phase of testing, may ultimately be of benefit in lesion types that are at the highest risk of restenosis.89 Deploying drug-eluting stents is feasible; the sirolimus-coated stent elicits minimal neointimal proliferation. Additional placebo-controlled trials are necessary to confirm the promising outcomes.90 Drug-Eluting Stents The SIRIUS trial enrolled patients with longer coronary lesions of 15 to 30 mm and allowed long sirolimus-eluting stent (SES) placement (maximum of two overlapping 18-mm SESs). The trial found a 9.2% restenosis rate.91 The Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) registry evaluated the efficacy of SES in 96 consecutive patients (102 lesions) with lesion lengths of more than 36 mm (mean stented length of 61.2 ± 21.4 mm; range, 41 to 134 mm). The binary restenosis rate was 11.9% and in-stent late loss was 0.13 ± 0.47 mm. At longterm follow-up (mean, 320 days), there were two deaths (2.1%), and the overall incidence of major cardiac events was 8.3%.92 Clinical trials using a paclitaxel-eluting stent (PES), such as the TAXUS V93 and TAXUS VI94 studies, evaluated the efficacy of PES for complex lesions. Both studies included more complex and longer lesion subsets compared with the previous TAXUS trials. In the TAXUS V study using a slow-release PES system, the average lengths of the lesion and total stented segment were 17.3 mm and 28.7 mm, respectively, in the Taxus stent group; 33% of lesions required multiple stents. In this study, the Taxus stent reduced the 9-month target lesion revascularization rate from

15.7% to 8.6% (P < .001) and target vessel revascularization from 17.3% to 12.1% (P = .02) compared with the control group. The angiographic restenosis rate was also lower in the Taxus group than in the control group (18.9% versus 33.9%, P < .001). However, incidence of cardiac death and stent thrombosis did not differ between the two groups. Tsagalou and colleagues95 reported the results of multiple DES implantations for 66 patients with diffuse LAD stenosis. In this study, 39 patients were treated with SES (average length of 84 ± 22 mm), and 27 patients were treated with PES (average length of 74 ± 14 mm). The number of stents implanted per patient was 2.8 ± 0.7, and the mean total stent length for the LAD treatment was 80 ± 20 mm. Procedural success was achieved in 95% of cases. Eleven (16.6%) patients had in-hospital non-Q-wave MIs (five SES and six PES), and one patient developed intraprocedural stent thrombosis. All patients had clinical follow-up, and 52 patients (79%) had angiographic follow-up at 6 months. The major adverse cardiac events rate was 15% (7.5% for SES and 7.5% for PES). No patient died, one patient had non-Q-wave MI, and 10 patients (15%) underwent target vessel revascularization.95 Aoki and associates96 reported the results of multiple DES implantations for 122 consecutive patients with de novo coronary lesions. In this study, 81 patients were treated with SES (average length of 77 mm), and 42 patients were treated with PES (average length of 84 mm). The number of stents implanted per patient was 3.3 ± 1.1, and the mean total stent length was 79 mm. Procedural success was achieved in 96% of cases. Seven (5.8%) patients had MIs within 30 days, and one patient developed subacute stent thrombosis. All patients had 1-year clinical follow-up. The major adverse cardiac events rate was 18% (18.5% for SES and 17.1% for PES). Five patients (4.1%) died, 12 patients (10%) had myocardial infarction, and 9 patients (7.5%) underwent target vessel revascularization.96 The Multicenter Prospective Nonrandomized Registry Study for Drug-Eluting Stents in Very Long Coronary Lesions (Cypher versus Taxus) (Long-DES Trial) was a nonrandomized registry comparing angiographic and clinical outcomes for 637 patients undergoing stent placement with a bare metal stent (BMS) (n = 177), SES (n = 294), or PES (n = 166) in long coronary lesions. Baseline characteristics were similar in the three arms, including the number of high-risk and diabetic patients enrolled. Average stent length was similar (42.8 mm for SES versus 43.1 mm for PES; P = NS). DES patients received more stents with longer lengths compared with BMS patients. Patients in the SES group had smaller reference vessel diameters (2.80 mm for SES versus 2.9 mm for PES). At the 6-month angiographic follow-up evaluation, completed for approximately 80% of all patients, there was 65% less in-stent late loss in the SES patients compared with the PES patients (0.27 versus 0.78 mm, P < .0001) and 65% less in-segment restenosis in SES patients compared with the PES patients (7.4% versus 21.3%, P < .001). The rate of

387

388

Coronary Intervention Table 21-7. Angiographic and Clinical Outcomes for Implantation of Three Types of Stents Outcomes Lesions Lesion length (mm) MLD before treatment (mm) MLD after treatment (mm) MLD at follow-up (mm) Acute gain (mm) Late loss (mm) Restenosis rate (%)

BMS

PES

SES

P Value

201 32.0 ± 12.3 0.78 ± 0.54 2.60 ± 0.50 1.56 ± 0.69 2.12 ± 0.67 1.02 ± 0.67 42.2

194 36.3 ± 14.5 0.77 ± 0.49 2.50 ± 0.45 1.90 ± 0.71 2.01 ± 0.57 0.56 ± 0.62 21.3

223 36.0 ± 14.9 0.76 ± 0.48 2.35 ± 0.43 2.21 ± 0.60 1.88 ± 0.59 0.14 ± 0.53 9.3

.002 NS 10

CPK-MB

Figure 27-8. Impact of embolic protection device use on periprocedural myocardial infarction (PMI) in vein graft PCI. Left, The Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) trial: cumulative distribution function curve of peak cardiac enzyme values after assignment to placebo (395 patients), GuardWire (406 patients), and the per-protocol subgroup with technically successful GuardWire use (366 patients). The creatine phosphokinase–myocardial band isoenzyme (CPK-MB) level is represented as multiples of the upper limit of normal. There is a significantly lower incidence of PMI of any size with GuardWire use. Right, The FilterWire EX Randomized Evaluation (FIRE) trial: a similar plot for patients randomized to distal protection with the FilterWire EX or to the GuardWire, showing noninferiority of the FilterWire. (Data from Baim DS, Wahr D, George B, et al: Randomized trial of a distal embolic protection device during percutaneous intervention of saphenous vein aorto-coronary bypass grafts. Circulation 2002;105:12851290; Stone GW, Rogers C, Hermiller J, et al: Randomized comparison of distal protection with a filter-based catheter and a balloon occlusion and aspiration system during percutaneous intervention of diseased saphenous vein aorto-coronary bypass grafts. Circulation 2003;108:548-553.)

pensating for most of the added expense of the EPD. The projected improved survival on the basis of reduced early complications (i.e., reduced PMI) was calculated to cost less than $4000 per year of life saved, which makes the use of EPD in vein graft PCI a very cost-effective strategy.82 The significant improvement in outcome with the use of the GuardWire occlusion device ushered a new era in which EPD use has become the standard of care with vein graft PCI. Because of ethical considerations, it was not feasible for developers of other EPD designs to test their devices against the conventional angioplasty wire, as was the case in the SAFER trial. The randomized, controlled trials leading to FDA approval of other EPDs for use in vein graft PCI were designed as noninferiority trials, with the GuardWire used in the active control arm. In a controlled trial, 651 patients undergoing vein graft PCI were randomized to receive the FilterWire EX or the GuardWire. Use of GP IIb/IIIa inhibitors was left to the discretion of the operators. The primary end point was a composite similar to that used in the SAFER trial. At 30 days, the incidence of any MI was 9% with the FilterWire versus 10% in the GuardWire arm (P = .49). There was a statistically insignificant reduction in the 1-year end point in the FilterWire arm (9.9% versus 11.6%, P = .53 for superiority; P = .0008 for noninferiority) (see Fig. 27-8).35 An examination of these results demonstrated a favorable interaction between the use of GP IIb/IIIa inhibitors and FilterWire use, but not with GuardWire use. This may be related to the improved flow seen with GP IIb/IIIa inhibitor administration in patients receiving FilterWire protection, probably due to reducing the degree of platelet aggregation and deposition on the surface of the filter.83

Similarly designed trials were conducted to test the efficacy of the Triactiv device (Kensey Nash), the Proxis system (St. Jude Medical), MedNova Emboshield (Abbott), and the Spider device (eV3, Inc.). In those trials, the 30-day primary end point was reached in 8% to 11% of patients, achieving the preset standard for noninferiority compared with the GuardWire or FilterWire in all studies.84 Percutaneous Coronary Intervention for Acute Myocardial Infarction The concept of EPD use in primary PCI is attractive and intuitive. These are the prototypical thrombotic lesions with a very high likelihood for distal embolization. The success of EPDs in vein graft PCI led to clinical trials examining the feasibility of the concept. In a small study of 72 patients with acute MI, the PercuSurge GuardWire was used during primary stenting of the infarct-related artery in 42 patients, with 30 patients undergoing primary stenting after thrombectomy with no EPD. The GuardWire group of patients had a significantly better corrected TIMI frame counts and TIMI myocardial perfusion grades. At 3 weeks, the GuardWire group also demonstrated a significantly better ejection fraction, improvement in ejection fraction, and improvement in wall motion abnormality.85 However, the larger randomized trials did not confirm the initial favorable impression regarding EPD use in primary PCI. The Enhanced Myocardial Efficacy and Removal by Aspiration of Liberated Debris (EMERALD) trial was an international multicenter, prospective, randomized trial enrolling 501 patients with ST-segment elevation MI undergoing primary or rescue PCI. Patients were randomized to

505

506

Coronary Intervention PCI with the PercuSurge GuardWire distal protection or to PCI without EPD. Two co-primary end points were prespecified: ST-segment resolution after PCI assessed by continuous Holter monitoring and infarct size measured by nuclear imaging between days 5 and 14. Secondary end points included major adverse cardiac events. Among 252 patients assigned to the GuardWire protection, debris was retrieved in 73% of cases. Disappointingly, there was no difference between the two groups in any of the primary or secondary end points (ST resolution: 63% versus 62%; infarct size: 12% versus 9.5%; P = NS for both). At 6 months, the frequency of major adverse cardiac events in the distal protection and control groups was similar (10.0% versus 11.0%, P = .66).86 Another trial tested the efficacy of the FilterWire distal protection in the setting of PCI for acute MI (i.e., Protection Devices in PCI treatment of Myocardial Infarction for Salvage of Endangered Myocardium [PROMISE]). However, the methodology of PROMISE was distinct from that of the EMERALD trial. Only 200 patients were randomized: 68% with ST-segment elevation MI, with the remainder diagnosed with non-ST-segment elevation MI. The primary end points were the coronary flow velocity measured by an intravascular Doppler wire and the size of the infarction measured by hyperenhancement on MRI scans 3 days after the procedure. Similar to the EMERALD trial, there was no difference between the patients randomized to the PCI with FilterWire protection and those who did not receive any EPD.87 There are several potential explanations for the disappointing results of the EMERALD and PROMISE trials. Using EPD may delay restoration of epicardial flow, and the devices may cause further embolization while crossing the lesion, negating any favorable effects of subsequent protection. The incomplete aspiration of liberated debris or leaking of vasoactive substances released from the ruptured plaques may lead to further downstream damage at the time of EPD removal. Embolization into side branches may play a role, particularly in cases with acute thrombotic occlusion, leading to initial TIMI flow grade 0 and absence of visualization of the distal artery at the time of EPD positioning. These results also indicate a relative underestimation of the degree of existing damage and the role of reperfusion injury in determining the final infarct size after primary PCI.88 Carotid Stenting Although the clinical implications of embolization were first elucidated for coronary interventions, the paradigm is applicable in angioplasty procedures in other arterial beds. Interventional procedures in the carotid and renal arteries are two areas where embolization may be particularly significant. Embolization appears to occur much more frequently after carotid stenting than carotid endarterectomy (CEA). Using transcranial Doppler (TCD) monitoring, microscopic embolization occurs at least eight times more fre-

Figure 27-9. Transcranial Doppler monitoring of middle cerebral artery flow during elective carotid artery stenting. The highintensity transients observed at the time of balloon deflation represent a surge of microemboli from the extracranial site of angioplasty to the intracranial circulation. (From Topol EJ, Yadav JS: Recognition of the importance of embolization in atherosclerotic vascular disease. Circulation 2000;101:570-580.)

quently with carotid angioplasty and stenting than with CEA.89 Most patients undergoing carotid stenting have TCD evidence of microembolization (Fig. 27-9). Similar to embolization related to coronary interventions, it appears that evidence of systemic inflammatory response can lead to more embolization. In a small study of 43 patients undergoing carotid stenting with TCD monitoring of the ipsilateral middle cerebral artery, there was a positive correlation between TCD-identified microembolism and the preprocedural leukocyte count, a marker of systemic inflammation. This correlation remained significant even after adjusting for age, gender, comorbidities, medical therapy, and use of EPDs.90 Potentially, even small embolic particles are poorly tolerated by the cerebral microcirculation.91 In an ex vivo model of carotid angioplasty, particles generated from human carotid plaques were injected into the cerebral circulation of rats. Stenting produced almost twice as much embolization as balloon angioplasty in this model; passage of the guidewire also produced embolization, although only about one fourth as many emboli as balloon angioplasty. Particles less than 200 µm in diameter did not cause cerebral ischemia during the first 3 days after the procedure, whereas particles 200 to 500 µm in diameter did cause neuronal death. However, at 7 days, injury was detected by fragments of both sizes. If smaller sizes of emboli are relevant in humans, an occlusion device may be better than a filter device. Although filters can be designed with smaller pore sizes, the disadvantage is that this can increase the risk of thrombosis by the filter itself and decrease distal flow. Several EPDs were designed for use in conjunction with carotid angioplasty and stenting in the hope of reducing the incidence of procedure-related strokes. By designing a safe and effective EPD, the high-risk

Periprocedural Myocardial Infarction and Embolism-Protection Devices percutaneous procedure can be converted into a lowrisk procedure comparable to, or even safer than, the current standard of care, carotid endarterectomy. Reimers and colleagues92 reported their initial experience with three filter designs (i.e., Angioguard, Neurosheild, and FilterWire) in 84 patients undergoing carotid stenting. Macroscopic debris was collected in 53% of filters, and histologic analysis of the debris revealed lipid-rich macrophages, fibrin, and cholesterol clefts. The early experience with the balloon occlusion–variety of EPDs (e.g., PercuSurge GuardWire) was reported for a series of 75 patients. In this series, macroscopic debris was collected from all cases (100%), and histologic analysis was very similar to particles obtained from filter devices.93 In addition to retrieval of macroscopic and microscopic debris, there is evidence that the use of EPDs during carotid stenting effectively reduces embolism to the cerebral microcirculation. These data have been gleaned from studies using magnetic resonance diffusion-weighted imaging (DWI), which is the most sensitive imaging modality for detection of early cerebral ischemia.94,95 Comparison of DWI scans before and after carotid stenting reveals that use of EPDs significantly reduces the incidence and number of new lesions identified on the postprocedural scan. Most new lesions were small (3 hr), older age (≥85 yr), and hypertension.

bolysis, such as subdural, subarachnoid, or parenchymal ICH. CT may also detect mass lesions or hemorrhagic infarctions. A CT without contrast is an excellent tool to rule out hemorrhage, but its sensitivity for discriminating between ischemic and infarcted brain is not as good as MRI. MRI perfusion and diffusion techniques can distinguish ischemic brain at risk from the infarcted core, thereby identifying salvageable brain tissue. Diffusion imaging is of value in the rapid detection of infarction and in the differentiation of new versus old brain infarctions. Perfusion imaging with contrast CT is also available, although most experts believe that MRI is superior.

MANAGEMENT OF PHYSIOLOGIC VARIABLES The management of acute stroke involves reducing the risk of recurrent events and minimizing the disability that occurs secondary to the established stroke. Acute therapy involves management of physiologic

NEW IMAGING TECHNIQUES

Table 46-3. Eligibility for Thrombolysis in Stroke Indication Ischemic stroke, within 3 hr since onset of symptoms Clinical Contraindications Any history of intracranial hemorrhage Blood pressure: systolic, >185 mm Hg; diastolic, >110 mm Hg Rapid improvement in neurologic status Mild neurologic impairment Symptoms of subarachnoid bleeding Stroke or head trauma within the last 3 mo Gastrointestinal/genitourinary hemorrhage within last 3 wk Major surgery within last 3 wk Recent heart attack Seizure with stroke Taking oral anticoagulants Received heparin within 48 hr Radiologic Contraindications Evidence of intracranial hemorrhage on computed tomography

The imaging choice for acute stroke patients is CT or magnetic resonance imaging (MRI). Imaging is the cornerstone for selecting candidates for stroke therapy. The purpose of the baseline CT is to detect conditions that make the patient ineligible for throm-

Laboratory Contraindications International normalized ratio (INR) >1.7 Platelet count LA pressure, shunts right to left; if LA pressure > RA pressure, may stay closed or shunt left to right Shunt is dynamic and bidirectional Shunt direction depends on RA-LA pressure difference, RVEDP-LVEDP difference, phase of respiration, and volume status

Usually RA pressure = LA pressure, but shunt is left to right because RV is more compliant; shunt reverses with Eisenmenger physiology Same as for PFO Same as for PFO

LA, left atrium; LVEDP, left ventricular end-diastolic pressure; PFO, patent foramen ovale; RA, right atrium; RVEDP, right ventricular end-diastolic pressure.

who had a CVA with an identifiable cause.14 Other researchers have noted a higher frequency of PFO in patients with a cryptogenic stroke irrespective of age (17.5 12.5-17.4 9-12.4 75 yr old offset reduction in ischemic stroke

SPAF III Warfarin INR 2-3 vs. aspirin plus low-intensity, fixed-dose warfarin in selected high-risk patients Aspirin-treated low-risk cohort

1993-1995 1993-1997

Warfarin INR 2-3 offers large benefits over aspirin plus low-intensity, fixed-dose warfarin for high-risk patients Patients whose stroke risk is low when given aspirin can be identified (validation of the SPAF risk stratification scheme)

INR, international normalized ratio.

The gold standard for treating cardiac thrombus has been oral anticoagulation. Many patients cannot or will not take anticoagulants, especially elderly patients. In this group, warfarin (Coumadin) has a narrow therapeutic window and high intracerebral bleeding rates, limiting it as a therapeutic option. The Stroke Prevention in Atrial Fibrillation (SPAF) studies assessed the value of antithrombotic therapies such as warfarin, aspirin, and their combination for preventing stroke in 3950 patients with nonvalvular AF. Table 49-1 summarizes the SPAF trials for cardioembolic events. SPAF II revealed that older patients with AF have a high stroke risk while taking aspirin. Yet the risk of cerebral hemorrhage is unacceptably high with warfarin therapy.8 It is for this reason that atrial appendage obliteration has attracted recent attention, especially as a percutaneous procedure. The first percutaneous LAA ablation using a catheter-based closure device was done by Nakai and colleagues, who demonstrated that transcatheter LA obliteration was feasible in an animal study.9 Benign healing occurred without adverse hemodynamic effects and without residual thrombus or tissue damage around the device. Other trials for stroke prevention in high-risk patients with nonrheumatic AF also used percutaneous LAA transcatheter occlusion devices. The feasibility of this treatment is established, and it suggests that percutaneous therapy is a clear therapeutic strategy for patients with AF and a contraindication to lifelong anticoagulation therapy.10

ANATOMY OF THE LEFT ATRIAL APPENDAGE Embryology The LAA develops from the primordial LA. Its internal surface arises from the primordial pulmonary vein and is quite uneven compared to the LA.11 The appendage develops during the third week of gestation, earlier than most smooth portions of the LA cavity.12 Most of the LA wall originates from primor-

dial pulmonary venous cells. The pulmonary veins form gradually and develop around the dorsal primordial LA wall, forming the four pulmonary veins with separate atrial orifices by about 8 weeks. In contrast, the remainder of the primordial LA wall becomes the LAA, which is tubular and attached to the LA.11 Fetal echocardiography has recently shown normal and abnormal structures of the developing heart at 16 weeks’ gestation, permitting identification of congenital abnormalities such as LA isomerism, which is highly correlated with LAA isomerism found at autopsy or surgery.13 Macroscopic and Microscopic Appearance Many macroscopic LAA features were described from both external and internal vantage points by Sharma and colleagues, who suggested that there was no crest (the “crista terminalis”) and that tapered pectinate muscles “spilled” out to the atrial septum. This is unlike the right atrial appendage, and it differentiates the LAA internal surface features from those of the right atrial appendage. Pectinate muscles are confined to the LAA interior, and there is no terminal crest within it. The external LAA morphology is tubular, with a hooked apex in most cases that points downward. By contrast, most of the right atrial appendage had a broad base and a hooked apex that in most cases points upward.14 Examination of 500 normal autopsy hearts for LAA lobe count found that two lobes were most often present (54% of cases) (Fig. 49-1). These lobes typically exist in different planes, so that imaging must be done in multiple planes to visualize the entire LAA body.15 Significant differences were found in mean length, width, and orifice size, which increase with time up to 20 years of age in both sexes. The investigators concluded that LAA dimensions change with individual age and sex. Ernst and associates described the LAA principal axis course by making casts and measuring minimal and maximal orifice diameters, length, width, and volume. The principal axis was tortuous and spiraled

The Left Atrial Appendage: Anatomy, Physiology, and Therapeutic Percutaneous Closure

2

3 1

A LPA

LSPV

2 LA

LIPV

LAA 1

B LS

P V

in 42% of hearts, extremely bent and slightly spiraled in 24%, slightly bent and extremely spiraled in 5%, and slightly bent and spiraled in 23%. Fifty-six percent had more than five branches (orifice area, >10 mm2), and 47% had more than 40 “twigs” (orifice area, 1-10 mm2). The mean minimal and maximal orifice diameters were 15 and 21 mm, respectively. Mean length (bottom to top) was 30 mm, mean width was 21 mm (at right angles), and mean volume was 5220 mm3 (5.22 mL).16 Blood supply to the LAA is typically provided by the left circumflex or right coronary arteries from positions in the left and right atrioventricular sulci.17 The microscopic appearance of the LAA in patients with chronic AF has been compared with that in patients with sinus rhythm.18 Patients with AF generally exhibit marked fibrous endocardial thickening and a much smoother internal LAA surface. It is unclear whether the LAA endocardium resembles the remainder of the heart in structure and function,19 although LAA myocardial cells are visually similar to those in other parts of the heart.20 The LAA epicardial thickness is greater for those portions overlying the ventricles. Morphologic changes occur in the LAA after exposure to ionizing radiation. Generalized collagen (fibrosis) develops in rat hearts exposed to radiation, with reduction of appendage volume and loss of elasticity. These changes appear to negatively influence ventricular function,21 because evidence suggests that the LAA plays a role in LV filling and contributes to normal cardiac function.22-25

PHYSIOLOGY OF THE LEFT ATRIAL APPENDAGE LA Oa

L

Oe

W

C Figure 49-1. A, Drawing of a left atrial appendage (LAA) anatomic specimen consisting of three lobes. Each protrusion from the body comprises a separate lobe. Directional changes or bends in the tail do not usually comprise new lobes. B, LAA with two lobes (1 and 2). C, Measurements of the LAA in a gross pathologic specimen. The echocardiographic orifice (Oe) is larger than the anatomic orifice (Oa). The length (L) of the appendage is a curvilinear distance (dashed line) from Oa to the tip of the tail, whereas the maximal width (W) is a straight-line measurement. Almost all appendages in the adult contain pectinate muscles greater than 1 mm in diameter. Oe is usually measured from the junction of the left superior pulmonary vein (LSPV) as it enters the left atrium (LA) to the junction of the LA and LAA. LIPV, left inferior pulmonary vein; LPA, left pulmonary artery. (From Veinot JP, Harrity PJ, Gentile F, et al: Anatomy of the normal left atrial appendage: A quantitative study of age-related changes in 500 autopsy hearts. Implications for echocardiographic examination. Circulation 1997;96:3112-3115.)

LAA distensibility is typically higher than that of the atrial body. If the LAA is compressed, LA dimensions and mean atrial pressures are increased, as are transmitral and pulmonary venous flow velocities, as shown by echocardiography.22 LAA distensibility lowers the LA pressure-volume relationship and augments hemodynamic function. In one preclinical study, LAA distension induced urine output, sodium excretion, and increased heart rate, a reflex presumably due to this diuresis. The atria and their appendages have a variety of innervations and receptors. Atrial innervation has both sympathetic and parasympathetic fibers. If both appendages are destroyed, parasympathetic afferent reflexes from the cervical vagal nerve and sympathetic efferent pathways are reduced.23 Myelinated and unmyelinated afferent fibers pass via the vagus to the brainstem or through sympathetic afferents to the spinal cord.24 The LAA has prominent muscular ridges. Its contraction is easily visible during open heart surgery,19 and it has inherent contractions. One case report described a giant LAA that moved more than 3 inches with each systolic beat. The LAA likely contributes to cardiac output. One study found that the cardiac output was halved in ligated compared to intact LAA in a guinea pig model.25 Such differences in cardiac

865

866

Intracardiac Intervention output suggest that the LAA functions as a contractile cardiac chamber that assists in LV filling. The LAA may stimulate thirst in hypovolemic patients, and it may impair the hemodynamic response to volume overload. Normal LV function in the dog has minimal contribution from the LAA. However, in LV dysfunction, LA and LAA dysfunction reduce cardiac output further.26 Another canine study showed that atrial appendectomy decreased cardiac output in experimental high-output heart failure.27 In patients with LV dysfunction, LAA function is also sometimes depressed. It can improve remarkably after treatment of the heart failure, as measured by LAA size, area, and emptying velocity. LAA size may decrease substantially more than LA size, suggesting that the LAA is more compliant than the LA itself. The LAA may thus be a volume reservoir that limits atrial pressure rise and may help protect against pulmonary congestion.28 Another important LAA physiologic function is hormone secretion, making it an endocrine organ. It releases both atrial natriuretic peptide (ANP) and brain natriuretic peptide.29,30 These have combined natriuretic, diabetic, and vasodilatory properties. In a canine model of experimental high-output heart failure, the atrial appendages contained approximately 30% of the total atrial ANP, and atrial appendectomy decreased the secretory function of ANP.27 In a human study, ANP concentration was fivefold to tenfold greater than that in the normal functioning heart, particularly the prohormones β-ANP and γ-ANP.31

IMAGING AND DETECTION OF THROMBUS Invasive Studies Initially, LAA thrombus was detected by invasive angiography, but this procedure required a transseptal approach. Alternatively, pulmonary arteriography was an invasive examination for visualizing the LAA in the levo phase.32 Since the advent of echocardiography, invasive angiography is rarely performed. Echocardiography Transthoracic echocardiography (TTE) is typically not sensitive enough for detecting LAA thrombus, compared with transesophageal echocardiography (TEE).33,34 New TTE systems may permit detection of LAA thrombi, and accurate determination of LAA function in most patients with neurologic deficits is feasible.35 Other studies show that several TTE variables are predictive of LAA thrombi, suggesting that TTE may now be useful for detecting LAA thrombi.36 TEE remains the gold standard for detecting LAA thrombus and usually provides high-quality images and physiologic information about appendage function. Two-dimensional and three-dimensional (3D) echocardiography and Doppler techniques are widely

available. Basic biplane TEE allows the horizontal short-axis view at the base of the heart37 and the twochamber longitudinal view of the LA and LV.38 Figure 49-2 shows multiplane TEE39 of an LAA with thrombi, before and after percutaneous device–based obliteration. TEE imaging allows great versatility in obtaining views with multiple anatomic planes.40 Two-dimensional imaging of the LAA is limited owing to its 3D anatomic complexity. Many reports have examined the relationship between LAA area and ejection fraction and LAA function.41-50 These studies suggested that functional measurement by LAA cross-sectional area has few advantages over measurement by Doppler echocardiography.51 Two-dimensional TEE reliably detects spontaneous echocardiographic contrast (SEC) and also semiquantitatively grades SEC.52,53 Multiplane TEE allows thrombus detection despite complex structural LAA features, although it also may result in misdiagnosis from oversensitivity54 or undersensitivity.55 LAA assessment by 3D echocardiography is now available, and a recent investigation revealed that this modality is promising for evaluating the LAA with or without thrombi. It is important to obtain good-quality images in TTE, because this real-time 3D modality may be used to screen the LAA for better understanding of its complicated 3D anatomy.56 Computed Tomography and Magnetic Resonance Imaging Recent advances in computed tomography (CT) of the cardiovascular system have introduced a new and exciting noninvasive imaging modalities that are complementary to other diagnostic modalities such as echocardiography and invasive methods. Multislice computed tomographic angiography (CTA) provides better image quality for intracardiac thrombus than does conventional CT,57 and several studies found it to be more sensitive and more specific than TTE in identifying LA thrombus.58-60 This is especially true for thrombi in the LAA and lateral wall of the LA, which are better detected by multislice CTA than by TTE.61 The accuracy of multislice CTA is comparable to that of TEE as a semi-invasive diagnostic tool for detecting LAA thrombi. In addition, multislice CTA can reveal the 3D structure of the LAA and its complicated anatomy, which TEE cannot demonstrate (Figs. 49-3 through 49-5). Electron-beam computed tomography (EBCT) allows high temporal resolution to visualize atrial thrombi in patients with AF better than with TEE.62 Multislice CT scanning provides higher resolution, better signal-to-noise ratio, and a shorter overall scan time than does EBCT, but at the expense of higher radiation doses. A comparison of the ability to identify LA thrombus by 3D CT scanning versus TEE in patients with persistent AF showed excellent agreement between TEE and CT for LAA size measurement in 31 patients.63 Multislice CTA may provide excellent image resolution, but it has limitations because of the irregular cycle length in AF.

The Left Atrial Appendage: Anatomy, Physiology, and Therapeutic Percutaneous Closure

Figure 49-2. A, Transesophageal echocardiogram (TEE) shows human left atrial appendage (LAA) containing multiple thrombi (arrow). B, TEE showing LAA obliteration using the WATCHMAN device after 45 days. The device is well seated in the LAA (right panel). The left panel shows an LAA angiogram from the femoral vein, indicating the location of the LAA and the closure device. The TEE probe in the esophagus is also seen.

Cardiac magnetic resonance imaging (MRI) has satisfactory temporal resolution for LAA imaging (Fig. 49-6), and contrast-enhanced MRI can detect thrombi in the LAA. MRI is also used for planning and follow-up of percutaneous LAA transcatheter occlusion.64

in volume and shape, and that this variability should be considered when interpreting LAA images, especially when diagnosing LAA thrombi.16 Better morphologic understanding is needed for future treatments that may use these new modalities.

Postmortem Studies

PATHOPHYSIOLOGY AND RELATED DISORDERS

Anatomic qualitative and quantitative LAA evaluations of many postmortem hearts were reported in the 1980s and 1990s. LAA morphology was found to be far more complicated than previously thought. Distinct variability was shown for LAA mean orifice size, length, and width. These dimensions increased at an average rate of 0.024, 0.041, and 0.030 cm/yr, respectively, during the first 20 years of life (P < .01). Length and width increased more slowly in females. The orifice size and width increased in individuals older than 20 years of age, at average rates of 0.0016 and 0.0019 cm/yr, but the length decreased, at an average rate of 0.0040 cm/yr.15 Other studies showed relative enlargement of the LAA volume with AF (7060 mm3 vs. 4645 mm3 in sinus rhythm; P < .01), LV hypertrophy (5740 mm3 vs. 4639 mm3 without hypertrophy; P < .01), myocardial scars (5923 mm3 vs. 4891 mm3 without scars; P < .05), closed foramen ovale (5515 mm3 vs. 4037 mm3 with patent foramen ovale; P < .01), and LAA thrombi (8566 mm3 vs. 5027 mm3 without thrombi; P < .01).16 The authors concluded that LAA configuration varies

Isomerism LAA isomerism is a rare condition that is often associated with other congenital anomalies and is known for its dire prognosis. The diagnosis is confirmed at surgery or autopsy and is defined by bilateral LAA and bilateral, bilobed lungs. The LAA arises from a narrow base, with a long, finger-like appearance compared with the right atrial appendage, which has a broad base and triangular shape. In isomerism, the LAA pectinate muscles are confined to only the appendage itself. It is often complicated by ventricular noncompaction. The echocardiographic diagnosis of LAA isomerism is made by concomitant findings of atrioventricular canal defects, interrupted inferior vena cava with azygos vein continuation, and heart block, usually in the presence of abdominal situs.65 LAA isomerism is also a diagnostic feature of LA isomerism, called the polysplenia syndrome. It may necessitate eventual cardiac transplantation because of the high mortality rate due to other cardiac and noncardiac anomalies.66,67

867

868

Intracardiac Intervention

A

B

C

D

Figure 49-3. Multislice computed tomographic angiography. Images were obtained with the volume-rendered technique (VRT) and show the location of the left atrial appendage (LAA) above the left main coronary artery. Multislice computed tomography allows threedimensional (3D) visualization of the LAA in living patients. These VRT images were reconstructed with the use of a 3D workstation (Vitrea2, Vital Images Inc., Minnetonka, MN). A, Exterior anterior view of the heart including the LAA (arrow), typically situated above and covering the left main coronary artery and its bifurcation. The coronary artery emerging from under the LAA is the left anterior descending, with several diagonal branches easily seen. B, Lateral view from left side of the heart after digital trimming of the left atrium (LA) and sectioning of the LAA in half. The LAA is located in the anterior portion of this image (arrow), and its orifice is separated from left superior pulmonary vein. Pectinate muscles can be seen inside the LAA. C, Cranial view from the heart after digital removal of the superior portion of the LA, pulmonary arteries, and veins. The ostium of the LAA is seen (arrow), and the pectinate muscles inside the LAA orifice are seen. D, After trimming of the left side of the LA and left ventricle (LV), the location of the LAA is easily visible. MV, mitral valve.

Juxtaposition The first case of levojuxtaposition of the right atrial appendage was reported by Birmingham.68 Since then, many cases of this abnormality have been described, often in association with other anomalies such as abnormal ventricular looping and abnormalities of the conus. In the 1950s, the term “atrial appendage juxtaposition” was introduced, describing its positional analysis. In atrial appendage positional right-sided juxtaposition, the right-sided structure is morphologically the LAA; this is the opposite of the more familiar situation of left-sided juxtaposition, in which the morphologically right atrial appendage is malpositioned.69 It is usually associated with other cardiac anomalies including hypoplastic LV and normal sinus. Juxtaposition of the right atrial appendage typically has a hypoplastic right ventricle and abnormal conus.70,71 Recently, fetal echocardiography has provided clues to its diagnosis before

birth. Abnormal vascular spaces are seen on the left side of the cross-sections and in the great arterial trunks in the case of atrial appendage juxtaposition.72 A preclinical model for left juxtaposition of the atrial appendage in chicks has been investigated to resolve several morphogenetic questions of human congenital cardiac malformations.73 Left Atrial Appendage Dysfunction and Myopathy AF is the most significant cause of emboli from the LAA. Poor LAA contraction causes thrombus formation not only in AF but also in sinus rhythm on occasion.74 LAA contraction is absent, comparable to LAA stunning, and produces an “acute LAA myopathy” that results in thrombus formation. On occasion, the cardiac musculature of some patients may gradually fatigue and develop a “delayed myopathy,” which creates susceptible environments for

The Left Atrial Appendage: Anatomy, Physiology, and Therapeutic Percutaneous Closure

Figure 49-4. A, Multislice computed tomographic angiography was used with volume-rendered technique (VRT) to derive this left anterior oblique view of the heart of a living patient. The complex structure of the left atrial appendage (LAA) is visualized (arrow). B, View of the LAA orifice from the left atrium, showing the orifice location next to the opening of left superior pulmonary vein (LSPV). C, Magnified lateral view of LAA internal anatomy, illustrating the complicated pectinate muscle anatomy (arrow). D, Isolated VRT image showing the structure of the LAA in a different patient. This LAA is smaller, but its complex three-dimensional structure can again be seen.

A

B

C

D

thrombus formation. TEE should be performed for patients with AF in whom there is a question of LAA myopathy and thrombus, to measure flow velocity and magnitude. It may allow early anticoagulation therapy even in patients with sinus rhythm. In such LAA dysfunction related to AF, atrial pacing at increased rates and isoproterenol may reverse mechanical atrial stunning associated with shortduration AF. In contrast, long-duration AF has an attenuated response to this therapy.75 Other Conditions: Aneurysm, Infarction, Cardiomyopathy, Amyloidosis Many LAA dysfunctional conditions and related syndromes have been described. LAA aneurysms are rare but are commonly associated with atrial tachyarrhythmias and thromboembolism.76 Aneurysm resection is the only reported curative strategy, but percutaneous LAA closure may become an alternative, similar to aortic aneurysms that are treated with percutaneous endovascular grafts.77 LAA infarction is also rare, but one report indicated appendiceal thrombosis in an infant after twisting of the appendage base with subsequent thrombosis within myocardial vessels.78 Isolated atrial amyloidosis is a cause of AF, and there is an inverse correlation with atrial fibrosis.79 In the setting of the hypertrophic and dilated cardiomyopathy, LAA function as a

means to measure pulmonary venous flow patterns may be useful to predict clinical outcomes.50,80 Cardiovascular and Cerebrovascular Events Several studies have suggested that measuring LAA function allows prediction of thrombotic or thromboembolic risk in patients with AF or atrial flutter. LAA function is often unrelated to global LA function.51 Reasons for this difference may be that the LAA and the main LA cavity originate from different embryologic sources. The trabecular LAA is a remnant of the embryonic LA, whereas the smooth LA cavity originates from an outgrowth of the pulmonary veins, as noted previously. Dissociation of LA and LAA mechanical activity typically occurs in patients who have undergone cardioversion, in whom organized LA mechanical activity may be present along with disorganized LAA contraction.81 High-velocity blood flow in the LAA is sometimes observed in patients with AF. Apparent differences between LA and LAA function may result from the multiple determinants of mitral inflow velocities that are independent of LA contractility, such as LV diastolic characteristics and loading conditions. In spite of these dissociations, LAA function may be a clinically applicable surrogate for overall LA function.51 In addition, many studies have revealed an association between LAA dysfunction and previous systemic embolic events, primarily cerebral emboli.82-86

869

870

Intracardiac Intervention

A

B Figure 49-5. A, Multislice computed tomographic angiography (CTA) image derived with volume-rendered technique (VRT) showing a magnified view of the left atrial appendage (LAA) orifice from the left atrium (arrow). The orifice forms a threedimensional spiral configuration rather than a single circular ostium. B, Magnified multislice CTA VRT image of the LAA from the external (left) and internal (right) views. Pectinate muscles appear as columns and can be seen in the inner portion of the LAA (arrow).

Left Ventricular Dysfunction and the Left Atrial Appendage Atrial function, including LAA emptying, is closely related to LV dysfunction87 in canine models. Atrial failure has little effect on cardiac output and rightsided heart pressures in normal LV function because of compensatory conduit function. If early LV dysfunction coexists, however, reservoir and conduit functions became insufficient to compensate for impaired atrial contraction. AF impairs LV cardiac function, a fact that has been known for a century. This includes not only diastolic dysfunction but also AF-mediated systolic dysfunction. Therefore, LAA function should always be checked in patients with AF.

Figure 49-6. Still frame from cine cardiac magnetic resonance imaging study showing the left atrial appendage (arrow) and its relation to the left atrium (LA). LV, left ventricle. (Courtesy Dr. John Lesser, Minneapolis Heart Institute Advanced Imaging Center.)

matic heart disease. This is true both in sinus and AF rhythms.88 Mitral stenosis especially limits normal augmentation of LAA flow during diastole in patients with AF, according to stenosis severity.89 Rheumatic AF with severe mitral stenosis demonstrates low-toabsent LAA velocity, in contrast to nonrheumatic AF, which shows a wide spectrum of flow velocities.82 Direct LA and LAA involvement in the rheumatic inflammatory process may elevate LAA pressure. Also, severe hemodynamic impairment from mitral stenosis occurs and also may elevate LA pressure. Atrial myopathy is a frequent cause of chronic LA pressure elevation. Similar phenomena have been noted in patients with mitral valve prostheses.90

Mitral Valve Disease Mitral stenosis is well known to increase resistance to LAA emptying, both actively and passively, resulting in lower LAA flow velocities despite normal cardiac rhythms. It is hemodynamically very important: one investigation suggested that LAA contraction velocity is significantly lower in patients with mitral stenosis compared to patients without rheu-

PATHOGENESIS OF STROKE AND TRANSIENT ISCHEMIC ATTACK Left Atrial Appendage Flow and Thrombus Formation in Atrial Fibrillation and Atrial Flutter Blood flow patterns in the LAA have been studied by TEE, and thromboembolism risk is greater with slower

The Left Atrial Appendage: Anatomy, Physiology, and Therapeutic Percutaneous Closure flow velocity. Low velocity in AF correlates well with thrombi in the appendage. It is often found in atrial stunning, such as after cardioversion, and in spontaneous recovery from AF. The magnitude of LAA filling and emptying in AF may not relate to ventricular rate and may be of limited relevance for preventing thromboembolism. LAA contraction velocities less than 20 cm/sec are associated with SEC in 75% of patients, significantly more than the 58% frequency observed in patient groups with higher-velocity flow, as was shown in the SPAF III TEE substudy. Significant LAA dysfunction is correlated with LAA thrombus formation,53,91 and anticoagulation decreases the chances of thrombus formation. LAA flow velocities independently predict thrombus formation even in various clinical hemostatic settings.92 The SPAF III trial showed a relationship between thrombus formation and LAA flow velocity. Namely, low-to-absent flow velocities in the LAA (≤20 cm/sec) indicated higher thrombus prevalence than in patients with high LAA velocity (17% vs. 5%, respectively).93 This study prospectively confirmed that LAA dysfunction is a risk factor for future embolic events. Patients with low flow velocities in the LAA (≤20 cm/s) have 2.6 times greater ischemic stroke risk than patients with higher LAA velocities. Decreased LAA velocities indicate increased clinical risk for thromboembolization.94 SEC and LAA dysfunction are strongly correlated with thrombus formation and thromboembolism in rheumatic heart disease, and more so in AF than in atrial flutter,43,82 because there is less LAA dysfunction in flutter. Decreased LAA flow velocities after catheter ablation of chronic atrial flutter recover slowly, and SEC typically resolves within 2 weeks in patients with preserved LV function.95 Imaging of Low-Flow “Smoke” SEC results from blood stagnation in the LA and LAA and carries a high incidence of thrombus and thrombogenic events. “Smoke-like” formation visualized by TTE has been known for more than 20 years. TTE remains the standard method used to detect SEC and evaluate the LAA. EBCT and multislice CT scanners are becoming widely available and have sufficient spatial resolution to assess fine LAA anatomic structure. LAA thrombus, even without AF, can be detected. TEE remains the gold standard for LAA evaluation, because multislice CT requires exposure to contrast agent and radiation. TEE is a semiinvasive tool, however, and it is not feasible in certain situations in which multislice CTA may be a reliable alternative because of shorter overall examination times and better images with less contrast utilization.96 MRI visualizes the LAA and thrombus even in patients with nonrheumatic AF.97 One study evaluated 50 subjects with nonrheumatic AF and a history of cardioembolic stroke with MRI and TEE. In all subjects, MRI allowed visualization of high-intensity masses in the LAA and clearly distinguished throm-

bus from LAA wall structure using triple-inversion recovery sequences. In the severe SEC cases, the LAA lumen was seen in great detail. With thrombus in SEC, a high-intensity mass was seen. Compared with TEE, MRI does not require esophageal intubation, and the detection of high-intensity masses between these two modalities is concordant.

CLOSURE OF THE LEFT ATRIAL APPENDAGE Concepts LAA obliteration is both theoretically and practically feasible, and it is commonly performed during cardiac surgery, because the LAA is generally responsible for thrombi.1 Anticoagulation is underused, has poor patient compliance, and is often contraindicated, so that prevention of LAA thrombus without warfarin therapy is of great interest for patients without other options. The concept of LAA obliteration has received increased attention as evidence has accumulated for its use as a replacement for anticoagulation. Studies to answer this question are currently underway.98 Surgical Closure and Results Patients undergoing surgical LAA closure have not been systematically evaluated. Nevertheless, surgical closure is universal during mitral valve surgery, presumably to decrease the risk of embolic events.4,99-102 Several investigations have suggested that surgical obliteration may fail to completely close the LAA.103,104 One study revealed that incomplete surgical LAA ligation was frequent, because the investigators found patent flow between LAA and LA by TEE in 50 patients undergoing concurrent mitral valve surgery. In their study, 18 (36%) of 50 patients had incomplete LAA ligation, and the incidence was no different between patients studied immediately or later after surgery. In addition, LA size, degree of mitral regurgitation, operative approach (sternotomy or port access), and type of surgery (replacement or plasty) did not correlate with these results. The important fact is that SEC or thrombus was detected within the appendages in 9 (50%) of 18 patients with insufficient closure, and 4 (22%) of the 18 patients had thromboembolic events after the procedure. The authors suggested that residual communication between the insufficiently ligated appendage and the LA body might produce an enhanced prothrombotic environment because of stagnant LAA blood flow and be a potential source of increased embolic events.104 A recent study suggested that better results were obtained with a stapling device than with sutures during coronary artery bypass grafting (CABG) in 77 patients. Only 45% (5/11) of patients demonstrated complete occlusion with the use of sutures, compared with 72% (24/33) when the surgical stapler was used. The rate of LAA occlusion by individual surgeons increased from 43% (9/21) to 87% (20/23) after

871

872

Intracardiac Intervention performing at least four cases (P = .0001). Stapling can be safely undertaken during CABG, although two patients (2.6%) had perioperative thromboembolic events: one was an intraoperative stroke and the other a TIA occurring on the third postoperative day.105

from the SPAF trials (P = .15). The authors concluded the LAPTONI procedure was technically feasible without immediate neurologic morbidity or mortality.107 Percutaneous Transcatheter Occlusion: WATCHMAN, PLAATO, and the Amplatzer Septal Occluder

Thoracoscopic Extracardiac Closure WATCHMAN Ligation of the LAA with an automatic surgical stapler makes LAA obliteration possible in open-chest procedures.101 This was accomplished in combination with a technique dealing with the pericardium, such as pericardiectomy, pericardiocentesis, or resection of pericardial cysts. This method allowed LAA obliteration via thoracoscopy in an animal study. Thoracoscopy showed rapid (21.3 ± 7.6 minutes) and successful obliteration in all cases, but with bleeding, pneumothorax, and fibrinous pericarditis as complications.106 A human study entitled Thoracoscopic Left Appendage Total Obliteration No cardiac Invasion (LAPTONI) revealed that 14 of 15 patients had successful procedures, although 1 patient required urgent open thoracotomy because of bleeding. Patients had a history of prior thromboembolism and were observed for 8 to 60 months (mean, 42 ± 14 months). One fatal stroke occurred at 55 months, and one disabling stroke occurred 3 months after the procedure. Two non–procedure-related deaths were observed, one after CABG and the other from hepatic failure. The annual rate of stroke was 5.2% per year (95% confidential interval [CI]: 1.3 to 21), compared with 13% per year for similar, aspirin-treated patients

A

C

B

The WATCHMAN LAA closure system (Atritech Inc., Plymouth, MN) is a device that is designed to seal the LAA orifice and allow endothelialization (Figs. 49-7 and 49-8). A self-expanding nitinol frame contains a porous 160-µm polyethylene terephthalate (PET) membrane on its proximal face that filters LAA blood entering and leaving the appendage. Fixation barbs surround the midportion of the device to engage the LAA wall. It is available in diameters from 21 to 33 mm according to the diameter of the LAA orifice. PLAATO The PLAATO device (ev3, Inc., Plymouth, MN) was the first percutaneously implanted occluder to find human use (see Fig. 49-8). Its framework is a nitinol basket with a tissue anchoring system on the struts designed to maintain position. A minimally thrombogenic expanded polytetrafluoroethylene (ePTFE) membrane covers the basket and is designed to seal the LAA orifice and allow endothelialization.

Figure 49-7. A, Postmortem human heart reveals the threedimensional internal anatomy of the left atrial appendage (LAA) orifice (arrow). B, Photograph of the LAA after deployment of the WATCHMAN device into a canine heart. This image reveals no injury as seen from external observation. C, Internal view of canine LAA orifice, showing complete neointimal covering on the surface of this canine device (left panel). Formalinfixed LAA reveals device attachment to the LAA from a longitudinal section (right panel).

The Left Atrial Appendage: Anatomy, Physiology, and Therapeutic Percutaneous Closure

Figure 49-8. Examples of percutaneous left atrial appendage obliteration systems in development. A, WATCHMAN Filter System, showing nitinol frame and a permeable 160-µm membrane covering the proximal atrial facing surface of the device. Fixation barbs are located around the midperimeter of the device. B, PLAATO system, showing the self-expandable nitinol cage with nonthrombogenic expanded polytetrafluoroethylene membrane covering. Each strut has three fixation anchors for stabilization. C, The Amplatzer Septal Occluder, showing nitinol wire frame mesh with incorporated Dacron patches to enhance occlusion.

Amplatzer Septal Occluder The Amplatzer septal occluder (AGA Medical Corporation) has also been used for LAA closure (see Fig. 49-8), although it too appears to have been suspended in favor of newer device designs. This device is a double-disk with a short, broad waist and is designed for closing atrial septal defects.108

length. These echocardiographic tools allow visualization of complicated structures, especially LAA neck anatomy, number of lobes, the anatomic relationship between the pulmonary veins and the LAA, and device position including leakage around the device after implantation. Endocarditis antibiotic prophylaxis is sometimes given and continued until echocardiographic follow-up.

Implantation Techniques

Preclinical Studies

Most percutaneous LAA device implant procedures have been performed with an international normalized ratio (INR) of less than 2.0. Antiplatelet therapy with aspirin and clopidogrel is typically begun at least 1 day before procedure. Aspirin is generally continued indefinitely, and clopidogrel is used for several months to prevent thrombosis around the device. Aspirin and warfarin also are often used for several months after implantation. The implant procedure is performed using femoral vein access, so that general anesthesia is not needed. Conventional transseptal catheterization should be performed unless a patent foramen ovale or atrial septal defect is present. A heparin bolus is usually administered after transseptal puncture, and the activated clotting time (ACT) generally is maintained at 200 seconds or longer during the procedure. A contrast “appendogram” is performed in at least two views as a hand injection of contrast medium after the catheter is advanced. TEE or intracardiac echocardiography (ICE) is also recommended to assess the possibility of mobile thrombi and to size the LAA orifice diameter and

A preclinical study using the PLAATO device for percutaneous transcatheter LAA closure was performed in 25 dogs.9 The device was implanted, and LAA sealing was confirmed by ICE and contrast fluoroscopy. The device could be replaced or recaptured if it did not fit properly in the LAA. After the procedure, the LAA was examined from 2 days to 6 months later for healing, migration, perforation, and any thrombus, both grossly and histologically. Healing occurred by 1 month in 90% and was complete by 3 months. The atrial-facing surface was studied, and no mobile thrombi around the device were observed; neither were embolic events found in other organs. The investigators concluded that benign healing without new thrombus or damage was seen around the structures. Human Clinical Studies A clinical study using PLAATO was carried out in nonrheumatic AF patients who had contraindications to long-term anticoagulation therapy and were

873

874

Intracardiac Intervention at high risk for thromboembolism. Their risk factors included congestive heart failure, diabetes mellitus, hypertension, history of TIA or stroke, and SEC in the LAA by TEE.109 Fifteen patients were recruited, and the PLAATO device was implanted. All implants were successful, although one patient had hemopericardium associated with LAA access before implantation, diagnosed with intraprocedure TEE. Pericardiocentesis was performed without sequelae, and the LAA was successfully occluded 4 weeks later. One month follow-up was performed by TEE and fluoroscopy, and no adverse events observed. Conclusions were that this technology was appropriate for patients with AF who could not have long-term anticoagulation therapy. Contraindications to warfarin were defined as cerebral hemorrhage, gastrointestinal bleeding, an unstable INR, and severe chronic liver disease. Another study reported LAA occlusion using the Amplatzer device. Sixteen patients were enrolled, ranging from 58 to 63 years of age. All were diagnosed with AF, although eight patients were in sinus rhythm at the time of implantation. There was one technical failure (device embolization) requiring surgery. No other complications occurred during an overall follow-up of 5 patient-years. The study concluded that the Amplatzer device could be delivered percutaneously by venous puncture under local anesthesia without echocardiographic guidance.110 A recent investigation of LAA occlusion by PLAATO reported on mitral valve and left upper pulmonary vein (LUPV) function. Eleven patients (mean age, 72 ± 7 years) were enrolled in the study. There were no significant changes in LUPV diameter at baseline, 1 month, and 6 months (mean: 1.55, 1.61, and 1.54 cm, respectively; P = .13), nor in peak systolic flow velocity (mean: 0.38, 0.34, and 0.31 m/sec; P = .72) or peak diastolic flow velocity (mean: 0.39, 0.40, and 0.42 m/sec; P = .46). The devices remained stable at the deployment site, with minimal residual flow around them. The authors concluded that the PLAATO device could achieve a satisfactory seal of the LAA neck without significant effects on LA or LUPV structure or function.111 Ostermayer and colleagues10 evaluated the feasibility of LAA percutaneous occlusion in two previously described prospective, multicenter trials109,111 using the PLAATO system. They showed that LAA obliteration was successful in 111 patients (mean age, 71 ± 9 years).10 Indications for the PLAATO device were at least one stroke risk factor, such as history of TIA or stroke, presence of congestive heart failure, low LV ejection fraction, hypertension, diabetes mellitus, age greater than 65 years, coronary artery disease, moderate or dense SEC, or less than or equal to 20 cm/sec blood flow velocity within LAA. The patients had nonrheumatic AF of at least 3 months’ duration and a contraindication for anticoagulation therapy. Patients were administered aspirin (300325 mg) and clopidogrel (75 mg) twice daily, starting 48 hours before the procedure. Antibiotics were given 1 hour before the intervention. The exclusion criteria

were LA or LAA thrombus, complex aortic plaque, mitral or aortic stenosis or regurgitation, LA diameter greater than 6.5 cm, acute coronary syndrome, recent stroke (1.5) or to unsatisfactory relief of the valve obstruction. The low incidence of embolism during follow-up, the progressive decrease in intensity or disappearance of spontaneous echocardiographic contrast, and the improved left atrial function after PMC suggest a beneficial effect of the procedure on left atrial blood stasis, from which a lower risk of thromboembolism may be expected.42 Finally, there is no direct evidence that PMC reduces the incidence of atrial fibrillation, even though it has a favorable influence on the predictors of atrial fibrillation (e.g., atrial size, degree of obstruction).43 To summarize the current experience, midterm follow-up data obtained after PMC are comparable to

887

888

Intracardiac Intervention those from closed commissurotomy, which was to be expected from the similar mechanisms of these two techniques.

PARTICULAR APPLICATIONS OF PERCUTANEOUS MITRAL COMMISSUROTOMY PMC after Surgical Commissurotomy Several series have reported the results of PMC in patients with previous surgical commissurotomy.44,45 This category of patients is of interest, because in Western countries recurrent mitral stenosis is becoming more frequent than primary mitral stenosis. Reoperation in this context is associated with a higher risk of morbidity and mortality and requires valve replacement in most cases. All of the series reported to date show that PMC is feasible in this setting, although the procedure may be technically difficult in the case of “funnel-shaped” stenosis, which is frequent in these circumstances. PMC significantly improves valve function. The risks appear to be low, on a par with those of initial procedures. Midterm results are also satisfactory. As illustrated in our series of 232 patients who underwent PMC a mean of 16 years after surgical commissurotomy, the 8-year survival rate without intervention and without symptoms was 48% for the total series and 58% after good initial results.44 On the whole, the results are good, even if slightly less satisfactory than those obtained in patients without previous commissurotomy; this probably can be attributed to less favorable characteristics observed in patients previously subjected to operation. These encouraging preliminary data suggest that PMC may well postpone reoperation in selected patients with restenosis after commissurotomy. The indications for PMC in this subgroup of patients are similar to those for “primary PMC,” but echocardiographic examination must be conducted with great care to exclude any patients in whom restenosis is due mainly to valve rigidity without significant commissural refusion. The latter mechanism could be responsible for the exceptional cases of mitral stenosis that develop in patients who have undergone mitral ring annuloplasty for correction of mitral regurgitation. PMC in Patients with High Surgical Risk Valvuloplasty is the only solution when surgery is contraindicated. It is also preferable to surgery, at least as the first attempt, in patients with an increased risk for surgery of cardiac origin, as in the following situations. Preliminary reports have suggested that valvuloplasty can be performed safely and effectively in patients with severe pulmonary hypertension.46 These results are encouraging even though they concern a limited number of patients. In Western countries, many patients with mitral stenosis have concomitant noncardiac disease, which

may also increase the risk of surgery.38 Valvuloplasty can be performed as a life-saving procedure in critically ill patients,47 as the sole treatment when there is an absolute contraindication to surgery, or as a “bridge” to surgery in other cases. In this context, dramatic improvement has been observed in young patients; on other hand, the outcome is very bad in elderly patients presenting with “end-stage” disease, who should probably be better treated conservatively. In elderly patients, valvuloplasty results in moderate but significant improvement in valve function at an acceptable risk, although subsequent functional deterioration is frequent.29,38 Therefore, valvuloplasty is a valid, if only a palliative, treatment for these patients. During pregnancy, surgery carries a substantial risk of fetal mortality and morbidity, especially if extracorporeal circulation is required. The experience of PMC during pregnancy is still limited48,49 but suggests the following. From a technical point of view, during the last weeks of pregnancy (which was the time of PMC in most cases), the procedure may be more difficult because of the enlarged uterus. The Inoue technique seems to be particularly attractive in this setting, because the fluoroscopy time is reduced and the short inflation-deflation cycle probably reduces the hemodynamic compromise. The procedure is effective and results in normal delivery in most cases. Regarding radiation exposure, PMC is safe for the fetus, provided that protection is provided by a shield that completely surrounds the patient’s abdomen and the procedure is performed after the 20th week. In addition to radiation, PMC carries the potential risk of related hypotension and the always-present risk of complications that require urgent surgery. These preliminary data, which now represent several hundreds of cases, suggest that PMC can be a useful technique in the treatment of pregnant patients with mitral stenosis and refractory heart failure despite medical treatment. PMC and Left Atrial Thrombosis Left atrial thrombosis is generally considered a contraindication to PMC. However, a few limited series have shown that PMC using the Inoue balloon is feasible and is not a cause of systemic embolization.50 In cases of left atrial thrombosis, if the clinical condition of the patient requires urgent treatment, the limited number of patients in these series does not allow us to recommend PMC if the patient is a candidate for surgery.51,52 This recommendation is selfevident if the thrombus is free-floating or is situated in the left atrial cavity; it also applies when the thrombus is located on the interatrial septum. If the thrombus is located in the left atrial appendage, it has not been shown to our satisfaction that the Inoue technique under transesophageal guidance precludes a risk of embolism. If the patient is clinically stable, as is the case for most patients with mitral stenosis, anticoagulant therapy can be given for 2 to 6

Mitral Valvuloplasty months53; then, if a new transesophageal examination shows that the thrombus has disappeared, PMC can be attempted.

SELECTION OF PATIENTS The application of PMC depends on four major factors: the patient’s clinical condition; the valve anatomy; the experience of the medical and surgical teams of the institution concerned; and the financial aspect.51,52 Evaluation of Patient’s Clinical Condition Evaluation must take into account the degree of functional disability, the presence of contraindications to transseptal catheterization, and the alternative risk of surgery as a function of the underlying cardiac and noncardiac status. Because of the small but definite risk inherent in the technique, truly asymptomatic patients with severe mitral stenosis (i.e., patients with normal physical working capacity on exercise testing) usually are not candidates for PMC, except in cases of urgent need for extracardiac surgery, to allow pregnancy in young women, or in patients with an increased risk of embolism, such as those with a previous history of embolism, heavy spontaneous contrast in the left atrium, or recurrent atrial arrhythmias. Finally, PMC can be proposed in patients who declare to be asymptomatic but who have pulmonary hypertension either at rest (systolic pulmonary pressure >50 mm Hg) or on exercise (>60 mm Hg), the thresholds of which should be refined by the increasing experience gained in exercise echocardiography. Under these conditions, PMC should be performed only by experienced interventionists when the anatomy is suitable, leading to a safe, effective procedure. Contraindications to transseptal catheterization include suspected left atrial thrombosis (Video 50-6), severe hemorrhagic disorder, and severe cardiothoracic deformity. Increased surgical risk of cardiac origin (previous surgical commissurotomy or aortic valve replacement) or extracardiac origin (respiratory insufficiency, old age) makes balloon valvuloplasty preferable to surgery, at least as the first attempt, or even as the only solution in case of a strict contraindication to surgery. The coexistence of moderate aortic valve disease and severe mitral stenosis is another situation in which PMC is preferable to postpone the inevitable later surgical treatment of both valves. Valve Anatomy The assessment of anatomy has several aims when establishing indications and prognostic considerations. It is critical to ensure that there are no anatomic contraindications to the technique (Table 50-5). The first of these is the presence of left atrial thrombosis, which must be excluded by systematic performance of TEE a few days before the procedure.

Table 50-5. Contraindications to Mitral Valvuloplasty Left atrial thrombosis Mitral regurgitation >2/4 Massive or bicommissural calcification Severe aortic valve disease, or severe tricuspid stenosis + regurgitation, associated with mitral stenosis Severe concomitant coronary artery disease requiring bypass surgery

The second is mitral regurgitation greater than grade 2/4, which contraindicates valvuloplasty. Third, in cases of combined mitral stenosis and severe aortic disease, the indication for surgery is obvious in the absence of contraindications. Fourth, the presence of combined severe tricuspid stenosis and tricuspid regurgitation with clinical signs of heart failure is an indication for surgery on both valves. On the other hand, the existence of tricuspid regurgitation is not a contraindication to the procedure even though it represents a negative prognostic factor.54 Our view on the performance of PMC in patients with only mild mitral stenosis (valve area >1.5 cm2) is that the risks probably outweigh the benefits, and these patients are usually well managed by medical treatment.51,52 For prognostic considerations, echocardiographic assessment allows the classification of patients into anatomic groups with a view to predicting the results. Most investigators use the Wilkins score (Table 50-6),28 whereas others, such as Cormier and colleagues,2 use a more general assessment of valve anatomy (Table 50-7). Controversy exists regarding the most effective echocardiography scoring system in the prediction of results of mitral valvuloplasty. In

Table 50-6. Anatomic Classification of the Mitral Valve (Massachusetts General Hospital, Boston) Leaflet Mobility Highly mobile valve with restriction of only the leaflet tips Midportion and base of leaflets have reduced mobility Valve leaflets move forward during diastole, mainly at the base No or minimal forward movement of the leaflets during diastole Valvular Thickening Leaflets near normal (4-5 mm) Midleaflet thickening, marked thickening of the margins Thickening extends through the entire leaflets (5-8 mm) Marked thickening of all leaflet tissue (>8-10 mm) Subvalvular Thickening Minimal thickening of chordal structures just below the valve Thickening of chordae extending up to one third of chordal length Thickening extending to the distal third of the chordae Extensive thickening and shortening of all chordae extending down to the papillary muscle Valvular Calcification Single area of increased echocardiographic brightness Scattered areas of brightness confined to leaflet margins Brightness extending into the midportion of leaflets Extensive brightness through most of the leaflet tissue Adapted from Abascal V, Wilkins GT, O’Shea JP, et al: Prediction of successful outcome in 130 patients undergoing percutaneous balloon mitral valvotomy. Circulation 1990;82:448-456.

889

890

Intracardiac Intervention Table 50-7. Anatomic Classification of the Mitral Valve (Bichat Hospital, Paris) Echocardiographic Group 1 2 3

Mitral Valve Anatomy Pliable noncalcified anterior mitral leaflet and mild subvalvular disease (i.e., thin chordae ≥10 mm long) Pliable noncalcified anterior mitral leaflet and severe subvalvular disease (i.e., thickened chordae 20 mm in length), primarily because

Medina 0,1,0 Duke type B Safian type IIIB Lefevre type 4a

Medina 0,0,1 Duke type E Safian type IV Lefevre type IVB

of the more extensive plaque burden in long lesions. Stents improve late outcome compared with balloon angioplasty, but stent and lesion length remain the most important predictors of restenosis in the stent era.19 Coronary stents have been used to treat suboptimal angiographic results (“spot stenting”) and dissections after balloon angioplasty of longer lesions, although the “full metal jacket” stent approach to diffuse disease is associated with a higher recurrence rate in the absence of complete stent expansion, particularly in smaller vessels. Overlapping sirolimus-eluting stents provide safe and effective treatment for long coronary lesions.20 Bifurcation Lesions The risk of side branch occlusion in bifurcation lesions relates to the extent of atherosclerotic involvement of the side branch within its origin from the parent vessel, which ranges from 14% to 27% in side branches with ostial involvement. To accurately assess the risk of side branch occlusion and avoid conflicting definitions of side branch and ostial stenosis, a number of classification systems for bifurcation stenoses have been proposed (Fig. 60-1).21-24 One stent is preferable to stents in both the parent vessel and the side branch, because subacute thrombosis and restenosis remain higher in bifurcation disease treated with coronary stents in both branches.25 If two stents are planned for the parent vessel and side branch, a number of stenting techniques are

1075

1076

Evaluation of Interventional Techniques possible, including simultaneous kissing stents, crush, coulotte, and T stenting. To date, the optimal technique has not been identified, although the use of DESs appears to reduce restenosis compared with bare metal stents. The origin of the side branch is the most common location of failure (recurrence) after bifurcation stenting.26 A number of dedicated bifurcation stents have been developed to provide adequate vessel coverage27 and side branch access28 during stent deployment. Common to all of these strategies is a final “kissing” balloon inflation in the parent vessel and side branch.29 Total Occlusion Total coronary occlusion is identified as an abrupt termination of the epicardial vessel; anterograde and retrograde collaterals may be present and are helpful in quantifying the length of the totally occluded segment. Coronary occlusions are common findings30 and often lead to the decision to perform coronary bypass surgery rather than PCI in the setting of multivessel disease.31,32 The success rate for recanalization depends on the occlusion duration and on certain lesion morphologic features, such as bridging collaterals, occlusion length greater than 15 mm, and absence of a “nipple” to guide wire advancement. Although newer technologies and techniques have been used to recanalize refractory occlusions,33,34 better guide wires and wire techniques have accounted for much of the improvement in crossing success over recent years.35 Simultaneous coronary injections are sometimes useful for identifying the length of the total occlusion (Fig. 60-2). Once the occlusion has been crossed, coronary stents, including DES,36,37 have been used to provide the best long-term outcomes. A key component to the assessment of total occlusion is definition of the collateral grades that provide blood flow to the jeopardized myocardium.38 The Rentrop classification system includes Rentrop grade 0 (no filling), Rentrop grade 1 (small side branches filled), Rentrop grade 2 (partial epicardial filling of the occluded artery), and Rentrop grade 3 (complete epicardial filling of the occluded artery). Anatomic collaterals summarized by the 26 potential pathways were consolidated into four groups: septal, intra-arterial (bridging), epicardial with proximal takeoff (atrial branches), and epicardial with distal takeoff.39 Finally, the size of the collateral connection can be quantified as group 0 (no continuous connection between donor and recipient artery), group 1 (continuous threadlike connection ≤0.3 mm), or group 2 (continuous small, branch-like collateral through its course ≥0.4 mm).39 Angiographic Complications After Percutaneous Coronary Intervention Although the frequency of angiographic complications during PCI has been reduced substantially with

the use of coronary stents, untoward effects resulting from disruption of the atherosclerotic plaque and embolization of atherosclerotic debris, thrombus, and vasoactive mediators still occurs during 5% to 10% of PCI procedures.

Coronary Dissection Plaque fracture is an integral component of balloon angioplasty, although significant vessel wall disruption resulting in reduced anterograde flow and lumen compromise is a relatively uncommon occurrence (50% when a 20 min) New ischemic ECG changes suggestive of acute ischemia Typical rise and fall in cardiac biomarkers Evidence of recent thrombus within the stent determined at autopsy or via examination of tissue retrieved after thrombectomy

Probable Any unexplained death within the first 30 days Irrespective of the time after the index procedure, any MI that is related to documented acute ischemia in the territory of the implanted stent without angiographic confirmation of stent thrombosis and in the absence of any other cause. Possible

Any unexplained death >30 days after intracoronary stenting

ECG, electrocardiographic; MI, myocardial infarction; TIMI, Thrombolysis in Myocardial Infarction.

document stent thrombosis in clinical studies. Timing of stent thrombosis is defined as acute (30 µg LCA 24-36 µg RCA

Dobutamine IV

20-40 µg/kg/min

Nitroprusside IC

0.3-0.9 µg/kg

Plateau (sec)

Half-life (min)

30-60

2

60-120

1-2

5-10

0.5-1

60-120

3-5

20

1

Side Effects Transient QT prolongation Torsades de pointes Decreased blood pressure (10%-15%), chest burning Transient AV block when injected into the dominant artery Tachycardia, increase in blood pressure Decreased blood pressure (20%)

AV, atrioventricular; IC, intracoronary; IV, intravenous; LCA, left coronary artery; RCA, right coronary artery.

Comments

Avoid in patients with history of bronchospasm Must repeat with escalating doses to ensure that maximal hyperemia is reached

1099

1100

Evaluation of Interventional Techniques QT prolongation and ventricular tachycardia or fibrillation.

is safely obtained with IC ATP doses greater than 15 µg.

Adenosine

Dobutamine Hyperemia

Both IC and IV adenosine have the advantage of a short half-life. The total duration of the hyperemic response to IC adenosine is only 25% that seen with papaverine or dipyridamole. Adenosine is benign in the appropriate dosages (18 to 24 µg in the RCA or 24 to 48 µg in the LCA, or infused IV at 140 µg/kg/ minute), although many have reported safety at much higher dosages. As described earlier, transient atrioventricular block and bradycardia may occur. IV administration tends to have a higher incidence of flushing, chest tightness, and atrioventricular block than IC dosing. Jeremias and colleagues9 examined differences in FFR between IC adenosine (15 to 20 µg in the RCA or 18 to 24 µg in the LCA) and IV adenosine (140 µg/ kg/minute) in 52 patients with 60 lesions. Mean percent stenosis was 56% ± 24% (range, 0% to 95%). The mean FFR was 0.78 ± 0.15, with a range of 0.41 to 0.98. There was a strong and linear relationship between IC and IV adenosine (r = .978 and P < .001). The mean measurement difference for FFR was −0.004 ± 0.03. A small random scatter in both directions of FFR was noted in 8.3% of stenoses, where the IC adenosine FFR value was 0.05 greater than the IV adenosine FFR value, suggesting a suboptimal IC hyperemic response. Changes in heart rate and blood pressure were significantly greater with IV adenosine. Two patients with IV, but none with IC, adenosine had side effects of bronchospasm and nausea. These data indicated that IC adenosine is equivalent to IV infusion for determination of FFR in most patients. However, in a small percentage of cases, coronary hyperemia was suspected to be suboptimal with IC adenosine, suggesting that a repeated, higher IC adenosine dose may be helpful.

Bartunek and associates10 examined FFR in response to IC adenosine and IV dobutamine (10 to 40 µg/kg/ minute) in 22 patients with single-vessel CAD. Peak dobutamine infusion produced similar distal coronary pressures and pressure ratios (Pd/Pa, 60 ± 18 and 59 ± 18 mm Hg; FFR, 0.68 ± 0.18 and 0.68 ± 0.17, respectively; all P = not significant [NS]). An additional bolus of IC adenosine given at peak dobutamine in nine patients failed to change the FFR. By angiography, high-dose IV dobutamine did not modify the area of the epicardial stenosis and, much like adenosine, fully exhausted myocardial resistance regardless of inducible left ventricular dysfunction. Sodium Nitroprusside IC nitroprusside may be an alternative to IC adenosine. Parham and associates11 examined coronary blood flow velocity, heart rate, and blood pressure in unobstructed LADs in 21 patients at rest, after IC adenosine (boluses of 30 to 50 µg), and after three serial doses of IC nitroprusside (boluses of 0.3, 0.6, and 0.9 µg/kg). IC nitroprusside produced equivalent coronary hyperemia with a longer duration (about 25%) compared with IC adenosine. IC nitroprusside (0.9 µg/kg) decreased systolic blood pressure by 20% with minimal change in heart rate, whereas IC adenosine had no effect on these parameters. FFR measurements with IC nitroprusside were identical to those obtained with IC adenosine (r = .97). IC nitroprusside, in doses commonly used for the treatment of the no-reflow phenomenon, can produce sustained coronary hyperemia without detrimental systemic hemodynamics. Sodium nitroprusside also appears to be a suitable hyperemic stimulus for coronary physiologic measurements.

Intracoronary Adenosine Triphosphate Although it is unavailable in the United States, adenosine triphosphate (ATP) may also be used to stimulate maximal hyperemia. Coronary flow velocity, hemodynamics, electrocardiography, and myocardial lactate metabolism before and after administration of 50 µg of IC ATP and 10 mg of papaverine into the LCA were examined in 18 patients with normal coronary arteries. Dose responses were obtained with IC ATP doses of 0.5, 5, 15, 30, and 50 µg and compared to the papaverine response in an additional seven patients. ATP did not produce significant hemodynamic or electrocardiographic changes. CFR was similar between ATP and papaverine. All patients showed lactate production after papaverine; only three patients showed lactate production after ATP (P < .001). There was a significant correlation between CFR with ATP and with papaverine, indicating that maximal coronary vasodilation

CLINICAL VALIDATION OF INTRACORONARY PRESSURE MEASUREMENTS A summary of physiologic threshold values for common clinical applications is provided in Table 61-2. In order to define the threshold of FFRmyo below which inducible ischemia is present, Pijls12 and De Bruyne13 and their colleagues conducted independent but parallel and complementary investigations. Pijls’ group studied 60 patients accepted for a singlevessel PTCA who had a positive exercise test in the preceding 24 hours. FFRmyo was measured before and 15 minutes after PTCA, and the exercise test was repeated after 1 week. If the second exercise test had reverted to normal after PTCA, FFRmyo values were associated with inducible ischemia. All except two FFRmyo measurements greater than 0.74 were not associated with ischemia, and all FFRmyo measure-

Intracoronary Pressure and Flow Measurements Table 61-2. Physiologic Criteria Associated with Clinical Applications Indication Ischemia detection Deferred angioplasty End point of angioplasty End point of stenting

CFR

rCFR

FFR

2.0 >2.0-2.5* —

0.90 >0.90

*With a 70% angiographic diameter stenosis), and vein grafts.132 However, controversial results were also reported in some IVUS-guided stent trials,145,146 presumably due

1127

1128

Evaluation of Interventional Techniques to differing procedural end points for IVUS-guided stenting, as well as various adjunctive treatment strategies that were used in these trials in response to suboptimal results (Table 62-1). Overall, a meta-analysis of 9 clinical studies (2972 patients) demonstrated that IVUS-guided stenting significantly lowers 6month angiographic restenosis (odds ratio [OR] = 0.75, 95% confidence interval [CI], 0.60 to 0.94; P = .01) and target vessel revascularizations (OR = 0.62; 95% CI, 0.49 to 0.78; P = .00003), with a neutral effect on death and nonfatal myocardial infarction, compared to an angiographic optimization.147 Insights into Long-Term Outcomes In-stent restenosis is primarily caused by intimal proliferation rather than chronic stent recoil.112,148

Growth of neointima is usually greatest in the areas with the largest plaque burden,64,149,150 and the intimal growth process seems to be more aggressive in diabetic patients.151 In the treatment of in-stent restenosis, IVUS can be helpful to differentiate pure intimal ingrowth from poor stent expansion, especially if ablative therapies are being considered. Using serial IVUS immediately before and after balloon angioplasty for in-stent restenosis, Castagna and colleagues152 showed, in 1090 consecutive in-stent restenosis lesions, that 38% of lesions had an MSA of less than 6.0 mm2. Stent underexpansion can result in clinically significant lumen compromise even with minimal neointimal hyperplasia. For this type of instent restenosis, mechanical optimization is appropriate in most cases. IVUS can also track the response to treatment, with evidence that angioplasty of in-stent restenosis

Table 62-1. IVUS Versus Angiographic Guidance of Bare Metal Stent Implantation Study* (Ref. No.)

N

Population

Study Design

Albiero et al. (128)

312

De novo, native

Multicenter, registry

Blasini et al. (129)

212

Single center, registry

Choi et al. (130)

278

De novo and restenotic, native and SVG De novo, native

Gaster et al. (131)

108

De novo and restenotic, native

Single center, randomized

AVID (132)

759

CRUISE (133)

499

OPTICUS (145)

550

De novo, native and SVG De novo and restenotic, native De novo and restenotic, native De novo and restenotic, native De novo, native De novo and restenotic, native Long lesions (>20 mm)

PRESTO (146)

9070

RESIST (134, 135)

155

SIPS (136)

269

TULIP (137)

144

IVUS Criteria for Optimal Expansion

Criteria Fulfilled

End Points

Results

NA

6-mo angiography

IVUS better (early phase)

50%

6-mo angiography

IVUS better

NA

Acute closure, 6-mo MACE

IVUS better

64%

IVUS better

NA

6-mo angiography, CFR, FFR, TVR; 2.5-yr MACE 12-mo TLR

—

9-mo TVR

Complete apposition, no ref disease, MSA ≥60% of average ref VA (early phase) or MSA ≥ distal ref LA (late phase) Complete apposition, no residual dissection, MSA >8 mm2 and/or 90% of average ref LA Complete apposition, no residual dissection, MSA ≥80% of distal ref LA MUSIC criteria

Multicenter, randomized

MUSIC criteria

Multicenter, nonrandomized

Discretion of individual institutional practice

Multicenter, randomized

MUSIC criteria

Multicenter, nonrandomized

Discretion of individual institutional practice

Multicenter, randomized

MSA >80% of average ref LA

80%

Single center, randomized

MLA >65% of average ref LA

69%

Single center, randomized

Complete apposition, MLD ≥80% of average ref diameter, MSA ≥ distal ref LA

89%

Single center, registry

56% —

IVUS better (subset analysis) IVUS better

6-mo angiography, 12-mo MACE 9-mo MACE

No difference

6-mo angiography, 18-mo MACE 6-mo angiography, 2-yr TLR 6-mo angiography, 12-mo MACE

IVUS better (nonsignificant reduction) IVUS better (2-year TLR)

No difference

IVUS better

*AVID, Angiography versus IVUS-Directed stent placement trial; CRUISE, Can Routine Ultrasound Influence Stent Expansion study; MUSIC, Multicenter Ultrasound guided Stent Implantation in Coronaries; OPTICUS, Optimization with ICUS to Reduce Stent Restenosis; PRESTO, Results of Prevention of REStenosis with Tranilast and its Outcomes Trial; RESIST, REStenosis after Intravascular ultrasound STenting; SIPS, Strategy for IVUS-Guided PTCA and Stenting; TULIP, Thrombocyte activity evaluation and effects of Ultrasound guidance in Long Intracoronary Stent Placement. CFR, coronary flow reserve; FFR, fractional flow reserve; IVUS, intravascular ultrasound; LA, lumen area; MACE, major adverse cardiac events; MLD, minimum lumen diameter; MSA, minimum stent area; ref, reference vessel; SVG, saphenous vein graft; TLR, target lesion revascularization; TVR, target vessel revascularization; VA, vessel area.

Intravascular Ultrasound

Figure 62-12. Arrows indicate unhealed dissection (A) and echolucent neointimal tissue, so-called acoustic hole or “black hole” (B), observed at follow-up after intracoronary radiation therapy.

is followed by early lumen loss due to decompression153 and/or reintrusion154 of tissue immediately after intervention. This phenomenon was more prominent in longer lesions and in those with greater in-stent tissue burden, which may partially account for the worse long-term outcomes in diffuse versus focal in-stent restenosis. Direct tissue removal, rather than tissue compression/extrusion through the stent struts, may help minimize early lumen loss due to this phenomenon. Several investigators have reported a considerable reduction in angiographic and/or clinical recurrence of in-stent restenosis in patients with diffuse in-stent restenosis treated with ablative therapies (DCA, rotational atherectomy, or laser angioplasty) compared with PTCA alone.155-161 Intracoronary Radiation Therapy Insights into Mechanism of Action ICRT, the first biologic treatment targeting excessive proliferative response of vascular smooth muscle cells to mechanical intervention, has undergone extensive clinical testing with IVUS characterization. Regardless of radiation source or delivery platform, most IVUS studies of ICRT in PTCA, stenting, and treatment of in-stent restenosis showed significant in-lesion or in-stent efficacy.162-171 Serial IVUS investigations confirmed that these beneficial effects are primarily derived from decreased neointimal hyperplasia, but there is also an effect from accelerated positive remodeling at the irradiated segment.115,162 Guidance of Procedures IVUS revealed that a combination of increased neointimal hyperplasia and either absence of positive remodeling or negative remodeling accounts for the unfavorable edge effect of ICRT.172-174 Because, in catheter-based ICRT (not radioactive stents), most edge effects are related to inadequate coverage of injured edge segments (so-called geographic miss),175

IVUS guidance may greatly enhance the safety and efficacy of this highly geographically specific technique. A combined ICRT/imaging device was developed to facilitate the fine-tuning of catheter positioning.176 This device also forms an asymmetric dose distribution to compensate for the eccentricity of the target plaque as well as the off-center position of the catheter in response to on-line IVUS guidance. Other investigator groups have developed detailed dosimetric analysis algorithms, based on dosevolume histograms derived from three-dimensional IVUS,177-183 for optimal dose prescription. Insights into Long-Term Outcomes Unusual IVUS observations related to ICRT include unhealed dissections, late-acquired incomplete stent apposition, and acoustic holes (or “black holes”)— that is, echolucent neointimal tissue. Delayed healing is commonly seen after brachytherapy in general. A substudy from the Beta Energy Restenosis Trial (BERT) reported that 8 (50%) of 16 dissections identified immediately after PTCA were still present at 6-month follow-up.184 Similarly, an IVUS analysis from the Stents and Radiation Therapy (START) 40/20 trial showed that 43% of dissections had partially healed and 7% were unchanged over time (Fig. 62-12A).185 Of note, previous nonradiation trials indicated that dissections seen after intervention normally heal within a 6-month follow-up period.186-188 Lateacquired incomplete stent apposition was reported with both catheter-based ICRT (beta and gamma radiation) and radioactive stents (Fig. 62-13). Detailed serial IVUS examination revealed that this phenomenon is a result of excessive positive remodeling of the underlying vessel wall combined with significant neointimal inhibition.189,190 Okura and colleagues also reported that late-acquired incomplete stent apposition appears to occur in segments with relatively little peri-stent plaque burden but high radiation dose exposure.190 Late-acquired incomplete stent apposition can also be seen after mechanical vessel injury in nonradiation interventions (e.g.,

1129

1130

Evaluation of Interventional Techniques

Figure 62-13. Classification of incomplete stent apposition (ISA). Baseline ISA can either be resolved (resolved ISA) or remain (persistent ISA) at follow-up. Late-acquired ISA without vessel expansion is typically seen in thrombuscontaining lesions, whereas late-acquired ISA with focal, positive vessel remodeling is more characteristic with brachytherapy and drug-eluting stents.

atherectomy plus stenting),191 although delayed endothelialization by ICRT may lead to different clinical implications. The acoustic hole or “black hole” consists of echolucent neointimal tissue and has been seen in all types of ICRT (see Fig. 62-12B).192 Post-ICRT specimens retrieved with DCA show large myxoid areas, with interspersed smooth muscle cells, scattered in a proteoglycan-containing extracellular matrix—findings that are compatible with a weak backscattering of echoes and a dark appearance on IVUS. These IVUS findings after ICRT are unique, but their exact clinical implications remain unknown. Drug-Eluting Stents

frequency distribution, with variable degrees of the tail ends. Because restenosis corresponds to the right tail end of the distribution curve, a discrepancy between mean neointimal volume and binary or clinical restenosis can occur in DES trials.194,198,199 Similarly, bare metal stents show a wide individual variation in geographic distribution of neointima along the stented segment,200 whereas some types of DES demonstrate predilection of in-stent neointimal hyperplasia for specific locations (e.g., proximal stent edge).201 In serial IVUS studies with multiple longterm follow-ups, neointima within nonrestenotic bare metal stents showed mild regression after 6 months.202 In contrast, both sirolimus- and paclitaxel-eluting stents showed a slight but continuous increase in neointimal hyperplasia for up to 4 years.203-206

Insights into Mechanism of Action IVUS observations from the clinical experience with antiproliferative drug-eluting stents (DES) have shown a striking inhibition of in-stent neointimal hyperplasia, whereas the mechanical performances of these new stents are similar to those of conventional bare metal stents.193-197 Additionally, both statistical and geographic distributions of neointimal hyperplasia can be significantly different between the biologic (DES) and mechanical (bare metal) stents. In general, neointimal volume (as a percentage of stent volume) within bare metal stents follows a near-Gaussian or normal frequency distribution, with a mean value of 30% to 35%. The standard deviation of this statistical distribution represents biologic variability in vascular response to acute and chronic vessel injury by interventions. In contrast, biologic modifications by DES often result in a non-Gaussian

Guidance of Procedures In DES, the fact that the drugs dramatically reduce the variability of the biologic response (neointimal proliferation) further strengthens the prognostic value of the MSA as a powerful predictor for in-stent restenosis.207-211 This was well illustrated in recent IVUS work by Sonoda and colleagues, in which sirolimus-eluting stents showed a stronger positive relation, with a greater correlation coefficient between baseline MSA and 8-month MLA, compared to control bare metal stents (0.8 versus 0.65 and 0.92 versus 0.59, respectively).207 The utility of IVUS to ensure adequate stent expansion cannot be overemphasized, particularly if there are clinical risk factors for DES failure (e.g., diabetes, renal failure).

Intravascular Ultrasound

Figure 62-14. Proximal disease development 8 months after drug-eluting stent implantation. In this example, the new stenosis at the proximal stent margin is primarily caused by plaque proliferation, despite minimal neointimal hyperplasia observed inside the stent. Baseline intravascular ultrasound (IVUS) reveals a significant residual plaque at the corresponding uncovered segment.

In this context, preinterventional IVUS can also provide useful information about plaque composition. In particular, calcified plaque is important to identify, because the presence, degree, and location of calcium within the target vessel can substantially affect the delivery and subsequent deployment of coronary stents (see Fig. 62-5).212,213 One important advantage of online IVUS guidance is the ability to assess the extent and distance from the lumen of calcium deposits within a plaque. For example, lesions with extensive superficial calcium may require rotational atherectomy before stenting.213,214 Conversely, apparently significant calcification on fluoroscopy may subsequently be found by IVUS to be distributed in a deep portion of the vessel wall or to have a lower degree of calcification (calcium arc 50% diameter stenosis) at 8 months had greater reference plaque burden (61%

versus 49%; P = .03) (Fig. 62-14) and a higher overexpansion index (maximum stent area/reference MLA, 1.8 versus 1.5; P = .03) at baseline, compared to those without edge stenosis.218 More recently, the Stent Deployment Techniques on Clinical Outcomes of Patients Treated with the Cypher Stent (STLLR) trial also demonstrated that geographic miss (defined as the length of injured or stenotic segment not fully covered by DES) had a significant negative impact on both clinical efficacy (target vessel and lesion revascularization) and safety (myocardial infarction) at 1 year after sirolimuseluting stent implantation.219 These findings suggest that less aggressive stent dilation and complete coverage of reference disease may be beneficial, as long as significant underexpansion and incomplete strut apposition are avoided. On-line IVUS guidance can facilitate both the determination of appropriate stent size and length and the achievement of optimal procedural end points, with the goal being to cover significant pathology with reasonable stent expansion while anchoring the stent ends in relatively plaquefree vessel segments. Insights into Long-Term Outcomes Because of the low incidence of DES failure, clarification of its exact mechanisms awaits the cumulative analysis of large clinical studies. Nevertheless, suboptimal deployment or mechanical problems appear to contribute to the development of both restenosis and thrombosis. Particularly, the most common mechanism is stent underexpansion, the incidence of which has been reported as 60% to 80% in DES failures.210,211,220 In a study of 670 native coronary lesions treated with sirolimus-eluting stents, the only independent predictors of angiographic restenosis were postprocedural final MSA and IVUS-measured stent length (OR = 0.586 and 1.029, respectively).208 Recurrent restenosis after DES implantation for bare metal

1131

1132

Evaluation of Interventional Techniques stent restenosis was also recently investigated using IVUS. In a series of 48 in-stent restenosis lesions treated with sirolimus-eluting stents, 82% of recurrent lesions had an MSA of less than 5.0 mm2, compared with only 26% of nonrecurrent lesions (P = .003).220 In addition, a gap between sirolimus-eluting stents was identified in 27% of recurrent lesions versus 5% of nonrecurrent lesions. These observations emphasize the importance of procedural optimization at DES implantation for both de novo and in-stent restenosis lesions. Although published data on DES thrombosis are further limited, one single-center IVUS study reported stent underexpansion (P = .03) and a significant residual reference segment stenosis (P = .02) as independent multivariate predictors of sirolimus-eluting stent thrombosis (median time, 14 days after implantation).221 For very late DES thrombosis (>12 months), another investigator group also suggested smaller stent expansion and incomplete stent apposition as possible risk factors.222 Late-acquired incomplete stent apposition with DES has been reported in both experimental (paclitaxel)223 and clinical (sirolimus and paclitaxel) studies (see Fig. 62-13).193,196,224,225 Several IVUS studies indicated that the main mechanism is focal, positive vessel remodeling, as in the case of brachytherapy.193,224 In addition, there is a strong suggestion that incompletely apposed struts are seen primarily in eccentric plaques, and that the gaps develop mainly on the disease-free side of the vessel wall. The combination of mechanical vessel injury during stent implantation and biologic vessel injury with pharmacologic agents or polymer in the setting of little underlying plaque may predispose the vessel wall to chronic, pathologic dilation. At present, however, no data directly link this finding with subsequent unfavorable clinical events such as late stent thrombosis.226 Nevertheless, the potential impact of this finding on long-term outcomes needs to be carefully assessed over an extended period. Other IVUS-detected conditions that may be of importance in DES include non-uniform stent strut distribution and strut fractures after implantation (Fig. 62-15).227-230 Theoretically, both abnormalities can reduce the local drug dose delivered to the arterial wall, as well as mechanical scaffolding of the affected lesion segment. A recent IVUS study of 24 sirolimus-eluting stent restenoses identified the number of visualized struts (normalized for the number of stent cells) and the maximum interstrut angle as independent multivariate IVUS predictors of both neointimal hyperplasia and MLA.228 In contrast, the exact incidence and clinical implications of strut fractures remain to be investigated.231,232

SAFETY As with other interventional procedures, the possibility of spasm, dissection, and thrombosis exists when intravascular imaging catheters are used. In a retrospective study of 2207 patients, Hausmann

and colleagues identified spasm in 2.9% of patients, and other complications, including dissection, thrombosis, and abrupt closure with “certain relation” to IVUS, in 0.4%.233 This study was performed with first-generation catheters in the early 1990s, and it is likely (although not documented) that the incidence of spasm and other complications is substantially lower with the current generation of catheters.

COSTS In the United States, current retail sales prices for the stand-alone single-use imaging catheters range between $600 and $850. The retail price of the IVUS imaging consoles is between $150,000 and $250,000, although the actual prices paid by hospitals vary widely depending on “bundling” deals and other special purchase arrangements. The U.S. Health Care Financing Administration approved reimbursement for the IVUS procedure and interpretation for Medicare patients in 1997, based on the number of vessels imaged. A number of other carriers have also approved reimbursement for IVUS, although payment is on a region-by-region basis. In most places, the total reimbursement is less than the cost of the catheters.

ONGOING AND ANTICIPATED TECHNICAL DEVELOPMENTS Three-Dimensional Transducer Tracking Compared with external imaging methods, one technical disadvantage common to catheter-based imaging modalities is difficulty in obtaining accurate spatial orientation of the investigated vessel segment. To overcome this limitation, three-dimensional transducer navigation or tracking techniques have been introduced by several investigator groups. In IVUS, this can be accomplished by reconstructing the three-dimensional pull back trajectory, using biplane x-ray recordings of the transducer,234 or by real-time tracking of a miniaturized electromagnetic position sensor mounted in the catheter tip.235 Computational algorithms and instrumentation have been developed to reduce significant artifacts induced by respiratory and cardiac motions. The geometrically correct three-dimensional IVUS reconstruction of vessel wall structure may not only facilitate routine percutaneous interventions in complex coronary anatomy, but also allow detailed in vivo profiling of intracoronary hemodynamics and endothelial shear stress.236,237 A number of pathologic and experimental studies have shown that inhomogeneities and irregularities of these factors play an important role in the initiation, localization, growth, composition, remodeling, and destabilization of atheromatous plaque.238 A recent human in vivo study with the spatially correct three-dimensional IVUS partly confirmed these observations in a clinical setting, directly relating local endothelial

Intravascular Ultrasound

Figure 62-15. Stent strut discontinuity (fracture) observed 8 months after deployment of three overlapping drug-eluting stents. On the cross-sectional intravascular ultrasound (IVUS) image (bottom, middle), an abnormal paucity of stent struts, not seen at implantation, is detected at a portion of the mid stent. The longitudinal IVUS image (top) shows an acute-angled bend at the corresponding segment. In this particular case, the strut discontinuity is not associated with increased intimal hyperplasia, at least at this time point.

shear stress at baseline to subsequent plaque progression and arterial remodeling at 6 months.237 Although the predictive value of this parameter in the context of vulnerable plaque requires further investigation, this technique, particularly if combined with advanced catheter-based diagnostic modalities, may offer further insight into the natural history of native coronary artery disease. In addition, vascular response to stent implantation is another potentially important field to investigate with this technology,239-241 because anatomically complex lesions, such as bifurcation and long lesions, are being more actively treated with DES that can significantly alter the vascular geometry and, consequently, local endothelial shear stress. Tissue Characterization Another intriguing area of current IVUS development is the attempt to identify tissue components using computer-assisted analysis of raw radiofrequency signals in the reflected ultrasound beam. This is pri-

marily based on the fact that there is greater information contained in the backscattered ultrasound signal than is revealed by the conventional amplitude-based image presentation alone. To date, a variety of signal parameters and mathematical modeling techniques have been proposed and have been shown to enhance the tissue discrimination. One investigator group demonstrated that integrated backscatter values, calculated as the average power of the backscattered ultrasound signal from a sample tissue volume, were significantly different among tissue types (calcification, fibrous tissue, mixed lesion, lipid core or intimal hyperplasia, and thrombus).242,243 Other investigators, including our own group, used unique ultrasound wave properties from different tissue types (e. g., signal attenuation slopes, statistical frequency distribution, angle-dependent echo-intensity variation).244-248 Wavelet analysis, a mathematical model for assessing local wave patterns within a complex signal, has also been proposed and has demonstrated accurate in vitro and in vivo discrimination of lipidrich from fibrous plaques.249

1133

1134

Evaluation of Interventional Techniques To date, one system (Virtual Histology, Volcano Therapeutics) has been commercialized in the United States. It uses spectral radiofrequency analyses with a classification algorithm developed from ex vivo coronary data sets.250 The pattern recognition algorithm generates color-mapped images of the vessel wall, with a distinct color for each of the fibrous, necrotic, calcific, and fibrofatty categories. Combined use with automated pull back and border detection techniques enables quantitative assessment of each tissue category over a three-dimensional volume of a coronary artery. An initial clinical study showed significant correlation between IVUS-determined plaque compositions and corresponding histopathology of the coronary specimens obtained by directional atherectomy.251 In another series of nonculprit, nonobstructive, de novo lesions, IVUS-derived thin-cap fibroatheroma (defined as focal, necrotic core–rich plaque without evident overlying fibrous tissue) was a more prevalent finding in acute coronary syndrome than in stable angina.252 Current technical limitations include limited spatial resolution (100 to 250 µm); no classifications for thrombus, blood, or intimal hyperplasia; and potential errors due to poor ultrasound penetration through extensive calcification. Several multicenter studies have been initiated worldwide, and the role of this system in the detection of rupture-prone plaques is yet to be established. Mechanical Strain Assessment IVUS elastography, which measures the mechanical strain property of the arterial wall, is an extension of radiofrequency signal analysis. Local tissue deformation is determined with the use of cross-correlation analysis of radiofrequency signals recorded at two different intravascular pressures during near enddiastole.253,254 The calculated local radial strain is displayed on the IVUS image, color coded on the plaque area (elastography) or the luminal boundary (palpography), on the IVUS image. An initial in vitro validation study with diseased human coronary and femoral arteries revealed different mean strain values among fibrous, fibrofatty, and fatty plaque components.253 A subsequent animal study using peripheral arteries demonstrated a high sensitivity to identify fatty plaque material from its increased mean strain value.255 These results were confirmed by a postmortem human coronary study that demonstrated 88% sensitivity and 89% specificity to detect histologic vulnerable plaque, defined as a high strain region at the plaque surface with adjacent low strain regions in elastography.256 In this study, the strain in caps also showed a high positive correlation with the amount of macrophages (P < .006), as well as an inverse relation with the amount of smooth muscle cells (P < .0001). One hypothesis for these observations is that active macrophage infiltration can cause local weakening of a fibrous cap, because macrophages play a key role in degradation of the extracellular matrix by secreting proteolytic enzymes.

Accordingly, a recent clinical study using threedimensional IVUS palpography showed that the number of highly deformable plaques correlated with both unstable clinical presentation and levels of C-reactive protein.257 The prognostic value of this modality is currently being investigated in prospective multicenter studies. Contrast Intravascular Ultrasound for Neovascular and Molecular Imaging The application of contrast ultrasound technology for catheter-based coronary imaging may add another dimension to the arena of atherosclerotic plaque assessment. For the past several decades, mounting evidence has linked neovascularization within the arterial wall (proliferation of the vasa vasorum) to the development of atherosclerosis as well as plaque vulnerability due to intraplaque hemorrhage and inflammation.258-261 Recent animal studies have also shown that suppression of plaque neovascularization with angiogenesis inhibitors or statins can significantly attenuate the progression of atherosclerosis.259,262 These pioneering works, however, were reported using histopathology, in vitro microcomputed tomography, or preclinical targeted-magnetic resonance imaging. Contrast-enhanced IVUS with microbubble agents may offer, for the first time, in vivo quantification of vasa vasorum density and plaque perfusion in a clinical setting. A preliminary human study using conventional IVUS systems demonstrated the feasibility to track the distribution and dynamic changes in the echo density within the coronary plaque and adventitial region after intracoronary injection of a clinically available microbubble agent.263 In this series, one case also showed persistent plaque enhancement (an IVUS “blush sign”), suggesting pathologic extravasation of microbubbles due to increased porosity of the neoangiogenic vessels. Although this analysis currently relies on the processing of gray-scale, phase-correlated IVUS images, use of raw radiofrequency signals may further improve the detection and quantification of subtle changes in the backscattered ultrasound intensity within the arterial wall. The utility of contrast IVUS imaging can be significantly enhanced by the introduction of contrast agents targeted to specific tissue components, cells, molecules, or biologic processes. This biomedical imaging technique, so-called molecular imaging, is also potentially applicable to site-targeted delivery of therapeutic agents with visual verification and quantification of the treatment. To date, several industrial and university-based research groups are developing molecular imaging agents for ultrasound applications. In addition to the traditional microbubblebased technology, nongaseous acoustically reflective nanoparticles have shown significant potentials,264 owing to their smaller particle sizes and greater in vivo stability.265 Whereas passive targeting uses primarily nonspecific uptake of the acoustic particles by

Intravascular Ultrasound

Figure 62-16. Conceptual diagrams of an active site-targeted ultrasound contrast agent. A variety of ligands conjugated to the particle surface can be used to provide site-specific ultrasound enhancement, offering the ability to detect and quantify the various cell-surface molecular signatures (e.g., intercellular adhesion molecule [ICAM]) of endothelium and atheroma components.

macrophages, active targeting agents provide sitespecific enhancement using a variety of ligands conjugated to the particle surface (Fig. 62-16). In a study using biotinylated, lipid-coated, perfluorocarbon emulsion nanoparticles, Lanza and colleagues first demonstrated marked acoustic enhancement of in vivo thrombi in a canine model.266 In a more recent animal study, the nanoemulsion was targeted to tissue factor (a 43-kDa transmembrane glycoprotein responsible for initiating coagulation cascade), confirming that the acoustic nanoparticles can infiltrate into arterial walls after balloon injury and detect the localized expression of tissue factor with conventional IVUS.267 Several other investigator groups have also developed echogenic liposomes conjugated with anti-intercellular adhesion molecule-1 (ICAM-1), anti-vascular cell adhesion molecule-1 (VCAM-1), anti-fibrin, anti-fibrinogen, and anti-tissue factor (TF) and have proven the in vivo feasibility of IVUS to detect and quantify the various cell-surface molecular signatures of endothelium and atheroma components.265,268 Furthermore, extended efforts are ongoing to integrate therapeutic agents, such as antiproliferative drugs and gene transfection agents, into these ligand-targeted ultrasound contrast systems.269

ACKNOWLEDGMENTS We appreciate Heidi N. Bonneau, RN, MS, for her expert review and editing advice, as well as Ichizo Tsujino, MD, PhD, and Atsushi Hirohata, MD, for their professional work with the figures. REFERENCES 1. Bom N, Lancee CT, Van Egmond FC: An ultrasonic intracardiac scanner. Ultrasonics 1972;10:72-76. 2. Bom N, ten Hoff H, Lancee CT, et al: Early and recent intraluminal ultrasound devices. Int J Card Imaging 1989;4: 79-88. 3. Yock PG, Johnson EL, Linker DT: Intravascular ultrasound: Development and clinical potential. Am J Cardiac Imag 1988;2:185-193. 4. Mallery JA, Tobis JM, Griffith J, et al: Assessment of normal and atherosclerotic arterial wall thickness with an intravascular ultrasound imaging catheter. Am Heart J 1990;119: 1392-1400. 5. Gussenhoven EJ, Essed CE, Lancee CT, et al: Arterial wall characteristics determined by intravascular ultrasound imaging: An in vitro study. J Am Coll Cardiol 1989;14:947952. 6. Pandian NG, Kreis A, Brockway B, et al: Ultrasound angioscopy: Real-time, two-dimensional, intraluminal ultrasound imaging of blood vessels. Am J Cardiol 1988;62:493-494.

1135

1136

Evaluation of Interventional Techniques 7. Potkin BN, Bartorelli AL, Gessert JM, et al: Coronary artery imaging with intravascular high-frequency ultrasound. Circulation 1990;81:1575-1585. 8. Yock P, Linker D, Saether O, et al: Intravascular two-dimensional catheter ultrasound: Initial clinical studies [Abstract]. Circulation 1988;78:II-21. 9. Marco J, Fajadet J, Robert G, et al: Intracoronary ultrasound imaging: Initial clinical trials [Abstract]. Circulation 1989;80: II-374. 10. Graham SP, Brands D, Sheehan H, et al: Assessment of arterial wall morphology using intravascular ultrasound in vitro and in patients [Abstract]. Circulation 1989;80:II-565. 11. Fuessl RT, Mintz GS, Pichard AD, et al: In vivo validation of intravascular ultrasound length measurements using a motorized transducer pullback system. Am J Cardiol 1996;77:1115-1118. 12. Tobis JM, Mallery J, Mahon D, et al: Intravascular ultrasound imaging of human coronary arteries in vivo: Analysis of tissue characterizations with comparison to in vitro histological specimens. Circulation 1991;83:913-926. 13. Siegel RJ, Chae JS, Maurer G, et al: Histopathologic correlation of the three-layered intravascular ultrasound appearance of normal adult human muscular arteries. Am Heart J 1993;126:872-878. 14. Nishimura RA, Edwards WD, Warnes CA, et al: Intravascular ultrasound imaging: In vitro validation and pathologic correlation. J Am Coll Cardiol 1990;16:145-154. 15. Meyer CR, Chiang EH, Fechner KP, et al: Feasibility of highresolution, intravascular ultrasonic imaging catheters. Radiology 1988;168:113-116. 16. Yock PG, Linker DT, Angelsen BA: Two-dimensional intravascular ultrasound: Technical development and initial clinical experience. J Am Soc Echocardiogr 1989;2:296-304. 17. Fitzgerald PJ, St Goar F, Connolly AJ, et al: Intravascular ultrasound imaging of coronary arteries: Is three layers the norm? Circulation 1992;86:154-158. 18. Maheswaran B, Leung CY, Gutfinger DE, et al: Intravascular ultrasound appearance of normal and mildly diseased coronary arteries: Correlation with histologic specimens. Am Heart J 1995;130:976-986. 19. Gussenhoven EJ, Frietman PA, The SH, et al: Assessment of medial thinning in atherosclerosis by intravascular ultrasound. Am J Cardiol 1991;68:1625-1632. 20. Isner JM, Donaldson RF, Fortin AH, et al: Attenuation of the media of coronary arteries in advanced atherosclerosis. Am J Cardiol 1986;58:937-939. 21. Mintz GS, Nissen SE, Anderson WD, et al: American College of Cardiology Clinical Expert Consensus Document on Standards for Acquisition, Measurement and Reporting of Intravascular Ultrasound Studies (IVUS): A report of the American College of Cardiology Task Force on Clinical Expert Consensus Documents. J Am Coll Cardiol 2001;37:1478-1492. 22. Rasheed Q, Dhawale PJ, Anderson J, et al: Intracoronary ultrasound-defined plaque composition: Computer-aided plaque characterization and correlation with histologic samples obtained during directional coronary atherectomy. Am Heart J 1995;129:631-637. 23. Di Mario C, The SH, Madretsma S, et al: Detection and characterization of vascular lesions by intravascular ultrasound: An in vitro study correlated with histology. J Am Soc Echocardiogr 1992;5:135-146. 24. Mintz GS, Popma JJ, Pichard AD, et al: Patterns of calcification in coronary artery disease: A statistical analysis of intravascular ultrasound and coronary angiography in 1155 lesions. Circulation 1995;91:1959-1965. 25. Mintz GS, Douek P, Pichard AD, et al: Target lesion calcification in coronary artery disease: An intravascular ultrasound study. J Am Coll Cardiol 1992;20:1149-1155. 26. Tuzcu EM, Berkalp B, De Franco A, et al: The dilemma of diagnosing coronary calcification: Angiography versus intravascular ultrasound. J Am Coll Cardiol 1996;27:832-838. 27. Yock PG, Fitzgerald PJ, Linker DT, et al: Intravascular ultrasound guidance for catheter-based coronary interventions. J Am Coll Cardiol 1991;17:39B-45B. 28. Pandian NG, Kreis A, Brockway B: Detection of intraarterial thrombus by intravascular high frequency two-dimensional

29.

30.

31.

32.

33.

34. 35. 36.

37.

38.

39.

40.

41.

42. 43. 44. 45.

46.

47.

48.

ultrasound imaging in vitro and in vivo studies. Am J Cardiol 1990;65:1280-1283. Nailon WH, McLaughlin S, Spencer T, et al: Can statistical texture analysis of unprocessed intravascular ultrasound (IVUS) signal discriminate red and white thrombi and plasma? [Abstract]. Circulation 1996;94:I-612. Fitzgerald PJ, Connolly AJ, Watkins RD, et al: Distinction between soft plaque and thrombus by intravascular tissue characterization [Abstract]. J Am Coll Cardiol 1991;17: 111A. Metz JA, Preuss P, Komiyama N, et al: Discrimination between soft plaque and thrombus based on radiofrequency analysis of intravascular ultrasound [Abstract]. J Am Coll Cardiol 1996;27:200A. Komiyama N, Chronos NA, Uren NG, et al: The progression of thrombus in an ex-vivo shunt model evaluated by intravascular ultrasound radiofrequency analysis. Ultrasound Med Biol 1999;25:561-566. Ramo MP, Spencer T, Kearney PP, et al: Characterisation of red and white thrombus by intravascular ultrasound using radiofrequency and videodensitometric data-based texture analysis. Ultrasound Med Biol 1997;23:1195-1199. Fitzgerald PJ, Yock C, Yock PG: Orientation of intracoronary ultrasonography: Looking beyond the artery. J Am Soc Echocardiogr 1998;11:13-19. Ge J, Liu F, Gorge G, et al: Angiographically “silent” plaque in the left main coronary artery detected by intravascular ultrasound. Coron Artery Dis 1995;6:805-810. Hausmann D, Johnson JA, Sudhir K, et al: Angiographically silent atherosclerosis detected by intravascular ultrasound in patients with familial hypercholesterolemia and familial combined hyperlipidemia: Correlation with high density lipoproteins. J Am Coll Cardiol 1996;27:1562-1570. Kimura BJ, Russo RJ, Bhargava V, et al: Atheroma morphology and distribution in proximal left anterior descending coronary artery: In vivo observations. J Am Coll Cardiol 1996;27:825-831. Mintz GS, Popma JJ, Pichard AD, et al: Limitations of angiography in the assessment of plaque distribution in coronary artery disease: A systematic study of target lesion eccentricity in 1446 lesions. Circulation 1996;93:924-931. Sheikh KH, Harrison JK, Harding MB, et al: Detection of angiographically silent coronary atherosclerosis by intracoronary ultrasonography. Am Heart J 1991;121: 1803-1807. St. Goar F, Pinto FJ, Alderman EL, et al: Intravascular ultrasound imaging of angiographically normal coronary arteries: An in vivo comparison with quantitative angiography. J Am Coll Cardiol 1991;18:952-958. Mintz GS, Painter JA, Pichard AD, et al: Atherosclerosis in angiographically “normal” coronary artery reference segments: An intravascular ultrasound study with clinical correlations. J Am Coll Cardiol 1995;25:1479-1485. Fitzgerald PJ, Yock PG: Mechanisms and outcomes of angioplasty and atherectomy assessed by intravascular ultrasound imaging. J Clin Ultrasound 1993;21:579-588. Glagov S, Weisenberg E, Zarins CK, et al: Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 1987;316:1371-1375. Hermiller JB, Tenaglia AN, Kisslo KB, et al: In vivo validation of compensatory enlargement of atherosclerotic coronary arteries. Am J Cardiol 1993;71:665-668. Schoenhagen P, Ziada KM, Vince DG, et al: Arterial remodeling and coronary artery disease: The concept of “dilated” versus “obstructive” coronary atherosclerosis. J Am Coll Cardiol 2001;38:297-306. Nishioka T, Luo H, Berglund H, et al: Absence of focal compensatory enlargement or constriction in diseased human coronary saphenous vein bypass grafts: An intravascular ultrasound study. Circulation 1996;93:683-690. Nishioka T, Luo H, Eigler NL, et al: Contribution of inadequate compensatory enlargement to development of human coronary artery stenosis: An in vivo intravascular ultrasound study. J Am Coll Cardiol 1996;27:1571-1576. Mintz GS, Kent KM, Pichard AD, et al: Contribution of inadequate arterial remodeling to the development of focal

Intravascular Ultrasound

49.

50.

51.

52.

53. 54. 55.

56.

57.

58.

59.

60.

61.

62.

63. 64.

65. 66.

67.

coronary artery stenoses: An intravascular ultrasound study. Circulation 1997;95:1791-1798. Pasterkamp G, Wensing PJ, Post MJ, et al: Paradoxical arterial wall shrinkage may contribute to luminal narrowing of human atherosclerotic femoral arteries. Circulation 1995;91:1444-1449. Pasterkamp G, Borst C, Post MJ, et al: Atherosclerotic arterial remodeling in the superficial femoral artery: Individual variation in local compensatory enlargement response. Circulation 1996;93:1818-1825. Hibi K, Ward MR, Honda Y, et al: Impact of different definitions on the interpretation of coronary remodeling determined by intravascular ultrasound. Catheter Cardiovasc Interv 2005;65:233-239. Schoenhagen P, Ziada KM, Kapadia SR, et al: Extent and direction of arterial remodeling in stable versus unstable coronary syndromes: An intravascular ultrasound study. Circulation 2000;101:598-603. Smits PC, Pasterkamp G, Quarles van Ufford MA, et al: Coronary artery disease: Arterial remodelling and clinical presentation. Heart 1999;82:461-464. Kaji S, Akasaka T, Hozumi T, et al: Compensatory enlargement of the coronary artery in acute myocardial infarction. Am J Cardiol 2000;85:1139-1141, A1139. Jeremias A, Spies C, Herity NA, et al: Coronary artery compliance and adaptive vessel remodelling in patients with stable and unstable coronary artery disease. Heart 2000;84:314319. Gyongyosi M, Yang P, Hassan A, et al: Arterial remodelling of native human coronary arteries in patients with unstable angina pectoris: A prospective intravascular ultrasound study. Heart 1999;82:68-74. Nakamura M, Nishikawa H, Mukai S, et al: Impact of coronary artery remodeling on clinical presentation of coronary artery disease: An intravascular ultrasound study. J Am Coll Cardiol 2001;37:63-69. Pasterkamp G, Schoneveld AH, van der Wal AC, et al: Relation of arterial geometry to luminal narrowing and histologic markers for plaque vulnerability: The remodeling paradox. J Am Coll Cardiol 1998;32:655-662. Mehran R, Dangas G, Mintz GS, et al: Atherosclerotic plaque burden and CK-MB enzyme elevation after coronary interventions: Intravascular ultrasound study of 2256 patients. Circulation 2000;101:604-610. Dangas G, Mintz GS, Mehran R, et al: Preintervention arterial remodeling as an independent predictor of target-lesion revascularization after nonstent coronary intervention: An analysis of 777 lesions with intravascular ultrasound imaging. Circulation 1999;99:3149-3154. Okura H, Hayase M, Shimodozono S, et al: Impact of pre-interventional arterial remodeling on subsequent vessel behavior after balloon angioplasty: A serial intravascular ultrasound study. J Am Coll Cardiol 2001;38:20012005. Okura H, Morino Y, Oshima A, et al: Preintervention arterial remodeling affects clinical outcome following stenting: An intravascular ultrasound study. J Am Coll Cardiol 2001;37:1031-1035. Endo A, Hirayama H, Yoshida O, et al: Arterial remodeling influences the development of intimal hyperplasia after stent implantation. J Am Coll Cardiol 2001;37:70-75. Shiran A, Weissman NJ, Leiboff B, et al: Effect of preintervention plaque burden on subsequent intimal hyperplasia in stented coronary artery lesions. Am J Cardiol 2000;86: 1318-1321. Hoffmann R, Mintz GS, Kent KM, et al: Serial intravascular ultrasound predictors of restenosis at the margins of PalmazSchatz stents. Am J Cardiol 1997;79:951-953. Hoffmann R, Mintz GS, Mehran R, et al: Intravascular ultrasound predictors of angiographic restenosis in lesions treated with Palmaz-Schatz stents. J Am Coll Cardiol 1998;31: 43-49. Dangas G, Mintz GS, Mehran R, et al: Stent implantation neutralizes the impact of preintervention arterial remodeling on subsequent target lesion revascularization. Am J Cardiol 2000;86:452-455.

68. Hong MK, Park SW, Lee CW, et al: Preintervention arterial remodeling as a predictor of intimal hyperplasia after intracoronary stenting: A serial intravascular ultrasound study. Clin Cardiol 2002;25:11-15. 69. Mintz GS, Kimura T, Nobuyoshi M, et al: Relation between preintervention remodeling and late arterial responses to coronary angioplasty or atherectomy. Am J Cardiol 2001;87: 392-396. 70. Abizaid A, Mintz GS, Pichard AD, et al: Clinical, intravascular ultrasound, and quantitative angiographic determinants of the coronary flow reserve before and after percutaneous transluminal coronary angioplasty. Am J Cardiol 1998;82: 423-428. 71. Takagi A, Tsurumi Y, Ishii Y, et al: Clinical potential of intravascular ultrasound for physiological assessment of coronary stenosis: relationship between quantitative ultrasound tomography and pressure-derived fractional flow reserve. Circulation 1999;100:250-255. 72. Briguori C, Anzuini A, Airoldi F, et al: Intravascular ultrasound criteria for the assessment of the functional significance of intermediate coronary artery stenoses and comparison with fractional flow reserve. Am J Cardiol 2001; 87:136-141. 73. Nishioka T, Amanullah AM, Luo H, et al: Clinical validation of intravascular ultrasound imaging for assessment of coronary stenosis severity: Comparison with stress myocardial perfusion imaging. J Am Coll Cardiol 1999;33:1870-1878. 74. Jasti V, Ivan E, Yalamanchili V, et al: Correlations between fractional flow reserve and intravascular ultrasound in patients with an ambiguous left main coronary artery stenosis. Circulation 2004;110:2831-2836. 75. Mintz GS, Pichard AD, Kovach JA, et al: Impact of preintervention intravascular ultrasound imaging on transcatheter treatment strategies in coronary artery disease. Am J Cardiol 1994;73:423-430. 76. Gorge G, Ge J, Erbel R: Role of intravascular ultrasound in the evaluation of mechanisms of coronary interventions and restenosis. Am J Cardiol 1998;81:91G-95G. 77. Tobis JM, Mallery JA, Gessert J, et al: Intravascular ultrasound cross-sectional arterial imaging before and after balloon angioplasty in vitro. Circulation 1989;80:873-882. 78. Baptista J, Umans VA, Di Mario C, et al: Mechanisms of luminal enlargement and quantification of vessel wall trauma following balloon coronary angioplasty and directional atherectomy. Eur Heart J 1995;16:1603-1612. 79. Baptista J, Di Mario C, Ozaki Y, et al: Impact of plaque morphology and composition on the mechanisms of lumen enlargement using intracoronary ultrasound and quantitative angiography after balloon angioplasty. Am J Cardiol 1996;77:115-121. 80. Braden GA, Herrington DM, Downes TR, et al: Qualitative and quantitative contrasts in the mechanisms of lumen enlargement by coronary balloon angioplasty and directional coronary atherectomy. J Am Coll Cardiol 1994;23:40-48. 81. Honye J, Mahon DJ, Jain A, et al: Morphological effects of coronary balloon angioplasty in vivo assessed by intravascular ultrasound imaging. Circulation 1992;85:1012-1025. 82. Buller CE, Davidson CJ, Virmani R, et al: Real-time assessment of experimental arterial angioplasty with transvenous intravascular ultrasound. J Am Coll Cardiol 1992;19: 217-222. 83. Athanasiadis A, Haase KK, Wullen B, et al: Lesion morphology assessed by pre-interventional intravascular ultrasound does not predict the incidence of severe coronary artery dissections. Eur Heart J 1998;19:870-878. 84. Fitzgerald PJ, Ports TA, Yock PG: Contribution of localized calcium deposits to dissection after angioplasty: An observational study using intravascular ultrasound. Circulation 1992;86:64-70. 85. Lee RT, Richardson SG, Loree HM, et al: Prediction of mechanical properties of human atherosclerotic tissue by high-frequency intravascular ultrasound imaging: An in vitro study. Arterioscler Thromb 1992;12:1-5. 86. Richardson PD, Davies MJ, Born GV: Influence of plaque configuration and stress distribution on fissuring of coronary atherosclerotic plaques. Lancet 1989;2:941-944.

1137

1138

Evaluation of Interventional Techniques 87. Shimodozono S, Okura H, Hayase M, et al: Influence of calcium on coronary dissection following cutting balloon angioplasty: An intravascular ultrasound study. J Am Coll Cardiol 2000;35:A18. 88. Stone GW, Hodgson JM, St Goar FG, et al: Improved procedural results of coronary angioplasty with intravascular ultrasound-guided balloon sizing: The CLOUT pilot trial. Circulation 1997;95:2044-2052. 89. Stone GW, Frey A, Linnemeier T, et al: 2.5 Year follow-up of the CLOUT study: Long-term implications for an aggressive IVUS guided balloon angioplasty strategy. J Am Coll Cardiol 1999;33:81A. 90. Schroeder S, Baumbach A, Haase KK, et al: Reduction of restenosis by vessel size adapted percutaneous transluminal coronary angioplasty using intravascular ultrasound. Am J Cardiol 1999;83:875-879. 91. Abizaid A, Pichard AD, Mintz GS, et al: Acute and long-term results of an intravascular ultrasound-guided percutaneous transluminal coronary angioplasty/provisional stent implantation strategy. Am J Cardiol 1999;84:1298-1303. 92. Peters RJ, Kok WE, Di Mario C, et al: Prediction of restenosis after coronary balloon angioplasty: Results of PICTURE (PostIntraCoronary Treatment Ultrasound Result Evaluation), a prospective multicenter intracoronary ultrasound imaging study. Circulation 1997;95:2254-2261. 93. Mintz GS, Popma JJ, Pichard AD, et al: Intravascular ultrasound predictors of restenosis after percutaneous transcatheter coronary revascularization. J Am Coll Cardiol 1996;27:1678-1687. 94. The GUIDE trial investigators: IVUS-determined predictors of restenosis in PTCA and DCA: Final report from the GUIDE trial, phase II [Abstract]. J Am Coll Cardiol 1996;27: 156A. 95. Yock PG, Fitzgerald PJ, Sykes C, et al: Morphologic features of successful coronary atherectomy determined by intravascular ultrasound imaging [Abstract]. Circulation 1990;82: III-676. 96. Safian RD, Gelbfish JS, Erny RE, et al: Coronary atherectomy: Clinical, angiographic, and histological findings and observations regarding potential mechanisms. Circulation 1990; 82:69-79. 97. Sharaf BL, Williams DO: “Dotter effect” contributes to angiographic improvement following directional coronary atherectomy [Abstract]. Circulation 1990;82:III-310. 98. Nakamura S, Mahon DJ, Leung CY, et al: Intracoronary ultrasound imaging before and after directional coronary atherectomy: in vitro and clinical observations. Am Heart J 1995;129:841-851. 99. Simonton CA, Leon MB, Baim DS, et al: “Optimal” directional coronary atherectomy: final results of the Optimal Atherectomy Restenosis Study (OARS). Circulation 1998;97: 332-339. 100. Matar FA, Mintz GS, Pinnow E, et al: Multivariate predictors of intravascular ultrasound end points after directional coronary atherectomy. J Am Coll Cardiol 1995;25:318-324. 101. Popma JJ, Mintz GS, Satler LF, et al: Clinical and angiographic outcome after directional coronary atherectomy: A qualitative and quantitative analysis using coronary arteriography and intravascular ultrasound. Am J Cardiol 1993;72: 55E-64E. 102. Umans VA, Baptista J, di Mario C, et al: Angiographic, ultrasonic, and angioscopic assessment of the coronary artery wall and lumen area configuration after directional atherectomy: The mechanism revisited. Am Heart J 1995;130:217-227. 103. Dussaillant GR, Mintz GS, Pichard AD, et al: Effect of rotational atherectomy in noncalcified atherosclerotic plaque: A volumetric intravascular ultrasound study. J Am Coll Cardiol 1996;28:856-860. 104. Kovach JA, Mintz GS, Pichard AD, et al: Sequential intravascular ultrasound characterization of the mechanisms of rotational atherectomy and adjunct balloon angioplasty. J Am Coll Cardiol 1993;22:1024-1032. 105. Mintz GS, Potkin BN, Keren G, et al: Intravascular ultrasound evaluation of the effect of rotational atherectomy in obstructive atherosclerotic coronary artery disease. Circulation 1992;86:1383-1393.

106. Kimura BJ, Fitzgerald PJ, Sudhir K, et al: Guidance of directed coronary atherectomy by intracoronary ultrasound imaging. Am Heart J 1992;124:1365-1369. 107. Topol EJ, Leya F, Pinkerton CA, et al: A comparison of directional atherectomy with coronary angioplasty in patients with coronary artery disease. The CAVEAT Study Group. N Engl J Med 1993;329:221-227. 108. Suzuki T, Hosokawa H, Katoh O, et al: Effects of adjunctive balloon angioplasty after intravascular ultrasound-guided optimal directional coronary atherectomy: The result of Adjunctive Balloon Angioplasty After Coronary Atherectomy Study (ABACAS). J Am Coll Cardiol 1999;34:1028-1035. 109. Tsuchikane E, Sumitsuji S, Awata N, et al: Final results of the STent versus directional coronary Atherectomy Randomized Trial (START). J Am Coll Cardiol 1999;34: 1050-1057. 110. Honda Y, Yock PG, Fitzgerald PJ: Impact of residual plaque burden on clinical outcomes of coronary interventions. Catheter Cardiovasc Interv 1999;46:265-276. 111. Mintz GS, Popma JJ, Pichard AD, et al: Arterial remodeling after coronary angioplasty: A serial intravascular ultrasound study. Circulation 1996;94:35-43. 112. Mintz GS, Popma JJ, Hong MK, et al: Intravascular ultrasound to discern device-specific effects and mechanisms of restenosis. Am J Cardiol 1996;78:18-22. 113. Lansky AJ, Mintz GS, Popma JJ, et al: Remodeling after directional coronary atherectomy (with and without adjunct percutaneous transluminal coronary angioplasty): A serial angiographic and intravascular ultrasound analysis from the Optimal Atherectomy Restenosis Study. J Am Coll Cardiol 1998;32:329-337. 114. Kimura T, Kaburagi S, Tamura T, et al: Remodeling of human coronary arteries undergoing coronary angioplasty or atherectomy. Circulation 1997;96:475-483. 115. Sabate M, Serruys PW, van der Giessen WJ, et al: Geometric vascular remodeling after balloon angioplasty and betaradiation therapy: A three-dimensional intravascular ultrasound study. Circulation 1999;100:1182-1188. 116. Honda Y, Yock CA, Hermiller JB, et al: Longitudinal redistribution of plaque is an important mechanism for lumen expansion in stenting [Abstract]. J Am Coll Cardiol 1997;29: 281A. 117. Ahmed JM, Mintz GS, Weissman NJ, et al: Mechanism of lumen enlargement during intracoronary stent implantation: An intravascular ultrasound study. Circulation 2000; 102:7-10. 118. Maehara A, Takagi A, Okura H, et al: Longitudinal plaque redistribution during stent expansion. Am J Cardiol 2000;86: 1069-1072. 119. von Birgelen C, Mintz GS, Eggebrecht H, et al: Preintervention arterial remodeling affects vessel stretch and plaque extrusion during coronary stent deployment as demonstrated by three-dimensional intravascular ultrasound. Am J Cardiol 2003;92:130-135. 120. Mintz GS, Pichard AD, Kent KM, et al: Axial plaque redistribution as a mechanism of percutaneous transluminal coronary angioplasty. Am J Cardiol 1996;77:427-430. 121. Nakamura S, Colombo A, Gaglione A, et al: Intracoronary ultrasound observations during stent implantation. Circulation 1994;89:2026-2034. 122. Colombo A, Hall P, Nakamura S, et al: Intracoronary stenting without anticoagulation accomplished with intravascular ultrasound guidance. Circulation 1995;91:1676-1688. 123. Goldberg SL, Colombo A, Nakamura S, et al: Benefit of intracoronary ultrasound in the deployment of Palmaz-Schatz stents. J Am Coll Cardiol 1994;24:996-1003. 124. Metz JA, Mooney MR, Walter PD, et al: Significance of edge tears in coronary stenting: Initial observations from the STRUT registry [Abstract]. Circulation 1995;92:I-546. 125. Schwarzacher SP, Metz JA, Yock PG, et al: Vessel tearing at the edge of intracoronary stents detected with intravascular ultrasound imaging. Cath Cardiovasc Diagn 1997;40: 152-155. 126. Hong MK, Park SW, Lee NH, et al: Long-term outcomes of minor dissection at the edge of stents detected with intravascular ultrasound. Am J Cardiol 2000;86:791-795, A799.

Intravascular Ultrasound 127. Nishida T, Colombo A, Briguori C, et al: Outcome of nonobstructive residual dissections detected by intravascular ultrasound following percutaneous coronary intervention. Am J Cardiol 2002;89:1257-1262. 128. Albiero R, Rau T, Schluter M, et al: Comparison of immediate and intermediate-term results of intravascular ultrasound versus angiography-guided Palmaz-Schatz stent implantation in matched lesions. Circulation 1997;96:2997-3005. 129. Blasini R, Neumann FJ, Schmitt C, et al: Restenosis rate after intravascular ultrasound-guided coronary stent implantation. Cathet Cardiovasc Diagn 1998;44:380-386. 130. Choi JW, Goodreau LM, Davidson CJ: Resource utilization and clinical outcomes of coronary stenting: A comparison of intravascular ultrasound and angiographical guided stent implantation. Am Heart J 2001;142:112-118. 131. Gaster AL, Slothuus Skjoldborg U, Larsen J, et al: Continued improvement of clinical outcome and cost effectiveness following intravascular ultrasound guided PCI: Insights from a prospective, randomised study. Heart 2003;89:1043-1049. 132. Russo RJ, Attubato MJ, Davidson CJ, et al: Angiography Versus Intravascular ultrasound-Directed stent placement: Final results from AVID [Abstract]. Circulation 1999;100: I-234. 133. Fitzgerald PJ, Oshima A, Hayase M, et al: Final results of the Can Routine Ultrasound Influence Stent Expansion (CRUISE) study. Circulation 2000;102:523-530. 134. Schiele F, Meneveau N, Seronde MF, et al: Medical costs of intravascular ultrasound optimization of stent deployment. Results of the multicenter randomized “REStenosis after Intravascular ultrasound STenting” (RESIST) study. Int J Cardiovasc Intervent 2000;3:207-213. 135. Schiele F, Meneveau N, Vuillemenot A, et al: Impact of intravascular ultrasound guidance in stent deployment on 6month restenosis rate: A multicenter, randomized study comparing two strategies—with and without intravascular ultrasound guidance. RESIST Study Group. REStenosis after Ivus guided STenting. J Am Coll Cardiol 1998;32: 320-328. 136. Frey AW, Hodgson JM, Muller C, et al: Ultrasound-guided strategy for provisional stenting with focal balloon combination catheter: Results from the randomized Strategy for Intracoronary Ultrasound-guided PTCA and Stenting (SIPS) trial. Circulation 2000;102:2497-2502. 137. Oemrawsingh PV, Mintz GS, Schalij MJ, et al: Intravascular ultrasound guidance improves angiographic and clinical outcome of stent implantation for long coronary artery stenoses: Final results of a randomized comparison with angiographic guidance (TULIP Study). Circulation 2003;107: 62-67. 138. de Jaegere P, Mudra H, Figulla H, et al: Intravascular ultrasound-guided optimized stent deployment: Immediate and 6 months clinical and angiographic results from the Multicenter Ultrasound Stenting in Coronaries Study (MUSIC Study). Eur Heart J 1998;19:1214-1223. 139. Uren NG, Schwarzacher SP, Metz JA, et al: Predictors and outcomes on stent thrombosis: An intravascular ultrasound registry. Eur Heart J 2002;23:124-132. 140. Cheneau E, Leborgne L, Mintz GS, et al: Predictors of subacute stent thrombosis: Results of a systematic intravascular ultrasound study. Circulation 2003;108:43-47. 141. Ziada KM, Tuzcu EM, De Franco AC, et al: Absolute, not relative, post-stent lumen area is a better predictor of clinical outcome [Abstract]. Circulation 1996;94:I-453. 142. Moussa I, Moses J, Di Mario C, et al: Does the specific intravascular ultrasound criterion used to optimize stent expansion have an impact on the probability of stent restenosis? Am J Cardiol 1999;83:1012-1017. 143. Kasaoka S, Tobis JM, Akiyama T, et al: Angiographic and intravascular ultrasound predictors of in-stent restenosis. J Am Coll Cardiol 1998;32:1630-1635. 144. Morino Y, Honda Y, Okura H, et al: An optimal diagnostic threshold for minimal stent area to predict target lesion revascularization following stent implantation in native coronary lesions. Am J Cardiol 2001;88:301-303. 145. Mudra H, di Mario C, de Jaegere P, et al: Randomized comparison of coronary stent implantation under ultrasound or

146.

147.

148. 149.

150.

151.

152.

153.

154. 155. 156.

157. 158.

159.

160.

161.

162.

163.

164.

angiographic guidance to reduce stent restenosis (OPTICUS Study). Circulation 2001;104:1343-1349. Orford JL, Denktas AE, Williams BA, et al: Routine intravascular ultrasound scanning guidance of coronary stenting is not associated with improved clinical outcomes. Am Heart J 2004;148:501-506. Casella G, Klauss V, Ottani F, et al: Impact of intravascular ultrasound-guided stenting on long-term clinical outcome: A meta-analysis of available studies comparing intravascular ultrasound-guided and angiographically guided stenting. Catheter Cardiovasc Interv 2003;59:314-321. Hoffmann R, Mintz GS, Dussaillant GR, et al: Patterns and mechanisms of in-stent restenosis: A serial intravascular ultrasound study. Circulation 1996;94:1247-1254. Prati F, Di Mario C, Moussa I, et al: In-stent neointimal proliferation correlates with the amount of residual plaque burden outside the stent: An intravascular ultrasound study. Circulation 1999;99:1011-1014. Hibi K, Suzuki T, Honda Y, et al: Quantitative and spatial relation of baseline atherosclerotic plaque burden and subsequent in-stent neointimal proliferation as determined by intravascular ultrasound. Am J Cardiol 2002;90:1164-1167. Kornowski R, Mintz GS, Kent KM, et al: Increased restenosis in diabetes mellitus after coronary interventions is due to exaggerated intimal hyperplasia: A serial intravascular ultrasound study. Circulation 1997;95:1366-1369. Castagna MT, Mintz GS, Leiboff BO, et al: The contribution of “mechanical” problems to in-stent restenosis: An intravascular ultrasonographic analysis of 1090 consecutive in-stent restenosis lesions. Am Heart J 2001;142:970-974. Albertal M, Abizaid A, Munoz JS, et al: A novel mechanism explaining early lumen loss following balloon angioplasty for the treatment of in-stent restenosis. Am J Cardiol 2005;95:751-754. Shiran A, Mintz GS, Waksman R, et al: Early lumen loss after treatment of in-stent restenosis: An intravascular ultrasound study. Circulation 1998;98:200-203. Dauerman HL, Baim DS, Cutlip DE, et al: Mechanical debulking versus balloon angioplasty for the treatment of diffuse in-stent restenosis. Am J Cardiol 1998;82:277-284. Mehran R, Mintz GS, Satler LF, et al: Treatment of in-stent restenosis with excimer laser coronary angioplasty: Mechanisms and results compared with PTCA alone. Circulation 1997;96:2183-2189. Mahdi NA, Pathan AZ, Harrell L, et al: Directional coronary atherectomy for the treatment of Palmaz-Schatz in-stent restenosis. Am J Cardiol 1998;82:1345-1351. Lee SG, Lee CW, Cheong SS, et al: Immediate and long-term outcomes of rotational atherectomy versus balloon angioplasty alone for treatment of diffuse in-stent restenosis. Am J Cardiol 1998;82:140-143. Sharma SK, Kini A, Mehran R, et al: Randomized trial of Rotational Atherectomy Versus Balloon Angioplasty for Diffuse In-stent Restenosis (ROSTER). Am Heart J 2004; 147:16-22. Radke PW, Klues HG, Haager PK, et al: Mechanisms of acute lumen gain and recurrent restenosis after rotational atherectomy of diffuse in-stent restenosis: A quantitative angiographic and intravascular ultrasound study. J Am Coll Cardiol 1999;34:33-39. Dahm JB, Kuon E: High-energy eccentric excimer laser angioplasty for debulking diffuse in-stent restenosis leads to better acute- and 6-month follow-up results. J Invasive Cardiol 2000;12:335-342. Kay IP, Sabate M, Costa MA, et al: Positive geometric vascular remodeling is seen after catheter-based radiation followed by conventional stent implantation but not after radioactive stent implantation. Circulation 2000;102:1434-1439. Albiero R, Adamian M, Kobayashi N, et al: Short- and intermediate-term results of (32)P radioactive beta-emitting stent implantation in patients with coronary artery disease: The Milan Dose-Response Study. Circulation 2000;101:18-26. Bhargava B, Mintz GS, Mehran R, et al: Serial volumetric intravascular ultrasound analysis of the efficacy of beta irradiation in preventing recurrent in-stent restenosis. Am J Cardiol 2000;85:651-653, A610.

1139

1140

Evaluation of Interventional Techniques 165. Mintz GS, Weissman NJ, Teirstein PS, et al: Effect of intracoronary gamma-radiation therapy on in-stent restenosis: An intravascular ultrasound analysis from the gamma-1 study. Circulation 2000;102:2915-2918. 166. Raizner AE, Oesterle SN, Waksman R, et al: Inhibition of restenosis with beta-emitting radiotherapy: Report of the Proliferation Reduction with Vascular Energy Trial (PREVENT). Circulation 2000;102:951-958. 167. Sabate M, Marijnissen JP, Carlier SG, et al: Residual plaque burden, delivered dose, and tissue composition predict 6month outcome after balloon angioplasty and beta-radiation therapy. Circulation 2000;101:2472-2477. 168. Popma JJ, Suntharalingam M, Lansky AJ, et al: Randomized trial of 90Sr/90Y beta-radiation versus placebo control for treatment of in-stent restenosis. Circulation 2002;106: 1090-1096. 169. Silber S, Popma JJ, Suntharalingam M, et al: Two-year clinical follow-up of 90Sr/90Y beta-radiation versus placebo control for the treatment of in-stent restenosis. Am Heart J 2005; 149:689-694. 170. Teirstein PS, Massullo V, Jani S, et al: Catheter-based radiotherapy to inhibit restenosis after coronary stenting. N Engl J Med 1997;336:1697-1703. 171. Waksman R, White RL, Chan RC, et al: Intracoronary gammaradiation therapy after angioplasty inhibits recurrence in patients with in-stent restenosis. Circulation 2000;101: 2165-2171. 172. Albiero R, Nishida T, Adamian M, et al: Edge restenosis after implantation of high activity (32)P radioactive beta-emitting stents. Circulation 2000;101:2454-2457. 173. Ahmed JM, Mintz GS, Waksman R, et al: Serial intravascular ultrasound analysis of edge recurrence after intracoronary gamma radiation treatment of native artery in-stent restenosis lesions. Am J Cardiol 2001;87:1145-1149. 174. Okura H, Lee DP, Handen CE, et al: Contribution of vessel remodeling to “edge effect” following intracoronary betaradiation therapy: A serial volumetric intravascular ultrasound study [Abstract]. Circulation 1999;100:I-511. 175. Sabate M, Costa MA, Kozuma K, et al: Geographic miss: A cause of treatment failure in radio-oncology applied to intracoronary radiation therapy. Circulation 2000;101: 2467-2471. 176. Ciezki JP: Brachytherapy guided by ultrasound. In Waksman R, Serruys PW (eds): Handbook of Vascular Brachytherapy, 2nd ed. London, Martin Dunitz, 2000, pp 153-162. 177. Kirisits C, Wexberg P, Gottsauner-Wolf M, et al: Dose-volume histograms based on serial intravascular ultrasound: A calculation model for radioactive stents. Radiother Oncol 2001;59:329-337. 178. Carlier SG, Marijnissen JP, Coen VL, et al: Guidance of intracoronary radiation therapy based on dose-volume histograms derived from quantitative intravascular ultrasound. IEEE Trans Med Imaging 1998;17:772-778. 179. Kozuma K, Costa MA, Sabate M, et al: Relationship between tensile stress and plaque growth after balloon angioplasty treated with and without intracoronary beta-brachytherapy. Eur Heart J 2000;21:2063-2070. 180. Crocker I, Fox T, Carlier SG: IVUS based dosimetry and treatment planning. J Invasive Cardiol 2000;12:643-648. 181. Morino Y, Kaneda H, Fox T, et al: Delivered dose and vascular response after beta-radiation for in-stent restenosis: Retrospective dosimetry and volumetric intravascular ultrasound analysis. Circulation 2002;106:2334-2339. 182. Kaneda H, Honda Y, Morino Y, et al: Safety of beta radiation exposure to the non-target segment: An intravascular ultrasound dosimetric analysis. J Invasive Cardiol 2006;18: 309-312. 183. Maehara A, Patel NS, Harrison LB, et al: Dose heterogeneity may not affect the neointimal proliferation after gamma radiation for in-stent restenosis: A volumetric intravascular ultrasound dosimetric study. J Am Coll Cardiol 2002;39: 1937-1942. 184. Kay IP, Sabate M, Van Langenhove G, et al: Outcome from balloon induced coronary artery dissection after intracoronary beta radiation. Heart 2000;83:332-337.

185. Morino Y, Bonneau HN, Fitzgerald PJ: Vascular brachytherapy: What have we learned from intravascular ultrasound? J Invasive Cardiol 2001;13:409-416. 186. Alfonso F, Hernandez R, Goicolea J, et al: Coronary stenting for acute coronary dissection after coronary angioplasty: Implications of residual dissection. J Am Coll Cardiol 1994;24:989-995. 187. Di Mario C, Gorge G, Peters R, et al: Clinical application and image interpretation in intracoronary ultrasound. Study Group on Intracoronary Imaging of the Working Group of Coronary Circulation and of the Subgroup on Intravascular Ultrasound of the Working Group of Echocardiography of the European Society of Cardiology. Eur Heart J 1998;19: 207-229. 188. Meerkin D, Tardif JC, Bertrand OF, et al: The effects of intracoronary brachytherapy on the natural history of postangioplasty dissections. J Am Coll Cardiol 2000;36:59-64. 189. Kozuma K, Costa MA, Sabate M, et al: Late stent malapposition occurring after intracoronary beta-irradiation detected by intravascular ultrasound. J Invasive Cardiol 1999;11: 651-655. 190. Okura H, Lee DP, Lo S, et al: Late incomplete apposition with excessive remodeling of the stented coronary artery following intravascular brachytherapy. Am J Cardiol 2003;92: 587-590. 191. Shah VM, Mintz GS, Apple S, et al: Background incidence of late malapposition after bare-metal stent implantation. Circulation 2002;106:1753-1755. 192. Kay IP, Ligthart JM, Virmani R, et al: The black hole: Echolucent tissue observed following intracoronary radiation. Int J Cardiovasc Intervent 2003;5:137-142. 193. Serruys PW, Degertekin M, Tanabe K, et al: Intravascular ultrasound findings in the multicenter, randomized, doubleblind RAVEL (RAndomized study with the sirolimus-eluting VElocity balloon-expandable stent in the treatment of patients with de novo native coronary artery Lesions) trial. Circulation 2002;106:798-803. 194. Moses JW, Leon MB, Popma JJ, et al: Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315-1323. 195. Tanabe K, Serruys PW, Degertekin M, et al: Chronic arterial responses to polymer-controlled paclitaxel-eluting stents: Comparison with bare metal stents by serial intravascular ultrasound analyses. Data from the randomized TAXUS-II trial. Circulation 2004;109:196-200. 196. Weissman NJ, Koglin J, Cox DA, et al: Polymer-based paclitaxel-eluting stents reduce in-stent neointimal tissue proliferation: A serial volumetric intravascular ultrasound analysis from the TAXUS-IV trial. J Am Coll Cardiol 2005;45: 1201-1205. 197. de Ribamar CJ Jr, Mintz GS, Carlier SG, et al: Intravascular ultrasound assessment of drug-eluting stent expansion. Am Heart J 2007;153:297-303. 198. Stone GW, Ellis SG, Cox DA, et al: A polymer-based, paclitaxel-eluting stent in patients with coronary artery disease. N Engl J Med 2004;350:221-231. 199. Fajadet J, Wijns W, Laarman GJ, et al: Randomized, doubleblind, multicenter study of the Endeavor zotarolimus-eluting phosphorylcholine-encapsulated stent for treatment of native coronary artery lesions: Clinical and angiographic results of the ENDEAVOR II trial. Circulation 2006;114: 798-806. 200. Weissman NJ, Wilensky RL, Tanguay JF, et al: Extent and distribution of in-stent intimal hyperplasia and edge effect in a non-radiation stent population. Am J Cardiol 2001; 88:248-252. 201. Hirohata A, Morino Y, Ako J, et al: Comparison of the efficacy of direct coronary stenting with sirolimus-eluting stents versus stenting with predilation by intravascular ultrasound imaging (from the DIRECT Trial). Am J Cardiol 2006;98: 1464-1467. 202. Kuroda N, Kobayashi Y, Nameki M, et al: Intimal hyperplasia regression from 6 to 12 months after stenting. Am J Cardiol 2002;89:869-872. 203. Sousa JE, Costa MA, Abizaid A, et al: Four-year angiographic and intravascular ultrasound follow-up of patients treated

Intravascular Ultrasound

204.

205.

206.

207.

208. 209. 210.

211.

212. 213. 214. 215. 216.

217.

218.

219.

220.

221.

222. 223.

with sirolimus-eluting stents. Circulation 2005;111:23262329. Sousa JE, Costa MA, Abizaid AC, et al: Sustained suppression of neointimal proliferation by sirolimus-eluting stents: Oneyear angiographic and intravascular ultrasound follow-up. Circulation 2001;104:2007-2011. Sousa JE, Costa MA, Sousa AG, et al: Two-year angiographic and intravascular ultrasound follow-up after implantation of sirolimus-eluting stents in human coronary arteries. Circulation 2003;107:381-383. Aoki J, Colombo A, Dudek D, et al: Peristent remodeling and neointimal suppression 2 years after polymer-based, paclitaxel-eluting stent implantation: Insights from serial intravascular ultrasound analysis in the TAXUS II study. Circulation 2005;112:3876-3883. Sonoda S, Morino Y, Ako J, et al: Impact of final stent dimensions on long-term results following sirolimus-eluting stent implantation: Serial intravascular ultrasound analysis from the sirius trial. J Am Coll Cardiol 2004;43:1959-1963. Hong MK, Mintz GS, Lee CW, et al: Intravascular ultrasound predictors of angiographic restenosis after sirolimus-eluting stent implantation. Eur Heart J 2006;27:1305-1310. Cheneau E, Pichard AD, Satler LF, et al: Intravascular ultrasound stent area of sirolimus-eluting stents and its impact on late outcome. Am J Cardiol 2005;95:1240-1242. Takebayashi H, Kobayashi Y, Mintz GS, et al: Intravascular ultrasound assessment of lesions with target vessel failure after sirolimus-eluting stent implantation. Am J Cardiol 2005;95:498-502. Kim SW, Mintz GS, Escolar E, et al: An intravascular ultrasound analysis of the mechanisms of restenosis comparing drug-eluting stents with brachytherapy. Am J Cardiol 2006;97:1292-1298. Hodgson JM: Oh no, even stenting is affected by calcium! Cathet Cardiovasc Diagn 1996;38:236-237. Hoffmann R, Mintz GS, Popma JJ, et al: Treatment of calcified coronary lesions with Palmaz-Schatz stents: An intravascular ultrasound study. Eur Heart J 1998;19:1224-1231. Henneke KH, Regar E, Konig A, et al: Impact of target lesion calcification on coronary stent expansion after rotational atherectomy. Am Heart J 1999;137:93-99. Kataoka T, Grube E, Honda Y, et al: Three-dimensional IVUS assessment of edge effects following drug-eluting stent implantation. J Am Coll Cardiol 2002;39:70A. Hong MK, Mintz GS, Lee CW, et al: Paclitaxel coating reduces in-stent intimal hyperplasia in human coronary arteries: A serial volumetric intravascular ultrasound analysis from the ASian Paclitaxel-Eluting Stent Clinical Trial (ASPECT). Circulation 2003;107:517-520. Serruys PW, Degertekin M, Tanabe K, et al: Vascular responses at proximal and distal edges of paclitaxel-eluting stents: Serial intravascular ultrasound analysis from the TAXUS II trial. Circulation 2004;109:627-633. Sakurai R, Ako J, Morino Y, et al: Predictors of edge stenosis following sirolimus-eluting stent deployment (a quantitative intravascular ultrasound analysis from the SIRIUS trial). Am J Cardiol 2005;96:1251-1253. Costa MA. Impact of stent deployment techniques on longterm clinical outcomes of patients treated with sirolimuseluting stents: Results of the multicenter prospective S. T.L.L.R. trial. Transcatheter Cardiovascular Therapeutics Convention. Washington, DC, 2006. Fujii K, Mintz GS, Kobayashi Y, et al: Contribution of stent underexpansion to recurrence after sirolimus-eluting stent implantation for in-stent restenosis. Circulation 2004;109: 1085-1088. Fujii K, Carlier SG, Mintz GS, et al: Stent underexpansion and residual reference segment stenosis are related to stent thrombosis after sirolimus-eluting stent implantation: An intravascular ultrasound study. J Am Coll Cardiol 2005; 45:995-998. Cook S, Wenaweser P, Togni M, et al: Incomplete stent apposition and very late stent thrombosis after drug-eluting stent implantation. Circulation 2007;115:2426-2434. Drachman DE, Edelman ER, Seifert P, et al: Neointimal thickening after stent delivery of paclitaxel: Change in composi-

224.

225.

226.

227.

228.

229.

230. 231. 232. 233. 234.

235. 236.

237.

238.

239.

240.

241.

242.

tion and arrest of growth over six months. J Am Coll Cardiol 2000;36:2325-2332. Ako J, Morino Y, Honda Y, et al: Late incomplete stent apposition after sirolimus-eluting stent implantation: A serial intravascular ultrasound analysis. J Am Coll Cardiol 2005;46:1002-1005. Tanabe K, Serruys PW, Degertekin M, et al: Incomplete stent apposition after implantation of paclitaxel-eluting stents or bare metal stents: Insights from the randomized TAXUS II trial. Circulation 2005;111:900-905. Hong MK, Mintz GS, Lee CW, et al: Late stent malapposition after drug-eluting stent implantation: An intravascular ultrasound analysis with long-term follow-up. Circulation 2006;113:414-419. Sano K, Mintz GS, Carlier SG, et al: Volumetric intravascular ultrasound assessment of neointimal hyperplasia and nonuniform stent strut distribution in sirolimus-eluting stent restenosis. Am J Cardiol 2006;98:1559-1562. Takebayashi H, Mintz GS, Carlier SG, et al: Nonuniform strut distribution correlates with more neointimal hyperplasia after sirolimus-eluting stent implantation. Circulation 2004;110:3430-3434. Lemos PA, Saia F, Ligthart JM, et al: Coronary restenosis after sirolimus-eluting stent implantation: Morphological description and mechanistic analysis from a consecutive series of cases. Circulation 2003;108:257-260. Halkin A, Carlier S, Leon MB: Late incomplete lesion coverage following Cypher stent deployment for diffuse right coronary artery stenosis. Heart 2004;90:e45. Aoki J, Nakazawa G, Tanabe K, et al: Incidence and clinical impact of coronary stent fracture after sirolimus-eluting stent implantation. Catheter Cardiovasc Interv 2007;69:380-386. Lee MS, Jurewitz D, Aragon J, et al: Stent fracture associated with drug-eluting stents: Clinical characteristics and implications. Catheter Cardiovasc Interv 2007;69:387-394. Hausmann D, Erbel R, Alibelli CM, et al: The safety of intracoronary ultrasound: A multicenter survey of 2207 examinations. Circulation 1995;91:623-630. Evans JL, Ng KH, Wiet SG, et al: Accurate three-dimensional reconstruction of intravascular ultrasound data: Spatially correct three-dimensional reconstructions. Circulation 1996; 93:567-576. Solomon SB, Dickfeld T, Calkins H: Real-time cardiac catheter navigation on three-dimensional CT images. J Interv Card Electrophysiol 2003;8:27-36. Feldman CL, Ilegbusi OJ, Hu Z, et al: Determination of in vivo velocity and endothelial shear stress patterns with phasic flow in human coronary arteries: A methodology to predict progression of coronary atherosclerosis. Am Heart J 2002;143:931-939. Stone PH, Coskun AU, Kinlay S, et al: Effect of endothelial shear stress on the progression of coronary artery disease, vascular remodeling, and in-stent restenosis in humans: In vivo 6-month follow-up study. Circulation 2003;108: 438-444. Stone PH, Coskun AU, Yeghiazarians Y, et al: Prediction of sites of coronary atherosclerosis progression: In vivo profiling of endothelial shear stress, lumen, and outer vessel wall characteristics to predict vascular behavior. Curr Opin Cardiol 2003;18:458-470. Wentzel JJ, Krams R, Schuurbiers JC, et al: Relationship between neointimal thickness and shear stress after Wallstent implantation in human coronary arteries. Circulation 2001;103:1740-1745. Gijsen FJ, Oortman RM, Wentzel JJ, et al: Usefulness of shear stress pattern in predicting neointima distribution in sirolimus-eluting stents in coronary arteries. Am J Cardiol 2003;92:1325-1328. Carlier SG, van Damme LC, Blommerde CP, et al: Augmentation of wall shear stress inhibits neointimal hyperplasia after stent implantation: Inhibition through reduction of inflammation? Circulation 2003;107:2741-2746. Kawasaki M, Takatsu H, Noda T, et al: Noninvasive quantitative tissue characterization and two-dimensional color-coded map of human atherosclerotic lesions using ultrasound integrated backscatter: Comparison between histology and

1141

1142

Evaluation of Interventional Techniques

243.

244.

245.

246.

247.

248.

249.

250. 251.

252.

253.

254.

255.

integrated backscatter images. J Am Coll Cardiol 2001;38: 486-492. Kawasaki M, Sano K, Okubo M, et al: Volumetric quantitative analysis of tissue characteristics of coronary plaques after statin therapy using three-dimensional integrated backscatter intravascular ultrasound. J Am Coll Cardiol 2005;45: 1946-1953. Wilson LS, Neale ML, Talhami HE, et al: Preliminary results from attenuation-slope mapping of plaque using intravascular ultrasound. Ultrasound Med Biol 1994;20: 529-542. Spencer T, Ramo MP, Salter DM, et al: Characterisation of atherosclerotic plaque by spectral analysis of intravascular ultrasound: An in vitro methodology. Ultrasound Med Biol 1997;23:191-203. Komiyama N, Berry GJ, Kolz ML, et al: Tissue characterization of atherosclerotic plaques by intravascular ultrasound radiofrequency signal analysis: An in vitro study of human coronary arteries. Am Heart J 2000;140:565-574. Hiro T, Fujii T, Yasumoto K, et al: Detection of fibrous cap in atherosclerotic plaque by intravascular ultrasound by use of color mapping of angle-dependent echo-intensity variation. Circulation 2001;103:1206-1211. Nair A, Kuban BD, Obuchowski N, et al: Assessing spectral algorithms to predict atherosclerotic plaque composition with normalized and raw intravascular ultrasound data. Ultrasound Med Biol 2001;27:1319-1331. Murashige A, Hiro T, Fujii T, et al: Detection of lipid-laden atherosclerotic plaque by wavelet analysis of radiofrequency intravascular ultrasound signals: In vitro validation and preliminary in vivo application. J Am Coll Cardiol 2005;45: 1954-1960. Nair A, Kuban BD, Tuzcu EM, et al: Coronary plaque classification with intravascular ultrasound radiofrequency data analysis. Circulation 2002;106:2200-2206. Nasu K, Tsuchikane E, Katoh O, et al: Accuracy of in vivo coronary plaque morphology assessment: A validation study of in vivo virtual histology compared with in vitro histopathology. J Am Coll Cardiol 2006;47:2405-2412. Rodriguez-Granillo GA, Garcia-Garcia HM, McFadden EP, et al: In vivo intravascular ultrasound-derived thin-cap fibroatheroma detection using ultrasound radiofrequency data analysis. J Am Coll Cardiol 2005;46:2038-2042. de Korte CL, Pasterkamp G, van der Steen AF, et al: Characterization of plaque components with intravascular ultrasound elastography in human femoral and coronary arteries in vitro. Circulation 2000;102:617-623. de Korte CL, van der Steen AF, Cepedes EI, et al: Characterization of plaque components and vulnerability with intravascular ultrasound elastography. Phys Med Biol 2000;45: 1465-1475. de Korte CL, Sierevogel MJ, Mastik F, et al: Identification of atherosclerotic plaque components with intravascular ultra-

256. 257.

258.

259.

260. 261.

262.

263. 264. 265. 266. 267.

268.

269.

sound elastography in vivo: A Yucatan pig study. Circulation 2002;105:1627-1630. Schaar JA, De Korte CL, Mastik F, et al: Characterizing vulnerable plaque features with intravascular elastography. Circulation 2003;108:2636-2641. Schaar JA, Regar E, Mastik F, et al: Incidence of high-strain patterns in human coronary arteries: assessment with threedimensional intravascular palpography and correlation with clinical presentation. Circulation 2004;109:2716-2719. Barger AC, Beeuwkes R, 3rd, Lainey LL, et al: Hypothesis: Vasa vasorum and neovascularization of human coronary arteries. A possible role in the pathophysiology of atherosclerosis. N Engl J Med 1984;310:175-177. Wilson SH, Herrmann J, Lerman LO, et al: Simvastatin preserves the structure of coronary adventitial vasa vasorum in experimental hypercholesterolemia independent of lipid lowering. Circulation 2002;105:415-418. Kolodgie FD, Gold HK, Burke AP, et al: Intraplaque hemorrhage and progression of coronary atheroma. N Engl J Med 2003;349:2316-2325. Moreno PR, Purushothaman KR, Fuster V, et al: Plaque neovascularization is increased in ruptured atherosclerotic lesions of human aorta: Implications for plaque vulnerability. Circulation 2004;110:2032-2038. Moulton KS, Vakili K, Zurakowski D, et al: Inhibition of plaque neovascularization reduces macrophage accumulation and progression of advanced atherosclerosis. Proc Natl Acad Sci U S A 2003;100:4736-4741. Carlier S, Kakadiaris IA, Dib N, et al: Vasa vasorum imaging: A new window to the clinical detection of vulnerable atherosclerotic plaques. Curr Atheroscler Rep 2005;7:164-169. Alkan-Onyuksel H, Demos SM, Lanza GM, et al: Development of inherently echogenic liposomes as an ultrasonic contrast agent. J Pharm Sci 1996;85:486-490. Hamilton AJ, Huang SL, Warnick D, et al: Intravascular ultrasound molecular imaging of atheroma components in vivo. J Am Coll Cardiol 2004;43:453-460. Lanza GM, Wallace KD, Scott MJ, et al: A novel site-targeted ultrasonic contrast agent with broad biomedical application. Circulation 1996;94:3334-3340. Lanza GM, Abendschein DR, Hall CS, et al: In vivo molecular imaging of stretch-induced tissue factor in carotid arteries with ligand-targeted nanoparticles. J Am Soc Echocardiogr 2000;13:608-614. Demos SM, Alkan-Onyuksel H, Kane BJ, et al: In vivo targeting of acoustically reflective liposomes for intravascular and transvascular ultrasonic enhancement. J Am Coll Cardiol 1999;33:867-875. Crowder KC, Hughes MS, Marsh JN, et al: Sonic activation of molecularly-targeted nanoparticles accelerates transmembrane lipid delivery to cancer cells through contact-mediated mechanisms: Implications for enhanced local drug delivery. Ultrasound Med Biol 2005;31:1693-1700.

CHAPTER

63 Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment Pedro R. Moreno, Javier Sanz, and Valentin Fuster KEY POINTS 䊏 Plaque instability can now be detected through a number of invasive and noninvasive technologies. 䊏 It is unclear which of these technologies will emerge as the most pragmatic and useful.

Independent of social status, race, or economical income, cardiovascular disease is by far the first cause of death, as shown in Figure 63-1.1 In the United States alone, someone is experiencing a heart attack every 30 seconds, on average, with more than 1 million attacks per year. Despite its deadly manifestations, coronary atherosclerosis is a condition that can remain asymptomatic for decades. The transition from asymptomatic, nonobstructive disease to symptomatic, occlusive disease is related to coronary thrombosis. This condition, known as atherothrombosis,2 is related mostly to plaque rupture or plaque erosion. Atherosclerotic plaques at increased risk for rupture are characterized by specific features suitable for proper identification by novel imaging techniques. Therefore, it is easy to understand why the high-risk, vulnerable plaque3 is considered by some the “holy grail” of cardiology today.4 However, several high-risk plaques can be present in the same patient, making it difficult for the interventionalist to identify and treat the single plaque that would trigger thrombosis. This fundamental issue may limit the scope of this field in the future. Nevertheless, tremendous advances are being made in the diagnosis and therapy of high-risk plaques. Therefore, it is crucial for interventional cardiologists to develop a comprehensive approach to plaque vulnerability.

䊏 Validation of plaque instability imaging via clinical followup to determine natural history will be quite important.

This chapter provides a systematic approach to high-risk vulnerable plaques and is divided into four sections. The first section is devoted to definition, incidence, location, risk factors, and clinical presentation. The second section is devoted to plaque composition, setting up the foundations for understanding plaque vulnerability. The third section summarizes the evolving field of invasive plaque imaging. The final section is devoted to therapy, from conservative, pharmacologic options to aggressive percutaneous coronary intervention (PCI) alternatives.

CLINICAL CHARACTERISTICS Definition and Clinical Evidence The concept of high-risk plaques evolved from pathologic studies that evaluated lesions responsible for thrombosis. Muller and colleagues originally introduced the term “vulnerable plaque” in 1985 to describe the potential of certain vessels to become vulnerable to thrombotic occlusion after plaque disruption in a circadian pattern.5 The adoption of the term “vulnerable plaque” by most investigators relies on the temporal associations among acute coronary thrombosis and certain types of morphologic features contained in the culprit atherosclerotic plaques (thin fibrous cap, large lipid core,

1143

Evaluation of Interventional Techniques 17,528,000

7,586,000

4,057,000

D ia be te s

1,125,000

C an ce r re C sp hr di ira on se to ic as ry es

M al ar C ia ar di ov di asc se u as lar es

2,830,000 1,607,000 883,000

H IV /A ID S Tu be rc ul os is

1144

Figure 63-1. Projected global deaths according to the World Health Organization. (Redrawn from Fuster V, Moreno PR, Fayad ZA, et al: Atherothrombosis and high-risk plaque. Part I: Evolving concepts. J Am Coll Cardiol 2005;46:937-954.)

macrophages),6 which are discussed later in this chapter. However, in daily practice in the cardiac catheterization laboratory, these morphologic features cannot be visualized. As a result, a histologic definition does not meet clinical standards. Therefore, we need a definition that can be useful in daily practice. For the interventionalist, the high-risk, vulnerable plaque is defined as a nonobstructive, silent coronary lesion that suddenly becomes obstructive and symptomatic, as shown in Figure 63-2. The clinical evidence of this disease was developed by Ambrose and Fuster in 1988, when they studied the baseline characteristics of lesions that evolved to produce acute myocardial infarction (AMI).7 The investigators found that, at baseline, the majority of these lesions were nonobstructive, with mean diameter stenosis of 48%. Multiple investigators reproduced this finding, creating the concept that most of the lesions responsible for infarction originate from nonobstructive coronary artery disease (CAD). Incidence In line with the Ambrose and Fuster criteria, the number of nonobstructive, asymptomatic lesions

A

Baseline

B

12 months

C

24 months

D

36 months

Figure 63-2. Rapid progression from nonobstructive, asymptomatic to severely obstructive and symptomatic disease, fulfilling the clinical definition of vulnerable plaques. Sequential coronary angiograms of the left anterior descending coronary artery (LAD) were performed at 12-month intervals. Arrows indicate locations of stenosis. A, Baseline angiogram showing 10% to 20% stenosis in the LAD in the proximal and middle segments. B, Progression to 30% to 40% stenosis in both segments, 12 months after the baseline angiogram. C, At 24 months, there is further progression to severe, 95% stenosis in the proximal segment and 50% stenosis in the middle segment. At that time, the patient was treated with a single drug-eluting stent covering both lesions (angiogram not shown). D, Stenosis is 10% to 20% in the same segments 12 months after local therapy with stent (36 months after baseline angiogram). (Courtesy of the cardiac catheterization laboratory at Mount Sinai Medical Center, New York, New York.)

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment 25%

20% Hazard rate

Figure 63-3. Hazard rates per year for target-lesion (blue) and nontarget-lesion (red) events derived from life table survival analysis. Target-lesion events include any repeat revascularization, death, myocardial infarction (MI), acute coronary syndrome (ACS), or congestive heart failure (CHF) attributed to the target lesion. Nontarget-lesion events include all repeat revascularizations involving the target vessel outside the target lesion, any nontarget-vessel revascularization, and any death, MI, ACS, or CHF that was clearly not attributable to the target lesion. (Adapted and redrawn from Cutlip DE, Chhabra AG, Baim DS, et al: Beyond restenosis: Five-year clinical outcomes from second-generation coronary stent trials. Circulation 2004;110:1226-1230.)

that progress to obstructive, symptomatic lesions can indicate the incidence of vulnerable plaques in patients with established CAD. Two studies have carefully addressed this issue. The first study included 1228 patients who underwent PCI for symptomatic CAD. The incidence of nonobstructive (also called nontarget) lesions that required additional PCI was 12.4% in the first year and 5% to 7% per year in years 2 to 5 after the initial procedure (Fig. 63-3).8 The second study included 3747 post-PCI patients from the National Heart, Lung, and Blood Institute (NHLBI) registry.9 The incidence of nonobstructive (nontarget) lesions that required additional PCI was 6% in the first year, ranging from 4.4% to 12.8% according to the number of vessels involved (Fig. 63-4). 16%

15%

18.3%

Restenosis

Target lesion event

Progression

Non-target lesion event

12.4%

10% 7.0%

5.6%

5.7%

6.7%

5% 2.3%

1.3%

1.5%

1.4%

Year 3

Year 4

Year 5

0% Year 1

Year 2

Risk Factors The risk factors associated with nonobstructive, silent lesions evolving into obstructive, symptomatic lesions requiring PCI were identified using multiple regression analysis and included multivessel CAD at baseline (three- and two-vessel CAD), previous PCI, and age younger than 65 years. Of note, treatment with statins failed to protect patients within the first year.9 Regarding biomarkers, a 2007 review by Koening and Khuseyinova summarized the current view of this field.10 Biomarkers per se cannot predict plaque instability but provide only likelihood ratios, c statistics, area under the curve data, and receiver operating characteristic analysis. Further research is needed to completely elucidate the role of

Single vessel (n
1520) Double vessel (n
1140) Triple vessel (n
839)

P .001 12.8%

New lesion PCI

12%

8% 7.1%

4.4%

4%

0% 0

60

120

180

240

300

360

Days from study entry

Figure 63-4. Kaplan-Meier analysis of repeat coronary intervention for rapid progression of nontarget lesions one year after coronary intervention. Note that the increased incidence of repeat intervention is related to the number of epicardial vessels involved during the first percutaneous revascularization. (Redrawn from Glaser R, Selzer F, Faxon DP, et al: Clinical progression of incidental, asymptomatic lesions discovered during culprit vessel coronary intervention. Circulation 2005;111:143-149.)

1145

1146

Evaluation of Interventional Techniques biomarkers in predicting plaque rupture and coronary events.10 Clinical Presentation Nonobstructive, silent lesions evolving into obstructive, symptomatic lesions requiring PCI trigger acute coronary events in 68.5% of the cases, the majority of patients presenting with unstable angina pectoris and 9.3% presenting with nonfatal myocardial infarction.9 Of note, because these studies excluded death, fatal events are probably underestimated. Anatomic Distribution The anatomic distribution of lesions evolving into acute coronary events was initially evaluated by Ambrose and Fuster, who identified proximal location as the only independent predictor for progression to AMI.7 More recently, Wang and associates properly characterized the anatomic distribution of coronary lesions responsible for AMI.11 As previously reported by other trials of AMI, the right coronary artery was the vessel most commonly involved (45%), followed by the left anterior descending coronary artery (LAD, 39%) and the left circumflex coronary artery (16%). The location within the vessel was dominated by the proximal segment, which harbored 80% of the lesions responsible for myocardial infarction in all three major vessels. For example, in the LAD, 90% of the high-risk plaques that evolved into AMI were located within the first 40 mm of the vessel. For all lesions, with each 10-mm increase in distance from the ostium, the risk of an acute coronary occlusion was significantly decreased, by 30% in the LAD, 26% in the left circumflex artery, and 13% in the right coronary artery. These data were recently reproduced by other investigators.12 As a result, proper evaluation and treatment of high-risk plaques in the proximal segments of the coronary tree has the potential of preventing 80% of AMIs. For the interventionalist, the incidence of highrisk, vulnerable plaques evolving into clinical events ranges from 4% per year, in patients with singlevessel disease, up to 13% in patients with triple-vessel disease. The lesions are usually located in the proximal segment of the coronary arteries, and the most common presentation is acute coronary syndrome (ACS). After a review of the clinical characteristics, the next section of this chapter provides the basis for understanding the pathophysiology of the disease. Most importantly, this offers the foundation to critically evaluate novel imaging techniques that claim effectiveness in the diagnosis of high-risk, vulnerable plaques.

PLAQUE COMPOSITION As discussed earlier, the mechanisms responsible for rapid progression to occlusion in atherothrombosis include plaque rupture and plaque erosion.2 Plaque

Figure 63-5. Cross-sectioned coronary artery containing a ruptured plaque at the shoulder of the fibrous cap with a nonocclusive thrombus superimposed. The large necrotic core can be identified by cholesterol crystals and extensive intraplaque hemorrhage secondary to plaque rupture. (Trichrome stain, rendering thrombus red, collagen blue, and lipid colorless.) (Courtesy of Dr. K-Raman Purushothaman, Mount Sinai Hospital, New York, New York.)

rupture is by far the most common cause of atherothrombosis, responsible for 70% to 75% of all events and up to 85% of events in hypercholesterolemic white males. In plaque rupture, disease progresses through lipid core expansion and macrophage accumulation at the edges of the plaque, leading to fibrous cap disruption (Fig. 63-5). As a result, identifying plaques at risk for rupture offers the possibility of preventing the most common substrate for coronary thrombosis. The second cause of atherothrombosis is plaque erosion. Included initially as “other causes of coronary thrombosis,” plaque erosion gained attention in the last decade as a significant substrate for coronary thrombosis and sudden cardiac death in premenopausal female patients.13 In contrast to plaque rupture, erosion occurs in plaques with no specific features suitable for detection. Most of these plaques exhibit histologic patterns similar to those of plaques responsible for stable angina. They are characterized by a thick, smooth muscle cell–rich, fibrous cap; reduced necrotic core areas; and a low degree of inflammation13 (Fig. 63-6). Plaque erosion is also associated with cigarette smoking, suggesting that thrombosis in these patients may be related to a systemic, prothrombogenic pathway rather than a local, atherothrombotic mechanism. Considering that plaques at risk for erosion cannot be differentiated from stable plaques, the imaging field has focused its attention on characterizing plaques at risk for rupture. In the absence of prospective, natural history studies, investigators have extrapolated the structural and chemical features of ruptured plaques to study this disease. The characteristic lesion is the thin-cap fibroatheroma (TCFA),

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

Figure 63-6. Plaque erosion. Cross-section of a coronary artery containing a stenotic atherosclerotic plaque with an occlusive thrombosis superimposed. The endothelium is missing at the plaque-thrombus interface, but the plaque surface is otherwise intact. (Trichrome stain, rendering thrombus red, collagen blue, and lipid colorless.) (Courtesy of Dr. Erling Falk, Aarhus, Denmark.)

Figure 63-7. Thin-cap fibroatheroma (TCFA), characterized by a very thin fibrous cap and a large necrotic core. (Trichrome stain.) (Courtesy of Dr. K-Raman Purushothaman, Mount Sinai Hospital, New York, New York.)

which is considered the hallmark of high-risk, vulnerable plaques14 (Fig. 63-7). The classic histologic patterns of TCFA include, but are not limited to (1) thin fibrous cap with increased stress/strain relationship; (2) large necrotic core with increased free-toesterified cholesterol ratio; (3) increased plaque inflammation; (4) positive vascular remodeling; (5) increased vasa vasorum neovascularization; and (6) intraplaque hemorrhage (IPH). Thin Fibrous Cap with Increased Stress/Strain Relationship Autopsy studies have shown that ruptured plaques are characterized by a very thin fibrous cap, measuring 23 ± 19 µm (microns) in thickness. Most importantly, 95% of ruptured caps measured 64 µm or less in the coronary arteries15 and 60 µm or less in the

aorta.16 As a result, the first and probably most important histologic feature of TCFA is a fibrous cap 65 µm or less in thickness. These thin caps are unable to withstand the circumferential tensile stress applied by the oscillations of arterial blood pressure. The ratio of the circumferential tensile stress to the radial strain of the fibrous cap equals the stiffness of the tissue. Hence, soft (fatty) tissue will be more strained than stiff (fibrous) tissue when equally stressed.17 Furthermore, as caps become thinner, the stress increases in an exponential pattern18 (Fig. 63-8). In addition, as lipid pools become larger, stress also increases. Therefore, the strength of a cap may be as important as the actual thickness of a fibrous cap. This stress/ strain relation in the fibrous cap is considered a feature for plaque vulnerability.17 New technology aiming at detecting TCFAs should have a radial resolution less than 65 µm and the 30

Figure 63-8. Relationship between circumferential stress (vertical axis) and fibrous cap thickness (horizontal axis). Note the exponential increase in circumferential stress when cap thickness is reduced to less than 200 µm. (Redrawn from Loree HM, Kamm RD, Stringfellow RG, Lee RT: Effects of fibrous cap thickness on peak circumferential stress in model atherosclerotic vessels. Circ Res 1992;71:850-858.)

Circumferential stress

25 20 15 10 5 0 100

150

200

300

400

Fibrous cap thickness (microns)

500

1147

1148

Evaluation of Interventional Techniques ability to quantify the stress/strain relationship in the fibrous cap. Large Necrotic Core with Increased Free-to-Esterified Cholesterol Ratio Modified oxidized low-density lipoprotein (oxLDL) is avidly taken up by macrophages via scavenger receptors, leading to cytoplasm overload with lipid droplets. Continuous inflow of oxLDL leads to cell death, with extracellular lipid accumulation within the matrix of the plaque. This is the basic mechanism of the necrotic core, which is formed after cell death by necrosis or apoptosis of lipid-laden macrophages, foam cells, and erythrocytes.19 Active collagen dissolution by metalloproteinases contributes to core expansion, which plays a major role in plaque vulnerability. Dr. Michael Davies suggested that coronary plaques with necrotic cores larger than 50% of the total plaque area are at risk for rupture and thrombosis.20 In the aorta, TCFAs exhibit necrotic core areas of 40%, and ruptured plaques up to 50%, of total plaque area.16 However, other studies in coronary arteries have shown lower necrotic areas, down to 24% and 34% in TCFAs and ruptured plaques, respectively.14 The composition of the core may influence the propensity to develop plaque rupture and thrombosis. A necrotic core with an increased free-toesterified cholesterol ratio favors plaque rupture.21 The composition of fatty acids at the rupture site may also determine the likelihood of local platelet thrombus and fibrin formation. Increased concentrations of pro-aggregatory saturated fatty acids and oxidized derivatives of polyunsaturated fatty acids favor thrombus formation. Most importantly, tissue factor activity within the necrotic core is currently considered the major source of thrombin generation leading to arterial thrombosis in humans.22 Of interest, the percentage of cholesterol clefts within the necrotic core is increased in ruptured plaques,23 and these clefts may even exert a direct local effect by tearing the fibrous cap, as recently documented by electron microscopy.24 Finally, the degree of calcification in the necrotic core is variable in TCFA. In a series of sudden coronary death cases, more than 50% of TCFAs showed a lack of calcification or only speckled calcification on postmortem radiographs of coronary arteries. In the remaining lesions, calcification was either fragmented or diffuse, suggesting a large variation in the degree of calcification within TCFAs.25 In contrast, 65% of acute ruptures demonstrated speckled calcification, with the remainder showing fragmented or diffuse calcification.25 To evaluate the role of necrotic core calcification in plaque instability, our group quantified the incidence of calcification in human aortic TCFAs (n = 42), compared with nonTCFA plaques (n = 128), using Von-Kossa staining.26 The incidences of early (speckled) and advanced (coarse) calcification were 62% and 82%, respectively (P = .006), suggesting that a significant number of TCFAs may not exhibit necrotic core calcification. As

a result, calcification is an equivocal variable in TCFA and may not be used as a surrogate for high-risk necrotic cores.27,28 New technology aiming at detecting TCFAs should precisely quantify necrotic core areas, which should be at least 24% of total plaque area. It should also provide free cholesterol and tissue factor content, independent of the degree of calcification. Increased Plaque Inflammation Macrophages and T cells play a major active role in the pathophysiology of TCFA. These cells are capable of degrading extracellular matrix by phagocytosis or secretion of proteolytic enzymes; thus, enzymes such as plasminogen activators and matrix metalloproteinases (MMPs)—including collagenases, elastases, gelatinases, and stromelysins—weaken the already thin fibrous cap and predispose it to rupture.29,30 The continuing entry, survival, and replication of monocyte-macrophages within plaques is mediated by a defense mechanism aimed to remove oxLDL and reduce the deleterious effects related to oxidation and reactive oxygen generation (ROS) products. Cytokines regulate oxLDL uptake, modulating the macrophage scavenger receptors. Most importantly, in situations where the macrophage scavenger capacity is overloaded, these cytokines also activate cell death by apoptosis, releasing MMPs and tissue factor.31,32 This link was elegantly documented by Hutter and colleagues, who showed excellent correlations between macrophage density, apoptosis, and tissue factor expression in human and mouse atherosclerotic lesions.33 Thus, macrophages, following what appears to be a defensive mechanism protecting the vessel wall, may eventually fail and undergo apoptosis, leading to plaque rupture and thrombosis.31,34,35 Of clinical relevance, multiple studies have shown that macrophage content is increased in plaques from patients with ACS, compared to plaques from patients with stable angina.36 Therefore, plaque inflammation is a pivotal feature of plaque vulnerability.2,37,38 New technology aiming at detecting TCFAs should have the resolution to identify and quantify macrophages in the fibrous cap and shoulders of the atherosclerotic plaque. Positive Vascular Remodeling The eccentric growth of atheroma away from the lumen was described by Glagov and colleagues in 1987.39 The vessel wall can expand significantly to harbor large atheromas without obstructing the lumen (Fig. 63-9). Since then, remodeling has been consistently identified in lesions responsible for ACS. Varnava and Davies studied the relationship between remodeling and plaque vulnerability. Of 108 coronary plaques analyzed, 64 (59%) had undergone no remodeling or positive remodeling, and 44 (41%) had undergone negative remodeling (vessel shrinkage).40 Lesions with positive remodeling, compared

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment Proximal reference

1149

Culprit lesion

Positive remodeling RR 1.05

EEM contour Proximal reference

Figure 63-9. Large human thrombotic left main coronary artery plaque with extensive remodeling containing a large necrotic core. (Courtesy of Dr. K-Raman Purushothaman, Mount Sinai Hospital, New York, New York.)

Culprit lesion

Negative remodeling RR 0.95

EEM contour Remodeling ratio (RR)
EEM area lesion / EEM area proximal reference

to lesions with vessel shrinkage, had a larger lipid core (mean, 39% ± 21% versus 22% ± 23%; P < .0001) and a higher macrophage count (16 ± 12 versus 9 ± 11; P = .005), explaining the common link between compensatory remodeling and plaque rupture. The mechanisms responsible for remodeling involve an inflammatory process at the base of the plaque, which leads to digestion of the internal elastic lamina (IEL) and involvement of the deeper layers of the vessel wall, including the tunica media and the adventitia. Several studies have shown increased expression of MMPs within the intimomedial interface of remodeled plaques.41 Furthermore, disruption of the IEL is associated with medial and adventitial inflammation.16 Concordantly, Burke and coworkers demonstrated that marked expansion of the IEL occurred also in plaque hemorrhage with or without rupture.42 Based on multivariate analysis, independent predictors of remodeling included macrophage infiltration, calcification, and lipid core area, providing further evidence linking remodeling to plaque vulnerability. The clinical relevance of remodeling was pioneered by Schoenhagen and colleagues, who studied 85 patients with unstable and 46 with stable coronary syndromes using intravascular ultrasound (IVUS) before coronary intervention.43 The remodeling ratio (RR) was defined as the area of the external elastic membrane at the lesion divided by the same area at the proximal reference site. Positive remodeling was defined as an RR greater than 1.05, and negative remodeling as an R lower than 0.95 (Fig. 63-10). The RR was higher at target lesions in patients with ACS than in patients with stable angina. As a result, positive remodeling was more frequent in ACS (51.8% versus 19.6%), whereas negative remodeling was more frequent in stable angina (56.5% versus 31.8%) (P = .001), confirming the histopathologic associations between plaque remodeling and vulnerability.43

Figure 63-10. Diagram of remodeling explaining the direction of positive and negative remodeling. See text for details. EEM, external elastic membrane; RR, remodeling ratio. (Redrawn from Schoenhagen P, Ziada KM, Kapadia SR, et al: Extent and direction of arterial remodeling in stable versus unstable coronary syndromes: An intravascular ultrasound study. Circulation 2000;101:598-603.)

New technology aiming at detecting TCFAs should provide exact measurements to quantify the degree of vascular remodeling. Of clinical relevance, Corti and colleagues were the first to document the same eccentric pattern for plaque regression after aggressive lipid therapy.44 Multiple studies have confirmed this observation45-47 (Fig. 63-11). Considering that lipid is the main plaque component that can be reversed with therapy, this eccentric pattern of plaque regression suggests an effective reverse lipid transport system through the deeper layers of the vessel wall, probably mediated by vasa vasorum neovascularization.2,48 Increased Vasa Vasorum Neovascularization Neovascularization is the process of generating new blood vessels to nurture the atherosclerotic plaque.48 Angiogenesis, the predominant form of neovascularization in atherosclerosis,49 is mediated by progenitor and/or endothelial cell sprouting from postcapillary venules, leading mostly to new capillaries.50 Atherosclerotic neovascularization evolves in early atherogenesis as a defense mechanism against hypoxia, which is generated by thickening of the tunica intima.51 In advanced disease, neovessels may play a defensive role, allowing for lipid removal from the plaque through the adventitia, leading to plaque regression, as described earlier. The adventitial vasa vasorum is the main source of neovascularization in atherosclerotic lesions (Fig. 63-12). Coronary vasa originate from bifurcation segments of epicardial vessels and selectively respond to sympathetic activity.51 Coronary plaque neovascularization in humans

1150

Evaluation of Interventional Techniques Baseline

EEM area 16.35 mm2

Follow-up

EEM area 11.77 mm 2 Atheroma area 10.16 mm2

Atheroma area 5.81 mm2

Lumen area 6.19 mm2

Lumen area 5.96 mm2

was elegantly delineated by Barger and colleagues52 using cinematography (Fig. 63-13). Neovascularization was distributed from the epicardial fat to the plaque throughout vessel wall.52 A decade later, Kumamoto and associates identified the vessel lumen as another source for microvessels.53 Nevertheless, neovessels from adventitial vasa were 28 times more numerous (96.5%) than those from luminal side (3.5%). Neovessels from adventitial vasa characterized severely stenotic lesions and correlated with the extent of inflammatory cell infiltration and lipid core size. On the other hand, neovessels from lumen origin were found in plaques with 40% and 50% stenosis and were associated more often with IPH or hemosiderin deposits.53 Neovessels may also serve as a pathway for leukocyte recruitment to high-risk areas of the plaque, including the cap and shoulders.54 The pivotal work of O’Brien and colleagues documented the mechanisms underlying neovessel recruitment of plaque leukocytes in human atherosclerosis. The expressions of vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1), and E-selectin were twofold to threefold higher on neoves-

Figure 63-11. The top panels show the baseline and followup intravascular ultrasound (IVUS) images of a single coronary cross-section after 24 months of rosuvastatin treatment. The bottom two panels illustrate measurements superimposed on the same cross-sections, demonstrating the reduction in atheroma area. EEM, external elastic membrane. (From Nissen SE, Nicholls SJ, Sipahi I, et al: Effect of very high-intensity statin therapy on regression of coronary atherosclerosis: The ASTEROID trial. JAMA 2006;295:1556-1565.)

sels compared to the arterial luminal endothelium, confirming the predominant role for neovessels as a pathway for leukocyte infiltration in human coronary plaques.55,56 More recently, our group documented histologic evidence of atherosclerotic neovascularization as a pathway for macrophage infiltration in advanced, lipid-rich plaques57 (Fig. 63-14). Of clinical relevance, neovessel content was significantly increased in plaques with severe inflammation. Moreover, ruptured plaques exhibited the highest degree of neovascularization.58 Further analysis of plaque angiogenesis in diabetes documented a complex morphology including sprouting, red blood cell (RBC) extravasation, and perivascular inflammation.59 New technology aimed at detecting TCFAs should quantify vasa vasorum neovascularization in the adventitia, in the tunica media, and within the atherosclerotic plaque. In summary, plaque neovascularization, in what appears to be a defensive mission to provide oxygen and remove lipid from the atherosclerotic lesion, may eventually fail, leading to extravasation of RBCs, perivascular inflammation, and IPH.

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

Figure 63-12. A, Volume-rendered high-resolution, threedimensional microcomputed tomographic image of the descending aorta vasa vasorum. B, Histologic cross-section demonstrates atherosclerotic lesion in the inferior vena cava (black arrow). C, Enlargement of microphotograph shown in B. D, Highlighted differentiated arterial (red) and venous (blue) vasa vasorum (box and white arrow correspond to same areas on A). (Masson Trichrome stain, bar 500 µm).

Intraplaque Hemorrhage Neovessel leakage leads to extravasation of plasma content and RBCs into the plaque, also known as IPH. Two main events occur after IPH. The first event includes ceroid accumulation, RBC membrane lysis, and lipid deposition. Perivascular foam cells frequently contain RBCs (hemoglobin [Hb], iron), suggesting that IPH initiates erythrocyte phagocytosis, leading to iron deposition and macrophage activation.60 In addition, the erythrocyte membrane is very rich in cholesterol, promoting lipid core expansion and increasing plaque vulnerability for rupture.61,62 The second event after IPH includes macrophage activation by free Hb. After RBC membrane lysis, extracorpuscular Hb can induce oxidative tissue damage by virtue of its heme iron,63 with subsequent

production of reactive oxygen species (ROS). Extracorpuscular Hb can also activate the pro-inflammatory transcription factor, nuclear factor-κB (NF-κB),64 leading to inflammation and angiogenesis.65 The primary defense mechanism against free Hb is haptoglobin (Hp), which rapidly and irreversibly binds to free Hb to form a Hp-Hb complex. In the vascular compartment, the Hp-Hb complex is cleared by two pathways, via the liver (90%) or the monocyte (10%).66,67 However, in extravascular sites such as atherosclerotic plaques, the only route for clearance of the Hp-Hb complex is via the macrophage.67 Most importantly, the ultimate effect of the Hp-Hb complex on iron deposition and macrophage activation may be determined by the Hp genotype.68 Two classes of alleles (Hp-1 and Hp-2) have been identified at the Hp locus at chromosome 16q22,66,69 The protein products of the two Hp alleles are structurally different,66,69 and the cardiovascular effects of this Hp polymorphism play a major role in patients with diabetes mellitus.70-73 Multiple independent epidemiologic studies examining incident cardiovascular disease have demonstrated that diabetic individuals who are homozygous for the Hp-2 allele have four to five times the risk for cardiovascular events as those individuals who are homozygous for the Hp-1 allele. An intermediate risk was found in individuals who were heterozygous.70,74-76 Remarkably, these studies demonstrated that the cardiovascular risk of Hp 1 homozygosity in individuals with diabetes mellitus was not significantly different from that found in individuals without diabetes; apparently the homozygous genotype mitigates the effect of diabetes on the development of cardiovascular disease. New technology aimed at detecting TCFAs should identify IPH, iron deposition, RBC membranes, and hemosiderin deposits in macrophages. In diabetic patients, Hp genotyping may offer additional prognostic value. Summary of Plaque Composition Atherosclerotic plaques at high risk for rupture and thrombosis are characterized by several features, including large lipid core, thin fibrous cap, macrophage infiltration, positive remodeling, vasa vasorum neovascularization, and increased IPH. These concepts apply only to lesions at risk for plaque rupture and thrombosis (i.e., TCFAs). Plaques at risk for erosion and thrombosis do not exhibit any specific morphologic feature, limiting their detection by any imaging technique. This is a significant limitation of the individual, lesion-oriented approach to the highrisk, vulnerable plaque hypothesis. Nevertheless, even if only TCFAs can be identified and treated, a significant reduction of clinical events can be achieved.

PLAQUE IMAGING Having reviewed the key features of plaque vulnerability, it is now easy for the clinician to understand

1151

1152

Evaluation of Interventional Techniques

Figure 63-13. Color prints taken during the injection of silicone polymer into the coronary arteries of cleared hearts, demonstrating regions of vessels with abundant neovascularization. The positive regions are composed of networks of small-caliber vessels coming from the adventitia and penetrating the tunica media into the atherosclerotic plaque. (From Barger AC, Beeuwkes R 3rd, Lainey LL, Silverman KJ: Hypothesis: Vasa vasorum and neovascularization of human coronary arteries. A possible role in the pathophysiology of atherosclerosis. N Engl J Med 1984;310:175-177. Copyright 1984 Massachusetts Medical Society. All rights reserved.)

Figure 63-14. Histologic evidence of atherosclerotic neovascularization as a pathway for macrophage infiltration in human aortic plaques obtained at autopsy. Bicolor, contrasting immunohistochemical technique showing microvessels in crosssections (red arrows) identified with the monoclonal endothelial cell marker CD34 linked to a blue chromogen and inflammatory cells identified with a combined macrophage/T-cell marker, CD68-CD3, linked to a red chromogen. (From Fuster V, Moreno PR, Fayad ZA, et al: Atherothrombosis and high-risk plaque. Part I: Evolving concepts. J Am Coll Cardiol 2005;46:937-954.)

the basic needs of any given technique to identify TCFA. Both noninvasive and invasive techniques are evolving rapidly. The noninvasive techniques were recently updated by our group and are beyond the scope of this review.77 This section concentrates in the invasive imaging techniques, as understood in 2007. Multiple intracoronary imaging techniques are under development to identify TCFAs. Of pivotal importance, every imaging technique should have proper validation by histologic investigation. Considering that most atherosclerotic plaques (TCFA and non-TCFA) have a certain degree of fibrous cap thickness, necrotic core area, macrophage area, positive remodeling, and vasa vasorum neovascularization, simply determining the presence or absence of these features (sensitivity and specificity) is not enough. Proper histologic validation must include the accuracy of the technique to predict the extent or degree of these components, which involves linear regression analysis. These validation processes should be confirmed in animal models of TCFA before being introduced in human coronary arteries, with the ultimate test being a natural history study showing that specific plaque components detected by the technique are associated with increased clinical events. Seven novel intracoronary techniques to detect TCFA are clinically relevant: IVUS, virtual histology

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment (VH), palpography, optical coherence tomography (OCT), intravascular magnetic resonance imaging (ivMRI), angioscopy, and spectroscopy. A summary of these techniques, the component detected, and the resolution or accuracy6 is presented in Table 63-1. The first five techniques, IVUS, VH, palpography, OCT, and ivMRI, are already available in the catheterization laboratory or are under active evaluation in humans. As a result, these five techniques require a critical approach. For the remaining two techniques, angioscopy and spectroscopy, a summarized, evidence-based approach is presented. The interventionalist must develop a critical approach to evaluate these novel techniques, understanding their potential but most importantly discerning their multiple limitations before considering them for clinical use. Intravascular Ultrasound After almost 40 years of evaluating the disease by lumen obstruction (angiography), IVUS allowed proper visualization the disease itself, and provided cross-sectional imaging of atherosclerotic plaques in vivo.48-50 IVUS is based on transmitting and receiving high frequency sound waves from tissue through a low profile catheter (∼1 mm), reaching a radial resolution between 100 and 250 µm.51 With the years, IVUS became a very friendly tool; it is safe, quick, and easy. Most importantly, IVUS allows you to identify hemodynamically significant lesions that may be underestimated by angiography. This is the most common indication for IVUS in clinical practice. In addition, it provides the degree of calcification, plaque density, and most importantly, the degree of arterial remodeling. In addition, IVUS can document and provide information about plaque regression. As a result, IVUS is a fundamental tool for the interventionalist interested in vulnerable plaque. Several studies have reported the IVUS characteristics of culprit lesions,78 and the presence of multiple ruptured plaques in patients with acute coronary events.79 Now we will review the ability of IVUS to identify TCFA, which can be summarized as follows.

Fibrous Cap Thickness

Considering that the resolution of IVUS is greater than that needed to detect TCFA, IVUS-based efforts at quantifying cap thickness will always provide information that overestimates the expected values by histology. For example, Ge and coworkers carefully quantified IVUS-derived fibrous cap thickness in ruptured and nonruptured plaques from 144 consecutive patients with angina pectoris.80 IVUS-derived ruptured plaques showed thinner caps compared with nonruptured plaques, with a mean cap thickness of 0.47 mm (470 µm), as shown in Figure 63-15. However, when ruptured plaques were evaluated by histology, the mean cap thickness was about 20 times lower, 23 ± 19 µm in the coronary arteries15 and 34 ± 16 µm in the aorta.16 As discussed earlier, this significant overestimation of cap thickness by IVUS is related to its poor axial resolution (100 to 200 µm), an inherent limitation impossible to overcome. As a result, it is very unlikely that IVUS or any IVUSrelated technology will detect TCFA, the more common form of high-risk, vulnerable plaque. Necrotic Core Area

Peters and colleagues evaluated the ability of IVUS to detect human coronary necrotic cores using in vitro video densitometry with a 30-MHz ultrasound catheter. Pixel gray-level distributions were represented as frequency histograms. The sensitivity of IVUS for necrotic core was 46%, and the specificity was 97%.81 Several studies were then performed in attempts to improve these results with the use of an integrated backscatter approach.82 The most recent study, by Sano and associates, showed encouraging results.83 Of clinical relevance, a three-vessel prospective IVUS study evaluated the association between large echolucent areas (possible necrotic cores) and risk of future coronary events in 106 patients, 80% of whom had chronic stable angina.84 Twelve patients had events 4 ± 3 months after IVUS evaluation. Most of these events (93%) occurred in plaques with large echolucent areas, suggesting a prognostic value for IVUS to predict future coronary events. Nevertheless, the total number of echolucent areas detected at baseline was not reported in the study, limiting the interpretation of the results.84 We conclude that,

Table 63-1. Summary of Current Invasive Detection Technologies Technology

Component Detected

Resolution/Accuracy

Intravascular ultrasound (IVUS) IVUS-Virtual Histology Optical coherence tomography (OCT) IVUS-Elastography Spectroscopy Thermography Angioscopy

Remodeling, calcium Calcific necrotic core, fibrous tissue, fibrofatty tissue, calcium Necrotic core, fibrous cap thickness, macrophages Plaque strain Necrotic core, fibrous cap thickness, macrophages Metabolic activity of the plaque Surface appearance of the plaque

100-250 µm 240-480 µm 5-20 µm 100-250 µm NA 0.05ºC accurate In vivo surface evaluation

NA, not applicable. Modified from Granada JF, Kaluza GL, Raizner AE, Moreno PR: Vulnerable plaque paradigm: Prediction of future clinical events based on a morphological definition. Catheter Cardiovasc Interv 2004;62:364-374.

1153

Evaluation of Interventional Techniques Intima Media Lumen Rupture

Fibrous cap Lipid

Calcification

Ultrasound catheter

Rupture

Lumen

Lipid 4.1 mm2

8

P .001

6 4 2 0 Rupture No rupture (n
32) (n
108)

2

Cap 0.7 mm P .01

1

0 Rupture No rupture (n
32) (n
108)

Lipid/plaque ratio (mm2)

10

Thickness of the fibrous cap (mm)

Fibrous cap

Lipid core size (mm2)

1154

100

Lipid 38%

80

P .001

60 40 20 0 Rupture No rupture (n
32) (n
108)

Figure 63-15. Intravascular ultrasound (IVUS) images of ruptured plaques, highlighting the fibrous cap and a large echolucent area under the cap suggestive of a large necrotic core (upper panel). Lipid area, cap thickness, and lipid percent area in ruptured versus nonruptured plaques are shown (lower panel). (Redrawn from Ge J, Chirillo F, Schwedtmann J, et al: Screening of ruptured plaques in patients with coronary artery disease by intravascular ultrasound. Heart 1999;81;621-627.)

although IVUS provides useful information about plaque echogenicity, the exact sensitivity and accuracy to identify necrotic cores are still unclear. The overall sense is that the resolution may not be enough to properly quantify this important feature of plaque vulnerability. Plaque Inflammation

Detection of macrophages within the fibrous cap requires a resolution within 10 to 20 µm. Considering that IVUS resolution is 10 to 20 times higher than that, it is impossible for IVUS to detect macrophages in atherosclerotic plaques.

positive remodeling in culprit lesions responsible for acute coronary events.85 IVUS-derived arterial remodeling allows understanding of the paradoxical relation between lumen and plaque size in ACS. It is clear that ruptured plaques are larger than fibrocalcific nonruptured plaques.16,78 However, the degree of arterial remodeling is so pronounced that even large plaques can appear as nonobstructive, “mild” stenosis on angiography.78 No other imaging modality can quantify remodeling better than IVUS. Therefore, IVUS will stand high in the search for the ideal combination technique that eventually can detect all features of plaque composition in the search for TCFAs.

Degree of Positive Remodeling

Contrary to the previous features of plaque composition, IVUS is an excellent tool to detect remodeling, a major feature of plaque vulnerability. Multiple studies have documented an increased prevalence of

Plaque Neovascularization

IVUS has the potential to detect flow within the plaque and therefore to identify functional neovessels. Whereas real-time IVUS is limited in the evalu-

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment ation of plaque perfusion, recent developments with contrast agents have dramatically improved the quality of Doppler ultrasound. Intravascular injection of microbubbles (i.e., small encapsulated air or gas bubbles) can boost the Doppler signal from blood vessels. Microbubbles can help in visualizing flow in smaller vessels, even at the capillary level, as previously shown within the myocardium using contrast-enhanced echocardiography (CEE).86,87 Direct visualization of atherosclerotic plaque microvessels using CEE was successfully done by Feinstein in carotid lesions before endarterectomy.88 In coronary arteries, IVUS-CEE successfully identified plaque neovessels with spatiotemporal changes and enhancement-detection techniques89,90 (Fig. 63-16). To improve resolution, an IVUS prototype using “harmonic” imaging with transmission of ultrasound at 20 MHz (fundamental) and detection of contrast signals at 40 MHz (second harmonic) was developed.91 Experiments showed improved detection of small vessels in harmonic mode relative to fundamental mode. Harmonic imaging improved contrast resolution outside the aortic lumen in atherosclerotic rabbits, consistent with the detection of adventitial microvessels.91 Most importantly, these microvessels were not detected in fundamental imaging mode, suggesting that harmonic imaging is needed for detection and quantification of atherosclerotic neovascularization. Virtual Histology Considering the significant limitations of traditional IVUS imaging for identification of necrotic cores, Nair and Vince at The Cleveland Clinic decided to evaluate the ultrasound scattered reflection wave as a possible alternative to improve tissue characterization by IVUS.92,93 This backscattered reflection wave is received by the transducer, where it is converted into voltage. This voltage is known as the backscattered radiofrequency (RF) data. Using a combination of previously identified spectral parameters, the researchers developed a classification scheme to construct an algorithm to test plaque composition ex vivo. Four major plaque components were tested: fibrotic tissue, fibrofatty tissue, calcific-necrotic core, and calcium. A color was assigned for each of these components and was displayed on the IVUS image (Fig. 63-17). Initial validation was performed on ex vivo human coronary specimens using 30-MHz, 2.9-Fr, mechanically rotating IVUS catheters (Boston Scientific Corp, Natick, MA). The Movat-stained histologic images identified homogeneous regions representing each of the four plaque components (Fig. 63-18). The unit of analysis (also called the “box”) was initially composed of 64 backscattered RF data samples 480 µm in length.93 In 2007, the unit of analysis comprises 32 backscattered RF data samples 240 µm in length (D.G. Vince, personal communication). The algorithm developed was validated ex vivo, with sensitivities and specificities between 79%

Figure 63-16. Differential intravascular ultrasound images to identify vasa vasorum, showing the subtraction of postinjection signals from baseline signals. A, Black and white image. B, Color coding applied to the image in panel A. C, Thresholding shows the most significant areas of enhancement. (Adapted from Vavuranakis M, Kakadiaris IA, O’Malley SM, et al: Images in cardiovascular medicine: Detection of luminal-intimal border and coronary wall enhancement in intravascular ultrasound imaging after injection of microbubbles and simultaneous sonication with transthoracic echocardiography. Circulation 1005;112:e1-e2.)

1155

1156

Evaluation of Interventional Techniques

Figure 63-17. Color-coded reproduction of intravascular ultrasound Virtual Histology (IVUS-VH) plaque composition displayed in vivo in the cardiac catheterization laboratory at Mount Sinai Hospital, New York, New York.

and 93% for all four-plaque components.93 Even though the initial work was performed with a 30MHz catheter, the current Food and Drug Administration (FDA)-approved catheter (Eagle-Eye Gold) is a 20 MHz device. Nevertheless, a 45-MHz catheter is under testing and may be available for commercial use soon. VH gained a lot of attention recently because of the Providing Regional Observations to Study Predictors of Events in the Coronary Tree (PROSPECT) Trial, the first prospective natural history study designed to evaluate the prognostic value of IVUS-VH-derived plaque composition in nonobstructive segments. Seven hundred patients with ACS underwent three-vessel IVUS-VH after successful PCI. Enrollment was completed by mid-2006 and a preliminary report on follow-up will be available by the end of 2007. Despite all of these promising facts on IVUS-VH, the interventionalist needs objective information regarding the ability to detect TCFA, which can be summarized as in the following paragraphs. Fibrous Cap Thickness

Considering that the resolution of IVUS-VH is greater than that needed to detect TCFA, IVUS-VH is limited in cap thickness evaluation. This was elegantly addressed in the initial publication by Nair and Vince93: “The window size currently applied for selection of regions of interest and eventual tissue map reconstructions is 480 microns in the radial direction. Therefore, detection of thin fibrous caps (≤65 microns below the resolution of IVUS) would be compromised, restricting the detection of vulnerable atheromas.” Despite this inherent limitation, investigators have proposed a classification of “IVUS-VHderived TCFA.”94,95 As a result, lesions with fibrous cap thickness greater than 65 µm will be incorrectly

classified as TCFAs, and the total number per patient will be overestimated. Still, if these data demonstrate prognostic value, we may have a useful tool for potential clinical use. Necrotic Core Area

IVUS-VH was initially developed to identify calcific necrotic cores. However, the incidence and degree of calcification in necrotic cores is variable, and necrotic cores without calcification may not be properly identified.28 Most importantly, the majority of advanced atherosclerotic lesions display a certain degree of necrotic core. As a result, when validating necrotic core using IVUS-VH, not only the presence/absence of the necrotic core is important (sensitivity/specificity)96 but also the area. IVUS-VH routinely reports necrotic core area (mm2) and percent of total plaque area when used in patients. However, as of yet, no proper validation of these areas in patients (linear regression analysis) has been published. The only study that performed correlations between necrotic core areas by IVUS-VH and histology was done in the swine model, and the regression curves showed no correlations at all.97 As discussed earlier, multiple pathologic studies have established the concept that necrotic core areas from patients with ACS are larger than those from patients with chronic stable angina.2,14,20 Conversely, fibrous plaque areas (collagen) have been found to be significantly lower in ACS patients.22 Recently, Surmely and coworkers quantified necrotic core and fibrous areas in patients with ACS using IVUS-VH and compared them with those of patients with chronic stable angina.98 Necrotic core areas were significantly lower in patients with ACS (6.8 mm2 ± 6% versus 11 mm2 ± 8%; P = .02). In addition, fibrous areas were

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

Figure 63-18. Definitions of the various plaque components obtained with intravascular ultrasound Virtual Histology (IVUS-VH). (Courtesy of D.G. Vince, The Cleveland Clinic, 2006).

higher in patients with ACS (66 mm2 ± 11% versus 61 mm2 ± 9%; P = .03). The authors concluded that plaque composition obtained by IVUS-VH is in contradiction to previously published histopathologic data.98 Contrary to this observation, RodriguezGranillo and colleagues identified larger necrotic areas in ruptured plaques99 and in nonculprit lesions100 from patients with ACS. No IVUS-VH coronary autopsy studies have characterized lesions by clinical syndrome, and proper histopathologic validation of IVUS-VH by clinical syndrome is urgently needed to elucidate this controversy. Plaque Inflammation

As discussed earlier, detection of macrophages within the fibrous cap requires a resolution within 10 to 20 µm. Considering that IVUS-VH resolution is at least 10 to 20 times higher, it is impossible for VH to detect macrophages in the fibrous cap of the plaque.

Degree of Positive Remodeling

IVUS is an excellent tool to detect remodeling, and IVUS-VH should preserve this advantage. As reviewed earlier, positive remodeling is related to large necrotic core areas, more frequently seen in patients with ACS. Plaques with negative or constrictive remodeling are associated with smaller necrotic core areas, usually seen in patients with chronic stable angina. Recent studies evaluated IVUS-VH necrotic core areas in plaques with positive and negative remodeling and found lower necrotic core areas in positively remodeled plaques.78 Other investigators101 confirmed these data. However, Rodriguez-Granillo and colleagues published opposite data showing larger IVUS-VH necrotic core areas in positively remodeled plaques, with an impressive correlation (r = .83; P < .0001).102 It is difficult to reconcile these controversial findings and therefore impossible to elucidate, at this time, the real clinical value of IVUS-VH-derived plaque composition in clinical practice. Finally, vasa vasorum neovascularization and IPH require very

1157

Evaluation of Interventional Techniques

3.0 Macrophages

Strain in %

2.5 2.0 1.5 1.0 0.5 0.0

None Minor Medium Heavy

3.0 SMCs

2.5 Strain in %

1158

2.0 1.5 1.0 0.5 0.0

None Minor Medium Heavy

Figure 63-19. Vulnerable plaque (arrows) marked by intravascular ultrasound (A), elastography (B), macrophage staining (C), and collagen staining (D). In the elastogram, a vulnerable plaque is indicated by a high strain on the surface. In the corresponding histologic images, a high number of macrophages are visible in C, with a thin cap and a lipid pool (LP) visible in D. Box plots correlate the amount of macrophages (upper right) and the smooth muscle cell (SMC) content (lower right) with the level of strain. High levels of strain are associated with significant increases in macrophage content and significant decreases in SMC content. (Redrawn from Schaar JA, De Korte CL, Mastik F, et al: Characterizing vulnerable plaque features with intravascular elastography. Circulation 2003;108:2636-2641.)

sophisticated technology and cannot be identified by IVUS-VH. Palpography The stress/strain relationship on coronary lesions may play a significant role in plaque rupture and can be identified by another IVUS-derived technique called elastography or palpography.17 Changes in blood pressure (stress) can induce deformation of the fibrous cap (strain) that can be quantified and displayed in a color-coded scale. Purple indicates a lowstrain region that is hard, stiff, and rigid, whereas yellow indicates a region of high strain that is soft, deformable, and therefore potentially suitable for plaque rupture.95 Studies at The Thoraxcenter have documented high sensitivity and specificity of this technique, with a deformation of more than 2% reflecting increased macrophage infiltration and reduced smooth muscle cell and collagen content3 (Fig. 63-19). Therefore, this technique may provide valuable information to detect TCFAs. To risk-stratify patients, The Thoraxcenter developed a scale called the Rotterdam Classification, which divides the strain into four subclasses, the worst being ROC IV, with a deformation of greater than 1.2%. Three clinical observations are consistent95: (1) the number of high-

strain spots rises in parallel with the level of Creactive protein; (2) strain is higher in ST-segment elevation myocardial infarction than in unstable or stable angina; and (3) aggressive treatment using statins, angiotensin-converting enzyme inhibitors, clopidogrel, and aspirin can reduce, over a period of 6 months, the intensity and frequency of these highstrain spots. This specific technique was not designed to evaluate any of the other components of TCFA. As a result, fibrous cap thickness, necrotic core, degree of remodeling, neovascularization, and IPH are not suitable for evaluation by palpography. Optical Coherence Tomography A novel high-resolution intravascular imaging technique, optical coherence tomography (OCT), offers great promise to identify TCFAs. It is catheter-based, measures back-reflected infrared light, and provides the highest resolutions of all invasive modalities (5 to 20 µm).103 Excellent histopathologic correlations have been performed, both for human coronary tissue and for animal models, highlighting a sensitivity and specificity of 92% and 94%, respectively, for lipid-rich plaque; 95% and 100% for fibrocalcific plaque; and 87% and 97% for fibrous plaque.103

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

Figure 63-20. In vivo optical coherence tomography (OCT) images of various coronary plaque types compared with intravascular ultrasound (IVUS) images of the same sites. A, Fibrous plaque. From 9 o’clock to 2 o’clock, the three-layer structure of a typical intimal hyperplasia is shown, and a magnified area is shown in the inset. A homogeneous, signal-rich pattern indicates fibrous plaque (F), which is partly obscured by a guidewire artifact (*). a, adventitia; i, intima; m, media. B, The IVUS image corresponding to A. C, Calcific plaque. A signal-poor region surrounded by sharp borders represents calcific plaque, which is clearly delineated on the OCT image (arrows). D, On the corresponding IVUS image, calcium is easily identified (arrows), but the strong signal obscures the structure in front of the calcium deposit, and a back-shadow artefact obscures that behind the deposit. E, Lipid-rich plaque. A signal-poor region (arrow in inset) surrounded by diffuse borders and separated by a thin cap (arrowheads in inset) is consistent with thin-cap fibroatheroma (TCFA). F, The corresponding IVUS image suggests a superficial echolucent region. (From Low AF, Tearney GJ, Bouma BE, Jang IK: Technology Insight: Optical coherence tomography—current status and future development. Nat Clin Pract Cardiovasc Med 2006;3:154-162.)

Superb resolution allows for detailed imaging (Fig. 63-20). The major limitation of OCT is the need to displace blood in the vessel with saline flush, which makes this technique difficult in the evaluation of long segments. However, recent advances using optical frequency domain imaging (OFDI) allow for high-speed comprehensive imaging, scanning up to 5 cm with a single saline flush (Fig. 63-21).104 Considering all these promising facts, the interventionalist needs objective information regarding the ability of OCT to detect TCFAs, which can be summarized as in the following paragraphs.

Fibrous Cap Thickness

OCT is the only imaging tool that can identify plaques with cap thickness of 65 µm or less. This was demonstrated with histologic studies using proper linear regression analysis, which showed excellent correlations between OCT and ocular micrometry (light microscopy; r = .89)105 as shown in Figure 63-22. Subsequently, in vivo studies with patients were performed to identify TCFA. Fibrous cap thickness was lowest in patients with AMI, intermediate in those with ACS, and highest in those with chronic stable angina106 (Fig. 63-23). Using this analysis, OCT can

1159

1160

Evaluation of Interventional Techniques

Figure 63-21. In vivo optical frequency domain images (OFDI) of a coronary stent deployed in the swine model. Normal endothelium is seen in red, dissections induced by the balloon during stent deployment in white, and stent struts in blue. Two sections of the image are enlarged below. (From Yun SH, Tearney GJ, Vakoc BJ, et al: Comprehensive volumetric optical microscopy in vivo. Nat Med 2006;12:1429-1433.)

500 450 400 350 300 Necrotic core

250 200 150 250 µm

A

y
1.02x  3.8 r
0.89 p 0.0001

100 50 0 0

C Necrotic core

B

250 µm

estimate the incidence of TCFA according to the clinical syndrome, as also illustrated in Figure 63-23. Necrotic Core Area

Because lipid pool and necrotic core are signal-poor, they are poorly delineated with respect to the sur-

100

200 300 OCT measurement (m)

400

500

Figure 63-22. Optical coherence tomography (OCT) evaluation of fibrous cap thickness. A, Minimal fibrous cap thickness measuring 44.1 µm by OCT (black arrow), obtained from plaque illustrated in B. B, Minimal fibrous cap thickness measuring 40.4 µm by histology (black arrow). Necrotic core is visualized underneath the fibrous cap. C, Linear regression analysis showing an excellent correlation between OCT and histology measurements in 29 human atherosclerotic plaques. (From Jang IK: Optical coherence tomography: Studies at MGH. Presented to the Transcatheter Cardiovascular Therapeutics Convention, Washington, DC, 2002.)

rounding tissue. The lipid-rich plaque can be recognized, therefore, by the presence of large areas of ill-defined, signal-poor regions that are evident to the naked eye, as seen in Figure 63-20E, F. On the other hand, when the cap is thick and the signal is strong, the operator can tell that the signal is coming from a fibrous plaque composed mostly of collagen, as

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment 160 *P .02 106.14

120 100 80

* 65.87

60 40

* 70.26

TCFA percent

Min. cap thickness

140

20 0

90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

*P .01

* 72% * 50%

AMI ACS SAP 20%

Figure 63-23. In vivo quantification of fibrous cap thickness by optical coherence tomography (OCT). Cap thickness was lowest in patients with acute myocardial infarction (AMI), intermediate in patients with acute coronary syndrome, and highest in patients with chronic stable angina. As a result, the incidence of thin-cap fibroatheroma was highest in AMI and lowest in chronic stable angina. (Redrawn from Jang IK, Tearney GJ, MacNeill B, et al: In vivo characterization of coronary atherosclerotic plaque by use of optical coherence tomography. Circulation 2005;111:1551-1555.)

illustrated in Figure 63-24 (see also Figs. 63-20A,B and 63-22). Of note, linear correlations of human coronary plaque collagen were recently performed, showing regression plots between OCT and measured collagen with a correlation value of 0.475 (P < .002); the predictive values for collagen were between 89% and 93%.107 Nevertheless, it is important to highlight that OCT has not validated necrotic core areas as well as cap thickness and collagen.

Plaque Inflammation

As discussed for cap thickness, OCT resolution allows for proper identification of macrophages in the atherosclerotic plaques.103 The first correlation in vitro was performed by Tearney and colleagues,108 who identified multiple strong back-reflections from caps with abundant macrophage infiltration, resulting in a relatively high OCT signal variance. This signal variance was processed using logarithmic transfor-

250 µm 35

G

30

A

D

% CD68 staining

25 20 15 10 5 0 0.02 5

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

10 OCT NSD (raw signal)

35

H

30

E

25 % CD68 staining

B

20 15 10 5 0 0.05 5

C

F

0.06

0.07

0.08

0.09

0.1

0.11

0.12

OCT NSD (base 10 logarithm)

Figure 63-24. Raw (A) and logarithm base 10 (B) optical coherence tomographic (OCT) images of a fibroatheroma with a low density of macrophages within the fibrous cap. C, Histologic section corresponding to A and B (CD68 immunoperoxidase stain; original magnification ×100). Raw (D) and logarithm base 10 (E) OCT images of a fibroatheroma with a high density of macrophages within the fibrous cap. F, Histologic section corresponding to D and E (CD68 immunoperoxidase; original magnification ×100). Correlation between the raw (G) and logarithm base 10 (H) OCT NSD and CD68 percent area staining (diamonds, NSD data; solid line, linear fit). (From Tearney GJ, Yabushita H, Houser SL, et al: Quantification of macrophage content in atherosclerotic plaques by optical coherence tomography. Circulation 2003;107:113-119.)

1161

Evaluation of Interventional Techniques

Stable Angina

Post-MI T LP

LP

LP

LP Lumen Lumen LP

8 7 6 5 4 3 2 1 0

*

*P .002

Rupture Non-rupture

Density (%)

Density (%)

1162

*

7 6 5 4 3 2 1 0

*P .001

Culprit Remote

STEMI

ACS

mation (see Fig. 63-24A-F). Linear regression analysis then showed correlations for raw and logarithmic OCT-derived and immunostained macrophages of 0.84 (P < .0001) and 0.47 (P < .05), respectively (see Fig. 63-24G,H). In vivo studies subsequently showed increased macrophage density in ruptured plaques from patients with ACS (at both culprit and nonculprit lesions), when compared with nonruptured plaques from patients with chronic stable angina109 (Fig. 63-25). Degree of Positive Remodeling

OCT has limited penetration and requires a bloodless field to obtain good images. Proper quantification of remodeling results from multiple images with precise identification of normal vessel wall segments proximal and distal to the stenosis, as discussed earlier. Considering that OCT images are difficult to obtain and plaque penetration is limited, OCT may not properly quantify remodeling. This is a significant limitation of OCT. Plaque Neovascularization and Intraplaque Hemorrhage

OCT is the gold standard to quantify neovascularization and the effects of novel anti-angiogenic therapies in other common diseases such as agerelated macular degeneration, extrafoveal choroidal neovascularization, and proliferative diabetic retinopathy.110-112 However, at the present time, no studies have tested OCT in the evaluation of athero-

SAP

Figure 63-25. Top, Optical coherence tomographic (OCT) images of plaques from patients with stable angina and postmyocardial infarction (post-MI). LP, lipid pool. Note increased signal area at the fibrous cap in post-MI figure (inset detail), highlighting areas of increased macrophage density. Bottom left, Increased macrophage density in ruptured versus nonruptured plaques. Bottom right, Macrophage density in culprit versus remote plaques from patients with STsegment elevation myocardial infarction (STEMI), acute coronary syndromes (ACS), and stable angina pectoris (SAP). Macrophage density is higher in both plaques (culprit and remote) among patients with STEMI and ACS, compared to patients with SAP. (From MacNeill BD, Jang IK, Bouma BE, et al: Focal and multi-focal plaque macrophage distributions in patients with acute and stable presentations of coronary artery disease. J Am Coll Cardiol 2004;44:972-979.)

sclerotic neovascularization or IPH. Nevertheless, it is likely that OCT will reproduce its resolution and clinical value in ophthalmology if tested to quantify atherosclerosis neovascularization in the cardiac catheterization laboratory. Intravascular Magnetic Resonance Imaging MRI is being increasingly recognized as a potential approach for the quantitation of atherosclerotic plaque burden and lesion composition.77 MRI allows for three-dimensional evaluation of vascular structures with outstanding depiction of various components of the atherothrombotic plaque, including lipid, fibrous tissue, calcium, and thrombus formation.113 In addition, MRI combined with cellular and molecular targeting is providing important data on the biologic activity of high-risk, vulnerable plaques, especially for carotid and aortic lesions.114 However, MRI characterization of coronary plaques is considerably more difficult. The deep intrathoracic position of the coronary arteries (4 to 10 cm from the surface), the smaller dimensions, and the tortuous and irregular course of these vessels under continuous movement further exacerbate the problem, resulting in a reduction in image quality.115 Whereas conventional MRI resolution is approximately 460 µm, the resolution of intravascular MRI (ivMRI) is improved to 250 µm. As a result, ivMRI may provide valuable information in plaque charac-

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment terization of coronary lesions. Pursuing this effort, ivMRI has undergone successful testing and is currently under aggressive clinical testing in the United States and Europe, with more than 100 patients enrolled. Therefore, MRI requires careful attention regarding the ability to detect TCFA, which can be summarized as in the following paragraphs. Fibrous Cap Thickness and Necrotic Core

As stated earlier, the resolution of ivMRI is higher than 65 µm. As a result, ivMRI is limited in the identification of TCFA. To overcome this limitation, the luminal surface of the plaque was simultaneously evaluated in two separate bands: a superficial, luminal band of 0 to 100 µm and a deeper band of 100 to 250 µm.116 TCFA was defined as the presence of an increased lipid fraction within the superficial band, which in turn denotes the presence of a thin fibrous cap, and increased lipid in the deep band, which indicates the presence of a necrotic core or an increased concentration of lipid-rich cells. Absence of lipid within the superficial band may be indicative of a thick fibrous cap, which is associated with more stable lesions.116 With the aid of a partially inflated balloon, the MRI catheter obtains images in four quadrants (Fig. 63-26). Within each quadrant, the percentage of lipid is assessed in both the superficial and deep bands simultaneously, and the data are integrated to produce a circular, color-coded display. The resulting image represents the sum of the four superficial and four deep bands from the four quadrants. Validation studies were performed using a total of 34 human plaques (aortic and coronary). Examples are shown in Figure 63-27. Results of histologic analysis in addition to the aortic data resulted in a sensitivity of 100% and a specificity of 89%.116

Plaque Inflammation

With the advantage of molecular targeting, MRI can image macrophages using various pathways.117 Several compounds, contrast agents, and nanoparticles are now available to target different molecules within macrophages.118 The ultrasmall superparamagnetic particles of iron oxide compounds (USPIOs, SPIOs), also called magnetic nanoparticles, are internalized by macrophage-receptors and can be properly imaged by MRI, as shown in Figure 63-28. In addition to macrophage-receptor uptake, inflammatory cells can be imaged by targeting metabolic and proteolytic activity. The combination of MRI and 18fluorodeoxyglucose positron emission tomography (FDG-PET) successfully images metabolic activity of plaque macrophages in patients with symptomatic carotid atherosclerosis (Fig. 63-29).119,120 Proteolytic activity can be identified by MRI targeting cathepsin-B, MMPs, and myeloperoxidase, as recently reviewed.118 Contrast agents including gadolinium-containing immunomicelles can also identify inflammatory cells in atherosclerosis.121 Finally, imaging of interleukin-2 presence122 and of apoptosis123 may also provide an estimation of macrophage content in atherosclerotic plaques. Of note, macrophage imaging by MRI is available only as a noninvasive tool. To become useful for coronary plaque imaging, ivMRI will require higher resolution and signal-to-noise ratios than can be obtained with ivMRI coils.116 Alternatively, newer magnetic nanoparticles, such as those that permit concomitant near-infrared fluorescence imaging,124 may allow for sequential optical imaging of coronary macrophages, also via intravascular catheters.118 Degree of Positive Remodeling

The evaluation of remodeling requires perfect delineation of the external elastic lamina, both at the site

MRI Diagram

1

4

MR probe

3

4

1

3

2

2

250µ

100µ

Figure 63-26. Left, Depiction of the magnetic resonance (MR) imaging catheter with imaging areas superimposed on a cross-section of a human coronary artery. Interrogation of the arterial wall takes place in four quadrants, each comprising a field of view. A single field of view is denoted by the white arrowhead. Right, The magnetic resonance imaging (MRI) diagram displays the lipid fraction in each quadrant assessed by the catheter. In this particular illustration, an increased lipid concentration is noted only in quadrant 3, which displays yellow. Quadrants 1, 2, and 4 are shown in blue, indicating a low lipid content or increased fibrous content.

1163

1164

Evaluation of Interventional Techniques

Movat’s

CD-68

Lipid fraction (%) 1 4 250µ

A

*

2 3 Thin cap fibroatheroma

B

5

Thick cap fibroatheroma

C

Stable plaque

Figure 63-27. The magnetic resonance imaging (MRI) scan demonstrates excellent correlation with histology. Coronary angiograms, MRI scans, and histologic cross-sections of three intermediate coronary lesions are shown. An arrow on each angiogram marks the site of interrogation. The corresponding MRI is shown in the second column, and histologic sections of the interrogated sites are shown in the third and fourth columns, with Movat’s pentachrome and anti-CD68 antibody staining, respectively. A, left to right, Thin-cap fibroatheroma in the proximal left anterior descending artery. The MRI display shows the presence of a high lipid content within three quadrants (2 to 4). Quadrant 1 has little lipid within the wall, as indicated by the lack of foam cells by Movat’s staining or macrophages by CD68 staining. Quadrant 2 has moderately increased lipid concentrations, as noted by an approximate lipid fractional index of 60%. Quadrant 3 has increased lipid only in the deep layer, whereas quadrant 4 has high lipid fractional indexes (±100%) within the superficial and deep layers. Approximately 75% of the arterial circumference is lipid-rich. The MRI display corresponds well with subsequent histology, because the Movat-stained section shows a large necrotic core (*) and a thin fibrous cap, and the adjoining immunohistochemical staining is markedly positive for CD68 in the area corresponding to quadrant 4 of the MRI display. B, Thick-cap fibroatheroma in the right coronary artery. The MRI display shows no lipid content within the superficial layer (blue); however, a mild degree of increased lipid concentration is observed within the deep band (>100 µm from the lumen) in quadrant 5 only. The lipid fractional index is about 50%. The corresponding histologic section shows a thick-cap fibroatheroma with a small, deep necrotic core (+), which is confirmed by the anti-CD68 staining, corresponding with the MRI image. Because there is little to no lipid within the superficial layer, this lesion is considered a thick fibroatheroma. C, Stable lesion in the intermediate branch of the left coronary artery. A mild stenosis is seen by angiography. The MRI display of the lesion shows no increased lipid concentration in the shallow or deep bands of any quadrant, indicating the presence of a fibrous lesion (hence, blue display). This diagnosis was confirmed by histology to be adaptive intimal hyperplasia, and the corresponding anti-CD68 staining was negative for foam cells or a necrotic core. (From Schneiderman J, Wilensky RL, Weiss A, et al: Diagnosis of thin-cap fibroatheromas by a self-contained intravascular magnetic resonance imaging probe in ex vivo human aortas and in situ coronary arteries. J Am Coll Cardiol 2005;45:1961-1969.)

of the plaque and at the reference segment. The current stage of ivMRI does not provide this degree of delineation and therefore cannot quantify remodeling. However, ongoing coronary studies with ivMRI document the degree of remodeling using concomitant IVUS.

Plaque Neovascularization

As a general principle, molecular imaging for angiogenesis targets the endothelial cell. Proliferating endothelial cells respond to signaling by adhesion molecules, including integrins and cadherins, along with signals generated by secreted growth factors.125

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment A

B

Pre

Post

C

D

Mac

Fe

E

F

G

Figure 63-28. A through D, Cellular magnetic resonance images (MRI) of macrophage (Mac) endocytosis in clinical and experimental atherosclerosis using magnetic nanoparticles. MRIs obtained before (A) and after (B) injection of dextrinated magnetic nanoparticles (ferumoxtran, 2.6 mg/kg) show focal signal loss (arrow) within a carotid plaque of a neurologically symptomatic patient. Histologic examination of the carotid endarterectomy specimen demonstrated colocalization of macrophages, revealed by anti-CD68 macrophage antibody staining (original magnification ×100) (C), and of iron (Fe), indicated by Perls iron stain with neutral red counterstain (original magnification ×400 (D). E through G, Multimodality MRI and near-infrared fluorescent imaging of murine atherosclerosis using magnetofluorescent nanoparticles. E, In vivo 9.4-T electrocardiogram- and respiratory-gated MRI of an apolipoprotein E-deficient (apoE– /–) mouse. Injection of a clinical-type near-infrared fluorescent dextrinated magnetic nanoparticle (15 mg/kg of iron, 24-hour circulation time) produced focal signal loss (arrow) in the aortic root, a known site of atherosclerosis in the apoE–/– mouse. F, Fluorescence reflectance imaging of the resected aorta confirms a focal near-infrared fluorescent signal within the aortic root (arrow). G, On fluorescence microscopy, the near-infrared magnetofluorescent nanoparticles accumulate in intimal macrophages (red, arrow) within aortic root plaque sections (original magnification ×200). In contrast, smooth muscle cells (stained here with a spectrally distinct α-actin fluorescent antibody, green) modestly colocalize with the magnetofluorescent nanoparticles. (From Jaffer FA, Libby P, Weissleder R: Molecular and cellular imaging of atherosclerosis: Emerging applications. J Am Coll Cardiol 2006;47:1328-1338.)

As a result, molecular imaging can target these molecules, such as the integrin avβ3, which promotes angiogenesis by signaling basic fibroblast growth factor (FGF), or avβ5, which promotes angiogenesis by signaling vascular endothelial growth factor (VEGF).

Winter and colleagues evaluated atherosclerotic neovessels using avβ3-targeted paramagnetic nanoparticles in hypercholesterolemic rabbits126 (Fig. 63-30). Other molecular targets have been successfully labeled for imaging of atherosclerosis. Matter and

1165

Evaluation of Interventional Techniques

FDG-PET

CTa

Fused

A 10

FDG accumulation (rate sec1)

1166

B

P .01

5

Symptomatic

Asymptomatic

18

Figure 63-29. F-fluorodeoxyglucose positron emission tomography (FDG-PET) and computed tomography angiography (CTa) images from patients with unstable carotid disease. A, FDG-PET (left), CTa (middle) images are combined to produce a fused image (right). Top row, Patient with symptomatic carotid stenosis. Bottom row, Patient with contralateral asymptomatic carotid stenosis. The arrows highlight areas of FDG uptake, corresponding to stenotic carotid plaque. B, A graph showing FDG accumulation rate in symptomatic versus asymptomatic carotid plaques. Note that FDG uptake into symptomatic plaque was significantly higher. (From Davies JR, Rudd JH, Weissberg PL, Narula J: Radionuclide imaging for the detection of inflammation in vulnerable plaques. J Am Coll Cardiol 2006;47: C57-C68.)

colleagues127 targeted a specific antibody against the fibronectin extradomain B, which was labeled with radioiodine and infrared fluorophores and injected intravenously into atherosclerotic ApoEknockout mice. Immunohistochemical studies revealed increased expression of extradomain B, not only in murine but also in human plaques, where it was found predominantly around vasa vasorum.127 Finally, VCAM-1, a critical component of the leukocyte-endothelial adhesion cascade, was successfully targeted using phage display-derived peptide sequences and multimodal nanoparticles for MRI and fluorescence molecular imaging in ApoE-knockout mice, adding an additional method to interrogate angiogenesis in atherosclerotic plaques.128 Intraplaque Hemorrhage

The evaluation of human carotid IPH was elegantly performed by Takaya and associates,129 as shown in Figure 63-31. Of clinical relevance, IPH detected by MRI is associated with a significant increase in sub-

sequent cerebrovascular events (hazard ratio, 5.2; P = .005).130 Other investigators have confirmed these findings,131 highlighting the value of MRI-detected IPH in complex atherosclerosis. Angioscopy Direct visualization of atherosclerotic plaques can provide information about plaque composition. Angioscopic classification of coronary lesions is performed by noting the color of plaques in a bloodless field. White, yellow, and glistening yellow plaques have been observed and studied in patients with CAD. Of significant clinical value, Uchida and coworkers evaluated the clinical prospective value of these three different types of plaques in a prospective, three-vessel angioscopic study including 157 patients with chronic stable angina.132 The incidence of ACS was evaluated 12 months later. The majority of patients (75%) had white plaques, which were

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment 100 80 60 40

A

B

C

20

I

M 20µ

D

E

Av

F

Av

20µ

Figure 63-30. Molecular magnetic resonance images of arterial neovessels in cholesterol-fed rabbits. Percent enhancement of adventitial signal (false-colored from blue to red) is shown in aortic segments at the renal artery (A), in mid-aorta (B), and at the diaphragm (C) 2 hours after administration of αvβ3-targeted gadolinium-loaded nanoparticles. D, Immunohistochemistry of αvβ3 integrin demonstrates thickened intima (I) and αvβ3 integrin staining in adventitial neovessels (black arrowheads). Immunostaining in aorta from the cholesterolfed animal in A (×600) delineates neovascular αvβ3 integrin (E, solid arrows) and expression of platelet and endothelial cell adhesion molecules (F, open arrows) at the interface between media (M) and adventitia (Av). (From Winter PM, Morawski AM, Caruthers SD, et al: Molecular imaging of angiogenesis in early-stage atherosclerosis with alpha(v)beta3-integrin-targeted nanoparticles. Circulation 2003;108:2270-2274.)

associated with a low incidence of ACS (3.3%). The second group of patients (18%) had yellow plaques, which were associated with an intermediate incidence of ACS (7.6%). Finally, the third group of patients (8%) had glistening yellow plaques, which were associated with an impressive incidence of ACS (68%), including death in 22% of the cases. Using autopsy information, Uchida evaluated fibrous cap thickness for each of the three plaque types (Fig. 63-32). White plaques were associated with thick caps (400 µm), yellow plaques with thinner caps (80 µm), and glistening yellow plaques with the thinnest caps (10 to 20 µm).132 Of note, Uchida also studied patients with chronic stable angina, who had a rather low incidence of yellow plaques. Other angioscopic studies in patients with ACS have shown a higher incidence of yellow plaques. Asakura and colleagues performed threevessel angioscopy in patients 1 month after myocardial infarction.133 Yellow plaques were detected in 90% of 21 culprit lesions. Yellow plaques were equally prevalent in infarct-related and noninfarct-related coronary arteries (3.7 ± 1.6 versus 3.4 ± 1.8 plaques per artery), suggesting a diffuse rather than a localized process in patients with myocardial infarction. To evaluate the predictive value of yellow plaques in clinical practice, Ohtani and associates performed culprit vessel angioscopy in 552 patients with chronic stable angina, ACS, or AMI.134 Yellow color intensity was also graded. The number of yellow plaques varied from 0 to more than 5 (Fig. 63-33). After 5 years, 7.1% of the patients developed ACS. The mean number of yellow plaques was higher in patients with an ACS event than in those without such an event (3.1 ± 1.8 versus 2.2 ± 1.5; P = .008). However,

the yellow color intensity scale was similar and not predictive. Spectroscopy Spectroscopy is a nondestructive optical technology by which the chemical composition of plaque components can be analyzed. After irradiation of tissue with a laser beam, scattered photons are acquired to identify specific features of plaque vulnerability.135 Two different modalities are currently under active evaluation for intravascular detection of high-risk, vulnerable plaques: near-infrared and Raman spectroscopy. Both techniques have good correlations with histologic analysis of coronary and aortic tissue.136,137 However, the complexity of signal analysis may force investigators to focus on only one or two features of plaque vulnerability. Both techniques face technological challenges in obtaining reliable signals through blood in the beating heart. Although this technique is promising, spectroscopy will require a significant effort to successfully overcome its limitations. Efforts are ongoing,138 and in the near future it will be possible to know whether spectroscopy can stand as a useful clinical tool in the catheterization laboratory. Before completing the imaging section of this review, it is important to mention that thermography was extensively evaluated in multiple settings with promising results.139 However, the cooling effect of the blood and other limitations significantly reduced the enthusiasm for thermography as a potentially useful tool for intravascular detection of highrisk plaques.

1167

1168

Evaluation of Interventional Techniques

Figure 63-31. A, Signal intensity of type II hemorrhage at baseline examination. Type II hemorrhage is identified by hyperintense signals on time of flight (TOF), T1-weighted (T1W), protein density weighted (PDW), and T2-weighted (T2W) magnetic resonance images of left internal carotid artery (arrows). Asterisks show location of lumen. ECA, external carotid artery; JV, jugular vein. B, Images from 18-month follow-up scans showed a similar signal intensity pattern in the same regions (arrowheads). C, Matched Mallory’s trichrome-stained sections from excised carotid endarterectomy specimen. D, High-power (×400) field taken from region (arrow in C) deep within necrotic core, shows hemorrhagic debris and cholesterol clefts. E, Glycophorin A immunostaining of the same region (×400) in an adjacent section shows extensive staining of hemorrhagic debris, indicating the presence of erythrocyte membranes. (From Takaya N, Yuan C, Chu B, et al: Presence of intraplaque hemorrhage stimulates progression of carotid atherosclerotic plaques: A high-resolution magnetic resonance imaging study. Circulation 2005;111:2768-2775.)

Summary of Intracoronary Imaging The development of new technologies for the purpose of detection of high-risk plaques is progressing rapidly. Although individual devices have reached a certain degree of technological sophistication, a combination these modalities may have a better future (e.g., OCT plus backscattered IVUS,140 IVUS plus Raman spectroscopy141). In the recently completed PROSPECT trial, three-vessel coronary imaging was performed in patients with ACS; the findings of this trial will certainly provide prognostic information related to invasive plaque imaging in CAD. If the event rate at follow-up is higher than expected with pharmacologic therapy, the scientific community will need to consider additional therapeutic strategies. A comprehensive approach regarding invasive therapy is becoming mandatory for the interventionalist who is interested in prevention of recurrent coronary events. The next section of this review sum-

marizes current and future therapies for high-risk, vulnerable plaques.

THERAPY Systemic Therapy The treatment of high-risk, vulnerable plaques relies on aggressive medical therapy, which has been shown to reduce coronary events and improve survival. As a result, systemic therapy is the cornerstone of plaque stabilization, with documented reductions in lipid content, inflammation, and vasa vasorum neovascularization.142 Intensive statin therapy has produced a significant decrease in coronary events in patients with stable disease (Treating New Targets study [TNT] and Incremental Decrease in Clinical Endpoints Through Aggressive Lipid Lowering [IDEAL]) and in patients with ACS (Aggrastat to Zocor study [A to Z] and Pravastatin or Atorvastatin Evaluation and Infection Therapy [PROVE IT-TIMI-22]), not only in the

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

A

B

C m

m

C

a

a

a C C

A

B

m

C

Figure 63-32. Angioscopic (top row) and histologic (bottom row) appearance of plaques at the time of autopsy. A, White plaque, associated with a thick cap. B, Yellow plaque, associated with a thinner cap. C, Glistening yellow plaque, associated with the thinnest cap. See text for details. (From Uchida Y, Nakamura F, Tomaru T, et al: Prediction of acute coronary syndromes by percutaneous coronary angioscopy in patients with stable angina. Am Heart J 1995;130:195-203.)

6

Figure 63-33. Angioscopic detection and grading of color of yellow plaques. A, Representative case in which no yellow plaque was detected in the right coronary artery (RCA): number of yellow plaques, 0; maximum color grade of yellow plaques, 0. B, Representative case in which multiple yellow plaques were detected in the RCA: number of yellow plaques, 3; maximum color grade of yellow plaques, 3. (From Ohtani T, Ueda Y, Mizote I, et al: Number of yellow plaques detected in a coronary artery is associated with future risk of acute coronary syndrome: Detection of vulnerable patients by angioscopy. J Am Coll Cardiol 2006;47:2194-2200.)

6

6

6 5 5

5

5

4 3

4

2 1

3 4

3

A

2

4

1

2

1

3

B

2

1

1169

Evaluation of Interventional Techniques 1.8 Median change in percent atheroma volume, %

1170

CAMELOT placebo

1.2 0.6 0

REVERSAL atorvastatin

0.6 ASTEROID rosuvastatin

REVERSAL pravastatin

A-Plus placebo

risk patients. Classified as regional and local therapies, these new options are under aggressive investigation and will need prospective, placebo-controlled, randomized trials before being considered for clinical use. Regional Therapy

r²
0.97 P .001

1.2 50 60 70 80 90 100 110 120 Mean low-density lipoprotein cholesterol, mg/dL

Figure 63-34. Median change in percent atheroma volume according to the type of statin used. A-Plus, Avasimbe and Progression of Lesions on UltraSound; ASTEROID, A Study to Evaluate the Effect of Rosuvastatin on Intravascular UltrasoundDerived Coronary Atheroma Burden; CAMELOT, Comparison of Amlodipine versus Enalapril to Limit Occurrences of Thrombosis; REVERSAL, Reversal of Atherosclerosis with Aggressive Lipid Lowering Therapy trial. (Redrawn from Nissen SE, Nicholls SJ, Sipahi I, et al: Effect of very high-intensity statin therapy on regression of coronary atherosclerosis: The ASTEROID trial. JAMA 2006;295:1556-1565.)

incidence of coronary death or myocardial infarction143 but also in development of heart failure, independent of recurrent infarct.144 Similarly, intensive statin therapy early after an ACS was associated with clinical benefits that became evident after 4 to 12 months.145 Most importantly, in A Study to Evaluate the Effect of Rosuvastatin on Intravascular Ultrasound-Derived Coronary Atheroma Burden (ASTEROID),47 aggressive therapy with rosuvastatin (40 mg/day) led to an absolute regression of atheroma volume (Fig. 63-34). Another potent antiatherogenic therapy is increasing high-density lipoprotein (HDL). Studies using bezafibrate, a peroxisome proliferator-activated receptor (PPAR)-α agonist, and fenofibrate demonstrated reduced events and plaque regression, respectively.146,147 These beneficial effects may be due not only to HDL augmentation and reverse cholesterol transport,148 but also to recruitment of endothelial progenitor cells into damaged endothelium.149 In addition to high-dose statin therapy, angiotensin-converting enzyme (ACE) inhibitors, β-blockers, and aspirin have demonstrated reductions in death and myocardial infarction and therefore are mandatory therapy in the stabilization of high-risk, vulnerable plaques.142,150 Despite the tremendous value of systemic therapy, patients still come back with recurrent events, and therefore prove to be resistant to systemic therapy. For example, the combination of ACE inhibitors, βblockers, aspirin, and high doses of atorvastatin (80 mg per day), still yields a 22% recurrent event rate within 2 years, as was shown in the PROVE IT trial.151 As a result, even the best combination of systemic therapy available today does not successfully prevent all episodes of plaque rupture and thrombosis. Therefore, new therapies are urgently needed as coadjuvants to systemic therapy in high-

Regional therapy is defined as the intravascular treatment of coronary segments with therapeutic agents that will stabilize high-risk, vulnerable plaques. Regional therapy includes photodynamic therapy (PDT),152 endoluminal phototherapy,153,154 and cryotherapy.155 Of these, PDT has gained more attention and is discussed here. PDT uses photosensitizing (light-sensitive) drugs, light, and tissue oxygen to treat targeted diseases, mostly in the field of cancer.156 Photosensitizing agents (porphyrins) are administered locally or parenterally. They are selectively absorbed and retained within tissues for targeted therapy. This differential selectivity offers selective therapeutic effects when the target tissue is exposed to light at an appropriate wavelength; the surrounding normal tissue is then spared from therapeutic injury.153 Activation of the photosensitizer within tissue induces the production of free radicals, leading to selective cytotoxic effects, mostly apoptosis (DNA fragmentation) or delayed necrosis. The application of PDT to atherosclerotic plaques was successfully performed in vivo by Waksman and colleagues in hypercholesterolemic rabbits.157 PDT induced a significant reduction (92% ± 6%) in the population of nuclei of all cell types in plaques relative to controls (P < .01). This effect was partly due to reduction of smooth muscle cells (α-actin) and macrophages (RAM-11), as shown in Figure 63-35. These results suggest that PDT can almost eliminate macrophages from atherosclerotic plaques and may provide a therapeutic alternative for high-risk, vulnerable plaques refractory to aggressive systemic therapy. The other two therapies, endoluminal phototherapy and cryotherapy, are also under investigation, but the experience is limited and escape the scope of this review. Local Therapy Coronary stents offers the possibility of stabilizing high-risk, vulnerable plaques by thickening the fibrous cap through the formation of neointimal hyperplasia. As predicted by Libby almost a decade ago,158 and more recently highlighted by Serruys,95 “If we could identify potentially unstable atheroma before they are evident, clinically we might even contemplate angioplasty on nonsignificant stenosis to induce smooth muscle cell proliferation and reinforce the plaque fibrous cap.”158 On the same subject, Davies wrote in an editorial, “the time for prophylactic angioplasty has not come yet, but it may.”159 There are some doubts, however, that better understanding of vulnerable plaques will do much to improve preventive therapy, particularly with stents. Atherosclerosis attacks the entire coronary system,

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

Figure 63-35. Histologic micrographs of the effects of photodynamic therapy on atherosclerotic plaques from hypercholesterolemic rabbits. A, Control, showing macrophages (black arrow). B, Seven days after photodynamic therapy, few macrophages are seen. (From Waksman R: Photodynamic Therapy. New York, Taylor & Francis, 2004.)

A

and there may be multiple TCFA plaques in the same patient. As a result, local therapy to reinforce TCFA may lead to a losing game of fixing one trouble spot, then finding many others.4 The controversy associated with late stent thrombosis is another disadvantage for the potential application of coronary stents in vulnerable plaques.160 On the other hand, proponents of the stent hypothesis argue that the clinical event rate associated with TCFA (4% to 13% per year, as discussed previously) will always be higher than the clinical event rate associated with stenting (4% to 6% for the first year and approximately 1% to 2% per year thereafter). Furthermore, new stent designs including biodegradable materials161 and self-expanding delivery systems, may reduce the long-term risk of stent thrombosis and preserve the integrity of the fibrous cap, with further reinforcement by neointimal tissue.162 The prediction of how we will treat vulnerable plaques resistant to aggressive medical therapy in the future is still a matter of great controversy, but it provides a wonderful opportunity for interventional cardiologists. Only prospective, randomized clinical trials can completely elucidate this issue.

SUMMARY AND FUTURE PREDICTIONS Atherosclerosis continues to evolve in diagnosis, imaging, and therapy. The concept of a high-risk, vulnerable plaque with specific histologic features continues to be the focus of many investigators, who are developing multiple imaging modalities for invasive diagnosis and therapy. This chapter reviews the state of the art in these techniques. However, the concept of vulnerable plaques may be outdated. The fact that many lesions exhibit the same morphology simultaneously in the same patient suggests that a single-plaque approach is narrow-minded and difficult to prove in clinical practice. A broader concept of disease burden is currently evolving, leading toward quantification by noninvasive

B

imaging techniques. Only randomized clinical trials will help to elucidate the risks and benefits of the proper approach, either invasive with local devices or noninvasive with aggressive systemic therapy. REFERENCES 1. Fuster V, Voute J. MDGs: Chronic diseases are not on the agenda. Lancet 2005;366:1512-1514. 2. Fuster V, Moreno PR, Fayad ZA, et al: Atherothrombosis and high-risk plaque. Part I: Evolving concepts. J Am Coll Cardiol 2005;46:937-954. 3. Schaar JA, De Korte CL, Mastik F, et al: Characterizing vulnerable plaque features with intravascular elastography. Circulation 2003;108:2636-2641. 4. Feder B: In Quest to Improve Heart Therapies: Plaque Gets a Fresh Look. New York Times. November 27, 2006:A1. 5. Muller JE, Stone PH, Turi ZG, et al: Circadian variation in the frequency of onset of acute myocardial infarction. N Engl J Med 1985;313:1315-1322. 6. Granada JF, Kaluza GL, Raizner AE, Moreno PR: Vulnerable plaque paradigm: Prediction of future clinical events based on a morphological definition. Catheter Cardiovasc Interv 2004;62:364-374. 7. Ambrose JA, Tannenbaum MA, Alexopoulos D, et al: Angiographic progression of coronary artery disease and the development of myocardial infarction. J Am Coll Cardiol 1988;12:56-62. 8. Cutlip DE, Chhabra AG, Baim DS, et al: Beyond restenosis: Five-year clinical outcomes from second-generation coronary stent trials. Circulation 2004;110:1226-1230. 9. Glaser R, Selzer F, Faxon DP, et al: Clinical progression of incidental, asymptomatic lesions discovered during culprit vessel coronary intervention. Circulation 2005;111: 143-149. 10. Koenig W, Khuseyinova N: Biomarkers of atherosclerotic plaque instability and rupture. Arterioscler Thromb Vasc Biol 2007;27:15-26. Epub 2006 Nov 2. 11. Wang JC, Normand SL, Mauri L, Kuntz RE: Coronary artery spatial distribution of acute myocardial infarction occlusions. Circulation 2004;110:278-284. 12. Valgimigli M, Rodriguez-Granillo GA, Garcia-Garcia HM, et al: Distance from the ostium as an independent determinant of coronary plaque composition in vivo: An intravascular ultrasound study based radiofrequency data analysis in humans. Eur Heart J 2006;27:655-663. 13. Farb A, Burke AP, Tang AL, et al: Coronary plaque erosion without rupture into a lipid core: A frequent cause of coro-

1171

1172

Evaluation of Interventional Techniques

14. 15. 16.

17. 18. 19. 20. 21.

22.

23.

24.

25. 26. 27. 28. 29.

30. 31.

32.

33.

34.

35.

nary thrombosis in sudden coronary death. Circulation 1996;93:1354-1363. Virmani R, Burke AP, Farb A, Kolodgie FD: Pathology of the vulnerable plaque. J Am Coll Cardiol 2006;47:C13-C18. Burke AP, Farb A, Malcom GT, et al: Coronary risk factors and plaque morphology in men with coronary disease who died suddenly. N Engl J Med 1997;336:1276-1282. Moreno PR, Purushothaman KR, Fuster V, O’Connor WN: Intimomedial interface damage and adventitial inflammation is increased beneath disrupted atherosclerosis in the aorta: Implications for plaque vulnerability. Circulation 2002;105:2504-2511. Schaar JA, van der Steen AF, Mastik F, et al: Intravascular palpography for vulnerable plaque assessment. J Am Coll Cardiol 2006;47:C86-C91. Loree HM, Kamm RD, Stringfellow RG, Lee RT: Effects of fibrous cap thickness on peak circumferential stress in model atherosclerotic vessels. Circ Res 1992;71:850-858. Oliver MF, Davies MJ: The atheromatous lipid core. Eur Heart J 1998;19:16-18. Davies MJ: Pathophysiology of acute coronary syndromes. Indian Heart J 2000;52:473-479. Felton CV, Crook D, Davies MJ, Oliver MF: Relation of plaque lipid composition and morphology to the stability of human aortic plaques. Arterioscler Thromb Vasc Biol 1997;17: 1337-1345. Moreno PR, Bernardi VH, Lopez-Cuellar J, et al: Macrophages, smooth muscle cells, and tissue factor in unstable angina: Implications for cell-mediated thrombogenicity in acute coronary syndromes. Circulation 1996;94:3090-3097. Virmani R, Kolodgie FD, Burke AP, et al: Lessons from sudden coronary death: A comprehensive morphological classification scheme for atherosclerotic lesions. Arterioscler Thromb Vasc Biol 2000;20:1262-1275. Abela GS, Aziz K: Cholesterol crystals rupture biological membranes and human plaques during acute cardiovascular events: A novel insight into plaque rupture by scanning electron microscopy. Scanning 2006;28:1-10. Burke AP, Weber DK, Kolodgie FD, et al: Pathophysiology of calcium deposition in coronary arteries. Herz 2001; 26:239-244. Moreno PR, Purushothaman KR, Fuster V, et al: Lack of association between aortic calcification and histologic signs of plaque rupture. J Am Coll Cardiol 2000;35:303A. Moreno P: The Role of Calcification in Plaque Vulnerability and Disruption. New York, Futura, 2002. Burke AP, Joner M, Virmani R: IVUS-VH: A predictor of plaque morphology? Eur Heart J 2006;27:1889-1890. Shah PK, Falk E, Badimon JJ, et al: Human monocyte-derived macrophages induce collagen breakdown in fibrous caps of atherosclerotic plaques: Potential role of matrix-degrading metalloproteinases and implications for plaque rupture. Circulation 1995;92:1565-1569. Shah PK, Galis ZS: Matrix metalloproteinase hypothesis of plaque rupture: Players keep piling up but questions remain. Circulation 2001;104:1878-1880. Steinberg D, Witztum JL: Is the oxidative modification hypothesis relevant to human atherosclerosis? Do the antioxidant trials conducted to date refute the hypothesis? Circulation 2002;105:2107-2111. Seshiah PN, Kereiakes DJ, Vasudevan SS, et al: Activated monocytes induce smooth muscle cell death: Role of macrophage colony-stimulating factor and cell contact. Circulation 2002;105:174-180. Hutter R, Valdiviezo C, Sauter BV, et al: Caspase-3 and tissue factor expression in lipid-rich plaque macrophages: Evidence for apoptosis as link between inflammation and atherothrombosis. Circulation 2004;109:2001-2008. Rosenfeld ME: Leukocyte recruitment into developing atherosclerotic lesions: The complex interaction between multiple molecules keeps getting more complex. Arterioscler Thromb Vasc Biol 2002;22:361-363. Mallat Z, Hugel B, Ohan J, et al: Shed membrane microparticles with procoagulant potential in human atherosclerotic plaques: A role for apoptosis in plaque thrombogenicity. Circulation 1999;99:348-353.

36. Moreno PR, Falk E, Palacios IF, et al: Macrophage infiltration in acute coronary syndromes: Implications for plaque rupture. Circulation 1994;90:775-778. 37. Galis ZS: Vulnerable plaque: The devil is in the details. Circulation 2004;110:244-246. 38. Libby P: Inflammation and cardiovascular disease mechanisms. Am J Clin Nutr 2006;83:456S-460S. 39. Glagov S, Weisenberg E, Zarins CK, et al: Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 1987;316:1371-1375. 40. Varnava AM, Mills PG, Davies MJ: Relationship between coronary artery remodeling and plaque vulnerability. Circulation 2002;105:939-943. 41. Tronc F, Mallat Z, Lehoux S, et al: Role of matrix metalloproteinases in blood flow-induced arterial enlargement: Interaction with NO. Arterioscler Thromb Vasc Biol 2000;20: E120-E126. 42. Burke AP, Kolodgie FD, Farb A, et al: Morphological predictors of arterial remodeling in coronary atherosclerosis. Circulation 2002;105:297-303. 43. Schoenhagen P, Ziada KM, Kapadia SR, et al: Extent and direction of arterial remodeling in stable versus unstable coronary syndromes: An intravascular ultrasound study. Circulation 2000;101:598-603. 44. Corti R, Fayad ZA, Fuster V, et al: Effects of lipid-lowering by simvastatin on human atherosclerotic lesions: A longitudinal study by high-resolution, noninvasive magnetic resonance imaging. Circulation 2001;104:249-252. 45. Schoenhagen P, Tuzcu EM, Apperson-Hansen C, et al: Determinants of arterial wall remodeling during lipid-lowering therapy: Serial intravascular ultrasound observations from the Reversal of Atherosclerosis with Aggressive Lipid Lowering Therapy (REVERSAL) trial. Circulation. 2006;113: 2826-2834. 46. Sipahi I, Nicholls SJ, Tuzcu EM, Nissen SE: Coronary atherosclerosis can regress with very intensive statin therapy. Cleve Clin J Med 2006;73:937-944. 47. Nissen SE, Nicholls SJ, Sipahi I, et al: Effect of very high-intensity statin therapy on regression of coronary atherosclerosis: The ASTEROID trial. JAMA 2006;295:15561565. 48. Moreno PR, Purushothaman KR, Zias E, et al: Neovascularization in human atherosclerosis. Curr Mol Med 2006;6: 457-477. 49. Simons M: Angiogenesis: Where do we stand now? Circulation 2005;111:1556-1566. 50. Carmeliet P: Angiogenesis in health and disease. Nat Med 2003;9:653-660. 51. Moreno PR, Purushothaman KR, Sirol M, et al: Neovascularization in human atherosclerosis. Circulation 2006;113:22452252. 52. Barger AC, Beeuwkes R 3rd, Lainey LL, Silverman KJ: Hypothesis: Vasa vasorum and neovascularization of human coronary arteries. A possible role in the pathophysiology of atherosclerosis. N Engl J Med 1984;310:175-177. 53. Kumamoto M, Nakashima Y, Sueishi K: Intimal neovascularization in human coronary atherosclerosis: Its origin and pathophysiological significance. Hum Pathol 1995;26: 450-456. 54. de Boer OJ, van der Wal AC, Teeling P, Becker AE: Leucocyte recruitment in rupture prone regions of lipid-rich plaques: A prominent role for neovascularization? Cardiovasc Res 1999;41:443-449. 55. O’Brien KD, Allen MD, McDonald TO, et al: Vascular cell adhesion molecule-1 is expressed in human coronary atherosclerotic plaques. J Clin Invest 1993;92:945-951. 56. O’Brien KD, McDonald TO, Chait A, et al: Neovascular expression of E-selectin, intercellular adhesion molecule-1, and vascular cell adhesion molecule-1 in human atherosclerosis and their relation to intimal leukocyte content. Circulation 1996;93:672-682. 57. Moreno PR, Fuster V: New aspects in the pathogenesis of diabetic atherothrombosis. J Am Coll Cardiol 2004;44: 2293-2300. 58. Moreno PR, Purushothaman KR, Fuster V, et al: Plaque neovascularization is increased in ruptured atherosclerotic

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

59.

60.

61.

62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72.

73. 74.

75.

76.

77.

78. 79. 80.

lesions of human aorta: Implications for plaque vulnerability. Circulation 2004;110:2032-2038. Moreno PR, Purushothaman KR, O’Connor WN, et al: Microvessel sprouting, red blood cell extravasation, and perivascular inflammation is increased in plaques from patients with diabetes mellitus. J Am Coll Cardiol 2005;45:430A. Kockx MM, Cromheeke KM, Knaapen MW, et al: Phagocytosis and macrophage activation associated with hemorrhagic microvessels in human atherosclerosis. Arterioscler Thromb Vasc Biol 2003;23:440-446. Arbustini E, Morbini P, D’Armini AM, et al: Plaque composition in plexogenic and thromboembolic pulmonary hypertension: The critical role of thrombotic material in pultaceous core formation. Heart 2002;88:177-182. Kolodgie FD, Gold HK, Burke AP, et al: Intraplaque hemorrhage and progression of coronary atheroma. N Engl J Med 2003;349:2316-2325. Vlodavsky I, Friedmann Y: Molecular properties and involvement of heparanase in cancer metastasis and angiogenesis. J Clin Invest 2001;108:341-347. Brownlee M: Biochemistry and molecular cell biology of diabetic complications. Nature 2001;414:813-820. De Martin R, Hoeth M, Hofer-Warbinek R, Schmid JA: The transcription factor NF-kB and the regulation of vascular cell function. Arterioscler Thromb Vasc Biol 2000;20:e83-e88. Langlois MR, Delanghe JR: Biological and clinical significance of haptoglobin polymorphism in humans. Clin Chem 1996;42:1589-1600. Graversen JH, Madsen M, Moestrup SK: CD163: A signal receptor scavenging haptoglobin-hemoglobin complexes from plasma. Int J Biochem Cell Biol 2002;34:309-314. Schaer DJ: The macrophage hemoglobin scavenger receptor CD163 as a genetically determined disease modifying pathway in atherosclerosis. Atherosclerosis 2002;163:199-201. Bowman BH, Kurosky A: Haptoglobin: The evolutionary product of duplication, unequal crossing over, and point mutation. Adv Hum Genet 1982;12:189-261, 453-454. Levy AP: Haptoglobin: A major susceptibility gene for diabetic cardiovascular disease. Isr Med Assoc J 2004;6: 308-310. Asleh R, Marsh S, Shilkrut M, et al: Genetically determined heterogeneity in hemoglobin scavenging and susceptibility to diabetic cardiovascular disease. Circ Res 2003;92:1193-1200. Asleh R, Guetta J, Kalet-Litman S, et al: Haptoglobin genotype- and diabetes-dependent differences in iron-mediated oxidative stress in vitro and in vivo. Circ Res 2005;96: 435-441. Levy AP, Roguin A, Hochberg I, et al: Haptoglobin phenotype and vascular complications in patients with diabetes. N Engl J Med 2000;343:969-970. Levy AP, Hochberg I, Jablonski K, et al: Haptoglobin phenotype is an independent risk factor for cardiovascular disease in individuals with diabetes: The Strong Heart Study. J Am Coll Cardiol 2002;40:1984-1990. Roguin A, Koch W, Kastrati A, et al: Haptoglobin genotype is predictive of major adverse cardiac events in the 1-year period after percutaneous transluminal coronary angioplasty in individuals with diabetes. Diabetes Care 2003;26: 2628-2631. Suleiman M, Aronson D, Asleh R, et al: Haptoglobin polymorphism predicts 30-day mortality and heart failure in patients with diabetes and acute myocardial infarction. Diabetes 2005;54:2802-2806. Fuster V, Fayad ZA, Moreno PR, et al: Atherothrombosis and high-risk plaque. Part II: Approaches by noninvasive computed tomographic/magnetic resonance imaging. J Am Coll Cardiol 2005;46:1209-1218. Fujii K, Mintz GS, Carlier SG, et al: Intravascular ultrasound profile analysis of ruptured coronary plaques. Am J Cardiol 2006;98:429-435. DeMaria AN, Narula J, Mahmud E, Tsimikas S: Imaging vulnerable plaque by ultrasound. J Am Coll Cardiol 2006;47: C32-C39. Ge J, Chirillo F, Schwedtmann J, et al: Screening of ruptured plaques in patients with coronary artery disease by intravascular ultrasound. Heart 1999;81:621-627.

81. Peters RJ, Kok WE, Havenith MG, et al: Histopathologic validation of intracoronary ultrasound imaging. J Am Soc Echocardiogr 1994;7:230-241. 82. Komiyama N, Berry GJ, Kolz ML, et al: Tissue characterization of atherosclerotic plaques by intravascular ultrasound radiofrequency signal analysis: An in vitro study of human coronary arteries. Am Heart J 2000;140:565-574. 83. Sano K, Kawasaki M, Ishihara Y, et al: Assessment of vulnerable plaques causing acute coronary syndrome using integrated backscatter intravascular ultrasound. J Am Coll Cardiol 2006;47:734-741. 84. Yamagishi M, Terashima M, Awano K, et al: Morphology of vulnerable coronary plaque: Insights from follow-up of patients examined by intravascular ultrasound before an acute coronary syndrome. J Am Coll Cardiol 2000;35: 106-111. 85. Schoenhagen P, Nissen SE, Tuzcu EM: Coronary arterial remodeling: From bench to bedside. Curr Atheroscler Rep 2003;5:150-154. 86. Kaul S, Ito H: Microvasculature in acute myocardial ischemia. Part I: Evolving concepts in pathophysiology, diagnosis, and treatment. Circulation 2004;109:146-149. 87. Kaul S, Ito H: Microvasculature in acute myocardial ischemia. Part II: Evolving concepts in pathophysiology, diagnosis, and treatment. Circulation 2004;109:310-315. 88. Feinstein SB: Contrast ultrasound imaging of the carotid artery vasa vasorum and atherosclerotic plaque neovascularization. J Am Coll Cardiol 2006;48:236-243. 89. O’Malley SM, Vavuranakis M, Naghavi M, Kakadiaris IA: Intravascular ultrasound-based imaging of vasa vasorum for the detection of vulnerable atherosclerotic plaque. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv 2005;8:343-351. 90. Carlier S, Kakadiaris IA, Dib N, et al: Vasa vasorum imaging: A new window to the clinical detection of vulnerable atherosclerotic plaques. Curr Atheroscler Rep 2005;7:164-169. 91. Goertz DE, Frijlink ME, Tempel D, et al: Contrast harmonic intravascular ultrasound: A feasibility study for vasa vasorum imaging. Invest Radiol 2006;41:631-638. 92. Nair A, Kuban BD, Obuchowski N, Vince DG: Assessing spectral algorithms to predict atherosclerotic plaque composition with normalized and raw intravascular ultrasound data. Ultrasound Med Biol 2001;27:1319-1331. 93. Nair A, Kuban BD, Tuzcu EM, et al: Coronary plaque classification with intravascular ultrasound radiofrequency data analysis. Circulation 2002;106:2200-2206. 94. Rodriguez-Granillo GA, Garcia-Garcia HM, McFadden EP, et al: In vivo intravascular ultrasound-derived thin-cap fibroatheroma detection using ultrasound radiofrequency data analysis. J Am Coll Cardiol 2005;46:2038-2042. 95. Serruys PW: Fourth annual American College of Cardiology international lecture: A journey in the interventional field. J Am Coll Cardiol 2006;47:1754-1768. 96. Nasu K, Tsuchikane E, Katoh O, et al: Accuracy of in vivo coronary plaque morphology assessment: A validation study of in vivo virtual histology compared with in vitro histopathology. J Am Coll Cardiol 2006;47:2405-2412. 97. Granada JF: In vivo plaque characterization using intravascular ultrasound-virtual histology in a porcine model of complex coronary lesions. Arterioscler Thromb Vasc Biol 2007;27:387-393. Epub 206 Nov 30. 98. Surmely JF, Nasu K, Fujita H, et al: Coronary plaque composition of culprit/target lesions according to the clinical presentation: A virtual histology intravascular ultrasound analysis. Eur Heart J 2006;27:2939-2944. Epub 2006 Oct 13. 99. Rodriguez-Granillo GA, Garcia-Garcia HM, Valgimigli M, et al: Global characterization of coronary plaque rupture phenotype using three-vessel intravascular ultrasound radiofrequency data analysis. Eur Heart J 2006;27: 1921-1927. 100. Rodriguez-Granillo GA, McFadden EP, Valgimigli M, et al: Coronary plaque composition of nonculprit lesions, assessed by in vivo intracoronary ultrasound radiofrequency data analysis, is related to clinical presentation. Am Heart J 2006;151:1020-1024.

1173

1174

Evaluation of Interventional Techniques 101. Surmely JF, Nasu K, Fujita H, et al: Association of coronary plaque composition and arterial remodeling: A virtual histology intravascular ultrasound analysis. Heart 2007;93:928932. 102. Rodriguez-Granillo GA, Serruys PW, Garcia-Garcia HM, et al: Coronary artery remodelling is related to plaque composition. Heart 2006;92:388-391. 103. Low AF, Tearney GJ, Bouma BE, Jang IK: Technology insight: Optical coherence tomography. Current status and future development. Nat Clin Pract Cardiovasc Med 2006;3:154162; quiz 172. 104. Bouma BE: New insights from OCT, polarization-sensitive OCT, and the emergence of OFDI. Presented at The Vulnerable Plaque Session: Pathophysiology, Detection, and Therapeutic Intervention. Transcatheter Therapeutic Intervention (TCT) Meeting, Washington, DC, 2006. 105. Jang IK: Optical coherence tomography: Studies at MGH. In Transcatheter Cardiovascular Therapeutics Convention, Washington, DC, 2002. 106. Jang IK, Tearney GJ, MacNeill B, et al: In vivo characterization of coronary atherosclerotic plaque by use of optical coherence tomography. Circulation 2005;111:1551-1555. 107. Giattina SD, Courtney BK, Herz PR, et al: Assessment of coronary plaque collagen with polarization sensitive optical coherence tomography (PS-OCT). Int J Cardiol 2006; 107:400-409. 108. Tearney GJ, Yabushita H, Houser SL, et al: Quantification of macrophage content in atherosclerotic plaques by optical coherence tomography. Circulation 2003;107:113-119. 109. MacNeill BD, Jang IK, Bouma BE, et al: Focal and multi-focal plaque macrophage distributions in patients with acute and stable presentations of coronary artery disease. J Am Coll Cardiol 2004;44:972-979. 110. Jonas JB, Harder B, Spandau UH, et al: Bevacizumab for occult subfoveal neovascularization in age-related macular degeneration. Eur J Ophthalmol 2006;16:774-775. 111. Moon SJ, Wirostko WJ: Photodynamic therapy for extrafoveal choroidal neovascularization associated with choroidal nevus. Retina 2006;26:477-479. 112. Avery RL, Pieramici DJ, Rabena MD, et al: Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006;113:363-372. 113. Fuster V, Corti R, Fayad ZA, et al: Integration of vascular biology and magnetic resonance imaging in the understanding of atherothrombosis and acute coronary syndromes. J Thromb Haemost 2003;1:1410-1421. 114. Sirol M, Fuster V, Fayad ZA: Plaque imaging and characterization using magnetic resonance imaging: Towards molecular assessment. Curr Mol Med 2006;6:541-548. 115. Wilensky RL, Song HK, Ferrari VA: Role of magnetic resonance and intravascular magnetic resonance in the detection of vulnerable plaques. J Am Coll Cardiol 2006;47: C48-C56. 116. Schneiderman J, Wilensky RL, Weiss A, et al: Diagnosis of thin-cap fibroatheromas by a self-contained intravascular magnetic resonance imaging probe in ex vivo human aortas and in situ coronary arteries. J Am Coll Cardiol 2005;45: 1961-1969. 117. Lipinski MJ, Frias JC, Fayad ZA: Advances in detection and characterization of atherosclerosis using contrast agents targeting the macrophage. J Nucl Cardiol 2006;13:699709. 118. Jaffer FA, Libby P, Weissleder R: Molecular and cellular imaging of atherosclerosis: Emerging applications. J Am Coll Cardiol 2006;47:1328-1338. 119. Rudd JH, Warburton EA, Fryer TD, et al: Imaging atherosclerotic plaque inflammation with [18F]-fluorodeoxyglucose positron emission tomography. Circulation 2002;105: 2708-2711. 120. Davies JR, Rudd JH, Weissberg PL, Narula J: Radionuclide imaging for the detection of inflammation in vulnerable plaques. J Am Coll Cardiol 2006;47:C57-C68. 121. Lipinski MJ, Amirbekian V, Frias JC, et al: MRI to detect atherosclerosis with gadolinium-containing immunomicelles targeting the macrophage scavenger receptor. Magn Reson Med 2006;56:601-610.

122. Fayad ZA, Amirbekian V, Toussaint JF, Fuster V: Identification of interleukin-2 for imaging atherosclerotic inflammation. Eur J Nucl Med Mol Imaging 2006;33:111-116. 123. Sosnovik DE, Schellenberger EA, Nahrendorf M, et al: Magnetic resonance imaging of cardiomyocyte apoptosis with a novel magneto-optical nanoparticle. Magn Reson Med 2005;54:718-724. 124. Jaffer FA, Sosnovik DE, Nahrendorf M, Weissleder R: Molecular imaging of myocardial infarction. J Mol Cell Cardiol 2006;41:921-933. 125. Purushothaman KR, Sanz J, Zias E, et al: Atherosclerosis neovascularization and imaging. Curr Mol Med 2006;6: 549-556. 126. Winter PM, Morawski AM, Caruthers SD, et al: Molecular imaging of angiogenesis in early-stage atherosclerosis with alpha(v)beta3-integrin-targeted nanoparticles. Circulation 2003;108:2270-2274. 127. Matter CM, Schuler PK, Alessi P, et al: Molecular imaging of atherosclerotic plaques using a human antibody against the extra-domain B of fibronectin. Circ Res 2004;95:1225-1233. 128. Kelly KA, Allport JR, Tsourkas A, et al: Detection of vascular adhesion molecule-1 expression using a novel multimodal nanoparticle. Circ Res 2005;96:327-336. 129. Takaya N, Yuan C, Chu B, et al: Presence of intraplaque hemorrhage stimulates progression of carotid atherosclerotic plaques: A high-resolution magnetic resonance imaging study. Circulation 2005;111:2768-2775. 130. Takaya N, Yuan C, Chu B, et al: Association between carotid plaque characteristics and subsequent ischemic cerebrovascular events: A prospective assessment with MRI—Initial results. Stroke 2006;37:818-823. 131. Puppini G, Furlan F, Cirota N, et al: Characterisation of carotid atherosclerotic plaque: Comparison between magnetic resonance imaging and histology. Radiol Med (Torino) 2006;111:921-930. 132. Uchida Y, Nakamura F, Tomaru T, et al: Prediction of acute coronary syndromes by percutaneous coronary angioscopy in patients with stable angina. Am Heart J 1995;130: 195-203. 133. Asakura M, Ueda Y, Yamaguchi O, et al: Extensive development of vulnerable plaques as a pan-coronary process in patients with myocardial infarction: An angioscopic study. J Am Coll Cardiol 2001;37:1284-1288. 134. Ohtani T, Ueda Y, Mizote I, et al: Number of yellow plaques detected in a coronary artery is associated with future risk of acute coronary syndrome: Detection of vulnerable patients by angioscopy. J Am Coll Cardiol 2006;47:2194-2200. 135. Moreno PR, Muller JE: Detection of high-risk atherosclerotic coronary plaques by intravascular spectroscopy. J Interv Cardiol 2003;16:243-252. 136. van de Poll SW, Kastelijn K, Bakker Schut TC, et al: On-line detection of cholesterol and calcification by catheter based Raman spectroscopy in human atherosclerotic plaque ex vivo. Heart 2003;89:1078-1082. 137. Moreno PR, Lodder RA, Purushothaman KR, et al: Detection of lipid pool, thin fibrous cap, and inflammatory cells in human aortic atherosclerotic plaques by near-infrared spectroscopy. Circulation 2002;105:923-927. 138. Caplan JD, Waxman S, Nesto RW, Muller JE: Near-infrared spectroscopy for the detection of vulnerable coronary artery plaques. J Am Coll Cardiol 2006;47:C92-C96. 139. Madjid M, Willerson JT, Casscells SW: Intracoronary thermography for detection of high-risk vulnerable plaques. J Am Coll Cardiol 2006;47:C80-C85. 140. Kawasaki M, Bouma BE, Bressner J, et al: Diagnostic accuracy of optical coherence tomography and integrated backscatter intravascular ultrasound images for tissue characterization of human coronary plaques. J Am Coll Cardiol 2006;48: 81-88. 141. Romer TJ, Brennan JF 3rd, Puppels GJ, et al: Intravascular ultrasound combined with Raman spectroscopy to localize and quantify cholesterol and calcium salts in atherosclerotic coronary arteries. Arterioscler Thromb Vasc Biol 2000;20: 478-483. 142. Ambrose JA, D’Agate DJ: Classification of systemic therapies for potential stabilization of the vulnerable plaque to prevent

Atherothrombosis and the High-Risk Plaque: Definition, Diagnosis, and Treatment

143.

144.

145.

146.

147.

148. 149.

150.

151.

acute myocardial infarction. Am J Cardiol 2005;95:379382. Cannon CP, Steinberg BA, Murphy SA, et al: Meta-analysis of cardiovascular outcomes trials comparing intensive versus moderate statin therapy. J Am Coll Cardiol 2006;48: 438-445. Scirica BM, Morrow DA, Cannon CP, et al: Intensive statin therapy and the risk of hospitalization for heart failure after an acute coronary syndrome in the PROVE IT-TIMI 22 study. J Am Coll Cardiol 2006;47:2326-2331. Hulten E, Jackson JL, Douglas K, et al: The effect of early, intensive statin therapy on acute coronary syndrome: A meta-analysis of randomized controlled trials. Arch Intern Med 2006;166:1814-1821. Goldenberg I, Goldbourt U, Boyko V, et al: Relation between on-treatment increments in serum high-density lipoprotein cholesterol levels and cardiac mortality in patients with coronary heart disease (from the Bezafibrate Infarction Prevention trial). Am J Cardiol 2006;97:466-471. Corti R, Osende J, Hutter R, et al: Fenofibrate induces plaque regression in hypercholesterolemic atherosclerotic rabbits: In vivo demonstration by high-resolution MRI. Atherosclerosis 2007;190:106-113. Epub 2006 Apr 5. Naik SU, Wang X, Da Silva JS, et al: Pharmacological activation of liver X receptors promotes reverse cholesterol transport in vivo. Circulation 2006;113:90-97. Tso C, Martinic G, Fan WH, et al: High-density lipoproteins enhance progenitor-mediated endothelium repair in mice. Arterioscler Thromb Vasc Biol 2006;26:11441149. Dagenais GR, Pogue J, Fox K, et al: Angiotensin-convertingenzyme inhibitors in stable vascular disease without left ventricular systolic dysfunction or heart failure: A combined analysis of three trials. Lancet 2006;368:581-588. Cannon CP, Braunwald E, McCabe CH, et al: Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004;350:1495-1504.

152. Waksman R, Leitch IM, Roessler J, et al: Intracoronary photodynamic therapy reduces neointimal growth without suppressing re-endothelialisation in a porcine model. Heart 2006;92:1138-1144. 153. Waksman R: Photodynamic Therapy. New York, Taylor & Francis, 2004. 154. Kipshidze N: Endoluminal phototherapy with low-power red light laser for vulnerable plaque passivation. Presented to the Transcatheter Cardiovascular Therapeutics Convention, Washington, DC, October 2006. 155. Dorval JF, Geoffroy P, Sirois MG, Tanguay JF: Endovascular cryotherapy accentuates the accumulation of the fibrillar collagen types I and III after percutaneous transluminal angioplasty in pigs. J Endovasc Ther 2006;13:104-110. 156. Triesscheijn M, Baas P, Schellens JH, Stewart FA: Photodynamic therapy in oncology. Oncologist 2006;11:1034-1044. 157. Waksman R: Photodynamic therapy. In Virmani R, Narula J, Leon MB, Willerson JT (eds): The Vulnerable Atherosclerotic Plaque: Strategies for Diagnosis and Management. Malden, MA, Blackwell Futura, 2007, pp. 321-330. 158. Lafont A, Libby P: The smooth muscle cell: Sinner or saint in restenosis and the acute coronary syndromes? J Am Coll Cardiol 1998;32:283-285. 159. Davies MJ: Detecting vulnerable coronary plaques. Lancet 1996;347:1422-1423. 160. Serruys PW, Kukreja N: Late stent thrombosis in drug-eluting stents: Return of the “VB syndrome.” Nat Clin Pract Cardiovasc Med 2006;3:637. 161. Waksman R: Biodegradable stents: They do their job and disappear. J Invasive Cardiol 2006;18:70-74. 162. Moreno PR: First report of a vulnerable plaque specific stent capable of focal passivation without fibrous cap rupture. Presented to the Transcatheter Cardiovascular Therapeutics Convention, Washington, DC, October 2006.

1175

CHAPTER

64 Cardiovascular Interventional Magnetic Resonance Imaging Robert J. Lederman KEY POINTS 䊏 Interventional cardiovascular magnetic resonance imaging (MRI) remains investigational. 䊏 Because it makes soft tissue and blood conspicuous, interventional MRI may enable nonsurgeons to conduct novel minimally invasive procedures. 䊏 Potential applications include radiation-sparing conventional catheter procedures, image-guided myocardial ablation or targeted drug or cell delivery, and catheterbased surgical procedures such as valve implantation. 䊏 Interventional XMR suites can be configured from MRI and X-ray systems that can be operated together or fully independently.

As this textbook attests, a remarkable range of minimally invasive procedures can be conducted safely and effectively under the guidance of X-ray imaging. For the most part, these procedures are confined to existing vascular lumens and chambers, and they are effective even though soft tissues such as myocardium are not readily conspicuous. Alternative image guidance modalities such as magnetic resonance imaging (MRI) may further broaden the range of therapeutic possibilities for catheter-based interventions. This may prove useful to avoid ionizing radiation, as during pediatric procedures; to view the tissue response to treatment interactively, as during myocardial ablation to treat rhythm disorders; and to cross ordinary tissue boundaries, as during extraanatomic bypass. Interventional MRI remains in its infancy. Provocative initial demonstrations have been conducted in animals. Initial human testing is underway throughout the world. This chapter reviews the challenges and future applications of this exciting new technology for interventional catheterization.

䊏 Fusion of prior MRI datasets with real-time X-ray, an intermediate step, may provide enhanced image guidance for conventional interventional procedures. 䊏 Coronary artery interventional procedures are not possible under MRI guidance, barring unforeseen technology breakthroughs. 䊏 The lack of availability of clinical-grade MRI catheter devices poses the chief obstacle to wider clinical translation of this technology.

HOW MRI CREATES PICTURES FOR INTERVENTION A brief review of MRI follows, but better introductions can be found elsewhere.1 An image is a twodimensional (2D) grid of pixels, each having numeric values that correspond to brightness, color, or both. Ultrasound, computed tomography (CT), and MRI images usually represent flattened tomographs (slices), and each contributing pixel actual represents a 3D volume (voxel) within that slice. Three-dimensional matrices can be acquired that contain image data, but usually these are computer-reconstructed into viewable images that are either thin-slice samples (such as 2D echocardiography) or projections (such as shadows). MRI technology and terminology can be offputting to cardiologists, but it should be remembered that cardiac ultrasound has comparable complexity and terminology and is merely in a more advanced stage of clinical development. Interventional cardiologists are accustomed to using images that are X-ray projections (shadows), in

1177

1178

Evaluation of Interventional Techniques which there is a direct relationship between the position in space of X-ray beam attenuation and the position in the image of the features (bone, tissue, catheter) that caused the beam attenuation. Unlike X-ray projections, images in MRI are acquired in the form of frequency information that corresponds to image features only after a complex mathematical transformation. These frequency datasets are not interpretable as images (Fig. 64-1). Such frequency encoding of information is similar in many ways to human hearing: audio information is encoded as an amalgam of sound frequencies with varying amplitudes that are decoded by the cochlea and transmitted to the brain. Similarly, video information in MRI is encoded as an amalgam of frequencies that are decoded by Fourier transformation into spatial equivalents that constitute an image.

MRI exploits the phenomenon of nuclear magnetic resonance (NMR, or MR). Protons within the hydrogen atoms of water have an intrinsic magnetism that causes them to interact with external magnetic fields. Because water is ubiquitous in biologic tissues, proton NMR is most widely used for MRI. For hydrogen protons in water, at a given magnetic field (e.g., 1.5 Tesla [T]), there is a characteristic Larmor frequency (63.87 MHz at 1.5 T) at which the protons absorb energy. Once energized, these protons “decay” and emit radiofrequency (RF) energy at the same Larmor frequency. Most important, if the magnetic field strength is altered, the Larmor frequency is also altered proportionately. MR images are created by exploiting this fundamental relationship between the Larmor frequency and the strength of the magnetic field. Specifically, a

Figure 64-1. Frequency characteristics of a classic image, Ansel Adams’ Moon Over Half Dome. Images in magnetic resonance imaging (MRI) are sampled as frequency or k-space data, then usually converted using a Fourier transformation into spatial representations. A, The Fourier transformation of the full-spectrum image (B). The starburst pattern indicates that most of the information is contained by the center frequencies. If the same frequency data are filtered to retain only the low frequencies (C), the image appears mostly intact but devoid of detail (D). The low frequencies are said to contain most of the image contrast data. One approach to real-time MRI is to sample only center frequencies, so as to create low-resolution images quickly. If the same frequency data are filtered to retain only the high frequencies (E), the resulting image (F) contain only the edges or spatial detail of the original image. (Source: U.S. Government work [Lederman, NHLBI], public domain.)

Cardiovascular Interventional Magnetic Resonance Imaging linear magnetic field gradient is created (i.e., the magnetic field strength changes according to position in space) at specific times. The position in space of excited spins is “encoded” onto their frequency by creating known magnetic field values in different positions. The RF frequency of the spins corresponds exactly to their position within the artificially created magnetic field. This characteristic is generally exploited to create images using a repeated two-step process. The 2D position is encoded onto spins, first when spins are excited using RF (one dimension) and second when spins emit RF. Each cycle creates a line in a 2D image. Repeated cycles generate multiple lines to create a complete 2D image. A computer coordinates the timing of RF excitation and “readout” of RF emissions, along with excitation magnetic field gradients and readout magnetic fields, so that the exact position of these lines is known in advance. More specifically, for every line contributing to a 2D image, MRI excites a slice of the patient in space by creating a known gradient and exposing the body to RF energy precisely tuned to the Larmor frequency of the desired slice. Once a 2D slice is excited, the entire slice begins to emit RF energy. During that emission, the MRI scanner rapidly creates a new magnetic field gradient so that an individual line within that slice can be “read” using a high-sensitivity radio receiver tuned to the characteristic (Larmor) frequency of spins within that selected line. Because the gradient strength is known, the location of the line is also known, using the Larmor relationship. The process is repeated for the additional lines within the selected slice, until sufficient frequency information is accumulated to create a complete 2D image. The computer “program” used to accumulate these image data is called a pulse-sequence. Manipulation of the timing, strength, speed, and phase of the RF energy and magnetic field gradients in different pulse-sequences can dramatically alter the appearance of different tissues. Other features can be imaged apart from the position of tissues in space, such as the velocity of excited spins using “phase-contrast” or “velocity-encoded” MRI. Exogenous contrast agents can be administered to alter image contrast; for example, to create contrast-enhanced MR angiograms or to enhance the MRI appearance of scarred or infarcted myocardium. Certain pulse-sequences are particularly useful for cardiovascular and interventional MRI. Steady-state free precession (SSFP, also known as Fiesta, TrueFISP, and balanced fast field echo) is a popular workhorse pulse-sequence because of high signal-to-noise, speed, and versatility. In SSFP, blood appears bright and myocardium appears gray.

CONFIGURING AND OPERATING AN iCMR LABORATORY An interventional cardiovascular magnetic resonance imaging (iCMR) laboratory can be configured and installed by commercial vendors with only minor

modification.2,3 Usually these are configured as “XMR” suites with adjoining X-ray and MRI bays. The two otherwise ordinary systems can be used independently as in any other hospital. Alternatively, they can be used together during interventional MRI procedures. Compared with two separate suites, the incremental capital cost is fairly low and consists primarily of the intermodality patient transfer system and the barrier doors. Hardware Real-time MRI hardware requires highly homogeneous magnetic fields that usually are available only on “closed-bore” systems. At higher field strengths, tissues yield stronger MRI signals. However, higher field strengths pose additional engineering challenges. For example, at 3 T, minor magnetic field inhomogeneities are amplified during rapid imaging in a way that dramatically degrades images. As a result, most interventional MRI applications today seem best addressed at field strengths limited to approximately 1.5 T. Indeed, contemporary gradient systems have approached a plateau in priceperformance ratio and have approached physiologic limits of heating or peripheral nerve stimulation. Vendors are introducing shorter and wider closed-bore MRI systems (one system is only 120 cm long, with a bore inner diameter of 70 cm). These are attractive for interventional MRI procedures, because the systems induce less claustrophobia, patients can more readily be monitored visually, and operators’ arms can directly reach the center “suite spot” of the magnet. Most interventional MRI investigators ensure that X-ray fluoroscopy systems are readily available if not adjoining, both to use for emergency bailout and for simple procedural steps such as vascular access. Conventional X-ray image intensifier systems are subject to image distortion by the high magnetic fields, but they usually operate satisfactorily if situated more than 4 or 5 m away from the MRI bore. Newer flat-panel detector systems are relatively insensitive to this distortion and have significant advantages when combining or “fusing” image data from the two modalities at once. One vendor introduced a 1.0-T MRI system in a “double-doughnut” configuration so that the central imaging field is exposed and directly accessible for interventional procedures.4 The Stanford team integrated a retractable flat-panel X-ray into this system, so that conventional fluoroscopy could be conducted without moving the patient.5 MRI systems are enclosed in an RF barrier so that ambient RF noise, including ordinary communication signals as well as noise emitted by the X-ray and other electronics systems, does not interfere with the highsensitivity MR system. Merely opening these RF doors creates significant artifacts during MRI. The Kings College London laboratories have implemented a double-door RF barrier entrance, analogous to an “air lock,” so that staff can enter and exit the laboratory without disturbing the MRI system (Video 64-1).6

1179

1180

Evaluation of Interventional Techniques Modifications for Interventional MRI Traditional MRI is designed around diagnostic applications, wherein the rate-limiting step usually is image acquisition, not image display. Diagnostic MRI workflows afford the “luxury” of slower image reconstruction for later off-line review by the clinician. Interventional MRI requires almost instantaneous or “real-time” MRI. This acquisition-to-display delay is barely noticeable when it is less than about 250 msec, compared with X-ray systems, which have a delay of about 100 msec. Real-time MRI also has utility in diagnostic settings with much nonperiodic motion, such as atrial fibrillation, noncooperative patients, or cardiac stress testing. Most commercial systems provide real-time MRI functionality, but we have found it more useful to extract raw data from the commercial MRI system during acquisition and to reconstruct and display it to the operator on an inexpensive, fast external computer. Indeed at least one vendor is using this approach in commercializing an interventional MRI package as an external computer add-on to their diagnostic MRI systems, attached conveniently via high-speed network. MRI requires a compromise between spatial resolution and speed; interventional MRI usually sacrifices image quality in favor of speed. Projection X-ray imaging uses a pixel matrix as large as 1024 × 1024 at 15 to 30 frames per second. Our typical interventional MRI pulse-sequences generate 192 × 128 pixels at 8 to 10 frames per second. The soft tissue content often makes the images comparably information-rich (Fig. 64-2). Many techniques are combined to accelerate MRI for intervention. Undersampling refers to the creation of images using incomplete data. This is analogous to video compression for Web broadcasts. For

example, the frequency data contributing to images have intrinsic symmetries that can be undersampled and computer-regenerated. Similar approaches exploit the fact that the low-frequency center of frequency space (k-space) is more information-rich than the higher frequency data. These data can be sampled radially so that each line intersects the center frequency. An image created using such radial sampling may contain information comparable to that of an image created using more lines in classic rectilinear (“Cartesian”) fashion but can be faster. Various research groups (e.g., Stanford) advocate other approaches to sampling frequency data, including “spiral” trajectories. Commercial MRI systems are beginning to employ so-called parallel imaging techniques (with commercial names such as SENSE, SMASH, GRAPPA, and BLAST) to increase speed. These exploit the unique differences between multiple surface receiver coils, analogous to the way that binoculars provide more information than monoculars. Parallel imaging requires optimized MRI surface coil hardware. Another popular image acceleration technique is echo-sharing, wherein lines are recycled over multiple images. This technique is used for broadcast television in the United States, in which updates of odd and even lines are alternated on images. Imaging speed is effectively doubled, but temporal resolution is not. Michael Guttman and Elliot McVeigh at the National Heart, Lung, and Blood Institute (NHLBI) have developed numerous additional features that we have found valuable in interventional MRI applications. Perhaps most important is the ability to process signals independently from “active” catheters containing embedded antennae (see later discussion). When so attached to separate MRI receiver channels,

Figure 64-2. A comparison of the information content of X-ray images (A) and real-time magnetic resonance images (MRI) (B) using a transaortic myocardial injection catheter. The X-ray pixel matrix is greater (512 × 512) than that of the MRI (192 × 128). Even at this reduced spatial resolution, the soft-tissue display in MRI can provide superior information. (Source: U.S. Government work [Lederman, NHLBI], public domain.)

Cardiovascular Interventional Magnetic Resonance Imaging the device-related images can be displayed in color, and their brightness can be adjusted independently. Other useful features include the ability to toggle electrocardiographic (ECG) gating to “suspend” cardiac motion and the ability to toggle “saturation” MRI pulses that turn the image dark except for gadolinium contrast used to enhance myocardial infarction or injected during procedures. They also have developed real-time multislice acquisitions, with instantaneous rotating display of the slices in their true 3D relation. Finally, Guttman developed a “projection-mode” feature that is useful when catheters are manipulated during tomographic imaging. As catheter parts move outside the selected slice, they become invisible. Projection-mode imaging resembles X-ray imaging, so that the catheter can be manipulated back into the desired slice. The groups from Case Western Reserve and Johns Hopkins University have pioneered automated slice prescription features such as the ability to have the MR imaging plane automatically track the position of active-tracking catheter devices (see later discussion). Similar techniques automatically create rapid low-resolution pictures when these catheters are moved rapidly and slower high-resolution pictures when the catheters are adjusted finely. This has created competing philosophies about how to select images during interventional procedures: some advocate images created in relation to the catheter; we advocate images focused on target pathology. Ideally both approaches should be available. Image Display Interventional images can be displayed to the operators inside the MRI laboratory using either RF-shielded liquid crystal diode (LCD) displays, which are expensive, or more affordable custom RF-shielded LCD projectors onto aluminum-framed screens. The screens and the LCDs can be mounted on movable booms, as in X-ray laboratories. Usually, separate screens are necessary for MRI system control, for real-time MRI, and for invasive hemodynamics. Communications MRI systems are almost as loud as airplanes, especially during rapid imaging. Most of the noise is generated when rapidly fluctuating magnetic fields force gradient coil elements inside the scanner to strike each other. Even newer “acoustically-shielded” MRI systems are loud enough to preclude verbal communications. Patients and staff must wear ear protection for safety and for comfort. Fiberoptic, pneumatic, and wireless transmission systems are available with noise cancellation technology to permit staff to communicate even during MRI. We recommend systems that permit “open” audio communication without speakers first needing to switch on their microphones. We also recommend a second audio channel to permit staff independently to communicate with or monitor the patient.

Patient Monitoring Commercial MRI patient monitoring systems are designed for use in diagnostic imaging. Available systems usually have limited display capability, with display tracings from only one ECG lead, one or two invasive blood pressure channels, noninvasive blood pressure monitoring, pulse-oximetry, and exhaled carbon dioxide and anesthesia gas concentrations. Because these are “monitoring” systems, they do not provide high-fidelity “recording” signals used in conventional cardiovascular catheterization suites. These are, however, safe for operation during interventional MRI. Absent commercial interventional MRI hemodynamics recording systems, we recommend that operators substitute biomedical research hemodynamics equipment (e.g., Powerlab, ADInstruments) and operate under Institutional Review Board supervision, which in the United States constitutes the Investigational Device Exemption in this setting. Such systems are easily configured to display highfidelity signals in a format that resembles commercial clinical hemodynamics recording systems. We prefer to use our commercial monitoring system in tandem with a research recording system; most can be configured to output monitoring signals such as oximetry and noninvasive blood pressure data to the research recording system. ECG is intrinsically limited during MRI. Blood flowing through the heart and aorta carries a charge that generates current in the high magnetic field, the so-called magnetohydrodynamic effect. This makes the ST-T wave segments of the ECG uninterpretable inside the MRI bore. The MRI RF and gradient systems contribute additional noise to the ECG signal. As a result, ECG cannot be used to detect myocardial ischemia or injury during MRI and is useful primarily for rhythm detection. Patient Transfer In an XMR configuration, patients are transferred bidirectionally between the X-ray and MRI systems for combined or bailout procedures (Fig. 64-3). Transfer systems are available from all major vendors that convey the patient between imaging systems while protecting from significant decelerations. The patient lies on a transfer shell that also carries the operators’ sterile field, interventional equipment, and transducers. A surprisingly large number of lines and devices may be attached to patients during interventional procedures: (1) oxygen gas; (2) endotracheal tube; (3) pulse oximetry and capnography probes; (4) multiple intra-arterial pressure lines and transducers; (5) multiple intravenous catheters and extension lines; (6) interventional catheterization manifold including contrast, flush, waste, pressure transducer, and catheter extension lines; (7) urinary catheter and collection bag; (8) ECG leads for MRI; (9) ECG leads for X-ray; (10) intravascular MRI coil connectors; and

1181

1182

Evaluation of Interventional Techniques

Figure 64-3. A combined X-ray and magnetic resonance imaging (“XMR”) interventional suite. A standard single-plane X-ray system has its table pedestal mounted on floor rails. With the X-ray and radiofrequency barrier door open, patients can be transferred rapidly and smoothly between the two modalities. With the barrier doors closed, the laboratories function independently. (Source: U.S. Government work [Lederman, NHLBI)] public domain.)

(11) surface MRI coils and connectors. These should be secured and organized at the beginning of the procedure in case emergency intermodality transfer is required. The operators should verify that conductive cables do not form loops and are kept away from the inner wall of the magnet bore (near the body/transmit coil); both predispose to heating during MRI. Safety and Sterility Practices Patient care practices during interventional MRI procedures combine standard practices from X-ray and from MRI suites. The MRI bore and surface are draped with sterile adhesive drapes. Before intermodality transfer, the sterile field is covered with an additional sterile drape. Depending on the field to be imaged (i.e., pelvis or chest), the MRI surface coils can be placed under the sterile drapes or can be enclosed in sterile bags and positioned on top of the sterile field. Intravascular receiver coil transmission lines are covered with sterile sheaths marketed for intravascular ultrasound probes (Fig. 64-4). Inside the MRI laboratory, ferrous materials can become missiles that endanger both patients and staff. This risk is highest during emergencies, and therefore emergency practices must be planned and rehearsed. All staff must be trained about MRI safety, including removal of ferrous materials before entering the laboratory. Some laboratories issue pocketless uniforms to further reduce this risk.6 Cardiac defibrillation is hazardous within the high magnetic field, because the high currents generate sufficient force to displace the electrodes and the operators (literally knocking staff off their feet). By convention,

Figure 64-4. Inside the interventional magnetic resonance imaging (MRI) laboratory. The operators communicate with the use of sound-suppressing headsets and fiberoptic microphones. In-room projectors display (clockwise from top) hemodynamics, scanner control, and real-time MRI data. (Source: U.S. Government work [Lederman, NHLBI)] public domain.)

the magnetic field is considered clinically negligible outside the 5 Gauss (0.0005 T) magnetic field line, which many laboratories mark on the floor and which should not encroach on the neighboring Xray laboratory. Adjunctive patient care devices are marketed specifically for operation within the high magnetic field and include mechanical ventilators, intravenous pumps, oxygen tanks, and mobile aluminum tables. These should be specifically labeled as “MRI-safe”; all other equipment should be assumed not to be safe for use in the MRI laboratory. All other ferrous materials that are deemed essential must be tethered to the wall and kept outside the 5-Gauss line. When equipment appears interchangeable, such as catheter preparation tables, it is prudent to use MRI-safe selections for both X-ray and MRI rooms.

CATHETER DEVICES FOR INTERVENTIONAL CARDIOVASCULAR MRI Catheter devices for X-ray procedures are usually conspicuous because they readily attenuate incident X-rays. X-ray catheter development primarily focuses on mechanical performance. MRI catheters, however, are not as simple and must be engineered not only for mechanical performance but also (1) to be conspicuous, (2) to avoid magnetic attraction and displacement, (3) to avoid distorting the image because of magnetic susceptibility effects of metal components, and (4) to avoid heating during exposure to the high RF energy required to create MR images. Most off-the-shelf clinical catheter devices suffer limitations in one of these areas. Indeed, these challenges pose the chief limitation to clinical development of interventional cardiovascular MRI.

Cardiovascular Interventional Magnetic Resonance Imaging Materials Considerations Ferrous materials are incompatible with MRI. The most common, 316L stainless steel, is not significantly attracted by magnetic fields. However, 316L and other steel stents create large (susceptibility) signal voids because they disrupt the nearby magnetic field during MRI. As a result, most commercial braided diagnostic and guiding catheters, most balloon-expandable stents, most guidewires, and most device delivery systems are not suitable for MRI. Other metals are better suited, including nitinol, many nickel-cobalt-chromium alloys (e.g., MP35N, L605), copper, gold, tungsten, and platinum, in that they create comparably minor magnetic susceptibility artifacts. Not all of these materials are biocompatible. Long conductive structures are prone to inductive heating during MRI, especially during RF excitation, and especially if longer than 50 to 80 cm, at 1.5T. This is similar to metal heating in a microwave oven. As a result, even unmodified nitinol guidewires can cause thrombosis or tissue damage during MRI interventional procedures. Various modifications can render conductive guidewires safe. Shorter metallic structures, including stents, are unlikely to heat in this way. MRI catheter devices have two basic designs. Passive catheters are conspicuous based on their intrinsic materials properties. Active catheters are conspicuous

because they contain embedded electronics that interact with the MRI system. The various approaches are summarized in Table 64-1 and Figure 64-5. Passive Catheter Devices The simplest passive catheters are unbraided polymers and are visible based on the absence of water. These are intrinsically safe. Unfortunately, such catheters appear “dark” under MRI, which makes them poorly conspicuous for a number of reasons. First, the darkness, or absence of signal, is easily diluted by water signal (“volume-averaged”) within the voxels constituting the picture. Second, many imaging artifacts also appear dark, so simple passive catheters do not necessarily have a specific “signature” on images. As a result, although simple passive catheters appear acceptably well when depicted in ex vivo imaging phantoms, they can be hard to see in vivo. Finally, magnetic susceptibility artifacts are usually much larger than the devices that generate them, so the “blooming” signal voids may obscure nearby tissue or other features of interest. Catheters filled with dilute gadolinium-based contrast agents appear bright and can be tracked under MRI. However, interventional devices displace the contrast agent and reduce catheter visibility using this approach. Unal and colleagues7 have reported effective hydrophilic catheter coatings that render devices bright under MRI by permitting a great deal

Table 64-1. Approaches to Make Catheters Conspicuous on MRI Approach

Advantages

Disadvantages

Examples

Passive catheters, visible based on intrinsic materials properties

Simple, inexpensive Can be used in combination with other approaches

Gadolinium-filled balloon dilatation catheters Nonbraided angiography catheters

Active “imaging” catheters, incorporating MRI antennae

Highly conspicuous Catheters can be depicted in color Versatile imaging approaches, including projection-mode Simple, inexpensive Marker points can be tracked without imaging, to increase speed or reduce heating Marker points can be used to automate scan plane adjustments Requires no physical connection

Usually visible based on absence of water, generating dark signals Dark signals are easily diluted on MRI “Compatible” conductive wires can heat Must not contain ferrous braids Complex, expensive Conductive wires can heat Blurry profile compared with X-ray catheters Dipole designs have poor distal tip visibility MRI pulse sequence must be modified Catheter locations are computersynthesized on image Conductive wires can heat

Active “tracking” catheters, incorporating MRI antennae

Wireless devices

Multispectral devices

Non-proton MRI species can be displayed in different colors from target tissue Passive imaging that does not require transmission lines or embedded electronics

Embeds electronics that might interfere with mechanical performance (i.e., on stents) Imaging compromises, such as reduced flip angle Additional hardware required to excite and detect the alternate compounds Hyperpolarized 13C requires constant replenishment using specialized generator hardware

Surgi-Vision Intercept 0.030″ dipole guide wire coil Boston Scientific MRI Stilletto endomyocardial injection system General Electric Massachusetts General/ St. Jude electrophysiologic catheter mapping system26

Essen wireless stents38 and catheters8

13

C selective coronary arteriography10 19 F catheter tracking9

1183

1184

Evaluation of Interventional Techniques

Figure 64-5. A comparison of representative catheter designs for interventional cardiovascular magnetic resonance imaging (iCMR). A, Magnetic resonance imaging (MRI) of a typical steel-braided X-ray catheter. The ferrous material destroys the MRI image, may generate force inside the high magnetic field, and may heat during MRI. B, The same catheter is shown without the steel braids. It is difficult to see even in vitro. C, An “imaging-active” version of the same catheter contains MRI antennae and is attached to the scanner hardware. Signal from excited spins can be depicted in color. Note that the signal “falls off” toward the tip of this simple dipole design. D, A CO2filled balloon catheter in a human right atrium during real-time MRI. It is visible only as a signal void (arrow). E, A 4-Fr catheter (arrows) coated with a hydrophilic gadolinium paint. It can be seen as a white line within the aorta of a large mammal. F, A “trackingdesign” active catheter has a microcoil embedded near the tip. G, The position of the microcoil catheter in F is displayed as green crosshairs inside the right ventricle of an animal. H, A “wireless” catheter is tuned to resonate passively because of embedded electronics. It appears bright (I) during low-flip-angle MRI pulse-sequences. (A-C, From U.S. Government work [Lederman, NHLBI], public domain. D, Courtesy of Reza Razavi. From Miquel ME, Hegde S, Muthurangu V, et al: Visualization and tracking of an inflatable balloon catheter using SSFP in a flow phantom and in the heart and great vessels of patients. Magn Reson Med 2004;51:988-995. E, Courtesy of Orhan Unal. From Unal O, Li J, Cheng W, et al: MR-visible coatings for endovascular device visualization. J Magn Reson Imaging 2006;23:763-769. F and G, Courtesy of Michael Bock and Sven Zuehlsdorff, DKFZ Heidelberg. Original work. Distinct from Figure 3G in Bock M, Muller S, Zuehlsdorff S, et al: Active catheter tracking using parallel MRI and real-time image reconstruction. Magn Reson Med 2006;55:1454-1459. H and I, Courtesy of Harald H. Quick. From Quick HH, Zenge MO, Kuehl H, et al: Interventional magnetic resonance angiography with no strings attached: Wireless active catheter visualization. Magn Reson Med 2005;53:446-455, Figure 2A&B.)

Cardiovascular Interventional Magnetic Resonance Imaging of gadolinium agent to interact with excited water spins (see Fig. 64-5). Simpler passive devices have been tracked under MRI based on their dark appearance, including polymer guidewires with embedded dysprosium markers (which create disproportionate signal voids), gas-filled balloon catheters (see Fig. 64-5D), and various nitinol and platinum implants. Such polymer devices also are resistant to RF-induced heating during MRI. Special MRI pulse-sequences can enhance the appearance of metal-induced susceptibility artifacts in passive catheters, but at present these compromise the imaging in other ways. Active Catheter Devices Active catheter devices contain integrated MRI receiver coils, or antennae, and are connected to the MRI system hardware. They are visualized using two general approaches. Imaging-active catheters are intrinsically visible on images because they contribute information used to create the image displayed to the operator. This approach is called “profiling,” because usually a length of the catheter “profile” also is visualized. An alternative is to embed point-markers on the active catheter, use special MRI pulse-sequences to detect the location of these points, and synthesize the location markers electronically on the image. These catheters are known as tracking-design active devices. We advocate the former, but compelling in vivo experiments have been conducted using both approaches. Imaging-active catheters detect nearby excited spins and usually appear as a nebulous bright field surrounding the catheter device, unlike the sharp profiles generated by catheters in X-ray images (see Fig. 64-5C). More importantly, because imagingactive catheters can be attached to separate RF receiver channels on the MRI system, they can be processed separately and displayed in color. We have found this method to be particularly helpful (Fig. 64-6B). The simplest imaging-active catheter design is the loop-

less dipole antenna. One iteration is approved for marketing as a 0.030-inch intravascular MRI guidewire (Intercept, Surgi-Vision, Baltimore, MD); it uses external detuning and decoupling circuitry to prevent heating. The dipole design, however, suffers sensitivity falloff near the guidewire distal tip, which limits its suitability for interventional procedures (see Fig. 64-5C). More sophisticated designs, with conspicuous tips, are under development. Imaging-active catheters are especially helpful when used in tandem with passive devices. For example, balloon angioplasty catheters filled with dilute gadolinium contrast are far more conspicuous when delivered over an imaging-active guidewire coil, especially when signal from that coil is color-coded (see Fig. 64-6). Tracking-design active catheters contain embedded point-marker receiver microcoils that also detect nearby excited spins. Usually, the MRI pulse-sequence is modified for short periodic RF pulses used to find these microcoils in space. Computer synthesized point- or line-markers are overlaid on the MR image to indicate the position of the catheter (see Fig. 645F,G). Active-tracking catheters are also useful when used with automated MRI pulse-sequences that interact with the catheter, for example to keep the imaged slice aligned with the catheter as it moves, or to speed up imaging when the catheter is moved rapidly. A simple extension of the active-tracking microcoil design, with two opposed solenoids wound in opposite directions, can be used for high-sensitivity MRI of nearby structures. Related approaches incorporate miniature magnetic field detectors into the catheter device and use the active microcoils to generate blinking susceptibility artifacts (dark spots). The transmission lines connecting active catheter devices to the MRI hardware are themselves conductive and prone to heating. This can be addressed using electronic decoupling and detuning circuitry, alternative transmission line designs that incorporate transformer devices, or nonconductive approaches such as fiberoptic lines.

Figure 64-6. The utility of combining active and passive devices. A, A platinum stent is premounted on a partially inflated balloon with dilute gadolinium contrast and advanced over a nitinol guidewire. The guidewire is not visible, and only the balloon “dumbbells” (arrowheads) are conspicuous. B, The same balloon and stent are advanced over an active-imaging guidewire. The devices are brighter and more conspicuous, and their signal can be displayed in color. (Source: U.S. Government work [Lederman, NHLBI], public domain.)

1185

1186

Evaluation of Interventional Techniques Other Novel MRI Catheter Designs One particularly attractive approach is to use wireless catheter devices or even stents, which couple inductively to the MRI system using embedded electronics.8 Present designs work best when imaging at low flip angles, which may otherwise compromise image quality (see Fig. 64-5H,I). Most designs also are not detected on separate receiver channels, so catheter signals cannot be depicted in separate colors from anatomic structures. Another creative approach is to coat or fill catheter devices or contrast media with non-proton materials such as 19fluorine9 or hyperpolarized 13carbon.10 Multispectral MRI techniques can image water proton spins and these other species either simultaneously or in alternating fashion. These solutions combine the safety and simplicity of passive designs with the ability to use color to distinguish catheter components from target tissues. The Topspin medical device is a wholly selfcontained MRI catheter device intended for intravascular atherosclerosis imaging. It operates within a conventional X-ray suite without a traditional MRI system. The catheter incorporates strong static magnets and radio transmitter-receiver coils and uses a technology called rotating-frame zeugmatography to create MRI. It can distinguish nearby plaque components.11

APPLICATIONS Atherosclerosis Imaging The earliest invasive MRI devices were developed to improve imaging of atherosclerotic lesions. The rationale was to improve signal-to-noise ratio, and thereby spatial resolution or speed, by bringing receiver coils closer to the tissue of interest. Larose and colleagues at Brigham and Women’s Hospital positioned active dipole guidewire coils under X-ray in iliac artery lesions before conducting MRI.12 Other laboratories have reported similar work using the same guidewire,3,13 but none has convincingly demonstrated the value of MRI in imaging atheromata. Highersensitivity devices are actively being developed.14 Macrophage uptake of iron-based contrast agents can be detected in patients using MRI,15,16 but such examinations can be accomplished noninvasively. Coronary Artery Interventions Are Not Feasible Using Interventional Proton MRI In healthy swine, Speuntrup and associates17 delivered stainless steel stents into normal coronary arteries using passive visualization. Both devices and the widely-patent coronary arteries were well visualized using SSFP MRI at 1.5 T, in spite of cardiac and respiratory motion. Similar work has not been reported in diseased coronary arteries. Active invasive devices are unlikely to help, because, as presently envisioned, they would need to be navigated through coronary

arteries before they can be expected to enhance spatial resolution. X-ray fluoroscopy can resolve points as close as 200 µm, with a frame rate up to 30 frames per second (33 msec/frame). Operators can navigate 0.014-inch guidewires through tortuous, even occluded, coronary arteries under X-ray based on subtle visual feedback. Proton-based MRI cannot approach this performance, because signal-to-noise is too low. Meaningful MRI-guided transluminal coronary artery interventional procedures would require a technical breakthrough not anticipated at present. Cell and Drug Delivery to the Myocardium Endoluminal catheter procedures on heart muscle are particularly well suited for interventional MRI applications. Despite significant cardiac and respiratory motion, hearts are sufficiently large and thickwalled that they can be imaged with good spatial resolution in real time. Workhorse pulse-sequences such as SSFP depict blood as bright, so the endocardial border is highly conspicuous. Thinner-wall structures such as the atria, however, are difficult to visualize under real-time MRI. Our laboratories3,18 and many others19-23 have conducted targeted cell delivery into specified targets in diseased heart models in swine. This has been particularly successful using two-channel active devices. A steerable guiding catheter system and a separate injection needle are attached to different receiver channels and displayed in color (Fig. 64-7). This can target delayed gadolinium hyperenhancement (infarcts), infarct borders, wall motion abnormalities, subvalvular or papillary structures, perfusion maps, strain maps, and other features of interest. Multiple cell preparations can be made conspicuous under MRI by labeling with iron, yet without apparently disturbing in vitro functional assessments. Labeled cells can be seen to accumulate within target territories interactively during injection. This can be useful to confirm successful delivery, to visualize previous injections and avoid inadvertent overlap, or to ensure confluence of treatment targets if desired. Operationally, we have found it useful to image multiple slices simultaneously and to render their 3D volumetric relationship during device manipulation. This provides superior imaging of the beating heart during catheter-based procedures. To date, MRIguided cell and drug delivery has advanced ahead of development of suitable agents that might require targeted delivery. Therapeutic Electrophysiology Procedures Catheter-based therapeutic cardiac electrophysiology procedures are traditionally conducted under functional guidance of local electrograms, with adjunctive projection X-ray imaging. Several guidance systems combine functional contact or noncontact maps with electromagnetic catheter positioning systems (e.g., CARTO [Biosense-Webster]), increas-

Cardiovascular Interventional Magnetic Resonance Imaging

Figure 64-7. Magnetic resonance imaging (MRI)-guided cell delivery to the borders of a small myocardial infarct. A, A small infarct is visualized in the distal septum based on delayed gadolinium enhancement. B, A two-channel active endomyocardial injection guiding catheter (green) and needle (red) are used to inject iron-labeled stromal cells (black) to the target. C, The two cell injections are visible as black spots along the proximal and distal borders of the infarct. D, View of the real-time multislice imaging used during this procedure. (Source: U.S. Government work [Lederman, NHLBI], public domain.) (See Video 64-2.)

ingly overlaid with static roadmaps imported from prior MRI or X-ray CT acquisitions. However, few such electrophysiology procedures are conducted under imaging guidance. By contrast, numerous effective therapeutic electrophysiology procedures are conducted fairly rapidly under direct visual guidance after surgical exposure. MRI might have value in specifying the instantaneous position of devices in relation to actual anatomic structures or in identifying tissue response to treatment (e.g., myocardial edema after RF ablation). In particular, MRI might be useful in identifying lines of continuity created by ablative energy, which might represent functional electrophysiologic block. Henry Halperin and colleagues at Johns Hopkins University successfully tracked active dipole catheters and acquired intracardiac electrograms during realtime MRI.24 They also characterized the MRI appearance of fresh RF ablation lesions.25 Vivek Reddy, Ehud Schmidt, Charles Dumoulin, and colleagues26 reported instantaneous positioning of cardiac electrophysiology catheters overlaid on high-resolution ECG-gated moving cardiac images. The cardiac images could be updated intermittently as needed.

prosthesis in the aortic position from a transfemoral approach in healthy swine. This demonstrated the value of combined device and tissue imaging using MRI. At present, the chief limitations to catheter-based prosthetic valve implantation relate to impaired performance of devices miniaturized for transcatheter deployment. One exciting application of MRI is to guide minimally invasive surgical deployment of valve prostheses. McVeigh, Horvath, and colleagues28 used minimally invasive access to the left ventricular apex (via a pursestring suture) with MRI-guided implantation of an aortic prosthesis in swine (Fig. 64-8). The apical access provided large-bore access to deliver a large-profile stent-mounted commercial bioprosthesis. MRI permitted precise orientation, axial positioning, and deployment of the prosthesis. Their subsequent work demonstrated the incremental value of active catheter devices over solely passive devices. Whether MRI has value in other minimally invasive cardiac valve annulus and leaflet interventions remains untested.

Valve Interventions and MRI-Guided Cardiac Surgery

Several groups have deployed passive nitinol occluder devices across atrial septal defects in swine using MRI guidance.29-31 MRI is useful for intraprocedural assessment of outcomes. Whether it has incremental value for this procedure over transesophageal or intracardiac ultrasound remains uncertain. MRI may prove

MRI might have value in positioning valve prostheses delivered from a transcatheter or minimally invasive approach. Kuehne and colleagues27 used MRI to deliver and position a passive nitinol-based valve

Connecting and Disconnecting Chambers and Vessels

1187

1188

Evaluation of Interventional Techniques

Figure 64-8. Magnetic resonance imaging (MRI)-guided transapical aortic valve implantation. A, With a trocar inserted across the left ventricular apex, a passive guidewire is advanced across the aortic valve. B, The prosthetic valve is introduced into the trocar, then into the left ventricular outflow tract (C), and is then positioned across the native aortic valve (D) and aligned with respect to the coronary ostia (not shown). The balloon-expandable stent is inflated with gadolinium contrast agent (E), and “saturation” imaging is toggled (F) to highlight this agent. The balloon is deflected (G) and withdrawn (H), along with the guidewire (I). (Source: U.S. Government work [McVeigh, NHLBI], public domain.) (See Videos 64-3 through 64-6.)

useful to deploy occlusion devices in more complex lesions such as ventricular septal defects, but few data are available. Interventional cardiovascular MRI starts to get interesting when it can enable operators to defy the normal confines of vascular chambers. A simple step in this direction is to connect adjacent vascular structures during atrial septal puncture. The interatrial septum in swine is easily and safely punctured and enlarged under MRI guidance using active Brockenbraugh-style needles and guidewires.32 Arepally and colleagues at Johns Hopkins University used a similar active needle to effect a transcatheter mesocaval shunt, connecting the vena cava and the superior mesenteric vein using a nitinol connector.33 The Stanford team, using the “doubledoughnut” MRI system described earlier, conducted effective transjugular intrahepatic portosystemic shunts (TIPS) in patients.34 MRI guidance reduced the number of unsuccessful punctures, which can be hazardous. MRI might enable nonsurgeons to establish connections between other nonadjacent vascular struc-

tures, for example connecting the subclavian and pulmonary arteries, or the aorta and femoral arteries. Catheter-based extra-anatomic bypass might be an attractive alternative to open-surgical bypass and is under active development. MRI should permit inspection of vascular and extravascular spaces and the trajectory of connecting devices, as well as instantaneous assessment of the adequacy of catheter-based anastomoses. Chronic Total Occlusion Another attractive application is to use MRI to negotiate intravascular trajectories that cannot be visualized by X-ray techniques. Chronic total occlusion (CTO) of peripheral arteries is one such setting. XRay fluoroscopic angiography with radiocontrast identifies the occluded inflow artery and possibly collateral-dependent outflow beyond the occlusion, but it cannot discriminate CTO arterial wall and lumen. In these cases, the entire occluded arterial segment remains invisible to the operator. Virtually “blind” device manipulation risks procedural failure

Cardiovascular Interventional Magnetic Resonance Imaging and arterial perforation. MRI can visualize even the occluded arterial segment, including arterial wall, lumen content, and adjacent structures, and therefore could guide recanalization of CTO. Raval and colleagues created an animal model of peripheral artery CTO in swine. They were able to recanalize these lesions, using homemade active devices under MRI guidance, but were largely unsuccessful using X-ray guidance and high-performance contemporary clinical guidewires.35 Clinical-grade devices are currently under development to translate this experience to human subjects. Aortic and Other Peripheral Artery Interventions Endovascular repair for abdominal aortic aneurysm (AAA) is limited in part by endoleak due to inflow or outflow malapposition. X-ray alone is imperfect in visualizing complex 3D aneurysms. Raman and colleagues36 used MRI to guide simple endograft repair of AAA in pig models. Homemade, temporarily active tube endograft devices, which appeared in color until they were disconnected after deployment, were more useful and successful than comparable passive endograft devices. MRI afforded precise positioning as well as immediate anatomic and functional evaluation of success by visualizing stent apposition, flow contours, and contrast exclusion. Eggebrecht and colleagues in Essen, Germany, demonstrated a wonderful application of MRI to guide deployment of unmodified passive, selfexpanding Gore endografts in an animal model of descending thoracic aortic dissection.37 MRI distinguished the true from the false lumen, guided device

positioning and deployment, and demonstrated obliteration of the dissection (Fig. 64-9). Straightforward clinical peripheral angioplasty and stenting is not likely to be a high-priority application for interventional MRI. MRI has been used to conduct angioplasty and stenting in animal models of iliac, renal, and carotid arteries38-40; to deploy vena cava stents41 and image captured embolic material; and to guide visceral embolization such as in kidney segments.42 These and many other applications reflect important technical advances. The team at the University of Regensberg have conducted MRI-guided angioplasty and stenting, using only passive devices, of human iliac43 and of femoropopliteal44 atherosclerotic lesions. Other laboratories are beginning similar work using passive and active devices. Pediatric and Other Congenital Applications There are settings where interventional MRI may have value even in straightforward catheter-based procedures. The advantages of MRI might include avoiding radiation or iodinated radiocontrast or reducing operator musculoskeletal injury from lead aprons. Children with congenital cardiovascular abnormalities are subjected to multiple catheter procedures during their lifetime and may accumulate a large radiation exposure with excess malignancy risk. Schalla, Higgins, Moore, and colleagues at the University of California San Francisco conducted comprehensive diagnostic cardiac catheterization procedures in swine with atrial septal defect (Fig. 64-10), using active-tracking catheters containing microcoils. They traversed right-sided and left-sided

Figure 64-9. Delivery of a passive aortic endograft (arrows) to treat a porcine model of aortic dissection. A, The dissection flap is visible (arrowheads). B through E, The endograft is positioned over the dissection origin. F, The dissection is obliterated by the self-expanding endograft. (Courtesy of Harald H. Quick. Modified slightly from Eggebrecht H, Kuhl H, Kaiser GM, et al: Feasibility of real-time magnetic resonance-guided stent-graft placement in a swine model of descending aortic dissection. Eur Heart J 2006;27:613-620, by permission of Oxford University Press.) (See Video 64-7.)

1189

1190

Evaluation of Interventional Techniques

Figure 64-10. Diagnostic cardiac catheterization using magnetic resonance imaging (MRI) and tracking-design active catheters in swine. The distal microcoil receiver is displayed as a dot as the catheter traverses from the inferior vena cava (A) to the right atrium (B and C) and across an atrial septal defect (D) into the left atrium (E) and left ventricle (F). Continuous pressure recordings are displayed during pull back at the bottom of the image, along with hemoglobin saturation values. (Courtesy of Simon Schalla. From Schalla S, Saeed M, Higgins CB, et al: Magnetic resonance-guided cardiac catheterization in a swine model of atrial septal defect. Circulation 2003;108:1865-1870.)

heart chambers with continuous intracavitary pressure measurement and blood sampling. They also used velocity-encoded MRI to make measurements such as shunt ratios. In landmark work, the team at Kings College London tested these techniques in humans using a combined XMR suite. They used MRI and passive gas-filled balloon-tipped catheters to conduct diagnostic cardiac catheterization in children, to avoid radiation exposure.6 In this and subsequent reports, they used MRI to make hemodynamic measurements after provocation, for selective gadolinium arteriography, and for measuring pulmonary artery compliance.45 We deployed stents to treat an animal model of aortic coarctation using commercially available clinical devices, including an active guidewire.46 Using a passive balloon filled with dilute gadolinium and an active guidewire, MRI provided continuous display of stent and target tissue during deployment (Fig. 64-11). Phase-contrast MRI provided hemodynamic results during the procedure. Perhaps more importantly, during deliberate balloon overstretch of the animal coarctation lesion, MRI provided instantaneous feedback about life-threatening aortic perforation. This might have incremental value over X-ray,

which would require contrast administration to demonstrate perforation. Krueger, Kuehne, and colleagues in Berlin performed balloon angioplasty of aortic coarctation lesions in patients during MRI in an important clinical first step toward wholly MRIguided treatment of this congenital lesion.47

X-RAY FLUOROSCOPY FUSED WITH MRI Clinical interventional MRI remains inaccessible to most clinicians, yet MRI acquisitions are readily available. One obvious intermediate step would be to combine prior MRI datasets with real-time X-ray fluoroscopy to guide therapeutic procedures (X-ray fused with MRI, or XFM). The challenge is to ensure that the images are properly aligned or “registered.” Images are registered by aligning common points of reference. There are two general approaches. The simpler approach is to assume a common image space; for example, in relation to a patient table. The table, bearing the patient, is moved from one imaging modality to another, and the table location is known precisely to the registration computer. As long as the patient does not move or change, images from one modality can be registered and combined precisely with images from another. The second general

Cardiovascular Interventional Magnetic Resonance Imaging

Figure 64-11. Delivery of a platinum stent into a porcine model of aortic coarctation.46 A, A platinum stent is hand-crimped on a delivery balloon that is partially inflated with dilute gadolinium. The balloon is advanced over an active-imaging guidewire using realtime magnetic resonance imaging (MRI). The signal from the active guidewire is colored green, which makes the balloon also appear green. The system is positioned in the coarctation (B), and the balloon is inflated (C). Both the devices and the target tissue are visible with the use of real-time MRI. Conventional (non-real-time) black-blood MRI shows the hourglass deformity of the coarctation lesion before (D) and after (E) stenting. F, Through-plane phase-contrast MRI just below the coarctation lesion shows turbulence-induced flow reversal. G, After stenting, laminar flow is restored. 1 and 3 indicate antegrade flow and 2 indicates flow reversal. (Source: U.S. Government work [Lederman, NHLBI], public domain.) (See Videos 64-8 through 64-10.)

approach uses so-called “fiducial” markers, which are uniquely identified in both imaging modalities. After corresponding fiducial markers are identified using each modality, computerized transformations can register and combine images from both. Fiducial markers may be simple beads attached to patients’ skin; better fiducial markers can be endogenous, such as bones or vascular bifurcations. Motion of the fidu-

cial markers causes misregistration. Combining X-ray fluoroscopy and MRI requires further computer manipulation in 3D-to-2D transformation. The Kings College London team used the first approach, opticallly tracking the patient transport shell.2,6 They overlaid MRI-derived endocardial surfaces with electrograms from contact mapping catheters during therapeutic electrophysiology procedures

1191

1192

Evaluation of Interventional Techniques

Figure 64-12. Two examples of X-ray fused with magnetic resonance imaging (XFA). A, Magnetic resonance angiography (MRA) contours of cavae, right atrium, right ventricle, and pulmonary arteries are combined with colored, computer-segmented X-ray images obtained with the use of an electrophysiology basket catheter. B, X-ray stent implantation for aortic coarctation under XFM, combining a magnetic resonance roadmap image. A, Adapted from Rhode KS, Sermesant M, Brogan D, et al: A system for real-time XMR guided cardiovascular intervention. IEEE Transactions on Medical Imaging 2005;24:1428-1440. B, Original work, courtesy of Reza Razavi, Kings College London.)

to enhance image guidance. They also have overlaid MR images with real-time X-ray during vascular interventions such as coarctation stenting (Fig. 64-12). Our laboratories use the second approach, affixing multimodality beads (containing X-ray and MRI contrast agents) to the skin of animals and patients during catheter procedures. These beads serve as fiducial markers to register prior MRI datasets, even ECGgated cinematic ones, with real-time X-ray images. Once registered, the combination of 2D (X-ray) and 3D (MRI) data is automatic irrespective of gantry or table motion. The combined XFM images can display features of interest, such as myocardial infarcts or zones of thinned myocardium. Using beating-heart endomyocardial cell injections to test the geometric accuracy of the XFM guidance system in vivo, we found a mean target registration error of about 3 mm (Fig. 64-13).48 An obvious next step is to combine ECG and respiratory gating for even finer combination of the two image datasets. Many other laboratories are developing combined MRI, CT, electroanatomic maps, and fluoroscopic images registered in many different ways to guide therapeutic procedures.

CONCLUSION This chapter outlined how an XMR facility can be configured during installation of two independent X-ray and MRI systems. The incremental cost of an intermodality transportation system and of barrier doors is relatively low. Existing technology already provides excellent real-time MR imaging sufficient to guide an array of investigational therapeutic procedures and to care for patients in this environment. The only barrier to clinical development of interven-

Figure 64-13. X-ray fused with magnetic resonance imaging (XFM) of myocardial cell injection in swine. Multimodality external fiducial markers (yellow arrows) are used to establish a common frame of reference to combine images from MRI onto Xray. Electrocardiographic (ECG)-gated moving features of interest from MRI can be combined with real-time X-ray images to provide some of the best capabilities of each independent modality. The red surface represents the MRI-derived left ventricular endocardium, blue the left ventricular epicardium, and green a large myocardial infarction. A transfemoral endomyocardial injection needle is shown delivering cells to desired targets (yellow dots). (Source: U.S. Government work [Lederman, NHLBI], public domain.)

Cardiovascular Interventional Magnetic Resonance Imaging tional cardiovascular MRI is the commercial unavailability of clinical-grade active catheter devices. Competing technologies are available but suffer important limitations. Electromagnetic position mapping relies on prior (roadmap) anatomic information. CT provides superb soft tissue imaging but requires prohibitive ionizing radiation exposure to guide interventional cardiovascular procedures. Three-dimensional ultrasound suffers limited acoustic windows and acoustic shadowing from interventional devices but will benefit from continued engineering development. MRI may prove superior because it provides “onestop shopping” in a single imaging modality, combining 3D anatomy, biochemical characterization, mechanical function, and hemodynamics. Real-time MRI provides surgical exposure to nonsurgeons conducting minimally invasive procedures and has the potential to revolutionize interventional cardiology.

16.

17. 18. 19. 20. 21.

22.

REFERENCES 23. 1. NessAiver M: All You Really Need to Know about MRI Physics. Baltimore, MD, Simply Physics, 1997. 2. Rhode KS, Sermesant M, Brogan D, et al: A system for real-time XMR guided cardiovascular intervention. IEEE Transactions on Medical Imaging 2005;24:1428-1440. 3. Dick AJ, Raman VK, Raval AN, et al: Invasive human magnetic resonance imaging during angioplasty: Feasibility in a combined XMR suite. Catheter Cardiovasc Interv 2005;64: 265-274. 4. Fahrig R, Butts K, Wen Z, et al: Truly hybrid interventional MR/X-ray system: Investigation of in vivo applications. Acad Radiol 2001;8:1200-1207. 5. Rieke V, Ganguly A, Daniel BL, et al: X-ray compatible radiofrequency coil for magnetic resonance imaging. Magn Reson Med 2005;53:1409-1414. 6. Razavi R, Hill DL, Keevil SF, et al: Cardiac catheterisation guided by MRI in children and adults with congenital heart disease. Lancet 2003;362:1877-1882. 7. Unal O, Li J, Cheng W, et al: MR-visible coatings for endovascular device visualization. J Magn Reson Imaging 2006;23:763-769. 8. Quick HH, Zenge MO, Kuehl H, et al: Interventional magnetic resonance angiography with no strings attached: Wireless active catheter visualization. Magn Reson Med 2005;53: 446-455. 9. Kozerke S, Hegde S, Schaeffter T, et al: Catheter tracking and visualization using 19F nuclear magnetic resonance. Magn Reson Med 2004;52:693-697. 10. Olsson LE, Chai CM, Axelsson O, et al: MR coronary angiography in pigs with intraarterial injections of a hyperpolarized 13C substance. Magn Reson Med 2006;55:731-737. 11. Schneiderman J, Wilensky RL, Weiss A, et al: Diagnosis of thin-cap fibroatheromas by a self-contained intravascular magnetic resonance imaging probe in ex vivo human aortas and in situ coronary arteries. J Am Coll Cardiol 2005;45: 1961-1969. 12. Larose E, Yeghiazarians Y, Libby P, et al: Characterization of human atherosclerotic plaques by intravascular magnetic resonance imaging. Circulation 2005;112:2324-2331. 13. Hofmann LV, Liddell RP, Eng J, et al: Human peripheral arteries: Feasibility of transvenous intravascular MR imaging of the arterial wall. Radiology 2005;235:617-622. 14. Hillenbrand CM, Elgort DR, Wong EY, et al: Active device tracking and high-resolution intravascular MRI using a novel catheter-based, opposed-solenoid phased array coil. Magn Reson Med 2004;51:668-675. 15. Kooi ME, Cappendijk VC, Cleutjens KB, et al: Accumulation of ultrasmall superparamagnetic particles of iron oxide in human atherosclerotic plaques can be detected by in vivo

24. 25.

26.

27.

28.

29.

30. 31.

32. 33. 34.

35.

36.

magnetic resonance imaging. Circulation 2003;107: 2453-2458. Trivedi RA, U-King-Im J, Graves MJ, et al: In vivo detection of macrophages in human carotid atheroma: Temporal dependence of ultrasmall superparamagnetic particles of iron oxideenhanced MRI. Stroke 2004;35:1631-1635. Spuentrup E, Ruebben A, Schaeffter T, et al: Magnetic resonance-guided coronary artery stent placement in a swine model. Circulation 2002;105:874-879. Lederman RJ, Guttman MA, Peters DC, et al: Catheter-based endomyocardial injection with real-time magnetic resonance imaging. Circulation 2002;105:1282-1284. Karmarkar PV, Kraitchman DL, Izbudak I, et al: MR-trackable intramyocardial injection catheter. Magn Reson Med 2004;51:1163-1172. Rickers C, Gallegos R, Seethamraju RT, et al: Applications of magnetic resonance imaging for cardiac stem cell therapy. J Intervent Cardiol 2004;17:37-46. Krombach GA, Pfeffer JG, Kinzel S, et al: MR-guided percutaneous intramyocardial injection with an MR-compatible catheter: Feasibility and changes in T1 values after injection of extracellular contrast medium in pigs. Radiology 2005;235:487-494. Corti R, Badimon J, Mizsei G, et al: Real time magnetic resonance guided endomyocardial local delivery. Heart 2005;91:348-353. Saeed M, Martin AJ, Lee RJ, et al: MR guidance of targeted injections into border and core of scarred myocardium in pigs. Radiology 2006;240:419-426. Epub 2006 Jun 26. Susil RC, Yeung CJ, Halperin HR, et al: Multifunctional interventional devices for MRI: A combined electrophysiology/MRI catheter. Magn Reson Med 2002;47:594-600. Lardo AC, McVeigh ER, Jumrussirikul P, et al: Visualization and temporal/spatial characterization of cardiac radiofrequency ablation lesions using magnetic resonance imaging. Circulation 2000;102:698-705. Reddy V, Malchano Z, Dukkipati S, et al: Interventional MRI: Electroanatomical mapping using real-time MR tracking of a deflectable catheter [Abstract]. Heart Rhythm 2005;2: S279-S280. Kuehne T, Yilmaz S, Meinus C, et al: Magnetic resonance imaging-guided transcatheter implantation of a prosthetic valve in aortic valve position: Feasibility study in swine. J Am Coll Cardiol 2004;44:2247-2249. McVeigh ER, Guttman MA, Lederman RJ, et al: Real-time interactive MRI-guided cardiac surgery: Aortic valve replacement using a direct apical approach. Magn Reson Med 2006;56:958-964. Buecker A, Spuentrup E, Grabitz R, et al: Magnetic resonanceguided placement of atrial septal closure device in animal model of patent foramen ovale. Circulation 2002;106: 511-515. Rickers C, Jerosch-Herold M, Hu X, et al: Magnetic resonance image-guided transcatheter closure of atrial septal defects. Circulation 2003;107:132-138. Schalla S, Saeed M, Higgins CB, et al: Balloon sizing and transcatheter closure of acute atrial septal defects guided by magnetic resonance fluoroscopy: Assessment and validation in a large animal model. J Magn Reson Imaging 2005;21: 204-211. Raval AN, Karmarkar PV, Guttman MA, et al: Real-time MRI guided atrial septal puncture and balloon septostomy in swine. Catheter Cardiovasc Interv 2006;67:637-643. Arepally A, Karmarkar PV, Weiss C, Atalar E: Percutaneous MR imaging-guided transvascular access of mesenteric venous system: Study in swine model. Radiology 2006;238:113-118. Kee ST, Ganguly A, Daniel BL, et al: MR-guided transjugular intrahepatic portosystemic shunt creation with use of a hybrid radiography/MR system. J Vasc Interv Radiol 2005;16(2 Pt 1):227-234. Raval AN, Karmarkar PV, Guttman MA, et al: Real-time magnetic resonance imaging-guided endovascular recanalization of chronic total arterial occlusion in a swine model. Circulation 2006;113:1101-1107. Raman VK, Karmarkar PV, Guttman MA, et al: Real-time magnetic resonance-guided endovascular repair of experimental

1193

1194

Evaluation of Interventional Techniques

37.

38. 39. 40.

41.

42.

43. 44.

abdominal aortic aneurysm in swine. J Am Coll Cardiol 2005;45:2069-2077. Eggebrecht H, Kuhl H, Kaiser GM, et al: Feasibility of real-time magnetic resonance-guided stent-graft placement in a swine model of descending aortic dissection. Eur Heart J 2006; 27:613-620. Quick HH, Kuehl H, Kaiser G, et al: Inductively coupled stent antennas in MRI. Magn Reson Med 2002;48:781-790. Elgort DR, Hillenbrand CM, Zhang S, et al: Image-guided and -monitored renal artery stenting using only MRI. J Magn Reson Imaging 2006;23:619-627. Omary RA, Gehl JA, Schirf BE, et al: MR imaging versus conventional X-ray fluoroscopy-guided renal angioplasty in swine: Prospective randomized comparison. Radiology 2006;238:489-496. Bartels LW, Bos C, van Der Weide R, et al: Placement of an inferior vena cava filter in a pig guided by high-resolution MR fluoroscopy at 1.5 T. J Magn Reson Imaging 2000;12: 599-605. Fink C, Bock M, Umathum R, et al: Renal embolization: Feasibility of magnetic resonance-guidance using active catheter tracking and intraarterial magnetic resonance angiography. Invest Radiol 2004;39:111-119. Manke C, Nitz WR, Djavidani B, et al: MR imaging-guided stent placement in iliac arterial stenoses: A feasibility study. Radiology 2001;219:527-534. Paetzel C, Zorger N, Bachthaler M, et al: Magnetic resonanceguided percutaneous angioplasty of femoral and popliteal artery stenoses using real-time imaging and intra-arterial con-

45.

46.

47. 48.

49.

50. 51.

trast-enhanced magnetic resonance angiography. Invest Radiol 2005;40:257-262. Muthurangu V, Atkinson D, Sermesant M, et al: Measurement of total pulmonary arterial compliance using invasive pressure monitoring and MR flow quantification during MR-guided cardiac catheterization. Am J Physiol 2005;289:H1301H1306. Raval AN, Telep JD, Guttman MA, et al: Real-time magnetic resonance imaging-guided stenting of aortic coarctation with commercially available catheter devices in swine. Circulation 2005;112:699-706. Krueger JJ, Ewert P, Yilmaz S, et al: Magnetic resonance imaging-guided balloon angioplasty of coarctation of the aorta: A pilot study. Circulation 2006;113:1093-1100. de Silva R, Gutierrez LF, Raval AN, et al: X-ray fused with magnetic resonance imaging (XFM) to target endomyocardial injections: Validation in a swine model of myocardial infarction. Circulation 2006;114:2342-2350. Miquel ME, Hegde S, Muthurangu V, et al: Visualization and tracking of an inflatable balloon catheter using SSFP in a flow phantom and in the heart and great vessels of patients. Magn Reson Med 2004;51:988-995. Bock M, Muller S, Zuehlsdorff S, et al: Active catheter tracking using parallel MRI and real-time image reconstruction. Magn Reson Med 2006;55:1454-1459. Schalla S, Saeed M, Higgins CB, et al: Magnetic resonanceguided cardiac catheterization in a swine model of atrial septal defect. Circulation 2003;108:1865-1870.

CHAPTER

65 Medical Economics in Interventional Cardiology Daniel B. Mark KEY POINTS 䊏 The most important determinant of the results of an economic analysis is the absolute magnitude of the clinical benefits being produced by the new therapy or strategy relative to the comparison therapy or strategy. 䊏 An economic analysis should examine the full spectrum of incremental clinical and cost outcomes of the therapy or strategy being studied over a time horizon, typically the lifetime of the cohort, that reasonably reflects the long-term consequences of that technology for the patients being treated. 䊏 Cost-effectiveness analysis assesses the extra or incremental cost for a new therapy or strategy to produce an incremental or extra unit of health benefits, such as an extra life-year or quality-adjusted life-year. The focus is on the

According to American Heart Association estimates, percutaneous coronary intervention (PCI) is now performed in the United States about 1,244,000 times each year on more than 600,000 patients.1 Despite this astonishing level of adoption into mainstream cardiovascular practice, many controversies persist about the appropriate indications for PCI and about its value provided for money spent. The purpose of the present chapter is to review the latter question. The first part of the chapter provides a brief overview of key economic concepts and approaches. The second part reviews empirical research data on the economics of interventional cardiology.

MEDICAL ECONOMICS CONCEPTS AND METHODS Medical Cost Definitions and Terminology The traditional economic view of “cost” is that it represents the consumption of societal resources that could have been used for another purpose.2 The term “opportunity cost” is sometimes used to indicate this particular meaning of cost. Society consumes resources to satisfy its wants, including those for food, housing, and recreation as well as health care.

assessment of efficiency of a particular investment relative to alternative ways to spend health care dollars. 䊏 Therapies with large up-front costs and delayed incremental clinical benefits, such as coronary artery bypass grafting (CABG), can be economically attractive if applied to sufficiently high-risk populations (and if the analysis examines an appropriately long time horizon). 䊏 Infrequently, a new therapy or strategy is so effective at reducing complications that it fully offsets its cost (a therapy that offers better clinical outcomes with lower cost is referred to as a “dominant therapy”), but most new therapies that improve outcomes are also more costly in the long run, necessitating cost-effectiveness analysis.

However, because resources are ultimately finite, society cannot satisfy all wants and is obliged to choose from among the potential alternative uses of its resources. Economics provides a set of tools and approaches to assist with the decisions regarding what health care to produce, in what quantity, and for whom. The classic illustration of the constrained resources concept is the “guns-versus-butter” example from freshman economics. Resources expended in the production of weapons cannot also be applied to the production of food; therefore, in a world of limited resources, more weapons may mean less food. At the societal level, more health care may ultimately translate into less investment in education, transportation, housing, or other societal priorities. In most businesses or industries, the market price of a product or service is equal to the cost of producing that item plus some amount of profit (typically reflecting a fair return on investment). In the U.S. medical sector, the discrepancy almost universally observed between prices (or charges) and costs (the true cost of providing a given medical service) is largely attributable to “cost shifting,” a set of accounting practices designed to shift costs from a variety of sources (Table 65-1) onto whichever group of payers is most willing and able to absorb them. The net

1197

1198

Outcome Effectiveness of Interventional Cardiology Table 65-1. Major Components of Hospital Charges for Medical Services

Table 65-2. Major Methodological Issues in Medical Cost Studies

Cost for given hospital service True costs to hospital of resources consumed (e.g., disposable supplies, personnel, equipment, allocated overhead)

Measurement of cost Categories of cost items to be included Disposable supplies Personnel (direct cost component) Department overhead (e.g., departmental administration, maintenance, equipment depreciation, utilities) Allocated hospital or clinic overhead (e.g., hospital administration, admissions, medical records) Focus for cost analysis Resources consumed/service provided (“bottom-up”) Billed charges/fees (“top-down”) Episode of care Historical data Structural framework of this study Randomized controlled trial Observational study Cost-effectiveness model Possible perspectives of the cost analysis Societal Medicare, managed care, other third-party payers Hospitals, clinics, physicians, other providers Patients Time effects on medical costs Inflation Discounting of future costs Geographic effects on medical costs Different practice settings Different geographic locations within a country

Charge or price for given hospital service Cost-shifting accounting maneuvers Bad debts Free services (e.g., indigent care, employees) Disallowed costs by third-party payers Replacement of existing equipment Acquisition of new technologies (e.g., magnetic resonance imaging) Budgeting for expansion of services (e.g., more inpatient beds, more outpatient clinics) From Mark DB, Jollis J: Economic aspects of therapy for acute myocardial infarction. In Bates ER (ed): Adjunctive Therapy for Acute Myocardial Infarction. New York, Marcel Dekker, 1991, 471-496.

effect of these cost-shifting practices is to distort the relationship between U.S. medical prices or charges and medical resource consumption. The concepts of marginal and incremental costs are commonly used in medical economics. Strictly speaking, marginal cost is defined as the cost of producing one additional (or one less) unit of product, such as a PCI or bypass surgery procedure. However, although the notion of one more or one less test or procedure is useful for some types of (usually theoretical) economic analyses, the more practical questions usually center on examining the costs of shifting groups of patients from one diagnostic or therapeutic strategy to another. For this type of analysis, the term incremental is often substituted for marginal. (Unfortunately, some leading medical economics researchers use “marginal” and “incremental” synonymously, whereas others do not, leading to potential confusion for those outside the field.) Incremental cost analysis is an important component of costeffectiveness analysis and is central to the economic notion of cost as a measure of alternative uses of scarce resources introduced at the beginning of this chapter. In the next section, we consider how incremental costs are actually measured and some of the problems involved in translating economic theory into practice. Induced costs (or savings) are the costs of the tests or therapies added or averted as a consequence of some initial management decision and/or resource use. A few examples will make the concept clear. The institution of an aggressive program of intravenous thrombolytic therapy in patients with acute myocardial infarction (MI) by a given hospital or physician practice group may be accompanied by a rise in the number of patients with major disabling strokes who need long-term care. That latter cost is a cost induced by the initial therapeutic decision. In the same way, performance of PCI induces the cost of repeat revascularization procedures to treat symptomatic restenosis. Use of statin therapy for secondary prevention induces a cost savings over the long term owing

to reduced need for cardiac procedures and hospitalizations. Finally, the term indirect costs is often used by health service researchers to discuss the societal costs associated with loss of employment or productivity due to morbidity. Because of the potential for confusion with the accounting meaning of indirect costs, the alternative term productivity costs is used by some. Methodologic Issues in Medical Cost Studies To perform a medical cost study, it is necessary to consider five major issues (Table 65-2): (1) the way cost is to be conceptualized and measured, (2) the type of study to be performed (the structural framework in which the cost analysis will be accomplished, discussed later), (3) the perspectives of the analysis, (4) the importance of cost variations over time, and (5) the importance of cost variations due to geographic and market factors. Cost Measurement In any clinical cost study, the investigators must decide at an early stage what categories of cost items they need to include in the analysis and at what level of detail they wish to focus (see Table 65-2). In practice, the types of detailed data required for marginal or incremental cost analysis are difficult to obtain unless the hospital involved has a computerized costaccounting system, and they are impractical (if not

Medical Economics in Interventional Cardiology Table 65-3. Five Estimates of Cost Savings Available by Shifting Patients from CABG to PTCA Cost Accounting Method* Component †

Total Difference ($) Room Procedure Blood Bank Electrocardiography Laboratory Pharmacy Respiratory Supplies Radiology

1

2

3

4

5

1,935 283 28 342 3 65 1,061 48 68 38

4,593 2,323 334 390 16 146 1,115 120 68 78

5,346 1,939 1,014 466 33 368 682 358 93 263

7,837 3,052 876 532 47 1,035 1,076 463 135 459

10,087 3,277 1,348 749 70 1,392 1,727 529 271 555

From Hlatky MA, Lipscomb J, Nelson C, et al: Resource use and cost of initial coronary revascularization: Coronary angioplasty versus coronary bypass surgery. Circulation 1990;82(Suppl IV):IV-208-IV-213. *Method 1, disposable supplies only; Method 2, supplies plus personnel; Method 3, costs from charges using department-level cost/charge ratios; Method 4, costs from charges using Medicare cost/charge ratios; Method 5, charges. † Cost differences are given in 1986 U.S. dollars. CABG, coronary artery bypass graft surgery; PTCA, coronary angioplasty.

impossible) to obtain for all participants in the typical large multicenter trial. Therefore, rather than adding up the individual resources being consumed (which might be termed the bottom-up approach), most U.S. cost studies start with an aggregated measure of costs, such as can be obtained from hospital or physician bills (a top-down analysis). Although the top-down approach is much more practical for many cost studies, especially multicenter studies, it does reduce the ability of the investigator to control the factors that are included as costs in the analysis. To evaluate the practical impact of using top-down versus bottom-up cost estimates, Hlatky and colleagues at Duke compared the magnitude of cost savings available by shifting from a more expensive treatment (i.e., coronary artery bypass grafting [CABG]) to a less expensive one (i.e., percutaneous transluminal coronary angioplasty [PTCA]) in 389 patients with coronary artery disease (CAD). Two bottom-up and three top-down cost estimates were examined (Table 65-3). Using only hospital charges (method 5), the cost savings was estimated at $10,000 per patient shifted to PTCA. However, if no hospital or departmental overhead is to be saved from this change in practice, then the true cost savings would be that estimated by method 1 or 2 : 20% to 46% of the amount estimated from charges. Methods 3 and 4, which include varying amounts of overhead, overestimate the short-term cost savings from the CABG to PTCA shift; they are more correctly viewed as providing an estimate of the difference in average cost. On the other hand, methods 1 and 2 indicate the marginal or incremental difference. Note that the difference between costs using the Medicare correction factors (method 4) and using charges (method 5) is due at least in part to the hospital’s shifting of costs from nonpaying patients to the paying segment. This “surtax” would, of course, never be recoverable by changes in patient management. For this reason alone, charges represent a poor

choice for evaluating the cost implications of different clinical strategies. A true bottom-up cost analysis, sometimes referred to as microcosting, is a complex, time-consuming process that requires identification of all the inputs into a health care service and the assignment of an appropriate cost to each. This is easiest for a relatively simple service, such as the administration of an antibiotic, the performance of a radiograph, or a laboratory test. A more complex hospital laboratory procedure, such as a coronary angioplasty, is a considerably greater challenge because of the large variability of inputs from one case to the next. Most complicated of all is an entire episode of care from admission to discharge, because this requires detailed cost and resource-use data from virtually every major hospital department. Of the top-down strategies for estimating costs, the most widely used in the United States involves converting hospital charges (taken from the hospital bill) to costs using the correction factors or ratios of costs to charges (RCCs) included in each hospital’s annual Medicare Cost Report. Medicare RCCs are largely a holdover from the era before prospective payment, when Medicare reimbursed hospitals on the basis of costs incurred. To do so, Medicare developed a method of deciding how to reimburse hospitals for the reasonable and necessary costs of providing care to its beneficiaries (i.e., true hospital cost), rather than paying the full charged amount. This method involved an elaborate reporting system that required each hospital to file with Centers for Medicare and Medicaid Services (CMS) each year. In this report, which is still required, the hospital details how expenses for direct patient care, overhead, capital equipment, and so forth relates to billed charges. To provide CMS with a means of converting charges to costs for its various ancillary services, each hospital includes in its report a set of ratios, the RCCs. Although not designed for research, the Medicare

1199

1200

Outcome Effectiveness of Interventional Cardiology RCCs represent a moderately standardized means of estimating cost across all the hospitals in the United States that file a Medicare Cost Report. Although no longer used for reimbursement, the Medicare Cost Report still serves as the primary source of government data on hospital costs. In addition, costs calculated with the RCC method are used to recalibrate diagnosis-related group (DRG) weights. Therefore, this method represents a valuable tool for multicenter cost research. There are three important limitations to the RCC method of cost estimation that should be noted. First, this approach does not separate out overhead and most other fixed costs and therefore provides an estimate of average rather than marginal cost; hence, it may overestimate potential cost savings. Second, Medicare Cost Reports have complex, detailed instructions for how they are to be filled out; as with the complex federal income tax reporting system, this means that hospitals may choose to interpret the instructions differently (just as different people choose to fill out their income tax forms differently). For this reason, the goal of uncovering a hospital’s true cost of providing care may be accomplished to a varying degree using this method. Finally, RCCs are themselves averages of all the cost/charge relationships within a large hospital revenue (ancillary) center such as the radiology, pharmacy, or laboratory departments. If an individual patient’s resource consumption pattern in a given cost center is not “average,” the Medicare RCCs may not be particularly accurate in converting those charges to costs. For the same reason, conversion of charges to costs for individual items on a detailed hospital bill may not be particularly accurate if the RCC for that item is not close to the average RCC of that cost center. By extension, the practice some have advocated of using one average RCC for an entire hospital may also be less accurate. Medicare DRG reimbursement rates provide an alternative top-down cost estimation method that does not depend on the vagaries of hospital bills. Once the patient’s DRG assignment is known, it becomes a simple matter to calculate the “hospital costs.” This system of cost estimation has a number of limitations, however. First, it is not sensitive to variations in resource-use intensity within a DRG. Thus, DRG reimbursement rates are averages in the sense that they represent the “average cost” for a particular DRG among all (elderly) patients in that DRG. Second, if CMS decides not to increase reimbursement to cover the costs of new technology, which has been the case with many new, expensive drug therapies, the DRG reimbursement is insensitive to large differences in resource costs. To take a more complex example, a patient who is admitted for unstable angina and undergoes a diagnostic cardiac catheterization, a PCI, a repeat PCI for abrupt closure, and then CABG will likely be coded out as DRG 106 (CABG with catheterization); from CMS’s point of view, the cost of the two PCIs is the hospital’s problem. For all these reasons, DRG reimbursement

is not a particularly good way of estimating costs in an economic analysis unless the analysis is being done from CMS’s perspective. The most approximate cost-estimation method used in clinical research involves counting only bigticket items consumed (such as number of coronary angioplasties, cardiac catheterizations, or CABGs; days in the intensive care unit; and total hospital length of stay) and assigning arbitrary unit prices to each item. The prices assigned are usually charges derived from a single institution or estimated from expert opinion. The resulting linear formula: Total cost = Σ price × quantity is simple and inexpensive to use (hence its appeal in clinical research), but it suffers from some important drawbacks. First, this approach has almost never been empirically validated within a given institution or (even more important) across institutions. Second, the appropriate set of big-ticket items necessary to estimate costs accurately using this method has never been rigorously defined. Third, the method usually treats the big-ticket inputs as though they were homogeneous. For example, an uncomplicated single-vessel PCI would typically be assigned the same price as a complex three-vessel PCI procedure complicated by abrupt closure. The true costs of these two procedures may in fact differ substantially. Assignment of costs to physician services in a cost analysis is usually done in one of two ways. In the past, physician fees (which are charges, analogous to hospital charges) have been used. Because most patients receive care from a variety of practitioners (each billing out of a separate office), collecting actual physician bills is several times more complicated than collecting hospital bills. Increasingly, however, physician fees have become a distorted measure of true resource inputs, because physicians have been forced to employ the same types of cost shifting used by hospitals to cover unreimbursed and underreimbursed services. Furthermore, it can be reasonably argued that physician fees in the “fee-for-service” era have never properly reflected a true market price for physician services. Unlike the situation for hospitals, however, Medicare has never required physicians to disclose their true costs in a cost report. Because of these distortions, the Medicare Fee Schedule—based on the resource-based relative-value scale (RBRVS) of Hsiao and colleagues3—has been adopted as a more appropriate method for assigning costs to physician services. The basic concept of the RBRVS is that the price of a service should reflect the (long-term) cost of providing that service. Medicare fees are tied to the American Medical Association Physician’s Current Procedural Terminology (CPT) classification system, so that to estimate physician costs from these fees, some map must be created between the CPT codes and the data available in the study database about physician services. Although the Medicare Fee Schedule is not an ideal measure of the consumption of physician work in

Medical Economics in Interventional Cardiology health care, it has the advantage of being more objective and consistent than charges or fees. Furthermore, the Medicare RBRVS payment schedule is now being used by many private insurers and managed care organizations, although with variations. Therefore, it represents the best available national fee schedule for physician work.

Importance of Perspective in Cost Analysis Cost is always defined (either explicitly or implicitly) in terms of specific buyers and sellers (or consumers and producers). Table 65-2 lists the different perspectives that can be used for a medical cost analysis. Most commonly, economists and health policy analysts advocate the use of a societal perspective, in which total health expenditures (public and private) are examined as a function of the benefits produced and the opportunities forgone across the economy. Such an analysis ideally includes hospital costs, physician fees, outpatient testing, outpatient drug therapy costs, non-medical direct expenses (e.g., transportation to the medical facility, child care, housekeeping), and the economic impact of lost productivity due to illness. In contrast, analysis from the perspective of specific payers or providers typically includes only a portion of the costs listed for a societal analysis. For CMS, for example, hospital costs are defined by the payments specified by the relevant DRG regardless of the amount of services provided (or their cost to the provider). The Medicare Fee Schedule performs a similar function for physician services. For payers other than CMS, costs are the amount they are actually required to pay (or agree to reimburse providers) for health care services. Large insurance companies and managed care plans usually are able to obtain significant discounts off the list price, whereas the individual or the small company that is self-insured may be required to pay total charges. As a practical matter, payer perspectives other than societal and that of CMS are infrequently used in cost studies in the medical literature.

Time Effects in Cost Analysis Time effects are important to medical cost analyses for two major reasons (see Table 65-2). First, inflationary forces in the economy cause the value of money to diminish over time, so that cost studies from different years should not be directly compared until differences due solely to inflation are accounted for. Although there are several ways to make this adjustment (none of them ideal), perhaps the most widely used is the medical care component of the Producer Price Index (available at http://www.bls. gov/ppi/). Independent of the effects of inflation, future medical expenditures are considered less costly than

current ones, because current expenditures take the money out of your pocket right away, whereas future expenditures allow you to hold onto your money for a period and invest it at the market rate of return. Assuming that the inflation-adjusted price for medical care does not change over time (i.e., no technologic advances that increase or decrease true costs), one can buy more medical care with a given nominal sum of money in 5 or 10 years than is possible in the present. For this reason, cost studies using a longterm perspective employ a technique called discounting to account for the differences between the present and future value of money.

Geographic and Market Factors Geographic and market economic factors also have important effects on medical care costs, although these have received little study in empirical medical cost research. Different practice settings (e.g., within a particular region of the country) can affect the cost of providing a given type of care owing to variations in case mix, different practice patterns of the health care team (e.g., physicians, nurses, administrators), and different levels of efficiency within each setting. For example, for a given patient, care in an academic tertiary care center and care in a large private community hospital in the same city may be associated with quite different hospital costs. First, the teaching hospital must add at least part of the cost of its resident staff, and because the latter must be supervised by an attending physician, total physician time is usually increased per unit of care in a teaching hospital. Furthermore, residents typically order more tests per patient encounter. Other cost differences could arise from differing levels of nursing intensity at each stage in the hospitalization, differing use of intensive and intermediate care beds, and different typical lengths of stay for particular problems. A comparison of patients in the Thrombosis in Myocardial Infarction (TIMI) II trial conservative arm initially admitted to a tertiary versus a community hospital revealed that tertiary centers used more coronary angiography, coronary angioplasty, and CABG for medically equivalent patients.4 The costs of material and labor inputs to medical care can vary substantially from one part of the country to another, creating true differences in the cost of providing a given medical service according to geographic factors. Labor costs (e.g., nursing salaries) are probably the most important of the geographic determinants of health care cost variations. Thus, comparison of cost studies from different regions of the country or different practice settings should include an adjustment for geographic cost differences. Several geographic adjustment indices are available, including the Medicare area wage index (for adjusting DRG reimbursement) and the Medicare Fee Schedule geographic adjustment factor.

1201

1202

Outcome Effectiveness of Interventional Cardiology Table 65-4. Cost Effectiveness, Cost Utility, and Cost Benefit: Sample Calculations

From Detsky AS, Naglie IG: A clinician’s guide to cost-effectiveness analysis. Ann Intern Med 1990;113:147-154, with permission.

Cost Study Structures

Cost-Effectiveness Analysis

Cost studies generally fall into one of three categories: randomized, controlled trials; observational studies; and cost-effectiveness models. Cost-effectiveness models are discussed in the next section. A cost study in a randomized clinical trial is usually ancillary to the primary objective of the trial. Typically, cost or resource consumption patterns are a secondary end point in a trial that has either a composite clinical or (preferably) a mortality primary end point. Some have argued that because randomized trials are rarely performed with cost as a primary end point, the trials are usually not optimized to answer the economic questions of greatest interest, except in so far as these questions parallel the primary clinical ones. In addition, requirements of the clinical portion of the study may distort the economic substudy. For example, follow-up protocol angiography to define restenosis leads to repeat revascularization procedures that would not otherwise have been done. Observational cost studies include both nonrandomized treatment comparisons and descriptive series without an intrinsic comparison group. Descriptive cost studies are useful in areas in which few empirical cost data have been published. Such data can be used to make sample size projections for randomized trials or to inform cost-effectiveness and other health policy studies (in conjunction with appropriate sensitivity analyses). Little has been done to date with observational treatment comparisons involving cost data. As with observational comparisons of medical outcomes, statistical adjustment techniques to “level the playing field” are critical. However, because medical costs are subject to variations over time and over geographic location and practice setting, it is still uncertain what boundaries exist for defining when a nonrandomized cost comparison can be appropriately performed and when it cannot.

In clinical medicine, the term cost-effective is frequently used synonymously with worthwhile to indicate an intuitive, unspecified threshold between useful and wasteful medical expenditures. In addition, the term cost-effective has a specific technical (and not particularly intuitive) meaning. Costeffectiveness analysis involves the explicit comparison of one option or program with at least one alternative investment of dollars, and it never indicates whether a given expenditure is worthwhile in an absolute sense but rather how it stands relative to other potential expenditures. Therefore, it is incorrect to speak of any medical practice in isolation as cost-effective. The primary objective of cost-effectiveness analysis is to evaluate disparate health care expenditures in common terms so that policy and other decision-makers can choose the most efficient method of producing extra health benefits from among the alternative ways that health care dollars can be distributed. The general term cost-effectiveness analysis actually refers to a family of methods for economic analysis (Table 65-4). For all methods, the final measure is expressed in ratio form, with incremental costs in the numerator and incremental health care benefits or outcomes in the denominator. The distinction among the methods derives primarily from how health benefits are measured. In cost-effectiveness analysis, the measure of incremental health effects chosen is typically the difference in life expectancy between the alternative strategies being evaluated (see Table 654). This is the most common type of economic health care analysis performed. In cost-utility analysis, remaining survival is adjusted for less than full quality (quality-adjusted life-years [QALYs]). Used much less often in medicine is cost-benefit analysis, probably because it requires measuring all health-related benefits of a program in monetary terms (see Table 65-4). Cost-benefit analysis permits comparison of medical

Medical Economics in Interventional Cardiology care expenditures with societal expenditures on education, defense, transportation, and so forth, whereas cost-effectiveness analysis is useful only in comparison of expenditures that produce the same type of outcome (e.g., QALYs). Table 65-4 compares two hypothetical treatment strategies (A and B) for a particular disease and summarizes the calculations involved in the different analyses. Treatment A costs twice as much as treatment B but also improves average life expectancy by 1 year. Thus, the cost-effectiveness ratio for A relative to B is $10,000 per life-year saved. Whether switching from B to A is “worthwhile” depends on the alternative health care expenditures (aside from A) available for $10,000 or less. This is the most common sort of problem faced in cost-effectiveness analysis: whether to fund a new program that provides more health benefits than the standard therapy but at a substantially increased cost. (It is theoretically possible to go in the other direction—to give up health benefits to save substantial health care dollars—but this is rarely politically viable.) QALYs allow us to factor in the value of the extended survival offered by a new program or alternative therapy to the patient, as well as its quantity. In the example in Table 65-4, strategy A improves life expectancy relative to B, but the average quality of life for survivors is lower. This could come about in several ways. For example, with strategy B the sickest patients could die, leaving a relatively healthier group of survivors. In contrast, strategy A saves these sick patients from dying but cannot restore them to the same level of health as other patients with lower disease severity. These sicker surviving patients lower the average quality of life for the group. Alternatively, there could be something about strategy A that negatively affects quality of life, such as the need for chronic medication that is associated with significant side effects and that is not required with strategy B. In this example, moving from cost-effectiveness to cost-utility analysis more than doubles the cost of an additional unit of (quality-adjusted) survival with strategy A. The underlying tenet of all these forms of economic analysis is that the analyst desires to determine the most efficient means of maximizing the net health benefits for a particular group or population under the constraint of limited resources (i.e., where it is not possible to provide every beneficial service to every potential recipient). Note that such economic analyses are neutral to the specific patients and diseases under study; the health benefits being maximized are abstractly conceptualized as belonging to a large group or population. In actual practice, however, political and other forces may play a large role in deciding where societal dollars are to be invested. The ultimate problem for economic analysis is that individual patients (and their physicians) wish to obtain all the health benefits that are available from modern medical technology. However, the collection of all patients taken together (i.e., society), does not

have enough resources to provide this for everyone, thereby forcing the need for difficult and potentially divisive choices. The more we do for selected segments of the population (e.g., chronic renal failure patients receiving dialysis, AIDS patients, patients with acute MI), the less we are able to do for the remainder. In the next section, we examine the costs of various coronary disease therapies. We then return to the issue of cost-effectiveness at the end of this chapter and examine the ways in which this tool can be applied (and misapplied) in the analysis of these difficult choices. General Issues An economic analysis comparing a new drug, device, or strategy with “conventional” or “usual” care starts with an exploration of the ways in which the new approach will alter costs for the patients involved. At the most basic level, this involves understanding the resource consumption patterns and associated incremental costs of the new approach or technology. For a new interventional procedure, this includes the costs of the equipment and supplies used and the personnel changes required. A careful economic analysis also must determine what diagnostic or therapeutic procedures and what complications are added or averted, along with the cost effects of these changes in practice and outcome. Understanding these relationships is often difficult in practice, and one of the major advantages of empirical data collection over armchair models for cost studies is the frequency with which actual medical practices and outcomes diverge from the expected ideal. In general, three major patterns of cost outcomes are possible when comparing alternative medical strategies or technologies (Fig. 65-1). 1. The new strategy or technology is associated with net higher costs but also provides additional clinical benefits. For example, CABG saves more lives than medical therapy in patients with severe CAD, but it costs more money. In such cases, economic analysis attempts to define the relationship between incremental costs and incremental health benefits, so that policy judgments can be made about whether the new strategy should be adopted. Cost-effectiveness analysis is the technique used to formalize this assessment. In some cases, the new strategy or technology may recoup some of its costs by reducing or preventing costly complications that occur with the comparison approach. For example, coronary stenting reduces the need for repeat revascularization procedures relative to conventional balloon angioplasty. These followup benefits provide at least a partial offset to the initially higher costs of stenting. However, because the total costs of the new strategy are often higher, cost-effectiveness analysis is still required to define its economic attractiveness.

1203

1204

Outcome Effectiveness of Interventional Cardiology New vs. comparison treatment Effectiveness analysis Better outcomes for new treatment

Same outcomes for new treatment

Cost analysis

Cost analysis

Higher costs for new treatment

Same or lower costs for new treatment

Higher costs for new treatment

Lower costs for new treatment

Reject

Economically attractive

Cost-effectiveness analysis Not economically attractive

Economically attractive

Dominant strategy

Figure 65-1. Schematic representation of patterns of outcome and cost differences that may result when a new therapy or strategy is compared with an existing standard. Effectiveness is always considered first, because if the new therapy is less effective, its cost is rarely of concern. If effectiveness is better, or at least equivalent, then costs are prepared. If the outcomes are better but the net costs are higher, cost-effectiveness analysis is then performed. If the costs are equivalent or lower, the therapy is said to be dominant (i.e., it becomes the preferred option). If outcomes are equivalent, cost analysis is used to select the more efficient, less costly option. This form of cost analysis is sometimes referred to as cost minimization or cost-efficiency analysis.

2. The new strategy or technology produces better outcomes and has lower net costs. Use of enoxaparin for patients with acute coronary syndrome (ACS) in the Low Molecular Weight Heparin (Enoxaparin) in the Management of Unstable Angina (ESSENCE) trial is an example. Such strategies are referred to by economists as dominant. 3. The new strategy or technology provides a less expensive, more efficient alternative to conventional therapy with the same benefits. For example, to the extent that PCI provides an equivalent revascularization option to CABG for some CAD patients, it may substantially reduce costs. Coronary Revascularization In 2003, approximately 1.4 million diagnostic cardiac catheterization procedures were done in the United States, followed by over 1.2 million PCIs and almost half a million CABG surgeries.1 Because a hospitalization for PCI costs from $8000 to $12,000 and a hospitalization for CABG costs $30,000 or more, coronary revascularization costs in aggregate probably exceed $30 billion per year. Although the procedures may be more efficient now than in the past owing to shortened hospital stays and lower complication rates, use in patients with more complex disease and advances in technology, especially for PCI, have tended to push costs back up. In this portion of the chapter, we review the available data addressing two key questions: First, what information do we have about how much these procedures cost? Second, what is the value of these procedures, where value refers to the balance between incremental costs and benefits? The literature on interventional cardiology tends to be divided into studies focused on the procedure, which may include both stable and

unstable patients, and studies focused on a portion of the clinical spectrum of CAD, which typically examine strategies of care. Therefore, this section starts with a general review of economics of coronary revascularization. Percutaneous Coronary Intervention The costs of PCI have changed considerably over time, as have the clinical and technical aspects of the procedures. Stenting is now used in more than 80% of PCI cases in the United States, and a similar proportion of patients also receive an intravenous glycoprotein (GP) IIb/IIIa inhibitor. Similar trends have been observed in Canada and Europe, although absolute rates are often lower, depending on the countries considered. Kugelmass and colleagues performed an analysis of the cost of Medicare beneficiaries undergoing PCI and observed the average cost of hospitalization in patients who did not develop complications to be $13,861, whereas patients undergoing PCI who developed complications had costs of $26,807.5 This analysis was based on the fiscal year 2002 Medicare Provider Analysis and Review file containing 336,373 hospital admissions with an International Classification of Diseases (ICD-9) procedure code indicating a PCI procedure during that admission. To understand what determines or “drives” these costs, we must examine four major categories of determinants: patient-specific, hospital-specific, treatment-specific, and geographic-economic. Patientspecific factors such as disease severity affect costs by influencing the type of procedure needed to treat the patient’s CAD, the associated likelihood of success, and the risks of short- and long-term complications (e.g., abrupt closure, restenosis). (Note that this discussion pertains to the cost of an episode of care, not

Medical Economics in Interventional Cardiology simply the procedural costs in the catheterization laboratory.) Procedures in patients with complex lesions (e.g., chronic total occlusions) are more costly, for example, because success rates are lower and long-term durability of successful dilations is reduced. The extent of CAD is also an important cost determinant: Costs are lowest in single-vessel disease and highest in three-vessel disease. In addition, diabetes and older age have been associated with higher costs. Among 1258 patients treated at the Cleveland Clinic from 1992 to 1993, costs of percutaneous revascularization were increased with acute MI (91% increase), recent MI (17% increase), more complex lesion morphology (≥12% increase), diabetes (12% increase), and number of diseased vessels (9% increase per vessel).6 Treatment-related factors include decisions made in the catheterization laboratory as well as management decisions before and after the procedures. In the large Cleveland Clinic series, physician decision delay (i.e., delay between admission and diagnostic catheterization or between catheterization and revascularization) increased hospital costs by 86%, and weekend delay (procedure postponed because of a weekend) increased costs by 61%.6 The cost of a combined diagnostic catheterization/coronary angioplasty procedure at University of California, San Francisco (UCSF) was $850 higher than when these two procedures were performed separately.7 An unsuccessful procedure, particularly with abrupt closure and need for bailout stenting or emergency CABG, is also associated with significantly higher costs. In the Cleveland Clinic series, need for urgent CABG increased costs by 83%, and need for intraaortic balloon pumping increased costs by 42%.6 Newer revascularization technologies are associated with significantly higher costs than those of balloon angioplasty. In addition, use of adjunctive pharmacotherapy also increases costs. In the Evaluation of Platelet IIb/IIIa Inhibitor for Stenting (EPISTENT) trial, use of bare metal stents added $2200 to the hospital costs and use of both abciximab and stents added $3600.8 In Evaluation in Percutaneous Transluminal Coronary Angioplasty to Improve LongTerm Outcome with Abciximab GP IIb/IIIa Blockade (EPILOG), use of abciximab added about $600 net to the hospital costs for balloon angioplasty. In Enhanced Suppression of the Platelet IIb/IIIa Receptor with Integrilin Therapy (ESPRIT), use of eptifibatide added about $300 net to the hospital costs for PCI.9 Less work has been done to define important provider-related cost determinants. In a large-claims database, Topol and colleagues found that care at a teaching hospital resulted in lower charges.10 In 250 patients treated at UCSF, the physician operator was a major cost determinant, with the highest-cost physician averaging $4400 more than the lowest-cost physician.7 This difference was due to a more resourceintensive style of practice and not to disease severity or procedure outcome differences. Another study found that high-volume operators (i.e., ≥50 cases per

year) had slightly lower hospital costs than lowvolume operators (approximately $300 difference, P = .07) despite performing more complex procedures on a higher-risk population.11 Finally, geographic factors have been infrequently studied but appear to explain modest differences in cost on a national level. In one study, costs in the West were highest and costs in the Midwest were lowest.10 Coronary Artery Bypass Graft Surgery In the recently completed Pexelizumab for the Reduction of Infarction and Mortality in Coronary ARtery Bypass Graft Surgery (PRIMO-CABG) trial, the hospital cost of a CABG was almost $30,000, with an additional $4000 in physician fees.12 As with the costs of PCI described in the previous section, the costs of CABG can be analyzed to identify their major determinants. A recent comparison of costs of CABG at five U.S. and four Canadian hospitals using the same cost accounting system showed that U.S. hospital costs were substantially higher ($20,673 versus $10,373; P < .001).13 These differences were not explained by clinical differences, and length of stay in Canada was 17% longer. Patient characteristics typically account for 25% or less of the variance in hospital costs. In 12,807 patients in New York State who underwent CABG in 1992, patient characteristics accounted for 23% of the variance in log costs.14 In a cohort of patients receiving their operation at Emory University, the major determinants of higher costs included worse angina class, previous MI, older age, heart failure, and more extensive CAD, with diabetes being a marginally significant predictor (P = .07).15 In contrast, in the Bypass Angioplasty Revascularization Investigation (BARI) study, CABG in diabetic patients had 5-year costs that were $15,000 higher than in nondiabetics.16 In the New York state data referred to earlier, the major predictors of extremely high costs of CABG included older age, nonelective procedure, ejection fraction less than 50%, repeat heart surgery, diabetes, chronic obstructive pulmonary disease, hepatic failure, chronic renal insufficiency, and transfer from another acute care facility. Three major procedural factors increase CABG costs: use of an internal mammary artery graft, use of cardiopulmonary bypass, and need for additional procedures such as valvular repair or replacement. Complications that occur after the operation (which may reflect a combination of patient, treatmentrelated, and provider factors) substantially increase costs. In the Emory study, important cost drivers of this sort included adult respiratory distress syndrome, septicemia, pneumonia, bleeding requiring reexploration, major arrhythmias, and neurologic events. Patients without any of these complications averaged hospital costs of $16,776 (1990 U.S. dollars); patients with one complication averaged $17,794; those with three complications, $23,624; and those

1205

1206

Outcome Effectiveness of Interventional Cardiology with five complications, $50,609.15 Similar findings were reported from 6791 CABG patients in the Massachusetts Health Data Consortium database.17 Hospital-level provider-related factors “explain” about 40% of the variance in the initial costs of CABG.14 In a Duke analysis, the attending surgeon was the most important determinant of cost among all factors examined.18 The most expensive surgeon had a median cost that was $4200 higher than the least expensive surgeon. As with PCI costs, the explanation for this appears to lie in a more resourceintensive style of practice rather than in a difference in disease severity or in complication rates. Use of the minimally invasive approaches to CABG and valvular surgery is now established as a viable alternative to standard operative approaches for selected patients. These procedures are performed by a subset of heart surgeons. Initial economic work in this area suggested that a minimally invasive direct coronary artery bypass (MIDCAB) procedure has costs that are about half those of a conventional CABG.19 Full sternotomy off-pump CABG is a recently developed variant of CABG that seeks to protect patients from complications attributed to cardiopulmonary bypass. At the Minneapolis Heart Institute, both MIDCAB and off-pump CABG required less operating room time than conventional CABG.20 Mean hospital costs were $17,000 for 44 MIDCAB procedures, $15,000 for 62 off-pump CABGs, and $19,000 for 243 conventional CABGs. In a trial conducted at a single academic medical center to assess clinical, economic, and quality-of-life outcomes, 200 patients were randomized to either elective off-pump CABG or CABG with pulmonary bypass.21 Follow-up out to 1 year showed similar graft patency; rates of death, MI, angina, stroke, and reintervention procedures; and quality-of-life measures. Initial hospitalization costs were $2272 higher in the CABG with bypass arm, and 1-year costs in this arm were $1955 higher. In the Canadian Off-pump CABG Registry, involving 1657 consecutive off-pump patients and 1693 consecutive on-pump patients, the initial hospital costs for off-pump CABG were significantly lower ($11,744 versus $13,720; P < .01).22 These differences persisted out to 1 year. Comparisons of Treatment Options for Coronary Artery Disease Revascularization versus Medical Therapy

For any CAD patient, medical therapy without revascularization is one important therapeutic option that must be considered. In the most recent meta-analysis on the trial data making this comparison, PCI demonstrated no prognostic advantage over initial medical therapy.23 Two recent trials have compared the costs of these two strategies. In the Trial of Invasive versus Medical therapy in Elderly patients (TIME) (which was not included in the previously described meta-analysis), a predefined subgroup of 188 chronic symptomatic CAD patients 75 years or older were

randomized to optimized medical therapy or coronary angiography followed by PCI or CABG at four hospitals in Northern Switzerland and monitored for resource utilization and quality-of-life status.24 At 1 year of follow-up, the rate of death or nonfatal MI was similar in the two arms, whereas the rate of major adverse cardiovascular events was significantly reduced in the invasive strategy arm. These outcomes were similar to the overall trial cohort. The invasive arm was found to have nonsignificantly increased total 1-year costs compared to the medical therapy arm and yielded an economically attractive costeffectiveness ratio. Sculpher and colleagues prospectively collected resource utilization data on 1018 patients in the second Randomized Intervention Treatment of Angina (RITA-2) trial.25 Patients with proven CAD were randomized to an initial strategy of PTCA or continued medical management, and follow-up was conducted out to 3 years. The composite end point of death or MI was higher in patients randomized to the PTCA arm (7.3% versus 4.1%; P = .025). Patients in the PTCA arm were also more likely to undergo subsequent coronary angiograms, whereas those in the medical therapy arm were more likely to undergo a nonrandomized PTCA procedure and use more antianginal medication. Mean resource utilization costs were £2685 higher per patient in the PTCA arm at 3 years. Although interventional therapy is expensive, especially initially, the cost involved with decades of medical therapy should not be underestimated. In the Women’s Ischemia Syndrome Evaluation (WISE), 833 women referred for angiography to evaluate chest pain were followed and their long-term costs were modeled.26 The 5-year cost averaged about $50,000, and patients with three-vessel disease had modestly higher costs than those with single-vessel disease. More than half of the expense was due to cost of hospital-based care, with outpatient medications accounting for an additional 25% to 30% of costs (Fig. 65-2). An important trial, the Clinical Outcomes Utilizing Revascularization and Aggressive DruG Evaluation (COURAGE) trial, randomized 2287 patients with evidence of ischemia and significant coronary artery disease to undergo either PCI with optimal medical therapy or optimal medical therapy alone. The patients were followed for a period of 2.5 to 7 years, with median follow-up of 4.6 years. In this cohort, the addition of PCI to optimal medical therapy did not reduce the risk of death, MI, or other major cardiovascular events. An economic analysis of this trial is forthcoming.27,28,28a The Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) is testing whether coronary revascularization added to state-of-the-art medical therapy in 2800 diabetics improves all-cause mortality. Collection of cost data out to 5 years and performance of a cost-effectiveness analysis are integral parts of the study program.29 The study is currently completing its follow-up phase.

Medical Economics in Interventional Cardiology 60,000 Observed direct costs (2003 US $)

P
.0001 50,000

40,000

30,000

20,000

10,000

0 Cost components Outpatient Drug Hospitalization

Nonobstructive CAD

1 vessel CAD

2 vessel CAD

3 vessel CAD

6.7% 32.6% 60.7%

5.8% 29.5% 64.7%

14.3% 27.0% 58.7%

14.8% 25.5% 59.7%

Figure 65-2. Five-year cardiovascular costs for women with nonobstructive and one- to three-vessel coronary artery disease in the Women’s Ischemia Syndrome Evaluation (WISE) study. (From Shaw LJ, Merz CN, Pepine CJ, et al: The economic burden of angina in women with suspected ischemic heart disease: Results from the National Institutes of Health–National Heart, Lung, and Blood Institute– sponsored Women’s Ischemia Syndrome Evaluation. Circulation 2006;114:894-904.)

PTCA versus CABG

Two major randomized trials have compared PTCA and CABG costs in U.S. patients. The Emory Angioplasty Surgery Trial (EAST) enrolled 392 multivessel CAD patients between 1987 and 1990. For the initial hospitalization, the PTCA patients averaged $11,684 ($16,223 including physician fees), whereas the CABG patients averaged $14,579 ($24,005 with physician fees) (all 1987 dollars).30 At the end of 3 years, the PTCA arm had cumulative costs of $23,734 versus $25,310 for the CABG arm (P < .001). Thus, PTCA was initially 32% less expensive than CABG, but after 3 years it was only 6% less expensive in this population of multivessel CAD patients. Between 3 and 8 years, the PTCA arm averaged $4700 in additional hospitalrelated costs compared with $2700 for the CABG arm. Thus, at the end of 8 years, cumulative costs were $44,491 for PTCA and $46,348 for CABG (P = .37).31 Of the trials comparing PTCA and CABG, BARI was the largest, enrolling 1829 multivessel CAD patients at 18 centers between 1988 and 1991. The BARI Substudy of Economics and Quality of Life (SEQOL) enrolled 934 of these patients at the seven largest enrolling sites.16 Initial costs for the PTCA arm were $14,415 in hospital costs and $6698 in physician fees. For the CABG arm, initial hospital costs were $21,534, with physician fees of $10,813. Total inpatient follow-up costs (hospital and physician fees) were $27,439 for PTCA and $19,529 for CABG. Outpatient follow-up care was equivalent in the two arms. The cumulative 5-year cost for medications was higher in the PTCA arm ($4948 versus $3670). Thus, at the end of 5 years, total discounted costs in the

PTCA arm were 5% lower than in the CABG arm ($56,225 versus $58,889). Five-year costs in the patients with two-vessel disease were significantly lower with angioplasty ($52,390 versus $58,498 for CABG), whereas in three-vessel disease, angioplasty was actually more expensive ($60,918 versus $59,430). At 12 years of follow-up in the BARI SEQOL study, cumulative costs had narrowed to $120,750 for the PTCA arm versus $123,000 in the CABG arm, and the difference was no longer significant (P = .55). CABG yielded a cost-effectiveness ratio of $14,300 per life year added when compared to PTCA at 12 years.32 The British Randomized Intervention Treatment of Angina (RITA) compared the costs of PTCA and CABG (using U.K. costs) in 1011 patients and confirmed the results of EAST and BARI.33 Initial costs of PTCA were half those of CABG, but at the end of 2 years, PTCA costs had risen to 80% of CABG costs. The Angina With Extremely Serious Operative Mortality Evaluation (AWESOME) trial randomized 454 high risk patients with medically refractory angina to either PCI or CABG.34 Costs out to 5 years were estimated using Medicare reimbursement rates. At 5 years, the PCI arm was $18,732 less expensive than the CABG arm. Survival at 5 years was 0.75 for PCI and 0.70 for CABG (P = .21). Bootstrap analysis demonstrated that the PCI arm was more effective clinically and less costly in 89% of repetitions. Bare Metal Stenting versus Balloon PTCA

Several randomized trials have compared primary coronary stenting with conventional balloon PTCA in native coronary vessels of stable patients. Metaanalysis of these data showed no evidence of a reduc-

1207

1208

Outcome Effectiveness of Interventional Cardiology tion in deaths or MIs with stenting.35 The trials also showed equivalent 8- to 12-month reduction in anginal symptoms. The major clinical advantage of stenting is a reduction in the need for repeat revascularization, primarily repeat percutaneous procedures, due to a reduced clinical restenosis rate. In the Belgium Netherlands Stent (BENESTENT) II trial, by 6 months, 14% of PTCA patients and 9% of stent patients had received a repeat revascularization procedure.36 Cohen and colleagues performed an economic analysis of BENESTENT II using U.S. costs along with the resource consumption patterns observed in the trial, which was conducted in Europe.37 They found that initial procedural costs were approximately $1900 higher with stenting and that post-procedure costs were equivalent. Follow-up costs were reduced by half in the stent arm, so that at the end of 6 months, stenting had a net incremental cost of $900 over PTCA. The Optimum Percutaneous Transluminal Coronary Angioplasty Compared with Routine Stent Strategy (OPUS) trial enrolled 479 patients with single-vessel disease.38 In the optimum PTCA arm, 37% of patients required stents. Use of abciximab was low in both arms (13% to 14%). Length of stay averaged 2.5 days and was similar in the two arms. The procedural costs were higher in the routine stenting arm ($5455 versus $4219; P < .001). Total hospital costs showed a similar pattern: $9234 for the stent arm versus $8434 for the PTCA arm (P < .001). However, in follow-up, the PTCA arm had a 15% repeat revascularization rate out to 6 months, compared with 5% in the stent arm. Thus, at 6 months, cumulative costs were $10,206 for stent and $10,490 for PTCA. In 1000 bootstrap replications, the routine stenting arm was less expensive by 6 months 69% of the time. Bare Metal Stenting versus CABG

Although comparisons of stenting with balloon PTCA provide useful information about the value that stents add to the interventionalist’s armamentarium, the more important comparison is between a primary percutaneous strategy and a primary surgical one for coronary revascularization. The Arterial Revascularization Therapies Study (ARTS) compared PCI with stenting versus CABG in 1205 multivessel disease patients.39 By 1 year, the major difference between the two arms was a higher repeat revascularization rate in the stent arm (21.2% versus 3.8% in the CABG arm), a difference that persisted at 3-year follow up (26.7% versus 6.6% in the CABG arm). One-year total costs were *11,117 for the stent group versus *13,896 for the CABG group, a difference of *2779. This difference in costs had narrowed to *1798 at the 3-year follow-up, with the stent group showing *14,302 in total costs versus *16,100 for the surgery group. These cost estimates, which were calculated from limited resource use data and European cost weights, appear to underestimate costs as typically seen in the United States. At 5 years, there was no difference in survival or freedom from death, MI, or stroke between the

two strategies, but repeat revascularization remained higher in the PCI arm (30% versus 9%; P < .001).40 The diabetic subgroup (208 patients) showed a nonsignificant trend toward higher mortality with PCI (13.4% versus 8.3%; P = .27). Between 1996 and 1999, the Canadian/European Surgery or Stent (SoS) trial randomly assigned 988 patients with typical angina and multivessel disease to either stent-assisted PCI or CABG. In an economic analysis of this trial, initial hospitalizations costs were higher in the CABG arm, and follow-up costs were higher in the PCI arm, consistent with previous trials.41 Total 1-year follow-up costs in the subgroup of patients with ACS were $6435 for the PCI arm and $8870 for the CABG arm. Total 1-year costs in the non-ACS subgroup were $5069 for the PCI arm and $7487 for the CABG arm. Patients with ACS had significantly higher initial costs and total 1-year costs. A meta-analysis of four trials comparing CABG versus PCI with mutivessel stenting showed no difference in the composite end point of death, MI, or stroke at 1 year.42 Repeat revascularization was more common among the PCI patients and freedom from angina was better in the CABG patients (82% versus 77%; P = .002). Drug-Eluting Stents versus Bare Metal Stents

Drug-eluting stents (DES) were received with much fanfare at the 2001 European Society of Cardiology meetings, where initial results were reported from the Randomized Study with the Sirolimus-Eluting Velocity Balloon-Expandable Stent (RAVEL) trial. Five years and millions of implanted stents later, at the 2006 European Society of Cardiology meetings a major subject of debate was the emerging reports of a late risk of catastrophic thrombosis with drug-eluting stents. During the intervening years, DES have become the de facto standard of care for PCI in many parts of the world. The recognition of a small but important late risk attached to these devices has stimulated a re-evaluation of the incremental benefits of the technology. Meta-analyses have repeatedly shown that DES, compared with bare metal stents, have no effect on death or nonfatal MI.43 The only end point consistently modified by DES is the need for repeat target lesion revascularization. Further, data from Emory showed that the occurrence of restenosis had no discernible effect on late survival.44 Several of the earlier DES trials have published an economic analysis. Van Hout and colleagues studied the economics of the RAVEL trial.45 RAVEL compared a sirolimus-eluting stent (SES) with a bare metal stent for single de novo coronary lesions in 238 symptomatic patients in Europe and Latin America. At 6 months, restenosis was 0% in the SES arm and 26% in the bare stent arm. The SES arm also had a substantially lower rate of major adverse clinical events. In the economic analysis, although the additional initial procedural cost of the SES was *1286, this difference had been reduced to *54 at 1 year owing to the reduced need for repeat revascularization.

Medical Economics in Interventional Cardiology In the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions (SIRIUS) trial, 1058 patients with complex coronary stenosis were randomized to receive either an SES or a bare metal stent. Cohen and colleagues performed a prospective economic analysis of this trial.46 They found initial hospital costs to be $2881 per patient higher in the SES arm, but this additional cost had attenuated to $309 at 1 year as costs in the SES arm were reduced because of significant reductions in repeat revascularization rates. The incremental cost-effectiveness ratio for SES versus bare metal stent was $1650 per repeat revascularization avoided. One important caveat about these results is that they were substantively influenced by the protocol 8-month repeat coronary angiography, which led to a doubling in the rate of late repeat revascularization procedures in the bare metal stent arm and a consequent narrowing in the cost difference between the two arms. If the extra procedures induced by the protocol angiograms are deleted from the calculations, the net cost of the DES strategy in this trial was $1300. The cost effectiveness of paclitaxel-eluting stents (PES) was prospectively examined by Bakhai and colleagues in the TAXUS-IV trial.47 In this study, 1314 patients undergoing PCI were randomly assigned to either PES or a bare metal stent, with follow-up out to 1 year. The PES arm index hospitalization costs were $2028 per patient higher when compared with the BMS arm. Cost savings of $1456 per patient in the PES arm were realized at 1-year follow-up, primarily owing to reductions in repeat revascularization procedures. The incremental cost effectiveness of PES was $4678 per target vessel revascularization avoided and $47,798 per QALY gained. In the randomized Basel Stent Kasten Effekivitäts Trial (BASKET), 826 consecutive, unselected patients were treated with a DES (both SES and PES were used) or a bare metal stent.48 After 6 months of follow-up, total costs in the DES arm were *10,544, compared with *9639 in the bare metal stent arm (P < .00001). The incremental cost-effectiveness ratio to avoid one major event (cardiac death, MI, or target vessel revascularization) was *18,311. Reperfusion and Revascularization of Acute Coronary Disease Thrombolytic Therapy In the United States, more than 200,000 patients each year get acute reperfusion therapy for ST-elevation acute myocardial infarction (STEMI). The use of thrombolytic therapy has diminished in recent years in favor of primary PCI, and the use of streptokinase is particularly uncommon. In contrast, in the remainder of the world, streptokinase continues to be the most commonly used thrombolytic. With regard to the cost of thrombolytic therapy relative to conventional therapy, the major variable costs are those of the drug itself, the pharmacy costs to ready the drug for administration, and the labor costs of administra-

Thrombolytic agent  Pharmacy preparation and handling Cost of thrombolytic treatment



 Disposable supplies  Nursing time to administer and monitor

Figure 65-3. Cost components of thrombolytic therapy.

tion and monitoring (Fig. 65-3). The drug costs of thrombolytic therapy are now well known: the average wholesale price for 1.5 million U of streptokinase is about $590, for 100 mg of tissue plasminogen activator (t-PA) it is about $2800, and the same price applied to both 30 mg of recombinant tissue plasminogen activator (rt-PA) and 50 mg of tenecteplase (TNK).49 From a research point of view, the most controversial aspect of the cost of thrombolytic therapy pertains to the induced costs and savings: the cost of those aspects of medical care that are added or averted specifically because the patient received thrombolytic therapy. No prospective empirical comparison of thrombolytic therapy versus no reperfusion therapy in the United States has been published. Therefore, although it is possible that the use of streptokinase induces cost savings relative to no reperfusion therapy by reducing postinfarction complications, no empirical data are available to support this hypothesis. In fact, Naylor and Jaglal found in pooled analysis of several small, blinded trials that use of thrombolytic therapy was associated with higher revascularization rates.50 In the absence of adequate empirical data on this question, most economic analysts have chosen to assume that thrombolytic therapy does not influence medical resource use, either in the short term (e.g., length of stay for the acute MI) or in the long term (e.g., rehospitalization). Naylor and colleagues calculated a cost-effectiveness ratio for streptokinase of $2000 to $4000 per life-year added relative to no reperfusion therapy, under the assumptions that each additional survivor produced by streptokinase would live an average of 10 years.51 In summary, these analyses show that the use of streptokinase in lieu of a management strategy without reperfusion therapy is quite economically attractive and might be reasonably considered one of medicine’s “best buys.” Although the cost-effectiveness of treating anterior MIs is most favorable, treatment of inferior MIs and treatment in elderly patients is also economically attractive using conventional benchmarks. In the Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO) I trial, t-PA (alteplase) was shown to save one extra life per 100 patients with acute MI shifted from streptokinase and to produce a higher proportion of TIMI grade 3

1209

1210

Outcome Effectiveness of Interventional Cardiology Table 65-5. Cost-Effectiveness Ratios for t-PA Compared with Streptokinase in the Primary Analysis and Selected Subgroups of Patients in GUSTO I Increased Life Expectancy with t-PA (Years of Life Saved) Patient Groups Primary Analysis Inferior MI ≤40 yr Inferior MI 41-60 yr Inferior MI 61-75 yr Inferior MI > 75 yr Anterior MI ≤40 yr Anterior MI 41-60 yr Anterior MI 61-75 yr Anterior MI > 75 yr

Undiscounted

Discounted

Cost Effectiveness Ratio (Dollars per LY Saved)

0.14 0.03 0.07 0.16 0.26 0.04 0.10 0.20 0.29

0.09 0.01 0.04 0.10 0.17 0.02 0.06 0.14 0.21

32,678 203,071 74,816 27,873 16,246 123,609 49,877 20,601 13,410

GUSTO, Global Use of Strategies to Open Occluded Coronary Arteries trial; LY, life year; MI, myocardial infarction; t-PA, tissue plasminogen activator. Data from Mark DB, Hlatky MA, Califf RM, et al. Cost effectiveness of thrombolytic therapy with tissue plasminogen activator as compared with streptokinase for acute myocardial infarction. N Engl J Med 1995;332:1418-1424.

coronary flow in the infarct vessel. In the prospective GUSTO I economic substudy, substitution of an accelerated t-PA regimen for intravenous streptokinase was economically attractive, with a costeffectiveness ratio of $27,000 to $33,000, depending on the specific assumptions used in the calculations.52 Subgroup analysis showed that t-PA was modestly more cost effective in anterior MIs but was considerably more effective in older patients (Table 65-5). These results were not substantially altered after taking into account the 1 per 1000 extra nonfatal disabling strokes produced by t-PA. In an effort to improve the efficacy of pharmacologic reperfusion, two studies have examined regimens that combine lower-dose thrombolytic therapy with intravenous GP IIb/IIIa inhibitors and with lowmolecular-weight heparin. GUSTO V compared halfdose rt-PA plus full-dose abciximab against full dose rt-PA in 16,588 patients.53 There was no difference in 30-day mortality, but nonfatal MIs were reduced by the combination regimen, as was urgent revascularization. The Assessment of the Safety and Efficacy of a New Treatment Strategy for Acute Myocardial Infarction 3 (ASSENT 3) trial randomized 6095 patients with acute MI to receive one of three regimens: TNK with unfractionated heparin, TNK with enoxaparin, or half-dose TNK with full-dose abciximab. Mortality at 30 days was 5.4% in the enoxaparin arm, 6.6% in the abciximab arm, and 6.0% in the unfractionated heparin arm (P = .25). Corresponding in-hospital reinfarction rates were 2.7%, 2.2%, and 4.2%, respectively (P = .0009). Both experimental arms also decreased in-hospital refractory ischemia (P < .0001) and increased major bleeding other than intracranial hemorrhage (P = .0005). Kaul and associates performed a prospective economic analysis of the ASSENT 3 study and found a trend toward lower costs in the TNK plus enoxaparin group at 30 days, but the difference was not significant.54 Another approach to improving the results of thrombolysis is to couple it with rescue PCI. In the recently published Rescue Angioplasty versus Con-

servative Treatment or Repeat Thrombolysis (REACT) trial, 427 patients with clinically failed thrombolysis were randomized to rescue PCI or conservative therapy.55 The rescue PCI arm had significantly improved event-free survival but no difference in allcause mortality. In a meta-analysis of five trials of rescue PCI, mortality risk was reduced but risk of heart failure and stroke were increased.56 No modern comparison of the economic consequences of these strategies is available. Primary Percutaneous Coronary Reperfusion Between 1990 and 2002, a total of 25 randomized trials compared primary PCI with thrombolysis. Pooling of patient-level data from 22 of these trials revealed that primary PCI was associated with a 37% reduction in 30-day mortality rates.57 This advantage was evident both in patients who presented early and in those who presented longer than 4 hours after symptom onset. Angiographic comparisons showed a higher early reperfusion rate and a higher proportion of patients achieving TIMI grade 3 flow. In addition, two early randomized trials suggested that direct PTCA might actually be a less expensive reperfusion strategy than t-PA.58 The contemporary economic consequences of primary PCI versus thrombolysis have not been studied. The Stenting versus Thrombolysis in Acute Myocardial Infarction (STAT) trial, a small Canadian study of 123 patients, reported initial hospital costs (in 1999 U.S. dollars) of $6354 for primary PCI and $7893 for t-PA (P < .01).59 The cost of t-PA in Canada at the time of the trial was about $1800, whereas the cost of the PCI using bare metal stents was approximately $2100. The study is difficult to interpret, both because it was unblended and small and also because 64% of the patients in the t-PA arm had an unscheduled coronary angiogram during the index hospitalization. An economic analysis of the Comparison of Angioplasty and Pre-hospital Thrombolysis in Acute Myocardial Infarction (CAPTIM) trial found no dif-

Medical Economics in Interventional Cardiology

Cost per quality-adjusted life-year saved (1993 US $)

200,000

150,000

Add night call Add needed lab Add redundant lab

100,000

50,000

0

0

50

100

150

200

250

300

350

400

Number of acute myocardial infarctions per year at admitting hospital

Figure 65-4. Cost per quality-adjusted life-year saved (1993 U.S. dollars) by employing a policy of direct angioplasty. Solid symbols indicate comparison with thrombolysis. Open symbols indicate comparison with no reperfusion therapy (angioplasty dominant over thrombolysis at those points). Solid lines give three hospital scenarios, with most favorable assumption of efficacy based on pre-GUSTO IIb randomized trials. Dashed lines give three hospital scenarios using effectiveness data from community-based observational studies. GUSTO IIb, Global Use of Strategies To Open Occluded Arteries in Acute Coronary Syndromes trial IIb. (From Lieu TA, Gurley J, Lundstrom RJ, et al: Projected cost-effectiveness of primary angioplasty for acute myocardial infarction. J Am Coll Cardiol 1997;30:17411750. Reprinted with permission from the American College of Cardiology.)

ference between the two reperfusion strategies in 1year outcomes but costs were lower for primary PCI.60 The thrombolysis strategy in this trial included the use of rescue PCI as needed. A model-based analysis of these two therapies using European costs concluded that primary PCI had both improved health benefits and lower long-term costs.61 In contrast, a model-based analysis from the U.K. perspective found that primary PCI improved clinical outcomes at a modest increase in cost and was economically attractive in cost-effectiveness analysis.62 Lieu and colleagues demonstrated that the costs of a primary PTCA strategy vary importantly with the structural details of individual hospital programs.63 Using a spreadsheet-based model and cost data obtained from a Kaiser Permanente Hospital, they showed that procedural costs varied from $1600 to $14,300, depending on extra costs for night call, annual procedural volume, and (particularly) the need to construct a new catheterization laboratory to handle the extra volume. Using these data in a decision analytic model along with effectiveness data from published clinical trials and other studies, Lieu and colleagues examined the cost-effectiveness of primary angioplasty.64 In the base case analysis, they considered the case of a hospital with an existing catheterization laboratory with night and weekend coverage that admitted 200 acute MI patients each year. Under these assumptions, direct PTCA was cost-saving compared with thrombolysis (either t-PA or streptokinase regimens) and had a cost of $12,000 per QALY relative to no reperfusion therapy. In sensitivity analyses, the need to build a new catheterization laboratory, the need to add the costs of night call, and a diminution of laboratory volume all significantly increased the cost-

effectiveness ratio (Fig. 65-4).This model demonstrates that the economic attractiveness of primary PTCA relative to thrombolysis depends not only on its incremental effectiveness in improving health outcomes but also on the availability of regional highvolume laboratories with experienced operators. Redundant low-volume laboratories led to worse outcomes and higher costs and made the procedure much less economically attractive. More recent trials have examined the use of stenting versus balloon PCI for acute MI as well as the benefits of adjunctive therapies with balloon- or stent-based PCI. The Primary Angioplasty in Myocardial infarction stent trial (PAMI-STENT) randomized 900 acute MI patients who had a native infarctrelated artery suitable for stenting to routine stenting or balloon PCI. Use of a GP IIb/IIIa inhibitor was low in both arms (5%). At 30 days, the stent arm had a trend toward increased mortality and decreased reinfarction. By 6 months, the stent arm had a 10 per 100 lower rate of target vessel repeat revascularization. In the economic analysis of this trial, the stent arm had a higher index hospitalization cost of about $2000 (due primarily to the cost of the stents used in PCI).65 Over the 12-month follow-up, reduced need for repeat procedures saved about half that amount, leaving a net cost for the stent arm of $1000. The higher mortality trend for stenting makes estimation of cost-effectiveness problematic. In order to improve the early patency associated with the primary PCI strategy, researchers have examined adjunctive use of both GP IIb/IIIa inhibitors and low-dose thrombolytics. Adding abciximab to primary stenting for acute MI improved early (pre-PCI) TIMI grade 3 flow and reduced target vessel repeat revascularization in Abciximab before Direct Angioplasty

1211

1212

Outcome Effectiveness of Interventional Cardiology and Stenting in Myocardial Infarction Regarding Acute and Long-term Follow-up study (ADMIRAL), a 300-patient trial. These results, including no differences in mortality or reinfarction, persisted out to 3 years.66 In the larger Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial, abciximab plus stenting showed a trend toward lower mortality and a reduction in ischemia and ischemic-driven repeat revascularization. Bakhai and coworkers performed a prospective cost-utility analysis in this study, with follow-up out to 1 year.67 Although initial procedural and hospitalization costs were higher in the stent arm compared with PTCA, at 1 year the total costs in both arms were similar, primarily owing to fewer repeat revascularizations in the stent arm. Stenting compared to PTCA produced a cost-effectiveness ratio of $11,237 per QALY gained in this acute MI cohort. Initial costs in the abciximab arm were comparable to those of the standard therapy arm, but abciximab had $1244 higher total costs at 1 year, and the cost-effectiveness of this therapy was uncertain. Early Invasive Versus Early Conservative Strategies In the United States, angiography is frequently performed after acute MI. In the GUSTO I trial, 71% of 21,772 U.S. STEMI patients had a diagnostic catheterization, whereas in the more recent GUSTO III trial, the rate was 78%.68,69 In the non-STEMI portion of the ACS spectrum, the rate is even higher: In the Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy (PURSUIT) trial, 88% of such patients underwent catheterization in the United States during their initial hospitalization. In a managed care environment, rates are lower and surprisingly variable, ranging from 30% to 77% in one study.70 A recent study from the Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the ACC/AHA guidelines (CRUSADE) registry found that only 45% of high risk non-STEMI ACS patients received the early invasive strategy.71 The decision to refer for diagnostic catheterization in acute coronary disease is at best modestly influenced by clinical factors but is substantially influenced by structural aspects of the practice environment, such as the availability of catheterization facilities at the admitting hospital, being admitted by a cardiologist rather than a generalist, and having an attending cardiologist who performs invasive procedures. A systematic review of invasive versus noninvasive management of uncomplicated STEMI identified six clinical trials published between 1988 and 2002.72 In aggregate, these trials did not demonstrate a prognostic benefit for the early invasive strategy. In addition, most of the trials reflected management that is no longer current. Observational studies of the same issue suggest that the early invasive approach may

improve quality-of-life end points, particularly functional status. The durability of that effect is uncertain. Kuntz and colleagues used a decision model to compare routine coronary angiography in the convalescent phase of acute MI hospitalization with medical therapy and exercise testing.73 Because of the lack of clinical trial data in acute coronary disease, the estimates needed for this analysis were collected from trials in chronic coronary disease and from a variety of literature and expert opinion sources. Long-term survival was projected using the Coronary Heart Disease Policy Model of Weinstein and colleagues. Cost data (given in 1994 U.S. dollars) were obtained from Medicare data. In this model-based analysis, routine coronary angiography increased quality-adjusted life expectancy (through its effect on subsequent revascularization) in almost all MI subgroups examined. Only women aged 35 to 44 years with normal ejection fractions, no post-MI angina, and a negative treadmill test appeared not to benefit. When costs were factored into the model, however, the cost to produce an extra unit of benefit varied substantially among subgroups. The most economically attractive cost-effectiveness ratios were obtained in patients with a history of a prior MI (presumably a surrogate marker for greater cumulative left ventricular dysfunction) and in those with inducible myocardial ischemia (Fig. 65-5). Cost-effectiveness ratios with both of these factors ranged between $44,000 and $17,000 per QALY added.73 Conversely, most patients with negative exercise tests had ratios exceeding $50,000 per added QALY. Several major randomized trials have compared early invasive with early conservative management strategies for ACS. The Treat Angina with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy (TACTICS-TIMI 18) trial reflects contemporary management, with coronary stents used in more than 80% of patients undergoing PCI and all patients receiving an intravenous GP IIb/IIIa inhibitor (tirofiban).74 Diagnostic catheterization was performed in 51% of patients receiving early conservative management and 97% of those receiving early invasive management. At 6 months, the early invasive arm had 2 per 1000 fewer deaths and 20 per 1000 fewer MIs. Economic analysis of this trial found that the early invasive arm had higher initial costs ($15,714 versus $14,047) and lower follow-up costs out to 6 months ($6098 versus $7180).75 Cumulative 6-month costs were very similar. Cost per life year added was estimated at $13,000. Fragmin and Revascularization during Instability in Coronary Artery Disease II (FRISC II) compared an invasive strategy (diagnostic catheterization rate, 96%) with a noninvasive strategy (diagnostic catheterization rate, 10%) in 2457 patients with ACS. At 1 year, the invasively treated group had a mortality rate of 2.2%, compared with 3.9% in the noninvasive strategy group (P = .016); the corresponding rates for MI were 8.6% and 11.6%, respectively (P = .015).76

Medical Economics in Interventional Cardiology

AMI

Yes

Angiography

Severe angina? No

Angiography

Yes

M: 45–64?

Yes

No

Prior AMI?

No No

Age 65?

F: 35–44 with ETT not  or not done? or LVEF 0.50 with ETT?

Figure 65-5. Cost-effective use of diagnostic coronary angiography after acute myocardial infarction (AMI). Angio, angiography; CHF, congestive heart failure; ETT, exercise treadmill test; F, female; LVEF, left ventricular ejection fraction; M, male. (From Kuntz KM, Tsevat J, Goldman L, Weinstein MC: Cost-effectiveness of routine coronary angiography after acute myocardial infarction. Circulation 1996;94:957-965.)

Yes

No

LVEF 0.20–0.49 with ETT? or LVEF 0.50 with ETT not  or not done?

No

Yes

Yes

Age 65?

No

Yes Angiography

Angiography

No

The invasive strategy was also associated with a reduction in readmissions (37% versus 57%; P < .001) and (repeat) revascularization (7.5% versus 31%; P < .001). An economic analysis of FRISC II from the Swedish perspective has been published. At the end of 1 year, cumulative costs in the invasive arm were SEK 201,622 (US $20,072) versus SEK 177,746 (US $16,939) for the noninvasive arm, leaving the invasive arm with a $3133 cost at the end of 1 year. Fiveyear follow-up from this trial has recently been published.77 Mortality was 9.7% in the invasive group and 10.1% in the noninvasive group (P = .69). Nonfatal MI rates were 12.9% in the invasive group and 17.7% in the noninvasive group (P = .002). In contrast with TACTICS-TIMI 18 and FRISC II, the Invasive versus Conservative Treatment in Unstable Coronary Syndromes (ICTUS) trial of 1200 ACS patients with an elevated troponins failed to find any benefit in hard cardiac events for the early invasive strategy.78 A meta-analysis of seven trials involving 8,375 patients and a mean follow-up of 2 years reported a 25% reduction in all-cause mortality and a 17% reduction in nonfatal MI for early invasive management.79 However, these results did not take account of the heterogeneous varieties of “early invasive” management represented in the seven trials, with hospital angiography rates ranging from 10% to

CHF and LVEF 0.2–0.49 and ETT not  or not done? Yes

No

CHF with LVEF 0.2–0.49 and age 75–84? or ETT? Yes

No coronary angiography

more than 50%. The more aggressive the “early conservative” arm, the more difficult it is to demonstrate any incremental benefit to routine early angiography. The most optimistic estimates of benefit for the early invasive strategy were seen in the smallest trial and in the trial with the most conservative version of the early conservative management strategy. Antiplatelet Therapy One of the most important advances in interventional cardiology has been the development of platelet GP IIb/IIIa receptor blockers. These drugs have been studied in three types of populations: acute MI, as an adjunct to thrombolytic therapy; non-STEMI ACS; and PCI. The GUSTO V and ASSENT 3 trials, which combined abciximab with thrombolytic therapy, were discussed earlier in this chapter. Use of intravenous GP IIb/IIIa blockers in ACS has now been extensively studied. Among the largest trials, two, Platelet Receptor Inhibition for Ischemic Syndrome Management (PRISM) and Platelet Receptor Inhibition for Ischemic Syndrome Management in Patients Limited by Unstable Signs and Symptoms (PRISM-PLUS), studied tirofiban, while PURSUIT used eptifibatide and GUSTO IV tested abciximab. Of these, only PURSUIT has published an economic

1213

1214

Outcome Effectiveness of Interventional Cardiology analysis.80 In the 3522 U.S. patients enrolled in the trial there was no evidence that eptifibatide use altered resource use or costs through either a positive or a negative effect on clinical course. The absence of an effect on resource use despite reduced ischemic complications may be related to the very high (85%) rate of routine diagnostic catheterization in the U.S. cohort. The cost of the eptifibatide regimen was $1014, and the 3.5% absolute reduction in death or MI at 30 days in the U.S. cohort translated into an incremental life expectancy of 0.11 life years per patient added by eptifibatide. The resulting cost-effectiveness ratio was $13,700 per added life year. In PCI populations, five trials with prospective economic analysis have been concluded, three with abciximab and one each with tirofiban and eptifibatide. In the Evaluation of 7E3 for the Prevention of Ischemic Complications (EPIC) trial, abciximab reduced 30-day ischemic endpoints after high-risk coronary angioplasty by 35% but increased in-hospital bleeding. Six-month ischemic episodes were reduced by 23%. A prospective economic substudy was conducted as part of the EPIC research effort.81 During the initial hospitalization, reduced ischemic events generated a potential cost savings of $622 per patient, but this was canceled out by an equivalent ($521) rise in costs due to care required for major bleeding episodes. Baseline medical costs (hospital plus physician) averaged $14,984 for the abciximab arm (which included $1407 for the bolus and infusion abciximab regimen) versus $13,434 for the placebo arm. During the 6-month study follow-up, the abciximab arm had a 23% decrease in rehospitalizations and a 22% decrease in repeat revascularizations, generating a mean cost savings of $1207 per patient (P = .02). Combining baseline and follow-up costs for each treatment arm yielded a net incremental 6-month cost for the abciximab strategy of $293 per patient. The Evaluation in Percutaneous Transluminal Coronary Angioplasty to Improve Long-Term Outcome with Abciximab GP IIb/IIIa Blockade (EPILOG) trial was conducted to validate and extend the findings of EPIC in a broader PTCA population. It also examined the benefits of using a weight-adjusted low-dose heparin regimen to reduce the early bleeding complications seen in EPIC. In 2792 patients, EPILOG demonstrated a 58% reduction in ischemic complications (death, MI, urgent revascularization) in the abciximab arm by 30 days, with elimination of excess major bleeding episodes by the modified heparin dosing regimen. Prospective cost analysis showed total baseline medical costs for the abciximab arm of $10,125 (including a $1457 cost for the abciximab regimen), compared with $9632 for placebo, resulting in a $583 excess cost for the abciximab strategy.82 EPILOG confirmed the prediction by EPIC that a reduction in ischemic complications without an excess of major bleeds would produce a cost offset of about $600 (EPIC estimate, $622; EPILOG cost saving, $583).

Follow-up data from EPILOG did not show the reduction in subsequent hospitalizations and procedures that had been observed in EPIC: Costs for the abciximab low-dose heparin arm to 6 months were $4221, compared with $3568 for placebo (P = .43). Cumulative 6-month incremental costs for the abciximab arm were $1236. The EPISTENT trial extended the EPILOG results to the use of stenting. Patients were randomized to stenting with abciximab, stenting with placebo, or balloon angioplasty with abciximab. Prospective economic analysis in the 1438 U.S. patients showed that costs for the baseline hospitalization differed only in the costs of the stents and abciximab used: $13,228 for stent plus abciximab, $11,923 for stent plus placebo, and $11,357 for balloon plus abciximab.8 Thus, stent plus abciximab therapy was $1300 more expensive than stent plus placebo and almost $1900 more expensive than balloon plus abciximab arm. In follow-up, the two stent groups had similar costs, whereas the balloon plus abciximab group exceeded these groups by more than $900. Therefore, at 1 year, the stent plus abciximab arm was $930 more expensive than balloon plus abciximab. At 1 year, 1.0% of the stent plus abciximab group had died, compared with 2.4% of the stent plus placebo patients and 2.1% of balloon plus abciximab patients. Incremental life expectancy was projected from these data and used to calculate cost-effectiveness ratios. For stent plus abciximab relative to stent plus placebo, the cost to add a life year was $6200; relative to balloon plus abciximab, the cost to add a life year was $5300. The Randomized Efficacy Study of Tirofiban for Outcomes and Restenosis (RESTORE) trial randomized 2212 patients with an ACS referred for PCI to tirofiban or placebo.83 At 30 days, the primary end point (death, MI, urgent revascularization) was reduced 16% by tirofiban (P = .16). The treatment effect was greatest within the first 48 hours and appeared to attenuate somewhat thereafter. There was no excess of major bleeding seen with tirofiban. Economic analysis of 820 patients at 30 U.S. sites showed a hospital cost of $12,145 for placebo versus $12,230 for tirofiban.83 The cost of the tirofiban regimen itself was estimated at $700, which was largely offset by a reduced need for repeat PTCA or for CABG. The ESPRIT trial (1999-2000) randomized 2064 patients undergoing PCI with stenting to either eptifibatide or placebo. The eptifibatide arm had reduced ischemic complications. The cost of the eptifibatide regimen was $495. Total hospital costs (including eptifibatide) were $10,721 for the eptifibatide arm versus $10,430 for placebo, an incremental cost of $291 per patient.9 Follow-up costs out to 1 year were $2121 for eptifibatide versus $2254 for placebo. Therefore, the net 1-year costs of the eptifibatide strategy were $146. At 1 year, eptifibatide was associated with 6 per 1000 fewer deaths (P = .28) and 35 per 1000 fewer MIs (P = .004). This translated into

Medical Economics in Interventional Cardiology 0.104 added life years per patient (discounted at 3%) and a cost-effectiveness ratio of $1407 per added life year. Several studies have examined the economics of clopidogrel use after PCI. In the PCI-CURE substudy of the Clopidogrel in Unstable angina to Prevent Recurrent Events (CURE) trial (1998-2000), 2658 patients were randomized to clopidogrel or placebo after PCI. Clopidogrel reduced the rate of cardiovascular death or MI by 25% but had no effect on cardiovascular death alone. The cost per day of the clopidogrel therapy was $3.22, which was largely offset by reductions in complications, so that, at the end of 1 year, the net cost of the clopidogrel strategy was between $250 and $425.84 Incremental cost effectiveness was less than $5000 per life year gained, making clopidogrel use in this setting highly economically attractive. Two decision-model-based analyses of the use of clopidogrel for 1 year after PCI came to similar conclusions with a cost of US $15,696 per life year saved in the United States and ä10,993 in Sweden.85,86 Antithrombin Therapy Based on several small trials showing a reduction in death and nonfatal MI, heparin has been standard therapy for moderate- or high-risk patients with unstable angina. Several studies have been conducted with low-molecular-weight heparins in patients with ACS. Enoxaparin was the first low-molecular-weight heparin to be tested in a clinical trial that incorporated a prospective economic analysis. The Efficacy and Safety of Subcutaneous Enoxaparin in Non-Q Wave Coronary Events (ESSENCE) trial randomized 3171 patients with non-STEMI ACS in North and South America and Europe. At the end of 14 days, the enoxaparin therapy had resulted in a significant 15% lower rate of death, MI, or refractory angina (the primary end point) relative to standard unfractionated heparin therapy. Economic analysis of this trial, performed in the 923 patients randomized in the United States, showed that the enoxaparin regimen cost $75 more than standard heparin therapy. At the end of the initial hospitalization, the enoxaparin arm had not only recouped this cost difference but had also produced a cost savings of more than $700 owing to reduced invasive cardiac procedures and a small concurrent reduction in length of stay. At 30 days, the cost advantage for the enoxaparin arm had risen to $1100. The Superior Yield of the New Strategy of Enoxaparin Revascularization and Glycoprotein IIb/IIIa Inhibitors (SYNERGY) trial compared subcutaneous enoxaparin versus intravenous heparin in 10,027 patients with non-STEMI ACS (2001-2003).87 No difference was observed in the primary end point of death or nonfatal MI. In addition, enoxaparin was associated with a moderate increase in major bleeding. Compared with earlier trials making this same comparison, the SYNERGY patients were older and both the medical and interventional therapies were

more aggressive. Whether these differences explain the lack of superiority for enoxaparin in SYNERGY remains unclear. The FRISC II investigators prospectively examined the cost-effectiveness of extended treatment with dalteparin in 2267 patients with unstable coronary artery disease.88 After a minimum of 5 days’ open label dalteparin treatment, patients were randomized to 3 months of subcutaneous dalteparin twice daily or placebo. The dalteparin arm produced a significant reduction in death or MI at 1 month, at an additional cost of SEK 849. The cost-effectiveness ratio was SEK 30,300 per death or MI avoided. Despite these early gains, no significant difference was detected in death or MI between the two arms at 3 months. Several newer and more potent antithrombin agents have been tested in clinical trials. At present, bivalirudin appears the most promising. Bivalirudin plus provisional GP IIb/IIIa inhibition was found to provide similar cardiac event rates out to 1 year while reducing the risk of severe bleeding events when compared with routine GP IIb/IIIa inhibition plus heparin in the Randomized Evaluation in PCI Linking Angiomax to Reduced Clinical Events (REPLACE)-2 trial, a randomized, double-blind trial of 4651 U.S. patients undergoing nonemergent PCI.89 Cohen and colleagues conducted a prospective economic evaluation of this study and found that bivalirudin produced a cost savings of $375 to $400 per patient at 30 days, primarily because of the reduction in bleeding episodes.90 Preventive Therapies A number of secondary prevention programs have been shown to be effective in large-scale clinical trials and have also been shown to be economically attractive. These include statin therapy for hypercholesterolemia, smoking cessation interventions, aspirin therapy, β-blockers for post-MI patients, and angiotensin-converting enzyme inhibitors for post-MI survivors with depressed left ventricular function. These data have been reviewed.2

COST EFFECTIVENESS AND HEALTH POLICY From 1994 to 1999, managed care had an almost unprecedented inhibitory effect on the growth of health care spending in the United States. During this period, the proportion of the Gross Domestic Product (GDP) devoted to health care remained stable at 13% to 13.3%. The era of managed care as a restraining influence on health care spending now appears finished. Growth in annual medical spending has gone back up, and current estimates are that the United States is devoting 16% of its GDP to health care. As of 2007, no significant federal efforts to control health care spending have been proposed. Nonetheless, experience from the last 40 years suggests that employers (who pay for much of the health insurance in the United States) and the federal government (who pays for the rest through the Medicare

1215

1216

Outcome Effectiveness of Interventional Cardiology and Medicaid programs) will not tolerate unchecked growth in medical spending. Any future health care reform that succeeds in restraining total annual U.S. medical care spending (public and private) will indirectly force patients to compete for health care resources and will directly force providers and policy makers to make more explicit decisions about which types of care are worthwhile. For example, should all PCI patients receive the next generation of more expensive DES? What about uncomplicated inferior MIs? Should patients older than 80 years of age be offered revascularization procedures or receive medical therapy only? As described in this chapter, defining the cost of medical care is an important first step. But when the overall goal is to decide which types of care to provide from among the available alternatives, cost must be explicitly tied to the health benefits produced, and the benefits must be valued in some uniform, comparable currency. A host of controversial and complex allocation decisions will be required if providers can no longer count on payers to cover all the care they believe is appropriate for their patients. In the view of some, cost-effectiveness analysis provides the soundest and most logical method for making decisions about health care allocation. Canada, the United Kingdom, and other countries that seek to maximize health benefits within a fixed budget are actively debating the use of cost-effectiveness as a decision aid. The Oregon Medicaid reform plan attempted to make cost-effectiveness the primary criterion for determining what to cover in the program. One important lesson from the Oregon experience is that current methodology and data are insufficient to permit a comprehensive cost-effectiveness ranking of all health care services. The insufficient database on cost-effectiveness is not a critical limitation; if desired, an aggressive empirical research program could rapidly improve the available information base. In addition, policy decisions would not depend on an encyclopedia covering all possible health care expenditures, as was attempted in Oregon. Some therapies would clearly be cost-effective (e.g., aspirin for acute MI), whereas others would clearly not be (e.g., a treatment that is both less effective and more costly than “usual care”). More serious challenges to the use of costeffectiveness analysis as the primary tool for health care spending decisions come from methodologic considerations. Principal among these is the absence of any uniform method for measuring either health care benefits or costs for use in cost-effectiveness analysis. On the effectiveness side, several major problems face the analyst. Chief among these is the difficulty in accurately estimating the change in life expectancy attributable to a new therapeutic strategy. Unless the disease process under study is rapidly fatal, it is virtually impossible to obtain timely empirical life-expectancy data for use in cost-effectiveness calculations. Most commonly, analysts use survival

or mortality rate measures at selected times, such as 30 days, 1 year, or 5 years, and attempt to project life expectancy from these figures with the aid of simple parametric survival modes. Examples include the Markov model and the declining exponential approximation of life expectancy (DEALE). Other, more empirical approaches have recently been developed.80,91 Problems also arise in interpreting statistical survival estimates in life-expectancy terms. For example, consider a therapy that lowers 30-day mortality rates for acute MI patients by 1%, with no additional therapeutic effect (i.e., survival curves are parallel after 30 days). Many would conclude that the therapy saves an average of 1 life per 100 treated. However, the “saved patient” is a statistical patient and cannot be clinically identified. Furthermore, it is uncertain whether only one patient benefits from therapy (by having his or her life prolonged) and 99 receive no benefits or whether some or all treated patients receive benefits, the average effect of which is to decrease the mortality rate by 1%. The distinction is relevant because life-expectancy projections must assume either that one patient is being saved and has some estimated number of life-years remaining or that all patients are being benefited somewhat and the additional life expectancy should be estimated for the group. Because there is no current way of validating such projections and no gold standard to aid in choosing from among competing methods for making them, different analysts may come to substantially different conclusions about the effects on life expectancy of a given therapy. Substantial additional difficulties arise if the analyst wishes to move from survival (cost-effectiveness analysis) to quality-adjusted survival (cost-utility analysis). No standard approach has been agreed on for quality adjustment of life expectancy figures, although QALYs have been used most often. One of the major assumptions of the QALY construct is that, for a given person, it is possible to identify some point of indifference between living x years with some disease state or health impairment and living y years (where y < x) in full health. Another assumption is that improving the quality of life for one type of patient may provide more societal “value” than saving the life of another type of patient. Three main approaches have been used to derive the quality weights or utilities employed in calculating QALYs: the standard reference gamble, the time trade-off technique, and category rating scales. In each case, the resulting measure is presumed to reflect patient preferences in a way that would allow prediction of their future economic behavior (specifically, their purchase of future health care). The advantage of the QALY for economic analysis is that it permits reduction of all survival, quality-oflife, and patient preference issues to a single measure. Whether such a reduction is valid or even appropriate continues to be vigorously debated among health policy investigators. Critics of the QALY construct have pointed out that some of its major assumptions

Medical Economics in Interventional Cardiology are probably invalid. For example, none of the current estimation methods take into account that good health is likely to be more highly valued at some stages in life (e.g., childrearing years) than at others. In addition, calculation of QALYs requires the assumption that patient preferences for different health states do not depend on the length of time spent in those states. Important technical issues relating to QALYs remain unsettled. For example, should ratings of disease states be obtained from patients who actually have the disease, from members of the general public asked to imagine that they have the disease, or from members of the medical profession familiar with the disease and its manifestations? Some data suggest that patient preferences (and consequently QALYs) may not be stable over time and that they are strongly influenced by the context of the assessment in which they are measured. Some have argued that QALYs and other utility scales are uniquely personal and cannot be averaged across a population, as is required in cost-effectiveness analysis. Furthermore, use of average utilities in cost-effectiveness analyses maximizes community preferences over those of the individual and raises important ethical questions. As might be expected, recent research has shown that different approaches to calculating quality-adjusted survival may yield significant differences in resulting cost-effectiveness ratios. Finally, cost-effectiveness studies may differ substantially in the methodology used to estimate costs. Some use carefully measured data obtained from randomized trials or observational studies, whereas others use expert opinion. As discussed earlier in the chapter, actually two factors must be estimated: what resources were consumed in what quantities and the associated unit costs. Often, none of the cost data

used in cost-effectiveness models is derived from empirical research. In addition, substantial assumptions are required to project the lifetime health care costs of a cohort of patients. Although the importance of such assumptions can be examined through sensitivity analyses, as with life expectancy estimates, these data are rarely tested against empirical observations. Thus, cost-effectiveness ratios are calculated by multiplying two imprecise measures—life expectancy gained and utilities of those years—and dividing the result by a third imprecise measure, the lifetime incremental costs. The resulting measure is necessarily imprecise. Despite these substantial limitations, which can severely limit the validity of comparing cost-effectiveness ratios from different studies, “league tables” providing such comparisons are common (Table 65-6). These figures, which are usually presented without any variability or distributional information, appear misleadingly precise and rigorous. Several groups have proposed standards for costeffectiveness analysis, thereby addressing some of the potential weaknesses described earlier. The Panel on Cost Effectiveness in Health and Medicine convened by the U.S. Public Health Service, in particular, has generated a carefully researched monograph that should help to advance work in this area.92 A number of commentators have noted that if cost-effectiveness assessment were an expensive new technology (which, in a sense, it is), health policy analysts would demand rigorous evaluations before it was unleashed on the public. Cost-effectiveness analysis is extremely useful when it focuses and informs professional and public debate, and much additional research is needed in this area. To use it as the primary method of medical resource allocation, however, implies that it possesses

Table 65-6. Cost-Effectiveness and Use of Selected Interventions in the Medicare Population* Intervention

Cost-Effectiveness (Cost/QALY)†

Influenza vaccine Pneumococcal vaccine β-Blockers after myocardial infarction Mammographic screening Colon-cancer screening Osteoporosis screening Management of antidepressant medications Hypertension medication (DBP > 105 mm Hg) Cholesterol management, as secondary prevention Implantable cardioverter-defibrillator Dialysis in end-stage renal disease Lung-volume-reduction surgery Left ventricular assist devices Positron emission tomography in Alzheimer’s disease

Cost saving Cost saving