Clinical Cardiac Pacing, Defibrillation and Resynchronization Therapy, Third Edition

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

CLINICAL CARDIAC PACING, DEFIBRILLATION, AND RESYNCHRONIZATION THERAPY Copyright © 2007 by Saunders, an imprint of Elsevier Inc.

ISBN-13: 978-1-4160-2536-8 ISBN-10: 1-4160-2536-7

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 Health Sciences Rights Department in Philadelphia, PA, USA: phone: (+1) 215 239 3804, fax: (+1) 215 239 3805, e-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://www.elsevier.com), by selecting “Customer Support” and then “Obtaining 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 their own 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 Editors assume 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 Previous editions copyrighted 2000, 1995 Library of Congress Cataloging-in-Publication Data Clinical cardiac pacing, defibrillation, and resynchronization therapy / [edited by] Kenneth A. Ellenbogen, G. Neal Kay, Chu-Pak Lau.—3rd ed. p. ; cm. Includes bibliographical references and index. ISBN 1-4160-2536-7 1. Cardiac pacing. 2. Defibrillators. I. Ellenbogen, Kenneth A. II. Kay, G. Neal. III. Lau, Chu-Pak. [DNLM: 1. Cardiac Pacing, Artificial. 2. Defibrillators, Implantable. 3. Pacemaker, Artificial. WG 168 C6413 2007] RC684.P3C54 2006 617.4′120645—dc22 2006040047

Executive Publisher: Natasha Andjelkovic Editorial Assistant: Katie Davenport Publishing Services Manager: Frank Polizzano Project Manager: Michael H. Goldberg Design Direction: Ellen Zanolle Multimedia Producer: Bruce Robison Marketing Manager: Dana Butler

Working together to grow libraries in developing countries Printed in China Last digit is the print number: 9 8 7 6 5 4 3 2 1

www.elsevier.com | www.bookaid.org | www.sabre.org

To my wife and family, Phyllis, Michael, Amy, and Bethany, for their patience, support, and love. To my parents, Roslyn and Leon, who instilled in me a thirst for learning. To my students, teachers, and colleagues, who make each day an absolute delight. KAE

To my teachers, colleagues, and students, who have taught me about cardiac pacing. I am also indebted to the many members of the industry who have dedicated their professional careers to the design and improvement of pacing technology. These individuals have greatly improved the therapy that clinicians can offer to their patients, undoubtedly resulting in an improvement in their lives. Perhaps most important, this book is dedicated to my wife, Linda, for her patience and understanding during its preparation. GNK

To my wife and family, Carven, Yuk-Fai, and Yuk-Ming, for their understanding, support, and love. To my teachers, patients, and colleagues, who are my source of inspiration and encouragement. CPL

To my wife, Ellyn, and children Jacob, Benjamin, Kara, and Ephram for their godly and inspirational patience and support. To my granddaughter, Isabelle, for life, hope, and love. To my parents, Harvey and Glenna, for their unconditional love and insights. To Yeshua, the Messiah, for His salvation, and His sustaining covenant love. And for the inspiration of His words in Proverbs 15:2: “The tongue of the wise makes knowledge acceptable.” May the words of this book prove to be wise and useful to the student of cardiac pacing, defibrillation, and heart failure device therapy. BLW

Contributors

Amin Al-Ahmad, MD Assistant Professor of Medicine, Stanford University School of Medicine; Associate Director, Stanford Hospital and Clinics, Stanford, California Timing Cycles of Implantable Devices

Angelo Auricchio, MD, PhD Fondazione Cardiocentro Ticino, Lugano, Switzerland Basic Physiology and Hemodynamics of Cardiac Pacing

Peter H. Belott, MD, FACC, FHRS Director, Electrophysiology, Sharp Grossmont Hospital, La Mesa, California Permanent Pacemaker and Implantable Cardioverter-Defibrillator Implantation

Alan D. Bernstein, MD Division of Surgical Research, New Jersey Pacemaker and Defibrillator Evaluation Center, Newark Beth Israel Medical Center, Newark, New Jersey Pacemaker, Defibrillator, and Lead Codes

Charles L. Byrd, MD Director of Electrophysiology Institute, Broward General Medical Center, Fort Lauderdale, Florida Managing Device-Related Complications and Transvenous Lead Extractions

Henry Chen, MD Cardiac Electrophysiology Fellow, Stanford University School of Medicine and Stanford Hospital and Clinics, Stanford, California Timing Cycles of Implantable Devices

Mina K. Chung, MD Associate Professor, Cleveland Clinic College of Medicine of Case Western Reserve University; Attending Physician, Cleveland Clinic, Cleveland, Ohio Imaging in Pacing and Defibrillation

Stuart J. Connolly, MD Professor, Faculty of Health Sciences, McMaster University; Chief of Cardiology, Hamilton Health Sciences, Hamilton, Ontario, Canada Clinical Trials of Pacing Modes vii

viii

Contributors

Ann M. Crespi, PhD Medtronic, Inc., Minneapolis, Minnesota Power Systems for Implantable Pacemakers, Cardioverters, and Defibrillators

Teresa De Marco, MD Professor of Medicine, University of California, San Francisco, School of Medicine; Director of Heart Failure and Pulmonary Hypertension Program and Medical Director of Heart Transplantation, UCSF Medical Center, San Francisco, California Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators

Kenneth A. Ellenbogen, MD Kontos Professor of Cardiology, Virginia Commonwealth University School of Medicine; Director, Electrophysiology and Pacing Laboratory, Medical College of Virginia and McGuire Veterans Affairs Medical Center, Richmond, Virginia Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators; Pacing for Atrioventricular Conduction System Disease

Andrew E. Epstein, MD Professor of Medicine, Division of Cardiovascular Disease, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama Troubleshooting of Implantable Cardioverter-Defibrillators

Derek V. Exner, MD, MPH, FRCPC Assistant Professor, Faculty of Medicine, University of Calgary; Libin Cardiovascular Institute of Alberta, Calgary, Alberta, Canada Clinical Trials of Defibrillator Therapy

Jeffrey M. Gillberg, MS Senior Staff Scientist, Medtronic, Inc., Minneapolis, Minnesota Sensing and Detection

Anne M. Gillis, MD, FRCPC Professor of Medicine, Faculty of Medicine, University of Calgary; Medical Director, Pacing and Electrophysiology, Calgary Health Region, Calgary, Alberta, Canada Pacing for Sinus Node Disease: Indications, Techniques, and Clinical Trials

Lorne J. Gula, MD, FRCPC Assistant Professor of Medicine, University of Western Ontario Faculty of Medicine and Dentistry; Electrophysiologist, Arrhythmia Service, London Health Sciences Centre, London, Ontario, Canada Follow-up and Interpretation of Implantable Syncope Monitors

Jeff S. Healey, MD, MSc Assistant Professor, Faculty of Health Sciences, McMaster University; Cardiologist and Electrophysiologist, Hamilton Health Sciences, Hamilton, Ontario, Canada Clinical Trials of Pacing Modes

Ahmad Hersi, MD Cardiac Electrophysiology Fellow, Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada Evolving Indications for Pacing: Hypertrophic Cardiomyopathy, Sleep Apnea, Long QT Syndromes, and Neurally Mediated Syncope Syndromes

Contributors

ix

Henry H. Hsia, MD Associate Director, Stanford University School of Medicine; Associate Director, Cardiac Electrophysiology Laboratory, Stanford Hospital and Clinics, Stanford, California Timing Cycles of Implantable Devices

Raymond E. Ideker, MD, PhD Professor, Division of Cardiology, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

Bharat K. Kantharia, MD, FRCP, FACC, FESC Associate The Ohio The Ohio Approach

Professor, Department of Internal Medicine, Division of Cardiology, State University; Director, Cardiac Electrophysiology Laboratories, State University Medical Center, Columbus, Ohio to Generator Change

G. Neal Kay, MD Professor of Medicine, Division of Cardiovascular Disease, and Director of Clinical Electrophysiology Section, University of Alabama at Birmingham, Birmingham, Alabama Cardiac Electrical Stimulation; Sensor Driven Pacing: Device Specifics

George J. Klein, MD, FRCPC, FACC Professor of Medicine, University of Western Ontario Faculty of Medicine and Dentistry; Chief, Division of Cardiology, London Health Sciences Centre, London, Ontario, Canada Follow-up and Interpretation of Implantable Syncope Monitors

Andrew D. Krahn, MD, FRCPC, FACC Professor of Medicine, University of Western Ontario Faculty of Medicine and Dentistry; Director of Education, Division of Cardiology, London Health Sciences Centre, London, Ontario, Canada Follow-up and Interpretation of Implantable Syncope Monitors

Mark W. Kroll, PhD Adjunct Professor of Biomedical Engineering, California Polytechnic State University, San Luis Obispo, California, and University of Minnesota, Minneapolis, Minnesota Pacemaker and Implantable Cardioverter-Defibrillator Circuitry; Testing and Programming of Implantable Defibrillator Functions at Implantation

Steven P. Kutalek, MD Associate Professor of Medicine and Pharmacology, Drexel University College of Medicine; Director of Clinical Cardiac Electrophysiology and Associate Chief of Division of Cardiology, Hahnemann University Hospital, Philadelphia, Pennsylvania Approach to Generator Change

Dhanunjaya R. Lakkireddy, MD Clinical Assistant Professor, Kansas University Medical Center; Staff Electrophysiologist, Mid America Cardiology at Kansas University Medical Center, Kansas City, Kansas Imaging in Pacing and Defibrillation

Chu-Pak Lau, MD William M. W. Mong Professor in Cardiology, Cardiology Division (Chief), The University of Hong Kong; Chief, Cardiology Division, Queen Mary Hospital, Hong Kong Sensors for Implantable Devices: Ideal Characteristics, Sensor Combination, and Automaticity; Sensor Driven Pacing: Device Specifics

x

Contributors

Sarah S. LeRoy, MSN Pediatric Nurse Practitioner, University of Michigan Congenital Heart Center, University of Michigan Health System, Ann Arbor, Michigan Pediatric Pacing and Defibrillator Use

Paul A. Levine, MD Vice President, Medical Services of the Cardiac Rhythm Management Division, St. Jude Medical, Inc., St. Paul, Minnesota Pacemaker and Implantable Cardioverter-Defibrillator Circuitry

Charles J. Love, MD Professor of Clinical Medicine, The Ohio State University College of Medicine, Columbus, Ohio Pacemaker Troubleshooting and Follow-up

J. March Maquilan, MD, FACS Clinical Senior Instructor, Hahnemann University Hospital, Philadelphia; Attending Cardiothoracic Surgeon, St. Mary Medical Center, Langhorne, Pennsylvania Approach to Generator Change

Francis E. Marchlinksi, MD Professor of Medicine, University of Pennsylvania School of Medicine; Director of Electrophysiology, University of Pennsylvania Health System, Philadelphia, Pennsylvania Engineering and Construction of Pacemaker and Implantable Cardioverter-Defibrillator Leads

Carlos A. Morillo, MD Professor, Faculty of Health Sciences, McMaster University; Cardiologist and Electrophysiologist, Hamilton Health Sciences, Hamilton, Ontario, Canada Clinical Trials of Pacing Modes

Hideo Okamura, MD Research Associate, Division of Cardiology, Department of Internal Medicine, National Cardiovascular Center, Suita, Osaka, Japan Timing Cycles of Implantable Devices

Walter H. Olson, MD Senior Research Fellow, Bakken Foundation, Tachyarrhythmia Research, Medtronic, Inc., Minneapolis, Minnesota Sensing and Detection

Victor Parsonnet, MD Medical Director, New Jersey Pacemaker and Defibrillator Evaluation Center; Director of Surgical Research, Newark Beth Israel Medical Center, Newark, New Jersey Pacemaker, Defibrillator, and Lead Codes

Sergio L. Pinski, MD Clinical Associate Professor of Medicine, University of South Florida College of Medicine, Tampa; Head, Section of Cardiac Pacing and Electrophysiology, Cleveland Clinic Florida, Weston, Florida Electromagnetic Interference and Implantable Devices

Contributors

xi

Frits Prinzen, PhD Associate Professor of Physiology, Maastricht University, Maastricht, The Netherlands Basic Physiology and Hemodynamics of Cardiac Pacing

Shahbudin H. Rahimtoola, MD Distinguished Professor and GC Griffith Professor of Cardiology; Chairman, Griffith Center, University of Southern California, Los Angeles, California Pacing for Atrioventricular Conduction System Disease

Dwight W. Reynolds, MD Professor and Chief, Cardiovascular Section, The University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma Permanent Pacemaker and Implantable Cardioverter-Defibrillator Implantation

Anthony Rorvick Medtronic, Inc., Minneapolis, Minnesota Power Systems for Implantable Pacemakers, Cardioverters, and Defibrillators

Andrea M. Russo, MD Clinical Associate Professor of Medicine, University of Pennsylvania School of Medicine; Director, Electrophysiology Laboratory, Penn-Presbyterian Medical Center, University of Pennsylvania Health System, Philadelphia, Pennsylvania Engineering and Construction of Pacemaker and Implantable Cardioverter-Defibrillator Leads

Elizabeth Saarel, MD Assistant Professor of Pediatric Cardiology, University of Utah School of Medicine; Attending Physician, Pediatric Cardiology, Primary Children’s Medical Center, Salt Lake City, Utah Imaging in Pacing and Defibrillation

Leslie A. Saxon, MD Professor of Medicine (Clinical Scholar), University of Southern California Keck School of Medicine; Director, Cardiac Electrophysiology, USC University Hospital, Los Angeles, California Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators

Craig L. Schmidt, PhD Medtronic, Inc., Minneapolis, Minnesota Power Systems for Implantable Pacemakers, Cardioverters, and Defibrillators

Gerald A. Serwer, MD Professor, University of Michigan Medical School; Director of Pacing Services, University of Michigan Congenital Heart Center, University of Michigan Health System, Ann Arbor, Michigan Pediatric Pacing and Defibrillator Use

Robert S. Sheldon, MD, PhD Professor of Cardiac Sciences and Associate Dean for Clinical Research, Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada Evolving Indications for Pacing: Hypertrophic Cardiomyopathy, Sleep Apnea, Long QT Syndromes, and Neurally Mediated Syncope Syndromes

xii

Contributors

Richard B. Shepard, MD Emeritus Professor, Division of Cardiothoracic Surgery, University of Alabama at Birmingham, Birmingham, Alabama Cardiac Electrical Stimulation

Allan C. Skanes, MD, FRCPC Assistant Professor of Medicine, University of Western Ontario Faculty of Medicine and Dentistry; Director, Electrophysiology Laboratory, Arrhythmia Service, London Health Sciences Centre, London, Ontario, Canada Follow-up and Interpretation of Implantable Syncope Monitors

Paul M. Skarstad, PhD Medtronic, Inc., Minneapolis, Minnesota Power Systems for Implantable Pacemakers, Cardioverters, and Defibrillators

Julio C. Spinelli, PhD Guidant Corporation, St. Paul, Minnesota Basic Physiology and Hemodynamics of Cardiac Pacing

Bruce S. Stambler, MD Professor of Medicine, Case Western Reserve University School of Medicine; Attending Physician, University Hospitals of Cleveland, Cleveland, Ohio Pacing for Atrioventricular Conduction System Disease

Michael O. Sweeney, MD Assistant Professor, Harvard Medical School; Attending Physician, Cardiac Arrhythmia Service, Brigham and Women’s Hospital, Boston, Massachusetts Programming and Follow-up of Cardiac Resynchronization Devices

Charles D. Swerdlow, MD Clinical Professor, Division of Cardiology, Department of Medicine, University of California, Los Angeles, School of Medicine; Division of Cardiology, Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, California Sensing and Detection

Patrick J. Tchou, MD Attending Physician, Department of Cardiovascular Medicine, Cardiac Electrophysiology and Pacing Section, Cleveland Clinic, Cleveland, Ohio Testing and Programming of Implantable Defibrillator Functions at Implantation

Hung-Fat Tse, MD Professor of Medicine, Cardiology Division, University of Hong Kong and Queen Mary Hospital, Hong Kong Sensors for Implantable Devices: Ideal Characteristics, Sensor Combination, and Automaticity; Sensor Driven Pacing: Device Specifics

Darrel F. Untereker, PhD Senior Director of Research and Technology, Medtronic, Inc., Minneapolis, Minnesota Power Systems for Implantable Pacemakers, Cardioverters, and Defibrillators

Gregory P. Walcott, MD Assistant Professor, Division of Cardiology, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

Contributors

xiii

Paul J. Wang, MD Professor of Medicine, Stanford University School of Medicine; Director, Cardiac Arrhythmia Service and Cardiac Electrophysiology Laboratory, Stanford Hospital and Clinics, Stanford, California Timing Cycles of Implantable Devices

Seth Worley, MD, FACC, FHRS Medical Director, Lancaster Heart and Stroke Foundation, Lancaster General Hospital, Lancaster, Pennsylvania Left Ventricular Lead Implantation

Raymond Yee, MD Professor of Medicine, University of Western Ontario Faculty of Medicine and Dentistry; Director, Arrhythmia Service, London Health Sciences Centre, London, Ontario, Canada Follow-up and Interpretation of Implantable Syncope Monitors

Preface

Cardiac pacing and implantable defibrillators have had a great impact on the treatment of patients with cardiac arrhythmias. The first pacing system was implanted in 1958, but the transformation of the technology, indications, and supporting clinical data in the almost five subsequent decades has been impressive. Particularly remarkable have been the recent technological developments and the growth in pacemaker and defibrillator implantation volume in the United States and around the world. It is hard to believe that six years have gone by since the second edition. With our third edition, the field of cardiac resynchronization therapy has added new complexity and understanding to the science and practice of device therapy. A great deal of effort has been expended to include new information on implantation, troubleshooting, physiology, clinical trials, and engineering of this new therapy. There are many consumers of pacemakers and defibrillator technology besides the patient. There are numerous physicians, including cardiologists, surgeons, internists, and emergency and family physicians, who care for these patients and need to evaluate the impact of this therapy on various medical conditions and treatments. In addition, nurses, engineers from pacemaker companies, and technical and sales representatives from these companies also interact with physicians and patients. Our philosophy in putting together the third edition remains the same as that of our first and second efforts. We have planned this book to emphasize the science of cardiac pacing and defibrillation and to underline the importance of the fact that it is an interdisciplinary field. Physicians are part of a web of health professionals who need increasing amounts of information about implantable devices. We have included a DVD with this edition that includes figures and movies not included in the paper version of our text, as well as much additional material. All of the figures from the text are included with this DVD. The evolution of cardiac pacing has inspired the publication of subspecialty journals, including Heart Rhythm, PACE, StimuCoeur, Journal of Cardiovascular Electrophysiology, Journal of Interventional Cardiac Electrophysiology, Cardiac Electrophysiology Review, and Europace, as well as monographs and a number of new books on implantable devices. Several national and international conferences on pacing and defibrillation take place every year. We have sought to meet the needs of many with this textbook. Clinicians, scientists, nurses, technicians, and engineers will find the information in these pages practical, authoritative, and helpful in better understanding this therapy. We are excited about the opportunity to present this material in a comprehensive scientific manner. We gratefully acknowledge the invaluable assistance and encouragement of Susan Pioli and Natasha Andjelkovic of the Health Sciences Division of Elsevier for all their help in keeping the third edition on track. We owe a great debt of gratitude to our colleagues from the Medical College of Virginia and the McGuire Veterans Affairs Medical Center, the University of Alabama, The University of Hong Kong and Queen Mary Hospital, and the Cleveland Clinic for their patience and support in shouldering the extra workload that allowed us to finish our chapters and editing on time. Most important, we cannot thank enough our many contributors and their colleagues, who labored extensively, often taking time from family and other projects to finish their xv

xvi

Preface

chapters. This large group of individuals deserves all the credit and thanks for making the third edition possible. Our wonderful secretaries, Vera Wilkerson (Virginia Commonwealth University/ Medical College of Virginia), Julie Griffis (Cleveland Clinic), Eleanor Lee (The University of Hong Kong), and Dorothy Welch (University of Alabama) were invaluable for their contributions to help complete this project. This textbook is designed to be a functional tool and reference, helping clinicians, scientists, and engineers make the decisions that improve patients’ lives. It is our desire that this book serve as a valuable resource to all of these people for many years to come. Kenneth A. Ellenbogen, MD G. Neal Kay, MD Chu-Pak Lau, MD Bruce L. Wilkoff, MD

Table of Contents Sect. 1

Basic principles of device therapy

1

Ch. 1

Cardiac electrical stimulation

3

Ch. 2

Principles of defibrillation : from cellular physiology to fields and waveforms

59

Ch. 3

Sensing and detection

75

Ch. 4

Engineering and construction of pacemaker and implantable cardioverterdefibrillator leads

161

Ch. 5

Sensors for implantable devices : ideal characteristics, sensor combination, and automaticity

201

Ch. 6

Power systems for implantable pacemakers, cardioverters, and defibrillators

235

Ch. 7

Pacemaker and implantable cardioverterdefibrillator circuitry

261

Ch. 8

Pacemaker, defibrillator, and lead codes

279

Sect. 2

Clinical concepts

289

Ch. 9

Basic physiology and hemodynamics of cardiac pacing

291

Ch. 10

Clinical trials of pacing modes

337

Ch. 11

Clinical trials of defibrillator therapy

357

Ch. 12

Clinical trials of cardiac resynchronization therapy : pacemakers and defibrillators

385

Ch. 13

Pacing for sinus node disease : indications, techniques, and clinical trials

407

Ch. 14

Pacing for atrioventricular conduction system disease

429

Ch. 15

Evolving indications for pacing : hypertrophic cardiomyopathy, sleep apnea, long QT syndromes, and neurally mediated syncope syndromes

473

Ch. 16

Sensor driven pacing : device specifics

499

Ch. 17

Testing and programming of implantable defibrillator functions at implantation

531

Sect. 3

Implantation techniques

559

Ch. 18

Permanent pacemaker and implantable cardioverter-defibrillator implantation

561

Ch. 19

Left ventricular lead implantation

653

Ch. 20

Approach to generator change

827

Ch. 21

Managing device-related complications and transvenous lead extractions

855

Ch. 22

Imaging in pacing and defibrillation

931

Sect. 4

Device electrocardiography

967

Ch. 23

Timing cycles of implantable devices

Ch. 24

Pacemaker troubleshooting and follow-up

1005

Ch. 25

Troubleshooting of implantable cardioverter-defibrillators

1063

Ch. 26

Programming and follow-up of cardiac resynchronization devices

1087

Ch. 27

Follow-up and interpretation of implantable syncope monitors

1141

Ch. 28

Electromagnetic interference and implantable devices

1150

Ch. 29

Pediatric pacing and defibrillator use

1177

969

Chapter 1

Cardiac Electrical Stimulation G. NEAL KAY • RICHARD B. SHEPARD

E lectrical stimulation is the fundamental principle supporting artificial cardiac pacing. An electrical stimulus interacts with myocardium through one or more electrodes. It produces a flow of electric current within the myocardium and blood pool. If the stimulus has the necessary characteristics, local action potentials occur. For the stimulus application to result in pumping of blood, the stimulus must start a local membrane depolarization process that becomes a self-propagating wavefront of myocyte contraction. Clinical application of cardiac pacing involves the placing of electrical stimulus sites and the timing of stimulus firings so that an efficient mechanical contraction sequence occurs. Each of these aspects of pacing must be managed to obtain hemodynamic and energy efficiency. This chapter is designed to introduce readers to fundamental concepts relevant to artificial electrical cardiac stimulation.

Concepts Related to Electrical Stimulation of the Heart Myocardial stimulation by means of pacemaker or defibrillator electrodes is a complex electrical, biophysical, and biochemical process. Brief definitions of some of the terms used are given below.

Stimulation For the purposes of this chapter, the term stimulation is defined as the initiation of a self-propagating wave of myocardial depolarization and contraction. An electrical stimulus that stimulates myocardium is often said to “capture” the chamber to which it is applied. Anisotropy Anisotropy is the existence of unequal physical properties along different axes. In the heart, myocardial conduction velocity is greater in the direction parallel to the long axis of myocardial fibers than along the transverse axis. Electric Circuit An electric circuit is an electrical charge–conducting pathway that ends at its beginning. Electric circuits involved in myocardial stimulation by pacemakers or implantable cardioverter-defibrillators (ICDs) include the pulse generator and its leads, electrodes, extracellular electrolytes, cell membranes with highly regulated transmission of charged ions in both directions through the membrane, and intracellular ions and charged molecules.* *Björn Nordenstrom, who was Professor of Diagnostic Radiology at Karolinska Institute, found much evidence for many usually unrecognized electric circuits in the body. See Nordenstrom BEJ: Biologically Closed Electric Circuits: Clinical, Experimental and Theoretical Evidence for an Additional Circulatory System. Stockholm, Nordic Medical Publications, 1983.

3

4

Section One: Basic Principles of Device Therapy

Series Circuit In a series circuit, or circuit module, the elements are connected one after another. Therefore, current must flow sequentially through each element in the circuit, one after another. The current flow through all elements is the same. Parallel Circuit In a parallel circuit, two or more elements are joined at each end to a common conductor. Therefore, current may flow from one of the common conductors to the other through any or all of the elements. The degree of current flow in each element is roughly inverse to the factors opposing the flow of electrical charge in that element. Most biologic circuits are made of various combinations of series and parallel modules or subcircuits. For example, because of electrochemical effects, an electrode placed in the heart may act like a capacitor in parallel with a resistor, both in series with the lead joining the pulse generator to the electrode. Electric Current An applied electric field gradient induces a net directional movement of electrical charge. In ordinary terminology, current is said to flow from positive to negative, as from the positive terminal of a battery through an external circuit to the negative terminal. However, electrons in the circuit external to the battery actually move from the negative terminal of the battery to the positive terminal. For clarity in this chapter, current is often stated in terms of electron or ion motion.

cathode. The consistency in the terminology is that, when oxidation occurs, it occurs at an anode, and in the circuitry an anode connects to a cathode that subsequently connects to another anode, and so on. Unipolar pacing is not to be confused with uniphasic pacing. In unipolar pacing, a single electrode (a cathode) is in contact with the heart. The anode is the pulse generator metal case or some other electrode away from the heart. Bipolar pacing is pacing with the cathode in contact with the myocardium and with the anode also in contact with the heart or blood within the heart. Uniphasic (monophasic) pacing is pacing with an electrical waveshape that, as measured at the pulse generator during the pulse, is entirely either positive or negative relative to zero current and voltage. Biphasic pacing has a waveshape that is initially either positive or negative and then reverses polarity during the final portion of the pulse. Capacitor A capacitor is an object that stores energy in an electric field by holding positive charges apart from closely approximated negative charges (unlike a battery, which stores energy in chemical form). The normally nonconducting material or space between the layers of negative and positive charges is the dielectric. A cell membrane acts as a leaky capacitor. Cell membranes have very high capacitance per unit area of cell membrane. Electrode-electrolyte interfaces act in part as capacitors. Helmholtz capacitor and Helmholtz capacitance are the terms used in this chapter for capacitor-like effects that occur at pacemaker and defibrillator electrode-electrolyte interfaces.

Electrode Polarity All defibrillator and pacemaker electric circuits have, during a stimulation pulse, both a positively charged electrode and a negatively charged electrode in contact with tissue. The negatively charged electrode (the catheter tip electrode in a bipolar pacing catheter) is a cathode. It receives electrons from the pulse generator and furnishes electrons to the electrode-tissue interface. Electrodes in a battery were named anode (electron sink) and cathode (electron source) by Michael Faraday after he received suggestions from a philologist friend.1 In a battery by Faraday’s definition, the electrode at which oxidation occurs (e.g., oxidation of Li to yield Li+ and an electron) is an anode. The anode, by continuing oxidation, furnishes electrons to the circuit external to it. Therefore, the negative terminal of the battery is the anode. From there, electrons go through the circuitry and eventually enter the pacemaker electrode that touches myocardium. This electrode, receiving electrons from the pulse generator and furnishing electrons to the tissue, is a cathode. The return electrode in the heart is an anode. It collects electrons from the tissue and returns them through the pulse generator circuitry to the positive electrode of the battery. There, reduction occurs (e.g., I2 + 2e− yields 2I−); this electrode is a

Capacitance Capacitance (C) is the term that specifies, for a given voltage applied across a capacitor, how much electrical charge (Q) can be stored by the capacitor. If V represents a steady voltage applied across the capacitor, then Q = CV. (If E is used instead of V as the symbol for electrical potential, then the relationship may be expressed as Q = CE.) The unit for capacitance is the farad. A farad is the capacitance of a capacitor that on being charged to 1 volt will have stored 1 coulomb of charge. A coulomb is the amount of charge delivered by 1 ampere flowing for 1 second. Coulombs delivered can be expressed as t

Qt = ∫ it dt 0

in which Qt is the total charge delivered between time 0 and time t, it is the instantaneous current at each tiny time segment between time 0 and time t, and the t

integral

∫ i dt is the net area under the instantaneous t

0

current versus time plot.

5

Chapter 1: Cardiac Electrical Stimulation

Inductor When electric current flows through a wire, a magnetic field surrounding the wire is induced. An inductor is an object that stores or releases energy in or from a changing magnetic field. The voltage difference across an inductor is proportional to the rate of change of current flowing through the inductor. Energy is stored during the formation of the magnetic field and is released when the magnetic field decreases or disappears. No energy is lost in a perfect inductor. Inductance Inductance is the term that specifies the relationship between the voltage across an inductor and the rate of change of current traversing the inductor. The magnitude of the inductance can be represented by the symbol L. If Vt represents the instantaneous voltage across the inductor and iL represents the instantaneous current flowing through the inductor, the relationship is given by the equation vt = L

diL dt

or

iL =

1 t vt dt L ∫0

Note that the voltage across the inductor is directly proportional to the rate of change of current flowing through the inductor. Cell membrane currents have some of the currentand voltage-versus-time characteristics of an inductance in parallel with a capacitance.2 This inductance-like effect is related to timing and magnitude of potassium movement into and out of the cell.3

biventricular pacing threshold measurements (see the discussion of biventricular pacing later in this chapter). Both pacing electrodes and cell membranes have reactance qualities, predominately capacitive. Although all physical and biologic electric circuits in principle exhibit all three qualities (resistance, capacitance, and inductance), one or more of the qualities is often so small as to be insignificant. For example, the cardiac action potential spreading throughout the heart generates a changing magnetic field, and a very small amount of energy is transiently stored in the magnetic field. However, the changing magnetic field generated by spread of the action potential is so small that it is not clinically significant except in the research setting.4 Resistance and Impedance Electrical resistance is the type of opposition to current flow in which energy is lost as heat. The instantaneous voltage developed across a perfect resistor is linearly proportional to the instantaneous current flow through the resistor. If a steady voltage across the resistor is represented by V, the current by I, and an unchanging resistance by R, then the relationship is V = IR (Ohm’s law). In reality, cardiac pacing is much more complex and involves many factors, including capacitance, resistance, and stop and start currents. For calculations involving pulsed (e.g., pacemaker stimulus) or alternating current (AC) circuits that contain reactive elements (e.g., electrodes in electrolytes), impedance (Z), a vector sum of resistance R and reactance X, must for accuracy be used in place of resistance. Impedance

Reactance When a changing electric potential is applied across a physical or biologic circuit, some of the opposition to current flow can occur because energy is being stored in an electric field or in a changing magnetic field. This opposition that represents energy storage is termed reactance (symbol X). Reactance and resistance are determining factors in impedance (see later discussion). For reactive elements connected in series, net reactance is the scalar sum of inductive reactance (positive in the mathematical complex plane) and capacitive reactance (negative in the mathematical complex plane). Pure reactance values are dependent on the rates of change of current and voltage, whereas pure resistance values are not. Ohm’s law–type calculations involving the relationships between current, voltage, and power dissipation in various portions of circuits that contain reactive components are more complicated than calculations for circuits involving resistance alone. For example, the peak voltages across individual reactive components in a series circuit may, if added in the ordinary arithmetical way, give a sum that is greater than the peak voltage driving the circuit. Phenomena of this type are caused by differences in the timing of the peaks (phase angles) of the voltages or currents in the various reactive components. These reactive effects can be important in

Impedance is a concept used to calculate the combined effects of resistance, capacitance, and inductance in opposition to current flow. Impedance comes into play when an applied current and/or voltage starts, stops, or changes over the duration of the stimulus. Both pacemaker pulses and defibrillator shocks start from zero current and voltage and go back to zero current and voltage, typically within a few milliseconds or less. At a pacemaker electrode in the heart, capacitance and resistance are major factors in impedance. Within the heart, inductance effects on pacemaker stimuli are too small to be important to clinicians. For circuits with resistance R, capacitance C, and inductance L in series, where qt represents the charge accumulated across the capacitance at any time t, and where the current through the group at time t is it, the voltage Vt across the whole group at time t is described by the equation t

q di vt = it R + t + L C dt

or

vt

∫ idt + L di =iR+ 0

t

C

dt

⎛ ⎜remembering that qt = ∫0 it dt and that the voltage ⎝ q across a capacitor at time t is t ⎞ . C⎠ These equations indicate that, for an instantaneous current it, the instantaneous voltage across the series t

6

Section One: Basic Principles of Device Therapy

circuit is the sum of the effects at that instant in time of the resistance, capacitance, and inductance of the circuit. Note especially that the instantaneous effects are highly related to what has already happened, for t example to the net amount of charge, qt = ∫ it dt , that 0 has accumulated in the capacitor from time 0 to the instantaneous time t. The equations show that the capacitance effect on a series circuit voltage magnitude decreases as the capacitance increases. This has clinical relevance in that, for example, the polarization voltage that interferes with autosensing pulse generators decreases as the electrode capacitance increases. That portion of a pulse generator circuit from the output connector through the lead and electrode into the tissue and then back to the pulse generator represents a set of series and series-parallel connections. For accuracy, this circuit must be viewed as an impedance, not just a resistance. The Helmholtz double-layer capacitance at the electrode-electrolyte junction, together with an ion diffusional effect (the Warburg impedance, described later), makes this necessary. When right ventricular (RV) and left ventricular (LV) pacing leads are connected to a pulse generator through a single ventricular output connector, the RV electrode and the LV electrode combinations are impedances in parallel. One possible effect of biventricular lead connections in parallel is a deceptive (if obtained in the ordinary way) set of threshold measurements.5 This is discussed further in the section on biventricular pacing.

(

)

Electrolyte In this chapter, electrolyte is used as a generic term for the extracellular and intracellular electrically conductive fluids near pacemaker or defibrillator electrodes and elsewhere in the heart. The electrolyte conducts ions but not electrons.

Cellular Aspects of Myocardial Stimulation The Phospholipid Bimembrane Living cells maintain or regenerate a difference in electric potential across the cell membrane. Excitable tissues such as myocardium respond to electrical stimulation of one or more cells with a wave of electrical depolarization, a transient reversal of the voltage gradient across the membrane, that can propagate from cell to cell. Excitable cells respond to a relatively small applied change in electric potential difference across the cell membrane by triggering a series of biochemical/biophysical events (described later). These depolarization events result in myocyte contraction. Depolarization propagates along excitable cell membranes, as in the transmission of an action potential along a squid axon, and from myocyte to myocyte through gap junctions.

Cell Membrane Characteristics Cell membrane characteristics are major determinants of tissue excitability. The membrane of the cardiac myocyte is composed principally of phospholipids, cholesterol, and proteins.6 The membrane phospholipids have a charged polar headgroup and two long hydrocarbon chains arranged as shown in Figure 1-1. The cell membrane comprises two layers of phospholipids with their hydrophobic aliphatic chains oriented toward the central portion of the bilayer membrane and their polar headgroup regions toward the outside boundaries of the membrane. Because the membrane is composed of two layers of phospholipids, the polar regions of the phospholipid molecules interface with the aqueous environments inside and outside the cell. The lipid-soluble hydrocarbon chains are forced away from the aqueous phase to form a nonpolar interior. Determinants of the Resting Transmembrane Potential Relatively large gradients of individual ion concentrations exist across the cardiac cell membrane.7 The gradient of sodium (Na+) ions across the membrane is approximately 145 mmol/L (outside) to 10 mmol/L (inside). In contrast, the potassium (K+) ion concentration outside the cell is approximately 4.5 mmol/L, whereas the inside concentration is 140 mmol/L. In the absence of a cell membrane, both Na+ and K+ would rapidly move in a direction determined by the concentration gradient. The diffusion force tending to move K+ out of and Na+ into the cell is proportional to the concentration gradients of those ions. The potential energy attributable to the diffusion force (PEd) tending to move K+ out of the cell is given by the following equation: ⎛ ⎛ [K ] ⎞ ⎞ PE d = RT ⎜ ln⎜ + i ⎟ ⎟ ⎝ ⎝ [ K ]o ⎠ ⎠ +

(1-1)

where R is the gas constant, T is the absolute temperature, ln is the natural logarithm operator, [K+]i is the concentration of potassium ion inside the cell, and [K+]o is the concentration of potassium ion outside the cell. If the ratio of [K+]i to [K+]o is large, the potential energy across the membrane is large. For each ion species, the difference in concentration between the inside and the outside of the cell results in that ion’s contribution to the difference in electric potential across the cell membrane. In resting cardiac cells, the intracellular cytoplasm has a measured potential of about −90 mV, relative to the extracellular fluid. This electric force tends to move positively charged ions such as K+ and Na+ to the inside of the cell and negatively charged ions such as chloride (Cl−) to the outside of the cell in proportion to the potential gradient. The potential energy attributable to the electric force (PEe) tending to move K+ into the cell is expressed as follows: PE e = zFVm

(1-2)

where z here is the valence (the number of positive or negative electrical charges) of the ion, F is the Faraday constant (96,500 coulombs/equivalent), and Vm is the

7

Chapter 1: Cardiac Electrical Stimulation Figure 1-1. Sarcolemmal structure and phospholipid composition. The lipid bilayer containing two membrane proteins is shown at the top of the figure. The detailed structure of the phospholipids is shown below. All of the aliphatic hydrocarbon groups on the sn-2 position are fatty acids that are covalently bound in the form of esters. The aliphatic hydrocarbon groups at the sn-1 position include O-acyl esters and vinyl ethers. (From Creer MH, Dabmeyer DJ, Corr PB: Amphipathic lipid metabolites and arrhythmias during myocardial ischemia. In Zipes DP, Jalife J [eds]: Cardiac Electrophysiology: From Cell to Bedside. Philadelphia, WB Saunders, 1990, pp 417-432.)

CH3  CH3 N CH3

Choline CH2

CH2

O O P

SN-1 H2C Plasmalogen Diacylphospholipid phospholipid

SN-2 CH CH2

O

O C CH2

HC

CH2

CH2 H2C

or H  H N H

H 2C

COO

O



or

HC

H2C

Serine CH CH2

CH2

CH2

OH

CH2 HC

OH

OH Inositol

OH

OH CH2

O

CH2

H2C H2C

O

O C O

HC



O

H  Ethanolamine H N CH2 CH2 O H

CH2 HC HC CH2

CH2 HC

H2C CH2 H2C CH2

HC CH2 HC HC

H2C CH2 H2C CH3

CH2

CH2

H2C CH3

transmembrane potential difference (measured in millivolts). During equilibrium, the total of the potential energies due to diffusion and electric forces is zero, and no net ionic movement occurs. Therefore, the sum of Equation 1-1 and Equation 1-2 may be set to zero. This yields the Nernst equation, which describes in measurable electrical units the potential that must exist for a single ionic species, here K+, to be in equilibrium across the membrane of a resting cardiac cell: ⎛ [K ] ⎞ Vm ( K + ) = −26.7 ln⎜ + i ⎟ ⎝ [ K ]o ⎠ +

(1-3)

⎛ [K ] ⎞ or, in log base 10 terms, Vm ( K + ) = −61.5 log ⎜ + i ⎟ . ⎝ [ K ]o ⎠ Using known values for extracellular K+, Vm(K+) = 90 mV. When Equation 1-3 is solved using Na+ concentrations, a Vm(Na+) of +50 mV is obtained. Therefore, it is the equilibrium potential for potassium ion (not sodium ion) that is the major factor responsible for the resting transmembrane potential. This suggests +

that the resting membrane is more permeable to K+ than to Na+. To calculate the transmembrane potential when multiple ionic species exist in different concentrations across the membrane, the Goldman constant field equation (modified by Hodgkin and Katz8) is used: + + + + − − − RT ⎞ ⎛ Pk [ K ]i + PNa [ Na ]i + PCl [Cl ]o + . . . ⎞ Vm = ⎛ ln⎜ + + ⎝ F ⎠ ⎝ Pk [ K ] + PNa+ [ Na + ] + PCl − [Cl − ] + . . .⎟⎠ o o i (1-4) + + − where PK , PNa , and PCl are the cell membrane permeabilities for the respective ions. At physiologic concentrations, this equation yields a transmembrane potential of −90 mV (the equilibrium potential for K+). Equation 1-4 describes how resting potentials vary as sodium and potassium ion concentrations are changed. Because there is a passive leak of charged ions through the membrane, the resting potential would not exist at the 90 mV level unless it were actively maintained. This is accomplished by two active transport mechanisms that exchange Na+ ions for K+ and Ca2+ ions.

8

Section One: Basic Principles of Device Therapy

One might ask whether, with all the potassium ions and other positively charged ions in the cell, and with a relatively small amount of negatively charged chloride ions in the cell, how can the interior of the cell be negative with respect to the outside? The answer is that an array of intracellular organic and inorganic anions inside the cell, molecules that do not cross the membrane, carry net negative charges sufficient to make the overall balance of charge negative.

Ion Channels Protein molecules embedded within the cell membrane have numerous functions, including those of being ion channels (Fig. 1-2) and signal transducers. The concept of ion channels was proposed in the 1950s by Hodgkin and Huxley.9 However, it was not until the introduction of the patch clamp technique by Neher and Sakmann in 1976 that the properties of these channels could be directly studied.10,11 There are two basic types of ion channels, distinguished by the factors that control opening and closing of the channel. Ion channels at muscle fiber end plates are chemically gated by specific transmitters. The opening of these channels is triggered by the binding of acetylcholine, and their closing is induced by its unbinding. In neuronal axons, conduction is mediated by faster, voltage-gated channels. These channels respond to differences in electric potential between the inside and outside of the cell, across the membrane. Voltage-gated channels for sodium, potassium, and calcium appear to operate in similar ways, sharing many of the same structural features. In addition, each type of channel can be subdivided into several subtypes with different conductance or gating properties. Voltage-gated Channels Voltage-gated channels open in response to an applied electric potential. The source of this voltage can be an action potential propagated from an adjacent cell or the electric field of an artificial pacemaker electrode. If depolarization of the membrane exceeds a threshold voltage, an action potential is triggered, resulting in a complex cascade of ionic currents flowing across the membrane into and out of the cell. As a result of this flow of charge across the membrane, the potential gradient across the membrane (Fig. 1-3) changes in a characteristic pattern of events that produce the cardiac action potential. Selective membrane-bound proteins (ion channels) determine the passive transmembrane flux of an individual ion species. The transmembrane currents determine or influence cellular polarization at rest, action potential depolarization and repolarization, conduction, excitation-contraction coupling, and myofibril contraction. The channels that regulate transmembrane conductance of Na+ and Ca2+ are voltage gated. The sodium channel is a large protein molecule composed of approximately 1830 amino acids.12 It contains four internally homologous repeating domains. These are believed to be arranged around a central water-filled

Figure 1-2. Ion channels underlie cardiac excitability. a, The key ion channels (and an electrogenic transporter) in cardiac cells. K+ channels (green) mediate K+ efflux from the cell; Na+ channels (purple) and Ca2+ channels (yellow) mediate Na+ and Ca2+ influx, respectively. The Na+/Ca2+ exchanger (red) is electrogenic, because it transports three Na+ ions for each Ca2+ ion across the surface membrane. b, Ionic currents and genes underlying the cardiac action potential. Top, depolarizing currents as functions of time, and their corresponding genes. Center, a ventricular action potential. Bottom, repolarizing currents and their corresponding genes. (From Marban E: Cardiac channelopathies. Nature 415:213-218, 2002, with permission. ©Nature Publishing Group, http://www.nature.com)

pore that is lined with hydrophilic amino acids. It is estimated that there are 5 to 10 Na+ channels per square micrometer of cell membrane. When an alteration changes the membrane potential to about −70 to −60 mV (the threshold potential), four to six positively charged amino acids move across the membrane in response to the change in electric field. This causes a change in the conformations of the channel proteins, resulting in opening of the channel. After a single

9

Maintenance of Resting Membrane Potential 50 Phase 1 Phase 2

50

Phase 0

0 Threshold

Transmembrane potential (millivolts)

Chapter 1: Cardiac Electrical Stimulation

Phase 3

Phase 4 100

100

200 300 400 Time (milliseconds)

500

Figure 1-3. A typical action potential, showing the various phases of depolarization and repolarization. In phase 0 (depolarization), sodium ions (Na +) rapidly enters the cell through fast channels. In phase 1, the initial repolarization is primarily the result of activation of a transient outward potassium ion (K +) current and inactivation of the fast Na+ current. In phase 2 (plateau), the net current is very small, although the individual Na+, Ca2+, and K+ currents are about an order of magnitude larger. Phase 3 (final repolarization) completes the cycle, with the Na +-K + pump bringing the membrane potential to a stable point at which inward and outward currents are again in balance. During phase 4 the cell is polarized and gradually undergoes slow depolarization. (From Stokes K, Bornzin G: The electrode-biointerface [stimulation]. In Barold SS [ed]: Modern Cardiac Pacing. Mt. Kisco, NY, Futura, 1985, pp 33-78.)

Na+ channel changes to the open conformation, about 104 Na+ ions enter the cell. On depolarization of the membrane, the Na+ channels remain open for less than 1 msec. After rapid depolarization of the membrane, the Na+ channel again changes to the closed conformation. In addition to the Na+ channel, specialized proteins are suspended in the cell membrane that have differential selectivity for K+, Ca2+, and Cl− ions, with markedly different time constants for activation and inactivation. Atrioventricular Node Cells In contrast to Purkinje fibers, sinus and atrioventricular (AV) nodal cells are characterized by action potentials with slower rates of depolarization. In these structures, depolarization is primarily mediated by inward Ca2+ conductance through specialized Ca2+ channels. There are two types of Ca2+ channel in the mammalian heart: the L type and the T type. The L-type channels are the major voltage-gated pathway for entry of Ca2+ into the myocyte, and they are heavily modulated by catecholamines.13 The T-type channels contribute to spontaneous depolarization of the cell associated with automaticity (pacemaker currents). The pore of the Ca2+ channel has a functional diameter of about 0.6 nm, larger than that of the Na+ channels (0.3 to 0.5 nm).14 The selectivity for Ca2+ is high, up to 10,000-fold greater than that for Na+ or K+. The key elements are highaffinity binding sites for Ca2+, positioned along a single file pore. “Elution” of a Ca2+ ion occurs when another Ca2+ ion enters and is selectively bound.

The resting membrane potential is maintained by the pumping of Na+ ions out of the cell and K+ ions into the cell. The Na+,K+-ATPase pump moves three Na+ ions out of the cell in exchange for two K+ ions moved into the cell.15-17 The basic unit of the Na+,K+-ATPase protein (pump) consists of one α- and one β-subunit. The βsubunit is large (1016 amino acids) and spans the entire membrane, whereas the β-subunit is a smaller glycoprotein. There appear to be about 1000 pump sites per square micrometer of cardiac cell membrane. The fully activated pump cycles about 50 to 70 times per second (an interval of 15 to 20 msec/cycle). Similarly, the Na+Ca2+ pump moves three Na+ ions out of the cell in exchange for one Ca2+ ion.18,19 Therefore, both transport mechanisms result in the net movement of one positive charge out of the cell, polarizing the membrane and maintaining a negatively charged interior. The function of both exchange mechanisms is dependent on the expenditure of energy in the form of high-energy phosphates and is susceptible to interruptions in aerobic cellular metabolism (e.g., during ischemia). The Cardiac Action Potential When the voltage gradient across the membrane of a myocyte decreases so that the inside of the cell becomes less negatively charged with respect to the outside of the cell, a critical transmembrane voltage difference is reached (the threshold voltage). At threshold, the cell membrane suddenly undergoes a further depolarization that is out of proportion to the intensity of the applied stimulus. This abrupt change in the potential across the membrane is the start of a cascade of inward and outward currents that together are known as an action potential.20 The cardiac action potential is an enormously complex event and consists of five phases:21 phase 0, the upstroke phase of rapid depolarization; phase 1, the overshoot phase of initial rapid repolarization; phase 2, the plateau phase; phase 3, the rapid repolarization phase (see Fig. 1-3); and phase 4, which in cells with spontaneous pacemaker activity is characterized by a slow, spontaneous depolarization of the membrane until the threshold potential is again reached and a new action potential is generated. Phase 0: Rapid Depolarization The upstroke of the action potential is triggered by a decrease in the potential gradient across the membrane to the threshold potential of −70 to −60 mV. On depolarization of the membrane to the less negative threshold voltage, the Na+ channels open, resulting in an influx of positively charged ions (the inward Na+ current) and rapid reversal of membrane polarity. The rate of depolarization in phase 0 ranges from 800 V/sec in Purkinje cells to 200 to 500 V/sec in atrial and ventricular myocytes. In these cells, the inward Na+ current is primarily responsible for phase 0 of the action

10

Section One: Basic Principles of Device Therapy

potential. In sinoatrial and AV nodal cells, where the inward Ca2+ current predominates, the upstroke velocity of phase 0 is much lower (20 to 50 V/sec). Phase 1: Initial Repolarization After voltage-dependent activation of the Na+ current in phase 0, the membrane potential rapidly changes from negative to positive. The increased conductance of Na+ is rapidly followed by voltage-dependent inactivation. Phase 1 is characterized by the transient outward K+ current (IKto). The outward movement of K+ is a major contributor to the various repolarization phases. It is complex and has a number of discrete pathways.22,23 Most K+ currents demonstrate rectification, that is, decreased K+ conductance with depolarization. The K+ currents include the instantaneous inward rectifier K+ current, the outward (delayed) rectifier K+ current, the transient outward currents, and ATP-, Na+-, and acetylcholine-regulated K+ currents. The initial repolarization, however, is mainly the result of activation of a transient outward K+ current and inactivation of the fast inward Na+ current. The transient outward K+ current has two components, one voltage gated and the other activated by a local rise in Ca2+.24 Phase 2: Plateau The net current during the plateau phase is apparently small, although the individual currents (inward Na+ and Ca2+ and outward K+) are each about an order of magnitude larger.25 Among the inward currents are the slowly activating Na+ current, a Ca2+ current, and an Na+-Ca2+ exchange current. Outward currents include a slowly activating K+ current (IKs), a Cl− current, a more rapidly activating K+ current (IKr), an ultra-rapidly activating K+ current (IKur), and the Na+-K+ electrogenic pump. During phase 2 of the action potential (the absolute refractory period), the cardiac cell cannot be excited by an electrical stimulus, regardless of its intensity.

Phase 4: Automaticity and the Conduction System Automaticity is the property of certain cells by which they are able to initiate an action potential spontaneously. It has been known for centuries that the heart can exhibit spontaneous contraction even when completely denervated. Leonardo da Vinci observed that the heart could “move by itself.”26 William Harvey reported that pieces of the heart could “contract and relax” separately.27 Many cells within the specialized conduction system have the potential for automaticity. Not all parts of the heart, however, possess this property. In fact, cells in different areas of the heart have different transmembrane potentials, thresholds, and action potentials. Fast responses are characteristic of ordinary working ventricular muscle cells and HisPurkinje fibers (resting membrane potentials of −70 to −90 mV, rapid conduction velocities). Normal sinus and AV nodal cells have slow responses, with resting potentials of −40 to −70 mV and slow conduction velocities. Those cells, or group of cells, with the fastest rate of spontaneous membrane depolarization during phase 4 are the first to reach threshold potential and initiate a propagated impulse. Therefore, cells with the steepest slope in phase 4 become the heart’s natural pacemaker. Ordinary working myocardial cells usually are not automatic. Normally, depolarization is initiated at the sinoatrial node (Fig. 1-4).28,29 Action potentials from an isolated sinoatrial node cell are shown in Figure 1-5.30 Rather than maintaining a stable resting membrane potential,

SA node

Phase 3: Final Repolarization Deactivation of inward Na+ and Ca2+ currents occurs earlier than for the K+ currents, favoring net repolarization of the membrane. When the membrane is sufficiently repolarized, an inward K+ rectifier current is progressively activated, resulting in a regenerative increase in outward currents and an increasing rate of repolarization. Repolarization is also accomplished by the function of the Na+,K+-ATPase pump. The membrane potential eventually becomes stable, so that inward and outward currents are again in balance and the resting potential reestablished. Between the end of the plateau phase and full repolarization, the cell is partially refractory to electrical stimulation. During this period (the relative refractory period), a greater stimulus intensity is required to generate an action potential than is required after full recovery of the resting membrane potential. For clinical measurement of myocardial refractoriness (the effective refractory period), stimulation of the heart is usually performed at twice the threshold current as determined during late diastole.

His bundle

Left bundle branch

AV node Right bundle branch

Figure 1-4. Schematic representation of the normal conduction system of the heart. The cycle begins at the sinoatrial (SA) node, propagating a wave of depolarization across the atrium. As the stimulus enters the atrioventricular (AV) node, its conduction slows. This allows complete contraction of the atria before the impulse reaches the ventricles. As the impulse enters the His bundle, conduction velocity increases. The impulse is then transmitted through the left and right bundle branches and the Purkinje fibers throughout the right and left ventricular endocardial shells.

Chapter 1: Cardiac Electrical Stimulation

0 mV 50 100 ms Figure 1-5. Spontaneous activity recorded in a single sinoatrial myocyte. The slow diastolic “pacemaker” depolarization extends from the maximum diastolic potential of −71 mV to the threshold for action potential onset, about −54 mV. (From DiFrancesco D: The hyperpolarization-activated current, If, and cardiac pacemaking. In Rosen MR, Jause ML, Wit AL [eds]: Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 117-132.)

the repolarization of the action potential is followed by a slow depolarization from about −71 to −54 mV, the threshold required to initiate another action potential. This slow, spontaneous depolarization drives cardiac automaticity and is related to a specialized current (the funny current, If). In the case of AV nodal cells, the fast upstroke is carried predominantly by an inward Ca2+ current. Repolarization is caused by delayed activation of the K+ current. The balance of inward and outward currents determines the net “pacemaker” current and is finely regulated by both adrenergic and cholinergic neurotransmitters. In the presence of AV block or abnormal sinoatrial nodal function, AV junctional cells in the region of the proximal penetrating bundle usually assume the role of pacemaker at rates slower than that of the sinus node. In the absence of disease in the AV junction, the escape rhythm occurs with a frequency that is about 67% of the sinus rate.31

Artificial Electrical Stimulation of Cardiac Tissue Artificial lipid membranes in their pure form are electrical insulators. The myocyte cell membrane (sarcolemma) is much more complicated. Specialized protein molecules in the membrane allow it to be conductive.32,33 These proteins, either singly or in certain groupings, form channels that open and close for transport of specific ions through the membrane in response to particular stimuli. The channel proteins are the end stages of processes that provides both active and passive transport of ions and molecules through the membrane. When application of a pacemaker or defibrillator pulse produces a local electric field gradient, ion drift in the extracellular fluid at that site, as well as ion flow within the cell and within the membrane, are affected by the field. The field-induced ion drift cannot be uniform within and outside the cell because of the different drift properties of different ion types, different

11

ion and protein concentrations within and without the cell, and the barrier impedance effect of connections between cells. The effect of the stimulus is to change the transmembrane voltage of nearby myocytes sufficiently so that depolarization begins in and spreads from those myocyte membranes. Propagation of the stimulus to nearby myocytes occurs because the local transmembrane depolarization changes the voltage gradient across adjacent membranes sufficiently to trigger depolarization of those membranes. The result is a self-regenerating action potential that progresses in a wavelike, relatively slow, manner beyond the local effect of the pacemaker stimulus. Away from immediate vicinity of the electrode, transmission of depolarization and its velocity are in part dependent on the resistance and capacitance properties of the membrane, on the opening and closing of ion channels, and on ion flows through the sarcolemma. Myocardial Cell Electrical Properties A single Purkinje fiber typically has an internal resistance that is two to three times greater than that of blood. The specific membrane resistance of a Purkinje fiber is on the order of 104 Ω-cm2. The time constant of the surface membrane is on the order of 10 msec, and the membrane capacity is about 1 μF/cm2. Gap junctions are intercellular channels that provide a pathway for electrical communication between myocytes. Their diameter is about 2 nm, and their length about 12 nm. Gap conductance is voltage sensitive.34 Under pathologic conditions such as severe hypoxia or ischemia, gap junctions will not function normally. Cable-like Properties in Depolarization Transmission of electric pulses can be thought of as occurring by two broad mechanisms. One is a regenerative mechanism, like the spread of depolarization over the heart by depolarization from cell to cell. The other is a nonregenerative mechanism. In this, the amplitude of the stimulation pulse decreases with distance from its origin and is not regenerated; this is electrotonic propagation.35 Transmission along a cable is nonregenerative transmission. The manner in which a voltage pulse travels along a Purkinje fiber is somewhat similar to transmission of a voltage pulse along a coaxial cable (Fig. 1-6).36 Propagation amplitude and velocity of a voltage pulse applied to the beginning of the cable are dependent on cable properties. These include the unit-length capacitance between the center wire and the shield, the resistance along the wire, and the nature of the insulation between the center wire and the shield. In a cable, the signal amplitude gradually decreases with distance from the source. Cable theory applied to a squid axon or a very long cell equates the bathing solution to ground (the shield of the cable) and the electrolyte inside the cell to the center wire of the cable. The cell membrane is assumed to be analogous to the insulating material—the dielectric

12

Section One: Basic Principles of Device Therapy

A

B

A rm B

cm ri

Figure 1-6. Equivalent circuit for a single fiber composed of electrically coupled single cells placed in an electrically conductive medium. A constant current is made to flow from A to B through the surface membrane and along the fiber core. Small voltage changes are produced to ensure constant values of the resistors. cm, capacitance across the membrane; ri, resistance in the intercellular space; rm resistance across the membrane. (From Weidmann S: Passive properties of cardiac fibers. In Rosen MR, Jause MJ, Wit AL [eds]: Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, p 30.)

substance—between the center wire and the shield of the cable. The cell membrane acts as an electrically leaky dielectric substance between two conductors, the electrolyte inside and that outside the cell. The membrane of a Purkinje fiber modeled as a cable (see Fig. 1-6) can be viewed as a series of modules placed sequentially.36 One end of each module is connected to the interstitial fluid, which is viewed as having negligible resistance. The other end of each module is connected to the interior of the cell. The electrolyte within the cell and junctions between cells are viewed as having resistance. Between the outside and inside of the cell is the sarcolemma, which is viewed as a series of modules connected in parallel between interstitial fluid and the cell interior. Each module spanning the membrane is made of a resistor and a parallel capacitor. In this schematic representation, the electric pulse being transmitted along the fiber is not regenerated. The electric pulse gradually dies away with distance from its origin at the beginning end of the fiber, as in a coaxial cable lacking booster amplifiers. Cable Theory Applied to Isolated Cardiac Fibers Is Useful But for the Whole Heart Is Inadequate One can immediately see that cable theory applied to the whole heart would be an oversimplification. Applied to isolated fibers, it is conceptually and experimentally useful. Knisley (paraphrasing Weidmann37) pointed out that, in isolated cardiac fibers, intracellular microelectrode techniques demonstrated a distribution of transmembrane voltage that agreed with cable model predictions. Transmembrane voltage differences induced by a negative-going stimulus decreased in an exponential manner with distance from the electrode. However, Knisley38 found that, in perfused rabbit hearts, polarization was not that predicted by cable

theory. He found that, in regions distant from the stimulating electrode in a direction parallel to the fibers, polarization changed sign, and in dog bone–shaped regions perpendicular to the fibers, no polarization sign change occurred. This and similar studies have demonstrated that cable theory is inadequate for predicting how depolarization spreads in a three-dimensional heart. A threedimensional heart not only has three dimensions but also has extracellular and intracellular electrical domains, each of which is anisotropic. Factors Determining Capture by an Electrical Stimulus In order for an electrical stimulus to stimulate (capture) myocardium, it must be applied with sufficient amplitude, for a sufficient duration (measured in milliseconds), at a time when the myocardium is electrically excitable. There are a great many clinical factors that determine whether a stimulus of a given amplitude will result in capture, including proximity of the electrode to the myocardium, pathology of the underlying cardiac tissue, size and shape of the electrode, and effects of drugs and hormones, as well as electrolytes. For routine clinical practice, the most important factor is that the lead must be positioned in close proximity to wellfunctioning myocardium in a secure manner. (It is also important that the lead does not stimulate the diaphragm or the phrenic nerve, that it is at a site that results in good hemodynamic function, and that the location does not allow the electrode to bump other electrodes.) After adequate lead positioning, the strength-duration relation is the next most important consideration. Current Density, Electric Field Gradient, and Propagation of Depolarization There are two approaches to thinking about how an electrical stimulus induces a self-propagating wavefront of depolarization within myocardium. The current density approach considers the magnitude of current flowing through a given mass of myocardium between the stimulating electrodes and finds this to be the critical factor required to induce a regenerative wavefront of depolarization. From this point of view, the stimulation threshold is a function of current density (amperes/cm2) in the excitable myocardium underlying the electrode.39-42 The electric field approach holds that the critical factor affecting myocardial depolarization is the magnitude of the electric field gradient (volts/cm in viable tissue) that is induced in the myocardium beneath the stimulating electrode.43,44 These approaches are fundamentally the same in that there is a mathematical relationship between the electric field gradient and the current density at the stimulating electrode. Because reactance as well as resistance are present, Ohm’s law in this context must be stated in terms of impedance, z: v = iz

(1-5)

Chapter 1: Cardiac Electrical Stimulation

where v is the stimulus voltage, i is the current, and z is the impedance to current flow. Note that each of these is a vector. Also note that z varies with current density at the electrode (because of interface properties), with direction of current flow (anisotropy), and with domain (extracellular or intracellular). The total energy of the pacing stimulus is determined by the applied voltage, the current, and the duration of the stimulus: t

J t = ∫ vt it dt o

where Jt is the energy delivered (expressed in joules) from time zero to time t during the pulse, vt is the voltage, and it is the current at the electrode at instantaneous time (t) during the pulse. This equation indicates that the total energy delivered during the pulse is proportional to the area under a curve. The curve is formed in the vertical dimension by the instantaneous product of voltage and current applied to the electrode, and in the horizontal dimension by the time elapsing during the pulse. When the time t reaches the programmed pulse duration, the pulse ends. The area under the curve then represents the electrical energy transferred during the time span of the pulse. The equation can easily be solved in real time with a small digital computer, provided that continuous phasic measurements of the voltage and current in the wire going to the electrode are available. (Note that this equation has blood flow and pressure energy analogs.45) For pacemakers in almost all clinical situations, it is more practical and very reasonable to calculate the energy delivered by a constant-voltage pulse through a pacemaker electrode by means of an approximation. The assumptions are (1) that the voltage during the pulse really is constant and (2) that the current flowing is linearly related to the voltage. Because neither of these is quite true, the questions become whether the information obtained proves useful and to what degree it can be misleading. If the assumptions are not too far from V V2 being correct, then the equation J = VIt = V t = t, R R in which J is the energy delivered, V is the “constant” voltage, I is the current, R is the resistance, and t is the stimulus duration, can be approximately correct and useful. This equation indicates that the total energy delivered can be estimated by multiplying the voltage reading displayed on the pacing systems analyzer (PSA) by the current reading on the PSA and multiplying that product by the stimulus duration. Alternatively, the voltage reading can be squared, then divided by the resistance reading, and the quotient multiplied by the stimulus duration. The virtue in this calculation is that ordinary equipment can provide a quick, approximate determination of the total energy delivered per stimulus. In clinical practice, however, the usual range of good or acceptable values for pacing threshold current and voltage is known, and there is rarely a need to calculate the delivered energy. The stimulation threshold of isolated cardiac myocytes has been shown to depend on their orienta-

13

tion within an electric field. The threshold is lowest when the myocytes are oriented parallel to the field and highest when the axis of the myocytes is perpendicular to the field.46 It is clear that myocardial stimulation may be induced with anodal or cathodal stimulation, or both, although with somewhat different characteristics. Various investigators have used several different parameters to express stimulation threshold, including current (mA), potential (V), energy (J), charge (Q), pulse width (t, msec), and voltage multiplied by stimulus duration (V-sec).47-53 For the purposes of this chapter, the myocardial stimulation threshold for pacing is defined as the minimum stimulus amplitude at any given pulse width required to consistently achieve myocardial depolarization outside the heart’s refractory period.54 Stimulation thresholds measured with a constant-voltage (CV) generator are stated in volts, and those of a constant-current generator (CI) are stated in milliamperes. Note that a true constant-voltage generator may yield a slightly different stimulation threshold than the pseudoconstant-voltage generators in ordinary use. Why must an electrical pacing stimulus rely on propagation in myocardium rather than directly exciting the entire heart? With the magnitude of a stimulus generated by a pacemaker pulse generator, the electric field gradient and ion current density near the electrode are great enough to trigger an action potential only very near the electrode. Unless propagation of the local depolarization occurs, no cardiac contraction will result. In contrast to cardiac pacing with direct production of only local depolarization and dependence on selfpropagation of depolarization, defibrillation involves direct depolarization or hyperpolarization of a large portion of the heart. This is necessary to provide at least the minimum local voltage gradient required to produce depolarization over major portions of the myocardium. To do so requires a stimulus intensity that is very much greater than that required for pacing55 (see Chapter 2 for a thorough discussion of the principles underlying defibrillation). Electric Potential Gradients for Stimulation and Defibrillation When the electric field exceeds approximately 1 V/cm in the extracellular space during diastole, myocardial stimulation (capture) results. When the electric field strength is increased to 6 V/cm, ventricular fibrillation may occur if the stimulus is applied during the vulnerable period (approximately the peak of the T wave). This same field strength (6 V/cm) is also required to interrupt ventricular fibrillation (defibrillation). However, defibrillation requires the minimum field intensity to be approximately 6 V/cm at almost all points in the myocardium. In order to achieve this minimum electric field gradient at the same instant over the entire heart, a very large shock current must be applied. This current is thousands of times greater than the current required for pacing.55 When a pacemaker pulse is applied, electrically excitable myocytes respond with a wave of depolarization

14

Section One: Basic Principles of Device Therapy

followed by repolarization in the myocardium. Depolarization of a small local group of myocytes begins the self-propagating process. The initial depolarization adjacent to the electrode produces a potential gradient great enough at neighboring myocytes to result in their depolarization. The process is a self-regenerating mechanism that requires time to spread, and, once established, it is largely independent of the distance from the stimulating electrode. Defibrillation is largely accomplished by providing a very large potential gradient between the defibrillation electrodes. This produces, at one instant within local myocardium everywhere in the heart, at least a 6 V/cm gradient. In contrast, pacing is accomplished by providing a gradient of approximately 1 V/cm or less at a local site and relies on self-propagation to spread the depolarization process throughout the myocardium. The Virtual Electrode Effect Virtual electrodes are important because they can serve as sites that initiate or prevent depolarization of myocytes.56 Newton and Knisley defined virtual electrodes as “experimentally observed regions of large delta Vm that arise distant from the stimulating electrode.”57 “Delta Vm” (ΔVm) here represents the change in voltage difference across the myocyte membrane during application of the stimulation current. A virtual electrode can be described as a collection of charge predominately of one sign at a site away from a regular electrode. If, in an electrically neutral solution, ions of one charge sign are moved away by an electric field, a relative excess of ions of the other charge sign will be left or will move in the opposite direction. For example, if the regular electrode is paced with a negative-going stimulus, a site elsewhere in the tissue (i.e., not at this physical electrode) may become transiently positive and alter the transmembrane potentials of cells at that site.58-61 In cardiac tissue, a virtual electrode can occur as an effect of anisotropy after a defibrillation shock and can initiate refibrillation. A virtual electrode can also occur as a result of ion redistribution patterns in a nonanisotropic medium. In anisotropic tissue, charge redistribution produced by a physical electrode can have a dog bone shape.62 The virtual electrodes occur at sites where ions flow into or out of cells by crossing the cell membrane. An applied unipolar stimulation current flows into some cells while simultaneously flowing out of others, thus producing negative and positive virtual electrodes at different locations. In an anisotropic medium, a cathodal pulse produces a virtual electrode with a dog bone shape oriented perpendicular to cardiac fibers and containing a large ΔVm, and a pair of virtual electrodes containing negative ΔVm at locations along the fibers. Theoretical models have shown that this effect can be attributed to unequal tissue anisotropy in the intracellular and extracellular spaces; that is, intracellular current at a given location favors the longitudinal direction 10 to 1, whereas extracellular current favors the longitudinal direction only 3 to 1 (Dr. S. B. Knisley, personal communication, July 2005). Knisley and Pollard63

studied the effects of electrode-myocardial separation on cardiac stimulation of rabbit hearts in conductive solution. The electrode-myocardial separation altered the spatial distribution of ΔVm and increased the pacing threshold. In regard to electrode-myocardial tissue separation, Shepard and colleagues,64 during 1980s transthoracic defibrillator implantation procedures, sewed the electrodes used for rate sensing onto the outer surface of the pericardium. Sensing voltages were normal. Pacing thresholds of these electrodes measured were high (3.7 ± 1.9 mA and 4.5 ± 2.19 V at 0.5 msec stimulus duration), as would be expected both from current density considerations and from Knisley’s findings. The initial impedance was 1209 ± 383 Ω, and the chronic impedance was 1550 ± 358 Ω at a median follow-up time of 964 days. The thresholds had by then decreased to 3.8 ± 2.07 V and 2.7 ± 1.8 mA. The transpericardial distance from underlying myocardium did result in high initial pacing thresholds (and in the special distribution of ΔVm near the electrodes). However the longterm tissue reaction of pericardium and underlying myocardium to the presence of these electrodes was not detrimental in terms of threshold evolution. Virtual electrodes normally exist near an electrode when a pacing pulse is applied. Nikolski and Sambelashvili,65 studying Langendorff-perfused rabbit hearts, found that stimuli of magnitude five times threshold produced “make” or stimulus-onset excitation from virtual cathodes, whereas near-threshold stimuli produced “break” or stimulus-termination excitation from virtual anodes. In studying how cardiac tissue damage at an electrode results in a pacing threshold increase, Sambelashvili and associates66 found that the virtual electrode effect was destroyed by very strong (40 mA, 4 msec, biphasic, rate 240/minute) pacing pulses applied for 5 minutes. Fluorescent optical mapping showed that decrease or loss of the virtual electrode polarization was associated with pacing threshold increase. Propidium iodide staining showed tissue damage within an area of about 1 mm diameter surrounding the electrode. Another possible way of looking at the effect of very strong stimuli as described in the previous paragraph is to note that local tissue damage at a pacing electrode increases the distance from the electrode to normal myocytes. Because electric field strength and current density decrease with distance from the electrode, an increase in pacing stimulus amplitude applied to the electrode is necessary to restore the stimulus current density to the pacing threshold level of myocytes at the outer edge of the damaged tissue.

Strength-Duration Relationships Chronaxie, Rheobase, Energy, and Pulse Duration Thresholds The intensity of an electrical stimulus (measured in volts or milliamperes) that is required to capture the

Chapter 1: Cardiac Electrical Stimulation

atrial or ventricular myocardium depends on the duration of the stimulus.67,68 The historical background and the relations between the various electrical factors have been reviewed, further studied, and very clearly stated by Blair,69 and by Geddes for electrode-electrolyte interface function and models.70 The interaction of stimulus amplitude and stimulus duration (pulse width) defines the strength-duration curve (Fig. 1-7). The voltage or current amplitude required for endocardial stimulation has an exponential relation to the pulse duration, with a relatively flat curve at durations of longer than 1 msec and a rapidly rising curve at durations of less than 0.25 msec. Because of this fundamental property, a stimulus of short pulse duration must be of much greater intensity to capture the myocardium than a longer-duration pulse. Conversely, increasing the pulse width to longer than 1 msec has little influence on the intensity of the stimulus required for capture. Therefore, if one defines the stimulation threshold in terms of pulse amplitude without also specifying the pulse duration, important information is neglected. Similarly, specifying the capture threshold in terms of pulse duration can be misleading if the stimulus amplitude setting is omitted or unknown. Hoorweg,71 in 1892, used a voltage source, a galvanometer, and low-leakage capacitors to conduct quantitative stimulation studies. He found that the voltage at which a capacitor must be charged to cause depolarization of nerves and muscles is an inverse function of the capacitance of the capacitor, as follows: VC = aR + b C

15

In this experimentally determined equation, Vc is the threshold voltage to which a capacitor of capacitance C must be charged to produce stimulation on discharge. R is the resistance of the circuit through which the capacitor is discharged, and a and b are coefficients that vary with the specimen (tissue) tested. Here, the experimentally determined constant a has the dimension amperes and the constant b has the dimension ampere-seconds. (The capacitance can be derived from the equation C=

Q V

where C is the capacitance, Q is the charge on either conductor, and V is the magnitude of the potential difference across the capacitor; the capacitance has the dimension ampere-seconds [or coulombs] per volt; 1 coulomb per volt = 1 farad). Hoorweg determined that there was only one specific capacitance value for which the threshold charge was a minimum. He also determined that the threshold charge was a linear function of the stimulus duration, “intersecting the y-axis above zero.” However, Hoorweg did not have the capability to measure thresholds at very short pulse durations. In reality, the threshold charge increases toward infinity as the pulse duration approaches zero. In 1901, Weiss72 reported that the threshold charge required for stimulation increases linearly with stimulus duration. He called this relationship the “formule fondamentale.” If I represents the magnitude of the current, t is the stimulus duration, and a and b are constants determined by analyzing the data, Weiss found that

(1-6)

t

∫ Idt = at + b

(1-7)

0

t

Threshold (volts, microjoules, microcoulombs)

or, because

∫ idt = Q, 0

x

5 Energy 4 x x

3

x x

x

x

x x

x

Charge

x

2 Potential

Rheobase

1 Chronaxie 0

.2

.4

.6 .8 1.0 Pulse width (milliseconds)

1.5

Figure 1-7. Relationships between chronic ventricular canine constant-voltage strength-duration curves expressed in terms of potential (V), charge (mC), and energy (mJ) for a tined unipolar lead with an 8-mm2 polished ring-tip electrode. Thresholds are measured at the point of gain of capture. Rheobase is the current or voltage threshold to the right that is independent of pulse width. Chronaxie is the pulse width at twice the rheobase. (From Stokes K, Bornzin G: The electrode-biointerface [stimulation]. In Barold SS [ed]: Modern Cardiac Pacing. Mt. Kisco, NY, Futura, 1985, pp 33-78.)

the charge in the capacitor at

threshold stimulus duration t, Q = at + b. For a constant-current stimulus, this can be stated in words as threshold charge requirement = a amperes times t seconds + b, where b is a constant with dimensional units of ampere-seconds. The values of the constants a and b vary with the tissue tested. The left-hand side of Equation 1-7 indicates that the charge delivered into the electrode is, for a constant current, the magnitude of the current multiplied by the duration of the current. The right-hand side says that the charge required to stimulate at threshold is a minimum of b ampere-seconds plus the product of the current level (a) and the stimulus duration (t). Weiss72 noted that, for various stimulus durations tested, the quantity of charge required to initiate depolarization remained constant. The statement that the threshold charge requirement does not change with stimulus duration is true only within a limited range. The limitation leads to the concept of rheobase, as stated in 1909 by Lapique.73 He pointed out that when pulse duration is increased beyond a limited range, the current requirement does not decrease further. The charge delivered continues to increase as pulse duration is increased. Lapique called this minimum current the rheobase. Rheobase can be defined as the lowest stimulus current that continues to

16

Section One: Basic Principles of Device Therapy

capture the heart when the stimulus duration is made very long. In this situation, further increases in stimulus duration no longer reduce the magnitude of current required to stimulate the heart. Note from Equation 1-7 that the ratio a/b has the dimensions amperes divided by (ampere-seconds); this expression reduces in dimensions to the reciprocal of the stimulus duration. Lapique called the time seen in the a/b ratio the chronaxie, specified in seconds. Chronaxie time experimentally turns out to be approximately the threshold pulse duration at twice rheobase amplitude, and it has become defined as such. This is illustrated in Figure 1-7. Note again that rheobase is specified in terms of current, and chronaxie is specified in terms of time. Lapique redefined the Weiss equation (Equation 1-7) as follows. The stimulation threshold charge Q in the capacitor at stimulation time t is equal to at + b, in which a and b are constants determined from threshold measurements. Then, because Q = It, It = at + b. Dividing both sides of the latter equation by t yields It = a +

b . t

(1-8)

The constant a has the same dimension as current, and the constant b has the same dimension as current × time, or charge. If the constant magnitude a is named the rheobase, the equation can be restated as I t = I rheobase +

⎛ I rheobase × t chronaxie ⎞ ⎝ t stimulusduration ⎠

in which the product Irheobase × tchronaxie has the numerical value of the constant b. Therefore, It = a +

t b = I rheobase ⎛1 + chronaxie ⎞ ⎝ t ⎠ t

where It is the current during the pulse duration t. This new equation specifies the threshold constant-current stimulus amplitude in terms of chronaxie time, stimulus duration, and rheobase current. If, for reasons discussed later, one moves the Irheobase term inside the parentheses, the equation becomes b ⎛ I ×t = I rheobase + rheobase chronaxie ⎞ ⎠ t ⎝ t Next, if both sides of this threshold current equation are multiplied by the stimulus duration, the equation becomes one of the charge delivered at that pulse duration: It = a +

Q = It = (t stimulusduration × I rheobase ) + t stimulusduration × ( I rheobase × t chronaxie t stimulusduration ) In the term on the right, the stimulus duration cancels out, and this equation becomes Q = (t stimulusduration × I rheobase ) + ( I rheobase × t chronaxie ) At very short stimulus durations, the product in the left-hand set of parentheses (stimulus duration times rheobase current) approaches zero. Therefore, when the stimulus duration is very short, the charge quantity represented by Equation 1-9 approaches the limit of chronaxie time multiplied by rheobase current. Q = It = ( I rheobase × t chronaxie )

(1-9)

The lower this product, the lower the charge necessary to stimulate the heart. Lapique74 also determined that stimulation at the chronaxie pulse width approaches the minimum threshold energy. This is a very useful clinical concept. Therefore, the two most important reference points on a current or voltage strength-duration curve are rheobase and chronaxie. Obtaining a true rheobase typically requires pulse widths of 10 msec or greater. For clinical purposes, rheobase can be approximately measured at pulse widths of 1.5 to 2 msec. The value obtained may be slightly greater than the true rheobase. Therefore, the chronaxie stimulus duration obtained by setting the stimulus current at twice the approximate rheobase current and then finding the minimum stimulus duration that will result in capture is slightly low. In a time-saving, useful, and reasonable clinical sense, one may empirically set the stimulus duration at a value determined from experience, such as 0.4 to 0.5 msec and then determine the current and voltage thresholds. Safety factor allowances of current and voltage are then added or subtracted based on the patient’s current and projected clinical status (see later discussion). The goal is to find the combination of pacing threshold stimulus current, voltage, and pulse width that results in minimal charge drain from the pulse generator battery at normal pulse rates. Figure 1-7 shows that, for capture to be accomplished, the least charge was required at the shortest pulse duration (0.1 msec in the figure), but the least energy was required at about 0.3 msec stimulus duration. A very short pulse duration puts thresholds close to the steeply ascending limb of the voltage or current curve, where slight fluctuations can risk loss of capture. For this reason, Irnich75,76 recommended that chronaxie, or the pulse duration slightly to the left of chronaxie, is the most efficient pulse duration for both pacing and defibrillation. In most cases, the pulse duration at chronaxie or slightly greater appears to represent the best overall compromise between adequate safety and generator longevity. One also considers the expected stability or instability of the patient’s status, patient compliance, and followup arrangements both short and long term. Pacing thresholds can also be specified in terms of only energy (microjoules) or only pulse duration (milliseconds). However, reliance on specifications in either of these units alone can be misleading. Doing so disregards the fact that, regardless of the pulse duration or energy used, the heart cannot be stimulated unless the stimulus amplitude exceeds the rheobase current.77 For example, if rheobase is achieved with a pulse amplitude of 0.5 V and 10 msec, the pulse at this point on the strength-duration curve will have 5 μJ of energy (current × voltage × time) with a 500-Ω lead (1 mA × 0.5 V × 10 msec = 5 μJ). A 0.4-V, 20-msec pulse on a 500-Ω lead has 6.4 μJ energy (28% greater than the “threshold” energy at 10 msec) but will not result in cardiac capture. Increasing the pulse duration to 100-msec at 0.4 V results in 32 μJ energy (540% greater than “threshold” energy at 10 msec) but will still not capture the heart, for the same reason. No matter how far the pulse duration is extended (with increased

Chapter 1: Cardiac Electrical Stimulation

Practical Application of the Strength-Duration Relationship to Threshold Measurement Determination of the strength-duration relationship requires that the clinician measure the stimulation threshold at specific amplitudes and pulse durations. However, the result obtained varies somewhat with the manner in which the threshold is measured. Usually, the threshold measured at a specific stimulus duration (e.g., 0.5 msec) will be slightly higher when the stimulus amplitude is gradually being increased than when it is gradually being decreased. The threshold can also in a sense be measured by holding the output voltage constant and changing the stimulus duration. The stimulation threshold then is defined as the lowest amplitude pulse duration that results in consistent capture of the myocardium. Threshold measured only in this way can be clinically useful, as noted earlier. It also can be deceptive, because the strength duration curve is not linear. For example, if the pulse duration threshold is 0.5 msec at an amplitude of 2 V, reprogramming the

10 9 8 7 6 Volts

energy), the myocardium will not be captured unless the stimulus amplitude is at least as great as the rheobase value. Therefore, calculation of the energy threshold at very long pulse durations does not provide clinically useful information.78 Coates and Thwaites79 studied the strength-duration curve in 229 patients with 325 leads. The mean atrial chronaxie (n = 101) was 0.24 ± 0.07 msec, and the mean ventricular chronaxie (n = 224) was 0.25 ± 0.07 msec. Mean atrial and ventricular rheobase values were 0.51 ± 0.2 V and 0.35 ± 0.14 V, respectively. Because the pulse generators were set at factory-nominal pulse durations of 0.45 to 0.50 msec, the authors concluded that pacing was suboptimal from the efficiency point of view. They pointed out that battery drain would be reduced by programming pulse duration to the chronaxie value and then programming the voltage to double the chronaxie value. In a study of excitability of rat atria during postnatal development up to 120 days, de Godoy and associates80 found that atrial rheobase decreased with animal age and was altered by electric field orientation. Atrial chronaxie increased only with age. The clinician might keep in mind that the chronaxie is influenced by not only electrode material, size, and stimulation mode but also by clinically varying biochemical factors. Certainly, marked variations in pacing threshold, either up or down, do occur months and years after implantation in some children and adults.81 As stated earlier, on the left side of the chronaxie point of the strength-duration curve the current and voltage rise rapidly as stimulus duration decreases. At pulse durations greater than chronaxie, the slope of the strength-duration curve gradually flattens. Small changes in stimulus amplitude are then less likely to result in loss of capture, especially if the threshold has increased for any one of several pathophysiologic reasons. This increased safety comes at the cost of some decrease in battery life.

17

D 2XV E

5 4

C

3 2

2XV 2 X PW

A

B

1 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

Pulse duration (msec) Figure 1-8. Programming of pulse amplitude and pulse duration based on analysis of the strength-duration curve in a patient evaluated at the time of pulse generator replacement (6 years after lead implantation). The rheobase voltage was 1.4 V, and the chronaxie pulse width (PW) was 0.30 msec. Note that the stimulation threshold, determined by decreasing the stimulus amplitude at a constant pulse duration of 0.5 msec, was about 2 V (point A). Doubling of the PW on the relatively flat portion of the strength-duration curve (point B) provides little safety margin for ventricular capture. In contrast, consider a threshold value on the steeply ascending portion of the strength-duration curve (point C). Doubling of the pulse amplitude doubles the safety margin on this portion of the curve, but it lies very close to the curve (point D). An appropriate setting for the chronic pacing pulse might be achieved by doubling the pulse amplitude from point A to point E. Also note that a similar programmed setting would have been obtained had the pulse duration been tripled from point C. Thus, the shape of the strength-duration curve has an important influence on the choice of the amplitude and duration of the pacing pulse.

pacemaker stimulus to a pulse duration of 1 or even 1.5 msec would provide a very small margin of safety (Fig. 1-8). Instead, doubling the stimulus amplitude to 4 V at a pulse duration of 0.5 msec would provide an adequate margin of safety. In contrast, consider that, in this same patient, the pulse duration threshold is 0.15 msec at a pulse amplitude of 3.5 V. Increasing the pulse width to 0.45 msec would also provide an adequate safety margin. The reason that a threefold increase in pulse width is not adequate in the first example but is acceptable in the second relates to the location of the threshold stimulus on the strengthduration curve. Thinking about programming the pulse generator to twice the voltage threshold found at the chronaxie pulse duration is only a starting point. In deciding how much safety margin is necessary for any particular patient, several physiologic factors come into play. These are discussed in the following paragraphs. Capture Hysteresis (Wedensky Effect) The threshold stimulus amplitude that is measured by decreasing the voltage or current until loss of capture occurs is sometimes less than that determined by

18

Section One: Basic Principles of Device Therapy 1.0 0.9

Rheobase (volts)

0.8 0.7 0.6 0.5

Increment

0.4

Decrement

0.3 0.2 0.1 0.0 2250 2000 1750 1500 1250 1000 750 Cycle length (msec)

500

250

Figure 1-9. Pacing thresholds determined by gradually incrementing and decrementing the pulse amplitude until gain and loss of capture, respectively, are demonstrated in a patient with complete atrioventricular block. The pacing threshold was determined at cycle lengths of 2000, 1500, 1000, 750, and 500 msec and with a constant pulse duration of 2.0 msec. To prevent variation in cycle length during incrementing and decrementing pulse amplitudes, a backup pulse was delivered at 25 msec. Note that the threshold values determined in this manner, with increments and decrements of the stimulus amplitude, are similar. Therefore, the Wedensky effect may be marginal when the pacing cycle length is maintained at a constant value.

increasing the stimulus intensity from below threshold until gain of capture occurs. This hysteresis-like phenomenon is the Wedensky effect.82 It is the effect of subthreshold stimulation on the subsequent suprathreshold stimulation when the stimulus amplitude is being increased. Figure 1-9 shows the Wedensky effect. Langberg and colleagues83 observed that there was no demonstrable capture hysteresis at pacing cycle lengths greater than 400 msec. They concluded that the Wedensky effect was related to asynchronous pacing in the relative refractory period when incrementing the stimulus intensity, as compared with synchronous late diastolic stimulation when decrementing the stimulus amplitude until loss of capture. Swerdlow and associates84 showed in a study of 40 patients that cardiovascular collapse occurs at AC current levels less than the ventricular fibrillation current threshold. They noted that the continuous capture threshold for AC current is less than the capture threshold for a single ordinary pacing stimulus. They suggested that continuous capture at low levels of AC current requires a cumulative effect of subthreshold stimuli. This is a variety of the Wedensky effect. They stated that the safety standard for 60-Hz leakage current lasting longer than 5 seconds should be 20 μA or less to avoid intermittent capture. Effect of Pacing Rate on Stimulation Threshold Hook and coworkers85 reported a significant increase in ventricular pacing threshold in 10 of 16 patients at 400 msec and in 15 of 16 at 300 msec (relative to a pacing cycle length of 600 msec). The phenomenon

was not observed at every trial (e.g., 12 of 72 trials at 400 msec). The patients were all candidates for ICDs, and 9 of 12 were receiving antiarrhythmic drugs. The leads were bipolar (a pair of epicardial corkscrews in 11 patients, an endocardial screw-in lead in 5 patients). The atrial stimulation threshold has also been shown to vary as a function of pacing rate.86,87 Katsumoto and associates88 reported that 29 of 36 patients exhibited constant-current atrial pacing threshold energy and current variations as a function of pacing rate in the range of 60 to 120 beats per minute (bpm). The pacing was done with activated vitreous carbon electrodes. Also, Kay and colleagues89 found significant human atrial threshold changes as a function of pacing rate (between 125 and 300 bpm) using constant-voltage stimulation. They found a significant increase in rheobase voltage, chronaxie, and minimum threshold energy at pacing rates greater than 225 bpm using platinized (low-polarizing) unipolar electrodes. They also determined strength-interval curves and found no correlation between atrial effective refractory period and rheobase voltage, chronaxie, or rate-dependent changes in either of these values. They concluded that the phenomenon is probably related to the “opposing effects of decreasing cycle length on the action potential duration and the slope of the strength-interval curve. Thus, if the pacing interval shortens to a greater extent than the refractory period and pacing stimuli are delivered during the ascending limb of the strength-interval curve (the relative refractory period), the diastolic threshold will increase.”89 The increase in stimulation threshold with increasing stimulation rate probably has minimal implications for bradycardia pacing. It is important for antitachycardia pacing, however, because threshold must be measured at rates required to interrupt the arrhythmia. In addition, the safety factors used with antitachycardia pacemakers must be based on thresholds measured at the clinically appropriate rates, rather than during pacing at resting rates.

Strength-Interval Relationships Voltage and current stimulation thresholds vary as a function of the coupling interval of the stimulus to prior beats and to the stimulation frequency used for the basic drive train. A typical ventricular constant-current strength-interval curve is shown in Figure 1-10. At relatively long extrastimulus coupling intervals (>270 msec), the intensity of the extrastimulus required for ventricular capture is relatively constant, approaching the rheobase value. At shorter extrastimulus coupling intervals (.5

6917

Epicardial helix

Myocardial helix

12

1.4 ± .53

17

>.5

Model No.

Stimulation Threshold (V)

*P < .05 considered statistically significant. From Stokes K, Bornzin G: The electrode-biointerface (stimulation). In Barold SS (ed): Modern Cardiac Pacing. Mt. Kisco, NY, Futura, 1985, p 40.

a branch of the cardiac venous circulation, usually the anterior interventricular or posterolateral vein. The electrical properties of chronic stimulation of both the right and left ventricles are complex. The complexity of dual ventricular stimulation has been reported in canine studies.242 In canine studies, bipolar leads were placed transvenously into a distal epicardial ventricular branch of the great cardiac vein. The distal electrode was a 5.8-mm2 porous, platinized hemisphere with a steroid coating. A transvenous lead was also placed at the RV apex. Two varieties of RV apex electrodes were used, one with a 4-mm2 platinized Target Tip and the other with a 1.2-mm2 steroid-eluting tip. Several modes of stimulation were compared (Tables 1-2, 1-3, and 1-4). One configuration (ventricular split bipole) used two unipolar leads and an adapter such that the electrode in one chamber was the cathode and the electrode in the other chamber served as the anode. The second configuration (dual-cathode) joined the LV and RV electrodes in parallel as a common cathode; the anode was the pulse generator can. The thresholds as a function of time after implantation are shown in Figure 1-36. The single-chamber chronic coronary venous LV thresholds were two to three times higher than the RV single-chamber values. The threshold for simultaneous capture of both the right and left ventricles was similar to the unipolar LV threshold. Similar findings were observed regardless of whether the anode was the pulse generator can or a ring electrode in the right ventricle. In contrast, with the ventricular split bipole configuration, the threshold for combined capture of both ventricles was much higher, leading to exit block (>5 V) in many cases. The pacing impedance with the dual-cathode configuration was 43% lower than the average of the two single-chamber values. When a small electrode with high impedance (1.2 mm2) was paired with a larger

Canine Thresholds (V) at 0.5 msec 12 Weeks after Implantation (n = 13) TABLE 1-2.

Chambers Paced 1

Unipolar LV

Unipolar RV

Unipolar DCO

1.8 ± 1.2

0.9 ± 0.2

0.9 ± 0.2 2.0 ± 1.4

2 1

2.6 ± 1.9

0.9 ± 0.5

1.1 ± 0.6 3.3 ± 2.0

2 Bipolar RV-, LV+

Bipolar LV-, RV+

Bipolar DCO-, RV Ring+

1

1.4 ± 0.5

2.0 ± 0.9

1.0 ± 0.4

2

7.2 ± 3.5

3.5 ± 1.0

1.7 ± 0.9

1

1.4 ± 1.0

3.0 ± 1.6

NA

2

8.5 ± 3.0

4.6 ± 2.3

NA

DCO, dual cathodal output; LV, left ventricular coronary vein; NA, not applicable; RV, right ventricular apex.

electrode (4 mm2), the dual-cathode impedance was 51% lower than the average of the two unipolar impedances values. Pacing impedance in the ventricular split bipole configuration was markedly higher than that for either electrode alone. The first generation of cardiac resynchronization therapy (CRT) devices used a bipolar RV pacing or ICD lead and a unipolar coronary venous lead with the ventricular output pulse delivered to both the coronary venous tip and the RV tip electrodes in a split-cathodal configuration. The anode was programmable to be either the pulse generator casing (unipolar split-cathodal configuration) or the ring electrode on the RV lead (bipolar split-cathodal configuration). As a result of splitting the

47

Chapter 1: Cardiac Electrical Stimulation

Canine Pacing Impedance (Ω) at 2.5V and 0.5 msec, 12 Weeks after Implantation TABLE 1-3.

Unipolar canine RVA and LVCV thresholds vs. time

Unipolar LV

Unipolar RV

Unipolar DCO

790 ± 118

734 ± 190

433 ± 68

829 ± 642

1193 ± 297

488 ± 157

Bipolar LV-, RV+

Bipolar DCO-, RV Ring+

1193 ± 186

1014 ± 157

410 ± 64

1863 ± 745

1216 ± 806

NA

Bipolar RV-, LV+

Thresholds (V/0.5 ms)

4 4.01.2 mm2 RVA 5.8 mm2 LVCV

3.5 3 2.5

N11

2 1.5 1

N11

0.5 0 0

Unipolar LV

Unipolar RV

Unipolar DCO

13 ± 4

26 ± 4

14 ± 2

13 ± 4

26 ± 4

14 ± 7

Bipolar RV-, LV+

Bipolar LV-, RV+

Bipolar DCO-, RV Ring+

29 ± 4

35 ± 7

NA

29 ± 4

29 ± 4

NA

DCO, dual cathodal output; LV, left ventricular coronary vein; NA, not applicable; RV, right ventricular apex.

pacing current between two cathodal electrodes, the current was delivered in parallel to both the right and left ventricles. The magnitude of the current flowing to either ventricular electrode was determined by the impedance of the two leads. The apparent pacing threshold in the left ventricle was dramatically affected by this split-cathodal configuration. For example, Mayhew and coworkers5 found that, when the unipolar LV (coronary venous) threshold was measured using the coronary venous tip electrode as the cathode and the pulse generator casing as the anode, the mean threshold was 0.7 ± 0.5 V at 0.5-msec pulse duration. Splitting the cathode between the LV tip and the RV tip electrodes increased the apparent LV threshold to 1.0 ± 0.8 V. When the anode was changed from the pulse generator casing to the RV ring electrode, the apparent LV threshold further increased to 1.3 ± 0.9 V. Therefore, a split-cathodal configuration markedly increased the apparent pacing threshold, compared with pacing the coronary venous lead alone. A further observation with the split-cathodal pacing configuration is that the impedance measured with a PSA does not behave as predicted by the simple application of Ohm’s law. For example, when the bipolar pacing configuration was used with the RV ring electrode as the anode,5 the pacing impedance was measured to be 705 Ω in the right ventricle and 874 Ω in the

Thresholds (V/0.5 ms)

Canine R-Wave Amplitudes (mV) 12 Weeks after Implantation

2

3

4 5 6 7 8 9 Implantation time (weeks)

10

11

12

Unipolar canine RV and LV thresholds for 5.8 mm2 CapSure SP electrodes

DCO, dual cathodal output; LV, left ventricular coronary vein; NA, not applicable; RV, right ventricular apex.

TABLE 1-4.

1

4 3.5 3

RVA LVCV

2.5 2 1.5 1 0.5 0 0

1

2

3

4 5 6 7 8 9 Implantation time (weeks)

10

11

12

Figure 1-36. Unipolar canine right ventricular apex (RVA) and left cardiac vein ventricular (LVCV) thresholds as functions of time after implantation. In the upper panel, the leads are paired in each animal with a 5.8-mm2 LV lead, and pooled data are shown in the RV from 1.2- and 4.0-mm2 electrodes. Pooling was allowed because there was no statistically significant difference in threshold between the two electrodes (P < .5). In the lower panel, the LV lead is compared with one of the same design in different animals. In both cases, the LV thresholds are about twice those of the right-sided leads.

left ventricle. Using a split-cathodal bipolar configuration, the measured impedance was 516 Ω, higher than would be predicted by Ohm’s law with both ventricles stimulated in parallel (390 Ω). The higher than predicted impedance is explained by the fact that the size and shape of the combined cathodes are different from those for either electrode alone. The combined cathodal configuration results in a Warburg resistance and capacitance. Combining electrodes of similar size essentially doubles the cathodal surface area. Because the electrode resistance of a hemispherical electrode is roughly proportional to the square root of the electrode surface area, doubling the size of the cathode by combining the RV and LV electrodes decreases the Warburg resistance by a factor of 1/√2. This helps to explain why the measured impedance using a split-cathodal configuration is not halved, as would be predicted by doubling the size of electrodes joined in parallel. Combining electrodes in parallel also increases the voltage droop during a constant-voltage pulse. Coronary venous pacing leads tend to result in higher impedance compared with endocardial RV leads (based on a lower volume of blood surrounding the electrode), and this further tends to shunt current from the higher-impedance

48

Section One: Basic Principles of Device Therapy

electrode to the lower-impedance electrode when both ventricles are stimulated in parallel. This factor serves to increase the apparent LV threshold with the splitcathodal configuration. These undesirable effects of a split-cathodal pacing configuration have largely been overcome by the addition of independent output circuits in newer CRT devices. However, because of the additional output circuit, there is additional battery drain with these devices. Another feature of CRT devices concerns the use of bipolar coronary venous leads. These leads differ from traditional right atrial or RV leads in that the surface area of the two coronary venous electrodes is quite similar. Therefore, when they are programmed to the bipolar pacing configuration, the impedance is much higher than with unipolar pacing, because current flows between two small electrodes rather than between one small and one large electrode. In addition, the bipolar pacing threshold is often considerably higher than the unipolar configuration. Bipolar leads have the advantage of offering the possibility of LV stimulation from either electrode, however. When combined with pulse generators capable of stimulating either of the two electrodes in a unipolar configuration, there may be a better chance of finding a coronary venous site that offers a low stimulation threshold. Another feature of bipolar coronary venous leads is that left phrenic nerve stimulation may be reduced if either electrode can be selected for unipolar pacing. However, when the electrodes are programmed to the bipolar configuration, the chances of left phrenic nerve stimulation may actually increase with a bipolar coronary venous lead, because there are two chances for an electrode to stimulate the phrenic nerve (one anodally and the other cathodally). Therefore, the chief advantage of the use of bipolar coronary venous leads in regard to stimulation is that there are two electrodes from which to choose for unipolar stimulation. Another advantage is that bipolar coronary venous stimulation avoids the chances of anodal RV stimulation when the RV ring electrode is used as the anode. The greatest advantages of bipolar leads in the coronary venous circulation relate to improved LV sensing.

Automated Capture Features In order to ensure ventricular capture and to allow programming of a low margin of safety, newer pacemakers use algorithms that automatically detect ventricular capture. Based on the measured stimulation threshold, the amplitude of the pacing stimulus is automatically adjusted to provide a programmed margin of safety. The Autocapture feature of St. Jude Medical (St. Paul, Minn.) pacemakers automatically adjusts the amplitude of the stimulation pulse by detecting capture in the ventricle from the evoked ventricular electrogram (the evoked response). The St. Jude Medical pacemakers require a bipolar ventricular pacing lead with low polarization properties for the distal electrode. The presence or absence of ventricular capture is determined by

sensing of the evoked response (ER) from the ring electrode. These devices automatically determine the evoked response gain and sensitivity levels by delivering five paired ventricular pulses of 4.5 V at a minimum pulse duration of 0.5 msec or the programmed value. The first of the paired pulses measures the evoked response, and the second pulse is delivered within 100 msec after the first (i.e., during the physiologic refractory period of the myocardium) to determine the level of polarization. If the amplitude of the evoked response is greater than 2.5 mV, the measured lead polarization is less than 4.0 mV, and the ratio of the evoked response to evoked response sensitivity is greater than 1.8 : 1, the device will automatically determine that the safety margin is acceptable and recommend Autocapture as a programmed feature. The Autocapture feature uses unipolar pacing from the tip electrode and determines capture on a beat-to-beat basis. If a ventricular stimulus is not followed by a detectable evoked response, a second test pulse is given at a value equal to 0.25 V greater than the last threshold measurement (a value known as the Automatic Pulse Amplitude, or APA). If a pulse is not followed by detectable capture, a backup pulse is delivered within 80 to 100 msec at an amplitude of 4.5 V. If two consecutive APA pulses are not followed by an evoked response, the threshold is measured to determine whether the APA needs adjustment. Specifically, the pulse is incremented 0.25 V above the last APA. If capture is not confirmed, the APA is repeated in 0.125 V increments until two consecutive captured events occur. For these devices, all loss-of-capture pulses are immediately followed by a backup pulse. A potential complicating factor with detection of the evoked response is differentiating fusion from capture. In the DDD(R) pacing mode, precisely timed intrinsic conduction can result in false detection of loss of capture. To differentiate true loss of capture from fusion, the AV delay is incremented by 100 msec after two consecutive loss-of-capture events to search for intrinsic conduction. If intrinsic conduction is indeed present during this extension of the AV delay, the backup pulse is eliminated. On the other hand, if subsequent backup pulses or APA increments are required due to loss of capture, the AV/PV delay is shortened to 50/25 msec. This sequence can introduce confusion into the interpretation of electrocardiographic tracings with irregular AV delays. However, knowledge of the function of the Autocapture algorithm allows recognition that this is a normal phenomenon. Automatic capture algorithms have also been applied to unipolar leads. The ELA Symphony pacemakers (ELA Medical, Sorin Group, Milan, Italy) detect the evoked response on the tip electrode of a unipolar lead, provided that the polarization properties of the electrode are favorable. Guidant (Boston Scientific, Natick, Mass.) pacemakers use a lower-capacitance output capacitor to minimize the afterpotentials on the ventricular lead as a method for improving detection of the evoked response. The smaller-output capacitor increases the droop of the pacing pulse, but the afterpotential is reduced. The smaller-output capacitor may have the effect of slightly increasing the apparent

Chapter 1: Cardiac Electrical Stimulation

stimulation threshold, an effect that is measurable only when the pacing threshold exceeds 2.5 V. Other manufacturers offer variants of the automatic capture algorithm that do not deliver backup pulses on a beat-to-beat basis but provide automatic determination of pacing threshold at programmed intervals during the day. These devices determine the pacing threshold and adjust the pacing amplitude to provide a programmed margin of safety. In general, automatic capture algorithms function quite effectively and may reduce the risk of loss of capture due to fluctuations in pacing threshold caused by drugs, metabolic derangement, or lead dislodgment. The capability for reducing the programmed margin of safety is effective for prolonging battery life and may reduce the frequency of clinic follow-up visits. The Medtronic Ventricular Capture Management feature determines a strengthduration threshold at a programmable interval (nominally, once per day). After the amplitude threshold is determined at a pulse duration of 0.4 msec, the pulse amplitude is doubled and a pulse duration threshold is measured. The permanent ventricular stimulation amplitude is then automatically reprogrammed using a programmable amplitude safety margin (usually twice the threshold) or a programmable minimum amplitude, whichever is higher. The nominal values for ventricular capture management are a safety margin of twice the threshold with a pulse duration of 0.4 msec and a minimum ventricular amplitude of 2.5 V. During measurement of the pacing threshold, each test pulse is followed by a backup pulse 110 msec later to ensure that a pacing pause does not occur. If the automatically measured ventricular stimulation threshold is greater than 2.5 V at 0.4 msec, the ventricular output is automatically programmed to 5.0 V and 1.0 msec. The Medtronic Atrial Capture Management feature is designed to periodically measure the atrial stimulation threshold and adapt the atrial output to a programmable amplitude safety margin. This feature does not use the evoked potential to determine the presence or absence of atrial capture. Rather, the pacemaker searches for evidence that atrial test pulses reset the sinus node (Atrial Chamber Reset Method) or observes the ventricular response to determine whether a captured atrial test pulse is conducted to the ventricles through the AV conduction system. The Atrial Capture Management feature performs an atrial amplitude threshold at 0.4-msec pulse duration and after loss of capture is detected (defined as two of three test pulses indicating loss of capture); the amplitude setting is increased until atrial capture is confirmed. Because this feature does not rely on detection of the evoked response, there is no restriction on the type of atrial lead that can be used. This feature will not measure thresholds if the sinus rate is consistently faster than 87 bpm.

Adequate Margin of Safety A pacemaker must be programmed with a safety factor that allows for changes in pacing threshold. One needs

49

to know how much thresholds actually change in the acute to chronic period as well as chronically on an hour-to-hour and day-to-day basis. Settings as low as 2.5 V and 0.5 msec at implantation appear to ensure capture for most adult patients with modern microporous, steroid-eluting leads. Higher values should be used with older technology leads. Some patients have large variations in threshold. The factors involved in estimating how great a margin of safety to allow in programming are the presently measured threshold, the probability of a catastrophe if pacing ceases, the ability of the patient to recognize intermittent loss of capture if it is occurring, the pharmacologic milieu, and how often the pacing threshold will be checked. In most patients, a ratio of output voltage (Vo) to threshold voltage (Vthr), or output current to threshold current, that is 1.5 : 1 or 2 : 1 provides a reasonable safety factor. The balance over years is that of safety factor beyond threshold versus how long the pulse generator can be used. Again, clinical factors and patient reliability in checking have to be considered. Children may experience a higher rate of exit block because of their active inflammatory responses. In all patients, but especially children, use of steroid-eluting electrodes is wise whenever possible.243-245 In a study reporting on 4953 threshold measurements made up to 20 years after implantation, long-term thresholds and threshold variations of non–steroid-eluting electrodes were found to be greater in children than in adults.81 The safety factors provided during the peak threshold phase are highest for steroid-eluting porous electrodes and lower for (in descending order) microporous electrodes, porous electrodes, and polished-tip electrodes, based on data represented in Figure 1-32. There may also be situations in which the patient’s pacemaker must be set at the maximum output at implantation (e.g., if the patient will not be available for follow-up). In these cases, an acceptable balance of safety margin and battery longevity might be obtained with a setting of 2.5 V and 0.5 msec for a steroid-eluting electrode, or a setting of up to 5 V and 0.5 msec with other leads. These numbers are at best only general guides. It is far better to arrange, by whatever means possible, for actual follow-up threshold measurements, both acutely and chronically. Modern perception of the range of chronic circadian threshold variation is based mainly on the work of Sowton and Norman246 (published in the mid-1960s), Preston and coworkers247 (in the late 1960s), and Westerholm248 (in 1971). The first two studies used constantcurrent generators and reported thresholds in terms of energy. Because the computed energy “thresholds” could have been a reflection of changes in impedance, the actual variation in voltage threshold cannot be determined from these studies. Westerholm, who reported both voltage and energy data, noted substantial circadian variations in both parameters. In all of these studies, the leads were primarily epicardial/ myocardial with polished electrodes. These human data, however, may be of marginal relevance to modern constant-voltage generators and porous, steroid-eluting electrodes.

50

Section One: Basic Principles of Device Therapy

Long-Term and Diurnal Variations in Pacing Threshold

18 Volts, microJ

McVenes and associates,249 in 1992, found no significant threshold changes in adult canines using then-modern chronic atrial or ventricular leads as a function of eating, sleeping, or exercise. This finding is supported by Kadish and coauthors,250 who found no changes in chronic human pacing thresholds during a 24-hour period in four of five patients studied. Although in one patient the threshold at 0.6 msec changed from 1 to 1.5 V between 3:00 and 6:00 PM, these investigators concluded that “ventricular pacing thresholds do not show substantial diurnal variability.”250 Grendahl and Schaanning also found minimal variation in pacing threshold during the day, after meals, or during sleeping or physical activity.192,251 Shepard and associates81 reviewed 4942 pacing threshold measurements they made in 257 patients with 312 non–steroid-eluting leads at up to 295 months after implantation. The median in-use time was 17 months. Of the measurements, 1053 were in children younger than 12 years of age. At stimulus durations of 0.5 ± 0.04 msec, for thresholds measured 1 month or more after implantation, the mean threshold was 1.2 ± 0.66 V for endocardial electrodes and 2.8 ± 1.39 V for epicardially applied electrodes. Highest mean thresholds were in the 6- to 12-year-old age group. In patients with five or more measurements after 3 months use, an increase in pacing threshold occurred after 3 months in 24%. An additional 21% had at least one threshold that exceeded the post-3-months individual patient mean by three standard deviations. Among other clinical events possibly related to threshold increases, one was the occurrence in a child of doubling of the threshold during two successive summers, at the times when symptoms resembling a mild cold began. The effects of various drugs on thresholds have been reported, but neither the test protocols nor the results have been consistent.252 Therefore, there appears to be little in the literature to support the statistical validity of any particular safety factor. On the basis of the earlier report of Preston and colleagues,247 Barold and associates253 suggested that Vo/Vthr must be at least 1.75 to ensure an adequate safety margin, assuming a 50% increase in energy at threshold throughout the day. Ohm and colleagues178,254 studied threshold evolution as a function of implant duration for an 8-mm2 polished platinum ring electrode and found that this lead had a peak threshold of 2.2 ± 0.75 V at a pulse duration of 0.5 msec measured 2 weeks after implantation. Assuming an output setting of 5 V and 0.5 msec pulse duration, the average patient had a Vo/Vthr of 2.3 : 1 during the peak threshold time. The 98th percentile patient (mean ± 2 standard deviations) had a Vo of 5 V and a Vthr of 2.2 ± 0.75 V, or a safety margin of about 1.35 : 1 at peak threshold. Unpredictable or unusual situations (e.g., myxedema) may occur that justify greater safety factor ratios.255 The important clinical point is that, for patients who are always or intermittently dependent on pacing to stay alive, a much greater pacing safety factor reduces the risk of otherwise unexplained sudden death.

22 20 Energy (2X voltage)

16 14 12 Energy (3X PW)

10 8

Energy (threshold)

6 4

Threshold (V)

2 0

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 PW (msec) Figure 1-37. Effect on current drain and safety margin of programming the stimulus voltage to twice the threshold (at a constant pulse duration) or programming the pulse width (PW) to three times the threshold value (at the threshold voltage). See text for discussion.

Programming Voltage Versus Pulse Width for Maximum Pulse Generator Longevity A common clinical concern for programming of the pulse generator to optimize battery longevity relates to whether it is more useful to program the amplitude or the duration of the output pulse. Based on examination of the strength-duration relationship, it is more efficient to reduce the Vo of the pulse, because the current drain varies as the square of voltage. Figure 1-37 illustrates the effect of doubling the Vthr (at a constant pulse width) or tripling the threshold pulse width (without changing voltage) on current drain. In this example, the rheobase voltage was determined to be 1 V and the chronaxie duration was 0.3 msec. The stimulation threshold was 4 V at a pulse duration of 0.1 msec or 2 V at 0.3 msec. Tripling the pulse duration at 4 V to 0.3 msec provided an adequate (2 : 1) safety margin with a current of 8 mA per pulse (4.8 μA continuous current) and a stimulus energy of 9.6 μJ. Similarly, doubling the Vthr at a pulse duration of 0.3 msec from 2 to 4 V yielded an identical current drain (8 mA/pulse, or 4.8 μA) and safety factor. If the patient had a higher threshold, for example 2 V at 1 msec, doubling of the voltage or tripling of the pulse width would still give the same current drain (12 μA), but the safety factor would be significantly different. Tripling the pulse width would provide a marginal (at best) safety margin, because threshold is approaching rheobase on the flat portion of the strength-duration curve. It would be necessary to double the voltage in this case to ensure a 2 : 1 safety margin. The foremost consideration for programming voltage and pulse duration is patient safety.

Summary Myocardial stimulation is the fundamental principle underlying artificial cardiac pacing. Perhaps the most

Chapter 1: Cardiac Electrical Stimulation

important concept for programming of an implantable pacing system is a thorough understanding of the strength-duration relationship. Pulse generators allow the clinician to program both the pulse amplitude (in volts) and the pulse duration (in milliseconds). The stimulation threshold is a function of both of these parameters. The exponential shape of the strengthduration curve must always be considered when programming the output pulse to ensure an adequate margin of safety between the delivered stimulus and the capture threshold. For example, pulse durations of 1 msec and greater are located on the flat portion of the strength-duration curve, whereas pulse durations of less than 0.15 msec are on the steeply rising portion of the curve. The practical importance of these facts can be appreciated by considering two points on the strength-duration curve shown in Figure 1-8. If the clinician determines the threshold to occur at point A (2 V and 0.5 msec) by decrementing the stimulus voltage at a constant pulse duration, programming of the pulse duration to 1 msec (point B) would provide very little margin of safety. Similarly, if the threshold is measured to be at point C (3.5 V and 0.15 msec) by decrementing the pulse duration at a constant voltage, doubling of the stimulation voltage to 7 V (point D) also would provide a poor safety margin. When one considers the shape of the strength-duration curve, a more appropriate programmed setting would be provided by doubling the threshold voltage at a pulse duration of 0.5 msec (point E, 4 V and 0.5 msec). As a general rule, if the threshold is determined by decrementing the stimulus voltage, an adequate margin of safety can be assumed by doubling the voltage if the pulse duration used was greater than 0.3 msec. The two most important points on the strengthduration curve (rheobase and chronaxie) are easily estimated with modern pulse generators (see Fig. 1-8). Rheobase can be estimated by decrementing the output voltage at a pulse duration of 1.5 to 2 msec. Chronaxie can then be estimated by determining the threshold pulse duration at twice the rheobase voltage. If instead the threshold is determined by decrementing the pulse duration, an adequate safety margin can be assumed by tripling the pulse duration only if the threshold is 0.15 msec or less. If the rheobase and chronaxie are measured, doubling of the threshold voltage at the chronaxie pulse duration provides an excellent method for programming a pacing system. However, in most circumstances, experienced clinicians will measure the pacing threshold at an initial stimulus duration of 0.4 or 0.5 msec and then make decisions based on their knowledge of the patient’s problems and medicines. What constitutes an adequate safety factor depends on knowledge of the patient’s status. How dependent on the pacemaker is the patient? How medically stable is the patient? What medicines that can influence the pacing threshold is this patient taking? Have important threshold variations been seen or are they anticipated in this patient after the initial stabilization period? Are there problems with the leads? How old is the pulse

51

generator, and what is the known history of this pulse generator and of the leads attached to it in other patients? These are important factors that move clinical judgment in regard to voltage or current safety factors, how often pacing and medical status will be checked, and whether there should be early or late replacement of the pulse generator and/or leads. The rules described in the previous paragraph are only a guide. Added to this must be clinical knowledge of the patient and of the particular pacing system. When programming the pulse generator at the time of implantation, the clinician must also consider the acute to chronic evolution of the stimulation threshold. Because there is typically an acute rise in threshold during the first several weeks after lead implantation, the voltage and pulse duration may need to be programmed to higher values than would be needed for chronic pacing. The physician is wise to re-evaluate the stimulation threshold after the acute rise (and sometimes subsequent fall) that may occur after implantation. For most patients, the pacing system can be programmed to chronic output settings at a follow-up evaluation about 6 weeks after lead implantation. Although these recommendations may not be as applicable to patients receiving a steroid-eluting lead, caution is still warranted. The importance of drug and electrolyte effects on the strength-duration curve should also be appreciated. For patients requiring antiarrhythmic drug therapy, the stimulation threshold should be measured a number of times after drug initiation to ensure an adequate margin of safety for pacing. Similarly, patients who are more likely to experience alterations in electrolyte concentration (e.g., patients with renal failure, patients taking potassium-wasting diuretics) may need their pacemakers to be programmed with a greater margin of safety. Perhaps most important, the degree to which the heart is dependent on pacing to sustain life or to prevent severe symptoms must be factored into the choice of a programmed margin of safety. For pacemaker-dependent patients, a pacing pulse that is at least 2.5 times the chronic capture threshold is usually recommended. In contrast, patients who are unlikely to experience severe symptoms should failure to capture occur may have their pacemaker programmed to a lower margin of safety (perhaps 2 times the threshold). The effect of pacing rate on the stimulation threshold should also be considered for patients who require antitachycardia pacing. The pacing threshold should be measured at all rates likely to be used for antitachycardia pacing. In the presence of high impedance due to lead fracture, the current output of a constant-voltage pulse generator decreases. Loss of capture can occur. If lead insulation failure occurs, the impedance as seen by the pulse generator may decrease because of current shunting to noncardiac tissue. This results in an increase in the current from the pulse generator without a change in the nominal output voltage. This change may not be detected early on if threshold is determined only by the voltage required for pacing capture. Of course, if one

52

Section One: Basic Principles of Device Therapy

Figure 1-38. Effects of bipolar constant–current pacing on voltage response and on after–current.

measures the voltage/current ratio, the nominal impedance and alterations in lead insulation or in wire continuity may be detectable. Because some wire fractures intermittently make and break contact, a normal impedance measurement does not always ensure that the lead is intact. Pacing impedance is determined by four factors: (1) resistance in the conductor wire pathways, (2) polarization at the electrode-tissue interfaces, (3) resistance (small geometric size for high resistance) at the electrode-tissue interface, and (4) impedance/resistance of the tissues between the electrodes. The first two of these factors are energy inefficient, decreasing the current available for stimulation, whereas the third factor decreases current drain without decreasing the efficiency of stimulation. An ideal electrode would have, among other attributes, high resistance and high capacitance (low polarization voltage) at the electrodetissue interface. Pacing with a monophasic stimulus is more energy efficient than pacing with a bipolar stimulus. One reason is that the pacing threshold is greater at normal stimulus durations for biphasic stimuli compared with uniphasic stimuli with the same total duration. For successful defibrillation, biphasic stimuli are more energy efficient. Figure 1-38 shows the relationships between a constant-current biphasic stimulus with microsecond-level delay between the phases and the electrode-electrolyte interface effects on voltage during the stimulus and on current flow after the stimulus stops. A biphasic stimulus with proper characteristics reduces the postpulse

ion rearrangements. Biphasic stimuli also may reverse some otherwise continuing local and undesirable chemical processes at the electrode. Electrical stimulation, not only of the heart but also of other kinds of tissue (e.g., brain), is being used clinically more and more. Increasing knowledge of fundamental factors in electrical stimulation and of what can be useful and practical or developed for particular clinical circumstances is desirable. Such knowledge will help cardiologists, as well as physicians and surgeons of other disciplines, make good decisions and expand the range of patient care. REFERENCES 1. Bockris JO’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum Publishers, 2000, p 1050. 2. Huxley Sir AH: Regarding Kenneth Stewart Cole, July 10, 1900April 18, 1984. In Huxley Sir AH: Biographical Memoirs). Washington, DC, National Academies Press, 1996, pp 24-45. Available at http://books.nap.edu/html/biomems/kcole.html 3. Cole KS: Membranes, Ions and Impulses: A Chapter of Classical Biophysics. Berkeley, University of California Press, 1968, p 173, figure 2:50, and chapter beginning on p 204. 4. Koch H: Recent advances in magnetocardiography. J Electrocardiol 37(Suppl):117-122, 2004. 5. Mayhew M, Johnson P, Slabaugh J, et al: Electrical characteristics of a split cathodal pacing configuration. PACE 26:22642271, 2003. 6. Corr PB: Contribution of lipid metabolites to arrhythmogenesis during early myocardial ischemia. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 720-722.

Chapter 1: Cardiac Electrical Stimulation 7. Kleber AG: Sodium-potassium pumping. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 37-54. 8. Hodgkin AL, Katz B: The effect of sodium ions on the electrical activity of the giant axon of the squid. J Physiol (Lond) 108:37, 1949. 9. Hodgkin AL, Huxley AF: A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol (Lond) 117:500-544, 1952. 10. Neher E, Sakmann B, Steinbach JH: The extracellular patch clamp: Method for resolving currents through individual open channels in biological membranes. Plugers Arch Eur J Physiol 375:219, 1978. 11. Neher E, Sakmann B: The patch clamp technique. Sci Am 266: 44-51, 1992. 12. Ebihara L: The sodium current. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 63-74. 13. Hartel HC, Duchatelle-Gourdon I: Structure and neural modulation of cardiac calcium channels. J Cardiovasc Electrophysiol 3:567-578, 1992. 14. Tsien RW: Calcium channels in the cardiovascular system. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 75-89. 15. Thomas RC: Electrogenic sodium pump in nerve and muscle cells. Physiol Rev 52:563-594, 1972. 16. Glitsch HG: Electrogenic Na pumping in the heart. Annu Rev Physiol 44:389-400, 1982. 17. Gadsky DC: The Na/K pump of cardiac cells. Annu Rev Biophys Bioeng 13:373-398, 1984. 18. Mullins IJ: The generation of electric currents in cardiac fibers by Na/Ca exchange. Am J Physiol 236:C103-C110, 1979. 19. Hilgemann DW: Numerical approximations of sodium-calcium exchange. Prog Biophys Mol Biol 51:1-45, 1988. 20. Carmeliet E: The cardiac action potential. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 55-62. 21. Hoffman BF, Cranefield PF: Electrophysiology of the Heart. New York, McGraw-Hill, 1960. 22. Binah O: The transient outward current in the mammalian heart. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 93-106. 23. Joho RH: Toward a molecular understanding of voltage-gated potassium channels. J Cardiovasc Electrophysiol 3:589-601, 1992. 24. Coraboeuf E, Carmeliet E: Existence of two transient outward currents in sheep cardiac Purkinje fibers. Pflugers Arch 392:352359, 1982. 25. Cohen IS, Datyner N: Repolarizing membrane currents. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 107-116. 26. Bottazzi F: Leonardo as physiologist. In Bottazzi F, Leonardo da Vinci. London, Leisure Arts, 1964, pp 373-387. 27. Harvey W: De Motu Cordis. [The Movement of the Heart and Blood]. Translated by D. Whiteridge. Oxford, Blackwell, 1976, 1628. 28. Keith A, Flack M: The form and nature of the muscular connections between the primary divisions of the vertebrate heart. J Anat Physiol 41:172-189, 1907. 29. Irisawa H: Comparative physiology of the cardiac pacemaker mechanism. Physiol Rev 58:461-498, 1978. 30. DiFrancesco D: The hyperpolarization-activated current, If, and cardiac pacemaking. In Rosen, MR, Janse ML, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, pp 117-132. 31. Urthaler F, Isobe JH, James TN: Comparative effects of glucagon on automaticity of the sinus node and atrioventricular junction. Am J Physiol 227:1415-1421, 1974. 32. Bean RC: Protein-mediated mechanisms of variable ion conductance in thin lipid membranes. Membranes 2:409-477, 1973.

53

33. Haydon DA, Hladky SB: Ion transport across thin lipid membranes: A critical discussion of mechanisms in selected systems. Q Rev Biophys 5:187-282, 1972. 34. Jongsma HJ, Rook MB: Biophysics of cardiac gap junction channels. In Zipes DP, Jalif J (eds): Cardiac Electrophysiology: From Cell to Bedside, 3rd ed. Philadelphia, WB Saunders, 2000, pp 119-125. 35. Bény J-L: Information networks in the arterial wall. News Physiol Sci 14:68-73, 1999. 36. Weidmann S: Passive properties of cardiac fibers. In Rosen MR, Janse MJ, Wit AL (eds): Cardiac Electrophysiology: A Textbook. Mt. Kisco, NY, Futura, 1990, 29-35. 37. Weidmann, S: Electrical constants of trabecular muscle from mammalian heart. J Physiol (Lond) 210:1041-1054, 1970. 38. Knisley SB: Transmembrane voltage changes during unipolar stimulation of rabbit ventricle. Circ Res 77:1229-1239, 1995. 39. Furman S, Parker B, Escher DJW: Decreasing electrode size and increasing efficiency of cardiac stimulation. J Surg Res 11:105, 1971. 40. Furman S, Hurzler P, Parker B: Clinical thresholds of endocardial cardiac stimulation: A long-term study. J Surg Res 19:149, 1975. 41. Angello DA, McAnulty JH, Dobbs J: Characterization of chronically implanted ventricular endocardial pacing leads. Am Heart J 107:1142-1145, 1984. 42. Geddes LA, Bourland JD: The strength-duration curve. IEEE Trans Biomed Eng BME-32:458-459, 1985. 43. Irnich W: Considerations in electrode design for permanent pacing. In Thalen HJT (ed): Cardiac Pacing. Proceedings of the IVth International Symposium on Cardiac Pacing. Assen, The Netherlands, Van Gorcum, 1973, p 268. 44. Irnich W: Engineering concepts of pacemaker electrodes. In Schaldach M, Furman S (eds): Advances in Pacemaker Technology. New York, Springer-Verlag, 1975, p 241. 45. Shepard RB: Invited letter concerning: The effect of extraanatomic bypass on aortic root impedance in open chest dogs: Should the vascular prosthesis be compliant to unload the left ventricle? J Thorac Cardiovasc Surg 104:1175-1177, 1992. 46. Bardou AL, Chenais J-M, Birkui PJ, et al: Directional variability of stimulation threshold measurements in isolated guinea pig cardiomyocytes: Relationship with orthogonal sequential defibrillating pulses. PACE 13:1590-1595, 1990. 47. Preston TA, Fletcher RD, Lucchesi BR, Judge RD: Changes in myocardial thresholds: Physiologic and pharmacologic factors in patients with implanted pacemakers. Am Heart J 74:235-242, 1967. 48. Katsumoto K, Niibori I, Takamatsu T, Kaibara M: Development of glassy carbon electrode (Dead Sea scroll) for low energy cardiac pacing. PACE 9(6 Pt 2):1220-1224, 1986. 49. Mond H, Stokes KB, Helland J, et al: The porous titanium steroid eluting electrode: A double blind study assessing the stimulation threshold effects of steroid. PACE 11:214-219, 1988. 50. Hill WE, Murray A, Bourks JP, et al: Minimum energy for cardiac pacing. Clin Phys Physiol Meas 9:41-46, 1988. 51. Breivik K, Ohm O-J, Engedal H: Acute and chronic pulse-width thresholds in solid versus porous tip electrodes. PACE 5:650657, 1982. 52. Irnich W: The chronaxie time and its practical importance. PACE 3:292-301, 1980. 53. Barold SS, Stokes KB, Byrd CL, McVenes R: Energy parameters in cardiac pacing should be abandoned. PACE 20:112-121, 1996. 54. Furman S, Hurzler P, Parker B: Clinical thresholds of endocardial cardiac stimulation: A long-term study. J Surg Res 19:149, 1975. 55. Ideker RE, Zhou X, Knisley SB: Correlation among fibrillation, defibrillation, and cardiac pacing. PACE 18:512-525, 1995. 56. Roth BJ: Virtual electrodes made simple: A cellular excitable medium modified for strong electrical stimuli. Rochester, MI,

54

57.

58. 59.

60.

61.

62.

63.

64.

65.

66.

67. 68.

69. 70. 71. 72.

73. 74. 75. 76. 77. 78. 79. 80.

Section One: Basic Principles of Device Therapy Department of Physics, Oakland University, 2002. Available at http://sprojects.mmi.mcgill.ca/heart/pages/rot/rothom.html Newton JC, Knisley SB: Review of mechanisms by which electrical stimulation alters the transmembrane potential. J Cardiovasc Electrophysiol 10:234-243, 1999. Knisley SB: Transmembrane voltage changes during unipolar stimulation of rabbit ventricle. Circ Res 77:1229-1239, 1995. Srinivasan R, Roth BJ: A mathematical model for electrical stimulation of a monolayer of cardiac cells. Biomed Eng Online 3:1, 2004. Wikswo JP Jr, Lin SF, Abbas RA: Virtual electrodes in cardiac tissue: A common mechanism for anodal and cathodal stimulation. Biophys J 69:2195-2210, 1995. Efimove IR, Gray RA, Roth BJ: Virtual electrodes and deexcitation: New insights into fibrillation induction and defibrillation. J Cardiovasc Electrophysiol 11:339-353, 2000. Knisley SB, Pollard AE, Fast VG: Effects of electrode-myocardial separation on cardiac stimulation in conductive solution. J Cardiovasc Electrophysiol 11:1132-1143, 2000. Knisley SB, Pollard AE: Effects of electrode-myocardial separation on cardiac stimulation in conductive solution. J Cardiovasc Electrophysiol 11:1132-1143, 2000. Shepard RB, Epstein AE, Kirklin JK, et al: Use of epipericardial rate-sensing/pacing electrodes in defibrillator implantation (avoid intra-pericardial scarring pretransplantation). PACE 15:575, 1992. Nikolski VP, Sambelashvili AT: Mechanisms of make and break excitation revisited: Paradoxical break excitation during diastolic stimulation. Am J Physiol Heart Circ Physiol 282:H565H575, 2002. Sambelashvili AT, Nikolski VP, Efimov IR: Virtual electrode theory explains pacing threshold increase caused by cardiac tissue damage. Am J Physiol Heart Circ Physiol 286:H2183H2194, 2004. Dressler L, Gruse G, von Knorre GH, et al: The optimization of the pulse delivered by the pacemaker. PACE 2:282, 1979. Chaptal AP, Ribot A: Statistical survey of strength-duration threshold curves with endocardial electrodes and long-term behavior of these electrodes. In Meere C (ed): Proceedings of the VI World Symposium on Cardiac Pacing. Montreal: PACESYMP 1979, pp 21-22. Blair HA: On the quantity of electricity and the energy in electrical stimulation. J Gen Physiol 19:951-964, 1936. Geddes LA: Historical evolution of circuit models for the electrode-electrolyte interface. Ann Biomed Eng 25:1-14, 1997. Hoorweg JL: Condensatorentladung und auseinandersetzung mit du Bois-Reymond. Pflugers Arch 52:87-108, 1892. Weiss G: Sur la possibilité de render comparable entre les appareils cervant à l’excitation électrique. Arch Ital Biol 35:413, 1901. Lapique L: Definition experimentale de l’excitabilité. C R Soc Bil 67:280, 1909. Lapique L: La chronaxie et ses applications physiologiques. Paris, Hermann et Cie, 1938. Irnich W: Elektrotherapie des herzens. Berlin, Fachverlag Schiele & Schon, 1976. Irnich W: The chronaxie time and its practical importance. PACE 3:292, 1980. Bernstein AD, Parsonnet V: Implications of constant energy pacing. PACE 6:1229-1233, 1983. Barold SS, Stokes KB, Byrd CL, McVenes R: Energy parameters in cardiac pacing should be abandoned. PACE 20:112-121, 1996. Coates S, Thwaites B: The strength-duration curve and its importance in pacing efficiency. PACE 23:1273-1277, 2000. Gurjao de Goday CM, de Magalhaes Galvao K, de Almeida Bacarin T, Franco GR: The effects of electrode position on the excitability of rat atria during postnatal development. Physiol Meas 23:649-659, 2002.

81. Shepard RB, Kim J, Colvin EC, et al: Pacing threshold spikes months and years after implant. PACE 14:1835-1841, 1991. 82. Wedensky NE: Uber die Beziehungzwischen Reizung und erregung im Tetanus. St. Petersburg, Ber Akad Wiss, 54:96, 1887. 83. Langberg JJ, Sousa J, El-Atassi R, et al: The mechanism of pacing capture hysteresis in humans [abstract]. PACE 15:577, 1992. 84. Swerdlow CD, Olson WH, O’Connor ME, et al: Cardiovascular collapse caused by electrocardiographically silent 60-Hz intracardiac leakage current. Circulation 99:2559-2564, 1999. 85. Hook BC, Perlman RL, Callans JD, et al: Acute and chronic cycle length dependent increase in ventricular pacing threshold. PACE 15:1437-1444, 1992. 86. Plumb VJ, Karp RB, James TN, Waldo AL: Atrial excitability and conduction during rapid atrial pacing. Circulation 63:1140-1149, 1981. 87. Buxton AE, Marchlinski FE, Miller JM, et al: The human atrial strength-interval relation: Influence of cycle length and procainamide. Circulation 79:271-280, 1989. 88. Katsumoto K, Niibori T, Watanabe Y: Rate dependent threshold changes during atrial pacing: Clinical and experimental studies. PACE 13:1009-1019, 1990. 89. Kay GN, Mulholland DH, Epstein AE, Plumb VJ: Effect of pacing rate on human atrial strength-duration curves. J Am Coll Cardiol 15:1618-1623, 1990. 90. Lindemans FW, Denier van der Gon JJ: Current thresholds and luminal size in excitation of heart muscle. Cardiovasc Res 12:477, 1977. 91. Irnich W, Gebhardt U: The pacemaker-electrode combination and its relationship to service life. In Thalen HJTh (ed): To Pace or Not to Pace: Controversial Subjects in Cardiac Pacing. The Hague, Martin Nijhoff, 1978, p 209. 92. Parsonnet V, Zucker IR, Kannerstein ML: The fate of permanent intracardiac electrodes. J Surg Res 6:285, 1966. 93. Thalen HJTh, Van den Berg JW: Threshold measurements and electrodes of the cardiac pacemaker. Acta Pharmacol Nederl 14:227, 1966. 94. Akyurekli Y, Taichman GC, White DL, et al: Myocardial responses to sutureless epicardial lead pacing. In Meere C (ed): Proceedings of the VI World Symposium on Cardiac Pacing. Montreal: PACESYMP 1979. 95. Stokes K, Bornzin G: The electrode-biointerface (stimulation). In Barold SS (ed): Modern Cardiac Pacing. Mt. Kisco, NY, Futura, 1985, pp 33-78. 96. Amundson DC, McArthur W, Moshaffafa M: The porous endocardial electrode. PACE 2:40, 1979. 97. Stokes KB, Frohlig G, Bird T, et al.: A new bipolar low threshold steroid eluting screw-in lead [abstract]. Eur JCPE 2:A89, 1992. 98. Greatbatch W: Metal electrodes in bioengineering. CRC Crit Rev Bioeng 5:1, 1981. 99. Horowitz P, Hill W: Section 1.18: Frequency analysis of reactive circuits. In Horowitz P, Hill W: The Art of Electronics. Cambridge, Cambridge University Press, 1980, pp 25-29. 100. Schmitt OH: Biological information processing using the concept of interpenetrating domains. In Leibovic KN (ed): Information Processing in the Nervous System. New York, SpringerVerlag, 1969, pp 325-331. 101. Geselowitz DB, Miller WT 3rd: A bidomain model for anisotropic cardiac muscle. Ann Biomed Eng.11:191-206, 1983. 102. Sepulveda NG, Roth BJ, Wikswo JP Jr: Current injection into a two-dimensional anisotropic bidomain. Biophys J 55:987-999, 1989. 103. Ashihara T, Trayanova NA: Asymmetry in membrane responses to electric shocks: Insights from bidomain simulations. Biophys J 87:2271-2282, 2004. 104. Roth BJ: The bidomain model: Two dimensional propagation in cardiac muscle. In Zipes DP, Jalife J (eds): Cardiac Electrophysiology: From Cell to Bedside, 3rd ed. Philadelphia, WB Saunders, 2000, pp 268-270.

Chapter 1: Cardiac Electrical Stimulation 105. Plonsey R, Barr RC: Inclusion of junction elements in a linear cardiac model through secondary sources: Application to defibrillation. Med Biol Eng Comput 24:137-144, 1986. 106. Roth BJ, Krassowska W: The induction of reentry in cardiac tissue. The missing link: How electric fields alter the transmembrane potential. Chaos 8:204-220, 1998. 107. Sharma V, Tung T: Theoretical and experimental study of sawtooth effect in isolated cardiac cell-pairs. J Cardiovasc Electrophysiol 12:1164-1173, 2001. 108. Sharma V, Tung L: Spatial heterogeneity of transmembrane potential responses of single guinea-pig cardiac cells during electric field stimulation. J Physiol 542:477-492, 2002. 109. Irnich W: The fundamental law of electrostimulation and its application to defibrillation. PACE 13(11 Pt 1):1433-1447, 1990. 110. Irnich W: George Weiss’ fundamental law of electrostimulation is 100 years old. PACE 25:245-248, 2002. 111. Thakor NV, Ranjan MS, Rajasekhar MS, et al: Effect of varying pacing waveform shapes on propagation and hemodynamics in the rabbit heart. Am J Cardiol 79:36-43, 1997. 112. Huang J, Ken-Knight BH, Walcott GP, et al: Effects of transvenous electrode polarity and waveform duration on the relationship between defibrillation threshold and upper limit of vulnerability. Circulation 96:1351-1359, 1997. 113. Wagner B: Electrodes, leads, and biocompatibility. In Webster J (ed): Design of Cardiac Pacemakers. Piscataway, NJ, IEEE Press, 1995, p 138. 114. Lilly JC, Hughes JR, Ellsworth CA, et al: Noninjurious electric waveform for stimulation of the brain. Science 121:468, 1955. 115. Moor WJ: Physical Chemistry. Englewood Cliffs, NJ, PrenticeHall, 1972, p 510. 116. von Helmholtz H: The modern development of Faraday’s conception of electricity, delivered before the Fellows of the Chemical Society in London on April 5, 1881. 117. Bockris J O’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum Publishers, 2000, p 885. 118. Bockris J O’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum Publishers, 2000, pp 1047-1049, 1122-1123. 119. Faraday M: On Electrical Decomposition. Philosophical Transactions of the Royal Society, 1834. 120. Bockris J O’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum Publishers, 2000, pp 771-1033. 121. Bockris J O’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum Publishers, 2000, pp 873-882. 122. Ragheb T, Geddes LA: Electrical properties of metallic electrodes. Biol Eng Comput 28:182-186, 1990. 123. Wagner B: Electrodes, leads, and biocompatibility. In Webster J (ed): Design of Cardiac Pacemakers. Piscataway, NJ, IEEE Press, 1995, p 138. 124. Bockris JO’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum, 2000, p 1455. 125. Kahn A, Greatbatch W: Physiologic electrodes. In Ray C (ed): Medical Engineering. Chicago, Year Book Medical, 1974, p 1073. 126. Mindt W, Schaldach M: Electrochemical aspects of pacing electrodes. In Schaldach M, Furman S (eds): Advances in Pacemaker Technology. New York, Springer-Verlag, 1975, p 297. 127. Knisley SB, Smith WM, Ideker RE: Effect of intrastimulus polarity reversal on electric field stimulation thresholds in frog and rabbit myocardium. J Cardiovasc Electrophysiol 3:239-254, 1992. 128. Recognizing a Warburg impedance. Research Solutions & Resources (Dr. Bob Rodgers). Available at http://www.consultrsr. com/resources/eis/warburg1.htm

55

129. Bockris JO’M, Reddy AKN, Gamboa-Aldeco M: Modern Electrochemistry 2A: Fundamentals of Electrodics, 2nd ed. New York, Kluwer Academic/Plenum Publishers, 2000, p 1133. 130. Ovadia M, Zavitz DH: Impedance spectroscopy of the electrodetissue interface of living heart with isoosmotic conductivity perturbation. Chem Phys Lett 390:445-453, 2004. 131. Rodman JE: Solution, Surface and Solid State Assembly of Porphyrins [PhD thesis]. Cambridge, University of Cambridge, 2000, Fig. 4.6. 132. Rho RW, Patel VV, Gerstenfeld EP, et al: Elevations in ventricular pacing threshold with the use of the Y Adapter: Implications for biventricular pacing. PACE 26:747-751, 2003. 133. Krassowska W, Neu JC: Response of a single cell to an external electric field. Biophys J 66:1768-1776, 1994. 134. Lu C-C, Kabakov A, Maarkin V: Membrane transport mechanisms probed by capacitance measurements with megahertz voltage clamp. Proc Natl Acad Sci U S A 92:11220-11224, 1995. 135. Mehra R, Furman S: Comparison of cathodal, anodal, and bipolar strength-interval curves with temporary and permanent pacing electrodes. Br Heart J 41:468-476, 1979. 136. Kay GN: Basic aspects of cardiac pacing. In Ellenbogen KA (ed): Cardiac Pacing. Cambridge, MA, Blackwell Scientific, 1992. 137. Preston TA: Anodal stimulation as a cause of pacemakerinduced ventricular fibrillation. Am Heart J 86:366-372, 1973. 138. Wiggers CJ, Wegria R, Pinera B: Effects of myocardial ischemia on fibrillation threshold: Mechanism of spontaneous ventricular fibrillation following coronary occlusion. Am J Physiol 131:309316, 1940. 139. Mehra R, Furman S, Crump JF: Vulnerability of the mildly ischemic ventricle to cathodal, anodal and bipolar stimulation. Circ Res 41:159-166, 1977. 140. Bilitch M, Cosby RS, Cafferry EA: Ventricular fibrillation and competitive pacing. N Engl J Med 276:598-604, 1967. 141. Stokes K, Bird T, Taepke R: A new low threshold, high impedance microelectrode. In Antonioli GE, Reufeut AE, Ector H (eds): Pacemaker Leads. Amsterdam, Elsevier, 1991, pp 543-548. 142. Bird T, Stokes KB: Ventricular electrode spacing and anode size [abstract 205]. Revue Europeene De Technologie Biomedicale 3:63, 1990. 143. De Caprio V: Endocardial Electrograms from Transvenous Pacemaker Electrodes [PhD thesis in biomedical engineering.] New York, Polytechnic Institute of New York, 1977. 144. Garberoglio B, Inguaggiato B, Chinaglia B, Cerise O: Initial results with an activated pyrolytic carbon tip electrode. PACE 6:440, 1982. 145. Jones JL, Jones RE, Balasky G: Improved cardiac cell excitation with symmetrical biphasic defibrillation waveforms. Am J Physiol 253:H1418-H1424, 1987. 146. Tse HF, Lau CP, Leung SK, et al: Single lead DDD system: A comparative evaluation of unipolar, bipolar and overlapping biphasic stimulation and the effects of right atrial floating electrode location on atrial pacing and sensing thresholds: PACE 19:1758-1763, 1996. 147. Izquierdo R, Rodrigo G, Pelegrin J, et al: Single lead DDD pacing using electrodes with longitudinal and diagonal atrial floating dipoles. PACE 25:1692-1698, 2002. 148. Ripart A, Fletcher R: Sensing. In Ellenbogen KA, Kay GN, Wilkoff B (eds): Clinical Cardiac Pacing. Philadelphia, WB Saunders, 1992. 149. Barold SS, Ong LS, Heinle RA: Stimulation and sensing thresholds for cardiac pacing: Electrophysiologic and technical aspects. Prog Cardiovasc Dis 24:1, 1981. 150. Barold SS, Ong LS, Heinle RA: Stimulation and sensing thresholds for cardiac pacing: Electrophysiologic and technical aspects. Prog Cardiovasc Dis 24:1, 1981. 151. Smyth NPD, Tarjan PP, Chernoff E, et al: The significance of electrode surface area and stimulating thresholds in permanent cardiac pacing. J Thorac Cardiovasc Surg 71:559, 1976.

56

Section One: Basic Principles of Device Therapy

152. Parsonnet V, Zucker IR, Kannerstein ML: The fate of permanent intracardiac electrodes. J Surg Res 6:285, 1966. 153. Thalen HJTh, Van den Berg JW: Threshold measurements and electrodes of the cardiac pacemaker. Acta Pharmacol Nederl 14:227, 1966. 154. Akyurekli Y, Taichman GC, White DL, et al: Myocardial responses to sutureless epicardial lead pacing. In Meere C (ed): Proceedings of the VI World Symposium on Cardiac Pacing. Montreal: PACESYMP 1979. 155. Furman S, Hurzler P, Parker B: Clinical thresholds of endocardial cardiac stimulation: A long-term study. J Surg Res 19:149, 1975. 156. Wilson GJ, MacGregor DC, Bobyn JD, et al: Tissue response to porous-surface electrodes: basis for a new atrial lead design. In Moore C (ed): Proceedings of the VI World Symposium on Cardiac Pacing: PACESYMP 1979. 157. Amundson D, McArthur W, MacCarter D, et al: Porous electrode-tissue interface. In Moore C (ed): Proceedings of the VI World Symposium on Cardiac Pacing: PACESYMP 1979. 158. Amundson DC, McArthur W, Moshaffafa M: The porous endocardial electrode. PACE 2:40, 1979. 159. MacGregor DC, Wilson GJ, Lixfeld W, et al: The porous surface electrode: A new concept in pacemaker lead design. J Thorac Cardiovasc Surg 78:281, 1979. 160. Breivik K, Ohm O-J, Engedahl H: Acute and chronic pulse-width thresholds in solid versus porous tip electrodes. PACE 5:650, 1982. 161. Berman ND, Dickson SE, Lipton IM: Acute and chronic clinical performance comparison of porous and solid electrode design. PACE 5:67, 1982. 162. Freud GE, Chinaglia B: Sintered platinum for cardiac pacing. Int J Artif Organs 4:238, 1981. 163. MacCarter DM, Lundberg KM, Corstjens JP: Porous electrodes: Concept, technology and results. PACE 6:427, 1983. 164. MacGregor DC, Pilliar RM, Wilson GJ, et al: Porous metal surfaces: A radical new concept in prosthetic heart valve design. Trans Am Soc Artif Intern Organs 22:646, 1976. 165. Elmqvist H, Schuller H, Richter G: The carbon tip electrode. PACE 6:436, 1983. 166. Stokes KB, Bornzin G: The electrode-biointerface (stimulation). In Barold SS (ed): Modern Cardiac Pacing. Mt. Kisco, NY, Futura, 1985, pp 33-78. 167. Cornacchia O, Maresta A, Nigro P, et al: Effect of propafenone on chronic ventricular pacing threshold in patients with steroideluting (capture) and conventional leads. Eur J Cardiac Pacing Electrophysiol 2:A88, 1992. 168. Pearce JA, Bourland JD, Neilsen W, et al: Myocardial stimulation with ultrashort duration current pulses. PACE 5:52-58, 1982. 169. Meyers GH, Parsonnet V: Engineering in the Heart and Blood Vessels. New York, Wiley-Interscience, 1989. 170. Hurzeler P, Furman S, Escher DJW: Cardiac pacemaker current thresholds versus pulse duration. In Silverman HT, Miller IF, Salkind AJ (eds): Electrochemical Bioscience and Bioengineering. Princeton, NJ, Electrochemical Society, 1973, p 124. 171. Barold SS, Winner JA: Techniques and significance of threshold measurement for cardiac pacing. Chest 70:760, 1976. 172. Brownlee WC, Hirst R: Six years experience with atrial leads. PACE 9(6 Pt 2):1239-1242, 1989. 173. Luceri RM, Furman S, Hurzeler P, et al: Threshold behavior of electrodes in long-term ventricular pacing. Am J Cardiol 40:184, 1977. 174. Bornzin GA, Stokes KB, Wiebusch WA: A low-threshold, lowpolarization platinized endocardial electrode [abstract]. PACE 6:A-70, 1983. 175. Elmqvist H, Schuller H, Richter G: The carbon tip electrode. PACE 6:436, 1983. 176. Mond H, Stokes KB: The electrode-tissue interface: The revolutionary role of steroid elution. PACE 15:95-107, 1992.

177. Ohm O-J, Breivik K: Pacing leads. In Gomez FP (ed): Cardiac Pacing, Electrophysiology, Tachyarrhythmias. Madrid, Editorial Group, 1985, pp 971-985. 178. Hoff PI, Breivik K, Tronstad A, et al: A new steroid-eluting electrode for low-threshold pacing. In Gomez FP (ed): Cardiac Pacing, Electrophysiology, Tachyarrhythmias. Mt. Kisco, NY, Futura, 1985, pp 1014-1019. 179. Anderson JM: Inflammatory response to implants. ASAIO Trans 34:101-107, 1988. 180. Henson PM: Mechanisms of exocytosis in phagocytic inflammatory cells. Am J Pathol 101:494-514, 1980. 181. Robinson TF, Cohen-Gould L, Factor SM: Skeletal framewok of mammalian heart muscle: Arrangement of inter- and pericellular connective tissue structures. Lab Invest 29:482-498, 1983. 182. Preston TA, Judge RD: Alteration of pacemaker threshold by drug and physiologic factors. Ann N Y Acad Sci 167:686-692, 1969. 183. Stokes KB, Anderson J: Low threshold leads: The effect of steroid elution. In Antonioli GE (ed): Pacemaker Leads. Amsterdam, Elsevier, 1991, pp 537-542. 184. Sibille Y, Reynolds HY: Macrophages and polymorphonuclear neutrophils in lung defense and injury. Am Rev Respir Dis 141:471-502, 1990. 185. Benditt DG, Kriett JS, Ryberg C, et al: Cellular electrophysiologic effects of dexamethasone sodium phosphate: Implications for cardiac stimulation with steroid-eluting electrodes. Int J Cardiol 22:67-73, 1989. 186. Stokes KB, Kriett JM, Gornick CA, et al: Low-threshold cardiac pacing electrodes. In Frontiers of Engineering in Health Care, 1983: Proceedings of the Fifth Annual Conference IEEE Engineering in Medicine and Biology Society, 1983. 187. Irnich W: Considerations in electrode design for permanent pacing. In Thalen HJT (ed): Cardiac Pacing. Proceedings of the IVth International Symposium on Cardiac Pacing. Assen, The Netherlands, Van Gorcum, 1973, p 268. 188. Irnich W: Engineering concepts of pacemaker electrodes. In Schaldach M, Furman S (eds): Advances in Pacemaker Technology. New York, Springer-Verlag, 1975, p 241. 189. Mond H, Sloman JG, Cowling R, et al: The small tined pacemaker lead: Absence of displacement. In Meere C (ed): Proceedings of the VI World Symposium on Cardiac Pacing, Montreal: PACESYMP 1979, pp 29-35. 190. Baker JH, Shepard RB, Plimb VJ, Kay GN: Effects of fixation mechanism and electrode material on atrial stimulation threshold: Long-term evaluation in 338 patients [abstract]. PACE 15:54, 1992. 191. Cornacchia D, Jacopi F, Fabbri M, et al: Comparison between active screw-in and passive leads for permanent transvenous ventricular pacing [abstract]. PACE 6:A56, 1983. 192. EI Gamal M, Van Gelder L, Bonnier J, et al: Comparison of transvenous atrial electrodes employing active (helicoidal) and passive (tined J-lead) fixation in 116 patients [abstract]. PACE 6:205, 1983. 193. Kay GN, Anderson K, Epstein AE, Plumb VJ: Active fixation atrial leads: Randomized comparison of two lead designs. PACE 12:1355-1361, 1989. 194. Rasor NS, Spickler JW, Clabaugh JW: Comparison of power sources for advanced pacemaker applications. In Proceedings of the 7th Intersociety Energy Conversion Engineering Conference, Washington, DC, American Chemical Society, 1972, p 752. 195. Hirshorn MS, Holley LK, Hales JR, et al: Screening of solid and porous materials for pacemaker electrodes. PACE 4:380, 1981. 196. Schaldah M: New pacemaker electrodes. Trans Am Soc Artif Intern Organs 17:29, 1971. 197. Helland J, Stokes KB: Nonfibrosing cardiac pacing electrode. U.S. Patent No. 4033357. February 17, 1976. 198. Elmqvist H, Schuller H, Richter G: The carbon tip electrode. PACE 6:436, 1983.

Chapter 1: Cardiac Electrical Stimulation 199. Thuesen L, Jensen PJ, Vejby-Christensen H, et al: Lower chronic stimulation threshold in the carbon-tip than in the platinum-tip endocardial electrode: A randomized study. PACE 12:1592-1599, 1989. 200. Bornzin GA, Stokes KB, Wiebush WA: A low threshold, low polarization, platonized endocardial electrode. PACE 6:A-70, 1983. 201. Mugica J, Duconge B, Henry L, et al: Clinical experience with new leads. PACE 11:1745-1752, 1988. 202. Djordjevic M, Stojanov P, Velimirovic D, et al: Target lead-low threshold electrode. PACE 9:1206-1210, 1986. 203. Amundson DC, McArthur W, Moshaffafa M: The porous endocardial electrode. PACE 2:40, 1979. 204. Timmis GC, Helland J, Westveer DC, et al: The evolution of low threshold leads. Clin Prog Pacing Electrophysiol 1:313, 1983. 205. Berman ND, Dickson SE, Lipton IM: Acute and chronic clinical performance comparison of porous and solid electrode design. PACE 5:67, 1982. 206. Freud GE, Chinaglia B: Sintered platinum for cardiac pacing. Int J Artif Organs 4:238, 1981. 207. Kay GN, Anderson K, Epstein AE, Plumb VJ: Active fixation atrial leads: Randomized comparison of two lead designs. PACE 12:1355-1361, 1989. 208. Stokes KB: Preliminary studies on a new steroid eluting epicardial electrode. PACE 11:1797-1803, 1988. 209. Hamilton R, Gow R, Bahoric B, et al: Steroid-eluting epicardial leads in pediatrics: Improve epicardial thresholds in the first year. PACE 14:2066, 1991. 210. Stokes KB, Frohling G, Bird T, et al: A new bipolar low threshold steroid eluting screw-in lead. Eur J Cardiac Pacing Electrophysiol 2:A89, 1992. 211. Schwaab B, Frohling G, Schwerdt H, et al: Long-term follow-up of a bipolar steroid eluting pacing lead with active and passive fixation. In Antoniolo GE (ed): Pacemaker Leads 1997. Bologna, Monduzzi Editore, 1997, pp 361-364. 212. Schwaab B, Frohling G, Schwerdt H, et al: Long-term follow-up of three microporous active fixation leads in atrial position. In Antoniolo GE (ed): Pacemaker Leads 1997. Bologna, Monduzzi Editore, 1997, pp 365-368. 213. Schwaab B, Frohling G, Schwerdt H, et al: Atrial and ventricular pacing characteristics of a steroid eluting screw-in lead. In Antoniolo GE (ed): Pacemaker Leads 1997. Bologna, Monduzzi Editore, 1997, pp 383-388. 214. Menozzi C: Comparison between latest generation steroideluting screw-in and tined leads: Long term follow-up. In Antoniolo GE (ed): Pacemaker Leads 1997. Bologna, Monduzzi Editore, 1997, pp 389-394. 215. Nurnberg JH, Schopper H, Busscher U, et al: Retrospective comparison of epicardial steroid-eluting and conventional leads for pacing after corrective surgery in congenital heart disease. PACE 20:1193, 1997. 216. Wallace AG, Cline RE, Sealy WC, et al: Electrophysiologic effects of quinidine. Circ Res 19:960-969, 1966. 217. Gay RJ, Brown DF: Pacemaker failure due to procainamide toxicity. Am J Cardiol 34:728-731, 1974. 218. Moss AJ, Goldstein S: Clinical and pharmacological factors associated with pacemaker latency and incomplete pacemaker capture. Br Heart J 31:112, 1969. 219. Hellestrand KJ, Burnett PJ, Milne JR, et al: Effect of the antiarrhythmic agent flecainide acetate on acute and chronic pacing thresholds. PACE 6:892, 1983. 220. Salel AF, Seagren SC, Pool PE: Effects on encainide on the function of implanted pacemakers. PACE 12:1439, 1989. 221. Montefoschi N, Boccadamo R: Propafenone induced acute variation of chronic atrial pacing threshold: A case report. PACE 13:480-483, 1990. 222. Huang SK, Hedberg PS, Marcus FI: Effects of antiarrhythmic drugs on the chronic pacing threshold and the endocardial R wave amplitude in the conscious dog. PACE 9:660, 1986.

57

223. Bianconi L, Boccadamo R, Toscano S, et al: Effects of oral propafenone therapy on chronic myocardial pacing threshold. PACE 15:148-154, 1992. 224. Kubler W, Sowton E: Influence of beta-blockade on myocardial threshold in patients with pacemakers. Lancet 2:67, 1970. 225. Irnich W, Gebhardt U: The pacemaker-electrode combination and its relationship to service life. In Thalen HJTh (ed): To Pace or Not to Pace: Controversial Subjects in Cardiac Pacing. The Hague, Martin Nijhoff, 1978, p 209. 226. Gay RJ, Brown DF: Pacemaker failure due to procainamide toxicity. Am J Cardiol 34:728-731, 1974. 227. Preston TA, Fletcher RD, Lucchesi BR, Judge RD: Changes in myocardial threshold: Physiologic and pharmacologic factors in patients with implanted pacemakers. Am Heart J 74:235, 1967. 228. Khastgir T, Lattuca J, Aarons D, et al: Ventricular pacing threshold and time to capture postdefibrillation in patients undergoing implantable cardioverter-defibrillator implantation. PACE 14:768-772, 1991. 229. Delmar M: Role of potassium currents on cell excitability in cardiac ventricular myocytes. J Cardiovasc Electrophysiol 3:474486, 1992. 230. Gettes LS, Shabetai R, Downs TA, et al: Effect of changes in potassium and calcium concentrations on diastolic threshold and strength-interval relationships of the human heart. Ann N Y Acad Sci 167:693-705, 1969. 231. Lee D, Greenspan K, Edmands RE, et al: The effect of electrolyte alteration on stimulus requirement of cardiac pacemakers. Circulation 38:124, 1968. 232. Walker WJ, Elkins JT, Wood LW, et al: Effect of potassium in restoring myocardial response to a subthreshold cardiac pacemaker. N Engl J Med 271:597, 1964. 233. Surawicz B, Chelbus H, Reeves JT, et al: Increase of ventricular excitability threshold by hyperpotassemia. JAMA 191:71-76, 1965. 234. Westerholm CJ: Threshold studies in transvenous cardiac pacemaker treatment. Scand J Thorac Cardiovasc Surg 8(Suppl):1, 1971. 235. Schlesinger Z, Rosenberg T, Stryjer D, et al: Exit block in myxedema, treated effectively by thyroid hormone replacement. PACE 3:737-739, 1980. 236. Basu D, Chatterjee K: Unusually high pacemaker threshold in severe myedema: Decrease with thyroid hormone therapy. Chest 70:677-679, 1976. 237. Nagatomo Y, Ogawa T, Kumagae H, et al: Pacing failure due to markedly increased stimulation threshold 2 years after implantation: Successful management with oral prednisolone. A case report. PACE 12:1034-1037, 1989. 238. Haywood J, Wyman MG: Effects of isoproterenol, ephedrine, and potassium on artificial pacemaker failure. Circulation 32(Suppl II):110, 1965. 239. Katz A, Knilans TK, Evans JJ, Prystowsky EN: The effects of isoproterenol on excitability, supranormal excitability and conduction in the human ventricle [abstract]. PACE 14:710, 1991. 240. Levick CE, Mizgala HF, Kerr CR: Failure to pace following high dose anti-arrhythmic therapy: Reversal with isoproterenol. PACE 7:252-256, 1984. 241. Daubert C, Ritter P, Cazeau S, et al: Permanent biventricular pacing in dilated cardiomyopathy: Is a totally endocardial approach technically feasible? PACE 19:699, 1996. 242. McVenes R, Stokes K: Alternative pacing sites: How the modern technology deals with this new challenge. In Antonioli GE (ed): Pacemaker Leads 1997. Bologna, Monduzi Editore, 1997, pp 223-228. 243. Hurzeler P, Furman S, Escher DJW: Cardiac pacemaker current thresholds versus pulse duration. In Silverman HT, Miller IF, Salkind AJ (eds): Electrochemical Bioscience and Bioengineering. Princeton, NJ, Electrochemical Society, 1973, p 124. 244. Stokes KB, Church T: The elimination of exit block as a pacing complication using a transvenous steroid-eluting lead. Proceed-

58

245. 246.

247.

248. 249.

Section One: Basic Principles of Device Therapy ings of the VIII World Symposium on Cardiac Pacing and Electrophysiology, Jerusalem, 1987. [abstract 475] PACE 10(3 Pt 2):748, 1987. Till JA, Jones S, Rowland E, et al: Clinical experience with a steroid eluting lead in children [abstract]. Circulation 80:389, 1989. Sowton E, Norman J: Variations in cardiac stimulation thresholds in patients with pacing electrodes. Digest of the 7th International Conference on Medical and Biological Engineering, Stockholm, 1967. Preston TA, Fletcher RD, Luchesi BR, et al: Changes in myocardial threshold: Physiologic and pharmacologic factors in patients with implanted pacemakers. Am Heart J 74:235-242, 1967. Westerholm C-J: Threshold studies in transvenous cardiac pacemaker treatment. Scand J Thorac Surg Suppl 8 (Suppl):1-35, 1971. McVenes R, Lahtinen S, Hansen N, Stokes K: Physiologic and drug induced changes in cardiac pacing and sensing parameters [abstract 324]. Eur JCPE 2:A86, 1992.

250. Kadish A, Kong T, Goldberger J: Diurnal variability in ventricular stimulation threshold and electrogram amplitude [abstract]. Eur JCPE 2:A86, 1992. 251. Grendahl H, Schaanning CG: Variations in pacing threshold. Acta Med Scand 187:75-78, 1970. 252. Barold S: Effect of drugs on pacing thresholds. In Antonioli GE, Aubert AE, Ector H (eds): Pacemaker Leads 1991. New York, Elsevier, 1991, pp 73-86. 253. Barold SS, Ong LS, Heinle RA: Stimulation and sensing thresholds for cardiac pacing: Electrophysiologic and technical aspects. Prog Cardiovasc Dis. 24:1, 1981. 254. Ohm O-J, Breivik K: Pacing leads. In Gomez FP (ed): Cardiac Pacing, Electrophysiology, Tachyarrhythmias. Madrid, Editorial Group, 1985, pp 971-985. 255. Basu D, Chatterjee K: Unusually high pacemaker threshold in severe myedema: Decrease with thyroid hormone therapy. Chest 70:677-679, 1976.

Chapter 2

Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms GREGORY P. WALCOTT • RAYMOND E. IDEKER

E lectrical defibrillation is the only practical means for halting ventricular fibrillation (VF). Although it has been known for more than a century that application of an electric shock directly to the myocardium causes VF and that the heart can be returned to normal rhythm by subsequent application of a shock of greater magnitude,1-3 knowledge of the mechanisms underlying the process of defibrillation was slow in developing. It was only with the relatively recent introduction of novel techniques for the analysis of action potentials and activation sequences4-8 that greater insight into the physiology of both fibrillation and defibrillation has been achieved. It is hoped that this insight will result in a higher success rate for external defibrillation and improved design of implantable cardioverter-defibrillators (ICDs). A large part of current research is dedicated to determining the underlying reason for the success or failure of a defibrillating shock. VF is maintained by multiple activation fronts that are constantly moving in a pattern of reentry. Characteristics of the activation pattern and action potential are believed to be important determinants of whether a shock will successfully defibrillate the heart. A successful defibrillating shock

is believed to extinguish most of these activation fronts, permitting the resumption of coordinated responsiveness.9-13 For the defibrillating shock to be completely successful, this must be accomplished without creating an environment that promotes susceptibility to reinitiation of fibrillation.12,13 It has been established that adequate distribution of the potential gradient (an estimate of local current flow) created by the shock throughout the ventricular myocardium is required for successful defibrillation.14-16 Fundamentally, defibrillation is believed to be realized through an electrical pulse that causes an alteration in the transmembrane potential of the myocyte. It most likely requires a rapid induction of changes in the transmembrane potential of the myocytes in a critical mass of myocardium (75% to 90% of the myocardium in dogs).11,14,16 Because this represents a large mass of tissue, depolarization must be achieved at a considerable distance from the stimulating electrode. To gain an understanding of this complex farfield process, various mathematical models have been generated, and the predictions of computer simulations have been compared with physiologic findings. Both discontinuities in the anisotropic properties of the 59

60

Section One: Basic Principles of Device Therapy

extracellular and intracellular domains, as described by the bidomain model,17,18 and highly resistive discontinuities in the intracellular space (e.g., collagenous septa), as described in the secondary source model,19-21 may contribute to the far-field changes in the potential gradient that halt the activation fronts of fibrillation. The mechanisms underlying degeneration into fibrillation in failed shocks remain incompletely understood. Residual wandering wavelets,22 nonuniform refractoriness,23 and areas of low potential gradient in which critical points (centers of reentrant circuits) form15 may be the sources of propagating wavefronts that can result in fibrillation through reentry. Centrifugal propagation from ectopic foci induced by the defibrillation shock may also play a role, especially in the atrium.13 This chapter expands on the subjects mentioned previously. Some of the characteristics of VF that are believed to be important to understanding defibrillation and some of the characteristics of shock that lead to successful defibrillation, such as waveform shape and electrode configuration, are discussed. Then, a shock is traced from its origin at the defibrillation electrodes to its distribution through the heart with a discussion of its affect on the transmembrane potential and how it leads to the successful cessation of fibrillation.

Fibrillation To understand defibrillation, it is necessary to have an understanding of fibrillation. Knowing the basic characteristics of VF and whether it is maintained by reentrant or focal activity, as well as the characteristics of the action potential and the excitability of the fibrillating tissue, helps to define the therapy that will be successful in stopping the arrhythmia. VF has been characterized as progressing through four stages based on high-speed cinematography of electrically induced fibrillation in dog hearts (Fig. 2-1).24 A brief undulatory, or tachysystolic, stage lasting only 1 to 2 seconds occurs first. It is characterized by three to six undulatory contractions that resemble a series of closely occurring systoles and involve the sequential contraction of large areas of the myocardium. This is followed by a second stage of convulsive incoordination, in which more frequent waves of contraction sweep over smaller regions of the myocardium. Because the contractions in each region are not in phase, the ventricles are pulled in a convulsive manner. It is during this stage of fibrillation that the ICD shocks are given—10 to 20 seconds after the onset of fibrillation. In the third stage of tremulous incoordination, the independently contracting areas of the ventricular surface become even smaller, giving the heart a tremulous appearance. Tremulous incoordination lasts for 2 to 4 minutes before the fourth and final stage of atonic fibrillation occurs. Atonic fibrillation develops within 2 to 5 minutes after the onset of fibrillation and is characterized by the slow passage of

B A

C

D

E

F

Figure 2-1. Diagrams indicating the spread of waves observed in the analysis of motion pictures obtained during the four stages of fibrillation described by Wiggers. A, Spread of wavefront during initial, undulatory stage. B, Theoretical passage of impulses from point x to form a wavefront at y. Panels C through F show the appearance of contraction waves in the small rectangular area W from panel A, magnified. C, Undulatory stage. D, Convulsive stage. E, Tremulous stage. F, Atonic stage. (From Wiggers CJ: Studies of ventricular fibrillation caused by electric shock. Cinematographic and electrocardiographic observations of the natural process in the dog’s heart: Its inhibition by potassium and the revival of coordinated beats by calcium. Am Heart J 5:351-365, 1930.)

feeble contraction wavelets over short distances. With time, the number of quiescent areas increases. Ischemia plays a role in the development of the third and fourth stages, because the fibrillating heart remains in the second stage if the coronary arteries are perfused with oxygenated blood.25,26 Driving the mechanical activity of the heart during fibrillation is the electrical activity of the myocardium. The electrical activity of the heart during fibrillation has been studied using both extracellular and optical recordings. Several groups have suggested that fibrillation is maintained by reentry. In most cases, reentry appears to be caused by “wandering wavelets” of activation, activation fronts that follow continually changing pathways from cycle to cycle. In some studies, the activation sequence appears moderately repeatable from cycle to cycle, following approximately the same pathway.27-29 Occasionally, a spiraling pattern of functional reentry emanates from the same region for several cycles. Sometimes, the central core of these spiral waves meanders across the heart.8 At other times, new reentrant activation fronts are generated when one front interacts with another during its vulnerable period. Study of VF has suggested that there is a level of organization to the seemingly random patterns of wandering wavelets. Two competing hypotheses have been proposed to explain this organization. The “mother rotor” hypothesis was first proposed by Sir Thomas Lewis in 192530 and was recently revived by Jalife and colleagues.33 This hypothesis proposes that a single stationary reentrant circuit or mother rotor, located in the fastest-activating region of the heart, “drives” VF by

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

giving rise to activation fronts that propagate throughout the remainder of the myocardium.29 These wavefronts propagate away from this fast-activating region, encounter areas of unidirectional block, and break up into smaller, slower-moving waveforms that resemble Wiggers’ wandering wavelets. Experiments that have best demonstrated the mother rotor have been performed in small hearts, from guinea pigs or rabbits. Studies of larger hearts have shown areas of faster and slower activation across the heart, but the existence of a single reentrant rotor that drives fibrillation has not been clearly demonstrated.31,32 In contrast to the mother rotor hypothesis is the “restitution” hypothesis. Restitution properties of the heart have been recognized for many years. Restitution in the heart refers to the relationship between the duration of an action potential in a particular cell and the duration of the previous diastolic or resting interval. If the previous diastolic interval (DI) is short, the current action potential duration (APD) will also be short. If the previous DI is long, then the current APD will be long. For a regular rhythm, the preceding DI is constant and, therefore, so is the duration of the subsequent APD. The relationship between an APD and the previous DI is often described graphically as a plot of APDn versus DIn−1 (Fig. 2-2). The steepness of the restitution

61

curve is an important characteristic of this curve, especially at short DIs. If this slope is greater than 1, then, at a constant cycle length, a single perturbation in DI will cause the ensuing APDs and DIs to oscillate, with the oscillations progressively increasing until the site is refractory at the time of the next cycle, causing conduction block and VF initiation. And during VF, it is hypothesized that, when the slope of the restitution curve is greater than 1, oscillations in DI and APD increase until block occurs and wavefronts break up. Figure 2-2 shows an example of a restitution curve recorded from the right ventricle of a pig. The relationship between a DI and the subsequent APD is well defined during paced rhythm but less well defined during VF. Understanding how VF is maintained may help develop new therapies that will make fibrillation easier to stop. If VF is maintained by a mother rotor, then targeting of electrical therapy, either shocks or pacing, to the region that contains the dominant reentrant circuit may be successful in halting VF. If VF is started and maintained primarily due to the DI restitution properties of the heart, then drugs that decrease the slope of the restitution curve, especially at short coupling intervals, may be successful in halting VF. The Cellular Action Potential and the Excitable Gap During Fibrillation

Figure 2-2. Dynamic action potential duration APD60 restitution relationship during pacing and ventricular fibrillation (VF) in one animal. (APD60 is the action potential duration at 60% of return to resting membrane voltage.) Open circles represent data from pacing, and solid squares represent data throughout 60 seconds of VF. Data were recorded from the anterior right ventricle using a floating microelectrode. The heart was stimulated using decremental pacing at an initial pacing rate of 1 pulse every 450 msec. The stimulus-to-stimulus interval was progressively shortened until either the heart was refractory to the stimulation or VF was induced. Note that the open circles form an exponential relationship between the diastolic interval (DI) and APD60, whereas the relationship between DI and APD60 is not well defined during VF. (From Huang J, Zhou X, Smith WM, Ideker RE: Restitution properties during ventricular fibrillation in the in situ swine heart. Circulation 110:3161-3167, 2004.)

In the past few years, knowledge of the characteristics of the action potentials during fibrillation has increased greatly. This is a direct result of the introduction of techniques for recording action potentials in whole hearts, either in vivo or in perfused, isolated hearts.4-6,8,33 During fibrillation, the action potentials are altered; the APD is decreased, the action potential upstroke is slowed (decreased first-order derivative, dV/dt) and of decreased magnitude, the plateau phase is abbreviated, and DIs are abbreviated or absent (Fig. 2-3). During the first few seconds of VF or of atrial fibrillation, the activation rate is quite rapid; the mean cycle length of VF in patients undergoing defibrillator implantation was measured to be 213 ± 27 msec.6 DIs are rarely seen during early fibrillation, and the upstroke of most action potentials occurs before the transmembrane potential has returned to baseline from the previous action potential. The demonstration of an excitable gap in fibrillating atrial tissue34 and evidence of an excitable gap in fibrillating ventricular tissue35 suggest that there are periods late in the action potential in the fibrillating myocardium during which an electrical stimulus can capture a portion of the fibrillating myocardium. Knowing that there is an excitable gap suggests that there is an opportunity to stimulate the tissue just in front of a fibrillating wavefront, to cause wavefront block. As described in Chapter 1, the electrical activity of the heart is controlled ultimately by ion channels located in the cell membrane of the myocyte. It has been established that both the voltage-gated fast channels (sodium [Na+]) and slow channels (Na+ and calcium [Ca2+]) are active during the first few seconds of VF.5 The fast channel activity is indicated by the rapidity of

62

Section One: Basic Principles of Device Therapy

Figure 2-3. Recording taken during ventricular fibrillation in a human. Leads I, II, and III are body surface electrocardiograms. Note that there is no period of diastole between action potentials. MAP, right ventricular monophasic action potentials; ECM, local bipolar electrogram. (From Swartz JF, Jones JL, Fletcher RD: Characterization of ventricular fibrillation based on monophasic action potential morphology in the human heart. Circulation 87:1907-1914, 1993.)

the upstroke of the action potential (phase 0) during early fibrillation and its sensitivity to administration of the Na+ channel blocker, tetrodotoxin. As fibrillation proceeds, the upstrokes of the action potentials become increasingly slower, with a decreased dV/dtmax, but the action potentials remain sensitive to tetrodotoxin until 1 to 5 minutes after initiation of fibrillation. A transition then occurs in which the action potential upstrokes become insensitive to tetrodotoxin. This suggests that the propagation of the action potential is no longer mediated primarily by the fast voltage-gated Na+ channels and may be mediated by slow voltage-gated Ca2+ channel activity in the later stages of fibrillation.4 The activation complexes recorded from the ventricular myocardium remain active only as long as the coronary arteries are perfused with oxygenated blood, suggesting that ischemia may be responsible for the loss of fast channel activation during prolonged VF.25

Defibrillation Successful defibrillation can reflect either the immediate cessation of all activation fronts or the cessation of activation fronts after two to three cycles,10,36 followed by coordinated beating of the heart. Unsuccessful defibrillation can reflect a failure to inhibit the fibrillating activation fronts or the resumption of fibrillating activation fronts after their initial inhibition. As previously mentioned, application of a powerful electrical shock to the heart is the only reliable means of stopping fibrillation. Waveforms, Current Strength, and Distribution during Defibrillation The two most common waveform shapes used clinically are the monophasic and biphasic waveforms. In monophasic waveforms, the polarity of the shock is unchanged at each electrode for the entire duration of the electrical shock. In biphasic waveforms, the polar-

ity of the shock reverses at each electrode partway through the defibrillation waveform. Many studies, in both animals and humans, have shown that biphasic waveforms can defibrillate with less current and energy than monophasic waveforms, in both internal and transthoracic defibrillation configurations.37-40 Within each type, waveforms can be described as truncated exponential or damped sinusoidal shapes. ICDs use truncated exponential biphasic waveforms. Until recently, most external defibrillators used damped sinusoidal waveforms. Because of the inductor necessary to shape the damped sinusoidal monophasic waveform, these defibrillators tend to be large and heavy. More recently, smaller, lighter external defibrillators have been developed that use truncated exponential biphasic waveforms similar to those used in ICDs. Damped sinusoidal biphasic waveforms are used in external defibrillators in Russia; similar to truncated exponential biphasic waveforms, they have been shown to have an improved efficacy over monophasic waveforms.41,42 Not all biphasic waveforms are superior to monophasic waveforms, however. For example, if the second phase of the biphasic waveform becomes much longer than the first phase, then the energy required for defibrillation increases and can eventually rise to a level greater than the energy required to defibrillate with a monophasic waveform with duration equal to the first phase of the biphasic waveform.40,43,44 The optimum duration of the two phases of the biphasic waveform depends on the electrode impedance and the defibrillator capacitance.45-48 Several groups have shown that defibrillation efficacy for square waveforms follows a strength-duration relationship similar to that for cardiac stimulation49,50; as the waveform gets longer, the average current at the 50% success point (the current when one half of delivered shocks will succeed) becomes progressively less, approaching an asymptote called the rheobase.51 On the basis of this observation, several groups have suggested that cardiac defibrillation can be mathematically modeled using a resistor-capacitor (RC) network to represent the heart (Fig. 2-4).46,52-54 Empirically, it has been determined that the time constant for the parallel RC network is in the range of 2.5 to 5 msec.46,47,53 In one version of the model,53 a current waveform is applied to the RC network. The voltage across the network is then calculated for each time point during the defibrillation pulse. The relative efficacies of different waveform shapes and durations can be compared by determining the current strength that is necessary to make the voltage across the RC network reach a particular value, called the defibrillation threshold. Several observations can be made from this model. First, for square waves, as the waveform duration gets longer, the voltage across the network increases, approaching an asymptote or rheobase. For truncated exponential waveforms, however, the model voltage rises, reaches a peak, and then, if the waveform is long enough, begins to decrease (see Fig. 2-4). Therefore, the model predicts that monophasic exponential waveforms should be truncated at a time when the peak voltage across the RC network is reached. Current or

A

Leading edge current

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

Model response

15.0

B

5.0

0.0

0

5

10

15 20 Time (msec)

25

30

Leading edge current

C

10.0

Model response

22

D

16 12 6 2 4 8

0

5

10 Time (msec)

15

20

Figure 2-4. The response of a parallel resistor-capacitor network representation of the heart to monophasic and biphasic truncated exponential waveforms with a time constant of 7 msec. The parallel resistor-capacitor network has a time constant of 2.8 msec. A, The input monophasic waveforms. The leading-edge current of the input waveform was 10 A. The waveforms were truncated at 1, 2, 3, 4, 5, 6, 8, and 10 msec. B, The model response, V(t). Initially, as the waveform gets longer, V(t) increases until it reaches a maximum at about 4 msec, after which it begins to decrease. C, The input biphasic waveforms. The leading-edge current was 10 A. Phase 1 was truncated at 6 msec. Phase 2 was truncated after 1, 2, 3, 4, 5, 6, 7, and 8 msec. D, The model response does not change polarity until the phase 2 duration is longer than 2 msec. (From Walcott GP, Walker RG, Cates AW, et al: Choosing the optimal monophasic and biphasic waveforms for ventricular defibrillation. J Cardiovasc Electrophysiol 6:737-750, 1995.)

energy delivered after that point is wasted. In supporting this prediction, strength-duration relationships measured in both animals53 and humans55 do not approach an asymptote but rather reach a minimum and remain constant as the waveform gets longer. This minimum occurs over a range of waveform durations and does not extend indefinitely. Schuder and col-

63

leagues56 showed that, if the duration of a waveform gets too long, defibrillation efficacy decreases. Second, the model predicts that the heart acts as a low-pass filter.54 Therefore, waveforms that rise gradually should have better efficacy than waveforms that turn on immediately. This prediction has been shown to hold true for external defibrillation,57 internal atrial defibrillation,58 and internal ventricular defibrillation.59 Ascending ramps defibrillate with a greater efficacy than do descending ramps.59,60 Third, several groups have suggested that the optimal first phase of a biphasic waveform is the optimal monophasic waveform.45,48 If this is true, then what does the model predict as the “best” second phase of a biphasic waveform? Empirically, it appears that the role of the second phase is to return the model voltage response back to zero as quickly as possible, to maximize the increased efficacy of the biphasic waveform over that of the monophasic waveform with the same duration as phase 1 of the biphasic waveform. If the network voltage does not reach zero or if it overshoots zero, efficacy is lost.45,53 Swerdlow and colleagues47 showed in humans that the “best” second phase of a biphasic waveform is one that returns the model response close to zero. Together, these ideas allow the clinician to choose optimal capacitor sizes and phase durations for truncated exponential biphasic waveforms, the waveforms most commonly used in ICDs. The capacitor must be large enough to be able to raise the network voltage to its threshold value and still hold enough charge to drive the network voltage back to zero. For a 40-Ω interelectrode impedance and a network time constant of 2.8 msec, the minimum capacitor that can accomplish this has a capacitance of 75 microfarads (μF). The location of the defibrillation electrodes affects the magnitude of the shock necessary to defibrillate the heart. Typically, 200 to 360 J of energy are necessary for successful defibrillation with the defibrillation electrodes located on the body surface, during transthoracic defibrillation with a damped sinusoidal monophasic waveform. Less energy is required for a truncated exponential biphasic waveform.61 However, only about 4% to 20% of the current that is delivered to transthoracic defibrillation electrodes ever reaches the heart.62,63 Indeed, when the defibrillation electrodes are placed in the heart itself, usually only 20 to 34 J of energy are required, and the requirement may be as low as only a few joules when very large, contoured epicardial electrodes are used.40,64 The strength of the shock also varies for different locations on or in the heart; epicardial patches defibrillate with a lower shock energy than do transvenous electrode configurations.65 Although defibrillation efficacy is usually described by some measure of defibrillation shock strength (energy, voltage, or current), little insight into the mechanisms of defibrillation can be obtained from these measures. Knowing how the current (or voltage) of a defibrillation shock is distributed over the heart allows a much deeper understanding of how defibrillation occurs. Several studies have been performed that measure the potential gradient distribution throughout

64

Section One: Basic Principles of Device Therapy

the heart during a defibrillation shock.14,66,67 The potential gradient is a measure of the spatial variation of shock voltage across the heart. The potential gradient is measured in volts per centimeter of tissue. In a region with a high potential gradient, the difference in voltage between a given point and an area 1 cm distant from that point is high. Regions of low potential gradient have a measured voltage that is similar to that of nearby points. These studies show, for most electrode configurations, an uneven distribution, with areas of high potential gradient near the defibrillation electrodes and areas of low potential gradient in regions distant from the defibrillating electrodes. It has been hypothesized that a minimum potential gradient must be attained for successful defibrillation to occur and that this requirement is independent of the current applied or the electrode configuration.15,16 After a shock that fails to defibrillate VF, the site of earliest activation immediately after the shock can be mapped and related to the electric field that was produced by the shock. For shocks near the defibrillation threshold, the sites of earliest activation after a failed shock occur in the areas of lowest potential gradient.15,16 The minimum potential gradient required for defibrillation is lower for biphasic than for monophasic waveforms (4 versus 6 V/cm).16 A minimum potential gradient of 6 V/cm was required for successful defibrillation using a 10-msec truncated exponential monophasic waveform in the open-chest dog model.16 Similar findings were observed using a 14-msec truncated exponential monophasic waveform and multiple electrode configurations.15 In contrast, a minimum potential gradient of 4 V/cm was required for successful defibrillation using a truncated exponential biphasic waveform. Because higher shock strengths are required to induce a higher potential gradient, biphasic shocks successfully defibrillate with lower energy than monophasic shocks (i.e., a lower-voltage gradient is required). The requirement for a minimum potential gradient may reflect the need for a shock to prevent the generation of new activation fronts that can result in reinitiation of fibrillation.68 Examination of activation patterns after failed defibrillation for progressively larger shock strengths indicate that postshock activation occurs at numerous sites throughout the ventricle, and reentry is common when the shock strength is much lower than that needed for defibrillation (Fig. 2-5).69 At shock strengths just lower than those required for defibrillation, postshock activation arises in a limited number of myocardial regions. The activation fronts then propagate to activate other regions of the myocardium for a few cycles before reentry occurs, activation becomes disorganized, and fibrillation is reinitiated. Although postshock activation sites can still arise in regions of lowest potential gradient after a shock slightly greater than that required for defibrillation, the cycles of activation that originate from these sites are slower. These activations terminate after a few cycles without reinitiating fibrillation.68-70 So far, we have discussed how defibrillation can fail because a shock is of insufficient strength. What

Preshock

Postshock 100V

200V

600V

800V

A

B

C

D

Figure 2-5. Phase maps of a single rabbit heart showing the last cycle before (left panel) and the first cycle after (right panel) a shock of 100 V (A), 200 V (B), 600 V (C), and 800 V (D) that failed to defibrillate. The defibrillation threshold was 800 ± 200 V in this series. Colors represent phase, and the symbols + and − indicate phase singularities or centers of reentrant circuits of opposite direction. Phase singularities as a marker of reentry were observed during ventricular fibrillation just before the shock in all cases. A and B, Postshock phase singularities were observed after failed 100- and 200V shocks. Visual analysis of animations of the optical recordings indicated that many of the phase singularities represented reentrant activations occurring immediately after the shock, so that the postshock interval was 0 msec. C and D, No phase singularities were observed after the 600- and 800-V shocks. For the 600-V shock, activation propagated away in all directions from two early sites, one at the apex and the other at the lateral base of the left ventricle, both of which appeared after a postshock interval of 42 msec. For the 800-V shock, a single wavefront of activation appeared at the apex and propagated away in all directions in a focal pattern after a postshock interval of 72 msec. (From Chattipakorn N, Banville I, Gray RA, Ideker RE. Effects of shock strengths on ventricular defibrillation failure. Cardiovasc Res 61:39-44, 2004.)

happens if a defibrillation shock gets very large? At high shock strengths, the probability of defibrillation success begins to decrease again. It is believed that, at large strengths, defibrillation shocks can have detrimental effects on the heart. Increasing the shock strength to very high levels (>1000 V with transvenous electrodes) can result in activation fronts arising from regions of high potential gradient that reinduce VF.71 Cates and coworkers72 showed that, for both monophasic and biphasic shocks, increasing shock strength does not always improve the probability of successful defib-

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

rillation and may in fact increase the incidence of postshock arrhythmias. Chapman and associates73 showed in dogs that the time required for the heart to recover hemodynamically after a defibrillation episode was shorter for biphasic than for monophasic shocks. Further, they showed that hemodynamic recovery took longer after high-energy shocks than after low-energy shocks. Reddy and colleagues74 showed that transthoracic defibrillation with biphasic shocks resulted in less postshock electrocardiographic evidence of myocardial dysfunction (injury or ischemia), than did standard monophasic damped sinusoidal waveforms, and without compromise of defibrillation efficacy. One mechanism that has been implicated in the means by which shocks cause damage to the myocardium is electroporation, the formation of holes or pores in the cell membrane. Electroporation may occur in regions in which the shock potential gradient is high (>50 to 70 V/cm) and may even occur in regions where the potential gradient is much less than 50 V/cm.75 The very high voltage can result in disruption of the phospholipid membrane bilayer and in the formation of pores that permit the free influx and efflux of ions and micromolecules. Electroporation can cause the transmembrane potential to change temporarily to a value almost equal to that of the plateau of the action potential. At this transmembrane potential, the cell is paralyzed electrically, being both unresponsive and unable to conduct an action potential. Exposure of the myocardium to yet higher potential gradients, probably greater than 150 V/cm, results in arrhythmic beating, and at very high potential gradients, necrosis may occur.76 The shape of the waveform alters the strength of the shock at which these detrimental effects occur. Use of a 10-msec truncated exponential monophasic waveform for VF in dogs resulted in conduction block in regions where the potential gradient was greater than 64 ± 4 V/cm.77 Shocks that created even higher potential gradients in the myocardium (71 ± 6 V/cm) were required for conduction block when a 5-msec/5-msec truncated exponential biphasic shock was used. Adding a second phase to a monophasic waveform, thereby making it a biphasic waveform, reduced the damage sustained by cultured chick myocytes compared with that induced by the monophasic waveform alone.78 Therefore, biphasic waveforms are less apt to cause damage or dysfunction in high-gradient regions than monophasic waveforms. Models Proposed to Explain the Induction of Changes in the Transmembrane Potential throughout the Heart during a Defibrillation Shock A shock in the form of a square wave given across the defibrillation electrodes appears almost immediately as a square wave in the extracellular space of the heart. There is no significant distortion, because the extracellular space throughout the body is primarily resistive, with little reactive component. Phase delays and alterations of the appearance of the shock wave occur in

65

Extracellular space

10 ms

Transmembrane potential

A

Extracellular space

10 ms

Transmembrane potential

B Figure 2-6. The effect of a square-wave shock on the extracellular potential and on the transmembrane potential. The square-wave shock appears immediately as a relatively undistorted square wave in the extracellular space. It appears as an exponentially increasing change in the transmembrane potential. When a shock of a given polarity is delivered during the action potential plateau (A), the depolarization obtained is different in magnitude and time course from the hyperpolarization obtained with a shock of the opposite polarity (B). (From Walcott GP, Knisley SB, Zhou X, et al: On the mechanism of ventricular defibrillation. PACE 20:422-431, 1997.)

the transmembrane potential, however, owing to the capacitance and ion channels of the membrane of the myocyte.79 Consequently, a square wave shock can elicit an exponential change with time in the transmembrane potential (Fig. 2-6). The nonlinear behavior of the membrane caused by the ion channels also affects the outcome of reversing the polarity of the defibrillation shock. Reversing the polarity may reverse the sign of the change in the transmembrane potential in some regions of the myocardium, and the nonlinear behavior of the membrane can alter the magnitude and the time course of this change. As discussed previously, reversing the polarity of the shock may not reverse the sign of the change in the transmembrane potential in all regions of the heart; some areas may be hyperpolarized with both shock polarities.80 This may reflect the nonlinear behavior of the membrane ion channels. Several models have been formulated in an attempt to explain the mechanisms by which the defibrillation shock is distributed throughout the myocardium to effectively restore coordinated, effective action potentials. As yet, none of the models adequately describes all of the experimental findings regarding the changes in the action potential that occur during defibrillation. It is well established experimentally that changes occur many centimeters from the defibrillating shock electrodes. These changes in transmembrane potential can result in new action potentials or prolongation of the action potential as described previously.23,81 Direct excitation can be observed23 even at great distances from the electrode (>30 mm)82 or across the entire heart.83

Section One: Basic Principles of Device Therapy

The Sawtooth Model The one-dimensional cable model posits two lowresistance continuous spaces that conduct current from the shock, the intracellular and the extracellular spaces, separated by a high-resistance cell membrane. In the sawtooth model, the intracellular space is divided by a series of high-resistance barriers, the gap junctions. Because of these high-resistance barriers, current moving in the intracellular space is forced to exit into the extracellular space and reenter the cell on the other side of the barrier. Exit and reentry of the current from the intracellular domain results in hyperpolarization near the end of the cell closest to the anode and depolarization near the end of the cell closest to the cathode. A tracing of the changes in transmembrane potential along a fiber during the shock should, therefore, resemble the teeth of a saw, with each tooth corresponding to an individual cell (Fig. 2-7).20,86-88 Increases in the junctional resistances are predicted to increase the

71.0 Transmembrane potential (mV)

Although the one-dimensional cable model described in Chapter 1 adequately describes the generation of self-propagating action potentials close to an electrode as required for pacing, it fails to account for the far-field changes observed during defibrillation. During stimulation or defibrillation, this model predicts that the tissue near the anode should be hyperpolarized, whereas the tissue near the cathode should be depolarized.84 The magnitude of the hyperpolarization or depolarization decreases exponentially with the distance from the electrodes according to the membrane space constant (the distance at which the hyperpolarization or depolarization has decreased by 63%). For cardiac tissue, the space constant is only 0.5 to 1 mm.84,85 Therefore, the one-dimensional cable equations predict that tissue more than 10 space constants (about 1 cm) distant from the defibrillation electrodes should not directly undergo changes in transmembrane potential because of the shock field. That is, new action potentials should not arise by direct excitation at distances greater than 1 cm from the electrodes. This model fails to describe the experimentally observed global distribution of action potentials during defibrillation. Therefore, several additional mathematical formulations have been proposed, including the sawtooth model,20,86-88 the bidomain model,89 and the formation of secondary sources at barriers in the myocardium,19,90 to explain how a defibrillation shock affects the transmembrane potential a long distance away from the shocking electrodes. In the simplest formulation of these models, the extracellular and intracellular spaces are considered to be low-resistance media and the membrane to be a high-resistance medium in parallel with capacitance. The simple case models incorporate only passive myocardial properties. The models have been rendered more realistic by the addition of active components to represent the ion channels in the membrane, gap junctions, and membrane discontinuities.91,92 By convention, the current is defined as the flow of positive ions from the anode to the cathode.

35.0

0.0

35.0

71.0 0.19

0.1

0.0 Fiber axis (cm)

0.1

0.19

A 71.0 Vm0(x) and Vm1(x,y) (mV)

66

Vm1

Vm0

35.0

0.0

35.0

71.0 0.19

0.1

0.0 Fiber axis (cm)

0.1

0.19

B Figure 2-7. The transmembrane potential during a shock, according to the sawtooth model. A, The transmembrane potential shown is a summation of the membrane potential profile of the cable model and the periodicity arising from the periodic changes in intracellular resistance. The anode is to the left and the cathode is to the right in this one-dimensional model. The fiber is divided into 31 cells of equal length separated by junctions of high resistance. In the figure, the junctional resistance is shown much greater than is believed to occur in cardiac fibers, to allow the sawtooth pattern to be seen. B, The two parts of the summation shown in panel A are shown. Vm0, the transmembrane potential profile of the cable model; Vm1, the periodicity arising from the periodic changes in intracellular resistance. (Adapted from Krassowska W, Pilkington TC, Ideker RE: Periodic conductivity as a mechanism for cardiac stimulation and defibrillation. IEEE Trans Biomed Eng 34:555-560, 1987.)

magnitude of the potential changes at the ends of the cell.20 Although gap junctions are of low resistivity, they can present significant junctional resistance under certain conditions, such as hypoxia93,94 and calcium depletion.95 As the resistance of the gap junctions increases, it is predicted that a greater fraction of current passes preferentially across the cell membrane rather than along the cell. The sawtooth model adequately describes the requirement for a minimum potential gradient, because the magnitude of the hyperpolarization and polarization at the ends of the cells is directly proportional to the strength of the stimulus. It also adequately

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

describes the generation of action potentials at a distance from the electrodes and the differences in threshold stimuli between cathodal and anodal stimulation. Sawtooth changes in transmembrane potential have been observed in preparations of isolated cardiomyocytes96,97; however, such a pattern in isolated cells would be consistent with the cable model. This pattern has not been observed in a syncytium of cardiac cells.19,98,99 Although the resistivity of the gap junctions at the boundaries between the cells may not adequately explain the physiologic effects of defibrillation, the resistivity of other intracellular discontinuities and interruptions may well play a role. Most theories concerning the generation of action potentials and their propagation across the ventricle, such as the bidomain theory described later, consider the myocardium to be a uniform electrical continuum. This assumption does not take into account the discontinuities of the intracellular domain, where the myocardium is interrupted by barriers such as connective tissue septa, blood vessels, and scar tissue. As described previously for the sawtooth model, the intracellular current, on encountering such a barrier, must leave the intracellular space, cross the barrier, and reenter the intracellular domain on the other side. Depolarization should occur on one side of the barrier and hyperpolarization on the other side. Therefore, the barrier acts as a set of electrodes during the shock, becoming a secondary source of action potentials (Fig. 2-8). These secondary sources are important causes of depolarization and hyperpolarization throughout the myocardial tissues during a shock.19 The resistive barriers act in a manner similar to that described for the sawtooth model. In this case, however, the resistive barriers represent larger discontinuities, which tend to increase with age and cardiac hypertrophy.100 Computer simulations have shown that the cathodal stimulation delivered to the myocardium near the oval scar results in three distinct activation fronts: the primary activation front and secondary fronts at the distal and proximal edges, which are generated by the exit and reentry of current from the intracellular and extracellular spaces (see Fig. 2-8).101 Optical recording techniques have been used to directly record changes in transmembrane potentials throughout a monolayer of ventricular myocytes.19 Localized regions of depolarization and hyperpolarization were observed that coincided with discontinuities in the monolayer that resulted in slow conductance. Microscopic regions of depolarization and hyperpolarization have also been observed in isolated slabs of left ventricle, although their correlation with anatomic structures was not possible using optical mapping techniques due to the light scattering properties of myocardium.102 The significance of secondary sources was demonstrated in whole hearts by mapping of potentials and determination of shock thresholds before and after the generation of a transmural lesion in the myocardial walls of dogs.90 Generation of the lesion resulted in the development of a region of direct activation in the area of the lesion, in addition to the region of direct activation resulting

67

Figure 2-8. Secondary sources adjacent to a scar elicited by a single large pacing pulse, as seen in a computer model. The area to the right represents myocardium that contains a rectangular scar (stippled region). To the left is the blood pool with a pacing catheter in it. Notice that a pacing pulse depolarizes not only the tissue near the cathode but also that near the scar. (From Street AM: Effects of Connective Tissue Embedded in Viable Cardiac Tissue on Propagation and Pacing: Implications for Arrhythmias. Durham, NC, Duke University, Department of Biomedical Engineering, 1996, p 134.)

from the stimulating electrode observed before the lesion. Furthermore, the strength of the shock required to cause direct activation in the area of the lesion was less than one half of that required before generation of the lesion (Fig. 2-9). The effects of secondary sources obviously have major implications for the probability of successful defibrillation at different shock strengths in individual patients, particularly elderly patients, as well as the potential for reentry. Furthermore, the size and placement of operative lesions may play significant roles in the success of subsequent defibrillation. The Bidomain Model The bidomain model is an extension of the onedimensional cable model into two or three dimensions. That is, the extracellular and intracellular spaces are represented as single, continuous domains that extend in two or three dimensions and are separated by the highly resistive cell membrane (Fig. 2-10).89 If the conductivities of the intracellular and extracellular spaces are constant in all directions, then the model collapses to the one-dimensional cable model. Anisotropy refers to the manner in which conductivities change with the direction of myocardial fiber orientation. Clerc103 showed that conductivity is higher in the direction parallel to the long axis of myocardial fibers (longitudinal) than in the direction perpendicular to the fibers (transverse) for both the intracellular and extracellular spaces. If conductivities change with direction but change the same for the intracellular and extracellular spaces, the bidomain model collapses to the onedimensional cable model.

68

Section One: Basic Principles of Device Therapy

Figure 2-9. Isochronal activation maps obtained after cathodal stimulation before and after creation of a transmural incision that caused secondary sources adjacent to the lesion. Isochrones are drawn at 5-msec intervals, timed from the onset of the S1 or S2 stimulus. Arrows represent direction of activation. Darkened regions represent areas directly activated by the stimulus. Black vertical bars represent the approximate location of the transmural incision. A, S1 stimulus delivered before incision. B, 75-mA S2 stimulus delivered before incision. C, 250-mA S2 stimulus delivered before incision. D, S1 stimulus delivered after incision. E, 75-mA S2 stimulus delivered after incision. F, Orientation of the long axis of myocardial fibers. (From White JB, Walcott GP, Pollard AE, et al: Myocardial discontinuities: A substrate for producing virtual electrodes to increase directly excited areas of the myocardium by shocks. Circulation 97:1738-1745, 1998.)

rey rex

riy

rix y

x Figure 2-10. A circuit diagram of a two-dimensional bidomain model. The top of the resistor network represents the extracellular space, and the bottom of the network represents the intracellular space. The symbol reDx represents the extracellular resistivity in the x direction; reDY represents the extracellular resistivity in the y direction; riDx and riDy represent the intracellular resistivities in the x and y directions. The rectangles represent the cell membrane. For a passive model, the rectangle would be replaced by a parallel resistor-capacitor network. For an active model, the rectangle would be replaced by a membrane ion model.

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

Studies have shown that the anisotropy ratio is about 3:1 in the extracellular space and 10:1 in the intracellular space. When anisotropy ratios are used, the bidomain model begins to give new insights into how shocks change the transmembrane potential. Similar to the one-dimensional cable model, the bidomain model predicts that hyperpolarization occurs in tissue that lies under the extracellular anodal electrode. Likewise, depolarization occurs in tissue under the extracellular cathodal electrode. Unlike the one-dimensional cable model, the bidomain model also predicts that depolarization occurs along the long axis of the myocardial fibers at distances just a few millimeters from the anode. A similar effect is predicted to occur at the cathode, with hyperpolarization at distances of a few millimeters.104,105 Therefore, the effect on the transmembrane potential near the shocking electrode is predicted to be much more complicated by the bidomain model than by the one-dimensional cable model. The power of the bidomain model, however, is that it hypothesizes that there should be changes in the transmembrane potential, either hyperpolarization or depolarization, across the entire heart. In this model, the change in transmembrane potential elicited by the

69

shock depends on the distribution of intracellular and extracellular current, which is affected by the change in potential gradient with distance, the distance from the electrode, and the orientation of the myocardial fibers. Experimental studies have shown that there is a complex pattern of transmembrane potential changes during the delivery of a defibrillation shock, similar to those predicted by the bidomain model.80,106,107 The transmembrane potential changes that occur during the delivery of a defibrillation shock can lead to the initiation of reentry and subsequent reinitiation of fibrillation after the shock. Reentrant circuits can be described by the mathematical concept of a phase singularity.108 Phase can be used to describe the cardiac action potential, with 0 phase assigned to the upstroke of the action potential and 2Π phase assigned to the end of the action potential. A reentrant circuit can be thought of as a circle of phase starting at 0 (excitation) and continuing to 2Π (recovery). The reentrant circuit moves around a central point, called a phase singularity. Efimov and colleagues106 showed that a defibrillation shock can impose changes on the transmembrane potential extending from 0 phase through 2Π phase (Fig. 2-11). Thus, a reentrant circuit is generated and

Figure 2-11. Creation of a shock-induced phase singularity. The upper left panel shows the change in transmembrane potential at the end of a +100/−200 V biphasic shock (i.e., at the 15th msec of a 16-msec shock), which resulted in a single extra beat. The scale is shown in millivolts, calibrated to a control 100-mV action potential. The point of phase singularity is indicated by the black circle. The upper middle panel shows a 5-msec isochronal map that depicts the initiation of the postshock spread of activation. The map starts at the onset of the 8-msec second phase of the shock (polarity reversal). The lower left and lower right panels show optical recordings from several recording sites used to reconstruct the activation maps. The upper right panel shows a continuation of the reentrant activation shown in the middle panel. Reentrant activity self-terminates after encountering refractory tissue in the lower right corner of the field of view (see traces in lower right panel). (From Efimov IR, Cheng Y, Van Wagoner DR, et al: Virtual electrode-induced phase singularity: A basic mechanism of defibrillation failure. Circ Res 82:918-925, 1998.)

70

Section One: Basic Principles of Device Therapy

fibrillation is induced. These same authors suggested that induced reentrant circuits may be one way that defibrillation shocks can fail. The secondary source and bidomain models may not be mutually exclusive; rather, both may contribute to the changes in transmembrane potential. The exact mechanism by which an electrical pulse results in defibrillation remains incompletely understood at the level of the cell membrane and the ion channels. As the transmembrane potential attains values closer to the typical resting transmembrane potential than the usual minimum of −65 mV observed in fibrillating myocytes, this may allow the voltage-gated Na+ channels to recover sufficiently and the myocytes to regain full excitability. The Effect of the Defibrillating Shock Field on the Cellular Action Potential The final common pathway of changes in the transmembrane potential caused by a defibrillation shock involves affects on the shape and duration of the cellular action potential. The shock can have one of three effects on the myocardium, depending on the local strength of the shock and its timing with respect to the

local action potential. If the shock is delivered during the early plateau, there will be little or no prolongation of the action potential. If the shock is strong enough and is delivered relatively late during the action potential, it will initiate a new action potential. A shock that is strong enough but is delivered during early phase 3 of the action potential will modify and prolong an ongoing action potential without initiating an entirely new action potential (Fig. 2-12). A defibrillating shock must do two things to defibrillate the heart successfully. First, it must stop most or all activation wavefronts on the heart. Second, it must not reinitiate fibrillation. The extension of refractoriness hypothesis helps to explain how a shock can stop fibrillation. A shock can prolong the refractory period of an action potential without triggering a new action potential if it is of sufficient strength and is delivered at an appropriate interval with respect to the upstroke.23,109 If the first activation front that forms after a defibrillation shock encounters tissue with an extended refractory period, the front will be stopped because it cannot propagate into the region of refractory tissue. Also, a defibrillation shock must not restart fibrillation. If only part of the front encounters tissue with an

Figure 2-12. Transmembrane recordings from a guinea pig papillary muscle showing an allor-none response to a weak field stimulus and action potential prolongation in response to a larger stimulus. A, Recordings that illustrate the response to an S2 stimulus of 1.6 V/cm oriented along the fibers. The S1-S2 stimulus interval for each response (milliseconds) is indicated to the right of the recordings. The responses are markedly different even though the change in S2 timing was only 3 msec. An S1-S2 interval of 222 msec caused almost no response, whereas an interval of 225 msec produced a new action potential. B, A range of action potential extensions produced by an S2 stimulus generating a potential gradient of 8.4 V/cm oriented along the long axis of the myofibers. The recordings were obtained from the same cell as in panel A. The action potential recordings, obtained from one cellular impairment, are aligned with the S2 time. An S1 stimulus was applied 3 msec before phase 0 of each recording. The longest and shortest S1-S2 intervals tested (230 and 90 msec, respectively) are indicated beneath their respective phase 0 depolarizations. The S1-S2 interval for each response is indicated to the right. (From Knisley SB, Smith WM, Ideker RE: Effect of field stimulation on cellular repolarization in rabbit myocardium: Implications for reentry induction. Circ Res 70:707-715, 1992.)

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms

extended refractory period, only part of the front will be halted. The rest of the activation front will propagate forward and will eventually move into the area that could not be stimulated or that would not allow propagation (unidirectional block). This process of stimulating some tissue and creating unidirectional block in adjacent regions creates a reentrant circuit that eventually breaks down into fibrillation. The critical point is that point at which a critical shock strength intersects a critical level of refractoriness, leading to the formation of a reentrant circuit.110-112 To understand better how reentrant circuits and critical points are formed after a defibrillation shock, studies have examined the behavior of the heart after delivery of a shock during a paced ventricular rhythm.111 Shocks were used to initiate VF during the vulnerable period of the paced rhythm. A large premature S2 stimulus was delivered through a long, narrow electrode oriented perpendicular to an activation front arising from an S1 stimulus (Fig. 2-13). S2 shocks were given to scan the vulnerable period after the last S1 stimulus. At an appropriate S1-to-S2 coupling interval, a reentrant circuit formed and continued for several cycles before breaking down into fibrillation. The initial postshock activation front circled a point at which a shock potential gradient field of 5 to 6 V/cm for a 10-msec monophasic shock intersected tissue that

Figure 2-13. Initiation of reentry and ventricular fibrillation after orthogonal interaction of myocardial refractoriness and the potential gradient field created by a large stimulus. A, Distribution of activation times during the last S1 beat (solid lines) and recovery times in relation to a local 2-mA stimulus (dashed lines) in milliseconds after this activation. B, S2 stimulus field (V/cm). C, Initial activation pattern just after the S2 stimulus. The hatched region is believed to be directly excited by the S2 stimulus field. (From Frazier DW, Wolf PD, Wharton JM, et al: Stimulus-induced critical point: Mechanism for electrical initiation of reentry in normal canine myocardium. J Clin Invest 83:1039-1052, 1989.)

71

was just passing out of its refractory period to a 2-mA local stimulus. This intersection formed a critical point. The intersection of a particular potential gradient level with a particular refractory state divides the tissue into four regions centered on the critical point (see Fig. 2-13C). The region to the left of the critical point was still in its refractory period and was not directly excited by the shock. The region to the right of the critical point had recovered enough to be directly excited by the S2 stimulus. Above the critical point, the shock had no effect on the refractory period of the tissue, whereas below the critical point, the shock prolonged the refractory period of the tissue. Therefore, an excitation wavefront propagated from the upper right quadrant of the mapped region across the top half of the plaque. Because of the prolongation of refractoriness in the tissue below the critical point, the excitation wavefront was unable to propagate across the bottom half of the plaque directly. As this tissue recovered, the excitation wavefront from the top half of the plaque entered the area at the bottom half and re-excited the tissue, creating a reentrant circuit around the critical point. Recent studies provide some insight into why the activation pattern observed in Figure 2-13 occurs. When the myocardium is stimulated with a shock field whose potential gradient is less than the critical value for that waveform, an all-or-none response is observed (see Fig. 2-11). If the stimulus is applied with a coupling interval greater than the refractory period, a new action potential is generated. If the coupling interval is shorter than the refractory period, almost no response is seen. When the shock field strength is greater than a critical value, a whole gradation of responses is observed, depending on the coupling interval. As the coupling interval is made shorter, a smaller response occurs. However, even these smaller responses prolong refractoriness. It is this prolongation of refractoriness adjacent to directly stimulated myocardium that leads to unidirectional block and ultimately to the formation of critical points and functional reentrant circuits. If fibrillation is initiated by the reentrant pathway formed whenever a critical point is created within the myocardium, then, to be successful, the strength of a defibrillation shock should be greater than that at which no critical points are formed. Ideker and associates113 examined the potential gradient and degree of refractoriness at the critical point for a series of monophasic and biphasic waveforms and compared these values with the defibrillation threshold. Monophasic waveforms lasting 1, 2, 3, 8, and 16 msec were delivered, as were biphasic waveforms in which both phases were of equal duration, the two phases totaling 2, 4, 8, and 16 msec. The defibrillation threshold decreased as the degree of refractoriness at the critical point decreased. The waveform that induced a critical point located where both the potential gradient and the degree of refractoriness were lowest (i.e., the 4/4-msec biphasic waveform) had the lowest defibrillation threshold. This observation may explain why some waveforms defibrillate at lower voltages than others. The mechanisms for the formation of critical points described here may be too simplistic in light of the

72

Section One: Basic Principles of Device Therapy

complex ways in which a shock can affect the transmembrane potential. Furthermore, the requirement that no critical point be formed for a shock to succeed may be too stringent. Several studies have shown that a defibrillation shock can still be successful even if it creates rapid postshock activation for one or two cycles, suggesting that it may not be necessary to prevent the creation of critical points and reentry. Rather, it is only necessary that the reentrant circuits die out in one or two cycles before secondary reentry occurs. Further research is necessary to understand these phenomena more completely.

Acknowledgment This work was supported in part by National Institutes of Health grant HL-42760. REFERENCES 1. Prevost JL, Battelli F: Sur quelques effets des décharges électriques sur le coeur des Mammifères. CRSAS 129:1267-1268, 1899. 2. Beck CS, Pritchard WH, Feil HS: Ventricular fibrillation of long duration abolished by electric shock. JAMA 135:985-986, 1947. 3. Zoll PM, Linenthal AJ, Gibson W, et al: Termination of ventricular fibrillation in man by externally applied electric countershock. N Engl J Med 254:727-732, 1956. 4. Akiyama T: Intracellular recording of in situ ventricular cells during ventricular fibrillation. Am J Physiol 240:H465-H471, 1981. 5. Zhou X, Guse P, Wolf PD, et al: Existence of both fast and slow channel activity during the early stage of ventricular fibrillation. Circ Res 70:773-786, 1992. 6. Swartz JF, Jones JL, Fletcher RD: Characterization of ventricular fibrillation based on monophasic action potential morphology in the human heart. Circulation 87:1907-1914, 1993. 7. Witkowski FX, Penkoske PS, Kavanagh KM: Activation patterns during ventricular fibrillation. In Zipes DP, Jalife J (eds): Cardiac Electrophysiology: From Cell to Bedside, 2nd ed. Philadelphia: WB Saunders, 1995, pp 539-544. 8. Gray RA, Jalife J, Panfilov AV, et al: Mechanisms of cardiac fibrillation: Drifting rotors as a mechanism of cardiac fibrillation. Science 270:1222-1225, 1995. 9. Wiggers CJ: The physiologic basis for cardiac resuscitation from ventricular fibrillation: Method for serial defibrillation. Am Heart J 20:413-422, 1940. 10. Mower MM, Mirowski M, Spear JF, Moore EN: Patterns of ventricular activity during catheter defibrillation. Circulation 49:858-861, 1974. 11. Zipes DP, Fischer J, King RM, et al: Termination of ventricular fibrillation in dogs by depolarizing a critical amount of myocardium. Am J Cardiol 36:37-44, 1975. 12. Chen P-S, Shibata N, Wolf PD, et al: Epicardial activation during successful and unsuccessful ventricular defibrillation in open chest dogs. Cardiovasc Rev Rep 7:625-648, 1986. 13. Gray RA, Ayers G, Jalife J: Video imaging of atrial defibrillation in the sheep heart. Circulation 95:1038-1047, 1997. 14. Chen P-S, Wolf PD, Claydon FJ III, et al: The potential gradient field created by epicardial defibrillation electrodes in dogs. Circulation 74:626-636, 1986. 15. Wharton JM, Wolf PD, Smith WM, et al: Cardiac potential and potential gradient fields generated by single, combined, and sequential shocks during ventricular defibrillation. Circulation 85:1510-1523, 1992.

16. Zhou X, Daubert JP, Wolf PD, et al: Epicardial mapping of ventricular defibrillation with monophasic and biphasic shocks in dogs. Circ Res 72:145-160, 1993. 17. Henriquez CS: Simulating the electrical behavior of cardiac muscle using the bidomain model. Crit Rev Biomed Eng 21:177, 1993. 18. Fishler MG, Sobie EA, Thakor NV, Tung L: Mechanisms of cardiac cell excitation with premature monophasic and biphasic field stimuli: A model study. Biophys J 70:1347-1362, 1996. 19. Gillis AM, Fast VG, Rohr S, Kléber AG: Spatial changes in transmembrane potential during extracellular electrical shocks in cultured monolayers of neonatal rat ventricular myocytes. Circ Res 79:676-690, 1996. 20. Plonsey R, Barr RC: Inclusion of junction elements in a linear cardiac model through secondary sources: Application to defibrillation. Med Biol Eng Comput 24:137-144, 1986. 21. Fast VG, Rohr S, Gillis AM, Kléber AG: Activation of cardiac tissue by extracellular electrical shocks: Formation of “secondary sources” at intercellular clefts in monolayers of cultured myocytes. Circ Res 82:375-385, 1998. 22. Moe GK, Abildskov JA, Han J: Factors responsible for the initiation and maintenance of ventricular fibrillation. In Surawicz B, Pellegrino ED (eds): Sudden Cardiac Death. New York: Grune & Stratton, 1964. 23. Dillon SM, Mehra R: Prolongation of ventricular refractoriness by defibrillation shocks may be due to additional depolarization of the action potential. J Cardiovasc Electrophysiol 3:442456, 1992. 24. Wiggers CJ: Studies of ventricular fibrillation caused by electric shock. Cinematographic and electrocardiographic observations of the natural process in the dog’s heart: Its inhibition by potassium and the revival of coordinated beats by calcium. Am Heart J 5:351-365, 1930. 25. Worley SJ, Swain JL, Colavita PG, et al: Development of an endocardial-epicardial gradient of activation rate during electrically induced, sustained ventricular fibrillation in the dog. Am J Cardiol 55:813-820, 1985. 26. Opthof T, Ramdat Misier AR, Coronel R, et al: Dispersion of refractoriness in canine ventricular myocardium: Effects of sympathetic stimulation. Circ Res 68:1204-1215, 1991. 27. Ideker RE, Klein GJ, Harrison L, et al: Epicardial mapping of the initiation of ventricular fibrillation induced by reperfusion following acute ischemia. Circulation 58:II-64, 1978. 28. Rogers J, Usui M, KenKnight B, et al: Recurrent wavefront morphologies: A method for quantifying the complexity of epicardial activation patterns. Ann Biomed Eng 25:761-768, 1997. 29. Samie FH, Berenfeld O, Anumonwo J, et al: Rectification of the background potassium current: A determinant of rotor dynamics in ventricular fibrillation. Circ Res 89:1216-1223, 2001. 30. Lewis T: The Mechanism and Registration of the Heart Beat. London, Shaw and Sons, 1925. 31. Newton JC, Evans FG, Chattipakorn N, et al: Peak frequency distribution across the whole fibrillating heart. PACE 23:617, 2000. 32. Newton JC, Ideker RE: Estimated global transmural distribution of activation rate and conduction block during porcine and canine ventricular fibrillation. Circ Res 94:836-842, 2004. 33. Gray RA, Jalife J: Self-organized drifting spiral waves as a mechanism for atrial fibrillation. Circulation 94:I-94, 1996. 34. Allessie M, Kirchhof C, Scheffer GJ, et al: Regional control of atrial fibrillation by rapid pacing in concious dogs. Circulation 84:1689-1697, 1991. 35. Ken-Knight BH, Bayly PV, Gerstle RJ, et al: Regional capture of fibrillating ventricular myocardium: Evidence of an excitable gap. Circ Res 77:849-855, 1995. 36. Witkowski FX, Penkoske PA, Plonsey R: Mechanism of cardiac defibrillation in open-chest dogs with unipolar DC-coupled simultaneous activation and shock potential recordings. Circulation 82:244-260, 1990.

Chapter 2: Principles of Defibrillation: From Cellular Physiology to Fields and Waveforms 37. Bardy GH, Ivey TD, Allen MD, et al: A prospective randomized evaluation of biphasic versus monophasic waveform pulses on defibrillation efficacy in humans. J Am Coll Cardiol 14:728-733, 1989. 38. Block M, Hammel D, Böcker D, et al: A prospective randomized cross-over comparison on mono- and biphasic defibrillation using nonthoracotomy lead configurations in humans. J Cardiovasc Electrophysiol 5:581-590, 1994. 39. Chapman PD, Vetter JW, Souza JJ, et al: Comparison of monophasic with single and dual capacitor biphasic waveforms for nonthoracotomy canine internal defibrillation. J Am Coll Cardiol 14:242-245, 1989. 40. Dixon EG, Tang ASL, Wolf PD, et al: Improved defibrillation thresholds with large contoured epicardial electrodes and biphasic waveforms. Circulation 76:1176-1184, 1987. 41. Gurvich NL, Markarychev VA: Defibrillation of the heart with biphasic electrical impulses. Kardiologiia 7:109-112, 1967. 42. Walcott GP, Melnick SB, Chapman FW, et al: Comparison of monophasic and biphasic waveforms for external defibrillation in an animal model of cardiac arrest and resuscitation. J Am Coll Cardiol 25:405A, 1995. 43. Tang ASL, Yabe S, Wharton JM, et al: Ventricular defibrillation using biphasic waveforms: The importance of phasic duration. J Am Coll Cardiol 13:207-214, 1989. 44. Feeser SA, Tang ASL, Kavanagh KM, et al: Strength-duration and probability of success curves for defibrillation with biphasic waveforms. Circulation 82:2128-2141, 1990. 45. Kroll MW: A minimal model of the single capacitor biphasic defibrillation waveform. PACE 17:1782-1792, 1994. 46. Kroll MW: A minimal model of the monophasic defibrillation pulse. PACE 16:769-777, 1993. 47. Swerdlow CD, Fan W, Brewer JE: Charge-burping theory correctly predicts optimal ratios of phase duration for biphasic defibrillation waveforms. Circulation 94:2278-2284, 1996. 48. Walcott GP, Walker RG, Krassowska W, et al: Choosing the optimum monophasic and biphasic waveforms for defibrillation. PACE 17:789, 1994. 49. Blair HA: On the intensity-time relations for stimulation by electric currents: II. JGENPH 15:731-755, 1932. 50. Lapicque L: L’Excitabilite en Fonction du Temps. Paris, Libraire J. Gilbert, 1926. 51. Mouchawar GA, Geddes LA, Bourland JD, Pearce JA: Ability of the Lapicque and Blair strength-duration curves to fit experimentally obtained data from the dog heart. IEEE Trans Biomed Eng 36:971-974, 1989. 52. Irnich W: The fundamental law of electrostimulation and its application to defibrillation. PACE 13:1433-1447, 1990. 53. Walcott GP, Walker RG, Cates AW, et al: Choosing the optimal monophasic and biphasic waveforms for ventricular defibrillation. J Cardiovasc Electrophysiol 6:737-750, 1995. 54. Sweeney RJ, Gill RM, Jones JL, Reid PR: Defibrillation using a high-frequency series of monophasic rectangular pulses: Observations and model predictions. J Cardiovasc Electrophysiol 7:134-143, 1996. 55. Gold MR, Shorofsky SR: Strength-duration relationship for human transvenous defibrillation. Circulation 96:3517-3520, 1997. 56. Schuder JC, Stoeckle H, West JA, Keskar PY: Transthoracic ventricular defibrillation in the dog with truncated and untruncated exponential stimuli. IEEE Trans Biomed Eng 18:410-415, 1971. 57. Walcott GP, Melnick SB, Chapman FW, et al: Comparison of damped sinusoidal and truncated exponential waveforms for external defibrillation. J Am Coll Cardiol 27:237A, 1996. 58. Harbinson MT, Allen JD, Imam Z, et al: Rounded biphasic waveform reduces energy requirements for transvenous catheter cardioversion of atrial fibrillation and flutter. PACE 20:226-229, 1997.

73

59. Hillsley RE, Walker RG, Swanson DK, et al: Is the second phase of a biphasic defibrillation waveform the defibrillating phase? PACE 16:1401-1411, 1993. 60. Schuder JC, Rahmoeller GA, Stoeckle H: Transthoracic ventricular defibrillation with triangular and trapezoidal waveforms. Circ Res 19:689-694, 1966. 61. Bardy GH, Marchlinski FE, Sharma AD, et al: Multicenter comparison of truncated biphasic shocks and standard damped sine wave monophasic shocks for transthoracic ventricular defibrillation. Transthoracic Investigators. Circulation 94:2507-2514, 1996. 62. Camacho MA, Lehr JL, Eisenberg SR: A three-dimensional finite element model of human transthoracic defibrillation: Paddle placement and size. IEEE Transact Biomed Eng 42:572-578, 1995. 63. Deale OC, Lerman BB: Intrathoracic current flow during transthoracic defibrillation in dogs: Transcardiac current fraction. Circ Res 67:1405-1419, 1990. 64. Karlon WJ, Eisenberg SR, Lehr JL: Effects of paddle placement and size on defibrillation current distribution: A three-dimensional finite element model. IEEE Trans Biomed Eng 40:246-255, 1993. 65. Block M, Hammel D, Isburch F, et al: Results and realistic expectations with transvenous lead systems. PACE 15:665-670, 1992. 66. Tang ASL, Wolf PD, Claydon FJ III, et al: Measurement of defibrillation shock potential distributions and activation sequences of the heart in three-dimensions. Proc IEEE 76:1176-1186, 1988. 67. Tang ASL, Wolf PD, Afework Y, et al: Three-dimensional potential gradient fields generated by intracardiac catheter and cutaneous patch electrodes. Circulation 85:1857-1864, 1992. 68. Chen P-S, Wolf PD, Melnick SD, et al: Comparison of activation during ventricular fibrillation and following unsuccessful defibrillation shocks in open chest dogs. Circ Res 66:1544-1560, 1990. 69. Chattipakorn N, Banville I, Gray RA, Ideker RE: Effects of shock strengths on ventricular defibrillation failure. Cardiovasc Res 61:39-44, 2004. 70. Shibata N, Chen P-S, Dixon EG, et al: Epicardial activation following unsuccessful defibrillation shocks in dogs. Am J Physiol 255:H902-H909, 1988. 71. Walker RG, Walcott GP, Smith WM, Ideker RE: Sites of earliest activation following transvenous defibrillation. Circulation 90:I447, 1994. 72. Cates AW, Wolf PD, Hillsley RE, et al: The probability of defibrillation success and the incidence of postshock arrhythmia as a function of shock strength. PACE 17:1208-1217, 1994. 73. Chapman FW, El-Abbady TZ, Walcott GP, et al: Dysfunction following transthoracic defibrillation shocks in dogs. PACE 20:1128, 1997. 74. Reddy RK, Gleva MJ, Gliner BE, et al: Biphasic transthoracic defibrillation causes fewer ECG ST-segment changes after shock. Ann Emerg Med 30:127-134, 1997. 75. DeBruin KA, Krassowska W: Electroporation and shock-induced transmembrane potential in a cardiac fiber during defibrillation strength shocks. Ann Biomed Eng 26:584-596, 1998. 76. Schuder JC, Gold JH, Stoeckle H, et al: Transthoracic ventricular defibrillation in the 100 kg calf with symmetrical one-cycle bidirectional rectangular wave stimuli. IEEE Trans Biomed Eng 30:415-422, 1983. 77. Yabe S, Smith WM, Daubert JP, et al: Conduction disturbances caused by high current density electric fields. Circ Res 66:11901203, 1990. 78. Jones JL, Jones RE: Decreased defibrillator-induced dysfunction with biphasic rectangular waveforms. An J Physiol 247:H792H796, 1984. 79. Walcott GP, Knisley SB, Zhou X, et al: On the mechanism of ventricular defibrillation. PACE 20:422-431, 1997.

74

Section One: Basic Principles of Device Therapy

80. Clark DM, Rogers JM, Ideker RE, Knisley SB: Intracardiac defibrillation-strength shocks produce large regions of hyperpolarization and depolarization. J Am Coll Cardiol 27:147A, 1996. 81. Zhou X, Wolf PD, Rollins DL, et al: Effects of monophasic and biphasic shocks on action potentials during ventricular fibrillation in dogs. Circ Res 73:325-334, 1993. 82. Daubert JP, Frazier DW, Wolf PD, et al: Response of relatively refractory canine myocardium to monophasic and biphasic shocks. Circulation 84:2522-2538, 1991. 83. Colavita PG, Wolf PD, Smith WM, et al: Determination of effects of internal countershock by direct cardiac recordings during normal rhythm. Am J Physiol 250:H736-H740, 1986. 84. Weidmann S: Electrical constants of trabecular muscle from mammalian heart. J Physiol 210:1041-1054, 1970. 85. Kléber AG, Riegger CB: Electrical constants of arterially perfused rabbit papillary muscle. J Physiol 385:307-324, 1987. 86. Plonsey R, Barr RC: Effect of microscopic and macroscopic discontinuities on the response of cardiac tissue to defibrillating (stimulating) currents. Med Biol Eng Comput 24:130-136, 1986. 87. Krassowska W, Frazier DW, Pilkington TC, Ideker RE: Potential distribution in three-dimensional periodic myocardium: Part II. Application to extracellular stimulation. IEEE Trans Biomed Eng 37:267-284, 1990. 88. Krassowska W, Pilkington TC, Ideker RE: Potential distribution in three-dimensional periodic myocardium: Part I. Solution with two-scale asymptotic analysis. IEEE Trans Biomed Eng 37:252266, 1990. 89. Tung L: A Bidomain Nodel for Describing Ischemic Myocardial DC Potentials. Cambridge, MA, Massechusetts Institute of Technology, 1978. 90. White JB, Walcott GP, Pollard AE, Ideker RE: Myocardial discontinuities: A substrate for producing virtual electrodes to increase directly excited areas of the myocardium by shocks. Circulation 97:1738-1745, 1998. 91. Trayanova N: Discrete versus syncytial tissue behavior in a model of cardiac stimulation. I: Mathematical formulation. IEEE Trans Biomed Eng 43:1129-1140, 1996. 92. Trayanova N: Discrete versus syncytial tissue behavior in a model of cardiac stimulation. II: Results of simulation. IEEE Trans Biomed Eng 43:1141-1150, 1996. 93. Kieval RS, Spear JF, Moore EN: Gap junctional conductance in ventricular myocyte pairs isolated from postischemic rabbit myocardium. Circ Res 71:127-136, 1992. 94. Shaw RM, Rudy Y: Electrophysiologic effects of acute myocardial ischemia: A mechanistic investigation of action potential conduction and conduction failure. Circ Res 80:124-138, 1997. 95. Shaw RM, Rudy Y: Ionic mechanisms of propagation in cardiac tissue: Roles of the sodium and L-type calcium currents during reduced excitability and decreased gap junction coupling. Circ Res 81:727-741, 1997. 96. Knisley SB, Blitchington TF, Hill BC, et al: Optical measurements of transmembrane potential changes during electric field stimulation of ventricular cells. Circ Res 72:255-270, 1993.

97. Windisch H, Ahammer H, Schaffer P, et al: Optical multisite monitoring of cell excitation phenomenon in isolated cardiomyocytes. Pflugers Arch 430:508-518, 1995. 98. Zhou X, Ideker RE, Blitchington TF, et al: Optical transmembrane potential measurements during defibrillation-strength shocks in perfused rabbit hearts. Circ Res 77:593-602, 1995. 99. Wikswo JP Jr, Lin S-F, Abbas RA: Virtual electrodes in cardiac tissue: A common mechanism for anodal and cathodal stimulation. Biophys J 69:2195-2210, 1995. 100. Sommer JR, Scherer B: Geometry of cell and bundle appositions in cardiac muscle: Light microscopy. Am J Physiol 248:H792H803, 1985. 101. Street AM, Plonsey R: Activation fronts elicited remote to the pacing site due to the presence of scar tissue. In Proceedings of the 18th Annual International Conference, IEEE Engineering Medical Biological Society. Amsterdam, The Netherlands, Institute of Electrical and Electronics Engineers. Available on CDROM (Piscataway, NJ, 1996, p 358). 102. Sharifov OF, Ideker RE, Fast VG: High-resolution optical mapping of intramural virtual electrodes in porcine left ventricular wall. Cardiovasc Res 64:448-456, 2004. 103. Clerc L: Directional differences of impulse spread in trabecular muscle from mammalian heart. J Physiol 255:335-346, 1976. 104. Wikswo JP Jr: Tissue anisotropy, the cardiac bidomain, and the virtual cathode effect. In Zipes DP, Jalife J (eds): Cardiac Electrophysiology: From Cell to Bedside, 2nd ed. Philadelphia: WB Saunders, 1995, pp 348-362. 105. Knisley SB: Transmembrane voltage changes during unipolar stimulation of rabbit ventricle. Circ Res 77:1229-1239, 1995. 106. Efimov IR, Cheng Y, Van Wagoner DR, et al: Virtual electrodeinduced phase singularity: A basic mechanism of defibrillation failure. Circ Res 82:918-925, 1998. 107. Efimov IR, Cheng YN, Biermann M, et al: Transmembrane voltage changes produced by real and virtual electrodes during monophasic defibrillation shocks delivered by an implantable electrode. J Cardiovasc Electrophysiol 8:1031-1045, 1997. 108. Iyer AN, Gray RA:. An experimentalist’s approach to accurate localization of phase singularities during reentry. Ann Biomed Eng 29:47-59, 2001. 109. Knisley SB, Smith WM, Ideker RE: Effect of field stimulation on cellular repolarization in rabbit myocardium: Implications for reentry induction. Circ Res 70:707-715, 1992. 110. Winfree AT: When Time Breaks Down: The Three-dimensional Dynamics of Electrochemical Waves and Cardiac Arrhythmias. Princeton, NJ: Princeton University Press; 1987. 111. Frazier DW, Wolf PD, Wharton JM, et al: Stimulus-induced critical point: Mechanism for electrical initiation of reentry in normal canine myocardium. J Clin Invest 83:1039-1052, 1989. 112. Chen P-S, Wolf PD, Dixon EG, et al: Mechanism of ventricular vulnerability to single premature stimuli in open-chest dogs. Circ Res 62:1191-1209, 1988. 113. Ideker RE, Alferness C, Hagler J, et al: Rotor site correlates with defibrillation waveform efficacy. Circulation 84:II-499, 1991.

Chapter 3

Sensing and Detection CHARLES D. SWERDLOW • JEFFREY M. GILLBERG • WALTER H. OLSON

T he electrical therapies of pacemakers and implantable cardioverter-defibrillators (ICDs) are controlled by sensing of cardiac depolarizations and detection of arrhythmias by analysis of the timing and morphology of sensed events. When a wavefront of depolarization passes the tip electrode of an intracardiac lead, a deflection in the continuous electrogram (EGM) signal travels instantaneously up the lead wire to the pacemaker or ICD. There, the signal is amplified, filtered, digitized, and processed by the sensing electronics. A sensed event is an instant in time when a pacemaker or ICD determines that an atrial or ventricular depolarization has occurred based on processing of the continuous EGM signal. Dual-chamber pacemakers and ICDs have separate sensing systems for the atrium and ventricle. Appropriate sensing results in one sensed event for each activation wavefront in the corresponding chamber. Failure to sense activation wavefronts results in undersensing, which can cause inappropriate pacing, failure to switch modes, or failure to detect a tachyarrhythmia. Undersensing occurs if the depolarization signal has insufficient amplitude or frequency content to be recognized as a sensed event or if a blanking period disables the sensing amplifier at the time of the event. Oversensing occurs when nonphysiologic signals are recorded or when physiologic signals that do not reflect local myocardial depolarization are recognized inappropriately as sensed events. Oversensing can cause inappropriate inhibition of pacing, inappropriate tracking, or inappropriate ICD therapy. Detection algorithms process sensed events to classify the atrial or ventricular rhythm. This classification is used to control beat-by-beat paced events, to change

the pacing mode, to store data regarding untreated tachyarrhythmias, and to terminate sustained tachyarrhythmias with antitachycardia pacing (ATP) or shocks.

Intracardiac Electrograms Surface Electrocardiogram Versus Intracardiac Electrogram An EGM is a graphic display of the potential difference between two points in space over time. The myocardium is composed of cells that maintain a resting potential across the membrane (i.e., the cell is polarized) such that the inside of the cell is electrically negative with respect to the outside of the cell. During the upstroke of the action potential, the inside of the cell abruptly changes from a negative potential (with respect to the outside of the cell) to a neutral or slightly positive potential. After a period of 300 to 400 msec, the cell membrane is then repolarized, with the inside of the cell returning to its resting, negatively charged state. Figure 3-1 illustrates how an EGM is recorded between two electrodes in contact with the myocardium. The electrocardiogram (ECG) is recorded from two electrodes on the surface of the body at some distance from the heart. The typical amplitude of its QRS complex is about 1 mV. The locations of the two electrodes determine the vectorial “viewpoint” from which the electrical activity of the entire heart is observed from the body surface. In contrast, the ventricular endocardial unipolar EGM typically is 5 to 20 mV in amplitude when recorded from a small electrode on the 75

76

Section One: Basic Principles of Device Therapy

tip of a lead placed in direct contact with the apex of the right ventricle (Fig. 3-2). The second electrode needed to record this unipolar EGM is the pacemaker or ICD metal can, which is located some distance from the heart. The location of this distant second electrode, sometimes called the indifferent electrode, has a much smaller effect on the signal’s properties, although it may record noncardiac electrical potentials (e.g., from the pectoral muscle). The ECG records electrical activity from the entire heart, whereas the EGM records only the local wavefronts of depolarization and repolarization that pass the tip electrode. The EGM depends on the viability of approximately 1 or 2 cm3 of myocardium immediately under the tip electrode,1,2 as depicted in Figure 3-2.

1

2

ⴙ ⴙ ⴙ ⴙ ⴙ ⴙ

ⴙ ⴙ ⴙ ⴙ ⴙ ⴙ

ⴚ ⴚ ⴚ ⴚ ⴚ ⴚ

ⴚ ⴚ ⴚ ⴚ ⴚ ⴚ

A

1

2 ⴙ ⴙ ⴙ

ⴙⴙ ⴙ ⴚ ⴚ ⴚ

Electrode Systems: Unipolar, Bipolar, Integrated Bipolar, Epicardial Figure 3-3 contrasts endocardial unipolar (tip-to-can), bipolar (tip-to-ring), and integrated bipolar (tip-to-coil) electrode systems, and Figure 3-4 shows representative examples. Epicardial electrode systems may be either unipolar (tip-to-can) or bipolar (tip-to-tip). These different electrode configurations have EGMs with similar R-wave amplitudes and slew rates, provided that the interelectrode spacing is at least 10 mm, as is true of almost all commercial pacemaker and defibrillator leads. Because unipolar electrode systems are more likely to oversense than are bipolar EGMs, they are contraindicated for ICDs and are used infrequently for modern pacemakers. Integrated bipolar electrodes used in ICDs have sensing characteristics that are more similar to the bipolar than the unipolar configuration. They are more likely to oversense, compared with true bipolar electrodes.3,4 In one study, oversensing occurred in 40% of patients with integrated bipolar sensing, compared with 8% of patients with true bipolar systems.4

ⴚ ⴚ ⴚ ⴚ ⴚ ⴚ

B

1

2 ⴙ ⴙ ⴙ

ⴙⴙ ⴙ ⴙ ⴙ ⴙ

ⴙ ⴙ ⴙ ⴚ ⴚ ⴚ

C

1

ⴙⴙ ⴙ ⴙ ⴙ ⴙ

2 ⴙ ⴙ ⴙ ⴙ ⴙ ⴙ

D

Amplitude, Slew Rate, and Waveshape of Electrograms The largest and steepest deflection on the local EGM, called the intrinsic deflection, occurs when the wavefront of depolarization passes the small-tip electrode. The EGM amplitude traditionally is defined as the peakto-peak amplitude, measured in millivolts, of the intrinsic deflection, as shown in Figure 3-5. The duration of a ventricular EGM usually is less than that of the QRS of the surface ECG, because the EGM is a local signal. The amplitude of an atrial or ventricular EGM is determined primarily by the excitable tissue near the tip electrode and therefore is usually similar for unipolar and bipolar signals. Typical amplitudes are 5 to 30 mV for ventricular EGMs and 1.5 to 6 mV for atrial EGMs.1,2 Increasing the size of the tip electrode in the range of 2 to 10 mm has minimal effect on atrial EGM amplitude but does increase EGM duration (Fig. 3-6). For short ventricular bipolar interelectrode spacing of 5 mm or less, the R-wave amplitude decreases, because the difference between the two unipolar EGMs from each electrode causes cancellation in the net bipolar signal.

ⴙ ⴙ ⴙ ⴙ ⴙ ⴙ

1

2

ⴙ ⴙ ⴙ ⴙ ⴙ ⴚⴚ ⴚ ⴚ ⴚ ⴙ

ⴙ ⴙ ⴙ ⴙ ⴙ ⴙ

E Figure 3-1. Illustration of how an EGM is recorded between two electrodes that are in contact with the myocardium. A, At rest, both electrodes record a similar charge, with no potential difference between them. B and C, As a wavefront of depolarization moves under electrode 1, a difference in electrical charge is generated such that electrode 1 becomes electrically negative with respect to electrode 2. D, As the wavefront propagates under electrode 2, no potential difference between the two electrodes is recorded. E, The depolarization wavefront is followed by a wavefront of repolarization, during which a potential difference of opposite polarity is recorded. Because the EGM is determined by the instantaneous potential difference between the electrodes, the amplitude and shape of the recorded signal are determined by the direction from which the wavefront approaches the electrodes. For example, if a wavefront of depolarization reached both electrodes at the same time, there would be no potential difference between the electrodes, and an EGM would not be inscribed.

Chapter 3: Sensing and Detection

77

RV-Coil

Tip Electrode Ring Tip 1 3

Voltage

2

1

3

2

Time Figure 3-2. This concept drawing indicates the spatial and temporal relationships for a unipolar endocardial EGM. The upper panel shows an anatomic drawing, and the lower panel shows the EGM recorded from a small-surface-area electrode at the tip of a pacemaker or defibrillator lead that makes direct contact with the endocardium in the right ventricular (RV) apex. The second electrode required to record an EGM is not shown, because it is a distant and indifferent electrode, usually the metal can of the pulse generator, and its location is not important provided that it is a substantial distance from the tip electrode. During a ventricular depolarization, the depolarization wavefront propagates from the septum, around the RV apex, and up the RV free wall (arrows). When the wavefront of depolarization arrives at location 1, just as it approaches the electrode, the initial positive deflection of the EGM occurs, at time 1. When the wavefront passes closest to the tip electrode at location 2, the major negative deflection on the EGM occurs, labeled as time 2. As the wavefront recedes from the electrode at location 3, the final portion of the EGM is inscribed at time 3. This local EGM is not affected by the depolarization wavefront as it travels further from the electrode. Therefore, the local EGM is shorter in duration than the surface electrocardiographic QRS complex.

The slew rate increases, because the time between arrival of the wavefront at the two electrodes decreases more than the EGM amplitude does. When two electrodes are widely separated, as in early Y-adapted cardiac resynchronization electrode systems, two distinct intrinsic deflections may be recorded on the EGM—one representing right ventricular (RV) activation and the other representing left ventricular (LV) activation. The interval between these deflections is determined by the conduction delay between the ventricles near the two electrodes.

Figure 3-3. The three practical endocardial electrode configurations used by most pacemakers and ICDs can be described with this figure. The distant indifferent or can electrode is not shown, because it is out of the field of view. The unipolar configuration used in Figure 3-1 to explain EGM formation simply records the signal between the tip electrode and the can. The tip electrode can be an active-fix screw or a small-surface-area tip electrode with various geometries. This unipolar configuration is subject to considerable noise and interference signals and is not suitable for ICDs. The bipolar configuration uses the Tip and Ring electrodes shown in the figure. The interelectrode spacing is typically 12 to 15 mm, and the ring electrode may or may not make contact with the endocardium. The integrated bipolar configuration uses the Tip and RV-Coil electrodes shown in the figure. EGMs recorded from bipolar and integrated bipolar configurations are very similar, and one less conductor is needed for the integrated bipolar configuration. The main disadvantages of the integrated bipolar configuration are susceptibility to diaphragmatic myopotentials, undesired atrial EGMs in small hearts, and slower postshock recovery times due to electrode polarization. RV, right ventricular.

The waveshapes of EGMs are quite variable (Fig. 3-7), probably because of the complex geometry of the trabecular endocardium adjacent to the tip electrode. In one study done at pacemaker lead implantation, 58% of unipolar EGMs were biphasic, with an initial upstroke followed by a roughly equal downstroke; 30% were predominantly monophasic negative, and 12% were predominantly monophasic positive.1 The maximum slope of the intrinsic deflection is the slew rate, measured in volts per second. It represents the maximum sustained rate of change of the EGM voltage. Mathematically, the slew rate is the first derivative of the voltage, dV/dt, so it depends on both the amplitude and the duration of the EGM. It is a crude representation of the frequency content of the EGM. The frequency content of ventricular and atrial EGMs is similar and in the range of 5 to 50 Hz. T waves and far-field R waves have lower frequency content, whereas most myopotentials and electromagnetic interference (EMI) have higher frequency content (Fig. 3-8). Typical values for slew rates are 2 to 3 V/sec for ventricular EGMs and 1 to 2 V/sec for atrial EGMs.1,2 Usually, an EGM with acceptable amplitude also has an

78

Section One: Basic Principles of Device Therapy

Figure 3-4. Ventricular electrocardiograms (ECGs) recorded from different electrode configurations in a single patient. The central panel shows a left anterior oblique radiograph of a cardiac resynchronization ICD system. Each of the four tracings shows surface ECG lead II, EGM markers, and one ventricular EGM during atrial pacing at a rate of 75 bpm. Top left, far-field EGM recorded between the right ventricular (RV) coil electrode and the electrically active ICD housing (CAN). Lower left, Integrated bipolar EGM recorded between RV tip and RV coil electrodes. Lower right, True bipolar EGM recorded between the RV tip and ring electrodes. Top right, Left ventricular (LV) unipolar EGM recorded between the LV tip electrode and CAN. EGM scale is 0.5 mV/mm, except for the LV unipolar EGM, which has a scale of 2 mV/mm. The downward EGM ventricular sense (VS) markers correspond to the time at which the true bipolar RV tip-ring EGM crosses the sensing threshold. Because the “field of view” of this EGM is local, its duration is short. It occurs early in the QRS complex of this patient with left bundle branch block. The integrated bipolar tip-coil EGM has a peak-to-peak amplitude and slew rate similar to those of the true bipolar EGM. However, its field of view is larger due to the size of the RV coil, and therefore the T wave is larger. Low-amplitude atrial EGMs are visible because of the proximity of the coil to the tricuspid annulus. Both the RV coil-CAN and the LV unipolar EGM are widely spaced, between an intracardiac electrode and the extracardiac CAN. Their duration is closer to that of the QRS complex. The intrinsicoid deflection of the LV unipolar electrode is late in the QRS complex, corresponding to late activation of the lateral LV. The greater amplitude of the LV unipolar EGM reflects the greater muscle mass of the LV. EGMs recorded from the superior vena cava (SVC) coil and from the atrial bipole (RA) are not shown. Radiograph and EGMs are from different patients. Radiograph is for illustrative purposes only.

Chapter 3: Sensing and Detection Figure 3-5. The major clinical descriptors of an intracardiac EGM are illustrated. The peakto-peak amplitude of the EGM is the difference in voltage recorded between two electrodes and is measured in millivolts (mV). The slew rate is equal to the first derivative of the EGM (dV/dt) and is a measure of the sharpness of the EGM and therefore its frequency content. Slew rate is measured in volts per second (V/sec). Usually, the amplitude of the EGM should be greater than 1.5 to 2.0 mV in the atrium and at least 5 to 6 mV in the ventricle at the time of lead implantation, to ensure adequate sensing. The slew rate should be at least 0.3 V/sec in the atrium and at least 1 V/sec in the ventricle.

Time (T)

Voltage (V)

200 ms

1 mV

1.7 mV

2 mm BIPOLAR CONTACT

32 msec 1.6 mV

10 mm BIPOLAR CONTACT

50 msec

1.0 mV

UNIPOLAR CONTACT

55 msec 1.3 mV

79

ORTHOGONAL CONTACT

20 msec ORTHOGONAL FLOATING 0.8 mV 20 msec Figure 3-6. Effects of electrode configurations on the atrial endocardial EGM. The EGMs were obtained from a single patient with two catheters placed simultaneously in the right atrial appendage. One catheter had 2-mm ring electrodes (top three tracings), and the other catheter had 1-mm orthogonal electrodes. The surface electrocardiogram tracing is shown at the top of the figure. Time and voltage amplitude scales are shown. For each electrode configuration (right), the corresponding EGM is shown (left), with the peak-to-peak amplitude and EGM duration labeled. “Contact” refers to electrodes in contact with the atrial tissue. “Floating” refers to noncontact electrodes in the atrial chamber. Note that greater ring electrode spacing, from 2 to 10 mm, prolongs EGM duration without altering the amplitude. The unipolar EGM shows a wider and diminished atrial EGM and a prominent far-field ventricular EGM as well. The orthogonal electrode configurations provide EGMs of lesser amplitude and shorter duration, compared with the ring electrodes.

Amplitude Slew rate =

Voltage (mV) (V) in Volts/sec (T)

SENSING

ATRIUM

VENTRICLE

Electrogram

1.5-2.0mV

5-6mV

Slew Rate (V/sec)

0.3

1.0

acceptable minimum slew rate (>1 V/sec for ventricular EGMs, >0.3 V/sec for atrial EGMs). EGMs with very low amplitude will not be sensed regardless of the slew rate. The ventricular depolarization recorded on the atrial electrode is referred to as the far-field R wave (FFRW). Oversensing of the FFRW confounds interpretation of the atrial rhythm. The amplitude of the FFRW depends strongly on the location of the atrial electrode. It is greatest near the septum, intermediate in the right atrial appendage, and least on the right atrial free wall. Even if the FFRW has comparable amplitude to the P wave, its slew rate usually is much lower. In one series, the mean slew rate was 1.2 V/sec for atrial EGMs and 0.13 V/sec for FFRWs.1 If an active-fixation, screw-in tip electrode is successfully attached to the myocardium, the acute ventricular EGM has a current of injury, with an elevated ST segment (Fig. 3-9) that is usually markedly reduced within 10 minutes after fixation. During the 10-minute period after electrode fixation, the EGM amplitude and slew rate usually do not change despite changes in waveshape, but the pacing threshold decreases by an average of 40% for EGMs.2 Acute to Chronic Electrogram Changes, Fixation Mechanism, and Steroid Elution The amplitude and slew rate of intracardiac EGMs typically decline during the first several days to weeks after lead implantation and then increase to chronic values that are slightly lower than those measured at implantation.5 The initial decrease in EGM amplitude is caused by the inflammatory response and edema at the electrodetissue interface. This gradually resolves and is followed by the development of a small, inexcitable fibrotic zone surrounding the electrode tip (Fig. 3-10). This inflammation and fibrotic tissue effectively increases the distance between the surface of the electrode and the excitable myocardium that generates the EGM signal. Although

80

Section One: Basic Principles of Device Therapy Figure 3-7. Similarities of unipolar and bipolar EGMs are shown by these examples recorded from a lead placed in the right atrial appendage in 10 patients. Note that the amplitudes of unipolar and bipolar EGMs are similar for each patient. The waveshapes of unipolar and bipolar EGMs for a given patient may be quite similar (patient 3) or quite different (patient 8), although these differences can be attributed to the relative size of the major inflections. Some of these differences may depend on whether the ring electrode for the bipolar recording makes contact with the myocardium. On the whole, intrapatient differences between unipolar and bipolar recordings appear to be less than interpatient differences.

Unipolar (mV)

5

0

5 1

2

3

4

7

8

9

5

Bipolar (mV)

5

0

0

5

100 (ms)

Unipolar (mV)

5

0

5 6

10

Bipolar (mV)

5

0 0

Signal amplitude (mV)

5

100 (ms)

NOT SENSED

SENSED

70 50 40 30 20 10

Amplifier Threshold R Waves

5 P Waves

4 3 2

EMI

T Waves Far-Field R Wave

1 0.7 1

2

5

10 20 Frequency (Hz)

Muscle Potentials 50

100

200

Figure 3-8. This plot of signal amplitude versus frequency shows the approximate characteristics of the P and R waves that pacemakers and ICDs are intended to sense and the approximate characteristics of the electromagnetic interference (EMI, muscle potentials), T waves, and far-field R waves that they are intended not to sense. The sense amplifier’s filters are designed to sense signals that are above the Ushaped amplifier threshold curve and to reject signals that are below the curve. P waves and R waves have similar frequency characteristics, but usually R waves have higher dominant frequency than P waves. Muscle potentials usually have higherfrequency components than intracardiac signals. T waves and far-field R waves have lower frequencies. As shown, there are some overlaps in these amplitude-frequency characteristics that cause oversensing or undersensing in particular situations. The ellipses representing the amplitude-frequency characteristics in this figure are conceptual and are not based on quantitative measurements.

chronic EGM amplitudes usually are reduced by less than 10% compared with acute amplitudes, chronic slew rates are reduced by 30% to 40%.6 The acute reduction in EGM amplitude is often greater with active-fixation leads than with passivefixation leads. Atrial undersensing can occur during the acute phase despite adequate EGM amplitudes at implantation. To account for these time-related changes in EGM amplitude, the filtered EGM recorded at lead implantation should be at least twice the sensitivity threshold that will be programmed in the pulse generator. Greater sensing safety margins are preferred for active-fixation leads. The method of lead tip stabilization, active screw-in or passive tines, has had no significant effect on sensing characteristics in most studies.7,8 Steroideluting electrodes reduce chronic pacing thresholds substantially but have no clinically significant effects on the sensing characteristics of endocardial leads.9-12 Metabolic, Ischemic, Aging, and Drug Effects on Electrograms The effects of metabolic abnormalities and drugs on pacing thresholds are well described. Much less information is available concerning their effects on EGMs and sensing functions of implantable pacemakers and ICDs. Factors that reduce EGM amplitude, slow con-

Chapter 3: Sensing and Detection Figure 3-9. Acute current of injury at implantation. Top panel, High-resolution recording shows marked ST-segment elevation, indicating the current of injury when an active fixation screw-in tip electrode is extended into the endocardium. Middle panel, After only 10 minutes, most of the ST-segment elevation in the signal has disappeared. Bottom panel, The EGM is not appreciably different at 1 hour after implantation.

81

20 mV Acute

18.4 mV

10 minutes

14.4 mV

1 hour

13.1 mV

10 mV

10 mV 20 mV 20 mV 10 mV

10 mV 20 mV 20 mV 10 mV

10 mV 20 mV

duction velocity, or diminish slew rate may produce either oversensing or undersensing. By prolonging the intracardiac EGM duration beyond blanking periods, ischemia or antiarrhythmic drugs can produce doublecounting of the QRS complex.13 Similarly, drugs that prolong the PR or QT interval beyond the refractory period may result in oversensing.14,15 Undersensing may result from reduction in EGM amplitude or slew rate after myocardial infarction at the electrode-tissue interface, from drug and electrolyte effects,14,15 or from progression of conduction system disease. Acute ischemia causes ST-segment changes that can be detected on ventricular EGMs. Monitoring of EGM ST-segment shifts has been proposed as a method for monitoring ischemia for pacemakers and ICDs.16 The likelihood of recording abnormal atrial EGMs (defined as ≥100 msec in duration or having ≥8 fragmented deflections) correlates with age of the patient (r = 0.34; P < .0005).17 Exercise, Respiratory, and Postural Effects on Electrograms The effect of exercise on the atrial EGM amplitude and slew rate is variable. Some studies have reported statistically significant decreases in amplitude that average 10% to 20% but may reach 40% in some patients.18,19

Other studies did not find significant changes between rest and exercise.20,21 Decreases in atrial EGM amplitude were not caused by atrial rate alone or by β-blockade.22 VDD/R lead studies with “floating” atrial electrodes showed particularly large decreases with exercise.23,24 These large decreases in atrial EGM amplitude for some patients support the value of programming a large safety margin for sensing at implantation when combined with the effects of lead maturation. P-wave amplitude increases significantly during full inspiration, during full expiration, and with erect posture.21 Substantial respiratory variations in the amplitude of the ventricular intrinsic deflection and slew rate are illustrated in Figure 3-11. Respiratory variation averaged 9.7% for unipolar atrial EGMs and 11.5% for bipolar atrial EGMs.24,25 The effect of respiration on ventricular EGMs was less, especially with the unipolar configuration.25 Ventricular Electrograms during Premature Ventricular Complexes, Ventricular Tachycardia, and Ventricular Fibrillation Premature ventricular complexes (PVCs) may have lower-amplitude R waves than sinus-rhythm R waves, as shown in Figure 3-12, but the reverse may also be

82

Section One: Basic Principles of Device Therapy

Approximate Outline of Lead Body and Helix Figure 3-10. The myocardium remaining after removal of an endocardial active-fixation screw lead is shown on this gross microscopic slide. The large dotted line shows where the lead body was located, and the colored staining shows a thin fibrotic sheath around the lead body. The approximate location of the helical screw-tip electrode is shown by the solid coiled line. The oval shape (dotted line) shows the size of the fibrotic capsule that formed around the helical extended-tip electrode. Most of the tissue outside the dotted lines stained red, indicating that it was active myocardium capable of conducting depolarizations. The tip region of this electrode is similar to that of the tip electrode in Figure 3-1, so propagation of depolarization wavefronts must travel around the tip electrode, in tissue largely out of the field of view on the right side of this figure.

true. For monomorphic ventricular tachycardia (VT), mean amplitude decreased only slightly from values in sinus rhythm—14% for epicardial EGMs and 5% endocardial EGMs.26 In contrast, EGM amplitudes during ventricular fibrillation (VF) decreased by 25% for epicardial and 41% for endocardial EGMs. More importantly, EGMs in VF often have low, highly variable, and rapidly changing amplitudes and slew rates. Figure 3-13 shows endocardial spontaneous VF EGMs from different patients, illustrating variability in intrinsic deflections, amplitudes, slew rates, and morphologies. In a study of induced VF reproducibility, 50% of the variability was due to interpatient differences and the other 50% was due to repeated episodes in the same patient.27 In another study, the ventricular EGM amplitude in VF was 1 mV or less in at least one VF episode in 29% of patients.26 When analyzing the sensed EGMs during induced VF at ICD implantation, care should be paid to variability in the beat-to-beat amplitude. If this variability in EGM amplitude is large, undersensing of VF may occur. Undersensing in VF may necessitate lead repositioning or insertion of another sensing lead despite an adequate EGM in sinus rhythm. See “Undersensing.” If VF persists for several minutes, the amplitude and slew rate of the EGMs deteriorate as shown in Figure 3-14. Atrial Electrograms during Rhythms Other Than Sinus Compared with sinus rhythm, atrial activation from ectopic sites or atrial arrhythmias can alter the ampli-

Voltage

Slew Rate (dV/dt)

1 sec.

0.2 sec.

Figure 3-11. Respiratory variations of intracardiac EGM amplitude and slew rate (dV/dt) are typically about 10% for the atrium and ventricle. These variations result from beat-to-beat changes in stroke volume caused by respiration or movement at the electrode-tissue interface. (From Furman S, Hurzeler P, De Caprio V: Cardiac pacing and pacemaker. III. Sensing the cardiac electrogram. Am Heart J 93:794-801, 1977, with permission.)

Chapter 3: Sensing and Detection

83

CHART SPEED 25.0 mm/s ECG

0.2 mV/mm LEAD II

V EGM

1.0 mV/mm

Slew Rate=dV/dt V

P-P Amplitude

t

MARKER CHANNEL

V 5

V 5

V 5

V 5

V 5

V 5

V 5

V 5

Figure 3-12. A surface electrocardiogram (ECG) lead II, a bipolar right ventricular (V) EGM, and event markers with downward pulses that show when sensing occurred. The QRS amplitude on the ECG is about 1 mV, which is typical. The peak-to-peak amplitude of the sinus R waves is about 10 to 12 mV, which is also typical. The slew rate is the maximum slope (dV/dt) of the EGM intrinsic deflection; it is difficult to measure with a paper speed this slow. The two premature ventricular complexes (PVCs) (fourth and sixth) have different amplitudes and shapes on both the ECG and the EGM. The sinus beat in the center, between the two PVCs, has its main intrinsic deflection during the last part of the ECG QRS complex. The left edge of the sense marker indicates the instant that sensing by the device occurred. Therefore, the EGM morphology and timing of the sense marker may not correspond to the start of the QRS on the ECG as electrocardiographers expect. Each R wave was sensed only once because sensing is blanked by the device for 120 msec after each ventricular sense (VS).

tude, frequency content, slew rate, and morphology of the atrial EGM. Retrograde atrial activation during ventricular pacing reduces atrial EGM amplitude and slew rate by up to 50%.28 These EGM changes are more pronounced in the high right atrium than in the right atrial appendage or low right atrium.29 The frequency content of the atrial EGM is not significantly altered by retrograde atrial activation.30 Analysis of EGM turning-point morphology or the first-differential coefficient of slew rate has been used to discriminate sinus EGMs from those recorded during retrograde and ectopic atrial activation in small groups of patients.31 Less sophisticated visual morphologic analysis did not effectively discriminate sinus from ectopic atrial electrical activity.28 Atrial EGMs during atrial fibrillation (AF) are characterized by extreme temporal and spatial variability. EGMs tend to be most organized in the trabeculated right atrial appendage and more disorganized in the smooth right atrium or coronary sinus.32-34 The amplitude, width, and morphology of atrial EGMs during AF vary markedly at various anatomic locations. The amplitude of chronic, unipolar pacemaker EGMs in AF was decreased by 40%, compared with sinus rhythm.35 The spectral components of EGMs from two separate atrial (or ventricular) sites show greatly reduced spectral coherence during fibrillation, as opposed to an atrial rhythm other than AF.36

A comparison of atrial EGM amplitudes in sinus rhythm, AF, and atrial flutter with temporary pacing catheters in the high right atrium or right atrial appendage showed that the mean sinus-rhythm EGM amplitude decreased only slightly in atrial flutter but decreased by about 50% in AF.34 The mean EGM amplitudes in both AF and atrial flutter were highly correlated to the amplitudes in sinus rhythm. The coefficient of variance of EGM amplitude was similar for sinus rhythm (19%) and atrial flutter (22%) but markedly increased for AF (42%, P < .0001 versus sinus). The likelihood of any patient demonstrating very-lowamplitude atrial EGMs ( V rate, PR Logic uses P-R patterns along with R-R regularity, P-R dissociation, and P-P regularity to ensure detection of double tachycardia (VT or VF during atrial fibrillation [AFib]) and to withhold therapy for 2:1 atrial flutter (AFlutter), rapid AFib, and ST with far-field R wave (FFRW) os. B, PR Logic computational flow diagram. On each ventricular event, PR Logic processes the new P-R, R-P, P-P, and R-R patterns and timing information for the building blocks. If VT/fast VT (FVT) or VF rate detection criteria are satisfied, the ventricular rate override criterion is checked first. If the median R-R interval is less than the supraventricular tachycardia (SVT) limit, detection occurs via the single-chamber detection criteria without considering the PR Logic discrimination algorithm. If the median R-R interval is greater than the SVT limit and double tachycardia (VT/FVT/VF + SVT) is not detected, then the three PR Logic criteria for identifying SVTs are tested in the order shown. If any one of the PR Logic SVT criteria is satisfied, inappropriate detection is avoided. If an SVT is not positively identified, VT/FVT/VF is detected when the R-R interval–based criterion is satisfied. If SVT is identified, the entire process repeats itself on each ventricular event until VT/FVT or VF is detected or the rhythm slows out of the ventricular rate detection zones.

No

Chapter 3: Sensing and Detection Figure 3-73. ELA PARAD+ algorithm. If most of the detected R-R intervals are in the ventricular tachycardia (VT) zone, ventricular interval stability is analyzed in a first step, using a histogram of R-R intervals. If the rhythm is irregular, atrial fibrillation (AF) is diagnosed and therapy is withheld. If the rhythm is regular, the atrioventricular (AV) association is assessed by comparing peak amplitudes of R-R and P-R interval histograms. If the rhythm is AV dissociated, VT is diagnosed, unless the Long Cycle Search is activated, which inhibits therapy if a long ventricular cycle (VTLC; characteristic of AF) is identified. If N:1 P-R association is identified, the rhythm is classified as supraventricular tachycardia (SVT). In the presence of 1:1 P-R association, the PARAD+ evaluates the rate of acceleration of the ventricular rate. If acceleration is gradual, sinus tachycardia (ST) is diagnosed. If acceleration is sudden, PARAD+ identifies the chamber of origin and withholds therapy if it is the atrium. AFlutter, atrial flutter; AT, atrial tachycardia.

131

Tachyarrhythmia Detected

RR Stability Unstable Diagnosis AF

Stable Associated PR No

Yes

VTLC No Diagnosis VT

Yes Diagnosis No AF Diagnosis AFlutter

PR Association 1:1 Acceleration Gradual

Sudden Origin of acceleration

Diagnosis ST Atrial Diagnosis AT

ELA118 and St. Jude119 algorithms demonstrated moderate superiority in SVT-VT discrimination for dualchamber algorithms. Active Discrimination Pacing maneuvers were not incorporated into early detection algorithms for many reasons, including concerns regarding proarrhythmia and the complexity of interpreting responses (Fig. 3-74). It is now recognized that conservative burst ventricular pacing at approximately 90% of the tachycardia cycle length has a low (1% to 2%) risk of proarrhythmia and that active discrimination provides unique advantages. In addition to terminating most monomorphic VTs, burst ventricular pacing terminates more than 50% of inappropriately detected pathologic SVTs with a 1:1 AV relationship.120 During other SVTs, concealed retrograde conduction of ventricular ATP may slow the ventricular response by causing AV conduction delay or block. Active discrimination represents a paradigm shift in the design of detection algorithms, from “diagnose before intervening” to “treat first; analyze only those tachycardias that persist after treatment.” To date, active discrimination methods have not been implemented in an ICD. Single-Chamber Analysis The number of pulses required for entrainment of a reentrant tachycardia is related to the conduction time from the pacing site to the tachycardia circuit. If the VT circuit is entrained, the first postpacing interval is independent of the number of pacing pulses delivered. There is a high probability of entraining reentrant VT with only a few pacing pulses. In contrast, nonreentrant SVT cannot be entrained, and reentrant SVT

Ventricular Diagnosis VT

usually requires about 10 ventricular pacing pulses for entrainment.101 Therefore, the difference between first postpacing intervals after the entrainment of VT by two sets of pacing bursts with different numbers of intervals is less than the comparable difference in postpacing intervals for bursts delivered during SVT. A limitation of this method is delay in detection of VT required to deliver and analyze the outcome of at least two sequences of burst pacing. Dual-Chamber Analysis In dual-chamber algorithms, the principal value of active discrimination applies to tachycardias with 1:1 AV association. Pacing is performed in one chamber with analysis in the opposite chamber. If the atrial cycle length is unchanged by ventricular burst pacing, the atrial rhythm does not depend on retrograde conduction, and the diagnosis is SVT. This feature has been tested but not yet implemented in an ICD. If the atrial rate accelerates to the ventricular rate during ventricular burst pacing, the response at the termination of unsuccessful pacing may be helpful: Two atrial events followed by a ventricular event (A-AV response) is quite suggestive of AT121 (Fig. 3-75), although a slowly conducting retrograde accessory pathway or the slow AV nodal pathway may give the same result. Rarely, the pattern of two ventricular events followed by an atrial event (V-V-A response) is diagnostic of VT. But a ventricular event followed by an atrial event followed by another ventricular event (V-A-V response) indicates VT if AV nodal and AV reentrant SVT are excluded (Fig. 3-76). Atrial extrastimuli or overdrive pacing may also discriminate between SVT and VT with 1:1 association. One method is to deliver a single atrial extrastimulus

132

Section One: Basic Principles of Device Therapy

RA ATP

ATP ATP

413

373

S

S

T

T

T

T

T

T

422 422 422 422 422 422 418 434 422 418 426 422 422 410

493

ATP ATP ATP ATP ATP ATP ATP ATP D 423 423 413 431 399 443 423

P ATP

R

ATP ATP

R

R

R

R

R

R

R

ATP ATP ATP ATP ATP ATP ATP ATP

RV 32

33

34

35

36

37

38

39

40

DDI

Proarrhythmia ATP in 1:1 SVT (rapidly conducted AF)

F

T

F

T

R

R

40 DDI

F

F

F

F

F

41

42

F

F

F

43

F

F

F

8V 328 293 301 289 305 371 426 344 430 414 397 P P P R R R R R R

48

49

* R

* R

* R

* R

F

F

F

F

* R

* R

* R

* R

* R

* R

* R

H V

* R

R R D 309 297 461 473 309 309 320 297 305 324 316 309 301 281 285 324 465 355 262 289 301 270 426 277 242 246 242 254 316 383 281 281 301 270 281 293 387 309 P R R R R R R R R R R R R R R R R R R R R R R

320 332 293 348 305 441 R

F

* R

50

* R

* R

* R

51

44

* R

* R

* R

45

* R

* R

* R

46

* R

* R

* R

47

* R

* R

* R

* R

48

* R

* R

D 26 293 289 309 305 230 297 320 277 293 273 277 285 273 293 258 324 336 418 328 293 297 273 258 309 305 324 277 313 35 P P R R R R R R R R R R R R R R R R

52

53

54

55

56

52

53

54

55

56

H V

* R

R

R 830V

6 262 445 281 648 309 289 2 324 285 254 285 285 266 297 313 273 R

48

R

R

R

49

R

R

R

50

R

R

51

Figure 3-74. Atrial proarrhythmia caused by ventricular antitachycardia pacing in 1:1 supraventricular tachycardia (SVT). Continuous stored EGM shows right atrial (RA) EGM, event markers, and true bipolar right ventricular (RV) EGM. In the top panel, antitachycardia pacing (ATP) is delivered after rate-only, inappropriate detection of sinus tachycardia as ventricular tachycardia. Ventricular-atrial conduction occurs, initiating atrial fibrillation (AF), which is detected as ventricular fibrillation and requires two shocks (HV) before termination. The first shock is delivered at the end of the middle panel and the second in the middle of the lower panel. Recording is suspended for approximately 1 second after each shock.

133

Chapter 3: Sensing and Detection Figure 3-75. Effect of ventricular antitachycardia pacing on supraventricular tachycardia (SVT) with 1:1 atrioventricular (AV) relationship. Atrial EGM (Atip to Aring), high-voltage ventricular EGM (Can to HVB), and dual-chamber event markers are shown. Burst ventricular antitachycardia pacing at cycle length 270 msec is applied to SVTs with cycle lengths of 310 to 320 msec. The atrial interval after the last paced beat is accelerated to the pacing rate, probably indicating entrainment of the atrium. The A-A-V response at termination of pacing is diagnostic of atrial tachycardia.

EGM1: Atip to Aring

EGM2: Can to HVB 3 1 0

A-A Interval (ms)

3 1 0

A A S S

3 1 0

A S

3 2 0

A S

3 2 0

3 3 3 2 2 2 0 0 0 A A A A S S S S

6 4 0

A S

3 2 0

A S

3 2 0

3 1 0 A A S S

2 7 0

3 6 0

A S

A S

3 1 0

3 1 0

A A S S

3 1 0

Marker Annotation T S

V-V Interval (ms)

3 1 0

T T T S S S 3 3 1 1 0 0

T T T T T T T S S D P P P P 3 3 3 2 2 2 2 1 2 2 7 7 7 7 0 0 0 0 0 0 0

T T T T P P P P 2 2 2 2 7 7 7 7 0 0 0 0

V S

7 1 0

3 6 0

T T S F 2 7 0

A-EGM

V-EGM Figure 3-76. Atrial response to ventricular antitachycardia pacing during ventricular tachycardia (VT) with 1:1 ventricular-atrial (VA) conduction. Stored atrial (A) and ventricular (V) EGMs are shown with event markers. During antitachycardia pacing, atrial rate accelerates to ventricular rate. The V-A-V response at termination of pacing is characteristic of VT, although atrioventricular (AV) nodal and AV reciprocating tachycardias cannot be excluded.

A R T S

A R T S

3 6 0

A R T S

3 6 0

A R T S

3 6 0

A R T S

3 6 0

A R T S

3 6 0

A R T S

3 6 0

3 4 0

3 7 0

3 6 0

3 6 0

3 6 0

3 6 0

3 6 0

3 6 0

A R T S

3 7 0

3 3 3 4 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 8 8 9 2 1 1 1 1 2 2 3 2 2 2 2 7 6 6 4 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 A A A A A A A A A A A A A A A A A A A A R R R R S S S S S S S S S S S S R R R R T S

3 6 0

and analyze the timing of the next ventricular event. Another is to deliver a train of atrial pacing and evaluate the ventricular response. Atrial extrastimuli usually shorten the next R-R interval in SVT, and atrial overdrive pacing usually accelerates the ventricular rate during SVT. This method may classify the rhythm incorrectly if the atrial stimuli block in the AV conduction system or conduct with too much decrement. It may also misclassify VT if ventricular capture occurs, usually in slower VTs. Atrial pacing may be proarrhythmic for either the atrium or the ventricle.122 Duration-Based “Safety-Net” Features to Override Discriminators These programmable features deliver therapy if an arrhythmia satisfies the ventricular rate criterion for a sufficiently long duration even if discriminators indicate SVT (Guidant Sustained-Duration Override, Medtronic High Rate Timeout, St. Jude Maximum Time to Diagnosis). The premise is that VT will continue to

T S 3 6 0

T S 3 7 0

T S 3 6 0

T P 3 7 0

T P 3 2 0

3 2 0

T T T T T T T T T T T P P P P P P P P P P P 4 3 3 3 3 3 3 3 3 3 3 9 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0 0 0 0 0 0 0

V S

T S 3 9 0

T S 3 8 0

3 8 0

satisfy the rate criterion for the programmed duration, whereas the ventricular rate during transient sinus tachycardia or AF will decrease below the VT rate boundary before the duration is exceeded. The limitation is delivery of inappropriate therapy when SVT exceeds the programmed duration, which occurs in approximately 10% of SVTs at 1 minute and 3% of SVTs at 3 minutes.103,111 The decision to use a discriminator override should be based on clinical factors, including the probability that discriminators will prevent detection of VT, the likely consequences of failure to detect VT, and the likelihood that SVT in the VT rate zone will persist long enough to trigger inappropriate therapy because of the override. For example, override features may be considered whenever a morphology algorithm is programmed without inducing VT at electrophysiologic study. The programmed duration in the Guidant Atrial View and St. Jude algorithms should be increased from the nominal value of 30 seconds to reduce inappropriate therapies. In most patients, a duration of 2 to 5 minutes is appropriate.

134

Section One: Basic Principles of Device Therapy

SVT-VT Discriminators in Redetection

equivalent SVT-VT discrimination in initial detection and redetection, except that algorithm building blocks related to tachycardia onset are disabled. Guidant algorithms permit programming discriminators after shocks, but not after ATP. In Medtronic ICDs, the single-chamber stability discriminator applies to redetection if it is ON for initial detection, but dual-chamber discrimination is not applied. St. Jude devices do not

SVT-VT discrimination during redetection serves two purposes: (1) to prevent inappropriate therapy for SVT after appropriate therapy for VT (Fig. 3-77) and (2) to provide a second chance for the algorithm to classify SVT correctly after inappropriate therapy. Biotronik ICDs (Fig. 3-78) and ELA ICDs provide essentially

RV Tip-RV Ring

RV Coil-Can VT T F 3 2 0

3 1 0

T F 3 2 0

T F 3 2 0

T F 3 3 0

T F 3 3 0

T F 3 3 0

T F 3 3 0

T F 3 4 0

T F 3 3 0

T F 3 5 0

T F 3 4 0

T F 3 3 0

T F 3 4 0

T F 3 3 0

T F 3 4 0

ATP

T F 3 6 0

T F 3 6 0

T P3 6 0

T P2 9 0

T T P2 P 2 9 9 0 0

T P2 9 0

T P2 9 0

Sinus Tachycardia T F2 9 0

T F 2 9 0

T P2 9 0

T P

T S

8 1 0

ATP T• F

4 3 0

T• F

T• F

4 3 0

4 4 0

T• F

T• F

4 4 0

4 5 0

T• F

T P

4 4 0

T T P P 3 3 6 6 0 0 FVT Rx 1/Sec 2

3 6 0

T P

T P

3 6 0

3 6 0

T P

T P

3 6 0

T F

4 6 2

T F

4 5 0

T F

4 5 0

T F 4 4 0

4 4 0

T F

Sinus Tachycardia T P

3 6 0

T F

6 0 0

3 6 0

T P

3 6 0

T P

T P

3 6 0

V S

8 7 0

4 1 0

T• F

V S

5 0 0

4 4 0

T• F

T• F

4 4 0

T• F

4 4 0

T• F

4 3 0

Shock 4 3 0

V S

4 4 0

V S

4 6 0

V S

4 6 0

V S

3 7 0

C E

5 3 0

V S

4 6 0

C D

8 3 0

V S

T• F

4 3 0

5 6 0

V S

5 8 0

V S

V S

8 1 0

4 1 0

T• F

5 2 0

V S

3 3 0

T• F

5 4 0

V S

4 2 0

T• F

4 2 0

T• F

4 1 0

T• F

4 1 0

T• F

4 1 0

14.9 J

A V-V 1,800

VF = 300 ms Burst

V-V interval (ms)

1,500

FVT = 360 ms 20.1 J

VT = 470 ms 35.1 J

34.8 J

14.9 J

35.1 J

1,200 900 600 Sinus Tach 400

200

40

VT

20

ATP

0

20

40

60

Time (sec) [0 = detection]

B

80

100

*

*

Figure 3-77. Redetection. Inappropriate therapy of sinus tachycardia after appropriate therapy for ventricular tachycardia (VT) in a single-chamber ICD. Ventricular sensing and high-voltage leads are shown with event markers. A, Continuous stored EGM strips after detection of VT. The top strip shows successful antitachycardia pacing (ATP) of VT followed by sinus tachycardia in the Fast VT (FVT) zone (TF on event markers). The second strip shows the first inappropriate sequence of ATP. The third strip shows the first of five shocks delivered after ATP. B, Interval plot shows the entire episode. No U.S. ICD manufacturer provides for VT-supraventricular tachycardia (SVT) discrimination after ATP. (From Swerdlow CD, Friedman PA: Advanced ICD Troubleshooting: Part II. PACE 29: 70-96, 2006, with permission.) RV, right ventricular; VF, ventricular fibrillation.

135

AFlut 399

AFlut 398

AFlut 399

VT1 390

VT1 398

VT1 399

AS 719 Ars 195 AS 203 AS 196 Ars 195 AS 203 AS 195 Ars 196 AS 203 Ars 195 AS 196 AS 203 Ars 195 VS 1844

AS Ars 109 20 J 81Ohm

VT2 320

25 AS 985 VS 984

AS 984 VS 984

AS 985 VS 985

AS 975

AS 985

Shock VS 977

VT2 328

VT2 321 AS 984

VS 984

VS 984

AS 984

AS 985 VS 985

VS 992

AFlut 391

AFlut 390

AFlut 391

AFlut 391

AFlut 398

AS 726

25 mm/s

AS 195 AS 204 Ars 195 AS 195 AS 203 Ars 196 AS 203 Ars 195 AS 195 AS 203 Ars AS 125 AFlut 398

VT2 328

VT2 351

Charge

VS 984

AS 352

AS 351

VT2 352

AS 351 VT1 351

VT2 352

VT2 352

AS 704

AS 1617 VT2 320

VT2 328

VT2 320

VT2 328

VT2 321

AS 586 VT2 320

VT2 351

AS 750 VT1

VT2 320

VT2 328

VT2 320

AS 1617

Chapter 3: Sensing and Detection

Figure 3-78. Supraventricular tachycardia/ventricular tachycardia (SVT-VT) discrimination in redetection. Dual-chamber EGM markers, atrial EGM, and true bipolar EGM are shown in this continuous strip from a Biotronik Lexos ICD, Model 347000. Rhythm at beginning of upper panel is VT with ventricular-atrial (VA) dissociation. Vertical dotted line indicates detection and start of capacitor charging (Charge). Black horizontal line indicates period of capacitor charging followed by successful shock. Postshock nonsustained atrial flutter begins toward right side of upper panel, continues into lower panel, and shows transient atrial flutter followed by conversion to sinus rhythm. Ventricular rhythm is classified by the lower row of markers as VT2 (faster VT zone) before shock, VT1 (slower VT zone) for the first four ventricular EGMs of atrial flutter, and conducted atrial flutter (Aflut) beginning at the fifth conducted ventricular EGM during atrial flutter (arrowhead). Without SVT-VT discrimination in redetection, VT would have been redetected after 10 intervals.

apply discriminators to redetection. Therefore, neither Medtronic nor St. Jude ICDs provide any single- or dual-chamber discriminators to reject sinus tachycardia after therapy. Measuring Performance of SVT-VT Discrimination Algorithms A comprehensive assessment requires analysis of all tachycardia episodes, including those that are not stored in ICD memory and those in the VF zone to which discriminators may not apply.62,63,114 Programmed detection parameters may influence reported algorithm performance.63 In most studies, a few patients contribute a large number of SVT or VT episodes. Therefore, statistical methods such as the generalized estimating equation (GEE)123 should be used to remove bias in raw performance measures introduced by these unusual patients. Although there are valid reasons to focus analysis on a limited class of tachycardias to evaluate the differential performance of specific algorithm building blocks, overall performance across the patient population must always be considered. Quantitative Considerations All detection algorithms must maintain almost 100% sensitivity for detection of VT. If patient populations and programmed detection boundaries are equivalent, positive predictive accuracy may be the most useful statistical measure of algorithm performance.62 It is highly dependent on the ratio of SVT to VT episodes and therefore on the programmed detection rate and patient population (prevalence of SVT and VT). Nevertheless, it is usually preferable to specificity, an commonly used

alternative measure. Regardless of what specific performance measure is employed, it is almost impossible to obtain clinically meaningful insights into the performance of two different detection algorithms based only on their performance numbers on different sets of data. Consider the following two examples. In the first example, the distribution of SVT versus VT and the distribution of “easy” versus “difficult” arrhythmias (e.g., AF with a highly irregular ventricular rate versus AF with a faster, regular ventricular rate) presented to an algorithm influences its performance.124,125 For instance, an algorithm that uses consecutive-interval counting rejects “easy” AF with highly irregular ventricular rate at the level of rate counting and does not store the episode in ICD memory. Only “difficult” AF with a more regular ventricular rhythm is presented to the discriminator and stored in the ICD. In contrast, algorithms that use X out of Y counting or interval averaging present most AF episodes with highly irregular ventricular rhythms to the discriminator for storage in ICD memory.96,126 However, if both ICDs delivered eight appropriate therapies for VT and two inappropriate therapies for AF, they would have the same positive predictive value, 80%. In the second example, single clinical SVTs frequently are counted as multiple SVT episodes in ICD data logs, because the ventricular rate varies spontaneously around the VT rate threshold. This is especially true of AF. Consider one clinical episode of AF that is rejected correctly nine times within a few minutes and then treated inappropriately with a shock that terminates it. By analysis of the ICD data log for ICD episodes, the specificity is 90%. By the patient’s reckoning, the inappropriate shock is the same whenever it happens, during the episode of AF or not. On the

136

Section One: Basic Principles of Device Therapy

other hand, credit is rarely assigned to algorithms for “partial success” or for success despite compromised sensing. An algorithm that detects VT inappropriately after 10 minutes of SVT has performed better than one that detects inappropriately after 10 seconds of the same SVT. An algorithm that classifies the atrial rhythm correctly with low-amplitude atrial EGMs and large FFRWs is more robust than one that makes the same classification from an exemplary atrial signal.

anticoagulated. Finally, early recurrence is common after transvenous cardioversion of AF.128 Experts differ about whether algorithmically inappropriate ventricular therapy of SVT may be clinically appropriate in specific clinical situations. ICDs designed to deliver both atrial and ventricular therapies may be implanted in patients who are likely to benefit from device-based therapy of SVT.

Inappropriate Detection Versus Inappropriate Therapy

Detection: Programming and Troubleshooting

ATP is delivered immediately after detection. Therefore, the numbers of detections and therapies are equivalent. In contrast, shock delivery requires capacitor charging, during which therapy may be aborted. However, delivery of an inappropriate shock results in a more adverse clinical outcome than delivery of inappropriate ATP. Active discrimination blurs the distinction between detection and therapy, introducing new complexity into the evaluation of algorithm performance. With active discrimination, only tachycardias that persist immediately after diagnostic pacing are classified. Further, concealed conduction into the AV node may cause transient slowing of the ventricular rate during atrial arrhythmias, resulting in repetitive ICD-defined detections and terminations of a single SVT episode.

Present ICDs have multiple programmable parameters that affect detection directly or indirectly. They provide opportunity both for customizing detection and for operator error. The trend of the 1990s for maximizing flexibility has been replaced by one to simplify the user interface while minimizing loss of automatic or operator-programmed versatility to address a wide range of clinical situations. A recent prospective study reported that empirical programming of dual-chamber ICDs provided comparable performance to individualized programming.115 Even if nearly universal nominal programming may soon be possible, it is important to understand the many tradeoffs and compromises inherent in programming and troubleshooting ICD detection.

Comment

Detection Zones and Duration

Valid assessments and comparisons of algorithm performance require consideration of multiple devicerelated and clinical factors. In isolation, statistical measures are rarely informative. Is Ventricular Therapy for Supraventricular Tachycardia Always Inappropriate? Persistent, rapidly-conducted atrial arrhythmias can cause hemodynamic compromise in patients with LV dysfunction or ischemia in patients with severe coronary artery disease. Because ventricular shocks often terminate AF and ventricular ATP often terminates 1:1 AT, algorithmically inappropriate ventricular therapy may fortuitously terminate clinically significant SVT. However, inappropriate ventricular therapy for SVT can have serious consequences. ATP may be proarrhythmic,127 and shocks for rapidly conducted AF have multiple drawbacks. First, AF in patients with ICDs is often paroxysmal, rapid conduction is often transient, and symptoms are usually mild, but ventricular shocks delivered shortly after detection do not permit spontaneous termination of AF or slowing of the ventricular rate. Therefore, they will be delivered for AF that would either have terminated spontaneously or have had only transient, rapid conduction. Second, detection algorithms in ventricular ICDs cannot use the total duration of (slowly conducted) AF to withhold shocks. Therefore, inappropriate shocks for AF may place patients at risk for thromboembolism if they are not

Zones and Zone Boundaries Typical values for rate zone boundaries are 500 to 360 msec for slower VT, 360 to 300 msec for faster VT, and 300 to 240 msec for VF. Two- or three-zone programming is preferred in most patients, even those undergoing implantation for secondary prevention and those whose only clinical arrhythmia is VF, because most spontaneous VF begins with rapid VT, and most rapid VT can be terminated by ATP.129-131 Three zones permit different ATP therapies for two distinct rates of VT, as well as ATP for monomorphic VT that overlaps in rate with polymorphic VT. Other experts recommend two rate zones, with the slowest at a cycle length of 340 to 320 msec, for primary prevention patients132 and those whose only spontaneous arrhythmia is VF. This approach lowers the risk of inappropriate therapy but increases the risk of not treating VT. In the largest primary prevention trial,132 ICDs were programmed to a single detection zone at a cycle length of 320 msec, without SVT-VT discriminators. Approximately one third of shocks were inappropriate; the incidence of sudden death from undetected VT was not reported. The sinus-VT rate boundary should be slow enough to ensure detection of all hemodynamically compromising VTs. To prevent underdetection of irregular VT, the VT detection interval should be set with a safety margin at least 40 to 50 msec longer than the slowest predicted VT for consecutive-interval counting and 30 to 40 msec longer for X out of Y or interval+intervalaverage counting.118 This safety margin should be a

Chapter 3: Sensing and Detection

longer cycle length if rapidly conducted SVT is unlikely or SVT-VT discrimination is reliable at long cycle lengths (ELA ICDs). The boundary between the two VT zones should be based on the cycle length at which different types or fewer trials of ATP are preferred. The VT-VF rate boundary is based on the cycle length below which ATP should not be delivered. In Medtronic ICDs, which use consecutive interval counting above the VF interval and X out of Y counting below it, this boundary should be set to prevent underdetection of irregular, polymorphic VT by consecutive interval counting. Monitor-only Zones If therapy is not programmed “on” for slow VT, the slowest rate zone may be programmed as a “monitoronly” zone, with detection “on” and therapies “off.” However, in Guidant, St. Jude, and older Medtronic ICDs (before Marquis), interactions between the counters in the monitor-only zone and the next zone may restrict use of SVT-VT discriminators or decrease the number of intervals required for detection in the therapy zone. Newer Medtronic ICDs (Marquis and later) provide independent monitor-only zones that avoid these limitations: because events in the “monitor” zone do not increment the detection counter, tachycardias in the monitor-only zone do not accelerate therapy even if a few intervals cross into the slowest therapy zone. The associated risk is delay in VT therapy if the VT cycle length fluctuates around the border between the monitor-only and therapy zones. Duration for Detection of Ventricular Tachycardia Detection duration before ATP should not be decreased from nominal values, because therapy is immediate after detection, and undersensing of monomorphic VT is rare. It should be increased in patients who have long episodes of nonsustained VT (e.g., long QT syndrome). Substantial increases in duration for detection of VT probably are safe in St. Jude and Guidant ICDs, which use counting methods that are insensitive to occasional long ventricular intervals or undersensing. In contrast, consecutive-interval counting used by Medtronic ICDs in the VT zone may underdetect VT if occasional long ventricular intervals or undersensing occurs.106 Unless VT is known to be highly regular, the number of intervals to detect VT probably should not be more than 50% greater than the nominal value in Medtronic ICDs. Duration for Detection of Ventricular Fibrillation Because the confirmation process is necessarily “trigger happy,” the first line of defense against inappropriate therapy for nonsustained VT or SVT in the VF zone is an appropriately long detection duration. Nominal values should be increased in patients who have long episodes of nonsustained device-detected VF (e.g., long QT syndrome). In Medtronic ICDs, nominal programming of the number of intervals for initial detection of

137

VF (18 of 24) substantially reduces inappropriate therapies without significantly delaying detection, compared with 12 of 16 intervals,130,133 another commonly used setting. Comparable settings for number of intervals or duration are available in ICDs from other manufacturers also. Duration for detection of VF should not be reduced from nominal if there is any alternative method to ensure reliable detection. Duration for Redetection Inappropriate therapy for nonsustained VT or SVT may be delivered after appropriate or inappropriate ATP or shocks (see Fig. 3-77). Increasing the duration for redetection may prevent inappropriate redetection of delayed termination of VT (type II break) or postshock nonsustained VT. However, excessive delays in detection or redetection may result in syncope, increase in defibrillation threshold, or undersensing caused by reduced amplitude and frequency of the sensed ventricular EGM. Fortunately, these adverse effects are rare for VF durations shorter than 30 seconds.134 ICDs misclassify effective therapy as ineffective if VT/VF recurs before the ICD identifies episode termination and reclassifies the post-therapy rhythm as sinus (Fig. 3-79). Misclassification may also occur due to limited SVT-VT discrimination during redetection if therapy successfully terminates VT during a double tachycardia (simultaneous SVT and VT) or if SVT begins after successful VT therapy but before episode termination because of frequent premature beats or nonsustained tachycardia. Decreasing the duration for redetection of sinus rhythm (St. Jude) may correct this classification error. However, this type of misclassification often does not constitute a clinical problem, whereas postshock detection of nonsustained VT does. An exception occurs when one iteration of ATP is programmed for the first therapy and a shock for the second. If ATP terminates VT, but it recurs before the rhythm is classified as sinus, the recurrent VT will receive a shock instead of a second potentially effective burst of ATP. Range of Cycle Lengths to which SVT-VT Discriminators Apply SVT-VT discriminators apply in a range of cycle lengths bounded on the slower end by the VT detection interval and on the faster end by a minimum cycle length that varies among manufacturers (see Fig. 3-49). Usually, SVT-VT discriminators will not withhold inappropriate therapy for SVT if the majority of ventricular intervals (typically 70% to 80%) are shorter than the SVT limit. Therefore, rapidly conducted AF may be classified as VT even if the mean cycle length is 20 to 40 msec longer than the SVT limit. Programming a sufficiently short minimum cycle length for SVT-VT discrimination is critical to reliable rejection of SVT. When discriminators are programmed, approximately 25% of inappropriate therapy is caused by SVT with ventricular cycle lengths shorter than that minimum cycle length.62,135

138

5 3 0

Section One: Basic Principles of Device Therapy

5 5 0

A R T F* 3 0 0

T F* 3 0 0

T F*

3 1 0

5 5 0

A S V P

4

5 5 0

A R

T F*

5 5 0

T F* 3 1 0

3 0 0

5 4 0

A S V P

5

5 2 0

A R

3 3 0

A R T S

0

T F*

5 2 0

A R 3 1 0

5 4 0

T F*

5 5 0

A R T P

A R

T P

T 2 P 6 0

FVT FVT Rx 1 Burst

A S V S

7 0 0

1

5 5 0

5 5 0

A S 6 2 0

V P

2

3 5 0

5 3 0

A R T S

0

5 0 0

V S

4 9 0 6 2 0

A R

5 5 0

2 T 5 F* 0

3 2 0

5 6 0

A S T P

1

5 6 0

T P

2

5 6 0

A R

2 T 2 8 F* 8 0 0

T 2 F* 9 0

A R T S

5 2 0

A S

1

5 3 0

A S 5 4 0

T S

3

5 5 0

A R

T 2 F* 9 0

T 3 F* 0 0

T S

3 4 0

A R V S

5 5 0

A S V P

7 1 0

0

5 5 0 T F

5 3 0

A R

1

5 4 0

A R

V S

V S

FVT FVT Rx 2 CV

3 5 0

T S

V S

V S

3 0 0

V S

5 9 0

1

5 6 0

A R

5 5 0

A R

3 7 0

0

5 2 0 3 0 0

5 3 0

A R

3 0 0

5 4 0

A R V S

3 0 0

6 5 0

3

1600

A R V S

V S

5 7 0

2

5 5 0

A R

V 1 C S 9 E 0

4 2 0

V S

3 0 0

C D 15.0

Figure 3-79. Failure to identify post-therapy sinus rhythm due to rapid reinitiation of ventricular tachycardia (VT) after successful therapy. Atrial EGM, ventricular true bipolar EGM, and EGM markers classify the rhythm as sinus (8 consecutive intervals). Therefore, VT is inappropriately redetected instead of being detected de novo for the second time. This results in delivery of the second programmed Fast VT (FVT) therapy, cardioversion, rather than repeat delivery of the previously successful antitachycardia pacing (ATP) stimulus, which is both painless and more energy efficient. Redetection marker (FV Rx 2 CV) indicates onset of capacitor charging for the second VT therapy. Shock is not shown. Large numbers below right side of upper panel and left side of lower panel show value of sinus rhythm counter, which increments for each interval in the sinus zone (VS) and is reset to zero by premature ventricular complexes (PVCs) with a coupling interval of less than the VT detection interval of 400 msec (TS). VS markers indicate unclassified ventricular intervals during capacitor charging. AS, atrial sense; AR, atrial intervals in pacing refractory period.

The performance of SVT-VT discriminators is linked explicitly or implicitly to boundaries between detection zones for ventricular arrhythmias. In Guidant ICDs, the link is explicit. Starting with the Vitality models (2004), SVT-VT discriminators are programmable to the entirety of either or both VT zones; previously, they were limited to the slower VT zone. In St. Jude ICDs, SVT-VT discriminators are programmable independently within the two VT zones but cannot be programmed in the VF zone. In Medtronic ICDs, the SVT Limit is programmable independently of VT/VF zone boundaries. However, the performance of the SVT rejection algorithm changes at the programmed VF detection interval, so that SVT with AV dissociation (AF) is not classified as SVT because it cannot be distinguished from VF. Further, Medtronic ICDs use consecutive-interval counting and other measures of R-R interval regularity to withhold inappropriate therapy for AF for rhythms with cycle length equal to or greater than the V-F interval. To ensure reliable detection of VF, they are not applied in the VF zone. Therefore, in the portion of the VF zone in which discriminators are applied, sinus tachycardia, AT, and 2:1 atrial flutter are rejected, but conducted AF is not. In nominal programming of Medtronic ICDs, the “VF Detection Interval” forms the boundary between the VT and Fast VT zones. Therefore, this degradation in discrimination of VT from rapidly conducted AF usually occurs between the VT and Fast VT zones. Conducted AF in the Fast VT zone may be classified correctly by programming “Fast VT via VT” rather than the

nominal “Fast VT via VF.” The risk is delay in detection of unusual, markedly irregular fast VT with occasional cycle lengths in the sinus zone. Programming of SVT-VT Discriminators Single-Chamber SVT-VT Discriminators Technical details vary among manufacturers, as do corresponding recommended programmed values, which are summarized in Table 3-5. See “Ventricular EGM Morphology for SVT-VT Discrimination” for programming and troubleshooting of morphology algorithms. Dual-Chamber SVT-VT Discriminators Dual-chamber algorithms should be programmed ON in any patient with intact AV conduction and a functioning atrial lead. Specific considerations for each manufacturer are summarized in Table 3-6. In Medtronic and Guidant algorithms, incremental addition of discriminators increases the likelihood that SVT will be classified correctly (specificity) but decreases the likelihood that VT will be classified correctly (sensitivity). In St. Jude ICDs, discriminators may be combined using either the “ANY” or “ALL” operators. Using ANY, the algorithm detects VT if any discriminator classifies the tachycardia as VT, resulting in higher sensitivity and lower specificity. Conversely, using ALL, the algorithm detects VT only if all discriminators classify the tachy-

Chapter 3: Sensing and Detection

139

Recommended Programming of SVT-VT Discriminators in Single-Chamber ICDs TABLE 3-5.

Medtronic

Guidant

St. Jude

Stability*

40-50 msec, NID = 16

24-40 msec, duration 2.5 sec

80 msec

Onset

84-88%

9%

150 msec

Morphology

3 of 8 electrograms ≥70% match

Not programmable†

5 of 8 electrograms ≥60% match

*Less strict values are required for patients taking type I or III antiarrhythmic drugs. † 3 of 10 electrograms with Feature Correlation Coefficient greater than threshold. NID, number of intervals to detect VT.

Recommended Programming of SVT-VT Discriminators in Dual-Chamber ICDs TABLE 3-6.

Guidant Medtronic PR Logic

Atrial View

Rhythm ID

St. Jude Rate Branch

AFib/AFlutter ON

AFib Rate Threshold

ON

Rate Branch ON

Sinus Tach ON

200 bpm

A = V Branch: Morphology

Other 1:1 SVTs

Onset 9%

A > V Branch Morphology; may combine Stability† with “ANY” logic

OFF 1:1 VT-ST

Inhibit If unstable 10%

Boundary 66%*

V rate > A rate ON Sustained Rate Duration 3 min

*Older models before Entrust (Model D153DRG). † Stability at 80 msec with AV Association of 60 msec.

cardia as VT, resulting in lower sensitivity and higher specificity. The ALL operator corresponds to addition of discriminators in other algorithms. St. Jude Rate Branch. Morphology should be programmed in both the V = A and V < A rate branches. Recommended programming adds the stability discriminator in the V < A branch using the ANY operator. This results in a minor increase in sensitivity for detection of VT (98% to 99%) and a similarly minor decrease in specificity (82% to 79%).102 Addition of the stability discriminator using the ALL operator reduces inappropriate detection of aberrantly conducted AF but may also reduce sensitivity for detection of VT.102 Guidant Rhythm ID. This algorithm uses atrial versus ventricular rate, EGM morphology, and interval stability to discriminate VT from SVT,110 the same three general features used by the St. Jude algorithm. It has no programmable features. The benefit of this algorithm’s design is that it requires no custom programming; the limitation is that troubleshooting is not possible. Guidant Atrial View. A major limitation of this earlier algorithm, inappropriate detection of rapidly conducted AF61,135,136 due to obligatory PVAB, may be ameliorated by programming the PVAB period to the minimum value of 45 msec, the “AFib Rate” to the minimum value of 200 bpm, and the “Stability” feature to the

highly specific value of 10%. This programming takes advantage of the fact that VT during AF or atrial flutter usually is highly regular and often is more regular than conducted 2:1 atrial flutter. However, highly regular 2:1 conduction of atrial flutter will be misclassified as VT, and slightly irregular VT that occurs during AF will be misclassified as SVT—especially in the setting of antiarrhythmic drugs.106,137 There are no programming solutions for inability to detect VT with both 1:1 VA conduction and a gradual onset or for inability to reject 1:1 AT with abrupt onset. Medtronic PR Logic. At implantation, rejection rules should be programmed “ON” for Sinus Tachycardia and Atrial Fibrillation/Flutter. The 1:1 SVT rejection rule should not be programmed until the atrial lead is stable, because its dislodgment to the ventricle may result in misclassification of VT as a 1:1 SVT. (This potential problem also applies to the St. Jude Rate Branch algorithm without additional discriminators.) PR Logic uses the patterns and rates of A-A, V-V, A-V, and V-A intervals to discriminate VT from SVT. This dependence, combined with the absence of atrial blanking periods after sensed ventricular events, made several generations of this algorithm susceptible to errors based on intermittent oversensing of FFRWs. These versions of PR Logic (still in use in cardiac resynchronization ICDs) also discriminate VT with 1:1 VA conduction from SVT,

140

Section One: Basic Principles of Device Therapy

based on the ratio of P-R to R-R intervals. Increasing the value from the nominal setting of 50% to 66% reduces inappropriate therapy for 1:1 SVT with long P-R intervals without significantly increasing the risk of failing to detect VT with 1:1 VA conduction.138 The present generation of Medtronic ICDs includes a new algorithm for identifying sinus tachycardia that reduces inappropriate detection of VT due to FFRWs and long P-R intervals in sinus tachycardia. It also reduces misclassification of VT with 1:1 VA conduction and a long R-P interval as sinus tachycardia.112

Undersensing and Underdetection Undersensing and underdetection may be caused by ICD system performance, programmed values (including human error), or a combination of the two. They result in failure to delivery therapy or delay in therapy. Undersensing VF may be undersensed due to combinations of programming (sensitivity, rate, or duration), low-amplitude EGMs, rapidly varying EGM amplitude, drug effects, and postshock tissue changes. Clinically significant undersensing of VF is rare in modern ICD systems if the baseline R-wave amplitude is 5 to 7 mV or higher.54 Postshock undersensing was an important clinical problem in older ICD systems that used integrated bipolar leads with closely spaced electrodes.139 Presently, the most common causes of VF undersensing are drug or hyperkalemic effects that slow VF into the VT zone, ischemia, and rapidly varying EGM amplitude (Fig. 3-80).48,65 ICDs that adjust dynamic range based on the amplitude of the sensed R wave (Guidant) may be the most vulnerable to the extremely rare problem of rapidly varying EGM amplitude48 (Fig. 3-81). Prolonged ischemia from sustained VT slower than the VT detection interval may cause deterioration of signal quality, resulting in undersensing of VF. Lead, connector, or generator problems may also manifest as undersensing. ICD Inactivation If detection is programmed OFF for surgery using electrocautery, reprogramming must be performed at the end of the procedure, a fact that is easily forgotten, especially without patient surgery. One study reported an unexplained 11% annual incidence of transient suspension of detection.140 This unfortunate problem is addressed in Medtronic Marquis and subsequent ICDs with an audible patient alert that sounds if programmed detection or therapy is “OFF” for longer than 6 hours. Ventricular Tachycardia Slower than the Programmed Detection Interval In most ICD patients, VT with cycle lengths greater than 400 to 450 msec are tolerated well, but repeated

inappropriate therapies are not. However, slow VT can be catastrophic in patients with severe LV dysfunction or ischemia.141 All SVT-VT discrimination algorithms (except those in ELA models65,118,142) deliver fewer inappropriate therapies if the VT detection interval is programmed to a shorter cycle length, simply because fewer SVTs are evaluated. A long VT detection interval is important in patients with advanced heart failure, in whom slow VT can be catastrophic141 (Fig. 3-82; see Fig. 3-14). The VT detection interval should be increased if antiarrhythmic drug therapy is initiated, particularly with amiodarone or a sodium-channel blocking (type 1A or 1C) drug.14,143 It may be prudent to measure the cycle length of induced VT at electrophysiologic testing after initiation of drug therapy.143 However, spontaneous VT often is slower than induced VT.144 SVT-VT Discriminators SVT-VT discriminators may prevent or delay therapy if they misclassify VT or VF as SVT.102,103,106,145 Discriminators that re-evaluate the rhythm diagnosis during an ongoing tachycardia (e.g., stability, most dual-chamber algorithms) reduce the risk of underdetection of VT compared with discriminators that withhold therapy if the rhythm is not classified correctly by the initial evaluation (e.g., onset, chamber of origin algorithms). The minimum cycle length for SVT-VT discrimination should be set to prevent clinically significant delay in detection of hemodynamically unstable VT. See earlier discussions of discriminators and programming. Pacemaker-ICD Interactions Although interactions between ICDs and separate pacemakers have become rare since ICDs incorporated dualchamber bradycardia pacing in the late 1990s, some combined systems have not been revised due to vascular access problems or other reasons. The multiple potential interactions have been reviewed, and testing protocols to detect them have been developed.146-148 The principal interaction that may delay or prevent ICD therapy is oversensing of high-amplitude pacemaker stimulus artifacts. If this occurs during VF, repetitive automatic adjustment of sensing threshold and/or gain may prevent detection of VF. Intradevice Interactions Presently, intradevice interactions, in which bradycardia pacing features of dual-chamber ICDs interact with and impair detection of VT or VF, pose a greater challenge than pacemaker-ICD interactions between separate devices.51 During high-rate, atrial or dual-chamber pacing, sensing may be restricted to short periods of the cardiac cycle because of the combined effects of ventricular blanking after ventricular pacing and crosschamber ventricular blanking after atrial pacing, which is needed to avoid crosstalk. If a sufficient fraction of the cardiac cycle is blanked, systematic undersensing of VT or VF may occur. When pacing and blanking

141

Chapter 3: Sensing and Detection

VF 248

VF 158

Pre-attempt EGM (10 sec max) VF Zone Initial Detection 176 bpm Pre-attempt AVS Rate

Near field V

VF 253 Epsd Grad 1

Far field V

VF 163

VF 265

VF 163

VF 263 VF 155

VF 280

VF 155

VF 270

VF 163

VF 270 VF 168

VF 265

VF 158

VF VF 283 270 VF VF 163 165

VF 278 VF 143

VF 303 VF 135

VF 305 VF 165

VF 280 VF 168

VF 278 VF 165

VF 290 VF 165

VF 278

VF 175

VF 278

VF 160

VF 300

VF 163

VF 298

VF 168

–– 533

VF 373 –– 98

VF 138 VS 465

VS 335 Dvrt

Chrg

End of Episode 00.25

Pre-attempt EGM (10 sec max) Initial Detection 218 bpm Pre-attempt AVS Rate

25 mm/s

VF –– 228 VF 165 230

VS 365

VS 478

VF 140

VS 345

VF 143

VS 345 VF 145

VS 353

VF 153

VS 353

VF 145

VT 310

VF 170

VS 355

VF 145

VS 368

VF 138

VT 315

VF 135 VS 425

VT 310

VF 173

VS 345

VF 145

VF 158 VS 355

VS 368

VS 543

VS 625

VS 578

A

VF 230

VF 228 Detct Chrg

VF 218

VF 190

VF 188

VF VF VF 135 223 225 VF VF VF VF 188 258 155 195

VS 383

VF 288

VF 243

VF 188

VF 178

VF 280

VF 218

VF 243

–– 170

Chrg

–– 223

VF 170

1

VF 260

VF 180

Shock

Attempt

VF 230

Attempt Type Elapsed Time Therapy Delivered

VF 208

14J Biphasic 00:01 VF Shock 1

Far field V

Pre-attempt EGM (10 sec max) Initial Detection 60 bpm Pre-attempt AVS Rate

Near field V

–– 1363

B Figure 3-80. Underdetection of ventricular fibrillation (VF) caused by hyperkalemia (potassium level, 6.7 mg/dL) in the setting of chronic amiodarone therapy. A, Near-field and far-field ventricular (V) EGM and event markers of a Guidant Prizm VR ICD are shown. The top panel shows a sinewave ventricular tachycardia (VT) in the far-field channel, which is detected as VF because local EGMs on the near-field channel are double-counted. At right of upper panel, four intervals greater than the programmed VF detection interval of 316 msec (190 bpm) result in an aborted shock. The lower panel shows persistence of VF after the shock is aborted. Too few intervals are sensed in the VF zone to permit detection of VF. The patient was resuscitated by external shock. B, ICD system testing after correction of hyperkalemia. Induced VF is sensed reliably and is terminated by an ICD shock. The near-field EGM is narrow, indicating that double EGMs present during the clinical arrhythmia were caused by functional conduction block. (Courtesy of Dr. Felix Schnoell.)

VS 1573

142

Section One: Basic Principles of Device Therapy

ECG

RA

RV (Tip-Coil)

(AS) 1343

PVC 303 VF 265

(AS) 1268

(AS) 1215 PVC 580

PVP

VS PVC 305 310

PVC 283

PVC 483

PVP

(AS) 1185

AS 1158

PVC VP-MT 320 VF 523 170 PVP

VP 555

PVC PVC 288 VF 330 273 PVP

VS 31

PVP

Figure 3-81. Undersensing of ventricular fibrillation (VF) despite normal R wave in sinus rhythm (18.5 mV). Programmer strip recorded at implantation testing of a Guidant Prizm ICD shows electrocardiogram (ECG), right atrial (RA) EGM, and integrated bipolar ventricular sensing EGM (RV TipCoil) during implantation testing. The VF EGMs have highly variable amplitudes, resulting in undersensing of low-amplitude EGMs immediately after high-amplitude ones (arrows). Intermittent ventricular pacing (VP) introduces postpacing blanking periods. Slow Automatic Gain Control may contribute to this type of undersensing. AS, atrial sensed event; PVC, premature ventricular complex; PVP→, extension of postventricular atrial refractory period (PVARP); VP, ventricular paced event. (Modified from Dekker LR, Schrama TA, Steinmetz FH, et al: Undersensing of VF in a patient with optimal R wave sensing during sinus rhythm. PACE 27(6 Pt 1):833-834, 2004, with permission.)

events occur at intervals that are multiples of a VT/VF cycle length, ventricular complexes are repeatedly undersensed, delaying or preventing detection (Fig. 3-83).49-51 Although intradevice interactions are uncommon, they have been reported most frequently with the use of the Rate Smoothing algorithm in ICDs.49-51 This algorithm is intended to prevent VT/VF initiated by sudden changes in ventricular rate.149 It prevents sudden changes in ventricular rate by pacing both the atrium and the ventricle at intervals based on the preceding (baseline) R-R interval. As an unintended consequence, it may prevent sensing of VT/VF in some patients, because it introduces repetitive postpacing blanking periods. The algorithm applies rate smoothing to baseline intervals independent of their cycle length, including intervals in the VT or VF zones. Intradevice interactions that result in delayed or absent detection of VT/VF are most common and most dangerous when VT is fast. The parameter interrelationships that result in delayed or absent detection of VT/VF are complex and difficult to predict, but they usually elicit a programmer warning. Generally, aggressive rate smoothing (a small allowable percentage change in R-R intervals), a high upper pacing rate, and a long and fixed AV interval favor undersensing and should be avoided. If rate smoothing is required, the AV delay should be dynamic, the upper rates should

be 125 bpm or less, and parameter combinations that result in warnings should be avoided. This programming reduces, but does not eliminate, the risk of undersensing.49-51,150

Detection of SVT and VT as a Diagnostic Tool and as a Basis for Atrial Antitachycardia Pacemakers and Atrial ICDs Detection of SVT and VT provides diagnostics that are useful for pacemaker management.80 Atrial ATP and atrial shocks provide a therapeutic option for some patients with paroxysmal atrial tachyarrhythmias. Appropriate delivery of this therapy requires accurate detection and discrimination of atrial tachycardia/ flutter (AT) and AF. Pacemaker Diagnostics Monitoring for Ventricular Tachycardia Some pacemakers have ventricular high-rate diagnostics that trigger EGM and/or marker storage of nonpaced rhythms that are faster than a specific rate. Triggers based on ventricular rate alone store episodes

Chapter 3: Sensing and Detection

143

Atrial EGM RV Coil-Can

Interval (ms) 1800 VF Detected

1720

3 5 0

1760

A P T S

6 8 0

VV SP

4 2 0

V S

F F F F F F S S S S S S 2 1 1 2 1 1 1 8 2 0 2 6 0 0 0 0 0 0

4 6 0

V S

2 1 0

1600

A P F S

7 2 0

VV SP

4 7 0

V S

1 7 0

F S

4 7 0

V F S S 1 3 0

4 6 0

V S

2 1 0

F S

1 8 0

F S

1 8 0

F S

2 3 0

F S

2 2 0

F S

F F F F V V S S S S S S 2 1 1 1 4 2 5 3 8 8 4 0 0 0 0 0 0 0 VF Rx 1 Defib

5 1 0

V S

4 6 0

V S

1400 1200 1000

3.7 minutes

800 600

TDI

400

FDI 200 0 600

400

200

Time Before VF Detection (Seconds)

0 VF Detected

Figure 3-82. Ventricular tachycardia (VT) slower than the programmed detection interval. The lower panel is a “Flashback Interval” plot of R-R interval cycle lengths before detection of ventricular fibrillation (VF), which occurs at the right side of each panel. The interval number before detection is plotted on the abscissa, and the corresponding interval is plotted on the ordinate. Horizontal lines indicate the VT detection interval (TDI) of 400 msec and the VF detection interval (FDI) of 320 msec. Shortly after the 500th interval preceding detection, regular tachycardia begins abruptly. The constant cycle length indicates reliable ventricular sensing. Atrial flashback intervals (not shown) demonstrated atrioventricular (AV) dissociation. This VT is not detected despite reliable sensing, because the cycle length is greater than the programmed TDI. VT persists for 3.7 minutes until approximately interval 280 before detection, when sensed intervals become highly variable. This indicates degeneration of the rhythm to VF with undersensing that delays detection. During VT and VF, atrial Flashback Intervals (not shown) indicated lower rate limit bradycardia pacing at 40 bpm (1500 msec). The upper panel shows stored atrial and far-field ventricular EGMs immediately before detection with atrial and ventricular channel showing event markers. Specific undersensed EGMs cannot be identified because the rate-sensing EGM was not recorded. However, long sensed R-R intervals ending with VS (ventricular sense) markers indicate undersensing and correspond to long intervals in the upper panel. “VF Therapy 1 Defib” at lower right (arrow) denotes detection of VF. AP, atrial pace; FD, VF detected; FS, intervals in VF zone. (From Swerdlow C, Friedman P: Advanced ICD Troubleshooting: Part I. PACE 28:1322-1346, 2005, with permission.)

that may be rapidly conducted SVT, ventricular oversensing, or VT. Studies of these diagnostics demonstrate the importance of stored EGMs to confirm the diagnosis, because the false-positive detection rate is high.82,151 Newer dual-chamber pacemakers and antitachycardia pacemakers incorporate tachyarrhythmia detection algorithms that are identical or substantially identical to those in dual-chamber ICDs (Fig. 3-84). The performance of these algorithms may be influenced by differences in pacemaker and ICD atrial sensing characteristics (fixed atrial sensitivity, automatically adjusting sensitivity, or an intermediate case) and atrial blanking periods (Figs. 3-85 and 3-86). Even with accurate atrial sensing, Bayes theorem predicts that the fraction of false-positive VT detections in

patients with pacemakers is large, because the incidence of true VT is low compared with most ICD patient groups. Monitoring for Atrial Tachycardia and Atrial Fibrillation Traditional pacemaker diagnostics for atrial high-rate episodes and automatic mode switches have been validated as a means of monitoring AT/AF.81,152 Diagnostic accuracy of these methods depends on the specific detection algorithm used and the accuracy of atrial sensing. Detection algorithms for atrial antiarrhythmic pacemakers and ICDs provide substantial additional information, including atrial EGMs, classification of AT

144

Section One: Basic Principles of Device Therapy Surface ECG Atrial EGM Vent EGM

AS 1875 VS 935

VP 350

VS 350

VP 350

VP 350

AS 970

VP 350

VP 350

Surface ECG Atrial EGM Vent EGM

25-FEB-00 19:30

AP 863 VF 283 VF 280

VP 350

AP 268 VP 560

AP 863 VF 293

VF 283

AP 1630 VP 350

VP 270

VP 250

VP 230

PVC 448

VS 575

AP 863 VF 283

AP 863 VS 575

VF 283

VF 285

Surface ECG Atrial EGM Vent EGM

25-FEB-00 19:30

VT 415

VF 308

AP 875

AP 723 VP 560

VF 305

VP 570

VF 280

AP 1118 VF 278

AP 860

AP 260 VP 560

VF 298

VP 570

VF 285

AP 850 VP 560

AP 1135

AP 865 VP 580

VF 288

GUIDANT

VP 570

VF 280

Surface ECG Atrial EGM Vent EGM

25-FEB-00 19:30

VENTAK PRIZM

AP 863 VP 570

25 mm/s

VENTAK PRIZM

PABP VBP

VF 280

VP 560

AP 870 VF 300

AP 855 VF 275

AP 853 VS 573

VF 278

VF 283

AP 858

AP 260 AP 560

VF 788

VP 570

External rescue

AP 878 VP 570

VF 298

VP 580

AP 880 VF 290

AP 8737 VS 583

VF 288

VF 293

AP 260 [VS]

VP 560

[AS] (AS) 733 VF 308

Figure 3-83. Failure to detect ventricular tachycardia (VT) due to an intradevice interaction. The rate-smoothing algorithm introduced atrial and ventricular pacing complexes with associated blanking periods that prevented detection of VT during postimplantation testing. An external rescue shock was required. Shown from top to bottom are the surface electrocardiogram (ECG), atrial EGM, ventricular EGM, and event markers. At top, VT is induced by programmed electrical stimulation with a drive cycle length of 350 msec and premature stimuli at 270, 250, and 230 msec (intervals labeled next to event markers). The first sensed ventricular event occurs 448 msec after the pacing drive (PVC 448). The rate-smoothing algorithm drives pacing to prevent a pause after the premature ventricular complex (PVC), labeled AP↓1638. A ventricular paced event does not follow the first AP↓ because a ventricular event is sensed (VT 415). Subsequent rate smoothing generated atrial and ventricular pacing pulses (indicated by AP↓ and VP↓ markers, respectively). The resultant postpacing blanking periods are shown on the figure as horizontal bars. PABP denotes cross-chamber (postatrial pacing) ventricular blanking period. VBP denotes same-chamber (postventricular pacing) blanking period. Together, they prevent approximately four of every six VT complexes from being sensed. Because the VT counter must accumulate 8 out of 10 consecutive complexes in the VT zone for detection of VT to occur, VT is not detected. (From Swerdlow C, Friedman P: Advanced ICD Troubleshooting: Part II. PACE 29:70-96, 2006, with permission.)

versus AF, and integrated displays of data from multiple episodes. These include histograms of episode duration and ventricular rate during AT/AF that may be of value in managing atrial antiarrhythmic drugs and AV nodal blocking drugs (Fig. 3-87). The former histogram, combined with patient-activated device interrogations (Fig. 3-88), may permit discontinuation of anticoagulation in some patients. Detection of Atrial Tachycardia and Atrial Fibrillation Principles Bradycardia pacemakers must detect AT/AF rapidly so that mode switching will avoid uncomfortable pacing at the upper rate limit. To determine the atrial rate and rhythm accurately, atrial blanking must be minimized. As long as the atrial rate exceeds the upper tracking limit, accurate determination of atrial rate and rhythm is not important. Ventricular ICDs must discriminate

VT from rapidly conducted AF quickly, to permit rapid detection of hemodynamically compromising VT. Discrimination of AT versus AF is less important and less urgent. In contrast, atrial antiarrhythmic pacemakers and ICDs must detect AF with high specificity to minimize painful and potentially proarrhythmic therapy. Rapid detection is not important, because AF usually is clinically stable and may terminate spontaneously after hours to days. Therefore, atrial antiarrhythmic devices should be capable of permitting therapy for long-duration AT/AF while withholding therapy from self-terminating AT/AF. To achieve this goal, they must be capable of detecting AF continuously for extended periods, in order to discriminate between repetitive self-terminating episodes and persistent episodes. Because atrial EGMs in AF have low and variable amplitudes and slew rates, antiarrhythmic devices must have high (automatically adjusting) atrial sensitivity and apply algorithms that are tolerant of some atrial undersensing. They must also discriminate between AT and AF to deliver ATP for AT.

Chapter 3: Sensing and Detection V-V

A-A

VTM=400 ms

Detection

Interval (ms)

1500 1200 900 600

Term. 9 sec

AEGM VEGM

400 A P

200

30

25

20

15

10

5

0

5

V S

0

A R

1180 5 6 0

Time (sec)

1020

V 3 V 4 V S 7 S 5 S 0 0

6 9 0

9 8 0

A S

A R

9 4 0

A R

9 6 0

A R

9 2 0

A R

8 9 0

A R

8 7 0

A R

EGM for

V 5 V 3 V 3 V 3 V 3 V3 V 3 V3 V 3 V 3 V 3 V 3 V 3V 3 V 3 V3 V 3 V S 0 S 8 S 9 S 7 S 8 S3 S 4 S5 S 4 S 6 S 6 S 7 S 4S 6 S 9 S3 S 6 T 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 VTM

A V-V

A-A

VTM=400 ms

Interval (ms)

Detection

Term. 1.4 min

1500 1200 900 600

AEGM

0.5mV

VEGM

400 2000

200 30

25

20

15

10

5

Time (sec)

V S 0

5

0

5 6 0

V S

5 1 0

A b

2000

A b

EGM for 1

V V V V V V V V V V V V V V V V V V S 4 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 S 3 T 1 7 8 7 9 8 7 7 6 8 7 9 5 0 0 8 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 VTM

B

% of False-Positive VT/VF episodes

Figure 3-84. Pacemaker diagnostics for monitoring of ventricular tachycardia (VT), showing stored intervals, atrial EGM (AEGM), ventricular EGM (VEGM), and markers from a patient with a Medtronic EnRhythm DR pacemaker. The left panels show the A-A and V-V intervals plotted against time before VT detection, which requires 16 consecutive intervals at less than the VT monitoring interval of 400 msec (horizontal line). The right panels show dual-chamber EGM and markers. A, Onset of spontaneous VT occurs with several premature ventricular complexes (PVCs) and sudden acceleration of ventricular rate with little or no change in A-A intervals (left panel). The VT cycle length is initially 340 msec and progressively lengthens to greater than 400 msec after several seconds. After 8 consecutive intervals of 400 msec or longer, the device declares the episode to be “terminated” despite clear atrioventricular (AV) dissociation and ongoing VT (right-hand side of interval plot). The duration of VT is reported as 9 seconds, representing the time that the R-R intervals remained less than 400 msec after initial detection. Dual-chamber EGM and markers leading up to detection of VT indicate appropriate sensing in both chambers and AV dissociation (right panel). B, Rapidly conducted atrial fibrillation with severe atrial undersensing reported as VT. The interval plot shows fast and irregular R-R intervals and long A-A intervals (1500 msec), with sporadic A-A intervals shorter than 200 msec. Dual-chamber EGMs indicate appropriate ventricular sensing but severe atrial undersensing of small-amplitude atrial fibrillation, resulting in falsepositive detection of VT. AB, atrial sense in postventricular atrial blanking period; AR, atrial refractory sense; AS, atrial sense; VS, ventricular sense; VT, VT detected.

100%

80%

Patients with AT/AF history and no VT/VF history

60%

40%

20%

Secondary prevention ICD patient population

0%

Higher incidence of true VT/VF relative to SVT Figure 3-85. Rate of inappropriate ventricular tachycardia/fibrillation (VT/VF) detection depends on the patient population. Bayes theorem predicts that patients with a higher incidence of true VT/VF relative to supraventricular tachycardia (SVT) will have fewer inappropriate detections than patients with a lower incidence of VT/VF. The receiver operator curve shows the Bayes theorem prediction of detection performance of an algorithm with a fixed sensitivity and specificity. The curve plots percentage of false-positive detections of VT/VF on the ordinate and estimated incidence of true VT/VF on abscissa. Two data points are plotted from clinical studies of two different patient populations with implanted devices running the same VT/SVT discrimination algorithm (Medtronic GEM DR62 and AT500 devices177 with PR Logic).

145

146

Section One: Basic Principles of Device Therapy

Lead II

Atip-Aring

AR

AR

FS FS FS RR PP Median Episode Atrial Evidence Rx Delivered VF Count

AR

FS

FS

AR

AR

AR

AR

AR

AR

AR

AR

AR

FS

AP

FS 290 300 0

FS 280 300 0

FS 300 300 0

FS 300 300 0

FS 300 300 0

FS 290 300 0

FS 290 300 0

FS 300 300 0

FS 300 300 0

FS 300 300 0

VS 610 300 1

24

24

24

24

24

24

24

24

24

24

23

Figure 3-86. Transient short ventricular cycle lengths before spontaneous termination of rapidly conducted atrial tachycardia/fibrillation (AT/AF). Telemetry Holter monitor shows electrocardiographic (ECG) lead II, atrial EGM, and dual-chamber markers along with multiple channels indicating rhythm classification and status of Medtronic Jewel AF Model 7250 atrial/ventricular ICD. Conduction of this AT episode was 1:1, driving ventricular rates up to 200 bpm before spontaneous termination. Most atrial antiarrhythmic devices detect atrial arrhythmias only if the atrial rate exceeds the ventricular rate. Therefore, tachycardias with 1:1 AV relationship are classified as ventricular in origin. In this patient, the ventricular fibrillation (VF) detection interval was programmed to 320 msec with VF therapies OFF. This rhythm was classified as VF, as indicated by the values of the Atrial Evidence Counter (zero), and the VF Count (24, maximum value). AR, atrial refractory event; FS, event in fibrillation zone of either channel; VS, ventricular sensed event.

Detection of Atrial Tachycardia/Fibrillation for Atrial Therapy High specificity in AT/AF detection has been achieved by multistep methods for detection of AT/AF. Initial detection is based on the presence of an atrial tachyarrhythmia and absence of VT. This is achieved by a measure of atrial rate combined with either comparison of atrial and ventricular rates (Guidant)153 or use of A:V patterns to identify N:1 rhythms (Medtronic) (Fig. 3-89). The tradeoff between atrial undersensing and oversensing of FFRWs that applies to dual-chamber ICD algorithms is even more important when considering detection of AT/AF for atrial therapy. Minimization of atrial blanking is important to prevent undersensing of AF, and rejection of FFRWs is important to prevent inappropriate detection of AT/AF. (See “Atrial Sensing in Dual-Chamber ICDs and Atrial ICDs.”) Once initial detection occurs (Fig. 3-90), the episode timer begins, initiating a sustained-detection mode in which AT/AF remains detected despite a moderate degree of undersensing (Fig. 3-91). When this timer expires, atrial therapy is delivered if AT/AF still persists (Fig. 3-92). Atrial episodes must be interrupted if true VT occurs (Fig. 3-93).

The final step in detection is discrimination of AT from AF or, alternatively, determining whether the rhythm is likely to respond to atrial ATP. Typically, there may be one timer for ATP and one for shock therapy. In newer Medtronic ICDs, the “reactive antitachycardia pacing” algorithm resets the ATP timer to permit repeat attempts at ATP if changes in atrial rate or atrial rhythm regularity are detected, or after a preprogrammed duration of sustained AT/AF. Changes in rate or regularity of the atrial rhythm are classified as a shift to a new rhythm, which may be more susceptible to ATP termination (Fig. 3-94). Additional ATP attempts hours after failed ATP may be more successful due to shifts in the patient’s autonomic tone or physiologic changes that render the rhythm more likely to be terminated with ATP. As shown in Figure 3-89, Medtronic atrial ICDs discriminate between AT and AF based on the median rate and regularity of the atrial rhythm. Guidant atrial ICDs use a more complex scheme based on maximum rate, standard deviation, and range of the 12 most recent atrial cycle lengths to plot a point in a three-dimensional space. A decision boundary divides the space into two regions: faster/irregular atrial cycle lengths (AF) and Text continued on p. 152

147

Chapter 3: Sensing and Detection

Ventricular Response (in AF) Long-Term Trends May 00

Jul 00

Program/Interrogate

Sep 00

# beats

Nov 00

P

Drug Change

Jan 01

Mar 01

P

May 01

80 bpm

0

(0%)

I

80-100 bpm

3

(0%)

100-120 bpm

124

(4%)

120-140 bpm

175

(6%)

140-160 bpm

344

(12%)

160-180 bpm

912

(33%)

180-200 bpm

880

(32%)

200

319

(12%)

Sotalol

CV/Ablation/Other AT/AF Patient Check AT/AF total hrs/day

24 20 16 12 8 4 0

AT/AF 25 20 episodes/day 15 10 5 0 ATP Change Treated AT/AF 25 20 episodes/day 15 10 5 0 % ATP Success

Data - AT/AF Episode Duration Histogram % of Episodes 100

100 75 50 25 0

ATP ON 80

Antici-Pace Change % Pacing/day — A.Total — A. ARS/APP

% Pacing/day

60 100 75 50 25 0 100 75 50 25 0

40 20 0 < 1 min 10 min 1 hr 4 hr 12 hr 24 hr 72 hr May 00

Jul 00

Sep 00

Nov 00

Jan 01

Mar 01

May 01

Episode Duration

Figure 3-87. Diagnostic data stored in atrial antiarrhythmic pacemakers and ICDs assist medical management of atrial arrhythmias. Data are taken from three patients with Medtronic AT500 pacemakers. Left panels, Long-term trend data over 1 year. Panels from top to bottom show total hours per day in atrial tachycardia/flutter (AT) or atrial fibrillation (AF), total AT/AF device-detected episodes per day, number of daily episodes treated by antitachycardia pacing (ATP), percentage of ATP therapies classified as “successful” by the device, percent atrial pacing, and percent ventricular pacing. ATP ON denotes activation of atrial ATP. Note that, although the many episodes of AT/AF are treated and this treatment is usually classified as successful, there is no detectable change in the number of hours per day of AT/AF until sotalol therapy is initiated. Sotalol decreases the total hourly “burden” of AT/AF more by shortening episodes rather than by preventing them. Percent ventricular pacing is high throughout but increases after sotalol treatment to almost 100%. Top right panel, Ventricular rate during AT/AF as percentage of time spent in AT/AF. Seventy-seven percent of intervals are shorter than intervals corresponding to 160 bpm, and 12% are shorter than intervals corresponding to 200 bpm, indicating inadequate control of ventricular rate. Lower right panel, Histogram of durations of AT/AF episodes since the last follow-up visit. This patient had 10 episodes of AF since the last visit, but only 1 episode lasted longer than 10 minutes.

>

148

Section One: Basic Principles of Device Therapy

Query AT/AF in progress? Logs date/time and V-rate (if AT/AF)

Figure 3-88. Handheld devices used by patients with combined atrial/ventricular ICDs communicates with the ICD via telemetry. They permit patients both to determine their atrial rhythm diagnosis and to deliver atrial shocks. To use this device in conjunction with the Medtronic GEM III AT ICD, the patient initiates an interrogation by pushing the blue query (“?”) button. The ICD responds by illuminating one of four colored light-emitting diodes (LEDs): Therapy Pending, AF present, No AF present, or Call Physician. On interrogation of the ICD with the standard programmer, a date/time log of patient queries is available to help correlate symptoms and rhythm. The log at right shows date/time of patient queries in the left column, ventricular rate in the center column, and presence or absence of an AT/AF episode in the right column. This patient, who was not anticoagulated, queried his ICD once or twice on most days.

Chapter 3: Sensing and Detection

Median P-P Interval

AF Detection Zone

Atrial Blanking (100 mS)

Auto Discrimination Zone AFDI: 280 mS AFDI: 320 mS

ATDImin: 180 mS AT Detection Zone

AF/AT Evidence Counter

Atrial Fibrillation

P R

A

B

 Count: 1 Count: 28

1 29

A=A range

1 30

1 29

1 30

1 31

1 32

SD

Figure 3-89. Criteria for detection of atrial tachyarrhythmias. A, Medtronic atrial therapy devices (Jewel AF, GEM III AT, AT500, and EnRhythm DR) use a combination of atrial cycle length and atrial/ventricular (A:V) patterns to detect atrial fibrillation and atrial flutter. The atrial tachycardia (AT) and atrial fibrillation (AF) detection zones are based on median atrial cycle length (12 P-P intervals) and may overlap. The overlap region is the “autodiscrimination” zone, where regularity of the P-P intervals determines whether rhythm classification is AT (regular P-P intervals) or AF (irregular P-P intervals). Detection of AT or AF requires (1) that the median P-P interval is in one of the detection zones AND (2) that the rhythm is N:1 as determined by the AF/AT evidence counter. The AF/AT evidence counter is an up-down counter (minimum value, 0; maximum value of detection threshold, +15). The counter increments by 1 on each ventricular event if there are two or more atrial events and there is no pattern-based evidence of far-field R-wave oversensing on the atrial channel (see Fig. 3-27A), and it decrements by 1 if there is strong evidence of lack of N:1 rhythm (e.g., two consecutive 1:1 beats). Isolated 1:1 beats contribute AF/AT evidence if they are preceded by a confirmed N:1 beat. The AF/AT counter detection threshold is between 24 and 32, depending on the specific device. AFDI, atrial fibrillation detection interval; mS, msec. B, The Atrial Rhythm Classification (ARC) algorithm (Guidant) discriminates AF from atrial flutter (AFL) based on the atrial rate and two measurements of variability of atrial rate: the range of the atrial intervals (i.e., the difference between the longest and shortest A-A intervals [A-A range]) and the standard deviation of A-A intervals (SD). The values of these three variables define a point in a three-dimensional space. A curved surface separates the AF region from the AFL region. Points in the AF region have higher atrial rates, a higher range of A-A intervals, and a greater standard deviation of A-A intervals. (From Morris MM, KenKnight BH, Lang DJ: Detection of atrial arrhythmia for cardiac rhythm management by implantable devices. J Electrocardiol 33(Suppl):133-139, 2000, with permission.)

149

150

Section One: Basic Principles of Device Therapy

ECG

AEGM 1 sec AS

AS

AR FS FS FS FS FS FS FS FS FS FS FS FS FS FS FS FS FS

Marker VS PP Median Atrial Evidence

VS ?? 0

VS ?? 0

VS ?? 0

VS 830 1

VS 810 2

VS 220 3

VS 190 4

VS 190 5

VS 190 6

AF – – MS 00

ECG

AEGM

AS FS FS

AS FS FS

AR

AR

AR FS FSFS

AS FS FS FSFS FD FD FD FD FD FD

Marker VP PP Median Atrial Evidence

VS 180 28

VS 180 27

VS 190 28

VS 190 29

VS 190 28

VS 200 29

VS 200 30

VS 210 31

VS 200 *A

VS 200

VS 200

Figure 3-90. Holter recording shows detection of spontaneous atrial fibrillation (AF). Electrocardiogram (ECG), telemetered atrial EGM (AEGM), and event markers are shown. The two lines below the event markers indicate the median P-P interval for the last 12 atrial events and the value of atrial tachycardia/fibrillation (AT/AF) evidence counter. Top panel shows onset of AF (arrow), and bottom panel shows AF detection (arrow). Panels are not continuous; bottom panel begins 10 seconds after top panel ends. Intermediate-height atrial markers (AS) correspond to sensed sinus P waves. Short markers (AR) indicate atrial events in the postventricular atrial refractory period for pacing. Short double markers (FS) indicate intervals below the programmed AF detection interval of 270 msec. Intermediate-height ventricular markers (VS) correspond to sensed R waves; tall ventricular markers (VP) correspond to ventricular paced events. Sinus rhythm is present at the beginning of the top panel. The AEGM shows a low-amplitude far-field R wave that is not sensed. The AT/AF evidence counter remains at 0 for the first four QRS complexes because a single atrial event exists in the preceding R-R interval. There are three atrial events in the R-R interval between the fourth and fifth complexes, and the counter first creates increments on the fifth complex. Absence of postventricular atrial blanking permits sensing of the first AF EGM, which follows the sensed ventricular event by only 30 msec. Intermittent undersensing occurs at the beginning of the bottom panel. The counter creates decrements from 28 to 27 on the fourth QRS complex and from 29 to 28 on the seventh complex. Detection of AF occurs when the count reaches 32 (arrow) and the P-P median is less than the AF detection interval. This starts the AT/AF duration timer and begins the atrial episode designated by *A. Short triple atrial markers indicate an ongoing AT/AF episode. (From Swerdlow CD, Schsls W, Dijkman B, et al: Detection of atrial fibrillation and flutter by a dual-chamber implantable cardioverter-defibrillator. For the Worldwide Jewel AF Investigators. Circulation 101:878-885, 2000, with permission.)

Figure 3-91. Holter recording illustrates continuous detection of atrial fibrillation (AF). A, Onset of AF and initial detection of AF after 32 ventricular events (arrow). Double atrial markers (FS) change to triple markers (FD), and the symbol *A appears on the atrial evidence channel to indicate AF episode in progress. B, Rhythms after 5 hours of continuous recording. AP, atrial-paced event; VP, ventricular-paced event; VS, sensed R wave. (From Swerdlow CD, Schsls W, Dijkman B, et al: Detection of atrial fibrillation and flutter by a dual-chamber implantable cardioverter-defibrillator. For the Worldwide Jewel AF Investigators. Circulation 101:878-885, with permission.)

13:26:26-1

AF - - MS delta

ECG

AEGM

AP

AP

AR FS FS FS FS FS

FS FS FS FS FS FS FS FS FD FD FD

Marker VP PP Median Atrial Evidence

VP 750 0

VS 750 1

VP 750 2

VS 200 30

VP 200 31

VP 210

VP 210 *A

A

ECG

AEGM 1 sec FD FD FD FD FD FD FD FD FD FD FD FD FD FD FD FD FD FD FD

FD FDFDFD FD FDFD FD FD FD FD

Marker VS VS PP Median Atrial Evidence

VS 170

VS 170

VP 170

VS 180 *A

VP 170

VS 170

VS 170

VS 170 *A

B Onset

Detection 12 sec

First Rx

Termination 19 sec

1 min

Marker 5 0 0

P P

V P

20

10

30

0

20

Time (sec)

10

2 1 1 1 1 1 1 1 1 7 9 8 9 9 8 9 7 9 A 0 T 0 T0 T 0 T0 T 0 T0 T0F0 T R S F F F F F F S F

4 4 0

0

V S

4 1 0

V S

T S

3 5 0

T S

3 7 0

3 5 0

T S

2 8 0

3 7 0

Onset

Atip-Vring EGM

Marker

1 2 1 1 1 1 1 1 1 2 2 1 1 2 1 1 2 1 2 1 1 1 1 1 1 2 2 2 8 1 9 9 9 9 9 8 7 1 0 9 9 0 9 9 0 9 0 8 7 6 5 5 5 6 2 2 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 T 0 A 0 A 0 A 0 A 0 A0 A 0 A 0 T 0 T 0 T D D D D D D D D D D D D D D D D D D D S P P P P P P S F F

4 4 0

V S

6 0 0

V P

4 4 0

V S

5 5 0

V S

6 2 0

V P

6 2 0

V S

4 8 0

V S

4 4 0

V S

First Rx

4 3 0

V S

4 1 0

V S

4 9 0

V S

3 4 0

6 7 0

A P

6 0 0

V P

6 6 0

A P

5 9 0

V S

A P

7 4 0

V P

Figure 3-92. Atrial tachycardia (AT) detection and therapy after 1 minute of sustained detection. Spontaneous episode of AT from a Medtronic GEM III AT device. Upper left panel, P-P intervals (open squares) and R-R intervals (closed circles) for this episode. Onset of the AT episode is labeled on the interval plot and on the stored marker channel (top right). The interval plot also labels the initial detection of AT (Detection), delivery of the first therapy (First ATP Rx), and device recognition of episode termination (Termination). In this case, therapies were programmed to begin 1 minute after initial detection. Top right panel, Dual-chamber EGM markers preceding detection of AT. Dots indicate discontinuous recording between top and bottom panels. Bottom panel, Composite EGM recorded between the atrial tip electrode (Atip) and the ventricular ring electrode (Vring) and dual-chamber EGM markers immediately preceding delivery of successful antitachycardia pacing (First Rx, fast AP events). TS, FS, and TF events refer to atrial intervals in the AT, atrial fibrillation (AF), and overlap zones, respectively. TD (atrial marker), atrial tachycardia detected; VP, ventricular paced event; VS, ventricular sensed event.

151

152

Section One: Basic Principles of Device Therapy

Atip-Vring EGM

F F F F F F F F F FF F F F F F F F D D D D D D D D D DD D D D D D D D

F F F F F F F F F F F F F F F F F F FF D D D D D D D D D D D D D D D D D D DD

Markers

V S 4 1 0

V P 6 0 0

V S 5 8 0

V S 5 6 0

F S 3 0 0

V S 3 2 0

F S 2 1 0

F S 1 8 0

V S 3 2 0

F S 2 4 0

F S 2 5 0

F S 2 5 0

F S 2 5 0

F S 2 5 0

F S 2 5 0

F S 2 5 0

F S 2 5 0

F D 2 5 0

R-R Intervals Figure 3-93. Spontaneous rapid ventricular tachycardia (VT) detected in the ventricular fibrillation (VF) zone during ongoing atrial fibrillation (AF) in this Medtronic Jewel AF Model 7250 atrial/ventricular ICD. Tracing shows composite EGM recorded between atrial tip electrode (Atip) and ventricular ring electrode (Vring), with dual-chamber EGM markers. Vertical arrows on left point to atrial component of the composite atrioventricular (AV) EGM. Horizontal arrow overlies the VT portion of the tracing. FD on atrial markers indicates ongoing AF episode. FD at right of tracing on ventricular channel indicates detection of VF. VS and FS are ventricular intervals in the sinus and VF zones, respectively. VP, ventricular paced interval.

Episode begins as AF EGM: Atip to Aring

A-A Interval (ms)

2 4 0

T F

2 5 0

T F

2 5 0

T F

2 5 0

T F

2 7 0

2 2 0

T T S F

2 8 0

T S

2 7 0

T S

3 1 0

T S

2 9 0

T S

2 6 0

T F

2 4 0

T F

2 4 0

2 1 2 2 3 2 2 4 5 8 9 0 7 6 0 0 0 0 0 0 0 T T F T T T T F F S S S S S

Marker Annotation V-V Interval (ms)

V S

V P

8 6 0

V P

8 6 0

V S

8 2 0

V P

8 6 0

V P

8 6 0

8 6 0

Transition to regular AT/Flutter

A-A Interval (ms)

T D

2 9 0

2 8 0

T D

T D

2 9 0

2 9 0

T D

T D

2 8 0

T D

2 8 0

T D

2 9 0

2 8 0

T D

T D

2 8 0

T D

2 8 0

T D

2 9 0

T D

2 8 0

T D

2 8 0

T D

2 8 0

T D

3 0 0

T D

3 2 0

T D

3 1 0

T D

Marker Annotation V-V Interval (ms)

V S

5 7 0

V S

5 8 0

V S

5 7 0

V S

5 7 0

V S

5 8 0

slower/regular cycle lengths (AT). Classifications are made on a sliding window of 12 consecutive cycles until the end of the episode is reached.153,154 Detection Considerations for Atrial Shocks In addition to permitting therapy after hours of continuous AF, atrial ICDs may withhold therapy if the episode duration is sufficiently long that patients may be at risk for thromboembolism if they are not adequately anticoagulated, typically 24 hours. R-wave synchronized atrial shock therapy is restricted to R-R intervals greater than 400 to 500 msec. This minimizes the risk that a therapeutic atrial shock can be delivered into the vulnerable period of the preceding

V S

5 7 0

V S

5 7 0

V S

6 0 0

V S

6 2 0

Figure 3-94. Atrial arrhythmia episode begins as atrial fibrillation (AF) (upper panel) but 1 hour later shows organization of local right atrial EGMs (lower panel), possibly indicating global transition to atrial flutter. One therapeutic strategy is to withhold antitachycardia pacing (ATP) until slowing and regularity of atrial sensed EGM indicates a high probability that it will be effective. This example is from a Medtronic AT500 pacemaker. Aring, atrial ring electrode; Atip, atrial tip electrode; TD, atrial tachycardia detected; VP, ventricular paced event; VS, ventricular sensed event.

cardiac cycle, but it may prevent shock therapy for rapidly conducted AF. Early or immediate recurrence of AF after shock is an important clinical problem in patients with atrial ICDs.128 Early recurrence of AF before postshock redetection of sinus rhythm will result in incorrect classification of shock success (Fig. 3-95).

Subcutaneous Electrocardiography The subcutaneous ECG is similar to the surface ECG because the two subcutaneous electrodes are suffi-

153

Chapter 3: Sensing and Detection Figure 3-95. Postshock ERAF recorded at implantation occurs three R-R intervals after shock. Because the ICD requires five consecutive postshock beats of sinus or atrial-paced rhythm to detect termination of atrial fibrillation (AF) and store intervals, no data were stored for this clinically unsuccessful shock. Lead II of the surface electrocardiogram, the atrial EGM, and the atrial and ventricular channels with event markers are shown. AR, atrial refractory event; AS, atrial sensed event; CD, charge delivered; CE, charge end; ERAF, early recurrence of atrial fibrillation; VP, ventricular paced event, VS, ventricular sensed event. (From Swerdlow CD, Schwartzman D, Hoyt R, et al: Determinants of first-shock success for atrial implantable cardioverter defibrillators. J Cardiovasc Electrophysiol 13:347-354, 2002, with permission.)

Lead II

Shock

Atip-Aring

ERAF

Markers A A A A AAA R R S R S RS

V S

V S

V C V S E S

A A A AC S S S SD

V R

ciently distant from the heart that they record electrical activity from the entire heart. Clinically, subcutaneous ECGs are used to detect arrhythmias in an implantable syncope monitor, to obviate the need for surface ECG electrodes during follow-up of pacemakers and ICDs, and to detect VT/VF in a subcutaneous ICD. Like the surface ECG, the amplitude of subcutaneous ECG signals usually is 1 mV or less. During development of an implantable syncope monitor (Medtronic Reveal), subcutaneous ECG signals with typical amplitudes of 0.25 mV were recorded between two electrodes separated by 3.2 cm on the surface of a pacemakersized device placed under the skin or muscle in the left pectoral region.53 Unlike EGMs, the amplitude of these subcutaneous ECGs increases over time to 0.30 mV at 2 to 3 months, and 0.35 mV at 4 to 6 months, and then remains stable.155 Practical implantation considerations usually limit the subcutaneous electrode separation distance to 4 to 8 cm, compared with the typical surface ECG limb lead electrode separation of 40 to 60 cm. The orientation of the two subcutaneous electrodes relative to the heart can affect the amplitude of the signal recorded. Mapping studies on the chest skin with the 4-cm electrode spacing of the implantable syncope monitor show larger intrinsic QRS amplitudes of 0.5 ± 0.1 mV for vertical orientation in the left parasternal zone and for horizontal orientation near the apex of the heart.156 These two locations had comparable amplitudes of subcutaneous ECGs. No significant differences were found for patients placed in five body positions (supine, left, right, sitting, and standing). Undersensing of intrinsic rhythm due to abrupt unexplained decreases in QRS amplitude, transient loss of signal, or baseline drift causes inappropriate detection of bradycardia by implantable syncope monitors. However, signal quality is usually sufficient to be diagnostic. Figures 3-96 to 3-98 show examples of rhythms recorded by implantable monitors, including bradycardia, VT, and skeletal myopotentials.

A S

A S

V P

V S

A S

A A A R R R

V S

V S

V S

A A A A A A AAAA A AA R R RS R RSRSR S RR

V S

V S

V S

V P

V S

V S

V P

An intraoperative study using three “button” electrodes mounted on an ICD can placed in the left infraclavicular region showed that electrode orientation had little effect. But P-wave amplitudes were only about 0.02 mV, and the ratio of P to QRS amplitude was about one half that of surface ECGs.157 Subcutaneous ECGs recorded between a defibrillation coil in the superior vena cava and the ICD can are a programmable option in some ICDs (Medtronic Marquis). These subcutaneous ECGs have properties similar to those of the syncope monitor. They are used to store an ECG-like signal during ICD episodes and in lieu of surface electrodes during ICD follow-up. Usable P waves are visible in about 80% of these patients.157

Future Directions Subcutaneous ICDs Subcutaneous ICDs with no transvenous or epicardial leads are undergoing early trials. They must rely on subcutaneous ECGs for sensing and detection of VF158 (Fig. 3-99). In one version, correlation waveform analysis using two different channels of subcutaneous ECG is employed to detect fast-VT/VF and to avoid inappropriate detection of myopotential noise and EMI.159 There are no published data on sensing and detection performance. Hemodynamic Sensors for ICDs Despite the substantial changes in ICD technology since the late 1980s, modern ICDs do not differentiate directly between hemodynamically stable and unstable tachycardias. Historically, detection durations and therapy sequences have been programmed aggressively to minimize the potential for syncope, but this results in more shocks being delivered than are necessary. Reduction of morbidity associated with

154

Section One: Basic Principles of Device Therapy Figure 3-96. Stored EGM recorded by an implantable loop recorder (Reveal, Medtronic) shows about 45 seconds of asystole, some myopotentials, perhaps one depolarization in the middle of the asystole, very small deflections throughout that are probably P waves, and relative bradycardia on the bottom panel. The amplitude of the subcutaneous electrocardiogram is approximately 0.25 mV.

11:32:07

11:32:20

11:32:33

11:32:46

11:32:59

12:22:47

P

12:23:00

230 BPM (260 ms)

unnecessary shocks for hemodynamically stable VT and for repetitive shocks due to SVT may be achieved through the use of implantable hemodynamic sensors integrated with the detection and therapy decision process. Mixed venous oxygen saturation, right atrial pressure, RV pressure, subcutaneous photoplethysmography, endocardial accelerometers, and impedance measurements have been proposed as methods of discriminating between hemodynamically stable and unstable tachycardias.52,160-168 Subcutaneous photoplethysmography technology has been described and tested in acute and chronic animal models. These studies have found a good correlation between mean arterial pressure and photoplethysmography pulse amplitude and have demonstrated the feasibility of discriminating perfusing (stable) from nonperfusing

Figure 3-97. Stored EGM recorded by an implantable loop recorder (Reveal, Medtronic) shows nonsustained tachycardia at about 230 bpm with a duration of 8 seconds. The underlying rhythm is reset by the high-rate segment. The patient did activate this stored strip, as indicated by the dark triangle and the “P” marker.

(unstable) tachycardias.169,170 Implantable systems for ambulatory hemodynamic monitoring using RV pressure and mixed venous oxygen sensors has been described.171 A recent prospective clinical study with RV pressure monitoring using the Medtronic Chronicle B reported the usefulness of the sensor for heart failure monitoring.172 This device has the capability of recording RV pressure waveforms during tachycardias detected using rate + interval analysis. Episodes of recorded spontaneous VF demonstrate substantial changes in RV pressure waveforms (Fig. 3-100). Many factors influence the hemodynamic stability of a tachyarrhythmia. Developing reliable metrics that discriminate stable versus unstable tachyarrhythmias for integration into ICD algorithms remains a major challenge.

Chapter 3: Sensing and Detection

16:20:15

16:20:28

16:20:41

16:20:54

16:21:07

16:21:20

Figure 3-98. Stored EGM recorded by an implantable loop recorder (Reveal, Medtronic) shows highfrequency electrical activity at a rate of about 800 bpm, probably myopotential artifacts. The rate is too fast to represent cardiac ventricular activation, and in some portions the underlying cardiac rhythm can be seen to be “marching through” the recording. Many artifacts are recorded by subcutaneous implantable loop recorders because the subcutaneous electrodes are closely spaced, greater than normal ECG amplification is required, and the frequency contents of myopotentials and electrocardiograms overlap extensively, so that filtering is of limited value in rejecting noise.

Amp mV1

3

0

3 0

0.05

0.1

0.15 Time

0.2

0.25

0.3

0.35

0.4

0.45 Time

0.5

0.55

0.6

Amp mV2

3

0

3 0.3

Figure 3-99. Ventricular fibrillation (VF) recorded from sensing electrodes of an investigational subcutaneous ICD (AXIOM Model 1010, Cameron Health, Inc., San Clemente, Calif.). Sensing occurs between a parasternal electrode and the implantable device placed in the left midaxillary region. VF is induced by DC stimulation and terminated by a shock. Shock artifact is not seen well in this tracing.

155

156

Section One: Basic Principles of Device Therapy Tachy Trigger Rate Max Heart Rate Min Heart Rate

150 bpm 300 bpm 46 bpm

Sampling Tachy Activity Tachy V. Beats to Detect

Evert beat 2 12 of 16

Chart speed: 25.0 mm/sec EGM (1 mV) Marker Annotation V S

V S

V S

V S

V S

V S

V S

V S

V S

V S

V S

V S

V V S S

V S

V S

V S

V V V V V S S S S S

V S

V S

30 mmHg RV 12 mmHg Pressure 5 mmHg

Figure 3-100. Spontaneous ventricular fibrillation (VF) recording from an implanted Chronicle B hemodynamic monitoring system (Medtronic). The criterion to trigger diagnostic storage for tachycardia episodes is 12 of 16 beats faster than 150 bpm in this example. The tracings shown are (from top to bottom) unipolar ventricular EGM, ventricular markers, and right ventricular (RV) pressure as measured from the chronic pressure sensor. Note the dramatic decrease in RV pulse pressure immediately after onset of VF. Undersensing of VF occurs due to fixed threshold sensing with a programmed sensitivity of 2.0 mV. VS, ventricular sensed event.

REFERENCES 1. Parsonnet V, Myers GH, Kresh YM: Characteristics of intracardiac electrograms II: Atrial endocardial electrograms. PACE 3:406-417, 1980. 2. Saxonhouse SJ, Conti JB, Curtis AB: Current of injury predicts adequate active lead fixation in permanent pacemaker/defibrillation leads. J Am Coll Cardiol 45:412-417, 2005. 3. Sweeney MO, Ellison KE, Shea JB, et al: Provoked and spontaneous high-frequency, low-amplitude, respirophasic noise transients in patients with implantable cardioverter defibrillators. J Cardiovasc Electrophysiol 12:402-410, 2001. 4. Weretka S, Michaelsen J, Becker R, et al: Ventricular oversensing: A study of 101 patients implanted with dual chamber defibrillators and two different lead systems. PACE 26(1 Pt 1):65-70, 2003. 5. Platia EV, Brinker JA: Time course of transvenous pacemaker stimulation impedance, capture threshold, and electrogram amplitude. PACE 9:620-625, 1986. 6. DeCaprio V, Hurzeler P, Furman S: A comparison of unipolar and bipolar electrograms for cardiac pacemaker sensing. Circulation 56:750-755, 1977. 7. Cornacchia D, Jacopi F, Fabbri M, et al: [Comparison between active screw type and passive electrodes in permanent intraventricular pacing]. Minerva Cardioangiol 32:101-103, 1984. 8. Ceviz N, Celiker A, Kucukosmanoglu O, et al: Comparison of mid-term clinical experience with steroid-eluting active and passive fixation ventricular electrodes in children. PACE 23:1245-1249, 2000. 9. Danilovic D, Ohm OJ: Pacing impedance variability in tined steroid eluting leads. PACE 21:1356-1363, 1998. 10. Hua W, Mond HG, Strathmore N: Chronic steroid-eluting lead performance: A comparison of atrial and ventricular pacing. PACE 20(1 Pt 1):17-24, 1997. 11. Crossley GH, Brinker JA, Reynolds D, et al: Steroid elution improves the stimulation threshold in an active-fixation atrial permanent pacing lead: A randomized, controlled study. Model 4068 Investigators. Circulation 92:2935-2939, 1995. 12. Wish M, Swartz J, Cohen A, et al: Steroid-tipped leads versus porous platinum permanent pacemaker leads: A controlled study. PACE 13(12 Pt 2):1887-1890, 1990. 13. Ellenbogen KA, Wood MA, Gilligan DM: Evaluation of “inappropriate” ICD shocks in an asymptomatic patient following myocardial infarction. PACE 19:254-255, 1996. 14. Goldschlager N, Epstein A, Friedman P, et al: Environmental and drug effects on patients with pacemakers and implantable cardioverter/defibrillators: A practical guide to patient treatment. Arch Intern Med 161:649-655, 2001. 15. Rajawat YS, Patel VV, Gerstenfeld EP, et al: Advantages and pitfalls of combining device-based and pharmacologic therapies

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

for the treatment of ventricular arrhythmias. PACE 27:16701681, 2004. Theres H, Stadler RW, Stylos L, et al: Comparison of electrocardiogram and intrathoracic electrogram signals for detection of ischemic ST segment changes during normal sinus and ventricular paced rhythms. J Cardiovasc Electrophysiol 13:990-995, 2002. Centurion OA, Isomoto S, Shimizu A, et al: The effects of aging on atrial endocardial electrograms in patients with paroxysmal atrial fibrillation. Clin Cardiol 26:435-438, 2003. Frohlig G, Schwerdt H, Schieffer H, et al: Atrial signal variations and pacemaker malsensing during exercise: A study in the time and frequency domain. J Am Coll Cardiol 11:806-813, 1988. Ross BA, Zeigler V, Zinner A, et al: The effect of exercise on the atrial electrogram voltage in young patients. PACE 14:20922097, 1991. Schuchert A, Kuck KH, Bleifeld W: Stability of pacing threshold, impedance, and R wave amplitude at rest and during exercise. PACE 13(12 Pt 1):1602-1608, 1990. Shandling AH, Florio J, Castellanet MJ, et al: Physical determinants of the endocardial P wave. PACE 13(12 Pt 1):1585-1589, 1990. Rosenheck S, Schmaltz S, Kadish AH, et al: Effect of rate augmentation and isoproterenol on the amplitude of atrial and ventricular electrograms. Am J Cardiol 66:101-102, 1990. Varriale P, Chryssos BE: Atrial sensing performance of the single-lead VDD pacemaker during exercise. J Am Coll Cardiol 22:1854-1857, 1993. Chan CC, Lau CP, Leung SK, et al: Comparative evaluation of bipolar atrial electrogram amplitude during everyday activities: Atrial active fixation versus two types of single pass VDD/R leads. PACE 17(11 Pt 2):1873-1877, 1994. Furman S, Hurzeler P, De Caprio V: Cardiac pacing and pacemaker. III. Sensing the cardiac electrogram. Am Heart J 93:794801, 1997. Ellenbogen KA, Wood MA, Stambler BS, et al: Measurement of ventricular electrogram amplitude during intraoperative induction of ventricular tachyarrhythmias. Am J Cardiol 70:1017-1022, 1992. Taneja T, Goldberger J, Parker MA, et al: Reproducibility of ventricular fibrillation characteristics in patients undergoing implantable cardioverter defibrillator implantation. J Cardiovasc Electrophysiol 8:1209-1217, 1997. McAlister HF, Klementowicz PT, Calderon EM, et al: Atrial electrogram analysis: Antegrade versus retrograde. PACE 11(11 Pt 2):1703-1707, 1988. Wainwright R, Davies W, Tooley M: Ideal atrial lead positioning to detect retrograde atrial depolarization by digitization and slope analysis of the atrial electrogram. PACE 7(6 Pt 2):11521158, 1984.

Chapter 3: Sensing and Detection 30. Timmis GC, Westveer DC, Bakalyar DM, et al: Discrimination of anterograde from retrograde atrial electrograms for physiologic pacing. PACE 11:130-140, 1988. 31. Davies DW, Wainwright RJ, Tooley MA, et al: Detection of pathological tachycardia by analysis of electrogram morphology. PACE 9:200-208, 1986. 32. Roithinger FX, SippensGroenewegen A, Karch MR, et al: Organized activation during atrial fibrillation in man: Endocardial and electrocardiographic manifestations. J Cardiovasc Electrophysiol 9:451-461, 1998. 33. Konings KT, Smeets JL, Penn OC, et al: Configuration of unipolar atrial electrograms during electrically induced atrial fibrillation in humans. Circulation 95:1231-1241, 1997. 34. Wood M, Moskovljevic P, Stambler B, et al: Comparison of bipolar atrial electrogram amplitude in sinus rhythm, atrial fibrillation, and atrial flutter. PACE 19:150-156, 1996. 35. Lewalter T, Schimpf R, Kulik D, et al: Comparison of spontaneous atrial fibrillation electrogram potentials with the P-wave electrogram amplitude in dual chamber pacing with unipolar atrial sensing. Europace 2:136-140, 2000. 36. Ropella KM, Sahakian AV, Baerman JM, et al: The coherence spectrum: A quantitative discriminator of fibrillatory and nonfibrillatory cardiac rhythms. Circulation 80:112-119, 1989. 37. Baerman JM, Ropella KM, Sahakian AV, et al: Effect of bipole configuration on atrial electrograms during atrial fibrillation. PACE 13:78-87, 1990. 38. Karagueuzian HS, Khan SS, Peters W, et al: Nonhomogeneous local atrial activity during acute atrial fibrillation: Spectral and dynamic analysis. PACE 13(12 Pt 2):1937-1942, 1990. 39. Ropella KM, Sahakian AV, Baerman JM, et al: Effects of procainamide on intra-atrial [corrected] electrograms during atrial fibrillation: Implications [corrected] for detection algorithms. Circulation 77:1047-1054, 1988. 40. Beeman AL, Deutsch G, Rea RF: Paradoxical undersensing due to quiet timer blanking. Heart Rhythm 1:345-347, 2004. 41. Willems R, Holemans P, Ector H, et al: Paradoxical undersensing at a high sensitivity in dual chamber pacemakers. PACE 24:308-315, 2001. 42. Castro A, Liebold A, Vincente J, et al: Evaluation of autosensing as an automatic means of maintaining a 2:1 sensing safety margin in an implanted pacemaker. Autosensing Investigation Team. PACE 19(11 Pt 2):1708-1713, 1996. 43. Nowak B, Kampmann C, Schmid FX, et al: Pacemaker therapy in premature children with high degree AV block. PACE 21:26952698, 1998. 44. Inama G, Santini M, Padeletti L, et al: Far-field R wave oversensing in dual chamber pacemakers designed for atrial arrhythmia management: Effect of pacing site and lead tip to ring distance. PACE 27:1221-1230, 2004. 45. Callans DJ, Hook BG, Kleiman RB, et al: Unique sensing errors in third-generation implantable cardioverter-defibrillators. J Am Coll Cardiol 22:1135-1140, 1993. 46. Callans DJ, Hook BG, Marchlinski FE: Paced beats following single nonsensed complexes in a “codependent” cardioverter defibrillator and bradycardia pacing system: Potential for ventricular tachycardia induction. PACE 14:1281-1287, 1991. 47. Callans DJ, Hook BG, Marchlinski FE: Effect of rate and coupling interval on endocardial R wave amplitude variability in permanent ventricular sensing lead systems. J Am Coll Cardiol 22:746-750, 1993. 48. Dekker LR, Schrama TA, Steinmetz FH, et al: Undersensing of VF in a patient with optimal R wave sensing during sinus rhythm. PACE 27(6 Pt 1):833-834, 2004. 49. Glikson M, Beeman AL, Luria DM, et al: Impaired detection of ventricular tachyarrhythmias by a rate-smoothing algorithm in dual-chamber implantable defibrillators: Intradevice interactions. J Cardiovasc Electrophysiol 13:312-318, 2002.

157

50. Cooper JM, Sauer W, Verdino R: Absent ventricular tachycardia detection in a biventricular implantable cardioverter-defibrillator due to intradevice interaction with a rate smoothing pacing algorithm. Heart Rhythm 1:728-731, 2004. 51. Shivkumar K, Feliciano Z, Boyle NG, et al: Intradevice interaction in a dual chamber implantable cardioverter defibrillator preventing ventricular tachyarrhythmia detection. J Cardiovasc Electrophysiol 11:1285-1288, 2000. 52. Ellenbogen KA, Wood MA, Kapadia K, et al: Short-term reproducibility over time of right ventricular pulse pressure as a potential hemodynamic sensor for ventricular tachyarrhythmias. PACE 15:971-974, 1992. 53. Leitch J, Klein G, Yee R, et al: Feasibility of an implantable arrhythmia monitor. PACE 15:2232-2235, 1992. 54. Swerdlow CD: Implantation of cardioverter defibrillators without induction of ventricular fibrillation. Circulation 103:2159-2164, 2001. 55. Tung L, Tovar O, Neunlist M, et al: Effects of strong electrical shock on cardiac muscle tissue. Ann N Y Acad Sci 720:160-175, 1994. 56. Winkle RA, Bach SM Jr, Echt DS, et al: The automatic implantable defibrillator: Local ventricular bipolar sensing to detect ventricular tachycardia and fibrillation. Am J Cardiol 52:265-270, 1983. 57. Goldberger JJ, Horvath G, Donovan D, et al: Detection of ventricular fibrillation by transvenous defibrillating leads: Integrated versus dedicated bipolar sensing. J Cardiovasc Electrophysiol 9:677-688, 1998. 58. Yee R, Jones D, Jarvis E, et al: Changes in pacing threshold and R wave amplitude after transvenous catheter countershock. J Am Coll Cardiol 4:543-549, 1984. 59. Jung W, Manz M, Moosdorf R, et al: Changes in the amplitude of endocardial electrograms following defibrillator discharge: Comparison of two lead systems. PACE 18(12 Pt 1):2163-2172, 1995. 60. Swerdlow CD, Brown ML, Lurie K, et al: Discrimination of ventricular tachycardia from supraventricular tachycardia by a downloaded wavelet-transform morphology algorithm: A paradigm for development of implantable cardioverter defibrillator detection algorithms. J Cardiovasc Electrophysiol 13:432-441, 2002. 61. Kuhlkamp V, Dornberger V, Mewis C, et al: Clinical experience with the new detection algorithms for atrial fibrillation of a defibrillator with dual chamber sensing and pacing. J Cardiovasc Electrophysiol 10:905-915, 1999. 62. Wilkoff BL, Kuhlkamp V, Volosin K, et al: Critical analysis of dual-chamber implantable cardioverter-defibrillator arrhythmia detection: Results and technical considerations. Circulation 103:381-386, 2001. 63. Swerdlow CD: Supraventricular tachycardia-ventricular tachycardia discrimination algorithms in implantable cardioverter defibrillators: State-of-the-art review. J Cardiovasc Electrophysiol 12:606-612, 2001. 64. Gunderson B, Patel A, Bounds C: Automatic identification of implantable cardioverter-defibrillator lead problems using intracardiac electrograms. Comput Cardiol 29:121-124, 2002. 65. Swerdlow C, Shivkumar K: Implantable cardioverter defibrillators: Clinical aspects. In Zipes DP, Jalife J (eds): Cardiac Electrophysiology: From Cell to Bedside, 4th ed. Philadephia, WB Saunders, 2004, pp 980-993. 66. Garcia-Moran E, Mont L, Brugada J: Inappropriate tachycardia detection by a biventricular implantable cardioverter defibrillator. PACE 25:123-124, 2002. 67. Lloyd M, Hayes D, Friedman P: Troubleshooting. In Hayes D, Lloyd M, Friedman P (eds): Cardiac Pacing and Defibrillation: A Clinical Approach. Armonk, NY, Futura, 2000, pp 347-452. 68. Ellenbogen KA, Wood MA, Shepard RK, et al: Detection and management of an implantable cardioverter defibrillator lead

158

69. 70.

71.

72.

73.

74.

75.

76.

77. 78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

Section One: Basic Principles of Device Therapy

failure: Incidence and clinical implications. J Am Coll Cardiol 41:73-80, 2003. Swerdlow C, Friedman P: Advanced ICD troubleshooting: Part I. PACE 28:1322-1346, 2005. Pinski SL: 2:1 Tracking of sinus rhythm in a patient with a dualchamber implantable cardioverter defibrillator: What is the mechanism? J Cardiovasc Electrophysiol 12:503-504, 2001. Hsu SS, Mohib S, Schroeder A, et al: T wave oversensing in implantable cardioverter defibrillators. J Interv Card Electrophysiol 11:67-72, 2004. Sweeney MO, Hellkamp AS, Ellenbogen KA, et al: Adverse effect of ventricular pacing on heart failure and atrial fibrillation among patients with normal baseline QRS duration in a clinical trial of pacemaker therapy for sinus node dysfunction. Circulation 107:2932-2937, 2003. Wilkoff BL, Cook JR, Epstein AE, et al: Dual-chamber pacing or ventricular backup pacing in patients with an implantable defibrillator: The Dual Chamber and VVI Implantable Defibrillator (DAVID) Trial [see comment]. JAMA 288:3115-3123, 2002. Pinski S, Egula L, Sgarbossa E, et al: Incidence and causes of commlted shocks in noncommited implantable defibrillators [abstract]. J Am Coll Cardiol 39:93A, 2002. Sweesy MW, Holland JL, Smith KW: Electromagnetic interference in cardiac rhythm management devices. AACN Clin Issues 15:391-403, 2004. Lau CP, Leung SK, Tse HF, et al: Automatic mode switching of implantable pacemakers: II. Clinical performance of current algorithms and their programming. PACE 25:1094-1113, 2002. Israel CW: Analysis of mode switching algorithms in dual chamber pacemakers. PACE 25:380-393, 2002. Lam CT, Lau CP, Leung SK, et al: Improved efficacy of mode switching during atrial fibrillation using automatic atrial sensitivity adjustment. PACE 22(1 Pt 1):17-25, 1999. Leung SK, Lau CP, Lam C, et al: Programmed atrial sensitivity: A critical determinant in atrial fibrillation detection and optimal automatic mode switching. PACE 21(11 Pt 2):2214-2219, 1998. Pollak WM, Simmons JD, Interian A Jr, et al: Pacemaker diagnostics: A critical appraisal of current technology. PACE 26(1 Pt 1):76-98, 2003. Passman RS, Weinberg KM, Freher M, et al: Accuracy of mode switch algorithms for detection of atrial tachyarrhythmias. J Cardiovasc Electrophysiol 15:773-777, 2004. Hammel E, Hudelo C, Mailland L, et al: Appropriate detection of Guidant pacemaker stored electrograms assessed by centralized arrhythmia workstation [abstract]. PACE 23:680, 2000. Israel CW, Gascon D, Nowak B, et al: Diagnostic value of stored electrograms in single-lead VDD systems. PACE 23(11 Pt 2): 1801-1803, 2000. Melzer C, Sowelam S, Sheldon TJ, et al: Reduction of right ventricular pacing in patients with sinus node dysfunction using an enhanced search AV algorithm. PACE 28:521-527, 2005. Sweeney M, Nsah E, McGrew F, et al: Reduction in ventricular pacing and its long term clinical outcomes: Preliminary results of the Save Pace Trial [abstract]. Heart Rhythm 2(Suppl):S322, 2005. Olshansky B, Day J, McGuire M, et al: Inhibition of unnecessary RV pacing with AV search hysteresis in ICDs (INTRINSIC RV): Design and clinical protocol. PACE 28:62-66, 2005. Deering T, Wilensky M, Tondato F, et al: Auto intrinsic conduction search algorithm: A prospective analysis [abstract 606]. PACE 26:1080, 2003. Savoure A, Frohlig G, Galley D, et al: A new dual-chamber pacing mode to minimize ventricular pacing. PACE 28(Suppl 1):S43-S46, 2005. Sweeney MO, Shea JB, Fox V, et al: Randomized pilot study of a new atrial-based minimal ventricular pacing mode in dualchamber implantable cardioverter-defibrillators. Heart Rhythm 1:160-167, 2004.

90. de Voogt WG, Vonk BF, Albers BA, et al: Understanding capture detection. Europace 6:561-569, 2004. 91. Splett V, Trusty JM, Hayes DL, et al: Determination of pacing capture in implantable defibrillators: Benefit of evoked response detection using RV coil to can vector. PACE 23(11 Pt 1):16451650, 2000. 92. Schuchert A, Frese J, Stammwitz E, et al: Low settings of the ventricular pacing output in patients dependent on a pacemaker: Are they really safe? Am Heart J 143:1009-1011, 2002. 93. Sperzel J, Binner L, Boriani G, et al: Evaluation of the atrial evoked response for capture detection with high-polarization leads. PACE 28(Suppl 1):S57-S62, 2005. 94. Sperzel J, Compton S, et al: Automatic measurement of atrial pacing thresholds in dual chamber pacemakers: Atrial Capture Management [abstract]. Heart Rhythm 1:S118, 2004. 95. Olson W: Safety margins for sensing and detection: Programming tradeoffs. In Kroll M, Lehmann M (eds): Implantable Cardioverter Defibrillator Therapy: The Engineering-Clinical Interface. Norwell, MA, Kluwer Academic Publishers, 1996, pp 389-420. 96. Anderson MH, Murgatroyd FD, Hnatkova K, et al: Performance of basic ventricular tachycardia detection algorithms in implantable cardioverter defibrillators: Implications for device programming. PACE 20(12 Pt 1):2975-2983, 1997. 97. Murgatroyd F, Anderson M, Hnatkova K, et al: Comparison of misdiagnosis of atrial fibrillation as ventricular tachycardia by algorithms employed in current implantable cardioverterdefibrillators. Circulation 88:I-353, 1993. 98. Grimm W, Menz V, Hoffmann J, et al: Failure of third-generation implantable cardioverter defibrillators to abort shock therapy for nonsustained ventricular tachycardia due to shortcomings of the VF confirmation algorithm. PACE 21(4 Pt 1):722727, 1998. 99. Raedle-Hurst TM, Wiecha J, Schwab JO, et al: Clinical performance of a specific algorithm to reconfirm self-terminating ventricular arrhythmias in current implantable cardioverterdefibrillators. Am J Cardiol 88:744-749, 2001. 100. Gillberg J, Olson W (eds): Dual-chamber sensing and detection for implantable cardioverter-defibrillators. In Singer I (ed): Interventional Electrophysiology. Philadelphia, Lippincott Williams & Wilkins, 2001. 101. Arenal A, Almendral J, Villacastin J, et al: First postpacing interval variability during right ventricular stimulation: A single algorithm for the differential diagnosis of regular tachycardias. Circulation 98:671-677, 1998. 102. Glikson M, Swerdlow CD, Gurevitz OT, et al: Optimal combination of discriminators for differentiating ventricular from supraventricular tachycardia by dual-chamber defibrillators. J Cardiovasc Electrophysiol 16:732-739, 2005. 103. Brugada J, Mont L, Figueiredo M, et al: Enhanced detection criteria in implantable defibrillators. J Cardiovasc Electrophysiol 9:261-268, 1998. 104. Le Franc P, Kus T, Vinet A, et al: Underdetection of ventricular tachycardia using a 40 ms stability criterion: Effect of antiarrythmic therapy. PACE 20:2882-2892, 1997. 105. Neuzner J, Pitschner HF, Schlepper M: Programmable VT detection enhancements in implantable cardioverter defibrillator therapy. PACE 18(3 Pt 2):539-547, 1995. 106. Swerdlow CD, Ahern T, Chen PS, et al: Underdetection of ventricular tachycardia by algorithms to enhance specificity in a tiered-therapy cardioverter-defibrillator. J Am Coll Cardiol 24:416-424, 1994. 107. Swerdlow CD, Chen PS, Kass RM, et al: Discrimination of ventricular tachycardia from sinus tachycardia and atrial fibrillation in a tiered-therapy cardioverter-defibrillator. J Am Coll Cardiol 23:1342-1355, 1994. 108. Gronefeld GC, Schulte B, Hohnloser SH, et al: Morphology discrimination: A beat-to-beat algorithm for the discrimination of

Chapter 3: Sensing and Detection ventricular from supraventricular tachycardia by implantable cardioverter defibrillators. PACE 24:1519-1524, 2001. 109. Boriani G, Biffi M, Frabetti L, et al: Clinical evaluation of morphology discrimination: An algorithm for rhythm discrimination in cardioverter defibrillators. PACE 24:994-1001, 2001. 110. Gold MR, Shorofsky SR, Thompson JA, et al: Advanced rhythm discrimination for implantable cardioverter defibrillators using electrogram vector timing and correlation. J Cardiovasc Electrophysiol 13:1092-1097, 2002. 111. Klein G, Manolis A, Viskin S, et al: Clinical performance of Wavelet™ morphology discrimination algorithm in a worldwide single chamber ICD population [abstract]. Circulation 110:III345, 2003. 112. Stadler RW, Gunderson BD, Gillberg JM: An adaptive intervalbased algorithm for withholding ICD therapy during sinus tachycardia. PACE 26:1189-1201, 2003. 113. Stein KM, Euler DE, Mehra R, et al: Do atrial tachyarrhythmias beget ventricular tachyarrhythmias in defibrillator recipients? J Am Coll Cardiol 40:335-340, 2002. 114. Aliot E, Nitzsche R, Ripart A: Arrhythmia detection by dualchamber implantable cardioverter defibrillators: A review of current algorithms. Europace 6:273-286, 2004. 115. Wilkoff B, Sterns L, Morgan JM, et al: Preventing shocks after ICD implantation: Can a strategy of standardized ICD programming match physician tailored? Heart Rhythm 2:1034, 2005. 116. Deisenhofer I, Kolb C, Ndrepepa G, et al: Do current dual chamber cardioverter defibrillators have advantages over conventional single chamber cardioverter defibrillators in reducing inappropriate therapies? A randomized, prospective study. J Cardiovasc Electrophysiol 12:134-142, 2001. 117. Theuns DA, Klootwijk AP, Goedhart DM, et al: Prevention of inappropriate therapy in implantable cardioverter-defibrillators: Results of a prospective, randomized study of tachyarrhythmia detection algorithms. J Am Coll Cardiol 44:2362-2367, 2004. 118. Bansch D, Steffgen F, Gronefeld G, et al: The 1+1 trial: A prospective trial of a dual- versus a single-chamber implantable defibrillator in patients with slow ventricular tachycardias. Circulation 110:1022-1029, 2004. 119. Friedman PA, McClelland RL, Bamlet WR, et al: Dual chamber versus single chamber detection enhancements for implantable defibrillator rhythm diagnosis: The detect supraventricular tachycardia study. Circulation 113:2871-2879, 2006. 120. Wathen MS, Volosin KJ, Sweeney MO, et al: Ventricular antitachycardia pacing by implantable cardioverter defibrillators reduces shocks for inappropriately detected supraventricular tachycardia [abstract]. Heart Rhythm 1:S148, 2004. 121. Knight BP, Zivin A, Souza J, et al: A technique for the rapid diagnosis of atrial tachycardia in the electrophysiology laboratory. J Am Coll Cardiol 33:775-781, 1999. 122. Paz O, Birgersdotter-Green U, Swerdlow C: Acceleration of ventricular rate following atrial antitachycardia pacing: What is the mechanism? J Cardiovasc Electrophysiol 14:1382-1384, 2003. 123. Zeger SL, Liang KY, Albert PS: Models for longitudinal data: A generalized estimating equation approach. Biometrics 44:10491060, 1998. 124. Hall GH: The clinical application of Bayes’ theorem. Lancet 2:555-557, 1967. 125. Malik M: Pitfalls of the concept of incremental specificity used in comparisons of dual chamber VT/VF detection algorithms. PACE 23:1166-1170, 2000. 126. Olson W, Gunderson BD, Fang-Yen M, et al: Properties and peformance of rate detection algorithms in three implantable cardioverter defibrillators. Comput Cardiol 65-68, 1994. 127. Birgersdotter-Green U, Rosenqvist M, Lindemans FW, et al: Holter documented sudden death in a patient with an implanted defibrillator. PACE 15:1008-1014, 1991.

159

128. Schwartzman D, Musley SK, Swerdlow C, et al: Early recurrence of atrial fibrillation after ambulatory shock conversion. J Am Coll Cardiol 40:93-99, 2002. 129. Schaumann A, von zur Muhlen F, Herse B, et al: Empirical versus tested antitachycardia pacing in implantable cardioverter defibrillators: A prospective study including 200 patients. Circulation 97:66-74, 1998. 130. Wathen MS, Sweeney MO, DeGroot PJ, et al: Shock reduction using antitachycardia pacing for spontaneous rapid ventricular tachycardia in patients with coronary artery disease. Circulation 104:796-801, 2001. 131. Wathen MS, DeGroot PJ, Sweeney MO, et al: Prospective randomized multicenter trial of empirical antitachycardia pacing versus shocks for spontaneous rapid ventricular tachycardia in patients with implantable cardioverter-defibrillators: Pacing Fast Ventricular Tachycardia Reduces Shock Therapies (PainFREE Rx II) trial results. Circulation 110:2591-2596, 2004. 132. Bardy GH, Lee KL, Mark DB, et al: Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med 352:225-237, 2005. 133. Gunderson BD, Gillberg JM, Olson WH, et al: Effect of programmed number of intervals to detect ventricular fibrillation on implantable cardioverter defibrillator longevity and unnecessary shocks [abstract]. Circulation 106:322, 2002. 134. Aminoff MJ, Scheinman MM, Griffin JC, et al: Electrocerebral accompaniments of syncope associated with malignant ventricular arrhythmias. Ann Intern Med 108:791-796, 1988. 135. Daubert JP, Wojciech Z, Cannom DC, et al: Frequency and mechanisms of inappropriate ICD therapy in MADIT II [abstract]. J Am Coll Cardiol 43:132A, 2004. 136. Kouakam C, Kacet S, Hazard JR, et al: Performance of a dualchamber implantable defibrillator algorithm for discrimination of ventricular from supraventricular tachycardia. Europace 6:3242, 2004. 137. Frang P, Kus T, Vinet A, et al: Underdetection of ventricular tachycardia using 40 ms stability criterion: Effect of antiarrhythmic thearpy. PACE 20:2882-2892, 1997. 138. Morgan JM, Sterns LD, Hanson JL, et al: A trial design for evaluation of empiric programming of implantable cardioverter defibrillators to improve patient management. Curr Control Trials Cardiovasc Med 5:12, 2004. 139. Panotopoulos P, Krum D, Axtell K, et al: Ventricular fibrillation sensing and detection by implantable defibrillators: Is one better than the others? A prospective, comparative study. J Cardiovasc Electrophysiol 12:445-452, 2001. 140. Kolb C, Deisenhofer I, Weyerbrock S, et al: Incidence of antitachycardia therapy suspension due to magnet reversion in implantable cardioverter defibrillators. PACE 27:221-223, 2004. 141. Bansch D, Castrucci M, Bocker D, et al: Ventricular tachycardias above the initially programmed tachycardia detection interval in patients with implantable cardioverter-defibrillators: Incidence, prediction and significance. J Am Coll Cardiol 36:557-565, 2000. 142. Shukla HH, Flaker GC, Jayam V, et al: High defibrillation thresholds in transvenous biphasic implantable defibrillators: Clinical predictors and prognostic implications. PACE 26(1 Pt 1):44-48, 2003. 143. Brode SE, Schwartzman D, Callans DJ, et al: ICD-antiarrhythmic drug and ICD-pacemaker interactions. J Cardiovasc Electrophysiol 8:830-842, 1997. 144. Monahan KM, Hadjis T, Hallett N, et al: Relation of induced to spontaneous ventricular tachycardia from analysis of stored far-field implantable defibrillator electrograms. Am J Cardiol 83:349-353, 1999. 145. Weber M, Bocker D, Bansch D, et al: Efficacy and safety of the initial use of stability and onset criteria in implantable cardioverter defibrillators. J Cardiovasc Electrophysiol 10:145-153, 1999.

160

Section One: Basic Principles of Device Therapy

146. Cohen AI, Wish MH, Fletcher RD, et al: The use and interaction of permanent pacemakers and the automatic implantable cardioverter defibrillator. PACE 11(6 Pt 1):704-711, 1988. 147. Geiger MJ, O’Neill P, Sharma A, et al: Interactions between transvenous nonthoracotomy cardioverter defibrillator systems and permanent transvenous endocardial pacemakers. PACE 20(3 Pt 1):624-630, 1997. 148. Glikson M, Trusty JM, Grice SK, et al: A stepwise testing protocol for modern implantable cardioverter-defibrillator systems to prevent pacemaker-implantable cardioverter-defibrillator interactions. Am J Cardiol 83:360-366, 1999. 149. Wietholt D, Kuehlkamp V, Meisel E, et al: Prevention of sustained ventricular tachyarrhythmias in patients with implantable cardioverter-defibrillators: The PREVENT study. J Interv Card Electrophysiol 9:383-389, 2003. 150. Shalaby AA: Delayed detection of ventricular tachycardia in a dual chamber rate adaptive pacing implantable cardioverter defibrillator: A case of intradevice interaction. PACE 2004;27: 1164-1166, 2004. 151. Pascal D, Hazard J, Besson B, et al: Contribution of pacemaker stored electrograms to patient management [abstract]. PACE 23(Part II):681, 2000. 152. Seidl K, Meisel E, VanAgt E, et al: Is the atrial high rate episode diagnostic feature reliable in detecting paroxysmal episodes of atrial tachyarrhythmias? PACE 21(4 Pt 1):694-700, 1998. 153. Morris MM, KenKnight BH, Lang DJ: Detection of atrial arrhythmia for cardiac rhythm management by implantable devices. J Electrocardiol 33(Suppl):133-139, 2000. 154. Schuchert A, Boriani G, Wollmann C, et al: Implantable dualchamber defibrillator for the selective treatment of spontaneous atrial and ventricular arrhythmias: Arrhythmia incidence and device performance. J Interv Card Electrophysiol 12:149-156, 2005. 155. Krahn AD, Klein GJ, Yee R, et al: Maturation of the sensed electrogram amplitude over time in a new subcutaneous implantable loop recorder. PACE 20:1686-1690, 1997. 156. Chrysostomakis SI, Klapsinos NC, Simantirakis EN, et al: Sensing issues related to the clinical use of implantable loop recorders. Europace 5:143-148, 2003. 157. Mazur A, Wang L, Anderson ME, et al: Functional similarity between electrograms recorded from an implantable cardioverter defibrillator emulator and the surface electrocardiogram. PACE 24:34-40, 2001. 158. Bardy G, Cappato R, Smith WM, et al: The totally subcutaneous ICD system (the S-ICD) [abstract]. PACE 25:222, 2002. 159. Bardy G: Subcutaneous implantable defibrillator. In Malik M (ed): Dynamic Electrocardiography. Armonk, NY, Futura, 2004. 160. Bordachar P, Garrigue S, Reuter S, et al: Hemodynamic assessment of right, left, and biventricular pacing by peak endocardial acceleration and echocardiography in patients with end-stage heart failure. PACE 23(11 Pt 2):1726-1730, 2000. 161. Hegbom F, Hoff PI, Oie B, et al: RV function in stable and unstable VT: Is there a need for hemodynamic monitoring in future defibrillators? PACE 24:172-182, 2001. 162. Kaye G, Astridge P, Perrins J: Tachycardia recognition and diagnosis from changes in right atrial pressure waveform: A feasibility study. PACE 14:1384-1392, 1991. 163. Khoury D, McAlister H, Wilkoff B, et al: Continuous right ventricular volume assessment by catheter measurement of imped-

164.

165.

166.

167.

168.

169.

170.

171.

172.

173.

174. 175. 176.

177.

178.

179.

ance for antitachycardia system control. PACE 12:1918-1926, 1989. Sharma AD, Bennett TD, Erickson M, et al: Right ventricular pressure during ventricular arrhythmias in humans: Potential implications for implantable antitachycardia devices. J Am Coll Cardiol 15:648-655, 1990. Wood M, Ellenbogen KA, Lu B, et al: A prospective study of right ventricular pulse pressure and dP/dt to discriminantinduced ventricular tachycardia from supraventricular and sinus tachycardia in man. PACE 13:1148-1157, 1990. Ellenbogen KA, Lu B, Kapadia K, et al: Usefulness of right ventricular pulse pressure as a potential sensor for hemodynamically unstable ventricular tachycardia. Am J Cardiol 65: 1105-1111, 1990. Plicchi G, Marcelli E, Marini S: An endocardial acceleration sensor for sustained ventricular tachycardia detection [abstract]. Europace Suppl 3:96, 2002. Whitman T, Sheldon T, McFadden S: Endocardial acceleration measurements in tachycardia induced heart failure in canines [abstract]. PACE 24:569, 2002. Nabutovsky Y, Pavek T, Wright G, Turcott R: Chronic performance of a subcutaneous photoplethysmography sensor [abstract]. Heart Rhythm 1:476, 2004. Turcott R: Detection of hemodynamically unstable arrhythmias using subcutaneous photoplethysmography [abstract]. Heart Rhythm 2:S83, 2005. Bennett T, Kjellstrom B, Taepke R, et al: Development of implantable devices for continuous ambulatory monitoring of central hemodynamic values in heart failure patients. PACE 28:573-584, 2005. Cleland JG, Coletta AP, Freemantle N, et al: Clinical trials update from the American College of Cardiology meeting: CARE-HF and the Remission of Heart Failure, Women’s Health Study, TNT, COMPASS-HF, VERITAS, CANPAP, PEECH and PREMIER. Eur J Heart Fail 7:931-936, 2005. Kay G: Troubleshooting and programming of cardiac resynchronization therapy. In Ellenbogen KA, Kay G, Wilkoff B (eds): Device Therapy for Congestive Heart Failure. Philadelphia, Elsevier, 2004, pp 232-293. Swerdlow C, Friedman P: Advanced ICD troubleshooting: Part II. PACE 29:70-96, 2006. Meyer Y: Wavelets: Algorithms and Applications. Philadelphia, Society for Industrial and Applied Mathematics, 1993. Morris M, Marcovecchio A, KenKnight B, et al: Retrospective evaluation of detection enhancements in a dual-chamber implantable cardioverter defibrillator: Implications for device programming [abstract]. PACE 22(4 Part II):849, 1999. Willems R, Morck ML, Exner DV, et al: Ventricular high-rate episodes in pacemaker diagnostics identify a high-risk subgroup of patients with tachy-brady syndrome. Heart Rhythm 1:414421, 2004. Swerdlow CD, Schsls W, Dijkman B, et al: Detection of atrial fibrillation and flutter by a dual-chamber implantable cardioverter-defibrillator. For the Worldwide Jewel AF Investigators. Circulation 101:878-885, 2000. Swerdlow CD, Schwartzman D, Hoyt R, et al: Determinants of first-shock success for atrial implantable cardioverter defibrillators. J Cardiovasc Electrophysiol 13:347-354, 2002.

Chapter 4

Engineering and Construction of Pacemaker and Implantable CardioverterDefibrillator Leads ANDREA M. RUSSO • FRANCIS E. MARCHLINSKI

T he role of the implantable cardioverter-defibrillator (ICD) in both primary and secondary prevention is now well established.1-5 ICD lead and device technology, as well as implantation approach, have changed substantially since approval of the first ICD system in 1985. Lead technology has had a significant impact on the sensing characteristics and defibrillation performance of ICD systems. This chapter describes the design and characteristics of ICD leads, focusing on the evolution of lead technology and implantation methods occurring over the past two decades. In addition, newer technology related to left ventricular (LV) pacing leads, designed for coronary sinus pacing and cardiac resynchronization therapy (CRT), is also discussed. IMPLANTABLE CARDIOVERTERDEFIBRILLATOR LEADS The lead is an essential part of the defibrillation system that conducts electrical impulses between the pulse generator and the patient. It is composed of one or more electrodes, conductors, insulation, connectors, and a fixation mechanism. The electrode is the portion of the

lead that is responsible for sensing, pacing, and defibrillation. Lead locations may be epicardial, endocardial, or extrathoracic, or they may include a combination of these systems. Epicardial systems were the first systems studied clinically, and they are only briefly discussed here, because they have been supplanted by less invasive endocardial systems. Current technology using endocardial lead systems is the main focus of this chapter.

Epicardial ICD Lead Systems: Historical Perspective Since the pioneering work of Michel Mirowski in the 1970s, rapid advancements in lead technology have occurred. With epicardial systems, the surgical procedures included median sternotomy, lateral thoracotomy, subxiphoid, and subcostal approaches, and transvenous access also was used for placement of an additional shocking lead in the superior vena cava (SVC) or an endocardial sensing lead in the right ventricle (RV). The first human implants used a titanium spring electrode positioned in the SVC and an epicardial patch.6,7 Initially, sensing was performed using the 161

162

Section One: Basic Principles of Device Therapy

high-voltage electrodes, but this resulted in sensing problems, leading to the use of separate epicardial or endocardial rate sensing leads to address this problem.8 Complications of the initially implanted defibrillation lead systems included migration of the SVC coil,9 which could have an impact on defibrillator function. The SVC

electrode/epicardial patch electrode configuration was supplanted by the epicardial patch/patch electrode configuration, which demonstrated better defibrillation efficacy and avoided the problem of SVC lead migration.10 Figure 4-1 illustrates a fully epicardial patch lead system, including epicardial rate sensing leads.

Defibrillation connector

Defibrillation patch electrode

A

B

C

Rate-sensing electrodes

Epicardial patch

D Figure 4-1. Epicardial patch lead systems. A, Schematic illustration of an oval-shaped epicardial patch electrode (St. Jude/Ventritex, Sunnyvale, Calif.). B, Photograph of a rectangular-shaped patch (Guidant/CPI). C, Lateral radiograph demonstrating epicardial defibrillation patches. D, Schematic drawing showing epicardial defibrillation patches and epicardial sensing leads. (From Accorti PR: Leads technology. In Singer I [ed]: Implantable Cardioverter Defibrillator. Armonk, NY, Futura, 1994, with permission.)

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

Epicardial lead implantation requires a thoracotomy and is associated with significant morbidity and mortality. The perioperative surgical mortality rate has been reported to be as high as 5%.11-16 Some of this mortality may be seen in patients undergoing other concomitant cardiac surgery at the time of patch lead placement. Other concerns include bacterial infections, which can spread directly from the pocket to leads on the heart, and subsequent difficulty removing epicardial patch leads due to fibrosis after long-term implantation. Later cardiac surgery, such as bypass graft surgery or valve replacement, may also be increasingly difficult if patches were placed directly on the heart, as opposed to outside the pericardium. In addition, patients with epicardial patches in place may be resistant to external defibrillation due to an increase in transthoracic impedance. A high rate of lead malfunction was identified with epicardial patch leads.17-19 Some patch leads had a failure rate of 28% at 4 years in one study.17 Many patients with fractured leads were asymptomatic and were identified during routine surveillance radiography and formal lead testing; they were most frequently detected more than 2 years after implantation.17 This finding highlighted the importance of regular lead testing in patients with epicardial systems. In addition to lead failures with identification of obvious fractures, fluid may sometimes be apparent within the lead insulation. The insulation of leads manufactured from silicone is permeable to serous fluid under normal circumstances.17 However, large elements such as cells or blood should not permeate silicone; if they do, it suggests the presence of a breach of insulation or leakage through the vulnerable parts of the leads, such as the joints or connector ends. If the electrical performance of the lead is normal, the clinical significance of the finding of blood or fluid beneath the lead is unclear as to whether this small breach could eventually result in lead dysfunction.17 Although no deaths were directly attributed to lead malfunction in two studies, unsuccessful defibrillation due to epicardial patch fracture can require external defibrillation.17,20 Symptomatic pericardial constriction may also be seen as a complication of epicardial patch placement.21 Malfunctions of epicardial pace/sense leads are common, occurring in 3.4% to 15.4% of patients,18-20,22,23 and they often manifest with inappropriate ICD discharges due to oversensing.17,23 This is frequently related to lead fracture, which may be detected by abnormally high impedance values, oversensing with delivery of inappropriate therapy, high pacing thresholds, or direct visualization.18,23 The presence of an adapter increases the risk of sensing lead problems in epicardial systems.22,24,25 There is also a high incidence of elevated pacing thresholds or complete loss of capture with epicardial pacing.26

Endocardial ICD Lead Systems Although the concept of transvenous defibrillation was first developed by the pioneering work of Michel

163

Mirowski27 and early studies established the feasibility of successful defibrillation using endocardial leads,28,29 widespread application was limited because of high defibrillation energy requirements and lead complications. These problems were later addressed by further advances in lead and especially device technology. The development of a lead that incorporates both shock and sensing functions permitted widespread use of transvenous systems. Initially, subcutaneous patches were routinely used as part of the defibrillation system. After lead design advances and improvements in defibrillation energy requirements with biphasic shocks became available, implantation of complete transvenous systems became more common. There are several potential advantages to an endocardial system. A transvenous or nonthoracotomy lead system obviates the need for a thoracotomy procedure and reduces morbidity and mortality associated with the procedure. Elimination of the need for a thoracotomy shortens hospital stay and makes convalescence easier, with reduced patient discomfort. Comparisons of epicardial and endocardial lead systems have demonstrated that endocardial systems are as effective as epicardial systems with respect to successful termination of spontaneous ventricular tachyarrhythmias.30,31 Although lead dislodgment and pocket infection were more frequent with endocardial systems, perioperative mortality was higher with epicardial systems.30 Endocardial systems can be implanted with a mortality rate of less than or equal to 1.0%,32-34 whereas the perioperative mortality rate with epicardial leads can be 2 to 5 times greater.12-16,26,34 General Description of ICD Lead Technology In addition to sensing and pacing, ICD leads have the function of delivering high-voltage shocks for defibrillation therapy. Despite some similarities, ICD leads must be designed somewhat differently from pacemaker leads. Like pacemaker leads, ICD leads are composed of electrodes, conductors, insulation, connectors, and a fixation mechanism. The initial endocardial leads were large in caliber,35 measuring 12F, and were subsequently reduced to 6.6F to 8.2F in size. One of the earliest endocardial leads was produced by Intec/CPI (Pittsburgh) and was tested in the 1980s.36 The Intec endovascular lead had a 16-mm2 distal platinum-iridium tip used for pacing and sensing functions. Distal 4.3-cm2 and proximal 8.5-cm2 platinum electrodes with helically wound coils were used for defibrillation, and this lead was the predecessor of the Endotak endocardial lead manufactured by Guidant/CPI (Boston Scientific, Natick, Mass.). Since the first implantations in 1988 and its approval in 1993, the Endotak lead has undergone several modifications. The current lead specifications are outlined in Table 4-1A. Early endocardial leads from three different manufacturers are demonstrated in Figure 4-2. Other manufacturers—including Medtronic, Inc. (Minneapolis, Minn.); St. Jude Medical (St. Paul, Minn.); Telectronics (now a division of St. Jude Medical,

164

Section One: Basic Principles of Device Therapy

TABLE 4-1A.

Specifications of Guidant/CPI Defibrillation Leads Endotak 60 0060/62/64

Endotak 70 0070/72/74

Endotak 0073/75 0113/15

Endotak DSP 90 0092/93/95

Endotak DSP 120 0123/25

Endotak Endurance 0134/35/36

Endotak Endurance Rx 0144/45/46

Endotak Endurance EZ 0154/55/56

Defibrillation coils

RV/SVC

RV/SVC

RV/SVC

RV/SVC

RV/SVC

RV/SVC

RV/SVC

RV/SVC

Fixation mechanism

Tines

Tines

Tines

Tines

Tines

Tines

Tines

Screw

Sensing

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Tip electrode mm2

9

9

9

8

8

2

2

6

Steroid

No

No

No

No

No

No

Yes

Yes

Insulation

Silicone

Silicone

Silicone

Silicone

Silicone

Silicone

Silicone

Silicone

Coil electrode coating

NA

NA

NA

NA

NA

NA

NA

NA

Lead lengths (cm)

100

100

100, 70

100

70

64, 70, 100

64, 70, 100

64, 70, 100

Lead body diameter (F)

9.7/12

9.7/12

9.7/12

8.2/10

8.2/10

8.2/10

8.2/10

8.2/10

Tube design

Coaxial

Coaxial

Coaxial

Multilumen

Multilumen Multilumen

Multilumen

Multilumen

Features

Connector terminal Pace/sense

4.75 mm

4.75 mm

4.75 mm

IS-1

IS-1

IS-1

IS-1

IS-1

High voltage

6.1 mm

6.1 mm

6.1 mm

DF-1

DF-1

DF-1

DF-1

DF-1

Interelectrode spacing Tip-Ring

NA

NA

NA

NA

NA

NA

NA

NA

Tip-RV coil (mm)

6

12

12

12

12

12

12

12

Tip-SVC coil (cm)

10/13/16

12/15/18

15/18

18

18

18

18

18

Electrode surface area RV (mm2)

295

379

379

450

450

450

450

450

SVC (mm2)

617

617

617

660

660

660

660

660

Integ, integrated; NA, not applicable; ePTFE, an expanded polytetrafluoroethylene polymer; RV, right ventricle; SVC, superior vena cava.

Sylmar, Calif.); Intermedics, Inc. (Boston Scientific, Natick, Mass.); and Biotronik GmbH & Co. (Berlin, Germany)—have also developed endocardial leads demonstrating defibrillation efficacy. The characteristics of these endocardial leads are summarized in Tables 4-1B through 4-1D. In addition to the Guidant/CPI Endotak lead, which is a single lead composed of two separate endocardial defibrillation coils and was the first nonthoracotomy lead approved in the United States, the use of two separate catheter endocardial braided defibrillation electrodes, one positioned at the RV apex and the other in the SVC or high right atrium also demonstrated defibrillation efficacy (Telectronics). These leads were used in conjunction with a subcutaneous patch electrode positioned on the left

lateral thorax at the midclavicular to midaxillary line in the second to sixth intercostal spaces.37 Other manufacturers, including Medtronic and St. Jude, have also marketed defibrillation systems with separate endocardial RV and SVC coils (see Table 4-1). Since the early 1990s, ICDs have been implanted like pacemakers in the infraclavicular position.38,39 This was the result of development of the transvenous lead system, as well as improvements in ICD technology that allowed a reduction in pulse generator size. Initially, ICDs were used as a therapy of last resort, but they are now considered the therapy of choice for treatment of life-threatening sustained ventricular arrhythmias, as well as in the prevention of sudden cardiac death in high-risk populations with underlying struc-

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

Endotak Reliance G 0174/75/76/77

165

Endotak Reliance G 0164/65/66

Endotak Reliance SG 0170/71/72

Endotak Reliance G 0160/61

Endotak Reliance 0157/58/59

Endotak Reliance S 0137/38

Endotak Reliance G 0184/85,86,87

Endotak Reliance SG 0180,81,82

RV/SVC

RV/SVC

RV

RV

RV/SVC

RV

RV/SVC

RV

Tines

Screw

Tines

Screw

Screw

Screw

Screw

Screw

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

Integ bipolar

2

5.7

2

5.7

5.7

5.7

5.7

5.7

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Silicone

Silicone

Silicone

Silicone

Silicone

Silicone

Silicone

Silicone

ePTFE

ePTFE

ePTFE

ePTFE

ePTFE

ePTFE

ePTFE

ePTFE

59, 64, 70, 90

59, 64, 70

59, 64, 70

59, 64

59, 64, 90

59, 64

59, 64, 70, 90

59, 64, 70

8.2

8.2

8.2

8.2

8.2

8.2

8.2

8.2

Multilumen

Multilumen

Multilumen

Multilumen

Multilumen

Multilumen

Multilumen

Multilumen

IS-1

IS-1

IS-1

IS-1

IS-1

IS-1

IS-1

IS-1

DF-1

DF-1

DF-1

DF-1

DF-1

DF-1

DF-1

DF-1

NA

NA

NA

NA

NA

NA

NA

NA

12

12

12

12

12

12

12

12

18

18

NA

NA

18

NA

18

NA

450

450

450

450

450

450

450

450

660

660

NA

NA

660

NA

660

NA

tural heart disease.40 The transvenous lead technology has had a major impact on this development. Integrated Bipolar Versus Dedicated Bipolar Leads The two types of endocardial defibrillation leads are true bipolar leads and integrated bipolar leads. ICD leads that have true bipolar sensing and pacing capabilities have separate tip and ring electrodes designed for sensing and pacing purposes, similar to permanent pacemaker leads. They have the best sensing and pacing performances and are less likely to manifest sensing errors. However, true bipolar leads have a longer distance from the tip to the distal defibrillation

coil in the RV, compared with integrated leads. ICD leads with integrated sensing and pacing capabilities have only a single tip electrode, and the distal defibrillation coil is used as the anode for sensing and pacing functions. The electrograms (EGMs) obtained from an integrated bipolar lead have a more “unipolar” appearance than those recorded from a dedicated bipolar lead. One advantage of ICD leads with integrated bipolar sensing and pacing functions is that there is a shorter tip-to-defibrillation coil distance. The placement of the defibrillation coil closer to the myocardium of the right ventricle and septum may represent an advantage with respect to defibrillation success. However, integrated bipolar leads are more susceptible to sensing problems.41 Text continued on p. 170

RV Tines True bipolar

Defibrillation coils

Fixation mechanism

Sensing

Yes Polyurethane Silicone 58, 65, 75, 100 6.6 Multilumen

Steroid

Insulation

Coil electrode coating

Lead lengths (cm)

Lead body diameter (F)

Tube design

NA

NA

513

NA

12

8

DF-1

IS-1

Multilumen

6.6

58, 65, 75, 100

Silicone

Polyurethane

Yes

4.2

True bipolar

Screw

RV

Sprint Fidelis 6931

NA, not applicable; RV, right ventricle; SVC, superior vena cava.

SVC (mm )

2

RV (mm2) 513

NA

Tip-SVC coil (cm)

Electrode surface area

12

Tip-RV coil (mm)

Tip-Ring (mm)

8

DF-1

High voltage

Interelectrode spacing

IS-1

Pace/sense

Connector terminal

2.5

Tip electrode (mm )

2

Sprint Fidelis 6930

663

513

18

12

8

DF-1

IS-1

Multilumen

6.6

58, 65, 75, 100

Silicone

Polyurethane

Yes

2.5

True bipolar

Tines

RV/SVC

Sprint Fidelis 6948

Specifications of Medtronic Defibrillation Leads

Features

TABLE 4-1B.

663

513

18

12

8

DF-1

IS-1

Multilumen

6.6

58, 65, 75, 100

Silicone

Polyurethane

Yes

4.2

True bipolar

Screw

RV/SVC

Sprint Fidelis 6949

819

585

18

12

8

DF-1

IS-1

Multilumen

8.2

58, 65, 75, 100

Silicone

Polyurethane

Yes

1.6

True bipolar

Tines

RV/SVC

Sprint Quattro 6944

860

614

18

12

8

DF-1

IS-1

Multilumen

8.6

58, 65, 75, 100

Silicone

Polyurethane

Yes

5.7

True bipolar

Screw

RV/SVC

Sprint Quattro Secure 6947

NA

426

NA

25

12

DF-1

IS-1

Coaxial

10.5

65, 75, 110

None

Polyurethane

—

10

True bipolar

Screw

RV

Transvene 6936

166 Section One: Basic Principles of Device Therapy

RV/SVC

Tines

True bipolar

5

Yes

Silicone

None

65

6.7

Multilumen

Defibrillation coils

Fixation mechanism

Sensing

Tip electrode (mm2)

Steroid

Insulation

Coil electrode coating

Lead lengths (cm)

Lead body diameter (F)

Tube design

DF-1

High voltage

17

1570:17 1571:21

Tip-RV coil (mm)

Tip-SVC coil (cm)

663

SVC (mm2) NA

414

NA

17

11

DF-1

IS-1

Multilumen

6.7

60, 65

None

Silicone

Yes

5

True bipolar

Tines

RV

Riata 1572

663

414

1580:17 1581:21

17

11

DF-1

IS-1

Multilumen

7.6

60, 65, 75

None

Silicone

Yes

8

True bipolar

Screw

RV/SVC

Riata 1580/81

NA

414

NA

17

11

DF-1

IS-1

Multilumen

7.6

60, 65, 75

None

Silicone

Yes

8

True bipolar

Screw

RV

Riata 1582

Integ, integrated; NA, not applicable; RV, right ventricle, SVC, superior vena cava.

414

RV (mm2)

Electrode surface area

11

Tip-Ring (mm)

Interelectrode spacing

IS-1

Pace/sense

Connector terminal

Riata 1570/71

663

414

1560:17 1561:21

11

NA

DF-1

IS-1

Multilumen

6.7

60, 65, 75

None

Silicone

Yes

5

Integ bipolar

Tines

RV/SVC

Riata i 1560/61

NA

414

NA

11

NA

DF-1

IS-1

Multilumen

6.7

60, 65, 75

None

Silicone

Yes

5

Integ bipolar

Tines

RV

Riata i 1562

Specifications of St. Jude/Ventritex Defibrillation Leads

Features

TABLE 4-1C.

663

414

1590:17 1591:21

11

NA

DF-1

IS-1

Multilumen

6.7

60, 65, 75

None

Silicone

Yes

8

Integ bipolar

Screw

RV/SVC

Riata i 1590/91

NA

414

NA

11

NA

DF-1

IS-1

Multilumen

6.7

60, 65, 75

None

Silicone

Yes

8

Integ bipolar

Screw

RV

Riata i 1592

NA

470

NA

11

NA

DF-1

IS-1

Coaxial

10.5

67

None

Silicone

No

6

Integ bipolar

Tines

RV

TVL RV02

NA

470

NA

11

NA

DF-1

IS-1

Coaxial

10.5

110

None

Silicone

No

6

Integ bipolar

Screw

RV

TVL RV01/1101

671

480

21

11

NA

DF-1

IS-1

Multilumen

6.7

70

None

Silicone

No

6

Integ bipolar

Tines

RV/SVC

SPL SP01/SP02

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

167

Tines

True bipolar

6

No

Silicone

None

75, 100

10.5

Multilumen

Fixation mechanism

Sensing

Tip electrode (mm2)

Steroid

Insulation

Coil electrode coating

Lead lengths (cm)

Lead body diameter (F)

Tube design

DF-1

High voltage

17

13, 16, 18

Tip-RV coil (mm)

Tip-SVC coil (cm)

600

2

NA

300

NA

15

9

DF-1

IS-1

Multilumen

8.7

75

None

Silicone

No

5

True bipolar

Tines

RV

Biotronik Kainox RV 124-005

NA

300

NA

21

14

DF-1

IS-1

Multilumen

10.5

74

None

Silicone

No

5.3

True bipolar

Screw

RV

Biotronik Kainox RV-S 124-574

NA

310

NA

16

9

DF-1

IS-1

Multilumen

9.3

65, 75

Yes

Silicone

Yes

1.8

True bipolar

Tines

RV

Biotronik Kentrox RV 348090,91

IROX, iridium oxide; NA, not applicable; RV, right ventricle; SVC, superior vena cava.

SVC (mm )

320

RV (mm2)

Electrode surface area

9

Tip-Ring (mm)

Interelectrode spacing

IS-1

Pace/sense

Connector terminal

RV/SVC

Biotronik Kainox SL 217,8,38,9,40

480

310

16, 18

16

9

DF-1

IS-1

Multilumen

9.3

65, 75, 100

Yes

Silicone

Yes

1.8

True bipolar

Tines

RV/SVC

Biotronik Kentrox SL 7351-4,9,50

NA

310

NA

18.5

11.5

DF-1

IS-1

Multilumen

9.3

65

Yes

Silicone

Yes

8.2

True bipolar

Screw

RV

Biotronik Kentrox RV-S 343080

Specifications of Biotronik and Intermedics Defibrillation Leads

Defibrillation coils

Features

TABLE 4-1D.

480

310

16, 18

18.5

11.5

DF-1

IS-1

Multilumen

9.3

65

Yes

Silicone

Yes

8.2

True bipolar

Screw

RV/SVC

Biotronik Kentrox SL-S 345988,9

NA

440/880

—

6 mm

—

4.0 mm

2 mm

—

11

100

None

Silicone

—

10

—

Screw

RV

Intermedics 497-05/06

NA

440

—

—

—

DF-1

IS-1

—

11

—

IROX

Silicone

—

8

—

Screw

RV

Intermedics 497-19/20

NA

440

—

8 mm/6 mm

—

DF-1

IS-1

—

11

60, 70, 100

IROX

Silicone

—

8

—

Screw

RV

Intermedics 497-23/24

168 Section One: Basic Principles of Device Therapy

B

D

Figure 4-2. Photographs and a radiograph of several early transvenous ICD leads from various manufacturers. A, Endotak DSP lead. (Courtesy of Guidant/CPI, Boston Scientific, Natick, Mass.) B, Sprint 6942 lead. (Courtesy of Medtronic, Inc., Minneapolis, Minn.) C, SPL lead. (Courtesy of St. Jude/ Ventritex, Sunnyvale, Calif.) D, Chest radiograph demonstrating the appearance of a dual-coil Guidant transvenous lead in a patient requiring ICD implantation.

C

A

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

169

170

Section One: Basic Principles of Device Therapy

Previous investigation demonstrated that the ranges of EGM amplitudes recorded in sinus rhythm and ventricular arrhythmias were not significantly different with true bipolar versus integrated bipolar leads.42 Filtering of the postshock EGM results in a dramatic reduction of R-wave amplitudes, whereas unfiltered amplitudes are similar to the preshock values. In contrast to integrated bipolar sensing, this effect is minimized for true bipolar sensing.42 Proximity of the defibrillation electrode to the sensing tip of the lead may result in conduction block during shock delivery.42,43 This could effect postshock sensing of ventricular fibrillation (VF).44 Delay with capture has also been a concern after shock delivery, because of a temporary increase in pacing threshold. Devices can now be programmed to higher pacing outputs after shock delivery to compensate for this effect. Single- and Dual-Coil Defibrillation Leads Single-coil defibrillation leads have one coil that lies in the right ventricle, which typically is used as the defibrillation cathode. When used in conjunction with an active or “hot” can defibrillator, the shell of the device is typically the defibrillation anode. However, polarity can be reversed noninvasively in the newer-generation ICDs. Dual-coil defibrillation leads have an additional proximal coil that typically lies within the junction of the high right atrium and the SVC. The proximal coil can be used alone as the defibrillation anode, or it can be combined with an active can as a combined defibrillation anode. One advantage of dual-coil systems may be an improvement in defibrillation threshold (DFT) compared with single-coil systems. Steroid-Eluting Leads The benefits of steroid-eluting leads have been demonstrated for permanent pacemaker leads.45-48 Steroid (e.g., dexamethasone sodium phosphate) elution at the electrode-tissue interface has helped maintain chronically low stimulation thresholds.45,47,48 This allows programmed devices to be set at a low pacing output, which has potential implications for device longevity.46 Steroid-eluting leads appear to reduce the inflammatory response and attenuate the connective tissue reaction at the tip of the lead, resulting in a decrease in the pacing threshold peak.45,49 In contrast to pacing threshold, steroid-eluting leads were not found to influence chronic ventricular EGM amplitude, compared with leads without steroid elution.50 Defibrillation lead systems now incorporate steroid elution. Sensing Oversensing and Undersensing In addition to the defibrillation function of ICD leads, it is equally important for the lead to provide reliable sensing of intracardiac signals, in order to detect and differentiate ventricular tachyarrhythmias from sinus rhythm. At the time of implantation, initial R-wave

testing in sinus rhythm is performed, aiming for an Rwave amplitude of 5 mV or greater in most instances. A significant relationship has been demonstrated between the ventricular EGM amplitude in sinus rhythm and the amplitude in VF.51 During epicardial device implantation, one study suggested that up to a fourfold decrease may be seen in mean ventricular EGM amplitude during VF compared with sinus rhythm, although a wide variation among individuals was seen.51 This may be useful in the initial assessment of sensing lead function at the time of implantation testing, and it has implications with respect to programming device sensitivity for detection of VF. The sensing issues related to defibrillation leads have been the focus of several papers.52-54 In addition to sensing issues related to lead problems, undersensing of ventricular tachycardia in one report53 was related to the ICD generator itself, with use of fixed-gain sensing. The system’s ability to sense depends on the quality of the EGM; the input amplifier, including filtering and gain; and the detection algorithm.42 Although it is crucial to sense small-amplitude signals during VF, it is also important for the defibrillator system to avoid oversensing of other intracardiac signals (e.g., T waves) or extracardiac signals (e.g., noise, myopotentials) during sinus rhythm. Although the role of the device amplifiers and detection algorithms is clearly crucial for appropriate sensing, high-quality lead performance and stability of the EGM are also essential. Endocardial leads may be associated with fibrotic reactions within the endocardium.55,56 This may result in a deterioration of the intracardiac EGMs. In addition, fibrotic reactions associated with endocardial leads and cumulative acute damage produced by defibrillator discharges may lead to new arrhythmogenic foci, changes in pacing threshold or DFTs, or future difficulty with lead extraction.55,56 An Internet-based registry of pacemaker and ICD pulse generators and leads revealed that disrupted insulation accounted for 29% of ICD lead failures.57 Failures of the conductor, electrode, fixation mechanism, or pin were identified as other causes of lead failure. Oversensing with delivery of inappropriate shocks was the most frequent consequence of ICD lead failure. Other signs and symptoms of lead failure included high or low impedance, noncapture, undersensing, and failure to defibrillate.57 Routine follow-up to document impedance and sense/pace functions may detect many unexpected lead failures before they are manifested clinically. One concern related to endocardial lead systems was the potential failure to redetect VF after an appropriately delivered but unsuccessful first shock.44,58-61 Redetection of VF after a failed first shock demonstrates longer redetection times with integrated lead sensing than with dedicated bipolar leads. This is probably due to the high-voltage field that results in local myocardial stunning, which is most apparent immediately after a shock.62 One study demonstrated that there are voltageand time-dependent reductions in postshock R-wave amplitude and that integrated bipolar systems appear to be more affected than true bipolar systems in this

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

regard. This may be a consequence of lead design, specifically the distance of the distal defibrillating coil from the rate-sensing cathode, with the distal defibrillating coil used as the rate-sensing anode.63 Rather than a uniform reduction of intracardiac EGM amplitude, other investigations have demonstrated that a rapid and repetitive change in amplitude occurred after a shock, resulting in failure to redetect VF.44,59,60 The problem appeared to improve when the spacing between the rate-sensing tip of the electrode and the distal coil was increased.44,59,62 The redetection malfunction was never seen in patients with integrated bipolar systems when the distal coil-to-ring electrode distance was greater than 6 mm.59 It should be noted that the distal end of the coil is separated from the tip electrode by only 6 mm in the Guidant Endotak C 60 series, by 12 mm in the Endotak 70 series, and by 11 mm in the Ventritex RV-1101 lead. The sensing algorithm in specific devices may also contribute to sensing problems, with marked variation of signal amplitude leading to inappropriate redetection of sinus rhythm in a device with an amplifier that has automatic gain control.60 The amplifier may be adjusted for optimal sensing of the predominate signal amplitude, and these large-amplitude signals may prevent the automatic gain amplifier from increasing the gain to a setting necessary to detect the low-amplitude signals, leading to signal dropout and failure to redetect VF after an unsuccessful first shock. Newer leads and detection algorithms may now reduce the chance of postshock undersensing. Early Detection of Sensing Problems The newer generation of ICDs may have several features designed to detect and alert the patient and physician to early lead problems. The use of short interval counters to monitor nonphysiologic R-R intervals ( 6 mo) 12 ± 8 14 ± 10

20

11 ± 7

Thakur et al. 1995 (39)

0

15 ± 6

Jones et al. 1995 (104)

15

Bardy et al. 1993 (33)

21 ± 10

*Specific lead complications typically include endocardial lead dislodgment, right ventricular perforation, insulation break, electrode/conductor fracture, connector failure or loose set-screw, adapter failure, patch migration, crinkled patch, and venous thrombosis. **High complication rates with old CPI Endotak C and SQ patch. † Compares abdominal implantation sites using integrated bipolar CPI Endotak lead (N = 140) for sensing and defibrillation with CPI BT-10 (N = 107) sensing lead, both tunneled to the abdomen. ‡ High failure rate refers to Medtronic 6936 coaxial, true bipolar, polyurethane active-fixation lead.

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

to be 2% to 4% in some large series.92-96 A higher complication rate of 7% to 16% involving endocardial leads or patches has also been reported in other studies.18,22,97-105 An exceptionally high complication rate of 56% was seen in a single study in which Endotak C leads were used in conjunction with subcutaneous patch electrodes.106 The reported incidence of lead-related complications varies widely and may be related to patient selection, lead used, mode of implantation, and particularly duration of follow-up. The risk of failure tended to increase over time93,99 and appeared to be particularly high during the first year after reoperation for battery depletion.99 Problems appeared to be more frequent when adapters between the pulse generator and the lead or extenders were used.22 Other factors potentially related to lead failure may include lead design, venous approach, type of insulation, abdominal location, and size of the pulse generator. Routine screening radiography appears to have a limited role in the diagnosis of lead complications after the first postimplantation month, and most complications are diagnosed clinically.97 It should also be noted that insulation failure may be associated with normal values of impedance, sensing, and pacing threshold at the time when lead failure is diagnosed.99 One study suggested that visible insulation breaks on silicone sensing leads may be repaired with the use of silicone sleeves and adhesive, demonstrating that there was no difference in subsequent lead survival in patients with lead repair compared with those who underwent insertion of new sensing leads.107 Acute Lead Failures Acute lead failures may involve lead dislodgment or high stimulation thresholds, often requiring surgical intervention to correct the problem. A higher dislodgment rate may be seen with transvenous systems that use a separate SVC coil, because these defibrillation coils do not incorporate a fixation mechanism.19 Dislodgment of the SVC lead may lead to an increase in the DFT19 or may be asymptomatic and detectable by routine radiography (Fig. 4-13). Earlier investigations reported lead migration or dislodgment occurring in 10% of cases.33 The incidence of lead dislodgment in patients who underwent implantation later in the study improved (3%), reflecting improved implantation technique and more experience with anchoring methods.33 A high incidence of lead dislodgment may be seen with defibrillation leads placed in the CS.33,108,109 Lead dislodgment may be asymptomatic, or it may result in oversensing or undersensing, as well as ineffective defibrillation. Acute lead failure may also be caused by cardiac perforation. In addition to symptoms of pain, which is often pleuritic, patients with perforation may develop high pacing thresholds or noncapture, undersensing, diaphragmatic pacing, or tamponade. Oversensing of noise may be seen with loose set-screws or loose connectors103 and may occur early after implantation. These problems can be easily treated and corrected, if detected early.

181

Figure 4-13. Radiograph demonstrating superior vena cava (SVC) lead dislodgment. The tip of the SVC and the proximal portion of the SVC coil are marked by arrows. The lead has dislodged into the inferior vena cava in this “hybrid” nonthoracotomy lead system, with a right ventricular (RV) sensing lead and subcutaneous patch also visualized on this radiograph.

Chronic Lead Failures Chronic lead failure is most often a result of insulation defects or fractured conductors. Signs and symptoms of lead failure are outlined in Table 4-5. Conductor fractures can lead to an increase in pacing impedance, intermittent loss of sensing or pacing, complete failure to sense or pace, or oversensing. Insulation defects can also lead to oversensing or undersensing problems. If the insulation defect involves the outer insulation and the stress point is located at an extracardiac site, stimulation of the pectoral muscle or diaphragm may also be seen. If the inner insulation is also involved, intermittent or complete loss of pacing may also occur.41 Due to similar clinical manifestations in some cases of lead and insulation problems, the exact cause of lead failure may not always be apparent, particularly before surgical intervention. Polyurethane leads with a coaxial design are susceptible to inner insulation failure and shorts between the conductors, often caused by metal ion oxidation.41 Silicone-coated leads with multilumen design are prone to erosions of the insulation, particularly in abdominal locations with larger devices. Friction of the lead due

182

Section One: Basic Principles of Device Therapy

Symptoms and Signs of Lead Problems TABLE 4-5.

I. Sensing lead problems A. Oversensing (most common) due to detection of noise Inappropriate shocks Inappropriate aborted shocks Inhibition of pacing Change in pacing impedance Prolonged oversensing immediately after shock delivery B. Undersensing in SR C. Undersensing or nonsensing of VT/VF, with failure to deliver therapy D. Elevated pacing thresholds or noncapture II. Defibrillation lead problems A. Ineffective defibrillation B. Change in high voltage impedance C. Sudden death

to contact with the ICD generator itself can result in abrasion of the insulation.41 Continuous or forceful traction of the lead may also result in thinning of the insulation and potential tearing of the insulation,41 emphasizing the importance of correct lead handling and the potential contribution of certain daily strenuous activities on lead reliability. Lead compression under the clavicle may result in conductor failure as well as damage to the insulation. A lateral or cephalic vein access approach may help reduce subclavian crush injuries or lead fractures.41,110,111 Data regarding the long-term reliability of multilumen defibrillator leads are limited, because these leads have been in use for less than 10 years. Previous reports regarding polyurethane coaxial leads with subpectoral ICDs showed an increase in failure rates after 4 to 5 years.99,112 The Medtronic Transvene single-coil, coaxial polyurethane leads (model numbers 6936/6966) comprised 54% of ICD lead failures detected by the multicenter registry published by Hauser and colleagues.112 The mean time to failure for these leads was 4.8 ± 2.1 years.112 Lead survival analysis of the Medtronic polyurethane coaxial 6963 lead revealed a cumulative failure probability of 37% at a mean follow-up of 69 months in another study.64 Insulation defects, pace/sense conductor fractures, or high-voltage coil fracture may be involved in the structural failure of these particular leads.112 (The middle insulation layer of the 6936 lead is 80A polyurethane, which is the same polyurethane used to manufacture Medtronic 4004 and 4012 leads, and insulation degradation presumably occurs over time in high-stress areas.) The incidence of lead failure increased over time with Medtronic coaxial polyurethane leads,99,112,64 with the risk of lead failure being particularly high during the first year after generator replacement due to battery depletion in one study.99 One study compared the incidence of structural failure in a dedicated bipolar sensing lead (CPI BT-10) with that in an integrated lead (CPI Endotak) in 247 patients undergoing abdominal ICD implantation.94

Over a mean follow-up of 29 ± 15 months, there were 17.8% lead failures with the BT-10 lead (occurring 261 to 1505 days after implantation) but only 4.3% with the Endotak lead (occurring 410 to 1211 days after implantation; P < .01).94 Lead failure was believed to be related to insulation defects in all cases, with the problem occurring in the proximal portion of the lead within the generator pocket in all but one case. The BT-10 lead is a standard silicone lead that is almost identical to those used in permanent pacing systems for years. Lead failure with the BT-10 lead continued to occur as long as 4 years after implantation, suggesting that the problem with lead failure may be progressive and that higher failure rates may be seen with longer follow-up. It should be noted that all implants were abdominal, and the tunneling technique was the same for both leads. One possible reason for lead failure is the mechanical force of the heavy pulse generator against a nonreinforced lead, resulting in insulation wear; a lower rate of problems is seen with the more reinforced, larger Endotak lead. Abdominal ICD implantations appear to have a higher associated lead failure rate, and the rate increases progressively from the time of implantation.41,113 Mechanical stresses placed on the lead from tunneling to the abdomen and wear on the lead from a heavy pulse generator may contribute to insulation problems.62,94,98 Other studies also demonstrated a higher failure rate in abdominal devices, with lead complications occurring in 15% to 16% of abdominal devices100,104 and 4% of pectoral devices,96 with higher rates of lead complications in devices with dualchamber pacing capabilities, during a mean follow-up period of only 21 months. The first-generation endocardial defibrillation lead system (CPI Endotak-C), often used in conjunction with a subcutaneous patch, was associated with an extremely high rate of complications.106 During a followup period of only 51 ± 36 weeks, 56% (5 of 9) patients developed lead-related complications. These complications included conductor fractures, leading to sudden death in one patient. Patch electrode conductor fractures were also observed. Because of an unacceptable incidence of lead fracture in the earlier versions of endocardial leads,106 the high-energy conductor was changed from a coiled wire to Teflon-coated braided flexible cable, which appeared to solve this problem.114 The management of sensing lead failure in patients with Endotak integrated bipolar endocardial defibrillator leads remains controversial. In one study, a new sensing lead alone was placed in 9 of 10 patients, assuming the “independent” function of the sensing component in the circuit based on lead design.115 However, failure of defibrillation occurred in one patient 3 years later, raising concern about the longterm reliability of defibrillation once a sensing lead problem is diagnosed in patients with integrated bipolar leads. Sensing lead complications may manifest with inappropriate shock therapy or aborted shocks due to sensing of noise, complete loss of pacing, or asynchronous pacing with undersensing during sinus rhythm,

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

or they may be initially asymptomatic and detected at the time of generator replacement or chest radiography.19,20,22,52,64,92,94,98,99,103,106,110-112,115 In extreme cases, complete loss of sensing may lead to nondetection of ventricular arrhythmias, requiring resuscitation and external defibrillation or sudden cardiac death.20,106 One study also described prolonged oversensing of electrical noise immediately after shock delivery.64 Defects with the defibrillation portion of the lead may result in failure to defibrillate ventricular arrhythmias.99 The true incidence of endocardial defibrillation lead failure will require continued follow-up, because more lead problems are likely to be detected over time. This emphasizes the need for continued close follow-up of patients with endocardial systems. Lead failure may result from conductor and/or insulation damage just beyond the stiff terminal ring of the IS-1 rate-sensing terminal connector of one particular Guidant Endotak lead (DSP model 0125).116 Lead failure occurred in 3.5% of patients who underwent pectoral ICD implantation using a submuscular approach, after a mean follow-up of only 16 ± 11 months.116 This was related to extension of the metal terminal ring beyond the header when a smaller pulse generator was used, creating a rigid fulcrum in this area. Flexing of the lead in this region could create sufficient stress to disrupt insulation or conductors at this point. Lead problems manifested with asymptomatic oversensing or inappropriate shocks in these patients. This lead was placed on advisory with a technical memorandum reporting the potential problems with the long IS-1 terminal pin connector in 1999. Other complications of endocardial lead systems include infections, with a reported incidence of up to 5%.20,34,93,97,98,100-102,108,117,118 Subclavian venous thrombosis or obstruction remains a potential long-term problem, particularly in patients who have multiple leads in place. Complications may also be seen with subcutaneous patches, including fractures, dislodgment, crinkling, erosion, and hematomas.19,20,30,100,101,109,117,118 Fractures of patch leads may be initially asymptomatic and may be detectable by chest radiography.19 Despite improvements in ICD lead technology, it is anticipated that the lead failure rate will progressively increase over time. As more patients have ICDs implanted for primary as well as secondary prevention, the need for lead removal will probably also increase over time. Reporting of Lead Failure At present, there are inadequate reporting mechanisms to identify lead failure in the ICD industry. Device manufacturers currently provide product performance reports that are designed to supply information to physicians regarding device system survivability and to alert clinicians to particular devices that may be failing at a higher rate than expected, perhaps warranting closer follow-up. Returned product analysis (RPA) provides a means for manufacturers to examine products that have been explanted and returned to the manufacturer to determine the mechanism of failure.

183

However, most leads that have failed are not explanted, so reports of failure will be grossly underestimated by this reporting mechanism.119 An ongoing prospective study, the Tachyarrhythmia Chronic System Study (TCSS) is being conducted at 11 medical centers in the United States, and specific follow-up findings and reports are required at 6-month intervals. Centers provide information regarding clinical observations related to the lead, as well as clinical responses, including whether the lead was explanted, abandoned, or replaced, or whether reprogramming was performed to rectify the problem. This study does not require return of the lead for analysis to be counted as a “failing lead.” The proportion of nonperforming leads detected by the RPA was 0.5%, compared with 2.2% in the TCSS (P < .001).119 The prospective follow-up of chronic leads by the TCSS, which includes identification of clinical performance at the time of follow-up, provides a more accurate basis for determination of ICD lead failure than the RPA reports do.

Biventricular Pacing: Cardiac Resynchronization Therapy Leads Multiple clinical trials have demonstrated the benefit of CRT in the treatment of patients with congestive heart failure, LV dysfunction, and a wide QRS, with and without ICD backup.120-127 Hemodynamics and heart failure symptoms are improved with CRT in patients with class III or IV heart failure and an ejection fraction of 35% or less. The branches of the CS are typically visualized through CS angiography, with contrast dye injected through a balloon-tipped catheter or specially designed guiding sheath (Fig. 4-14). LV leads are placed through the CS into venous branches to pace the LV and resynchronize myocardial contraction. The site of LV pacing may play an important role in the efficacy of CRT.128,129 Therefore, it is important to be able to direct these leads into specific branches. This can be accomplished by using leads that have a variety of shapes, which can be altered by use of a stylet-driven lead system. An example of a Medtronic unipolar LV lead with a preshaped distal curve to increase stability in the CS is shown in Figure 4-15. An over-the-wire lead system can be used.130,131 When the stylet is inserted into the distal tip of the lead, the lead is straightened. As the stylet is withdrawn, the lead assumes its shape with a curve on the end. Some leads have a combination of stylet-driven and over-the-wire capabilities. Table 4-6 outlines the currently available LV leads. It should be noted that LV leads may have a higher rate of lead dislodgment than RV or RA leads. This is probably related to the absence of any “active” fixation mechanism. Some LV leads do have “tines” (Fig. 4-16), but the absence of trabeculation within the venous branches may still limit stability of these leads, compared with RV leads. The size of the venous branch may also affect lead stability. In addition to 7F and 6F leads, some LV leads are now available in 4F to accommodate smaller venous branches. More recently

184

Section One: Basic Principles of Device Therapy

Attain OTW

A

Guidant Easytrak

Figure 4-14. Coronary sinus angiography. Contrast dye is injected through a balloon-tipped catheter to enable visualization of the venous branches of the coronary sinus, to help guide left ventricular lead placement.

B Figure 4-16. Comparison of left ventricular leads from various manufacturers. A, The Medtronic Attain over-the-wire lead is shown on top of the Guidant Easytrak lead. This particular Guidant lead is straight and has “tines” for stability. B, The St. Jude Aescula 1055K lead has a multicurved or “S” shape to help with lead stability in the coronary sinus. (Courtesy of Medtronic, Inc., Minneapolis, Minn.)

Figure 4-15. Over-the-wire left ventricular lead (Attain OTW 4194). The preshaped distal curve is constructed to increase stability. When a stylet is introduced into the lead, the lead is straightened. As the stylet is withdrawn, the lead assumes its shape with the curve on the end. (Courtesy of Medtronic, Inc., Minneapolis, Minn.)

designed leads assume a variety of shapes after the stylet is removed, which may help with lead stability. Figure 4-16 demonstrates the variety of different lead shapes available from various manufacturers. In addition to lead dislodgment, another complication of LV lead placement is diaphragmatic pacing. This

is related to stimulation of the phrenic nerve, which may be in close proximity to the more lateral venous branches. The newer, smaller-caliber leads tend to have lower pacing thresholds, and therefore the pacing output may be minimized, often reducing or eliminating diaphragmatic pacing. The initial LV leads were unipolar, with one electrode on the distal tip of the lead. Pacing occurred between the LV tip and the RV coil. This sometimes resulted in anodal capture. Newer LV lead technology includes bipolar LV pacing capabilities, with a more proximal ring electrode on the LV lead. This allows multiple pacing configurations, which can be programmed noninvasively through the newer ICDs with CRT capabilities, leaving additional options in an attempt to reduce diaphragmatic pacing or improve LV pacing thresholds (Fig. 4-17). Because some earlier ICDs did not have the availability of separate RV sensing, or an external adapter was used to connect RV and LV leads for pacing, the sensing circuit included both RV and LV EGMs. This could lead to “double-counting” of the RV and LV EGMs, with inappropriate shock delivery for sinus

6 4.8 Straight Tines Unipolar LV-1 (10-13) IS-1 (37, 38) Silicone with poly No Yes 65, 72, 80, 90 3.5 NA Platinum iridium

Delivery

Proximal lead body diameter (F)

Distal lead body diameter (F)

Distal shape

Fixation

Electrode configuration/polarity

Connector terminal

Insulation

Tip seal

Steroid

Lead lengths (cm)

Electrode surface area (mm2)

Electrode spacing (mm)

Tip electrode material

Platinum iridium

11

4.2 (proximal) 4 (distal)

65, 80, 90, 100

Yes

No

Silicone with eTFE/poly

LV-1 (15-20) IS-1 (42-44)

Bipolar

Tines

Straight

5.4

6

OTW

Platinum iridium

11

9 (proximal) 8.5 (distal)

65, 80, 90, 100

Yes

No

Silicone with eTFE/poly

LV-1 (22-27) IS-1 (48-50)

Bipolar

3D spiral

Spiral

5.7

6.3

OTW

Guidant Easy Trak 3 4522/24/25/ 27/48/49/50

Platinum

NA

5.8

58, 65, 75

No

NA

Poly 55D

IS-1

Unipolar

Curved

Curved

—

6

Stylet

Medtronic Attain 2187

Platinum

NA

5.8

78, 88, 103

Yes

Yes

Poly 55D

IS-1

Unipolar

Angled

Angled

5.4

4

OTW and Stylet

Medtronic Attain 4193

Platinum

11

5.8 (proximal) 38 (distal)

78, 88

Yes

Yes

Poly 55D

IS-1

Bipolar

Angled

Angled

5.4

6.2

OTW and Stylet

Medtronic Attain 4194

eTFE, expanded tetrafluoroethylene; NA, not applicable; OTW, over the wire; poly, polyurethane; Pt/Ir, platinum-iridium; TiN, titanium nitride fractal coating.

OTW

Features

Guidant Easy Trak 2 4515/17/18/ 20/42/43/44

Specifications of Left Ventricular Pacing Leads Guidant Easy Trak 4510/11/12/ 13/37/38

TABLE 4-6.

TiN

NA

6.8

75

No

NA

Silicone

IS-1

Unipolar

Curved

S curved

4.5

4.7

Stylet

St. Jude Aescula 1055 K

Pt/Ir,TiN

NA

4.8

75, 86

Yes

No

Silicone with poly

IS-1

Unipolar

Curved

S curved

5.6

5

OTW and Stylet

St. Jude QuickSite 1056 K

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

185

186

Section One: Basic Principles of Device Therapy Figure 4-17. Bipolar left ventricular (LV) lead. The availability of the more proximal ring on the LV lead (Attain 4194) allows a variety of pacing options using enhanced programmability within the InSync pulse generator. Pacing may still be performed between the LV tip and the right ventricular coil. In addition, pacing can be performed from the LV tip to the LV ring. (Courtesy of Medtronic, Inc., Minneapolis, Minn.)

tachycardia or atrial arrhythmias (Fig. 4-18). With the current generation of CRT devices, separate RV sensing has eliminated this problem. As noted earlier, steroid elution is frequently used with permanent pacing leads and with ICD leads. Steroid-eluting CS leads are also available. Satisfactory pacing and sensing parameters have been demonstrated acutely and during follow-up with these leads. The pacing threshold stabilized 2 weeks after implantation, and the sensing threshold remained stable from implantation to 4-month follow-up in these patients.132

Lead Follow-up Regular outpatient follow-up is necessary to identify and anticipate potential problems with ICD leads. Routine follow-up every 3 to 4 months is recommended, with additional follow-up as needed if ICD therapy is delivered or if symptoms, such as syncope or presyncope, develop in the interim. Interrogation reveals pacing lead impedance values, and pacing threshold and sensing can be evaluated. Current devices also have the ability to perform a lead integrity test to measure the impedance within the high-voltage lead system. Highvoltage values greater than 100 Ω suggest conductor fracture, and values lower than 20 Ω suggest insulation failure. In addition, after shock delivery, a large disparity between the programmed energy and the delivered energy suggests lead conductor failure. The current generation of devices also has the availability of intracardiac EGMs that can be reviewed in real time to exclude evidence for “noise,” which may be apparent in patients with insulation breaks or conductor fracture (Fig. 4-19). Chest radiographs, although

they may play a limited role, can be used to exclude lead dislodgment, in the event that a significant change in sensing or pacing parameters is noted. Follow-up DFT testing may also be considered in patients who are found to have significant changes from baseline in certain lead parameters, such as sensing and impedance values, to confirm lead integrity and appropriate sensing of ventricular tachyarrhythmias.

Lead Extraction With the increasing numbers of ICDs implanted, it is anticipated that there will be an increasing need for ICD lead extraction. ICD lead extraction may be more complex than pacemaker lead extraction, because ICD leads are bulkier and shocking electrodes may uncoil with extraction techniques.133 There may also be a greater fibrotic reaction around the shocking coils, due to the irregularity of the coils, leaving more space for inflammation and fibrous tissue ingrowth (Fig. 4-20A). Effective lead extraction has been demonstrated with a variety of extraction systems, including intravascular traction techniques133,134 and laser extraction.134,135 Improvements in lead technology may also play a role in lead extraction. It is believed that the ease of extraction may be improved with new Gore (expanded polytetrafluoroethylene [ePTFE]–coated leads, due to the reduction in the degree of fibrous ingrowth (see Fig. 4-20B). This lead has a coating of ePTFE, an expanded polytetrafluoroethylene polymer, which is electrically inert without having any effect on DFTs. This coating prevents tissue ingrowth at the shocking electrode surfaces. This is evident by the absence of fibroblastic growth in the outer coil filars during preclinical studies (see Fig. 4-20B).

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

187

Figure 4-18. Double-counting of the ventricular EGM during sinus tachycardia. The earlier ICDs did not have separate right ventricular (RV) sensing, or an external adapter was sometimes used to connect the RV and left ventricular (LV) leads for pacing. This could lead to “double-counting” of the RV and LV EGMs, with inappropriate shock delivery for sinus tachycardia. With the current cardiac resynchronization therapy (CRT) devices, separate RV sensing is available, eliminating this problem. AS, atrial sensed event; VF, ventricular fibrillation zone sensed event; VT, ventricular tachycardia zone sensed event.

Figure 4-19. Intracardiac EGMs with electrical noise. Very short R-R intervals with electrical noise, shown on the stored intracardiac EGMs, demonstrate evidence for a lead integrity problem (e.g., lead fracture) resulting in ICD shock delivery. CD, charge delivered; CE, charge ended; FD, fibrillation detection; FS, fibrillation zone sensed event; TS, ventricular tachycardia zone sensed event; VS, ventricular sensed event.

Future Advances Lead development has focused on new materials and designs, size reductions, improved steering and handling, and maintenance of long-term reliability. The IS-1 pacing and DF-1 shocking standards for the lead/device interface may be updated to an IS-4 terminal, to allow a reduction in connector/header size. Newer leads have high-impedance electrode technology, which may help to maximize the longevity of ICD

generators.136 Additional investigation should help determine ways to further increase the long-term reliability of endocardial leads, and newer designs may aid in easier lead extraction.

Summary Over the past decade, ICDs have become the treatment of choice for patients with sustained ventricular

188

Section One: Basic Principles of Device Therapy

One possible solution to this problem is development of a leadless ICD system. A prototype of this system (Cameron Health, Inc., San Clemente, Calif.) is currently undergoing investigational studies abroad and should be available for investigational use soon in the United States.137 Alternatively, development of lead systems with better durability and greater resistance to damage is needed. This subchapter highlights the need for close clinical follow-up of all current ICD systems and the importance of follow-up performed by skilled personnel who have extensive training to ensure early detection of lead problems. This may affect not only patient quality of life but also long-term morbidity and mortality in the ICD population. A

PERMANENT PACEMAKER LEADS: GENERAL CONCEPTS

B

100 mm Figure 4-20. Fibrous ingrowth and lead histology. A, There appears to be a marked fibrous reaction around the shocking coils of defibrillation leads. The irregularity of the shocking coils leaves more space for inflammation and fibrous tissue ingrowth. B, The new Gore-coated (ePTFE) defibrillation lead (Guidant) appears to have less fibrous ingrowth, which may enhance the ease of lead extraction. (Courtesy of Guidant, Boston Scientific, Natick, Mass.)

tachyarrhythmias, in addition to playing a very important role in the primary prevention of sudden cardiac death. Technological advances in ICD therapy, including the development and refinement of ICD lead technology, have been largely responsible for these changes in the treatment of ventricular arrhythmias. However, lead failures are still an important complication of ICD therapy, despite technological advances. Insulation defects and fractured conductors or electrodes are common lead complications that have important clinical implications with respect to ICD efficacy as well as patient comfort and quality of life. The high failure rate of endocardial leads adds greatly to follow-up costs, including costs related to rehospitalization and reoperation. Although advancements have been made in the field of lead extraction, removal of any chronically implanted lead still carries a risk of significant complications. Additional advancements in lead technology are still needed, especially because many devices are implanted prophylactically, and more patients receiving devices are younger and will require multiple generator replacements throughout their lifetime.

Pacemaker lead development and advances in technology preceded the development of ICD leads by many years, setting the stage for current advances with the more complex ICD leads, which require defibrillation circuits in addition to bradycardia stimulation. Both permanent pacing systems and defibrillation lead systems need to offer reliable sensing and pacing. A variety of factors influence lead performance and pacing threshold. In addition to lead and electrode factors, myocardial factors (e.g., fibrosis, infarction), drugs, and electrolytes can influence pacing thresholds. Lead and electrode factors to be considered include the time since the electrode was implanted, unipolar versus bipolar electrode, electrode surface area and shape, electrode material, lead insulation, lead fixation, and steroid elution. Lead and electrode factors are discussed in more detail in the following sections.

Lead Polarity: Unipolar Versus Bipolar Electrodes According to strict definition, all pacemaker circuits are bipolar, and electrons flow from the cathode to the anode. In reference to pacemaker leads, the terms “unipolar” and “bipolar” refer to the number of electrodes that have contact within the heart.138 Figure 4-21 demonstrates that a unipolar lead has one electrode at the tip of the lead (the cathode or negatively charged end), which is in contact with the heart muscle. The pulse generator serves as the anode (or positively charged end). Current flows from the negatively charged tip of the lead to the heart muscle and then to the positively charged pulse generator, completing the circuit. Figure 4-21 also demonstrates a bipolar pacemaker lead, in which both electrodes are in the heart. The tip electrode is the negatively charged cathode. A proximal electrode ring, which is a short distance from the tip electrode, serves as the positively charged anode.

189

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads Figure 4-21. Unipolar and bipolar pacemaker leads. A unipolar lead has only one electrode, the cathode, at the tip of the lead, which lies within the heart. The anode lies on the surface of the pulse generator. A bipolar lead has both poles on the end of the lead, located close to each other. In this diagram, the cathode is at the tip of the lead, and the anode is slightly proximal to the cathode. (From Mond HG: Unipolar versus bipolar pacing: Poles apart. PACE 14:1411, 1991, with permission.)

⫺

ANODE

Cathode

Unipolar

⫹ Anode

⫺ Cathode

Bipolar

There has been a continued trend toward use of bipolar pacing leads in the United States. Initially, bipolar leads were large, stiff, and difficult to implant. However, with advances in technology, bipolar leads are now smaller and easier to implant, similar to unipolar leads.138 Both types of leads have demonstrated reliability with pacing and low pacing thresholds.139 With modern pacemaker systems, similar ventricular EGM amplitudes have been seen with unipolar and bipolar sensing configurations.139,140 Bipolar sensing has been shown to be similar to unipolar R-wave sensing during acute testing or at follow-up in most patients.140 The unipolar system has a larger interelectrode distance, which is the distance between the distal tip electrode and the pulse generator. For this reason, the unipolar sensing system will “see” more of the heart to detect the intracardiac signal. Unipolar systems are more susceptible to crosstalk, such as inappropriate sensing of the atrial stimulus artifact by the ventricular channel in a dual-chamber pacing system, which can result in the inhibition of the ventricular output. In a pacemaker-dependent patient, this could be lifethreatening.138 Crosstalk is less common with bipolar systems because of the smaller amplitude of the pacing stimulus.138 Unipolar pacing systems are also more likely than bipolar systems to result in oversensing of skeletal muscle myopotentials, resulting in inhibition of pacing.139-143 A bipolar lead, with its more closely spaced electrodes, is less likely to record far-field electrical signals than a unipolar lead. For example, a bipolar lead is less likely to oversense ventricular EGMs in the atrial channel. These sensed far-field electrical signals may result in inappropriate mode switching with current pacemaker algorithms. Because of the proximity of skeletal muscle to the pulse generator in a unipolar pacing system, skeletal muscle stimulation can occur during unipolar pacing, especially at high outputs, which is not typically seen with bipolar pacing systems.138,139 It is particularly important for ICD systems, which now have complete bradycardia sensing and pacing capabilities, to be bipolar. This minimizes inappropri-

ate oversensing of far-field intracardiac or extracardiac signals in order to prevent inappropriate defibrillator therapy. Previously, a separate pacemaker was implanted with an ICD to provide bradycardia backup pacing. This required special testing at the time of implantation to evaluate for interactions between the ICD and the permanent pacemaker, to assure the absence of oversensing of pacemaker spikes, which could lead to the withholding of defibrillator therapy during VF. After delivery of an ICD shock, some permanent pacemaker pulse generators revert to unipolar pacing. Therefore, only dedicated bipolar pulse generators were recommended for use in the patient who has an implanted defibrillator. A unipolar pacing stimulus artifact is much larger than a bipolar pacing artifact on surface electrocardiography, as shown in Figure 4-22. At times, the bipolar pacing stimulus may be difficult to see on the surface 12-lead ECG (see Fig. 4-22B). The bipolar pacing stimulus may be particularly difficult to see on transtelephonic monitoring performed for pacemaker follow-up, because usually only one surface lead recording is available, and the paced beats are sometimes difficult to distinguish from underlying conducted beats with a wide QRS.

Electrode Size The earlier pacemaker lead electrodes had a large surface area in contact with the endocardium, resulting in a low pacing lead impedance and high current drain. The current was distributed over a large surface area, often resulting in a higher pacing threshold. With the goal of increasing pulse generator longevity, the cathodal surface area was reduced in size. As the cathodal surface area was reduced, higher current densities resulted, and a lower pacing threshold was obtained with a higher impedance (400-800 Ω) at the electrodetissue interface.144-146 Subsequently, leads with smaller distal electrodes (4-6 mm2) and higher impedance

190

Section One: Basic Principles of Device Therapy

values were developed.147-148 Eventually, distal electrodes with very small surface areas (1.2 mm2) and even higher impedance values (>1000 Ω) were developed.149-151 (Reducing the size of the electrode results in an increased pacing impedance. A decrease in current is predicted, as predicted by Ohm’s law, I = V/R.) These

newer, high-impedance leads also had very low pacing thresholds and resulted in less current drain (Fig. 4-23),149,150,152,153 which was expected to lead to better pacemaker generator longevity. The improved pacemaker battery longevity with high-impedance leads, compared with standard impedance leads, has been confirmed.152 Sensing characteristics of these high-impedance leads are similar to those of standard pacing leads.149,152,153 A relatively low dislodgment rate of the small-surface (1.2 mm2), highimpedance pacing lead was reported in one investigation in which careful lead positioning and securing techniques were used.151

Polarization

A

B Figure 4-22. Unipolar and bipolar pacing stimuli. A, A large pacing artifact is visualized on 12-lead electrocardiography in a unipolar pacing system. B, Small pacing artifacts are seen in bipolar pacing systems.

Polarization refers to the electrochemical impedance generated at the electrode-tissue interface. The ideal lead has a low polarization, in addition to a high electrode impedance and low resistance of the conductor.154 The polarization effect increases as the electrode surface area is reduced. The accumulation of ions in the myocardium gives rise to an afterpotential, and this is recorded after a pacing stimulus.154 Polarization occurs at the electrode-myocardium interface due to the electrical current from the tip of the electrode to the tissue. After the stimulus is delivered, electrical signals are generated due to ion movement in the extracellular spaces around the electrodes. Certain pacemaker pulse generators have a function called AutoCapture, which allows automatic measurement of pacing thresholds. A low-amplitude pacing stimulus (just above the pacing threshold) is delivered while the device assesses whether the stimulus has captured the myocardium. By maintaining a stimulus amplitude that is only 0.5 to 0.7 V above threshold, the

Ventricular Current Drain at 2.5V 6.0 Control High

5.5 5.0

Current Drain (mA)

4.5 4.0

* **

*

*

*

3.5 3.0 2.5 2.0 1.5 1.0 0.5

p⬍0.0005

0.0 0 2 4

12

26 Follow-up Time in Weeks

52

Figure 4-23. Ventricular current drain at 2.5 V, comparing a high-impedance lead (solid line) with a standard lead (dashed line). A higher current drain was demonstrated at each follow-up visit up to 1 year in control leads having a larger (5.8 mm2) distal electrode, compared with the high-impedance lead with a smaller (1.2 mm2) distal electrode. *P < .001. (From Ellenbogen KA, Wood MA, Gilliagan DM, et al, and the Capsure Z Investigators: Steroideluting high-impedance pacing leads decrease short and long-term current drain: Results from a multicenter clinical trial. PACE 22:3948, 1999, with permission.)

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

longevity of the pacemaker generator is increased. To allow a reliable determination of capture, the device must be able to distinguish the myocardial depolarization-evoked response (ER) from the electrode polarization signal (PS). The polarization signal (PS) could be detected by the pacemaker and confused with myocardial depolarization, leading to pacing at an output below the true threshold. Autocapture is most reliable when low-polarization leads are used. Use of non–lowpolarization leads can lead to incorrect threshold determination and noncapture,155 which could be a serious problem in pacemaker-dependent patients. An additional schema that has been used to reduce the polarization artifact is to reduce the capacitance of the stimulus output capacitor, thereby improving the signal-to-noise ratio between the evoked response and the polarization artifact.

Electrode Design Porous electrodes were designed to help reduce lead dislodgment and to reduce excessive threshold rises compared with standard smooth-tip electrodes. An electrode with a porous electrode tip, composed of sintered platinum-iridium fibers having a 20-μm diameter with hemispherical shape, was examined in dogs and compared with a standard solid electrode of similar size.156 Porous electrodes, which have this complex surface structure, have low-polarization properties.156 Histologic examination demonstrated tissue ingrowth throughout the electrode interior and a fibrotic capsule that was about one half the thickness of the capsule seen with solid electrodes. R-wave stability and improved anchoring were also seen in dogs with the porous electrode.156 This design allows the geometric area to be small, with a large total electrode surface area that includes the internal spaces within the electrode pores (Fig. 4-24).157 Electrodes with porous designs have been shown to have better stimulation thresholds in humans.158 Active-fixation leads may also include a

Figure 4-24. The sintered porous electrode (left) and the laser porous electrode (right). (From Hirshorn MS, Holley LK, Skalsky M, et al: Characteristics of advanced porous and textured surface pacemaker electrodes. PACE 6:525-536, 1983, with permission.)

191

porous electrode used as a surface cathode. The screw is used as the fixation mechanism, but it may be either electrically active or inert, being used or not used for sensing or pacing.159 Porous electrode tips can also be created by fabricating a laser-drilled platinum cap, which is treated to increase the external surface area, as shown in Figure 4-24. Lead tips are texturized or etched to enlarge the active surface area and reduce polarization properties.157 The laser porous electrode demonstrated improved fixation by tissue ingrowth, when compared with smooth electrodes, in sheep.157

Electrode Material The choice of metals used for electrodes is important in terms of long-term function, including minimizing corrosion and avoiding an excessive foreign-body reaction and excessive fibrosis. Platinum and platinumiridium electrodes have been used. An alloy of cobalt, iron, chromium, molybdenum, nickel, and manganese (Elgiloy) has been used for cathodes. Microscopic corrosion has been noted with platinum-iridium and Elgiloy electrodes, but no consistent clinical adverse effects were observed.160 Corrosion was directly related to the duration of implantation of the electrode.160 Carbon has also been used for cathodes.161 Vitreous carbon has excellent mechanical strength and biocompatibility, and is inert in body tissues.161 Carbon-tipped electrodes resulted in lower chronic stimulation thresholds when compared with platinum or Elgiloy.161,162 Excellent stimulation thresholds have been achieved with carbon electrodes.163,164

Lead Conductors The conductor of the lead is a wire that conducts electrical current from the pacemaker generator to the tip electrode to allow pacing. Unipolar leads have one conductor, and bipolar leads have two. Details regarding conductor materials and standard design were discussed in the ICD section. Bipolar leads require two conductors. The original design used a parallel arrangement; subsequent advancements in technology included the coaxial design. The coaxial design allowed introduction of a stylet which is passed through the inner lumen (Fig. 4-25). A more recent development of coated wire technology has allowed the development of “thin bipolar leads.” The thin lead has an outside diameter similar to that of a unipolar lead. This is possible because the single strands of conductor wire are coated with a thin layer of ethylene tetrafluoroethylene (ETFE) insulation. Therefore, two or more conductors can be placed into a single coil to form the lead body165,166 (see Fig. 4-25). Excellent acute and chronic stimulation thresholds at follow-up are obtained, with a relatively low short-term failure rate.166

192

Section One: Basic Principles of Device Therapy

Thin line

Multifilar single coil Outer insulation

Anode coil

Integral insulation Inner insulation

Coaxial bipolar lead Conventional bipolar lead

Cathode coil

Figure 4-25. Multifilar, single-coil, coated lead (top) and conventional bipolar lead (bottom). (From Mond HG, Grenz D: Implantable transvenous pacing leads: The shape of things to come. PACE 27:887-893, 2004, with permission.)

Insulation Insulation material extends from the tip of the lead to the lead connector. As with defibrillator leads, pacemaker leads are composed of polyurethane or silicone. Silicone was used for pacemaker leads in the 1960s, and polyurethane was introduced in 1979.167 Silicone Rubber Silicone has a high biocompatibility and biostability but a relatively low tear strength.154 Therefore, it can easily be damaged during implantation, especially by sharp tools or tight sutures. Therefore, a thick layer of silicone is necessary, making the lead diameter much larger than leads composed of polyethylene.167 Silicone also has a high coefficient of friction, making it difficult to pass one lead next to another during implantation from a single subclavian stick with the original version of the silicone lead.167 To overcome this, leads can be coated with a more lubricious material to allow easier handling. A stronger silicone was manufactured in later years, allowing silicone leads to be thinner than originally constructed. Polyurethanes Polyurethanes have an excellent biocompatibility and were initially believed to have a good biostability.167 However, leads with polyurethane insulation are stiff. Polyurethane has high tensile and tear strengths, making it possible to construct leads of smaller diameter than silicone leads.167 Unlike silicone, polyurethane also has a high flexibility and a low coefficient of friction,154,168 so that two leads can be passed more easily side-by-side during implantation. A poor long-term performance has been noted with some chemical varieties of polyurethane, whereas excellent long-term reliability has been observed with other forms.168-173 The probability of failure of one Medtronic bipolar polyurethane ventricular lead

(Model 4012 insulated with Pellethane 80A) was 20.9% at 6 years after implantation by Kaplan-Meier analysis.171 Surface cracking and insulation failure have been noted. Clinically, insulation failure may manifest with muscle stimulation, oversensing, undersensing, loss of capture, reduced lead impedance, or premature battery depletion.169,172-174 An early sign of diminishing EGM amplitude predicted lead failure several months before clinical manifestations.174 Leads composed of Pellethane 80A, a polyurethane insulation material, were shown to have insulation degradation, or environmental stress cracking, when explanted.169 Environmental stress cracking is caused by the tendency of soft and hard segments of the lead to separate, causing damage to the surface.167 More severe damage may occur in areas of the lead that are exposed to stress or strain, such as the region near the lead connector or the site of the ligature.175 Another area of stress is the region between the clavicle and the first rib, when a subclavian puncture is used for access.176 Leads with Pellethane 80A insulation are more prone to insulation failure when the subclavian venous approach is used, rather than the cephalic vein approach.177 It was suggested that pacemakerdependent patients with earlier implanted polyurethane leads should be given consideration for prophylactic replacement because of the high failure rate.171,172 Metal-induced oxidation is another mechanism of failure for polyurethane leads. An adverse chemical reaction may occur when the conductor produces corrosion products, which can lead to oxidative degradation of polyurethane insulation.154 These adverse reactions appear to be design and model specific; they are not generic to all polyurethane leads.178 This can be minimized by coating the conductor.154 Current polyurethane leads now typically use Pellethane 55D for insulation, which is stiffer and harder than Pellethane 80A. Long-term studies of leads using Pellethane 55D have shown better performance compared with Pellethane 80A.154,179 The newer leads appear to be less susceptible to environmental stress cracking.167 However, because of the stiffness of Pellethane 55D leads, they may be more prone to RV perforation.179 Leads may now also have a combination of silicone and polyurethane insulation materials. To avoid excessive stiffness at the tip of the lead, the distal end of the lead may include silicone for insulation. In addition, silicone may be placed between the conductor and the polyurethane, in an attempt to minimize metal-induced oxidation.154 Subsequent studies suggested improved reliability for other leads made from polyurethane insulation.180 It has been suggested that the early failure rate was related to the design of particular leads, rather than the insulation material itself.180,181 Low-stress designs using noncorrosive coil wire have a good reported clinical history, without the high failure rate of designs in which tensile stress was not minimized.181 Nevertheless, the preponderance of data suggests that Pellethane 80A is an unacceptable insulation material for pacing and ICD leads. A clinically significant difference related to lead insulation is the response of the lead to electrocautery.

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

Silicone rubber is very resistant to the thermal effects of electrocautery, a major advantage at the time of pulse generator replacement. In contrast, polyurethane may melt when exposed to electrocautery, and great care must be taken at the time of operation to minimize heating of the insulation, to avoid insulation breakdown.

Fixation Mechanisms Endocardial leads may have active- or passive-fixation mechanisms to allow attachment of the lead to the endocardial surface. The current passive-fixation leads typically use tines, located immediately behind the electrode, to allow attachment to the endocardium. The tines are extensions of the insulation material; they protrude backward just proximal to the tip and are designed to become entrapped within the trabeculae in the heart (Fig. 4-26). In addition to tines, other passivefixation mechanisms tried in the past included wedge tips or flanges, wings, and fins.154,167 Tined leads have a low rate of lead dislodgment.182 Compared with olderstyle wedge-tip leads, tined ventricular leads were superior with respect to a significantly lower dislodgment rate and fewer reoperations.183 Active-fixation leads have a screw at the distal end. The screw in these leads may or may not be an electrically active component of the lead. The original screws were fixed and unprotected, placing venous and endocardial structures at risk of damage as the lead was passed through the vasculature and heart. To minimize this hazard, a mannitol covering over the screw was used by one manufacturer, which was designed to dissolve within 5 minutes. After the capsule dissolves, the screw on the tip of the lead can be fixed in the atrium or ventricle by rotating the lead over the stylet.184,185 The rate of dissolution varied as a function of flow, taking approximately 3 minutes to dissolve in high-flow areas and up to 10 minutes in the case of stagnant flow.185 The more common active-fixation mechanism now involves a retractable screw184,186 (Fig. 4-27). A clothespin-type instrument is used to extend and retract the screw during implantation by turning the connector pin. All active-fixation leads result in some trauma to

the endocardium and myocardium, and steroid elution at the base of the screw has resulted in significantly improved stimulation thresholds in active-fixation leads,187,188 similar to the improvement seen with passive-fixation leads. Advantages of active-fixation leads include the ability to place the lead at various sites in the ventricle or atrium and a low dislodgment rate.167,188 Activefixation leads are also easier to extract than passivefixation leads. However, both types perform well, with low rates of dislodgment with current systems. The selection of an active or passive mechanism is now largely based on physician preference.

Electrode-Tissue Interface A normal rise in stimulation threshold occurs after lead implantation. This is caused by an inflammatory response at the electrode-tissue interface.189 A higher stimulus output may be programmed within the first 3 months after implantation to allow an acceptable safety margin for pacing in case an excessive increase in stimulation threshold is noted. A fibrous capsule forms at the electrode-tissue interface, and the stimulation threshold eventually falls, but typically it remains at a level higher than the implantation value. The design of the pacemaker lead is one important factor in determining the amount of inflammatory response at the electrode-tissue interface. It is important that the electrode be in a stable position against the tissue surface to avoid excessive irritation, which should help to minimize the inflammatory response. A smaller distal tip of the lead is also important in reducing the inflammatory response. Bipolar leads are stiffer than unipolar leads, and bipolar leads made with polyurethane insulation are much stiffer than bipolar leads made from silicone rubber.179 The stiffer the distal segment of the lead, the greater the force on the endocardium, which may increase the risk of perforation.179 Softer insulation materials at the tip of the lead might also help to reduce trauma and therefore reduce the inflammatory response. Steroid-eluting electrodes have been effective in reducing inflammation and reducing the peak and chronic stimulation thresholds.

Platinum/iridium ring-tip electrode

Figure 4-26. A schematic representation of a tined lead. The small tines are extensions of the insulation that are located just proximal to the tip electrode. (From Mond H, Sloman G: The small tined pacemaker lead: Absence of dislodgment. PACE 3:171-177, 1980, with permission.)

193

Silicone rubber insulation

Silicone rubber Silicone rubber tines

194

Section One: Basic Principles of Device Therapy

6 5 7 3

4

2 1 Figure 4-27. A schematic representation of a retractable screw-in lead. On the left portion of the diagram the tip is protracted, and on the right it is retracted. 1, helical electrode; 2, front section; 3, radiopaque ring; 4, crimp bus; 5, sealing ring; 6, conductor coil; 7, insulation tubing. (From Bisping HJ, Kreuzer J, Birkenheier H: Three year clinical experience with a new endocardial screwin lead with introduction protection for use in the atrium and ventricle. PACE 3:424-435, 1980, with permission.)

Steroid-Eluting Leads Steroid-eluting leads were designed to reduce or eliminate the early and late rises in pacing threshold that were seen with permanent pacemaker leads. Dexamethasone sodium phosphate has been used in the steroid-eluting electrode systems.189,190 Multiple studies have demonstrated the effectiveness of steroid elution in obtaining very low acute and chronic stimulation thresholds in the atrium and ventricle, with nearelimination of the acute rise in stimulation threshold early after implantation.187,190-192 The steroid effect is believed to result from a reduction in the inflammatory response at the electrode tip. Both passive- and active-fixation leads may incorporate steroid elution.187,193 A passive-fixation lead (Medtronic CapSure SP) with steroid elution demonstrated a much improved and “flat” response with respect to pacing threshold during 18 months of followup, when compared with another passive-fixation, non–steroid-eluting lead (Telectronics Encor) and with active-fixation leads (Medtronic Bisping, electrically active; Telectronics Accufix, electrically inactive, non–steroid-eluting)159 (Fig. 4-28). The ability to pace at low outputs chronically has clear ramifications with respect to current drain and pulse generator longevity. Excellent sensing characteristics have also been noted with steroid-eluting leads.192-194

Lead Recalls: Accufix Issues In November of 1994, a Telectronics atrial lead was recalled because of reported fractures noted in the J-

shaped retention wire used in Accufix leads (Model 330-801 and 329-701). The electrically inactive Jretention wire lies beneath the outer insulation; it is welded at the distal end to the anode and is free at its proximal end. If a fracture occurs, the lead typically functions normally. However, the retention wire may protrude through the lead insulation and may cause damage to cardiac or other vascular structures. In addition, this wire may embolize to the lung vasculature. Cardiac or other vascular tears have led to deaths related to this lead. Periodic fluoroscopy was recommended to evaluate the lead and exclude protrusion of the retention wire.195-196 Clinical manifestations of retention wire fracture included pericardial effusion, tamponade, intracardiac thrombosis, pulmonary embolism, tricuspid valve laceration, aortic erosion, and migration of the fractured retention wire to the lungs.197-199 It has been suggested that the shape of the Accufix lead after implantation is a strong predictor of fracture, and that a lower threshold for lead extraction is indicated in patients with non–J-shaped leads.196,200,201 It has also been suggested that leads with retention-wire fracture should be extracted, with continued regular fluoroscopic screening for those leads that demonstrated a normal fluoroscopic appearance.202 One study reported that the incidence of retention-wire fracture in the Accufix 330801 lead was 25.6%.202 It was initially suggested that many of these leads should be extracted, particularly in younger patients and in those with a more open appearance of the J wire.203 The Multicenter Clinical Study of patients with Accufix leads was designed to determine the rate of injury due to the J retention wire and the outcome of lead extraction. In this study of 2589 patients with Accufix atrial leads who were monitored by cinefluo-

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads Figure 4-28. Bipolar stimulation thresholds for specific active- and passivefixation leads. The Medtronic CapSure SP lead (passive fixation, steroid-eluting) had better pacing thresholds (P < .05) after implantation at all periods of follow-up, compared with another passive, non–steroid-eluting lead (Telectronics Encor) and with active fixation leads (Medtronic Bisping and Telectronics Accufix). (From Mond H, Hua W, Wang CC: Atrial pacing leads: The clinical contribution of steroid elution. PACE 18:1601, 1995, with permission.)

Volts

195

215 Leads

4.0 3.5

Telectronics 330-801 (ACCUFIX) 2.63

3.0

3.08

3.92

3.40

1.31

2.5

Medtronics 4058 (BISPING) 0.45

2.0

x 1.5

0.65

1.0 x

0.09

0.76

0.58

1.00 0.80

x 0.14

0.11

x

x

0.12 0.13 Telectronics 330-854 (ENCOR) 0.04

0.04

x 0.17 0.09

0.5 Medtronic 5524M (CAPSURE SP) 0.0 0

6

12

18

Months Postimplantation

roscopic imaging, the risk of J-wire fracture was approximately 5.6% per year at 5 years and 4.7% per year at 10 years after implantation. Lead extraction was found to carry higher risks, suggesting that a conservative approach is indicated for most patients.195 The overall advantage of a more conservative approach was confirmed in another study.196 The hazard of J-wire fracture decreased over time after implantation.201 Telectronics Encor leads also have J-shaped retention wires. However, the wires are located deeper, within the cathode conductor, as opposed to the wire of the Accufix leads, which lies just beneath the outer insulation.154 Fracture and protrusion of the wire in the Encor leads appear to have been related to the way the lead was implanted, with trauma during implantation potentially leading to deformity of the lead. Fluoroscopy was recommended on at least one occasion for these leads.203

Conclusions Considerable advances have been made in the construction and design of permanent pacemaker leads. Improvements include the development of small, porous, steroid-eluting leads, which result in better long-term pacing thresholds, as well as leads with better reliability and long-term safety. Future advances may include additional improvements in lead insulation and advances that enhance the ease of lead extraction for older leads that are no longer functional. Continued improvements in lead technology are still warranted to further increase the reliability and longterm durability of endocardial leads, particularly for leads implanted in younger patients, who will require multiple generator replacements throughout their lifetime.

REFERENCES 1. Moss AJ, Hall WJ, Cannom DS, et al: Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. N Engl J Med 335:19331940, 1996. 2. Buxton AE, Lee KL, Fisher JD, et al., and the Multicenter Unsustained Tachycardia Trial Investigators: A randomized study of the prevention of sudden death in patients with coronary artery disease. N Engl J Med 341:1882-1890, 1999. 3. Moss AJ, Zareba W, Hall J, et al: Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction (The Multicenter Automatic Defibrillator Implantation Trial II). N Engl J Med 346:877-883, 2002. 4. Antiarrhythmics Versus Implantable Defibrillator (AVID) Investigators: A comparison of antiarrhythmic drug therapy with implantable defibrillators in patients resuscitated from near fatal ventricular arrhythmias. N Engl J Med 337:1576-1583, 1997. 5. Connolly SJ, Hallstrom AP, Cappato R, et al: Meta-analysis of the implantable cardioverter defibrillator secondary prevention trials. AVID, CASH and CIDS studies. Antiarrhythmics Versus Implantable Defibrillator study. Cardiac Arrest Study Hamburg. Canadian Implantable Defibrillator Study. Eur Heart J 21:20712078, 2000. 6. Mirowski M, Mower MM, Reid PR, Watkins L Jr: Implantable automatic defibrillators: Their potential in prevention of sudden coronary death. Ann N Y Acad Sci 382:371-380, 1982. 7. Mirowski M, Reid PR, Mower MM, et al: Termination of malignant ventricular arrhythmias with an implanted automatic defibrillator in human beings. N Engl J Med 303:322-324, 1980. 8. Reid PR, Mirowski M, Mower MM, et al: Clinical evaluation of the internal automatic cardioverter defibrillator in survivors of sudden cardiac death. Am J Cardiol 51:1608-1613, 1983. 9. Marchlinski FE, Flores BT, Buxton AE, et al: Automatic implantable cardioverter defibrillator: Efficacy, complicationas and device failures. Ann Intern Med 104:481, 1986. 10. Troup PJ, Chapman PD, Olinger GN, Kleinman LH: The implanted defibrillator: Relation of defibrillating lead configuration and clinical variables to defibrillation threshold. J Am Coll Cardiol 6:1315-1321, 1985. 11. Mirowski M: The automatic implantable cardioverter defibrillator: An overview. J Am Coll Cardiol 6:461-466, 1985. 12. Kelly PA, Cannom DS, Garan H, et al: The automatic implantable cardioverter defibrillator: Efficacy, complications and

196

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27. 28.

29.

30.

Section One: Basic Principles of Device Therapy

survival in patients with malignant ventricular arrhythmias. J Am Coll Cardiol 11:1278-1286, 1988. Gartman DM, Bardy GH, Allen MD, et al: Short-term morbidity and mortality of implantation of automatic implantable cardioverter defibrillator. J Thorac Cardiovasc Surg 100:353-359, 1990. Echt DS, Armstrong K, Schmidt P, et al: Clinical experience, complications and survival in 70 patients with the automatic implantable cardioverter defibrillator. Circulation 71:289-296, 1985. Fromer M, Brachmann J, Block M, et al: Efficacy of automatic multimodal device therapy for ventricular tachyarrhythmias as delivered by a new implantable pacing cardioverter defibrillator: Results of a European multicenter study of 102 implants. Circulation 86:363-374, 1992. Saksena S: Defibrillation threshold and perioperative mortality associated with either endocardial and epicardial defibrillation lead systems. The PCD Investigators and participating institutions. PACE 16:202-207, 1993. Brady PA, Friedman PA, Trusty JM, et al: High failure rate of epicardial implantable cardioverter defibrillator lead: Implications for long-term follow-up of patients with an implantable cardioverter defibrillator. J Am Coll Cardiol 31:616-622, 1998. Almassi GH, Olinger GN, Wetherbee JN, Fehl G: Long-term complications of implantable cardioverter defibrillator lead systems. Ann Thorac Surg 55:888-892, 1993. Korte T, Jung W, Spehl S, et al: Incidence of ICD lead related complications during long-term follow-up: Comparison of epicardial and endocardial electrode systems. PACE 18:2053-2061, 1995. Mattke S, Muller D, Markewitz A, et al: Failures of epicardial and transvenous leads for implantable cardioverter defibrilators. Am Heart J 130:1040-1044, 1995. Chevalier P, Mancada E, Canu G, et al: Symptomatic pericardial disease associated with patch electrodes of automatic implantable cardioverter defibrillator: An underestimated complication? PACE 19:2150-2152, 1996. Stambler BS, Wood MA, Damiano RJ, et al: Sensing/pacing lead complications with a newer generation implantable cardioverter defibrillator: Worldwide experience from the Guardian ATP 4210 clinical trial. J Am Coll Cardiol 23:123-132, 1994. Daoud EG, Kirsh MM, Bolling SF, et al: Incidence, presentation, diagnosis, and management of malfunctioning implantable cardioverter defibrillator rate sensing leads. Am Heart J 128:892895, 1994. Grimm W, Flores BF, Marchlinski FE: Complications of implantable cardioverter defibrillator therapy: Follow-up of 241 patients. PACE 16:218-222, 1993. Sgarbossa EB, Shewchik J, Pinski SL: Performance of implantable defibrillator pacing/sensing lead adapters. PACE 19:811814, 1996. Saksena S, Poczobutt-Johanos M, Castle LW, et al, for the Guardian Multicenter Investigators Group. Long-term multicenter experience with second generation implantable pacemaker defibrillator in patients with malignant ventricular tachyarrhythmias. J Am Coll Cardiol 19:490-499, 1992. Kastor JS: Michel Mirowski and the automatic implantable defibrillator. Am J Cardiol 63:1121-1126, 1989. Mirowski M, Mower MM, Gott VL, Brawley RK: Feasibility and effectiveness of low energy catheter defibrillation in man. Circulation 47:79-85, 1973. Zipes DP, Jackman WM, Heger JJ, et al: Clinical transvenous cardioversion of recurrent life-threatening tachyarrhythmias: Low energy synchronized cardioversion of ventricular tachycardia and termination of ventricular fibrillation in patients using a catheter electrode. Am Heart J 103:789-794, 1982. Zipes DP, Roberts D, for the Pacemaker Cardioverter-Defibrillator Investigators. Results of the international study of the implant-

31.

32.

33. 34. 35.

36.

37.

38.

39. 40.

41. 42. 43.

44.

45. 46.

47.

48. 49.

50.

51.

52.

able pacemaker cardioverter defibrillator: A comparison of epicardial and endocardial lead systems. Circulation 92:59-65, 1995. Trappe HJ, Fieguth HG, Pfitzner P, et al: Epicardial and nonthoracotomy lead systems combined with a cardioverter defibrillator. PACE 18:127-132, 1995. Bocker D, Block M, Isbruch F, et al: Do patients with an implantable defibrillator live longer? J Am Coll Cardiol 21:16381644, 1993. Bardy GH, Hofer B, Johnson G, et al: Implantable transvenous cardioverter defibrillators. Circulation 87:1152-1168, 1993. Hauser RG, Kurschinski DT, McVeigh K, et al: Clinical results with nonthoracotomy ICD systems. PACE 16:141-148, 1993. Saksena S, An H: Clinical efficacy of dual electrode systems for endocardial cardioversion of ventricular tachycardia: A prospective randomized crossover trial. Am Heart J 119:15-22, 1990. Winkle RA, Bach SM, Mead RH, et al: Comparison of defibrillation efficacy in humans using a new catheter and superior vena cava spring left ventricular patch electrode. J Am Coll Cardiol 11;365-370, 1988. Saksena S, Luceri R, Krol RB, et al: Endocardial pacing, cardioversion and defibrillation using a braided endocardial lead system. Am J Cardiol 71:834-841, 1993. Hammel D, Block M, Borggrefe M, et al: Implantation of a cardioverter/defibrillator in the subpectoral region combined with a nonthoracotomy lead system. PACE 15:367-368, 1992. Thakur RK, Ip JH, Mehta D, et al: Subpectoral implantation of ICD generators: Long-term follow-up. PACE 18:159-162, 1995. Gregoratos G, Cheitlin MD, Conill A, et al: ACC/AHA guidelines for implantation of cardiac pacemakers and antiarrhythmia devices: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelyines (Committee on Pacemaker Implantation). J Am Coll Cardiol 31:1175-1209, 1998. Gradaus R, Breithardt G, Bocker D: ICD leads: Design and chronic dysfunctions. PACE 26:649-657, 2003. Accorti PR: Leads technology. In Singer I (ed): Implantable Cardioverter Defibrillator. Armonk NY, Futura, 1994. Yabe S, Smith WM, Daubert JP, et al: Conduction disturbances caused by high current density electric fields. Circ Res 66:11901203, 1990. Callans DJ, Swarna US, Schwartzman D, et al: Postshock sensing performance in transvenous defibrillation lead systems: Analysis of detection and redetection of ventricular fibrillation. J Cardiovasc Electrophysiol 6:604-612, 1995. Mond H, Stokes K: The electrode-tissue interface: The revolutionary role of steroid elution. PACE 15:95-107, 1992. Stamato N, O’Toole MF, Fetter JG, Enger EL: The safety and efficacy of chronic ventricular pacing at 1.6 volts using a steroid eluting lead. PACE 15:248-251, 1992. Mond H, Stokes K, Helland J, et al: The porous titanium steroid eluting electrode: A double blind study assessing the stimulation threshold effects of steroid. PACE 11:214-219, 1988. Kruse M: Long-term performance of endocardial leads with steroid-eluting electrodes. PACE 9:1217-1219, 1986. Radovsky AS, van Vleet JF, Stokes KB, Tacker WA Jr: Paired comparisons of steroid-eluting and nonsteriod endocardial pacemaker leads in dogs: Electrical performance and morphologic alterations. PACE 11:1085-1094, 1988. Schuchert A, Hopf M, Kuck KH, Bleifeld W: Chronic ventricular electrograms: Do steroid-eluting leads differ from conventional leads? PACE 13:1879-1882, 1990. Leitch JW, Yee R, Klein GJ, et al: Correclation between the ventricular electrogram amplitude in sinus rhythm and in ventricular fibrillation. PACE 13:1105-1109, 1990. Almeida HF, Buckingham TA: Inappropriate implantable cardioverter defibrillator shocks secondary to sensing lead failure: Utility of stored electrograms. PACE 16:407-410, 1993.

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads 53. Sperry RE, Ellenbogen KA, Wood MA, et al: Failure of a second and third generation implantable cardioverter defibrillator to sense ventricular tachycardia: Implications for fixed-gain sensing devices. PACE 15:749-755, 1992. 54. Callans DJ, Hook BG, Kleiman RB, et al: Unique sensing errors in third-generation implantable cardioverter defibrillators. J Am Coll Cardiol 22:1135-1140, 1993. 55. Epstein AE, Kay GN, Plumb VJ, et al: Gross and microscopic pathological changes associated with nonthoracotomy implantable defibrillator leads. Circulation 98:1517-1524, 1998. 56. Epstein A, Anderson P, Kay GN, et al: Gross and microscopic changes associated with a nonthoracotomy implantable cardioverter defibrillators. PACE 15:382-386, 1992. 57. Hauser R, Hayes D, Parsonnet V, et al: Feasibility and initial results of an Internet-based pacemaker and ICD pulse generator and lead registry. PACE 24:82-87, 2001. 58. Jung W, Manz M, Moosdorf R, Luderitz B: Failure of an implantable cardioverter defibrillator to redetect ventricular fibrillation in patients with a nonthoracotomy lead system. Circulation 86:1217-1222, 1992. 59. Natale A, Sra J, Axtell K, et al: Undetected ventricular fibrillation in transvenous implantable cardioverter defibrillators: Prosepective comparison of different lead system-device combinations. Circulation 93:91-98, 1996. 60. Berul CI, Callans DJ, Schwartzman DS, et al: Comparison of initial detection and redetection of ventricular fibrillation in a transvenous defibrillator system with automatic gain control. J Am Coll Cardiol 25:431-436, 1995. 61. Goldberger JJ, Horvath G, Donovan D, et al: Detection of ventricular fibrillation by transvenous defibrillating leads: Integrated versus dedicated bipolar sensing. J Cardiovasc Electrophysiol 9:677-688, 1998. 62. Gold MR, Shorofsky SR: Transvenous defibrillation lead systems. J Cardiovasc Electrophysiol 7:570-580, 1996. 63. Gottlieb CD, Schwartzman DS, Callans DJ, et al: Effects of high and low shock energies on sinus electrograms recorded via integrated and true bipolar nonthoracotomy lead systems. J Cardiovasc Electrophysiol 7:189-196, 1996. 64. Ellenbogen KA, Wood MA, Shepard RK, et al: Detection and management of an implantable cardioverter defibrillator lead failure: Incidence and clinical implications. J Am Coll Cardiol 41:73-80, 2003. 65. Nelson RS, Gilman BL, Shapland JE, Lehmann MH: Leads for the ICD. In Kroll MW, Lehmann MH (eds): Implantable Cardioverter Defibrillator Therapy. Norwell, Mass., Kluwer Academic, 1996. 66. Hayes DL, Graham KJ, Irwin M, et al: A multicenter experience with a bipolar tined polyurethane ventricular lead. PACE 15:1033-1039, 1992. 67. Fotuhi P, Alt E, Callihan R, et al: Endocardial carbon braid electrodes: New approach to lowering defibrillation thresholds. PACE 16:1919A, 1993. 68. Alt E, Theres H, Heinz M, et al: A new approach towards defibrillation electrodes: Highly conductive isotropic carbon fibers. PACE 14:1923-1928, 1991. 69. Jacobs DM, Fink AS, Miller RP, et al: Anatomical and morphological evaluation of pacemaker lead compression. PACE 16:434444, 1993. 70. Magney JE, Flynn DM, Parsons JA, et al: Anatomical mechnaisms explaining damage to pacemaker leads, defibrillator leads, and failure of central venous catheters adjacent to the sternoclavicular joint. PACE 16:445-457, 1993. 71. Doring J, Flink R: The impact of pending technologies on a universal connector standard. PACE 9:1186-1190, 1986. 72. Schaumann A, von zur Muhlen F, Gonska BD, Kreuzer H: Enhanced detection criteria in implantable cardioverterdefibrillators to avoid inappropriate therapy. Am J Cardiol 78:42-50, 1996.

197

73. Butter C, Auricchio A, Schwarz T, et al: Clinical evaluation of a prototype passive fixation dual chamber single pass lead for dual chamber ICD systems. PACE 22:169-173, 1999. 74. Saksena S, DeGroot P, Krol RB, et al: Low energy endocardial defibrillation using an axillary or a pectoral thoracic electrode location. Circulation 88:2655-2660, 1993. 75. Higgins SL, Alexander DC, Kuypers CJ, Brewster SA: The subcutaneous array: A new lead adjunct for the transvenous ICD to lower defibrillation thresholds. PACE 18:1540-1548, 1995. 76. Yee R, Klein GJ, Leitch JW, et al: A permanent transvenous lead system for an implantable pacemaker cardioverter defibrillator: Nonthoracotomy approach to implantation. Circulation 85:196204, 1992. 77. Venditti FJ, Martin DT, Vassolas G, Bowen S: Rise in chronic defibrillation thresholds in nonthroacotomy implantable defibrillators. Circulation 89:216-223, 1994. 78. Strickberger SA, Hummel JD, Horwood LE, et al: Effect of shock polarity on ventricular defibrillation threshold using a transvenous lead system. J Am Coll Cardiol 24:1069-1072, 1994. 79. Natale A, Sra J, Axtell K, et al: Preliminary experience with a hybrid nonthoracotomy defibrillating system that includes a biphasic device: Comparison with a standard monophasic device using the same lead system. J Am Coll Cardiol 24:406412, 1994. 80. Bardy GH, Ivey TD, Allen MD, et al: A prospective randomized evaluation of biphasic versus monophasic waveform pulses on defibrillation efficacy in humans. J Am Coll Cardiol 14:728-733, 1989. 81. Block M, Hammel D, Bocker D, et al: A prospective randomized cross-over comparison of mono- and biphasic defibrillation using nonthoracotomy lead configurations in humans. J Cardiovasc Electrophysiol 5:581-590, 1994. 82. Natale A, Sra J, Krum D, et al: Comparsion of biphasic and monophasic pulses: Does the advantage of biphasic shocks depend on the waveshape? PACE 18:1354-1361, 1995. 83. Saksena S, An H, Mehara R, et al: Prospective comparison of biphasic and monophasic shocks for implantable cardioverter defibrillators using endocardial leads. Am J Cardiol 70:304-310, 1992. 84. Saksena S, Scott SE, Accorti PR, et al: Efficacy and safety of monophasic and biphasic waveform shocks using a braided endocardial defibrillation lead system. Am Heart J 20:13421347, 1990. 85. Neuzner J, Pitschner HF, Huth C, Schlepper M: Effect of biphasic waveform pulse on endocardial defibrillation efficacy in humans. PACE 17:207-212, 1994. 86. Bardy GH, Johnson G, Poole JE, et al: A simplified, single-lead unipolar transvenous cardioversion defibrillation system. Circulation 88:543-547, 1993. 87. Mattke S, Fiek M, Markewitz A, et al: Comparison of a unipolar defibrillation system with a dual lead system using an enlarged defibrillation anode. PACE 19:2083-2088, 1996. 88. Lang DJ, Heil JE, Hahn SJ, et al: Implantable cardioverter defibrillator lead technology: Improved performance and lower defibrillation thresholds. PACE 18:548-559, 1995. 89. Niebauer MJ, Yamanouchi Y, Hills D, et al: Voltage dependence of ICD lead polarization and the effect of iridium oxide coating. PACE 23:818-823, 2000. 90. Niebauer MJ, Wilkoff B, Yamanouchi Y, et al: Iridium oxidecoated defibrillation electrode: Reduced shock polarization and improved defibrillation efficacy. Circulation 96:3732-3736, 1997. 91. Gradaus R, Bocker D, Dorszewski A, et al: Fractally coated defibrillation electrodes: Is an improvement in defibrillation threshold possible? Europace 2:154-159, 2000. 92. Peters RW, Foster AH, Shorofsky SR, et al: Spurious discharges due to late insulation break in endocardial sensing leads for cardioverter defibrillators. PACE 18:478-481, 1995.

198

Section One: Basic Principles of Device Therapy

93. Kron J, Herre J, Renfroe EG, et al., and the AVID Investigators. Lead and device-related complications in the Antiarrhythmics Versus Implantable Defibrillators Trial. Am Heart J 141:92-98, 2001. 94. Degeratu FT, Khalighi K, Peters RW, et al: Sensing lead failure in implantable defibrillators: A comparison of two commonly used leads. J Cardiovasc Electrophysiol 11:21-24, 2000. 95. Gold MR, Peters RW, Johnson JW, Shorofsky SR: Complications associated with pctoral implantation of cardioverter defibrillators. World-Wide Jewel Investigators. PACE 20:208-211, 1997. 96. Grimm W, Menz J, Hoffmann J, et al: Complications of thirdgeneration implantable cardioverter defibrillator therapy. PACE 22:206-211, 1999. 97. Gupta A, Zegel HG, Dravid VS, et al: Value of radiography in diagnosing complications of cardioverter defibrillators implanted without thoracotomy in 437 patients. Am J Radiol 168:105-108, 1997. 98. Mehta D, Nayak HM, Signson M, et al: Late complications in patients with pectoral defibrillator implants with transvenous defibrillator lead systems: High incidence of insulation breakdown. PACE 21:1893-1900, 1998. 99. Dorwarth U, Frey B, Dugas M, et al: Transvenous defibrillatin leads: High incidence of failure during long-term follow-up. J Cardiovasc Electrophysiol 14:38-43, 2003. 100. Schwartzman D, Nallamouthu N, Callans DJ, et al: Postoperative lead-related complications in patients with nonthoracotomy defibrillator lead systems. J Am Coll Cardiol 26:776-786, 1995. 101. Brooks R, Garan H, Torchiana D, et al: Determinants of successful nonthoracotomy cardioverter defibrillator implantation: Experience in 101 patients using two different lead systems. J Am Coll Cardiol 22:1835-1842m 1993. 102. Zipes DP, Roberts D: Results of the international study of the implantable pacemaker cardioverter defibrillator: Intentionto-treat comparison of lnical outcomes. Circulation 92:59-65, 1995. 103. Raviele A, Gasparini G, for the Italian Endotak Investigator Group. Italian multicenter clinical experience with endocardial defibrillation: Acute and long-term results in 307 patients. PACE 18:599-608, 1995. 104. Jones GK, Bardy GH, Kudenchuk PJ, et al: Mechanical complications after implantation of multiple lead nonthoracotomy defibrillator system: Implications for management and future system design. Am Heart J 130:327-333, 1995. 105. Lawton JS, Wood MA, Gilligan DM, et al: Implantable transvenous cardioverter defibrillator leads: The dark side. PACE 19:1273-1278, 1996. 106. Tullo NG, Saksena S, Krol RB, et al: Management of complications associated with a first-generation endocardial defibrillation lead system for implantable cardioverter defibrillators. Am J Cardiol 66:411-415, 1990. 107. Mahapatra S, Homound MK, Wang PJ, et al: Durability of repaired sensing leads equivalent to that of new leads in implantable cardioverter defibrillator patients with sensing abnormalities. PACE 26:2225-2229, 2003. 108. Sra JS, Natale A, Axtell K, et al: Experience with two different nonthoracotomy systems for implantable defibrillators in 170 patients. PACE 17:1741-1750, 1994. 109. Fahy GJ, Kleman JM, Wilkoff BL, et al: Low incidence of lead related complications associated with nonthoracotomy implantable cardioverter defibrillator systems. PACE 18:172-178, 1995. 110. Roelke M, O’Nunain SS, Osswald S, et al: Subclavian crush syndrome complicationg transvenous caredioverter defibrillator systems. PACE 18:973-979, 1995. 111. Gallik DKM, Ben-zur UM, Gross JN, Furman S: Lead fracture in cephalic versus subclavian approach with transvenous implantable cardioverter defibrillator systems. PACE 19:10891094, 1996.

112. Hauser RG, Cannom D, Hayes DL, et al: Long-term structural failure of coaxial polyurethane implantable cardioverter defibrillator leads. PACE 25:879-882, 2002. 113. Peralta AO, John RM, Martin DT, Venditti FJ: Long term performance of the Endotak C defibrillator lead [abstract] Circulation 98:1-787, 1998. 114. Moore SL, Maloney JD, Edel TB, et al: Implantable cardioverter defibrillator implanted by nonthoracotomy approach: Initial clinical experience with the redesigned transvenous lead system. PACE 14:1865-1869, 1991. 115. Lawton JS, Ellenbogen KA, Wood MA, et al: Sensing lead-related complications in patients with transvenous implantable cardioverter defibrillators. Am J Cardiol 78:647-651, 1996. 116. Mera F, Delurgio DB, Langberg JJ, et al: Transvenous cardioverter defibrillator lead malfunction due to terminal connector damage in pectoral implants. PACE 22:1797-1801, 1999. 117. Block M, Hammel D, Isbruch F, et al: Results and realistic expectations with transvenous lead systems. PACE 15:665-670, 1992. 118. Schwartzman D, Nallamothu N, Callans DJ, et al: Postoperative lead-related complications in patients with nonthoractomy defibrillation lead systems. J Am Coll Cardiol 26:776-786, 1995. 119. Pratt TR, Pulling CC, Stanton MS: Prospective postmarket device studies versus returned product analysis as a predictor of system survival. PACE 23:1150-1155, 2000. 120. Gras D, Leclereq C, Tang AS, et al: Cardiac resynchronization therapy in advanced heart failure: The multicenter InSync clinical study. Eur J Heart Fail 4:311-320, 2002. 121. Auricchio A, Stellbrink C, Sack S, et al., and the Pacing Therapies in Congestive Heart Failure (PATH-CHF) Study Group. Long-term clinical effect of hemodynamically optimized cardiac resychronization therapy in patients with heart failure and ventricular conduction delay. J Am Coll Card 39:2026-2033, 2002. 122. Cazeau S, Leclercq C, Lavergne T, et al., and the Multisite Stimulation in Cardiomyopathies (MUSTIC) Study Investigators. Effects of multisite biventricular pacing in patients with heart failure and intraventricular conduction delay. N Engl J Med 344:873-880, 2001. 123. Linde C, Leclercq C, Rex S, et al: Long-term benefits of biventricular pacing in congestive heart failure: Results from the Multisite stimulation in caredioomyopathy (MUSTIC) study. J Am Coll Card 40:111-118, 2002. 124. Abraham WT, Fisher WG, Smith AL, et al, and the MIRACLE Study Group Multicenter InSync Randomized Clinical Evaluation. Cardiac resynchronization in chronic heart failure. N Engl J Med 346:1845-1853, 2002. 125. Blanc JJ, Etienne Y, Giloard M, et al: Evaluation of different ventricular pacing sites in patients with severe heart failure: Results of an acute hemodynamic study. Circulation 96:3273-3277, 1997. 126. Young JB, Abraham WT, Smith AL, et al., and the Multicenter InSync ICD Randomized Clinical Evaluation (MIRACLE ICD) Trial Investigators. Combined cardiac resynchronization and implantable cardioversion defibrillation in advanced chronic heart failure: The MIRACLE ICD Trial. JAMA 289:2685-2694, 2003. 127. Bristow MR, Saxon LA, Boehmeer J, et al., and the Comparison of Medical Therapy, Pacing and Defibrillation in Heart Failure (COMPANION) Investigators. N Engl J Med 350:21402150, 2004. 128. Rossillo A, Verma A, Saad EB, et al: Impact of coronary sinus lead position on biventricular pacing: Mortality and echocardiographic evaluation during long-term follow-up. J Cardiovasc Electrophysiol 15:1120-1125, 2004. 129. Butter C, Auricchio A, Stellbrink C, et al: Should stimulation site be tailored in the individual heart failure patient? Am J Cardiol 86:144K-151K, 2000. 130. Purerfellner H, Nesser HJ, Winter S, et al., for the EASYTRAK Clinical Investigation Study Group and the European EASY-

Chapter 4: Engineering and Construction of Pacemaker and ICD Leads

131.

132.

133.

134.

135. 136.

137. 138. 139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150.

TRAK Registry. Transvenous left venticular lead implantation with the EASYTRAK lead system: The European experience. Am J Cardiol 86:157K-164K, 2000. Sack S, Heinzel F, Dagres N, et al: Stimulation of the left ventricle through the coronary sinus with a newly developed “over the wire” lead system: Early experiences with lead handling and positioning. Europace 3:317-323, 2001. Achtelik M, Bocchiardo M, Trappe HJ, et al., on behalf of the Ventak CHF/Contak CD Clinical Investigation Study Group. Performance of a new steroid-eluting coronary sinus lead designed for lefter ventricular pacing. PACE 23:1741-1743, 2000. Kantharia BK, Padder FA, Pennington JC, et al: Feasibility, safety, and determinants of extraction time of percutaneous extraction of endocardial implantable cardioverter defibrillator leads by intravascular countertraction method. Am J Cardiol 85:593-597, 2000. Saad EB, Saliba WI, Schweikert RA, et al: Nonthoracotomy implantable defibrillator lead extraction: Results and comparison with extraction of pacemaker leads. Pacing Clin Electrophysiol 26:1944-1950, 2003. Parsonnet V, Roelke M, Trivedi A, et al: Laser extraction of entrapped leads. PACE 24:329-332, 2001. Morris MM, KenKnight BH, Warren JA, Lang DJ: A preview of implantable cardioverter derfibrillator systems in the next millennium: An integrative cardiac rhythm management approach. Am J Cardiol 83:48D-54D, 1999. Bardy GH, Cappato R, Smith WM, et al: The totally subcutaneous ICD system (the S-ICD) [abstract]. PACE 24:578, 2002. Mond HG: Unipolar versus bipolar pacing: Poles apart. PACE 14:1411, 1991. Wiegand UK, Bode F, Bonnemeier H, et al: Incidence and predictors of pacemaker dysfunction with unipolar ventricular lead configuration: Can we identify patients who benefit from bipolar electrodes? PACE 24:1383-1388, 2001. Breivik K, Ohm OJ, Engedal H: Long-term comparison of unipolar and bipolar pacing and sensing, using a new multiprogrammable pacemaker system. PACE 6:592-600, 1983. Secemsky SI, Hauser RG, Denes P, Edwards LM: Unipolar sensing abnormalities: Incidence and clinical significance of skeletal muscle interference and undersensing in 228 patients. PACE 5:10-19, 1982. Echeverria HJ, Luceri RM, Thurber RJ, Castellanos A: Myopotential inhibition of unipolar AV sequential (DVI) pacemaker. PACE 5:20-22, 1982. Zimmern SH, Clark MF, Austin WK, et al: Characteristics and clinical effects of myopotential signals in a unipolar DDD pacemaker population. PACE 9:1019, 1986. Smyth NP, Tarjan PP, Chernoff E, Baker N: The significance of electrode surface area and stimulation thresholds in permanent cardiac pacing. J Thorac Cardiovasc Surg 71:559-65, 1976. Furman S, Garvey J, Hurzeler P: Pulse duration variation and electrode size as factors in pacemaker longevity. J Thorac Cardiovasc Surg 69:382-389, 1975. Barold SS, Ong LS, Heinle RA: Stimulation and sensing thresholds for cardiac pacing: Electrophysiologic and technical aspects. Prog Cardiovasc Dis 1981;24:1-24. Mond H, Holley L, Hirshorn M: The high impedance dish electrode: Clinical experience with a new tined lead. PACE 5:529534, 1982. Schuchert A, Kuch KH: Benefits of smaller electrode surface area (4 mm2) on steroid-eluting leads. PACE 14:2098-2104, 1991. Ellenbogen KA, Wood MA, Gilliagan DM, et al, and the CapSure Z Investigators. Steroid-eluting high-impedance pacing leads decrease short and long-term current drain: Results from a multicenter clinical trial. PACE 22:39-48, 1999. Moracchini PV, Cornacchia D, Bernasconi M, et al: High impedance low energy pacing leads: Long-term results with a very

151.

152.

153.

154.

155.

156. 157.

158. 159. 160. 161. 162.

163.

164. 165. 166. 167. 168. 169. 170.

171.

172.

173.

174.

175.

176.

199

small surface area, steroid-eluting lead compared to three conventional electrodes. PACE 22:326-334, 1999. Deshmukh P, Casavant D, Anderson K, Romanyshyn M: Stable electrical performance of high efficiency pacing leads having small surface, steroid-eluting pacing electrodes. PACE 22:15991603, 1999. Berger T, Roithinger FX, Antretter H, et al: The influence of high versus normal impedance ventricular leads on pacemaker generator longevity. PACE 26:2116-2120, 2003. Scherer M, Ezziddin K, Klesius A, et al: Extension of generator longevity by use of high impedance ventricular leads. PACE 24:206-211, 2001. Mond HG. Engineering and clinical aspects of pacing leads. In Ellenbogen K, Kay G, Wilkoff B (eds.): Clincial Cardiac Pacing and Defibrillation, 2nd ed. Philadelphia, WB Saunders, 2000. Luria D, Gurevitz O, Bar Lev D, et al: Use of automatic threshold tracking function with non-low polarization leads. PACE 27:453-459, 2004. Amundson DC, McArthur W, Mosharrafa M: The porous endocardial electrode. PACE 2:40-50, 1979. Hirshorn MS, Holley LK, Skalsky M, et al: Characteristics of advanced porous and textured surface pacemaker electrodes. PACE 6:525-536, 1983. Djordjevic M, Stojanov P, Velimirovic D: Target lead: Low threshold electrode. PACE 9:1206, 1986. Mond H, Hua W, Wang CC: Atrial pacing leads: The clinical contribution of steroid elution. PACE 18:1601, 1995. Parsonnet V, Villaneuva A, Driller J, Bernstein AD: Corrosion of pacemaker electrodes. PACE 4:289-296, 1981. Elmqvist H, Schueller H, Richter G: The carbon tip electrode. PACE 6:436-439, 1983. Mugica J, Henry L, Attuel P, et al: Clinical experience with 910 carbon tip leads: Comparison with polished platinum leads. PACE 9:1230-1238, 1986. Gargeroglio B, Inguaggiato B, Chinaglia B, Cerise O: Initial results with an activated carbon tip electrode. PACE 6:440-448, 1983. Pioger G, Ripart A: Clinical results of low energy unipolar or bipolar activated carbon tip leads. PACE 9:1243-1248, 1986. Mond HG, Grenz D: Implantable transvenous pacing leads: The shape of things to come. PACE 27:887-893, 2004. Breivik K, Danilovic D, Ohm OJ, et al: Clinical evaluation of thin bipolar pacing lead. PACE 20:637-646, 1997. Crossley GH: Cardiac pacing leads. Cardiol Clin 18:95-112, 2000. Scheuer-Lesser M, Irnich W, Kreuzer J: Polyurethane leads: Facts and controversy. PACE 6:454-458, 1983. Byrd CL, McArthur W, Stokes K, et al: Implant experience with unipolar polyurethane pacing leads. PACE 6:868-882, 1983. Woscoboinik JR, Maloney JD, Helguera ME, et al: Pacing lead survival: Performance of different models. PACE 15:1991-1995, 1992. Hayes DL, Graham KJ, Irwin M, et al: A multicenter experience with a bipolar tined polyurethane ventricular lead. PACE 15:1033-1039, 1992. Sweesy MW, Forney CC, Hayes DL, et al: Evaluation of an inline bipolar polyurethane ventricular pacing lead. PACE 15:1982-1985, 1992. Hayes DL, Graham KJ, Irwin M, et al: Multicenter experience with a bipolar tined polyurethane ventricular lead. PACE 18:9991004, 1995. Van Beek GJ, Den Dulk K, Lindemans FW, et al: Detection of insulation failure by gradual reduction in noninvasively measured electrogram amplitudes. PACE 9:772-775, 1986. Pirzada FA, Seltzer JP, Blair-Saletin D, Killian M: Five-year performance of the Medtronic 6971 polyurethane endocardial electrode. PACE 9:1173-1180, 1986. Stokes KB, Church T: Ten-year experience with implanted polyurethane lead insulation. PACE 9:1160-1164, 1986.

200

Section One: Basic Principles of Device Therapy

177. Antonelli D, Rosenfeld T, Freedberg NA, et al: Insulation lead failure: Is it a matter of insulation coating, venous approach, or both? PACE 21:418-421, 1998. 178. Philips R, Frey M, Martin RO: Long-term performance of polyurethane pacing leads: Mechanisms of design-related failures. PACE 9:1166-1172, 1986. 179. Cameron J, Mond H, Ciddor G, et al: Stiffness of the distal tip of bipolar pacemaker leads. PACE 13:1915-1920, 1990. 180. Mugica J, Daubert JC, Lazarus B, et al: Is polyurethane lead insulation still controversial? PACE 15:1967-1970, 1992. 181. Phillips R, Frey M, Martin RO: Long-term performance of polyurethane pacing leads: Mechanisms of design-related failures. PACE 9:1166-1172, 1986. 182. Mond H, Sloman G: The small tined pacemaker lead: Absence of dislodgement. PACE 3:171-177, 1980. 183. Kertes P, Mond H, Sloman G, et al: Comparison of lead complications with polyurethane tined, silicone rubber tined, and wedge tip leads: Clinical experience with 822 ventricular endocardial leads. PACE 6:957-962, 1983. 184. Stokes KB: Recent advances in lead technology. In Barold SS, Mugica J (eds): New Perspectives in Cardiac Pacing. Mount Kisco, NY, Futura, 1988, pp 217-227. 185. Ormerod D, Walgren S, Berglund J, Heil R: Design and evaluation of a low threshold porous tip lead with a mannitol coated screw-in tip (“Sweet Tip”). PACE 11:1784-1790, 1988. 186. Bisping HJ, Kreuzer J, Birkenheier H: Three year clinical experience with a new endocardial screw-in lead with introduction protection for use in the atrium and ventricle. PACE 3:424-435, 1980. 187. Crossley GH, Brinkler JA, Reynolds D, et al: Steroid elution improves the stimulation threshold in an active-fixation atrial permanent pacing lead. Circulation 92:2935-2939, 1995. 188. Hidden-Lucet F, Halimi F, Gallais Y, et al: Low chronic pacing thresholds of steroid-eluting active-fixation ventricular pacemaker leads: A useful alternative to passive-fixation leads. PACE 23:1798-1800, 2000. 189. Mond H, Stokes KB: The electrode-tissue interface: The revolutionary role of steroid elution. PACE 15:95-107, 1992. 190. Mond H, Stokes K, Helland J, et al: The porous titanium steroid eluting electrode: A double blind study assessing

191. 192. 193.

194.

195.

196.

197.

198.

199.

200. 201.

202.

203.

the stimulation threshold effects of steroid. PACE 11:214-219, 1988. Mond H, Stokes KB: The steroid-eluting electrode: A 10 year experience. PACE 19:1016-1020, 1997. Kruse IM: Long-term performance of endocardial leads with steroid-eluting electrodes. PACE 9:1217-1219, 1986. Hua W, Mond H, Sparks P: The clinical performance of three designs of atrial pacing leads from a single manufacturer: The value of steroid elution. Eur J Cardiac Pacing Electrophysiol 6:99-103, 1996. Schuchert A, Hopf M, Kuck KH, et al: Chronic ventricular electrograms: Do steroid eluting leads differ from conventional leads? PACE 13:1879, 1990. Kay GN, Brinker JA, Kawanishi DT, et al: Risks of spontaneous injury and extraction of an active fixation pacemaker lead: report of the Accufix Multicenter Clinical Study and Worldwide Registry. Circulation 100:2344-2352, 1999. Parsonnet V, Roelke M, Bernstein AD, Stern M: Reduced frequency of retention wire fractures suggests that elective explantation of affected atrial leads is no longer indicated. PACE 23:380-383, 2000. Lau C, Nishimura SC, Oxorn D, Goldman BS: Is this the natural history of the retention wire? A case report. PACE 20:1373-1376, 1997. Kao HL, Wang SS, Chen WJ, et al: Migration of a fractured retention wire in the pulmonary artery from an active fixation atrial lead. PACE 18:1966, 1995. Tatou E, Lefex C, Reybet-Degat O, et al: Intrapulmonary artery and intrabronchial migration and extraction of a fragment of Jshaped atrial pacing catheter. PACE 22:1829-1830, 1999. Saliba BC, Ardesia RJ, John RM, et al: Predictors of fracture in the Accufix atrial “J” lead. Am J Cardiol 80:229-231, 1997. Kawanishi DT, Brinker JA, Reeves R, et al: Cumulative hazard analysis of J-wire fracture in the Accufix series of atrial permanent pacemaker leads. PACE 21:2322-2326, 1998. Lloyd MA, Hayes D, Holmes DR Jr: Atrial J-pacing lead retention wire fracture: Radiographic assessment, incidence of fracture and clinical management. PACE 18:958-964, 1995. Important Information about Telectronics Accufix and Encor Atrial “J” Leads. Accufix Research Institute, April 2005.

Chapter 5

Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity HUNG-FAT TSE • CHU PAK LAU

P hysiologic pacing aims to restore the rate and sequence of cardiac activation in the presence of abnormal cardiac automaticity and conduction. The atrial electrogram can be used for rate control if sinus node (SN) function is adequate. However, a high proportion of pacemaker recipients have abnormal SN function, either at rest or during exercise. Such chronotropic incompetence may be the result of medications or intrinsically abnormal SN function. In addition, in patients whose atrium is unreliable for sensing or pacing (e.g., during atrial fibrillation), an alternative means to simulate SN responsiveness is required. These problems with the pacemaker function of the SN prompted the development of artificial implantable sensors for cardiac pacing. Although the optimal rate adaptation in a pacemaker recipient may be different from that of a healthy subject, it is assumed that these sensors should mimic the behavior of the healthy SN response to exercise and nonexercise needs. In addition, atrioventricular (AV) conduction is normally under the control of the autonomic nervous system, and a sensor may also contribute to adapting the AV interval to changes in atrial rate.

The role of sensors has also been expanded to include functions other than rate augmentation—such as detection of ventricular capture and monitoring for heart failure, sleep apnea, and hemodynamic status. In this chapter, the basic principles and the clinical applications of sensors in implantable devices are reviewed.

Historical Landmarks of RateAdaptive Pacing The limitations of dual-chamber pacing in the setting of SN dysfunction led to the development of nonatrial sensors. This development was possible because of the recognition that during exercise an increase in heart rate (HR), rather than the maintenance of AV synchrony, is the main determinant of increases in cardiac output.1 Therefore, a single-chamber rate-adaptive pacemaker that varies the pacing rate with exercise according to a nonatrial sensor can achieve near-normal exercise physiology in patients with bradycardia. 201

202

Section One: Basic Principles of Device Therapy

Cammilli and associates2,3 implanted the first ratevariable single-chamber pacemaker, which sensed changes in blood pH during exercise. In 1981, a “physiologically adaptive” cardiac pacemaker responding to changes in the QT interval during exercise was described by Rickards and Norman.4 In the same year, Wirtzfeld and coworkers5 reported the use of central –o2) for the control of autovenous oxygen saturation (Sv matic rate-responsive pacing. Despite the fact that respiratory changes during exercise were proposed as a physiologic parameter to be sensed by a rate-adaptive pacemaker as early as 1975,6 a rate-adaptive pacemaker capable of detecting the respiratory rate was introduced by Rossi and colleagues7 only in 1982. With continuing research, the number of sensors available for rateadaptive pacing steadily increased. In particular, activity sensors8 and minute ventilation sensors9,10 have been used with extensive clinical application. Although single-chamber ventricular rate-adaptive pacing was originally meant to replace dual-chamber pacing, the additional benefits of atrial sensing and pacing became well recognized. Dual-chamber rateadaptive pacing became available as early as 1986.11 Since then, virtually all pacemaker companies have introduced their own versions of rate-adaptive dualchamber devices. Technical improvement in the sensing of atrial electrograms with “floating,” diagonally opposed atrial electrodes or closely spaced rings has enabled the use of a single-pass lead for VDD pacing. When combined with a rate-adaptive sensor, VDDR pacing with a single-pass lead has become a possibility.12 With the proliferation of sensor technology, it soon became apparent that none of the sensors could simulate the normal SN function in all aspects,13 although different sensors were better in some areas of performance. Therefore, it was logical to combine sensors for optimal rate adaptation. Although this idea was not new, investigational units became available only in 1988,14 first in the form of combined activity and QT-interval sensing, and later with combined activity and minute-ventilation sensors.15 The increasing sophistication of sensors and their combinations prompted the use of automatic sensor calibration for optimal programming.16-18 Currently, rate-adaptive pacing has become a standard component of implantable devices and was incorporated into the pacing code after the joint effort between the North American Society of Pacing and Electrophysiology Mode Code Committee and the British Pacing and Electrophysiology Group. The original three-letter code for pacemaker mode proposed in 1974 was revised to the five-letter pacing code, and the rate-adaptive function is now denoted by the use of the letter R in the fourth position.19-21

Normal Heart Rate and Respiratory Responses to Exercise and Nonexercise Needs Cardiac output is the product of HR and stroke volume. Stroke volume is enhanced during exercise when

venous return increases cardiac filling, and cardiac contractility is augmented in response to sympathetic stimulation and a more vigorous pumping action of the skeletal muscles. Because of the difference in venous return, the changes in HR and cardiac hemodynamics are highly influenced by whether the exercise is carried out in the upright or supine posture. Normal Heart Rate Response during Exercise An anticipatory response of the HR occurs in many patients before exercise. With both supine and upright isotonic exercise, HR and cardiac output increase within 10 seconds after the onset of exercise.22-24 This initial increase in HR is mediated by parasympathetic withdrawal rather than sympathetic stimulation. The cardiac output may increase by as much as 40% within three heartbeats after the onset of vigorous muscular exercise. Both cardiac output and sinus rate increase exponentially, with a half-time that ranges from 10 to 45 seconds (Fig. 5-1A), the rate of rise being proportional to the intensity of work (see Fig. 5-1B).22 At the termination of upright exercise, there is a delay of about 5 to 10 seconds before cardiac output starts to decrease, followed by an exponential fall with a half-time of 25 to 60 seconds. The recovery time is related to age, work intensity, total work performed, and physical condition of the patient.25 Although the optimal rate onset and decay kinetics for patients with pacemakers have not been firmly established, artificial sensors for rateadaptive pacing should probably simulate the onset and recovery kinetics of the normal SN as the ideal physiologic standard. Respiratory Changes during Exercise The change in HR during muscular exercise is linearly related to oxygen consumption and workload. Because of the relationship between oxygen uptake and ventilatory volume during aerobic metabolism, minute ventilation is closely related to HR during exercise. At rest, the respiratory rate is typically between 10 and 20 breaths/min. During low-intensity exercise, an increase in tidal volume is the primary respiratory adaptation.26 At higher work levels, a further increase in tidal volume of up to 50% of the vital capacity occurs, together with an increase in breathing frequency. The relationship between breathing rate and tidal volume varies considerably among individuals. In addition, the breathing rate is often synchronized with the work rhythm (e.g., the walking pace). However, because respiration is carefully controlled to maintain the concentration of arterial carbon dioxide within a narrow physiologic range, compensatory adjustments in tidal volume ensure an appropriate minute ventilation and gas exchange. Minute ventilation (the product of breathing rate and tidal volume) is linearly related to the rate of carbon dioxide production. At low- and medium-intensity workloads, minute ventilation is also linearly related to oxygen consumption during incremental exercise.

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

203

Proportionality 100

Speed 100 T 90

75

%HR

%HR (bpm)

90

T 1/2

50

50

25 DT 0

0 0

A

30

60

90

120

150

0

25

50

75

% Workload 100

Time(s) 180

FIXED WORKLOAD

B

GRADED WORKLOAD

Figure 5-1. Quantification of sensor response: speed and proportionality. A, The heart rate (HR) changes at the onset of exercise. The normal sinus rate responds almost immediately; one half of the change is achieved in less than 30 seconds, and most of the change is achieved within 1 minute. This speed of response was quantified by the response times; DT is the delay time, and T 1/2 and T90 represent times needed to reach 50% and 90% of maximum HR, respectively. %ΔHR is the change in HR expressed as a percentage of the maximum increase in HR. B, Relationship between exercise workload and the HR increase during incremental exercise. The workload is expressed as a percentage of the maximum workload (% workload) at each quartile of exercise. The %ΔHR is linearly related to % workload, with a slope that is almost 1.

Resting values for minute ventilation are about 6 L/min and can increase to 100 L/min during exercise for normal men and up to 200 L/min for trained athletes. Tidal volume may rise from a resting value of 0.5 L/min to 3 L/min at maximal exercise in normal individuals. The breathing rate may increase from a resting value of 12 to 16 breaths/min to 40 to 50 breaths/min during peak exercise. In normal children and some adults with restrictive lung disease, the respiratory rate may exceed 60 breaths/min. At about 70% of the maximum oxygen uptake, the rate of tissue metabolism outstrips the rate of tissue oxygen delivery. Anaerobic metabolism of carbohydrate, fat, and protein is accelerated, resulting in an accumulation of lactic acid, the so-called anaerobic threshold. Acidosis is initially prevented by plasma buffers, but continuous lactate production eventually leads to excess production of carbon dioxide from plasma bicarbonate. This excess carbon dioxide stimulates the respiratory center, leading to respiratory compensation (increased minute ventilation). Hence, carbon dioxide production and minute ventilation increase in a manner that is disproportionate to the increase in oxygen consumption at workloads that surpass the anaerobic threshold. The anaerobic threshold is defined more clearly using a protocol in which the work intensity is increased more rapidly than with a gradual exercise protocol. It has been customary to use the slope of the curve of HR versus minute ventilation as a measure of the

appropriateness of rate adaptation by an artificial sensor, with the regression line at 1 to 2 bpm/L. In addition, this slope is said to be relatively independent of the functional class or the degree of left ventricular dysfunction.27 At workloads that surpass the anaerobic threshold, however, oxygen consumption asymptotically reaches its maximum value, whereas carbon dioxide production and minute ventilation increase disproportionately to oxygen consumption. Because of the disproportionate increase in minute ventilation above the anaerobic threshold, the HR/minute ventilation slope is reduced. The reduction in the HR/minute ventilation slope above the anaerobic threshold, compared with below it, is an average of 29% for women and 26% for men.28 This fact may be important, both in assessment of rate adaptation and in the design of the minute ventilation–sensing algorithms. Oxygen Uptake Kinetics In healthy subjects, exercise-induced cardiac and respiratory changes occur simultaneously to provide blood flow commensurate with the increase in oxygen consumption in the skeletal muscles. Figure 5-2 shows the oxygen uptake curve during exercise at a constant workload. When exercise begins, oxygen uptake increases gradually, following an exponential time course because of the slow adjustment of respiration and circulation, until it reaches a steady state at which oxygen uptake corresponds to the demands of the

204

Section One: Basic Principles of Device Therapy

Oxygen Consumption 1000 Maximum VO2

Stop Exercise

Oxygen Deficit

800 600

Total VO2 400 200

Oxygen Debt

Basal VO2 Rest

Start Exercise

Recovery

0

Figure 5-2. Oxygen kinetics during a steadystate exercise test. An oxygen deficit is incurred as the subject increases oxygen consumption (Vo2) to the steady state, which usually takes 3 to 5 minutes. The incurred deficit is repaid at the termination of exercise as the oxygen debt. The oxygen deficit is closely related to the alactic oxygen debt; therefore, the oxygen debt, which includes the fast alactic and the slower lactic oxygen debt, is a better measurement of oxygen kinetics. (From Leung SK, Lau CP, Wu CW, et al: Quantitative comparison of rate response and oxygen uptake kinetics between different sensor modes in multisensor rate adaptive pacing. PACE 17:1920-1927, 1994).

Time

tissues. Because oxygen uptake does not reach the required steady state immediately, the inadequate supply of energy from aerobic sources during the first 2 to 3 minutes must be met largely by use of creatine phosphate high-energy stores.29 This inadequacy of oxygen use at the onset of exercise is described as the oxygen deficit. After exercise stops, replenishment of energy stores requires oxygen consumption in excess of the baseline recovery conditions before the oxygen uptake gradually decreases to the resting level. This elevated postexercise oxygen uptake repays the initial deficit and is described as repaying the “oxygen debt.”30 The oxygen debt includes a fast alactic phase during moderate exercise. The magnitude of the oxygen deficit is usually found to be about equivalent to that of the oxygen debt in submaximal exercise before the anaerobic threshold is reached.31 If the level of exertion remains relatively modest, the oxygen deficit may be paid back during exercise, so that the debt measured after exercise is actually less than the deficit at the onset. In contrast, during more strenuous exercise, the oxygen debt has both a fast alactic phase and a slower lactic phase, and the oxygen debt exceeds the size of the deficit.32 The rate-limiting step for the increase in oxygen uptake at the onset of exercise is the rate of oxygen transport; this is primarily a cardiac function that is dependent on the change in pulmonary blood flow.33,34 Therefore, an appropriate rate response behavior is the best way to ensure optimal oxygen delivery with either submaximal or maximal exercise. Because the oxygen debt may require a long period for repayment and may be affected by postexercise oxygen consumption for processes other than the simple deficit repayment,35 and because of variability in recovery of the baseline oxygen consumption,36 the oxygen deficit is used to assess the contribution of rate response to oxygen transport. By artificially increasing the rate response of a sensor beyond that expected of the SN, it was shown that oxygen deficit was reduced, compared with an optimal sensor setting. The dependence on a quick initial response to minimize oxygen debt was demonstrated in studies of minute ventilation and activity and dP/dt sensors,37,38 and oxygen uptake kinetics is often used as a sensitive indicator of appropriate rateadaptive response at the submaximal workload.

Heart Rate Response during Exercise in Patients with Heart Disease In patients with significant heart disease (e.g., heart failure, ischemic heart disease), pharmacologic treatment with β-blockers39 and/or coexisting chronotropic incompetence40-43 frequently limits the increase in HR during exercise, which may have a negative effect on exercise capacity. In patients with heart failure, the ability to augment left ventricular stroke volume and ventricular filling without a concomitant increase in left atrial pressure during exercise may be lost. Because the left atrial pressure is already high, the atrial contribution to left ventricular stroke volume is small. In addition, maintaining left ventricular stroke volume during exercise by increasing myocardial contractility is also markedly attenuated in patients with heart disease. HR augmentation is therefore a major determinant of cardiac output during exercise. Appropriate rate adaptation during exercise may provide incremental benefit to patients with heart disease. Conversely, inappropriate use of rate-adaptive pacing with excessive tachycardia in patients with heart disease may lead to an adverse outcome.44 Heart Rate Modulation for Nonexercise Needs Exercise is but one of the many physiologic requirements for variation in HR (Table 5-1). Emotions such as anxiety may trigger a substantial change in HR. The sinus rate is higher when a person moves from the supine to the upright posture, and cardiac output decreases. Isometric exercise also results in an increase in cardiac output and HR.45 An appropriate compensatory HR response is especially important in pathologic conditions such as anemia, acute blood loss, or other causes of hypovolemia. The normal resting HR peaks during the day and reaches a trough in the early morning hours during sleep. A sensor that is responsive only to exercise will increase the pacing rate during periods of physiologic stress, but the lower rate remains fixed at the programmed base rate. This faster programmed lower rate, which is normal for the day, may be too rapid during periods of rest. The lack of an ability to decrease the

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

Some Physiologic Factors that Affect the Responses of the Sinus Node to Body Requirements TABLE 5-1.

Exercise—isotonic and isometric, during and after exercise Postural changes Anxiety and stress Postprandial changes Vagal maneuvers Circadian changes

205

ator must have an algorithm that relates changes in the sensed parameter to a change in pacing rate. The design of the rate-control algorithm can have a profound impact on the overall rate-response characteristics of a pacing system. Third, because the magnitude of the physical or physiologic changes that are monitored by a sensor differs among patients, physician input is usually necessary to adjust the algorithm (usually by programming one or more rate-responsive variables) to achieve the clinically desired rate response.48 Classification of Sensors

Fever

Physiologic Classification

pacing rate in patients with VVI pacemakers during sleep may result in palpitations and sleep disturbances.46 By means of sensors that are always active in monitoring the level of metabolic demand, the appropriate lower rate can be calculated and adapted to individual need (see later discussion). HR increases during febrile episodes may be detected by a sensor that measures the central venous temperature.47

Components of a Rate-Adaptive Pacing System At least three aspects of a rate-adaptive pacing system influence its rate-modulating characteristics (Fig. 5-3). First, a sensor (or a combination of sensors) must detect a physical or physiologic parameter that is either directly or indirectly related to metabolic demand. Second, the rate-modulating circuit in the pulse gener-

A primary sensor is defined as one that detects the physiologic factors that control the normal SN during varying metabolic needs.49 Primary sensed parameters include circulating catecholamines and autonomic nervous system activities. Although these parameters may be the most physiologically accurate indicators for use by a rate-adaptive pacing system, technical realization of a rate-adaptive pacemaker that uses a primary sensor has yet to be achieved. The bulk of rate-adaptive sensors that have been proposed belong to the class of secondary sensors, those that detect physiologic parameters that are a consequence of exercise. Some of these parameters, such as the QT interval,4 respiratory rate or minute ventilation,9,10,50-52 average atrial rate,53 central venous temperature,47,54,55 venous blood pH,2,3 right ventricular stroke –o2,5,59 volume,56 pre-ejection interval57 or pressure,58 Sv and ventricular inotropic indices (ventricular inotropic parameter60,61 and peak endocardial acceleration62,63) have been developed for either clinical or investigational pacing systems. Each of these physiologic variables responds to the onset of exercise with its own

Closed Loop

Open Loop

Sensor

Sensor

 Physiologic parameters

 Physiologic/ Physical parameters

Algorithm

Ph ys Inp ician ut

Ph ys Inp ician ut  Rate

Algorithm

 Rate

Figure 5-3. Design of a rate-adaptive system. Open loop: The physiologic or physical change detected by the sensor is converted to a change in rate using an algorithm. The resultant rate change does not have a negative feedback effect on the physiologic/physical parameter. Closed loop: The physiologic change detected by the sensor is converted to a change in rate using an algorithm. The resultant rate change induces a change in the physiologic parameter in the opposite direction, thereby establishing a negative feedback loop.

206

Section One: Basic Principles of Device Therapy

kinetics and has a different proportionality to exercise workload. The tertiary sensors detect external changes that result from exercise. An example of a tertiary sensed parameter is body movement.64 As expected, the relationship between exercise workload and these tertiary variables is less tightly linked, and there is often greater susceptibility to environmental influences, such as vibration. Other measures, such as the use of a 24-hour clock to vary the lower pacing rate, can be considered tertiary sensors. Likewise, in the most primitive rateadaptive pacemaker, the pacing rate was changed by the patient, using a hand-held programmer, before beginning exercise.65 Technical Classification Although conceptually attractive, the physiologic classification does not adequately separate the bulk of the socalled secondary sensors. A more practical classification is to categorize sensors according to the technical methods that are used to measure the sensed parameter (Table 5-2). During isotonic exercise, body movements (especially those produced by heel strike during walking) result in changes in acceleration forces that are transmitted to the pacemaker. Sensors that are capable of measuring the acceleration or vibration forces in the pulse generator are broadly referred to as activity sensors. The sensing of body vibrations is therefore a

simple way to indicate the onset of exercise. Technically, detection of body movement can be achieved with the use of a piezoelectric crystal, an accelerometer, a tilt switch, or an inductive sensor. Each of these devices transduces motion of the sensor, either directly into voltage, or indirectly into measurable changes in the electrical resistance of a piezoresistive crystal. Although activity sensing is a tertiary sensor, it is the most widely used control parameter in rate-adaptive pacing because of its ease of implementation and its compatibility with all standard unipolar and bipolar pacing leads. Impedance is a measure of all factors that oppose the flow of electric current and is derived by measuring resistivity to an injected electric current across a tissue. The impedance principle has been used extensively for measuring respiratory parameters66,67 and parameters associated with right ventricular contractility, such as relative stroke volume or the pre-ejection interval,68 in situations involving invasive monitoring. The elegant simplicity of impedance has enabled it to be used with implantable pacing leads, including both standard pacing leads and specialized multielectrode catheters. The pulse generator casing has been used as one electrode for the measurement of impedance in most of these pacing systems. Impedance can be used to detect relative changes in ventilatory mechanics, right ventricular mechanical function, or the combination of these parameters. Relative motions between electrodes for impedance sensing also lead to changes in imped-

Major Classes of Sensors Used in Rate-Adaptive Pacing, Classified According to Method of Technical Realization TABLE 5-2.

Methods

Physiologic Parameter

Models

Manufacturers*

Vibration sensing

Body movement

Sigma, Kappa, Enpulse Diamond, Clarity, Selection AF Talent Miniswing, Neway Insignia, Pulsar Max, Discovery Actros, Protos, Philos Identity, Integrity, Affinity

Medtronic Vitatron ELA Medical Sorin Biomedica Guidant Biotronik St. Jude Medical

Impedance sensing

Minute ventilation Ventricular inotropic parameter Pre-ejection interval Lung fluid status

Kappa Talent Insignia, Pulsar Max Protos, Inos Precept† InSync Sentry

Medtronic ELA Medical Guidant Biotronik Guidant Medtronic

Evoked QT interval

Diamond, Clarity, Selection AF

Vitatron

Thermos† Deltatrax† Chronicle‡ — Best-Living system

Biotronik Medtronic Medtronic — Sorin Biomedica

— OxyElite† —

— Medtronic —

Ventricular evoked response Special sensors on pacing electrode

Physical parameters Central venous temperature dP/dt Right ventricular pressure Pulmonary arterial pressure Peak endocardial acceleration Chemical parameters pH Mixed venous oxygen saturation Catecholamine

*Biotronik GmbH & Co., Berlin; Guidant, Boston Scientific, Natick, Mass.; ELA Medical, Sorin Group, Milan; Medtronic, Inc., Minneapolis, Minn.; Sorin Biomedica, Sorin Group, Milan; Vitatron, Dieren, the Netherlands. † Not commercially available. ‡ Investigational device.

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

ance, and this is inversely related to the number of electrodes used to measure impedance.69 In rate-adaptive pacemakers, motion artifacts are usually the result of arm movements that cause the pulse generator to move within the prepectoral pocket,70,71 thereby changing the relative electrode separation between the pacemaker and the intracardiac electrodes. Because arm movement accompanies normal walking, these artifacts in the impedance signal occur with both walking and upper limb exercises. The intracardiac ventricular electrograms resulting from a suprathreshold pacing stimulus have been used to provide several parameters that can guide rate modulation. The area under the curve inscribed by the depolarization phase of the paced ventricular electrogram (the intracardiac R wave) has been termed the ventricular depolarization gradient or paced depolarization integral (PDI).72 In addition to depolarization, the total duration of depolarization and repolarization can be estimated by the interval from the pacing stimulus to the intracardiac T wave (the QT or stimulus-toT interval). Both of these parameters are sensitive to changes in HR and circulating catecholamines and can be derived from the paced intracardiac electrogram with conventional pacing electrodes. Because a large polarization effect occurs after a pacing stimulus, a modified waveform of the output pulse that compensates for afterpotentials is needed to eliminate this effect, so that these parameters can be accurately measured. The interaction of the “square wave” output pulse with the endocardium of a ventricular pacemaker leads to a distortion of the pacing waveform,73,74 and the “shape” of the resultant output pulse has been proposed to be useful in estimating intracardiac volume. The last group of sensors are those that are incorporated into a specialized pacing lead. Examples of these leads include thermistors (used to measure blood temperature), piezoelectric crystals (used to measure right ventricular pressure), optical sensors (used to –o2), and accelerometers at the tip of pacing measure Sv leads. Some of these sensors measure highly physio–o2 is closely related to logic parameters. For example, Sv oxygen consumption during exercise. Physical activities increase cardiac output and oxygen extraction from the blood, and a widening of the tissue arteriovenous oxygen difference occurs during exercise.59,75,76 How–o2 is not linearly related to the workload, and ever, Sv –o occurs during the first minute most of the drop in Sv 2 –o2 with increased of exercise, with less decrease in Sv workload. The preliminary clinical experience with –o2 sensors showed a rate response proimplanted Sv portional to exercise level.77-80 In one study, the rate –o2 sensor (OxyElite, Medtronic, Inc., response of an Sv Minneapolis, Minn.) was compared with that of a conventional piezoelectric activity sensor during activities of daily living (ADLs) and non–exercise-related physi–o sensor showed a better proologic changes. The Sv 2 portionality of rate response than the activity sensor and occurred at a comparable speed of onset.78,79 The –o2 sensor is its long-term stamain concern with the Sv bility, which may be affected by fibrin coating on the sensor. Although to a certain extent this instability is

207

–o2 reduced by using two different wavelengths for Sv measurements, the sensor’s ability to function in the long term remains an issue.79 Right ventricular pressure can be detected by a hermetically sealed pressure sensor containing a piezoelectric crystal and electronic circuitry incorporated into a pacing lead. The first derivative of the right ventricular pressure (dP/dt) is influenced by the contractile state of the heart, the ventricular filling pressure, and the HR, with a positive correlation between maximum dP/dt and sinus rate in healthy subjects.80 The change in the maximum value of dP/dt is a sensitive indicator of the change in right ventricular contractility and is directly proportional to change in sympathetic tone.81,82 The dP/dt sensing principle is highly proportional to workload, as shown in a limited number of investigational implants (Deltatrax, DPDT, Medtronic). The pacing rate achieved with the dP/dt sensor was reported to correlate well with estimated oxygen consumption during exercise (r = 0.93).13,83,84 The increases in pacing rate paralleled the HR that was expected from the metabolic reserve during treadmill testing in the VVIR mode. Exercise time was significantly prolonged in the VVIR mode compared with the VVI mode during paired exercise testing.85 An accelerometer incorporated at the tip of a unipolar ventricular electrode can be used to assess the contractile state indirectly from the endocardial vibration generated during isovolumic contraction of the heart, a parameter known as peak endocardial acceleration (PEA).86,87 Such a microaccelerometer has been developed by Sorin Biomedica (Saluggia, Italy) and has a frequency response of up to 1 kHz and a sensitivity of 5 mV/G (1 G = 9.8 m/sec). In preliminary experience in sheep under basal conditions using an external system and an implantable radiotelemetry system, the PEA was not affected by HR but was significantly increased by emotional stress, exercise, and natural inotropic stimulation.86 This parameter follows the changes in the maximum left ventricular dP/dt and apparently measures the global left ventricular contractile performance rather than the regional mechanical function of the right ventricle.88,89 In preliminary studies on the PEA-driven rate-adaptive pacemaker, there was a good correlation between the sinus rate and the PEA-indicated rate during ADLs and a submaximal stress test.90,91 A potential role of the PEA is its ability to optimize the A-V interval automatically. Limitations with this sensor system are several. First, the effects of the relative contributions of valvular movement and change in preload to the myocardial vibrations and the measured PEA are uncertain. Second, this system requires a dedicated lead with unproved long-term reliability and stability, although an initial study on sheep showed acceptable medium-term results.87 Closed-loop versus Open-loop Sensors A rate-adaptive pacing system can operate in either a closed-loop or an open-loop manner. In a completely closed-loop system (see Fig. 5-3), the physiologic parameter that is monitored is used to effect a change

208

Section One: Basic Principles of Device Therapy

in the pacing rate. Changes in pacing rate in turn induce a physiologic change in the sensed parameter in the opposite direction. Therefore, closed-loop pacing systems have negative feedback, such that the sensed physiologic variable tends to return toward its baseline value in the presence of an appropriately modulated pacing rate. A partial degree of closed-loop negative feedback control is observed with pacemakers that use –o2 as the rate-control parameter.59,75,76 Exercise in the Sv absence of adequate cardiac output (as in a patient with a fixed-rate pacemaker) increases tissue oxygen extraction from arterial blood, thereby decreasing the content of oxygen in the returning venous blood. This –o2 can be measured and used to increase decrease in Sv the pacing rate, thereby increasing the cardiac output to a value that is optimal for the level of exercise workload, resulting in improved oxygen delivery to the tissues. Under conditions of equilibrium, the pacing –o2 rate is adjusted to maintain the maximum possible Sv for any level of metabolic demand. The PDI has also been advocated as a closed-loop sensed parameter.72 An increase in sympathetic activity decreases the PDI, whereas an increase in HR increases it, thereby establishing a negative feedback loop that tends to maintain the sensed parameter at a relatively constant value during exercise. However, true closed-loop performance was not achieved in clinical trials with this sensor. Theoretically, the physician input required for a closed-loop system should be minimal, because the system is designed to be fully automatic. In practice, a rate-adaptive algorithm is still necessary, because the available pacing systems provide only partial closed-loop negative feedback. In an ideal closed-loop system, the sensor automatically takes into account any changes in the patient’s cardiovascular condition. Apart from setting the lower and upper rate limits, the physician can indirectly control the rate changes in a closed-loop system by determining the speed at which the pacing rate adjusts to return the sensed parameter to its baseline value. Although a closed-loop sensor is theoretically attractive, the practical application of this concept has been less than ideal. Normal control of HR involves multiple parameters and feedback measures, and it is unlikely that a single sensor can accurately control HR in all clinical circumstances. In addition, a closed-loop sensor involved in rate control may be affected by factors other than metabolic demand. Therefore, at present, the potential of closed-loop sensors remains unrealized. Open-loop logic is employed in most available sensors that measure either physiologic or biophysical parameters (see Fig. 5-3). In such open-loop systems, a change in the HR does not result in a negative feedback effect on the physiologic or physical parameter used to modulate the pacing rate. Therefore, open-loop algorithms require the physician to prescribe the relationship between the parameter monitored by the sensors and the desired change in pacing rate. An example of this is an activity-sensing pacing system that detects body movements. Physical exercise results in acceleration forces on the pulse generator that can

be used to increase the pacing rate.8,64 The resultant increases in pacing rate usually have minimal effects on body movement. In the extreme case, a positive feedback of HR on the rate-control parameter might occur, as exemplified by the old version of the QT interval sensing pacemaker. The QT interval shortens during exercise. However, an increase in the pacing rate itself induces further shortening of the QT interval, especially if a linear slope is used to relate changes in the QT interval to changes in pacing rate. Increases in pacing rate during exercise could shorten the QT interval excessively, leading to an excessive increase in rate.92 This type of “sensor feedback tachycardia”93 has also been described with rate-adaptive pacemakers that detect minute ventilation94 or body activity.95 The mixed venous oxygen sensor also has the potential for positive feedback during exercise-induced myocardial –o2 decreases in response ischemia. For example, the Sv to either reduced cardiac output (e.g., during ischemia) or increased oxygen consumption. If exercise-induced –o ischemia is the cause, the resultant decrease in Sv 2 might trigger a further increase in pacing rate, which may further exacerbate myocardial ischemia. Such a scenario has been observed with this sensor.

Rate-Control Algorithms and Rate-Response Curves The term algorithm refers to the way in which the raw sensor data are converted to a change in pacing rate. Typically, sensor data are first filtered to exclude unwanted signals (e.g., signals outside the frequency range of the rate-control parameter). The changes in the sensor signal over an averaged baseline (or, on rare occasions, the absolute sensor signals) are used for further processing. The filtered signals are appropriately modified through rectification and gain control (Fig. 5-4). The processed signals are then used to modulate pacing rate through the application of a ratecontrol algorithm. The physician must determine the ultimate rate response that will be observed by choosing the lower pacing rate, the upper pacing rate, and a rate-response curve that determines the slope of the sensor-pacing rate relationship. In some pacemakers, the physician can also modify the rate response by changing the “filter” used to process the raw sensor signal. An example of such filtering is the threshold feature of activity-sensing pacemakers. The relationship between the processed signals and pacing rate can be linear, curvilinear, or a more complex function (see Fig. 5-4). An example of an even more complex rate-response curve is the biphasic pattern of response of a more recent minute ventilation sensor. This algorithm provides a steeper slope of the pacing rate–minute ventilation relationship at the beginning of exercise than at the end of the exercise. The initial aggressive slope is made possible by a separately programmable “rate augmentation factor,” so that different slopes control

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity Figure 5-4. Types of rate-responsive curves used in rate-adaptive pacemakers employing a single sensed parameter (sensor level). An appropriate filter (F) eliminates unwanted raw signals (e.g., high- or low-pass frequency filters and thresholds). The filtered signals are then appropriately modified (M) (e.g., with gains and rectification) before being converted to a rate change. The curves can be linear (A), curvilinear (B), or complex (C and D). The physician can select an appropriate rate-responsive slope from a family of curves. See text for further discussion. IR, interim rate.

Curvilinear

Linear

 Rate

F

209

 Rate

M

 Sensor

A

F

M

 Sensor

B

Complex

 Rate

 Rate

IR

 F

M

C

the first and second halves of the pacing range. Furthermore, a linear relationship is maintained within the aerobic range of exercise but a less aggressive slope is used above the anaerobic threshold, where minute ventilation increases disproportionately to the HR change. Perhaps the most sophisticated rate-response curves are those proposed for temperature sensing (see Fig. 5-4D). In one algorithm, a curvilinear relation is employed when the temperature increases. Because a temperature “dip” characteristically occurs at the beginning of exercise, this algorithm responds to a rapid decline in temperature with a rapid increase in the pacing rate to an interim value. Subsequent increases in temperature are used to modulate further increases in pacing rate. In addition to programming the rate-response curve, temperature-sensing pacing systems provide for a gradual decrease in the lower pacing rate in response to diurnal variation in blood temperature (bidirectional arrow in Fig 5-4D). Thus, the rate of change in the rate-control parameter (in this case, a slow decrease in temperature within predefined limits during periods of rest) is used to vary the lower rate limit. Although the rate-response curves discussed

 Sensor

F

0

 M

 Sensor

D

previously define the relationship between sensor output and increases in pacing rate during exercise, many rate-adaptive pacemakers use a different set of curves to control decreases in pacing rate during the recovery phase. Note that the rate-response curves that have been discussed relate changes in the sensed parameter to changes in the desired pacing rate. In addition to these sensor-desired, pacing rate-response curves, other factors control the speed (or time constant) with which the sensor-indicated desired pacing rate is translated into a change in the actual pacing rate. Obviously, an abrupt change in the sensed parameter must be translated into a gradual increase (or decrease) in the pacing rate. For example, if the activity signals of a pacemaker were to double abruptly, the rate-response curve might indicate that the pacing rate should be increased from 70 to 100 bpm. The interval required for the pacing rate to increase from 70 to 100 bpm is a programmable feature in some rate-adaptive pacing systems and has a fixed time constant in others. Similarly, if the sensor indicates that the pacing rate should be decreased from 100 bpm to 70 bpm, a time constant is required to

210

Section One: Basic Principles of Device Therapy

translate this desired change into an actual decline in pacing rate. These “attack” and “decay” constants can have a major effect on the chronotropic response characteristics of rate-adaptive pacemakers.

Characteristics of an Ideal Sensor for Rate-Adaptive Pacing TABLE 5-3.

Considerations

Examples and Remarks

Sensor consideration

Characteristics of an Ideal RateAdaptive Pacing System The normal human SN increases the rate of its spontaneous depolarization during exercise in a manner that is linearly related to Vo2. Because this response undoubtedly has evolutionary advantages, the goal of rate-adaptive pacemakers that modulate pacing rate by artificial sensors has been to simulate the chronotropic characteristics of the SN. It is uncertain, however, whether the SN provides the ideal rate response in patients who require permanent pacemakers. Nevertheless, until there is evidence indicating otherwise, rate-adaptive pacemakers will strive to reproduce this physiologic standard. Keeping these uncertainties in mind, the ideal rate-adaptive pacing system should provide pacing rates that are proportional to the level of metabolic demand. In addition, the change in pacing rate should occur with kinetics (or speed of response) similar to those of the SN. The artificial sensor should be sensitive enough to detect both exercise and nonexercise needs for changes in HR and yet be specific enough not to be affected by unrelated signals arising from both the internal and the external environments. Although the ideal sensor should provide these functional characteristics, it must also be technically feasible to implement with a reliability that is acceptable with modern implantable pacemakers (Table 5-3).96

Principles Used for Comparing and Evaluating Rate-Adaptive Systems Proportionality of Rate Response One of the best indicators of sensor proportionality is the correlation between the sensor-indicated pacing rate and the level of oxygen consumption during exercise.13 In general, parameters such as minute ventilation and the paced QT interval are proportional sensors. Some sensors for which specialized pacing leads are –o2 is used are also highly proportional. For example, Sv closely related to oxygen consumption during exercise. To assess the proportionality of chronotropic response during exertion, the exercise workload should be increased gradually. Traditional treadmill exercise protocols used to evaluate coronary artery disease usually aim to reach maximum HR rapidly and tend to skip the lower workloads (3-5 Mets) that are performed normally by pacemaker recipients in their daily lives. Exercise protocols with gradually increasing workloads, such as the Chronotropic Assessment Exercise Protocol, are probably more appropriate for assessing the

Proportionality

Oxygen saturation sensing has good proportionality

Speed of response

Activity sensing has the best speed of response

Sensitivity

QT sensing can detect non–exerciserelated changes, such as anxiety reaction

Specificity

Activity sensing is affected by environmental vibration Respiratory sensing is affected by voluntary hyperventilation

Technical consideration Stability

Stability of early pH sensor was a problem

Size

Large size or requirement for additional electrodes may be a problem Energy consumption must not harm pacemaker longevity unduly

Biocompatibility

Important for sensor in direct contact with the bloodstream

Ease of programming

Difficult programming in early QT-sensing pacemakers

rate response of a pacemaker over the wider range of workloads (and oxygen consumption) that are relevant to these patients.97 Graded exercise testing to maximal tolerated workload may be impractical for some patients for assessing the function of a rate-adaptive pacemaker. Brief, submaximal ramp exercise tests are especially valuable for assessing the proportionality of current rate-adaptive pacing systems.13 These tests can be “informal,” such as asking the patient to walk at varying speeds or to ascend and descend stairs. In addition, monitoring of pacemaker function during ADLs may provide the most clinically relevant method of evaluating an elderly patient. Alternatively, submaximal exercises, such as treadmill tests at a low speed and grade, may be performed to assess the sensor response. These tests show that walking at a faster speed increases the pacing rate of most rate-adaptive pacemakers. However, the rate is not necessarily increased by walking up a slope in patients with activity-sensing pacemakers, which respond according to the pattern of body motion or vibration associated with each type of activity. For example, with many activity-sensing pacemakers, ascending stairs is associated with a lower pacing rate than is descending stairs, because the intensity of the heel strike is less walking upstairs than downstairs. These findings suggest that there is only a

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

moderate correlation between the rate achieved by activity-sensing pacemakers and exercise workload. These differences in chronotropic response may not be detected with graded treadmill exercise. Ambulatory electrocardiographic (ECG) monitoring, rate histograms, or stored rate trends may provide useful methods for evaluating the chronotropic response of rate-adaptive pacing systems in patients who are less active or who cannot exercise. Furthermore, very few patients with pacemakers (or, indeed, in the general population) exercise to maximal levels of workload on a regular basis. Therefore, formal exercise testing may have little clinical relevance in these patients. Speed of Onset of Rate Response and Recovery from Exercise An appropriate speed of response of the pacing rate to the onset of and recovery from exercise is an essential feature of a rate-adaptive pacing system. The onset kinetics are best assessed during treadmill exercise, such as walking at a fixed speed on the treadmill. From ECG monitoring, the delay time for the onset of rate response, the time required to reach one half of the maximum change in pacing rate during exercise (halftime), and the time required to reach 90% of the maximum response can be derived and used as a basis for comparison. The exercise responses of six different types of rate-adaptive pacemakers (with sensors for activity, QT interval, respiratory rate, minute ventilation, and right ventricular dP/dt) were compared with normal sinus rate in one study.13 The results demonstrated that the activity-sensing pacemakers best simulated the normal speed of rate response at the start of exercise. The rate response of activity sensors is usually immediate (no delay time), and the half-time was within 45 seconds from the onset of exercise. The maximum change in pacing rate was reached within 2 minutes after beginning an ordinary activity, such as walking. The respiratory rate and the right ventricular dP/dt sensors had a longer delay time (about 30 seconds) and half-time (1 to 2 minutes), although the maximum change in rate was still attained within 2 to 3 minutes after beginning exercise. The slowest sensor to respond to exercise was an early version of the QTsensing pacemaker, which required up to 1 minute to initiate a rate response, and the maximum change in pacing rate was attained only in the recovery period after a short duration of exercise. The onset of rate response and proportionality to workload of the QTsensing pacemaker was in sharp contrast to those of activity-sensing pacemakers. The speed of onset of the newer generation of QTsensing pacemakers has been significantly improved by the use of a linear (rather than curvilinear) rateresponse slope that produces a larger change in pacing rate per unit change in QT interval at the onset of exercise (slow HRs) than at higher workloads.98 These minute ventilation–sensing pacemakers produce an increase in pacing rate that is linearly related to minute ventilation throughout exercise, providing the effect of shortening the rate-response half-time. Furthermore,

211

the speed of onset of rate response is programmable in these newer generations of minute ventilation sensors. After termination of exercise, body movement decreases, and the pacing rate of an activity-sensing pacemaker returns toward the resting level based on an arbitrary rate-decay curve.99 If the rate decay is faster than is physiologically appropriate, adverse hemodynamic consequences may occur in the presence of a substantial decrease in HR. In one study in which pacing rate was reduced either abruptly or gradually after identical exercise, it was shown that an appropriately modulated rate recovery was associated with a higher cardiac output, lower sinus rate, and faster lactate clearance, compared with a nonphysiologic rate-recovery pattern.100 Appropriate adjustment of the rate-recovery curve is important to enhance recovery from exercise. Sensitivity of a Rate-Adaptive Pacing System to Changes in Exercise Workload and Other Physiologic Stresses Table 5-4 shows the factors to which some rate-adaptive pacemakers are sensitive. Rate-adaptive pacemakers that are controlled by ventricular-evoked response and intracardiac hemodynamic parameters are able to respond to emotional stresses. A reverse rate response has been observed during the Valsalva maneuver in patients with respiratory sensing pacing systems and some dP/dt sensing pacemakers. None of the available pacemakers reliably detects changes in posture, although several sensors, such as intracardiac impedance or accelerometer designs, have the potential to do so (Fig. 5-5).79,101,102 A paradoxical decrease in HR may be observed during movement to the upright position with pacemakers that sense the ventricular depolarization gradient.103 A varying postural drop in HR has also been reported with a rate-adaptive pacemaker that detects the dP/dt.89,90 A diurnal rate variation is possible with temperature-sensing and QT-sensing pacemakers. The clinical implication of some of these rate changes to nonexercise stimuli remains to be determined. Specificity of Rate-Adaptive Pacing Systems One of the main limitations of activity-sensing pacemakers is their susceptibility to extraneous vibrations. This typically occurs during various forms of transport. The degree of susceptibility to extraneous vibrations may vary with different types of activity sensors. For example, an accelerometer using a tilt switch (Swing, Sorin Biomedica) may be one of the most susceptible types. The QT interval and the PDI may be significantly affected by such factors as cardioactive medications and myocardial ischemia. Ischemia may result in shortening of the QT interval, leading to an increase in pacing rate and further myocardial ischemia.104,105 Sensors that use impedance to measure respiratory mechanics or the right ventricular pre-ejection interval are susceptible to artifacts produced by arm movement, hyperventilation, and speech.70,71,106 Electric diathermy is likely to cause inappropriate changes in pacing rate in pacemakers that

212

Section One: Basic Principles of Device Therapy

Physiologic Sensitivity of Some Currently Available RateAdaptive Pacemakers TABLE 5-4.

Exercise Physiologic Measure

Isotonic

Isometric

Emotion

Valsalva Maneuver

Posture

Diurnal Variation

Sinus

+

+

+

+

+

+

Activity

+

Respiration

+

QT interval

+

Gradient

+

Temperature

+

PEI

+

+

+

dP/dt

+

+

+

–O Sv 2

+

+

PEA

+

+

VIP

+

+

R + +

+

+

R

±

+ V V

V

+

dP/dt, maximum first derivation of right ventricular pressure; PEA, peak endocardial acceleration; PEI, pre-ejection interval; R, reversed response; –O , mixed venous oxygen saturation; V, variable; VIP, ventricular inotropic parameter. Sv 2

measure impedance.107 The same problem may be expected to occur during radiofrequency ablation in impedance-sensing pacemakers. In addition, external temperature change can significantly affect the pacing rate of temperature-sensing pacemakers.

Clinical Contraindications to Specific Rate-Adaptive Sensors A number of clinical factors may preclude the use of some sensors for an individual patient (Table 5-5). When a parameter can be detected only at the ventricular level, the sensor cannot be used in an AAIR pacing system. The use of antiarrhythmic medications,

the use of β-blockers, and the presence of myocardial ischemia may interfere with the detection of the QT and affect its duration. The QT system is sensitive to adrenergic stimulation and response to emotional stress.108,109 In patients with ischemic heart disease, this response to psychological stress may occasionally be excessive, with an undesirable rate increase, thereby precipitating angina. Excessive adrenergic tone may cause a QT pacemaker to pace at the upper rate limit in the setting of acute myocardial infarction, creating an undesirable and potentially harmful response.104,105 On the other hand, the dP/dt sensor does not appear to be adversely affected by myocardial ischemia.110 The clinical impact of ischemia in patients with rate-adaptive pacemakers controlled by oxygen saturation remains to be determined.78

count

ActVar

Act standing

Act Threshold ActVar Threshold

resting time

Figure 5-5. Schematic representation of the Orthostatic Response (OSR) algorithm for detection of posture change. The accelerometer sensor provides two outputs: the instantaneous activity level (Act, solid line) and the moving average of the absolute difference between Act levels (ActVar, dotted line). If both signals are lower than the corresponding threshold (horizontal lines), prolonged rest is detected. With changes in posture, the Act rises above the Act threshold, ActVar remains below ActVar threshold, and OSR pacing is triggered. (From Tse HF, Siu CW, Tsang V, et al: Blood pressure response to transition from supine to standing posture using an orthostatic response algorithm. PACE 28:S242-S245, 2005.)

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

213

Clinical Factors Contraindicating Use of Some Currently Available Rate-Adaptive Pacemakers TABLE 5-5.

Factor

Atrial Pacing

Antiarrhythmic Drugs

Myocardial Ischemia

Young Children or Respiratory Disease

Standard Unipolar Lead

Activity

Exposure to High-Vibration Environment —

Respiration

—

—

QT interval

—

—

—

Gradient

—

—

—

Temperature

±

PEI

—

—

—

—

dP/dt

—

±

—

—

–O Sv 2

±

PEA

—

±

±

VIP

—

±

—

— —

—

— —

±

–O , mixed venous dP/dt, maximum first derivation of right ventricular pressure; PEA, peak endocardial acceleration; PEI, pre-ejection interval; Sv 2 oxygen saturation; VIP, ventricular inotropic parameter; —, unsuitable; ±, feasibility remains to be validated.

The minute ventilation–sensing pacemakers are best avoided in young children, who may have very rapid respiratory rates during exercise that exceed the range detected by the pacemaker. Inappropriate tachycardia may occur in patients with advanced heart failure and the rapid breathing phase of Cheyne-Stokes dyspnea.94 During cesarean section111 and general anesthesia,112 passive hyperventilation may induce pacemaker tachycardia; hence, a non–rate-adaptive mode is preferred when these patients undergo general anesthesia. Electrocautery in the thoracic area may affect impedancesensing pacemakers and may lead to upper rate pacing; therefore, it is preferable to program the pacemaker to the non–rate-adaptive mode when electrocautery or radiofrequency ablation is used.107 Occupations associated with exposure to vibrations in the environment, such as horseback riding,113 and various types of transportation that may cause rate acceleration are a relative contraindication to the use of an activity sensor. In patients with heart failure, the use of temperature sensors tends to be difficult because a prolonged temperature fall occurs at the onset of exercise in some of these patients.114 In addition, replacing or upgrading a pulse generator requires that the new sensor be compatible with the existing pacing lead, unless a new lead is contemplated. Therefore, if a unipolar ventricular pacing lead is to be used, a minute-ventilation pacemaker that requires a bipolar lead in the ventricle is not feasible.

Sensor Combinations The only sensors that are presently in clinical use are those that can be used with a standard lead. The

instability of sensors requiring a special lead prevents their widespread use for rate adaptation. The surviving sensors include activity, minute ventilation, intracardiac impedance, and QT sensors (used only in combinations). Combining artificial sensors is feasible and may be superior to rate adaptation with a single sensor. In addition, automatic programming of sensors is feasible and effective, making the complexity of multisensor pacemaker programming an insignificant issue. Justification for Sensor Combinations There has been significant improvement in instrumentation, and rate-adaptive algorithms have been incorporated in the “clinical sensors” to address the issues of speed of onset, proportionality, specificity, and sensitivity of sensor response. For example, piezoelectric crystals for activity sensing using a “peak counting” algorithm are limited by the relatively poor ability of the sensor to differentiate between different levels of workload99 and their susceptibility to external vibration. Some of these aspects are improved by the use of an accelerometer and an algorithm that integrates the activity signals to determine the sensor-indicated rate.115-117 Because body movement has no direct relationship to metabolic requirement, however, this sensor remains inadequate to detect isometric exercise, nonexercise needs, and exercises that do not result in significant vibration (e.g., bicycle riding). Minute ventilation as measured by impedance delivers appropriate rate-adaptive therapy that is proportional to workload.9,10 Criticism has centered on the slower HR response of minute ventilation sensors to the

214

Section One: Basic Principles of Device Therapy

TABLE 5-6. Speed of Response Activity

Components of a Multisensor System for Rate-Adaptive Pacing Proportionality

Sensitivity

Specificity

“Energy Saver”

Minute ventilation

Diurnal variation: 24-hour clock Diurnal activity change Diurnal minute ventilation change

Minute ventilation

Detection of capture: Evoked QRS Stroke volume

QT

Emotional response: QT PEA

QT

Rate reduction during sleep

Activity

Minimize myocardial oxygen consumption

PEA

PEA, peak endocardial acceleration.

onset of exercise, as compared with activity sensors, with a 30-second delay when compared with the response of the SN.118 This response is also potentially influenced by conditions that may not be directly relevant to cardiac output, such as talking or voluntary respiration.119 New, faster algorithms with a programmable rate augmentation factor and speed of response have improved this slow response during the early stages of exercise.120 This, however, leads to a more rapid recovery, with a significantly shortened recovery time at the end of exercise. The main limitation of the QT sensor is the relatively slow speed of onset of the rate response121 and the susceptibility of the QT interval to drugs and ischemia. With the use of curvilinear rate-response curves that have a higher slope at the onset of exercise, the lag in the onset of rate response is reduced,98 but the speed of rate response is still too slow during brief periods of exercise.122 Therefore, the new generation of clinical sensors remains imperfect, even when only speed and proportionality of rate response are considered. In addition, apart from the QT and the ventricular impedance sensors, which react to emotional changes,123 none of the other clinical sensors is sensitive to nonexercise needs such as changes induced by postural, postprandial, and vagal maneuvers; fever; and circadian variations. These limitations of the available sensors mean that none of them is suitable for every patient under all cir-

TABLE 5-7.

cumstances. Despite the fact that the response of a sensor can be significantly enhanced by fine-tuning the characteristics of the sensor and the algorithms used to translate sensor output into modulation of pacing rate, the “clinical” sensors are mainly limited because a fastresponding sensor is not proportional, whereas a proportional sensor is relatively slow (Table 5-6). In addition, an activity sensor is relatively insensitive to nonexercise stress and is nonspecific and liable to external interference. Dual sensor combinations are aimed to create an integrated sensor that simulates the SN response of healthy individuals by combining the strong points and eliminating the weak points of the individual sensors (Table 5-7). The sensor combination aims to improve the speed of rate response, proportionality to workload, sensitivity to physiologic changes induced by exercise and nonexercise requirements, and specificity in rate adaptation (Fig. 5-6). A sensor that is more specific to the onset of exercise can be used to prevent false-positive rate acceleration caused by a more sensitive yet relatively nonspecific sensor. In the absence of the specific sensor indicating that exercise is occurring, the HR response of the other sensor can either be nullified or restrained. Multisensor pacing may also offer the possibility of selecting an alternative sensor should one sensor fail or become inappropriate for an individual patient. In addition, an appropriate rate of recovery can shorten

Relative Advantages of Clinically Used Sensors Speed

Proportionality

Specificity

Sensitivity

Activity

High

Low

Low

Low

Minute ventilation

Moderate

High

Moderate

Low

QT interval

Low

Moderate

High

Moderate

PEA

Moderate

Moderate

Moderate

High

VIP

Moderate

Moderate

Moderate

High

PEA, peak endocardial acceleration; VIP, ventricular inotropic parameter.

215

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

repayment of oxygen debt and promote lactate clearance.100 The potential for combining sensors for purposes other than rate modulation during exercise is a strong incentive for the development of multisensor pacemakers.

SR

S1

S2

S1S2 Exercise

Nonexercise stress

Interference

Proportionality and Speed

Sensitivity

Specificity

Figure 5-6. Algorithms for sensor combinations needed to achieve better (1) proportionality and speed of response, (2) sensitivity, and (3) specificity. The graphs (top to bottom) depict the responses of the sinus node (SR), sensor 1 (S1), sensor 2 (S2), and combined rate profile of S1 and S2. SR shows ideal proportionality, speed of rate response, and freedom from interference. S1 is a rapidly responding sensor, although it is neither proportional nor sensitive and is susceptible to interference. S2 is a proportional and sensitive sensor, although it has a slow response. It is also specific to exercise. Note the improved ability of the combined sensor approach in simulating the sinus rate under different conditions.

Additive

Principles for Integrating Rate-Adaptive Sensors Algorithms for Combining Rate-Adaptive Sensors Two basic methods for combining sensors to control chronotropic response during exercise have been used. The types of sensor combination can be “faster-win” or “blending” (Fig. 5-7A and B). In the faster-win form, the inputs from two sensors are compared, and the sensor indicating the faster rate is chosen to regulate the pacing rate. A differential combination (blending) combines the input of two sensors, either as a fixed ratio of one sensor to the other or as a variable ratio that changes in relation to the HR. For example, a fastresponding sensor (such as an activity sensor) can be used to modulate pacing rate at the onset of exercise, with a second sensor (such as a minute ventilation or QT sensor) to modulate the pacing rate during more prolonged exercise. The pacing rate may increase to an “interim” or intermediate value when the faster sensor detects the onset of exercise; a more proportional rate increase will occur when the slower, more proportional sensor “catches up.” A variation of this approach is to calculate the output of each sensor in a relative proportion, so that the ultimate rate profile is a blend of both. The pacing rate can be controlled by two sensors that have different sensitivities to exercise and nonexercise physiologic stresses, so that the system can respond to both exertional and emotional needs. It is conceivable that separate rate-adaptive slopes (or different upper and lower rates) can be programmed for modulation of rate in response to exertional versus

Blending

Cross-Check S1 S2

Work Exercise

Sensor indicated rate

Sensor indicated rate

Sensor indicated rate

S1S2

Exercise Interference

A

B

C

Figure 5-7. Algorithms to combine sensors. A, In the faster-win algorithm, the faster rate, from either sensor 1 (S1) or sensor 2 (S2), is chosen. B, In the blending algorithm, the sensor rate responses from S1 and S2 are blended to give an intermediate rate. C, In the sensor crosschecking algorithm, the more specific sensor S2 cross-checks the less specific fast sensor S1 when S2 does not indicate the presence of exercise.

216

Section One: Basic Principles of Device Therapy

emotional needs. The algorithm can be designed to weigh the input of both sensors to diagnose a nonexercise physiologic stress and provide a different pattern of rate adaptation. Sensor Cross-Checking Algorithms for Enhancing Specificity The response of one sensor can also be checked against the output from another sensor to improve the specificity of the chronotropic response (see Fig. 5-7C). A more specific sensor may be used to cross-check a nonspecific sensor, thereby avoiding inappropriate rate acceleration. In this instrumentation, the rate adaptation of a less specific sensor is allowed to increase the pacing rate only over a restricted range of HRs and for a limited duration. In the absence of a determination of exercise by the other sensor, the diagnosis of falsepositive rate acceleration with the first sensor is made, and the pacing rate returns to the baseline, so that prolonged high rate pacing is avoided. Such sensor crosschecking can be reciprocal between the two sensors, so that either sensor may limit the chronotropic response that results from the other. In practice, cross-checking is usually applied to limit the less specific of the two sensors. Possible Sensor Combinations A number of possible sensor combinations have been suggested (see Table 5-6). One of the simplest is a 24hour clock that is used to vary the lower pacing rate. The normal diurnal variation in HR is well recognized, and an automatic decrease in the lower rate during the hours of sleep is physiologically appropriate. Battery consumption can also be reduced by this reduction in average pacing rate. Because of its simplicity, reliability, and compatibility with any pacing lead, an activity sensor that has a fast onset of response to exercise may be used as one sensor in combination with another sensor that provides a more proportional response to workload. An activity sensor can be easily added to the pulse generator. It requires minimal energy consumption to operate and is compatible with other sensors. Activity has been combined with central venous temperature,124 a parameter that is more proportional to metabolic need during prolonged exertion than is activity. Similarly, the combination of QT interval and activity sensing (e.g., Diamond, Vitatron B.V., Arnhem, The Netherlands) enhances the speed of response compared with a QT sensor alone.14,125-127 An activity sensor has been combined with a minute ventilation sensor in single- and dualchamber pacemakers (e.g., Kappa 400, Medtronic; Pulsar Max and Insignia, Guidant, Boston Scientific, Natick, Mass.; Symphony and Talent, ELA Medical, Sorin Group, Milan). The sensing of intracardiac impedance is one of the simplest ways to combine sensors. Despite the many possible sensor combinations, only three clinical sensors have been used clinically in sensor combination (activity, minute ventilation, and QT).

Dual-sensor Rate-adaptive Pacemakers QT and Activity In the Topaz and Diamond pacemakers (Vitatron), a piezoelectric sensor is used for activity sensing. The algorithms for combining the activity and QT sensors are both blending and cross-checking. Activity and QT input can be programmed at different contribution levels: activity < QT, activity = QT, or activity > QT, representing ratios of 30:70, 50:50, and 70:30, respectively. To avoid false rate acceleration by the activity sensor, the pacemaker allows activity rate response for only a short duration unless confirmed by changes in QT sensor (sensor cross-checking). The blending of the QT and activity sensors shows a quick rate response at exercise onset and a more proportional rate response during the latter part of exercise and during the recovery period.125 The fast rate-adaptive response during the first stage of exercise is due primarily to the activity sensing, with a high correlation between the activity sensor counts and the mean pacing rate (r = 0.94). The QT sensor predominates during higher levels of exertion, with a low correlation between activity sensor counts and the pacing rate (r = 0.14).125 In a multicenter study of 79 patients with the Topaz pacemaker, exercise in the dual-sensor mode produced a more gradual rate response than it did with the activity mode alone. The rate profile during treadmill exercise testing with dual-sensor pacing was improved over that of single-sensor pacing.38,126 In one study, simultaneous recording and comparison of combined sensor pacing and the sinus rate during ADLs and standardized exercise testing were performed in 12 patients. There was an improved correlation between the dual-sensor-indicated rate and the sinus rate with the combination, compared with either individual sensor alone (Fig. 5-8).127 Furthermore, an inappropriate high rate response from the activity sensor caused by external vibrations could be limited by sensor blending and cross-checking (Fig. 5-9).128,129 With a too-sensitive activity sensor setting, however, activity counts may be registered at rest when the QT sensor is inactive. This may result in crosschecking of the activity sensor by the QT sensor, which could delay the speed of the dual-sensor rate response when exercise begins.129 Minute Ventilation and Activity A piezoelectric activity sensor has been combined with a minute ventilation sensor to improve the initial response time while allowing a proportional rate response to higher workload. In the Kappa 400 pacemaker (Medtronic), the pacing rate is determined by automatic blending of the activity and minute ventilation sensors, using daily activities as a guide. Both sensors in the Kappa 400 contribute to the sensorindicated rate between the lower and an interim rate limit, the so-called ADL rate.130 The influence of the activity sensor diminishes and shifts toward the minute ventilation sensor as the integrated sensor-indicated rate increases toward the ADL rate. At HRs greater than

Chapter 5: Sensors for Implantable Devices: Ideal Characteristics, Sensor Combinations, and Automaticity

Low Moderate High

P 150 msec, no indication for pacemaker or ICD HF hospitalization in the past year

Combined all-cause mortality and allcause hospitalization† QOL, functional capacity, peak exercise performance, cardiac morbidity

1520

Cardiac Resynchronization in Heart Failure (CARE-HF) (Europe)

Open-label, randomized, controlled trial of CRT + optimal medical therapy vs. optimal medical therapy alone

2001-2003 (2004)

NYHA class III or IV, LVEF ≤ 0.35, LVEDD ≥ 30 mm/m (height), QRS > 50 msec or QRS ≥ 120 msec + echocardiographic criteria of dyssynchrony; stable optimal medical therapy

All-cause mortality or unplanned cardiovascular hospitalization† All-cause mortality, all-cause mortality or hospitalization for HF, NYHA class, QOL, echocardiographic LV function, neurohormone levels, economic impact

800

Study (Location)

Inclusion Criteria

Endpoints

N*

Results Stopped early owing to reduced all-cause mortality and hospitalization with CRT; reduced allcause mortality with CRT-D

Improvements in morbidity and mortality/ cardiovascular hospitalization

*Accrual or accrual goals. † Primary endpoint. 6MWD, 6-minute walk distance; AT, anaerobic threshold; AV, atrioventricular; BiV, biventricular; CRT, cardiac resynchronization therapy; CRT-D, CRT with defibrillator; ICD, implantable cardioverter-defibrillator; HF, heart failure; ICD, implantable cardioverter-defibrillator; LV, left ventricle; LVEDD, left ventricular end-diastolic diameter; NYHA class, New York Heart Association functional class; QOL, quality of life; RV, right ventricle; VO2, oxygen uptake; VF, ventricular fibrillation; VT, ventricular tachycardia. Adapted from Saxon LA, De Marco T, Prystowsky EN, et al: Executive Summary: Resynchronization Therapy for Heart Failure. Executive Consensus Conference, May 8, 2002. Available online at http://www.hrsonline.org/positionDocs/CRT_12_3.pdf/

endpoints, the initial trials assessed only measures of heart failure functional status, LV systolic function (LVEF), and LV remodeling (LV end-systolic and diastolic dimensions). The later, larger studies targeted mortality and hospitalization endpoints (Table 12-3). The use of these multiple endpoint measures is standard for heart failure trials evaluating medical therapies and has created an interesting issue with regard to defining benefit from CRT.1,6 One can define response as consisting only of symptom improvement or can require that all three of the measures of heart failure show benefit, as outlined in Table 12-3. To further complicate the issue, there does not appear to be a 1:1 correlation between these measures of response. As already described, the CONTAK CD and MIRACLE ICD studies enrolled patients with NYHA class II in addition to those with NYHA classes III and IV status. In the NYHA II group, significant improvements in measures of functional status were not uniformly observed, yet some of these patients experienced a positive reverse remodeling response.7,14 Clearly, the therapy has a positive effect on decreasing LV size, but it does not affect symptom status that was not severely compromised at baseline. In spite of this improvement, the less symptomatic NYHA II subset of patients has not been well studied and is currently not included in FDA-labeled indications for CRT devices.56

TABLE 12-3. Study Endpoints Evaluation in Trials of Cardiac Resynchronization Therapy Devices Measures of functional status

Quality of life 6-minute walk distance Cardiopulmonary exercise test

Measures of heart failure progression

Left ventricular ejection fraction, ventricular volume, dimension Mitral regurgitation Serum catecholamines, brain natriuretic protein, heart rate variability

Measures of heart failure outcome

Hospitalization Mortality

The Multisite Stimulation in Cardiomyopathy Studies Published in 2001 and 2002, the MUSTIC European studies provided the first long-term controlled trial data on the efficacy of CRT, delivered as BiV stimulation, for 3-month intervals, in comparison with normal sinus rhythm or continuous RV-based pacing in AF.3,54

390

Section Two: Clinical Concepts Inactive pacing

Implantation

CO1

CO2

Baseline Randomization Active pacing

4 weeks

2 weeks

12 weeks

12 weeks

Figure 12-2. Patients were randomly assigned to 3 months each of inactive pacing (ventricular, inhibited at a basic rate of 40 beats per minute [bpm]) and active pacing (atriobiventricular). CO1 denotes the end of crossover period 1, and CO2 the end of crossover period 2. (From Cazeau S, Leclercq C, Lavergne T, et al: Multisite Stimulation in Cardiomyopathies (MUSTIC) Study Investigators: Effects of multisite biventricular pacing in patients with heart failure and intraventricular conduction delay. N Engl J Med Mar 22;344:873-880, 2001.)

Figure 12-2 illustrates the crossover study design of the MUSTIC and MUSTIC AF studies. Although only 48 patients completed the two 3-month crossover study periods, all leads placed in the trial were transvenous and there were no significant safety issues. Eligibility for patient enrollment included NYHA class III with a QRS duration longer than 150 msec. In those patients with normal sinus rhythm, quality of life (QOL) score improved by 32%, 6-minute walk distance (6MWD) improved by 23% and peak Vo2 improved by 8%. Although the study was not statistically powered to determine a reduction in rate of hospitalization, hospitalizations after CRT initiation decreased by two thirds.3 At the end of the crossover phase, patients (who were blinded to treatment) were asked to choose which 3month period they preferred; 85% chose the pacing period during which they had been assigned to VDD, 10% had no preference, and 4% chose ODO (no pacing). Four patients had severe episodes of congestive heart failure exacerbation during the ODO pacing period. In the patients with permanent AF and continuous RV pacing, 37 of 59 who underwent CRT implantation completed both 3-month crossover phases and had documentation of 97% to 100% CRT delivery. Because of the significant number of dropouts (42%), the intention-to-treat analysis did not show a significant improvement with CRT. In the 37 patients with a complete data set and documentation of CRT, QOL measure did not improve, but 6MWD and peak Vo2 increased significantly by 9% (P = .05) and 13% (P =.04), respectively.54 When the entire 6-month crossover phase is considered, 10 of 44 patients were hospitalized for heart failure decompensation during RV pacing, whereas only 3 patients were hospitalized for heart failure during the CRT period. Eighty-five percent of patients preferred the CRT period. There was a trend toward a better QOL among patients with CRT (11% improvement; P = .09). Subsequent uncontrolled trials in patients with permanent AF and continuous RV pacing have shown a more robust improvement in these measures as well as demonstrating a reverse

remodeling response with CRT compared with RV pacing alone.57,58 The PAVE trial (see later) also showed improvement with CRT but did not target patients with heart failure per se for enrollment.55 Pacing Therapies in Congestive Heart Failure The two Pacing Therapies in Congestive Heart Failure (PATH I and II) European studies were groundbreaking in that chronic device programming was based on acute hemodynamic measures of cardiac performance. In addition, these same measures were used to optimize AV delay programming.52,53 These trials were begun in 1995 and completed in 1998. They were performed in Europe and enrolled patients with NYHA III or IV congestive heart failure, sinus rate higher than 55 beats per minute (bpm), and QRS duration longer than 120 msec. In PATH I, patients were crossed over between LV and BiV stimulation with a 1-month interval of no stimulation. A second study phase lasted 9 months and used the CRT mode that achieved what the follow-up physician determined was the most optimal mode. There were no differences in the acute or chronic response whether patients were programmed to LV or BiV CRT. Statistically significant improvements in peak Vo2 anaerobic threshold (24% improvement; P < =.001), 6MWD (25% improvement; P < .001) and QOL (59% improvement; P < .001) were observed at 3 months and 12 months during follow-up. Twenty-one of 29 patients followed up to 12 months improved from NYHA class III or IV to class I or II. Heart failure hospitalizations were decreased from 76% in the year prior to implantation to 31% during the year after implantation. It should be noted that LBBB was the type of conduction delay in more than 87% of patients; most CRT trials enroll up to 30% of patients with either interventricular condition delay (IVCD) or RBBB.4-6,9 This difference may account for the fact that LV stimulation in PATH I resulted only in an equivalent response to BiV stimulation to achieve CRT, although the study was not statistically powered to demonstrate a difference

Chapter 12: Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators

between the two modalities and the long-term data were pooled from both modes. The best that one can say is that in small numbers of patients who undergo hemodynamic optimization programming during implantation that shows equivalence between LV and BiV stimulation to achieve CRT, long-term symptom responses appear to be equivalent. Extending the observations from PATH I, PATH II evaluated LV-only CRT compared with no CRT in a 3month crossover design.53 In all patients with LBBB (88% of subjects), LV pacing was identified as the optimal single-chamber pacing mode (compared with RV only) on the basis of immediate hemodynamic response, and AV delay timing was optimized in all patients. Patients were further divided by QRS duration according to whether the QRS was more than 120 msec but less than 150 msec (“short QRS”) or more than 150 msec (“long QRS”). Unfortunately, only 35 patients, slightly less than one half of all patients enrolled, completed both 3-month crossover intervals. Nonetheless, the study did demonstrate improvements in peak Vo2, anaerobic threshold, 6MWD, and QOL score in the patients with long QRS duration. For example, 71% of patients in the long QRS group and 38% of patients in the short QRS group had an increase in the peak oxygen uptake of more than 1 mL/kg/min with active pacing. The short QRS group did not have an improvement in peak oxygen uptake or any other endpoint measure. This was the first study to demonstrate that QRS duration predicts the magnitude of symptom response to CRT delivered as LV-only stimulation. Subgroup analysis of all but one of the larger U.S. long-term studies of BiV CRT has also suggested that the magnitude of

391

benefit may be greater in patients with longer QRS duration at baseline.5,7,8-10 Multicenter InSync Randomized Clinical Evaluation The Multicenter InSync Randomized Clinical Evaluation (MIRACLE) study was the only U.S. trial of CRT for heart failure that used a CRT device only.5 The numbers of patients randomly allocated in U.S. clinical trials was much greater those in the European trials until the CARE-HF trial (see later). All 453 patients enrolled in the MIRACLE study underwent implantation of the CRT device and were then randomly assigned to “CRT on” or “CRT off” status for a period of 6 months. Figure 12-3 illustrates the study design of the MIRACLE, MIRACLE ICD, and CONTAK CD U.S. trials. Unlike the COMPANION trial (see later), in which patients were randomly assigned after consent was obtained and before device implantation, these earlier U.S. trials randomly assigned patients only after a successful CRT implant. The U.S. trials also employed strict protocol-mandated criteria relative to appropriate and stable heart failure medical regimen requirements prior to consent and device implantation. The primary endpoints, including 6MWD, QOL score, and NYHA class, were all favorably influenced by CRT, and the effects of CRT were apparent as early as 1 month after therapy initiation. Patients who underwent CRT showed 13% improvement in 6MWD, 13% improvement in QOL, a roughly 1 mL/kg/min improvement in exercise capacity, and an increase in total exercise time of approximately 60 seconds. Unlike in the

CRT CONTAK CD Baseline 6MW, CPX

Implantation

⬎30 days

6 months

CRT

No CRT

CRT MIRACLE ICD Baseline 6MW

Baseline CPX

Implantation 0-7 days

Figure 12-3. Study design of three U.S. trials: CONTAK CD Biventricular Pacing Study (CONTAK CD), Multicenter InSync ICD [implantable cardioverter-defibrillator] Randomized Clinical Evaluation (MIRACLE ICD), and Multicenter InSync Randomized Clinical Evaluation (MIRACLE). 6MW, 6-mile walk distance; CPX, cardiopulmonary exercise test; CRT, cardiac resynchronization therapy.

6 months

CRT

No CRT

CRT MIRACLE Baseline 6MW, CPX

Implantation

6 months

No CRT

CRT

392

Section Two: Clinical Concepts

European and acute hemodynamic studies, neither baseline QRS duration nor type of bundle branch block influenced response to CRT in the MIRACLE study. The secondary endpoints, of Vo2 and LVEF, also improved with CRT, as did episodes of heart failure worsening, including heart failure hospitalizations. At 6 months, CRT was associated with a reduction in LV enddiastolic and end-systolic volumes, reduced LV mass, increased ejection fraction (+3.6%), reduction in mitral regurgitation jet area (−2.5 cm2), and improvement in the myocardial index. Improvements in LV end-diastolic volume and ejection fraction were twofold greater in patients with nonischemic cardiomyopathy.

effects occur as early as 4 weeks and are sustained for a time even after CRT is suspended, indicating that CRT affects cardiac structure.61 Another measure of heart failure progression, level of plasma neurohormones, did not improve or worsen with CRT in the MIRACLE ICD or MIRACLE study. This neutral effect may be due to the fact that medical therapy with neurohormonal antagonists was optimized before device implantation or the possibility that the duration of follow-up may have been inadequate.7

CONTAK CD and Multicenter InSync ICD Randomized Clinical Evaluation

The Comparison of Medical Therapy, Pacing, and Defibrillation on Heart Failure (COMPANION) study was the first and only U.S. trial statistically powered to assess the impact of CRT on hospitalization and mortality endpoints.9,62 At the time of the COMPANION study design, it was unclear whether ICD therapy in addition to CRT would reduce mortality in advanced heart failure in comparison with medical therapy. Therefore, the study randomly assigned patients to optimal medical therapy for heart failure (OPT), a CRT device alone, or a CRT with an ICD (CRT-D). The patients were assigned in a 1:2:2 ratio, respectively, to maximize the number of patients receiving devices. There was not sufficient statistical power to directly compare CRT with CDT-D (both were compared with OPT), but the highest-order secondary endpoint was mortality. In order to enrich the anticipated event rate in the trial, patients were additionally required to have been hospitalized for heart failure in the year prior to enrollment but to be receiving stable medical therapy at enrollment and to have no history of hospitalization in the month preceding enrollment. Unlike the prior trials of CRT, patients were assigned for therapy and data were analyzed after they had provided informed consent, not after they had undergone a successful implantation. The design of the COMPANION study is provided in Figure 12-4. The clinical characteristics of COMPANION patients are listed in Table 12-4. Like MIRACLE study patients, and unlike the patients in the CRT-D studies, an equal proportion of patients in the COMPANION study had both ischemic and nonischemic etiologies for LV dysfunction. This was the first trial to enroll patients with advanced heart failure who were medically treated with “triple therapy,” consisting of angiotensin-converting enzyme inhibitors, β-receptor blockers, and aldosterone antagonists. The primary study endpoint was a composite of allcause hospitalization and mortality. The primary secondary endpoint was mortality. A total of 1520 patients were enrolled. Figure 12-5 shows the event-free survival curves for the primary and secondary endpoints and demonstrates the 20% 12-month reductions in death or hospitalization from any cause observed with both CRT and CRT-D devices compared with OPT. It is also noteworthy that the risk of one of these events was 68% in OPT patients, attesting to the severity of heart failure in this population. The CRT device did reduce mortality by 24%, but this reduction was not statistically significant

Concurrent with the MIRACLE trial enrollment, two large-scale trials of CRT-D for patients with heart failure and either primary or secondary indications for an ICD were also enrolling subjects. These were the Multicenter InSync ICD Randomized Clinical Evaluation (MIRACLE ICD) and the CONTAK CD Biventricular Pacing Study. Unlike in the MIRACLE study, the 950 patients randomly assigned to different therapies in the CRT-D studies had primarily ischemic cardiomyopathy (61% to 75%,7,8) and roughly one half of the patients had a secondary indication for the ICD. In patients with NYHA class III to IV status, both studies demonstrated improvements in functional measures of heart failure status. In the CONTAK CD study, patients with NYHA class II status did not experience benefit in all measures of functional status. Neither study showed a difference in the incidence of treated episodes of VT/VF with CRT on or off, indicating a neutral effect of CRT on the arrhythmia substrate early after device implantation. A subsequent analysis of the MIRACLE ICD study data indicated that patients with secondary ICD indications experienced more ICD therapies for VT, whereas those with primary ICD indications had more therapy for VF.60 The incidence of ICD therapy was, as expected, higher in those with secondary prevention indications. In CONTAK CD, the incidence of ICD therapy over the 6-month follow-up interval was 16% for both VT and VF. Improvements in LVEF and ventricular size and dimension and degree of mitral regurgitation were noted in the VIGOR-CHF, MIRACLE, MIRACLE ICD, and CONTAK CD studies, all of which had central core echocardiographic laboratories performing echocardiographic analysis with excellent intraobserver and interobserver variability.19,41 As mentioned earlier, even the patients with NYHA class II benefited from CRT in terms of an echocardiographic response of reverse remodeling.7,14 These CRT-related effects were independent of the use of β-blocker therapy.19 This finding suggests that CRT can exert beneficial effects on the remodeling process across a spectrum of heart failure severity, much like that observed with angiotensin-converting enzyme (ACE) inhibitor therapy.1 Subsequent study of the effects of CRT on ventricular function, volume, and dimension have shown that the beneficial

Comparison of Medical Therapy, Pacing, and Defibrillation on Heart Failure Study

Chapter 12: Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators

nⴝ1520

Follow up

Follow up

Follow up

Follow up

Follow up

Follow up

Follow up

Follow up

393

1

• OPT

Randomization Stratification

Baseline

Implantation

2 ⴙ • OPT • Resynchronization Therapy Implantation

Follow up

2 ⴙ • OPT • Resynchronization Therapy w/ICD backup

⬍5 days

0

⬍2 working days

one week

one month

quarterly

Time Post Randomization Figure 12-4. Study design of the Comparison of Medical Therapy, Pacing, and Defibrillation on Heart Failure (COMPANION) study. ICD, implantable cardioverter-defibrillator; OPT, optimal medical therapy.

(P = .06). The CRT-D device alone reduced mortality by 36% (P = .003) compared with OPT. Close inspection of the mortality curves shows that CRT survival parallels OPT survival until 6 months, when CRT shows benefit. In contrast, the CRT-D and OPT curves separate immediately. These observations may be interpreted as follows: The ICD portion of the CRT-D has an immediate effect to prevent sudden arrhythmic deaths, whereas the reductions in sudden death with CRT alone may be mediated through stabilization of heart failure status, which may be time dependent. Subsequent analysis showed that the reduction in mortality with CRT-D was due to a reduction in sudden cardiac death as adjudicated by an events committee.62 Subgroup analyses were remarkably consistent in demonstrating CRT benefit in all patient subgroups. There appeared to be equal benefit in women and men, in either ischemic or nonischemic etiologies of heart failure, regardless of LVEF greater or less than 20% (0.20) and LV size greater or less than 67 mm. Those patients with longer QRS duration did appear to experience greater benefit with CRT, as did those with LBBB rather than RBBB or IVCD. Figure 12-6 provides the subgroup analysis from the COMPANION study, comparing CRT and CRT-D with OPT therapy for the primary and secondary endpoints.

The COMPANION data expand the role of CRT to achieve the three primary therapeutic goals in treating patients with heart failure—to improve symptoms, to retard disease progression, and to reduce rates of hospitalization and mortality. There are two primary reasons for selecting a CRT-D over CRT alone. The COMPANION data support early sudden death protection with CRT-D, and the majority of patients with CRT indications also have ICD indications in single-chamber primary prevention ICD trials.16,17 In an abstract presentation of the ICD discharge rate in COMPANION, the 12-month appropriate ICD therapy rate was reported to be 15%, suggesting that the risk of development of a potentially fatal, ventricular arrhythmia is significant in the first year after CRT.63 Cardiac Resynchronization–Heart Failure Study The Cardiac Resynchronization–Heart Failure (CAREHF) study, enrolling 813 patients at 82 European centers, compared CRT only with optimal medical therapy.10 Over a mean follow-up of 29 months, CRT resulted in significant reductions in the primary composite endpoint of death or cardiovascular hospitalization. Additionally, reductions in the secondary

394

Section Two: Clinical Concepts

TABLE 12-4.

Clinical Characteristics of the 1520 COMPANION Study Patients* Cardiac Resynchronization Therapy

Characteristic

Optimal Pharmacologic Therapy (N = 308)

Pacemaker (N = 617)

PacemakerDefibrillator (N = 595)

Age (yrs)

68

67

66

Male sex (%)

69

67

67

NYHA class III (%)

82

87

86

Duration of heart failure (yrs)

3.6

3.7

3.5

LV ejection fraction

0.22

0.20

0.22

LV end-diastolic dimension (mm)

67

68

67

Heart rate (bpm)

72

72

72

Blood pressure (mm Hg): Systolic Diastolic

112 64

110 68

112 68

Distance walked in 6 min (m)

244

274

258

QRS interval (msec)

158

160

160

Ischemic cardiomyopathy (%)

59

54

55

Diabetes (%)

45

39

41

Bundle branch block (%): Left Right

70 9

69 12

73 10

69 89

70 89

69 90

66 94 55

68 94 53

68 97 55

Pharmacologic therapy (%): ACE inhibitor† ACE inhibitor or angiotensinreceptor blocker† β-blocker Loop diuretic Spironolactone

*Median values are given for continuous measures. There were no significant differences among the groups. † Patients who could not tolerate an ACE inhibitor received an angiotensin-receptor blocker. ACE, angiotensin-converting enzyme; COMPANION, Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure trial; LV, left ventricular; NYHA class, New York Heart Association functional class. Adapted from Bristow MR, Saxon LA, Boehmer J, et al: Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators: Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med 350:2140-2150, 2004.

endpoint of mortality were achieved. Figure 12-7 illustrates the Kaplan-Meier curves for these endpoints. Comparing CARE-HF with COMPANION, we observe that the COMPANION patients appear to have been a somewhat sicker group, perhaps owing to the requirement for a heart failure hospitalization in the year prior to enrollment.9 The 12-month mortality rate in the medical therapy group in CARE-HF was 12.6%, versus 19% in COMPANION. Only 38% of the patients in CARE-HF had coronary artery disease, compared with 56% in COMPANION; the mean LV ejection fraction was 25% in CARE-HF, compared with 21% in COMPANION; and a higher percentage of patients in COMPANION had NYHA class IV status (16% vs. 6.5% in CARE-HF). Another issue is the positive effect of CRT on mortality in CARE-HF. The relative risk reduction with CRT in CARE-HF patients was equivalent to that

of the COMPANION CRT-D patients (36%). The CRT group in the COMPANION study did show a 24% reduction in mortality, but this was not statistically significant (P = .06). These differences may be due to the shorter duration of follow-up in the COMPANION study, meaning that if follow-up had been longer in COMPANION, the value would have reached significance. Another important question is the risk of sudden death in the CARE-HF patients given CRT. Although only 7% of deaths in CRT patients in CARE-HF were adjudicated as sudden, these accounted for 37% of all deaths, and COMPANION data indicate such deaths may have been prevented with a CRT-D device.62 Several other important issues were answered by CARE-HF. The advantages of CRT over best medical management appear to increase over time, at least to 18 months. Before this trial, follow-up past 12 months was

Chapter 12: Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators

A

B

Primary Endpoint

Secondary Endpoint 100

Pacemaker (414 events, P⫽0.014) Pacemakerdefibrillator (390 events, P⫽0.010)

80 60 40

Event-free Survival (%)

Event-free Survival (%)

100

Pharmacologic therapy (216 events)

20 0 0

Pacemaker-defibrillator (105 events, P⫽0.003)

90 80 Pharmacologic therapy (77 events)

70 60

Pacemaker (131 events, P⫽0.059)

50

120 240 360 480 600 720 840 960 1080 Days after Randomization

No. at Risk Pharmacologic 308 176 115 72 46 24 therapy Pacemaker 617 384 294 228 146 73 Pacemaker595 385 283 217 128 61 defibrillator

16

6

1

36 25

14 8

3 0

0

90 180 270 360 450 540 630 720 810 900 990 1080

Days after Randomization No. at Risk Pharmacologic 308 284 255 217 186 141 94 57 45 25 4 2 therapy 617 579 520 488 439 355 251 164 104 60 25 5 Pacemaker 595 555 517 470 420 331 219 148 95 47 21 1 Pacemakerdefibrillator

C

D

Death from or Hospitalization for Cardiovascular Causes

Death from or Hospitalization for Heart Failure 100

80 Pacemaker-defibrillator (312 events, P⬍0.001)

60 40

Pharmacologic therapy (188 events)

20

Pacemaker (338 events, P⫽0.002)

0

Event-free Survival (%)

100 Event-free Survival (%)

395

Pacemaker-defibrillator (212 events, P⬍0.001)

80 60 Pharmacologic therapy (145 events)

40 20

Pacemaker (237 events, P⫽0.002)

0 0

120 240 360 480 600 720 840 960 1080 Days after Randomization

No. at Risk Pharmacologic 308 199 134 91 56 29 therapy Pacemaker 617 431 349 282 194 102 Pacemaker595 425 341 274 167 89 defibrillator

20

8

2

51 45

22 20

5 3

0

120 240 360 480 600 720 840 960 1080 Days after Randomization

No. at Risk Pharmacologic 308 216 161 118 76 39 therapy Pacemaker 617 498 422 355 258 142 Pacemaker595 497 411 343 228 131 defibrillator

28

11

2

75 71

35 27

9 5

Figure 12-5. Kaplan-Meier estimates of the time to the primary endpoint of death from or hospitalization for any cause (A), the time to the secondary endpoint of death from any cause (B), the time to death from or hospitalization for cardiovascular causes (C), and the time to death from or hospitalization for heart failure (D) in the Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) study. In A, The 12-month rates of death from or hospitalization for any cause—the primary endpoint—were 68% in the pharmacologic therapy group, 56% in the group that received a pacemaker as part of cardiac resynchronization therapy, and 56% in the group that received a pacemaker-defibrillator as part of cardiac resynchronization therapy. In B, the 12-month rates of death from any cause—the secondary endpoint—were 19% in the pharmacologic therapy group, 15% in the pacemaker group, and 12% in the pacemaker-defibrillator group. In Panel C, the 12-month rates of death from or hospitalization for cardiovascular causes were 60% in the pharmacologic therapy group, 45% in the pacemaker group, and 44% in the pacemakerdefibrillator group. In Panel D, the 12-month rates of death from or hospitalization for heart failure were 45% in the pharmacologic therapy group, 31% in the pacemaker group, and 29% in the pacemaker-defibrillator group. In the pharmacologic therapy group, death from heart failure made up 24% of the events, hospitalization for heart failure 72% of events, and the intravenous administration of inotropes or vasoactive drugs for more than 4 hours 4% of events. P values are for comparison with optimal pharmacologic therapy. (From Bristow MR, Saxon LA, Boehmer J, et al: Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators: Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med 20;350:2140-2150, 2004.)

396

Section Two: Clinical Concepts

Variable

Hazard Ratio for Death from or Hospitalization for Any Cause

No. of Patients Pharmacologic Pacetherapy maker (n⫽308) (n⫽617)

Age ⱕ65 yr ⬎65 yr Sex Male Female Cardiomyopathy Ischemic Nonischemic NYHA class III IV LVEF ⱕ20% ⬎20% LVEDD ⱕ67 mm ⬎67 mm QRS interval width ⱕ147 msec 148–168 msec ⬎168 Bundle branch block Left Other Heart ratio ⱕ72 beats/min ⬎72 beats/min Systolic BP ⱕ112 mm Hg ⬎112 mm Hg Diastolic BP ⱕ68 mm Hg ⬎68 mm Hg ACE inhibitor use No Yes β-blocker use No Yes Loop diuretic use No Yes Spironolactone use No Yes

Hazard Ratio for Death from or Hospitalization for Any Cause

Hazard Ratio for Death from Any Cause

Pacemakerdefibrillator (n⫽595)

123 185

272 345

272 323

211 97

415 202

401 194

181 127

331 285

325 270

253 55

537 80

512 83

143 165

324 293

282 313

133 122

257 266

248 237

115 111 82

209 203 205

178 232 185

215 93

426 190

434 161

161 147

318 299

315 280

164 144

347 270

307 288

178 130

328 289

316 279

96 212

186 431

183 412

104 204

196 421

193 402

17 291

36 581

20 575

139 169

288 329

267 328 0.0

0.5 Pacemaker Better

1.0

1.5 0.0

Pharmacologic Therapy Better

0.5

1.0

1.5 0.0

Pacemaker- Pharmacologic Defibrillator Therapy Better Better

0.5

1.0

PacemakerDefibrillator Better

1.5

2.0

Pharmacologic Therapy Better

Figure 12-6. Hazard ratios and 95% confidence intervals for the primary endpoint, death from or hospitalization for any cause, and the secondary endpoint, death from any cause, according to the baseline characteristics of the patients in the Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) study. Echocardiographically determined values for left ventricular end-diastolic dimension (LVEDD) were not available for all patients. ACE, angiotensinconverting enzyme; BP, blood pressure; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association functional class. (From Bristow MR, Saxon LA, Boehmer J, et al: Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators: Cardiacresynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med 350:2140-2150, 2004.)

2.5

Chapter 12: Clinical Trials of Cardiac Resynchronization Therapy: Pacemakers and Defibrillators

A

100

Percentage of Patients Free of Death from Any Cause or Unplanned Hospitalization for a Major Cardiovascular Event

Figure 12-7. Kaplan-Meier estimates of the time to the primary endpoint (A) and the principal secondary outcome (B) in the Cardiac Resynchronization–Heart Failure (CARE-HF) study. The primary outcome was death from any cause or an unplanned hospitalization for a major cardiovascular event. The principal secondary outcome was death from any cause. (From Cleland JG, Daubert JC, Erdmann E, et al: Cardiac ResynchronizationHeart Failure [CARE-HF] Study Investigators: The effect of cardiac resynchronization on morbidity and mortality in heart failure. N Engl J Med 352:15391549, 2005.)

397

75 Cardiac resynchronization 50

Medical therapy

25

P⬍0.001 0 0

500

1500

1000 Days

No. at Risk Cardiac resynchronization Medical therapy

B

409

323

273

166

68

7

404

292

232

118

48

3

100

Percentage of Patients Free of Death from Any Cause

Cardiac resynchronization 75

Medical therapy

50

25

P⬍0.002 0 0

500

1500

1000 Days

No. at Risk Cardiac resynchronization Medical therapy

409

376

351

213

89

8

404

365

321

192

71

5

available for too few patients undergoing CRT. This is the first large trial to demonstrate benefit with respect to biomarker measurements as a surrogate of congestive heart failure severity with CRT. A nonsignificant decrease in Nterminal pro-brain natriuretic peptide (NT-BNP) was noted at 3 months, but by 18 months, there was a dramatic decrease of more than 1100 pg/mL (P = .0016). Finally, CARE-HF was the first large clinical trial to select patients on the basis of either a wide QRS >149 msec or a wide QRS >120 msec and the presence of additional dyssynchrony markers. It is likely that the selection of

patients for CRT using measures of dyssynchrony served to better identify candidates for CRT and improve the response rate and outcome for patients in this trial.

Procedural Safety and LV lead Performance In all of the clinical trials subsequent to the PATH I study and in patients enrolled early in the CONTAK CD study,

398

Section Two: Clinical Concepts

LV lead placement was performed through a coronary sinus branch vein with an over-the-wire unipolar lead. In general, lateral LV wall sites were chosen on the basis of results of trials of short-term implants demonstrating that the most robust hemodynamic response is located at this site in patients with LBBB.48 Figures 12-8 and 12-9 illustrate radiographs of CRT systems implanted long-term using two over-the-wire LV leads from two different manufacturers with a CRT and a CRT-D device. In both, the LV lead is in a lateral branch vein. In Figure 12-9, two RV defibrillating leads are shown because there was an insulation break in the long-term RV lead and a new one was placed at the time of CRT implantation. Despite initial concerns about the additional risk of LV branch vein lead placement, particularly given the additional skills required for the procedure and extensive anatomic remodeling that occurs in the setting of advanced heart failure, which renders access to the coronary sinus and its branch veins more challenging, the procedural safety and LV lead performance have been very good. Table 12-5 lists the procedural safety and LV lead performance data observed in the longterm CRT studies. A learning curve does appear to be involved in CRT implantation, and the CONTAK CD study found the threshold for attaining a higher than 95% rate of successful LV implantation was 15 cases.64 The most common complication of LV lead placement is coronary sinus trauma, which occurs at a rate of 2% to 4%; and the incidence of true perforation resulting in cardiac tamponade is less than 1%. LV lead dislodgement requiring repositioning occurs in 1% to 3% of implants. Diaphragmatic stimulation requiring reoperation for lead repositioning occurs in 1% to 4% of cases, but this risk has been lowered with the introduction of bipolar LV leads.

LV Epi RA

RV

LAO

RA

LV Epi

Trials of Long-Term Cardiac Resynchronization Therapy in Special Populations RV

The Post AV Nodal Ablation Evaluation Study The Post AV Nodal Ablation Evaluation (PAVE) study compared CRT with RV pacing in patients with permanent AF who were undergoing AV nodal ablation.55 Entry criteria for the study did not require patients to have symptomatic heart failure or systolic dysfunction. However, patients were required to have exercise limitation, defined as the inability to walk farther than 450 meters during a 6MWD test, which was also the primary study endpoint. Peak Vo2 and QOL were assessed as secondary endpoints. A total of 252 patients were randomly assigned to either CRT or RV pacing; 205 patients (mean age 69 years, 64% male, 51% NYHA class II, 32% NYHA class III, 46% with LVEF 6 months) efficacy of the algorithms. Future studies also need to demonstrate that any incremental reductions in AF burden are associated with a clinically relevant outcome, such as better QOL, improved functional capacity, or reduction in health care utilization. Pacing Algorithms for Termination of Atrial Fibrillation Episodes of atrial tachycardia (AT) and atrial flutter occur commonly in patients with AF, and AT or atrial flutter frequently transition between episodes of AF.57,63,98-100 Figure 13-10 shows an example of AF that organizes into atrial flutter and is then effectively terminated by atrial ATP. Atrial ATP therapy has been incorporated into some pacemakers and implantable

6

Overdrive Pacing OFF Overdrive Pacing ON 4.44%

4 3.19% 2.63% 1.93%

2

Mode-Switch Duration (hr)

1.73%

1.37%

0 1 Month

Figure 13-9. Time spent in symptomatic atrial fibrillation (AF) (top) and AF burden estimated from the mode-switch duration counters retrieved by pacemaker interrogation (bottom) at 1-, 3-, and 6-month follow-ups in 399 patients randomly allocated to DDDR pacing with dynamic atrial overdrive (DAO) programmed “on” (yellow) or “off” (blue). (Data from Carlson MD, Ip J, Messenger J, et al: Atrial Dynamic Overdrive Pacing Trial [ADOPT] Investigators: A new pacemaker algorithm for the treatment of atrial fibrillation: Results of the Atrial Dynamic Overdrive Pacing Trial [ADOPT]. J Am Coll Cardiol 42:627633, 2003.)

417

3 Months

6 Months

3 Months

6 Months

1800 1600 1400 1200 1000 800 600 400 200 0 1 Month

Time Following Randomization

418

Section Two: Clinical Concepts

cations to atrial prevention and termination therapies “on” (n = 199) or “off” (n = 206). Patients were monitored for 7 months. The mean AT/AF burden was 4.3 ± 20.0 hr/month in patients with AT/AF prevention and termination therapies programmed “on,” compared with 9.0 ± 50.0 hr/month in the group with these therapies programmed “off” (P = .11). The discrepancy between the reported high atrial ATP efficacy for termination of AT and the failure to demonstrate a significant reduction in AT/AF burden may be due to several factors.53 Device-classified efficacy may be exaggerated because of spontaneous termination of many episodes of AT.64 Indeed, in ATTEST102 and in a trial reported by Gillis and colleagues,64 which evaluated the GEM III AT (Medtronic, Inc.), approximately one half of the episodes that the pacemakers classified as atrial tachyarrhythmias were less than 10 minutes in duration. By design, these devices define ATP efficacy if sinus or atrial paced rhythm occurs before redetection of atrial tachyarrhythmia.64 Although the redetection time is usually less than 1 minute, these devices may allow up to 3 minutes from the last-delivered ATP therapy for rede-

defibrillators. A number of clinical studies have reported the efficacy of atrial ATP for termination of atrial tachycardia and atrial flutter to range from 30% to 54%.100-105 In select individuals, atrial ATP therapy has been reported to reduce atrial tachyarrhythmia burden over time.57,103,106 The hypothesis that successful pace termination of atrial tachycardia or atrial flutter would prevent the development of AF over time was tested in the Atrial Therapy Efficacy and Safety Trial (ATTEST).102 In a parallel study design, 370 patients received a Medtronic AT 500 (Medtronic, Inc., Minneapolis, Minn.) and were randomly assigned to either DDDR pacing or DDDR pacing with atrial ATP therapies and atrial pace prevention therapies programmed “on.” Over a 3month follow-up period, 15,000 episodes of an atrial tachyarrhythmia were treated by atrial ATP therapies, and the device-classified efficacy was 41%. However, AF frequency and AF burden (Fig. 13-11) were not reduced by the delivery of prevention therapies or atrial ATP therapies. Friedman and coworkers107 randomly allocated 405 patients with a history of AF who had received an implantable cardioverter-defibrillator for standard indi-

7 5 0

A P

V P

7 5 0

F D

V P

2 1 0

F D

2 1 0

8 6 0

F D

2 2 0

F D

A P

V P

7 5 0

2 2 0

V P

F D

7 5 0

2 2 0

F D

8 6 0

2 1 0

F D

2 2 0

7 5 0

A P

F D

2 3 0

V P

F D

A P

V P

7 5 0

2 1 0

F D

8 6 0

2 1 0

F D

V P

7 5 0

2 1 0

7 5 0

F D

2 1 0

V P

A S

1 9 0

A P

7 5 0

1 1 1 2 8 7 6 8 0 0 0 0 A A A A P P P P

1170

4 5 0

2 1 1 1 1 1 1 1 1 1 1 1 0 7 7 6 5 5 7 6 6 5 3 4 0 0 0 0 0 0 0 0 0 0 0 0 A F F F F F F F F F F F F S S S S S S S S S S S S S

V P

7 5 0

2 0 0

1 1 8 8 0 0 F F F S S S

T D

V S

7 5 0

V P

8 4 0

V S

V P

7 5 0

8 6 0

A P

8 2 0

V P

7 5 0

V P

A P

8 6 0

First Rx Figure 13-10. An example of atrial fibrillation (AF) organizing into atrial flutter. The atrial electrogram (EGM) and the annotated markers indicating how the pacemaker classifies each atrial and ventricular event as well as the cycle length (in msec) of each interval are displayed in each panel. The initial rhythm is AF (upper panel), which subsequently organizes into atrial flutter (cycle length 210 msec). The atrial flutter is terminated by a ramp atrial antitachycardia pacing (ATP) therapy that restores atrial paced rhythm. The marker channel notations indicate how the device classifies each beat. Interbeat intervals are also shown (in msec). AP, atrial paced event; VP ventricular paced event; AR, atrial event sensed in atrial refractory period; FS, AF sensed event; TD, tachycardia detected; TS, tachycardia sensed event; Rx, pharmaceutical therapy.

V P

Chapter 13: Pacing for Sinus Node Disease: Indications, Techniques, and Clinical Trials 5 p⫽0.65

1.0 3

2 0.5 1

Atrial Therapies ON

Symptoms of Pacemaker

Syndrome 4

0

TABLE 13-5.

p⫽0.20 AF Burden (median hr/mo)

AF Frequency (median episodes/mo)

1.5

419

0 Atrial Therapies OFF

Figure 13-11. Left, Atrial fibrillation (AF) frequency in patients randomly allocated to atrial antitachycardia pacing (ATP) therapy and three AF pace prevention therapies “on” or “off” in the Atrial Therapy Efficacy and Safety Trial (ATTEST). Patients were followed for 3 months. Right, AF burden over 3 months of follow-up in the two treatment groups. Median data are shown. (Data from Lee MA, Weachter R, Pollak S, et al: ATTEST Investigators: The effect of atrial pacing therapies on atrial tachyarrhythmia burden and frequency: Results of a randomized trial in patients with bradycardia and atrial tachyarrhythmias. J Am Coll Cardiol 41:1926-1932, 2003.)

tection to occur. Gillis and colleagues64 showed that if a more conservative definition of efficacy—termination of AT or AF within 20 seconds of delivery of atrial ATP therapy—is used, atrial ATP efficacy is lower than previously reported; these researchers found that atrial ATP terminated only 26% of all atrial tachyarrhythmias and 32% of AT episodes. Furthermore, the incorporation of 500-Hz burst pacing algorithms into atrial defibrillators has not been shown to terminate AF.64 In the studies evaluating atrial ATP efficacy for prevention of AF, not all patients received maintenance class I/III antiarrhythmic drug therapy, which may be important in facilitating ATP therapy and preventing early recurrence of atrial tachyarrhythmia.108,109 Some episodes of AT are not reentrant in mechanism, and some episodes classified as AT may have been AF, which cannot be pace terminated.110 Certain patients do benefit from atrial ATP therapy for prevention of AF (see Fig. 13-8).57,106 Patients with high atrial ATP efficacy for termination of AT (>60% of all treated episodes effectively terminated) experience a significant reduction in AF burden.111,112 As many as 30% of patients with SND and paroxysmal AF may benefit from atrial ATP therapy. Also, more aggressive programming of atrial ATP therapy, rather than use of the nominal values in these devices, may improve the efficacy of therapy.113 Pacing and Pacemaker Syndrome in Sinus Node Disease The pacemaker syndrome consists of a constellation of signs and symptoms that occur due to the loss of AV synchrony during ventricular pacing (Table 13-5).2,47 The definition and diagnostic criteria of pacemaker syndrome have varied, but symptoms include fatigue, dyspnea on

Severe

Syncope Presyncope

Moderate

Dizziness Dyspnea Chest pain Jaw pain Confusion

Mild

Venous pulsation in neck Fatigue Weakness Palpitations Fullness in chest

exertion, paroxysmal nocturnal dyspnea, orthopnea, orthostatic hypotension, and syncope. In MOST, pacemaker syndrome was prospectively defined as either (1) new or worsened dyspnea, orthopnea, elevated jugular venous pressure, rales, and edema with ventriculoatrial conduction during ventricular pacing or (2) symptoms of dizziness, weakness, presyncope or syncope, and a reduction in systolic blood pressure of more than 20 mm Hg during VVIR pacing in comparison with atrial pacing or sinus rhythm.43,114 The incidence of pacemaker syndrome was 13.8% at 6 months, 16.0% at 1 year, 17.7% at 2 years, 19.0% at 3 years, and 19.7% at 4 years. Univariate predictors of pacemaker syndrome were a higher percentage of ventricular paced beats, a higher programmed lower pacemaker rate, and a slower underlying sinus heart rate. However, only a higher percentage of ventricular pacing was an independent predictor of developing pacemaker syndrome. QOL, measured with a variety of metrics, diminished in association with the diagnosis of pacemaker syndrome and improved after the pacemaker was reprogrammed to a physiologic mode.114 Of the 204 patients randomly allocated to VVIR pacing mode in the PASE trial, 26% crossed over to DDDR mode because of intolerance to ventricular pacing.40 The incidence of pacemaker syndrome was reported to be 2% in the Danish study39 and 2.7% at 3 years in CTOPP.41 The incidence of pacemaker syndrome was likely underestimated in these last two trials, because treatment would have required a surgical intervention. It is possible that pacemaker syndrome was overestimated in MOST and the PASE trial, because it was easy to cross patients over to the DDDR pacing mode.47 Pacing and Congestive Heart Failure in Sinus Node Disease In the Danish study, New York Heart Association functional class and diuretic use were significantly higher during follow-up in the ventricular pacing group than in the atrial pacing group.39 In CTOPP, the annual incidences of hospitalization for CHF were similar in the ventricular pacing group (3.5%) and the atrial pacing group (3.1% relative risk reduction 7.9%; 95% CI, –18.5% to 28.3%; P = .52).41 In MOST, the annual inci-

420

Section Two: Clinical Concepts

dences of hospitalization for heart failure were similar in patients receiving dual-chamber pacing (3.7%) and those receiving ventricular pacing (4.4%; hazard ratio, 0.82; 95% CI, 0.63 to 1.06; P = .13).43 During follow-up, patients receiving dual-chamber pacing had a lower heart failure score than those receiving ventricular pacing (average points per visit during follow-up: ventricular pacing, 1.75; dual-chamber pacing, 1.49; P < .001). Pacing and Quality of Life in Sinus Node Disease Dual-chamber pacing has been reported to be associated with better physiologic parameters, including cardiac output, exercise capacity, and exercise oxygen consumption, than ventricular pacing.47 However, none of the large clinical trials conducted to date has demonstrated substantial improvements in QOL measures associated with atrial and ventricular pacing.115,116 The PASE trial investigators did not show higher values for measures of QOL in patients treated with dual-chamber pacing than in those treated with ventricular pacing, although they did report a modest improvement in some scales of the Medical Outcomes Study 36-Item Short Form Health Survey (SF-36) at 3 months of follow-up in the subgroup with SND.40 In MOST, dual-chamber pacing resulted in a small but measurably higher QOL than ventricular pacing.43 In CTOPP, QOL improved after pacemaker implantation in both atrial pacing and ventricular pacing groups, but no significant healthrelated QOL difference was observed between the two groups.115 Thus, atrial pacing appears to confer some improvements in QOL in patients with SND, particularly the group at risk for pacemaker syndrome.114

Treatment of Sinus Node Disease Pacing Modalities in Sinus Node Disease The indications for pacing in the setting of SND are shown in Table 13-6. Although none of the large ran-

TABLE 13-6.

domized clinical trials has shown a survival benefit of atrial pacing, dual-chamber pacemakers are commonly implanted in North America.1 Atrial pacing should be considered in the setting of SND for prevention of AF and pacemaker syndrome.47,116,117 There is some controversy about the American College of Cardiology/ American Heart Association/North American Society for Pacing and Electrophysiology (ACC/AHA/NASPE) classification of these recommendations. Some experts have assigned a class I indication to the selection of dual-chamber pacing over ventricular pacing for prevention of AF in patients with SND (level of evidence A).47 In contrast, this recommendation was assigned a class IIa recommendation by a Canadian Cardiovascular Society–sponsored Consensus Conference on Atrial Fibrillation.117 This lower class was assigned on the basis of the absence of data showing that prevention of AF in this population is associated with significant clinical benefit, such as reduction in rates of stroke or mortality.118 Certainly, on the basis of all the available data from multiple clinical studies, an AAIR system should be considered if the patient has intact AV conduction.39,75,76 A cost-effectiveness analysis performed in the MOST population has reported that during the first 4 years of the trial, dual-chamber pacemakers increased quality-adjusted life expectancy by 0.013 year per subject at an incremental cost-effectiveness ratio of $53,000 per quality-adjusted year of life gained.80 This cost could be further reduced by increasing the use of AAIR pacing over that of DDDR pacing in select patients. AAIR Pacing As many as 20% of patients with symptomatic SND are potential candidates for AAIR pacing systems.1,39,54 Because of the high incidence of chronotropic incompetence in patients with SND, a rate-adaptive pulse generator should be considered.47 Although the most economical approach to providing atrial pacing in this population, this modality is used for less than 1% of implants in North America.1 Concerns about progression of AV block and the development of chronic AF

Indications for Pacing in Sinus Node Disease

Class I

General consensus that pacing is indicated

Sinus node dysfunction with documented symptomatic bradycardia Symptomatic chronotropic incompetence

Class II: IIa IIb

Divergence of opinion on need for pacing

Sinus node dysfunction with heart rate 0.30 sec) on conducted beats during 2:1 AV block with a narrow QRS complex is suggestive of an AV nodal site of block, whereas a normal P-R interval favors intra-Hisian block. In general, the response of the block, particularly 2:1 AV block, to pharmacologic agents may help to determine the site of the block. Atropine generally

1 2 3 V1 HRA A HBE T

A

650

A 240H 40

v

A

660

A 230H

740

A

v

240H

40

v

A

780

A

A A

v 200H

300 msec

40

improves AV conduction in patients with AV nodal block; however, atropine is expected to worsen conduction in patients with block localized to the HisPurkinje system, owing to its effect on increasing sinus rates without improving His-Purkinje conduction (Fig. 14-8). Carotid sinus stimulation is expected to worsen block localized to the AV node, whereas it either has no effect or improves conduction in patients with HisPurkinje system disease by causing sinus node slowing. The effect of any given drug, however, may be difficult to predict because its effect on the sinus node may be greater than its effect on the AV node. For example, atropine may improve AV node conduction, but if

Lead II (continuous)

40

Figure 14-3. Example of atypical or uncommon type I seconddegree atrioventricular (AV) block in the AV node. Surface leads I (1), II (2), III (3), and V1 are displayed with intracardiac electrograms recorded from the high right atrium (HRA), His bundle (HBE), and right ventricular apex (RV). There is little alteration in the A-H interval before the fourth atrial complex not conducting to the ventricle. The true nature of this arrhythmia is revealed by the first-conducted P wave (A, atrial electrogram) after the pause, which is associated with substantial shortening of the A-H interval from 230 to 240 msec to 200 msec. (From Josephson ME: Clinical Cardiac Electrophysiology: Techniques and Interpretations, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2002, pp 92-109.)

0.19

0.19

0.74

0.20

0.97

.18

.18

.18

0.84

Figure 14-4. Example of type II second-degree atrioventricular (AV) block showing repeating episodes of block. Note that the P-R interval is constant both before and after the nonconducted P wave and that there is associated bundle branch block.

0.78

0.20

0.80

0.20

1.04

0.83

0.20

0.92

0.20

1.06

0.20

0.85

0.20

1.02

0.28

0.96

0.97

0.84

1550

HBE T

A

785

A

785 A V H 270

0.21

0.94

0.88

V1 A

0.94

1.00

0.18

3

775 V H 290

0.94

0.20

2

HRA A A

0.20

Figure 14-5. Rhythm strip demonstrating vagally mediated second-degree atrioventricular (AV) block. Note the progressive slowing of the heart rate before the first episode of block, although no change in P-R interval can be discerned. During the second episode of block, both sinus slowing and P-R interval prolongation occur, confirming the presence of type I second-degree AV block.

1540

1

0.50

0.20

0.74

0.20

0.16 0.19

.18

0.20

A A

765

A

785 A V H 270

A A

500 msec

Figure 14-6. Spontaneous 2:1 (high-grade) atrioventricular (AV) block localized to the AV node. Surface leads I (1), II (2), III (3), and V1 are displayed with intracardiac electrograms recorded from the high right atrium (HRA), His bundle (HBE), and right ventricular apex (RV). Alternate atrial depolarizations (A) are not followed by either a His bundle or a ventricular depolarization. On the basis of a surface electrocardiogram, the finding of 2:1 AV block with a narrow QRS is compatible with a block at either the AV node or an infra-Hisian bundle site. The intracardiac recordings localize the site of block to the AV node. (From Josephson ME: Clinical Cardiac Electrophysiology: Techniques and Interpretations, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2002, pp 92-109.)

Chapter 14: Pacing for Atrioventricular Conduction System Disease

433

Control

4021 I II

I

III V1 HBE

II A H

AH

A H

AH

A H

AH

III V1

RV HRA A Figure 14-7. Intracardiac tracing of 2:1 second-degree atrioventricular (AV) block located in the His-Purkinje system. Sinus rhythm with left bundle branch block is present. Surface leads I, II, III, and V1 are displayed with intracardiac electrograms recorded from the high right atrium (HRA), His bundle (HBE), and right ventricular apex (RV). The A-H intervals are constant, but every other atrial complex fails to activate the ventricle even though each atrial depolarization is followed by a His bundle deflection. This finding shows that the site of AV block is within the His-Purkinje system. (From Josephson ME: Clinical Cardiac Electrophysiology: Techniques and Interpretations, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2002, pp 92-109.)

H

HBE

1175 AH 100 HV 70

A H

A Atropine 1 mg

IV

V1

HRA A

H

V

770 AH 60 HV 80

A

H

V

HBE

B Atropine 1.5 mg IV

atropine causes excessive sinus node acceleration, AV conduction may improve marginally or not at all. The response to infusion of isoproterenol is less clear. Isoproterenol may improve conduction disorders localized in the AV node as well as occasionally in the HisPurkinje system. The diagnosis of complete heart block (CHB) rests on demonstration of complete dissociation between atrial and ventricular activation. Care must be taken to distinguish transient AV dissociation due to competing atrial and junctional or ventricular rhythms with similar rates (so-called isorhythmic AV dissociation). If sufficiently long monitoring strips are available, intermittent conduction of appropriately timed atrial events is seen. Temporary atrial pacing can be performed to accelerate the atrial rate to overdrive the competing junctional or ventricular arrhythmia, demonstrating intact AV conduction. In the presence of atrial fibrillation (AF), CHB can be inferred when the ventricular rate becomes regular rather than the typical irregular ventricular response (Fig. 14-9). Digoxin toxicity may be the cause of heart block with AF, and this and other drug toxicity should be ruled out before one assumes that structural AV conduction disease is present. In patients with AF, regular R-R interval may on occasion be due to “concealed” sinus rhythm and not heart block.17 The escape rhythm in CHB may be generated by the AV junction, His bundle, bundle branches, or distal conduction system. Rarely, the underlying rhythm arises from the ventricular myocardium or, for all practical purposes, is absent. The site of AV block is important, in that it determines to a great extent the rate and reliability of the underlying escape rhythm. The site of origin of the escape rhythm in cases of advanced AV block is more important than the escape rate itself.10,18 For example, in heart block associated with inferior AMI or congenital CHB, the escape rhythm is usually generated by the AV junction, and permanent pacing

V1 A HBE

C

RV

H

AA 550 AH 75 A H

AH 75 HV 75 AH

1100

Figure 14-8. His-Purkinje atrioventricular block after an atropine-induced increase in sinus rate. Surface leads I, II, III, and V1 are displayed with intracardiac electrograms recorded from the high right atrium (HRA), His bundle (HBE), and right ventricular apex (RV). A, Sinus rhythm at a cycle length of 1175 msec with 1:1 AV conduction. Left bundle branch block is present, and the H-V interval is slightly prolonged, to 70 msec. B, After injection of 1 mg of atropine, the sinus rate speeds up to 770 msec and the H-V interval increases to 80 msec, but 1:1 AV conduction is still present. The A-H interval shortens despite the faster sinus rate, because of the direct effect of atropine on AV nodal conduction. C, After injection of 1.5 mg of atropine, the sinus cycle length decreases further to 550 msec, and 2:1 AV block occurs below the His bundle. Atropine worsens AV conduction in this patient, not through a direct drug effect but because the improvement in AV nodal conduction caused by atropine stresses the already abnormal His-Purkinje system. (From Miles WM, Klein LS: Sinus nodal dysfunction and atrioventricular conduction disturbances. In Naccarelli GV: Cardiac Arrhythmias: A Practical Approach. Mt Kisco, NY, Futura, 1991, pp 243-282.)

may not be required. Nevertheless, it is worth emphasizing that symptomatic AV block requires pacing regardless of the site, morphology, or rate of the escape rhythm. Trifascicular block is present when bifascicular block is associated with HV prolongation. Trifascicular block, however, is often applied loosely to the electrocardiographic patterns of bifascicular block (RBB block [RBBB] plus left anterior hemiblock, RBBB plus left posterior hemiblock, or LBB block [LBBB]) plus firstdegree AV block. The use of the term trifascicular block to describe these AV conduction disturbances on the ECG is misleading, because the site of block in such

434

Section Two: Clinical Concepts I

aVR

V1

V4

I II

II

aVL

V2

V5

III

aVF

V3

V6

III V1 HRA Mp Md

Figure 14-9. Twelve-lead electrocardiogram from a patient with recent aortic valve surgery and atrial fibrillation treated with digoxin. Surface leads I, II, III, and V1 are displayed with intracardiac electrograms recorded from the high right atrium (HRA), His bundle (HBE), and right ventricular apex (RV). The QRS complexes are narrow and occur at a regular rate of 56 bpm, illustrating complete heart block with a junctional escape rhythm in the setting of atrial fibrillation. Despite discontinuation of digoxin, complete heart block persisted in this patient.

cases may be located in the AV node or His-Purkinje system.12 The P-R interval does not identify those patients who have prolonged H-V intervals in such cases. Up to 50% of patients with bifascicular block and prolonged P-R intervals have prolongation of the A-H interval (i.e., AV nodal conduction time).19 Trifascicular block should be used only to refer to alternating RBBB and LBBB, RBBB with a prolonged H-V interval (regardless of the presence or absence of left anterior or posterior fascicular block), and LBBB with a prolonged H-V interval. In addition, the term can be used in a patient with second- or third-degree AV block in the His-Purkinje system with permanent block in all three fascicles, permanent block in two fascicles with intermittent conduction in the third, permanent block in one fascicle with intermittent block in the other two fascicles, or intermittent block in all three fascicles. Thus, according to its strict definition, when one is interpreting an ECG in the absence of a His bundle recording, trifascicular block should be applied only to the patterns of alternating RBBB and LBBB or RBBB with intermittent left anterior and posterior hemiblocks. These situations are class I indications for permanent pacing even in asymptomatic individuals. Electrophysiologic Study Invasive electrophysiologic study (e.g., His bundle recording) is a useful means of evaluating AV conduction in patients who have symptoms and in whom the need for permanent pacing is not obvious (Fig. 14-10). Electrophysiologic studies (strictly for evaluation of the conduction system and site of block) are not required in patients with symptomatic high-grade or complete AV block recorded on surface ECG tracings, ambulatory Holter monitoring, or transtelephonic recordings. The need for permanent pacing has already been established in these patients (class I indication). However electrophysiologic studies may be indicated in patients with

RV Figure 14-10. Surface leads I, II, III, and V1 reveal sinus rhythm with 2:1 AV conduction, marked P-R interval prolongation during conducted beats, and right bundle branch block. Intracardiac electrograms from the high right atrium (HRA), proximal and distal poles of the His bundle catheter (Mp and Md, respectively), and the right ventricle (RV) are shown. The distal pole of the His bundle catheter registers a His bundle potential, which can be seen as a discrete sharp potential between the atrial and the ventricular electrograms on that channel. During nonconducted atrial complexes, spontaneous infra-Hisian block is apparent because the His bundle potential (arrow) is not followed by a ventricular electrogram.

high-grade AV block if another arrhythmia is suspected or is a likely cause of symptoms. For example, even if high-grade AV block is documented on spontaneous recordings, ventricular tachycardia may still be the cause of syncope in patients who have suffered extensive MI. In patients with alternating BBB, electrophysiologic testing almost invariably demonstrates a high degree of His-Purkinje system disease. These patients typically have very long H-V intervals and are at very high risk of progression to CHB in a short time. Pacing in these patients is indicated on clinical grounds, and electrophysiologic testing may not be necessary.10 Electrophysiologic studies are also not indicated in patients whose symptoms are shown to not be associated with a conduction abnormality or block. In addition, patients without symptoms who have intermittent AV block associated with sinus slowing, gradual PR prolongation before a nonconducted P wave, and a narrow QRS complex should not undergo electrophysiologic study, given the benign prognosis of these findings. The incidence of progression of bifascicular block to CHB is variable, ranging from 2% to 6% per year. The method of patient selection affects this incidence, with patients who have asymptomatic bifascicular block progressing to CHB at a rate of 2% per year, and patients with symptoms (e.g., syncope or presyncope) progressing at a rate closer to 6% per year.20-22 Many of these studies emphasize the high mortality associated with BBB and bifascicular block. It is worth emphasizing that the mortality associated with the presence of structural heart disease predominantly reflects death due to AMI, heart failure, or ventricular tachyarrhythmias rather than bradyarrhythmias. Three large studies of patients with chronic BBB have been performed to assess the role of His bundle

Chapter 14: Pacing for Atrioventricular Conduction System Disease

conduction (e.g., H-V interval) measurements in predicting progression to CHB. The measurement of the H-V interval represents the conduction time through the His bundle and bundle branches until ventricular activation begins. Because these studies included both asymptomatic and symptomatic patients, care must be taken to ensure that similar patient populations are compared for proper interpretation of these results. Dhingra and colleagues20 prospectively followed 517 patients with BBB and measured the time required for progression to second- and third-degree block. In their study, only 13% of patients presented with syncope; the remainder did not have symptoms. The cumulative 7year incidence of progression to AV block was 10% in the group with a normal H-V interval and 20% in the group with H-V interval prolongation. The cumulative mortality rate at 7 years was 48% in patients with a normal H-V interval and 66% in patients with a prolonged H-V interval.20 This study emphasized that despite the high mortality associated with the presence of bifascicular block, there is only a low rate of progression to more advanced AV block. McAnulty and associates23 studied 554 patients with “high-risk” BBB, defined as LBBB, RBBB and left- or right-axis, RBBB with alternating left and right axis, or alternating RBBB and LBBB. The cumulative incidence of AV block, either type II second-degree or CHB, was 4.9%, or 1% per year, in patients with a prolonged H-V interval and 1.9% in patients with a normal H-V interval (difference not significant). H-V interval prolongation did not predict a higher risk of development of CHB. In this study, 8.5% of patients experienced syncope after entry into the study. The incidence of complete AV block was 17% in patients with syncope, compared with 2% in patients without a history of syncope. Scheinman and colleagues21 studied 401 patients with chronic BBB for about 30 months. This study, in contrast to the Dhingra20 and McAnulty23 studies, primarily included patients with symptoms referred for electrophysiologic study. About 40% of the patients in the Scheinman study had a history of syncope.21 In patients with an H-V interval of more than 70 msec, the incidence of progression to spontaneous second- or thirddegree AV block was 12%. The incidence of complete AV block was 25% for those with an H-V interval of 100 msec or greater. The yearly incidence of spontaneous AV block was 3% in those with a normal H-V interval and 3.5% in those with a prolonged H-V interval. Thus, these findings suggest that there is a relationship between a prolonged H-V interval and the development of CHB during the ensuing years in patients with intraventricular conduction disturbances. It also seems that the risk varies directly with the extent of H-V prolongation. Symptomatic patients with syncope or presyncope are at much higher risk than asymptomatic patients. The overall risk of CHB in an unselected, symptom-free group of patients with chronic BBB is low (40%). Methods Used to Identify Patients at Risk for Atrioventricular Block In patients with symptomatic BBB or bifascicular block, electrophysiologic testing with measurement of H-V intervals has been used for several decades. A markedly prolonged H-V interval (100 msec or longer) is predictive of development of symptomatic heart block. As noted earlier, Scheinman and colleagues21 demonstrated that an H-V interval of 100 msec or longer identified a group of patients who had a 25% risk of development of heart block over a mean follow-up of 22 months. According to current guidelines published by the American College of Cardiology/American Heart Association

436

Section Two: Clinical Concepts

Task Force on Practice Guidelines/North American Society for Pacing and Electrophysiology (ACC/AHA/ NASPE), an H-V interval of 100 msec or longer in an asymptomatic patient documented as an incidental finding on an electrophysiologic study is a class IIa indication for permanent implantation of a pacemaker. Although the finding of a markedly prolonged H-V interval is quite specific, it is very insensitive because H-V intervals of 100 msec or longer are very uncommon. Atrial pacing to stress the His-Purkinje system may provide additional information to identify patients at risk of spontaneous AV block. Most healthy subjects do not experience second- or third-degree infra-Hisian block during atrial pacing when the atrial rate is gradually increased, as would occur spontaneously. Certain pacing protocols with abrupt onset of pacing at rapid rates are more likely to induce infra-Hisian block, even in healthy subjects, but this rarely occurs at pacing rates below 150 beats per minute (bpm). Because AV nodal dysfunction is frequently seen in patients with significant His-Purkinje system disease, AV nodal block may occur at lower pacing rates than those necessary to demonstrate infra-Hisian block. This “protective” effect of AV nodal dysfunction during resting states may lead to the incorrect conclusion that significant His-Purkinje disease is not present. A second trial of atrial pacing after administration of atropine or isoproterenol to facilitate AV nodal conduction, however, may demonstrate infra-Hisian block. Dhingra and colleagues27 reported a 50% rate of progression to type II or complete AV block in patients in whom block develops distal to the His bundle at paced rates of less than 150 bpm. In a later study, Petrac and coworkers28 evaluated 192 patients with chronic BBB and syncope, of whom 18 (9%) had incremental atrial pacing–induced infra-Hisian seconddegree AV block at a paced rate of 150 bpm or less (mean pacing rate, 112 ± 10 bpm). During a mean follow-up of 68 ± 35 months, 14 of the 18 patients (78%) demonstrated spontaneous second- or third-degree AV block, confirming that this abnormal finding identifies a subgroup at a high risk for development of heart block. However, like an H-V interval of 100 msec or less, His-Purkinje block during incremental atrial pacing at physiologic rates is an uncommon finding in patients with BBB and syncope. According to current guidelines, if atrial pacing–induced infra-Hisian block that is “not physiologic” is demonstrated as an incidental finding on an electrophysiologic study, permanent pacing is recommended (class IIa indication). Provocative drug tests have been suggested as another means of evaluating the distal conduction system (Fig. 14-11).29 Pharmacologic stress testing is often considered in patients with BBB and syncope in whom the H-V interval at baseline is 70 msec or higher but less than 100 msec and infra-Hisian block is not demonstrated. There are only limited available data describing the experience with intravenous type Ia (procainamide, ajmaline, or disopyramide) or type Ic (flecainide) antiarrhythmic drugs.20,30-32 Only intravenous procainamide is available in the United States. Administration of these agents may result in a marked increase in the H-V interval (>15-20 msec), an H-V

interval greater than 100 msec, or precipitation of spontaneous type II second- or third-degree AV block, all of which may indicate a higher risk for development of CHB. Intravenous disopyramide has the potential benefit of facilitating AV nodal conduction by its anticholinergic properties while accentuating underlying infra-Hisian disease through its membrane-stabilizing effects.31 Tonkin and associates32 administered procainamide at a dose of up to 10 mg/kg to 42 patients with BBB and syncope and produced intermittent second- or third-degree His-Purkinje block or H-V prolongation to more than 15 msec during sinus rhythm in 11 patients (26%).33 However only 2 of the 11 patients (18%) with a positive result had documented high-grade AV block during 38 months of follow-up, and 3 of 5 asymptomatic control patients with BBB (60%) had a positive procainamide challenge test result. Other studies of class I antiarrhythmic drug testing to stress the HisPurkinje system likewise suffer from limited patient numbers, lack adequate control groups or follow-up, and seem to indicate that this test has low predictive value. Current permanent pacing guidelines do not include a recommendation regarding the need for permanent pacing on the basis of results of pharmacologic stress testing of the His-Purkinje system. Electrophysiologic testing has recognized limitations for identifying patients with significant AV nodal or His-Purkinje dysfunction.34 Although finding an abnormality on an electrophysiologic study may be helpful, the sensitivity of electrophysiologic testing is low and cannot be used to entirely exclude a significant AV conduction disturbance. In a small study conducted by Fujimura and associates,34 13 patients with documented symptomatic transient second- or third-degree AV block referred for implantation of a permanent pacemaker underwent AV conduction testing at the time of pacemaker insertion. These tests included facilitation of AV nodal conduction with atropine and pharmacologic stress of His-Purkinje conduction with low doses of procainamide. Surprisingly, only 2 of the 13 patients showed significant abnormalities in the AV conduction system (inducible infra-Hisian block in both cases) during electrophysiologic testing, yielding a sensitivity of 15.4%. Two other patients had moderately prolonged H-V intervals, although they were much shorter than 100 msec. If these 2 patients are included in the diagnostic data, the sensitivity of electrophysiologic testing is increased to 46%. Thus, this study raised important questions about the sensitivity of electrophysiologic testing for identifying patients at risk for development of symptomatic AV block.

Classification, Epidemiology, and Natural History of Atrioventricular Conduction Disturbances AV block can be classified clinically on the basis of electrocardiographic findings, anatomic site of block, onset,

Chapter 14: Pacing for Atrioventricular Conduction System Disease Figure 14-11. A, Electrophysiologic testing in the baseline state in a patient with syncope reveals a left bundle branch block pattern during sinus rhythm. There is 1:1 atrioventricular (AV) conduction with an H-V interval of 60 msec. B, After a loading dose of procainamide, 2:1 AV conduction develops. The first and third atrial activations are conducted to the ventricles with an H-V interval of 110 msec. The third and fourth atrial activations conduct through the AV node to generate a His bundle potential without subsequent ventricular activation. Thus, this test illustrates spontaneous infra-Hisian block induced by procainamide. HBE1, HBE2, and HBE3 are proximal, middle, and distal His bundle catheter recordings, respectively; RA, right atrial recordings; RV, right ventricular recording.

437

I II III V1 RA RA RV HBE1 HBE2 HV⫽60ms

HBE3

A I II III V1 RA RA RV HBE1 HBE2 HBE3

A

H

V

A

H

HV⫽110ms

B

extent of severity, clinical presentation, underlying etiology, or associated conditions. Each of these classifications provides insight into the basis and management of these clinical disorders. Patients who present with symptomatic first-, second-, or third-degree AV block usually have symptoms of syncope, dizziness, decreased energy, palpitations, or recurrent presyncope or dizziness. Other symptoms, which primarily reflect inadequate cardiac output or tissue perfusion, are fatigue, angina, and congestive heart failure. The most severe symptom is recurrent Stokes-Adams attacks or documented episodes of polymorphic ventricular tachycardia. Patients with long P-R intervals may have symptoms suggestive of pacemaker syndrome and may demonstrate resolution of symptoms with institution of dualchamber pacing. It is important to emphasize that symptoms can be subtle or nonspecific in some patients or may be of sufficiently long duration that a high index of suspicion is warranted. Some clinicians recommend temporary pacing to document improvement of symptoms or reversal of long-standing problems, but the use-

fulness of this intervention has not been demonstrated in prospective studies. First-Degree AV Block The prognosis and natural history in patients with primary first-degree AV block and moderate PR prolongation are almost always benign.35 Progression to CHB over time occurred in about 4% of patients in the study by Mymin and coworkers.35 Most of the patients (66%) had only mild to moderate PR prolongation, to about 0.22 to 0.23 seconds. In the great majority of subjects, the P-R interval remained within a narrow range, changing by less than 0.04 seconds. Patients with markedly prolonged P-R interval may or may not be symptomatic at rest, but may demonstrate a pseudo-pacemaker syndrome due to AV dyssynchrony, particularly during exertion. They are more likely to become symptomatic during exercise, because the P-R interval may not shorten appropriately as the R-R interval decreases. Zornosa and colleagues36 described occurrence of PR prolongation after radiofrequency ablation as a result of

438

Section Two: Clinical Concepts

injury of the fast pathway in patients with AV nodal reentry who were undergoing ablation. Symptoms due to long P-R intervals resolved after DDD pacing was performed. Kim and associates37 also described a patient with intermittent failure of fast pathway conduction who experienced light-headedness, weakness, and chest fullness when the P-R interval suddenly shifted from between 160 and 180 msec to 360 msec; this patient’s condition improved after pacing.36,37 Implantation of permanent pacemakers is recommended in patients with first-degree AV block with symptoms similar to those of pacemaker syndrome (class IIa indication). Previous versions of the ACC/AHA/NASPE guidelines required documentation of alleviation of symptoms with temporary AV pacing prior to permanent pacing. However, this requirement was removed from the latest (2002) revision of the guidelines, in part because a temporary AV pacing study may not demonstrate symptomatic improvement at rest and is often impractical to perform during exercise. Thus, it is reasonable to institute permanent pacing in symptomatic patients with very long P-R intervals (≥0.30 sec) that do not shorten during exercise. Current guidelines also recommend permanent pacemaker implantation in patients with LV dysfunction, congestive heart failure, and marked first-degree AV block (>0.30 sec) in whom a shorter A-V interval results in hemodynamic improvement (class IIb indication). A study in patients with first-degree AV block suggested that systolic performance measured using a Doppler echocardiography–derived aortic flow time velocity integral improved after institution of DDD pacing at a rate of 70 bpm if the intrinsic AV conduction time (A-R interval) was longer than 0.27 seconds.38 The optimal AV delay in this study was 159 ± 22 msec, which is consistent with that in most other studies, in that the optimal AV delay is around 150 msec at rest. Second-Degree AV Block Controversy exists regarding the prognosis and need for permanent pacing in patients with chronic type I second-degree AV block in the presence of a narrow QRS complex. Some consider this condition benign only in young people or athletes without organic heart disease. The natural history of 56 patients with chronic type I second-degree AV block, some of whom were younger than 35 years or were well-trained athletes, was described by Strasberg and colleagues in 1981.39 They concluded that progression to CHB is relatively uncommon in this patient population and that this finding carries a benign prognosis in the absence of structural heart disease. A 1985 report by Shaw and coworkers40 in 1985 suggested that patients with type I second-degree AV block have a worse prognosis than age- and sex-matched individuals unless the patients already had permanently implanted pacemakers. Their patient population consisted of 214 patients with chronic second-degree AV block with a mean age of 72 years who were monitored over a 14-year period; the patients were divided into three groups—those with type I block (77 patients), those with type II block (86

patients), and those with 2:1 or 3:1 block (51 patients). The 3- and 5-year survival times were similarly poor regardless of the type of AV block. Patients with type I block without BBB did not fare any better than those with type II block. Patients with type I second-degree AV block who received permanent pacemakers had a survival similar to that of an age- and sex-matched control population. In 1991, a working group of the British Pacing and Electrophysiology Group (BPEG) suggested that pacing should be considered in adults in whom type I seconddegree AV block occurs during much of the day or night irrespective of the presence or absence of symptoms.41 In 2004, Shaw and associates42 reported again on the prognosis of patients with type I second-degree AV block and once again concluded that type I seconddegree AV block is not a benign condition in patients 45 years or older. The majority of their patients with type I second-degree AV block who were 45 years or older progressed to higher-degree AV block, experienced symptomatic bradycardia, or died prematurely if they did not receive pacemakers. These investigators recommended pacemaker implantation in patients with type I seconddegree AV block even in the absence of symptoms or structural heart disease, except in those younger than 45 years. According to current ACC/AHA/NASPE guidelines however, permanent pacemaker implantation in asymptomatic type I second-degree AV block that is at the supra-His (AV node) level or is not known to be intra-Hisian or infra-Hisian is considered insupportable by current evidence (class III indication). It seems prudent, on the basis of available data, to at least monitor closely any elderly patient with asymptomatic type I AV block or 2:1 AV block with narrow QRS complexes, because these ECG abnormalities may be markers for progressive conduction system disease. The natural history of asymptomatic type II seconddegree AV block initially was addressed in a study from the University of Illinois reported in 1974.43 Most patients monitored in this study were found to experience symptoms within a relatively short period. In the study by Shaw and colleagues40 reported in 1985, 86 patients (mean age 74 years) seen and monitored between 1968 and 1982 with chronic Mobitz type II second-degree heart block had a 5-year survival rate of 61%. The five-year survival of those who underwent permanent pacemaker implantation was significantly better than those who did not. These observations form the basis for recommendations to institute permanent pacing in all patients with type II second-degree AV block regardless of symptoms (class IIa indication). Complete Heart Block The natural history of spontaneously developing asymptomatic CHB in adult life dates back to the days before pacemaker therapy was available.44-46 Today, almost all adult patients with CHB eventually have symptoms and undergo pacemaker placement. Several studies published in the 1960s emphasized the poor prognosis of patients with CHB. The 1-year survival rate of patients who experienced Stokes-Adams attacks due to CHB and

Chapter 14: Pacing for Atrioventricular Conduction System Disease

did not receive pacing was only 50% to 75%, significantly less than that of a sex- and age-matched control population.45-47 The “best” prognosis was in patients with an idiopathic or unknown cause of CHB. At least 33% of deaths were related to CHB and Stokes-Adams attacks. These differences in survival persisted even after 15 years of follow-up and appear to be related to the considerably higher incidence of sudden death.46 There is some debate about whether the presence of syncope is associated with a worse prognosis in patients with documented CHB. The prognosis for transient CHB was poor as well, with a 36% 1-year mortality rate reported in at least one study from the 1970s.45 Whether this poor prognosis applies today to patients with transient CHB who have not received pacing is unknown. Edhag and Swahn45,46 reported in the 1960s and 1970s on the long-term prognosis of 248 patients with high-grade AV block, most of whom had complete heart AV block, with a mean 6.5 years of follow-up. The mean age at pacemaker implantation was 66 years, and the 1-year survival rate of patients who received pacemakers was 86%, slightly lower than the 95% 1-year survival rate of an age- and sex-matched group of Swedish patients. After the first year, survival in the patients with pacemakers was similar to that in the general population. Edhag46 compared survival in different age groups, found no difference in survival between elderly patients with heart block who underwent permanent pacing and the age- and sex-matched general population. In contrast, younger patients with heart block had an increased mortality even after pacing than sex- and age-matched controls from the general population.46 It is likely that this higher mortality is a reflection of the underlying structural heart disease that is responsible for high-grade AV block. CHB can be described as acute or chronic depending on its onset. Acute AV block associated with myocardial ischemia is rare but may occur and result in transient AV block. High-grade AV block is strictly defined as 3:1, 4:1, or higher AV ratios in which AV synchrony is intermittently present. As in complete AV block, block may be localized anywhere in the conduction system (Fig. 14-12). In some patients, block may be present at multiple levels in the conduction system. Generically, the term high-grade AV block has been used to describe any form of AV block that suggests an increased risk for CHB or symptomatic bradycardia. This typically includes type II second-degree block, 2:1 AV block, strictly defined high-grade AV block, and CHB. The generic use of the term high-grade AV block may be best avoided because the multiple forms of AV block included have variable pathogeneses and prognoses that blur the clinical usefulness of the term. Paroxysmal AV Block Paroxysmal AV block is defined as the sudden occurrence, during a period of 1:1 AV conduction, of block of sequential atrial impulses resulting in a transient total interruption of AV conduction.49,50 It is thus the onset of a paroxysm of high-grade AV block associated with a period of ventricular asystole before conduction

439

returns or a subsidiary pacemaker escapes. Paroxysmal AV block may occur in a variety of clinical conditions but has been described most often in association with vagal reactions, such as during vomiting, coughing, or swallowing, after urination, or with abdominal pain, carotid sinus massage, coronary angiography, or headup tilt table testing. Patients with neurally mediated syncopal syndromes may have transient heart block, typically with associated sinus slowing. It may occur in some patients with tachycardia-dependent AV block in the His-Purkinje system (phase 3 block), during or after exertion, with the abrupt onset of bradycardia (phase 4 block), and in type II second-degree AV block. One report described the clinical experience in 20 patients (mean age, 63 ± 14 years) with paroxysmal AV block seen at a single institution over a 12-year period.51 Paroxysmal AV block in these patients was related to a vagal reaction, AV-blocking drugs, or distal conduction disease. The AV block in these patients lasted from 2.2 to 36 seconds. Fifteen patients experienced syncope, and one patient had bradycardia-induced polymorphic ventricular tachycardia that required electrical cardioversion. About one half the patients had structural heart disease and a wide QRS duration. Complete AV block can be classified as congenital or acquired. In patients with acquired complete AV block, the site of block is localized distal to the His bundle in about 70% to 90% of patients, to the His bundle in 15% to 20%, and within the AV node in 16% to 25%.10 In patients with congenital heart block, the escape rhythm is more often found in the proximal His bundle or AV node. Bundle-Branch Block Most patients with chronic BBB or bifascicular block have underlying structural heart disease, the prevalence of which ranges from 50% to 80%.20,21,23,25 Historically, it was believed that progression from chronic bifascicular block to trifascicular block was common. Retrospective studies in patients with chronic bifascicular block suggested that the risk of progression to complete AV block was 5% to 10% per year. In the early 1980s, the results of several large prospective studies questioned assumptions about the incidence and clinical implications of the progression of conduction system disease in this patient population. Prospective studies of groups of symptom-free patients with bifascicular block who were found to have prolonged H-V intervals on electrophysiologic study showed that such patients are at increased risk for CHB but that the absolute risk remains very low, about 1% to 2% per year.20,23 In the study by McAnulty and colleagues,23 the risk for development of CHB was 5% in 5 years. A prolonged H-V interval was associated with higher values for both total cardiovascular mortality and sudden death. It is likely that a prolonged H-V interval is associated with more extensive structural heart disease. Furthermore, these studies demonstrated that in the absence of symptoms, routine His bundle recordings are of limited usefulness in patients with bifascicular block. Asymptomatic individuals with chronic BBB need no further evaluation than an occasional ECG.

440

Section Two: Clinical Concepts

Figure 14-12. Rhythm strip of high-grade atrioventricular (AV) block. The baseline rhythm is sinus tachycardia, in which the P wave occurs simultaneously with the T wave. After the fifth QRS complex, there is an abrupt, or paroxysmal, onset of AV block with four consecutive P waves that do not conduct to the ventricles. The sixth QRS complex probably represents a junctional escape rhythm followed by three conducted complexes and then a longer episode of high-grade AV block. Before both episodes of high-grade AV block, there is no obvious P-R interval prolongation, nor is there slowing of the sinus rate to suggest hypervagotonia as a cause of this patient’s block.

On the other hand, patients with syncope and bifascicular block represent a different clinical problem. If a thorough clinical evaluation, including a history, physical examination, and ECG, do not uncover a cause of syncope, an electrophysiologic study may be useful.25,26 Linzer and associates48 found that the presence of first-degree AV block or BBB increased the odds ratio of finding abnormalities suggesting risk of bradyarrhythmia (predominantly heart block) by three- to eightfold during electrophysiologic studies in patients with unexplained syncope (Table 14-1). Electrophysiologic studies may uncover other causes of syncope, such as sinus node dysfunction, rapid supraventricular tachycardias, and inducible monomorphic ventricular tachycardia. In some studies, monomorphic ventricular

tachycardia was inducible in at least 30% of patients with BBB and syncope.25,26 A minority of patients are found to have a markedly prolonged H-V interval, abnormal or fragmented His bundle electrogram, or block distal to the His bundle with atrial pacing, suggesting the need for permanent implantation of a pacemaker. Congenital AV Block Congenital CHB traditionally was diagnosed in the first month after birth in a child in whom a slow heart rate was detected and certain infectious etiologies rarely seen today, such as diphtheria, rheumatic fever, and congenital syphilis, were excluded.52,53 Currently, with fetal echocardiography, many cases are diagnosed in utero and, if associated with structural heart disease, are associated with a high rate of fetal death. The incidence of congenital CHB is estimated to be 1 in 15,000 to 1 in 22,000 live births. More than one half of fetuses found to have congenital CHB have structural heart disease, including congenitally corrected transposition of the great arteries, and often have a poor prognosis in infancy. When congenital CHB is detected in utero in a child with a structurally normal heart, the condition is frequently associated with intrauterine exposure to maternal autoantibodies to Ro and La (i.e., neonatal lupus); this situation has a better prognosis than congenital heart disease. The development of AV block in a child with a structurally normal heart is uncommon but should not be confused with congenital CHB. Childhood-onset heart block often is presumed to be due to viral myocarditis.

Odds Ratio for Abnormality on Electrophysiologic Testing in Patients with Syncope TABLE 14-1.

Multivariable

95% Confidence Interval for Multivariable

Age

1.01

0.99-1.03

Duration (months)

1.00

0.98-1.02

Sex (male)

1.76

0.79-3.93

Organic heart disease

1.53

0.71-3.33

Sudden loss of consciousness

1.93

0.89-4.16

Left ventricular ejection fraction

0.99

0.93-1.06

ECG variables: Bundle branch block Sinus bradycardia First-degree heart block Premature ventricular contractions (PVSs)

2.97* 3.47* 7.89† 1.47

1.23-7.21 1.12-10.71 2.12-29.31 0.37-5.82

Holter monitoring variables: Sinus bradycardia Sinus pause Mobitz I atrioventricular block PVCs

0.68 1.04 0.63 0.87

0.21-2.23 0.26-4.23 0.06-6.33 0.35-2.13

Clinical Variable

*P < .05. † P < .001. Adapted from Linzer M, Prystowsky EN, Divine GW, et al: Predicting the outcomes of electrophysiologic studies of patients with unexplained syncope: Preliminary validation of a derived model. J Gen Intern Med 6:113, 1991.

Chapter 14: Pacing for Atrioventricular Conduction System Disease

In patients with congenital CHB, the mean resting heart rate is between 40 and 60 bpm but decreases with age. The indications for permanent pacemaker implantation are controversial in these young patients but may include symptomatic bradycardia manifested by long naps or nightmares, heart rate of less than 55 bpm in a neonate, an average heart rate of less than 50 bpm in a child, pauses in heart beat of more than 3 seconds while a child is awake or more than 5 seconds during sleep, associated congenital heart disease or ventricular dysfunction, exercise intolerance due to chronotropic incompetence, prolonged QTc, wide ventricular escape rhythm, and complex ventricular ectopy. A 2004 study demonstrated, however, that long-term transvenous RV apical pacing was associated with deleterious LV remodeling and reduced exercise capacity after 10 ± 3 years of follow-up in patients with congenital CHB.54 Thus, alternative ventricular pacing sites, including RV septum and outflow tract and LV, have been proposed in patients with congenital CHB, who may require many decades of ventricular pacing. Inherited causes of AV conduction disturbances have been recognized, representing an exciting new development in our understanding of the AV conduction system.55 Insights have also been provided from genetic animal models in which molecular genetic causes of AV block are beginning to be identified. Many of the identified cases of inherited AV conduction disease were previously classified as idiopathic or having an unknown cause. Familial clustering in some cases has been seen with an autosomal dominant pattern of inheritance and associated congenital heart malformations and cardiomyopathy. For example, mutations of lamin A/C gene (LMNA), referred to as “laminopathy,” cause dilated cardiomyopathy with AV conduction defects. A large Japanese family with 21 of 224 members affected by this genetic defect, which manifests clinically as progressive AV block, dilated cardiomyopathy, progressive heart failure, and sudden death, has been described.56 Electrophysiologic studies in affected individuals demonstrated AV nodal dysfunction (marked prolongation of A-H interval) and normal H-V and QRS durations. Histologic evaluation of postmortem heart specimens from affected members showed preferential degeneration of the AV nodal region.

Causes of Acquired Atrioventricular (AV) Block TABLE 14-2.

Idiopathic fibrodegenerative diseases

Lev’s disease Lenègre’s disease

Ischemic heart diseases

Myocardial infarction Ischemic cardiomyopathy

Nonischemic cardiomyopathies

Myocarditis Idiopathic dilated cardiomyopathies Hypertensive heart disease

Cardiac surgery

Coronary artery bypass Aortic, mitral, or tricuspid valve replacement/repair Ventricular septal defect repair Congenital heart disease repair Septal myomectomy

Ablation

His bundle ablation Ablation of septal accessory connections Ablation of slow or fast AV nodal pathway for AV nodal reentrant tachycardia Catheter-based septal ablation for hypertrophic cardiomyopathy

Trauma

Chest trauma

Infections

Endocarditis Chagas’ disease Lyme disease Acute rheumatic fever Other: bacterial, viral, rickettsial, fungal

Neuromuscular diseases

Myotonic dystrophy Fascioscapulohumeral dystrophy Other muscular dystrophies Kearns-Sayre syndrome Friedreich’s ataxia

Infiltrative diseases

Amyloidosis Sarcoidosis Hemochromatosis Carcinoid

Neoplastic diseases

Postradiation therapy Primary and metastatic tumors

Connective tissue diseases

Rheumatoid arthritis Systemic lupus erythematosus Systemic scleroderma Ankylosing spondylitis Others

Drugs

β-blockers Calcium channel antagonists Digoxin and other cardiac glycosides (e.g., oleandrin) Amiodarone Procainamide Flecainide Adenosine Chemotherapeutic agents (arsenic trioxide) Antimalarials (chloroquine) Tricyclic antidepressants Phenothioazines Donepezil

Miscellaneous causes

Exercise Vagal mediation (hypervagatonia, neurocardiac, vasovagal) Temporal lobe epilepsy Hyperkalemia

Acquired Causes of AV Block Acquired AV block may be secondary to a number of causes of generalized myocardial scarring (Table 14-2). These causes include atherosclerosis, dilated cardiomyopathy, hypertension, infiltrative cardiomyopathies, inflammatory disorders, and infectious diseases. In most cases, the specific etiology is clinically unknown and, with few exceptions (e.g., Lyme disease, endocarditis, sarcoidosis), is relatively unimportant from a therapeutic point of view. The most common cause of chronic acquired AV block is related to aging of the cardiac cytoskeleton. An entity known as idiopathic bilateral bundle branch fibrosis, or Lev’s disease, is characterized by slowly progressive replacement of specialized conduction tissue by fibrosis, resulting in progressive

441

442

Section Two: Clinical Concepts

fascicular block and BBB.57 Lev57 proposed that damage to the proximal LBB and adjacent main bundle or main bundle alone is the result of an aging process exaggerated by hypertension and arteriosclerosis of the blood vessels supplying the conduction system. Another variant of idiopathic conduction system disorder is Lenègre’s disease, which occurs in the younger population and is characterized by loss of conduction tissue, predominantly in the peripheral parts of the bundle branch.58 Patients with sick sinus syndrome are known to be at risk for concomitant symptomatic heart block.59,60 The block may be due to progressive fibrodegenerative disease extending from the sinus node region to the AV conduction system. The relative frequency of this association varies between studies. Rosenqvist and Obel61 pooled the data from 28 published studies and reported a mean incidence of development of second- or thirddegree AV block to be 0.6% per year (range, 0%/yr to 4.5%/yr) in patients in whom permanent atrial pacemakers have been implanted for symptomatic sinus node dysfunction. The total prevalence of second- or third-degree AV block was 2.1% (range, 0% to 11%). In a retrospective review of 1395 patients with sick sinus syndrome who were monitored for a mean of 34 months, Sutton and Kenny62 estimated that the development of conduction system diseases had an annual incidence of 3%; such conduction diseases included significant first-degree AV block, BBB, H-V prolongation, and a low Wenckebach heart rate. Thus, AV conduction system disease occurs relatively frequently in patients with sinus node dysfunction. Similarly, sinus node dysfunction, particularly chronotropic incompetence, may occur commonly in patients with acquired CHB. Complete AV block may occur after cardiac surgical or interventional catheter-based procedures. CHB occurs more commonly (3%-6% incidence) after replacement of aortic, mitral, or tricuspid valves, given the proximity of their anuli to the AV junction, than after isolated coronary artery bypass graft surgery, in which the incidence is less than 1% to 2%.63,64 CHB is seen commonly after surgical procedures to repair ventricular septal defects, tetralogy of Fallot, AV canal defects, or subvalvular aortic stenosis.65,66 Heart block is also a potential complication of septal myomectomy and catheter-based septal ablation to relieve LV outflow tract obstruction in hypertrophic cardiomyopathy.67,68 The incidence of heart block requiring permanent pacing after alcohol septal ablation is reported to vary from 10% to 33%. The time course of conduction defects after bypass surgery was investigated in one study.69 Operative technique consisted of cold, hyperkalemic cardioplegia, and conduction defects resolved partially or completely in 50% of patients. Patients with conduction defects generally had longer cardiopulmonary bypass times, longer aortic cross-clamp times, and more vessels requiring bypass. In three of the four patients with CHB, the heart block eventually resolved after discharge and implantation of a permanent pacemaker. Reasons for conduction abnormalities after cardiac surgery include ischemic injury to the conduction system, direct surgical manipulation or trauma to conduction tissue, traumatic disruption of the distal conduction

system, edema, dissecting hematomas, and alterations in conduction caused by cardioplegia. Surgery for correction of valvular heart disease commonly leads to conduction defects. After discontinuation of cardiopulmonary bypass, a variety of cardiac rhythm disturbances may be seen, including sinus arrest, junctional rhythm, BBB, AV block, and sinus bradycardia. Many of these rhythm disturbances are transient, resolving within 5 to 7 days. Transient BBB is quite common, occurring in 4% to 35% of patients and generally resolving within 12 to 24 hours.63 In one study, newly acquired, persistent BBB developed in 15.6% of patients after aortic valve replacement and was associated with a higher adverse event rate.70 The investigators recommended that prophylactic implantation of a pacemaker be considered soon after surgery in patients who demonstrate persistent BBB. Conduction disturbances are particularly common both in patients with aortic valve disease and after aortic valve replacement, with 5% to 30% of patients experiencing some conduction abnormality after valve replacement. Most of these abnormalities are transient; however, chronic CHB may occur. Postoperative AV block that is not expected to resolve after cardiac surgery is a class I indication for permanent pacing. The incidence of conduction disorders requiring permanent pacing in patients after aortic valve replacement is reported to be 3% to 6%.71 Intraoperative heart block does not predict the need for permanent pacing.72 One study found that risk factors for irreversible AV block requiring permanent pacemaker implantation after aortic valve replacement were previous aortic regurgitation, MI, pulmonary hypertension, and postoperative electrolyte imbalance.71 In another study, Koplan and associates73 developed and validated a preoperative risk score to predict the need for permanent pacing after cardiac valve surgery, through the use of a large database from patients undergoing cardiac valve surgery at a single institution from 1992 to 2002.73 Preoperative predictors of the need for permanent pacing after cardiac valve surgery were age more than 70 years, previous valve surgery, multiple valve surgery (especially involving the tricuspid valve), preoperative BBB (especially RBBB), and first-degree AV block. The incidence of permanent pacemaker implantation ranged from 25% in high-risk patients to 3.6% in low-risk patients, with risk based on preoperative variables. Glikson and colleagues74 showed that postoperative complete AV block was the most important predictor of subsequent pacemaker dependency. They recommended earlier decisions on the timing of permanent pacemaker implantation, by the sixth postoperative day in patients with a wide QRS complex escape rhythm and by the ninth postoperative day in patients with a narrow QRS complex escape rhythm. In most institutions, permanent pacing would be instituted earlier, probably by the fourth to sixth postoperative day. Persistent heart block occurs infrequently (0.5%-2% of patients) after radiofrequency catheter ablation of septal accessory pathways or the slow AV nodal pathway in patients with AV nodal reentrant tachycardia.75 Transient, intraprocedural AV block during abla-

Chapter 14: Pacing for Atrioventricular Conduction System Disease

tion performed in close proximity to the AV septum does not necessarily indicate that permanent pacemaker implantation is required. Ablation of the fast pathway in AV nodal reentrant tachycardia in the anterior septum is associated with a higher risk of transient AV block than ablation of the slow pathway in the posterior septum. In one large series of more than 500 patients, transient second- or third-degree AV block was seen in 20% of patients during fast pathway ablation, in 2.3% during slow pathway ablation, and in 42% during combined fast and slow pathway attempted ablation.76 Within 7 days after the ablation procedure, however, persistent AV block was seen in 3.4%, 0.2% and 0% of patients in these groups, respectively. Late occurrence of unexpected heart block after radiofrequency catheter ablation of AV nodal reentry (using a posterior approach) or posteroseptal accessory pathways is rare (100 msec) in a symptomfree patient (level of evidence: B) Incidental finding at electrophysiologic study of atrial pacing–induced infra-Hisian block that is not physiologic (level of evidence: B) Neuromuscular diseases, such as myotonic muscular dystrophy, Kearns-Sayre syndrome, Erb’s (limb-girdle) dystrophy, and peroneal muscular atrophy with any degree of fascicular block with or without symptoms, because there may be unpredictable progression of AV conduction disease (level of evidence: C).

Class IIb

Class III

Fascicular block without AV block or symptoms (level of evidence: B) Fascicular block with first-degree AV block without symptoms (level of evidence: B)

AV, atrioventricular; BBB, bundle branch block.

ratio = 1.44) but a tendency toward a lower rate of occurrence of BBB (odds ratio = 0.68).132 Another study suggested that thrombolytic therapy was associated with a tendency toward a lower rate of third-degree AV block in anterior AMI but a higher rate in inferior AMI.133 Among 6657 patients admitted with AMI

between 1990 and 1992 and included in the TRAndolapril Cardiac Evaluation (TRACE) randomized trial in Denmark, 340 (5.1%) experienced third-degree AV block during their hospitalization.133 The incidence of third-degree AV block was higher among patients with inferior AMI (193 of 2061 [9.4%]) than among those

448

Section Two: Clinical Concepts

with anterior AMI (44 of 1747 [2.5%]).133 Likewise, in pooled data from 75,993 patients with ST-segment elevation AMI treated with thrombolytic therapy, 5251 patients (6.9%) had second- or third-degree AV block. AV block occurred in 9.8% of those with inferior AMI and in 3.2% of those with anterior AMI.134 The onset of AV block usually occurs 2 to 3 days after the infarction but has a range of a few hours to 10 days. The mean duration is usually 2 to 3 days and the range of duration is 12 hours to 16 days. In one large study, third-degree AV block occurred within 48 hours of symptom onset in 81% of patients and a trend was observed toward later onset of third-degree AV block in anterior rather than in inferior AMI.134 Clinicopathologic studies indicate that there is a relationship between the anatomic location of an AMI and involvement of the conduction system.8,135 The development of AV and intraventricular blocks during anterior AMI is related to the extent of the ischemic/ infarcted area. AV block in patients with inferior AMI more often results from vagal reflexes or local metabolites occurring early within the AV node in a transient fashion. Several mechanisms have been proposed for AV block in the presence of inferior AMI. These include Bezold-Jarisch reflex, reversible ischemia or injury of the conduction system, local accumulation of adenosine or its metabolites, and local AV nodal hyperkalemia. Stimulation of the Bezold-Jarisch reflex causes an abnormally increased output of vagal nerve traffic; it is initiated by ischemia of the afferent nerves in the area of the inferoposterior LV. Reperfusion of the RCA with thrombolytic agents is a strong stimulus for the BezoldJarisch reflex.136,137 Despite this, the second Thrombolysis in Myocardial Infarction (TIMI II) study did not show an increase in AV block in patients with inferior AMI who received thrombolytic therapy and had a patent infarct-related artery.138 In inferior or posterior AMI, obstruction of the RCA produces reversible ischemia of the AV node. In patients who experienced AV block after inferior AMI, pathologic studies demonstrated little or no necrosis, structural damage, or histologic degenerative changes in the conduction system in most cases.8,135,139 Bilbao and colleagues,140 however, identified a subgroup of patients with fatal inferior or posterior AMI and AV block who had necrosis of the prenodal atrial myocardial fibers. These necrotic fibers were absent in patients without AV block. Clinically, the transient nature of the AV block supports the concept that injury to the AV node is reversible.141 The anatomic data reported by Bassan and associates128 support the concept that the blood supply of the AV node is dual. In their prospective study, 11 of 51 patients who survived an inferior AMI had some degree of transient AV block, and about 90% of the patients with AV block had simultaneous obstruction of the RCA (or left coronary artery [LCA] when it was dominant) and the proximal segment of the LAD artery. Moreover, patients with inferior AMI and LAD artery obstruction had a sixfold higher risk for development of AV block than those without LAD artery obstruction. The TIMI II data do not support this finding, however;

in study patients with inferior AMI and AV block, the incidence of disease in the LAD was low and was similar to that in patients with inferior AMI without AV block.138 The Thrombolysis and Angioplasty in Acute Myocardial Infarction (TAMI) study group also showed no increase in incidence of LAD disease in patients with inferior AMI and complete AV block.142 Local accumulation of endogenous adenosine or its metabolites also has been suggested to play a mechanistic role in AV block occurring as an early complication of inferior AMI.143 Patients with “early” AV block (occurring less than 24 hours into their hospital course) are less likely to show response to atropine.144 Several case reports or small series have suggested that aminophylline, a competitive adenosine antagonist, reverses atropine-resistant AV block in patients with inferior AMI.145 Current practice guidelines for management of ST-segment elevation AMI, however, recommend against giving aminophylline to treat bradyarrhythmias because it increases myocardial oxygen demand and is arrhythmogenic.146 A higher level of potassium was found in the lymph draining from the infarcted inferior and posterior cardiac walls of dogs after experimental RCA occlusion, suggesting that local AV nodal hyperkalemia may play a role in the development of AV block in the presence of inferior AMI.147 Sugiura and colleagues148 found that serum potassium value was an independent predictor of the occurrence of fascicular blocks in anteroseptal AMI. Anterior or anteroseptal AMI results from obstruction of the LAD artery. Occurrence of AV block and BBB in patients with anterior AMI is usually the result of necrosis of the septum and the conduction system below the AV node and reflects more extensive and permanent myocardial damage with severe LV dysfunction.139,149,150 Wilber and colleagues,151 however, reported on two patients with anterior AMI and complete AV block in whom 1:1 conduction returned within minutes after late reperfusion (>40 hours) with angioplasty. Their experience suggests that reversible ischemia rather than necrosis of the conduction system occurs in some patients. Some experimental studies in dogs with anterior AMI suggest that extensive but reversible ischemia of the infranodal conduction tissue occurs, as evidenced by recovery from complete AV block. Several clinical studies also report that most patients with anterior AMI and high-grade AV block who are discharged from the hospital show late recovery of 1:1 AV conduction.152-156 Atrioventricular Block without Bundle Branch Block Incidence In a series of 684 consecutive patients with AMI admitted to the Los Angeles County–University of Southern California Medical Center (LAC-USCMC) Coronary Care Unit (CCU) between 1966 and 1970, 110 had AV block (16%); 79 of 110 patients (72%) with AV block did not have BBB.157 The total percentages of patients who had

Chapter 14: Pacing for Atrioventricular Conduction System Disease TABLE 14-5. Atrioventricular (AV) Block in Acute Myocardial Infarction (AMI) without Bundle Branch Block* Incidence (%): First-degree AV block Second-degree AV block Third-degree AV block

12 (79/684 patients) 6 (44/684 patients) 7 (50/684 patients) 4 (29/684 patients)

Site of infarction (%): Inferior Anterior Combined

79 18 6

Progression (%): First-degree AV block to secondor third-degree AV block Second-degree AV block to third-degree AV block Outcome (%): Hospital mortality Return to 1:1 conduction in survivors

59

449

series, of the 79 patients with AV block who did not have BBB, 60 (76%) had an inferior infarction, 14 (18%) an anterior infarction, and 5 (6%) a combined infarction (Table 14-6). Progression of Atrioventricular Block In patients with inferior AMI, progression of AV block commonly occurs in stages, whereas in those with anterior AMI, it may occur in stages, or third-degree AV block or ventricular asystole may develop suddenly (Fig. 14-15).141 Outcome

36

29 95

*Data from 684 consecutive patients with AMI at Los Angeles County–University of Southern California Medical Center (LACUSCMC), Los Angeles. Adapted from de Guzman M, Rahimtoola SH: What is the role of pacemakers in patients with coronary artery disease and conduction abnormalities? In Rahimtoola SH (ed): Controversies in Coronary Artery Disease. Philadelphia, FA Davis, 1983, pp 191-207.

AV block complicating AMI is associated with a high mortality rate (24% to 48%), two to three times that for AMI without AV block (9% to 16%). Even with the use of thrombolytic therapy and primary percutaneous coronary interventions, if AV block occurs in the setting of AMI, mortality remains high, especially in anterior MI.138,142,159,160 This poor prognosis generally reflects the larger ischemic/infarcted region associated with devel-

first-, second-, or third-degree AV block at some time were 6%, 7%, and 4%, respectively (Table 14-5). Site of Infarction AV block is more commonly associated with inferior infarction, and in those who experience second- and third-degree blocks, inferior AMI is present two to four times as often as anterior AMI. The site of block in inferior infarction is above the His bundle in about 90% of patients, whereas in anterior infarction, the conduction abnormality is usually localized below the His bundle in the distal conducting system.158 In the LAC-USCMC

Figure 14-15. Lead II. Sudden ventricular asystole in a patient with acute myocardial infarction complicated by right bundle branch block and left axis deviation. (From de Guzman M, Rahimtoola SH: What is the role of pacemakers in patients with coronary artery disease and conduction abnormalities? In Rahimtoola SH: Controversies in Coronary Artery Disease. Philadelphia, FA Davis, 1983.)

Atrioventricular (AV) Block in Anterior and Inferior-Posterior Acute Myocardial Infarction (AMI) TABLE 14-6.

Feature

Anterior AMI

Inferior-Posterior AMI

Pathophysiology

Extensive necrosis of septum

Reversible ischemia, injury of conduction system

Site of block

Infranodal

Intranodal

Frequency

Less frequent

Two to four times more frequent

Progression to complete AV block

Sudden

Gradual

Intraventricular conduction defect

Common

Rare

Escape focus

Ventricular

Junctional

Escape rate (per minute)

20-40

40-60

Prognosis

High mortality

Lower mortality

450

Section Two: Clinical Concepts

opment of AV block in the setting of AMI. Although AV block that occurs during inferior AMI predicts a higher risk of in-hospital death, it may be less predictive of long-term mortality in patients who survive to hospital discharge.138,142 Before the reperfusion era, Tans and coworkers130 reported that in patients with inferior AMI, those with high-grade AV block and no severe pump failure had a higher in-hospital mortality rate (17%) than those without high-grade AV block (9%).130 The major cause of death in patients who have AV block in the setting of AMI is pump failure.161 In the LAC-USCMC series, the hospital mortality rate was 29% (32 of 110 patients); 29 of the 32 deaths (91%) were due to pump failure. Survival, therefore, is greatly influenced by the severity of the hemodynamic disturbance and is less dependent on the degree of heart block. Death is related primarily to extensive myocardial damage, but in an important minority of patients, it can be attributed to sudden ventricular asystole or severe bradycardia. Atrioventricular Block with Bundle Branch Block Prior to the widespread use of thrombolytic therapy, BBB was present during hospitalization in 8% to 18% of patients with AMI.162-167 The presence of a persistent intraventricular conduction defect during the hospitalization increases the risk of high-grade AV block as well as other complications and is associated with poor survival in patients with AMI. In patients receiving thrombolytic therapy, the incidence of persistent intraventricular conduction defects appearing during the hospitalization is reduced to about 4% to 9%.131-133,168 However, the adverse risk associated with intraventricular conduction defects (except for isolated left anterior hemiblock) has persisted even in the modern era of AMI reperfusion therapy. In an older series of 2779 patients with AMI admitted from 1966 to 1977 to the LAC-USCMC coronary care unit, 257 (9%) had BBB (Table 14-7).157 Of the 257 patients, 83 (32%) had LBBB, 80 (31%) had RBBB, 72 (28%) had RBBB plus left-axis deviation, 21 (9%) had RBBB plus right-axis deviation, and one had alternating BBB. The conduction abnormality was “new” in 60%—that is, the BBB developed during the infarction and was documented by serial ECGs or was present on admission and was not seen on previous ECGs or reverted to normal conduction later as documented by serial ECGs. When the site of infarction was not obscured by the BBB in the LAC-USCMC series, the block was associated with anterior AMI about three times as often as with inferior AMI. Progression of Atrioventricular Block In the LAC-USCMC series conducted before the thrombolytic era, progression of AV block occurred in 75 of the 257 patients (29%) with AMI and BBB (see Table 14-7). Of the 28 patients with AV block and BBB who initially had a normal P-R interval, 13 (46%) experienced first-degree AV block. Nine of the 13 (69%) remained in

Bundle Branch Block in Acute Myocardial Infarction TABLE 14-7.

Incidence (%): LBBB RBBB RBBB + LAD obstruction RBBB + RAD obstruction

9 (257/2779 patients)* 32 (83/257 patients) 31 (80/257 patients) 28 (72/257 patients) 9 (21/257 patients)

Onset of BBB (%): New Old

60 40

Site of infarction (%): Inferior Anterior Combined Indeterminate Nontransmural

21 52 4 18 5

Incidence of AV block (%): First-degree Second-degree Third-degree

29 (75/257 patients) 10 (25/257 patients) 5 (13/257 patients) 14 (37/257 patients)

Progression of AV block (%): First-degree AV block to secondor third-degree AV block Second-degree AV block to third-degree AV block Progression to high-grade AV block Bilateral BBB + first-degree AV block New bilateral BBB + firstdegree AV block First-degree AV block New BBB + first-degree AV block Bilateral BBB New BBB New bilateral BBB Outcome (%): Hospital mortality Return to 1:1 conduction in survivors

32 46 18 (46/257 patients) 50 43 30 29 18 16 15 20 89

AMI, acute myocardial infarction; AV, atrioventricular; BBB, bundle branch block; L, left; LAD, left anterior descending (artery); R, right; RAD, right anterior descending (artery). *Data from 2779 AMI patients seen from October 1966 to March 1977 at Los Angeles County–University of Southern California Medical Center (LAC-USCMC), Los Angeles. Adapted from de Guzman M, Rahimtoola SH: What is the role of pacemakers in patients with coronary artery disease and conduction abnormalities? In Rahimtoola SH (ed): Controversies in Coronary Artery Disease. Philadelphia, FA Davis, 1983, pp 191-207.

first-degree block, 2 (15%) progressed to second-degree block type II only, 1 (8%) had type II second-degree block that progressed to third-degree AV block, and 1 had third-degree block without first having seconddegree block. Six patients had type II second-degree block without initially having first-degree AV block; 4 of the 6 (67%) went on to have third-degree AV block. Of the total 41 patients who had first-degree AV block and BBB, 13 (32%) progressed to high-grade block. Twenty-eight (68%) were admitted in first-degree AV block. Sixteen of the 28 (57%) remained in first-degree block; 3 (11%) progressed to type II

Chapter 14: Pacing for Atrioventricular Conduction System Disease

second-degree AV block, 1 of whom went on to thirddegree block; 5 (18%) progressed to type I seconddegree block, 2 of whom went on to third-degree block; and 4 (14%) patients had third-degree block without initially having second-degree AV block. Of the total 24 patients who had second-degree AV block and BBB, 11 (46%) progressed to third-degree block. Six had second-degree AV block on admission; of these, type II block occurred in 4, 1 of whom progressed to third-degree block, and type I occurred in 2, 1 of whom progressed to third-degree block. There were 37 patients with third-degree AV block and BBB, 13 (35%) of whom were admitted in thirddegree block. Of the 24 who progressed to third-degree block, 11 (46%) had demonstrated second-degree block; 7 of the 11 had type II second-degree block. Consistent with the LAC-USCMC series, AV block occurred in other studies in about one third of patients with AMI and BBB.157,169-172 Two large, older studies from the 1970s and 1980s developed a data bank on patients with BBB in association with MI. One was a large collaborative multicenter study involving five centers,156 and the other was a study conducted at LAC-USCMC.157 Both studies have limitations because (1) the data were obtained retrospectively and (2) at the time, no guidelines existed pertaining to pacemaker insertion, which was performed at the physician’s discretion in all cases. Thus, although these studies are unable to provide definitive answers about the natural history of BBB in association with AMI, they nevertheless do offer valuable clinical information. In the multicenter study reported by Hindman and coworkers,156 high-grade AV block (third- or seconddegree block with a type II pattern) occurred in 55 of 432 patients (22%). To determine which patients were at considerable risk for development of high-grade AV block while hospitalized with AMI, several variables were analyzed. Combinations of the three following ECG findings identified high-risk patients: (1) firstdegree AV block, (2) bilateral BBB (if both bundle branches were involved [e.g., RBBB plus left- or rightaxis deviation] or alternating RBBB and LBBB), and (3) “new” BBB. The absence of all variables or the presence of only one of the three defined variables was associated with a lower risk (10% to 13%) for development of high-grade AV block during hospitalization. The risk was moderate for patients with first-degree AV block with either new BBB or bilateral BBB (19% to 20%), and highest (31% to 38%) for new bilateral BBB regardless of the P-R interval (Fig. 14-16). In the LAC-USCMC study, high-grade AV block occurred in 46 of 257 patients (18%). The absence of all three variables (first-degree AV block, bilateral BBB, and new BBB) or the presence of either bilateral BBB or new BBB or new bilateral BBB was associated with the lowest risk (10% to 18%) for development of highgrade AV block during hospitalization with AMI.157 The risk was moderate for first-degree AV block with or without new BBB (29% to 30%) and highest (50%) for bilateral BBB plus first-degree AV block, regardless of whether the BBB was old or new (Fig. 14-17).

451

Figure 14-16. Venn diagram for 432 patients in the multicenter study, depicting the risk for high-grade atrioventricular block (AVB) in patients with acute myocardial infarction (AMI) and bundle branch block (BBB). (From Hindman MC, Wagner GS, JaRo M, et al: The clinical significance of bundle branch block complicating acute myocardial infarction. 2: Indications for temporary and permanent pacemakers. Circulation 58:689, 1978. Copyright 1978, American Heart Association.)

Figure 14-17. Venn diagram for 257 patients in the Los Angeles County–University of Southern California Medical Center (LAC-USCMC) study, depicting the risk for high-grade atrioventricular block (AVB) in patients with acute myocardial infarction (AMI) and bundle branch block (BBB). (From de Guzman M, Rahimtoola SH: What is the role of pacemakers in patients with coronary artery disease and conduction abnormalities? In Rahimtoola SH: Controversies in Coronary Artery Disease. Philadelphia, FA Davis, 1983.)

Despite some differences in the findings between the two studies, both studies found that the following subgroups of patients were at highest risk for high-grade AV block: (1) those with new bilateral BBB plus first-degree AV block (risks, 38% in the Hindman156 study and 43% in the LAC-USCMC157 study), (2) those with bilateral BBB plus first-degree AV block (risks, 20% and 50%, respectively), and (3) new BBB plus first-degree AV block (risks, 19% and 29%, respectively). Subgroups in whom the findings of the two studies show different risks can be considered to be at moderate risk for high-grade AV

452

Section Two: Clinical Concepts

I

aVR

V1

V4

I

aVR

V1

V4

II

aVL

V2

V5

II

aVL

V2

V5

III

aVF

V3

V6

III

aVF

V3

V6

A

B Figure 14-18. A and B, Electrocardiograms of a patient with anterior acute myocardial infarction (AMI) and development of “new” bilateral bundle branch block (BBB)—left BBB and right BBB with right-axis deviation. This patient had sudden ventricular asystole.

50 Risk of Endpoint (%)

block. These subgroups include patients with (1) new bilateral BBB (risks, 31% and 15%, respectively) (Fig. 14-18), (2) first-degree AV block (risks, 13% and 30%, respectively), and (3) bilateral BBB (risks, 10% and 18%, respectively). The remaining subgroups of patients with AMI and BBB can be considered to be at lowest risk (≤10%) for development of high-grade AV block. The database assembled by the Multicenter Investigation of the Limitation of Infarct Size (MILIS) was used to develop a simplified method of predicting the occurrence of CHB. Data from 698 patients with proved MI were analyzed, and the presence or absence of ECG abnormalities of AV or intraventricular conduction was determined for each patient. Risk factors for development of CHB were as follows: first-degree AV block, Mobitz type I AV block, Mobitz type II AV block, left anterior hemiblock, left posterior hemiblock, RBBB, and LBBB. A risk score for the development of CHB was devised that consisted of the sum of each patient’s individual risk factors. Incidences of CHB of 1.2%, 7.8%, 25%, and 36% were associated with risk scores of 0, 1, 2, and 3 or more, respectively (Fig. 14-19). The risk score was subsequently tested on the published results of six studies for a combined total of 2151 patients.173 The limitations of this scoring system include the lack of differentiation between newly appearing and old BBB, a factor that has been shown to be of predictive value. It is likely that consideration of such factors would further improve the accuracy of the scoring system, but it would also add to its complexity. Another criticism of the scoring system is that it would assign a risk score of only 1 to a patient with isolated Mobitz type II AV block, a disorder usually believed to be highly predictive of progression to CHB. Isolated Mobitz type II AV block is, however, relatively rare. It must be recognized that the algorithms and clinical databases used to estimate the risk for occurrence of CHB in the setting of AMI were developed in the prethrombolytic era and, thus, must be interpreted cautiously when applied to the modern post-AMI patient population. The availability of transcutaneous pacing and early primary percutaneous coronary interventions has had an important effect on the clinical management

Duke data MILIS Prior data 25

0

1

2

ⱖ3

CHB Risk Score Figure 14-19. Comparison of the incidence of complete heart block (CHB) predicted by the CHB risk score method (blue bars), the observed incidence of CHB in the Duke University myocardial infarction database (orange bars), and the observed incidence of CHB or CHB and Mobitz II atrioventricular (AV) block in six reported studies (yellow bars). MILIS, Multicenter Investigation of the Limitation of Infarct Size. (From Lamas GA, Muller JE, Turi AG, et al: A simplified method to predict occurrence of complete heart block during acute myocardial infarction. Am J Cardiol 57:1213, 1986.)

of these patients in the setting of AMI in the current era. The actual clinical decision whether to institute prophylactic temporary pacing should be individualized according to the patient-related risk factors and the available personnel and equipment at each institution. Outcome The short- and long-term mortality and sudden death rates are higher in patients with AMI and BBB (25% to 50%) than in those without BBB (15%).130,162-166,174 The one exception is the isolated finding of left anterior fascicular block in patients with AMI, which appears not to carry an unfavorable prognosis. When the infarction is extensive and produces diffuse conduction system abnormalities progressing to high-grade AV block, it is also extensive enough to damage a large amount of myocardial muscle. Therefore, affected patients often die from pump failure and from ventricular

Chapter 14: Pacing for Atrioventricular Conduction System Disease

tachyarrhythmias, and the adverse prognosis is not necessarily due to development of high-grade AV block. Nevertheless, some of these patients do not die from heart failure or ventricular arrhythmias, and in these patients, the conduction abnormality may be contributory and can be the major cause of death if prophylactic pacing is not undertaken. In some patients, sudden third-degree AV block or asystole is abrupt and fatal if untreated. It is interesting that in the LACUSCMC study, 75% of patients with BBB and AMI had either no heart failure or, at worst, mild heart failure.157 These results as well as those of Hindman and coworkers156 showed that high-grade AV block influenced hospital mortality independently of pump failure. Impact of Thrombolytic Therapy Several prospective trials involving thrombolytic therapy of AMI provide data pertaining to the effect of such therapy on the development of high-grade (second- or third-degree) AV block and BBB.138,142 The study by Clemmensen and associates142 was designed to examine the effect of thrombolytic therapy and adjunctive angioplasty as a treatment strategy for AMI (TAMI trial) after inferior AMI.142 In all patients, treatment was initiated with thrombolytic agents within 6 hours of symptom onset. There were 373 patients with an inferior AMI, of whom 50 (13%) had complete AV block; 54% of these patients had complete AV block on admission. In all but 2 patients, the block was manifest within 72 hours of onset of symptoms. The duration of block was less than 1 hour in 25% and less than 12 hours in 15%; the median duration of block was 2.5 hours. There was no difference in the rate of infarct vessel patency between those with and those without AV block (90% and 91%, respectively). A precipitating clinical event—vessel reperfusion, performance of percutaneous transluminal coronary angioplasty (PTCA)—or vessel reocclusion was identifiable in 38% of instances of complete AV block. At the predischarge angiogram, the vessel patency rate was 11% lower in the group with AV block than in the group without it (71% vs. 82%, respectively). Those in whom AV block developed showed a decrease in ejection fraction between the early post-thrombolytic angiogram and the predischarge angiogram. Also, those who experienced AV block had a higher in-hospital mortality, 10 of 50 (20%) versus 12 of 323 (4%; P < .001). When age, LV ejection fraction in the acute phase, number of diseased vessels, and grade of blood flow through the culprit lesion were entered into a multivariate model, the development of complete AV block still contributed significantly to the risk for in-hospital death. After a median follow-up period of 22 months, mortality rates for patients with and without AV block were equivalent (2%). These data suggest that, compared with the prethrombolytic era, use of thrombolytics and angioplasty has not altered either the incidence of complete AV block nor the associated greater ventricular dysfunction or in-hospital mortality of inferior MI. In another study of 1786 patients with inferior AMI who received recombinant tissue-type plasminogen

453

activator (rt-PA) within 4 hours of symptom onset, high-grade (second- or third-degree) AV block developed in 214 (12%) (TIMI II trial).138 Of the group who had AV block, 113 (6.3% of the total, or 52% of those who ever had AV block) had this finding on admission. The remaining 101 patients (5.7%) experienced heart block during the 24 hours after treatment with thrombolytics. Patients who already had high-grade AV block before receiving thrombolytic therapy tended to be older and had a higher prevalence of cardiogenic shock than those without heart block. Nevertheless, the presence of heart block did not carry a higher 21-day mortality rate independent of other variables such as shock, and the 1-year mortality rate was similar to that in the group without heart block. Patients in this study were randomly assigned to coronary arteriography 18 to 48 hours after admission. Among those who had heart block after admission, the infarct-related artery was less frequently patent than in those without heart block (28 of 39 [72%] vs. 611 of 723 [84.5%]; P = .04]). The RCA was the infarct-related artery more often in patients who had heart block than in those who did not (36 of 69 [92.3%] vs. 542 of 723 [75.1%], respectively; P = .04). Among patients without heart block at the time of hospital admission, death occurred within 48 hours in 4 of 9 patients (44%) with new heart block and in 8 of 68 (12%) without new heart block at 24 hours. The 21-day mortality rate was higher in the group with AV block than in the group without block (10 of 101 [9.9%] versus 35 of 1572 [2.2%], respectively; P < .001), as was the 1-year mortality rate (15 of 101 [14.9%] versus 65 of 1572 [4.2%], respectively; P = .001). A temporary pacemaker was inserted in about one third of patients who had heart block on admission and in almost 30% of patients who experienced heart block after institution of thrombolytic therapy, whereas only 6.5% of patients without heart block received temporary pacemakers. None of the patients who had heart block on admission or who experienced heart block later received permanent pacemakers, but 4 patients without heart block at 24 hours went on to receive permanent pacemakers. Heart block was not listed as a primary or contributing cause of death in any patient. The data from these two studies of thrombolytic therapy suggest that aggressive treatment with thrombolytic agents or thrombolytic therapy plus angioplasty is not associated with a lower incidence of high-grade or complete AV block in patients with inferior AMI than was seen in the prethrombolytic era; the incidence remains about 10% to 13%, with about one half of cases appearing as new AV block during hospitalization. The infarct-related vessel is more often the RCA, and there is a lower vessel patency rate after thrombolysis among patients in whom AV block complicates inferior AMI. In-hospital and early posthospitalization mortality rates are higher in patients with than in those without AV block among patients treated with thrombolytics, with or without angioplasty. It has not been clear, however, that patients with acute AV block continue to be at greater risk of death over the long term if they survive the initial hospitalization. A study that pooled data

454

Section Two: Clinical Concepts

rate of heart failure, greater chance of needing a permanent pacemaker, and a higher 1-year mortality rate.176

50 45

Mortality (%)

40 35

Bundle Branch Block after Recovery from Acute Myocardial Infarction

30 25 20 15 10 5 0 30 days

6 months

All with AVB All without AVB Anterior MI with AVB

1 year

Anterior MI without AVB Inferior MI with AVB Inferior MI without AVB

Figure 14-20. Unadjusted mortality rates in patients with and without atrioventricular block (AVB). MI, myocardial infarction. (From Meine TJ, Al-Khatib SM, Alexander JH, et al: Incidence, predictors, and outcomes of high-degree atrioventricular block complicating acute myocardial infarction treated with thrombolytic therapy. Am Heart J 149:670, 2005.)

from four large randomized clinical trials involving 70,000 patients with AMI treated with thrombolytic therapy evaluated the short- and long-term mortality rates associated with second- and third-degree AV blocks.134 Compared with patients with AMI and no AV block, patients with AMI and AV block were more than three times more likely to die within 30 days and 1.5 times more likely to die during 1 year of follow-up. The higher short- and long-term mortality rates were observed in the setting of inferior as well as anterior AMI (Fig. 14-20). The presumption remains, as before the era of thrombolytic therapy, that the presence or development of AV block is associated with a higher mortality because it tends to indicate the presence of more extensive infarction or injury. In trials conducted with thrombolytic therapy in AMI, BBB is reported present on admission in up to 2% to 4% of patients.173 In the first Global Utilization of Streptokinase and tissue-type plasminogen activator for Occluded Coronary Arteries (GUSTO-I) trial, among 26,003 North American patients, 420 (1.6%) had left (n = 131) or right (n = 289) BBB on admission ECG.175 Interestingly, reversion of BBB occurred in 24% of patients during hospitalization and was associated with a 50% relative risk reduction in 30-day mortality, from 20% to 10%. Prognosis for patients who recovered normal intraventricular conduction (i.e., transient BBB) was similar to that for patients who never had BBB. Another study examined the significance of RBBB in AMI in the prethrombolytic era and compared it with the incidence and prognostic significance of new RBBB in the thrombolytic era.176 In a multicenter prospective study of 1238 patients with 1-year follow-up, a higher rate of new and transient RBBB and lower rate of bifascicular block were found in patients receiving thrombolytic therapy. The overall prognostic implications of RBBB, however, were unchanged and included a higher

The subset of patients with persistent BBB and transient high-grade AV block during AMI are at increased risk of late mortality.156,177,178 It is now recognized that most of these deaths are sudden and result from ventricular tachyarrhythmias. In a previous era, however, whether these patients were at high risk for sudden death from a bradyarrhythmia was debated and controversial.154,172,179-181 A number of investigators suggested that patients with persistent BBB plus transient AV block during the acute infarction had a higher risk of dying suddenly as a result of CHB. These investigators attempted to identify the subset of patients at highest risk of late sudden death due to AV block in order to maximize the therapeutic benefit of permanently implanting a pacemaker. The multicenter study by Hindman and coworkers156 supported the previous reports by Atkins and colleagues177 and Ritter and associates,178 who found that the subset of patients with chronic BBB and transient high-grade AV block during AMI were at increased risk of late sudden death. These data showed that patients who did not receive pacing had a higher incidence of sudden death or recurrent high-grade AV block during follow-up (65%) than those who were given permanent pacing (10%), suggesting that implantation of a permanent pacemaker protected against sudden death in these patients. Waugh and coworkers182 likewise recommended permanent pacemaker therapy to prevent syncope or sudden death in another group of high-risk patients, those with bilateral BBB plus transient highgrade AV block (type II progression). Other studies from the same era, however, questioned whether these patients are at high risk for sudden death from a bradyarrhythmia.154,174,179-181 For example, in the study by Nimetz and associates,172 late sudden death occurred in 4 of 13 (31%) survivors with BBB and second- or third-degree AV block and in 14 of 41 (34%) survivors without AV block. In a study by Ginks and coworkers,154 of patients with anterior MI complicated by CHB with return to normal sinus rhythm but with persistent BBB, 4 of 14 hospital survivors (29%) with anterior MI, persistent BBB, and transient AV block died, and 2 of 4 (50%) with permanent pacemakers died during a follow-up period averaging 49 months. In a study by Murphy and associates180 of patients surviving AMI complicated by BBB, none of the deaths resulted from heart block, even in patients with transient AV block during the AMI. Lie and colleagues181 reported on a group of 47 patients who had survived anterior infarction complicated by BBB and who were kept for 6 weeks in the monitoring area; 17 of the 47 patients (36%) sustained late ventricular fibrillations. Likewise, the Birmingham Trial of permanent pacing in patients with persistent intraventricular defects after AMI showed no significant

455

Chapter 14: Pacing for Atrioventricular Conduction System Disease

difference in survival between patients with and without heart block during up to 5 years of follow-up.183 Finally, in a prospective long-term study by Talwar and colleagues,184 18 patients with anterior AMI, intraventricular defects, and transient complete AV block were monitored for a mean of 2 years; pacemakers were permanently implanted in 8 patients. There was only one death, in the unpaced group, and it was due to a cerebrovascular accident. Clearly, these older studies were limited by the small number of patients with AMI enrolled and monitored and may not be applicable to the current era of post-AMI management. Despite the controversy regarding whether late sudden death in patients with BBB is caused by heart block, permanent ventricular pacing is indicated for transient advanced second- or third-degree infranodal AV block and associated BBB after AMI, according to current practice guidelines (class I indication).127 Role of Electrophysiologic Studies in Atrioventricular Block and Bundle Branch Block in Acute Myocardial Infarction Electrophysiologic studies with recording of the His bundle electrogram (HBE) are not performed routinely today for risk stratification of patients with AV block after AMI. HBE studies after AMI were used primarily in the 1970s and 1980s to identify the sites of AV conduction disturbances, which were shown to be either in the AV node (proximal block) or in the distal conduction system (distal block). The presence of distal block identifies patients at high risk for development of high-grade AV block. In individual cases in which the diagnosis is uncertain (e.g., type I second-degree AV block in patients with BBB) or infranodal or multilevel block is suspected, an electrophysiologic study is helpful in identifying the sites of AV block (Fig. 14-21). In patients with inferior AMI, the site of AV block is usually proximal. Harper and colleagues185 showed that 30 of 32 patients (94%) with inferior AMI and thirddegree AV block had AV nodal block during HBE; the remaining two patients were in normal sinus rhythm during the study and had a normal P-R interval and normal A-H and H-V intervals. Thus, in this group of patients, HBE offered no advantage over conventional ECG criteria in localizing the site of block. In anterior AMI, the block is frequently in the distal conduction system. In the study by Harper and colleagues,185 50% of patients (9 of 18) with BBB and a normal P-R interval on ECG had a prolonged H-V interval.185 Of the 22 patients who experienced AV block and BBB, 5 had proximal block, 14 had distal block, and 3 had both proximal and distal blocks. Thus, in both groups of patients, HBE was the only means of localizing the block in the proximal or distal portion of the conduction system. Distal block indicates disease in either the His bundle or the remaining bundle branches; clinically, this finding is a common antecedent to sudden asystole and a poorer prognosis. Despite the fact that a prolonged H-V interval could identify a group of patients who may be at high risk for high-

I aVF V1 H

48 ms

A

S1

HisM

V

S1 137ms

HisP

V1 S1

S1

S1

S1

S1

S1

HisM 168 ms

178 ms

203 ms

233 ms

97 ms

HisP Figure 14-21. Simultaneous surface electrocardiogram (leads I, aVF, and V1) and intracardiac recordings from the His bundle region (HisM and HisP) in a 65-year-old patient with ischemic cardiac disease and a history of syncope. Left upper panel shows sinus rhythm with normal baseline H-V interval (48 msec) and a narrow QRS complex with a left anterior fascicle block pattern. Right upper panel shows prolonged H-V interval (137 msec) during atrial overdrive pacing (S1S2 = 650 msec). Lower panel shows H-V Wenckebach conduction during atrial overdrive pacing (S1S1 = 600 msec). This case and figure show that serious infranodal conduction disease may be localized to within the His bundle and may be “masked” by a normal QRS as well as a normal P-R interval during sinus rhythm. (Courtesy of Drs. Yanfei Yang and Melvin Scheinman, University of California San Francisco/Cardiac Electrophysiology, San Francisco.)

grade AV block, several studies have shown that it does not help in assessing the short- or long-term prognosis of patients after AMI.185,186 Intracardiac electrophysiologic studies have been evaluated as a means of attempting to predict which patients with MI and BBB are most likely to die. Harper and associates179 reported on 72 patients with AMI complicated by AV block, BBB, or both who underwent His bundle recording or electrophysiologic (HBE) studies during their coronary care unit stay. Thirty of 32 patients (94%) with AV block and narrow QRS complexes had a proximal block. Hospital mortality was low (13%), and HBE studies provided no information additional to that obtained from the surface ECG. Of 18 patients with BBB and a normal P-R interval, 9 had distal block, but there were no hospital deaths in this group of patients. Of 22 patients with BBB and AV block, 5 had proximal block, 14 distal block, and 3 proximal and distal blocks. Hospital mortality in these patients, who progressed to second- or third-degree AV block, was higher (9 of 12 patients, 75%) than in those who remained in first-degree AV block (2 of 10 patients, 20%). Lichstein and colleagues187 and Lie and associates155 also concluded that patients with BBB and AMI who had HBE evidence of a distal block had higher hospital mortality (73% and 81%, respectively) than those with normal H-V intervals (25% and 47%, respectively). On the other hand, in the study by Gould and coworkers,188 the presence or absence of a prolonged H-V interval did not affect mortality.

456

Section Two: Clinical Concepts

Indications for Pacing after Acute Myocardial Infarction Temporary Pacing in Acute Myocardial Infarction The use of temporary transvenous pacing in the postAMI period has diminished in recent years with the greater and more widespread reliance on transcutaneous pacing. Situations in which temporary transvenous pacing is recommended or should be considered according to practice guidelines are listed in Table 14-8.146 Current practice guidelines and recommendations for temporary pacing in the setting of AMI, however, are based primarily on clinical experience rather than well-controlled clinical trials. Essentially no trials have evaluated the risks versus the benefits of temporary pacing during the current era of AMI treatment. Furthermore, RV pacing (VVI) may have potential deleterious hemodynamic effects even when compared with spontaneous intrinsic bradycardic rhythms in the absence of BBB (e.g., sinus node dysfunction or heart block with junctional escape rhythms). In addition, there is little scientific evidence of an advantage of temporary RV pacing over intrinsic rhythm in patients with bradycardia after AMI. Thus, temporary ventricular pacing should not be used for hemodynamic support but should be used primarily as “backup pacing” for prophylactic indications to prevent catastrophic brady-

TABLE 14-8. Recommendations for Temporary Pacing in Patients with Acute Myocardial Infarction Symptomatic AV block

Marked bradycardia (ventricular rate 30 mm Hg at rest or >50 mm Hg provoked

Patient selection

Medically refractory heart failure with significant gradient at rest or provoked Consider the benefit of myectomy

Programming considerations

Use echocardiography to optimize atrioventricular (AV) delay Shortest AV delay that is associated with the greatest gradient reduction Pace/sense offset feature and rateresponsive AV delay

Programming The pacemaker should have the following features: (1) flexible programming to permit the AV delay to be short, (2) a pace/sense offset feature, and (3) a rate-responsive AV delay feature. The programming of the AV delay should be individualized. The AV delay should be the longest interval that results in full apical preexcitation both at rest and during exercise.8,12 The AV delay should be 40 to 80 msec shorter than the PR interval. The AV delay should also be greater than or equal to the P wave duration to prevent interference with left atrial emptying, and thus to avoid a decrease in stroke volume with short AV delay pacing. The use of echocardiography may help select the shortest AV delay that is associated with the greatest gradient reduction and that also does not adversely affect left atrial emptying. In some cases, normal AV conduction is very rapid, possibly leading to failure to capture the ventricle. This in turn may cause an incomplete response to pacing. The problem might be overcome through the use of AV nodal ablation. Gadler and colleagues18 reported that in a series of 6 patients, AV nodal modification could be used to permanently prolong the native AV interval sufficiently to facilitate full apical pre-excitation without using prohibitively short AV delays. The positioning of the ventricular lead is crucial. The hemodynamic outcome is better when pacing is performed from the RV apex than from the RV mid-septum or outflow tract.19,20 Gadler and colleagues20 showed that RV apical pacing is important for gradient reduction and does not reduce cardiac output. During followup visits, pacemaker programming should be optimized to achieve the best possible hemodynamic results. The AV delay may require adjustment with time, especially with a change in medication that slows AV conduction. Finally, in all the pacing trials, optimal medical

Chapter 15: Evolving Indications for Pacing

therapy was established for all patients prior to device implantation.

Long QT Syndrome Clinical Perspective Patients with congenital long QT syndrome (LQTS) have a diverse group of myocardial repolarization disorders that makes their diagnosis, risk stratification, and therapy interesting and challenging. The prevalence of congenital LQTSs in the United States is about 1 in 7000 to 10,000.21 The syndromes are usually caused by autosomal dominant genetic mutations. Long QT types LQT1, LQT2, and LQT3 account for more than 90% of cases of congenital LQTS, with LQT1 and LQT2 more common than LQT3.22,23 Seven causal genes have been identified; five are associated with potassium channels, one with ankyrin B, and one with the sodium channel. These patients are at high risk of syncope and sudden death, usually due to a polymorphic ventricular tachycardia called torsades de pointes. This tachycardia is typically preceded by pauses or bradycardia.

477

Syndrome Registry31 suggest that cardiac pacing, with concomitant β-blocker therapy, may reduce the rates of recurrent syncope and sudden death. The registry reported that in 124 patients who underwent pacing for long QT syndrome, there was an approximately 50% reduction in the incidence of cardiac events. Many of the patients received concomitant β-blocker therapy, making interpretation of results difficult. There were 30 patients, however, who received a pacemaker after failure of β-blockers without a subsequent increase in drug dose. These patients experienced a significant reduction in the incidence of syncope. However, pacing should not be implemented without concomitant βblocker therapy and β-blocker therapy should not be stopped after a pacemaker is implanted. Of the 10 patients in whom β-blockers were withdrawn after pacemaker implantation, 3 died suddenly during the 2 years of follow-up. The benefit of pacing in such patients may be due to the prevention of bradycardia and pauses together with rate-related shortening of the QT interval. Randomized Studies There are no randomized controlled trials of pacing in LQTS.

Rationale for Pacing Pacing is used in LQTS with the intent of preventing potentially arrhythmogenic pauses and bradycardia. Interestingly, there is an association between the specific genotypes and different situational triggers of this arrhythmia.23 Exercise, emotion, and noise are more likely to trigger ventricular tachycardia in LQT1 and LQT2 than in LQT3 (85% and 67% vs. 33%, respectively). Sensitivity of patients with LQT1 to exercise may be due to prolongation of the QT interval during exercise. Patients with LQT3 are at highest risk of event at rest or sleep.23 Arrhythmias that occur during sleep are more often pause-dependent torsades de pointes.24 Schwartz and colleagues23 reported that patients with LQT3 may have fewer events with exercise or stress because they significantly shorten their action potentials during faster rates and therefore become less susceptible to catecholamine-induced arrhythmia. These findings are consistent with the clinical observation that these patients derive no benefit from therapy with β-blockers.25 Although these findings suggest that patients with LQT3 might derive particular benefit from pacing therapy not available to patients with LQT1 and LQT2 mutations, there is no published evidence for this. The pause dependence of torsades de pointes does suggest, however, that permanent pacing might prevent it in many patients. Evidence of Clinical Benefit Observational Studies There is strong evidence that in acquired LQTS, transvenous pacing is a life-saving procedure.26,27 Similarly, observational studies28-30 and the International Long QT

Patient Selection There is no evidence that pacing is more useful in patients with specific genotypes of LQTS or that patients with specific repolarization abnormalities or family histories respond better than others (Table 15-2). Pacing should not be used in patients who have had sustained ventricular tachyarrhythmias, who prob-

TABLE 15-2.

Pacing and Long QT Syndrome

Goal

Prevent bradycardia-related QT prolongation, torsades de pointes VT, and sudden death

Level of evidence for success

Observational only

Consensus recommendations

None

Patient selection

Definite congenital long QT syndromes Patients with pause-related or bradycardia-dependent QT prolongation Syncope due to torsades de pointes Exclude cardiac arrest survivors, who should receive a defibrillator

Programming considerations

Dual-chamber pacemaker Bradycardia support rate >70 bpm Disable features that might result in transient pauses or bradycardias Avoid oversensing Consider post-extrasystole ratesmoothing algorithms

478

Section Two: Clinical Concepts

ably would do better with an implanted defibrillator. Finally, many clinicians choose a DDD defibrillator rather than a DDD pacemaker for patients with LQTS because of the unpredictable risk of sudden death. Programming Insertion of a dual-chamber pacemaker (rather than a single-chamber device) is the norm for patients with LQTS. Although sinus bradycardia is very common in such patients,32 there have been several reports of functional AV block immediately before the onset of torsades de pointes ventricular tachycardia.33,34 Careful attention to pacemaker programming is essential.35 More rapid pacing both shortens the QT interval and reduces dispersion of repolarization. Hence high lower rates—higher than 80 bpm36—may be required, with the attendant disadvantages of reduced battery life and the potential risk of tachycardia-induced cardiomyopathy.37 Patients with a lower rate limit—lower than 70 bpm—have a higher chance of recurrence of symptoms, suggesting that pacing rates lower than 70 bpm may not confer optimal protection.29 Hysteresis and sleep function, which permit slowing of the heart rate below a safe lower rate limit, should be turned off. Meticulous programming is essential to prevent failure to capture or oversensing. Other features that require attention are rate search hysteresis and algorithms that extend the postventricular atrial refractory period, because both can cause pauses. Rate-smoothing algorithms may prevent pausedependent ventricular tachycardia.38 When rate smoothing is programmed on, extrasystoles trigger pacing at a relatively fast rate for a few beats, followed by a gradual slowing until the lower rate limit is reached. This arrangement prevents post-extrasystolic pauses and relative bradycardias.

Sleep Apnea Clinical Perspective The obstructive sleep apnea syndrome is usually defined by daytime sleepiness and other sequelae attributable to frequent obstructive apneas or hypopneas during sleep. It occurs in 9% of adults, most of whom have obstructive rather than central sleep apnea. Patients with this condition are at increased risk of systemic hypertension, cardiovascular disease, including bradyarrhythmias and atrial fibrillation, and exacerbation of congestive heart failure. Several studies report improvement in sleep-disordered breathing after pacemaker implantation in patients with sinus node dysfunction and AV block. Rationale for Pacing The mechanism underlying the apparent improvement with pacing in sleep apnea is unclear. It might be that greater sympathetic activity during pacing might coun-

teract sustained increases in vagal tone or that there is a central mechanism affecting both respiratory rhythm and pharyngeal motor neuron activity.39 The studies reported results with patients who had conventional reasons for pacing, and it is unknown whether patients with sleep apnea who do not have a conventional indication for pacing would have a similar benefit. Evidence of Clinical Benefit Observational Studies Permanent pacing may improve sleep apnea in patients with sinus node dysfunction and AV block.40,41 Mizutani and coworkers40 examined the relationship of pacing mode and sleep in a total of 16 patients (8 men and 8 women; mean age 72 years) with DDD pacemakers. Of these patients, 8 had complete AV block and 8 had sick sinus syndrome. The recording was done twice in VVI and DDD modes. Between VVI mode and DDD mode, sleep latency time, frequency of temporary waking, the number of episodes of apnea, the apneahypopnea index, and efficacy of sleep were all significantly lower when the patients received dual-chamber pacing. There was no significant difference in total sleep time or in total duration of temporary waking between the two groups. This study showed a significant reduction in sleep disturbance when dual-chamber pacing rather than single-chamber ventricular pacing was used. Randomized Studies Atrial overdrive pacing has been assessed in a randomized crossover trial.42 The trial involved 15 patients with sleep apnea confirmed by polysomnography who had a bradycardia indication for pacing. Patients underwent sleep studies on three consecutive nights. The first night provided baseline evaluation; during the second and third nights, the patients were randomly assigned to either atrial overdrive pacing or backup ventricular pacing in a crossover design (Figs. 15-2 and 15-3). The hypopnea index was 9 ± 4 during spontaneous rhythm and 3 ± 3 during atrial overdrive pacing (P < .001). The combined apnea and hypopnea index was 28 ± 22 during spontaneous rhythm and 11 ± 14 (P < .001) during atrial overdrive pacing. Two groups have performed randomized studies of atrially based dual-chamber pacing in patients with sleep apnea. Preliminary results of these two randomized trials do not show any benefit from pacing on indices of sleep disturbance, hypopnea/apnea episodes, or heart failure in patients in whom pacemakers were implanted for the treatment of bradycardia. Although there were differences in the patient populations studied in all these trials, the evidence that pacing improves sleep apnea is at best controversial. Cardiac resynchronization therapy may also reduce central sleep apnea (in patients with heart failure and sleep-related breathing disorders). In a 2004 study, cardiac resynchronization therapy improved sleep apnea and sleep quality in patients with central sleep

Chapter 15: Evolving Indications for Pacing No-Pacing Phase Oxyhemoglobin Saturation (%)

100

80

100 Heart Rate (beats/min) 50

15 Central Apnea 0 15 Obstructive Apnea 0 15 Hypopnea 0

A

1 hr

Pacing Phase Oxyhemoglobin Saturation (%)

100

80

100 Heart Rate (beats/min) 50

15 Central Apnea 0 15 Obstructive Apnea 0 15 Hypopnea 0

B

1 hr

Figure 15-2. Polysomnographic recordings in a patient with predominantly central apnea obtained with the pacemaker programmed to a fixed basic rate of 40 beats per minute (bpm) (no-pacing phase) (A) and with atrial overdrive pacing at 72 bpm (pacing phase) (B).The patient presented with very frequent episodes of central sleep apnea. Each episode of apnea or hypopnea is represented by a vertical line; the height of the line indicates the duration of the episode in seconds. (From Garrigue S, Bordier P, Jais PM, et al: Benefit of atrial pacing in sleep apnea syndrome. N Engl J Med 346:404, 2002.).

479

480

Section Two: Clinical Concepts

Central-Apnea Index

50 35 20

TABLE 15-3.

No-pacing phase Pacing phase

Pacing and Sleep Apnea

Goal

Reduce apneic episodes and their sequelae

Level of evidence for success

Early randomized studies of pacemaker patients

Consensus recommendations

None

6 4

Patient selection

Patients with sleep apnea and another indication for pacing Evidence better for central sleep apnea

Programming considerations

Use atrially based pacing (usually dual-chamber)

12 10 8

2 0

Obstructive-Apnea Index

A

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15

16 14 12 10 8 6 4 2 0 1

2

3

4

B

5

6

7 8 9 10 11 12 13 14 15 Patient No.

Figure 15-3. Effect in 15 patients of atrial overdrive pacing on central sleep apnea (A) and on episodes of obstructive sleep apnea (B). The central-apnea and obstructive-apnea indexes were calculated as the number of episodes divided by the number of hours of sleep. (From Garrigue S, Bordier P, Jais PM, et al: Benefit of atrial pacing in sleep apnea syndrome. N Engl J Med. 346:404, 2002.)

apnea, but not in patients without central sleep apnea.43 This finding is best explained as simply an improvement in congestive heart failure. Patient Selection There is no published evidence that pacing improves sleep apnea in the absence of another indication for pacemaker therapy (Table 15-3). Programming The evidence published to date has centered on dualchamber pacing in patients with sleep apnea.

Neurally Mediated Syncope Syndromes Syncope is the transient loss of consciousness with subsequent complete resolution and without focal neurologic deficits, resulting from cerebral hypoperfusion, and not requiring specific resuscitative measures. The neurally mediated syncope syndromes are a collection of clinical disorders of heart rate and blood pressure

regulation caused by autonomic reflexes.44,45 These often include bradycardia, which has led to attempts to use cardiac pacing as a therapy for carotid sinus syncope (CSS), and vasovagal syncope, the most common form of the neurally mediated syncopes. Vasovagal syncope generally begins at a much younger age, usually occurs in the absence of any underlying structural heart disease, and can have a long and sporadic course lasting decades. Terminology is still variable, and diagnostic synonyms include ventricular syncope, empty heart syndrome, neurally mediated syncope, cardioneurogenic syncope, neurocardiogenic syncope, and neurally mediated hypotension bradycardia. We prefer the term vasovagal syncope, partly for historical deference, partly because of its descriptive accuracy, and partly in the absence of a compelling reason to adopt another term. In contrast to vasovagal syncope, CSS occurs in the elderly and is often associated with hypertension, peripheral vascular disease, or coronary artery disease. There are now several expert consensus conferences and position papers on these syndromes.46,47 Carotid Sinus Syncope Clinical Perspective Carotid sinus syncope is a syndrome of syncope associated with a consistent clinical history, carotid sinus hypersensitivity, and the absence of other potential causes of syncope. Historical features that suggest the diagnosis are syncope or presyncope occurring with carotid sinus stimulation that reproduces clinical symptoms, and fortuitous Holter monitoring or other documentation of asystole during syncope after maneuvers that could presumably stimulate the carotid sinus.48-52 The incidence of CSS is low, being perhaps 35 per million cases per year.53 CSS occurs in older patients, mainly in men. It tends to occur abruptly, with little prodrome, and only one half of patients may recognize a precipitating event. Such events most typically are wearing tight collars, shaving, head turning (as in looking to back up a car), coughing, heavy lifting, and looking up.

Chapter 15: Evolving Indications for Pacing

Symptoms of CSS range from mild presyncope to profound loss of consciousness, occasionally with significant injuries. Some patients may not recall losing consciousness, instead presenting with unexplained falls. In Britain, fits, faints, and falls are often investigated in an integrated setting with a comprehensive clinical pathway. Elderly patients with unexplained falls may have positive responses to carotid sinus massage (CSM), suggesting that CSS is responsible for many unexplained or recurrent falls.54,55 However, physiologic carotid sinus hypersensitivity is far more common than CSS, and care should be taken in the interpretation of these results.

481

of high salt intake, of volume expanders such as fludrocortisone (Florinef),62 or of oral vasopressors such as midodrine hydrochloride (ProAmatine) may be helpful, but such measures are frequently limited in older patients by comorbidities such as hypertension and heart failure. In the pre-pacemaker era, recalcitrant cases of CSS were treated with carotid sinus denervation by surgical technique.63 Surgical sinus denervation is currently reserved for cases that are secondary to head or neck tumors or lymphadenopathy, or is performed in conjunction with carotid endarterectomy or in patients with severe refractory CSS of the pure vasodepressor variety.

Natural History Little is known about the natural history of CSS. Even though it may have a substantial effect on quality of life, it has not been shown to significantly affect mortality, and patients with CSS who receive therapy do not appear to have worse prognoses than the general population. Even in the absence of pacing, only 25% of patients may have a syncope recurrence.56,57 Rationale for Pacing: Physiology The carotid sinus reflex is an integral component of the homeostatic mechanisms of blood pressure regulation.58 Increases in intrasinus pressure stimulate mechanoreceptors, which participate in an afferent arc terminating in the brainstem. The efferent arc travels to peripheral end organs through vagal efferents, which augment cardiac vagal input and slow heart rate, and through the spinal cord to inhibit peripheral sympathetic activity in skeletal vasculature, resulting in peripheral vasodilation. This reflex maintains blood pressure within a narrow range. An abnormal carotid sinus reflex can cause exaggerated responses of heart rate and blood pressure. There is some evidence that the major defect in carotid sinus hypersensitivity does not reside in the carotid sinus, in its neural efferents,60 or in the brainstem. Rather, the neuromuscular structures surrounding the carotid sinus may be involved in CSS. Blanc and colleagues61 found similar results in 30 patients without known carotid sinus hypersensitivity or syncope. Abnormal sternocleidomastoid electromyogram findings were associated with abnormal responses to CSM. It may be that because the denervated sternocleidomastoid muscle cannot provide or contribute information to the central nervous system baroreflex centers, any output from the carotid sinus is inappropriately interpreted as heightened blood pressure. Other Therapies When CSS is the likely cause of syncopal episodes, the initial treatment recommendation should be simple elimination of any recognized maneuvers that may precipitate an event. Discontinuation of the wearing of tight collars and ties and shaving more carefully may help. Hypovolemia should be corrected. The addition

Evidence of Clinical Benefit Observational Studies. Although permanent pacing is almost universally accepted for the treatment of CSS, there are no randomized, placebo-controlled trials of pacemaker therapy, patient selection, or pacemaker mode. A comprehensive summary of studies of pacing for CSS is shown in Table 15-4.56,64-68 Earlier studies tended to be retrospective reports of pacing practices for CSS and therefore were inherently biased toward patients with a clear diagnosis of CSS who would truly benefit from pacing. Randomized Studies. Later prospective, randomized trials examined outcomes on the basis of presence of pacing and mode. A prospective randomized trial from Reggio, Italy, reaffirmed the important role of permanent pacing for CSS.57 Sixty patients with CSS were randomly allocated either to pacing (32 patients) or to no pacing (28 patients). During a follow-up of about 3 years, syncope recurred in 16 patients in the nonpacing group (51%) and in 3 (9%) of the pacing group (P = .002). This observation to some extent confirms the usefulness of pacing for the prevention of CSS, although a significant placebo effect cannot be excluded (see later). Finally, Kenny and associates58a reported the SAFE PACE trial, in which older patients (mean age: 73 ± 10 years, 60% women) with unexplained falls and cardioinhibitory response to carotid sinus massage were randomly assigned to receive either a dual-chamber pacemaker with rate-drop responsiveness or no placebo. This was an open-label trial involving 187 subjects. The investigators found that patients who received a pacemaker had a highly significant (58%) reduction in falls and a 40% reduction in syncope. Injurious events were reduced by 70% (202 in the control group, compared with 61 in the pacing group). Although these results suggest that many unexplained falls in the elderly are due to CSS and that they can be prevented with pacing, one must remember that this was an open-label trial. Trials of similar design involving patients with vasovagal syncope have been fraught with large placebo effects (see later). Finally, Kenny has also suggested that amnesia for loss of consciousness is the most likely reason why elderly patients with CSS and syncope present with falls rather than syncope.68a

482

Section Two: Clinical Concepts

TABLE 15-4.

Clinical Studies of Pacing for Carotid Sinus Syncope*

Study and Treatment

Study Type

Sugrue et al:65 No pacing Pacing

OBS

Huang et al:56 No pacing VVI DDD

OBS

Morley et al:66 VVI DVI DDD

OBS

Brignole et al:68 No pacing Pacing VVI DDD

RCT

Brignole et al:67 VVI DDD

Crossover

Follow-Up (mos)

Patients (n)

SyncopeFree (%)

39 23

11 23

73 91

42 42 42

8 9 4

88 100 100

18 18 18

54 13 3

89 92 66

8.4 7.2 7.2 7.2

19 16 11 5

53 100 100 100

2 2

26 26

92 100

*No differences were statistically significant. The most striking feature is how well patients appear to do regardless of attempted therapy. OBS, observation; RCT, randomized controlled trial.

Patient Selection Careful patient selection may help provide effective and efficient therapy for CSS (Table 15-5). Permanent pacemaker therapy is indicated for patients with recurrent, frequent, or severe CSS, and in particular for predominantly cardioinhibitory syncope.69,70 Predictors of success with permanent pacing include multiple episodes before implantation, episodes that occur while the patient is upright or sitting, and episodes that are preceded by a recognized stimulus.71 When syncope

TABLE 15-5.

Pacing and Carotid Sinus

Syncope Goal

Prevent reflex bradycardia and compensate for reflex hypotension Prevent syncope

Level of evidence for success

Observational and open-label randomized controlled studies No double-blind studies

Consensus recommendations

Class I: Recurrent syncope, with syncope induced by carotid sinus massage Class IIa: Recurrent syncope, with profound bradycardia induced by carotid sinus massage

Patient selection

Syncope and positive carotid sinus massage

Programming considerations

Atrioventricular sequential pacing

recurs after implantation of a permanent pacemaker, it may be due to a major persistent vasodepressor component. Physical Diagnosis with Carotid Sinus Massage The carotid sinus is located high in the neck below the angle of the mandible. Carotid sinus massage (CSM) is contraindicated in the presence of bruits or a history of cerebral vascular disease, transient ischemic attacks, or carotid endarterectomies. Sequential applications of carotid sinus massage to the left and right carotid arteries should be performed with at least 10 to 20 seconds between applications. The duration of CSM should be 5 to 10 seconds, and the massage should be terminated with the onset of characteristic asystole or severe presyncope. In most series, the predominant responses to carotid massage are obtained on the right side.48 CSM should be performed while the patient is both supine and upright—either sitting or while secured safely on a tilt table. It may be difficult to document transient hypotension with standard sphygmomanometric methods, and noninvasive continuous digital plethysmography is often used. Physiologic Responses. Carotid sinus massage (CSM) elicits both cardioinhibitory and vasodepressor responses (Fig. 15-4). A cardioinhibitory response to CSM is defined as 3 seconds or longer of ventricular standstill or asystole. Ventricular asystole usually occurs as a consequence of a sinus pause due to sinus node exit block72 but can be due to AV block as well. A vasodepressor response to CSM, defined as a drop in systolic blood pressure of 50 mm Hg or more during

Chapter 15: Evolving Indications for Pacing

483

Figure 15-4. Combined cardioinhibitory and vasodepressor response to carotid sinus massage (CSM). Note the slow return of blood pressure despite the resolution of asystole. (From Almquist A, Gornick C, Benson W, et al: Carotid sinus hypersensitivity: Evaluation of the vasodepressor component. Circulation 71:927, 1985. Copyright 1985 American Heart Association.)

massage; this response may be difficult to demonstrate in patients who have a significant cardioinhibitory component. In contrast to the induced cardioinhibitory component of carotid sinus hypersensitivity, the vasodepressor response may have a more insidious, slower onset and a more prolonged resolution. Carotid Sinus Hypersensitivity and Carotid Sinus Syncope. Carotid sinus hypersensitivity denotes the abnormal physiologic responses, either cardioinhibitory or vasodepressor or both, to CSM. The presence of asymptomatic carotid sinus hypersensitivity is quite common in populations of older adults. For example, a positive cardioinhibitory response to CSM was noted in 32% of patients undergoing coronary angiography.73 CSS is the syndrome of syncope in association with carotid sinus hypersensitivity and in the absence of other apparent causes of syncope.

Complications. CSM is quite safe if done carefully. It is contraindicated in patients with a history of cerebrovascular disease or carotid bruits, in whom it can cause cerebrovascular accidents. In a review of 3100 episodes of CSM performed on 1600 patients, there were seven complications (0.14%), all of which were neurologic and transient.74 Rare arrhythmic complications include asystole and ventricular fibrillation.75 Programming AAI pacing is contraindicated in CSS because many patients may eventually demonstrate associated reflex AV block.76 In general, patients appear to benefit most from AV sequential pacing, even when a significant component of vasodepressor CSS is present. VVI pacing should not be used in patients with intact VA conduction,77 because of possible pacemaker syndrome. Lack

Section Two: Clinical Concepts

Vasovagal Syncope Clinical Perspective Vasovagal syncope is the most common of the neurally mediated syncopal syndromes (Table 15-6). Most people who faint probably do not seek medical attention for isolated events. Prolonged standing, sight of blood, pain, and fear are common precipitating stimuli for this, the common faint. Patients experience nausea, diaphoresis, pallor, and loss of consciousness as a result of hypotension with or without significant bradycardia. Return to consciousness after seconds to 1 or 2 minutes is the norm. Those with adequate warning may be able to use physical counterpressure maneuvers, or simply sit or lie down, to prevent a full faint. For some patients, however, there is little or no prodrome or recognized precipitating stimulus and no marked bradycardia accompanies the faint.79-83 These patients have sparked interest in the use of permanent pacing as a therapy. Principles of Management of Vasovagal Syncope TABLE 15-6.

Diagnosis and prognosis

Assess patient needs

Conservative advice

Medical options

Permanent pacing

Confirm diagnosis with history, tilt-table tests, loop recorder Assess likelihood of syncope recurrence (>2 spells or recent worsening) Insight into diagnosis Cause of syncope Probability of syncope recurrences Treatment options Limit salt and fluid intake Physical counterpressure maneuvers Driving and reporting to authorities Avoidance and management of triggers Fludrocortisone (weak evidence) Midrodine HCl (good evidence) Serotonin reuptake inhibitors (weak evidence) β-Blockers in patients >42 years old (modest evidence)

Although we prefer the term vasovagal syncope to signify reflex fainting due to bradycardia and/or hypotension, there are numerous synonyms—emotional faint, reflex syncope, empty heart syndrome, neurally mediated syncope, situational syncope, vasomotor syncope, ventricular syncope, neurocardiogenic syncope, hypotension-bradycardia syndrome, and autonomic syncope. In addition, the terms convulsive syncope and venipuncture fits have been used to describe those patients with vasovagal syncope who experience generalized muscle movements that may resemble epilepsy.84,85 Epidemiology of Vasovagal Syncope. About 40% of people faint at least once in their lives, and at least 20% of adults faint more than once.86,87 Fainters usually present first in their teens and twenties and may faint sporadically for decades. This long, usually benign, and sporadic history can make for difficult decisions about therapy. Syncope is responsible for 1% to 6% of emergency room visits and 1% to 3% of hospital admissions.88-90 Tilt table tests are commonly used as a diagnostic tool, although they are limited by difficulties with sensitivity, specificity, reproducibility, and little evidence-based agreement on methodologic details and outcome criteria. Positive tilt table test results (Fig. 15-5) are characterized by presyncope, syncope, bradycardia, and hypotension, and a reproduction of the patient’s pre-syncope symptoms.91,92 Although many patients of all ages simply have vasovagal syncope, clinicians must remain vigilant and look for other causes, including valvular and structural heart disease, sick sinus syndrome, CSS, and orthostatic hypotension. Symptom Burden and Quality of Life. The vasovagal syncope syndrome has an extremely wide range of symptom burden, from a single syncopal spell in a lifetime to daily faints. Some patients have very sporadic presentations, with periods of intense symptoms interspersed with long periods of quiescence. Several observational studies and randomized clinical trials reported that patients have a median of 5 to 15 syncopal spells 120 MAP, mm Hg

of VA conduction at a given point in time, however, does not ensure against its future development. Therefore, we recommend dual-chamber pacemakers for patients with CSS and normal sinus rhythm. Few studies have examined the role of rate-responsive pacing in CSS. Patients are generally older and therefore may have bradycardia comorbidities such as sick sinus syndrome and chronotropic incompetence, either intrinsic or due to medications. Therefore, rateresponsive pacing might be beneficial. Similarly, few studies have prospectively examined pacing with ratedrop or hysteresis capabilities, which has the theoretical advantage of providing rapid, higher-rate, AV sequential pacing to counteract the vasodepressor component during attacks of CSS.78

100 80 60 40 20 120

HR, beats/min

484

100 80 60 40 20 300

Supine

Head up tilt SYNCOPE

400

500

600 700 800 Time, seconds

900 1000 1100

Weak evidence Figure 15-5. Hypotension and bradycardia induced during a positive drug-free passive tilt table test response.

Chapter 15: Evolving Indications for Pacing

and a duration of fainting of 2 to 60 years.93-96 Patients with recurrent syncope are impaired much like those with severe rheumatoid arthritis or chronic low back pain or like psychiatric inpatients.93 The quality of life decreases as the frequency of syncopal spells rises.94 After clinical assessment, many patients continue to do poorly. After 1, 2, and 3 years, 28%, 38%, and 49% of patients, respectively, faint again.97 Interestingly, several studies reported a 90% reduction in the total number of faints in the population after a tilt table test. The cause for this apparently great reduction in syncope frequency after assessment is unknown, but it does leave a large number of patients who request further treatment. Therefore, when assessing syncope patients, clinicians must be alert to the surprising impairment of quality of life that many patients endure, must provide a perspective that lasts decades, and must remember that the clinical state of such a patient will probably fluctuate. Rationale for Pacing Physiology of Vasovagal Syncope. Syncope is a transient loss of neurologic function due to a global reduction of cerebral blood flow. Sudden cessation of cerebral blood flow results in loss of consciousness within 4 to 10 seconds.98 Lesser reductions in blood flow may result in presyncope. Almost all vasovagal syncope occurs while the patient is in an upright position and is usually associated with heightened physiologic or psychological stress, such as prolonged orthostatic stress, arising quickly and walking, pain, fear, or other strong emotion, seeing blood or medical procedures, or heavy exercise. There is no unified hypothesis that explains all the aspects of the pathophysiology of vasovagal syncope. The classic explanation, advanced by Sharpey-Shafer, is based on animal models suggesting that LV mechanoreceptors located primarily in the inferoposterolateral LV can trigger vagally mediated bradycardia and decrease sympathetic output to the peripheral arteriolar vasculature, resulting in vasodilation and hypotension. This model is based on the Bezold-Jarisch reflex: Orthostatic stress increases sympathetic tone, and the resultant increase in β-adrenoreceptor stimulation causes either increased contractility or increased gain in the LV baroreceptors. Thus, a relatively hypovolemic, vigorously contracting ventricle is the presumed trigger of the reflex. There are three main concerns with this model. The first is the uncertainty that the particular animal models and measurements appropriately illuminate the human condition. Second, vasovagal syncope can be provoked in patients with heart transplants, which are centrally denervated for at least one year after transplantation.99,100 Third, there is relatively little evidence from echocardiographic studies of loss of LV volume preceding the onset of the vasovagal reaction on tilt table tests. Other important factors may include reduced plasma volume and salt sensitivity,101 blunting of vagal and sympathetic baroreflex response to orthostatic stress,102,103

485

altered peripheral vascular and endothelial reflex responses,104 and impaired cerebral autoregulation.105,106 Transient hypotension is the most common hemodynamic manifestation of vasovagal syncope. Many patients have inappropriate peripheral sympathetic responses to physiologic and psychological stressors. Under conditions in which the normal response may be vasoconstriction, patients with syncope usually fail to demonstrate it. Abnormalities have been documented in arteriolar vasoconstriction, splenic venoconstriction, and venous capacitance. The ultimate cause of hypotension is an abrupt cessation of vascular sympathetic traffic, causing withdrawal of αadrenergic tone. Evidence for Bradycardia in Vasovagal Syncope. Permanent pacemaker therapy could be effective if bradycardia is a common and symptomatically important feature of vasovagal syncope. The evidence for clinically important bradycardia comes from studies that have used tilt table tests, pacemaker memory, and implantable loop recorders to record heart rate during syncopal spells. Tilt Table Tests. Bradycardia frequently occurs during vasovagal syncope induced by tilt table testing.107,108 The mean heart rate during syncope induced by passive head-up tilt table tests is 30 bpm, and asystole longer than 3 seconds is often documented. However, there is uncertainty about the relationship between hemodynamics observed on tilt table testing and clinical vasovagal syncope. For example, the investigators in the International Study on Syncope of Uncertain Etiology (ISSUE; see later) found no relationship between the heart rate during syncope on tilt table testing and the heart rate during syncope occurring spontaneously in the community.109 Patients with tilt table test–induced bradycardia frequently do not have bradycardia during clinical syncope.102,103 Therefore, although bradycardia is the rule rather than the exception during a positive tilt table test response, the bradycardia evoked on a tilt table test may not resemble the hemodynamics during syncope in that patient during day-to-day living. Pacemaker Memory. Is asystole in patients in the community commonly associated with syncope? Evaluation of frequent fainters with pacemakers programmed to act as event recorders demonstrated that although transient asystole is common during documented syncope, many other asymptomatic asystolic episodes also occurred. Only 0.7% of asystolic events lasting 3 to 6 seconds and 43% of events lasting more than 6 seconds resulted in symptoms of presyncope or syncope.110 Therefore, even asystole of several seconds’ duration does not necessarily cause syncope. Implantable Recorders. The implantable loop recorder (ILR) permits prolonged electrocardiographic monitoring, and its use is a reasonable approach to the diagnosis of patients with infrequent syncope. Current ILRs weigh only 17 grams and have battery lives of 14 months. The electrocardiographic signal is stored in a buffer that can be frozen with a manual activator. The ILR has

486

Section Two: Clinical Concepts

programmable automatic rate detection parameters for high and low rates as well as pause detection. In a Canadian study of 206 patients with syncope, symptoms recurred in 69% of patients; bradycardia was detected more frequently than tachycardia (17% vs. 6%).111,112 The ISSUE investigators studied 111 patients with syncope who had previously undergone tilt table testing. Not all tilt table test responses were positive.113 In both patient groups, those with positive and those with negative tilt table test results, clinical events occurred in 34% of patients group over a follow-up of 3 to 15 months. Marked sinus bradycardia or asystole was detected during syncope (46% and 62%, respectively). Therefore, quite a wide range of bradycardic episodes is reported during syncope: Between 17% and 62% of patients with vasovagal syncope had significant bradycardia during syncope in the ILR studies.111-113 The heart rate response during tilt table testing did not predict spontaneous heart rate response, asystole being observed more frequently than expected on the basis of the tilt table test response. Thus, many patients with positive tilt table test results may have some degree of bradycardia at the time of presyncope or syncope, and pacing may be a plausible treatment. Conservative Therapy Pacing should be tried only in patients with vasovagal syncope who have not shown responses to other treatments or are not candidates for them. Currently, most clinicians first teach the patients about the causes of syncope, encourage fluid and salt intake, and coach physical counterpressure maneuvers. If this initial approach is unsuccessful, attempts at pharmacologic therapy with drugs such as fludrocortisone, midodrine, β-blockers, and serotonin reuptake inhibitors are made. Only after these options have been explored should permanent pacing be considered. Reassurance and Education. Most patients with vasovagal syncope simply require reassurance and education. Many of the patients with more frequent syncope also benefit from these conservative measures. Patients have four broad areas of learning needs—etiology, management, natural history, and prognosis.114 These areas should be covered in efforts to provide patients with reassurance and education. In particular, patients should be taught to recognize their prodromal symptoms and to sit or lie down as quickly as possible to minimize injury during recurrences and lessen the severity of attacks. Salt and Fluids. Blood volume is an important factor in the pathophysiology of vasovagal syncope. Syncope almost always occurs in the upright position, and many patients faint solely from orthostatic stress in situations such as attendance at religious services, cadet parades, outdoor band practices, and showering. Prolonged drug-free head-up tilt provokes syncope in a large number of patients with syncope, depending on the duration and angle of the head-up tilt.115 Finally, many patients report avoiding dietary salt and have a low daily urinary sodium excretion.101

Two studies reported that acute volume loading prevented syncope on tilt table testing in adolescents with vasovagal syncope and a previously positive tilt table test response. A combined 53 of 62 patients had a subsequent negative response after receiving 10 to 15 mL/kg of saline intravenously.116,117 El-Syed and Hainsworth101 reported a small, placebo-controlled study of salt supplements in patients with vasovagal syncope. Eight weeks of salt supplements increased plasma volume and urinary sodium excretion and also increased time to presyncope during a combination of head-up tilt and lower-body negative pressure. Similarly, Claydon and Hainsworth118 showed that orthostatic cerebral and vascular control improved after the administration of 100 mmol/day of slow-release sodium for 2 months. Thus, orthostatic stress causes vasovagal syncope, and salt supplements and acute volume loading help prevent it. Unless contraindicated, patients should drink at least 2 liters of water per day and should consume high-salt meals. Salt tablet administration can be useful in patients with a urinary sodium excretion of less than 170 mmol per day. Salt supplementation should be avoided in patients with a history of hypertension or heart failure. Physical Counterpressure Maneuvers. Physical counterpressure maneuvers may be quite helpful, although no blinded, controlled studies confirming their usefulness have been reported. Patients must have a prodrome long enough that they can react by isometrically tightening muscles using maneuvers such as squatting, leg crossing, and fist clenching. Brignole and associates119 showed that vigorous isometric arm tensing around a small ball raised blood pressure 30% to 40% and prevented 80% to 90% of presyncope and syncope responses on tilt table testing. Similarly, leg crossing raises blood pressure 50% to 60% and eliminates symptoms on tilt table testing.120 There are two remaining problems with counterpressure techniques. Patients must have enough warning to act on them, and symptoms, including syncope, often recur when the muscles are relaxed. Whether these techniques will help during daily life is unknown. A randomized clinical trial, the Physical Counterpressure Trial, has finished and is under review. Medical Therapy There are no therapies that have been proved in large randomized clinical trials to prevent vasovagal syncope. Few have been subjected to rigorous clinical trials, and when interpreting results of open-label studies, one should remember that most patients appear to improve after assessment. There is an estimated 90% reduction in syncope in the population after tilt table testing.121 The four major drug classes that are used are α1-adrenergic agonists, β-adrenergic blockers, serotonin-specific reuptake inhibitors, and salt-retaining mineralocorticoids. Vasopressors. There is reasonable evidence for the effectiveness of the α1-adrenergic agonist midodrine. A pro-drug, midodrine was shown to reduce symptoms

Chapter 15: Evolving Indications for Pacing

of syncope and presyncope in three small, randomized clinical trials. Ward and associates122 reported that compared with placebo, midodrine significantly increased the number of symptom-free days in 16 highly symptomatic patients. Subsequently, Perez-Lugones and colleagues123 reported an open-label randomized study of 61 patients showing that patients who received midodrine had significantly fewer syncopal spells than patients who did not.123 This study was not blinded, and thus, a placebo effect cannot be excluded. Finally, Kaufmann and associates124 reported that midodrine reduced the likelihood of a positive tilt table test result to 17%, compared with a 67% likelihood in patients receiving placebo. The major limitations of midodrine are the need for frequent dosing and its tendency to raise supine blood pressure. The latter side effect is usually seen at higher doses (>30 mg/day). Midodrine should not be used in patients with hypertension. It also causes piloerection and crawling paresthesias in the scalp. Serotonin-Specific Reuptake Inhibitors. Numerous small open-label studies of serotonin-specific reuptake inhibitors in the early 1990s reported that serotoninspecific reuptake inhibitors prevented the induction of syncope on tilt table tests and reduced symptoms in patients in the community. Paroxetine, a selective serotonin reuptake inhibitor, was found to be effective in preventing syncope in one randomized placebo-controlled study125 and numerous case report series. In contrast, Takata and associates126 reported that paroxetine did not block the vasovagal reaction elicited by lowerbody negative pressure. These drugs do not appear to be used widely for the prevention of syncope. b-Blockers. The evidence for the effectiveness of βblocker therapy is mixed. It has a strong physiologic rationale, as well as positive results in two and negative results in one open-label study involving 42 to 153 patients.127-129 Five randomized clinical trials have evaluated the efficacy or effectiveness of β-adrenergic blockers for the prevention of syncope.130-134 Although the results are not completely consistent, they do suggest strongly that β-blocker therapy does not prevent vasovagal syncope. Mahanonda and coworkers,130 studying 42 patients with an unspecified mix of historical presyncope and syncope and positive tilt table test results, reported that after 1 month, 71% of patients receiving atenolol and only 29% of patients receiving placebo reported feeling better and had fewer combined presyncopal and syncopal spells. Madrid and associates131 randomly assigned 50 patients with vasovagal syncope to either treatment with atenolol 50 mg daily, or placebo.131 Patients were monitored for up to 1 year, during which a nonsignificantly higher number of patients in the group taking atenolol had recurrent syncope than those taking placebo. Flevari and colleagues132 performed a prospective, randomized crossover study of nadolol, propranolol, and placebo in 30 patients with recurrent vasovagal syncope and positive tilt table test responses. These investigators found a remarkable 80% to 90% reduction in all measures of presyncope and syncope in all

487

three treatment arms, with no significant difference among the three arms. Ventura and coworkers133 randomly allocated 56 patients with recurrent syncope either to treatment with β-blockers or to no treatment. In a 1-year follow-up, syncope recurred in 71% of untreated patients and only 29% of patients who received β-blockers. The strength of the conclusions of this study is weakened by its lack of placebo control and blinding. Sheldon and colleagues134 performed the Prevention of Syncope Trial, whose design has been described elsewhere. It was a randomized, placebo-controlled, doubleblind trial designed to assess the effects of metoprolol in vasovagal syncope over a 1-year treatment period. Nearly 40% of the subjects had at least one recurrence of syncope during the 1-year observation period, as predicted in the initial power calculations. Metoprolol was no more effective than placebo in preventing vasovagal syncope in the study population as a whole. Taken together with the results of the previous four smaller studies, these data indicate that metoprolol and atenolol, and possibly β-adrenergic receptor blockade in general, are ineffective in preventing vasovagal syncope in the broad patient population. A substudy showed a possible benefit in middle-aged and elderly patients. Fludrocortisone Acetate. Fludrocortisone acetate has mineralocorticoid activity without appreciable glucocorticoid effect at doses up to 0.2 mg, which are the commonly used clinical doses for various disorders.135 The immediate actions of fludrocortisone acetate are sodium and water retention at the expense of urinary potassium excretion. Two open-label trials examined fludrocortisone in neurocardiogenic syncope. Both demonstrated clinical improvement but neither was placebo controlled.136,137 Salim and Di Sessa138 reported a small, placebo-controlled, randomized clinical trial of fludrocortisone acetate (Florinef) in children with vasovagal syncope. Patients taking the drug did significantly worse (P < .04) than those taking placebo. Other drugs have been studied, but not in adequately powered, placebo-controlled, randomized clinical trials. Evidence of Clinical Benefit of Pacing in Vasovagal Syncope Four groups assessed the effect of pacing in preventing syncope induced by tilt table testing (Table 15-7). Fortyone patients with a positive initial tilt table test result and a marked bradycardia underwent a second tilt table test with temporary pacing at rates around 85 to 100 bpm.107,108,110,138-143 Taken together, the studies showed that temporary dual-chamber pacing prevented the development of syncope in 24 of 41 subjects (57%).139-141 However almost all the conscious patients experienced the vasovagal reaction and had significant presyncope. Temporary pacing may be partly effective in preventing vasovagal syncope but does not prevent presyncope. Rate Drop–Responsive Pacemakers. The strategy that emerged in the early 1990s was to attempt to use cardiac pacing not only to overcome transient brady-

488

Section Two: Clinical Concepts

Clinical Studies of Rate Drop–Responsive Pacing in Vasovagal Syncope

TABLE 15-7.

Study

Type 144

Patients (n)

Treatment Tested

Results

Petersen et al

2-period

31

Rate hysteresis

62% stopped fainting

78

2-period

28

Rate drop

78% stopped fainting

Sheldon et al145

2-period

12

Rate smoothing

50% stopped fainting

First North American Vasovagal Pacemaker Study (VPS I)146

Unblinded RCT

54

Rate drop

Hazard ratio down 85% 78% stopped fainting

Vasovagal Syncope International Study (VASIS)147

Unblinded RCT

42

Rate hysteresis

Hazard ratio down 80% 95% stopped fainting

Syncope Diagnosis and Treatment (SYDAT) Study148

Unblinded RCT

93

Rate drop vs. atenolol

Hazard ratio down 87% 95% stopped fainting

Ammirati et al149

Unblinded RCT

20

Rate drop vs. rate hysteresis

Rate drop better than rate hysteresis

VPS II154

Blinded RCT

100

DDD-rate drop vs. ODO

No significant difference

McLeod et al153

Blinded RCT

12

No pacing vs. VVI– rate hysteresis vs. DDD–rate drop

Pacing better than not; pacing modes equivalent

Vasovagal Syncope and Pacing Trial (SYNPACE)155

Blinded RCT

29

DDD–rate drop vs. ODD

No difference

Benditt et al

RCT, randomized controlled trial.

cardia during syncope but also to provide enough heart rate support to compensate for the transient hypotension that is part of the vasovagal reflex. The major issues in the development of pacing strategies are the detection of the onset of syncope and whether pacing is effective at all during the episode. Given the modest efficacy of simple bradycardia support in preventing syncope induced by tilt table tests, investigators soon turned to more flexible and more sophisticated modes of sensing and pacing. Sensors of Transient Heart Rate Drops. Three combined sensing and pacing options have been assessed clinically in some detail: rate hysteresis, rate smoothing, and rate-drop sensing. Rate hysteresis is the oldest. With this approach, pacing at a relatively high rate is triggered by a drop in heart rate below a low detect rate. An example of settings might be a sensing rate of 40 bpm and a pacing response rate of 90 bpm. Typically the high rate pacing might be interrupted periodically to permit the pacemaker to detect an underlying rate, and might cease pacing if the rate is high enough. Rate smoothing is used to avoid abrupt and sustained changes in heart rate. An abrupt drop in heart rate over only one to three beats elicits a pacing response at a slightly lower rate, and this pacing rate gradually slows until it is below the accelerating intrinsic heart rate. Rate drop–responsive pacing is a more sensitive and sophisticated variant of rate hysteresis. An abrupt drop in

heart rate of a programmable magnitude over a programmable duration elicits a burst of high rate pacing for 1 to 2 minutes. The drop in heart rate need only be on the order of 10 bpm, over a few seconds, and the high rate might be around 90 to 110 bpm. Observational Studies. Three groups reported studies on the usefulness of long-term pacing in the prevention of vasovagal syncope. Petersen and colleagues144 reported the first clinical study of dual-chamber pacing with rate hysteresis in 37 patients with syncope. The patients had experienced a median of six syncopal spells, and a positive tilt table test response with bradycardia. Of the 37 patients, 31 received pacemakers with rate hysteresis. Over a mean follow-up of 50 months 62% of the patients remained free of syncope, and the number of syncopal spells in the total population fell from an expected number of 136 to only 11. Benditt and associates78 reported equally encouraging results in a study of 36 patients with predominantly vasovagal syncope. The patients were very symptomatic, with a median of 10 syncopal spells over about 2 years, or about 5 spells yearly. All patients received a pacemaker with rate drop responsiveness. The patients were monitored for a mean of 6 months. During this time, syncope recurred in only 6 patients, compared with expected recurrences in about 30 patients. Therefore, in this relatively short-term study, pacing may have benefited about 80% of patients.

Chapter 15: Evolving Indications for Pacing

Sheldon and associates145 studied 12 extremely symptomatic patients who had had a median syncope frequency of three spells per month.145 All had a positive tilt table test response and recurrent syncope while receiving medical therapy. All received a pacemaker with a rate-smoothing feature but without a high rate response. After implantation of the pacemaker, the actuarial syncope-free survival increased 20-fold, the syncope frequency dropped by 93%, and the improvement in quality of life was highly significant. These were all sequential design studies, with no control for time-dependent effects or for the placebo effect of pacemaker implantation. Open-Label Randomized Studies of Rate-Drop Responsiveness. The first North American Vasovagal Pacemaker Study (VPS I) tested whether permanent pacing with rate-drop responsiveness would reduce the likelihood of syncope in patients with frequent vasovagal syncope.146 Patients were eligible if (1) they had fainted six or more times before tilt table testing or they had fainted within the first year after a positive tilt table test result and (2) they had a predefined degree of bradycardia. Patients were randomly allocated either to receive a pacemaker with automatic rate-drop responsiveness or to receive the best medical therapy according to their treating physicians. The 54 patients were randomly allocated evenly to pacemaker or no pacemaker. There was a lower rate of syncope recurrence in the pacemaker patients (6/27) than in the medically treated patients (19/27). The hazard ratio for a recurrence of syncope in the pacing group compared with the medically treated group was 0.087 (P = .000016). The likelihood of a first syncope recurrence among patients randomly assigned to receive a pacemaker (or not) in VPS I is shown in Figure 15-6.

100

Cumulative Risk (%)

90 80

No Pacemaker

70 60 2P⫽0.000022

50 40 30

Pacemaker

20 10 0 0

3

6

9

12

15

1 11

0 8

Time in Months Number C 27 At Risk P 27

9 21

4 17

2 12

Figure 15-6. Cumulative likelihood of a recurrence of syncope in patients randomly assigned to receive or not to receive a pacemaker in the North American Vasovagal Pacemaker Study. C, control; P, pacemaker. (From Connolly SJ, Sheldon RS, Roberts RS, Gent M: The North American Vasovagal Pacemaker Study: A randomized trial of permanent cardiac pacing for the prevention of vasovagal syncope. J Am Coll Cardiol 33:16, 1999.)

489

Although VPS I was the first randomized trial to show benefit from pacing, it appeared to have a number of limitations. First, it was an open-label study. The investigators believed that there was insufficient evidence of pacemaker effectiveness to enable them to implant a device in all subjects and therefore selected an openlabel design. Second, the patients were highly select: They had all fainted frequently, had positive tilt table test responses with the development of bradycardia, and agreed to participate in a study with only a 50% chance of receiving new therapy—that is, a pacemaker. The pacemaker could have benefited these patients through either conventional bradycardia support or by the sophisticated rate-drop responsiveness algorithm. Finally, medical therapy was not standardized. Ammirati and colleagues148 performed a small, randomized clinical trial comparing rate hysteresis and rate drop responsiveness. Twenty patients with moderately frequent syncope received a pacemaker with either rate hysteresis or rate-drop responsiveness. Three patients with rate hysteresis fainted during follow-up, whereas no patients with rate-drop responsiveness did so (P < .05). This small study suggested that rate-drop responsiveness is superior to rate hysteresis in preventing syncope, and therefore that not all of the pacemaker effect was due to placebo. In the Vasovagal Syncope International Study (VASIS), 19 patients were randomly assigned to receive a dual-chamber pacemaker with rate hysteresis, and 23 patients to no pacemaker implant.147 The patients all had experienced three or more syncopal spells over the preceding 2 years, with a median of 6 spells, and a cardioinhibitory response to tilt table testing. The VASIS subjects had a lower syncope burden than the subjects in VPS I. During a mean follow-up of 3.7 ± 2.2 years there was a lower likelihood of a syncope recurrence in the pacemaker group than in the no pacemaker group (5% vs. 61%; P = .0006). The intention-to-treat results are shown in Figure 15-7. Similar to VPS I, VASIS was an open-label study that involved highly select patients and could not eliminate a placebo effect. The Syncope Diagnosis and Treatment (SYDAT) trial tested whether pacemakers or atenolol better prevented vasovagal syncope.148 Ninety-three patients were randomly assigned to receive a DDD pacemaker with ratedrop responsiveness (n = 46) or to atenolol therapy (n = 47). All patients were older than 35 years, had had more than 3 syncopal spells in the preceding 2 years, and had a positive tilt table test response with a trough heart rate lower than 60 bpm. There was at least one syncope recurrence in 4.3% of the pacing group, compared with 26% of the atenolol group (Odds Ratio 0.13; P = .004). This was another open-label study of pacing in vasovagal syncope using a highly select population. One confounding issue is a possible deleterious effect of the atenolol rather than a beneficial effect of pacemaker implantation. This possibility seems unlikely, given the overall neutral effect of β-blockers in the randomized clinical trials summarized in the preceding section. In summary, observational reports78,144,145 and four randomized, open-label, controlled studies146,147,149 sug-

490

Section Two: Clinical Concepts

Figure 15-7. Cumulative likelihood of a recurrence of syncope in patients randomly assigned to receive or not to receive a pacemaker in the Vasovagal Syncope International Study. (From Sutton R, Brignole M, Menozzi C, et al: Dualchamber pacing in the treatment of neurally mediated tilt-positive cardioinhibitory syncope: Pacemaker versus no therapy: A multicenter randomized study. The Vasovagal Syncope International Study (VASIS) Investigators. Circulation 102:294, 2000.)

gested strongly that patients have less syncope after they receive a permanent pacemaker. But is this effect real? All of these studies were unblinded for both patients and physicians. Syncope is an outcome that can be difficult to verify objectively. Surgical procedures can have a placebo effect.150-152 Patients receiving a pacemaker may have benefited from the psychological effects of receiving a surgical procedure from enthusiastic health professionals. Given these uncertainties, the invasiveness and cost of pacing mandated placebocontrolled or blinded trials to determine the true beneficial effect of pacing in vasovagal syncope. Blinded Randomized Studies of Rate-Drop Responsiveness. Three blinded studies have compared the benefit of rate-drop responsive pacemakers with what was anticipated to be lesser therapy. McLeod and associates153 reported the efficacy of rate drop–responsive, dual-chamber pacing in the prevention of vasovagal syncope in 12 highly symptomatic young children who had frequent syncope associated with asystolic pauses longer than 4 seconds. This was a three-way, doubleblind, randomized crossover study in which the pacemakers were programmed to no active pacing, ventricular pacing with rate hysteresis, or dualchamber pacing with rate-drop responsiveness. The two pacing modes were equivalently more effective than no pacing in preventing syncope, and dualchamber pacing was superior to ventricular pacing in preventing presyncope. This small study concluded that rate drop–responsive pacing was more efficacious than no pacing in preventing vasovagal syncope in children. To ascertain the true therapeutic effect size of permanent pacing in vasovagal syncope, Connolly and associates154 performed the second North American Vasovagal Pacemaker Study (VPS II), a larger double-blind trial.154

The investigators expected that the risk of syncope in the control group would be reduced to some extent by the placebo effect of device implantation, so they increased the study sample size accordingly. VPS II was a multicenter, double-blind, placebo-controlled randomized clinical trial. Patients were eligible if they had recurrent vasovagal syncope with at least six lifetime syncope spells, or at least three spells in the 2 years prior to enrollment, and a positive response on a tilt table test performed according to the protocol in use in each center. A requirement for a specific degree of bradycardia during tilt table testing was not included because trough heart rate during tilt table testing did not correlate in patients with heart rate during clinical syncope,115 and because trough heart rate during tilt table testing did not appear to predict response to pacing.5,24 After receiving dual-chamber pacemakers, all patients were randomly assigned either to dualchamber pacing with rate-drop responsiveness or to sensing without pacing. The patients’ health care providers remained blinded to treatment allocation, except for an unblinded nurse or physician who did all the programming but disclosed no details. The study was designed to have 80% power to detect a 50% relative reduction in the risk of recurrent syncope from a rate of 60% in the control group to 30% in the treatment group. There were 100 patients in the study, who were evenly divided into active pacing and sensing only groups. A total of 38 patients had syncope during the 6-month follow-up period. Twenty-two of 52 patients randomly assigned to sensing-only mode had recurrent syncope within 6 months, compared with 16 of 48 in the active pacing group. The cumulative risk of syncope at 6 months was 40% (95% confidence interval [CI] = 25% to 52%) for the sensing-only group and 31% (95% CI = 17% to 43%) for the rate drop–responsive group (Fig. 15-8). The relative risk reduction in time to syncope with active pacing was 30% (95% CI = –33% to 63%; one-sided P = .14). A retrospective analysis did not identify any variable that predicted benefit from pacing except in patients who received isoproterenol during the tilt table test. Most importantly, VPS II found no statistically significant benefit in favor of pacemaker therapy for the prevention of syncope. The most important difference between the results of VPS I and VPS II is the observed risk of syncope in the nonpacing group. In VPS I, almost 80% of control patients fainted within 6 months, whereas in VPS II, only 41% of control patients fainted within 6 months. In contrast, the 6-month likelihoods of syncope in the patients receiving active pacing therapy were similar in the two studies: 20% in VPS I and 31% in VPS II. The Vasovagal Syncope and Pacing (SYNPACE) trial involved 29 patients who had had a median of 12 lifetime syncopal spells, a positive tilt table test response, and bradycardia during the syncope induced by the tilt table test.155 They received a dual-chamber rate drop– responsive pacemaker and were randomly allocated to either active pacing or no pacing modes. The trial was stopped early, after the VPS II results were released. Thirteen patients had at least one syncope recurrence,

Chapter 15: Evolving Indications for Pacing 1.0 Cumulative Risk

Figure 15-8. Kaplan-Meier plots of the time to the first recurrence of syncope among 48 patients randomly assigned to receive active dual-chamber pacing and 52 patients randomly assigned to receive a pacemaker in the sense-only mode, by intention-to-treat analysis in the second Vasovagal Syncope Pacemaker Study (VPS 2). (From Connolly SJ, Sheldon R, Thorpe KE, et al; on behalf of the VPS II Investigators: The Second Vasovagal Pacemaker Study [VPS II]: A double-blind randomized controlled trial of pacemaker therapy for the prevention of syncope in patients with recurrent severe vasovagal syncope. JAMA 289:2224, 2003.)

491

Only Sensing Without Racing (ODO) Dual-Chamber Pacing (DDD)

0.8 0.6 0.4 0.2 0 0

1

2

3

4

5

6

Months Since Randomization No. at Risk Only Sensing Without Pacing 52 Dual-Chamber Pacing 48

and there was no benefit from active pacing with ratedrop responsiveness. Although extremely underpowered, the SYNPACE trial did not provide any support for the usefulness of pacing in the prevention of vasovagal syncope. Why Did VPS II and SYNPACE Not Show Benefit for Pacing? Both VPS II and the SYNPACE trial did not demonstrate a statistically significant benefit of pacemaker therapy for the prevention of vasovagal syncope. Although the SYNPACE trial was underpowered because of early termination, VPS II was the largest randomized trial of pacemaker therapy for vasovagal syncope, whether open label or double blinded. The researchers made a considerable effort to maintain blindedness, and there were no known protocol violations. This strict adherence sets the standard for randomized trials of treatments for vasovagal syncope. The possible reasons for the negative outcomes are (1) a simple play of chance, (2) early termination of previous studies, (3) the lack of bradycardia on tilt table testing as an inclusion criterion, (4) insufficiently accurate patient selection, (5) inadequate sensing with the rate drop criterion, (6) a placebo effect in open-label studies, and (7) inability to overcome vasodepression with high rate pacing. Play of Chance. VPS II was designed to detect a relative risk reduction of 50% due to pacing. The observed relative risk reduction was 30% with a wide 95% confidence interval. Nevertheless, the very large relative risk reductions in the four unblinded randomized trials are well outside the 95% confidence interval of the relative risk reduction seen in VPS II. This trial did have reasonable power to detect a relative risk reduction of 50%, which may be the minimum effect size that would justify permanent pacing. Early Termination of Open-Label Studies. An important difference between VPS II and the four open-label trials was that three of the four previous trials were stopped prematurely. Early termination of a trial for unexpected efficacy tends to overestimate the treatment effect. Lack of Demonstrated Bradycardia at Baseline. A common speculation about the interpretation of the

37 37

35 35

32 34

31 34

21 18

results of VPS II is that the lack of a requirement for a prespecified bradycardia provoked by tilt table testing may have contributed to selecting a population less amenable to the benefits of pacing. This seems unlikely for several reasons. Patients in the VASIS and the SYDAT studies had more pronounced bradycardia and documented asystole.147 This raises the possibility that the patient selection was insufficiently accurate and that a prespecified bradycardia during syncope on tilt table testing might identify patients more likely to have a response. However the tilt table–induced bradycardia noted during syncope on the baseline test was similar in VPS I, and heart rate changes during tilt table testing neither correlate with heart rate changes during clinical syncope nor predict responses to pacing. As well, tilt table test bradycardias during VPS I and VPS II were quite similar. For example, in VPS I, 12 of 54 patients enrolled (22%) had a rate lower than 40 bpm during tilt, compared with 19 of 100 patients (19%) in VPS II. Therefore, numbers of patients with extreme bradycardia at the time of positive tilt table test response in the two studies were similar, and this minor difference in study design is probably not the reason for the different results of the two studies. Furthermore, the SYNPACE trial, which required a prespecified bradycardia during syncope on tilt table testing, also had a negative result. Indeed, the predictive ability of bradycardia on tilt table testing may not be useful. For example, trough heart rate during tilt table testing does not predict improvement after pacemaker insertion. Petersen and colleagues144 found that the extent of cardioinhibition did not correlate with the level of benefit from permanent pacing. The ISSUE investigators also reported that asystole during tilt table testing did not predict asystole during follow-up143; indeed, some patients with tilt table–induced asystole had syncope without bradycardia (vasodepressor syncope). An asystolic response during recurrent syncope was found even if the patient had a vasodepressor response during the tilt table test. Taken together, these data suggest that the hemodynamic response during tilt table testing (including trough heart rate) does not predict the hemodynamic responses during spontaneous

492

Section Two: Clinical Concepts

syncope and does not predict the response to permanent pacemaker insertion. Strictly speaking, VPS II and the SYNPACE trial simply failed to demonstrate a significant benefit, and it remains possible that there is a small benefit, either in each patient or in a fraction of the population. We recently examined the long-term benefit from pacemaker insertion for vasovagal syncope. This openlabeled, observational study evaluated 40 patients with severe syncope. Thirty-five patients had positive tilt table testing. Pacemakers were programmed to rate drop-responsiveness or rate-smoothing algorithms in equal numbers. Rate-drop responsive pacemakers were implanted in 20 patients, and 20 patients received ratesmoothing pacemakers. Raj and associates156 found an overall 87% decrease in the median frequency of syncope over an approximate 5-year follow-up period (0.46 vs. 0.06 spells/ month; P = .04). However, only 32.5% of their subjects continued to be syncope-free at 60 months. Patients were labeled “responders” to pacing therapy if they experienced reduction of syncopal episodes by more than 75% after implantation. With use of this definition, 55% of patients were long-term “responders.” There was no difference in response according to pacing algorithm, as was previously expected. Patient characteristics of “responders” and “nonresponders” were similar, thus making predictions about which patients might benefit from pacing unlikely. Of particular relevance to this discussion, trough heart rate during tilt table testing was found not to predict longterm response to cardiac pacing. The findings of this study were entirely consistent with the overall negative results of VPS II and the SYNPACE trial. Vasodepression Cannot Be Paced. Prevention of bradycardia is the main physiologic mechanism by which a pacemaker can prevent attacks of syncope. During positive tilt table test responses, however, reductions in blood pressure begin earlier than the development of bradycardia.140,141 Pacing therapy might not help patients with hypotension due to vasodepression even if bradycardia or asystole also occurs at the time of syncope. The results of VPS II and the SYNPACE trial suggest that most episodes of vasovagal syncope may be associated with profound vasodepression as the cause of syncope, rather than simply bradycardia. In this light, pacing per se may simply be ineffective in the setting of profound vasodepression, and future progress in device therapy might best target implantable drug delivery systems. Placebo Effect in Open-Label Studies. The history of attempts to treat patients with implanted devices has other examples of an initial promise of therapeutic success followed by subsequent well-controlled studies with negative results. For example, open-label studies suggested that dual-chamber pacing causes a marked improvement in the hemodynamics and functional status of patients with HCM. Later randomized, controlled, blinded studies revealed evidence of a much smaller effect size. Similarly, preliminary open-label studies suggested that atrially based pacing might

prevent atrial fibrillation, but well-controlled, randomized, crossover trials showed much less benefit from conventional atrially based pacing for the prevention of atrial fibrillation.157,158 Finally, large, open-label studies provided strong evidence for the ability of atrially based pacing to reduce stroke and death in patients with pacemakers. A large, randomized, blinded, controlled study showed that patients with atrially based pacemakers derived no benefit with respect to death, stroke, quality of life, or exercise tolerance for several years after implantation in comparison with patients with singlelead ventricular pacemakers.159 From this experience, it appears that care should be taken in the assessment of results of open-label or nonrandomized pacemaker studies. The placebo effect can be substantial. There are several reasons why pacemaker therapy may be associated with an initial spurious benefit. First, patients receiving expensive or invasive therapy may be loath to admit the possibility that such a therapy might be ineffective. Particularly with surgical procedures, the placebo effect can be pronounced.150-152 Second, many patients with vasovagal syncope appear to improve spontaneously after tilt table testing.90,95,96 This effect may account for up to 90% of the apparent benefit. The mechanism is unknown but may involve the counseling received at the time of the clinic visit, a regression to the mean, and the sporadic nature of the timing of manifestations of vasovagal syncope. This effect is similar in magnitude to the beneficial effect of pacing in sequential design trials. It is possible, in the unblinded studies, that some patients, hoping to have received a pacemaker and disappointed by being randomly allocated not to receive one, may have been more prone to report syncope. Conversely, patients receiving a pacemaker may have benefited from the psychological effects of receiving a surgical procedure from enthusiastic health care professionals. The doubleblind trial design, to a considerable extent, removes this type of potential bias. Patient Selection Patients should have a definite history of vasovagal syncope based on a positive tilt table test response, suggestive loop recorder findings, or scrupulous history (Table 15-8). Given the weakness of the evidence supporting the efficacy of pacing, it is reasonable to institute pacing only in patients with documented profound bradycardia or asystole during syncope. Only patients at high risk of syncope recurrences should undergo pacing. The frequency and number of syncopal spells preceding positive tilt table test responses are independent risk factors that predict an early recurrence of syncope after the test. Criteria such as having more than six syncopal spells over any duration of time, or at least two syncopal spells in the preceding year, or any recurrence of syncope within a year of tilt table testing, generally select patients who have a more than 50% risk of at least one syncopal spell in the next 2 years. Otherwise, similar patients with syncope and either negative or positive tilt table test responses have similar likelihoods of syncope after assessment.160,161

Chapter 15: Evolving Indications for Pacing TABLE 15-8.

Pacing and Vasovagal Syncope

Goal

Prevent reflex bradycardia and compensate for reflex hypotension Prevent syncope

Level of evidence for success

Limited evidence for benefit based on double-blind randomized controlled trials May be a subset of patients with proved bradycardia who benefit

Consensus recommendations

Class IIa. Recurrent vasovagal syncope with clinically documented bradycardia, or bradycardia induced on tilt test

Patient selection

Medically refractory, frequent, disabling vasovagal syncope Documented pauses during syncope Tilt test results not helpful

Programming considerations

Dual-chamber pacemaker Benefit from specific sensor to drive rate response or pacing algorithm (rate-drop response, ventricular impedance) not proved

The result of the tilt table test (negative versus positive) does not predict subsequent clinical outcome. Similarly, the lowest heart rate (including asystole) during tilt table testing does not predict the eventual likelihood of syncope in clinical follow-up.115,144 Programming There is an inexorable trade-off between sensitivity and specificity in programming rate-responsive pacemakers to detect the early stages of syncope. The main detection features are the range over which heart rate must fall, the time interval during which the drop must take place, and the number of confirmation beats below the minimum detection heart rate. Decreasing the specified heart rate range or number of confirmation beats improves the sensitivity, as does lengthening the time during which the heart rate fall must occur. Generally, the pacemakers greatly overdetect, with therapies delivered many times daily. This arrangement is usually well tolerated, particularly if the patient perceives a benefit in syncope prevention, but in some patients, the palpitations are intolerable. This can be particularly true at night, when respiratory sinus arrhythmia is larger and the patient is quieter. Although there is an understandable tendency to program the rate-drop feature “off” at night, doing so affords the patient no potential protection at night should he or she arise. This can be a problem, because patients have usually not had any fluids for hours and can have syncope provoked by either orthostatic changes or micturition. Pacing therapy is usually a relative high rate burst. The pacemaker should be programmed to deliver atrially based pacing at rates of 90 to 110 bpm for 1 to 2 minutes. There is no evidence of greater benefit from any particular rate or duration.

493

Evolving Paradigms and Technology Contractility Sensors. The search continues for alternative sensing strategies, such as QT interval, respiratory volume or frequency, RV pressure transduction (dP/dt), and indices of contractility. These are intended to sense either early hypovolemia or early rises in sympathetic activity that may precede frank syncope. There is some evidence that contractility can be estimated with measures of endocardial acceleration, with use of a microaccelerometer in the pacemaker lead to estimate RV myocardial contractility,162,163 or with intracardiac impedance measurements.164,165 The theory behind contractility sensors is that vasovagal syncope might be preceded by small but significant increases in contractility due to a sympathetic surge. These devices increase pacing rates in response to increases in apparent contractility, and then slowly decrease their rates after contractility subsides toward baseline. Discouragingly, Brignole and colleagues162 found that endocardial acceleration did not predict the occurrence of tilt table– induced syncope. The use of closed-loop stimulation (CLS) was evaluated in a preliminary study in 2002.165 CLS pacemaker technology reacts to a change in the RV intracardiac impedance, which is believed to be a surrogate measure of contractility and, therefore, sympathetic tone. This study of 22 patients demonstrated that syncope is predicted by impedance changes and could be prevented on tilt table testing of patients with cardioinhibitory syncope. A subsequent study used CLS prospectively in 34 patients with recurrent vasovagal syncope.166 Over 12 to 50 months of follow-up, 30 of 34 patients had not experienced a recurrent syncopal event. On the basis of this pilot study, a larger, multicenter randomized trial, the Inotropy Controlled Pacing in Vasovagal Syncope (INVASY) study, was planned and partially carried out.167 Twenty-six patients with recurrent vasovagal syncope and a positive tilt table test response with induced bradycardia received dual-chamber pacemakers with CLS. Asymmetric randomization assigned patients to simple DDI pacing (9 of 26) and/or to dualchamber pacing with CLS. This was an open-label trial. It was stopped early, after a preliminary analysis (Fig. 15-9) at a mean of 19 months showed that 7 of the 9 patients in the DDI group and none of the 17 patients in the CLS group had syncope recurrences (P < .0001). These positive results suggest three possible conclusions: (1) pacing can be useful in patients with vasovagal syncope, (2) CLS based on RV impedance changes is effective sensing for syncope, and (3) this may be the placebo effect once again. A larger and properly blinded study is being planned. Automatic Drug Delivery Systems. Current pacemaker therapies focus only on heart rate support. This might not be as useful in patients with a predominantly vasodepressor response. Giada and coworkers168 described a study assessing a novel implantable system that delivers phenylephrine when activated at the onset of syncope (prodrome with a drop in blood pressure). When treated with the phenylephrine, 15 of 16 patients had an immediate rise in blood pressure and a termination

494

Section Two: Clinical Concepts Figure 15-9. Probability of remaining free of syncopal recurrences according to KaplanMeier estimation in 41 patients in the closedloop stimulation arm and 9 patients in the control group in the Inotropy Controlled Pacing in Vasovagal Syncope (INVASY) study. (From Occhetta E, Bortnik M, Audoglio R, Vassanelli C; INVASY Study Investigators: Closed loop stimulation in prevention of vasovagal syncope. Inotropy Controlled Pacing in Vasovagal Syncope (INVASY): A multicentre randomized, single blind, controlled study. Europace 6:538, 2004.)

CLS 100

% Syncope-free

80 P⬍0.0001 60

40 Placebo 20 Time since randomization

N. of pts in CLS arm

3m 6m 9m 1y

2y

3y

50

29

4

2

0

0

50

N. of pts 9 in placebo arm

5

4

3

of their tilt table–induced syncope, despite ongoing tilt table testing. In contrast, none of the patients was able to abort the episodes when a placebo infusion was delivered. The one patient who fainted despite the phenylephrine experienced a severe cardioinhibitory response to the tilt table test. This small study suggests the promise of a combined approach using pacing for the cardioinhibitory component and immediate pharmacologic support for the vasodepressor component of syncope. Implications for Future Trials Several lessons can be learned from this discussion. First, clinical trials of vasovagal syncope must be placebo-controlled and double-blinded to mitigate the sizable potential for the placebo effect. Second, the potential for the placebo effect must be considered when population sample is calculated. Third, initial efforts might be directed toward understanding whether there are physiologic differences between patients who faint during pacing and those who do not. Fourth, accurate patient selection may be important. Guidelines for Pacing in Vasovagal Syncope At the least, pacing should not be used early in patients with vasovagal syncope. First, it should be reserved for patients with frequent, highly symptomatic vasovagal syncope whose quality of life is markedly diminished and whose syncope has not responded to lifestyle and dietary changes, education, reassurance, the use of counterpressure maneuvers,46,47 and at least three medication attempts. Second, pacing might be more effectively used in patients with documented asystole at the time of syncope. Third, all patients should have frank and detailed education about the limited (at best) evidence of its efficacy. In light of this, a revision of the ACC/AHA/NASPE pacing guidelines issued in 2002 should be reconsidered,17 because they recommended

permanent pacemaker implantation for patients with symptomatic and recurrent neurocardiogenic syncope associated with bradycardia documented spontaneously or at the time of tilt table testing. REFERENCES 1. Maron BJ: Hypertrophic cardiomyopathy. Lancet 350:127, 1997. 2. Maron BJ, Shen WK, Link, MS, et al: Efficacy of implantable cardioverter defibrillation for the prevention of sudden death with hypertrophic cardiomyopathy. N Engl J Med 342:365, 2000. 3. Elliott PM, Sharma S, Varnava, A, et al: Survival after cardiac arrest or sustained ventricular tachycardia in patients with hypertrophic cardiomyopathy. J Am Coll Cardiol 33:1596, 1999. 4. Silka MJ, Kron J, Dunnigan A, Dick M 2nd: Sudden cardiac death and the use of implantable cardioverter defibrillator in pediatric patients: The Pediatric Electrophysiology Society. Circulation 87:800, 1993. 5. McDonald KM, Maurer B: Permanent pacing as treatment for hypertrophic cardiomyopathy. Am J Cardiol 68:108, 1991. 6. McDonald K, McWilliams E, O`Keeffe B, et al: Functional assessment of patients treated with permanent dual chamber pacing as a primary treatment for hypertrophic cardiomyopathy. Ann Thorac Surg 47:236, 1988. 7. Kappenberger L, Linde C, Daubert C, et al: Pacing in hypertrophic obstructive cardiomyopathy: A randomized crossover study. Eur Heart J 18:1249, 1997. 8. Fananapazir L, Cannon RO, Tripodi D, et al: Impact of dual chamber permanent pacing in patients with obstructive hypertrophic cardiomyopathy with symptoms refractory to verapamil and beta-blocker therapy. Circulation 85:2149, 1992. 9. Slade AKB, Sadoul N, Shapiro L, et al: DDD pacing in hypertrophic cardiomyopathy, a multicenter clinical experience. Heart 75:44, 1996. 10. Gadler F, Linde C, Juhli-Dannfeldt A, et al: Long-term effects of dual chamber pacing in patients with hypertrophic cardiomyopathy without outflow tract obstruction at rest. Eur Heart J 18:636, 1997. 11. Pak PH, Maughan WL, Baughman KL, et al: Mechanism of acute mechanical benefit from VDD pacing in hypertrophied heart: Similarity of responses in hypertrophic cardiomyopathy and hypertensive heart disease. Circulation 98:242, 1998.

Chapter 15: Evolving Indications for Pacing 12. Jeanrenaud X, Goy J, Kappenberger L, et al: Effects of dualchamber pacing in hypertrophic obstructive cardiomyopathy. Lancet 339:1318, 1992. 13. Fananapazir L, Epstein ND, Curiel RV, et al: Long-term results of dual-chamber (DDD) pacing in obstructive hypertrophic cardiomyopathy: Evidence for progressive symptomatic and hemodynamic improvement and reduction of left ventricular hypertrophy. Circulation 90:2731, 1994. 14. Nishimura RA Trusty JM, Hayes DL, et al: Dual-chamber pacing for hypertrophic obstructive cardiomyopathy: A randomized, double-blind crossover trial. J Am Coll Cardiol 29:435, 1997. 15. Maron BJ, Nishimura RA, McKenna WJ, et al: Assessment of permanent dual-chamber pacing as a treatment for drug-refractory symptomatic patients with obstructive hypertrophic cardiomyopathy: A randomized, double-blind, crossover study (M-PATHY). Circulation 99:2927, 1999. 16. Ommen SR, Nishimura RA, Squires RW, et al: Comparison of dual-chamber pacing versus septal myectomy for treatment of patients with hypertrophic obstructive cardiomyopathy: A comparison of objective hemodynamic and exercise end points. J Am Coll Cardiol 34:191, 1999. 17. Gregoratos G, Abrams J, Epstein AE, et al; American College of Cardiology/American Heart Association Task Force on Practice Guidelines/North American Society for Pacing and Electrophysiology Committee: ACC/AHA/NASPE 2002 guideline update for implantation of cardiac pacemakers and antiarrhythmia devices: Summary article: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/NASPE Committee to Update the 1998 Pacemaker Guidelines). Circulation 106:2145, 2002. 18. Gadler F, Linde C, Darpo B: Modification of atrioventricular conduction as adjunct therapy for pacemaker-treated patients with hypertrophic obstructive cardiomyopathy. Eur Heart J 19:132, 1998. 19. Chang AC, Atiga WL, McArevey D, et al: Relief of left ventricular outflow tract obstruction following inadvertent left ventricular apical pacing in a patient with hypertrophic cardiomyopathy. Pacing Clin Electrophysiol 18:1450, 1995. 20. Gadler F, Linde C, Juhli-Dannfeldt A, et al: Influence of right ventricular pacing site on left ventricular outflow tract obstruction in patients with hypertrophic obstructive cardiomyopathy. J Am Coll Cardiol 27:1219, 1996. 21. Schwartz PJ: The long QT syndrome. Curr Probl Cardiol 22:297, 1997. 22. Splawski I, Shen J, Timothy KW, et al: Spectrum of mutations in long-QT syndrome genes: KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 102:1178, 2000. 23. Schwartz PJ, Prior SG, Spazzolini C, et al: Genotype-phenotype correlation in long-QT syndrome: Gene-specific triggers for lifethreatening arrhythmia. Circulation 103:89, 2000. 24. Benhorin J, Medina A: Congenital long QT syndrome. N Engl J Med 336:1568, 1997. 25. Moss AJ, Zareba W, Hall WJ, et al: Effectiveness and limitations of beta-blocker therapy in congenital long-QT syndrome. Circulation 101:616, 2000. 26. Sclarovsky S, Strsberg B, Lewin R, et al: Polymorphous ventricular tachycardia: Clinical features and treatment. Am J Cardiol 44:339, 1979. 27. DiSegni E, Klein HO, David D, et al: Overdrive pacing in quinidine syncope and other long QT-interval syndromes. Arch Intern Med 140:1036, 1980. 28. Elder M, Griffin JG, Hare V, et al: Combined use of beta adrenergic blocking agents and long-term cardiac pacing for patients with long QT syndrome. J Am Coll Cardiol 20:830, 1992. 29. Moss AJ, Liu JE, Gottlieb S, et al: Efficacy of permanent pacing in the long QT syndrome. Circulation 84:1524, 1991. 30. Eldar M, Griffin JG, Abbott JA, et al: Permanent cardiac pacing in patients with the long QT syndrome. J Am Coll Cardiol 10:600, 1987.

495

31. Zareba W, Priori SG, et al: Permanent pacing in the long QT syndrome patients. Pacing Clin Electrophysiol 20:1097, 1997. 32. Schwartz PJ, Periti M, Malliani A: The long QT syndrome. Am Heart J 89:378, 1975. 33. Van Hare GF, Franz MR, Roge C, et al: Persistent functional atrioventricular block in two patients with prolonged QT interval: Elucidation of the mechanism of block. Pacing Clin Electrophysiol 13:608, 1990. 34. Scott W, Dick M: Two to one atrioventricular block in infants with congenital long QT syndrome. Am J Cardiol 60:1409, 1987. 35. Viskin S: Cardiac pacing in the long QT syndrome: Review of available data and practical recommendation. J Cardiovasc Electrophysiol 11:593, 1999. 36. Viskin S, Alla SR, Baron HV, et al: Mode of onset of torsades de points in congenital long QT. J Am Coll Cardiol 28:1262, 1996. 37. Klein H, Levi A, Kaplinsky E, et al: Congenital long QT syndrome: Deleterious effect of long-term high rate ventricular pacing and definitive treatment by cardiac transplantation. Am Heart J 132:1079, 1996. 38. Viskin S, Fish R, Roth A, Copperman Y: Prevention of torsades de points in the congenital long QT syndrome: Use of a pause prevention pacing algorithm. Heart 79:417, 1998. 39. Gottlieb DJ: Cardiac pacing—a novel therapy for sleep apnea? N Engl J Med 346:444, 2002. 40. Mizutani N, Waseda K, Asai K, et al: Effect of pacing mode on sleep disturbance. J Artif Organs 6:106, 2003. 41. Kato I, Shiomi T, Sasanabe R, et al: Effect of physiological cardiac pacing on sleep-disordered breathing in patients with chronic bradydysrhythmias. Psychiatry Clin Neurosci 55:257, 2001. 42. Garrigue S, Bordier P, Jais O, et al: Benefit of atrial pacing in sleep apnea syndrome. N Engl J Med 346:404, 2002. 43. Sinha AM, Skobel EC, Breithardt OA, et al: Cardiac resynchronization therapy improves central sleep apnea and CheyneStoke respiration in patients with chronic heart failure. J Am Coll Cardiol 44:68, 2004. 44. Benditt DG, Remole S, Milstein S, et al: Causes, clinical evaluation and current therapy. Ann Rev Med 43:283, 1992. 45. Brignole M, Alboni P, Benditt DG, et al: Guidelines on management (diagnosis and treatment) of syncope: Update 2004: Executive Summary. Eur Heart J 25:2054, 2004. 46. Brignole M, Alboni P, Benditt DG, et al; Task Force on Syncope, European Society of Cardiology: Guidelines on management (diagnosis and treatment) of syncope: Executive Summary. Eur Heart J 22:1256, 2001. 47. Gregoratos G, Cheitlin MD, Conill A, et al: ACC/AHA/ Guidelines for Implantation of Cardiac Pacemakers and Antiarrhythmic Devices: Executive Summary—a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Pacemaker Implantation). Circulation 97:1325, 1998. 48. Thomas JE: Hyperactive carotid sinus reflex and CSS. Mayo Clin Proc 13:2065, 1969. 49. Volkmann H, Schnerch B, Kuhnert B: Diagnosis value of carotid sinus hypersensitivity. Pacing Clin Electrophysiol 13:2065, 1990. 50. Hartzler GO, Maloney JD: Cardioinhibitory carotid sinus hypersensitivity. Arch Intern Med 137:727, 1977. 51. Weiss S, Baker JP: The carotid sinus reflex in health and disease: Its role in the causation of fainting and convulsions. Medicine 12:297, 1933. 52. Zee-Cheng CS, Gibbs HR: Pure vasodepressor carotid sinus hypersensitivity. Am J Med 81:1095, 1986. 53. Morley CA, Sutton R: Carotid sinus syncope. Int J Cardiol 6:287, 1984. 54. Kenny RA, Traynor G: Carotid sinus syndrome: Clinical characteristics in elderly patients. Age Ageing 20:499, 1991.

496

Section Two: Clinical Concepts

55. Richardson DA, Bexton RS, Shaw FE, Kenny RA: Prevalence of cardioinhibitory carotid sinus hypersensitivity in patients 50 years or over presenting to the accident and emergency department with “unexplained” or “recurrent” falls. Pacing Clin Electrophysiol 15:820, 1997. 56. Huang SKS, Ezi MD, Hauser RG: Carotid sinus hypersensitivity in patients with unexplained syncope: Clinical, electrophysiological, and long term follow-up observations. Am Heart J 116:989, 1988. 57. Brignole M, Menossi C, Lolli G: Long-term outcome of paced and non-paced patients with severe carotid sinus syndrome. Am J Cardiol 69:1039, 1992. 58. Lown B, Levine SA: The carotid sinus: Clinical value of its stimulation. Circulation 23:766, 1961. 58a. Kenny RA, Richardson DA, Steen N, et al: Carotid sinus syndrome: A modifiable risk factor for nonaccidental falls in older adults (SAFE PACE). J Am Coll Cardiol 38:1491, 2001. 59. Baig MW, Kaye GC, Perrins EJ: Can central neuropeptides be implicated in carotid sinus reflex hypersensitivity? Med Hypotheses 28:255, 1989. 60. Strasburg B, Sagie A, Erdman S: Carotid sinus hypersensitivity and carotid sinus syndrome. Prog Cardiovasc Dis 31:376, 1989. 61. Blanc JJ, L`Heveder G, Mansourati J: Assessment of a newly recognized association, carotid sinus hypersensitivity and denervation of sternocleidomastoid muscles. Circulation 95:2548, 1997. 62. DaCosta D, Mclntoch S, Kenny RA: Benefits of fludrocortisone in the treatment of symptomatic vasodepressor carotid sinus syndrome. Br Heart J 69:308, 1993. 63. Trout HH, Brown LL, Thompson JE: Carotid sinus syndrome: Treatment by carotid sinus denervation. Ann Surg 189:575, 1979. 64. Katritsis D, Ward DE, Camm AJ: Can we treat carotid sinus syndrome? Pacing Clin Electrophysiol 14:1367, 1991. 65. Sugrue DD, Gersh BJ, Holmes DR: Symptomatic “isolated” carotid sinus hypersensitivity: Natural history and results of treatment with anticholinergic drugs or pacemaker. J Am Coll Cardiol 158:1986, 1986. 66. Morley CA, Perrins EJ, Grant P: Carotid sinus syncope treated by pacing. Br Heart J 47:411, 1982. 67. Brignole M, Menozzi C, Lolli G: Validation of a method for choice of pacing mode in carotid sinus syndrome with or without sinus bradycardia. Pacing Clin Electrophysiol 14:196, 1991. 68. Brignole M, Menossi C, Lolli G: Natural and unnatural history of patients with severe carotid sinus hypersensitivity: A preliminary study. Pacing Clin Electrophysiol 11:1678, 1988. 68a. Parry SW, Steen IN, Bapfist M, Kenny RA: Amnesia for loss of consciousness in carotid sinus syncope: Implications for presentation with falls. J Am Coll Cardiol 45:1840, 2005. 69. Madigan NP, Flaker GC, Curtis JJ: Carotid sinus hypersensitivity: Beneficial effects of dual-chamber pacing. Am J Cardiol 53:1034, 1984. 70. Stryjer D, Friedensohn A, Schlesinger Z: Ventricular pacing as the preferable mode for long-term pacing in patients with carotid sinus syncope of the cardio inhibitory type. Pacing Clin Electrophysiol 9:705, 1986. 71. Walter F, Crawley IS, Dorney ER: Carotid sinus hypersensitivity and syncope. Am J Cardiol 42:396, 1978. 72. Gang ES, Oseran DS, Mandel WJ: Sinus node electrogram in patients with the hypersensitivity carotid syndrome. J Am Coll Cardiol 5:1484, 1985. 73. Brown KA, Maloney JD, Smith HC: Carotid sinus reflex in patients undergoing coronary angiography: Relationship of degree and location of coronary artery disease to response to carotid sinus massage. Circulation 62:697, 1980. 74. Munro NC, McLintoch S, Lawson J: Incidence of complications after carotid sinus massage in older patients with syncope. J Am Geriatr Soc 42:1248, 1994.

75. Alexander S, Ding WC: Fatal ventricular fibrillation during carotid sinus stimulation. Am J Cardiol 18:289, 1966. 76. Probst P, Muhlberger V, Lederbauer M, et al: Electrophysiologic findings in carotid sinus massage. Pacing Clin Electrophysiol 6:689, 1983. 77. Alicandri C, Fouad FM, Tarazi RC: Three cases of hypotension and syncope with ventricular pacing. Am J Cardiol 42:137, 1978. 78. Benditt DG, Sutton R, Gammage MD, et al: Clinical experience with Thera DR rate-drop response pacing algorithm in carotid sinus syndrome and vasovagal syncope. Pacing Clin Electrophysiol 20:832, 1997. 79. Maloney JD, Jaeger FJ, Fouad-Tarazi FM: Malignant vasovagal syncope: Prolonged asystole provoked by head-up tilt. Case report and review of diagnosis, pathophysiology and therapy. Cleve Clin J Med 55:543, 1988. 80. Sutton R: Vasovagal syncope: Could it be malignant? Eur Heart J 2:89, 1992. 81. Fitzpatrick AP, Ahmed R, Williams S: A randomized trial of medical therapy in “malignant vasovagal syndrome” or “ neurally-mediated bradycardia/hypotension syndrome.” Eur J Cardiac Pacing Electrophysiol 2:99, 1991. 82. Fitzpatrick AP, Sutton R: Tilting toward a diagnosis in recurrent unexplained syncope. Lancet 1(8639):658, 1989. 83. Grubb BP, Temesy-Armos P, Moore J: Head-upright tilt-table testing in the evaluation and management of the malignant vasovagal syndrome. Am J Cardiol 69:904, 1990. 84. Jaeger FJ, Schneider L, Maloney JD: Vasovagal syncope: Diagnostic role of head-up tilt test in patients with positive ocular compression test. Pacing Clin Electrophysiol 13:1416, 1990. 85. Grubb BP, Gerard JG, Roush K: Differentiation of convulsive syncope and epilepsy with head-up tilt testing. Ann Intern Med 115:871, 1991. 86. Ganzeboom KS, Colman N, Reitsma JB, et al: Prevalence and triggers of syncope in medical students. Am J Cardiol 91:1006, 2003. 87. Sheldon R, Sheldon A, Connolly SJ, et al; for the Investigators of the Syncope Symptom Study and the Prevention of Syncope Trial: Age of first faint in patients with vasovagal syncope. J Cardiovasc Electrophysiol 17:49, 2006. 88. Day SC, Cook EF, Funkenstein H, Goldman L: Evaluation and outcome of emergency room patients with transient loss of consciousness. Am J Med 73:15, 1982. 89. Savage DD, Corwin L, McGee DL, et al: Epidemiologic feature of isolated syncope: The Framingham Study. Stroke 16:626, 1985. 90. Dermkasian G, Lamb LE: Syncope in a population of healthy young adults. JAMA 168:1200, 1958. 91. Benditt DG, Ferguson DW, Grubb BP, et al: Tilt table testing for assessing syncope. J Am Coll Cardiol 28:263, 1996. 92. Mosqueda-Garcia R, Furlan R, Tank J, Femandez-Violante R: The elusive pathophysiology of neurally mediated syncope. Circulation 102:2898, 2000. 93. Linzer M, Pontinen M, Gold DT, et al: Impairment of physical and psychosocial function in recurrent syncope. J Clin Epidemiol 44:1037, 1991. 94. Rose MS, Koshman ML, Spreng S, Sheldon RS: The relationship between health related quality of life and frequency of spells in patients with syncope. J Clin Epidemiol 53:1209, 2000. 95. Lamb L, Green HC, Combs JJ, et al: Incidence of loss of consciousness in 1980 Air Force personnel. Aerospace Med 12:973, 1960. 96. Murdoch BD: Loss of consciousness in healthy South African men: Incidence, causes and relationship to EEG abnormality. South African Med J 57:771, 1980. 97. Sheldon R, Flanagan P, Koshman ML, Killam S: Risk factors for syncope recurrence after a positive tilt-table test in patients with syncope. Circulation 93:973, 1996. 98. Rossen R, Kabat H, Anderson JP: Acute arrest of cerebral circulation in man. Arch Neurol Psych 50:510, 1943.

Chapter 15: Evolving Indications for Pacing 99. Giannattasio C, Grassi C, Mancia G: Vasovagal syncope with bradycardia during lower body negative pressure in a heart transplant recipient. Blood Press 2:309, 1993. 100. Scherrer, Vissing S, Morgan BJ, et al: Vasovagal syncope after infusion of a vasodilator in a heart-transplant recipient. N Engl J Med 322:602, 1990. 101. El-Syed H, Hainsworth R: Salt supplement increases plasma volume and orthostatic tolerance in patients with unexplained syncope. Heart 75:134, 1996. 102. Thomson HL, Atherton JJ, Khafagi FA, Frenneaux MP: Failure of reflex venoconstriction during exercise in patients with vasovagal syncope. Circulation 93:953, 1996. 103. Thomson HL, Wright K, Frenneaux MP: Baroreflex sensitivity in patients with vasovagal syncope. Circulation 95:395, 1997. 104. Manyari DE, Rose S, Tyberg JV, Sheldon RS: Abnormal reflex venous function in patients with neuromediated syncope. J Am Coll Cardiol 27:1730, 1996. 105. Bechir M, Binggeli C, Corti R: Dysfunctional baroreflex regulation of sympathetic nerve activity in patients with vasovagal syncope. Circulation 107:1620, 2003. 106. Dietz NM, Halliwill JR, Speilman JM: Sympathetic withdrawal and forearm vasodilation during vasovagal syncope in humans. J Appl Physiol 82:1785, 1997. 107. Morillo CA, Eckberg DL, Ellenbogen KA, et al: Vagal and sympathetic mechanisms in patients with orthostatic vasovagal syncope. Circulation 96:2509, 1997. 108. Mosqueda-Garcia R, Furlan R, Tank J, et al: Sympathetic and baroreceptor reflex function in neurally mediated syncope evoked by tilt. J Clin Invest 11:2736, 1997. 109. Moya A, Brignole M, Menozzi C, et al; International Study on Syncope of Uncertain Etiology (ISSUE) Investigators: Mechanism of syncope in patients with isolated syncope and in patients with tilt-positive syncope. Circulation 104:1261, 2001. 110. Menozzi C, Brignole M, Lolli G, et al: Follow-up of asystolic episodes in patients with cardioinhibitory, neurally mediated syncope and VVI pacemaker. Am J Cardiol 72:1152, 1993. 111. Krahn AD, Klein GJ, Fitzpatrick A, et al: Predicting the outcome of patients with unexplained syncope undergoing prolonged monitoring. Pacing Clin Electrophysiol 25:37, 2002. 112. Krahn AD, Klein GJ, Yee R: Randomized assessment of syncope trial: Conventional diagnostic testing versus a prolonged monitoring strategy. Circulation 104:46, 2001. 113. Krahn AD, Klein GJ, Yee R, et al: Cost implication of testing in patients with syncope: Randomized assessment of syncope trial. J Am Coll Cardiol 42:495, 2003. 114. White WD, Sheldon RS, Ritchie DA: Learning needs in patients with vasovagal syncope. Can J Cardiovasc Nurs 13:26, 2003. 115. Sheldon R: Tilt testing for syncope: A reappraisal. Curr Opin Cardiol 1:38, 2005. 116. Burklow TR, Moak JP, Bailley JJ, Makhlouf FT: Neurally mediated cardiac syncope: Autonomic modulation after normal saline infusion. J Am Coll Cardiol 33:2059, 1999. 117. Mangru NN, Young ML, Mas MS, et al: Usefulness of tilt table test with normal saline infusion in management of neurocardiogenic syncope in children. Am Heart J 131:953, 1996. 118. Claydon VE, Hainsworth R: Salt supplementation improves orthostatic cerebral peripheral vascular control in patients with syncope. Hypertension 43:809, 2004. 119. Brignole M, Menozzi C, Solano A, et al: Isometric arm counterpressure maneuvers to abort impending vasovagal syncope. J Am Coll Cardiol 40:2053, 2002. 120. Krediet CT, van Dijk N, Linzer M, et al: Management of vasovagal syncope: Controlling or aborting faints by leg crossing and muscle-tensing. Circulation 106:1684, 2002. 121. Sheldon RS, Rose S, Flanagan P, et al: Risk factors for syncope recurrence after a positive tilt-table test in patients with syncope. Circulation 93:973, 1996.

497

122. Ward CR, Gray JC, Gilroy JJ, Kenny RA: Midodrine: A role in the management of neurocardiogenic syncope. Heart 79:45, 1998. 123. Perez-Lugones A, Schweikert R, Pavia S, et al: Usefulness of midodrine in patients with severely symptomatic neurocardiogenic syncope: A randomized control study. J Cardiovasc Electrophysiol 12:935, 2001. 124. Kaufmann H, Saadia D, Voustianiouk A: Midodrine in neurally mediated syncope double-blind, randomized, crossover study. Ann Neurol 52:342, 2002. 125. Di Girolamo E, Di Lorio C, Sabatini P, et al: Effects of paroxetine hydrochloride, a selective serotonin reuptake inhibitor, on refractory vasovagal syncope: A randomized, double-blind, placebo-controlled study. J Am Coll Cardiol 33:1227, 1999. 126. Takata TS, Wasmund SL: Serotonin reuptake inhibitor (Paxil) does not prevent vasovagal reaction association with carotid sinus massage and/or lower body negative pressure in healthy volunteers. Circulation 106:1500, 2002. 127. Sheldon R, Rose S, Flanagan P, et al: Effect of beta blockers on the time to first syncope recurrence in patients after a positive isoproterenol tilt table test. Am J Cardiol 78:536, 1996. 128. Cox MM, Perlman BA, Mayor MR, et al: Acute and long term beta-adrenergic blockade for patients with neurocardiogenic syncope. J Am Coll Cardiol 26:1293, 1995. 129. Alegria JR, Gersh BJ, Scott CG, et al: Comparison of frequency of recurrent syncope after beta-blocker therapy versus conservative management for patients with vasovagal syncope. Am J Cardiol 92:82, 2003. 130. Mahanonda N, Bhuripanyo K, Kangkagate C: Randomized, double-blind, placebo-controlled trial of oral atenolol in patients with unexplained syncope and positive upright tilt table test results. Am Heart J 130:1250, 1995. 131. Madrid AH, Ortega J, Rebollo JG, et al: Lack of efficacy of atenolol for the prevention of neurally mediated syncope in a highly symptomatic population: A prospective, double-blind, randomized and placebo controlled study. J Am Coll Cardiol 37:5554, 2001. 132. Flevari P, Livanis EG, Theodorakis GN, et al: Vasovagal syncope: A prospective, randomized, crossover, evaluation of the effect of propranolol, nadolol and placebo on syncope recurrence and patients’ well being. J Am Coll Cardiol 40:499, 2002. 133. Ventura R, Maas R, Zeidler D: A randomized and controlled pilot trial of beta-blockers for the treatment of recurrent syncope in patients with a positive or negative response to head-up tilt test. Pacing Clin Electrophysiol 25:816, 2002. 134. Sheldon RS, Connolly S, Rose S, et al: The Prevention of Syncope Trial (POST): A randomized, placebo-controlled study of metoprolol in the prevention of vasovagal syncope. Circulation 113:1164-1170, 2006. 135. Schimmer BP, Parker KL: Adrenocorticotropic hormone: Adrenocortical steroids and their synthetic analogue; inhibitors of the synthesis and actions of adrenocortical hormones. In Hardman JG, Limbird LE, Gilman AG (eds): Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 10th ed. New York, McGraw-Hill, 2001, pp 1649-1677. 136. Balaji S, Oslizlok PC, Allen MC, et al: Neurocardiogenic syncope in children with a normal heart. J Am Coll Cardiol 23:779, 1994. 137. Scott WA, Pongiglione G, Bromberg BI, et al: Randomized comparison of atenolol and fludrocortisone acetate in the treatment of pediatric neurally mediated syncope. Am J Cardiol 76:400, 1995. 138. Salim MA, Disessa TG: Effectiveness of fludrocortisone and salt in preventing syncope recurrence in children. J Am Coll Cardiol 45:484, 2005. 139. el-Bedawi KM, Wahba MA, Hainsworth R: Cardiac pacing does not improve orthostatic tolerance in patients with vasovagal syncope. Clin Auton Res 4:233, 1994.

498

Section Two: Clinical Concepts

140. Fitzpatrick AP, Theodorakis G, Ahmed R, et al: Dual chamber pacing aborts vasovagal syncope induced by head-up 60 degrees tilt. Pacing Clin Electrophysiol 14:13, 1991. 141. Samoil D, Grubb BP, Brewster P, et al: Comparison of single and dual chamber pacing techniques in prevention of upright tilt induced vasovagal syncope. Eur J Cardiac Pacing Electrophysiol 1:36, 1991. 142. Sra JS, Jazayeri MR, Avitall B: Comparison of cardiac pacing with drug therapy in the treatment of neurocardiogenic (vasovagal) syncope with bradycardia or asystole. N Engl J Med 328:1085, 1993. 143. Moya A, Brignole M, Menozzi C, et al; International Study on Syncope of Uncertain Etiology (ISSUE) Investigators: Mechanism of syncope in patients with isolated syncope and in patients with tilt-positive syncope. Circulation 104:1261, 2001. 144. Petersen ME, Chamberlain-Webber R, Fitzpatrick AP, et al: Permanent pacing for cardioinhibitory malignant vasovagal syndrome. Br Heart J 71:274, 1994. 145. Sheldon R, Koshman ML, Wilson W, et al: Effect of dualchamber pacing automatic rate-drop sensing on recurrent neurally mediated syncope. Am J Cardiol 81:158, 1998. 146. Connolly SJ, Sheldon RS, Roberts RS, Gent M: The North American Vasovagal Pacemaker Study. J Am Coll Cardiol 33:16, 1999. 147. Sutton R, Brignole M, Menozzi C, et al: Dual-chamber pacing in the treatment of neurally mediated tilt-positive cardioinhibitory syncope: Pacemaker versus no therapy: A multicenter randomized study. The Vasovagal Syncope International Study (VASIS). Circulation 102:294, 2000. 148. Ammirati F, Colivicchi F, Santini M: Permanent cardiac pacing versus medical treatment for the prevention of recurrent vasovagal syncope: A multicenter randomized, controlled trial. Circulation 104:52, 2001. 149. Ammirati F, Colivicchi F, Toscano S: DDD pacing with rate drop response function versus DDI with rate hysteresis pacing for cardioinhibitory vasovagal syncope. Pacing Clin Electrophysiol 21:2178, 1998. 150. Beecher HK: Surgery as placebo. JAMA 176:1102, 1961. 151. Moerman DE, Jonas WB: Deconstructing the placebo effect and finding the meaning response. Ann Intern Med 136:471, 2002. 152. Redelmeier DA, Tu JV, Schull MJ, et al: Problems for clinical judgement. 2: Obtaining a reliable past medical history. Can Med Assoc J 164:809, 2001. 153. McLeod KA, Wilson N, Hewitt J: Cardiac pacing for severe childhood neurally mediated syncope with reflex anoxic seizures. Heart 82:721, 1999. 154. Connolly SJ, Sheldon R, Thorpe KE, et al; VPS II Investigators: Pacemaker therapy for prevention of syncope in patients with recurrent severe vasovagal syncope. Second Vasovagal Pacemaker Study (VPS II): A randomized trial. JAMA 289:2224, 2003. 155. Raviele A, Giada F, Menpzzi C, et al: A randomized, doubleblind, placebo-controlled study of permanent cardiac pacing for the treatment of recurrent tilt-induced vasovagal syncope: The

156.

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

vasovagal syncope and pacing trial (SYNPACE). Eur Heart J 25:1741, 2004. Raj SR, Koshman ML, Sheldon R: Outcomes of patients with dual-chamber pacemakers implanted for the prevention of neurally mediated syncope. Am J Cardiol 91:565, 2003. Gillis AM, Connolly SJ, Lacombe P, et al: Randomized crossover comparison of DDDR versus VDD pacing after atrioventricular junction ablation for prevention of atrial fibrillation. The Atrial Pacing Peri-Ablation for Paroxysmal Atrial Fibrillation Study investigators. Circulation 102:736, 2000. Gillis AM, Connolly SJ, Lacombe P, et al: Atrial pacing periablation for prevention of paroxysmal atrial fibrillation. Circulation 99:2553, 1999. Connolly SJ, Kerr CR, Gent M, et al: Effects of physiological pacing versus ventricular pacing on the risk of stroke and death due to cardiovascular causes. Canadian Trial of Physiologic Pacing Investigators. N Engl J Med 342:1385, 2000. Sheldon R, Rose S, Koshman ML: Comparison of patients with syncope of unknown cause having negative or positive tilt-table tests. Am J Cardiol 80:581, 1997. Grimm W, Degenhardt M, Hoffman J, et al: Syncope recurrence can better be predicted by history than by head-up tilt testing in untreated patients with suspected neurally mediated syncope. Pacing Clin Electrophysiol 18:1465, 1997. Brignole M, Menozzi C, Corbucci G: Detecting incipient vasovagal syncope: Intraventricular acceleration. Pacing Clin Electrophysiol 20:801, 1997. Osswald S, Cron T, Gradel C: Closed-loop stimulation using intracardiac impedance as a sensor principle: Correlation of right ventricular dp/dtmax and intracardiac impedance during dobutamine stress test. Pacing Clin Electrophysiol 23:1502, 2000. Binggeli C, Duru F, Corti R: Autonomic nervous system-controlled cardiac pacing: A comparison between intracardiac impedance signal and muscle sympathetic nerve activity. Pacing Clin Electrophysiol 23:1632, 2000. Griesbach L, Huber T, Knote B, et al: Closed loop stimulation: Therapy for malignant neurocardiogenic syncope. Prog Biomed Res 7:242, 2002. Occhetta E, Bortnik M, Vassanelli C: The DDDR closed loop stimulation for the prevention of vasovagal syncope: Results from the INVASY prospective feasibility registry. Europace 5:153, 2003. Ochetta E, Bortnik M, Audoglio R, Vassanelli C; for the INVASY Study Investigators: Closed loop stimulation in prevention of vasovagal syncope. Inotropy Controlled Pacing in Vasovagal Syncope (INVASY): A multicentre randomized, single blind, controlled study. Europace 6:538, 2004. Giada F, Raviele A, Gasparini G: Efficacy of a patient-activity drug delivery system using phenylephrine as active drug in aborting tilt-induced syncope [abstract]. Pacing Clin Electrophysiol 24:573, 2001.

Chapter 16

Sensor-Driven Pacing: Device Specifics CHU-PAK LAU • HUNG-FAT TSE • G. NEAL KAY

T he ideal characteristics of an implantable sensor for rate adaptation have been addressed in Chapter 5. In clinical practice, many of the special lead sensors, such as the oxygen saturation sensor, have not been used for rate-adaptive pacing. Clinicians remained concerned about the stability and reliability of special leads and the future implications of their replacement. Thus the majority of clinically used rate-adaptive devices employ standard pacing leads. These include body movement sensing, minute ventilation sensing, QT interval sensing, and their combinations. The sensing of peak endocardial acceleration (PEA) and regional impedance changes at the pacemaker lead, used by the so-called closed-loop stimulation (CLS) sensors, are the more recently available methods of rate adaptation. Activity-Sensing and Accelerometer-Based Pacemakers Activity sensing has achieved wide clinical acceptance as a rate-controlling parameter for implantable cardiac pacemakers. The rate-adaptive pacemaker incorporating sensors that monitor body activity was the first to be implanted and is the standard used for rateadaptive pacemakers in all companies.1,2 Activity sensing is also the sensor method that is combined with other sensors. Because activity-based pacemakers are operationally simple and do not require a special sensor outside the pulse generator casing, they work with any

type of pacing lead, have excellent long-term stability, and are highly reliable. Implantation of activity-guided pacemakers is no different from that of conventional pacemakers. Although they may not be excellent proportional sensors, activity sensors react promptly to the start and end of physical exercise. The first activity sensors were piezoelectric crystals that responded mostly to the frequency of vibrations that were transmitted to the pulse generator. The specific use of an activity sensor for pacemaker rate augmentation was described first by Dahl3 in 1979 (an accelerometer configuration) and then by Anderson and colleagues4 in 1983 (a pressure-vibration configuration). In 1987, the possibility of using accelerometer-based activity sensing for pacing rate control was reported for the first time.5,6 Activity sensing has potential application for detecting daily activity, as in monitoring for heart failure, to detect the posture of the patient, and to adjust the lower pacing rate by defining the resting state. Physical Principle of Activity Sensing Body movements, especially walking, result in vibrations that are transmitted to the upper chest or generate acceleration forces in the body. In the pacemaker scenario, acceleration forces acting on the body during exercise are best detected by a device inside the pacemaker box. With triaxially mounted accelerometers placed on the surface of an externally attached pacemaker, acceleration signals during a variety of exercises 499

500

Section Two: Clinical Concepts

have been measured, permitting the study of the acceleration forces during these exercises.5 The axes used are referred to as “anteroposterior” (x-axis), “lateral” (y-axis), and “vertical” (z-axis). Because of the sloping of the chest and the swaying of the body (and hence the pacemaker) during walking, these axes are not the same as true horizontal (x or y) or true vertical (z) axes. The acceleration signals are transformed by fast Fourier method with respect to the frequency. The root mean square value of acceleration is used to quantify the acceleration force. The following findings have been reported:5 • Axes most relevant to detect walking: A recording

of acceleration signals in a typical subject during walking is shown in Figure 16-1. It is apparent that either the x-axis or z-axis is useful to detect the acceleration forces during walking. On the other hand, the y-axis is useful only to detect body swinging. In the construction of an activity sensor in an implanted pacemaker, the anteroposterior (x) axis would be more practical than the z-axis because the “top” of the pacemaker can vary

according to how the pacemaker is implanted and is likely to be influenced by subsequent pacemaker rotation in the pocket, whereas the anteroposterior axis remains relatively fixed over time. • Effects of walking speed and gradient on the accel-

eration signals: Acceleration forces are represented by the integrated root mean square value of accelerations. Walking at a higher speed will induce significant increase in acceleration signal (Fig. 16-2). Although walking upslope also increases the acceleration forces, the increase is less than that induced by walking faster. Thus, activity sensors as a whole are less sensitive to this form of exercise. • Frequency range of acceleration forces during

walking: During normal walking, the fast Fourier transformed acceleration shows that the majority of the signal is less than 4 Hz (see Fig. 16-1). Relatively little signal is lost by low-pass filtering at 4 Hz, and filtering may also improve the proportionality of the acceleration force to the level of workload (Fig. 16-3).

S5 AW 1.2 mph 15% Acceleration

Amplitude Spectrum

m/sˆ2

m/sˆ2 1

4 2 X

0 ⫺2

.1

⫺4

.04 sec

⫹ 5

⫹ 10

⫹ 15

⫹ 20

⫹ 25 Hz

⫹ 5

⫹ 10

⫹ 15

⫹ 20

⫹ 25 Hz

⫹ 5

⫹ 10

⫹ 15

⫹ 20

⫹ 25 Hz

1

4 2 Y

0 ⫺2

.1

⫺4

.04 sec 1

4 2 Z

0 ⫺2

.1

⫺4

.04 sec

Figure 16-1. Representative acceleration signals recorded in a typical subject during walking at 1.2 mph at a 15% gradient on a treadmill. Each strip represents 10 seconds’ duration on the treadmill, and each peak of the curves in the x- and z-axes on the left side represents one step on the treadmill. Left, The number of peaks in the lateral (y-) axis is one half of either that of the x- or z-axis. This represents the swaying of the shoulder, which occurs once per complete walking cycle, which this axis principally detects. Right, Fourier-transformed acceleration amplitudes at different frequencies are shown graphically. Most of the acceleration forces are less than 4 Hz. (From Lau CP, Stott JR, Toff WD, et al: Selective vibration sensing: New design of activity-sensing rate-responsive pacing. PACE 11:1299, 1988.)

Chapter 16: Sensor-Driven Pacing: Device Specifics

tRMS (x) (ms⫺2)

Sensors

2.5 mph 1.5 mph

1

0.5

0 0

501

5 10 Percentage gradient

15

Figure 16-2. Total root mean square acceleration (tRMS) in the x-axis during walking at different speeds and gradients. tRMS increases as a function of both speeds and gradients (P < .001). Each error bar represents 1 SEM. (From Lau CP, Stott JR, Toff WD, et al: Selective vibration sensing: New design of activity-sensing rate-responsive pacing. PACE 11:1299, 1988.)

• Other forms of exercise: Appropriate increase in

acceleration force occurs during running (Fig. 16-4). However, because the acceleration force encountered during arm exercise is small in the pectoral area, particularly if the x-axis is used, the forces measured during cycling and weightlifting would be lower than would be expected from the amount of workload. Although accelerations during arm movements are better detected in the horizontal axis, this direction is difficult to use in a pacemaker, which is liable to change its position inside the pocket after implantation.

1.8

The three different types of clinically used activity sensors are a piezoelectric sensor and an accelerometer using either piezoelectric or piezoresistive materials (Fig. 16-5). The piezoelectric crystal is attached to the inside of the pacemaker casing, and pressure waves initiated in the skeleton and soft body tissues by physical activity result in a physical deformation of the piezoelectric element (see Fig. 16-5A).4,7 Because deformation of the piezoelectric sensor induces a voltage that is proportional to the amount of structural disturbance, measurement of these induced voltages permits estimation of the level of physical activity. Because the piezoelectric element is usually attached to the posterior surface of the pulse generator can during manufacturing, it is typically positioned directly against the pectoralis major muscle to ensure good physical contact with the skeletal muscles. The pacemaker can be implanted with the sensor facing away from the muscle activity if the activity threshold or the rate response can be programmed to compensate for the reduced signal amplitude with this orientation. Generally, the piezoelectric element produces potentials in the range of 5 to 50 mV during rest and as much as 200 mV during vigorous activity. The range of frequencies to which these systems are most sensitive is generally about 10 Hz, close to the typical resonant frequency of the human body.5,6 Given these signal characteristics, activity-based pacemakers that use a piezoelectric element bonded to the inside of the pacemaker case appear to offer good correlation with upright physical movement involving walking or running. An accelerometer is a sensor designed to measure acceleration, defined as the rate of change in velocity. Accelerometers can be made of piezoelectric crystal, mounted on a cantilever mounted on the circuit board (see Fig. 16-5B), or as an integrated circuit using silicon wafers sandwiching a suspended mass, the so-called piezoresistive accelerometer (see Fig. 16-5C). Physiologic studies have shown that rhythmic body motions, such as walking and riding a bicycle, fall within a narrow

tRMS ⬍4 Hz

1.6

RMS (m/s/s)

1.4 1.2 1 0.8 0.6 0.4 Figure 16-3. Total root mean square acceleration (tRMS) and low-pass RMS (23% of the heart rate reserve (HRR) over time. Note that the sensor-driven upper rate limit (SURL) may not be reached at a low slope setting. LRL, lower rate limit.

SURL Response Factor LRL Day 1

Day 2

Day 3

Day 4

Day 5

Day 6

Day 7

SURL Response Factor LRL Day 1

Day 2

Day 3

Day 4

Day 5

Day 6

Day 7

Figure 16-15. Automatic evaluation of rate response in the Guidant Insignia pacemaker. The “Response Factor” will be scaled upward or downward according to the average peak daily rates and compared with the programmed “Sensor Rate Target” accelerations above a moving average. LRL, lower rate limit; SURL, sensor-driven upper rate limit.

Guidant: Insignia, Pulsar Max In the devices manufactured by Guidant (Boston Scientific, Natick, Mass.), an accelerometer mounted in the circuit board is used to detect anteroposterior acceleration. Four parameters are used to determine the rate response: response factor (1 = least sensitive, 16 = most sensitive), activity threshold (very low, low, medium-low, medium-high, high, and very high), reaction time, and recovery time. Programming can be effected either manually or automatically with the “Automatic Response Factor.” In this method, the physician uses the “Expert Ease” feature in the programmer to decide the LRL, SURL, and the so-called Sensor Rate Target, on the basis of the patient’s age, gender, exercise frequency, and target heart rate during exercise from a

population average. After these parameters are input, an initial nominal slope, such as 8 (1 through 16 are available), is programmed. The “Automatic Response Factor” will track the maximum sensor rate each day for a week, and the average of this maximum will be compared with the “Sensor Rate Target.” The response factor will then be scaled to effect either a more aggressive or a more conservative response, depending on whether the average rate falls below or exceeds the “Sensor Rate Target” by 5 bpm (Fig. 16-15). Other Manufacturers Sorin Biomedica CRM (Sorin Group, Milan) has introduced a gravitational sensor that is used either alone (Swing) or in combination with a PEA sensor (MiniLiving).

508

Section Two: Clinical Concepts

The gravitational sensor uses the vibrations from a mercury ball to measure body activity. ELA Medical (Sorin Group, Milan) has an accelerometer activity sensor (Opus G) that uses a half-bridge variable-capacitance accelerometer. The accelerometer detects anteroposterior axis at a frequency range of 0.6 to 6 Hz, and samples acceleration signals every 1.56 seconds. The device can be programmed manually or automatically using the “Autocalibration.” The level of acceleration at rest is reset daily according to the lowest mean acceleration of 64 consecutive measurements. The maximum mean acceleration of 8 consecutive accelerations is used to match the programmed upper sensor rate. In 43 patients with Opus G rate-adaptive pacing, the rate-response was reported to be proportional to workload on walking and descending stairs in comparison with normal and vibration sensors, but the rate-response on stair climbing remained inadequate.18 Clinical Results Clinical studies have convincingly demonstrated that activity-based pacing systems offer the potential for greater exercise capacity and fewer exertionally related symptoms than do fixed-rate (VVI) pacemakers. In an early clinical study of a piezoelectric vibration-based, adaptive-rate pacemaker, Benditt and colleagues12 used cardiopulmonary treadmill exercise tests to compare exercise tolerance during fixed-rate VVI pacing with that during VVIR pacing. Adaptive-rate pacing prolonged exercise duration by 35% and led to similar improvements in peak oxygen consumption and oxygen consumption at anaerobic threshold. Adaptiverate pacing also reduced the patient’s perception of exertion at comparable exercise levels,12 and the benefit was sustained when exercise testing was repeated after an average of 5 months of follow-up. Furthermore, at the time of follow-up exercise testing, reversion of the pacing system to a fixed rate (VVI) mode resulted in prompt deterioration of both observed oxygen consumption and exercise duration. Thus, the ability of a single-chamber piezoelectric vibration-based pacing system to provide immediate and long-term improvements in exercise tolerance was clearly demonstrated. Other investigators have also reported better exercise capacity with adaptive-rate pacemakers based on activity sensors, although the results from studies using bicycle exercise have been less dramatic than those observed with treadmill exercise.19,20 The effect of ventricular function on exercise responses of 16 patients with piezoelectric vibration-based, adaptive-rate pacemakers in the VVI and the VVIR modes has been reported.21 The findings indicate that the provision of appropriate heart rate responsiveness by this technique resulted in a substantial increase in cardiac index that was independent of baseline ejection fraction. Another study using mainly activity-sensing devices found that the potential hemodynamic benefits of adaptive-rate pacing in 22 patients was proportional to the extent of systolic dysfunction, with a greater benefit in those with poorer ventricular function.22

Activity-based VVIR pacing has also been compared with atrial-tracking, dual-chamber pacing modes (VDD, DDD),23 which showed similar exercise tolerance in the two modes (VVIR, 68 ± 15 W/min vs. DDD, 70 ± 18 W/min), but more patients preferred the DDD mode, suggesting the importance of AV synchrony. On the other hand, other studies found no significant differences between the DDD and VVIR modes with respect to symptom scores, maximal exercise performance (treadmill), or plasma concentrations of epinephrine, norepinephrine, and atrial natriuretic peptide.24,25 Surprisingly, venous epinephrine and norepinephrine levels were not higher during exercise in the VVIR mode, as might have been expected given comparable exercise levels. These data suggest that the role of atrial contribution is less in patients with heart block. Several studies comparing the behavior of accelerometer-based devices with that of piezoelectric crystal devices reported that the accelerometer devices showed a better response to walking, jogging, and standing.9,17,26 It was observed that the subject’s footwear had no significant effect on the results seen with the accelerometer, as opposed to the results obtained with piezoelectric vibrational devices. Increasing grade of the treadmill had a significant effect on pacing rate with the accelerometer device, whereas there was no change in pacing rate with the piezoelectric vibrational sensors. The investigators concluded that, compared with the vibrational device, the accelerometer sensorcontrolled devices showed a better rate-response and were less susceptible to direct pressure or to tapping on the pulse generator, unlike what was observed with the piezoelectric crystal.26,27 Using a strapped-on accelerometer, Charles and colleagues27 found that the response of an accelerometerbased device (CPI Excel; CPI is now part of Guidant, St. Paul, Minn.) to graded treadmill testing was more strongly correlated with the patient’s intrinsic heart rate (r = 0.80) than that of a vibrational adaptive-rate device (r = 0.27). The accelerometer responded appropriately when subjects walked up stairs (103 bpm) and walked down stairs (98 bpm). The response of the vibrational devices was paradoxical, giving a slower pacing rate when subjects walked up stairs (83 bpm) than when they walked down stairs (89 bpm). This multicenter study on the Excel has documented good response during daily activities.27 Schuster and associates28 evaluated the efficacy of automatic “Rate Profile Optimization” over time. Eleven patients with Kappa 700 pacemakers performed treadmill testing at 1 month, 1 year, and 2 years after implantation. On the basis of the sinus profile at followup, the investigators found that the use of a more aggressive slope was needed to match the sinus rate profile. This required the change of ADL response from 3 to 4 and of ER from 3 to 4 at 1 year, and the change of activity threshold from medium/low to low at 2 years. These adjustments enabled better approximation of pacing rate to sinus rate during treadmill exercise. Exercise capacity was maintained during the 2 years. In another study, activity level variation (“Activity Variance”), as detected by an accelerometer, was used

Chapter 16: Sensor-Driven Pacing: Device Specifics

to determine sleeping time in devices manufactured by St. Jude Medical.29 The results showed good agreement with an actigraph that recorded patient movement externally. However, minor movement of the subject during sleep would reactivate the device, so that long periods of sleep rate pacing were not possible with the current algorithm.29 Raj and coworkers30 examined the ability of the activity sensor to simulate heart rate variability, an important measure in heart failure.30 They found that activity response contributed to long-term measures of heart rate variability through heart rate changes during exercise but had no effect during short-term measures, which depend on autonomic changes unrelated to exercise. This study suggested that rate modulation by activity sensors during exercise is an important element of heart rate variability. Limitations Activity sensing may give inappropriate heart rateresponses when subjected to environmental vibrations, such as those induced by the movement of a motor vehicle over rough terrain or those resulting from air travel or the use of appliances or machinery. The piezoelectric vibrational sensor also responds to the application of static pressure on the pulse generator, which may be important when the patient is prone. This falsepositive response to pressure is less of a problem with pacemakers that incorporate an accelerometer.31 Matula and colleagues32 assessed the effects of various means of locomotion on pacing rate for different activity-based pacemakers. Three different activitybased pacing systems (peak counting algorithm, integration type, and accelerometers) were strapped to the chests of volunteers. Bicycling on the street resulted in higher pacing rates than stationary bicycling for each type of pacemaker, although none of the pacemakers reached the heart rate achieved by the normal sinus node. During driving, the pacemakers raised the pacing rate, although the intrinsic sinus rate continued to be higher. In passively riding passengers, the pacemakers tended to produce a higher pacing rate than that of the normal sinus node. Of interest, the accelerometer-based system responded mainly to acceleration and curves, whereas vibration sensors responded primarily to vibrations and rough roads. Independent of the sensor, activity-initiated rate-response depends on the manner in which activity is being carried out rather than on the exercise workload, and proportionality is generally limited. Activity sensors may manifest paradoxically slower heart rates during walking uphill than during walking downhill. Non–exercise-related stresses such as emotional changes are not detected, limiting the sensor’s sensitivity.

Minute Ventilation Sensing Of all impedance-based sensors, the sensing of minute ventilation (MV) is the most commonly used technique

509

and has been the concept for a physiologic system since 1966.33,34 Although a respiratory rate–sensing pacemaker had been in use as early as 1983, it was limited by the need of an auxiliary subcutaneous electrode and easy interference by arm movement because of unipolar impedance sensing. All subsequent generations of respiratory sensors detect MV for rate adaptation. Physiologic Principle Relationship Between Heart Rate and Respiratory Parameters during Exercise Changes in heart rate during exercise are closely related to changes in oxygen uptake (Vo2) at all levels of exertion. At metabolic workloads of less than anaerobic threshold, Vo2 and heart rate are also directly proportional to MV, with correlation coefficients greater than 0.9 in most studies.35-39 On the other hand, the correlation between respiratory rate and oxygen uptake during submaximal exercise is less than 0.54.35-37 The reason is that the ventilatory response at the onset of exercise is predominantly due to a change in tidal volume rather than in respiratory rate.38,39 In a number of studies, it was noted that the tidal volume increased to a plateau within 2 minutes after the onset of exercise. The relative speed of changes in respiratory rate and MV (the product of tidal volume and respiratory rate) during exertion is shown in Figure 16-16.39 The respiratory rate not only rises slowly during exercise but also declines faster than tidal volume at the cessation of exercise. Anaerobic Threshold With more strenuous exercise, the heart is unable to meet the increased oxygen demand of the working muscles completely, and anaerobic metabolism is initiated, resulting in greater production of lactic acid. Lactic acid dissociates into lactate and H+, which is buffered by bicarbonate, resulting in an abrupt rise in carbon dioxide production. Because MV is largely controlled by carbon dioxide production and blood pH rather than Vo2, the higher rate of carbon dioxide production induces a rise in MV that is out of proportion to the increase in Vo2. Vo2 is linearly related to the normal sinus rate throughout exercise, so at workloads above anaerobic threshold, MV will increase disproportionately relative to Vo2 and the sinus rate. This has important implications in an MV-controlled pacemaker, which needs special rate-adaptive curves to avoid overpacing above the anaerobic threshold. Effect of Pulmonary Disease and Congestive Heart Failure on Minute Ventilation Sensing In patients with chronic obstructive pulmonary and restrictive lung diseases, it is the pulmonary system that usually limits exercise, not the cardiovascular system. In a healthy patient at peak exercise, MV is about 50% of the maximum voluntary ventilation. In a patient with pulmonary disease, this reserve is smaller and, in many

510

Section Two: Clinical Concepts Figure 16-16. Changes in respiratory variables during progressive exercise. Minute ventilation (RMV, in L/min) closely parallels the change in heart rate (HR, in bpm), whereas there is little change in respiratory rate (RR, in breaths/min) at the beginning of exercise. (From Alt E, Heinz M, Hirgestetter C, et al: Control of pacemaker rate by impedance-based respiratory minute ventilation. Chest 92:247, 1987.)

cases, disappears before the anaerobic threshold, indicating that the patient will not reach this condition. The transthoracic estimate of MV is often higher for the same Vo2 in patients with pulmonary disease, providing a strong signal for the implantable device to control heart rate, although the pacing rate change has to be limited. To date, only a limited number of case reports have suggested the safety and feasibility of using MV sensors in lung disease.40 In the era of biventricular pacing for heart failure, the use of MV rate-adaptive sensors during exercise and for monitoring heart failure is of interest. Like the resting heart rate, MV is higher at rest but achieves a lower maximum level in patients with congestive heart failure.41 It has been suggested that the ratio of peak MV to resting MV in healthy patients is close to twice that in patients with congestive heart failure, thus providing a natural limit to the extent of rate adaptation. When patients with congestive heart failure undergo a regimen of exercise training, their peak MV increases, as does their maximum heart rate. On the cardiovascular side, patients with heart failure often have associated chronotropic incompetence, but the range of heart rate changes is small given the higher resting heart rate. Because of the lesser importance of atrial contribution during exercise in patients with heart failure, rate adaptation assumes a greater significance.22 Patients with biventricular pacing may still derive benefit from rate-adaptive pacing if their maximum predicted heart rate during exercise is below 70% of the age-predicted maximum.42 Cheyne-Stokes respiration is a waxing and waning of tidal volume and respiratory rate with a periodicity of about 0.02 Hz. More than 60% of the patients with congestive heart failure show overt Cheyne-Stokes respiration or some form of pulsed breathing during sleep,43 owing to prolonged transit time of blood flow to the brain. Manifestation of Cheyne-Stokes respiration during exercise44 and activities of daily living45

tends to carry a poor prognosis. Sleep apnea and Cheyne-Stokes respiration can be effective indices of congestive heart failure that a pacemaker, measuring MV, could monitor to guide therapy. Despite these interesting data, MV sensing is best avoided in patients with significant lung disease. The usefulness of MV sensors in patients with heart failure and cardiac resynchronization therapy remains to be tested. No MV sensor–driven biventricular device has yet been introduced. Sensors The measurement of impedance in biomedical applications is commonly referred to as plethysmography and has its basis in Ohm’s law. This law states that the ratio of the applied voltage (V) to the current (I) flowing is as follows: R = V/I where R is the impedance. If the current (I) is kept constant, then the voltage (V) that is measured will reflect changes in resistance. The value of R is related to the resistivity (ρ) of the medium (blood and tissue, etc.), by the length of the path (L), and inversely by the cross-sectional area (A + ΔA) of the conducting medium, as indicated by the following equation: R = ρ(L/A + ΔA) Note that the cross-sectional area is displayed as having two components, a constant component (A) and a dynamic component (ΔA) that changes with respiration (and other factors). Placing any two electrodes subcutaneously across the human torso results in the impedances shown in Figure 16-17. A measurement of this type is termed a bipolar measurement. The resistances, R1 and R2, and the capacitances, C1 and C2, are due to the effects of polarization at the electrode-electrolyte interface (polar-

Chapter 16: Sensor-Driven Pacing: Device Specifics

511

IMP Q1 E1 R1

Rp

Atrial electrode

Q2 Rm

Rp

R2

E2

T

R I

C1

Pacemaker

C2 Electrical Model

R1 Rp

Rm

Rp

R2

R

I IMP

T Figure 16-17. Constituents of impedance measurement across the human torso. When voltage is applied between electrodes E1 and E2, separate measurement points Q1 and Q2 give respiration changes independent of contact resistances. R1, C1, R2, and C2 are polarization elements; RP, contact resistance; RM, torso resistance containing respiration signal.

ization effects). The values of these parameters depend on the frequency of the measurement current. At frequencies above a few thousand hertz, their contribution becomes negligible. For this reason, as well as for the purposes of minimizing battery drain and maintaining patient safety, the measurements in implantable devices are performed with high frequencies or very narrow pulse widths. As an example, a current pulse used in one early MV pacing system is 0.015 msec wide, roughly equivalent to a frequency of 33 kHz (Meta MA, Telectronics Pacing Systems, Sylmar, Calif.). This frequency eliminates all polarization effects. A fluid-flow analogy can be used to understand the bipolar impedance measurement. With this model, the impedance can be conceptualized to comprise three “conduits” that impede the flow of electric current. The narrow conduits (Rp) are related to contact resistance at the electrode-tissue interfaces and have a considerably larger impedance than the wider conduit Rm, which is related to resistance across the torso (the two Rp values are shown as equal for convenience). The impedance Rm contains the respiration signal that we want to measure to control the pacing rate of a rateadaptive pacemaker. As a rule, the values of Rp are greater than those of Rm, especially for electrodes with a small surface area. This makes the regions around the electrodes prone to artifacts related to movement of the skin and underlying tissues. To eliminate this source of inaccuracy, normally either a four-electrode (quadripolar) system, as shown by the measurement points Q1 and Q2 in Figure 16-17, is used, or the electrodes are made large, thereby minimizing this funneling. The usual tip electrode-tissue impedance of a pacing lead is on the order of 400 ohms, and that around a pacemaker case is usually less than 20 ohms. The impedance across the torso is on the order of 50 ohms. The change in Rm that is related to respiration (the value that we want to measure) is in the vicinity of 1 ohm. To detect small changes in minute ventilation, a resolution of 0.06 ohms is required. This

Ventricular electrode Figure 16-18. Measurements explored for conventional pacing electrode configurations for minute volume (MV) measurement. Measurements in both chambers of the heart are shown. I, constant-current pulse; IMP, impedance measured; R, ring; T, tip.

example gives some indication of how sensitive the measurement system has to be. Rossi and colleagues34 used an auxiliary lead tunneled across the chest as one pole of a bipolar system. This arrangement was successful in detecting respiratory rate, but the auxiliary lead is prone to erosion and movement artifacts from the chest cage. With these disadvantages in mind, Nappholz, in collaboration with Maloney and Simmons of the Cleveland Clinic, carried out a series of studies to explore the use of transvenous electrodes to measure MV in exercising patients.46,47 The impedance measurements were made using a quadripolar system in the superior vena cava at first in dogs and subsequently in patients. The correlation of changes in impedance with actual changes in MV was excellent (r > 0.9). In human studies, a cutaneous defibrillation pad was placed over the prepectoral region (the site of the pacemaker case), and the measurement current was generated between the right ventricle and the cutaneous pad. The results confirmed that the use of a common electrode for both generating the current pulses and measuring the impedance was appropriate and that the impedance of a cutaneous pad was about the same as that for a pulse generator case (0.9 in first-, second-, and third-order polynomial equations; Fig. 16-21).58 Furthermore, the calibration between measured MV and impedance MV changed over time (between 1 week and 1 month).59 However, the changes were not correlative. This finding has implications for the need for continual automatic adaptation and the potential use of the sensor for MV monitoring.

Chapter 16: Sensor-Driven Pacing: Device Specifics 1.0 Total Area 89% Observed

0.75 Expected Quartile 1 137%

0.25 Quartile 2 81% 0.00 0.00 Min VO2

A

Quartile 3 66%

0.4

Observed

0.50

0.75

1.0

1.00 Max VO2

0.8

0.6

0.4

0.2

0.0

Metabolic Equivalents (normalized)

B

1.0 Total Area 122%

0.8

Expected

Observed 0.6 0.4 Quartile 1 352%

0.0 0.00 Min VO2

Quartile 2 166%

Total Area 30%

0.8 HR (Normalized)

HR (Normalized)

Expected

0.0 0.25

1.0

C

0.6

0.2

Quartile 4 102%

Metabolic Equivalents (normalized)

0.2

Total Area 85%

0.8 HR (Normalized)

Heart Rate (Normalized)

1.00

0.50

513

Quartile 3 100%

0.6 Expected 0.4

Observed

0.2

Quartile 4 75%

0.0 0.25

0.50

0.75

1.00 Max VO2

Metabolic Equivalents (normalized)

1.0

D

0.8

0.6

0.4

0.2

0.0

Metabolic Equivalents (normalized)

Figure 16-20. Quartification of pacing percentage of minute volume (MV) and activity pacer during graded treadmill exercise. A shows the results of the normalized heart rate and workload in one patient. There is overpacing in quartile 1 (137%), underpacing in quartiles 2 and 3 (81% and 66%, respectively), and near-ideal pacing in quartile 4. B shows the near-ideal rate recovery of the MV sensor. C and D show the exercise response of an activity sensor with overpacing during most of the time, but inadequate rate recovery. (From Kay GN: Quartification of chronotropic response: Comparison of methods for rate-modulated permanent pacemakers. J Am Coll Cardiol 20:1533, 1992.)

Guidant Devices In the minute ventilation–sensing devices manufactured by Guidant, the Pulsar Max, and the Insignia, MV is available in conjunction with an accelerometer sensor, although the rate-adaptive function can be used either alone or blended with the accelerometer. MV collected by either the atrial or ventricular lead (programmable) over 24 hours is used as an average against which future changes in MV are compared for a rate response. The minimum time for achieving a baseline is 4 minutes (thus 4 → On). A linear response curve is used, and a total of 16 response factors can be chosen below the judged pacing rate for anaerobic threshold. This rate is also programmable. In between the pacing rate for anaerobic threshold and the maximum predicted heart rate, a gentler “High Rate Response” can be programmed. The adaptation can be made faster by activating the 4-minute walk

within 30-minute option, in which the subject is instructed to exercise to achieve the maximal MV change. Telemetered MV impedance signals from a Guidant MV device have been compared with measured MV in 20 patients.60 Respiratory rate was accurately measured by the device during hyperventilation, with a difference of less than 0.2 breaths/minute. During 10-minute cycle ergometry at 50 W, the correlation between MV measured directly and that measured by the device was 0.99. There are large individual variations between the measured and impedance MV slopes, requiring specific rateresponse curves for particular patients. Sorin Group–ELA Medical Minute Ventilation–Sensing Devices The MV devices manufactured by ELA Medical and its parent company, Sorin Group, are the Chorus, Talent,

Section Two: Clinical Concepts Ohms/min

100 90

VE (liters/min)

80 70 60

140

140

120

120

100

100

80

50 40

60

30

40

IMV (Ohms/min)

Liters/min

IMV (Ohms/min)

514

20 10 0 0

A

5

10

15

80 60 40

20

20

0

0

20

Minutes

15

B

20

40

60

80

100

VE (liters/min)

Figure 16-21. A, Transthoracic impedance minute ventilation (IMV) and measured minute ventilation (VE) by duration of the exercise test and recovery for an individual subject. The two parameters show good correlation. B, Regression analysis for the same subject showing first-, second-, and third-order regression lines, which are practically superimposed. (From Cole CR, Jensen DN, Cho Y, et al: Correlation of impedance minute ventilation with measured minute ventilation in a rate responsive pacemaker. PACE 24:989, 2001.)

Opus, Synphony, and Rhapsody. The automatic slope algorithm in the Chorus determines the resting and maximum MV values on a daily basis. The device calculates the exercise MV signal by looking for the maximal MV signal and recalculates this value every eighth cycle during which a change has occurred. The exercise MV value is increased or decreased in 6% intervals. The mean resting and exercise MV values are used to adjust the rate-response slope automatically over the range of values from 1 to 15 in steps of 0.1. By automatically adjusting the rate-response slope in this manner, the Chorus may be able to provide rateresponse that is individualized for each patient and can vary it as physiologic conditions evolve. The algorithm used by the Chorus has now been tested extensively and found to be effective.61,62 Furthermore, the use of different impedance-sensing electrode configurations has been evaluated—atrial bipolar, ventricular bipolar, double unipolar, and “floating” configurations in a unipolar lead.63 The mean correlations for these four configurations between impedance-based MV values and MV values measured during exercise were 0.89 ± 0.08, 0.95 ± 0.05, 0.87 ± 0.14, and 0.88 ± 0.05, respectively, suggesting the possibility of using the MV sensor in a wide range of populations. In addition, if the MV value over 128 cycles remains below the 24-hour mean MV, the lower rate decreases to the rest rate, which is set below the lower rate limit. In a Holter monitoring study of 46 patients with the Chorus RM pacemaker,64 the MV successfully decreased the base rate during the sleep period, with the diurnal pacing rate at 68 ± 5 bpm and a nocturnal rate of 60 ± 4 bpm. Similarly, at the upper rate, if MV continues to increase, the device automatically decreases the linear rate-response slope to reach the higher MV on a subsequent exercise. In another study, the response of the Talent DR (ELA Medical) during exercise was

compared in 81 patients. The correlation coefficient between the sinus rate and the programmer-derived sensor rate was 0.983 ± 0.005, and a linear relationship was observed between heart rate reserve and MV reserve.61 Limitations The use of MV devices in patients with lung disease and heart failure remains controversial. Because of filtering of the cardiac component, the MV sensor does not accurately reflect rate adaptation at high respiratory rates, which occur commonly in children. A 1998 study found good correlation in 5 of 11 children between sinus rate and MV-driven rates.65 In a later study, Cabrera and associates66 measured MV in 38 healthy children and used computer simulation of the Kappa MV algorithm to define the intrinsic heart rate. They found good correlation between predicted heart rate within 80% of heart rate reserve. However, particularly in small children with body surface areas of less than 1.1 m2, inadequate rate response occurred (26 ± 16 bpm lower than expected) because the respiratory rate at 65 breaths/ minute exceeds the pacemaker MV processing capability, which is set at 48 breaths/minute. Thus, in children, MV is a feasible alternative with appropriate selection of subjects. In conjunction with an activity sensor, MV sensing may enhance metabolic response of patients with pacemakers, a capability that is particularly relevant in active children. The speed of rate response of the earlier-generation MV sensing pacemakers was slow and had a delay of 30 to 45 seconds compared with an activity-sensing pacemaker. This slowness was due partly to the MV-averaging algorithm used and the curvilinear rateadaptive curve, which resulted in a slow increase of rate at the onset but reached the maximal rate earlier

Chapter 16: Sensor-Driven Pacing: Device Specifics

during exercise compared with the sinus rate. The pacing rate may also remain high after exercise for 1 to 2 minutes before returning gradually to the baseline. This limitation is addressed with the use of more complex curves in modern MV devices. A bipolar atrial/ventricular lead is needed for MV sensing. The battery current for MV sensing may take up to about 2% of the total current of a dual-chamber pacemaker. It is possible for some sense amplifiers to sense the small impedance pulses unless some preventive steps are taken. In some products, a blanking period of about 1 msec is applied to the amplifiers to “blind” them to these pulses. In other cases, special balancing of the pulse is carried out, achieving the same objective. Sensing of the impedance pulses by a surface electrocardiogram monitor is always a possibility and depends on the sensitivity of the electrocardiogram machine. The pulse width and the balancing of the impedance pulse influence this possibility. The impedance is measured only for the duration of the narrow, microsecond pulses. As long as the electric signal does not change during this short interval, interference from underlying electric signals is rejected. Sixty hertz takes about 7000 msec to change from its minimum to its maximum value. In the 15 msec required to make the impedance measurement, an intracardiac signal, obviously, changes very little. Hence, the effect of an intracardiac electrogram on the impedance pulse is negligible. Frequencies above a few kilohertz, such as those generated by electrocautery and electrosurgery, are detected by the rate-response circuitry and could drive a pacemaker to its maximum rate, so it is recommended that the rate-response function of a pacemaker be turned off whenever a patient is to undergo electrosurgery. Respiration is also potentially influenced by phonation and coughing, which have no relevance to cardiac output.

Evoked QT Interval–Based Pacemakers In 1920, Bazett67 showed that changes in heart rate induced by exercise result in a progressive shortening

515

of the QT interval on the surface electrocardiogram. The normal QT interval was found to be longer at relatively slow heart rates than at faster rates, and a nonlinear formula to correct the QT interval for changes in heart rate was proposed. In 1981, Rickards and Norman68,69 found that QT interval shortening during exercise consisted of two components, an effect induced by exercise alone and an effect of an increased heart rate. They measured QT intervals during exercise in patients with normal sinus rhythm (in whom the QT interval was influenced by both factors), during atrial pacing at different rates with the patients at rest (a pure heart rate influence), and during exercise in patients with VVI pacemakers (a pure exercise influence). These observations led to the design of a cardiac pacemaker that uses the QT interval to modulate the pacing rate.70 The first QT interval–driven, rate-responsive pacemaker, the TX1(Vitatron Medical, Dieren, The Netherlands), was implanted in 1982. Experience gained with this and later models have proved the clinical applicability of this concept and have led to a series of pacemakers characterized by progressive improvement in rate-modulating behavior (Table 16-2). Physical Principle With exercise or psychological stress, the metabolism of the myocardium and the heart rate (sinus node) rise, mainly as a result of adrenergic stimulation. Available data indicate a relatively strong linear correlation between atrial rate and cardiac sympathetic activation. The ionic currents responsible for cardiac depolarization and repolarization periods parallel these changes. At low levels of exercise, the catecholamine release is relatively low, and the cardiac rate increase is primarily due to vagal withdrawal. This implies that as a biosensor, the QT interval (catecholamine influence) may be slow to respond at the start of exercise, although its major dependence on the sympathetic nervous system should result in its being a specific sensor. Indeed, a correlation between the QT driven pacing rate and the level of circulating adrenaline of more than 0.9 was found in a group of 9 patients in one study.71 To study the influence of various factors (e.g., drugs, autonomic influence, heart rate) on cardiac repolarization, investigators have looked for formulas to describe

Evolutions of Different Generations of QT Interval–Driven Rate-Responsive Pacemakers TABLE 16-2.

Year

Model*

Improvement

Technology

1985

Quintech TX

T-wave detection T-wave detection

Fast-recharge pulse Dual fast-recharge pulse

1988

Rhythmyx

Slope programming Onset of rate response

QT interval combined with activity

1992 Current device

Topaz Diamond, Selection

Sensor specificity

Sensor cross-checking

*All manufactured by Vitatron (a subsidiary of Medtronic, Inc., Minneapolis, Minn.).

516

Section Two: Clinical Concepts

the QT interval–heart rate relationship under these different circumstances. Several studies have shown that Bazett’s formula is relatively accurate at heart rates between 60 and 100 bpm but is not correct in describing the relationship between the QT interval and heart rate over a wider range of rates because of the adrenergic influence on the QT interval.72-74 Figure 16-22 shows electrocardiographic data obtained with the atrium paced at a constant rate of 130 bpm, and the QT interval measured before and after the administration of isoproterenol. The QT interval clearly shortens, independent of the heart rate, after the administration of the drug. A different QT interval–heart rate relationship was observed after administration of propranolol,

I

II

III 390 msec

460 msec

V1

Control

Isoproterenol 4 μg/min

High Right Atrial Pacing PCL⫽450 msec Figure 16-22. Fixed-rate atrial pacing at a pacing cycle length (PCL) of 450 msec and the influence of a catecholaminemimicking drug, isoprenaline, on the length of the QT interval. (From Browne KF, Prystowsky E, Heger JJ, Zipes DP: Modulation of the Q-T interval by the automatic nervous system. PACE 6:1050, 1983.)

a β-adrenergic blocker.74 There are also large interindividual differences in the QT interval–heart rate relationship. A later study reported a curvilinear relationship between QT interval and heart rate, with a small QT interval change at low heart rates and a larger QT change at high heart rates.75 Sensor The QT interval in rate-adaptive pacemakers is defined as the interval between the pacing stimulus and the evoked endocardial T wave. Detection of the evoked T wave should occur by means of the same electrode that is used for pacing. After a conventional pacing stimulus, a slowly decaying voltage can be observed with an amplitude of several hundred millivolts that gradually dissipates over a period of more than 300 msec, interfering with normal assessment of the T wave. These polarization after-potentials can be minimized by using a fast recharge. After a blanking period of 200 msec to avoid detection of an evoked R wave, the first negative derivative of the endocardial signal allows the pacemaker to sense the downslope of the evoked T wave (Fig. 16-23).68,69,76 Boute and colleagues77 used this method to evaluate the reliability of evoked T-wave sensing. T-wave sensing was possible in 99.5% of patients (n = 368). Mean evoked T-wave amplitude was 3.0 ± 1.3 mV at implantation and 2.2 ± 0.9 mV 3 months later, thus allowing for reliable sensing with a maximum T-wave sensitivity of 0.5 mV. Older, longterm leads tend to show slightly smaller T-wave amplitudes (1.6 ± 0.6 mV, vs. 2.3 ± 0.9 mV with newly implanted leads), mainly as a result of their electrode characteristics, such as larger surface area and nonporous surface structure.77 The QT Interval Rate-Adaptive Algorithm In 1987, Baig and associates,75,78-80 conducting a reappraisal of the relationship between the evoked QT interval and ventricular pacing rate, found a nonlinear relationship between pacing and evoked QT intervals in individual patients. They found that the extent of QT interval shortening is least at low heart rates. This finding resulted in the development of a new rate-adaptive algorithm that featured a rate-dependent slope—

Paced evoked response First derivative T-wave sensitivity Blanking/T-wave sensing window

2 mV 100 ms

Figure 16-23. Recording of the paced evoked response from the pacing electrode using a dualfast-recharge technique to eliminate polarization afterpotentials. The first derivative of the evoked T wave is compared with the programmed T-wave sensitivity.

517

Chapter 16: Sensor-Driven Pacing: Device Specifics

that is, the slope is highest at low heart rates and decreases gradually as the heart rate increases.78 The slope setting for low rates is adjusted automatically every night by measuring the QT interval at two different rates near the lower rate limit (daily learning). At the upper rate, the slope is adjusted in such a way that pacing at the upper rate occurs at the patient’s shortest QT interval. Further shortening of the QT interval during pacing at the upper rate, an indication that the patient reached the upper rate at submaximal exercise levels, causes the slope at high rates to decrease (see Fig. 5-17). Evaluation of the effectiveness of this new algorithm showed a faster initial acceleration of the pacing rate at the onset of exercise (a 10-bpm rate increase was obtained after 126 seconds, versus 255 seconds with the linear algorithm; n = 11; P = .02) and fewer instances of rate instability.79,80 Baig and associates81 also evaluated the long-term stability of the automatic slope adjustments. Slope settings were found to change considerably in the first 2 weeks, from relatively low settings initially to steeper values 2 weeks after implantation (n = 17; pacing cycle length–QT interval slope at lower rate limit changing from 3.7 msec/ msec at implantation to 5.8 msec/ msec after 2 weeks; P < .001). During the 1-year follow-up period, only minor slope adjustments were found, resulting in satisfactory and reproducible rate modulation. Vitatron Combined QT Interval–Activity Devices: Topaz, Diamond, Selection

110

110

105

105

100

100

95

95

90 85 80

Too fast (QT⫽ACT; n⫽3) Good (QT⫽ACT; n⫽30) Too slow (QT⫽ACT; n⫽12)

75 70 65 0

1

2

3

6 4 5 Time (min)

7

8

9

QT⫽ACT; n⫽30

QT⬍ACT; n⫽12

90

QT⫽ACT; n⫽3

85 80 75 70 65

Exercise level

60

Rate (/min)

Rate (/min)

The main limitation of the QT sensor is its slow rate to initiate a rate-response. Despite algorithmic improvement, the QT sensor is still limited by the relatively slow onset of rate response. In one study, rate response was observed in the recovery phase of a short exercise.82 On the other hand, an activity sensor gives an immediate rate response but is not proportional to the workload. Thus, it is logical to combine these two

sensors to give a proportional and rapid response kinetics. In the currently available QT devices, the QT sensor is used with a piezoelectric activity sensor. In addition to improving the pattern of rate adaptation, the overall sensor specificity can be improved by continuous crosschecking of the information from the two sensors. If the two sensors provide consistent information, either exercise or recovery is confirmed, and the pacing rate increases or decreases, respectively. If false-positive activity signals are received (rises in the activity counts without a change in the QT interval), the pacemaker initially increases the pacing rate. If the QT interval still does not indicate an exercise condition after about 1 minute, a function called “sensor cross checking” is activated. This slowly decreases the pacing rate toward the QT interval–indicated rate. In case of false-positive activity sensing, the pacing rate gradually returns to the lower rate limit. Conversely, when the QT interval shortens while no activity is detected, mental stress or isometric exercise is most probable. Under these circumstances, the pacemaker is designed to increase the pacing rate, although its magnitude is limited. The Topaz (Vitatron Medical) is the first dual-sensor VVIR device; QT and activity can be blended as follows: QT < Activity (ACT), QT = ACT, or ACT > QT.82 In one study, 45 patients exercised according to a modified chronotropic assessment exercise protocol (CAEP) (stage 1 was made identical to stage 2) (Fig. 16-24).83 In 30 patients, the rate adaptation was judged to be appropriate; that is, heart rate increased after 1 minute of exercise between 25% and 50% of the rate range and after 2 minutes between 50% and 75% of the rate range. In 12 patients, the initial response was too slow, and their pacemakers were reprogrammed to a blending pattern of QT < ACT; in three other patients, however, the initial response was judged to be too aggressive; these patients underwent reprogramming to QT > ACT. Sharp and coworkers84 evaluated heart rate and oxygen uptake at rest and at low exercise levels in

10

Exercise level

60 0

1

2

3

4 5 6 Time (min)

7

8

Figure 16-24. Effect of reprogramming “Sensor Blending” on the rate-response pattern. Initially, all patients exercised in the QT = ACT (activity) setting (left). Two subgroups underwent pacemaker reprogramming because of a suboptimal rate response. A second exercise test (right) confirmed that the desired effect was actually obtained.

9

10

518

Section Two: Clinical Concepts

patients with left ventricular dysfunction in the fixedrate VVI mode and compared them with VVIR pacing based on the activity sensor only, the QT sensor only, and blending of both sensors. The dual-sensor chronotropic response reproduced the theoretical linear relationship among metabolic workload, heart rate, and oxygen uptake, suggesting the usefulness of this sensor in patients with stable left ventricular dysfunction. The effectiveness of “sensor cross checking” was tested during continuous levels of false-positive activity: gentle and vigorous tapping on the pacemaker case and applying massage equipment over the pacemaker, which created excessive activity signals.82 In all cases, “sensor cross checking” prevented unphysiologic rate accelerations. The time taken for the pacing rate to decrease from its peak (85 ± 8 bpm) to about the lower rate limit was 2.7 ± 1.7 minutes with gentle tapping and 8.3 ± 2.4 minutes with the massage equipment. The ability of the combined sensor to simulate the normal sinus activity on a daily basis was studied in patients with the DDDR Diamond pacemaker.85 A special software contained an additional diagnostic feature that continuously stored the difference between the sensor and the sinus rate during a 1-month ambulatory period. Sinus and dual-sensor rates were significantly correlated (P < .001; correlation coefficients > 0.90; mean difference throughout exercise and recovery 2.8 ± 6.1 bpm). During the ambulatory period, sensor and sinus rate differences were classified according to three activity levels (see Fig. 5-13). Nearly 90% of the sensor-driven beats were within 8 bpm of the sinus rate at medium and low levels of exercise. However, the difference was larger at higher levels of exercise. Other Applications of the Evoked QT Interval Sensor The evoked QT interval sensor is one of the few clinically available sensors that reflect cardiac metabolism. Several investigators have indicated other applications for the evoked QT interval sensor, as described here. Dynamic Pace Refractory Period Because the T wave marks the end of cardiac refractoriness, it can be used to match the pacemaker’s paced ventricular refractory period, automatically and dynamically, to cardiac refractoriness.86 This behavior is implemented in the current pacemaker models and provides optimized detection of early premature ventricular contractions, especially during exercise. Under exercise conditions, the ventricular pace refractory period automatically shortens with the measured QT interval, thus allowing earlier ventricular sensing. Optimal Atrioventricular Interval In patients at rest and during pacing in the atrium at a fixed rate, a positive correlation was observed between the longest QT interval, the highest cardiac output or cardiac index, and the programmed atrioventricular

(AV) delay.87 In the implanted Diamond pacemaker, the evoked QT interval value, derived from a downloadable software, was found to be longest at the optimal AV delay as assessed by the maximum Doppler mitral valve inflow.88 Furthermore, QT interval–determined AV interval also minimized mitral regurgitation. It is of interest to see whether the optimal AV interval in cardiac resynchronization therapy can be similarly determined. Recognition of Ventricular Fusion Beats Boute and associates89 demonstrated in dogs that the amplitude of the evoked T wave significantly decreases during fusion. Recognition of ventricular fusion is important for reliable and effective operations such as automatic capture detection as well as to avoid unnecessary fusion pacing, which may influence the ventricular contraction pattern and battery current wastage. On detection of fusion, a dual-chamber pacemaker could automatically extend its AV interval to allow the conducted R waves to prevail. Finally, in patients with hypertrophic obstructive cardiomyopathy, one would like to maintain ventricular pacing to obtain consistent septal preexcitation. In these patients, detection of ventricular fusion may shorten its AV interval, thereby providing an effective therapy through full ventricular capture. Other Applications A circadian variation in the QT interval has been reported.90 The difference in the QT interval at 60 bpm between being awake and being asleep was 19 ± 7 msec,91 reflecting greater vagal tone or sympathetic withdrawal. This difference could be used to automatically decrease the LRL during those hours of the day that the patient is asleep. The evoked T-wave amplitude after cardiac transplantation has been tested in 13 patients during the immediate post-transplantation period. In 11 patients, the initial biopsy that proved rejection was associated with a significant decrease in the evoked T-wave amplitude from 1.3 to 0.6 mV (P < .005), which began 1 to 4 days before the biopsy. The paced evoked response can be used to evaluate drug-induced changes in myocardial repolarization.93 Furthermore, Donaldson and coworkers94 showed that subendocardial ischemia can be detected with the paced evoked response.

Unipolar Ventricular Impedance (Closed-Loop Stimulation Sensor) Physiologic Principle Cardiac contractility of the ventricle increases during catecholamine stimulation, as occurs during exercise and emotional stresses. In the absence of an adequate rate response, exercise induces a higher contractility, which decreases when rate response is adequate, thus establishing a negative feedback loop and a new

Chapter 16: Sensor-Driven Pacing: Device Specifics

519

difference between the exercise and baseline impedance waveforms is converted to a pacing rate using an “Auto Response Factor, ” until SURL is reached for the first time. This “Auto Response Factor” is continually adaptive and is patient specific. Thereafter, the response is determined by a programmable “Exertion Threshold Rate” (ETR) [very low, low, medium, high, and very high]. The ETR, acting through the “Auto Response Factor,” determines that 80% of the heart rate will occur below the ETR and 20% above the ETR (see Fig. 16-26B). A young and active individual will probably require a higher ETR than an inactive sedentary elderly person. Again, the type of rate profile attained will be determined by the cardiac condition, the patient’s physical state, and the patient-specific “Auto Response Factor,” even though the ETR is programmed similarly. The maximum ETR is limited to less than 80 bpm above the LRL. Although this is strictly not a dual-sensor pacemaker, the incorporated accelerometer performs an “on-off” function to decide on the acquisition of baseline CLS waveform and the CLS-driven response. Full CLS driven-rate according to the programmed ETR will be allowed only if the accelerometer registers exercise; otherwise only 20 bpm above the LRL is allowed to enable a nonexercise increase rate. In case of neurocardiogenic syncope, this rate-limiting algorithm can be inactivated for full overdrive response, as occurs during a syncopal episode.

steady-contractility state. On the other hand, rises in pacing rate per se can increase contractility, the socalled Treppe effect, although this effect has not been important in clinical practice. Sensor and Algorithm The CLS sensor is based on unipolar impedance at the tip of a pacing lead.95 Subthreshold pulses of automatically selected outputs (ranging from 100 to 400 μA), with a biphasic duration of 46 msec, are emitted 50 to 300 msec after a sensed (Vs) or paced (Vp) ventricular event. Because two pulses are required for an impedance measurement, eight samples are taken per cardiac cycle (Fig. 16-25A). During diastole (immediately after a Vp or Vs), there is significant amount of blood around the electrode tip, and the impedance is low. On the other hand, as contraction occurs, the walls surrounding the electrode tip get closer, and impedance rises. A baseline waveform occurs, which depends on the conduction state of the heart: AsVs, AsVp, ApVs, ApVp, where As and Ap refer to atrial sensing and pacing, respectively. In addition, the impedence waveform changes with the heart rate and the time of the cardiac cycle. Because the field strength falls off rapidly with distance from the lead tip, approximately 90% of impedance is reflected in a diameter of 1 cm from the tip. Thus, the effect of respiration is limited. Baseline CLS waveforms are acquired only when the associated accelerometer indicates no activity, and a waveform is discarded within 48 hours if not referenced. An average template of the baseline CLS waveform will take 2 to 3 days to optimize. As contractility increases during exercise, unipolar impedance changes (Fig. 16-26A). The time-integrated

Biotronik Closed-Loop Stimulation Pacemakers: Inos, Protos Schaldach and Hutten95 took immediate measurements of the CLS parameter (previously known as ventricular Time (ms) 240 resting

Units

200 160 120

exercise

80 40 50

A

100

150

200

250

300

Milliseconds after a Sensed or Paced Ventricular Event 8% ETR⫽73

7%

ETR⫽80

ETR⫽92

Figure 16-25. A, Changes in closed-loop stimulation (CLS) parameter during exercise. The hatched area represents the difference between baseline and exercise CLS waveforms, and is converted to a rate using the “Exertion Transfer Rate” (ETR). B, The impact of ETR on the rate response of CLS. “Medium” ETR corresponds to a rate of 80 bpm (20 bpm above the lower rate limit). Programming this rate will ensure a rate response of 80% less than this rate and 20% above this rate. A higher or lower ETR will result in different rate responses.

% Rate

6% Very low

5% 4%

Medium

3% Very low

2% 1% 0%

B

60

70

80

90

100

110

120

130

520

Section Two: Clinical Concepts VVI-paced rhythm

Impedance (a.u.)

Intrinsic rhythm 150

180

140

160

130

140

120

120

20 μg/kg/min 10 μg/kg/min

110

100

5 μg/kg/min

100

80

Baseline

60

90

40

80 50

100

A

150

200

250

50

300

100

150

200

250

Time after stimulus (ms)

Time after event (ms)

Intrinsic events

VVI-paced events

1.0

CLS (normalized)

0.8

0.6

0.4

0.2

300

Figure 16-26. A, The curves on the left show typical impedance curves at different dobutamine doses in the same patient during intrinsic rhythm, which in this case was sinus rhythm with a narrow QRS complex. Note that the shapes of these impedance curves were substantially different from those during ventricular paced rhythm (right) in the same patient. However, in both situations, a region of interest could be defined that provided an optimal curve separation. For intrinsic and VVI-paced rhythm, the difference in slope (measured slope – baseline slope of the corresponding curve pattern) was used to calculate a rhythmspecific sensor signal for each inotropic state of the heart. a.u. = arbitrary units. B, These graphs show the correlation between CLS and dP/dtmax separated for intrinsic (left) and VVIpaced rhythm (right). There was a comparably strong correlation between the CLS signal and right ventricular dP/dtmax for both types of rhythm. (From Osswald S, Cron T, Gradel P, et al: Closedloop stimulation using intracardiac impedance as a sensor principle: Correlation of right ventricular dP/dtmax and intracardiac impedance during dobutamine stress test. PACE 23:1502, 2000.)

y⫽0.9672x⫹0.0186 r2⫽0.9247

y⫽0.9385x⫹0.0036 r2⫽0.9252 0 0.2

B

0.4

0.6

0.8

1.0

0.2

0.4

0.6

0.8

1.0

dP/dt (normalized)

inotropic index) in 82 patients with long-term implanted unipolar ventricular leads at the time of pulse generator replacement. A wide fluctuation of baseline impedance was observed (500 Ω to 1500 Ω), although CLS fluctuated by about 4 Ω to 25 Ω, with a good correlation between CLS and the baseline impedance. Using an investigational VVIR pacemaker with telemetry (Biotronik Neos-PEP, Biotronik, Lake Oswego, Oregon) in 158 patients, these investigators95 demonstrated that rate adaptation can be achieved with this sensor. In individual patients, it was reported that rate adaptation close to that of the sinus node was observed with this sensor during exercise, although it was necessary to individually adjust the CLS detection algorithm (Fig. 16-27). A delay in the onset of an increase in pacing rate was observed, compared with onset in the normal sinus rhythm in some patients, possibly because the normal sinus rate at the onset of exercise is due to parasympathetic withdrawal rather than sympathetic increase. A clinical study involving 205 patients was performed to evaluate the CLS pacemaker.96 A significant

proportion of these patients were young subjects with complete AV block due to Chagas’ disease. Satisfactory rate modulation was reported in 93% of the patients. In the remaining 7% of patients, rate adaptation could not be achieved because of such factors as poor exercise tolerance, severe myocardial dysfunction, and intermittent AV conduction. In a multicenter study that involved 178 VVIR (Biotronik Neos-PEP) and 84 DDDR pacemakers (Biotronik Diplos-PEP and Inos2DR), physiologic rate adaptation was possible in 93% and 96% of patients with these devices, respectively.97 Apart from exercise rate response, this study also involved mental stress testing using color-word matching and the infusion of inotropic agents. A moderate level of rate response was documented in some patients with CLS pacemakers during these nonexercise stresses. During stress echocardiography, the increase in CLS-driven pacing rate was similar to the changes observed with sinus rhythm.98 In some patients with the CLS pacemaker undergoing angioplasty, balloon inflation in the artery supplying the myocardium around the pacing electrode led to

Chapter 16: Sensor-Driven Pacing: Device Specifics Sinus rate

Pacing rate

120

130 Exercise

Recovery

120 Pacing Rate (bpm)

110 HR (1/min)

100 90 80 70 60 0

2

56W

25W 4

6

8

A

75 W 10 12

100 90 80 70 60

14

16

18

20

22

50 50

Rest Supine Sitting Upright Supine

Recovery Supine

HR (1/min)

HR (1/min)

Exercise

100

80

60

0

2

4

6

56W 28W 8 10 12

70

80 90 100 Sinus Rate (bpm)

CNS-compiled Pacing Rate

Pacing rate

140

120

60

B

Time (min)

Sinus rate

110

14

16

Time (min)

C

D

125 120 115 110 105 100 95 90 85 80 75 70 65 60

Reset

110

120

130

Mean Blood Pressure

Exercise

1 2 3 4 5 6 7

Recovery

1 2 3 4 5 6 7

120 110 100 90 80 70

MBP (mm Hg)

Rest

50

521

25 W 50 W 75 W 100 W

Time (min) Figure 16-27. Simultaneous recording of the sinus-derived and closed-loop stimulation (CLS)–derived pacing rates in a patient with atrioventricular block with a Neos-VVIR pacemaker (Biotronik GmbH & Co., Berlin). A, Appropriate rate response during graded cycle ergometry. B, Good correlation of sinus rate and CLS-driven pacing rate. C, Inappropriately adjusted CLS detection with failure of rate response. D, Maintenance of constant mean arterial pressure during exercise with appropriate rate response. HR, heart rate; MBP, mean blood pressure. (From Schaldach M, Hutten H: Intracardiac impedance to determine sympathetic activity in rate responsive pacing. PACE 15:1778, 1992.)

a decrease in CLS, suggesting that CLS reflected a change in local contractility of the myocardium. Within an increase in pacing rate from 70 to 90 bpm at rest, there was no observable change in CLS. This finding suggests that within the rate range studied, there is no potential positive-feedback loop. Osswald and associates99 reported that the intrinsic and ventricular paced QRS has substantial influence on the rate response. In their study, infusion of dobutamine increased both intrinsic and paced CLS parameters in a dose-dependent manner, and the changes in CLS parameter in either case were proportional to the measured dP/dt. This finding suggests that rate adaptation is possible for both intrinsic and paced QRS complexes, provided that a correct reference waveform is used. However, the rate-adaptive behaviors of intrinsic and paced acquired CLS conditions have not been formally compared. During cardiopulmonary testing with the older version of a CLS device that has an activity

sensor (Inos), Cook and coworkers100 found excellent correlation between the pacing rate with the measured Vo2 and cardiac output (P < .01 in both instances). The rate responses during daily activities have also reported to be appropriate in a multicenter study,101 with rate responses of 104 ± 18 bpm, 95 ± 15 bpm, and 88 ± 11 bpm during ascending stairs, descending stairs, and walking, respectively. The CLS level is also correlated with skeletal muscle sympathetic nerve activity.102 Furthermore, rate-adaptive pacing using the CLS sensor has been reported to simulate the rate response during phases II and IV of the Valsalva maneuver.103 On the other hand, a delayed increase of heart rate was observed after nitroglycerine infusion.103 These data indicate that the CLS sensor responds, at least in part, to the changes in autonomic tone. There is interest in the possible use of this sensor to detect posture. Passive upright tilt, which depletes intravascular volume, increases the inotropic state of

522

Section Two: Clinical Concepts

the heart.104 This may cause a false-positive rate increase. This is now prevented by using an associated accelerometer to detect exercise. In a multicenter study entitled INotropy Controlled pacing in Vasovagal SYncope (INVASY), 105 50 patients with severe vasovagal syncope and positive head-up tilt test response were randomly assigned to DDD-CLS or DDI mode at 40 bpm.105 Seven of the 9 patients in the DDI arm experienced syncope within 1 year, whereas only 4 of 41 patients in the DDD-CLS arm had presyncope. The investigators recommended the efficacy of this approach, although a placebo effect of pacing was suspected to have occurred in 22% of patients. Of special interest is the use of CLS sensor rate-adaptive pacing in patients who had bradycardia and heart failure. From a preliminary study,106 in which the pacing rate was “titrated” according to the cardiac contractility, Bailey and Hull106 suggested that CLS pacing may avoid (or even prevent) worsening of heart failure, preserving ejection fraction, and improving functioned class. This possibility has to be confirmed in further studies. The use of CLS in the left ventricle, as in a cardiac resynchronization therapy device, would be of interest. Advantages and Limitations The CLS sensor appears to be an interesting sensor to measure contractility of the heart, and CLS can be achieved using a conventional ventricular pacing electrode. The demands on pacing energy are acceptable. As a contractility sensor, it is sensitive not only to exercise but also to nonexercise requirements, and it may therefore be used for monitoring cardiac contractility for non-rate augmentation purposes. Like the QT sensor, the CLS sensor can be used only in a pacing mode that incorporates a ventricular lead. The effects of pacing rate on CLS have not been completely studied, and the difference in changes in preload and the CLS during pacing and intrinsic conduction may affect the resultant rate response. It is likely that CLS is affected by right ventricular ischemia or cardioactive medications. These factors may influence rate adaptation with the CLS sensor, although automatic adjustment may permit long-term function. CLS sensors are not suitable for a small proportion of patients because of severely impaired right ventricular function,

although it may be possible to identify such patients preoperatively.

Peak Endocardial Acceleration Sensing Sensor and Algorithm The contractile state of the heart can be identified from the maximal velocity of shortening of unloaded myocardial contractile elements, which can be measured with a catheter-tip accelerometer attached to the ventricular wall. The PEA represents the endocardial vibration measured by the accelerometer in the right ventricle during the isovolumetric contraction phase of the ventricles. This signal is in close relationship to the intensity of the first heart sound. The sensor, developed by Sorin Biomedica Cardio, S.p.A., is termed the BEST (Biomechanical Endocardial Sorin Transducer) sensor. The microaccelerometer consists of an acceleration sensor built into an indeformable capsule located on the tip of a standard unipolar ventricular pacing lead. The lead is placed against the right ventricular wall so as to be sensitive to its acceleration and insensitive to the pressure of blood and myocardium (Fig. 16-28). This system has a frequency response of up to 1 kHz and a sensitivity of 5 mV/G (1 G = 9.8 m/sec/sec). In preliminary experience in animals, and using an external system and an implantable radiotelemetry system, Occhetta and colleagues107 found that the PEA was not affected by heart rate but was significantly raised by emotional stress, exercise stress testing, and inotropic stimulation. The PEA signal changes in parallel to the maximal left ventricular dP/dt and appears to measure the global left ventricular contractile performance rather than the regional mechanical function.108,109 The PEA signal occurs at 150 msec after the R wave and corresponds to the isovolumetric contraction phase of the left ventricle (Fig. 16-29). This is also called the PEA-1, is the signal for cardiac contractility, and is proportional to the positive dP/dt during inotropic stimulation (r = 0.83).110 A smaller signal also occurs in the 100-msec period after the T wave, the socalled PEA-2, which corresponds to the isovolumetric left ventricular relaxation. PEA-2 is related to peak

Micro Mass PEA Pace/Sense Electronics Figure 16-28. The Biomechanical Endocardial Sorin Transducer (BEST) sensor (Sorin Biomedica Cardio, S.p.A., Sorin Group, Milan) consists of a microaccelerometer located inside a rigid capsule at the tip of a standard ventricular pacing lead. This is connected to a triple header of the device. PEA, peak endocardial acceleration.

523

Chapter 16: Sensor-Driven Pacing: Device Specifics

negative dP/dt (r = 0.92) and aortic diastolic pressure (r = 0.91).110 The BEST Sensor Rate-Adaptive Pacemakers: MiniLiving D and MiniLiving S Sorin manufactures the MiniLiving D and MiniLiving S pacemakers, which use a dedicated unipolar ventricu-

ECG

LV dP/dt

Ao Press

LV Press

EA PEA-II

PEA-I

Figure 16-29. Electrocardiographic (ECG), left ventricular dP/dt (LV dP/dt), aortic pressure (Ao press), and E: accelerometer tracing (EA) recordings in a sheep with the Biomechanical Endocardial Sorin Transducer (BEST) sensor. The peak of the accelerometer tracing during the isovolumic diastole (PEAII) occurs in a 100-msec period after the T wave on ECG. (From Plicchi G, Mercelli E, Parlapiano M, et al: PEA I and PEA II based implantable haemodynamic monitor: Preclinical studies in sheep. Europace 4:49, 2002.)

lar pace/sense lead, that is IS1 compatible. Together with the two electrodes for PEA signals, a tripolar connector is necessary. These devices also incorporate an activity sensor using the vibration from a mercury ball (gravitational sensor). Vibrations during exercise result in greater excursion frequency and magnitude of the mercury ball, leading to an increase in output from the vibration sensor. The MiniLiving pacemakers can be programmed in either the single-sensor or dual-sensor mode. In singlesensor, PEA mode, continuous PEA signal collected over time is used as a reference, with which the change in PEA is compared. A linear pacing rate to PEA signal rate-adaptive curve is used to control the rate response. Six different rate-response curves are available. If the maximum level of PEA occurring over time is below the maximum PEA signal change permitted by the curve, a more aggressive slope is advised. On the other hand, if the PEA exceeds the maximum change in the curve, a gentler slope will need to be programmed. Optimal PEA programming occurs when the maximum PEA change equals the maximum PEA level allowed for that rate-response slope (Fig. 16-30A). In dual-sensor mode, the combined output from the two sensors is used to drive the rate to a middle rate level; further rate increase will be effected with output from the PEA sensor alone (see Fig. 16-30B). The rise time and recovery times are separately programmable. In addition, prolonged absence (about 45 minutes) of signals from both the gravitational and PEA sensors allows the basic rate to drop to the programmable rest rate. On the other hand, once the sensors are active, approach to rate response will be rapid. The device also incorporates short-term (20 minutes) and long-term (3, 6, 12, and 24 hours) trends of both the PEA and heart rate signals to assess rate adaptation. In addition, the automatic AV search function

PEA  Gravitational Sensor

PEA Single Sensor

140

140

120

120

100

Pacing rate (bpm)

Pacing rate (bpm)

More PEA less aggressive Peak aggressive

PEA reference

80 60 40 20

PEA only

100

Middle rate

PEA and Gravitational sensor

80

Basic rate

60 40 20

0

A

SURL

1

2

3 PEA(g)

4

5

0

B

1

2 3 Time (min)

4

Figure 16-30. A, Rate-adaptive algorithm of the MiniLiving (Sorin Biomedica CRM) at nominal settings. The ΔPEA (change in peak endocardial acceleration) allowed is 3 g to reach the sensor-driven upper rate limit (SURL) of 140 bpm at about 4 g. A more aggressive slope allows the SURL to be reached at a ΔPEA of 3 g, whereas a gentler curve may require a ΔPEA of 5g to reach the SURL. B, In a combined-sensor mode, initial exercise to a middle rate is contributed by the combined sensors, whereas the approach to the upper rate is contributed by the PEA sensor alone.

5

524

Section Two: Clinical Concepts

allows the AV interval that corresponds to the minimum level of PEA to be derived automatically. Clinical Results Preliminary studies of pacemakers using the PEA sensor have shown a good correlation between the sinus rate and PEA sensor–indicated rate during daily life activities and submaximal stress test.111,112 Similar results were obtained in patients being tested during electrophysiologic studies through the use of an external system; the changes in PEA were found to be linearly related to the right ventricular dP/dt during dobutamine infusion.107,113 PEA is not affected by ischemia of the area in which the senor is attached. The performance of the PEA sensor with exercise-related and non–exercise-related stress was studied in 17 patients. The PEA sensor showed a quick response to exercise, and it also responded to both hand grip and the Valsalva maneuver.114 During maximal treadmill exercise testing in 15 patients, the rise in PEA was found to have a good correlation with the increase in exercise workload (change in PEA to Mets, r = 0.97).115 The correlation was best at the higher stages of exercise; the PEA changes at lower levels of exercise are less discriminative.114,115 PEA signals have been used to monitor hemodynamic function and for programming the AV interval. In 13 patients with end-stage heart failure in whom DDDPEA devices with custom lead arrangements were implanted,116 PEA level during right ventricular, left ventricular, and biventricular pacing were compared. Both left ventricular and biventricular pacing resulted in higher stroke volume than right ventricular pacing (+21 and +37%, respectively), and mean PEA changes over a 15-minute duration were also higher (+43 and +38%, respectively). In addition, there appears to be a minimum PEA level at the optimal AV interval, and this finding has shown some promise for automatic detection of the optimum AV interval in a dual-chamber device.117 An increase in PEA during head-up tilting has been observed, and the use of PEA-driven overdrive pacing in patients with vasovagal syncope has been reported.118 Patients randomly assigned to DDDR pacing have a lower frequency of syncope than those assigned to DDI pacing. These data suggest the potential use of PEA sensor for hemodynamic monitoring. The role of PEA-2 in assessing diastolic function and aortic pressure remains to be investigated.110 Advantages and Limitations The PEA sensor is a proportional sensor that shows good correlation to workload, especially at the higher ranges. During daily activities in one study, the rate response was correlated with sinus rate, but the actual rates achieved were significantly lower.119 Apart from using a higher baseline rate, the combination with the gravitational sensor may be advantageous for lower levels of exercise loads. There are many preliminary communications on the ability of the PEA sensor to monitor hemodynamics. Although these reports are interesting and potentially important, they must be validated by larger

trials. The PEA sensor is limited by the need for a specialized lead, as well as by concern about its longer-term stability and the use of this lead at the time of a pacemaker replacement in which a different sensor is used.

Current Combined-Sensor Devices Experience with sensors has suggested that fastresponding sensors such as activity sensors are not proportional at higher levels of workload, whereas a proportional sensor is usually slow in response. Furthermore, single sensors may be limited by insensitivity to non-exercise stress and are liable to be interfered with by nonphysical causes. Thus, it is logical to enhance the rate-response profiles of the various sensors by combining two or more sensors in a single pacemaker. There are two principles of sensor combination, sensor blending and sensor cross-checking (see Chapter 5). Sensor blending involves combining the sensor-driven rates from individual sensors in a certain ratio. This can be the “faster win” method in which the higher rate is chosen as the dual-sensor rate, or ratios of the individual rates are added together to compile the ultimate rate response. Sensor cross-checking enhances the specificity of each sensor. If a more specific sensor registers no exercise or physiologic stresses, changes in the other, less specific sensor are ignored or its response is attenuated. In some situations, sensor blending and cross-checking may not be clear-cut but occur in combination to derive the rate adaptation. The instrumentation of current dual-sensor devices is summarized in Table 16-3. Details of the combinations of QT, activity, CLS, and PEA sensors have already been presented. Kappa 400 In Medtronic’s Kappa 400 pacemaker, a piezoelectric sensor is used for activity sensing, and MV is sensed from a bipolar ventricular lead. Differential sensor blending is used. Up to the ADL rate, activity input predominates, whereas MV-driven pacing will predominate at the SURL (see Fig. 5-15). Activity and MV sensors are checked against each other. In the absence of piezoelectric sensor indications of exercise, MV pacing will only reach the ADL rate, and vice versa. Only when both the MV and activity sensors signify exercise will pacing above the ADL rate occur. The activity sensor can be programmed with the use of the conventional threshold and slope. In the dualsensor mode, rate adaptation is achieved automatically using the “Rate Profile Optimization.” This requires the input of the ADL and ER rates, together with the percentage of time spent in each rate (range 1 to 5), similar to the programming of the Kappa accelerometer devices. Clinical Performance The dual-sensor rate response has been reported to be reliable for both maximal and submaximal activities,

Chapter 16: Sensor-Driven Pacing: Device Specifics

525

TABLE 16-3.

Types of Dual-Sensor Pacemakers in Current Use

Sensors

Manufacturer

Models

Sensors

Algorithms

Cross-checking

Automaticity

Medtronic, Inc.

Kappa 400

ACT = piezoelectric MV = impedance

Pulsar Max Insignia

ACT = accelerometer MV = impedance

ACT(0) and MV(+): up to ADL rate; ACT (+) and MV(0): up to ADL rate ACT(0) and MV(+): MV rate; ACT(+) and MV(0): limited rate

“Rate Profile Optimization”

Guidant

ELA Medical

Chorus Talent Symphony Rhapsody

ACT = accelerometer MV = impedance

Blending: ≤ADL range: ACT + MV; ADLER range: mainly ACT Blending: Low heart rate: ACT 80%, MV 20%; High heart rate: ACT 40%, MV 60% No Blending: MV-determined rate response if ACT indicates exercise

ACT(+) and MV(0): initial limited rate response; ACT(0) and MV(+): rate recovery

Automatic matching MV sensor to LRL and SURL

ACT + QT

Vitatron

Topaz Diamond Selection

ACT = accelerometer QT = unipolar evoked QT

Blending: ACT > QT ACT = QT ACT < QT

ACT(0) and QT(+): limited rate ACT(+) and QT(0): decrease to LRL

Automatic matching QT sensor to LRL and SURL

CLS + ACT

Biotronik GmBH

Inos Protos

CLS = unipolar ventricular impedance ACT = accelerometer

No blending: No ACT rate contribution Rate response determined by CLS only

ACT(0) and “Auto Response CLS(+): Factor” adjusts limited rate CLS data to reach response; ACT(+) rate distribution and CLS(0): no determined by rate response the programmed “Exertion Threshold Rate”

PEA + ACT

Sorin Biomedica CRM

MiniLiving

PEA = accelerometer at ventricular lead tip ACT = gravitational sensor

Blending: Up to Middle rate: PEA + ACT > Middle rate: PEA only

Nil

ACT + MV

“AutoLife style”

Manual adjustment to match peak PEA from trend data to the desired SURL

ACT, activity sensor; CLS, closed-loop stimulation sensor; LRL, lower rate limit; MV, minute ventilation sensor; PEA, peak endocardial acceleration sensor; QT, QT interval–driven sensor; SURL, sensor-driven upper rate limit.

and resistant to non-physiologic interference.120 Compared with MV sensor alone, dual sensor mode reduces oxygen deficit acquired during exercise by enhancing the initial rate response.121 “Rate Profile Optimization” was found to be a useful method for rate-adaptive programming, and comparable with manual programming.122 On the other hand, in a series of 11 patients followed up for 3 years with the accelerometer programmed by “Rate Profile Optimisation,”28 repeated programming was still necessary to optimize the response of the activity sensor to treadmill exercise. Insignia and Pulsar Max An accelerometer activity sensor is integrated with the MV sensor in Guidant’s Insignia and Pulsar Max pacemakers. A differential sensor blending is used. At low heart rate, approximately 80% of the blended sensor rate is contributed by the accelerometer, and 20% by MV sensor. These proportions change to 40% and 60%, respectively, near the SURL. In addition, if the MV-

indicated rate is higher than the accelerometer rate, the dual-sensor rate follows the MV level. Cross-checking occurs only against the activity sensor, because MV is considered to be more specific. An interim rate is allowed only in the event that the activity sensor alone indicates exercise and the MV sensor is inactive. The programming of the dual sensor involves either a manual adjustment or use of the automatic sensor adjustment. In brief, after a “Sensor Rate Target” is chosen on the basis of the “Expert Ease” system, the combined sensor is adjusted according to the average of the maximum combined sensor activities and its difference from the “Sensor Rate Target” (see Fig. 16-15). Individual adjustments can be made with the use of trend data during a structured exercise test. In the Insignia Ultra pacemaker, automatic adjustment of the dual sensor (or the MV only) is available. The “MV Response Factor” (10 nonprogrammable slopes) is adjusted on a weekly basis, on the basis of the detected “MV Max Long-term” value. The latter is

526

Section Two: Clinical Concepts

derived from the maximum impedance–derived MV that has been confirmed by the accelerometer. The accelerometer further identifies this MV as occurring during mild, moderate, or vigorous exercise before “deciding” whether the MV is the maximum that occurs during vigorous exertion. The “MV Max Long-term” value is updated by 10% if it exceeds the value from last week, or decreased by 15% if it is above. Note that the “MV Max Long-term” value is linked to the age-predicted heart rate and is independent of the programmed SURL. In the MV mode, the accelerometer contributes only to the “MV Max Long-term” measurements. In the dual-sensor mode, the accelerator’s input is differentially blended as previously described (Fig. 16-31). Clinical Results A preliminary study has addressed the use of automatic sensor adjustment in the Pulsar Max. The sensitivity of the MV sensor was changed in 36% of patients, whereas the accelerometer reactivity was increased in almost all patients. In another study, Pieragnoli and colleagues124 randomly assigned 120 patients with Insignia pacemakers to accelerometer single-sensor mode, MV singlesensor mode, or dual-sensor mode, each for a 3-month period. Using the implanted “Activity Log” to determine the mean percentage and intensity of activity, these investigators assessed quality of life and New York Heart Association (NYHA) functional classes at the end of each period. Overall, either single-sensor DDDR mode led to better “Activity Log,” quality of life, and NYHA scores than DDD pacing, but there was no difference between the two sensors. Dual-sensor mode did not provide additional benefit. Results of this study, which may have been limited by the prolonged triple-crossover design, nevertheless suggest that rate-adaptive pacing is more beneficial than DDD mode, but the clinical differences between sensors and their combinations in effecting a better adaptation are likely to be small. Chorus, Talent, Symphony, Rhapsody An accelerometer and a MV sensor are combined in the ELA Medical’s dual-sensor pacemakers. Sensor blend-

ing and cross-checking are both operative to effect rate adaptation during exercise. When the accelerometer is active but MV has not increased, as may occur at the beginning of exercise, rate response occurs according to a fixed activity response curve to a limited rate (Fig. 16-32A). When MV increases, rate response follows the MV sensor–driven rate. Persistent absence of accelerometer signal is regarded as cessation of exercise and this drives the pacing rate in a recovery curve to the LRL even though MV remains higher than baseline. An “Autocalibration” function enables automatic matching of the MV sensor data to the upper and lower rate limits (see Fig. 16-32B). If the SURL is reached but the sensor level is still rising, the rate response is reduced in steps so that the SURL is reached later. Lower rate is considered when the MV value over 128 respiratory cycles is below the mean MV over the last 24 hours. In addition, persistent absence of either sensor signal forces the rate response to return to the LRL. Clinical Results Results of a multicenter study of 81 patients with the Talent pacemaker have been reported.61 In patients who underwent exercise stress testing at 1 month of follow up, sensor-driven rate was found to be well correlated with the sinus rate (r = 0.92 ± 0.07; P < .001), with the slope of linearity of 1.0 ± 0.2. With metabolic reserve used to relate to heart rate reserve, a slope of 1.1 ± 0.2 was obtained, suggesting a close relationship between the dual-sensor rate and the metabolic workload.

Conclusions Current rate-adaptive devices have evolved to incorporate newer sensor technologies and increasingly sophisticated algorithms to enhance rate adaptation. Because single-sensor rate response may be limited by the speed of response, lack of proportionality, sensitivity, and specificity, sensors have been combined to enhance the overall rate-adaptive behavior. Almost all single- and

220-Age

Pacing rate (bpm)

Normal heart rate zone

MV Max Long term

SURL Fine adjustment More response

Coarse adjustment Starting response

LRL

Less response Rest

Maximum Exercise

Figure 16-31. “AutoLife style” sensor adjustment of the Insignia Ultra Maximum MV (Guidant) will be updated on a weekly basis and confirmed by the accelerometer, indicating vigorous exercise. Note that the sensordriven upper rate limit (SURL) is independent of the “MV Max Long-term.” LRL, lower rate limit; MV, minute ventilation. See text for further discussion.

Chapter 16: Sensor-Driven Pacing: Device Specifics

527

Autocalibration

Sensor Blending

Y Rate

SURL

Rate

SURL

LRL

X LRL

Rest

Exercise

Rest MV/ACT

Rest

Exercise MV/ACT

ACT

MV

A

B

Figure 16-32. Dual-sensor algorithm of the Symphony (ELA Medical, Sorin Group). A, A combination of activity (ACT) and minute ventilation (MV) inputs during exercise. B, “Autocalibration” of the sensor. If the maximum sensor level is reached before the sensor-driven upper rate limit (SURL; indicate by X), then the slope is adjusted in steps to Y to match the SURL to the maximum sensor level. LRL, lower rate limit.

dual-sensor devices are now self-programmable, and physician input can be kept to a minimum. On the other hand, understanding of the interplay between sensors and their algorithms will be needed to fine-tune rate response in some patients. Apart from avoiding false rate acceleration, the clinical role of dual-sensors over single-sensor pacing remains to be addressed. REFERENCES 1. Richards AF, Donaldson RM: Rate-responsive pacing. Clin Prog Pacing Electrophysiol 1:12, 1983. 2. Benditt DG, Milstein S, Buetikofer J, et al: Sensor-triggered, rate-variable cardiac pacing: Current technologies and clinical implications. Ann Intern Med 107:714, 1987. 3. Dahl JD, inventor: Variable rate timer for a cardiac pacemaker. US patent 4 140, 32. February 20, 1979. 4. Anderson K, Humen D, Klein GJ, et al: A rate variable pacemaker which automatically adjusts for physical activity. PACE 6:12, 1983. 5. Lau CP, Stott JR, Toff WD, et al: Selective vibration sensing: New design of activity-sensing rate-responsive pacing. PACE 11:1299, 1988. 6. Alt E, Matula M, Theres H, et al: The basis for activity controlled rate variable cardiac pacemakers: An analysis of mechanical forces on the human body induced by exercise and environment. PACE 12:1667, 1989. 7. Humen DP, Kostuk WJ, Klein GJ: Activity-sensing, rateresponsive pacing: Improvement in myocardial performance with exercise. PACE 8:52, 1985. 8. Lau CP, Mehta D, Toff W, et al: Limitations of rate response of activity-sensing rate-responsive pacing to different forms of activity. PACE 11:141, 1988. 9. Stangl K, Wirtzfeld A, Lochschmidt O, et al: Physical movementsensitive pacing: Comparison of two “activity”-triggered pacing systems. PACE 12:102, 1989.

10. Lipkin DP, Buller N, Frenneaux M, et al: Randomized crossover trial of rate response Activittrax and conventional fixed rate ventricular pacing. Br Heart 58:613, 1987. 11. Lindemans FW, Rankin IR, Murtaugh R, et al: Clinical experience with an activity-sensing pacemaker. PACE 9:978, 1986. 12. Benditt DG, Mianulli M, Fetter J, et al: Single-chamber cardiac pacing with activity-initiated chronotropic response: Evaluation by cardiopulmonary exercise testing. Circulation 75:184, 1987. 13. Perrins EJ, Morley CA, Chan SL, et al: Randomized controlled trial of physiological and ventricular pacing. Br Heart J 50:112, 1983. 14. Kubisch K, Peters W, Chiladakis I, et al: Clinical experience with the rate-responsive pacemaker Sensolog 703. PACE 11:1829, 1988. 15. Lau CP, Tse WS, Camm AJ: Clinical experience with Sensolog 703: A new activity-sensing rate-responsive pacemaker. PACE 11: 1444, 1988. 16. Ahern T, Nydegger C, Mc Cormick DJ, et al: Incidence and timing of activity parameter changes in activity response pacing system. PACE 15:762, 1992. 17. Lau CP, Tai YT, Fong PC, et al: Clinical experience with an accelerometer based activity sensing dual chamber rate adaptive pacemaker. PACE 15:334, 1992. 18. Garrigue S, Gentilini C, Hofgartner F, et al: Performance of a rate responsive accelerometer-based pacemaker with autocalibratoin during standardized exercise and recovery. PACE 25:883, 2002. 19. Smedgard P, Kristensson BE, Kruse I, et al: Rate-responsive pacing by means of activity sensing versus single rate ventricular pacing: A double-blind cross-over study. PACE 10: 902, 1987. 20. Lau CP, Wong C-K, Leung W-H, et al: Superior cardiac hemodynamics of atrioventricular synchrony over rate-responsive pacing at submaximal exercise: Observations in activity-sensing DDDR pacemakers. PACE 13:1832, 1990. 21. Buckingham TA, Wodruf RC, Pennington G, et al: Effect of ventricular function on the exercise hemodynamics of variable rate pacing. J Am Coll Cardiol 11:1269, 1988.

528

Section Two: Clinical Concepts

22. Lau CP, Camm AJ: Role of left ventricular function and Doppler derived variables in predicting hemodynamic benefits of rateresponsive pacing. Am J Cardiol 62:906, 1988. 23. Menozzi C, Brignole M, Moracchini PV, et al: Intrapatient comparison between chronic VVIR and DDD pacing in patients affected by high-degree AV block without heart failure. PACE 13:1816, 1990. 24. Oldroyd KG, Rae AP, Carter R, et al: Double-blind crossover comparison of the effects of dual-chamber pacing (DDD) and ventricular rate-adaptive (VVIR) pacing on neuroendocrine variables, exercise performance, and symptoms in complete heart block. Br Heart J 65:188, 1991. 25. Linde-Edelstam C, Hjemdahl P, Pehrsson SK, et al: Is DDD pacing superior to VVIR? A study on cardiac sympathetic nerve activity and myocardial oxygen consumption at rest and during exercise. PACE 15:425, 1992. 26. Millerhagen J, Bacharach D, Street G, et al: A comparison study of two activity pacemakers: An accelerometer versus piezoelectric crystal device. PACE 14:665, 1991. 27. Charles RG, Heemels JP, Westrum BL, et al: Accelerometerbased pacing: A multi-center study. PACE 16:418, 1993. 28. Schuster P, Faerestrand S, Ohm OJ et al: Proportionality of rate response to metabolic workload provided by a rate adaptive pacemaker with automatic rate profile optimization. Europace 7:54, 2005. 29. Duru F, Block KE, Weilenmann D, et al: Clinical evaluation of a pacemaker algorithm that adjusts the pacing rate during sleep using activity variance. PACE 23:1509, 2000. 30. Raj SR, Boach DE, Koshmen ML, et al: Activity-responsive pacing products long-term heart rate variability. J Cardiovasc Electrophysiol 15:179, 2004. 31. Wilkoff BL, Shimokochi DD, Schaal SF: Pacing rate increase due to application of steady external pressure on an activity-sensing pacemaker. PACE 10:423, 1987. 32. Matula M, Alt E, Fotuhi P, et al: Rate adaptation of activity pacemakers under various types of means of locomotion. Eur J Cardiac Pacing Electrophysiol 2:49, 1992. 33. Krasner JL, Voukydis PC, Nardella PC: A physiologically controlled cardiac pacemaker. J Assoc Adv Med Instrum 1:14, 1966. 34. Rossi P, Rognoni G, Occhetta E, et al. Respiration-dependent ventricular pacing compared with ventricular and atrialventricular synchronous pacing: Aerobic and hemodynamic variables. J Am Coll Cardiol 6:646, 1985. 35. Weber KT, Kinasewitz GT, Janicki JS, Fishmann AP: Oxygen utilization and ventilation during exercise in patients with chronic cardiac failure. Circulation 65:1213, 1982. 36. McElroy P, Weber KT, Nappholz TA: Heart rate, ventilation, mixed venous temperature, pH, and oxygen saturation during incremental upright exercise. Presented at Third Asian Pacific Symposium on Cardiac Pacing and Electrophysiology, Melbourne, Australia, October 1985. 37. Vai F, Bonnet JL, Ritter PH, Pioger G: Relationship between heart rate and minute ventilation, tidal volume and respiratory rate during brief and low level exercise. PACE 11:1860, 1988. 38. Beaver WL, Wassermann K: Tidal volume and respiratory rate change at start and end of exercise. J Appl Physiol 29:872, 1970. 39. Alt E, Heinz M, Hirgestetter C, et al: Control of pacemaker rate by impedance-based respiratory minute ventilation. Chest 92: 247, 1987. 40. Kay GN, Bubien SR, Epsten AE, Plumb VJ: Rate-modulated pacing based on transthoracic impedance measurements of minute ventilation: Correlation with exercise gas exchange. J Am Coll Cardiol 15:1283, 1989. 41. Sullivan MJ, Higginbotham MB, Cobb FR: The anaerobic threshold in chronic heart failure. Circulation 81(Suppl II):II-47, 1990. 42. Tse HF, Siu CW, Lee KLF, et al: The incremental benefit of rateadaptive pacing on exercise performance during cardiac resynchronization therapy. J Am Coll Cardiol 46:2292, 2005.

43. Mortara A, Sleight P, Pinna GD, et al: Abnormal awake respiratory patterns are common in chronic heart failure and may prevent evaluation of automatic tone by measures of heart rate variability. Circulation 96:246, 1997. 44. Ben-Dov I, Sietsema KE, Casaburi R, Wasserman K: Evidence that circulatory oscillations accompany ventilatory oscillations during exercise in patients with heart failure. Am Rev Respir Dis 145:776, 1992. 45. Andreas S, Hagenah G, Möller C, et al: Cheyne-Stokes respiration and prognosis in congestive heart failure. Am J Cardiol 78:1260, 1996. 46. Nppholz TA, Valenta H, Maloney J, Simmons A: Electrode configurations for a respiratory impedance measurement suitable for rate-responsive pacing. PACE 9:960, 1986. 47. Simmons A, Maloney, Abi-Samra F, et al: Exercise-responsive intravascular impedance changes as a rate controller for cardiac pacing. PACE 9:285, 1986. 48. Pioger G, et al: Comparison of different electrode configurations in minute ventilation measurement. Eur J Cardiol Pacing 6(1): 1996. 49. Lau CP, Antoniou A, Ward DE, Camm AJ: Reliability of minute ventilation as a parameter for rate-responsive pacing. PACE 12:321, 1989. 50. Mond HG, Kertes PJ: Rate responsive pacing using a minute ventilation sensor. PACE 11:1866, 1988. 51. Lau CP, Ward DE, Camm AJ: Single-chamber cardiac pacing with two forms of respiration-controlled rate-responsive pacemaker. Chest 95:352, 1989. 52. Kay GN, Bubien RS, Epstein AE, et al: Rate-modulated cardiac pacing based on transthoracic impedance measurements of minute ventilation: Correlation with exercise gas exchange. J Am Coll Cardiol 14:1283, 1999. 53. Lau CP, Butrous GS, Ward DE, et al: Comparison of exercise performance of six rate-adaptive right ventricular cardiac pacemakers. Am J Cardiol 63:833, 1989. 54. Lau CP, Antoniou A, Ward DE, et al: Initial clinical experience with a minute ventilation sensing rate modulated pacemaker: Improvements in exercise capacity and symptomatology. PACE 11:1815, 1998. 55. Candinas R, Eugster W, MacCarter D, et al: Does rate modulation with a minute ventilation pacemaker simulate the intrinsic heart rate response observed during representative patient daily activities? Eur J Cardiac Pacing Electrophysiol 2:89, 1995 56. Kay GN: Quantitation of chronotropic response. Comparison of methods for rate-modulated permanent pacemakers. Am Coll Cardiol 20:1533, 1992. 57. Duru F, Radicke D, Wilkoff BL, et al: Influence of posture, breathing pattern, and type of exercise on minute ventilation estimation by a pacemaker transthoracic impedance sensor. PACE 23:1767, 2000. 58. Cole CR, Jensen DN, Cho Y, et al: Correlation of impedance minute ventilation with measured minute ventilation in a rate responsive pacemaker. PACE 24:989, 2001. 59. Duru F, Cho Y, Wilkoff BL, et al: Rate responsive pacing using transthoracic impedance minute ventilation sensors: A multicenter study on calibration stability. PACE 25:1679, 2002. 60. Simon R, Ni Q, Willems R, et al: Comparison of impedance minute ventilation and direct measured minute ventilation in a rate adaptive pacemaker. PACE 26:2127, 2003. 61. Bonnet JL, Geroux L, Cazeau S; on behalf of the French Talent DR Pacemaker Investigators: Evaluation of a dual sensor rate responsive pacing system based on a new concept. PACE 25:2198, 1998. 62. Le Helloco A, et al: Optimal rate modulation slope provided by an automatic function in a DDDR pacemaker. Eur J Cardiac Pacing Electrophysiol 6:172, 1996. 63. Bonnet JL, Ritter P, Pioger G: Measurement of minute ventilation with different DDDR pacemaker electrode configurations.

Chapter 16: Sensor-Driven Pacing: Device Specifics

64.

65.

66.

67. 68.

69. 70. 71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

Investigators of a Multicenter Study Evaluating the Chorus RM and Opus RM Pacemakers. PACE 21:4, 1998. Defaye P, Pepin JL, Poezevara Y, et al: Automatic recognition of abnormal resporatory events during sleep by a pacemaker transthoracic impedance sensor. J Cardiovasc Electrophysiol 15:1034, 2004. Celiker A, Ceviz N, Alehan D, et al: Comparison of normal sinus rhythm and pacing rate in children with minute ventilation single chamber rate adaptive permanent pacemakers. PACE 21:2100, 1998. Cebrera ME, Portzline G, Aach S, et al: Can current minute ventilation rate adaptive pacemakers provide appropriate chronotropic response in pediatric patients? PACE 25:907, 2002. Bazett HC: An analysis of time relations of electrocardiograms. Heart 7:353, 1920. Rickards AF, Norman J: Relation between QT interval and heart rate: New design of a physiologically adaptive cardiac pacemaker. Br Heart J 45:56, 1981. Rickards AF, Norman J: The use of stimulus-T interval to determine cardiac pacing rate. Am J Cardiol 47:435, 1981. Donaldson RM, Rickards AF: Initial experience with a physiological, rate responsive pacemaker. Br Med J 286:667, 1983. Jordaens L, Backers J, Moerman E, Clement D: Catecholamine levels and pacing behavior of QT-driven pacemakers during exercise. PACE 13:603, 1990. Milne JR, Ward DE, Spurrel RAJ, Camm AJ: The ventricular paced QT interval: The effects of rate and exercise. PACE 5:352, 1982. Browne KF, Prystowsky E, Heger JJ, Zipes DP: Modulation of the Q-T interval by the autonomic nervous system. PACE 6:1050, 1983. Sarma JSM, Venkataraman K, Samant DR, Gadgil UG: Effect of propranolol on the QT intervals of normal individuals during exercise: A new method for studying interventions. Br Heart J 60:434, 1988. Baig MW, Boute W, Begemann M, Perrins EJ: Nonlinear relationship between pacing and evoked QT intervals. PACE 11:753, 1988. Brouwer J, Nagelkerke D, De Jongste MJL, et al: Analysis of the morphology of the unipolar endocardial paced evoked response. PACE 13:302, 1990. Boute W, Derrien Y, Wittkampf FHM: Reliability of evoked endocardial T-wave sensing in 1, 500 pacemaker patients. PACE 9:948, 1986. Boute W, Gebhardt U, Begemann MJS: Introduction of an automatic QT interval driven rate responsive pacemaker. PACE 11:1804, 1988. Baig MW, Wilson J, Boute W, et al: Improved pattern of rate responsiveness with dynamic slope setting for the QT sensing pacemaker. PACE 12:311, 1989. Baig MW, Green A, Wade G, et al: A randomized double-blind, cross-over study of the linear and nonlinear algorithms for the QT sensing rate adaptive pacemaker. PACE 13:1802, 1990. Baig MW, Boute W, Begemann M, Perrins EJ: One-year followup of automatic adaptation of the rate response algorithm of the QT sensing, rate adaptive pacemaker. PACE 14:1598, 1991. Landman MA, Senden PJ, van Rooijen H, van Hemel NM: Initial experience with rate adaptive cardiac pacing using two sensors simultaneously. PACE 13:1615, 1990. Connelly DT and the Topaz Study Group: Initial experience with a new single chamber, dual sensor rate responsive pacemaker. PACE 16:1833, 1993. Sharp C, Busse E, Burgess J, Haennel R: Non-linearity of the oxygen uptake: Heart rate relationship in pacemaker patients with left ventricular dysfunction. In Vardas PE (ed): Proceedings of Europace 97. Athens, Monduzzi Editore, 1997, p 517. Lau CP, Leung SK, Guerola M, Crijns HJGM: Comparison of continuously recorded sensor and sinus rates during daily life activ-

86. 87.

88.

89. 90.

91. 92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

529

ities and standard exercise testing: Efficacy of automatically optimized rate adaptive dual sensor pacing to simulate sinus rhythm. PACE 19:1672, 1996. Begemann MJS, Boute W: Automatic refractory period. PACE 11:1684, 1988. Sugano T, Ishikawa T, Ogawa H, et al: Relationship between atrioventricular delay and QT interval or cardiac function in patients with implanted DDD pacemakers. PACE 20:1544, 1997. Ishikawa T, Sugeno T, Sumitas et al: Optimal atrioventricular delay setting determined by QT sensor of implanted DDDR pacemaker. PACE 35:195, 2002. Boute W, Cals GLM, den Heijer P, Wittkampf FHM: Morphology of endocardial T-waves of fusion beats. PACE 11:1693, 1988. Djordjevic M, Kocovic D, Pavlovic S, et al: Circadian variations of heart rate and stim-T interval: Adaptation of nighttime pacing. PACE 12:1757, 1989. Browne KF, Prystowsky E, Heger JJ, et al: Prolongation of the Q-T interval in man during sleep. Am J Cardiol 52:55, 1983. Grace AA, Newell SA, Cary NRB, et al: Diagnosis of early cardiac transplant rejection by fall in evoked T wave amplitude measured using an externalized QT driven rate responsive pacemaker. PACE 14:1024, 1991. Donaldson RM, Rickards AF: Evaluation of drug-induced changes in myocardial repolarization using the paced evoked response. Br Heart J 48:381, 1982. Donaldson RM, Taggart P, Swanton H, et al: Intracardiac electrode detection of early ischaemia in man. Br Heart J 50:213, 1983. Schaldach M, Hutten H: Intracardiac impedance to determine sympathetic activity in rate responsive pacing. PACE 15:1778, 1992. Pichlmaier AM, Braile D, Ebner E, et al: Autonomic nervous system controlled closed loop cardiac pacing. PACE 15:1787, 1992. Witte J, Reibis R, Pichlmaier AM, et al: ANS-controlled rateadaptive pacing: Clinical evaluation. Eur J Cardiac Pacing Electrophysiol 6:53, 1996. Christ T, Shier M, Brattstrom A, et al: Rate-adaptive pacing using intracardiac impedance shows no evidence of positive feeback duing dobutanine stress rest. Europace 4:311, 2002. Osswald S, Cron T, Gradel P, et al: Closed-loop stimulation using intracardiac impedance as a sensor principle: Correlation of right ventricular dP/dtmax and intracardiac impedance during dobutamine stress test. PACE 23:1502, 2000. Cook L, Hamilton D, Busse E, et al: Impact of adaptive rate pacing controlled by a right ventricular impedance sensor on cardiac output response to exercise. PACE 26:244, 2003. Griesbach L, Gestrich B, Wojciechowski D, et al: Clinical performance of automatic closed-loop stimulation systems. PACE 26:1432, 2003. Bingelli C, Duru F, Corti R, et al: Autonomic nervous systemcontrolled cardiac pacing: A comparison between intracardiac impedance signal and much sympathetic nerve activity. PACE 23:1632, 2000. Filho MM, Nishioke SAD, Lopes M, et al: Neurohumoral behaviour in recipients of cardiac pacemakers controlled by closed-loop autonomic nervous system-driven sensor. PACE 23:1778, 2003. Cron TZ, Hilti P, Schächiger H, et al: Rate response of a closedloop stimulation pacing system to changing preload and afterload conditions. PACE 26:1504, 2003. Occhetta E, Bortnik M, Audoglio R, et al: Closed loop stimulation in prevention of vasovagal syncope. Inotropy Controlled Pacing in Vasovagal Syncope (INVASY): A multicenter randomised, single blind, controlled study. Europace 6:538, 2004. Bailey WM, Hull D: Closed loop stimulation improves ejection fraction and NYHA class in patients with congestive heart failure and/or ejection fraction ≤40%. Heart Rhythm 2:S285, 2005.

530

Section Two: Clinical Concepts

107. Occhetta E, Perucca A, Rognoni G, et al: Experience with a new myocardial acceleration sensor during dobutamine infusion and exercise test. Eur J Cardiac Pacing Electrophysiol 5:204, 1995. 108. Bongiorni MG, Soldati E, Arena G, et al: Is local myocardial contractility related to endocardial acceleration signals detected by a transvenous pacing lead? PACE 19:1682, 1996. 109. Wood JC, Festen MP, Lim MJ, et al: Regional effects of myocardial ischemia on epicardially recorded canine first heart sounds. J Appl Physiol 76:291, 1994. 110. Plicchi G, Mercelli E, Parlapiano M, et al: PEA I and PEA II based implantable haemodynamic monitor: Preclinical studies in sheep. Europace 4:49, 2002. 111. Clementy J: Dual chamber rate responsive pacing system driven by contractility: Final assessment after 1-year follow-up. The European PEA Clinical Investigation Group. PACE 21:2192, 1998. 112. Langenfeld H, Krein A, Kirstein M, et al: Peak endocardial acceleration-based clinical testing of the “BEST” DDDR pacemaker. European PEA Clinical Investigation Group. PACE 21:2187, 1998. 113. Rickards AF, Bombardini T, Corbucci G, et al: An implantable intracardiac accelerometer for monitoring myocardial contractility. The Multicenter PEA Study Group. PACE 19:2066, 1996 114. Leung SK, Lau CP, Lam C, et al: Performance of a sensor measuring intracardiac cardiac acceleration signals during submaximal exercise. Pacing Electrophysiol 22:A106, 1999. 115. Greo EM, Ferrario M, Romano S: Clinical evaluation of peak endocardial acceleration as a sensor for rate response pacing. PACE 26:812, 2003. 116. Bordachar P, Garrigue S, Reuter S, et al: Hemodynamic assessment of right, left and biventricular pacing by peak endocardial

117.

118.

119.

120.

121.

122.

123.

124.

acceleration and echocardiography in patients with end-stage heart failure. PACE 23:1726, 2000. Leung SK, Lau CP, Lau CT: Automatic optimization of resting a peak endocardial acceleration sensor: Validation with Doppler echocardiography and direct cardiac output measurements. PACE 23:1762, 2000. Deharo JC, Brunetto A, et al: DDDR pacing driven by contractility versus DDI pacing in vasovagal syncope: A multicenter, randomized study. PACE 26:447, 2003. Clementy J, Kobeissi A, Gamigue S, et al: Validation by serial standardization testiny of a new rate-responsive pacemaker sensor based on variations in myocardial contractility. Europace 3:124, 2001. Leung SK, Lau CP, Tang MO, et al: New integrated sensor pacemaker: Comparison of the rate responses between an integrated minute ventilation and activity sensor and single sensor modes during exercise and daily activities and non-physiological interference. PACE 19:1664, 1996. Leung SK, Lau CP, Tang MO: Cardiac output is a sensitive indicator of difference in exercise performance between single and dual sensor pacemakers. PACE 21:35, 1998. Leung SK, Lau CP, Tang MO, et al: An integrated dual sensor system automaticity optimized by target rate histogram. PACE 21:1559, 1998. Boland J, Scherer M, Hartnung W: Clinical evaluation of an automatic sensor response algorithm in patients with DR pacemakers: A multicenter study. PACE 22:A102, 1999. Pieragnoli P, Colella A, Moro E, et al: Blended dual sensor does not give additional benefits to single sensor in DDDR PM patients: Results from the DUSISLOG study. Heart Rhythm 2:S40, 2005.

Chapter 17

Testing and Programming of Implantable Defibrillator Functions at Implantation MARK W. KROLL • PATRICK J. TCHOU

S ince the inception of clinical implantation of automatic defibrillators in the early 1980s, testing of the device function at implantation has been an integral part of the surgical procedure. Implantable defibrillators serve to terminate life-threatening cardiac arrhythmias, so it is imperative that such a device sense a tachyarrhythmia appropriately and terminate it successfully. Verification of these functions at implantation involves measurement of the detected ventricular and⎯more recently⎯atrial electrograms, both during the normal rhythm and during ventricular fibrillation (VF) and ventricular tachycardia (VT). Testing of sensing and detection is especially important during VF, in which the amplitudes of the recorded ventricular complexes can vary greatly from beat to beat. Such variation can cause failure of arrhythmia detection as a result of automatic adjustment of sensitivity in the recording amplifiers or detection thresholds. Dropout of electrograms during detection may result in prolonged delays in tachycardia detection and termination of VF. A measurement of defibrillation threshold (DFT) is important to ensuring that the implanted device has the shock strength to terminate VF in a reliable manner. Although biphasic shocks and newer electrode systems

are more reliable in defibrillation, one occasionally still encounters a patient in whom the DFT is high. An implanting physician must be familiar with an approach to such a patient. Antitachycardia pacing (ATP) plays a critical role in the treatment of VT by terminating VT without a shock. Thus, ATP therapy may be tested at implantation.

Strength, Duration, and Probabilistic Nature of the Shock Response The most important function of the implantable cardioverter-defibrillator (ICD) is its ability to reliably terminate a life-threatening ventricular tachyarrhythmia. Thus, testing of this function is an important aspect of implanting an ICD. The process of terminating VF can best be described in a three-dimensional relationship of shock strength (voltage or current), shock duration (milliseconds; msec) and the probability of success of the shock. Conceptually, a defibrillation shock has certain similarities to a pacing impulse. The strengthduration relationship of pacing has been well defined. 531

532

Section Two: Clinical Concepts Figure 17-1. A generalized strength duration curve for defibrillation. The rheobase current in the human heart is generally in the range of 2-6 A, whereas the chronaxie is typically 2-4 msec. Optical mapping measurements of the transmembrane response to a shock show that the membrane time constant is in the same approximate range as the chronaxie.

20

Required Current (A)

Chronaxie 15

10

5 Rheobase 0 1

2

3

4

5

6

7

8 9 10 11 12 13 14 15 16 17 18 19 20 Pulse Duration (ms)

Similarly, there is a strength-duration relationship for defibrillation shocks.1-7 This relationship, shown in Figure 17-1, applies to monophasic shocks and the first phase of biphasic shocks. If one considers a rectangular monophasic shock waveform, there exists a rheobase voltage/current below which pulses will not defibrillate regardless of pulse duration. As one shortens the pulse width, the voltage/current required to defibrillate the heart increases. This relationship of pulse width to voltage and current has been well established in experimental models. The slope of this curve rises as the pulse width shortens, such that the voltage/current rises asymptotically when the pulse width approaches zero. For pacing, such a strength-duration curve gives a good description of myocardial capture, because the threshold of myocardial capture is essentially an all-ornone phenomenon—that is, the probability of capture rises very steeply as the parameters of a pacing impulse (strength and duration) traverse the strength-duration curve. One essentially finds 100% capture on one side of the curve and no capture on the other side of the curve except for the well-known hysteresis effect. Such is not the case for defibrillation. The probabilistic nature of defibrillation cannot be adequately described with a single strength-duration curve. The strength-duration relationship of defibrillation is best described as a family of curves in which each curve has a particular probability of successful defibrillation. Thus, another manner of looking at the defibrillation phenomenon is described by the defibrillation success curve. This sigmoidal dose-response curve has been commonly described as a logistic regression curve.8 In fact, this would suggest that at very low levels one could occasionally “get lucky” and defibrillate the heart. However, a more appropriate curve is one that rises steeply from a 0% probability and then asymptotically approaches 100%, as shown in Figure 17-2.9,10 The use of extremely high energies can also reduce the efficacy of a defibrillation shock because of the deleterious effects of high energies, as discussed later. This prob-

lem is relatively rare at the voltages used with current ICDs, especially with anodal biphasic shocks. The postulated mechanisms by which a shock terminates VF have undergone modifications over the years. These have been reviewed elsewhere, and a thorough discussion is beyond the scope of this chapter.11,12 Although the defibrillation phenomenon and its probabilistic nature are still incompletely understood, several factors contribute to the difference between a pacing threshold and a defibrillation threshold. A pacing impulse is delivered at a time when the myocardium is in a relatively homogenous state⎯ diastole. Furthermore, pacing capture occurs within a small volume of myocardium in the immediate vicinity of the pacing electrode. Thus, the electrical state of the myocardium within this small volume at the time of impulse delivery is highly homogenous. In contrast, defibrillation involves the influence of an electrical shock across the entire myocardium. Reproducibly

100% 90% Probability of Success

0

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

5

10 15 Shock Energy (J)

20

25

Figure 17-2. A typical defibrillation success curve showing the probability of successful defibrillation according to the amount of energy used. This type of success curve can just as easily be described for the average current or voltage of the shock. More effective waveforms would shift this curve to the left and generate a steeper slope.

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

successful defibrillation shocks must achieve minimum voltage gradients of about 5 V/cm at nearly all myocardial sites.13 The defibrillation shock is delivered at a time when the myocardium is in a highly nonhomogenous state of activation. There may be several reasons why defibrillation behaves in a probabilistic fashion. The number and size of wavelets or rotors existing at any time in the myocardium during VF can vary. The distance of any dominant wavefronts from the shocking electrode can vary. Furthermore, this nonhomogenous state is changing rapidly over time. Thus, the timing of the shock is not synchronized to dominant wavelets or rotors. Experiments in which the defibrillating shock was synchronized to a large-amplitude wave on a surface electrocardiogram (ECG) during VF, presumably a time when there is a large dominant wave, have demonstrated lowered DFTs.14 Thus, spatial and temporal heterogeneity of ventricular myocardium is a likely explanation for the probabilistic nature of defibrillation.

Why Perform Defibrillation Threshold Testing? With proper implantation of the ICD leads, today’s biphasic waveforms in implantable defibrillators can achieve reliable defibrillation in a large percentage of patients. However, some form of testing to verify that an implanted device is capable of accomplishing the task of defibrillation reliably is needed at the time of implantation. This testing allows one to ensure that the device can deliver a shock waveform of sufficient amplitude to accomplish reliable defibrillation and provides an opportunity to make changes in the device to improve the defibrillation efficacy, if needed. Because defibrillation is a probabilistic phenomenon, a single success at a shock strength near the maximum output of the device is not adequate assurance that the device will work well in clinical circumstances when defibrillation is needed. Furthermore, clinical circumstances may be complicated by additional variables that may alter defibrillation energy requirements. The variability of serum electrolyte levels, changes in sympathetic tone, the rise and fall of serum antiarrhythmic drug levels, and the diastolic filling pressures of the heart are factors that may influence defibrillation energy requirements. When testing demonstrates that defibrillation is reproducibly successful at relatively low shock strengths, one can program the device to deliver the first shock at a lower shock strength. This lower-energy output would require a shorter capacitor charge time and shorten the time from the initiation of ventricular tachyarrhythmia to the delivery of the first shock. Such a shortened device response time can mean the difference between maintenance and loss of consciousness. Sudden and unexpected loss of consciousness is usually an undesirable occurrence because it can increase patient morbidity and even cause death. Although awareness of the psychological stresses that

533

may complicate sensed shocks, especially multiple sequential shocks, is growing, one must balance the potential generation of anxiety and depression against the real possibility of injuries to the patient and others brought on by loss of consciousness.15 The lower shock strengths that can be programmed after defibrillation testing may also have the advantage of decreasing the probability of myocardial damage, especially if the patient experiences circumstances that produce multiple shocks over a short period, because of either a “storm” of ventricular arrhythmias or inappropriate device discharges in response to atrial fibrillation with rapid ventricular response or device malfunction. The use of a lower shock strength can also accelerate post-shock hemodynamic recovery. One study showed that shocks stronger than 10 joules (J) tend to depress hemodynamic recovery by about 20% for as long as 4 minutes.16 Lower shock strengths can also lengthen longevity of the device battery. Another advantage of determining the actual DFT is that the parameter can be monitored for increases during follow-up, providing an indication of changes in the lead position or myocardial substrate. However, a thorough understanding of the DFT would require multiple fibrillation/shock sequences during implantation testing. The value of an accurate DFT measurement is further diminished by the knowledge that this threshold can vary from day to day in response to clinical changes in a patient’s physiologic status and medication levels. Patients who receive defibrillator implants frequently have markedly depressed ventricular function as a reflection of the underlying substrate that puts them at risk for ventricular tachyarrhythmias. In such patients, multiple shocks for determining the DFT may cause deterioration of ventricular function during the implantation and thus may raise the risk of the procedure.17 During clinical implantation of a defibrillator, therefore, one needs to strike a balance among the safety of the patient, the adequate performance of the implanted device, and a sufficient understanding of the DFT to enable one to program the first shock strength appropriately. The most common definition of defibrillation threshold is the shock amplitude that provides a 50% chance of success. This is the definition we use as the default. When there is danger of confusion, we use DFT50 to refer to this parameter. Other useful DFTs are DFT70, DFT80, and DFT90 (see Fig. 17-2). Alternatively, one can simply verify that the device, at its maximum output, has an adequate safety margin to defibrillate the heart reliably. The role of verification is to merely “verify” that the DFT is below a certain level and that the device can thus be safely implanted.10,18 The advantage of a verification approach is that one minimizes the number of shocks during the implantation and shortens the duration of the procedure. If the device to be implanted is capable of a maximum energy shock of 30 J and if a single shock at 15 J or two successive shocks at 20 J are successful in converting VF, one can have some confidence that 30 J will be sufficient. The advantages of this approach are simplicity and brevity. In patients in whom one wishes to

534

Section Two: Clinical Concepts

minimize the number of VF inductions and shocks, this approach would allow the implantation of a device with a minimal amount of testing. The drawbacks of using a verification approach are that the first shock energy may be set unnecessarily high and that there is no baseline DFT for tracking changes. There will also be no DFT data for scientific comparison. Thus, this approach is inadequate for comparative evaluation of different waveforms or shock electrode positions. The main method for determining the DFT during ICD implantation or for postimplantation testing is to induce VF in the patient, wait approximately 10 seconds, and deliver a shock of a given energy. If this shock is successful, VF is induced again, and a lowerenergy shock is tried. If the shock was unsuccessful, a rescue shock at a higher energy or an external shock is delivered. Next, a shock of energy intermediate between those of the last successful shock and the failed shock can be tried. Various protocols, discussed later, are used to determine the sequence of shocks and the number of shocks as well as to calculate the DFT. In experimental studies, a longer sequence of shocks may be necessary to provide an estimate of the DFT50 that is accurate enough for comparison purposes. However, during standard clinical implantation of an ICD, a step-down or step-up approach, starting at 15 J output with increments of 5 J, usually provides an adequate endpoint. Animal studies have shown that prolonged DFT testing with multiple inductions of VF can lead to hemodynamic compromise and diminished cardiac function.19 Human results are mixed, however, and there is no clear consensus. One study showed no decrease in the mean ejection fraction (EF) but reported one patient whose EF fell from 0.20 to 0.11.20 Two studies found the only hemodynamic damage from VF testing to be a significant impairment in diastolic filling.21,22 Another study found that, in the first hour after VF, there is no hemodynamic deterioration and, in fact, diastolic filling is enhanced.23 Other studies have suggested that there is no morbidity from VF inductions for patients with left ventricular EF (LVEF) values in excess of 0.30.17 Patients with lower EFs can have a serious reduction in cardiac index, however, with one report of prolonged inotropic support being necessary. Another human study, which looked specifically at patients with an LVEF of 0.35 or less, found no deterioration even with an average of nine inductions each.24 In the Medtronic PCD trials, physicians were encouraged to find a combination of leads that would allow the implantation of the monophasic sequential shock device. Many of these patients received more than 50 shocks with no reported trends in mortality and morbidity. Thus, there is considerable variation in the tolerance of repeated VF induction and defibrillations even in patients with reduced EFs. However, most implanting physicians have encountered the occasional patient— usually with very poor LV function⎯who showed marked hemodynamic deterioration mostly of limited duration following several VF inductions and shock sequences. It is important to use clinical judgment at the time of implantation to assess the safety of repeated VF inductions. Patients with EFs of less than 0.3 may

need closer hemodynamic monitoring during and immediately after the implantation procedure, but the vast majority of patients undergo DFT testing with no problem at all. When defibrillation is followed by a prolonged episode of hypotension or marked bradycardia, it would be prudent to limit DFT testing to a verification protocol. Initial electroencephalogram (EEG) data suggested that exceeding six inductions can cause transient cerebral dysfunction (defined as an increase in delta-wave power longer than 2.5 minutes).25 Later studies have shown that the duration of EEG changes are correlated with VF duration.26,27 These changes occur within 8 to 12 seconds of induction and last about 1 minute on average.26 A study of 36 patients with 286 inductions of VF found that although EEG recovery time was correlated with VF duration, it was negatively correlated with the number of inductions.28 Thus, there was no indication that any cumulative effect on the EEG occurred with repeated induction of VF. Defibrillation testing involves delivery of high-voltage shocks through the ICD leads. Thus, patients should be well sedated during the process to avoid undue stress.

Methods of Determining Defibrillation Threshold Induction Methods Before one can directly test a defibrillation shock, one must induce fibrillation. The four common methods for inducing VF are: (1) stimulating at extremely high rates, (2) delivering a shock on the T-wave, (3) applying direct current (DC) across the defibrillation electrodes, and (4) applying alternating current (AC). The classic approaches to inducing VF were to perform high-rate pacing and to deliver alternating current through an external testing device. One method used today is extremely high-rate stimulation on the order of power-line frequencies. Pulses at a rate of 50 Hz (ICDs manufactured by Medtronic, Inc., Minneapolis, Minn.) or of 20 or 33 Hz (ICDs manufactured by Guidant [Boston Scientific, Natick, Mass.]), or programmable rates from 10 to 50 Hz (ICDs manufactured by St. Jude Medical, St. Paul, Minn.) are delivered until VF is induced. Direct delivery of the pulses through the ICD pacing leads has the advantage of being painless. However, the efficacy of inducing VF may be limited. A popular method now used is to deliver a shock on the peak of the T-wave. This can be performed during sinus rhythm or ventricular pacing. It has the advantage of more reliably inducing VF. This approach is based on the principle of the “zone of vulnerability” surrounding the T-wave, where a range of shock strengths will initiate VF.29 The upper limit of this zone of shock strength correlates well with reliable defibrillation. T-wave shocking is available with all current ICDs. In clinical practice, VF is generally initiated through delivery of a low-level shock, around 1 J, timed

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

to the peak of the T-wave during a paced ventricular rhythm. A commonly used approach is to pace the ventricle at 400 msec for eight beats, followed by a 1-J shock at around 300 msec after the last paced impulse. The coupling interval of the shock can be adjusted if VF is not reliably induced. At times, several adjustments of the timing and, occasionally, the energy may be needed to achieve induction of VF. In general, shocks delivered before the peak of the T-wave are most reliable for inducing VF. This approach is available with all current ICD models. A subtlety of this approach is that fibrillation is induced most reliably with a cathodal right ventricular (RV) shock, yet defibrillation is most reliable with anodal RV polarity.30 The third method is to apply a low DC voltage, around 5 to 12 V, across the defibrillation electrodes. This was classically done with an external source of DC, such as a 9-volt battery, connected to the implanted epicardial leads. Current implantation techniques generally do not allow the use of an external source of DC. Thus, one uses the internal power of the ICD generator to deliver this current. One advantage of this approach is the high success rate for induction without the need for alignment to a T-wave. The method also appears to out-perform the T-wave shock in reliably inducing VF.31 This DC fibrillation feature is currently available in St. Jude Medical devices. The final method for induction is the use of lowvoltage AC coupled directly to the defibrillation leads. Either 50- or 60-Hz AC (depending on the continent) has proved to be extremely effective in this role.32 Transformer reduction to a 3-V level provides sufficient current to induce VF within 2 to 4 seconds in almost any subject. This method is now seldom used in ICD implantation, however, because it is not available for delivery through the devices. It may still be used occasionally when an ICD is implanted in a patient undergoing open-chest surgery for some other reason. Defining the Defibrillation Threshold It is important to understand the DFT both for historical and clinical reasons. The DFT must not exceed the capability of the device. In addition, even when the DFT is not evaluated explicitly, it must be evaluated implicitly through assessment of an upper bound so as to verify that the device has a shock of sufficient magnitude to defibrillate the heart reliably. This is what Singer and Lang18 refer to as a “verification approach.” Finally, an explicit DFT must often be calculated for research purposes to demonstrate that a certain waveform or lead system has advantages over others. The definition of defibrillation threshold is deceptively simple: It is the electrical dosage required to defibrillate the heart. This definition, however, is complicated by two major subtleties. The first one is the choice of units for the dosage. This issue is dealt with later in the chapter; see “The Energy Crisis.” Historically, energy has been chosen to compare efficiency in comparisons of the DFT values obtained in two different defibrillation systems. Although the use of energy is an acceptable means of comparing DFTs, one must

535

be aware that just raising the amount of “delivered” energy does not necessarily translate to a better safety margin for defibrillation. For example, increasing the delivered energy by delaying the truncation to prolong the duration of a truncated exponential waveform may actually lower the defibrillation efficacy. Because of the ease with which the shock energy is measured and the historical precedent of reporting defibrillation success as shock energy, however, it has become accepted practice to use energy to describe DFTs and to discuss safety margins in units of energy. Fortunately, in current devices, increases in energy are tied to increases in voltage and current and not just to prolongation of pulse duration. Thus, the discussion of DFT in this chapter is focused on joules or energy for defibrillation. However, it is crucial to understand that DFT is most closely correlated to the current delivered through the leads. The second subtlety in the definition of the DFT is the apparently probabilistic nature of the dose-response curve for defibrillation.33 The probabilistic character of defibrillation efficacy inherent in this curve may be due to random variation in the size of the myocardial mass,34 to conductive properties of myocardial cells,35 or to systematic alteration of cellular or tissue electrophysiologic characteristics involved in the initiation and perpetuation of VF.36 The “Progressive Depolarization” hypothesis offered by Dillon and Kwaku11 holds that progressively stronger shocks depolarize progressively more refractory myocardium, to progressively prevent postshock wavefronts, and prolong and synchronize postshock repolarization, in a progressively larger volume of the ventricle, to progressively decrease the probability of fibrillation after the shock. These multiple functions, requiring different current densities acting on different phases of cellular activation potentials, which are distributed differently with wavefront positions, generate multiple degrees of freedom that help explain the apparent probabilistic nature of defibrillation. One might say “apparently” probabilistic because no one has been able to establish that the dose-response relationship is truly probabilistic on a level comparable with that of, say, electron position in quantum mechanics. In fact, evidence suggests that a large portion of the “probabilistic” nature of the shock response is actually due to our inability to better time the shock.37-39 Defibrillation Threshold Testing Protocols The simplest protocol for determining the DFT and the one most commonly used in clinical practice is to reduce the shock amplitude until a shock fails to defibrillate the heart; this is known as a step-down protocol. The lowest successful shock level is then called the DFT. This approach gives, on average, a DFT estimate that approximates the 70% success (DFT70) level.40-42 However, given the probabilistic nature of a shock’s success in converting VF, such a determination of DFT may have (occasionally) as low an overall success rate as 25%. Thus, one would need a device that has a maximum output of around twice the energy of the DFT to be confident that it would defibrillate reliably.

536

Section Two: Clinical Concepts

For example, a DFT of 15 J would require that the device have a maximum output of 30 J. When the DFT measured in this manner is 20 J, one does not have the confidence that this device would convert VF at all times. One means of reducing the need for such a large safety margin is to require two consecutive successful shocks at the lowest successful energy level. This DFT has been referred to as a DFT+, indicating the extra success of shocks at this energy level.17 With a DFT+ determination, there is a greater likelihood that this DFT is near the upper end of the DFT curve. Thus, one can reduce the required safety margin. For a patient with a DFT+ of 20 J, one might implant a 35-J device. Similarly, with three consecutive successful shocks at 25 J (DFT++), one might consider implanting a device with a 35-J maximum output, or a device with a maximum output 10 J higher than the DFT++. To arrive at a more accurate estimate of DFT, in experimental protocols, for example, one could average a step-down value and a step-up value.43 The most popular method for determining a DFT has been the classic step-up/step-down method first popularized by Purdue University researchers. With this protocol, the attempted shock strength begins at a high level. It is then reduced at fixed steps until failure occurs. Typical fixed steps are 2 J or 100 V. After failure to defibrillate occurs, the direction of changes is reversed, and the shock strengths are raised by finer steps, such as 1 J or 50 V. This protocol is simple to perform, and the DFT is defined as the lowest amplitude at which a shock is successful. The disadvantage of this protocol is that it can require many shocks and is not very accurate; the mean estimate tends to overestimate the DFT50 with a fairly wide error band. A variant on the step-up/step-down approach is the three-reversal approach.44 With this approach, the step sizes are kept fixed. After stepping the shock strength down to defibrillation failure, the direction of change is reversed, and the strength is raised in same fixed steps until defibrillation is successful. At this point, the shock strength is changed again, and stepped down until defibrillation failure occurs. At this point, the third reversal is assumed “on paper”—that is, one more stepup would result in success. All of the shock energies from the lowest successful shock energy in the first step-down to the final “paper” shock are then averaged. This approach has the advantage of giving a more accurate estimate of the DFT and has proved very useful in research studies comparing the relative benefits of different waveforms.45 The disadvantage of this reversal approach is that it requires numerous shocks. An example of the three-reversal approach is as follows. Shock energies are reduced by 1 J at a time, with the lowest successful shock being 8 J. A reduction of shock strength to 7 J finds failure. The first reversal is then performed, and the shock strength is increased up to 8 J (another failure) and then to 9 J before defibrillation occurs. The second reversal is then performed, the shock being reduced to 8 J, which fails to terminate VT. The final reversal then is the “paper shock,” which assumes that a 9-J shock would be successful. The DFT estimate then is the average of the shocks (in order) of

8, 7, 8, 9, 8, and 9 J, which is 8.2 J. The inclusion of the higher “paper shock” value balances the downward bias from inclusion of the failure shocks and results in an accurate, unbiased estimation of DFT50. During clinical ICD implantation, it is typical to use a step-down method with very coarse steps such as 5 J. This allows the determination of an approximate DFT while exposing the patient to a minimum number of shocks. In an attempt to achieve a high-resolution DFT estimate while staying with a lower number of shocks, the binary search technique was suggested. With this approach, the range of possible values for the DFT is continually cut in half to rapidly focus in on the DFT. This approach is best explained with an example: For a 32-J device, the first attempt at a shock is 16 J. If that shock is successful, the energy range is bifurcated, at 8 J. Assuming that the 8 J shock is a failure, the next energy attempted is halfway between 8 and 16 J, or 12 J. If the 12 J shock is a failure, the energy range is cut in half again, and 14 J is tried. The procedure is continued until the desired resolution is obtained. The advantage of a binary search is that it gives high resolution with very few shocks. The disadvantage is that the high resolution may be misleading in that it is not necessarily associated with high accuracy. Malkin has derived a unique approach, which would appear to combine the best features of all of the DFT protocol techniques in his Bayesian technique.46 This approach has also proved useful in scientific studies.47 Depending on the number of shocks one is willing to use, Malkin and colleagues supply tables and formulas to give the exact sequence of shocks. Then, depending on the success and failure of each of the shocks, there is an optimal Bayesian estimate of the actual DFT. One version combines this with the upper limit of vulnerability (ULV; see later) approach.48 Another approach possible with Guidant ICDs is the step-up. A series of five, increasing energy shocks are programmed in for a therapy regimen. VF is induced, and then the shocks of increasing energy are given until defibrillation occurs. This approach allows for a reasonably accurate DFT determination without multiple inductions. The drawbacks are the length of the individual fibrillation episode and the possible influence of VF duration on the DFT. Upper Limit of Vulnerability Approach to Determining the Defibrillation Threshold There is a time interval near the peak of the T-wave (within 40 msec) known as the vulnerable period, during which a shock of sufficient but not too great magnitude induces VF (Fig. 17-3). According to the ULV theory of defibrillation, when such a shock generates an electrical field with its voltage gradient not parallel to a repolarizing wavefront, it initiates a spiraling wavefront around a critical point that can initiate VF. As one raises the shock strength, there is a limit above which VF can no longer be induced. This upper limit of vulnerability to VF has been demonstrated to correlate well with the DFT; that is, shocks with energies above the ULV should defibrillate the heart as well.49 This concept can also

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

Shock strength

ULV

Vulnerable zone

VF Threshold

537

vantage of the ULV approach is the number of shocks to be delivered for a reliable determination of defibrillation efficacy. The advantage is that one can minimize the number of inductions of VF. The ULV and DFT are both predictors of a shock strength that defibrillates the heart with a given probability of success. The prediction of defibrillation efficacy is actually more important than the precise correlation between ULV and DFT. Verification

Figure 17-3. The vulnerable zone is a region of shock amplitude and time in which shocks will usually produce VF. The time period is within the T-wave and the shock strength lies between the VF threshold and the upper limit of vulnerability.

explain the probabilistic nature of defibrillation. In the myocardial regions of relative low voltage gradients, the exact orientation of a repolarizing wavefront with the local voltage gradient generated by a shock is a random phenomenon. Thus, the vulnerable zone of shock strengths for initiating (or reinitiating) VF in that region should follow a probabilistic distribution. It should be noted, however, that other theories also account for the clinical ULV phenomenon that do not rely on the ULV theory of defibrillation.11 The clinical attraction of the ULV approach is that one could perform implantation testing of DFT with the use of only one actual VF induction. Shocks are delivered synchronized with the T-wave (in sinus or paced rhythm) at a high energy and decreased in a stepwise fashion with successive shocks. If the 20-J shock fails to induce VF, that level must be higher than the ULV and hence higher than the DFT. Note that the patient has not been put into fibrillation with that shock. The shock strength then is reduced successively until either a sufficiently weak shock does not induce VF (which means that there was no induction of VF) or the first episode of VF occurs, which in turn has significance similar to that of a failed defibrillation shock. The timing of the shock within the T-wave is critical. The peak ULV within the vulnerable zone is not necessarily at the peak of the T-wave and can vary from patient to patient. In isolated rabbit hearts, accuracy with monophasic shocks appears to be optimized with a shock delivery during the T-wave upslope, whereas shocks at or after the peak may be better for biphasic waveforms.50 Accuracy and repeatability are improved with multiple shock positions within the T-wave. For example, the shock can be delivered at the peak of the T-wave and 20 msec before.51 One can add a third shock positioned 40 msec before the T-wave peak.52,53 To use this method as a reliable approach to predict successful defibrillation, one should apply this approach at varying intervals before the peak of the T-wave, for example, 0, 20, and 40 msec. Although there are occasional significant differences between the ULV and DFT,54 the two values correlate well regardless of the pacing site,55 the presence of ischemia,56 whether testing is acute or chronic,57 electrode polarity,58 and waveform durations.58 The disad-

With the verification approach, one ensures that the maximum output of the ICD is well above the DFT without actually determining a DFT. Singer and Lang18 have provided a classification of various approaches to verification. With their one-shock verification protocol, one low-energy shock was given which was successful in defibrillating the heart. This protocol assumes that this shock was so low in strength that even though it was tested only once, there was a large margin between that low-energy shock and the much higher output of the device. For example, if the low-energy shock of 10 J terminated VF, then one can have reasonable confidence that a device output of 27 J will defibrillate reliably. Similarly, one could use a two-shock protocol, in which two shocks of identical energy are used to defibrillate the heart. With this “2S” protocol, one can reduce the margin between the successful shock energy and the ICD output. A more detailed discussion of this topic can be found in the literature.17, 59 Equipment for Defibrillation Threshold Testing Classically, external implant support devices were used to induce VF and provide defibrillation shocks. With the advent of the pectoral implantation approach, however, external device–based testing is rarely used today.60,61 The advantages of device-based testing rather than external support device testing are numerous, as follows: (1) one can be sure that there is a perfect match between the waveform for the testing and the waveform that will be used for therapy; (2) one can be certain that the measurements of shock amplitude and impedance are identical; (3) the lead connections are all being tested with each delivered shock; (4) the same sensing circuit and algorithms are tested; and (5) storage and maintenance of the external device are avoided. Major disadvantages of the external devices are that they are not updated as frequently as the actual ICDs and that the sensing amplifiers, filters, and sensing algorithms are rarely identical to those found in the ICD being currently implanted. Thus, in many cases, sensing performance is not adequately tested with the external system. Also, the waveforms have been, in some cases, significantly different with external device–based testing. The advantages of using the external device are as follows: (1) the ICD longevity is enhanced because it is not used to deliver any shocks during testing; (2) one has the capability of higher-energy rescue shocks, should “bailout” be necessary; and (3) the use of the

538

Section Two: Clinical Concepts

external support device for defibrillation testing gives the physician the flexibility of choosing a lower-energy ICD (presumably much smaller) for the final implant. This last procedure can be performed without breaking the sterility of the smaller device, which may be a risk if it is unable to perform the defibrillation. Regardless of the device used for testing DFT, backup defibrillation is essential. The most common approach is to use a biphasic external defibrillator with flexible adhesive patches attached to the patient’s chest. Two defibrillators with a biphasic waveform should be available during DFT testing. It is important to verify proper placement of the patches, especially in larger patients in whom the required defibrillation energy may be near the maximal output of the defibrillator. The adhesive patch electrodes are relatively radiolucent and generally do not obstruct the fluoroscopic view during the implantation procedure. A typical anterior-posterior placement of these electrodes keeps them out of the operating field. Optimal anterior placement is over the apex of the heart, and optimal posterior placement is over the spine at the upper border of the heart shadow in the AP view with fluoroscopy. It is important to verify the proper placement of the internal defibrillation leads as well as the absence of a pneumothorax prior to initiation of VF for DFT testing. Occasionally, with a pneumothorax, even the external device does not successfully defibrillate the heart.62 Pneumothorax defibrillation failure is very rare now with external biphasic defibrillators, but when external defibrillation is unsuccessful, one may consider—in addition to poor external electrode positions and pneumothorax—whether intramyocardial current is being shunted in some manner. Such shunting can occur, for example, if a posterior electrode such as a subcutaneous coil, a coronary sinus electrode, or an azygous shocking coil is connected to the superior vena cava (SVC) electrode port of the ICD header. Although this shunting is usually not of concern, it may make the difference between success and failure of defibrillation in a patient with high transthoracic impedance and high external DFT. Disconnecting the ICD electrode from the ICD header under those circumstances would solve the problem. In desperate circumstances, one can consider using the external biphasic defibrillator by connecting the apical patch electrode to the RV ICD electrode and use the posterior patch to deliver larger energy shocks in the 50- to 100-J range. Finally, should all external defibrillating attempts fail, extracorporeal cardiopulmonary bypass using peripheral vessel access can be tried if such equipment is readily available. How Much Shock Energy Is Enough? There are two main issues regarding the adequacy of energy output, as follows: 1. Does the device have an adequate safety margin with respect to its output above the DFT to ensure reliable therapy for the patient? 2. At what energy output should the first (and maybe second) shock amplitude be set?

Although there is some overlap in the consideration of these two questions, they really are quite different. For the first question, let us presume that the DFT or an upper bound for the DFT has been determined through a verification protocol. How much additional energy must the device have in order to give an adequate safety margin for implant? Early systems occasionally had DFT shifts greater than 10 J, and patient deaths resulted or electrode revisions were required.63,64 The classic rule of thumb was that a 10-J safety margin was sufficient, and this rule has achieved the status of accepted medical practice. Others have championed a safety margin equal to the DFT.88 In other words, the device’s maximal shock capability must be equal to twice the DFT for successful implantation. Originally this was a significantly more conservative approach than the 10-J safety margin. However, with thresholds now typically in the single digits, this rule is actually more liberal than the 10-J safety margin.65 In one study, with a DFT of less than 6 J, a shock at twice the DFT gave a 95% success rate, whereas a classic DFT plus 10 J rule yielded a 99.5% rate for successful defibrillation.65 When the DFT is very low, such as 4 J or less, using twice the DFT energy may yield only a 67% successful first shock rate.66 Thus, when the DFT is less than 5 J, it would be more reliable to set the first shock energy with a safety margin of at least 7 J to achieve a first shock defibrillation success rate of 96%. Finally, one should not take excessive comfort from a low DFT, because such numbers may have less stability (or confidence) associated with them. In another study, the shock of twice the energy yielded a success rate of 98% for patients with a DFT higher than 4 J but of only 67% for those with a DFT of less than 4 J.67 The Malkin Bayesian approach may be used to calculate the shock strength required to achieve a conversion rate of 95%.68,69 The safety margin is needed to cover two basic problems. First, because of the probabilistic nature of the defibrillation dose-response curve, a shock must have higher energy than the DFT to give a reasonable confidence of defibrillation success (i.e., the shock strength should be the value that would suggest a near 100% probability of success). Second, it is difficult to estimate the required level of “insurance” against long-term drift in the DFT. Long-term rises in DFT have been reported by some investigators with the early monophasic devices,70-73 but others found no increase in DFTs with these devices.74 It is not clear that this drift is a problem with modern biphasic transvenous devices.75-77 One study has reported a long-term rise in biphasic thresholds,78 but the typical result is a rise over the first month or two, followed by a gradual return to the implant values. Although an early study in an animal model using epicardial patches suggested that rapid pacing–induced cardiomyopathy would elevate DFT,79 a later animal study of transvenous endocardial defibrillation has shown that progressive heart failure does not increase the DFT.80 The mean values of DFT in these studies using transvenous lead systems do not seem to show marked increases. However, there clearly are individual cases in which significant rises in DFTs

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

do occur. In fact, one study found a DFT rise of 10 J or greater at 1- or 2-year follow-up testing in 15% of patients.81 The reasons for long-term changes in DFT are not well understood. One can imagine that fibrosis might tend to increase resistance, and it often does. However, these increases in resistance have not been correlated with the rises in DFT that one might expect. Lead microdislodgement can significantly affect the DFT, although confirmed cases of a long-term DFT being affected by a change in lead position (short of a macro-dislodgment) are relatively rare. There are also circadian changes in the DFT, with a somewhat higher value in the morning than in the afternoon corresponding to the peak incidence of failed first shocks in response to clinical tachycardias.82 The need for antiarrhythmic drug therapy could arise in patients who have ICDs either for the treatment of atrial arrhythmias or for suppression of frequent ventricular arrhythmias. These can also affect the DFT, as discussed later. A pneumothorax can lead to an increase in the DFT.62,83,84 Thus, multiple factors can lead to changes in DFT after implantation that would make an adequate safety margin of the shock energy important. The Low Energy Safety Study (LESS) is often miscited in support of the use of small safety margins such as 5 J.85,86 This study excluded 10% of the enrolled patients—largely for high DFTs. Thus, by design, the study clearly did not address the patient with a high DFT. In this study, the DFT was defined as the DFT++. Because this definition required three consecutive successful shocks at the DFT, it was essentially a DFT88 (0.88 = 1 − 2−3, which is the probability of failing to throw three out of three “heads” in a coin toss.). For a typical patient, as shown in Figure 17-2, the DFT++ is about 6 J above the DFT50; thus, adding 5 J for the first two shock energies in LESS actually resulted in a setting about 11 J above the DFT50. In spite of that, 4% of the patients required one to three maximal-energy rescue shocks for their spontaneous VF episodes. Table 17-1, which shows the predicted performances of various verification techniques for defibrillator implantation testing, gives a sobering view and very well explains why a small number of patients who pass standard defibrillation testing may require multiple shocks at maximum output to terminate spontaneous VF. Using a large database of successful and unsuccessful defibrillation shocks, Smits and DeGroot87 evalTABLE 17-1. Protocol

539

uated the popular ICD implantation defibrillation testing techniques for performance. Table 17-1 reflects the results of their computer simulation. They assumed a device with a 35-J maximum output. Patients with high DFTs would have a less than 90% chance of defibrillation with a single shock of 35 J. Those having a greater than 90% chance would be regarded as having low DFTs. The “Criterion”(column 2 of Table 17-1) is the lowest successful energy output for the testing protocol used. This is best explained by example. The row with protocol S1 and a criterion of 15 J means that a single shock verification at 15 J will be successful (Pass) in 91% of patients. The sensitivity value 94% means that this test will detect 94% of the patients with low DFTs. However, the specificity of 52% means that defibrillation with a 35-J shock failed in only 52% of patients with high DFTs; therefore, the test will miss 48% of patients with high DFTs. Approximately 3.4% of the patients who “pass” this test nevertheless belong in the high-DFT group. Setting the First Shock Energy Traditionally, it was believed that the shock energies for VF should all be set to the maximum output of the device. This was reflected by the philosophy that one “just could not take a chance.” However, now with typical thresholds below 10 J, this practice has come into question. There are several problems with using excessive shock energy. The first is that the charge time is (at least) increasing proportionally with the shock energy. Thus, the charge time for a 10-J shock will be less than one third of that for a 30-J shock. This delayed therapy can potentially increase DFTs, exacerbate postshock dysfunction, and significantly raise the chance of syncope and accompanying sequelae. In addition, the idea that a maximum-energy shock has a better chance of success than a low-energy shock is not universally accurate. The dose-response curve for defibrillation is not monotonic—especially for monophasic shocks. With shock energy above a certain level, the success rate may begin to go down from near 100%. The output of clinical biphasic devices does not typically generate this phenomenon, but it may still be seen clinically from time to time.88 A report comparing the setting of the first shock to twice the DFT with the device maximum of 34 J demonstrated no significant

Predicted Performance of Various Verification Techniques Criterion (J)

Pass (%)

Sensitivity (%)

Specificity (%)

S2

24

93

96

53

S1

15 12

91 87

94 90

52 61

Step-down

≤24 ≤18

96 87

98 91

32 74

Binary

≤24 ≤12

99 87

100 90

11 61

Data from Smits KF, DeGroot P: A Bayesian approach to reduced implant testing of a ventricular defibrillator: A computer simulation. Europace 6(Suppl):97, 2004.

540

Section Two: Clinical Concepts

difference in incidences of first biphasic shock conversion of spontaneous ventricular tachyarrhythmia. The lower setting actually trended toward a higher first shock conversion rate (98.5% vs. 92%).88 The third problem with the use of maximum-energy shocks is that higher-energy shocks can temporarily cause depressed ventricular function. Although seen only occasionally, this stunning may last for multiple seconds, delaying the hemodynamic recovery from VF.16 Thus, if the goal is to have the patient regain good cardiac output and consciousness as soon as possible, the use of a maximum-energy shock could actually be counterproductive.

Factors that Affect the Defibrillation Threshold Lead Systems The current generation of defibrillation leads typically has either a single shocking electrode located in the right RV or two shocking electrodes, the second of which is located on the proximal portion of the lead, positioning it in the SVC. The advantage of the twocoil lead is its ease of use. It has the pace/sense electrodes and both shocking electrodes on one lead, allowing implantation with a single insertion (the single-pass lead). However, separate leads for the two shocking electrodes may allow better positioning of the proximal electrode. Occasionally, better results are reported for placement of the proximal electrode in the innominate vein89,90 or the subclavian vein.91 In general, placement in the left brachiocephalic vein does not appear to offer better DFTs.92 In various patients, DFTs are lowest with this proximal lead in different locations.93 When the proximal electrode is positioned too close to the RV, in the right atrium, for example, there may be energy shunting that could reduce defibrillation efficacy. Thus, locating this lead in the SVC or higher would optimize the position. In larger patients, leads that provide various positionings of the SVC electrode may be of help in ensuring that the proximal electrode is properly positioned. Pectoral implantation of the ICD generator has become standard. Inclusion of the housing electrode has reduced DFTs by about 30% in comparison with previous, purely transvenous lead systems.94,95 In fact, these so-called unipolar systems now offer thresholds that are comparable with those seen with monophasic shocks using epicardial patches. In most patients, using the RV shocking electrode in conjunction with the can electrode provides an adequate DFT for implantation of a device without the need for another electrode. This configuration can allow use of a thinner transvenous lead. Furthermore, such a unipolar lead may be more reliable because no high-voltage differences would exist within the lead body itself. Adding a patch, subaxillary array, or a subcutaneous coil can result in lower DFTs than with the use of a simple transvenous system.96,97 A subcutaneous patch

may actually be as effective as an epicardial patch.98 Because of crinkling problems, such patches are almost never used today. The addition of an array may be helpful in reducing DFT through two primary mechanisms. The first and most direct mechanism is that the overall system impedance decreases, and therefore, for a given maximum voltage from the ICD, the current increases. For a tilt-based waveform, the resistance lowering of the array will reduce the shock time constant, diminish the pulse widths, and possibly improve the efficiency of the waveform. Second, with a more posterior placement of the coils, the current from the RV coil may be more posteriorly directed, improving voltage gradients in the posterior LV portion of the myocardium. Thus, a subcutaneous coil must be inserted as posterior as possible at a level near the base of the ventricular shadow. This position may, however, increase the chance of back pain. The most important electrode in transvenous defibrillation systems is the one placed in the RV, although it obviously cannot function in isolation. Although this coil is typically placed in the RV apex, there is some evidence that even better DFTs can be obtained with placement of the coil in the septum, near the outflow tract.99,100Assuming an RV coil and a can electrode in the left pectoral region, a natural question arises, “To what extent will the addition of an SVC lead reduce DFTs?” One study suggested that it probably makes no difference for most patients.101 Practically, that may be the case. Later studies have demonstrated that inclusion of an SVC electrode can lower energy and voltage DFTs.102,103 However, these studies were performed with fixed-tilt waveforms, so the automatic reductions in the pulse durations may have exerted a dominating influence by bringing the waveform closer to optimal durations. One report demonstrated that the addition of an SVC electrode lowers DFT energy but increases the current at DFT.103 Because an increased current is associated with higher tissue voltage gradients, the lower DFT energy was actually associated with a higher DFT tissue gradient, implying a less efficient distribution of voltage gradients. DFTs using the RV coil–pectoral can defibrillation pathway typically are 10 J or less, so it is unclear whether using a “single-pass, two-electrode” lead has any significant advantage over using a lead with only one RV shocking electrode. However, in patients with higher DFTs, use of an SVC coil may improve DFTs enough to provide an adequate safety margin. In certain clinical circumstances, the left pectoral region is not available for implantation of an ICD generator. Infection, a preexisting pacemaker, or other anatomic problems (e.g., ipsilateral breast cancer) may prevent implantation in that region. Although a left-sided approach gives the best DFTs for the endocardial-can electrode systems,104 data from animal105 and clinical106-108 studies suggest that placement of the generator on the right side or even over the abdomen105,109 allows achievement of reasonable DFTs. With the abdominal implantation site, the use of an SVC lead appears to offer significant advantage and should probably be used routinely even with a “hot” can.110 With the continued

541

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation 400 350 300 Shock Voltage

down-sizing of the generator that has occurred since the early 1990s, one may question whether reduction of the can electrode size may influence DFT. However, the can size, at least within the anticipated range of potential reductions, does not appear to significantly influence the DFT.111,112 Interestingly, the radius of the RV electrode113 and the surface area of the RV electrode,114 within reasonable bounds, have little effect on the DFT. One study found a small benefit in using an 11F rather than a 7F SVC electrode.115 The polarization of the RV electrode during shock delivery may contribute to greater impedance and may also affect DFT.

250 Membrane response

200 150 100

Wasted energy

50 0 0

Shock Waveforms Because of the clearly demonstrated benefits of biphasic waveforms for converting VF, all ICDs today have biphasic waveforms.116 However, several factors affect the efficiency of biphasic waveforms. Nevertheless, one must note that there is no evidence of the superiority of biphasic waveforms for converting VT.

1

2

3

4 5 Time (ms)

6

7

8

Figure 17-4. A 400-volt capacitive discharge shock waveform is shown as the purple line. A typical passive cardiac membrane response is identified by the arrow and the red line. Note that the energy delivered does not help charge the cardiac membrane and is actually counterproductive. This applies to a monophasic shock as well as to the first phase of a biphasic shock.

A Primer on Defibrillation 800 700

22-J low capacitance waveform

600 500 400 Voltage

Defibrillation involves more questions than answers. However, the knowledge of the basic principles of defibrillation has advanced to the point that waveforms may be optimized with the use of simple first-order theories. The recitation and use of these theories to optimize defibrillation are by no means meant to give the impression that these simple models represent the state of knowledge of defibrillation. The monophasic shock acts as a super pacing pulse that synchronizes the fibrillating heart. (This is also true of the first phase of the biphasic shock.) During fibrillation, most cells are in phase 2 or 3; thus, the shock typically interacts with the membrane passive response, as seen in Figure 17-4. The membrane is charged up until the new activation potential begins. Numerous optical and electrode mapping studies have demonstrated this membrane response. Indirect human data have shown that the typical time constant for human myocardial cells is about 3.5 msec. In the situation represented by Figure 17-4, the capacitor is charged to 400 V and then delivers its charge to the heart. The cell membrane is charged up as the capacitor is discharged. Note that the membrane response peaks at 4 msec, yet the shock is prolonged to “deliver more energy” or “achieve a certain tilt.” However, the membrane response clearly shows that the energy delivered after 4 msec is wasted. Actually, it is worse than wasted—it is actually counterproductive because it is reducing the final membrane response. This simple analysis shows why shock truncation can have such a dramatic effect on DFTs. Figure 17-5 shows a low-capacitance waveform and a high-capacitance waveform. The 22-J low-capacitance waveform has the same delivered charge (area under the curve) in the critical first 4 msec as the 27-J high-capacitance waveform. The high capacitance (160-μF) system has a maximum voltage of around 600 V. Because of this low voltage, it requires a great

27-J high capacitance waveform

300 200 100 0

⫺100

2

4

6

8

10

12

14

⫺200 ⫺300 Time (ms) Figure 17-5. The waveforms from the highest and lowest capacitance ICDs available clinically. Note that the 22-J low capacitance waveform (yellow) has the same delivered charge (area under the curve) in the critical first 4 msec as does the 27-J high capacitance waveform (blue). About 3 of the 5 J delivered by the high capacitance waveform between 4 and 7.3 msec is wasted. Hence the low capacitance waveform has at least a 3 J DFT advantage over the high capacitance waveform.

deal more time to deliver its energy. Also, the durations are not programmable but are fixed with a 60% phase 1 tilt and a 50% phase 2 tilt. For a patient with a resistance of 50 Ω, durations would be 7.3 and 4.9 msec, respectively. These durations are about twice optimal. In fact, the first phase is almost as long as the entire duration of the low-capacitance waveform. During the “golden” 4 msec, however, the high-capacitance shock delivers about 22 J. Thus, about 3 of the 5 J of the

542

Section Two: Clinical Concepts

shock is wasted. The excessively long second phase further reduces efficacy. This is the simple reason why the high-capacitance, lower-voltage shock is less efficient and has higher-energy DFTs. Effect of Capacitance One now has a choice of defibrillation capacitance values over a range of 85 to 160 μF. Theoretical models of defibrillation all show that the optimal capacitance is inversely related to the inter-electrode resistance.117-119 To be precise, the product of the resistance and the capacitance should be about 80% of the chronaxie value.117 Thus, for a patient with a chronaxie of 5 msec120 and a resistance of 50 Ω, the optimal capacitance is 80 μF. Numerous animal and clinical studies have shown the benefits of reductions in the capacitance value from the conventional values of 140 to 150 μF.45,121-124 The benefit of the reduced capacitance values is, of course, more dramatic in patients with higher resistance.125 The converse is also true. With extremely-lowresistance pathways, there is little to no benefit. For example, one study with an average resistance of 32 Ω found no difference between a 125-μF and 450-μF capacitor⎯although the tilts and durations were not held constant.126 The benefits of the smaller capacitance values probably accrue from the shortening of the first phase so that its duration is closer to the defibrillation chronaxie. In addition, the second phase is closer to the passive membrane time constant required for optimum membrane discharge, as described by the “burping theory.”127 A simple example explains why this is so critical for the patient with a high resistance. Imagine a device with a capacitance of 140 μF and a patient with an impedance of 70 Ω who underwent implantation of an ICD with a 65% tilt for both phases. The first phase duration is 10 msec and a second phase duration is also about 10 msec. These durations are significantly longer than the chronaxie (3-5 msec) and passive membrane time constant (2-4 msec), respectively. If, in the same situation, the output capacitance were 70 μF, these durations would be halved and much closer to optimal. For the average patient, the best capacitance value is probably about 90 μF.128 One concern about using lower capacitance values is that⎯even though they lower energy DFTs⎯they store less energy for a given fixed voltage level. With the limits of capacitor voltages under current technology, the maximum stored energy would be reduced in proportion to the capacitance. Thus, the resulting safety margin may not, in fact, be improved. However, if future technology would permit higher voltages in capacitors, smaller capacitors, in the range of 50 to 60 μF, may be optimal. In addition, for patients with DFTs below 10 J, a maximum energy of 18 to 20 J can provide appropriate safety margins. With current and improving lead systems, a majority of patients may belong in this category. Thus, for these patients, a smaller capacitance would allow the use of a smaller device. Last, the use of a smaller capacitance

tends to improve the DFT in patients with high shock electrode impedance at implantation. Because the impedance stays relatively constant with implanted lead systems, smaller capacitance values may be most helpful in these patients.117,128,129 For a given stored energy, the capacitance is inversely related to the voltage. Hence the data supporting the small capacitance waveforms also directly support the use of a higher-voltage shock. Current ICDs use shock voltages ranging from 600 to 830 V. If the shock voltage is not given on the labeling or the programmer, the implanting physician should insist on obtaining the voltage value (and by implication, the capacitance value) from the manufacturer’s representative before attempting an implantation in a patient with a potentially high DFT. Effect of Pulse Durations Let us assume that the ICD already has the optimum capacitance for defibrillation. It is now well understood that a major function of the first phase is to act as a monophasic shock designed to synchronize the vast majority of the myocytes by extending their refractory periods.127,130 Thus, the first phase should be set at between 3 and 5 msec, the typical range for the human chronaxie.120,128 However, if one has a large capacitance value (i.e., >120 μF), a compromise is in order. A duration roughly equal to the average of the chronaxie and the shock time constant is required, so as to achieve a balance between the chronaxie and the need to deliver a charge from the capacitor. Consider a patient with a resistance of 50 Ω and a 140-μF capacitor. This time constant is simply the product of these two variables, or 7 msec. If one chooses a 4-msec chronaxie, the optimal first phase duration should then be the average of 4 msec and 7 msec, or about 5.5 msec. The second phase is far simpler. Regardless of the shock time constant or the impedance, the second phase duration should be set at slightly less than the passive membrane time constant in order to actively “burp” the cell membrane.47,127,131 The burping theory of the biphasic shock holds that the function of the first phase is to extinguish as many wavefronts as possible by “capturing” a majority of the cells in the broad sense. The function of the second phase is to counter the three side effects of the first phase—marginal stimulation, electroporation, and cathodal launching of new wavefronts. The term burping came from the analogy of the removal of excess gas from a baby. This duration should be in the order of 2.5 to 3 msec for optimal performance. The burping model is illustrated in Figure 17-6. The first phase charges up the membrane of each cell. If the cell is captured, there is an extended activation potential, and everything is wonderful. There is nothing for phase two to do. If the cell is only marginally charged up, the second phase will remove the charge, and the cell will go back to normal. If the cell is electroporated, the second phase, by quickly removing the excess charge sitting on the membrane, immediately will heal the cell. Finally, the

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation 1 0.9 0.8 0.7 0.6

Relative Voltage

0.5 0.4

Membrane response

0.3 0.2 0.1 0

⫺0.1

1

2

3

4

5

6

7

8

9

⫺0.2 ⫺0.3 ⫺0.4 ⫺0.5 Time (ms) Figure 17-6. The biphasic shock voltage is shown in blue. The typical membrane response is shown in yellow. The burping theory of the biphasic shock holds that the function of the first phase is to maximally charge the membrane, and the function of the second phase is to discharge the membranes of the uncaptured cells back to zero. In this example, the second phase is too long as the membrane is pulled past zero.

second phase will tend to discharge virtual electrodes and reduce the launching of new wavefronts.132 In the example shown in Figure 17-6, we can see that the second phase, at 3.5 msec, is too long and the membrane is actually discharged and taken slightly negative; this result is suboptimal. But a 2.5-msec duration would have been optimal. This is how one can calculate the optimal durations for the second phase. All predictions of the burping theory have been verified in animal or clinical studies.133-136 Nevertheless, an unwritten myth persists, in some circles, that the second phase somehow captures cells that were missed by the first phase. That spoken myth may have impeded proper programming of the second phase. Simply reducing the second (and first) phase duration from “standard” settings tends to lower DFTs in clinical studies.134,137-139 The effect of optimizing the first and second phase durations is sometimes quite dramatic, as can be seen in studies using devices in which these durations are programmable.140 Some devices have their phases determined only in terms of the classic tilt instead of fixed durations in milliseconds. This approach arose from the invention of the truncated capacitive waveform by Schuder and colleagues.141 They used a single time constant for truncation (which one would now refer to as a 63% tilt). This gave a duration for the monophasic shock close to optimum for the low-resistance epicardial patches then in use. Unfortunately, this one lucky datum led to the

543

false belief that tilt-based waveforms were somehow optimal. It is commonly believed that the dependence of the tilt-based durations on resistance is helpful because it was hoped that the duration changes are a possibly correct adjustment. However, the first phase significantly over-adjusts to resistance changes, and the second phase is actually adjusted in the wrong direction.127 Imagine the case of a patient’s impedance rising to 100 Ω with a 140-μF capacitor and 65% tilt. This would result in a first phase duration of 14 msec, at least triple the chronaxie. For the second phase, use of the tiltbased duration is highly nonoptimal according to current understanding of the operation of the biphasic waveform. The tilt approach would tend to make the second phase increase in proportion to impedance, whereas the burping theory suggests that the second phase duration should actually decrease slightly with increases in resistance.127,131 More complex waveforms currently under study may lower thresholds even further than those achieved with the use of a lower capacitance. These so-called parallel-series waveforms achieve a more ascending first phase while leaving enough charge for a fully functioning second phase to operate. Such a waveform operates by running multiple capacitors in parallel for a few milliseconds and then, after they have discharged sufficiently, switching them to series, which brings the upper voltage back up to near the initial voltage.142,143 The use of fixed pulse durations dramatically reduces the DFT with the use of tilt. Denman reported on a series of 56 patients in whom DFTs were determined both for tilt and msec pulse durations.140 The mean DFT was reduced by 20% and the DFT for those starting over 15 J was reduced by 30%. Impressively, the population peak DFT was reduced by 40%. Keane144 studied a series of 17 patients with DFTs of 30 J or higher who would have had an insufficient safety margin without an array—with tilt-based waveforms. By using msec pulse durations instead of tilt, Keane was able to reduce the DFT to 25 J or less in all 17 cases and thereby remove the need for an array. Thus, in patients with high DFTs, especially those with higher shock impedance, it is advantageous to use ICDs with independently adjustable first and second phase pulse durations (or tilts). Doing so allows optimization of the first and second phases separately rather than locking the two phases to the same pulse width. This concept is illustrated by the following case study. An 18-year-old man with a nonischemic cardiomyopathy underwent implantation of a single-chamber ICD (Contour, St. Jude Medical) with a 150-μF capacitance and 750-V maximum voltage shock. The impedance was 38 Ω. A classic 65% tilt-equivalent waveform was attempted. This gave phase durations of 6 msec for both phases. A maximum energy (38 J delivered) shock succeeded once but then failed. Application of the burping theory suggested 5 and 3 msec as optimal phase durations. The implanter was reluctant to shorten the waveform so drastically, because it would “deliver less energy.” So only the second phase was shortened to 3 msec for the first iteration. There were

544

Section Two: Clinical Concepts

two successful defibrillations at 27 J, followed by two more successes at 22 J. No further waveform adjustments were made. Shock Polarity The defibrillating shock polarity can influence the DFT, especially with suboptimal biphasic and monophasic shocks. Unfortunately, the polarity is typically set incorrectly in most implants. Because most ventricular pacing is done with the tip as a cathode, this polarity was naturally assumed to be appropriate for defibrillation. However, it is clear that with monophasic waveforms, the use of “anodal” defibrillation leads to significantly lower thresholds than those found with cathodal defibrillation.145,146 The same changes have been found, although not as dramatic or as consistent, for biphasic waveforms. A listing of reports on the influence of the first phase polarity of biphasic shocks on DFT is shown in Table 17-2. These reports involved 110 patients and showed an average reduction in DFTs of 16% when the RV electrode started the shock as an anode. This was the best polarity for 46% of patients, whereas 42% had equal DFTs with either polarity. Animal studies with more “optimal” biphasic waveforms, those using shorter phase 1 and phase 2 durations closer to optimal “burping,” have shown no superiority of either polarity.147 It appears that polarity makes little difference in “optimal” biphasic waveforms as the anodal polarity tends to prevent the main first phase side effect, which is treated by the second phase—namely, cathodal waveform launching. Observations by Efimov and associates148 of cells surrounding the RV shock electrode provide some support for this hypothesis. The passive membrane response to a shock is the opposite of the electrode polarity (e.g., it is negative for cells nearer the anode and positive for cells nearer the cathode). The cells act as if they are either near the anode or near the cathode, with essentially no graded response; this is referred to as the virtual electrode effect. During fibrillation, most cells are in their plateau phase. The transmembrane potential of such cells in the virtual anode tends to be

TABLE 17-2.

reduced from near zero to, say, −80 mV. Thus, they are, at least temporarily, masquerading as repolarized cells and very amenable to capture. On the other hand, cathodal shocks generate positive transmembrane potentials, which tend to “pace” the cells in the virtual anode. Hence, the virtual cathode launches new wavefronts into the virtual anode.132 If the RV coil is an anode, these wavefronts merely go toward the coil and are usually extinguished. If the RV coil is cathodal, however, the wavefronts are launched into the rest of the ventricle and are proarrhythmic. Biphasic shocks, through the “burping” function, appear to counteract the persistence of the virtual electrode effect. If this were to be the primary mechanism by which biphasic shocks are superior to monophasic ones, optimized biphasic shocks would not demonstrate a polarity preference in defibrillation because they would eliminate or minimize the virtual electrode effect. In fact, Mowry and colleagues132 have demonstrated a second phase burping of the virtual electrode potential back to near zero. A common clinical practice, especially in a patient with a high DFT, of programming five or six cathodal shocks and then finally making the polarity anodal has no scientific support. A common practice at the Cleveland Clinic is to program the ICD to anodal shock configuration before it is tested. Given that one cannot truly determine whether a particular biphasic waveform is optimized for a given patient, it would make sense to test DFT in the anodal configuration. It is important to appreciate the different polarities of the various ICD models. St. Jude devices are shipped anodal and thus require no polarity “reversal.” Guidant devices are always cathodal and thus should be reversed before usage. Older Medtronic models require reversal, but the Intrinsic (and presumably later models) are already anodal polarity. Timing of the Shock Timing of the shock during VF may be important. At present, a shock is synchronized to ventricular activation at the sensing electrode, usually located at the RV

Influence of Right Ventricular Electrode Polarity on Defibrillation Threshold N

RV+ (J)

RV- (J)

Reduction in Mean (%)

Lower RV+

Equal DFT

Lower RV+

Schauerte et al252

27

11.1

13.3

17

10

14

3

Shorofsky & Gold253

26

11.1

12.2

9

Natale et al254

20

16.3

21.5

24

12

6

2

15

9.9

9.5

−4

3

9

3

10 12

6.6 12

10.8 16.3

39 26

7 7

3 3

0 2

16

39 46

35 42

10 12

Study

Strickberger et al Keelan et al

255

256

Total or merged % of patients DFT, defibrillation threshold.

110

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

apex or sometimes in the RV outflow or mid-septum. However, in the future, this timing may change. VF may have periods of greater organization or susceptibility to shock termination. For example, timing the shock to large amplitude points on the ECG may reduce DFT.149,150 The absence of such timing in shock delivery during DFT determination and in current ICDs may be a factor contributing to the probabilistic nature of defibrillation.38,39,150,151 Greater coherence of myocardial activation suggests fewer random wavefronts. Shocks delivered during periods of fibrillation with greater coherence of ventricular activation appear to have a better chance of success, that is, lower DFT, than shocks delivered during other periods.152 The influence of VF duration on DFT is somewhat controversial. Many studies have shown that the DFT rises steadily over time for monophasic shocks.153,154 This can be shown for shocks delivered anywhere from 2 seconds to 9 minutes after induction of VF and has been attributed to rises in adenosine levels.155,156 However, the situation for biphasic shocks is much more controversial. Although early animal studies suggested that the DFT for biphasic shocks also rises with time, later papers have challenged this concept and suggested that the threshold may actually dip to a minimum at about 20 seconds after VF initiation.157 In the interval between onset of VF and delivery of first or even second shock from an ICD, however, it is unlikely that VF duration would affect defibrillation efficacy. Finally, the time of day affects the DFT; DFTs are higher in the morning,81 which is unfortunately when the incidence of tachyarrhythmias also peaks.158 Effects of Drugs on the Defibrillation Threshold Most antiarrhythmic and anesthetic drugs can affect the DFT. However, reports on these influences are mixed and results may depend on whether a monophasic or a biphasic waveform was used in the study. The discrepancies are, at least in part, related to the lack of standardized testing from one report to another. Details of the studies are listed in Table 17-3. Pharmacologic therapy is often used in patients with ICDs to minimize the frequency of therapy delivery by the device. Because some antiarrhythmic drugs also affect DFT, the effects of antiarrhythmic therapy on DFT must be considered when safety margins are being established during ICD implantation and/or testing and when the pharmacologic therapy is modified. Drugs that Raise the Defibrillation Threshold Class Ic drugs, such as encainide, can increase DFT.159 Fain and colleagues160 reported on intravenous encainide and the two metabolites, O-dimethyl encainide (ODE) and 3-methoxy-ODE (MODE). Intravenous encainide and ODE increased the average DFT in dogs by 129% and 76%, respectively, from control values. The DFT returned to normal after drug washout. No significant drop in DFT was reported with MODE. For another class Ic drug, flecainide, widely divergent effects on DFT were

545

Antiarrhythmic Drugs: Effects on Defibrillation Threshold (DFT)* TABLE 17-3.

Drugs that increase DFT

Ajmaline257 Amiodarone167,168,170,172,173,174,177 Atropine214 Bidisomide166 Diltiazem257 Encainide160 ODE160 Recainam165 Verapamil179 Carvedilol180

Drugs that cause no change in DFT

Atenolol216 Disopyramide258,191 MODE160 Phenylephrine216 Phentolamine216 Procainamide182,184,193 Propafenone194 Azimilide212

Drugs for which reports on effect on DFT are conflicting

Amiodarone161,167,171,259,260 Bretylium179,181,197,198 Flecainide159,161-164,205,191 Isoproterenol214-216 Lidocaine184,187,193,195,257,261 Moricizine190-192 Mexiletine187-189 Quinidine180,185,186

Drugs that decrease DFT

Clofilium262 E-403205,206 Ibutilide203,204 LY-190147263 MS-551207,264 Sotalol209-211 Tedisamil208 Dofetilide199

*Superscript numbers indicate chapter references.

reported in different animal species. In dogs, this agent markedly raised DFTs so that defibrillation was nearly impossible with the existing equipment.161 Another study, however, reported flecainide to have no effect on DFT in pigs.162,163 Flecainide has also been found to reduce the atrial DFT in humans.164 The investigational class Ic drug recainam has been shown to elevate DFT.165 Bidisomide (SC-40230), an investigational class Ia/Ib agent, raises the DFT in dogs.166 Amiodarone, a multiple-class drug, tends to elevate DFTs. Short-term administration of amiodarone has been reported to both lower the DFT167 and to raise it.168 Long-term oral administration of amiodarone tends to raise DFT.169-171 Guarnieri and associates found larger increases in the DFT in patients receiving amiodarone who were undergoing generator change.172 DFTs rose from 10.9 ± 4.3 J to 20 ± 4.7 J. The mean DFT decreased for patients taking no antiarrhythmics or only class Ia agents. Pelosi and colleagues found a 60% DFT increase in patients after an average of 73 days of amiodarone therapy, a finding that showed a does-response relationship (r2 = 0.36).173 Epstein and associates174 found that 52% of patients with high DFTs (>25 J) at implantation were taking amiodarone. However, such elevated

546

Section Two: Clinical Concepts

DFTs may reflect preselection of patients with poor prognosis rather than an intrinsic effect of amiodarone on the DFT. For example, one clinical study found no difference with long-term amiodarone use.171 These differences may be caused by a metabolite and the total body load of amiodarone.175 Amiodarone and its metabolite, desethylamiodarone, are stored in cardiac tissue, delaying the drug’s action on the myocardium; such storage can explain the increase in DFTs in patients taking the drug long term.176 Zhou and coworkers177 found levels of desethylamiodarone to have a larger impact on the DFT than amiodarone itself. The selective potassium blocker barium lowers the DFT.178 Ajmaline and calcium channel blockers tend to increase the DFT.179 The combined β1, β2, and α receptor blocker carvedilol is widely used for the treatment of heart failure and does not appear to affect the mean DFT in most patients, although there are case reports of significant DFT increases in individual patients.180

refractory period extension associated with a defibrillating shock.202 Ibutilide significantly lowers DFTs and occasionally causes spontaneous defibrillation.203,204 The research class III drug E-4031 cut DFTs in half in a dog study,205 even in the presence of isoproterenol.206 The experimental class III agent LY-190147 also lowers the DFT,201 as does MS-551 (nifekelant hydrochloride). Murakawa and associates207 found that although MS-551 lowered the DFT50, it had minimal effect on the DFT90. Tedisimal increases the electrogram coherence and reduces the DFT.208 D-Sotalol and DL-sotalol decrease the DFT in humans.209-211 Azimilide appears to reduce the DFT by increasing spatial organization.212 What is the effect of β-adrenergic modulation? A dog study showed that isoproterenol reduced the DFT, which returned to baseline value after β-blockade.213 However, others have reported a rise in DFT with isoproterenol in dogs214,215 but no change of DFT in pigs.216 Aminophylline has been reported to lower DFTs.217

Drugs with Minimal Effects or with Conflicting Reports of Effect

Anesthetic Agents

Class Ia drugs have minimal effects on DFT. Dorian and associates181 found no changes in DFT with quinidine infusion. In dogs and pigs, intravenous infusion of procainamide (15 mg/kg) had no significant influence on DFT.182,183 Procainamide generally has no effect on DFT in humans at the usual therapeutic doses.184 Guarnieri and associates reported that DFTs were actually lower in patients taking class Ia agents at the time of generator replacement than at the time of device implantation.172 Rises in DFT were reported by Woolfolk and colleagues185 and Babbs and coworkers186 with the use of very high doses of quinidine in animal studies. The class Ib drug mexiletine increased DFTs in a case report.187 Animal studies, however, have shown little or no increase in the DFT with its use.188,189 Moricizine does not affect DFTs in pigs190 but increases DFTs in dogs,191 especially in the presence of lidocaine.192 One clinical report found a decrease in DFT, whereas another found an increase.190,193 Oral propafenone does not affect human DFT.194 Lidocaine raises the DFT in dogs,193,195 especially with monophasic waveforms.196 One study in dogs found that lidocaine did not increase DFT with the use of chloralose for anesthesia but a large increase with the use of pentobarbital.193 Bretylium did not affect DFT in two animal studies.181,197 Interestingly, in another study in dogs, the DFT was reported to be lowered 15 to 90 minutes after intravenous injection of bretylium tosylate (10 mg/kg IV).198 A pig study also showed a reduction in DFTs.179 Drugs that Decrease DFT Pure class III agents tend to decrease DFT.193,199,200 This effect is most likely due to lengthening of the refractory period.201 Clofilium blocks outward potassium current and can directly defibrillate the heart. Tacker and colleagues198 found that clofilium lowers the DFT in dogs,198 possibly owing to its ability to lengthen the

Fentanyl can increase the DFT in humans by 41% compared with nitrogen dioxide.218 However, fentanyl reduced the DFT in animals, unlike pentobarbital and enflurane.219 The common inhalation agents and barbiturates, however, do not appear to have significant effects on measured DFTs in dog studies,220 with the possible exception that pentobarbital may interact with lidocaine to increase DFT as noted previously. Noncardiac Drugs A 2005 study found that 100 mg IV of sildenafil citrate in 20- to 30-kg swine raised the mean DFT from 12 to 19 J.221 Although the applied dosage was rather large (typical adult human dosage is 50-100 mg), patients with low shock safety margins should be counseled regarding the use of sildenafil. The risk of VF induction by sexual activity does suggest that the probability of an interaction leading to a failure to defibrillate is real.

The Energy Crisis Energy is used as the primary defibrillation dosage unit for unfortunate historical and nonscientific reasons. Although it is the dosage unit of common medical practice, energy simply does not defibrillate. Defibrillation requires the generation of an adequate voltage gradient across the myocardium. A myocardial voltage gradient, given a stable myocardial impedance, will generate a current. Thus, current is a good reflection of defibrillation efficacy for a constant set of electrode positions. For relatively stable electrode impedances, voltage correlates with current and can also be a good reflection of defibrillation efficacy. Consideration of shock energy may be useful in considering battery longevity, but it does not necessarily correlate to defibrillation efficacy.

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

A simple gedanken (thought) experiment shows why that is the case. Imagine that one could merely turn on a pacemaker battery output of 6 V for 120 seconds. Assuming an impedance of 100 Ω, total energy delivered would be 21.6 J. This would eliminate the need for the bulky capacitors and inverter that determine the size of the present ICD. This is energy delivered directly to the heart, yet no one would ever imagine that it would defibrillate the heart.222 In fact, delivery of energy at a constant low voltage is an excellent way to induce fibrillation.223 The confusion of energy with defibrillation has led to some accepted practices that are not justified. Scientifically, for example, the idea that energy defibrillates the heart led to the corollary that the capacitance probably did not matter. As long as the capacitor stored enough energy, it did not matter whether it was a smallvalue capacitance or a large-value capacitance. This assumption delayed the arrival of the smaller-capacitance, higher-efficiency waveforms. As unfortunate as the concept of energy for defibrillation is for understanding and optimization, the concept of “delivered” energy is even more damaging. This concept held that the more energy delivered to the heart, the better. A related theme held that some systems are not subject to the vagaries of electrode impedance changes because they are able to “guarantee” a certain energy delivery. Simple reflection demonstrates that the concept of delivered energy has no scientific basis. It was demonstrated more than 20 years ago that truncating the shock (consider the monophasic waveform momentarily) significantly decreases the DFT—in fact by up to 50%.141 More recently, shortening of the exponential shock waveform by truncation did not demonstrate any deterioration in the DFTs of delivered energy until the pulse width was shortened to the range of 5 to 6 msec. Truncation of the shock, however, means less energy is delivered, violating the philosophy that delivered energy is the critical parameter for defibrillation.117 Transthoracic modeling studies have shown that current is far superior to an energy-based dosage parameter.224 This finding is confirmed in actual clinical studies.225 Current, of course, has a direct relationship with tissue voltage gradient. That is, for a fixed electrode configuration, a higher current is associated with a higher tissue voltage gradient. In some situations, an increase in delivered energy may actually have a negative relationship with the average current of the pulse. A more pernicious and damaging effect of the concept that more delivered energy would defibrillate better and provide greater safety margin is the idea of increasing delivered energy by lengthening the waveform. This approach, which has been blamed for patient deaths, can actually generate negative safety margins. In an older investigational device, high-energy shocks of 21, 24, 28, and 33 J were available. Although the voltage rose very slightly between the lowest and the highest of these four high-energy shocks, the pulse widths increased dramatically from 4 to 12 msec. However, the 33-J shock did in fact “deliver” the most energy, but its average current was as low as or lower

547

than any of the other shocks. Thus, patients whose DFT was found to be equal to one of the lower applied shocks (i.e., 21 J) could actually have a zero or negative safety margin when the device was set at the 33 J. A further harmful effect comes from the attempt to apply the delivered energy concept to biphasic waveforms. By confusing the role of the second phase as just another means of delivering additional energy, some devices extend the second phase out proportionally to the resistance (by maintaining the same tilt) in order to guarantee the “delivered energy” of the whole waveform. This results in second-phase durations that are significantly suboptimal, especially for patients with high resistance. In spite of the significant problems in the use of energy as the defibrillation dosage measurement, it is supported by a wealth of published literature and is given by every manufacturer. It has become accepted practice to measure thresholds in terms of joules. It is acceptable to use energy as a dosage measurement so long as the model of device, the lead resistance, the capacitance value, and the pulse duration are not changing.226 (If the model changes, then one might assume that the capacitance values and durations are also changing.) If any of those four items is changing and one needs to make a DFT comparison, the average current should be calculated. Average current is very easy to calculate, as follows: Average Capacitance × Peak voltage × Tilt = current Pulse (msec ) This calculation applies to the first phase. Dosage calculations for the second phase are irrelevant and misleading. As recent literature would support, the small amount of charge delivered in a second phase merely functions as a counterbalance (albeit a very important one) for the deleterious actions of the first phase,47,127,131 including the virtual electrode effects.148 A capacitor is best viewed as a tank holding a “gas of electrons.” One can store a gas in a large tank at low pressure or in a small tank at high pressure. The advantage is with the low capacitance, because it delivers its charge in a time much closer to the optimal timing of the cardiac cells. For the same delivered energy, the shock from a high-capacitance, low-voltage shock is less efficacious. For example, for a 30-J device, a shock from a 600-V, 160-μF capacitor will tend to have DFTs that are higher than those from an 830-V, 85-μF capacitor (see Fig. 17-5). Use of such low-voltage ICDs resulted in significant increases in the number of patients with high DFTs in the Multi-Center Inductionless Defibrillator Implant Study (MIDIS).227

Approach to the Patient with a High Defibrillation Threshold Although certain clinical characteristics may be associated with high DFTs, the predictive value of these

548

Section Two: Clinical Concepts

characteristics are generally not very accurate because of the wide variation of patients. A number of studies have attempted to predict the DFT from clinical data available before implantation. The results are mixed. Several studies have found no clinical predictors of high DFT, possibly owing to limited sample size.228-230 Other studies have found statistically significant predictors of a high DFT.231-236 The most common predictors are large cardiac size, large body size, wide QRS, high New York Heart Association (NYHA) functional class, VF as the manifesting arrhythmia, and low EF. These characteristics are helpful in alerting the implanting physician to the potential for a higher DFT, but their predictive value in an individual is not good enough to be reliable. Thus, a large patient with an enlarged heart may not necessarily have a high DFT. Another and perhaps more reliable item to consider is the obvious one of the previously determined threshold. If the DFTs at prior implantations in a patient were high, assuming that the electrodes were positioned in appropriate locations, one should expect a high DFT at a device change-out or when implanting a new system in that patient. In a typical pectoral implantation of a “hot can” system (in which the device housing is a defibrillation electrode), one occasionally encounters a patient with high DFTs. Several approaches can be used to lower the DFT. The first is to check that the RV electrode is in a reasonable position. Attempts can be made to reposition the RV electrode as far into the apex as possible. This may increase the voltage gradients in the area of low voltage gradient located at the apical region of the LV free wall. An alternative is to position the tip of the electrode in the high mid-septum/RV outflow tract region and the proximal end of the electrode toward the apex. This configuration may also improve the DFT by bringing the main body of the electrode closer to the septum. The addition of a right atrial/SVC lead can sometimes improve the DFT.103,237 The mechanism of such lowering consists primarily of lowering the shock impedance, increasing the current of the shock delivered through the RV coil and reducing pulse width. However, this benefit has not been established for millisecond waveforms. Although pectoral implantation of an ICD is clearly the standard, clinical circumstances occasionally prevent pectoral implantation. Implantation in the abdominal area may still be necessary in a few patients. In the past, such implantation used a long “single-pass” lead containing both the RV electrode and an SVC electrode. Incorporating the device canister into the shocking configuration by connecting the SVC and the abdominal can, so as to have the same polarity can significantly improve DFT.105 When an independent SVC electrode is available, one can consider positioning that electrode higher into the left innominate vein85 or into the azygos vein238 to attempt to improve the DFT.90 However, withdrawing the SVC electrode too far into the brachiocephalic region may reduce any benefit of such an electrode because of higher impedance in this location.92 Similarly, use of the subcutaneous array tends to lower the DFT. The use of a long subcutaneous electrode such as the model made by Medtronic or Guidant

may help lower DFT through more than one mechanism. This electrode is designed to be inserted at the lateral edge of the pectoral pocket and tracked subcutaneously along the back to a position near the spine. The tip of the electrode should be near the base of the ventricular shadow on an AP fluoroscopic view. This large surface electrode will reduce impedance of the shock as well as potentially redirect current posteriorly. Two other locations of an additional coil could be of help, the high lateral coronary sinus and the azygos vein. It may be more difficult to anchor the electrode in the coronary sinus. However, a coronary sinus lead would provide an excellent vector in combination with a RV apical electrode. The azygos vein, however, can provide an excellent position for a shock vector from the RV as well as a stable location. The azygos vein enters the SVC near the right atrial junction posteriorly. The ostium can be searched with a curved sheath or even a curved stylet inserted into the lead. Another potential placement for an SVC lead to improve DFTs can be exploited in patients with persistent left-sided SVC. Insertion of the ICD lead through the left-sided SVC, across the coronary sinus, and into the RV usually paces the proximal “SVC” coil in the coronary sinus. This would provide an excellent vector for defibrillating the heart. Alternatively, a separate SVC coil can be placed in the lower part of the leftsided SVC, positioning it near the lateral portion of the coronary sinus. Lastly, the hemi-azygos vein, when present in adequate size, can serve as the location of a proximal coil. Because of the more posterior nature of these veins, positioning a proximal coil in any of them would provide an excellent shock vector for improving DFTs. The presence of epicardial patches from a prior ICD implantation may affect DFT. Animal studies have suggested that locating an inactive epicardial patch in lowvoltage-gradient areas of the apical LV free wall, a not unusual area for such a patch, can markedly increase the DFT.239,240 However, a clinical study in patients in whom pectoral “hot can” systems were used did not demonstrate an unusually high DFT.241 For every trick to lower the patient’s DFT, there is a cost and a benefit. There is certainly a cost in terms of physician time, but there are also going to be material costs. Probably more significantly, there are patient costs with each trial at threshold lowering, resulting from additional time in VF and delivery of possibly unwanted additional shocks. Longer procedure time may also be associated with higher infection risks. Thus, one should try to follow an optimal path toward obtaining an acceptable DFT. Additional leads bring additional potential complications.242 Table 17-4 summarizes the various approaches to lowering DFT, listing them in order from least costly to most costly.242a In our experience, these maneuvers seldom fail to achieve an adequate DFT in patients demonstrating high DFTs when implanted with a standard transvenous ICD. However, if all of these maneuvers fail to provide an adequate safety margin for defibrillation, one can still consider implanting the device if defibrillation can be achieved with the highest output of the

Chapter 17: Testing and Programming of Implantable Defibrillator Functions at Implantation

549

Approaches to Reducing Defibrillation Threshold in Patients Undergoing Implantation of a Pectoral Device* TABLE 17-4.

Approach

Benefit(s)

Drawback(s)

Note(s)

Verify RV coil polarity is positive (anodal)

Lower or equal DFT in 88% of patients

Must be changed in some models

No “downside”

Add SVC coil

—

Dangerous to remove in presence of infection

May not help with programmable outputs

Use high-voltage device

DFT reduction ≤25%

Some devices are not labeled for voltage

Refers to benefit of going from 600 to 800 V

Program output durations to theoretical optimal

DFT reduction ≤40%

—

Only in devices with programmable pulse widths

Add subcutaneous coils

Reduces durations for nonprogrammable devices

Infection risk, back pain Difficult to implant

—

Add lead in coronary sinus or azygos or hemi-azygos vein

Good current vector

Possibly difficult to implant

—

DFT, defibrillation threshold; RV, right ventricular; SVC, superior vena cava. *Listed in order from least costly to most costly.

device. Clinical evidence would suggest that such patients nevertheless benefit from such an implant.243 The alternative at present is to proceed to an epicardial or pericardial implantation.

Evaluation of Sensing A most critical function of the ICD, of course, is to sense VF. Inappropriate shocks quickly destroy the patient’s quality of life. Unlike ventricular electrograms during sinus rhythm, electrograms during VF can vary widely in amplitude. ICD manufacturers have different engineering approaches to accommodate this variability. Thus, the best approach to assessment of VF detection is to actually induce VF and assess this function of the implanted device. The availability of annotation on the stored or telemetered electrogram indicating detection of ventricular beats is helpful in analyzing the reliability of detection. Owing to the variability of VF amplitude, some dropout of detection from a beat-to-beat basis may be present in all devices. However, one must be assured that such dropout will not be long enough to divert the device from progressing to delivery of therapy. A second sensing function that must be tested is interference from electrical pulses delivered from another device, typically a pacemaker. Although this problem may become less important as dual-chamber ICDs become more commonly used, such testing must ensure that the pacemaker will not interfere with the ICD’s detection of VF. For implantation of an ICD lead into the RV where a previous pacemaker lead is already present, it is ideal to implant the ICD lead remote from

the pacemaker lead and to have its sensing electrode oriented more or less perpendicular to the pacer electrodes.244 With bipolar pacemaker electrodes, one can generally reduce the pacing artifact to a small enough amplitude to minimize cross-sensing. An additional improvement can be obtained by using ICD leads with dedicated bipolar sensing rather than “integrated bipolar” sensing, which incorporates the RV shocking electrode into the sensing circuit.245,246 The dedicated bipolar sensing leads use two closely spaced electrodes at the lead tip for sensing, thus minimizing any far-field sensing. The disadvantage of such a lead is that the RV shocking electrode is moved further away from the tip of the lead to accommodate the sensing electrodes, possibly reducing its efficacy somewhat. In most circumstances, however, when one wants to use such a lead, there is a pacemaker lead in the RV apex, and the ICD lead should be place higher on the septum. With such placement, the position of the shocking coil near the tip of the lead may not be important. A further advantage of such a dedicated bipolar lead is that it may be possible to implant it even with a unipolar pacemaker, a circumstance generally considered a contraindication to ICD implantation. Because of the dedicated bipolar sensing of such a lead, even unipolar pacing artifacts may be small enough to not interfere with VF sensing. Such a system, of course, should be thoroughly tested at implantation to ensure lack of interference by programming the pacemaker to its maximal output in the VOO or DOO mode during VF induction. One must not forget about the potential sensing interference that the atrial lead may generate when a dual-chamber pacemaker is present. The output of the leads from both chambers should be

550

Section Two: Clinical Concepts

maximized during VF testing to ensure that they will not interfere with ICD sensing of VF. With the advent of dual-chamber defibrillators, sensing in the atrium may also play a role in detection of tachycardia and delivery of therapy. Because automatic gain controls may amplify the atrial signals considerably, it is important to locate the atrial lead in an area where minimal or no ventricular electrogram is detectable even with high gain. This would avoid potential confounding of the detection enhancements available with such devices, which incorporate atrial sensing as part of a detection algorithm. Such optimal placement would also be important in functions such as atrial tachycardia detection and mode switching. Testing and Programming of Antitachycardia Pacing ATP is an important therapeutic modality of the ICD. Shocks from the ICD even at very low energy outputs, such as 2 to 5 J, are usually perceived as painful. Most clinical shocks from the device are received in the conscious state, sometimes causing marked anxiety in the patient, so the ability to terminate tachycardia with pacing offers a relatively pain-free approach that is sometime imperceptible to the patient. The role of ATP can vary considerably from patient to patient. Testing of ATP may be important in those patients in whom empirically programmed approaches were clinically unsuccessful in terminating tachycardia. In patients with frequently occurring clinical VT, testing of ATP may also be useful in establishing the most effective means of terminating the tachycardia without promoting acceleration. However, testing in the electrophysiology laboratory with the patient sedated may not be the same as clinical tachycardia occurring during daily life, in which there may be varying degrees of sympathetic tone. Even for faster tachycardia, empirically programmed ATP can be quite useful in terminating a majority of clinical events. Several reports have indicated the usefulness of activating the ATP feature of the ICD even in the patient who has not yet experienced an episode of clinical sustained ventricular tachyarrhythmia. Retrospective analyses of stored electrogram data in the ICDs of patients with recurrent episodes of VT showed that about 90% of tachycardia can be terminated with ATP.247,248 It had been assumed that faster VTs were less likely to respond to ATP, having a higher chance of accelerating with the pacing. Thus, in general, ATP has not been commonly used for tachycardia faster than rates of 180 to 200 bpm. However, two clinical trials have demonstrated the efficacy of ATP in tachycardia with rates up to 250 bpm. The PAcing Fast VT REduces Shock ThErapies (PainFREE Rx) clinical trial enrolled 220 patients with coronary artery disease who underwent ICD implantation for standard clinical indications.249 A fast VT zone was programmed in all patients to allow ATP for tachycardia with a CL shorter than 320 msec but longer than 240 msec. Two bursts of ATP, 8 beats each, were programmed for this zone. The first burst was programmed at 88% of tachycardia CL,

whereas the CL of the second burst was shortened by 10 msec. Over a mean follow-up of 6.9 months, 52 patients experienced 446 episodes of fast VT. Eighty-five percent of the VT episodes were terminated with the trial prescribed ATP therapy in this zone. Approximately three fourths of the patients who experienced fast VT experienced successful termination of their fast VT episodes and did not receive a shock. Several patients did experience syncope. However, owing to the nonrandomized nature of this pilot study, it was unclear whether shock therapy could have prevented syncope. The PainFREE Rx II trial was a randomized study in which patients were assigned to either initial ATP therapy for fast VT or to immediate shock therapy.250 The ATP therapy was programmed as a single ATP sequence, 8 beats, at 88% of the tachycardia CL. The results indicated that more than 70% of tachycardia episodes with rates in the range of 188 to 250 bpm could be terminated by one burst of ATP. The incidence of syncope was no higher in patients programmed to receive this empiric ATP therapy than in those who were programmed to receive immediate shocks. In this trial, median numbers of VT episodes were similar for the patients undergoing primary prevention (248 of the 582 patients) and those undergoing secondary prevention.251 Thus, patients receiving ICDs for primary prevention of ventricular arrhythmia–related deaths, in whom the first episode of tachycardia could reasonably be expected to have a faster rate, can nevertheless benefit from empiric programming of a tachycardia zone where ATP is used. Avoidance of shocks reduces the morbidity of ICD therapy in this patient population. Quality-of-life assessment of patients in the PainFREE Rx II trial demonstrated better physical and mental outlooks in those receiving the ATP therapy than in those receiving shocks. Clinical programming of an empiric tachycardia zone should take into account several factors, including the expected heart rates during exercise, any known history of atrial fibrillation, the use of β-blockers, and any prior documented sustained VT. In a young patient ( 60 sec BV < 2.62 V or CT > 7.5 sec

St. Jude Medical, St. Paul, Minn.

Profile MD V-186 Atlas + DR V-243 Epic HF V-337

BV < 2.55 V BV < 2.45 V BV < 2.45 V

Telectronics Pacing Systems, Englewood, Col. (now Pacesetter, part of St. Jude Medical, Sylmar, Calif.)

Guardian ATP, 4210 Guardian ATPII, ATPIII, 4211, 4215

Remaining battery life 20 sec BV > 4.90 V Programmer indicated

*After capacitor reform. BOL, beginning of life; BV, battery voltage; CT, charge time;

evidence does not exclude lead fracture. A break in the connection of the lead to the generator, or within the lead itself, can produce intermittent loss of energy delivery to the heart, which in turn results in absence of pacemaker spikes. Undersensing, or oversensing due to chatter, may also occur with lead conductor fracture. Lead dislodgement produces intermittent noncapture or failure to sense that may be related to respiration. Pacing thresholds needed to achieve consistent capture may rise significantly. Lead impedance increases or remains unchanged. Fluoroscopy may demonstrate a loose or displaced lead tip but is not always diagnostic. Special Issues Regarding Implantable Cardioverter-Defibrillator Leads Evaluation of the ICD generator and its lead system poses a special problem in patients who remain free of arrhythmic events after ICD implantation. The ICD lead remains the weak link in the ICD system for the patient. Oversensing due to diaphragmatic impulses or extraneous signals may inhibit pacing therapy or lead to inappropriate delivery of “treatment” for presumed ventricular tachyarrhythmias that actually represent noise sensing (Figs. 20-4 and 20-5).52 Further, although fracture and degradation of transvenous leads have become less common with transvenous, as opposed to

epicardial, ICD lead systems, they can nevertheless occur with some frequency, necessitating reprogramming or reoperation.53 Depleted battery status is readily evident on routine ICD follow-up (as described previously and in Table 20-6), and integrity of the ICD shocking conductors can be evaluated easily in most current devices. High shocking electrode impedance measurements may indicate lead discontinuity due to conductor fracture or a lead-generator interface problem. Measuring high-energy electrode impedance traditionally required delivery of a shock, either for a clinical tachyarrhythmia or as part of a noninvasive testing protocol. In the absence of consensus or guidelines for performing noninvasive programmed stimulation routinely during follow-up (during which shock electrode integrity may be documented),54 it was not uncommon in patients with early ICD devices and infrequent shocks for the first documentation of high-energy conductor fracture to occur when they presented for generator replacement. About 10% of patients undergoing ICD generator replacement due to battery depletion were found to have a previously undetected sensing or defibrillation system failure.55 The operator therefore should test the lead system carefully during the replacement procedure and should be prepared to deal with malfunctioning leads at the time of generator change.

Chapter 20: Approach to Generator Change Pacing rate Pacing interval Average cell voltage Cell impedance Sensitivity Lead impedance Pulse amplitude Pulse width Output current Energy delivered Charge delivered Tachycardia detected

68 873 2.63 1.00 8 41 7.42 0.45 96.5 173.2 42.85 No

ppm ms volts KOhms mV ohms volts ms mA UJ UC

Figure 20-3. Acquired telemetry data from an Intermedics, Inc. (Angleton, Tex.), Intertach 262-14 VVICP antitachycardia pacemaker connected through a bipolar coaxial lead to the right ventricular apex. Battery voltage and cell impedance are normal. Measured lead impedance of 41 W is, however, extremely low. In this patient, measured lead impedance was normal in the supine position but decreased with sitting or when the device was pulled inferiorly. This behavior indicates a break in the inner insulator between the coaxial conductor strands in the area of the clavicle. With movement, the conductors contact each other, resulting in low impedance and preventing delivery of electric current to the heart. Because output voltage is fixed, the low lead impedance results in a high delivered current (96.5 mA) and energy.

25 mm/s

S 70

VS 650 VP 857

Newer systems automatically measure high-energy lead impedance at device interrogation in a manner similar to that used for standard pacing and sensing electrodes. The lower-energy impulses delivered by these devices may be more sensitive to the detection of microfractures than would higher-energy shocks. Determination of Pulse Generator–Lead Interface Malfunction Pulse generator–lead interface problems may be grouped into the following three categories: (1) loose, incomplete, or uninsulated connections, (2) reversal of atrial and ventricular leads in the pulse generator connector block (for ICDs, reversal of shocking electrode polarity may also occur), and (3) pulse generator–lead mismatch. A loose pace/sense lead connection should become apparent with noninvasive testing. The device may fail to deliver pacing spikes when appropriate, it may intermittently fail to sense, and/or it may oversense as a result of chatter due to intermittent contact with the 07-AUG-98 16:43

VF 176 VS 648

833

Surface ECG Rate EGM Markers

VF 146 VS 729

VP 857 VP 857

VP 857

Figure 20-4. Sensing of diaphragmatic myopotential during periods of deep inspiration can lead to inappropriate triggering of the antitachycardia functions of an implantable cardioverterdefibrillator (ICD) (Endotak, CPI, St. Paul, Minn). These electrograms are recorded from an ICD placed with a passive-fixation endocardial lead that incorporates integrated bipolar sensing and high-energy shocking coils in the right ventricular apex and the superior vena cava. Surface electrocardiogram (ECG), rate-sensing electrograms (EGMs), and marker channels all record spurious signals that represent inappropriate sensing of extracardiac electrical potentials. The frequency of these signals is high, and they occur after paced events as well as after sensed events. Pacing increases the gain of the device to avoid undersensing of low-amplitude signals of ventricular fibrillation. An underlying paced rhythm exists at a cycle length of 857 msec, but even this is altered by oversensing. The first paced complex (VP 857) is followed by two inappropriately sensed events (VS 650 and VS 648) that inhibit ventricular pacing output. Because the next native QRS complex occurs close to an inappropriately sensed signal, it is interpreted by the device to represent sensing in the ventricular fibrillation zone (VF 176). After that, another myopotential is inappropriately sensed (VS 729), and the native QRS is again sensed in the VF zone (VF 146). Finally, three sequential paced events occur at intervals of 857 msec, despite the presence of spurious electrical signals, which are not of sufficient amplitude to trigger sensing. Repetitive events such as these could lead to inappropriate antitachycardia therapies or prolonged periods of inhibition of pacing. This lead was extracted and replaced with an active-fixation endocardial defibrillation lead positioned distally on the lower region of the interventricular septum. (Photography by Todd Forkin, Hahnemann University Hospital, Philadelphia.)

834

Section Three: Implantation Techniques

V S

V S 3 2 0

6 5 0

V S 1 6 0

F FF S SS 11 2 22 7 00 0

F F V V F F F F S S S S S S S S 1 3 4 2 1 2 2 5 7 4 2 1 7 6 6 0 0 0 0 0 0 0 0 0

V S

FF F SS S 2 1 2 6 2 4 0 0 0

6 3 0

V S 2 5 0

FF SS 1 5 2 4 0 0

VF F F F VV SS S SD SS 1 2 2 1 3 1 1 2 8 3 2 0 2 9 0 0 0 0 0 0 0 VF Rx 1

V V V S S S 2 3 5 5 7 2 0 0 0

V S

Figure 20-5. Conductor fracture with intermittent contact of the broken ends of the lead wire may result in inappropriate sensing of noise chatter, as demonstrated in this example. Underlying native QRS complexes are difficult to discern. Noise sense intervals vary from 120 to 650 msec. The high degree of variability of sensed intervals, as well as frequent nonphysiologic intervals shorter than 200 msec, lead to the diagnosis of noise sensing. In this example, the number of sensed intervals in the ventricular fibrillation zone is great enough to trigger the VF detection algorithm of the device; this is recorded by the marker channel (FD). Ventricular fibrillation therapy was inappropriately delivered (VF Rx 1). Lead replacement, or placement of a new rate-sensing lead, is indicated. The top tracing is a ventricular electrogram, and the bottom tracing is an interpretation channel with interval measurements. (Photography by Todd Forkin, Hahnemann University Hospital, Philadelphia.)

setscrew. Oversensing can result in inappropriately high tracking rates or inhibition of ventricular output. Capture or sensing problems may be exacerbated by manipulation of the device. An uninsulated connection most commonly produces current leakage (an electrical short circuit in the system) that inhibits pacing or sensing. Leakage can occur if a setscrew is not properly insulated or tightened or if sealing rings on the lead header do not prevent body fluid from oozing into the pulse generator connector block around a loosely fitting lead. Leakage around lead header sealing rings may result from a loose lead connection or lead–pulse generator mismatch. Lead impedance in pulse generator–lead interface problems varies, depending on the specific situation. A loose, unconnected lead that remains in the pulse generator connector block, so that lead header sealing rings prevent fluid from entering, causes a break in the electric circuit and a very high impedance. If fluid enters the pulse generator connector block around a loose lead or at the level of a setscrew and maintains contact with body fluids, however, the resultant electric short circuit can produce very low measured impedance. As with lead fractures, impedance can vary with manipulation of the device. Reversed lead connections (i.e., atrial lead in the ventricular port, and vice versa) should be evident before the patient leaves the implantation laboratory, allowing immediate correction. Some atrial and ventricular leads are marked to enable easy identification; however, it is not uncommon to place “generic” leads into both chambers, especially straight screw-in leads, which may not be marked. Likewise, atrial and possibly the left and right ventricular lead pace/sense headers for insertion into ICD ports are both of the International Standard IS-1, so the implanter must exercise care in placing these leads properly into the appro-

priate locations in the ICD generator connector block. It is also possible, in patients in whom a pacemaker or ICD remains inhibited because of native electrical activity, to see no pacing spikes initially after implantation. To be certain that the pulse generator–lead system functions appropriately immediately after implantation, the device should be programmed to an atrioventricular (AV) delay shorter than the intrinsic PR interval, the device should be checked with a programmer or (for a pacemaker) one should place a magnet over the device after the leads are attached to document appropriate function before the pocket is closed. Caution exercised at implantation should avoid reversed leads; for example, we always connect the ventricular lead first to ensure pacing in the proper chamber. Beyond ensuring the presence of adequate and appropriate lead connections to the pulse generator, the battery connector block and leads must be compatible (see later).56-58 This issue is especially important with older lead models for device upgrades or generator replacements. Incompatibility can result in fluid leakage or loose connections, with resultant loss of pace/sense or shocking capabilities, requiring reoperation. Detection of Need for Reoperation for Other Reasons Other indications for pacemaker or ICD generator replacement or lead revision (see Table 20-1) generally become apparent through careful patient evaluation. Abrupt pulse generator failure with no antecedent sign of battery depletion is rare but can occur, producing symptoms in pacemaker-dependent patients. In others, abnormal pacing output or rate, lack of pacing output, or inappropriate sensing due to generator malfunction may be detected by remote interrogation at home or in the

Chapter 20: Approach to Generator Change

physician’s office.17 Of particular importance to patients with ICDs are the possibilities of no output when required to terminate tachyarrhythmias, inappropriate shocks due to oversensing of diaphragmatic or lead chatter artifact (see Figs. 20-4 and 20-5), and oversensing of extraneous electromagnetic signals, such as surveillance systems or high-voltage generators, that can be sensed as ventricular fibrillation or can inhibit ventricular pacing output. Cellular telephones rarely present substantial interference due to variations in signal frequency.59,60 Development of pacemaker syndrome in patients with implanted ventricular demand (VVI), ventricular rate-responsive (VVIR), or atrial rate-responsive (AAIR)22 pacemakers presents another indication for device revision. This need should be apparent from history and physical examination, although confirmatory blood pressure or cardiac output measurements may be required. Documentation of hemodynamic improvement with dual-chamber synchronization may require placement of a temporary atrial lead before reoperation for upgrading to a dual-chamber system. Pacemaker syndrome occurring with an implanted functioning dual-chamber pacemaker must be managed by reprogramming.61,62 Interchangeability of Products from Different Manufacturers Unlike pacemaker leads, most early ICD leads from different manufacturers were compatible only with ICD pulse generators from the same manufacturer. For later models, manufacturers have adhered to standard header designs for ICDs, including IS-1 ports for the pace/sense lead heads from both atrial and ventricular leads. Defibrillation ports now also follow a standard for defibrillation lead headers, DF-1, a 3.2-mm unipolar lead head with sealing rings. The newest agreed-on IS-4 standard, which provides four electrical connections combining the functions of a bipolar pace/sense connection with up to two high-voltage connections, should reduce some of the confusion. However, there are two IS-4 connections, one for devices with and one for devices without high-voltage (defibrillation) capacity. In the end, this situation should simplify connections except when extra leads for defibrillation are adapted into the lead system. For procedures involving old, nonstandard ICD connector blocks, however, the operator must be familiar with the existing system of leads and generator in the patient before surgery, and technical support from the manufacturer may be required at the time of the operation. A full range of adapters, or various header designs, to mate a replacement generator to the existing leads must be available. Ensuring tight and proper connections between the generator and the lead, and any adapters and lead extenders, avoids malfunction and current leak. Although older adapters used an uncured medical adhesive to seal set-screws in the connector block of the device, some newer adapters use set-screw seals similar to those found in pacemaker pulse generators.

835

Special Indications for Replacement of Implantable Cardioverter-Defibrillator Generators In addition to the indications outlined in Table 20-1, the ICD generator may need to be replaced for the reasons discussed here. Malfunction of the Generator with or without Lead Malfunction. Hardware or software errors in the ICD generator—or, more commonly, malfunctioning ICD leads—may result in the need to revise the ICD system. The overall reported incidence of lead-related complications has ranged from 2% to 28%.63,64 These complications commonly manifest as inappropriate shocks resulting from oversensing of noise (see Fig. 20-5) or as ineffective shocks from the shunting of defibrillation energy due to an inner insulator breach. Upgrading the Device to Incorporate Tiered Therapy and Multiple Zones of Therapy. Older devices incorporated monophasic shock energy for defibrillation, which was more likely to achieve an inadequate defibrillation threshold (DFT) than biphasic lead systems. The success of current ICD generators is attributed primarily to the use of biphasic shocking waveforms65-68 and tiered therapy, incorporating antitachycardia pacing. Certain older ICD systems and “stripped-down” shock-only rescue units were designed to deliver only shock therapy. In the rare instance that a shock-only ICD is still in place and requires replacement by an ICD that offers more sophisticated therapies, including tiered therapies for different arrhythmia zones, reoperation is required. Upgrading to a Higher-Energy Device or Addition of Hardware for an Inadequate Defibrillation Threshold. Occasionally, through invasive or noninvasive testing, the physician determines that the best function of the device may be achieved through a change in its hardware configuration. This may involve reoperation to place a generator capable of delivering higher defibrillation energy to respond to an elevated DFT, repositioning the right ventricular apical (RVA) shocking electrode, or the addition of various other lead systems, including superior vena cava coils and subcutaneous coils, arrays, or patches. Clearly, location of the RVA lead as close to the cardiac apex as possible affords the lowest DFT. Similarly, addition of various other leads to better distribute current around the heart can also reduce the DFT,69,70 often concomitantly lowering shocking electrode impedance for higher current delivery. Finally, if these invasive adjustments fail to reduce the DFT, waveform adjustments65 or replacement of the pulse generator with a higher-energy system may be warranted. Many patients with ICDs continue to require treatment with antiarrhythmic medications, which can affect the appropriate functioning of ICDs. The most common antiarrhythmic medication–ICD interaction observed is that of an elevated DFT. This is particularly evident with potent sodium channel–blocking drugs and with amiodarone.71,72 In this regard, reoperation may be required for antiarrhythmic drug changes that lead to substantial alterations in the DFT, although

836

Section Three: Implantation Techniques

elimination of the offending medication provides a more straightforward solution. When that is not possible, the physician may consider use of a device that delivers higher energy for defibrillation. Addition of a superior vena cava (SVC) coil or subcutaneous array may be indicated. Upgrading to Incorporate Dual-Chamber Pacing Capability. With an established role for β-blockers in the treatment of congestive heart failure and coronary artery disease, as well as a baseline frequency of developing sinus node dysfunction or AV conduction disorders, substantial numbers of patients with implanted defibrillators require dual-chamber bradycardia pacing backup. This problem is easily resolved in the new implant but can require considerable deliberation when substantial hardware is already in place. Several scenarios may be encountered, each with unique potential solutions. Some of these situations, with possible approaches, are described here. The patient may have a previously implanted abdominal single-chamber ICD with a fully functional epicardial lead system. In this situation, the operator has three primary options, as follows: (1) to place an endocardial atrial pacing lead through the subclavian system and tunnel the lead subcutaneously to the abdominal pocket, while upgrading the device to a dual-chamber ICD, (2) to abandon the abdominal ICD and place an entirely new AV sequential ICD system in the pectoral area, and (3) to place a totally separate dual-chamber permanent pacemaker system with atrial and ventricular leads and perform device interaction testing to ensure that neither of the two devices inhibits the other. There are various advantages and disadvantages to each of these techniques. For the first option, the advantage of long-term stability of thresholds for endocardial pace and sense leads speaks for the approach of adding an endocardial atrial lead and tunneling it to the abdomen, but it also requires that the lead be long enough to tunnel to the abdominal site. This makes manipulation and positioning of the lead in the atrium more challenging. Alternatively, a lead extender may be attached to a shorter lead, but this arrangement adds another weak link in the system in the form of an adapter. Finally, this approach requires opening both the abdominal pocket and the subclavian site simultaneously, which could raise the risk for cross-infection of the abdominal site. Accordingly, we prefer not to have two pockets open at the same time, especially when one involves an epicardial lead system, where infection could be disastrous. Abandoning the abdominal site altogether, the second option, may be the preferred technique because it eliminates the need to open two pockets simultaneously; it also eliminates the need to depend on any epicardial leads, which have a higher failure rate than endocardial leads. The new pulse generator and lead system are placed in the standard manner in the pectoral area, and DFT testing is performed at implantation; the previous abdominal pocket remains closed during this operation, eliminating the possibility of cross-infection between sites. Electrical shielding

afforded by the epicardial patches may affect the endocardial DFT. This problem may well be offset, however, by the improved long-term reliability of the endocardial lead system. The abdominal generator may be turned off and left in place, or it can be removed after implantation of the new system, preferably during a separate procedure. This approach is particularly useful in patients in whom a new high-energy ventricular coil is needed, which can be placed endocardially. The third option, placement of a totally endocardial dual-chamber pacemaker system with atrial and ventricular bradycardia pacing leads, also eliminates the need to open the two pockets simultaneously. It also affords the advantages of prolonging ICD battery life (through a separate pacemaker battery) and provides consistent and separate bradycardia pacing backup through a device designed primarily for bradycardia support. Implantation of a separate pacemaker and an ICD may, however, lead to various device-device interactions, including undersensing of ventricular fibrillation.72 The availability of combined dual-chamber pacemaker and ICD systems should reduce the physician’s concern about device interaction. This option is our least-favored approach. Upgrade to a Biventricular, Cardiac Resynchronization System. Upgrade to a biventricular (BiV), cardiac resynchronization system has become one of the most common indications for ICD reoperation, either as a de novo device upgrade or at the time of generator replacement.73 Upgrade to a biventricular system requires device generator replacement and insertion of a new coronary sinus/LV electrode. We perform venography to ensure patency of the vasculature for the new lead if any difficulty is encountered in accessing the axillary vein. Implantation of the new lead and device often requires a pocket revision to accommodate the larger generator. Change of Implantation Site. Older ICD generators, because of their bulky nature, were routinely implanted in the abdominal wall. With the availability of active can electrodes and ICD generators of smaller size, consideration should be given to changing the implantation site to the pectoral area when leads need to be revised. This issue was addressed previously with respect to placement of atrial leads; it should also be considered when ventricular lead malfunction necessitates reoperation, especially with a long nonthoracotomy ICD lead tunneled from the subclavian area to the abdominal wall. Another example is malfunction of epicardial ventricular sensing electrodes due to conductor fracture, high pacing threshold, or oversensing. Options for such a situation are placing a ventricular endocardial sensing electrode and tunneling it to the abdominal pocket or abandoning the abdominal pocket to place an entirely new completely endocardial ICD system in a pectoral location. We prefer the latter approach, again because it eliminates the need to open the two pockets simultaneously in a patient with epicardial leads and patches, in whom infection could be devastating, and because it eliminates the danger of fracture of the epicardial leads due to wear.

Chapter 20: Approach to Generator Change

The implantation site may also need to be changed in the event of incipient erosion, or outright infection, of the device site. Clearly, a staged approach is most useful here. If the site is infected with a malignant organism such as Staphylococcus aureus, the pulse generator and leads should be removed, with use of appropriate antibiotic coverage; a separate surgical procedure will be needed for implantation of a pectoral, contralateral ICD system. The two pockets should not be opened at the same operation. Most commonly, removal of the entire system is required for cure of the infection. Tunneling a lead from a location in the abdomen to one in the pectoral area can damage the tunneled lead, especially the header. There is no definitive way to avoid this risk, although gentleness with respect to lead manipulation and use of a standard dilator or tunneling tool reduces the tendency to damage. The lead needs first to be positioned in the ventricle through manipulation at the shoulder, because a long lead would be difficult to maneuver from an abdominal location through a tunnel. Only after it is carefully positioned can it be tunneled; therefore, the connector head of the lead will be tunneled from the shoulder to the abdominal pocket. Such tunneling can place significant stress on the lead connector, which could damage it. When preexisting tunneled leads must be replaced, the tract can be dilated with extraction sheaths, although this procedure is time-consuming and complex. As noted previously, whenever possible, adapters should be avoided because they merely add another weak link in the chain of possibilities for lead malfunction. Complications of Pacemaker or Implantable CardioverterDefibrillator Implantation that Require Reoperation. Reoperation may be required for complications resulting from the initial implantation procedure.10,15,16,74-80 Decisions about surgery in patients with large pocket hematomas or effusions, cardiac chamber perforation by a lead, or a need to reposition the pulse generator must be made on an individual basis. Most small to moderate hematomas resolve; the risk of secondarily introducing infection through reoperation or aspiration can thus be avoided. Large hematomas or effusions that do not resolve and that compromise the blood supply through pressure on the overlying skin of the pacemaker or ICD pocket require evacuation followed

837

by primary closure, because the pocket cannot be left open with a device in place. Bolus dosing of heparin, use of enoxaparin, and large loading doses of warfarin should be avoided whenever possible to reduce hematoma risk. Pocket twitch (due to lead insulation break, loose lead-generator connection, exposed set-screw, battery insulation break, inverted unipolar insulated pacemaker pulse generator, or the need for high-output pacing in a unipolar system), diaphragmatic pacing, or skeletal muscle stimulation or myopotential inhibition81 (Fig. 20-6) may require surgical intervention if such problems cannot be solved by reprogramming. Identification of Pulse Generator Make and Model The most straightforward means of identifying a pulse generator showing signs of malfunction or operating in a mode that suggests end of service is to obtain information directly from the patient (Table 20-7). An identification card is provided by the manufacturer for each pacemaker and patient with an ICD, specifying the type of device, model and serial number, implantation date, name of implanting or monitoring physician, and, often, lead model and serial numbers. This information may also be obtained from records kept by the manufacturer, the implanting physician, the monitoring physician, the transtelephonic service that monitors the patient, or the institution at which the device was placed. If none of these sources of information is helpful, alternative methods must be used to identify the pulse generator. Identification of the make and model of the existing pulse generator is crucial to determining its true functional status and, with older leads, to have the necessary information to select a compatible replacement or upgraded device. In the rare instance in which a pulse generator cannot be identified before surgery, the implanting physician must have a full array of leads, generators, and adapters available at the time of reoperation. Magnet Response. The response of a bradycardia pacemaker pulse generator to placement of a magnet can assist in the identification of its manufacturer (Fig. 20-7; see Table 20-5). Most pacemaker pulse generators

1 sec Figure 20-6. Myopotential inhibition in a unipolar dual-chamber pacing system induced by pectoral muscle contraction in a patient who noted recurrent lightheadedness with activity. Atrial sensitivity is programmed to 0.5 mV, and ventricular sensitivity to 2.5 mV. Two atrially tracked complexes are followed by a ventricular premature depolarization. The fourth QRS complex occurs early as a result of atrial myopotential tracking. Thereafter, ventricular pacing output is inhibited by myopotentials for nearly 6 seconds, after which normal DDD function resumes. Intrinsic QRS complexes occurring during the period of inhibition may be obscured by myopotential activity. Programming the ventricular sensitivity to 5 mV avoided myopotential inhibition and the need for lead revision. Although atrial sensitivity could not be adjusted because of the low intrinsic P wave amplitude, atrial myopotential tracking remained asymptomatic.

838

Section Three: Implantation Techniques

TABLE 20-7.

Identification of the Pulse

Generator Manufacturer code and serial number code Implantation data Identification card Transtelephonic monitoring records Manufacturer’s implantation records Monitoring physician’s records Noninvasive testing Size, shape, thickness Magnet response (pacemakers) Interrogation (if manufacturer identified) Fluoroscopy: Size, shape Connector block Unipolar or bipolar (pacing) Single- or dual-chamber Number of ports for implantable cardioverter-defibrillator leads Identifying markings/codes Lead: Unipolar or bipolar (pacing) Active or passive fixation Number of high-energy coils Invasive testing: Direct identification of pulse generator Lead—manufacturer code and serial number code Type of connector Size of lead header

respond to magnet application by entering a fixedrate single-chamber or dual-chamber pacing mode corresponding to the type of generator and the programmed mode. Magnet rates vary among manufacturers and may provide a clue to the origin of the device. To undergo a magnet-activated test, the patient must be connected to an electrocardiographic recorder before the magnet is applied and must remain connected until after the magnet is removed. The first few paced complexes after magnet application may occur at a rate or output other than that seen later in the recording, providing identification data as well as information regarding the integrity of the pulse generator and lead system (e.g., the delivered pulse width may be reduced during the first few paced complexes to ensure that capture still occurs with an adequate safety margin). Furthermore, with constant magnet application over the pacemaker, some devices continue to pace at a fixed rate, whereas others cease pacing after a programmed number of intervals. Devices temporarily reprogrammed to a backup mode by electrical interference (e.g., electrocautery during surgery) may exhibit magnet responses that vary from the standard for such.

Radiographic or Fluoroscopic Identification of the Pulse Generator. Most pulse generators—both pacemakers and ICDs—can be identified from their appearance under radiography. This is the most helpful method of identifying unknown devices. The shape and size of the generator may characterize a particular manufacturer (e.g., square, oval, elongated ellipsoid, round). Pulse generator shape can vary significantly from one device model to another, however, even when produced by the same manufacturer. Considering that the life span of some pacemaker devices may exceed 10 or 12 years, various shapes and sizes may be encountered. More specific to identification of the pulse generator are radiopaque markings placed near the connector block that code for manufacturer and device model. These markings appear most clearly under magnified fluoroscopic or radiographic examination when the device is positioned perpendicular to the x-ray beam (Fig. 20-8). The shape and orientation of internal components, which can often be identified radiographically, provide further clues to the device type, manufacturer, and model. Comparison of these radiographic features (size and shape, identification markings, internal components) with compiled x-ray photographs available from manufacturers facilitates identification of the pulse generator. Finally, an attempt to interrogate a pulse generator with an “army” of different programmers may identify the pulse generator, unless the battery is so depleted that telemetry communication is not possible. Radiographic or Fluoroscopic Identification of Leads. Radiographic examination of leads serves two purposes.82 First, it allows the physician to ascertain the presence of unipolar versus bipolar distal electrodes and, frequently, the fixation mechanism. Distal activefixation screws may often be seen directly on radiography, whereas passive-fixation leads often have a bulbous tip. Second, radiographic examination identifies lead conductor fractures in which the conductor has clearly separated, leaving a gap; this identification may require magnified views. Lead information of this sort is important for programming (bipolar vs. unipolar), for selecting an appropriately compatible generator, and for identifying leads for extraction. Fluoroscopy also gives some indication of the degree of fibrosis evident through the real-time motion of the lead, information that could be useful if extraction is required. Radiography of leads involves an examination of the insertion site (e.g., subclavian, axillary, cephalic, jugular, or epicardial), acute bends or fractures in the lead, the location of lead coils beneath the pulse generator in the event that they need to be freed for lead repositioning or extraction, the position of the pulse generator connector block, and a general preview of the character of the connector block–lead interface (see Fig. 20-8). The lead should be examined fluoroscopically throughout its course for kinking, fracture, or excessive tension as well as for fixation at the distal tip. A thorough radiographic examination of lead integrity and pulse generator–lead interface before reoperation

Chapter 20: Approach to Generator Change Jun 18 1992

A

MAGNET

839

MAGNET

DEMAND DEMAND

B

C Figure 20-7. Examples of normal pacemaker generator magnet response. Tracings A and B were recorded transtelephonically during routine follow-up. Upper tracings in both A and B show magnet response (DOO), and lower tracings show demand mode with the magnet off. Tracing C was recorded by real-time surface electrocardiography at follow-up in the office. A, Representative tracings from Medtronic, Inc. (Minneapolis, Minn.), dual-chamber pacing systems, here from a model 7070 Synergyst II device programmed DDD. The first three atrioventricular (AV) sequentially paced complexes are delivered at a rate 10% higher than the magnet rate; the AV interval is shortened to ensure both atrial and ventricular capture. The first two AV complexes are delivered at programmed atrial and ventricular outputs, whereas the delivered pulse width of the third complex is reduced by 25% (“threshold margin test”). Thereafter, the device delivers AV sequentially paced complexes at a fixed rate of approximately 85 bpm at programmed output. The device in this example remains entirely inhibited in the demand mode. B, Representative tracings from ELA Medical (Sorin Group, Milan) dual-chamber pacing systems, here from a model 6034 Chorus device programmed DDD. Magnet application results in fixed-rate DOO pacing at a rate of 96 bpm at the programmed AV interval. The AV interval in this example is short owing to activation of the rateadaptive AV delay. Output during magnet application may be increased but reverts to programmed levels for six complexes after magnet withdrawal (see Table 20-5). The pacemaker appropriately tracks in the demand mode. C, Representative electrocardiographic recording from Intermedics, Inc. (Angleton, Tex.), dual-chamber pacing systems, here from a model 294-03 Relay device programmed DDD. The first four complexes are delivered at 90 bpm with a shortened AV delay. Programmed pulse width for the fourth complex is reduced by 50%. Thereafter, the device delivers AV sequentially paced complexes in a fixed mode (DOO) at the programmed rate, while the magnet remains in place for a total of 60 AV complexes in the DOO mode.

in pacemaker and ICD patients saves much distress when the pocket is opened. Invasive Evaluation After as much information as possible has been gathered noninvasively about the hardware of the pacing system and the functional status of all its components, further invasive evaluation may proceed at the time of reoperation. Invasive evaluation does not supplant noninvasive analysis but adds to it. Invasive evaluation involves (1) measuring the functional capacity of implanted leads, (2) examining the structural integrity of leads and the lead-generator interface, and (3) venography. Measuring the Functional Capacity of Implanted Leads Figure 20-8. Radiographic identification of a pulse generator. The unit is clearly connected to two leads, and each lead has only a single electrically active pole (i.e., unipolar). Although the shape and arrangement of electronic components assist in identification, the specific radiopaque code block inside the pulse generator provides the primary means of identifying the device. Here, a Medtronic logo, followed by S W 2, indicates that the pulse generator is a Medtronic, Inc. (Minneapolis, Minn.), Synergyst II model 7071 DDDR unit with a connector block that will accept two (atrial and ventricular) 5-mm or 6mm unipolar leads.

By far one of the most crucial parts of invasive analysis during reoperation involves measurement of pacing and sensing capabilities in existing long-term leads. Vigorous noninvasive evaluation should provide the operator with a significant amount of information regarding lead viability and functional status as well as a determination of pulse generator end of life.28,29,35,43,51 Verification of lead integrity and precise DFT determination must, however, be performed at reoperation. If

840

Section Three: Implantation Techniques

surgery is undertaken for pulse generator replacement, demonstrating viability of existing leads is vital to the appropriate long-term performance of the new battery. Surgery for lead repair or revision itself involves extensive testing of long-term leads to confirm the lead as the source of malfunction, ensure normal operation of other leads, and evaluate new leads for optimal positioning inside the heart. After the pacemaker or ICD pocket is opened, the pulse generator is disconnected from the leads to enable testing of lead sensing and pacing functional capacity.83 The lead must be disconnected from the pulse generator cautiously in pacemaker-dependent patients; to avoid prolonged ventricular asystole, the operator must be prepared to connect the lead immediately to a cable attached to a functioning external pacing system. The external device should be activated and should be delivering pacing impulses before the ventricular lead is disconnected from the pulse generator in a pacemaker-dependent patient. Alternatively, although it is not usually necessary, a temporary pacing wire may be placed before lead disconnection in a pacemaker-dependent patient; such additional instrumentation, however, may raise the risk of infection. Of course, the operator must exercise care in the removal of the pulse generator and dissection of the lead to maintain lead integrity. Invasive Testing of the Sensing and Pacing Capabilities of Leads that Have Been Implanted for a Long Time One of the most important aspects of invasive testing involves measurement of pacing and sensing thresholds in long-term pacemaker and ICD leads. Pace and sense lead thresholds rarely remain as low as those at initial lead implantation. Most leads show some deterioration in pacing and sensing thresholds during the first 4 to 8 weeks after implantation, then reach a relatively stable level for the long term.23,50 It is possible, however, for thresholds to continue to increase over time, a change that may not be recognized by transtelephonic monitoring alone. The change in threshold from baseline appears greatest with active-fixation, non–steroid-eluting leads; threshold increases are reduced with passive-fixation and steroid-eluting leads, even if the steroid is applied to active-fixation systems. Noninvasive testing should give the operator some clues to the usefulness of long-term leads, but invasive testing confirms their functional utility. Both atrial and ventricular leads must be tested. If bipolar, they should be evaluated in both unipolar and bipolar configurations. The external pacing analyzer is connected to the lead; pacing and sensing thresholds and lead impedance are determined. The voltage pacing threshold at a fixed pulse width is recorded as the threshold that produces reliable capture. Delivered current can then be measured. Pacing lead impedance is best determined at an increased output voltage (e.g., 5 V) to ensure accuracy. Low-voltage pacing thresholds are desirable for longterm leads. This allows programming of the pulse gen-

erator output to a reduced level, enhancing battery longevity. For leads that have been in place for several years, the operator may decide to accept a pacing threshold (at 0.5-msec pulse width) of up to 2.5 V because this provides a two-times pacing safety margin for most pulse generators and because long-term leads generally show little additional increase in threshold over time. However, a pacing threshold of 2.5 V that occurs early after implantation (e.g., within 6 months) may not be acceptable. This situation suggests excessive early fibrosis around the lead tip and the possibility that exit block and noncapture will develop in the future if the pacing threshold continues to increase. Care at initial implantation helps ensure lower long-term pacing thresholds and improved sensing capabilities. Thresholds for sensing likewise tend to increase after lead implantation. Acceptable measurable intracardiac electrogram amplitudes depend on the maximum programmable sensitivity of the new pulse generator. For most systems, a P wave amplitude of 1 mV or more and R wave amplitude of 3 mV or more constitute minimally acceptable long-term values. Such low amplitudes, however, leave little room for further deterioration in lead function. Values of 1.5 mV or more and of 4 mV or more for P and R sensing, respectively, provide an additional safety margin. If atrial or ventricular ectopy is present, the operator should determine electrogram amplitude of ectopic complexes to ensure appropriate sensing by the pacemaker. In patients with paroxysmal atrial fibrillation, excellent atrial sensing may be required to detect atrial fibrillation reliably when it occurs, without signal dropout. Higher-amplitude electrograms are required for unipolar (versus bipolar) leads to allow programming of lower sensitivities to avoid myopotential sensing interference. Inadequate sensing or pacing thresholds at the time of generator replacement are indications for placement of a new lead in the affected chamber. This may entail either capping an old lead and leaving it in place or removing it. The new lead can usually be placed through the same subclavian or axillary vein, although it is preferable to avoid having too many leads (especially more than three) pass through the same vessel, to reduce the chance of venous occlusion and thrombosis. A single new lead may also be placed through the internal jugular vein, external jugular vein, or contralateral subclavian or axillary vein. The proximal tip can be tunneled to the original pocket to meet a second, functional long-term lead for a dual-chamber pacemaker system if required. Alternatively, an entirely new generator or lead system may be placed on the contralateral side.84 Invasive Testing of the Defibrillation Capacity of High-Energy Leads that Have Been Implanted for a Long Time DFT testing of ICD leads can be performed after evaluation of the pace and sense functions of the multifunctional lead and proceeds according to standard DFT testing techniques. Stability of the intracardiac

Chapter 20: Approach to Generator Change

electrogram recording should be ensured, and visual inspection should demonstrate no significant abnormalities in lead appearance, consistent with lead structural integrity. Examining the Structural Integrity of Leads and the Lead-Generator Interface Visual inspection at surgery provides clues to lead integrity. Fluid inside the lead body suggests an outer insulation break but, especially in low-voltage leads, does not mandate lead replacement. Undue tension on the lead near the fixation site may cause kinking, conductor uncoiling, conductor fracture, or thinning of the electric insulator. A hazy appearance of the insulator surrounding an area of tension or repeated stress is common in older leads. This appearance represents surface erosion of the lead insulator and does not itself imply lead malfunction. The finding should, however, alert the operator to the possibility of lead damage in areas of stress to the insulation. An examination of the suture location ensures that the ligature remains around the suture sleeve, and gentle tension on the lead body ensures its fixation at the venous entry site. Visual inspection of the specific course of a coiled lead in the pocket may be hampered by a significant thickness of overlying capsule scar; fluoroscopy can assist in this regard.39-42 Direct examination of the lead connector can assist in the identification of the lead model if not previously known.56-58 This issue is particularly important for lead models that have been found to have excessive premature failure rates; such leads in the ventricular position should be replaced in pacemaker-dependent patients. Venography Venography is becoming more commonly required as part of the device replacement procedure. It plays an important role when insertion of replacement leads into the subclavian vein is rendered difficult, because the evaluation can ensure patency of the subclavian and SVC systems, demonstrate the point of venous occlusion, or show the course of the axillary venous system for direct access. Venography is indicated when the subclavian, axillary, or cephalic veins cannot be accessed (to demonstrate their locations) and when the veins are accessed but a guidewire cannot be passed into the SVC. Inability to access the subclavian vein that carries a previously implanted lead suggests either an incorrect needle insertion angle or an occluded subclavian or brachiocephalic venous system.85-89 Finding an appropriate location to insert the access needle can be facilitated by advancing the needle fluoroscopically in the direction of the chronic electrode under the clavicle, with care taken not to damage the implanted lead. The vein should be approached with the bevel of the needle facing the implanted lead. If access is not possible, venography may provide better delineation of the course of the axillary or subclavian vein. In this situa-

841

tion, radiopaque dye must be injected distal to the veins to be visualized, that is, into the basilic or median cubital vein. Occasionally, access to the axillary or subclavian vein is possible, but the guidewire will not pass freely to the SVC. If needle placement in the vessel is adequate, failure to pass suggests proximal venous occlusion.90-94 Venography demonstrates whether occlusion is indeed present and, if so, its site. Chronic venous occlusion may occur asymptomatically in conjunction with the development of collateral venous circulation around the shoulder. Delineation of the location and length of occlusion indicates to the operator an appropriate needle insertion site for placement of a new lead. It also ensures patency of the SVC. Dye is injected directly into the subclavian vein. Occlusion of the subclavian system proximal to the junction of the internal jugular vein excludes the ipsilateral jugular system as an alternative site for a new lead. Alternatively, if the subclavian vein is occluded and the internal jugular vein remains patent, a new lead may still be placed using the jugular approach. Occlusion of the SVC precludes the use of any new endocardial lead placed from a superior site unless the vessel is dilated after lead extraction.95 Although venoplasty is acceptable, stents should never be placed into veins without removal of the leads, so as to avoid trapping of the leads between a stent and the vein wall. Anomalies of the left SVC (which usually drain into the coronary sinus) make placement of right ventricular endocardial leads difficult or impossible.96,97 Venography defines the anatomy of the venous system in such a situation, which may be suggested by an unusual intravascular guidewire course. Finally, leakage of venography dye into perivascular tissues or into the pericardial space suggests vessel or cardiac chamber perforation, respectively. The technique is performed by injection of 10 to 20 mL of radiopaque dye (a 50% dilution generally suffices) into a vein peripheral to the occlusion site. The dye may be injected directly into the subclavian vein or, if subclavian access is not possible, into the antecubital or brachial vein. Fluoroscopy with permanent storage of cine images is necessary to evaluate flow.

Generator Lead Adaptability Lead Connectors Replacing a pulse generator onto one or more long-term leads requires that the pulse generator connector block be compatible with the proximal lead tip.56-58 Through years of development by multiple manufacturers, pacemaker leads have evolved to an IS-1 proximal lead connector configuration, which consists of (1) a 3.2-mm lead connector with a short pin that is electrically connected to the distal electrode tip, (2) a lead connector ring wired to the proximal pacing pole (the ring is electrically active in a bipolar lead), and (3) sealing rings. The proximal lead connector configuration is the same

842

Section Three: Implantation Techniques

TABLE 20-8.

Common Configurations for Endocardial Lead Connectors Pin L/S

Connector Diameter (mm)

Sealing Rings? (Y/N)

Fixation (A/P)

Atrial J Available? (Y/N)

L

S

3.2

Y

A or P

Y

U

—

S

3.2

Y

—

N

VS-1

U or B

L

S or L

3.2

Y

A or P

Y

3.2-mm low-profile

U or B

L

L

3.2

Y or N

A or P

Y

5-mm

U

—

L

5.4

Y

A or P

Y

5-mm

B

B

L

5.4

Y

A or P

Y

6-mm

U or B

L or B

S or L

6.4

Y

A or P

Y

Unipolar (U) or Bipolar (B)?

Linear (L) or Bifurcated (B)?

IS-1

U or B

DF-1

Designation

A/P, active or passive; DF-1, Standard for high-energy defibrillator lead connectors; IS-1, International Standard for pacemaker lead connectors; L/S, long or short; VS-1, Voluntary Standard for pacemaker lead connectors; Y/N, yes or no.

for unipolar and bipolar leads, except that the ring is inactive in unipolar leads. Modern pulse generators have connector block specifications that conform to IS1 leads and that also fit the prior Voluntary Standard (VS-1) lead type. Thus, generator replacement onto implanted leads of either of these two types poses no difficulty because of the wide array of compatible pulse generators from multiple manufacturers. In a similar manner, ICD leads have evolved to incorporate IS-1 pace/sense connectors with concomitant IS1 atrial and ventricular ports in the generator connector block to attach these leads. Early-generation ICDs, however, used a wide variety of lead port sizes and configurations, requiring the operator to have available generators with various header port sizes, adapters, and upsizing sleeves for reoperation of old ICD generators implanted before the adoption of a uniform standard. Even coronary sinus lead ports have evolved from an LV-1 configuration without sealing rings to the IS-1 configuration. As discussed previously, high-energy defibrillation lead headers have also evolved into a standard configuration, DF-1, which consists of (1) a lead pin that is electrically connected to the corresponding high-energy coil or patch, (2) sealing rings, and (3) a single lead head for each coil of an endocardial lead to allow various hardwire configurations. A single DF-1 connector attaches to all three coils of a subcutaneous array or to a single patch. As with pace/sense leads, defibrillator lead heads had a variety of different end configurations before the adoption of the DF-1 standard, resulting in greater complexity at the time of generator or lead replacement. Because a variety of other lead connector configurations had previously been developed (Table 20-8; Fig. 20-9) and because implanted leads may remain useful for many years, a number of these older lead connector configurations remain in use. If sensing and pacing thresholds are adequate, old leads with such configurations may be used, but a pacemaker or ICD pulse gen-

erator with a compatible connector block should be selected. Like pacemakers, ICD generator connector blocks from some manufacturers are available in a variety of configurations and port sizes to attach directly to existing implanted leads (Fig. 20-10). An alternative (but less desirable) approach involves using an adapter to fit odd-sized lead connectors into available ports on an ICD header block. Most old pacemaker leads are unipolar; upgrading to a bipolar system should be considered but is not always necessary.

Figure 20-9. A variety of pacemaker lead connectors (top), nonconducting adapters (bottom right five), and lead caps (bottom left two). The lead connectors shown are (clockwise from bottom left) Intermedics, Inc. (Angleton, Tex.), 6-mm linear bipolar; Intermedics 6-mm unipolar; Medtronic, Inc. (Minneapolis, Minn.), 5-mm unipolar; Cordis (Miami) 3.2-mm linear bipolar, Pacesetter IS-1; Medtronic atrial 5-mm unipolar; and Medtronic IS-1. Lead caps are 5-mm cap (bottom left) and 3.2-mm low-profile or IS-1/VS-1 (bottom left, center). Upsizing sleeves are 5-mm to 6-mm (top left, center), 3.2-mm to 5-mm (top right, center), and 3.2-mm to 6-mm (bottom right, center). Unipolarizing sleeves for 6-mm bipolar leads are shown on bottom right. (Photography by Andrew C. Floyd, Hahnemann University Hospital, Philadelphia.)

Chapter 20: Approach to Generator Change

843

Figure 20-10. Examples of a variety of implantable cardioverter-defibrillator (ICD) header connector block configurations. These various styles of header configurations are designed to connect directly to different types of implanted rate-sensing and pacing leads (pace and sense, P/S) and to high-voltage lead connectors (HVA, HVB, HVX). All rate/sense lead ports are of the IS-1 configuration. This means that any rate/sense leads that are of different connector configurations must be adapted to fit into these ports. Likewise, high-energy lead ports are of the DF-1 or 3.2-mm configuration. Adapters may be required here as well to provide a direct connection into the ICD header. Many forms of adapters are available for pace and sense leads (see Fig. 20-14). A variety of adapters for high-energy leads are also available (see Fig. 20-15). (From ICD Replacement Guide. Minneapolis, Minn., Medtronic, Inc., 1997.)

Before the availability of IS-1 and VS-1 lead connector configurations,56,57 the most commonly used pacemaker leads were 3.2-mm low-profile leads (unipolar or linear bipolar), 5-mm or 6-mm unipolar and linear bipolar leads, and bifurcated bipolar systems. Pacemaker pulse generators available from some manufacturers remain compatible with each of these lead models, especially 3.2-mm and 5- or 6-mm linear bipolar and unipolar leads. Precise compatibility, however, is essential to ensure that no fluid leaks into the pulse generator connector block and that electrical continuity to proximal and distal poles remains intact. The physician must be particularly cautious to ensure that sealing rings are located either on the lead connector or in the pulse generator connector block, because not all older lead models had sealing rings placed on the lead connector itself. Review of manufacturers’ specifications of devices should provide the necessary details regarding lead and pulse generator compatibility. Because the lead model may not always be known before reoperation and because it may not be determined even with visual inspection, careful evaluation of the lead connector configuration may be required in the laboratory after the old pulse generator has been removed. Although the lead and pulse generator should have been compatible at the initial implantation, the physician cannot make this assumption without visual inspection of the type of lead connector at reoperation. Lead-generator incompatibility may be the cause of presumed lead malfunction or premature generator depletion. Bifurcated bipolar leads were implanted primarily with ventricular demand pacing systems with 5- to 6-

mm lead connectors that plugged side by side into the pulse generator connector block. Some of these leads continue to function because they are so heavily insulated. Replacement of the pulse generator entails selection of a new battery with a compatible connector block to maintain bipolar capability, use of an adapter to convert the bifurcated bipolar lead to a linear bipolar system, or conversion to unipolar through the use of only one pole of the bifurcated lead, with the other capped. The last option is particularly useful in the event that only one pole of a bifurcated bipolar lead functions well. Either of the last two options allows conversion to a dual-chamber pacing system, if desired, by placement of a new IS-1 atrial bipolar lead and adaptation of the functional pole of the old lead in a unipolar configuration to an IS-1 connector. If the bifurcated bipolar lead is adapted to linear bipolar, placement of a linear bipolar atrial lead can convert the system to bipolar DDD. Defibrillator (high voltage) and pacemaker (low voltage) standards continue to progress with the proposal of a quadripolar 3.2-mm standard. Likely this will be introduced as DF-4 for combined pacing and shocking applications and IS-4 for dedicated low-voltage applications (Figs. 20-11 and 20-12). The terminal pin and first ring will carry low-voltage impulses in both applications and for all four poles in the low-voltage application. The most anticipated application, DF-4 replaces the trifurcated ICD lead combining the function of one IS-1 and up to two DF-1 connectors (Fig. 20-13). However, the introduction of this connector creates new problems with needs for new adapters and need for care during intraoperative testing with correct placement of the alligator clips.

844

Section Three: Implantation Techniques Figure 20-11. Drawing of the proposed quadripolar low- and high-voltage lead pin configuration and header configuration. The specifications for the pins and the header are extremely precise, permitting the interchange of leads between models and manufacturers. Although both configurations are quadripolar, the low-voltage leads will not fit into the cavity of the highvoltage cavity. Although both leads are 3.2 mm in diameter, the diameter of the distal pin is larger for the low-voltage leads. (From Proposed IS-4 Standard. American Association of Medical Instrumentation presentation at Cardiostim, Nice, France, June 2004.)

Figure 20-12. Drawing of three high-voltage and two low-voltage lead configurations as proposed by the American Association of Medical Instrumentation. In all configurations, the connections flow from the distal extent of the lead proximally and connect at the pin, most distal to most proximal, and so forth. In all configurations, the pin and next electrode are reserved for low-voltage functions. Each pole is labeled either L or H, depending on its function. (From Proposed IS-4 Standard. American Association of Medical Instrumentation presentation at Cardiostim, Nice, France, June 2004.)

Adapters The following two general categories of adapters are available: (1) electrically conducting units that change the size or configuration of lead connectors to fit specific pulse generators and (2) upsizing sleeves that allow IS-1 or 3.2-mm, low-profile leads to fit into 5- or 6-mm pulse generator connector blocks while maintaining a fluid seal (Table 20-9 and Fig. 20-14). Electrically conducting adapters necessarily contain wires attached on one end to a lead pin to enter the new pulse generator and a socket on the other to accept the old lead as well as a mechanism to connect the old lead, generally a set-screw. Produced by most manufacturers, adapters are available in an array of types (see Table 20-9). The most common pacemaker

adapters downsize 5- or 6-mm leads to IS-1 unipolar or bipolar configurations or adapt 3.2-mm low-profile connectors to the IS-1 variety. Adapters are also available to convert bifurcated bipolar leads to the linear IS-1 bipolar configuration; they may be particularly useful in a patient who requires upgrading from VVI to DDD pacing when maintenance of bipolar pacing is important, as occurs when a permanent pacemaker is placed in conjunction with an implanted defibrillator (rarely done now, as it would be better to place an entirely new ICD system), or to avoid myopotential sensing. Adapters for ICD leads may be used either for the rate-sensing (pace/sense) leads or for the high-energy leads, patches, or arrays (Fig. 20-15). These small units are most helpful for adapting epicardial lead connectors found in an abdominal pocket to newer-generation

Chapter 20: Approach to Generator Change

Figure 20-13. Pictures of the proximal portions of leads intended for a dual-coil implantable cardioverter-defibrillator lead. The traditional connection with a yolk dividing into two DF-1 connectors and one IS-1 connector is replaced with a single-connection, 3.2-mm in diameter, permitting all four electrical connections and a simplified lead configuration proposed to be labeled DF4. (From Proposed IS-4 Standard. American Association of Medical Instrumentation presentation at Cardiostim, Nice, France, June 2004.)

Common Configurations for Pacemaker Adapters* TABLE 20-9.

From (Lead) Conducting adapters: 6-mm UNI 6-mm BIF 6-mm BI 5-mm BIF 5-mm BIF 5-mm UNI 3.2-mm LP BI 3.2-mm LP

845

Figure 20-14. Eight pacing lead adapters: four conducting (left) and four nonconducting (right). Conducting adapters shown are (top to bottom) 6-mm lead pin replacement; 3.2mm low-profile linear bipolar (to accept a lead connector without sealing rings) to VS-1 linear bipolar; 6-mm linear bipolar to 3.2-mm linear bipolar; and 3.2-mm low-profile linear bipolar to VS-1 linear bipolar. Nonconducting adapters shown are (top to bottom): unipolarizing sleeve for 6-mm bipolar lead; upsizing from 5 to 6 mm; upsizing unipolarizing sleeve from 3.2 to 6 mm; upsizing unipolarizing sleeve from 3.2 to 5 mm. (Photography by Andrew C. Floyd, Hahnemann University Hospital, Philadelphia.)

To (Generator) 5-mm UNI 3.2-mm LP BI 3.2-mm LP BI 3.2-mm LP BI IS-1 BI IS-1 UNI 5-mm BIF IS-1 BI

Nonconducting upsizing sleeve adapters: 3.2-mm LP BI 5-mm or 6-mm UNI (± pin extender) IS-1 UNI or BI 5-mm or 6-mm UNI 5-mm UNI 6-mm UNI LV-1 IS-1 *Adapters shown in italics are those most commonly used. BI, linear bipolar; BIF, bifurcated bipolar; LP, low profile; UNI, unipolar.

Despite the available variety of electrically conducting adapters, these units prove bulky in the pacemaker or ICD pocket. Furthermore, they provide another weak link, that is, one additional set of connections in the pacing circuit for delivery of current to the patient and for sensing, increasing the chance of malfunction, compared with direct attachment of a lead into a pulse generator connector block. Some adapter set-screws must be sealed with medical adhesive after being fastened to the lead; a poor seal can result in a short circuit in the system. Because of internal connections, not all adapters have the reliability inherent in most pacemaker or ICD leads directly connected to the generator. Tools

ICD batteries, that is, attaching nonstandard lead connectors to IS-1 pacing and DF-1 shocking ports in an ICD header block. The adapters may also be used when older transvenous leads must be attached to newer, standard-connect ICD pulse generators. One additional special use for ICD lead adapters involves connecting high-energy leads in parallel to enable them to function as a single unit with the same polarity. For example, a subcutaneous patch or array may have to be added to a system because of a high DFT. This additional hardware can be connected in parallel with a proximal high-energy coil located in the SVC. Lead connectors from both the subcutaneous lead and the proximal coil are inserted side by side into an adapter, which then attaches to a single port in the ICD header. This lead system thereby functions as a single electrically connected unit with the same polarity.

Several specially designed tools assist the operator in replacing pacemaker and ICD pulse generators and repairing leads (Table 20-10 and Fig. 20-16). Most important are wrenches to loosen set-screws in the pulse generator connector block to allow the old lead to be withdrawn. Most set-screw sizes are now standardized, but if the old generator manufacturer and model are known, a specific wrench may be required. If these are not known, it is best to have available an array of small Allen wrenches to remove the hexagonal set-screw. Some pulse generators must be removed from the lead with a small flat screwdriver. Some pulse generators are connected to the lead without set-screws through pressing of an attachment unit into place; to loosen this unit requires that a small probe be inserted into the side of the connector block to push open the locking mechanism. It is unusual to lose set-screws

846

Section Three: Implantation Techniques

Commonly Used Tools for Generator Replacement, Lead Replacement, Lead Revision, or Lead Repair TABLE 20-10.

Probes (to unlock lead connector block in some devices) Allen wrenches*: 0.035-inch (No. 0.050-inch (No. 0.062-inch (No. 0.093-inch (No.

2) 4) 6) 10)

Screwdrivers: 0.100-inch 0.200-inch Set-screws Anchoring sleeves Lead repair kit: Conductor with crimp ends Crimping tool Insulating sleeve Medical adhesive

Figure 20-15. A variety of implantable cardioverterdefibrillator (ICD) lead adapters. As indicated, several different sizes of lead connectors can be inserted side by side to be adapted to the appropriate size to fit directly into the ICD header port. Additionally, linear adapters used merely to change the size of a lead connector may also be required, as indicated. Lead extenders are used to attach rate-sensing and high-energy leads placed through the subclavian system to an ICD pulse generator inserted in the abdomen. As the size and weight of ICDs have been reduced, however, the need to place the ICD pulse generator in the anterior abdominal wall has diminished markedly. (From Multiple Options for Customized Therapy. Guidant Corporation, Cardiac Pacemakers [CPI], St. Paul, Minn., 1996.)

because they are generally held in place by a seal. It is advisable, however, to have additional set-screws available in a busy pacing laboratory. Occasionally, repair can salvage an old lead, as long as the conductor fracture or insulation break is accessible at least several centimeters from the point at which the lead enters the vascular system. Lead insulation breaks can be repaired by gluing on a polymeric silicone (Silastic) sleeve with medical adhesive. The operator can repair a conductor fracture by (1) severing the lead, (2) placing the two cut conductor ends into an electrically conducting sleeve, which is crimped down onto both ends of the lead conductor, and (3) gluing a silicone sleeve over the insulator with medical adhesive. This procedure is recommended only for a lead on which the patient is not dependent, because recurrent conductor fracture may occur. It is also not recommended for repair of high-energy defibrillation lead conductors. Repair of polyurethane leads can prove functionally inadequate because adhesive may not bond properly with the lead insulator, as it does with silicone. A more viable approach in any of these

Lead end-caps: 6.5-mm 5-mm 3.2-mm LP IS-1 *Specific torque wrenches may be available from some manufacturers; these are especially important for tightening set-screws with appropriate force. Wrench numbers are standardized for ease of identification.

situations may be to extract or cap the culprit lead and replace it entirely.

Surgical Considerations Device replacement or revision in a tertiary care institution with an active electrophysiology service and longterm follow-up may account for more than one quarter of all pacemaker or ICD procedures (Table 20-11). The timing of intervention depends on the specific indication. Most patients require reoperation for elective battery replacement or battery or lead revision, whereas 1% to 6% of patients return to the laboratory for other problems, such as pocket hematoma, pocket twitch, diaphragmatic pacing, and pocket relocation (Table 20-12). Elective Device Replacement or Revision Most reimplantation procedures are either elective or performed for repair or replacement of prior devices. Preoperative blood analysis is performed. In our laboratory, aspirin and clopidogrel are not stopped prior to the procedure. Warfarin should generally be discontinued for 3 to 5 days for procedures in which the major vessels will be instrumented, although not all laboratories follow this approach. Because the risk of hematoma development is much higher with heparin

Chapter 20: Approach to Generator Change

A

847

B Figure 20-16. Tools commonly required for reoperation. A, Three nondeformable Allen wrenches (top left) of various sizes; a pinch-on tool (Medtronic, Inc., Minneapolis, Minn.) (top right) to extend and retract the distal screw of an active-fixation lead; three wrenches (bottom left); two torque wrenches (Medtronic and CPI/Guidant, St. Paul, Minn.) (bottom right) on either side of a probe (Intermedics, Inc., Angleton, Tex.) used to unlock a pacemaker connector block. Some wrenches are deformable to avoid placing excess torque on the set-screw, whereas others are not; caution is required in use of the various systems (see text). B (left to right), Intermedics ratchet torque wrench; Intermedics flat-bladed ratchet torque screwdriver; Cordis (Miami) No. 6 Allen wrench; unlocking probe; two Allen wrenches with handles (not deformable). (Photography by Andrew C. Floyd, Hahnemann University Hospital, Philadelphia.)

Rates of Pacemaker Pulse Generator Implantations, Revisions, and Replacements* TABLE 20-11.

Procedure

Rate (%)

DDD implantation

51

VVI implantation

23

Revision or replacement

26

*The frequency of dual-chamber (DDD-R) and single-chamber (VVI-R) pulse generator implantations and battery or lead revisions or replacements over a 51/2-year period at Hahnemann University Hospital, Philadelphia.

and enoxaparin than with warfarin, we may elect to perform simple device replacements in patients with full warfarin anticoagulation. The patient fasts from midnight and receives preoperative antibiotics, most commonly being admitted on the day of the procedure. Elevated coagulation times may be corrected with fresh-frozen plasma if absolutely necessary. Procedures are routinely performed with local or regional anesthesia, supplemented by intravenous conscious sedation. For ICDs, the patient is given general anesthesia for ventricular fibrillation induction and testing of shock therapies. Most institutions use a combination of a short-acting, amnestic benzodiazepine such as midazolam together with an intravenous narcotic for analgesia. Continuous electrocardiographic monitoring, pulse oximetry, and sterile preparation and draping are standard procedures. Preoperative antibiotics are administered intravenously.

Pacemaker Reoperations: Frequency of Various Indications TABLE 20-12.

Indications

Frequency (%)

Generator indications: Battery end of service Battery insulation break DDD to VVIR conversion

53 1.1 1.1

Lead indications: Lead revision (exit block or lead injury) Diaphragmatic pacing Lead fracture Lead insulation break

17 2.2 1.1 1.1

Battery or lead indications: Battery end of service and lead replacement VVI to DDD conversion Unipolar DDD to bipolar DDD Pocket twitch Surgical indications: Pocket relocation Pocket effusion

6.8 4.4 1.1 1.1 5.6 2.2

The frequency of indications for pacemaker reoperation over a 51/2year period at Hahnemann University Hospital, Philadelphia.

General Guidelines and Techniques There is no substitute for careful surgical planning in approaching the established pacemaker pocket and gentle handling of the tissues. Perfect hemostasis, avoidance of a tight-fitting pacemaker or ICD pocket, and multilayered incision closure are the basic principles that help prevent future difficulties. These

848

Section Three: Implantation Techniques

principles are similar to those required at initial implantation (see Table 20-2). To avoid induction of ventricular fibrillation, development of fibrosis at the lead tip, and damage to the generator itself, electrocautery must not be used directly over an implanted pulse generator with unipolar leads. This issue has become much less of a problem with bipolar leads. Electrocautery also must not be used during battery changes with the generator disconnected when pacemaker leads are grounded to the patient for testing, because current may be shunted directly to the heart. Hemostasis at reoperation can usually be secured with electrocautery or direct ligature. Use of surgical absorbable cellulose or topical thrombin assists in treating persistently oozy pockets. Clinical judgment should be used in the application of various technical approaches (Fig. 20-17).

Local anesthesia is administered most commonly as 1% lidocaine (Xylocaine) infiltrated into the scar line from the previous procedure; additional lidocaine may be given under direct vision once the capsule of the pocket has been defined. The surgical incision is placed directly over the previous incision. The skin and subcutaneous tissues are opened with sharp dissection, which is required to penetrate the tough scar tissue and dermal layer. Deeper dissection with Metzenbaum scissors is carried out to delineate the pacemaker capsule. Once the pocket is reached, the fibrous capsule is sharply incised with a blade to make a small opening, which is then extended with scissors under direct visualization of the implanted pulse generator and leads (Fig. 20-18). Alternatively, electrocautery may be used to open the pacemaker capsule. The capsule must be opened far enough to allow extraction of the pulse generator and lead connector assembly without undue force. The posterior capsule

may have to be carefully dissected away from the leads to allow mobility. Access to leads and generator may be facilitated through the use of self-retaining retractors. Extreme care is required throughout the procedure to preserve the integrity of the leads and lead connectors; they must not be punctured with anesthetic needles or cut with blades or scissors. If electrocautery is used to remove tissue from the leads in dissecting them from scar tissue in the posterior capsule, the probe must keep moving over the lead so as to not overheat the lead insulation and thereby damage it. Leads with very thin insulation, including most coronary sinus electrodes used with biventricular systems, are more prone to heat damage from electrocautery. Once the generator is delivered out of the pocket, the leads are disconnected and analyzed, as described earlier. Leads from pacemaker-dependent patients need to be expeditiously reconnected to an external pacemaker (Fig. 20-19). Unipolar pacemaker leads require direct grounding to subcutaneous tissue; the active part of the unipolar generator must remain in contact with the patient before the lead is disconnected. Grounding can best be accomplished through a large surface area ground electrode placed directly into the open pocket. Making contact with this electrode onto the active surface of a unipolar pulse generator allows the generator to be removed safely from the pocket before the lead is disconnected, even in a pacemaker-dependent patient. After being secured to temporary pacing cables, leads can be completely freed of adhesions up to their entry point into the subclavian vein, if necessary, to examine lead integrity or for extraction. We use lowenergy electrocautery sparingly to dissect the leads free of adhesions because the scar tissue could be especially tough and adherent to lead structures. If lead replacement or repair is not necessary, and if the function of previously implanted leads is adequate, dissection of the complete course of each lead may not be necessary.

Figure 20-17. Preparation for surgery. A typical array of instruments required for reoperation. Foreground, Retractors, scissors, forceps, scalpel, syringes filled with lidocaine anesthetic, sterile saline solution, sponges, and a variety of wrenches. Background, Hemostats and absorbable and nonabsorbable sutures. (Photography by Andrew C. Floyd, Hahnemann University Hospital, Philadelphia.)

Figure 20-18. Opening the capsule of the long-term generator pocket to expose the pulse generator. (Photography by Andrew C. Floyd, Hahnemann University Hospital, Philadelphia.)

Specific Techniques

Chapter 20: Approach to Generator Change

Figure 20-19. Disconnecting the previously implanted lead. The active surface of the unipolar pacemaker pulse generator remains in contact with the open pocket to maintain pacing output until the lead is withdrawn from the connector block. The surgeon has adequately mobilized the proximal portion of the 5-mm lead to allow rapid connection to an external pacing cable, one end of which has already been securely grounded to the patient. A ligature previously placed around the lead connector block entry post has been removed. After the external pacemaker has been activated, the surgeon will loosen the set-screw (here covered by a seal on the top right side of the connector block) and withdraw the lead from the pulse generator, immediately connecting it to the negative pole of the external pacing cable. (Photography by Andrew C. Floyd, Hahnemann University Hospital, Philadelphia.)

The physician, however, must ascertain whether the lead connector mobility is sufficient to attach it to a new pulse generator without tension. Inadequate implanted lead pacing or sensing thresholds may require placement of one or two new leads. Upgrade from a single-chamber to a DDD pacemaker system or to a biventricular system may require placement of an additional lead. If a previously implanted lead is extracted through a dilating sheath, a guidewire can usually be inserted into the vascular system through the extraction sheath to maintain a conduit for replacement. In other cases, repeated axillary or subclavian vein puncture, brachiocephalic cutdown, or an internal jugular approach provides an alternative means of inserting a new lead. If new leads are placed through the same subclavian system by direct puncture, the operator must be careful to avoid lead damage. After the old pulse generator has been detached from leads and the lead integrity and functional status have been ascertained, a new pulse generator can be attached. The principles of generator-lead compatibility must be maintained (Fig. 20-20). Redundant lead coils are placed posterior to the pulse generator, and the pocket is closed with three layers of absorbable suture—two subcutaneous and one subcuticular. ICD leads may be tested for defibrillation threshold before, or concomitant with, final pocket closure. We also open the capsule inside the device pocket, usually in a medial and inferior direction, for two reasons. First, a new device, even if an identical model to the one

849

Figure 20-20. Adapting and inserting rate-sensing and highvoltage defibrillator leads into the implantable cardioverterdefibrillator (ICD) connector block. The incision is made in the left upper quadrant of the abdomen over the scar line from the previous device implant. The old ICD generator has been removed, and the leads have been freed up by gentle electrocautery to allow them to be easily manipulated and to be placed into the header of the new pulse generator without undue tension or bending. Basic pacing and sensing thresholds have been tested; this testing also includes delivery of low-energy (5-V) pacing impulses through the high-energy leads to measure lead impedance and thus ensure structural integrity of the high-voltage conductors. The new ICD device to be implanted has been placed on the surgical field, and its connection to the lead system has begun. A plug is first inserted into the upper right port because the port will not be used. The set-screw has already been tightened onto this plug, and a required cap has been placed over the set-screw. The surgeon is inserting a unipolar lead connector from the ratesensing portion of the defibrillator lead into the ICD connector block. This lead head is connected to the distal tip of a right ventricular endocardial integrated lead system (497-05 lead, Intermedics, Inc., Angleton, Tex.); it is inserted into the appropriate sensing port of the ICD. Sealing O-rings are evident on the ICD header and are used in this device to prevent fluid leakage around a cap placed over the set-screw. Also evident just within the forceps is a thickened area of the lead that represents a sleeve glued in place with medical adhesive to cover a minor break in the outer insulator of the lead (i.e., an insulator repair). A plug has been inserted into an unused port of the ICD connector block; the sealing cap has already been placed over the set-screw for this plug. The dangling high-voltage lead connector has been inserted through an upsizing sleeve before being placed in the ICD connector block. The original lead connector is a 4-mm configuration; the sleeve passes over the insulator on the head to increase its diameter to fit into a 6-mm high-voltage port in the ICD connector block. The remaining lead connector evident on the patient attaches to a subcutaneous high-voltage patch used in the defibrillation circuit. The tip of this lead is also adapted through the use of a similar sleeve to fit a 6-mm port. (Photography courtesy of Todd Forkin, Hahnemann University Hospital, Philadelphia.)

removed, will never fit perfectly in the original pocket, and second, doing so allows for absorption of fluid and fresh blood flow, which are not possible if the relatively avascular capsule is left intact. For generator replacement, the previous pacemaker or ICD pocket location is used most commonly, usually opposite the patient’s dominant side. Because most pacemakers and most ICDs are placed inferior to the clavicle in a subcutaneous location anterior to the

850

Section Three: Implantation Techniques

pectoralis fascia, the location of replacement devices is similar. Various modifications suit individual patient needs. In very thin patients, subpectoral or axillary locations may be required. One can access the subpectoral plane by locating the junction between the sternal and clavicular heads of the pectoralis major muscle and making entry at that point, taking care to avoid damage to penetrating neurovascular bundles. Alternatively, the muscle fibers of the pectoralis major can be teased apart longitudinally to allow entry to the subpectoral plane, which can also be accessed through the deltopectoral groove. Axillary subcutaneous placement of a pacing device is generally avoided because of the possibility of lateral migration of the device, which can be uncomfortable for the patient and can, especially with larger ICDs, lead to erosion. When required, however, the axillary location can be entered through direct extension from a subclavian pocket or through a separate axillary incision. The device may be placed in a subpectoral location at that site for more stability. The abdominal wall, subcostal, and intrathoracic positions represent other alternatives for a replacement pulse generator. Nevertheless, a subcutaneous prepectoral approach is appropriate in most patients for both pacemaker and ICD reimplantations. For replacement of an abdominal ICD, dissection can be difficult if the device is located behind the rectus abdominis muscle. We usually begin over the area where the device is most easily palpated, carefully spreading the muscle apart in the area of prior scarring. The scar tissue is carefully divided to expose the pulse generator, which is then handled in the usual manner. The physician must be cautious not to exert undue force to remove the generator in order to avoid rupture into the peritoneal cavity. Electrocautery is avoided over the muscle if the patient is not under general anesthesia. Approach to the Eroding Device Although relatively uncommon, chronic erosion through the skin by the pulse generator or lead can occur.1,10,77 Incipient erosion manifests as localized erythema in an area of thinned skin that is adherent to the underlying device. The area gradually becomes necrotic and may drain serosanguineous fluid. Outright erosion and drainage necessarily imply that the pacemaker pocket is no longer sterile1,98; in such instances, the system (generator and leads) should be removed if possible.5,8 Occasionally, the pocket heals with removal of the pulse generator alone, but only when skin integrity has not been breached. After removal of the pulse generator and leads, eroded pockets are packed for secondary closure or they can be fully débrided and closed primarily, leaving a drain in place for 2 to 3 days. Administration of intravenous antibiotics proceeds for 1 to 6 weeks (the longer duration if bacteremia has occurred). A new device should be implanted on the contralateral side only after all signs of infection have resolved at the old pacemaker site, if the patient has not experienced recurrent fever, and if there is no elevation of the white blood cell count. Five to 7 days of

intravenous antibiotic administration appear sufficient before device replacement, as long as bacteremia has resolved and there are no large intracardiac vegetations. Replacement of the pulse generator on the original side is not recommended but may be possible if complete erosion did not occur, the pocket could be closed after primary removal of the generator, and there are no signs of active infection after antibiotics have been discontinued. Alternative approaches are discussed in Chapter 21. Antibiotic Prophylaxis Compared with pacemaker generator change, the overall procedure time for ICD generator change may be lengthier because of the larger size of the device and greater extent of dissection, the frequent need for upgrade, and the added time for DFT testing during the procedure. The surgical wound may, therefore, be open for a longer time, although closure may be started before all testing is complete. Furthermore, the generator change could involve a larger incision in an abdominal site. We recommend routine use of broadspectrum antibiotics for antimicrobial prophylaxis during all procedures involving generator or lead revision, especially because these procedures use combinations of previously implanted hardware with new equipment. Whether and how long to use postoperative antibiotic prophylaxis has been debated.98 Intervention for Acute Problems Indications for acute intervention include primary complications of pacemaker or ICD implantation (e.g., pocket hematoma, infection,1-3 or cardiac perforation73-77) as well as other, less crucial indications, such as iatrogenic lead damage and lead dislodgement. Pocket hematomas occur most commonly in patients receiving anticoagulants, especially heparin and enoxaparin, and in patients with platelet dysfunction, which is common in those undergoing long-term hemodialysis. The range in hematoma size varies from a contained, small amount of fluctuance and ecchymosis to a large hematoma that may drain through the skin. A minor hematoma requires only observation, whereas a breach of skin integrity after operation may require evacuation of the hematoma or, if it has become secondarily infected, complete removal of the generator and lead system. If the patient remains pacemaker dependent, a temporary wire must be placed when the original system is removed; after an appropriate course of intravenous antibiotic therapy, a new pacemaker can be placed on the contralateral side. Prolonged antibiotic therapy may be required in some cases. Antibiotic therapy alone and conservative surgical approaches other than complete removal of an eroded or infected generator and leads prove unsatisfactory.5-8 Immediate reoperation may also be required for cardiac perforation.73-77,79 Perforation is suggested by curvature of the lead beyond the confines of the right ventricular apex, an abrupt rise in pacing threshold or

Chapter 20: Approach to Generator Change

deterioration of sensing, precordial pain that increases with inspiration, hypotension, and hemodynamic collapse. Although most perforations close spontaneously, development of a large pericardial bleed or tamponade requires immediate intervention.79 Pericardiocentesis usually suffices, but occasionally, a subxiphoid approach to pericardial drainage is necessary. Proper lead selection to match the patient’s anatomy and gentle technique are vital to avoiding acute perforation. Subcostal placement of epicardial screw-in leads has been associated with a high incidence of serious or fatal ventricular perforations; chronic perforation by endocardial leads is distinctly rare.24 Early surgical exploration is indicated to confirm the diagnosis of iatrogenic lead insulation damage. This is an uncommon complication that manifests early in the form of pocket twitch,81 failure to capture, or failure to sense, with associated low measured lead impedance.51 Chronic lead damage has been associated with excessively tight anchoring sutures, especially if they are placed around the lead and not the anchoring sleeve. The damaged lead, whether passive or active fixation, should be removed and replaced, if possible; alternatively, it may be repaired, although repair is difficult if damage has occurred near the venous insertion site. Lead dislodgement occurs most commonly during the first 24 to 48 hours after system implantation.26,27 It can occur later, however, as a result of a loose anchoring sleeve, incomplete fixation of the distal lead tip, excessive diaphragmatic motion, or patient manipulation of the device (i.e., twiddler’s syndrome).49 Before the development of leads with active fixation or a finlike mechanism at the distal tip, the incidence of lead dislodgement ranged as high as 5% to 18%. With careful technique and selection among a variety of active-fixation and passive-fixation leads, the incidence should range no higher than 1% to 2%.30,31,99 Most spontaneous dislodgements occur with atrial passive-fixation leads. The diagnosis may be facilitated by chest radiography or fluoroscopy; pacing analysis reveals an increased pacing threshold with, usually, normal lead impedance. The operator has the option of repositioning or replacing the lead. If a distinct cause cannot be identified, placement of an active-fixation lead may avoid a second dislodgement. To prevent recurrent lead dislodgement in twiddler’s syndrome, leads must be sutured to prepectoral fascia or firm pacemaker pocket fibrous tissue with nonabsorbable sutures around anchoring sleeves at more than two points; the pacemaker connector block may also be anchored to the pectoralis fascia. A polyester (Dacron) pouch has been used in the past to improve device stability in this syndrome and in patients with very loose subcutaneous tissue,100 but it is rarely required today. Interval or Unscheduled Intervention In the course of pacemaker or ICD follow-up and before the patient requires elective replacement, interval intervention may be needed to correct other complications. These include pulse generator migration,15 lead dislodgement,16,23,26,27,31,49 high pacing thresholds,23,50

851

pocket twitch or diaphragmatic pacing,24,81 lead insulation break or lead fracture,11,13,26 premature generator failure (which could be due to intrinsic component failure or the result of externally induced failure, such as that caused by electrocautery, irradiation, or cardioversion),19,20 and the need for upgrade of the system.73 The pacemaker clinic may prove particularly useful in recognizing early surgical or functional problems.37 Evaluation and technique follow the principles described earlier.

Summary Successful pacemaker or ICD replacement is the result of accurate preoperative evaluation and careful surgical intervention. The preoperative status of the pulse generator battery and lead pacing and sensing function as well as the appropriateness of both to future pacing or defibrillating systems need to be determined to enable surgical planning. There should be no surgical surprises, and all the tools, adapters, leads, and generators should be ready for the intervention. The goal should be to avoid reoperation for as long as possible with careful initial implantation and programming. When properly planned, surgery is likely to proceed smoothly. REFERENCES 1. Bonchek LI: New methods in the management of extruded and infected cardiac pacemakers. Ann Surg 176:686, 1972. 2. Corman LC, Levison ME: Sustained bacteremia and transvenous cardiac pacemakers. JAMA 233:264, 1975. 3. Morgan G, Ginks W, Siddons H, Leatham A: Septicemia in patients with an endocardial pacemaker. Am J Cardiol 44:221, 1979. 4. Wohl B, Peters RW, Carliner N, et al: Late unheralded pacemaker pocket infection due to Staphylococcus epidermidis: A new clinical entity. PACE 5:190, 1982. 5. Prager PI, Kay RH, Somberg E, et al: Pacemaker remnantsanother source of infections. PACE 7:763, 1984. 6. Mansour KA, Kauten JR, Hatcher CR Jr: Management of the infected pacemaker: Explantation, sterilization, and reimplantation. Ann Thorac Surg 40:617, 1985. 7. Buch J, Mortensen SA: Late infections of pacemaker units due to silicone rubber insulation boots. PACE 8:494, 1985. 8. Ruiter JH, Degener JE, Van Mechelen R, Bos R: Late purulent pacemaker pocket infection caused by Staphylococcus epidermidis: Serious complications of in situ management. PACE 8:903, 1985. 9. Vilacosta I, Zamorano J, Camino A, et al: Infected transvenous permanent pacemakers: Role of transesophageal echocardiography. Am Heart J 125:904, 1993. 10. Garcia-Rinaldi R, Revuelta JM, Bonnington L, Soltero-Harrington L: The exposed cardiac pacemaker: Treatment by subfacial pocket relocation. J Thorac Cardiovasc Surg 89:136, 1985. 11. Kronzon I, Mehta SS: Broken pacemaker wire in multiple trauma: A case report. J Trauma 14:82, 1974. 12. Tegtmeyer CJ, Bezirdjian DR, Irani FA, Landis JD: Cardiac pacemaker failure: A complication of trauma. South Med J 74:378, 1981. 13. Grieco JG, Scanlon PJ, Pifarre R: Pacing lead fracture after a deceleration injury. Ann Thorac Surg 47:453, 1989.

852

Section Three: Implantation Techniques

14. Wallace WA, Abelmann WH, Norman JC: Runaway demand pacemaker: Report, in vitro reproduction, and review. Ann Thorac Surg 9:209, 1970. 15. Bello A, Yepez CG, Barcelo JE: Retroperitoneal migration of a pacemaker generator: An unusual complication. J Cardiovasc Surg 15:256, 1974. 16. Kim GE, Haveson S, Imparato AM: Late displacement of cardiac pacemaker electrode due to heavyweight pulse generator. JAMA 228:74, 1974. 17. Austin SM, Kim CS, Solis A: Electrical alternans of pacemaker spike amplitude: An unusual manifestation of permanent pacemaker generator malfunction. PACE 4:313, 1981. 18. Campo A, Nowak R, Magilligan D, Tomlanovich M: Runaway pacemaker. Ann Emerg Med 12:32, 1983. 19. Venselaar JL, Van Kerkeorle HL, Vet AJ: Radiation damage to pacemakers from radiotherapy. PACE 19:538, 1987. 20. Lewinn AA, Serago CF, Schwade JG, et al: Radiation-induced failures of complementary metal oxide semiconductor containing pacemakers: A potentially lethal complication. Int J Radiol Oncol Biol Phys 19:1967, 1984. 21. Halperin JL, Camunas JL, Stern EH, et al: Myopotential interference with DDD pacemakers: Endocardial electrographic telemetry in the diagnosis of pacemaker-related arrhythmias. Am J Cardiol 54:97, 1984. 22. den Dulk K, Lindemans FW, Brugada P, et al: Pacemaker syndrome with AAI rate-variable pacing: Importance of atrioventricular conduction properties, medication, and pacemaker programmability. PACE 11:1226, 1988. 23. Aris A, Shebairo RA, Lepley D Jr: Increasing myocardial thresholds to pacing after cardiac surgery. Surg Forum 24:167, 1973. 24. Gaidula JJ, Barold SS: Elimination of diaphragmatic contractions from chronic pacing catheter perforation of the heart by conversion to a unipolar system. Chest 66:86, 1974. 25. Contini C, Papi L, Pesola A, et al: Tissue reaction to intracavitary electrodes: Effect on duration and efficiency of unipolar pacing in patients with A-V block. J Cardiovasc Surg 14:282, 1973. 26. Holmes DR Jr, Nissen RG, Maloney JD, et al: Transvenous tined electrode systems: An approach to acute dislodgement. May Clin Proc 54:219, 1979. 27. Snow N: Elimination of lead dislodgment by the use of tined transvenous electrodes. PACE 5:571, 1982. 28. Alt E, Volker R, Blomer H: Lead fracture in pacemaker patients. Thorac Cardiovasc Surg 35:101, 1987. 29. Woscoboinik JR, Maloney JD, Helguera ME, et al: Pacing lead survival: Performance of different models. PACE 15:1991, 1992. 30. Morse D, Yankaskas M, Johnson B, et al: Transvenous pacemaker insertion with a zero dislodgment rate. PACE 6:283, 1983. 31. Hakki AH, Horowitz LN, Reiser J, Mundth ED: Improved pacemaker fixation and performance using a modified finned porous surfaced tip lead. Int Surg 69:291, 1984. 32. Mirowski M, Reid PR, Mower MM, et al: Termination of malignant ventricular arrhythmias with an implanted automatic defibrillator in human beings. N Engl J Med 303:322, 1980. 33. Mond H, Twentyman R, Smith D, Sloman G: The pacemaker clinic. Cardiology 57:262, 1972. 34. Starr A, Dobbs J, Dabolt J, Pierie W: Ventricular tracking pacemaker and teletransmitter follow-up system. Am J Cardiol 32:956, 1973. 35. Janosik DL, Redd RM, Buckingham TA, et al: Utility of ambulatory electrocardiography in detecting pacemaker dysfunction in the early postimplantation period. Am J Cardiol 60:1030, 1987. 36. Mugica J, Henry L, Rollet M, et al: The clinical utility of pacemaker follow-up visits. PACE 9:1249, 1986. 37. Joseph GK, Wilkoff BL, Dresing T, et al: Remote interrogation and monitoring of implantable cardioverter defibrillators. J Interv Card Electrophysiol 11:161, 2004.

38. Kertes P, Mond H, Sloman G, et al: Comparison of lead complications with polyurethane tined, silicone rubber tined, and wedge tip leads: Clinical experience with 822 ventricular endocardial leads. PACE 6:957, 1983. 39. van Gelder LM, El Gamal MI: False inhibition of an atrial demand pacemaker caused by an insulation defect in a polyurethane lead. PACE 6:834, 1983. 40. Sanford CF: Self-inhibition of an AV sequential demand (DVI) pulse generator due to polyurethane lead insulation disruption. PACE 6:840, 1983. 41. Timmis GC, Westveer DC, Martin R, Gordon S: The significance of surface changes on explanted polyurethane pacemaker leads. PACE 6:845, 1983. 42. Chawla AS, Blais P, Hinberg I, Johnson D: Degradation of explanted polyurethane cardiac pacing leads and of polyurethane. Biomater Artif Cells Artif Organs 16:785, 1988. 43. Van Beek GJ, den Dulk K, Lindemans FW, Wellens HJ: Detection of insulation failure by gradual reduction in noninvasively measured electrogram amplitudes. PACE 9:772, 1986. 44. Stokes KB, Church T: Ten-year experience with implanted polyurethane lead insulation. PACE 9:1160, 1986. 45. Phillips R, Frey M, Martin RO: Long-term performance of polyurethane pacing leads: Mechanisms of design-related failures. PACE 9:1166, 1986. 46. Barold SS, Gaidula JJ: Demand pacemaker arrhythmias from intermittent internal short circuit in bipolar electrode. Chest 63:165, 1973. 47. Adler SC, Foster AJ, Sanders RS, Wuu E: Thin bipolar leads: A solution to problems with coaxial bipolar designs. PACE 15:1986, 1992. 48. Barold SS, Scovil J, Ong LS, Heinle RA: Periodic pacemaker spike attenuation with preservation of capture: An unusual electrocardiographic manifestation of partial pacing electrode fracture. PACE 1:375, 1978. 49. Bayliss, CE, Beanlands DS, Baird RJ: The pacemaker-twiddler’s syndrome: A new complication of implantable transvenous pacemakers. Can Med Assoc J 99:371, 1968. 50. Starr DS, Lawrie GM, Morris GC Jr: Acute and chronic stimulation thresholds of intramyocardial screw-in pacemaker electrodes. Ann Thorac Surg 31:334, 1981. 51. Ferek B, Pasini M, Pustisek S, et al: Noninvasive detection of insulation break. PACE 7:1063, 1984. 52. Sandler MJ, Kutalek SP: Inappropriate discharge by an implantable cardioverter defibrillator: Recognition of myopotential sensing using telemetered intracardiac electrograms. PACE 17:665, 1994. 53. Korte T, Jung W, Spehl S, et al: Incidence of ICD lead related complications during long-term follow-up: Comparison of epicardial and endocardial electrode systems. PACE 18:2053, 1995. 54. Schwartzman D, Callans DJ, Gottlieb CD, et al: Early postoperative rise in defibrillation threshold in patients with nonthoracotodefibrillation lead systems: Attenuation with biphasic shock waveforms. J Cardiovasc Electrophysiol 7:483, 1996. 55. Goyal R, Harvey M, Horwood L, et al: Incidence of lead system malfunction detected during implantable defibrillator generator replacement. PACE 19:1143, 1996. 56. Calfee RV, Saulson SH: A voluntary standard for 3.2-mm unipolar and bipolar pacemaker leads and connectors. PACE 9:1181, 1986. 57. Doring J, Flink R: The impact of pending technologies on a universal connector standard. PACE 9:1186, 1986. 58. Tyers GF, Sanders R, Mills P, Clark J: Analysis of setscrew and sidelock connector reliability. PACE 15:2000, 1992. 59. Hayes DL, Wang PJ, Reynolds DW, et al: Interference with cardiac pacemakers by cellular telephones. N Engl J Med 336:1473, 1997. 60. Fetter JG, Ivans V, Benditt DG, Collins J: Digital cellular telephone interaction with implantable cardioverter-defibrillators. J Am Coll Cardiol 31:623, 1998.

Chapter 20: Approach to Generator Change 61. Torresani J, Ebagosti A, Allard-Latour G: Pacemaker syndrome with DDD pacing. PACE 7:1183, 1984. 62. Cunningham TM: Pacemaker syndrome due to retrograde conduction in DDI pacemaker. Am Heart J 115:478, 1988. 63. Schwartzman D, Nallamothu N, Callans DJ, et al: Postoperative lead-related complications in patients with nonthoracotomy defibrillation lead system. J Am Coll Cardiol 26:776, 1995. 64. Lawton JS, Wood MA, Gilligan DM, et al: Implantable cardioverter defibrillator leads: The dark side. PACE 19:1273, 1996. 65. Swartz JF, Fletcher RD, Karasik PE: Optimization of biphasic waveforms for human nonthoracotomy defibrillation. Circulation 88:2646, 1993. 66. Neuzner J, Pitschner HF, Huth C, et al: Effect of biphasic waveform pulse on endocardial defibrillation efficacy in humans. PACE 17:207, 1994. 67. Natale S, Sra J, Axtell K, et al: Preliminary experience with a hybrid nonthoracotomy defibrillating system that includes a biphasic device: Comparison with a standard monophasic device using the same lead system. J Am Coll Cardiol 24:406, 1994. 68. Block M, Hammel D, Bocker D, et al: A prospective randomized cross-over comparison of mono- and biphasic defibrillation using nonthoracotomy lead configuration in humans. J Cardiovasc Electrophysiol 5:581, 1994. 69. Gold MR, Foster AH, Shorofsky SR: Lead system optimization for transvenous defibrillation. Am J Cardiol 80:1163, 1997. 70. Bardy GH, Dolack GL, Kudenchuck PJ, et al: Prospective comparison in humans of a unipolar defibrillation system with that using an additional superior vena cava electrode. Circulation 89:1090, 1994. 71. Almassi GH, Olinger GN, Wetherbee JN, et al: Long-term complications of implantable cardioverter defibrillator lead system. Ann Thorac Surg 55:888, 1993. 72. Brode SE, Schwartzman D, Callans DJ, et al: ICD-antiarrhythmic drug and ICD-pacemaker interactions. J Cardiovasc Electrophysiol 8:830, 1997. 73. Baker CM, Christopher TJ, Smith PF, et al: Addition of a left ventricular lead to conventional pacing systems in patients with congestive heart failure: Feasibility, safety, and early results in 60 consecutive patients. PACE 25:166, 2002. 74. Peters RW, Scheinman MM, Raskin S, Thomas AN: Unusual complications of epicardial pacemaker: Recurrent pericarditis, cardiac tamponade and pericardial constriction. Am J Cardiol 45:1088, 1980. 75. Phibbs B, Marriott HJ: Complications of permanent transvenous pacing. N Engl J Med 312:1428, 1985. 76. Villaneuva FS, Heinsiner JA, Burkman MH, et al: Echocardiographic detection of perforation of the cardiac ventricular septum by a permanent pacemaker lead. Am J Cardiol 59:370, 1987. 77. Hill PE: Complications of permanent transvenous cardiac pacing: A 14-year review of all transvenous pacemakers inserted at one community hospital. PACE 10:564, 1987. 78. Pizzarelli G, Dernevik L: Inadvertent transarterial pacemaker insertion: An unusual complication. PACE 10:951, 1987. 79. Sandler MA, Wertheimer JH, Kotler MN: Pericardial tamponade associated with pacemaker catheter manipulation. PACE 12: 1085, 1989. 80. Mueller X, Sadeghi H, Kappenberger L: Complications after single- versus dual-chamber pacemaker implantation. PACE 13:711, 1990.

853

81. Ekbom K, Nilsson BY, Edhag O, Olin C: Rhythmic shoulder girdle muscle contractions as a complication in pacemaker treatment. Chest 66:599, 1974. 82. Chun PK: Characteristics of commonly utilized permanent endocardial and epicardial pacemaker electrode systems: Method of radiologic identification. Am Heart J 102:404, 1981. 83. Angello DA: Principles of electrical testing for analysis of ventricular endocardial pacing leads. Prog Cardiovasc Dis 27:57, 1984. 84. Kemler RL: A simple method for exposing the external jugular vein for placement of a permanent transvenous pacing catheter electrode. Ann Thorac Surg 26:266, 1978. 85. Sethi GK, Bhayana JN, Scott SM: Innominate venous thrombosis: A rare complication of transvenous pacemaker electrodes. Am Heart J 87:770, 1974. 86. Fritz T, Richeson JF, Fitzpatrick P, Wilson G: Venous obstruction: A potential complication of transvenous pacemaker electrodes. Chest 83:534, 1983. 87. Sharma S, Kaul U, Rajani M: Digital subtraction venography for assessment of deep venous thrombosis in the arms following pacemaker implantation. Int J Cardiol 23:135, 1989. 88. Antonelli D, Turgeman Y, Kaveh Z, et al: Short-term thrombosis after transvenous permanent pacemaker insertion. PACE 12:280, 1989. 89. Spittell PC, Vlietstra RE, Hayes DL, Higano ST: Venous obstruction due to permanent transvenous pacemaker electrodes: Treatment with percutaneous transluminal balloon venoplasty. PACE 13:271, 1990. 90. Wertheimer M, Hughes RK, Castle CH: Superior vena cava syndrome: Complication of permanent transvenous endocardial cardiac pacing. JAMA 224:1172, 1973. 91. Toumbouras M, Spanos P, Konstantaras C, Lazarides DP: Inferior vena cava thrombosis due to migration of retained functionless pacemaker electrode. Chest 82:785, 1982. 92. Blackburn T, Dunn M: Pacemaker-induced superior vena cava syndrome: Consideration of management. Am Heart J 116:893, 1988. 93. Goudevenos JA, Reid PG, Adams PC, et al: Pacemaker-induced superior vena cava syndrome: Report of four cases and review of the literature. PACE 12:1890, 1989. 94. Mazzetti H, Dussaut A, Tentori C, et al: Superior vena cava occlusion and/or syndrome related to pacemaker leads. Am Heart J 125:831, 1993. 95. Pace JN, Maquilan M, Hessen SE, et al: Extraction and replacement of permanent pacemaker leads through occluded vessels: Use of extraction sheaths as conduits-balloon venoplasty as an adjunct. J Interv Cardiol Electrophysiol 1:271, 1997. 96. Chaithiraphan S, Goldberg E, Wolff W, et al: Massive thrombosis of the coronary sinus as an unusual complication of transvenous pacemaker insertion in a patient with persistent left, and no right superior vena cava. J Am Geriatr Soc 22:79, 1974. 97. Kennelly BM: Permanent pacemaker implantation in the absence of a right superior vena cava: A case report. S Afr Med J 55:1043, 1979. 98. Wade JS, Cobbs CG: Infections in cardiac pacemakers. Curr Clin Topics Infect Dis 9:44, 1988. 99. Boake WC, Kroncke GM: Pacemaker Complications: Cardiac Pacing. Philadelphia, Lea & Febiger, 1979. 100. Parsonnet V: A stretch fabric pouch for implanted pacemakers. Arch Surg 105:654, 1972.

Chapter 21

Managing Device-Related Complications and Transvenous Lead Extraction CHARLES L. BYRD

M anagement of device-related complications is an established and essential branch of electrophysiology. The prerequisites include proficiency in device implantation; knowledge of the etiology, pathophysiology, and treatment of each type of complication; and acquisition of the procedural skills necessary to administer the treatment. In this chapter, emphasis is placed on both procedure-related technical skills and management of individual complications. A divergence is made from the classic historical description of topics in order to emphasize the current clinical relevance of the various types of complications. The two essential skills are lead extraction and lead implantation. In the past, extraction procedures were on the leading edge of technical development, evolving rapidly, and potentially dangerous; patient management revolved around the lead extraction procedure. Today, lead extraction plays the same essential role in managing device-related complications as does lead implantation for insertion of a device. Mastery of extraction and implantation skills is a necessary prerequisite to managing device-related complications. Complications A device-related complication is a potential or actual morbid process or event associated with an implanted

device. This definition of a device-related complication mirrors the definition for any medical complication, except for the identification of a “potential process or event.” Potential is used because, in some situations, the probability of occurrence of a morbid process or event elevates it to the level of a complication. A morbid process pertains to a disease or clinical event. By this definition, a device-related event (DRE) may not be a complication. For example, a device component failure does not qualify as a complication unless it causes a potential or actual clinical event. Classifications of Device-Related Complications A uniform classification of device-related complications does not exist. Device-related complications encompass a diverse collection of DREs, clinically ranging from trivial to life-threatening, and some may not require therapeutic intervention. DREs involve the device itself, the biophysical interface between the device and tissue, and communication between the device and the heart and blood vessels. For a complication to exist, the DRE must cause a local clinical event involving heart rhythm, mechanically induced hemodynamic events, and/or electric field effects. The local clinical event may cause systemic clinical events, which involve organ perfusion and compensatory mechanics. The name given to a complication should reflect the associated DRE, the local clinical event, and the 855

856

Section Three: Implantation Techniques

systemic clinical event. The challenge is to construct a logical classification system that meets this objective. It is not surprising that current practices label a device-related complication with a superficial descriptive name associated with an event, such as loss of capture, loss of sensing, venous occlusion, phrenic nerve stimulation, or infection. Although some of these labels are unique, relate to symptoms, reflect the magnitude of the clinical sequelae, and define the cause, most do not. For example, the descriptive devicerelated complication name “loss of pacing” signifies that there is no evidence the device is pacing the heart. It does not indicate whether pacing spikes are present or why the loss of capture occurred (e.g., conductor coil fracture, battery failure, component failure in the pulse generator, exit block, low output voltage, oversensing). The label is obviously not unique. “Loss of capture” implies that a pacing spike was present and narrows the list to conductor coil fracture, exit block, and low output voltage. To further define the problem, additional test data, such as impedance and stimulation threshold data, are needed, along with more descriptive labels to reflect the testing. In addition, more labels are required to describe the magnitude of the problem: signs and symptoms related to the heart, central nervous system (CNS), pulmonary system, renal system, and so on. Assume all of these labels qualify as a complication. Which label is to be listed as the complication? This is not a logical or uniform classification system. A pictorial classification is descriptive and complete, and its meaning is intuitively obvious. Pictures can easily demonstrate any complicated scenario. Unfortunately, unless the picture can be standardized, incorporated into a classification scheme, or communicated in a simple fashion, its integration into a logical uniform classification system is too complicated. An abstract classification tries to integrate the complexities depicted in a picture into a simple, uniform, and logical classification using words and/or symbols. The reverse is also true: an abstract description can also be pre-

sented as a picture. A classification of a device-related complication must include all of the DREs (Fig. 21-1). Device-Related Events DREs are separated into problems with the physical device components (e.g., pulse generator, leads, adaptors), interaction of the device with the biophysical interface, and communication of the device with tissues (cardiac, muscular, neural, and vascular). An abstract classification based on components, biophysical interface, and communication physically defines all DREs, and these events can be presented as pictures. All complications are DREs, but not all DREs are complications. A DRE must have actual or potential clinical sequelae to qualify as a complication. DREs must be an integral part of any classification system. Component events are issues associated with the device hardware. Examples of component events are pulse generator failures (battery and pulse generator circuit) and lead failures (insulation and conductor coil). Biophysical interface events are stress related. Physical, metabolic, and chemical stresses injure the tissue, causing inflammatory tissue reactions (encapsulating fibrous tissue, exit block, venous occlusion, and vegetation on leads) and mechanical tissue disruption (resulting in lead dislodgment, perforation, and mechanical stimulation of heart). Physical stress refers to mechanical stresses (traction, shearing, and compression forces) that injure the surrounding tissue. Chemical stress relates primarily to toxins associated with bacterial infections. Communication events involve the electrical signals delivered to and received from the cardiac, muscular, and neural tissues. It also involves the logic of the commands sent to the heart and how they interact. Examples include programming variables (voltage output, sensing levels, and refractory period), appropriateness of device logic interactions (conduction system and pathologic rhythms), interaction with skeletal muscle (stimulation and oversensing), and nerve stimulation (phrenic nerve stimulation). Clearly, an abstract separation into components, biophysical interface, and communication events is allinclusive with respect to DREs. Despite the fact that this abstract presentation defines DREs and reflects the pictorial information, clinical information is needed for an event to qualify as a device-related complication. Local Clinical Events

Figure 21-1. Device-related complications. Device-related complications are separated into device-related events (DREs), local events, and systemic events. DREs include all events and may or may not cause a clinical complication. DREs are separated into component failures, biophysical interface issues, and problems of communication between the pulse generator and the heart. A local event is a complication caused by a DRE and is an actual complication. Local events manifest as abnormal rhythms, hemodynamic parameters, and electrical signals. A systemic event is the systemic manifestation of a local complication. Systemic events are the signs, symptoms, and organ failures caused by perfusion or compensatory mechanisms.

Local clinical events (see Fig. 21-1) include the rhythm, hemodynamic, and electric field effects caused by DREs. A rhythm event is the resultant rhythm caused by a DRE. A rhythm event can be described using the same classic terminology used for all rhythms. Sinus bradycardia, Mobitz II atrioventricular block, and pacemaker-mediated tachycardia are all acceptable descriptions for a rhythm event. Hemodynamic clinical events are the hemodynamic sequelae of a DRE. Examples are swelling from venous occlusion, shunting from pulmonary emboli, ascites from tricuspid valve insufficiency, and cardiac tamponade from disruption of the superior vena cava (SVC).

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

An electric field effect is a DRE caused by a remote action of the electric field. Examples are crosstalk between the atrial and ventricular electrodes, stimulation of the phrenic nerve, stimulation of the pectoral muscle, and sensing of myopotentials. The local clinical events (rhythm, hemodynamic, and electric field effects) include all of the DREs that qualify as a complication. For a complication name (label) to be descriptive, both the local clinical event and the DRE must be combined in some fashion. Although the combined label is a satisfactory description of a devicerelated complication, system-related events and the clinical magnitude of an event are not known. Systemic Clinical Events Systemic clinical events are caused by local clinical events superimposed on the system’s underlying pathophysiology. Acute effects are related to perfusion and chronic effects to compensation. Perfusion issues are controlled by blood flow and pressure gradients. Tissue perfusion is intuitively obvious. A low cardiac output can result in decreased blood pressure and/or elevated venous pressure, both of which can result in poor tissue perfusion. Examples of poor perfusion to the brain are dizziness, syncope, and coma; to the kidneys, prerenal failure; and to the skeletal muscle, fatigue. Examples of elevated venous pressure are stiff lungs and pulmonary edema. “Compensatory mechanism” is a name given for the cardiovascular system’s attempt to correct a perfusion problem. Compensatory mechanisms all represent pathologic states, even though they may restore satisfactory tissue perfusion. For example, a dilated cardiomyopathy is the heart’s compensatory mechanism for low output (bradycardia, tachycardia, decreased muscle power, and/or valvular stenosis and insufficiency) and for conduction abnormalities altering the filling of the left ventricle (right ventricular apical pacing and left bundle branch block). Formation of collateral veins or arteries is the compensatory mechanism for venous or arterial occlusion. The labeling for a systemic clinical event should reflect the magnitude of the insult and/or its potential risk. This includes both the heart and the vascular system. Performing this task for each organ system is not practical. Use of accepted integrated clinical labels such as signs and symptoms will suffice. In describing an acute poor perfusion state, statements such as cardiovascular collapse, low-output syndrome, weakness and fatigue, syncope, dizziness, and asymptomatic are all acceptable. If multiple signs and symptoms apply, the one reflecting a worst-case scenario should be used. For compensatory states, the situation is less clear. The patient may be asymptomatic even though the pathologic state caused by the local clinical event is significant. In these cases, the label should include the compensatory mechanism modified by symptoms (e.g., symptomatic dilated cardiomyopathy, heart failure with dilated cardiomyopathy). Compensatory states are usually reversible with correction of the local clinical event. For example, a dilated cardiomyopathy with

857

failure symptoms secondary to a chronic heart rate of 30 beats per minute will reverse itself once the heart rate is returned to normal. Summary of Abstract Classification The abstract classification presented here defines a device-related complication. The word presentation of events (device-related, local, systemic) gives a complete picture that is intuitively obvious; examples include conductor coil fracture, loss of pacing and sensing, asymptomatic or exit block, loss of capture, and asymptomatic dilated cardiomyopathy. Although these labels are easy to understand when read, their formulation is complicated. Facility with this type of classification forces one to develop a clear understanding and organization of device-related complications. Once understood, the uniformity of the classification and the completeness of the definitions are rewards for the effort. The author has no delusions that this is the only way or even the best way to construct a uniform classification system. The usefulness of such a classification will be apparent as the chapter progresses.

Device-Related Events DREs constitute the basis for all complications. Understanding a DRE is necessary for both diagnosis and management of a device-related complication. DREs are separated into events related to device components, device interaction at the biophysical interface, and communication between the device and the tissue. The first two types of DREs will be discussed here. The third type is the subject of the chapters on stimulation, sensing, and arrhythmia detection (Chapters 1 and 3). Component-Related Problems Battery Failure and Pulse Generator Circuit Failure A battery and/or pulse generator circuit failure is an industry-related misadventure. The physician can neither cause nor prevent these types of complications. Pulse generator reimplantation is the only definitive treatment for these failures. Reimplantation is simple and direct and relieves the problem. The company should be notified and the pulse generator returned. Even if this was a known random failure mode, it is best to be safe and notify the company. No further action is needed. If the mode of failure is unknown or new, it may require some form of advisory or recall action. In both circumstances, the pulse generator should be returned to the manufacturer. A company’s advisory or recall notice usually states that reimplantation should be performed immediately should a failure occur, if there is a high probability of failure occurring, or if there is a chance that failure may be life-threatening. Reimplantation for generator failure or high probability of failure is obvious. The language “chance of a life-threatening failure” leaves the decision

858

Section Three: Implantation Techniques

to replace the pulse generator in the hands of the physician. Two variables are implied in this subjective language: the incidence of failure and the patient’s intrinsic conduction. If the incidence of failure is low and/or the patient is not “pacemaker dependent,” continued close follow-up is recommended, as well as notification and counseling of the patient. In those situations where the manufacturer can be precise and recommend a logical plan of action, most physicians will follow those recommendations. If the recommendations are not precise and the physician has to make the reimplantation decision, confusion and insecurity often ensue, and the pulse generator is replaced. Although replacement resolves the pulse generator problem, the probability of a complication associated with replacement, such as infection, lead damage requiring a corrective action, or physiologic decompensation or death, may be greater than the risk of continued close patient monitoring. A brief discussion of pulse generator failure is in order. It will be separated into battery failure and failure of other circuit components. Battery failure seems to be a trivial event, because all batteries fail with normal depletion of their energy stores. This is not a component failure as long as the battery depletes with time as designed. A true battery failure is rare, but when it does occur, it causes global pulse generator failure. The most common forms of battery failure are caused by some form of short circuit in the system. For example, a short circuit caused by a circuit component failure may manifest as a battery failure. If the system is short circuited and the battery overheats, the temperature change is felt by the patient. There are examples in the past of overheated batteries rupturing their containers, but tissue damage has not been reported. Examples of circuit failure are shunting of the highenergy circuit to a low-energy circuit, reed switch failure, grommet failure, and other random individual component failures. Stress-related issues usually occur over time. Examples of time-dependent stress-related failures are battery failure in the Marquis family implantable cardioverter-defibrillator (ICD) and cardiac resynchronization therapy defibrillator (CRT-D) (Medtronic, Minneapolis, Minn.), which was caused by a manufacturing error,1 and shunting of energy from the highenergy circuit to the low-energy circuit in the PRIZM 2 ICD (Guidant, Boston Scientific, Natick, Mass.), with destruction of the low-energy circuit.2 A brief discussion and recognition of causes of past failures are informative and may be useful if similar mechanisms occur in the future. A circuit failure mode that plagued all manufacturers in the past was the presence of dendritic growth of conductive crystals causing a short circuit. This unique failure mode was prevented by hermetically sealing the circuit. Another form of battery failure was caused by failure of the artificial baffles placed between the chemical components within the battery. This is not an issue with the lithium/iodine battery currently in use in pacemakers. The baffle is formed by the reaction of the two components, lithium and iodine. This reaction generates the charge and forms the baffle at the same time. Rechargeable batteries failed not because of battery issues, but

because they failed to gain patient acceptance due to the necessity of frequently recharging the battery. The most odious battery technology was the nuclear battery. A nuclear cell generated heat, which was converted into electricity. This battery was cloaked in red tape by the nuclear regulatory agency and was expensive. Red tape, expense, and the frequent advances in pulse generator performance negated the advantages of a long-acting battery, dooming it to failure. Lead Failure Lead failure is a common complication. A pacemaker lead failure is defined as failure of 1% per year for 5 years (i.e., 95% lead survival at 5 years).1 ICD leads are more complicated and fail at a higher rate, and their failure rate has not been well defined. Despite their complexity, both pacemaker and ICD leads should be judged by the same standards. Lead failures include conductor coil fractures and loss of integrity of the inner and/or outer insulation. The mechanisms of failure are well defined for both types of leads. The mechanisms for lead failure relating to the conductor coil include binding and excessive torque. The mechanisms for conductor coil failure are mechanical and are best illustrated by clinical examples. Failures relating to the insulation include environmental stress cracking (ESC), metal ion–induced oxidation (MIO), compression erosion or abrasion, cold flow, and tears. Insulation failures have general mechanical and chemical properties common to all polymers. Clinical examples specific to polyurethane and silicone rubber are discussed in more detail later. Two types of polyurethane are still used today: Pellethane 80A and Pellethane 55D. The difference between these two polymers is the percentage of short-chain polyurethane. Pellethane 80A has 70% short chain and is soft and supple, making it ideal for bipolar leads. Because of its insulation properties, bipolar leads could be made small for the first time. Pellethane 55D has 30% short chain and is harder and less supple. These leads were too stiff for bipolar configuration but were ideal for unipolar leads. Conductor Coil Fracture Conductor coil fracture is a break of one or more of the wires going from the pulse generator to the electrode (Fig. 21-2). A mental picture of a lead failure with the two ends separated or, at best, making intermittent contact is valid immediately after the break. The space between the broken ends acts as a perfect insulator with an infinite impedance (i.e., no pacing or sensing). Chronic fractures are different. The space at the fracture site is filled with an electrolytic fluid within a short time period. It then behaves like two electrodes in an electrolytic medium (leaking capacitor), allowing some current to flow and generating a voltage at the pacing electrode. If the voltage is high enough, it can continue to pace the heart. The impedance is high, in the range of 1000 to 2000 ohms. Intermittent contact can cause a transmitted current spike, resulting in a voltage sufficient to pace or generate an electrical signal (a form of make-break signal) which, if seen by the sense ampli-

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

859

tor coil fracture. The excessive torque at the terminal pin is usually caused by the conductor coil’s binding against the stylet along a circuitous route. Similarly, excessive torque can be applied to the distal crimp and weld joint at the helix when the helix is bound inside the housing. Breakage at the terminal pin is more common and is associated with a circuitous path, whereas breakage at the helix is related to a manufacturing defect or to blood and tissue ingress that wedges the helix in the housing. Insulation Failure

Figure 21-2. Conductor coil fracture. Conductor coil fractures can have an acute or a chronic presentation. A, An acute fracture and separation of the conductor coil result in no current flow to the electrode and an infinite impedance. The result is no pacing spike on electrocardiogram (ECG) and loss of pacing. B, A chronic fracture has accumulated electrolytic fluid in the fracture space. The ends of the fractured wire behave as electrodes, and current can flow between the two ends of conductor coils at a high impedance. In many cases, the charge density on the electrodes is sufficient to pace, and a pacing spike is present on ECG.

fier, will be acted on, causing a sensing abnormality. In some situations, both the inner and outer coils break, along with the insulation, leaving the entire lead in two separate pieces. In regard to mechanisms of failure, the first example involves the medial placement of leads between the clavicle and the first rib in the thoracic inlet. This complication is related to lead binding. Passing the introducer needle through the subclavius muscle or the costoclavicular ligament can result in calcification of those structures. Lead binding occurs when the lead body becomes entrapped in the encapsulating calcifying fibrous tissue. If a lead with a helical conductor coil is rigidly bound, the helix cannot bend to relieve deformation stresses when the conductor coil is flexed on either side of the binding site. This situation is similar to a wire being bent back and forth in a narrow site. The stress at the flex site causes heat and, finally, breakage (fracture) of the wire. This same wire, configured as a helix, will bend back and forth without causing stress at the flex site. The helix is flexed, preventing deformation of the wire itself. Binding of the helix prevents it from flexing, causing the wire to be deformed at the junction between the bound and unbound helix, resulting in conductor coil fracture. Another mechanism of failure involves the crimp and weld joints. These types of conductor coil fractures are caused by the torque and motion stresses applied to the joints by extension and retraction of the helix on an active fixation lead. Turning the inner conductor coil is a common mechanism used to extend and retract an active fixation helix. Excessive torque can break a crimp and weld joint between the conductor coil and the terminal pin or the helix (or both), resulting in a loss of integrity of the wire and mimicking a conduc-

Lead insulation failure (Fig. 21-3A) became a significant clinical issue in the late 1970s with the use of polyurethane (Pellethane 80A).3-6 Failure of the polyurethane insulation in unipolar leads was shown to be caused by ESC. Once the residual strains were relieved by changing the manufacturing practices, this problem was resolved. Normal stresses applied at the biophysical interface will cause some minor ESC. A far more sinister problem, MIO, also arose and is still with us today. MIO is the breakage of the short-chain ether linkages in Pellethane 80A by metal ions diffusing from the conductor coil into the insulation. MIO causes dissolution of the insulation at the molecular level. Current state-of-the-art leads have minimized, but not eliminated, the problem. Pellethane 55D is primarily long-chain polyurethane and is more resistant to MIO than Pellethane 80A. Cold flow is a property of plastics caused by compression. The polymer flows away from the compression site, thinning the insulation. Cold flow can cause a loss of the insulation’s integrity. This is a problem for both polyurethane and silicone. It is seen at sites of compression, such as suture tie sites on the suture sleeve, points where two leads press against one another, between the clavicle and first rib, and at points where the pulse generator is pinning the lead against the bone in a submuscular pocket. Formulations of silicone designed to withstand compression are more resistant to cold flow than silicones designed to stretch. The effects of loss of insulation integrity depend on the arrangement of the conductor coils. Each conductor coil needs an insulator. A unipolar lead with one insulated conductor coil has only one configuration. Most bipolar leads with two helical conductor coils are coaxial, with the insulation around the inner and outer coils. Failure modes for the unipolar and coaxial bipolar leads are well known. Clinical experience with ICD leads that have multiple individually insulated conductor coils or individually insulated high-voltage coils has been good. There are few insulation failures. Although the failure modes for new leads with more sophisticated arrangements are unknown, failures caused by known mechanisms should still be apparent. An outer insulation covers both unipolar and bipolar leads. Unipolar leads have one conductor coil connecting the pulse generator to the electrode (see Fig. 21-3B). A loss of insulation integrity creates an alternative current pathway between the lead and the pulse generator. These defects behave the same as another

860

Section Three: Implantation Techniques

B A

C

negative electrode capable of stimulating muscle and/or nerve tissue and sensing myopotential signals. The electric circuit created by the defect is in parallel with the pacing circuit. The resultant circuit obeys the laws governing a parallel circuit. The impedance across the parallel circuit decreases as the size of the defect gets larger and the current flow increases. Because the pulse generator’s output current from the discharge of the capacitor is limited (approximately 17 mA), at some point the current to the pacing electrode will be insufficient to generate a threshold voltage, causing failure to pace. Also, the increased current drainage depletes the battery, shortening its life. If the electrode at the site of the defect is near skeletal muscle, myopotential signals can be detected by the sense amplifier, causing inhibition of pacing (same as a bipolar lead). Multiple defects in the insulation create multiple electrodes and multiple parallel circuits. Except for differences in location, multiple defects behave like one large alternative site electrode in parallel with the pacing electrode. Bipolar leads have both outer and inner insulation. A loss of integrity of the outer insulation in a bipolar lead exposes the outer conductor coil to the tissue in the same manner as for a unipolar lead. This might

Figure 21-3. Lead insulation failure. A, Outer insulation failure exposes the conductor coil to the surrounding tissue. Inner insulation failure allows the two conductor coils to communicate. B, Unipolar configuration has only one conductor coil and an outer insulation. Defects in the insulation act as electrodes, with the potential of simulating tissue (muscle and nerve stimulation) and/or sensing electrical activity (myopotential inhibition). The stimulation circuit is parallel to the pacing circuit, resulting in a decrease in impedance. Current flow may exceed the maximum discharge current from the capacitor, causing a decrease in voltage to the pacing electrode (loss of capture). C, Both outer and inner insulation failure can occur in bipolar configuration. Outer insulation failure in the bipolar configuration differs from that in the unipolar configuration. Sensing abnormalities can still occur, because electrical activity can be seen by the sensing circuit. Stimulation pathways similar to the unipolar configuration do not exist. Inner insulation failure has the potential to create makebreak signals and to short-circuit the pacing circuit. Small defects in the inner insulation act as a small capacitor between the inner and outer coils, charging and discharging. These discharge signals can be seen by the sense amplifier, causing inhibition. A large defect causing shunting of current between the inner and outer coil lowers the impedance and potentially causes loss of capture (exceeding the output current of the capacitor).

lower the impedance slightly, but it does not change the pacing threshold. A conductor coil exposed to skeletal muscle acts as a sensing electrode and sends myopotential signals to the sense amplifier, causing oversensing and inhibition of pacing. The voltage is usually too low on the outer conductor coil to stimulate muscle or nerve tissue. Inner insulation failure has meaning only for bipolar leads (see Fig. 21-3C). Loss of integrity of the inner insulation creates a pathway to the outer conductor coil (electrical short-circuit). The size of the defect determines the current flow through the defect and its influence on the pacing electrode’s voltage. In addition, a small defect behaves like a capacitor with the two wires separated by a dielectric. As the voltage builds across the defect, the dielectric breaks down, shunting current from the inner to the outer coil. The sense amplifier may see this make-break signal. Sensing abnormalities caused by make-break signals are an early sign of inner insulation failure. As the defects become larger, they no longer act as a capacitor, and current is shunted from the inner to the outer coils. This defect creates low-impedance parallel circuits. The increased current decreases battery longevity. Once the pulse generator’s current output

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

from its capacitor is maxed out, voltage on the pacing electrode will decrease, just as for a unipolar lead. The first example of this mechanism of failure is related to the lead’s being pinched (crush injury) between the clavicle and the first rib. This type of compression injury damages the lead body insulation. The loss of integrity of the polymers caused by compression injury results in inner insulation failure, resulting in make-break signals and short-circuiting between the inner and outer coils. Outer insulation failure is of consequence only if the exposed wire senses electrical signals and causes oversensing. Silicone rubber and polyurethane are the two common polymers used for lead body insulation. Silicone deforms, cracks, and breaks when subjected to physical stress. Deformation by cold flow is the most common type of silicone compression, resulting in a flow (cold flow) of the polymer away from the compression stress. Once the stress is relieved, the flow stops. Continued stress will result in thinning to the point of loss of integrity. Excessive shearing and traction stresses result in cracking (breaking of molecular bonds) along stress lines, which can lead to loss of polymer and/or loss of lead integrity. Polyurethane deforms, cracks, and breaks when subjected to physical stress in a fashion similar to silicone. However, it seems to be more sensitive to stress. Compression easily causes cold flow, and excessive stress (shearing, traction, and compression) causes cracking (breaking of ether bonds) along stress lines ESC. ESC occurs at sites of applied stresses to the lead. The initial failure mechanism for Pellethane 80A was massive ESC caused by residual manufacturing stress left in the insulation from a solvent expansion-contraction technique used to place the insulation over the conductor coil. In the early 1980s, a new failure mechanism was found. Metal ions that diffused into the Pellethane 80A polymer broke the ether linkages in the polyurethane polymer, causing dissolution (crumbling) of the insulation. This was called MIO of the polymer, and it caused a non–stressrelated failure with loss of polymer integrity. The insulation seemed to just melt away. Historically, a potentially dangerous type of lead failure occurred due to protrusion of a retention wire through the insulation and into the surrounding tissue. This type of failure was the first example of a destructive lead failure. It had the potential to be destructive (i.e., to penetrate or tear tissue locally or remotely after migration), but it did not cause a problem with the electrical performance of the leads. The Telectronics 330801 and 330-854 leads (St. Jude Medical, Sylmar, Calif.) had a component called a retention wire that was designed to hold a J configuration in the atrium. This wire had the potential of breaking (fracture) and eroding through the insulation without a loss of its electrical integrity; pacing and sensing characteristics were unchanged. The eroded retention wire could penetrate the superior veins, atrium, or aorta, causing a cardiovascular emergency. This failure mechanism had a particularly important historical impact on the development of lead extraction expertise in the physician community. Because of the

861

potential for fatal bleeding, hundreds of physicians were motivated to use these techniques. However, not everyone who extracted these leads was experienced or well trained, and the disadvantage of extraction became immediately apparent as the risk of extraction in the hands of some operators greatly exceeded the risk of fatal intrathoracic bleeding from penetration of the aorta by the J wire.7-10 Although the specific techniques required for the removal of these leads are of historical interest (almost all were removed or the patients have died of old age), the risk-benefit analysis done for this situation is core to the decisions patients and physicians make every day for lead extraction or even for device change in the event of a device notification (recall).11 The bottom line is that lead extraction training is essential, and tailoring extraction techniques to the specific lead construction and failure mechanism is crucial. Biophysical Interface The biophysical interface is the boundary between an implanted device and the body’s tissue. If the device injures the tissue, there is an interaction between device and tissue called the inflammatory reaction.12 The author’s modeling of the interactions between implanted devices and surrounding tissue is based on general knowledge and experiences. Depending on the magnitude of the inflammatory reaction, it is considered a normal event, in the same sense as wound healing (primary intention) by an inflammatory reaction is considered normal. For example, the electrode is the perfect example of a critical device component whose function is dependent on a normal inflammatory reaction at the biophysical interface. Device-related tissue injury is caused by applied stresses, including mechanical stresses (pressure and traction), traumatic disruptions (penetration, tears, and ruptures), metabolic insults (supply of oxygen and other nutrients), and chemical toxins (infections). Applied stresses are separated naturally into physical, metabolic, and chemical stresses (Fig. 21-4). Allergic reactions to device components have been postulated, speculated, and alleged but never proved to exist. The applied stresses cause an inflammatory reaction ranging from minor tissue injury to tissue disruption. To help avoid confusion, discussion of injury-related events at the biophysical interface are separated into inflammatory reactions and applied stresses. Inflammatory Reactions The inflammatory reaction is the body’s physiologic response to tissue injury or death caused by any form of stress applied to the tissue. This includes the organization of blood clots in the bloodstream into fibrous tissue. Inflammatory reaction is hypothesized to be the same for all tissue, but there are various tissue-related presentations. In the vascular space, clot plays a predominant role. The following is a pictorial discussion modeling an inflammatory reaction (Fig. 21-5). If one assumes an event happening in extravascular tissue (e.g., soft tissue), the initiation of an inflamma-

862

Section Three: Implantation Techniques

Figure 21-4. Biophysical interface. Applied stresses at the biophysical interface are caused by physical stresses (mechanical forces), metabolic issues, and chemical toxins (bacterial toxins). Applied stresses injure tissue, initiating an inflammatory reaction. Both the applied stresses and the inflammatory reaction cause physiologic events such as cardiac rhythm disorders, hemodynamic insults, and electric field complications (phrenic nerve stimulation).

tory reaction begins with emigration of cellular elements in the blood into the extravascular space. In normal laminar blood flow, cellular elements, such as red and white blood cells, are located in the center cellular zone, and the smaller blood components, such as proteins, are in the peripheral plasmatic zone (see Fig. 21-5A). Normally, there are no cellular elements in the

plasmatic zone near the vessel walls. Imagine this arrangement as a natural consequence of blood flowing above a certain velocity. The known sequence of events after injury to a mass of tissue in which the blood vessels remain structurally intact includes multiple steps. The first event is the dilation and subsequent increase in flow of small blood vessels in and near the injured tissue. The initial cause of this dilation is not clear. At this point, except for dilated vessels and increased flow, the physiology at the capillary level is still normal: the arterial pressure (hydrostatic pressure) forces electrolytic fluid out of the first part of the capillaries and into the extravascular space. This increases the concentration of the proteins, increasing the intravascular oncotic pressure and viscosity. The blood flow slows, and fluid is pulled back into the capillaries on the venous side (see Fig. 21-5B). Ideally, equal volumes are forced out and pulled back in. The next event is a swelling of the endothelial cells that line the vessels in the capillary bed. This allows a fluid exudate containing both electrolytic fluids and protein material to move into the extravascular space, resulting in edema and dilution. The protein loss decreases the oncotic pressure and the intravascular volume, causing further slowing of blood flow. With this loss of fluid exudate and slowing of blood flow, the blood cells migrate from the center of the bloodstream into the plasmatic zone and to the wall (see Fig. 21-5C). The vessels then become more porous, resulting in neutrophil emigration and forcing out of red blood cells by hydrostatic pressure. The presence of fluid exudate and neutrophils marks the beginning of the full

Cellular Zone

A

B Figure 21-5. Inflammatory reaction (normal physiology). Inflammatory reaction is the body’s response to tissue injury. The extravascular inflammatory reaction represents the microbiologic and macrobiologic physiologies of an inflammatory reaction. A, Normal physiology at a microbiologic level involves the flow of blood through small blood vessels and capillaries supplying the metabolic needs of uninjured tissue and removing metabolic waste. Blood flow through these vessels is laminar, with blood cells flowing in the central cellular zone and proteinaceous material flowing in the peripheral plasmatic zone. Hydrostatic pressure forces transudate out of the blood vessels with the nutrients, and oncotic pressure draws the transudate back into the blood vessels with the metabolic waste. B, Tissue injury starts the transudate phase. Chemical mediators and direct injury to the vessel walls cause the vessels to become more porous. The hydrostatic pressure forces more fluid out than the oncotic pressure draws back in, resulting in tissue edema.

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

863

D C

E

F Figure 21-5, cont’d. C, The exudative phase occurs when cellular elements pass through the blood vessel wall and into the injured tissue. As the injury persists, the vessel walls become more porous, and cellular elements migrate to the walls. Emigration of neutrophils and the hydrostatic pressure force red blood cells out of the blood vessels and into the edematous tissue exudate. This marks the beginning of the full inflammatory reaction. The intensity of the inflammatory reaction depends on the magnitude of the tissue injury. D, The organizational phase is a dynamic event, with formation and drainage of exudate and debris. In the early organization phase, an exudative effusion occurs as the result of poor drainage. A transition zone between the injury zone and normal tissue has capillary and fibrous tissue formation. The transition zone creates a barrier between the injury zone and normal tissue. Barrier formation consists of contraction and cross-linkage of the fibrous tissue, which increases the tensile strength. Contraction also decreases the vascularity in the transition zone as the tissue injury subsides. Continued tissue injury decreases the contraction, increasing the vascularization and formation of fibrous tissue. This dynamic reaction persists until the barrier is protective. This reaction is most intense at the interface between the transition zone and the injury zone. E, The late organization phase is characterized by maturation of the interface between the transition zone and injury zone into a granulation tissue zone. Granulation tissue is the vascular fibrous tissue barrier that separates the injury zone from normal tissue. As the barrier becomes more effective, the transition zone decreases. F, In the normal early healing stage, scar tissue is still present and is continuously remodeling. Natural healing is complete resolution of scar tissue, leaving normal tissue. Natural healing occurs when the scar tissue mass is not large and tissue stresses are not present. Bonding scar (healed incision) occurs when the tensile strength of the normal tissue is insufficient to hold it together without stress and strain, injuring normal tissue. A permanent bonding scar supplies the tensile strength necessary to remove stress from normal tissue. A residual scar exists when the mass of scar tissue is so large and dense that it stresses the surrounding tissue, causing a low-grade tissue injury that perpetuates the scar. This scar mass usually resolves over time.

inflammatory reaction, with plasma factors such as fibrinogen and chemical mediators being released from the cells. The continued evolution of the inflammatory reaction depends on the tissue involved and the magnitude of the injury. Now picture a lead located in the middle of the bloodstream, with the cellular elements and blood

components flowing past at some velocity, and assume that a clot forms on the lead. (The terms “clot” and “thrombus” are the same and are used interchangeably; thrombus is used clinically to describe events associated with clot.) Clot can lyse in two ways. It can lyse completely without an inflammatory reaction, or it can lyse as a part of an inflammatory reaction. For an

864

Section Three: Implantation Techniques

inflammatory reaction to take place, the clot and lead must first contact the vessel wall, injuring the tissue. The inflammatory reaction in the vessel wall is the same as an extravascular reaction. The inflammatory reaction extends into the clot, with lysis of the clot or organization of the clot to form a mural thrombus (organized clot attached to the vessel wall). Except for drainage, most of the other factors are similar. If the cleaning phase is incomplete (i.e., lysis of clot and fibrin), the remaining tissue organizes. This granulation tissue acts as a barrier, isolating the thrombus from the bloodstream. The granulation tissue transforms the thrombus into the encapsulating fibrous tissue surrounding the lead or becomes a mural thrombus. Now picture the lead touching the wall, applying enough pressure to injure the tissue. The injured tissue initiates the inflammatory reaction. Clot forms on the injured tissue and around the lead as a result of chemical mediators associated with the inflammatory reaction and other factors, such as negative surface charge, turbulence, and stasis. It does not matter whether free clot is contiguous with the vessel wall or the lead touches and injures the wall—the end result is the same. The clot is completely dissolved, or it is organized into encapsulating fibrous tissue and/or a mural thrombus. The properties of an inflammatory reaction are dependent on the tissue type and magnitude of the injury. From a macrobiologic point of view, the inflammatory reaction cleans and drains the area, organizes fibrous tissue to wall off the cause of injury, bonds tissue together, and promotes cellular regeneration where applicable. To clean the injury site, phagocytosis kills bacteria and degrades necrotic cellular debris, and enzymes lyse fibrin and clot to facilitate lymphatic drainage. Organization is the formation of granulation tissue at the edges of the damaged tissue (see Fig. 21-5D). Granulation tissue is composed of endothelial cells forming capillaries and fibroblasts laying down a fibrous tissue network. The granulation tissue protects normal tissue from injury by constructing a fibrous tissue barrier at the biophysical interface (see Fig. 21-5E). It uses its tensile strength to bind tissue and its contractile properties to obliterate dead spaces. An example of regeneration is new skin covering the injury site. Soft tissue and cardiovascular tissues do not regenerate. Applied Stresses Physical Stresses. To avoid confusion, physical stresses are separated into extravascular and intravascular stresses. There are significant differences in the initiation and presentation of an intravascular inflammatory reaction. Intravascular inflammatory reaction is complicated by the clotting mechanism, blood flow, and the circulating cellular and protein components in the bloodstream. The physiologic effects of physical stresses are common to both extravascular and intravascular inflammatory tissue reactions. Implantation of a pulse generator, leads, adaptors, arrays, and so on, in the soft tissue causes an inflammatory reaction. The physical stresses create a tissue injury zone by tissue disruption (surgically forming the

pocket) and by pressure (compression) on surrounding tissue at the biophysical interface. The pressure is caused by the exudative fluid in the pocket (hemorrhagic fluid) and by the implanted devices themselves. The tissue disruption far exceeds the criteria set in the microbiologic view for initiating an inflammatory reaction. The tissue is disrupted by both sharp and blunt dissection, leaving a surface of exposed tissue with exposed ends of blood vessels and lymph channels, drainage of extracellular fluid, exposed injured and dying cells, and an injury zone extending from the pocket edge. The tissue injury and exudative fluid media mark the beginning of the inflammatory reaction. Excessive bleeding into the cavity with clotting creates a biodegradable foreign body which, along with the cellular and proteinaceous debris, must be dissolved and drained. The inflammatory reaction in the injury zone is caused by pressure applied to the tissues by the implanted device, the diffusion of chemical mediators released from injured and dying cells after tissue disruption, and the fluid exudate. Surgical closure of the incision leaves in progress an inflammatory reaction associated with tissue-to-tissue healing by primary intention and an inflammatory reaction associated with a cavity containing an exudate and a foreign body, hopefully healing by secondary intention. The inflammatory reaction associated with closure of a wound by primary intention is called wound healing. Lymphatic drainage removes the cellular debris, proteinaceous material, and blood products. Organization by a thin layer of granulation tissue between the two layers vascularizes the tissue and bonds the two sides with fibrous tissue. As the fibrous tissue matures, its tensile strength increases, by way of collagen cross-linkage, and becomes greater than that of the native tissue. The skin regenerates on the thin scar. The cavity contains a device, exudate, and a cavity wall of injured tissue. Drainage is the same as for wound healing. The difference is in the amount of material to be drained, the time it takes to prepare the material for drainage, and the capacity of the lymphatics. Organization takes place along the cavity wall. Granulation tissue forms on the edge and functions as a barrier, relieving stress to the tissue by forming a fibrous tissue interface (pocket or biophysical interface). The biophysical interface is, in effect, a protective barrier of encapsulating fibrous tissue that separates and protects the normal tissue from device-related physical stresses. Because the goal of the barrier is to eliminate the physical stress to the tissue, the thickness of the biophysical interface is determined by the physical stress. Without this stimulus, new granulation tissue is not formed and the fibrous tissue contracts, constricting blood vessels. The resultant devascularized fibrous tissue is called scar tissue when it is acting primarily as a bonding agent, as in wound healing. When it acts as a protective barrier, as in a device pocket, it is called encapsulating fibrous tissue. If the inflammatory reaction is minimal, tissue bonding and a protective barrier are not required. Fibrous tissue is absorbed, leaving normal tissue. This is called natural healing (see Fig. 21-5F).

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

Scar tissue and encapsulating fibrous tissue are continuously remodeled along stress lines. A simple explanation for this is that a dynamic equilibrium exists at the biophysical interface between the device and the surrounding tissue. The inflammatory reaction caused by tissue injury results in organization, reinforcing fibrous tissue, and the absence of injury results in dissolution of the fibrous tissue. This dynamic process continuously remodels the fibrous tissue along lines of stress. Intravascular leads implanted in a vein and in the right side of the heart demonstrate different inflammatory reaction properties than those in soft tissue. In the bloodstream, lead-related injury to the endothelial cells lining the vein and heart wall may be a dominant factor contributing to the encapsulating fibrous tissue and to mural thrombus formation (defined as a thrombus attached to the vein or heart wall). However, blood flow factors (stasis and turbulence) and surface properties of the lead also contribute to encapsulating fibrous tissue and thrombus formation. Inflammatory reaction is also involved with incorporation of the lead into the vein wall and, in some cases, exclusion of a segment of the lead from the vein. Assume that a small area of vein wall is traumatically injured, with damage to the endothelial cells. Because the wall of the vein does not contain a vascular component, the normal sequence for injury, small vessel dilatation, and fluid exudation cannot take place. However, the injured cells do swell, change polarity to positive, and liberate clotting factors. Platelets drawn to the wall by opposite charge or by random migration into the plasmatic zone adhere to the injured site. Once a small mound of platelets forms, the local turbulence and stasis cause more platelets to migrate into the plasmatic zone and adhere to the mounded platelets, releasing more clotting factors. Once red blood cells migrate to the area and fibrinogen is converted into fibrin by the clotting factors, a clot is formed. This clot is called a mural thrombus. Depending on the magnitude of the injury, blood flow, and the effectiveness of clot lysis, the thrombus will grow, remain stable, or dissolve (Fig. 21-6). A persistent mural thrombus will eventually attract other cells, including fibroblasts. Organization of the thrombus by an inflammatory reaction will replace the clot with fibrous tissue. Unless further clot forms on the fibrous tissue mass, it will remodel along stress lines with time, leaving a contracted scar at the injury site. Continuous clot formation will result, with continuous organization in a large pedunculated fibrous tissue mass. Now assume that a small area is injured by an implanted lead that is applying pressure to the wall. A mural thrombus can form, as described, from the injury to the wall. In addition, the presence of the lead can cause turbulent flow, accelerating the migration of platelets and red blood cells to the wall. Also, stasis of flow can occur between the vein wall and segments of the lead. This further aids in the migration of platelets, red blood cells, and fibroblasts to the wall. The influence of continued pressure causes injury to the wall; turbulence and stasis practically guarantee mural thrombus formation, and that is exactly what

A

B

C

D

865

Figure 21-6. Intravascular inflammatory reactions. Intravascular inflammatory reactions involve clot and differ from extravascular reactions. Clot can form without causing an inflammatory reaction, or the clot (clot and lead) can injure the vein wall and cause an inflammatory reaction. A, Clot forms when leads injure the vascular wall, cause stasis and turbulence of flow, and/or have thrombogenic surface texture properties. B, Clot can lyse, removing all clot from the lead surface, or it can become involved in an inflammatory reaction at the vessel wall, forming a mural thrombus, which can organize into encapsulating fibrous tissue. C, Secondary clot formation on the organizing thrombus occludes veins and causes thrombi in the atrium. Lead pressure on the wall causes inclusion of the lead into vessel or heart wall. D, Thrombosed veins can recanalize into patent veins or be obliterated by the inflammatory reaction into scar tissue surrounding the lead.

happens clinically. The organization of the thrombus into encapsulating fibrous tissue surrounding the lead and involving the vein wall can persist in some form for the duration of the implant (dynamic modeling). If clot formation persists because of turbulence, stasis, and/or surface issues, organization can result in exuberant encapsulating fibrous tissue (Fig. 21-7). Also, if the exuberant fibrous tissue mass causes tissue injury, a recursive reaction can be started. The mass of fibrous tissue injures the vein wall, generating more fibrous tissue, which causes further clotting, perpetuating the process. In time, fibrous tissue increases in tensile strength, and at some sites it mineralizes (primarily calcium deposits), creating a bonelike structure. In children and young adults, mineralization occurs after 4 to 5 years, and in the elderly after 8 to 10 years.13 A lead in the bloodstream may be free of encapsulating fibrous tissue, leaving the blood as the biophysical interface. However, other leads in this same situation can be covered with a thin layer of encapsulating fibrous tissue. Chemical structure and the texture of the lead’s surface can stimulate fibrous tissue formation on the surface of the lead, even if these leads are in a high flow area in the center of the blood stream.14 The cellular elements are also in the center of flow, and the surface properties can attract platelets and other cell elements. Microturbulence can increase their contact with the lead’s surface. For example, the old rough silicone surfaces were usually covered with a thin layer of encapsulated fibrous tissue, whereas the smooth polyurethane surfaces were free (Fig. 21-8).9 Also, the interstices of the conductor coils on ICD leads

866

Section Three: Implantation Techniques

B

A

C

are ideal for fibrous tissue ingrowth. Conditioning the surface to make it smooth or changing its chemical properties is becoming a priority with industry. Covering conductor coils with a conductive material such as ePTFE (expanded polytetrafluoroethylene), known as Gore-Tex (Guidant), and back filling with silicone (Medtronic) are two current examples. Multiple leads increase the vein wall injury sites, and the lead configuration can increase turbulence and create areas of stasis. Mural thrombus formation occurs at the contact points along the walls. Metabolic Stresses. Metabolic stresses clinically refer to those situations that cause oxygen deprivation to

Figure 21-7. Thrombus formation. A, Acute thrombus formation 1 week after implantation in a dog. Thrombus has formed at each site where the lead touches the wall and in some regions where there was stasis or eddying of blood flow. Most of this early thrombus lyses. It persists and organizes at sites of continued pressure on the wall and in areas with stagnant blood flow. Inset demonstrates a close-up of thrombus. B, Organized fibrous tissue encapsulating the lead forms chronically at sites of pressure. It progresses from fibrous tissue to cartilaginous tissue and, finally, to bone. C, Magnified view of chronic encapsulation shows the magnitude of the fibrous tissue to be ablated in extracting the lead.

cells. Local ischemia is caused by entrapment of the arterial blood vessels in encapsulating scar tissue, which obstructs blood flow. Acute oxygen deprivation injures or kills cells, liberating chemical mediators and initiating an inflammatory reaction. Complications such as migration, erosion, and tissue gangrene are caused by metabolic stresses. Chemical Toxins (Infection). Bacteria can liberate chemical toxins (exotoxins and endotoxins) that injure and kill tissue, initiating an inflammatory reaction. Exotoxins are classified as neurotoxins (nerve), enterotoxins (intestine), and cytotoxins (all tissue). Cytotoxins are of primary interest with device-related complica-

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

867

B

A

Figure 21-8. Lead surface reaction canine model. Reaction to polyurethane (A) and silicone rubber (B) in leads implanted for 12 weeks. A, The polyurethane lead has minimal encapsulation, whereas the silicone rubber lead is completely encapsulated. B, A surface thrombogenicity phenomenon not related to the mechanical properties of the lead. The magnitude of the surface encapsulation is probably responsible in part for the greater difficulty of extraction of silicone leads in the first 2 to 3 years after implantation, compared with polyurethane.

tions, although Staphylococcus aureus can produce all three toxins, causing a lethal infection if left untreated. Endotoxins are released by gram-negative bacteria and are not a common problem with device-related infections. Endotoxins usually activate the Hageman factor (factor XII), which can activate the coagulation, fibrinolytic, complement, and/or kininogen systems. By definition, an infection is just the presence and multiplication of bacteria or any other organism. If a symbiont bacterium causes harm or lives at the expense of the host, it is classified as parasitic. A parasitic bacterium that causes harm is called a pathogen, and the infection becomes an infectious disease. All bacteria associated with an implanted device are parasitic, and, in this chapter, infection refers to an infectious disease caused by pathogenic parasitic bacteria, whether the infection is active or dormant (latent or clinically occult). Pathogenic bacteria have properties such as virulence (intensity), invasiveness (ability to spread, grow, or reproduce), and infectivity (ability to establish a focal point). These properties and the organism’s pathogenic potential (morbidity caused by its toxigenicity)

determine the magnitude of the infection and its clinical sequelae. However, toxigenicity is usually not a major issue in device-related infections. Other properties, such as adhesins, may be just as important as pathogenicity. Examples of adhesin properties are slime layer, capsule, and S layer, which facilitate the adherence of bacteria to smooth surfaces, especially implant devices. Slime production may be the primary criterion for the pathogenicity of S. aureus. It is a factor in both the initiation and the persistence of a device-related infection. For example, the inability to cure a device-related infection with antibiotics may be related to presence of adhesins.

Clinical Concepts Requisite Skills Prerequisites for managing device-related complications include a general knowledge base, patient

868

Section Three: Implantation Techniques

preparation, anesthesia, procedure room issues, soft tissue surgical skills, lead extraction skills, and lead implantation skills. Most physicians practicing electrophysiology accept lead extraction as a skill to be learned. However, they may not recognize the need to expand and perfect their soft tissue surgical and implantation skills to perform at the advanced level necessary to manage device-related complications. Patient Information and Preparation All patients presenting for a procedure must have a standard history and physical examination with a detailed description of all DREs, including the company and the model of the pulse generator and all leads. The procedure notes should be carefully reviewed for any problems observed during the implantation, such as difficult access. Also, if infection is present, detailed information reflecting the time line, organism, susceptibilities, and antibiotic therapy is needed. Basic laboratory work including white blood cell count, hematocrit and hemoglobin, platelet count, blood urea nitrogen, creatinine, potassium, sodium, liver profile, and prothrombin time is generally indicated. The patient’s blood should have been typed and crossmatched for a possible blood transfusion. A current chest X-ray and electrocardiogram (ECG) are mandatory. An echocardiogram is mandatory for two groups of patients before the procedure, even if transesophageal echocardiography (TEE) is routinely available in the procedure room: those with infection, to rule out vegetation in the right atrium; and those with heart failure, to define cardiac function. The patients are given antibiotics before the procedure. Antibiotic coverage ranges from a cephalosporin to combinations such as vancomycin and gentamicin, covering both gram-positive and gram-negative organisms. The author uses vancomycin and gentamicin in an attempt to cover most of the common organisms infecting implantable devices. If the infecting organism and susceptibilities are known, an antibiotic specific to the known organism is administered. Although some question the need for prophylactic antibiotics, most physicians believe they are beneficial. Da Costa and colleagues15 performed a meta-analysis that demonstrated a favorable effect of antibiotic prophylaxis in pacemaker implantation. There were seven available randomized trials for new implants or replacement pacemaker procedures. The incidence of endpoint events in the control groups ranged from 0% to 12%. The meta-analysis suggested a consistent protective effect of antibiotic pretreatment (P = .0046). Overall, the surgical literature emphasizes the importance of having the antibiotics administered so that good levels are present at the moment that the incision is made. This has become the standard of care. The patient is continuously monitored via ECG, an arterial pressure line, oxygen saturation, and Foley catheter. A reliable intravenous line should be placed. The author prefers a triple-lumen catheter placed in a femoral vein (or internal jugular vein) for all extraction procedures. In addition, a femoral arterial line is used

for patients needing continuous vasopressor support and for those being sent to an intensive care unit. Patients are prepared from chin to midthigh for all transvenous and cardiac surgical approaches, including an emergency cardiac surgical procedure, if needed. Anesthesia The types of anesthesia available include local anesthesia, conscious sedation, managed anesthesia care (MAC), laryngeal mask anesthesia (LMA), and general endotracheal anesthesia. The rationale for using a specific type of anesthesia is based on factors such as the type of procedure, physician comfort level with general anesthesia, perceived risk of a given type of anesthesia, and availability of general anesthesia. General Endotracheal Anesthesia General endotracheal anesthesia is the only type of anesthesia that is suitable for all procedures, and it is essential for some. General anesthesia consists of an “anesthesia package:” anesthesiologist, compliance with preoperative anesthesia protocols, anesthesia and monitoring machines, and general anesthetic agents and gases. It requires procedure room space, scheduling, and an anesthesia recovery room. In addition, the electrophysiologist (EP) and anesthesiologist must work as a team to manage the patient. Most surgeons acquire this teamwork skill early in their training; a medical EP may have to learn it. The merits of placing a patient at any desired level of anesthesia and providing a satisfactory environment to perform any type of surgical procedure are obvious and are accepted by all when presented in this abstract fashion. However, the practical demands of the anesthesia package and the fundamental questions relating to the safety of general anesthesia limit its use. The perceived risk of anesthesia causes insecurity and a lack of confidence. The potential risks associated with general anesthesia range from small to negligible. If the risk of anesthesia is combined with the risk of performing a more complicated implantation, a lead extraction, or a cardiac surgical procedure, the risk is not trivial. The procedure risks should remain separate from the actual technical risks associated with anesthesia. The procedure risk is associated with the EP procedure and patient management, including coordination between the anesthesiologist and the EP. The anesthesiologist is a trained professional in possession of certain technical skills. Giving anesthesia for EP procedures is complicated and stressful and frequently involves patients with significant compromise to their cardiovascular system. This type of procedure is not for everyone, just as cardiac surgical procedures are not for everyone. Consequently, choosing an anesthesiologist is an important decision, one that can potentially affect the outcome of the procedure. The anesthesiologist’s goal is to provide an optimal anesthetic for both the patient and the surgeon. These professionals are presented with patients whose physical status ranges from American Society of Anesthesi-

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

ologists (ASA) class I through V for the full spectrum of surgical procedures, and they often have the feeling that keeping the patient alive is their sole responsibility. In many cases, this feeling is justified. They do what is necessary to manage the patient, and give the surgeon only the information they think he or she needs to know. Even in cardiac surgical procedures, the anesthesiologist manages the renal, respiratory, and metabolic status in most situations and shares cardiovascular management issues with the cardiac surgeon and pump technician. In EP procedures, management is generally more challenging: there is no specific EP training for these procedures, the rhythm and filling pressures are changing continuously, and metabolic insults caused by low cardiac output are rapidly precipitated. To safely manage these events, the EP and anesthesiologist must communicate and share the management responsibilities. The EP cannot delegate this responsibility. The author’s experience with general anesthesia is positive and is presented here to highlight important issues. All procedures, including pacemaker implantation, are performed under some form of general anesthesia. The risk of anesthesia is negligible, if it exists at all. This statement requires a detailed explanation, because it is based on rigid guidelines for giving anesthesia, a patient management philosophy, and a defining of the responsibilities of the anesthesiologist and EP. There is no dispute regarding performance by the EP or anesthesiologist of technical procedures related to their specialty. However, protocols or guidelines for adjusting anesthesia for EP procedures are in order. The author strongly believes that use of central venous lines and continuous arterial pressure monitoring are warranted, and, once the patient is intubated, paralytic agents are discontinued. Also, fundamental decisions must be made with respect to patient management responsibilities. The need for central venous and arterial lines is intuitively obvious. Views on paralytic agents and management responsibilities require some explanation. Most anesthesiologists feel uncomfortable operating without paralytic agents. The indications for paralytic agents are intubation and relaxation of skeletal muscle. Fortunately, when the chest is opened, breathing can be suppressed with other agents. In some situations, the combination of paralytic and amnestic agents allows the anesthesiologist to decrease the level of inhalation agents, making it easier to manage a hemodynamically compromised patient. This approach results in less than optimal anesthesia, potentially masking serious hemodynamic issues that could cause cardiovascular collapse or other hemodynamic sequelae later in the case. For example, if a patient requires norepinephrine (Levophed) and/or high-dose dopamine while paralyzed and receives only an amnestic agent, marked vasoconstriction may develop, causing a low cardiac output, poor tissue perfusion, and acidosis resulting in cardiovascular collapse. Also, nerve stimulation from electrosurgery or pacing stimuli is masked by paralytic agents, increasing the risk for destruction or inadvertent chronic stimulation of the

869

phrenic nerve. Both of these complications can be prevented by avoiding paralytic agents. To give anesthesia without paralytic agents requires a change in philosophy and reliance on inhalation agents, narcotics, and short-term agents such as propofol (Diprivan). Separating anesthesia management from cardiovascular management is essential. The same level of anesthesia should be given despite the presence of cardiovascular issues. These issues are treated separately. For example, suppression of the myocardium and vascular dilatation are considered the “cost of giving anesthesia” and not a reason to modify the anesthetic regimen. It is not a complication to give an appropriate vasopressor by injection or infusion as needed to compensate for these expected events. It is the EP’s responsibility to manage the cardiovascular system. The EP’s actions during the procedure are continuously influencing the rhythm and filling pressures; the sequelae of these actions, superimposed on the intrinsic cardiac function, are best managed primarily by the EP. Many of the maneuvers performed during a lead extraction can reduce filling pressure. Traction on an atrial lead may block the SVC and reduce blood flow to the heart. Traction on a ventricular lead reduces the compliance of the chamber wall during diastole or, if strong enough, can pull the wall to the tricuspid valve, reducing blood flow. Immediate injections of a short-term α-adrenergic stimulant such phenylephrine (Neo-Synephrine) or Levophed constrict the cardiovascular system, causing an increase in both filling pressure and systemic blood pressure. This is frequently required throughout the case to compensate for these transient iatrogenic insults. Their use in no way reflects on the safety and efficacy of general anesthesia. In conclusion, the risk of anesthesia is related to the procedure and not the general anesthesia. Laryngeal Mask Anesthesia Laryngeal mask anesthesia is general anesthesia without the use of an endotracheal tube or paralytic agents. For less complicated procedures in which patients are allowed to breathe spontaneously, LMA is an excellent form of general anesthesia. Inhalant anesthesia is uniform, eliminating problems encountered with MAC or conscious sedation. For procedures such as pacemaker or ICD implantation, it is ideal. Contraindications are airway obstruction, history of esophageal regurgitation, peptic ulcer disease, and the need for TEE. At the end of the procedure, the inhalation agents are turned off and the patient wakes up. Managed Anesthesia Care and Conscious Sedation MAC is essentially conscious sedation managed by an anesthesiologist. Conscious sedation is anesthesia managed by the physician who is performing the procedure. The rationale for use of this type of anesthetic includes the following points: it is effective for most procedures, it removes the general “anesthesia package” from the procedure room, and the option of

870

Section Three: Implantation Techniques

general anesthesia is available. MAC and conscious sedation are satisfactory for less extensive procedures and for those procedures that seem less likely to involve a procedure-related complication. Short-acting drugs such as midazolam (Versed) and Diprivan, given intravenously, provide sedation and cause amnesia. Combined with a local anesthesia, they are effective for most procedures. The issue with this type of anesthesia is the maintenance of a uniform level of sedation. When patients become too “light,” they may behave like a “bad drunk,” becoming impossible to control and possibly making the procedure dangerous. When patients get too deeply sedated, they begin to have difficulties with ventilation and/or snoring. A potentially lethal volume of air may be sucked into the heart during snoring or deep breathing. If an air lock develops in the right ventricle, blood flow will be compromised. The cardiovascular system must be supported, and rolling the patient onto the left side may facilitate the passage of air out into the lung field. Compromise of cardiac output, perfusion to the alveoli, and air exchange in the lungs may be marked, requiring time to resolve. The most effective treatment is to pass a catheter into the right side of the heart and out into the pulmonary artery, sucking out the air as you go. In less severe cases, although it takes time to clear, support of blood pressure and high-pressure hyperventilation will help. The author has not had to take additional steps to resolve this problem. Local Anesthesia Performing device-related complication procedures using only a local anesthetic is a rare occurrence today. Once an intravenous sedative of any kind is given, it becomes conscious sedation. With the anesthetic agents available today, it is difficult to find an indication for using only local anesthesia. It is effective for local pain control in a normal tissue environment. The vascular environment associated with an acute inflammatory reaction diminishes the effectiveness, and the large doses needed increase the potential for an overdose complication. For example, high doses of lidocaine can cause a catatonic CNS reaction, incapacitating the patient and usually terminating the procedure. Procedure Room EP procedures are currently performed in general operating room suites, device implantation procedure rooms, EP procedure rooms, catheterization laboratories, and fully equipped cardiovascular operating rooms. The room must be large enough to support the procedure. Small procedure rooms are sufficient for a device implantation but not large enough for a complicated lead extraction procedure. There should also be space to accommodate emergency procedures and/or a more extensive surgical EP procedure. Procedures should not be performed in smaller rooms without a contingency plan for an emergency. The author’s vision of a complete procedure room suitable for handling the most complicated procedure

is described. Scaling down for lesser procedures should be intuitively obvious. The ideal procedure room should meet most of the requirements for an operating room, especially those requirements related to room cleaning, patient draping, gown and gloving, and instrument sterility. It should have the full “anesthesia package” including continuous monitoring of ECG, arterial pressures, and oxygen saturation. Specialty equipment such as fluoroscopy, TEE, pacemaker system analyzer, and other external EP devices to ensure pacing and defibrillation; electrosurgery; and an excimer laser and/or dedicated electrosurgical unit for lead extraction are essential. In addition, for minimally invasive cardiac surgical procedures, lighted retractors, access to thoracoscopy equipment, and an emergency tray to open the chest should be available. Additional safety devices to help protect physicians and nurses include lead drapes for radiation protection, smoke evacuators, and chairs for sitting when appropriate during the procedure. Physicians wearing lead aprons, thyroid collars, or a lead face shield have less musculoskeletal issues if they can sit during long procedures. Pocket Surgery Pocket surgery is a requisite skill that is essential for both implantation and explantation procedures. It is soft tissue surgery and centers on the creation, debridement, and abandonment of a pocket. Some pocket procedures, such as extensive debridement or revision, can be challenging for experienced surgeons. In extreme tissue debridement and reconstruction situations, it is the EP’s responsibility to provide an experienced surgeon and to ensure that the procedure is performed in a proper fashion. These procedures can be performed in any procedure room, with the patient prepared and draped in a sterile fashion. Instruments needed include basic surgical instruments, suction, sutures, ties and staples (optional), an electrosurgical unit, and an assortment of retractors. Also needed is an antibiotic irrigation solution. Two additional pieces of equipment that are useful for these extensive procedures are a lighted retractor to see clearly under subcutaneous tissue and muscle flaps, and a smoke evacuator (same as for laser and orthopedic surgery) to remove the airborne debris. In the author’s opinion, a smoke evacuator is helpful for extensive debridement of noninfected pockets and for all infected pockets. Pocket Location A pulse generator can, in theory, be implanted almost anywhere, limited only by anatomic and device-related constraints. Proved and popular locations are on the left or right anterior surface of the chest wall and in the upper quadrants of the anterior abdominal wall. Other locations are not used routinely because they are uncomfortable, are inconvenient, require special equipment such as longer leads, and/or have poor durability. Anterior Chest Wall. Implantation on the anterior chest wall is a natural location. It is close to the vein

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

entry site and is not influenced by body movement. For example, shoulder motion does not influence the pocket or the lead tunneled to the vein entry site. Because of the variations in chest wall anatomy, it is not always intuitively obvious where to place the pocket on the anterior chest wall. There are two general approaches. One approach is to make an incision convenient to the vein entry site and intended pocket location. The second approach is to uncouple the vein entry and pocket procedures. If the vein entry procedure cannot be performed from within the pocket incision, a separate vein entry incision is made. The author uses the second approach, placing the pocket in a predetermined location (Fig. 21-9). The medial aspect of the incision is made at the intersection of two landmarks. Imagine a horizontal line crossing at a point 5 cm below the sternal notch and a vertical line drawn inferior from the middle of the clavicle. Placement of the medial margin of the incision near this site compensates for variations in the chest wall and provides uniformity in pocket location. Also, for most chest wall configurations, the vein can be cannulated from within the incision using a percutaneous introducer approach. There are no rigid rules, but several principles should be followed. Do not place the pocket near the deltopectoral groove. Frequent contact between the head of the humerus and the pulse generator causes pain. Do not place the pocket immediately below the clavicle, because the pulse generator’s contact with the clavicle causes pain. A minor consideration is a location remote from the vein entry site. If the leads are long enough and the tunnel is deep, a remote site is not a problem. Caution should be used in placing the

Figure 21-9. Skin incision site. Placing the skin incision in a uniform location on the chest wall is difficult because of the variations in the chest wall. The sternal notch and the clavicle are two relatively constant landmarks that can be used for reference. Placing the medial aspect of the skin incision at the intersection of a horizontal line approximately 5 cm below the sternal notch and a vertical line extending down from the clavicle is one technique to achieve uniformity.

871

pocket beneath the breast in a female or beneath the nipple in a male patient. The anatomic considerations for this location are a potential issue. Part of the pocket will be below the origin of the pectoralis muscle on the chest wall, and in males the subcutaneous tissue may be of insufficient thickness. This is not an issue for females, and many young women request that the generator be placed in the tissue space beneath the breast for cosmetic reasons. Regardless of gender, the pulse generator must be carefully anchored, especially in women, to prevent migration and erosion through the skin below the breast. Subcutaneous Tissue Pocket. Historically, the classic location is in normal tissue on the anterior superior portion of the chest wall, in the relatively avascular fascial plane between the subcutaneous tissue and the surface of the pectoralis major muscle. The pocket is made with sharp and blunt dissection, and/or by electrocautery. If the fascia is cut or torn, the muscle frequently separates in the direction of its fibers, exposing the intramural portion of the muscle. Torn muscle should be sutured back together using an absorbable suture. Separation of the muscle fibers is the most common cause of bleeding into the pocket, and it is caused by the small arteries that run in the direction of the muscle fibers on or just beneath the surface. For safety, all arterial bleeding sites in the subcutaneous tissue should be suture-ligated. The pocket can migrate within the fascial plane. A suture securing the generator to the muscle prevents migration within the plane. It usually takes a traumatic or vascular event for it to break out of the plane and migrate to the surface. Submuscular Pocket. Construction of a pocket in the tissue plane posterior to the pectoralis major and anterior to pectoralis minor is another natural location. Historically, submuscular pockets were used as an alternative location. Indications were related to the size of the generator, the condition of an old pocket, erosion of an old pocket, and cosmetic considerations. Today, the indications are essentially the same. If the thickness of the subcutaneous tissue is considered inadequate for the size of the pulse generator, the physician’s concern for migration, erosion, or ischemic injury may motivate the change to a submuscular pocket. During reimplantation of a pulse generator, issues may be discovered with the old pocket, such as exuberant fibrous tissue, granulation tissue, calcification, or a gelatinous material from an old clot. If the pocket is not suitable for reimplantation after debridement, a submuscular pocket is used. Erosion with secondary infection is one of the most common complications. Once the infection is treated and the old pocket abandoned, the new pocket is frequently placed beneath the muscle at a remote site as protection against erosion at the new site. Today, pulse generators are smaller, but many patients do not want the device showing on their chest wall. A submuscular pocket may still show a bulge, but the outline cannot be seen. Cosmetic concerns are becoming more common. Four approaches are described to construct a submuscular pocket; these are the clavicular, second

872

Section Three: Implantation Techniques

intercostal, third intercostal, and axillary approaches. The author uses the second and third intercostal approaches. The standard pocket incision is over the second intercostal space and is the preferred approach. An incision over the third intercostal space is used for pocket creation and epicardial lead implantation. An incision below the clavicle, near the vein entry site, is at a natural separation between the clavicular and sternocostal origins of the pectoralis major muscle. The lateral approach is a plastic surgical approach. An incision is made in the axilla, and the pocket is created by a lateral dissection onto the chest wall. For the clavicular, second intercostal, and third intercostal approaches, the pectoralis muscle is split in the direction of its fibers, using blunt dissection. The pocket is constructed in the avascular plane beneath the pectoralis major muscle and anterior to the pectoralis minor. The pectoralis minor helps to prevent the projection of the pocket into the axilla when the arm is raised. In the clavicular approach, the muscle is separated between the clavicular and sternocostal portion of the pectoralis major muscle, exposing the same avascular plane beneath the pectoralis major and minor. This is a satisfactory approach if the pocket is placed inferior enough to prevent the pulse generator from being pushed superiorly against the clavicle and laterally into the deltopectoral groove. The pulse generator must be secured to full-thickness muscle to prevent migration. The axillary approach is a surgical variation using a lateral dissection into the same fascial plane. Short-term cosmetic issues need to be convincing for the author to use this approach. Abdominal Pockets. Abdominal pockets are still used today. In the past, large pulse generators were routinely placed into abdominal pockets. Once their size was reduced, they met the subjective criteria for a pectoral pocket near to the vein entry site. However, there are still indications for an abdominal pocket: infection, mastectomy, injury, pain, convenience, and patient preference. The pockets for short-term epicardial implants in patients being treated for vegetative endocarditis are usually placed in the abdomen. Pockets for femoral vein implants are also placed in the abdomen. Leads are tunneled from the implant site to the abdominal pocket. Tunneling should be deep in the subcutaneous or submuscular tissues whenever possible. Tunneling substernal, within the chest and/or pericardium, is also acceptable, especially for cardiac surgical procedures. For example, transatrial and epicardial leads can be tunneled to any convenient location through the pericardium and/or chest. The same rules apply to subcutaneous tissue implants in the abdomen as in the pectoral region. If the subcutaneous tissue is adequate for the pulse generator, the pocket is best placed in the fascial plane between the subcutaneous tissue and muscle. The pocket is usually placed in the left or right upper quadrant. The pocket must be placed inferior to the costal margins and away from the belt line. This is relatively easy to do, because body motion and position are predictable in the upper quadrant. Placement in a lower

quadrant is necessary for femoral implants. Lower quadrant pockets need to be selected with care. Body motion, especially positions such as bending and sitting, influence these pockets. Placement away from the iliac crest prevents painful contacts. In some cases, pockets cannot be placed out of harm’s way in the lower abdomen because of the extreme ranges of pocket motion. Complications seen by the author with lower quadrant pockets include pain and erosion of the pocket into the abdomen. If a lower quadrant implant is necessary (e.g., femoral vein implant), awareness and avoidance of these issues are essential. A submuscular pocket in the abdomen can safely be made under the rectus abdominus muscle. A device under the muscle in any other location can erode into the abdomen. It is safe to place the pocket within the rectus sheath, under the rectus abdominus muscle in the upper quadrant. The sheath is strong enough to prevent erosion into the abdomen. This is not true for the lower quadrants. The plane between the muscle and the posterior sheath in the upper quadrant is traversed by large neurovascular bundles, passing from the abdomen to the muscle. To prevent a hematoma, these bundles should be suture ligated. An arterial bleed causes a dramatic hematoma. If it is not corrected immediately, the pressurized pocket enlarges by dissection within the rectus sheath or decompresses by rupturing through the incision. Also, do not open the posterior rectus sheath; it could result in herniation of the omental apron or bowel into the pocket. Tears or cuts must be closed with suture material. Reimplantation or re-exploration in an abdominal pocket is a potentially dangerous operation. The inflammatory reaction involving the posterior sheath frequently causes exuberant fibrous tissue reactions, especially in large pulse generator pockets. Attempts to modify and/or debride this suboptimal pocket is technically challenging. Misadventures resulting in loss of integrity of the posterior wall can damage the colon and small bowel contiguous to this site. Loss of integrity of the colon or small bowel wall requires an extensive emergency surgical procedure such as a colostomy. All device hardware must be removed from this site and temporary devices implanted at a remote site. Loss of integrity of the colon or small bowel that is not recognized during the procedure can cause a lethal infection. All defects in the wall must be repaired to avoid herniation. Consequently, it is recommended that only surgical EPs, qualified surgeons, or medical EPs with extensive experience perform these procedures. The pocket is closed with the use of conventional surgical techniques. First, the pocket is irrigated extensively with an antibiotic solution. If a submuscular pocket was used, the pocket is closed with permanent suture material. The subcutaneous and subcuticular tissues are then closed with suture material that dissolves by hydrolysis (e.g., Dexon, Vicryl). Skin closures are made according to the physician’s preference, and almost anything works. Subcuticular closures have plastic surgery appeal. They work well as long as there is no reaction to the suture or infection associated with the suture. In those cases, the suture line becomes red, swollen, warm

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

and may or may not become suppurative. Sutures are placed perpendicular to the incision and are, in effect, a sinus tract through the subcutaneous tissue that can cause a pressure injury to the surface of the skin. Staples are an alternative and are used by the author. Staples are fast and can be applied to any incision regardless of the condition of the tissue. An occlusive dressing should be applied until the suture line seals (about 24 hours). Liquid skin adhesives are used by some; the author does not believe they are acceptable for closure when there has been a pocket complication. Pocket Complications Pocket complications, summarized in Table 21-1, include pocket hematoma, wound dehiscence, migration, erosion, pain, and infection. Hematoma formation and wound dehiscence are acute events and are usually related to implantation technique. A pocket hematoma may form late in anticoagulated patients, especially those treated with heparin or enoxaparin (Lovenox), or if the pocket is subjected to trauma. Migration, erosion, and pain are related to device–tissue interaction. Infection is caused by contamination of the pocket and is associated with implantation technique and tissue made susceptible by an abnormal device–tissue interaction. Pocket Hematoma. A hematoma that develops immediately after an implantation procedure is one of the

TABLE 21-1.

873

most common complications associated with a device implant. Although this is a technique-related complication, experienced implanters occasionally have difficulty obtaining hemostasis. Three conditions predispose to hematoma formation: a tear outside the fascial plane, arterial bleeding, and extrusion of venous blood back along the leads and into the pocket. Tears outside the fascial plane were discussed in the section on pocket construction. Arterial bleeding within the pocket causes the most dramatic hematoma and should be considered an urgent or emergent situation, depending on how fast the hematoma is expanding. A hematoma develops rapidly; if not corrected immediately, the pressurized pocket enlarges by dissection into the tissue planes or decompresses by rupturing through the incision. The arterial pressure can cause extensive tissue dissection, enlarging the pocket area multiple times. If a larger artery is torn, the expanding hematoma can contain more than 500 mL of clot and blood. If it is left unchecked for a period of time, the tissue pressure and resultant ischemia can cause tissue necrosis and/or disruption of a recent incision. A small artery running in the direction of the pectoralis muscle fibers, on or just beneath the surface, and the small artery running parallel to the cephalic vein in the deltopectoral groove are the two most common causes of arterial bleeding. External forces applied to the pulse generator can cause a traumatic rupture of the pectoralis muscle, causing

Pocket Complications

Complication

Predisposing Factors or Causes

Treatment

Pocket hematoma

Tear outside fascial plane Arterial bleeding Extrusion of venous blood along leads

If it is large enough to be palpated: Remove clot and debris Obtain hemostasis Reduce pocket size if necessary Use closed drainage system if hemostasis is difficult to achieve Avoid repeated needle aspirations

Wound dehiscence

Excessive stress on suture line by hematoma, hemorrhagic effusion, or trauma Error in surgical technique

Immediate: attempt to salvage site Delayed: treat as infected pocket

Migration

Unknown

No treatment unless another complication exists

Erosion

Device implanted outside correct plane Sustained insult forcing device out of correct plane (compromised blood supply, trauma, sequestered effusion)

Before pocket sticks to skin: debride and relocate pocket After pocket sticks to skin or skin is broken: treat as infected (abandon site)

Pain

Unknown

Relocate pocket if necessary

Infection

Perioperative contamination Chronic site may have poorer defenses against infection Metastatic (seeding from remote infection or procedure such as teeth cleaning) Chronic occult infection becomes acute Note: pocket infection may manifest as respiratory distress if infection decompresses into venous system

Most infections: Antibiotic treatment and abandon site (removing device and leads) If no septicemia, no inflammatory buildup around leads near insertion site, and >2.5 cm from pocket to suture sleeve, antibiotic treatment and abandonment of pocket may be sufficient

874

Section Three: Implantation Techniques

rupture of an artery when a sedated patient changes position in bed or uses the ipsilateral arm to lift a heavy object. Blood forced retrograde out of the implant vein along the leads and into the pocket causes a hematoma. Although the pocket may be dry, an elevated venous pressure from heart failure, Valsalva maneuver, or coughing can extrude blood along the leads, filling the pocket with venous blood. With an introducer approach, extrusion of blood is prevented by placing a suture around the leads at the muscle entry site. If the cephalic vein is used, a suture around the vein and lead (or leads) at the entry site will suffice. The suture also prevents debris collected within the pocket from being forced back into the venous circulation. Once clot forms in the pocket, it is subjected to both lysis and organization. Lysis creates particulate debris, increases the osmotic pressure, and pulls fluid into the pocket, thereby creating a hemorrhagic effusion. As the hemorrhagic effusion increases, the resultant tension on the pocket wall continues to enlarge the pocket by dissection or ruptures a recent incision. Both of these complications require immediate surgical intervention for correction. Excessive granulation tissue, formed during the organization of a large clot, may be present years later. In addition, only the surface of a massive clot may organize, leaving a residual gelatinous portion of partially organized clot debris. These pockets are not healthy, behave similar to pockets with exuberant fibrous tissue, and should be debrided. A pocket hematoma that is large enough to be palpated should be treated. The only successful treatment is immediate pocket exploration. Prolonged observation and procrastination must be avoided. This is especially true if the pocket wall is tense. Wound dehiscence with spontaneous evacuation of the hematoma can occur if the pressure generated by the effusion is great enough. Other complications of an untreated hematoma include chronic wound dehiscence, migration or erosion, and infection. The goal is to remove all clot and tissue debris and obtain hemostasis. The pocket must be opened and the hemorrhagic effusion, clot, and tissue debris removed. Needle aspiration is ineffectual and dangerous, because clot cannot be aspirated through a needle, and percutaneous needle puncture is a potential source for introduction of bacteria. If the pocket has been enlarged, the dissected region should be excluded with suture material, leaving only an appropriate-sized pocket. If adequate hemostasis is difficult to achieve, a closed drainage system (e.g., Jackson-Pratt) is placed in the excluded pocket to prevent the hematoma from reoccurring. In the author’s experience, the immediate surgical correction of a hematoma and placement of a closed drainage system do not adversely affect the healing of the pocket. Wound Dehiscence. Wound dehiscence is a rare event. It occurs within the first few days or weeks after implantation (Fig. 21-10). During the acute wound healing phase, suture material is required to reapprox-

Figure 21-10. Wound dehiscence. A wound dehiscence left unattended for more than a week. The pacemaker pocket developed a hematoma that evacuated spontaneously, rupturing the incision. At this point, the pocket is considered infected and is treated as such.

imate and reinforce the tissue. Most wound dehiscences are caused by excessive stress placed on the suture line by a hematoma, effusion, or trauma. Traumatic disruption is rare. Dehiscence, without a predisposing cause, is caused by an error in surgical technique. Treatment consists of salvaging the site by intervening immediately (within hours) after a dehiscence, similar to treating a hematoma. It is usually successful and, considering the consequences, worth the attempt. A delayed intervention allows gross contamination, and an infection is likely to develop. If intervention is delayed, the case is treated as an infected pocket, with the removal and discard of expensive leads and pulse generator, debridement of the pocket, insertion of a closed drainage system, and abandonment of the pocket. Migration. Migration is the movement of a device through the surrounding tissue. In the past, migration was more common due to the large size and pointed shape of some devices, even when contained within a Dacron (Parsonnet) pouch. Most migrations are slow, occurring over a period of years; most move in an inferior-lateral direction and do not cause a complication. The author does not know the exact mechanism for migration, but size and weight of the pulse generator are factors. One possible scenario relates to the motion between the device and the musculoskeletal system, which creates directed forces (pressures) that compress and/or stretch the fibrous capsule and surrounding tissues. This would cause a low-grade inflammatory reaction, initiating a cycle of fibrous tissue lysis and formation (organization), with remodeling of the tissue to relieve the stress on the wall. This could result in migration of the pocket through the tissue. A migrated device is not usually treated unless the potential for an actual complication exists. Erosion. Erosion is the exteriorization of the device after loss of skin wall integrity. Infection is the most

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

common cause of erosion and is discussed later. For all other erosions, the pacemaker pocket sticks to the skin before erosion occurs (pre-erosion). Before the pocket becomes adherent to the skin, it moves freely. At this point, if infection is not present, debridement and relocation of the pocket are usually successful. However, once an inflammatory reaction begins, the skin sticks to the pocket, skin integrity is lost, and bacteria cross the skin, contaminating the pocket. Pockets treated after the skin sticks or after frank erosion are treated as infected pockets.

tissue is not sufficient to prevent free migration when the device is subjected to an applied force. Another factor is significant weight loss. Loss of the subcutaneous tissue barrier between the pulse generator and the skin can result in erosion, especially if the pocket is subjected to some form of trauma. Most of the eroding pockets seen by the author were placed in the subcutaneous fatty tissue away from the tissue planes. Pacemaker pulse generators placed under the pectoralis major muscle sometimes protrude into the axilla when the muscle is flexed. The resulting migration into the axilla can cause pain or erosion. This complication is avoided by placing the device more medially under the pectoralis major, over the pectoralis minor, and anchoring the device by placing a suture through the full thickness of the muscle. Compromise of the blood supply causes tissue loss. Blood supply is compromised by mechanical factors and by exuberant fibrous tissue, which causes pressure and/or constricts the entrapped vessels. If the blood supply to a region of subcutaneous tissue and skin is compromised, the resultant dissolution of the subcutaneous tissue (pre-erosion) is caused by lack of nutrition (Fig. 21-11A). If this is severe enough, tissue necrosis develops, leaving only ischemic or gangrenous skin (see Fig. 21-11B). Trauma and sequestered effusion may rupture the fibrous capsule barrier around the device. A traumatic rupture is intuitively obvious, resulting in device migration through the rupture site. A sequestered effusion or a hematoma can generate sufficient pressure to erode through the subcutaneous tissue and skin, draining to the outside like a decompressing abscess. Other con-

Implant Fascial Plane. Flat devices, such as pulse generators, do well when implanted in a natural tissue plane (fascial plane) between two tissue types (subcutaneous tissue and pectoralis muscle or pectoralis muscle and chest wall). Minimal pressures are exerted by the flat surfaces against the two tissue types. The device’s curved surfaces apply the greatest pressures. It may migrate in the plane, but it will not erode. For an erosion to occur, the pulse generator would have to be implanted with a portion of a curved surface extending out of the fascial plane, or it would have to sustain an insult forcing it out of the plane. Such insults include compromise of the blood supply with loss of tissue, trauma with tissue disruption, and a sequestered effusion applying pressure to the pocket wall. In these situations, erosions occurring within the first few months after an initial implantation are not uncommon. A current trend is to place the pocket within the subcutaneous tissue. This pocket must be perfect and without applied stress to prevent migration and erosion. Anchoring the pulse generator to the soft

A

875

B Figure 21-11. Compromised blood supply. Compromise of the blood supply to the subcutaneous tissue and skin causes a decrease in the supply of nutrients to the cells. A, In the early stages (preerosion), the fatty tissue begins to lose mass. During the latter pre-erosion stages, the tissue becomes ischemic, and skin changes occur. The last stage before erosion is adhesion of the skin to the fibrous tissue pocket. B, Complete loss of blood supply to the skin results in gangrene of the skin before erosion.

876

Section Three: Implantation Techniques

tributing mechanisms include tissue ischemia and an intense inflammatory reaction. Pain. Occasionally, patients complain of pain in or near the pocket. History and physical examination are the only diagnostic tools available. Pain is a complication regardless of the cause. Pain is a difficult complication to manage. Acutely, some patients heal without any complaints of pain. Others have severe pain. The cause of pain may be nerve entrapment, inflammation of the scar tissue, migration, or injury to the musculoskeletal system. Pain associated with no physical signs may be related to nerve entrapment. Because nerves cannot be seen, a diagnostic exploration of the pocket is not an option. The best chance for a successful treatment is to remove the generator and leads, abandon the site, and reimplant on the opposite side. A worst-case scenario for a normal healing process is a patient involved in strenuous physical activity. This type of patient should be given more time for postimplantation pain to subside. If the pain is caused by such activities, it will subside when the body adjusts to the applied stresses causing the pain. The author believes that patients are going to subject themselves to these activities at some point and encourages them to continue their activities. If a normal activity causes pain, the patient is treated with pain medication and the activity is allowed to continue. If the pain does not subside, the pocket is relocated; a subcutaneous tissue pocket is changed to a submuscular pocket, or vice versa. The pocket area must eventually be made pain free, regardless of the patient’s activity. Limitation of activity is not an option. Chronic pain is not normal, and complaints of pain should be taken seriously. An implanted device is like a watch: the patient should be aware of its existence only when something draws attention to it. Pain on the chest wall can cause other symptoms. For example, chronic chest wall pain can cause muscle spasm. Inflammation of the scar tissue is usually obvious. The scar is exuberant, red, and painful to the touch. It has the appearance of a red keloid. Injection with steroid reduces the inflammation and will eventually alleviate the symptoms. Pain associated with injury to the musculoskeletal system can be related to the initial placement of the pocket, migration, or trauma. Pain is usually caused, not by the migration per se, but by some traumatic event associated with the new anatomic position. Trauma to a rib or at the costochondral junctions is the most common pain complaint. This pain can be similar to a broken rib: intense and even debilitating. Injury to the pectoralis major muscle from sutures, tears, or erosions into the muscle can also cause pain. In most cases, the pain is typical of muscle spasm and the insertion site on the humerus is tender (bursitis). Spasm of the neck muscle is not uncommon in this situation. This pain usually goes away when the muscle injury heals. Placement of the pocket near the deltopectoral groove can injure the head of the humerus, mimicking bursitis. A similar, but less specific cause of pain is

trauma to the tissue beneath the clavicle. Pocket relocation to another subcutaneous area or placement beneath the pectoralis major muscle may be effective. Tissue Debridement Tissue debridement is the surgical removal of all inflammatory and damaged native tissues, leaving only normal native tissue behind (Fig. 21-12). The need for tissue debridement in normal pockets seen on routine reimplantation is minimal. On opening an old pocket, the debridement goal is to remove any exuberant inflammatory tissue (fibrous tissue), leaving only thin healthy fibrous tissue (biophysical interface) or normal native tissue behind. The fibrous tissue present is involved in the chronic remodeling inflammatory reaction; rarely is an acute inflammatory reaction present. This is important, because exuberant fibrous tissue masses, which are the result of an inflammatory reaction, can injure adjacent tissue, continuing the inflammatory reaction (recursive reaction). Also, if this tissue is contaminated by bacteria, it becomes a nidus for infection (bacteria stick to the smooth surface of the exuberant encapsulating tissue, and the body’s immunodefense mechanisms cannot reach the bacteria). Leads entrapped in the encapsulating fibrous tissue may be under undue stress if the pulse generator is not placed back in the same position. Tissue debridement and freeing of the leads rectify this situation. Initially, tissue debridement was performed using a scalpel. This was tedious, bloody, and time-consuming, deterring all but the most dedicated. Fortunately, modern electrosurgical units provide the frequency options and control of power needed to cut and coagulate in an efficient, tissue-friendly manner. Tissue debridement with a modern electrosurgical unit is recommended. Tissue debridement, even in infected cases, is rarely extensive and can be treated as described earlier. However, there is occasionally a situation, usually involving an infection, in which the pocket must be extensively debrided and then abandoned. If an acute inflammatory reaction is extensive, debridement is tedious because of the presence of acute and chronic inflammatory material and/or proximity of large blood vessels and important nerves. Knowledge of local anatomy is mandatory in this situation. In some cases, the inflammatory tissue can be more than 2.5 cm thick, extending above and below the pectoralis major muscle with fingers to the clavicle, and damaging a large amount of skin. In these cases, skin loss is significant, and reapproximation of skin edges is challenging. Once the debridement is complete, hemostasis is difficult to achieve, especially on the muscles. Whenever possible, muscle fascia should be reapproximated. Tissue Closure Healing by primary intention is the closing of an open wound by reapproximation of tissue (muscle, subcutaneous tissue, and skin) using suture material. The suture material holds the tissue in place until the tensile

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

877

A

B

C

Figure 21-12. Debridement of a chronically infected pocket. A, Chronic draining sinus tracking to an infected pocket lined with granulation tissue. B, Surgical debridement. To close the pocket, with healing by primary intention, the pocket must be completely debrided of all inflammatory material, including the tissue encapsulating the lead near the vein entry site. The excised mass includes a thin layer of normal tissue, organizing inflammatory tissue protecting the normal tissue, and the inner surface of granulation tissue. The swab stick marks the location of the sinus track. C, Extravascular encapsulating fibrous tissue. Leads are dissected free, demonstrating the tissue surrounding the leads.

strength of the bonding fibrous tissue produced by the inflammatory reaction is sufficient to permanently keep the tissue together. All initial and reimplanted pockets are closed in a conventional fashion by reapproximation of the tissue with suture material and allowed to heal by primary intention. Chronic pockets are debrided of all exuberant inflammatory material before closure. The author also closes those pockets that are not suitable for reimplantation, because of inadequate tissue or infection, before abandoning the pocket. These pockets are debrided and closed using a closed drainage system (Jackson-Pratt) placed in the debrided pocket, applying suction to prevent the development of effusions or clot, and keeping normal tissue contiguous with normal tissue. This is the author’s method for closure of all complicated pockets and has been successful. The author has not left a pocket or any other type of soft tissue wound open since the early 1970s. A large defect, resulting from loss of tissue during debridement, is sometimes a challenge to close. In extreme cases, a major defect requires an experienced surgeon or a plastic surgeon, especially if a flap is needed. Skin grafts are not used, because the areas of tissue loss are made permanent, creating cosmetic issues. Another philosophy is that pockets that have debridement defects or are infected should be left open. Once the device (foreign body) is removed, the infection will heal, so why not let it heal by secondary intention? This rationale shortens the procedure and is an accepted method of healing. Infection is not an issue,

and it avoids the need for acquiring surgical debridement and closure skills. Also, a qualified surgeon may be hard to find on short notice. Concerns such as morbidity, extensive healing time, long-term antibiotic therapy, the requirement for constant professional care of the wound, and the possible need for some form of surgical intervention, including skin grafting, are considered the natural cost of healing the wound. Healing by primary intention, on the other hand, does not have these issues. Healing by secondary intention was used a long time ago as the only way to heal an open wound. The healing stages were well documented: suppuration, granulation, closure of defect, and, finally, skin closure. This was the recommended method of treating contaminated wounds by surgeons up and until the 1980s. Since the 1970s, however, primary closure of debrided infected pockets has been successfully and almost exclusively used to manage device-related infections. Lead Extraction Skills Lead extraction is a fundamental skill that is required to manage device-related complications. Lead extraction, like lead implantation, is a requisite skill with predictable and expected results. This was not always the case. From its inception in the early 1980s until the late 1990s, the procedures evolved rapidly. The management of a device-related complication centered on the lead extraction procedure, overshadowing all other

878

Section Three: Implantation Techniques

aspects of management. This was because unexpected tears in the SVC and heart wall sometimes occurred without warning, despite the rigid protocols followed. The technology and those rigid protocols have evolved into today’s procedures. The procedures are less stressful and have predictable results. Predictability allows an extractor to recognize an approach that has a potential for a bad result and change to an approach with a predictably good outcome. Lead extraction is discussed in its entirety in this chapter, including indications, extraction techniques, and medical/surgical EP approaches. A goal is to show that once lead extraction is mastered, it becomes a requisite skill and, like lead implantation, a routine component of the procedure. Indications range from the author’s simplistic list to a more extensive list with conditional and subjective statements. An attempt is made to add perspective to this controversy. Extraction techniques are much simpler today and are explained in detail. Unfortunately, some of the old, more complicated techniques are also needed to manage an occasional rare situation. Consequently, some of these techniques are presented in the same detail as the more modern techniques. Presentation of the extraction techniques and approaches from the medical and surgical EP’s point of view has been challenging. The common ground is that most of the extractions involve transvenous leads. These are extracted using the techniques and approaches common to both groups. The alternative approaches are sometimes different for surgical and medical EPs. For example, in addition to the transvenous approaches, surgeons have the option of a transatrial or epicardial approach. Although these surgical approaches are not mainstream, they are essential to the management of certain complications. Though medical EPs cannot perform these procedures, they must be able to direct a non-EP surgeon enlisted to perform the procedure. Indications for Extraction There is considerable divergence of opinion in regard to the indications for lead extraction. In addition, there has been significant evolution of the tools, techniques, and number of experienced physicians since the lead extraction policy conference was convened in May of 1997.16 The author’s indications for lead extraction are the presence of a device-related infection, creation of a conduit, and superfluous leads. For the indications to be accepted, the risk of not extracting a lead in a specific patient must be greater than the risk of extracting (risk-benefit ratio). The author’s earlier classification of indications into mandatory, necessary, and discretionary indications reflected the magnitude of the risk (Fig. 21-13).17 The author’s current descriptive classification of indications into infection, creation of a conduit, and superfluous leads does not reflect actual or potential risk. The relation between the two is straightforward. Infection is a mandatory indication, and creation of a conduit is a necessary indication. Superfluous leads, based on the earlier discussion, become a discretionary indication. The change from a risk-based to a descriptive indication evolved over the

A

B

Figure 21-13. Indications for lead extraction. Indications are difficult to describe because of the complexities associated with device-related complications and the procedure-related risk. A, Magnitude of risk. The author’s first attempt was to classify indications based on the magnitude of the perceived risk. This approach presents a classification based on risk of the complication relative to the perceived risk of the procedure. B, Descriptive. The author’s second attempt is a descriptive classification based on three clinical situations. The rationale for these indications is clinical and is not based on procedure risk. The author assumes that the risks are negligible for an experienced physician trained in the management of device-related complications.

years as the risk and morbidity associated with lead extraction have become less dominant factors. In other words, the concept of “arm twisting” to justify the risk of extracting leads has changed to the concept of “managing an actual or potential complication.” The potential dangers associated with the procedure have not changed. However, today, a large number of procedures have been performed, and an individual extractor’s experience is known. The risk of extraction decreases to an acceptable level with training and experience. Consequently, the focus has shifted from risk to management of the potential or actual complication. Infection. Infection is an intuitively obvious indication for lead extraction. Most physicians accept the hypothesis that antibiotic therapy, pocket debridement, and local relocation are palliative and that all device components, including leads, must be removed to cure the infection. Also, the morbidity associated with a local pocket infection, the lethal sequelae of septicemia, and the potential risk of infected thrombus formation in the heart are well known. Because leaving an infected lead in the body is potentially lethal, the risk of the procedure is clearly less than the risk of lead extraction (i.e., the risk of not extracting far exceeds the risk of extracting). The risk of S. aureus device infection without extraction was supported by a series of 33 patients from the Duke Medical Center, in which 10 (47.6%) of 21 patients died without lead extraction, and 2 (16.7%) of 12 died despite lead extraction, and none from lead extraction.18 The safety and efficacy of complete lead extraction, with debridement and delayed reimplantation at a remote anatomic site, were demonstrated in 123 patients at the Cleveland Clinic Foundation with device infection. Despite infections with a wide range of bacterial organisms, mostly coagulase-negative staphylococci and S. aureus, extraction was associated with no major complications. Infection reoccurred only in those four patients who had incomplete extraction or reimplantation concurrent with the extraction.19

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

879

A

B

C

D

Figure 21-14. Creation of a conduit. A, Atrial lead failure caused by a conductor coil fracture and an occluded ipsilateral subclavian-brachiocephalic vein. B, Extraction of the old atrial lead and implantation of a new lead through the extraction site (two functioning leads). C, Insertion of a new lead through the contralateral subclavian vein, tunneled to the old pocket (two functioning leads and one superfluous lead). D, Insertion of an atrial lead, a ventricular lead, and a new pocket on the contralateral side (two functioning leads and two superfluous leads).

Creation of a Conduit. Creation of a conduit is a more subtle indication for lead extraction. The rationale for this indication is applicable to component failures. An example is the best way to define this problem. Consider a patient with a dual-chamber pacemaker and an atrial lead conductor coil fracture. This patient has a normal ventricular lead, an occlusion of the brachiocephalic vein, and the need to implant a new atrial lead. The only way to insert the new lead through the same vein entry site is to extract one of the old leads and reinsert two new leads through the extraction conduit (Fig. 21-14A and B). This same logic would apply to other situations, such as the addition of a new lead (Fig. 21-15A and B). For example, a patient with a dual-chamber ICD needs a cardiac vein implant for biventricular pacing. The alternatives are doing nothing, implanting the new lead through a contralateral vein or a transfemoral vein, and using a cardiac surgical approach. Sometimes, when there is severe stenosis with or without symptoms from the obstruction of flow, physicians have initiated balloon venoplasty and stenting without extraction of the leads. This produces a particularly difficult scenario if either infection or reocclusion occurs, because extraction now becomes impossible without extensive open surgery. A more appropriate approach includes extraction, venoplasty, stenting, and reimplantation through the stent, as reported by Chan and associates20 in a subclavian occlusion that progressed to an SVC occlusion.

Doing nothing is an option in some cases, such as with patients who rarely use their device, who have chronic atrial fibrillation or flutter, who have complete heart block, or who have high filling pressures where the atrial contraction has negligible effect. In these patients, loss of an atrial lead will go unnoticed, cause occasional palpitations, or result in an unnoticed chronic compromise in hemodynamic function. If the example had been loss of the ventricular pacing lead in a patient with frequent ventricular pacing, doing nothing would not be an option. Implantation through a contralateral vein seems logical, especially if it is the implanter’s only skill level option. The simplest alternative is to abandon the two old leads on the ipsilateral side and implant the new leads on the contralateral side. In the example of a conductor coil fracture, the patient has the risk of having two functioning and two superfluous leads in the heart and the risk of instrumentation of the superior veins on the opposite side (see Fig. 21-14D). In the example of adding a biventricular pacing lead, the patient has the risk of having three functioning leads and two superfluous leads in the heart and the risk of instrumentation of the superior veins on the opposite side (see Fig. 21-15D). These risks are weighed against the risk of extracting the one atrial lead. The resultant risk of having a complication is the sum of the individual risk factors. For these two examples, another approach would be to implant the new atrial lead or biventricular lead

880

Section Three: Implantation Techniques

A

B

C

D

Figure 21-15. Creation of a conduit. A, Addition of a cardiac vein lead in a patient with an implantable cardioverter-defibrillator (ICD), an occluded subclavian-brachiocephalic vein, and heart failure. B, Extraction of the atrial lead and implantation of a new atrial and cardiac vein lead (three functioning leads). C, Insertion of the new lead through the contralateral subclavian vein, tunneled to the old pocket (two functioning leads and one superfluous lead). D, Insertion of a new atrial lead, ICD lead, cardiac vein lead, and new pocket on the contralateral side (three functioning leads and two superfluous leads).

on the contralateral side and tunnel it across to the pulse generator and ventricular lead. In addition to the risk of instrumentation of the superior veins and tunneling, the patient with conductor coil fracture will have two functioning leads and one superfluous lead in the heart (see Fig. 21-14C), and the patient with the biventricular lead will have three functioning leads in the heart (see Fig. 21-15C). In the author’s opinion, the combined risks associated with not extracting often exceed the risk of extracting in both these cases, but clearly the risks will vary in individual patients. Total bilateral occlusion of the superior veins further complicates the problem. The transfemoral approach is the only approach available to the medical EP. The surgical EP has the options of a transatrial or epicardial approach. The risks associated with these choices make the decision to extract the atrial lead easier. However, without extraction skills, the medical and surgical EPs may choose these alternatives. The risk/benefit ratio is crucial to the rationale for extracting these leads. For example, the life-threatening risk associated with infection in effect forces an EP to extract the lead and abandon the pocket. The risk of not creating a conduit to insert new leads is a potential risk for a future complication related to bilateral implants, superfluous leads, multiple implanted leads, and/or tunneling. This risk is obviously less than the

life-threatening risk of infection. In the situation of lead failure, the alternatives presented provide an acceptable short-term solution and can be performed by implanters without lead extraction skills. Potential risks are not as compelling a reason for action as the immediate risks associated with lead extraction. However, physicians with experience in lead extraction may not be happy with the scenario of abandoning two leads (i.e., creating two superfluous leads) and leaving a total of four or five leads implanted. Many would probably accept the tunneling of a lead from the opposite side, leaving only three leads in the heart. Although three leads may be acceptable to many, there is still a stigma associated with four leads. The discussion comparing the risk of not extracting with the risk of extracting may be helpful in resolving these issues. Superfluous Leads. A rationale for extraction of superfluous leads is not easy to construct. This situation differs from creation of a conduit. For example, if the ipsilateral vein is patent, insertion of a new biventricular lead should be uneventful. However, the addition of a new atrial lead and abandonment of the old atrial lead create a superfluous lead. The disposition of superfluous leads causes controversy, confusion, and, at times, an emotional debate. Many questions need to be answered. Why should a lead be removed, if the lead itself is not causing a problem such as penetration

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

or perforation, is not dislodged, or is not broken?21 A discussion of the risks of lead extraction ensues. Risk of Extracting Versus Risk of Not Extracting The risks associated with lead extraction are tamponade (tearing of the SVC and/or heart), hemothorax (tearing into the thorax), arteriovenous fistula and/or dissecting hematoma (tear of the aortic arch), and failure to extract the lead (attempted lead extraction). The latter is usually not considered a risk; however, a failed lead extraction may lead to additional procedures or may be a precursor for dangerous situations in the future. There are two situations in which the risk of tearing the SVC and/or heart is negligible to nonexistent. The first situation is those patients who have previously undergone an open heart surgical procedure. The pericardial space has been obliterated, and fibrous tissue reinforces the SVC and heart wall. The author has not had a cardiac tamponade in a patient with an obliterated pericardial space. The second situation is an implant of short duration: less than 2 years for pacemaker leads, or less than 1 year for ICD leads. The forces involved in freeing these leads usually are not sufficient to tear the SVC or heart. Extraction centers from the continental United States and Hawaii voluntarily submitted data for a national registry between December 1988 and December 1999.22,23 The most recent published report, from 1996, included data from 226 centers, 2338 patients and 3540 leads and demonstrated major complications in 1.4% of the cases (300 extraction procedures).24 The total U.S. data, including 7823 extraction procedures and 12,833 leads, were presented at Cardiostim in June of 2000. Multivariate analysis of the data from 1994 through 1999 demonstrated four predictors of major complications (1.6%): (1) implant duration of oldest lead, (2) female gender, (3) ICD lead removal, and (4) use of laser extraction technique. Major complications were (1) death, 0.3%; (2) nonfatal hemopericardium or tamponade, 0.7%; (3) nonfatal hemothorax, 0.2%; (4) transfusion for bleeding/ hypotension, 0.1%; (5) pneumothorax requiring a chest tube, 0.1%; and (6) other nonfatal events, 0.2% (including 4 arteriovenous fistulae, 2 pulmonary embolisms, 2 thoracotomies for defibrillator leads trapped in sheaths, 2 respiratory arrests, 2 strokes, 2 cases of renal failure, 1 anoxic encephalopathy, and 1 open surgical retrieval of a device fragment). It is otherwise clear that the risk of lead extraction is dependent on the extractor’s experience, duration of implant, age of patient, and condition of patient. There is no ongoing national database or registry, and the risks depend on the individual, the assembled team, and the institution. For example, the experiences and opinions of this author, a cardiac surgeon, must be judged accordingly. If a complication occurs, it is managed by the author. The most reliable indicator of risk is the individual extractor’s personal statistics. It is important that each institution and individual keep track of complications and effectiveness.

881

Risks are caused by maturation of the encapsulating fibrous tissue, which is related to the duration of the implant and the patient’s age. With time, the tensile strength of the encapsulating fibrous tissue increases; it may calcify in 3 to 4 years in children and in 8 to 10 years in older adults. Sedentary patients increase their tensile strength more slowly, and the tissue takes longer to mineralize. In sedentary elderly patients, the tensile strength seems to decrease with time. The influences of duration and age are apparent in the extremes. Also, patients with calcium metabolism abnormalities can calcify at any age in a short duration. Although the properties of encapsulated fibrous tissue are known, it is difficult to apply general principles to a specific patient and assign a risk. Condition refers to the patient’s physical status and mental state. It is a dominant factor in assessing the risk of the procedure for a given patient. Condition relates to the capability of a patient to survive a worstcase complication, such as a tear of the SVC or heart, or management of a difficult medical condition, such as refractory heart failure. Patients who have an ejection fraction of less than 10% or who have ascites and renal failure caused by cardiac cirrhosis are considered end-stage. The risk of a procedure in such an individual is extremely high. The risk is related to the postoperative management of these patients and not to the procedure itself. In the author’s experience, except for the two high-risk examples, patients survive. Condition also influences procedure decisions in patients who have an altered mental state or function, especially those confined to an extended care facility. Decisions in regard to these philosophical and social issues, although influenced by the physician, are usually based on family preference, medical ethics, and legal factors. Potentially lethal complications requiring extensive surgical procedures include tear of the vein and heart wall causing tamponade, arterial tears causing arterialvenous fistulae and/or dissecting hematoma, and tears into the thoracic cavity causing a hemothorax (Fig. 21-16). The procedure-related complications are discussed in detail later. Time and the surgeon’s experience are the two factors related to survival. The influence of time is obvious: low blood pressure and poor tissue perfusion are time-dependent events. Being prepared for a cardiovascular emergency is the only way to meet time constraints. This includes having a cardiovascular surgeon available, along with the proper instrumentation and experienced support personnel. A cardiovascular surgeon has the technical skill to manage these complications but may need direction from the extractor on the proper approach. Once a complication resulting in poor or no perfusion is recognized, the repair should begin immediately. The fear of needlessly subjecting the patient to extensive surgery and morbidity pales in comparison to that of applying the therapy late because of confusion or procrastination. Failure to recognize the complication in a timely fashion or the lack of access to qualified personnel, not patient condition, is the cause of a lethal outcome. In regard to the risk of not extracting, a decision to abandon a lead creates a superfluous lead—one of the

882

Section Three: Implantation Techniques

Figure 21-16. Potentially lethal complications. These potentially lethal complications all result from vascular tissue disruption during lead extraction. Disruptions are caused by tears, cutting, perforations, or avulsions of a vascular wall. Tearing or cutting of the superior vena cava (SVC) or atrial wall (A) is the most common complication resulting in cardiac tamponade. Leads that are contiguous with or embedded in the subclavian artery, innominate artery, or aorta can tear these vessels during lead extraction (B). Acutely, a dissecting hematoma develops, having the potential of rupture into the thorax or, chronically, of developing an arteriovenous fistula. Disruption of the right brachiocephalic vein into the right thoracic cavity (C) is insidious; it occurs during a difficult lead extraction when the vein and pleura are adherent or when a dissecting hematoma ruptures the pleura.

author’s indications for lead extraction. It must be stated that the author’s indication classification is not accepted by all. In fact, a superfluous lead is considered a class II or III indication for lead extraction in the guidelines published by the North American Society of Pacing and Electrophysiology in 2000.16 Using the logic applied in the previous section to justify extraction of these leads, the risk of not extracting has to be greater than the risk of extracting. The risk of extracting is not known, except as it pertains to the skill and experience of the individual extractor, and data do not exist on the risk of not extracting. Acquiring some knowledge of the risk of not extracting a lead is totally different from the knowledge base for extracting a lead. Each physician performing lead extraction must draw on his or her own experience. Although individual statistics are the most important for a given lead extraction, statistics relative to the population of lead extractors are not known. In contrast, if you make a decision not to extract a lead, creating a superfluous lead, it is unlikely that the outcome of that decision will be known in the short term, if at all. Lack of detailed outcome data and issues with logic make it difficult to establish the presence of a superfluous lead as an indication for lead extraction. Despite this dilemma, an attempt will be made. The only tools available are logical discussions based on the author’s experiences and the consequences to the biophysical interface of a chronic lead implant. Complications related to superfluous leads are the same as for a functioning lead. To justify the prophy-

lactic removal of a superfluous lead, a potential biophysical interface problem relating to the presence of multiple leads and the duration of the implant must exist. Arguments not related to the biophysical interface involving lead component and communication failures do not apply to superfluous leads. Questions such as, “What is the evidence that a biophysical interface issue will develop over time?” and “If three functioning leads are acceptable, why are two functioning leads and one superfluous lead not acceptable?” must be answered. Statements such as, “It is intuitively obvious,” “The X-ray looks better,” and “What would you have done?” cannot be used. Also not to be used are arguments suggesting that the superfluous lead may interfere with the electrical performance of the new lead (e.g., electrical noise created by contact between two ICD leads). Biophysical interface issues are known to occur with multiple leads over time. They include thrombus formation in the SVC and right atrium, damage to the tricuspid valve, and lead removal issues at lead–lead binding sites caused by an organized thrombus. These events have the same potential of forming regardless of the number of leads, functional or superfluous. For example, outcome with three functioning leads is the same as with combinations of functional and superfluous leads, such as one functioning and two superfluous leads or two functioning and one superfluous lead. The only difference is that the functioning leads provide a benefit justifying the potential risk. Based on this line of reasoning, there is no need to remove superfluous leads. The author’s view is based on experiences gained in dealing with the complications associated with multiple leads. Insight into how multiple leads increase the potential for a complication to occur and the added difficulty of managing those complications, including lead removal, are discussed. To avoid confusion, complications known to occur at the biophysical interface are presented for each vein and cardiac region in both single and multiple leads. The complications known to occur include clot formation, thrombus formation, encapsulating fibrous tissue, vascular occlusion and obliteration, embedding of the lead into the vascular wall, lead exclusion from a vein, tricuspid valve insufficiency, and, rarely, tricuspid valve stenosis. For a lead or leads passing through the ipsilateral axillary-subclavian-brachiocephalic veins, all of the complications listed have occurred, regardless of the number of implanted leads. The incidence of a complication with a single lead is not known, and it is only assumed that the probability of a complication increases with multiple leads. The clinical sequelae of these complications are rarely significant and are the same regardless of the number of leads involved. Consequently, the complications associated with multiple leads (used or superfluous) in these ipsilateral veins cannot be used as an argument. The same is not true of an occlusion or obliteration of both the ipsilateral and contralateral brachiocephalic veins. The occlusion of the brachiocephalic vein also occludes the internal jugular and collateral veins draining into the brachiocephalic vein. Bilateral occlusion of the brachiocephalic

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

veins is equivalent to occlusion of the proximal SVC. A proximal SVC occlusion causes an SVC syndrome, which is incapacitating in an active patient. Contralateral vein implants should be avoided whenever possible. This is especially true if the removal of superfluous leads prevents a contralateral vein implant from being used. The interaction in the SVC of a lead with the vein wall, resulting in thrombus formation, maturation of the thrombus into encapsulating fibrous tissue, and embedding in or exclusion from the vein wall, have been discussed for both single and multiple leads. It is hard to make an argument that extracting a superfluous lead will prevent a future embedding and exclusion complication, because the potential for this to occur is unknown. More importantly, embedding and exclusion have not caused a known complication. The presence of embedding or exclusion is not apparent unless a complication occurs in extracting these leads. Lead–lead interactions cause complications such as thrombus formation that can result in stenosis or occlusion of both the proximal and distal SVC. The author believes that the presence of more leads increases the probability of such an event. This is especially true over time. Changes in flow and lead binding cause stasis and turbulence. It can be argued that increasing the number of leads passing through the SVC increases the potential for a complication. The distal SVC, at the junction between the SVC and atrium, has the same potential for a complication as the proximal SVC. Multiple lead interactions in this narrowed area cause stenosis or occlusion resulting in SVC syndrome. The potential for this complication increases with the presence of multiple leads. Thrombus formation in the atrium is common. Progression of thrombus to occlusion of the atrium has been seen by the author only in association with severe infection. The reason for an increased incidence of thrombus formation in the atrium is probably related to its larger size and lower flow. The potential for this complication increases with the presence of multiple leads. A lead passing through the tricuspid valve damages the valve by interaction with the valve leaflets. Thrombus formation between multiple leads makes the situation worse. The clinical sequela of tricuspid valvular insufficiency caused by leaflet damage is hemodynamic dysfunction. Stenosis of the tricuspid valve is a rare complication seen only once by the author. This is caused by organization of a lead–lead thrombus. As the scar tissue contracts, the stenosis worsens. It is logical to assume that the potential for this complication increases with the presence of multiple leads. Multiple lead implants in the right ventricle cause a change in compliance of the ventricular wall. Multiple stiff leads can decrease the compliance of the ventricular wall without lead–lead thrombus interactions. The sequelae of thrombus formation and organization with scarring only make the situation worse. To compensate for the resultant decrease in chamber volume and wall compliance, the filling pressure increases. The potential for this complication increases with the presence of multiple leads.

883

Multiple lead implants have the potential for obstructing venous return with or without lead interactions and thrombus formation. Examples of initial compensatory mechanisms are formation of collateral venous drainage and elevated venous pressure. Successful compensation for these insults may leave the patient asymptomatic, especially if the insults occur over an extended period. The gradual decrease in activity may go unnoticed or misdiagnosed. Although the number of patients with venous occlusions is unknown, the author’s experiences suggest that it is a significant chronic implant problem. Lead Extraction Techniques Segments of chronically implanted leads are encased in encapsulating fibrous tissue and bound to the vein and/or heart wall, bound to another lead, or both. Lead extraction is the removal of chronically implanted leads from these binding sites. Because the tensile strength of encapsulating fibrous tissue is greater than that of the surrounding tissue, leads cannot easily be removed without risking a tear or avulsion of the vein or heart wall. The word ablation best describes the removal, separation, and freeing of leads from encapsulating fibrous tissue. Ablation techniques include traction, countertraction, counterpressure, and tissue disruption cutting locally with an instrument, laser, or electrosurgery unit. Lead extraction techniques are designed to free the lead from the encapsulating fibrous tissue (countertraction) or to free the encapsulating fibrous tissue (counterpressure) from the vein or heart wall.25 Telescoping sheaths are used to remotely apply countertraction and counterpressure at the selected binding sites. Lead extraction procedures are those procedures used to apply the sheaths and remove the lead in a safe and efficacious manner. Traction, Countertraction, and Counterpressure. In the 1960s and early 1970s, transvenous leads were large, bipolar, and without fixation devices. These leads were usually isodiametric, implanted for 1 to 2 years, and removed by traction. With the introduction of tines and pulse generators lasting 4 to 6 years, leads were entrapped in encapsulating fibrous tissue with significant tensile strength. These leads could not be safely removed by traction alone. Traction is the force exerted on the lead by pulling. Applying traction to the lead pulls directly on the binding site. Once the encapsulating tissue has a greater tensile strength than the venous or cardiac tissue, the tissue will tear or avulse. Disruption of the vein or heart wall can be lethal. Because the relative tensile strengths are not known, traction should be used with caution. This being said, traction is an acceptable extraction technique if it is applied in a judicious manner. The force applied when pulling on a lead is related to lead size, lead tensile strength, how the lead is grasped, use of locking stylets, and, most importantly, the extracting physician’s catecholamine level. The catecholamine level is an important factor in determining the actual traction force. When the extractor is

884

Section Three: Implantation Techniques

Figure 21-17. Direct traction. Direct traction is being applied by pulling on the proximal portion of the lead. Rubber bands are used in this case. Variations include using free weights and applying weights via an orthopedic traction apparatus. If a locking stylet is inserted, the traction point is moved distal to the locking site, usually near the electrode.

relaxed and calm, the perceived force may be realistic. When he or she is upset, agitated, or mad, the force is much greater than perceived. Consequently, “giving it a little tug” is not wise. Direct Traction. All current lead extraction procedures use some form of traction, or pulling force (Fig. 21-17).26 Pulling on leads was a successful method of extracting leads during the early years of pacing, when leads lacked efficient fixation devices and were implanted for short periods of time. Traction was applied manually for minutes or applied using various weights or elastic bands for days. Traction proved unsafe and had a high incidence of failure when applied to leads with efficient fixation devices and leads implanted for longer periods of time. The amount of traction required increases and becomes more dangerous as the duration of the implant and the tensile strength of the fibrous tissue increase. Leads with efficient passive fixation devices may be difficult to remove 4 to 6 months after implantation. A failed previous attempt to extract a lead frequently damaged the lead, making future extraction attempts more difficult. Traction must be applied judiciously to minimize the risk to the patient. The pulling force applied to the proximal portion of the lead is distributed to sites where fibrous tissue binds the lead or electrode and makes contact with the vein or heart wall. Multiple leads may be bound to the vein or heart wall and to each other. Because the pulling force is not focused, the distribution of force to the binding sites is unknown. It is possible to inadvertently tear a vein or the heart wall. Traction, in some form, is integral to lead extraction. The physician must consciously limit the pulling force

and apply a continuous, steady traction. Never jerk the lead, because impulse forces tear. Accidents are not predictable and frequently happen without warning, in part because it is impossible to accurately judge the level of force applied to the lead. In an attempt to gauge the applied force, most direct traction techniques try applying sufficient force to feel the rhythmic tugging of the heart without producing arrhythmias, hypotension, or chest pain. These are crude and unreliable end points and are not reflective of the tensile strength of the lead or the tissue. Breaking the lead, tearing a vein, or avulsing or tearing the heart wall all represent complications. Although applying “just a little tug” to see if the lead will come out may not be safe, following basic principles and guidelines acquired from practical experience will help minimize the risk. It is important to understand the difference between pulling from above and pulling from below. The mediastinal structures are not bound from below. If you pull upward on the heart, it will move, along with the lungs, diaphragm, and rest of the mediastinal contents, in that direction. If you pull downward, the superior veins and surrounding structures are bound to the musculoskeletal system and do not move downward. Assume that traction is applied from above to a lead implanted in the right ventricle. With continuous traction, the right ventricle starts to evaginate, decreasing the compliance of the wall and finally obstructing the flow of blood through the tricuspid valve. At the same time, the heart is pulled into the superior portion of the mediastinum. This maneuver is safe as long as the hemodynamic status is monitored and the process is reversible. Reversibility means that traction-induced deterioration in hemodynamic function is corrected on cessation of the traction, and the right ventricle, along with remainder of the mediastinal structures, returns to its normal position. Most of the time, the structures do return to normal. The danger is slippage of the lead body through a binding site. On release of the traction, the mediastinal forces pulling downward are insufficient to pull the lead back through the binding site. The forces required to elevate the mediastinum to this position are still applied to the heart. Any hemodynamic compromise such as decreased blood flow through the right ventricle persists, creating a cardiovascular emergency. If the lead body cannot be released by manipulation, including use of a stiff stylet to help push it through the binding site, an emergency median sternotomy is required to manually retract the heart. Once a lead is freed from a distal binding site, the lead and associated fibrous material can then become wedged in a more proximal binding site. For example, a lead removed from the right ventricle can become wedged at a binding site in the atrium, SVC, or axillary-subclavian-brachiocephalic veins. With direct traction, the lead is freed from the binding sites, distal to proximal. The strongest binding site determines the outcome of an extraction attempt, regardless of its proximal or distal location. Indirect Traction. Elevation of the mediastinum with traction is the main reason why traction from above is

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

not as effective at pulling a lead through a binding site as is pulling from below. When traction is applied from below, the mediastinum is not pulled down (inferior), because the superior veins and surrounding tissues are bound to the musculoskeletal system. Traction forces applied to the binding sites are directed at freeing the lead, and not at moving structures. The limit to the force applied to the axillary-subclavian-brachiocephalic veins is determined by the tensile strength of the lead. If the lead binding in these veins is extensive enough to require that kind of force, the veins are probably occluded, atretic, or encapsulating sheaths. Disruption of these veins by tearing and/or avulsion is of no consequence. Binding sites in the SVC, however, must be treated in the same manner as in the heart. Disruption of the wall of the SVC has the same consequences as disruption of the heart wall. The safety and efficacy (higher success rate) of applying indirect traction from below should be apparent from these comments. Indirect traction is traction applied by an instrument, such as a snare passed into the heart, usually through a femoral vein. The lead is entrapped in the snare, and traction is applied by pulling or pushing. The difficulty is in grasping the lead in a fashion that allows sufficient traction to be applied. Only a few snares, such as the Dotter basket snare (Cook), have sufficient strength to support extraction forces. The lead must first be freed from the superior veins and then from the heart. The lead is pulled out of the superior veins and into the atrium or inferior vena cava (IVC). It can be regrasped, if necessary, and traction can be applied to the heart. The techniques for applying indirect traction are the same as for grasping and manipulating the leads in other approaches, such as applying countertraction sheaths, and are described later. The risks eliminated by indirect traction are tearing of superior veins, wedging of the lead in the atrium or in a superior vein, and creating a low cardiac output caused by failure of the lead to return to its original position after traction. Indirect traction has the same potential for breaking the lead or tearing the heart wall as direct traction, if their tensile strengths are exceeded. Countertraction. Countertraction is the technique used to free the lead from compliant encapsulated fibrous tissue. Countertraction was first used to remove a lead from an implantation site in the right ventricle or atrium. Although the technique for extracting leads from the heart wall is discussed first, this is the last step in a normal lead extraction procedure, using any type of sheath. Extraction sheaths free leads from binding sites, proximal to distal (Fig. 21-18). Once the sheath is passed over the lead and down to the implantation site, traction on the lead pulls the site to the sheath (Fig. 21-19). The traction force is countered by the circumference of the sheath. The countertraction sheath focuses the traction force at the tip of the sheath, limiting the excursion of the heart wall. This prevents compliance changes and blockage of the tricuspid valve with possible perforation, tearing, and avulsion of the heart wall. The countertraction forces

885

Figure 21-18. Countertraction. Telescoping sheaths are passed over the lead and maneuvered from binding site to binding site, breaking through each one primarily by countertraction. When necessary, the application of counterpressure peels the encapsulating fibrous tissue from the wall. At the electrode myocardial interface, countertraction is the only safe method that can be used to remove the lead.

Figure 21-19. Countertraction. Countertraction is a safe method extirpating a lead or electrode from a vein or heart wall. Traction on the lead body with evagination of the right ventricular wall (left) decreases the compliance on the wall and pulls it toward the tricuspid valve. Complications include decreased flow through the right ventricle and disruption of the ventricular wall instead of the scar tissue, with tearing of the wall and/or avulsion of tissue. Placing a sheath near the heart wall applying traction (right) causes the traction force to be countered by the sheath. The limit on the traction force applied is the tensile strength of the lead. The lead is extirpated from the scar tissue, and the heart returns to its normal position.

are limited by the tensile strength of the lead. At some point, the electrode is freed from the encapsulating fibrous tissue, allowing the heart wall to fall away and the electrode to be pulled out of the sheath. The way countertraction actually frees the lead is postulated but not known. It is imagined that the traction force wedges the lead against the countertraction sheath. The pulling force on the electrode tries to evaginate the encapsulating fibrous tissue. The electrode is then imagined to be freed either by a plastic deforma-

886

Section Three: Implantation Techniques

tion of the tissue that allows it to slide out of the encapsulating tissue as the countertraction sheath peels the tissue off the electrode or by an actual disruption or bursting of the encapsulating tissue that frees the electrode, or both. For a passive electrode, the tines are removed intact with the electrode; for an active fixation electrode, the fixation mechanism is ideally retracted or unscrewed before countertraction is applied. In some cases, continued “unscrewing” of an active fixation lead results in complete lead removal without the need for countertraction because of the absence of significant binding at other sites along the lead.27 If the helix will not retract, the electrode and fixation mechanism are removed together. The same scenario is envisioned for removal of electrodes from the atrial wall. Countertraction is also used to free the lead from the encapsulating fibrous tissue at binding sites along the vein and heart wall (see Fig. 21-19). This is possible only if the encapsulating fibrous tissue still has plastic qualities (compliant). The tissue at the binding site is pulled against or into the sheath and is removed by evagination, peeling, or tissue rupture. Countertraction can be performed with either the inner or the outer sheath. Counterpressure. Counterpressure was the name given to the removal technique used for noncompliant encapsulating fibrous tissue (mineralized tissue). A sheath larger than the solid encapsulating tissue is used, and the tissue is pulled into the sheath. The encapsulating fibrous tissue is usually attached to the vein, tricuspid valve, or heart wall. The sheath counters the traction force applied to the tissue mass by converting this force into a pressure concentrated locally between the edge of the sheath and the vein wall (counterpressure sheath). This local action peels the calcified mass off the vein or heart wall. The encapsulating tissue is included with the lead inside the sheath (inclusion). The force applied is limited by the tensile strength of the lead and/or the wall. Because the magnitude of the counterpressure force actually focused on the wall is unknown, application of force is subjective. Counterpressure is potentially dangerous and should be approached with caution. It is believed by some that mineralization of tissue may lead to a higher risk associated with lead extraction. The inability to safely pass a binding site using counterpressure is the primary reason for abandoning this approach and changing to a transfemoral or transatrial approach. These approaches allow the lead to be pulled out of the superior veins from below, through the binding site. It is unknown whether the lead is being removed by countertraction or counterpressure (Fig. 21-20). In the past, counterpressure was used to describe the removal of tissue from all sites other than the electrode implantation site. In most cases, removal is still credited to counterpressure; compliant tissue is removed primarily by countertraction, and noncompliant tissue by counterpressure. Not discussed are leads bound to one another by the calcified encapsulating tissue. Separation of the two leads is safe, and the traction force is limited only by the tensile strength of the lead.

Figure 21-20. Mechanical and powered sheaths. Telescoping sheaths are inserted transvenously and passed to the first binding site. Countertraction, counterpressure, and/or power ablation is applied to extirpate the lead. Countertraction dilates and/or ruptures the binding site, liberating the lead. The laser and electrosurgical dissection sheath (EDS) is a powered sheath that vaporizes the tissue. Counterpressure is used to pass the sheath over the entire calcified encapsulating fibrous tissue mass (inclusion). Counterpressure is the actual manipulation of the telescoping sheaths to peel the encapsulating tissue off the vein or heart wall.

Extraction Instruments. Lead extraction instruments are separated into mechanical sheaths, powered mechanical sheaths, and snares. Mechanical Sheaths. Mechanical sheaths are telescoping sheaths made of Teflon, polypropylene, or stainless steel (Fig. 21-21). These telescoping sheaths are designed to pass over the lead, which acts as a rail guiding the sheaths through the veins and down to the heart wall. Countertraction and counterpressure are applied as the sheaths move down the lead from one binding site to another. The outer sheath also acts as a workstation. As a workstation, it facilitates the free movement of the inner sheath and lead by eliminating binding, and it protects the surrounding vascular structures. The leading edges of the sheaths are beveled. The rotation of the beveled tips facilitates maneuvering past obstructions and through the narrow channels along the tortuous paths surrounding the lead body. This is especially true in the superior veins. For the sheaths to pass down the lead in a true fashion, the lead must be stiff enough to act as a guide rail. The lead is stiffened by pulling it taut. The lead must be stiff enough to resist bending or kinking as the sheaths are passed over it (lead stiffness > sheath stiffness). The telescoping action of the sheaths allows the more supple inner sheath to track over the lead. The larger outer sheath is then advanced using the combination of taut lead and inner sheath as the guide rail. The lead is made taut by traction (pulling on it). A locking stylet is inserted, and a suture is usually tied to the lead, acting as both an extender and a traction handle. As described earlier, experience and judgment are required to avoid tearing and avulsing vein and heart wall tissues. Also, if the traction force exceeds the lead’s tensile strength, it can cause lead disruption

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

887

A

B

C

Figure 21-21. Mechanical sheaths. Mechanical telescoping sheaths have a beveled tip and come in various sizes. A, Teflon sheaths are softer and more supple; they were the original sheaths. B, Polypropylene sheaths are stiffer and allow more force to be applied at the binding site. C, Stainless steel sheaths are rigid and are used to bore through bone or calcified tissue at the vein entry site. The stainless steal sheaths should not be passed into the brachiocephalic vein.

and/or breakage. Insertion of a locking stylet adds some stiffness to the lead and focuses the traction force to a locking site near the electrode. Except for simple cases, in which leads have been implanted for a short duration, a locking stylet should be inserted. Despite these efforts, as the tensile strength increases it becomes more difficult and more time-consuming to free the lead from its binding site. The forces involved frequently exceed the tensile strength of the lead, resulting in lead disruption and breakage. Powered Mechanical Sheaths. Powered cutting tips positioned at the leading edge of the mechanical sheath have changed the nature of lead extraction. The ability to free the leads by cutting tissue significantly decreases the countertraction forces. Reduction in the applied force has made lead breakage and separation of the lead body from the distal electrode rare events. The expectation is that any lead should be completely extracted regardless of its tensile strength. Before the advent of powered sheaths, a lead breakage rate of 15% to 20% was common for leads with satisfactory tensile strengths. The first powered sheath was the excimer laser sheath (Spectranetics, Colorado Springs, Colo.) developed in the mid-1990s. The second was the electrosurgical dissection sheath (EDS; Cook) in the early 2000s. Both of these sheaths are used successfully today. Excimer Laser Sheath. The development of the excimer laser was a milestone for lead extraction (Fig. 21-22).

The excimer laser generated a high-energy 308-nm laser beam known to disrupt tissue (both cells and hydrated proteins) by an explosive vaporization of intracellular water. The rapid vaporization helped to cool the site. These were appealing properties for lead extraction. Unfortunately, the laser did not ablate mineralized tissue. If the laser could not be maneuvered through this tissue in grinding fashion, counterpressure techniques had to be used. The development of the excimer laser sheath was a technical achievement. It required expertise in polymer chemistry and optic fibers to develop a small-diameter, flexible sheath capable of withstanding the excessive forces applied during a lead extraction. At the time, optic fibers were bundled in a circumferential fashion inside a cylindrical metal housing. The metal housing protected the optic fibers from the applied forces generated while maneuvering the tip through the encapsulating tissue. Initially, the only sheath meeting all the clinical requirements was a 12F sheath that was interchangeable with the 12/16F mechanical Teflon sheaths. In time, larger 14F and 16F sheaths were perfected. These sizes were sufficient to manage all sizes of pacing leads up to the largest ICD leads. The last iteration of the laser sheath was to place a 15-degree bevel at the tip. It is the largest angle permitted by the circular configuration of the optic fibers at the electrode. The laser is controlled by a foot switch. By design, the laser is on for 10 seconds and off for 5 seconds. The

888

Section Three: Implantation Techniques

Figure 21-22. Excimer laser sheath. The laser sheath plugs into the CVX-300 excimer laser system with a black connector at the proximal end. A blue fiberoptic cable conducts the pulsed ultraviolet light to the working section, which is tubular in shape. The inset gives an idea of how the laser sheath is designed to slide over the lead body as it threads its way through the veins to the heart. A micrograph of the tip shows how the 83 fibers are arranged in a single circumferential row at the tip. When the laser light comes out of these fibers, it cuts the fibrotic tissue that binds the lead to the vein walls.

sound caused by the rapid pulsing of the laser furnishes a unique sound indicating the laser is on. The laser beam is a light cone that ablates tissue up to a distance of 1 mm. The water vapor generates bubbles that are clearly visible on echocardiography. Although the bubbles and other particulate debris are filtered out in the lungs, there are no apparent clinical sequelae. The cutting action of the laser can disrupt the SVC or the atrial wall if the lead is embedded in the wall. Because there is no way to know when a lead is embedded in the wall, the same emergency precautions apply to the laser as to the mechanical sheaths. The laser sheath technique was evaluated prospectively in two clinical trials. The first, the Pacing Lead Extraction with the Excimer Sheath (PLEXES) trial, included only the initial version of the 12F sheath. Although there have been substantial subsequent improvements in the 12F sheath, including an outer sheath, better mechanical properties to prevent crushing of the optical fibers, lubrication, a flexible distal and a more still proximal segment, and certainly better understanding of how to use the tool, the PLEXES trial was a dramatic success. This was a randomized clinical trial comparing mechanical extraction tools with laserassisted lead extraction, and it was used to support the clinical release of this technology. The complete lead removal rate was 94% in the laser group and 64% in the nonlaser group (P = .001). Failed nonlaser extraction

was completed with the laser tools 88% of the time. The mean time to achieve a successful lead extraction was significantly reduced for patients randomized to the laser tools: 10.1 ± 11.5 minutes compared with 12.9 ± 19.2 minutes for the nonlaser techniques (P < .04). There was only one death, but it was in the laser group; and there were two other potentially life-threatening bleeding episodes in the laser group.28 After the trial with the 12F sheaths, a second, nonrandomized cohort trial was done with 14F and 16F sheaths. This was particularly important, because implantable defibrillator leads required the 16F sheaths, and many of the bipolar leads (almost all) were better approached with the 14F sheath. In contrast to other, nonlaser sheaths, upsizing of the laser sheath to pass over (include) the fibrosis or calcification is frequently a very effective maneuver. In this study, 863 patients underwent extraction of 1285 leads. Expanding the number of research sites from fewer than 10 to 52 gave a broader view of this tool in general practice. The patients treated with the 14F device tended to have older leads than patients in the 12F population; the 16F population, composed mostly of defibrillator patients, were younger, had younger leads, and were more often male than the 12F population. Clinical success (extraction of the entire lead or of the lead body minus the distal electrode) was observed in 91% to 92% of cases for all device sizes. The overall

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

complication rate was 3.6%, with a 0.8% perioperative mortality rate. The incidence of complications was independent of laser sheath size.29 Ultimately, a cohort comparison trial of defibrillator and pacemaker leads extracted with laser assistance was done at the Cleveland Clinic Foundation. ICD extraction results were compared with the results for a matched cohort of patients undergoing extraction of ventricular pacemaker leads from a national registry and with the experience with pacemaker lead extraction at the Cleveland Clinic Foundation. Successful complete extraction of ventricular nonthoracotomy implantable defibrillator leads, in the absence of major complications, was achieved in 96.9% of attempts to extract leads from 161 patients. Clinical success was achieved in 98.1% of patients. There were three major complications, including one death. ICD lead extraction was done at an experienced center with equal risk and no significant difference in procedure or fluoroscopy time.30 The total investigational experience with laser sheaths was also reported, encompassing the period from October 1995 to December 1999, including 2561 pacing and defibrillator leads in 1684 patients at 89 sites in the United States. Of these leads, 90% were completely removed, 3% were partially removed, and 7% were failures. Major perioperative complications (tamponade, hemothorax, pulmonary embolism, lead migration, and death) were observed in 1.9% of patients, with in-hospital deaths in 13 (0.8%). Minor complications were seen in an additional 1.4% of patients. Multivariate analysis showed that implant duration was the only preoperative independent predictor of failure, and female sex was the only multivariate predictor of complications. Success and complications were not dependent on laser sheath size. At follow-up, various extraction-related complications were observed in 2% of patients. The learning curve showed a trend toward fewer complications with experience.31 A similar experience was observed in Europe.32 Electrosurgical Dissection Sheaths. The EDS has two bipolar electrodes positioned at the tip of the bevel (Fig. 21-23). The sheath is connected to an interface plate inserted on a conventional electrosurgery unit (Valley Lab Force V; PEMED, Denver, Colo.), placed in a bipolar cutting configuration, and activated with a foot switch. The interface plate is attached to the front panel of the electrosurgery unit to ensure that the EDS is connected in a bipolar configuration. The interface also has an attachment to pulse the electrosurgery unit 80 times per minute. A plasma arc is generated between the electrodes. The plasma arc extends out from the electrodes and vaporizes the tissue to a depth of about 1 mm. On continuous discharge, desiccated tissue debris shunts the arc between the electrodes, preventing it from cutting. Also, on continuous discharge, if one of the electrodes touches a conductor coil, a parallel alternate current (AC) circuit is created consisting of the EDS electrode in contact with the conductor coil, the conductor coil down to an electrode in the heart, and back to the other EDS electrode. An AC current applied to the heart in a unipolar configuration can fibrillate the heart. To ensure cutting and avoid fibrillating the heart,

889

Figure 21-23. Electrosurgical dissection sheath (EDS). The EDS is a telescoping, Teflon, beveled sheath set with two electrodes in a bipolar configuration at the tip of the inner sheath. Encapsulating tissue is ablated with the use of radiofrequency energy delivered as needed in a pulsed fashion at 80 Hz. The electrodes are radiopaque, allowing precise placement.

the EDS is operated in a pulsed mode at 80 pulses/min. In the pulsed mode, if a conductor coil is touched, it paces the heart. The EDS is a conventional Teflon mechanical sheath with two tungsten electrodes embedded in the polymer. Placement of the EDS electrodes in a bipolar configuration at the tip of the inner telescoping sheath endows the EDS sheath with properties of mechanical sheaths: it can maneuver through tortuous veins, and it can be used to apply countertraction and counterpressure. These sheaths are currently available in sizes 7F, 9F, 11F, and 13F (circumference of the inner sheath). The electrodes and sheath are radiopaque, allowing visualization during fluoroscopy. The electrodes’ positions are continuously adjusted by rotation of the sheath. For example, around a curve, the electrodes are placed on the inner curvature passing down the lead. Also, the electrodes are rotated away from skeletal muscle and nerves to avoid stimulation. Stimulation of skeletal muscle and/or the phrenic nerve does not harm the patient and is not an issue for patients under general anesthesia. However, skeletal muscle and diaphragmatic contractions can be discomforting and even frightening if the patient is awake. At present, this is the only clinical issue associated with the EDS. The same emergency precautions applicable to mechanical and laser sheaths also apply to the EDS. Clinical evaluation of the EDS has been formally published only in an observational study from five centers, involving 265 patients with extraction of 459 leads.33 During the investigation, only the 9F and 11F sheaths were used, excluding almost all ICD leads from consideration for extraction. As in all extraction series, some of the leads came out easily and others were more difficult to remove, and the techniques consisted more of an approach than of universal use of one tool to remove all leads. In this case, 542 leads were potentially presented for extraction, but about 15% were removed with direct traction, yielding 459 for which the EDS was employed. The laser tool was used in fewer than 3% of the leads. The average implant duration of

890

Section Three: Implantation Techniques

the patient’s oldest lead was 8.4 ± 5.0 years; 31% of patients had leads implanted for longer than 10 years. Major complications occurred in 2.6% of patients, including cardiac tamponade in 4 patients (1 surgical repair, 1 after switching to a femoral approach), 1 hemothorax, 1 arteriovenous fistula (surgical repair), and 1 death that was associated with the mechanical removal of an oversized SVC lead for which the EDS was not used. For the 459 leads with attempted removal by the EDS, 99.4% were removed (95.9% completely, 3.5% subtotally with ≤4 cm of lead remaining), and only 0.6% were not removed. Overall, the experience with the EDS has been good. The application of 7F and 13F sheaths with intermittent pulsing of the electrosurgical energy has been useful in removing smaller and larger leads and in making the sheath more powerful in cutting through the fibrosis. EVOLUTION Mechanical Dilation Sheath. Cook Vascular has recently introduced a new mechanically powered lead extraction sheath set, the EVOLUTION Mechanical Dilation Sheath. A rotating inner sheath with a threaded barrel metal tip was designed to function as a dilating drill. This bores through the encapsulating fibrous tissue as it advances down the lead through the binding sites. The outer sheath is a conventional beveled plastic sheath. The rotation of the inner sheath is powered by a pistol grip handle squeezed by the operator (mechanical power). Multiple inner diameter sizes (7, 9, 11 and 13 French) are available. In the author’s initial clinical experience with 19 patients (leads implanted from 1 to 17 years), the leads were successfully removed. In the absence of calcified encapsulating tissue, the device moves at an impressive rate along the lead. Although this tool has not been tested by multiple operators, it appears

A

C

to be useful and an advance in the technology of lead extraction. Locking Stylets. Locking stylets were developed after the mechanical sheaths. From the beginning, the major pacing companies (Medtronic, Cordis [St. Jude Medical], and Pacesetter [Guidant]) all attempted to make a universal locking stylet (one size stylet to fit all). These initial attempts were unsuccessful because of breakage of the locking mechanism. The tensile strength was inadequate to withstand the traction forces. Cook Pacemaker, Inc., (now Cook Vascular Inc., Vandergrift, Penna.) took another approach; they abandoned the concept of a universal stylet. Their first-generation locking stylet came in various sizes to fit a variety of conductor coil diameters (Fig. 21-24). The conductor coil had to be measured before selecting the locking stylet. This locking mechanism was a small wire welded to the tip of the stylet and wrapped around the lead. Once the lead was passed down the conductor coil to the electrode, it was rotated counterclockwise, bundling the free wire and causing it to bind against the conductor coil. The greater the traction, the greater the binding force. The locking stylet bound to the inner conductor coil, ideally at the distal electrode, functioned as a lead extender for applying traction and focused the traction force at the binding site. Focusing the traction force helped maintain the integrity of the lead but did not prevent lead disruption if excessive force was applied or if the lead had poor tensile strength. Also, the locking stylet was conductive, and the heart could be paced during parts of the lead extraction procedure if needed. This was the first effective locking stylet. Cook’s second-generation locking stylet (Wilkoff Locking Stylet) had a different locking mechanism (Fig. 21-25). These stylets had a small flange at the tip that was designed to stay flat against the stylet until the

B

Figure 21-24. A, Cook first-generation locking stylet. Ideally, the locking stylet is passed to the electrode and secured. The locking mechanism is a wire that is secured to the tip and wrapped around the stylet. B, When the stylet is turned counterclockwise, the wire bundles together, binding the stylet to the conductor coil. The locking stylet acts primarily as a lead extender to apply traction and secondarily to keep the lead intact during the extraction. When the stylet is positioned at the electrode, the lead has its best chance of remaining intact. If the stylet is positioned near the proximal end, the fragile leads will be pulled apart when traction is applied. C, Clockwise from lower left: locking stylet, sizing pin, lead cutter, corical coil expander, soft grip hemostat, standard stylets, and polypropylene telescoping sheaths.

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

Figure 21-25. Cook second-generation locking stylet (Wilkoff Locking Stylet). The Wilkoff Locking Stylet employs a different mechanism for applying traction to the lead conductor at the tip of the lead. It also functions as a lead extender to lengthen the lead so that the sheaths can be advanced to the endocardial surface. The locking mechanism does not require precise measurement of the internal diameter of the conductor coil, and each size of the Wilkoff stylet bridges three sizes of the original locking stylet. The mechanism is activated by advancement of the thin cylinder, which nudges the hook to the side, engaging it between the coils of the conductor. This mechanism can be reversed and relocked but does not tolerate rotation. The response to rotation can be an advantage if the stylet needs to be removed.

A

C

891

preloaded thin cylinder was advanced; the cylinder deflected the flange to lock into the conductor coil. This was an efficient locking stylet that was easy to implant and could be removed by rotating the stylet, breaking the flange. However, it could not be used if the conductor coil diameter was 0.016 inches or less, and it could not be used with extendable/retractable screw-in leads. Spectranetics made the first near-universal locking stylet (Lead Locking Devices 1, 2, and 3), with three sizes to fit all leads. It used a long wire mesh that bundled and bound the stylet to the conductor coil. This type of locking stylet was efficient and functioned well. Cook’s most resent locking stylet (Liberator Locking Stylet) is a true universal locking stylet (one size fits all). It uses a wire coil, which is compressed by a reloaded cylinder, expanding the wire coil and binding it against the conductor coil (Fig. 21-26). This is the only universal stylet on the market, and it has excellent performance.34 As mentioned earlier, the goals for inserting the locking stylet were to provide a lead extender and to focus the traction force at the tip to help maintain lead integrity. Although these goals were met, the insulation still has to be secured with a suture. Without traction to the insulation, it can “snow plow” and tear more easily, making passage of the extraction sheaths difficult, and in some cases preventing their passage. Therefore, traction should be applied to both the suture

B

D

Figure 21-26. Cook third-generation locking stylet (universal locking stylet). The universal locking stylet works by binding the locking stylet to the conductor coil. A wire coil is ideally positioned at the tip of the lead, near the distal electrode. The coil is expanded, binding the stylet to the lead’s inner conductor coil. Traction increases the binding force. A universal locking stylet is possible today because of advances in technology. It is now possible to manufacture a stylet to fit the smallest conductor coil and still have sufficient tensile strength to support a wire coil large enough to bind to the largest diameter conductor coil. A, Lead conductor without the locking stylet. B, The liberator locket stylet has been advanced to the tip of the lead. C, The cylinder is pushed to crush the wire coil at the tip of the lead, locking to the conductor. D, Comparison of the inner wire with and without the wire coil at the tip of the universal or liberator locking stylet.

892

Section Three: Implantation Techniques

(insulation) and the locking stylet (conductor coil) to be most effective.

Creating a reversible loop using two snares is a complicated maneuver requiring practice to perfect. One technique is to use a Cook deflecting wire guide and a Dotter basket snare. The Cook deflecting wire guide is wrapped more than 360 degrees around the lead body. Next, the tip of the deflection catheter is passed into the Dotter basket snare. When the basket snare is pulled into the workstation, the basket closes, grasping the deflection catheter and completing the loop. The loop is pulled into the workstation, tightly binding the lead body to the workstation, and traction is applied. If needed, the loop is relaxed and repositioned on the lead body. This sequence is repeated until the lead is pulled out of the superior veins, through the right atrium, and into the IVC. The reversible loop is then released, and the deflection catheter is removed. The Needle’s Eye Snare is a more efficient method of grasping the lead in a reversible fashion (Fig. 21-28). This snare has a wire loop that is passed over the lead body. A small wire loop tongue is then passed over the opposite side of the lead and into the larger wire loop tongue. Pulling this apparatus into the workstation binds the lead in a reversible manner for safe indirect traction. Also, the binding forces are more diffuse, resulting in fewer lead breakages. The Goose Neck Snare is a radiopaque noose that is slipped over the free proximal end of a lead (Fig. 21-29). In situations in which the proximal end is floating in the SVC, heart, or a pulmonary vein, the free end can be lassoed and extracted.

Snares. Snares are used to grasp leads and to remove tissue in the bloodstream from the vein or transatrial entry site. The vein entry sites commonly used are the subclavian, internal jugular, and femoral veins. Only a few snares are safe to maneuver in the cardiovascular system or have the tensile strength to support the forces involved in a lead extraction procedure. The Dotter snare, Needle’s Eye Snare (Cook), and Amplatz Goose Neck snare (ev3/Vasocare, Seoul, Korea) are discussed as examples of the types of snares available. The Dotter snare, together with a deflection catheter, is prepackaged in the femoral workstation. Before the availability of powered sheaths, lead breakage and the need to use a femoral approach were common. The only substitute for a snare is a cardiac surgical procedure. Consequently, facility with snares still is a requisite skill. Also, there are still extractors who do not use powered sheaths, relying only on the mechanical sheath extraction, and snares are integral to these procedures. Reversible Loop. With the Dotter snare, a reversible loop is created around the lead body to pull the proximal end of the lead out of the superior veins into the IVC, without placing traction on the electrode myocardial interface (Fig. 21-27). A loop must be created and bound to the lead body. It is mandatory that the binding of the loop be reversible. Irreversibly binding the lead, or inability to remove the loop from around the lead, may result in dangerous traction maneuvers being performed in desperation while trying to extract the lead and snare. Failure to extract the lead subjects the patient to more invasive procedures to remove both the lead and snare.

A

E

B

Extraction Procedure Approaches Extraction procedure approaches involve transvenous and cardiac surgical extraction techniques. Transve-

C

F

D

G

H

Figure 21-27. Cook femoral workstation (Byrd WorkStation). The workstation’s two snares (Dotter basket and Cook deflection) were the first found by the author to have the versatility and tensile strength suitable for lead extraction. A, Telescoping Teflon sheaths and Dotter snare advanced to the right atrium with lead body entangled in the snare. Pulling the lead into the sheaths for lead extraction is possible but not recommended because it is not reversible. B, Dotter snare and deflection catheter in atrium. C, Looping the lead body with the deflection catheter. D, Entangling the tip of the deflection catheter in the Dotter snare. E to G, Tightening the loop and pulling it into the Teflon sheaths. H, Sliding the telescoping sheaths to the electrode for removal using countertraction. At any time, the process can be reversed, freeing the lead.

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

B

A Figure 21-28. Needle’s Eye Snare. Cook’s Needle’s Eye Snare is the only reversible snare designed specifically for removing leads. A, The lead body is hooked, usually in the atrium, and the tongue is extended, forming a reversible loop. B, The lead body is pulled into the telescoping sheaths for lead removal.

A

B

C

D

Figure 21-29. Amplatz Goose Neck snare. A to D, The Goose Neck snare places a noose around the end of a lead and is pulled taut. The Goose Neck assembly is then pulled into telescoping sheaths for lead removal. This snare is reversible unless the assembly binds in the telescoping sheaths. In the author’s experience, this snare is most effective for removal of small lead segments (e.g., electrodes, lead body segment) from small vessels such as hepatic veins, pulmonary arteries, azygos veins, and pelvic veins.

893

894

Section Three: Implantation Techniques

nous approaches are usually performed through the lead implant site, but any other vein suitable for the extraction instruments may be used. Suitable veins include the axillary-subclavian-brachiocephalic, external jugular, internal jugular, and femoral-iliac. Cardiac surgical approaches to intravascular leads are transatrial, right ventriculotomy, and open heart surgery. All epicardial lead extractions are cardiac surgical procedures. In the past, procedure approaches were separated into SVC and IVC approaches. An SVC approach is defined as the passage of lead extraction instruments through the vein entry site and down the SVC to the heart. The IVC approach was defined as a transfemoral approach using remote extraction instruments inserted in a femoral vein and passed through the IVC into the heart. This classification was based on the two types of extraction tools available at that time. Today, a more descriptive classification is to separate the approaches into transvenous and cardiac surgical procedures. Transvenous procedures are further divided into procedures through the vein entry site and those through a remote vein. Cardiac surgical procedures are separated into the individual cardiac surgical approaches: transatrial, ventriculotomy, and open heart. The approach selected depends on the experience, extraction skills, and extraction instruments available to the physician; the reason for the lead extraction; and situations that arise during lead extraction. For example, a transvenous lead implanted in the right ventricle with a conductor coil fracture is ideally extracted from the vein entry site, using a powered sheath. A change in approach must be considered if this same lead is found to be attached to the lateral wall in the distal half of the SVC by a calcified mass of encapsulating tissue. A determination to continue the approach or a decision to choose a new one depends on experience and extraction skills. Two safe approaches are through the femoral vein using snares and via a transatrial cardiac surgical approach. If this same lead is found to have large solid thrombi measuring 4 × 8 cm, it is best extracted through a transatrial or an open heart procedure. Extracting leads is dangerous if it is not properly performed. Procedures with these inherent dangers can be made safe only by knowledge of the pathology, pathophysiology, extraction skills, and available approaches. An example illustrating a potentially dangerous situation made safe by knowledge, training, and experience is flying. Flying an airplane has inherent, potentially life-threatening risks, but with proper training, experience, and “flying by the numbers,” it is safe. A lead extraction performed in an organized step-by-step fashion, following known principles and proved guidelines, is equivalent to “flying by the numbers.” The antithesis is “flying by the seat of your pants” or extracting a lead in a reckless, cavalier fashion, hoping for the best. While reading this section, remember that life-threatening complications are rare. During the evolution of lead extraction, the author and many others helped define the procedures, techniques, and approaches to minimize and hopefully eliminate com-

plications of all types. The material presented represents more than 25 years of experiences that form the basis for current lead extraction procedures. Transvenous Approaches Lead Vein Entry Site. A transvenous approach through the lead vein entry site is a natural approach. After the pocket debridement procedure, which includes freeing the leads to near the vein entry site, it is natural to continue and remove the lead from this site. In some situations, the lead is broken or cut and retracted into the axillary-subclavian-brachiocephalic veins. If this lead can be grasped by an instrument passed into the vein and exteriorized, it is then considered to be a lead passing through the vein entry site for the purposes of this discussion. The lead can be extracted by direct traction, if applicable, or by a mechanical or powered telescopic sheath. This natural approach is used by most extractors today. Working through the vein entry site is efficient and subjects the patient and physician to the least amount of radiation. Efficiency includes the least number of maneuvers and the least amount of time required. Radiation exposure is related to the amount of fluoroscopy time. Also, once the lead is removed, a new lead may be readily inserted through the conduit created during the extraction. Direct traction is the term used when traction is applied by pulling directly on the lead manually, with a suture or locking stylet or both. This is in contrast to indirect traction, which refers to grasping the lead with a snare and applying traction by the snare. Some leads are easily removed by direct traction. The vein entry site is not always easily accessed. In some situations, the lead is entrapped in calcified encapsulating tissue and cannot be entered using the nonmetal extraction sheaths, with or without power. The cause of the calcified encapsulating tissue at the vein site is usually related to introducer technique. If the introducer needle scores the clavicle or first rib, elevating the periosteum, the periosteum will reform about the lead, entrapping it in a bone sheath. If the introducer needle passes through the costoclavicular ligament, this damaged tissue will mineralize and entrap the lead. It can also be postulated that natural maturation of a thrombus into fibrous tissue, which mineralizes with time, also leads to this problem. Although this can occur, the problem usually is associated with the periosteum and/or costoclavicular ligament. In this situation, metal sheaths are used. Telescoping stainless steel metal sheaths look dangerous but are safe and effective if properly used (Fig. 21-30). The principle is simple: keep the sheaths tracking true over the lead, and apply the force needed to destroy this tissue. A combination of pushing and rotation of the beveled tip is most effective. Once the metal sheaths break into the vein, they are removed and replaced by plastic extraction sheaths. Tracking true is a principle used for all sheath maneuvers. All sheath maneuvering must be performed under fluoroscopy. This is to ensure that the sheaths are tracking over the lead. Any kinking or other deviation of the vein course

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

Figure 21-30. Stainless steel sheaths. Stainless steel sheaths are invaluable for freeing leads at the vein entry site that are entrapped in calcified tissue or bone (clavicle). They are remarkably safe when fluoroscopy is used to monitor all maneuvers and the sheath is not inserted past the end of the subclavian vein.

is dangerous. It creates a false passage that may be extravascular and may damage nearby structures. Complications. Telescoping sheaths use a combination of countertraction, counterpressure, and tissue ablation (powered sheaths) to maneuver past the intravenous and intracardiac binding sites to the heart wall. These techniques, used with various types of sheaths, have already been described in detail. The maneuvering of these sheaths over the lead and down to the heart is separated into the three anatomic regions: axillarysubclavian-brachiocephalic veins, SVC, and intracardiac. This is the most dangerous portion of the procedure, and a detailed understanding of the issues unique to these three regions is essential for a safe and successful lead extraction. Complications associated with lead extraction involving loss of vein or heart wall integrity are tearing (including bone spicules), avulsion, rupturing, penetration, embedding of lead into the wall, and exclusion of the lead from the vascular space. The sequelae of these events are related to the anatomy of the region. The presence of fibrous tissue must be borne in mind when considering the anatomy of the region. Inflammatory reaction and the resultant fibrous tissue formation increase the tensile strength of the tissue. For example, a dissecting arterial hematoma or pericardial effusion is rarely seen in patients with previous open heart surgery, because scar tissue has the positive effect of reinforcing the tensile strength of the surrounding tissue. The forces required to tear aorta, subclavian, or innominate arteries or heart wall reinforced with scar tissue exceed those normally applied during lead extraction. The axillary veins are surrounded by soft tissue on both sides, and the loss of vein wall integrity causes a

895

low-pressure extravasation of blood, which is usually of no consequence. The subclavian-brachiocephalic veins are surrounded by structures which, if damaged, could lead to life-threatening consequences. On the left side, the vein is contiguous with the subclavian artery and aorta. During an inflammatory reaction, if the lead becomes embedded in the vein wall and involves the wall of the subclavian artery, innominate artery, or aorta, passage of extraction sheaths over the lead could tear these contiguous vessels, causing an arteriovenous fistula or a dissecting hematoma or both. The arteriovenous fistula could cause high-output heart failure, and the dissecting hematoma could cause major blood loss. If the hematoma ruptures into the left thoracic cavity, the resultant hemorrhage could be lethal. Immediate surgical intervention is required for a dissecting hematoma. An emergency median sternotomy provides satisfactory surgical exposure. If time permits, a minimally invasive approach may be preferable, especially for correction of a small arteriovenous fistula. Also on the left side, it is common to maneuver in and out of the subclavian-brachiocephalic veins, causing low-pressure extravasation. After vein occlusion and organization, frequently all that is left is an atretic encapsulating fibrous tissue sheath (sometimes mineralized). For a single lead, the telescoping sheaths are, at times, larger than the circumference of this fibrous tissue sheath, and the entire capsule may be included along with the lead in the telescoping sheaths. If two or more leads are present, the circumference of the telescoping sheaths will remove at least one wall of the encapsulating sheath entirely or in segments. The resulting extravasation is not clinically significant with reasonable venous pressures. In those exceptional cases in which the venous pressure reaches a systemic level of 70 to 90 mm Hg, a dissecting hematoma would ensue, causing the same clinical scenario as described earlier. The right-sided vein is contiguous with the right pleura. If the encapsulating fibrous tissue and embedded lead constitute the inferior vein wall, a tear in the pleura will result in hemorrhage into the right thoracic cavity. This can be lethal without immediate surgical intervention. The area is hard to reach through a median sternotomy or an anterior thoracotomy. Because of the time constraints, a median sternotomy with elevation of the clavicle and right anterior chest wall is radical but provides adequate exposure. In less urgent situations, the patient may be repositioned for a less extensive approach. Venous bleeding into the right chest, if not massive, can be insidious and may not be recognized until the onset of cardiovascular collapse. It is not painful, and the signs and symptoms associated with blood volume depletion can be masked by anesthesia or by increased catecholamines compensating for the low filling pressures. With sufficient blood loss, compensation is no longer possible, blood pressure falls, metabolic acidosis develops, and cardiovascular collapse ensues. The SVC passes from the brachiocephalic veins to the right atrium. Along the upper half of the SVC, the lateral surface is contiguous with the right pleura, and

896

Section Three: Implantation Techniques

the lower half is within the pericardium. Loss of integrity of the upper half results in blood loss into the right thoracic cavity; in the lower half, it causes cardiac tamponade. The sequelae of blood loss into the right chest are identical to those described for the right subclavian-brachiocephalic veins. In an emergency, a median sternotomy provides adequate exposure to either the upper or lower half of the SVC. Cardiac Tamponade. Bleeding into the pericardial space causes a cardiac tamponade (pathologic cardiac compression). Small amounts of blood (approximately 200 mL), accumulating rapidly in the pericardium, will cause symptoms. Rapid accumulation increases ventricular wall compliance, decreasing filling of the ventricles and the resultant stroke volume. Rapidly accumulating blood in the pericardial space clots, whereas a slow accumulation does not. Clot formation can localize the compression, and some parts of the heart are more sensitive to compression than others. A 1-cm tear causes immediate tamponade with instantaneous decrease in systolic pressure. A precipitous drop of systolic blood pressure with failure to recover within 2 to 3 minutes is an emergency requiring immediate surgical intervention. The author uses this short time to confirm that the drop in pressure is real and not secondary to a technical issue by viewing the TEE and checking the pressure lines. The patient and operating field are then prepared for a median sternotomy. The TEE is usually definite, but if it is not, and no other cause is apparent, a tamponade must be considered the cause. The time constraints force a decision to be made within 2 to 3 minutes to prevent irreversible brain damage. A median sternotomy is used to decompress the pericardium, manually remove the clot, and surgically repair the tear. Needle aspiration and tube drainage are ineffectual for removing clot. A tear in the lower half of the SVC measuring 2 mm is an example of a slow-onset pericardial effusion. A slow onset of tamponade, while maintaining a blood pressure, allows more time to confirm the tamponade and to make the decision of whether to use a less invasive pericardial drainage procedure. Pericardial drainage tubes inserted into a pericardial effusion are therapeutic in relieving the compression, with immediate restoration of blood pressure. It provides a drainage system for monitoring bleeding and in some cases for blood replacement, cell saver, or direct reinfusion, and it buys time to set up for a corrective surgical procedure. A rushed insertion of a percutaneous pericardial tube without an effusion being present can be a disaster. This is especially true in an enlarged heart if the diaphragm is penetrated and torn during insertion of the tube, creating the need for surgical correction. The only safe way to insert a pericardial tube without clear confirmation of the effusion is under direct vision. The safest approach is through a small subxyphoid incision that opens the pericardium under direct vision. Although the incision is slightly larger than for a percutaneous approach, it is safe. Other conditions mimicking a tamponade are mechanical occlusion of the SVC, lead traction applied

to the right ventricular wall decreasing compliance and filling, tachyarrhythmia, and metabolic acidosis from poor perfusion. Reflex therapeutic actions for a drop in blood pressure during lead removal include pausing lead extraction maneuvers (including lead traction), immediate intravenous administration of a vasopressor (Neo-Synephrine or Levophed), cardioversion of arrhythmias, and administration of sodium bicarbonate in low cardiac output situations. These reflex actions are used throughout any procedure, at any time, to compensate for transient decreases in filling pressure from the causes mentioned. They should be considered routine maintenance and not emergency treatment. The right atrium and ventricle are also contained within the pericardium, and loss of wall integrity from a penetration, tear, or tissue avulsion can cause cardiac tamponade as described earlier. The treatment is the same. There are four mechanisms for tearing the wall not previously discussed: bone spicule laceration of the right atrium, chronic penetration of the right ventricle, fibrous tissue between the electrode and the heart wall, and size disparity between the electrode and the sheath. The first mechanism involves a variation of the encapsulating tissue seen only in some extreme cases of mineralization. An example is a large silicone lead (>10F) that was in place for more than 20 years and making contact with the inferior-lateral atrial wall. The encapsulating tissue was completely mineralized, with “bonelike spicules” embedded within the wall. With any traction on the encapsulating sheath, the “bonelike spicules” acts as a knife blade, causing a surgical incision in the wall. This scenario is rare, but the author has seen it on two occasions. It has not been seen with any of the silicone leads manufactured since the late 1970s. To be safe, a lead bound to the lateral wall of the atrium should be removed with caution. A second mechanism involves the application of countertraction in removing a lead from the ventricular wall. For example, an ICD lead implanted for 12 years is attached to the thin anterior wall of the right ventricle. The tissue at the electrode removed during the extraction contains epicardial fat, indicating a defect in the ventricular wall and suggesting that the electrode had penetrated the wall, most likely during implantation. Unless the lead has perforated and is positioned within the pericardial space, it is undetectable. Fortunately, this is a rare occurrence. This mechanism for tearing the heart has been observed twice by the author. A third mechanism that could potentially tear the heart wall has not been seen by the author. It involves a band of tissue attached to the extracted electrode which is still connected to the heart wall. Continued traction on the freed lead could avulse the tissue or tear the heart wall. When this is seen, the sheath (preferably powered) is maneuvered until the band is cut. With mechanical sheaths, care must be taken with the force used to break the band. This mechanism is readily on fluoroscopy, and once it is recognized, care is taken. A fourth mechanism is caused by a large size disparity between the extraction sheath and the lead.

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

Although this mechanism has been reported, it has not been seen by the author. If an attempt is made to apply countertraction with a sheath larger than the lead, the electrode and the ventricular wall will both be pulled into the sheath, potentially tearing the wall. Once the ventricular wall tissue is pulled into the sheath, it becomes a form of direct traction and not countertraction. For this to happen, the size difference between the lead and the sheath must be sufficient to accommodate the electrode-tissue mass. The cardiac tissue being pulled into the sheath cannot be seen. The only protection is avoiding large size differences between the lead and the sheath. Unfortunately, this is not always possible. For example, the smallest laser sheath is 12F with a 16F outer sheath, so extracting a 4F lead involves a sizable mismatch, and care must be taken. Also, a 9F sheath may have to be changed to an 11F sheath to include calcified encapsulating tissue, resulting in a size mismatch during countertraction. Today, the vast majority of leads are successfully removed through the vein entry site. The current procedures and extraction tools provide a safe, reliable, and effective method of removing these leads from most of the pathologic environments created at the biophysical interface. Recognizing the limits of an extraction approach (i.e., a dangerous situation) and changing to a more appropriate, safe approach is the key to eliminating complications. A national lead extraction database no longer exists, and extraction data are now kept by individual practitioners. Individual data are considered anecdotal, and the data from one individual cannot be extrapolated to another. Consequently, it is impossible to quote meaningful lead extraction statistics, and the author’s experiences must be placed in this context. These comments apply to most of the lead extraction discussions. Remote Vein Sites. A remote vein site is a vein other than the lead vein entry site. A remote vein site requires a remote instrument, such as some type of snare, to manipulate the lead. The femoral vein was the first remote vein site used. During the early evolution of lead extraction, removing leads with mechanical sheaths from the superior veins via the vein entry site was frequently a failure and potentially dangerous. A more difficult approach through a femoral or other remote vein site was more effective and safer. The problems were related to maneuvering the long telescoping mechanical sheaths from the femoral vein to the heart wall and the complexities associated with grasping the lead body. With the advent of powered sheaths, the safety and efficacy of removing leads from the vein entry site improved dramatically. Although most of the leads can be removed from the vein entry site, there are still dangerous situations and lead breakages. In these cases, a remote vein or a transatrial cardiac surgical procedure are the only options. In many cases, the femoral vein approach is still the best remote vein approach. However, other remote veins (contralateral external jugular, internal jugular, or axillary-subclavian-brachiocephalic) may be more suitable and easier to use. Some physicians have developed combination approaches,

897

applying mechanical sheaths from both the vein entry site and a remote vein site. This combination constitutes a safe and efficacious approach that is still used by some extractors today. For cases such as lead breakage, use of a snare to grasp the lead and the subsequent removal of the lead using direct traction or a powered sheath is advantageous. Powered sheaths are usually not long enough for a femoral approach. Consequently, these approaches are confined to the superior veins, via both remote and vein entry sites. The techniques used involve manipulation of snares to grasp the lead body. Although these techniques are simple in principle, they can be difficult to achieve in a timely fashion. Initially, there were no snares designed for lead removal, so a suitable off-the-shelf (“off-label”) snare designed for other purposes had to be found. The author’s criteria for such a snare included the tensile strength to withstand the forces applied during lead extraction, safety of the surrounding tissue when maneuvering the snare, and reversibility of the grasping mechanism. The Dotter snare was the first snare to meet the tensile strength requirements. When the Dotter snare was combined with the Cook deflection snare, the safety and reversibility requirements were met. Later, two snares were manufactured meeting these requirements: Cook developed the Needle’s Eye snare and the Amplatz Goose Neck snare, which are still in use today. A classification of remote vein approaches deviates from the historic use of the IVC (femoral vein) approach. A natural classification is to employ the removed vein (e.g., femoral approach, internal jugular approach). A detailed discussion of remote vein site usage will be made based on this natural classification. The goal is the same as for the vein entry discussion: to define the approach, procedure techniques, and expected results in relation to the pathology and pathophysiology associated with the implanted lead. Femoral Approach. Before the advent of powered sheaths, the femoral approach was used extensively. The indications for its use were failure to extract leads from the superior veins by any technique, lead breakage, and avoidance of the application of excessive force with the mechanical sheaths. The techniques used have not evolved significantly over the past 15 years. The transvenous approach through a femoral vein requires a special sheath set (Byrd WorkStation) that functions as an introducer, as a workstation for manipulation of snares, and as countertraction sheaths (Fig. 21-31). The set consists of an introducer needle, a guidewire, a 16F workstation, an 11F tapered dilator, an 11F telescoping sheath, a Cook deflection snare, and a Dotter basket snare. The workstation serves many functions. Initially, it acts as a protective sheath. The outer sheath prevents the insertion, withdrawal, and manipulation of the inner sheath and snares from damaging the veins or heart. To prevent clot formation, the workstation has a valve (Check-Flo) to continuously irrigate the sheath. The workstation and snares form a reversible loop to pull the proximal portion of the lead out of the superior veins; the workstation also acts as the outer

898

Section Three: Implantation Techniques

B

A

Figure 21-31. Byrd Femoral WorkStation. Telescoping sheaths are passed to the heart. The outer sheath functions as a workstation protecting the vein and heart wall during maneuvering of the inner sheath. Once the lead is secured by the reversible loop, the sheaths are advanced to the electrode to apply countertraction. A, The workstation is packaged with telescoping sheaths, guidewire, and dilator to be inserted like an introducer. The dilator is removed and the snares inserted. B, Check-Flo valve with the Dotter snare and deflection catheter. The deflection catheter requires a special, nondisposable handle to manipulate the tip of the deflection catheter, forming a reversible loop. The Dotter snare can be maneuvered using a plastic knob or a surgical clamp.

telescoping countertraction sheath. The safe insertion and removal of the workstation is a prerequisite for an extraction procedure through a femoral vein. The workstation is 16F and must be inserted with care. Fluoroscopic monitoring is mandatory. Once the guidewire is passed into the heart, the workstation with its tapered dilator must be maneuvered through the iliac vein and IVC and into the right atrium. The route can be circuitous, especially from the left side. In rare cases, the curvature may be too sharp for the stiff dilator to follow the guidewire. Forcing the dilator in this situation is unsafe, and the approach should be abandoned. A torn retroperitoneal iliac vein or IVC is a serious complication. Once the workstation is inserted, irrigation fluids are run continuously through the Check-Flo valve to prevent clotting. Pulling leads down and out of the superior veins has been remarkably successful. Many leads freed from the heart cannot be pulled up through these same veins. The lead binding forces caused by the circumferential bands of fibrous tissue are the same in both directions. The author’s hypothesis for this difference is the free upward mobility of the mediastinal structures and the inability to pull the superior veins downward. The mediastinum is easily pulled upward, compressing the veins. This compression of the fibrous tissue bands around the lead as they bunch together is postulated to increase the binding strength. The superior veins cannot be pulled downward, and irreversible lead slippage through the binding sites is not an issue. Cardiac function is not influenced by pulling leads downward out of the superior veins. Failures to extract the proximal lead from the superior veins using this technique are rare and are usually caused by other complicating factors. Examples of complicating factors are thrombosis of the superior veins and excessive fibrosis around the lead caused by a previous extraction attempt that

left the conductor coil exposed or pulled the lead taut against the heart and vein wall. Such leads are removed using approaches such as the transatrial approach, which are reserved for more complicated extractions. To apply countertraction through the workstation, the proximal end of the lead must be entangled in the basket snare. This is accomplished by placing the basket snare in close proximity to the lead and rotating it slowly. The lead will flip into the basket. The basket closes when the 11F sheath is advanced over the snare. For most leads, the snare and lead are pulled into the 11F sheath. The workstation and 11F sheaths are then worked in a telescoping fashion to a point near the electrode. At this point, countertraction is applied to extract the electrode, as previously described. Removal of the workstation must be carefully performed. Once the lead is extracted, clot and debris may be attached to the end of the tubing. If this material dislodges in the femoral vein entry site, it can act as a nidus, forming a thrombus or initiating thrombophlebitis and its sequelae. To prevent this complication, blood is aspirated during the withdrawal of the workstation. If the entry site does not bleed freely after withdrawal of the workstation, a surgical exploration of the vein is recommended. Bleeding is controlled by applying pressure over the vein entry site after withdrawal of the sheath and during Valsalva maneuvers induced by the anesthesiologist. A suture or staple is required to close the skin. A potential complication is thrombophlebitis and pulmonary embolus. Fear of this complication was the incentive for the workstation. Postoperatively, anti-emboli stockings (pneumatic, if possible) and subcutaneous heparin (5000 units twice a day) are the only precautions taken. The technique for grasping the lead in a reversible loop using a Cook deflection catheter and a Dotter snare is more complicated. Once mastered, it is an

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

effective method of performing precision extraction. For example, a patient may have six leads in the heart. Two new leads are connected to a pacemaker and are to be saved. Four leads are abandoned and are to be extracted. The abandoned atrial and ventricular leads can be extracted, leaving the newly implanted leads intact. The IVC approach allows this level of precision. Superior Vein Approaches. The other transvenous approaches include the external, internal jugular, and axillary-subclavian-brachiocephalic veins, usually on the contralateral side. The telescoping sheaths used with the femoral approach are mechanical (powered sheaths are not available). The creation of a conduit by the powered sheaths from the vein entry site encourages the use of snares passed through these sheaths to remove leads (e.g., broken free floating leads). A snare is used to grasp the lead, and, if needed, sheaths are passed over the snare to the lead. The difference is in the extraction tools available. For example, the femoral workstation is too long and is not applicable for the superior vein approaches. Using standard introducer techniques, a guidewire is inserted and telescoping mechanical sheaths are passed into the SVC or heart. These sheaths then act as a workstation for passage and manipulation of a snare. The same snares used for the femoral approach are usable with the superior vein approaches. The Needle’s Eye Snare has a shorter, easier to use version for the superior veins. Once snared, the lead can be removed using indirect traction, mechanical sheaths, or powered sheaths. The noose snares are also effective from the superior approaches. All of the principles and guidelines for lead extraction apply to these approaches. Combined Vein Entry and Remote Entry Approaches. To some degree, most remote approaches are a combination of a vein entry site and a femoral vein approach. An extensive effort may have been used to free the lead from the superior veins; this was always in preparation for the intended use of the remote approach. From its inception, some physicians championed the use of a combined approach. Initially, the vein entry approach and femoral vein approach were combined, sometimes with two teams working, one for each approach. The goal was to free the lead by applying the extraction techniques from both approaches and removing the lead in an opportunistic fashion (the resultant easiest approach). It was not uncommon to have the EP working from the vein entry site and a radiologist from the femoral vein site. Although this approach was successful, it was overkill for most lead extractions. This was especially true once powered sheaths became available. A combination of the vein entry site approach and a right internal jugular approach is used by some physicians as their primary approach to lead extraction. They use snares and/or a grasping instrument passed through the internal jugular vein into the SVC to pull the lead out of the axillary-subclavian-brachiocephalic veins and, if necessary, to pull the lead into the atrium. Mechanical sheaths are then applied to complete the extraction (to date, powered sheaths are not used).35

899

The techniques used with remote approaches can be applied through the vein entry site. A lead may have broken or been cut, retracting into the axillarysubclavian-brachiocephalic veins, floating in the SVC or heart, or migrating into the pulmonary veins. If the lead can be reached and grasped inside the axillarysubclavian-brachiocephalic veins by a surgical instrument, it can be pulled out of the vein entry site and secured with a suture or a locking stylet, or both. The lead can then be extracted using the vein entry approach. If the lead cannot be grasped by an instrument but a snare can be passed into the SVC from the vein entry site or a contralateral vein, the lead is grasped and removed using mechanical and/or powered sheaths. Before powered sheaths became available, the femoral approach was commonly used by the author, but today it is rarely used. Surgical EPs have the transatrial approach as a viable alternative. Regardless of the frequency of use today, the transfemoral approach is a requisite procedure required for managing devicerelated complications. Surgical Approaches The surgical approach to lead extraction is a cardiac surgical procedure. There are three procedures used for transvenous endocardial implants: the transatrial approach, right ventriculotomy, and an open heart procedure using cardiopulmonary bypass (CPB). In the author’s opinion, these procedures should be performed only by experienced cardiac surgeons. The technical and patient management skills possessed by an experienced cardiac surgeon negate the normal risk associated with the increased magnitude and complexity of the surgical procedure. The transatrial procedure is a general procedure that can be used for both extraction and implantation of leads. It can be used instead of the transvenous remote approach for lead extraction. It has the added advantage of being an implantation site that bypasses the SVC and IVC. The right ventriculotomy is a technique for removing leads from the right ventricle in special situations. Today, an open heart procedure is reserved for removing thrombotic material (usually infected) from the right atrium and ventricle and for removing leads implanted in the left ventricle. Transatrial Approach. The transatrial approach was first described by the author in 1985.36 This surgical EP procedure is suitable for intracardiac EP implantation, explantation, and ablation procedures. The only disadvantages are the morbidity associated with surgical thoracic pain and the fact that a medical EP must work with a cardiac surgeon to perform this procedure. It is a primary approach for noninfected patients who are candidates for a transatrial lead implantation. Younger patients with occlusion of one brachiocephalic vein or with an SVC syndrome have the old leads extracted through a transatrial approach, followed by implantation of new leads. The advantage of the transatrial approach is the ability to remove leads that are not accessible or removable by the SVC or IVC approach.

900

Section Three: Implantation Techniques

procedure, or the pursestring is tightened, tied, and abandoned. After transatrial extraction, patient management is more involved than for the transvenous extraction techniques. The pericardium must be drained, in most cases by a closed drainage system such as a chest tube, if the pleural space is free; by a mediastinal tube, if the pleural space is obliterated; or by a Jackson-Pratt closed drainage system, if both the pleural and pericardial spaces are obliterated. The thoracic cavity is occasionally entered and a chest tube inserted to drain both the pericardium and pleura. These drainage tubes are removed in 2 to 3 days. The procedure-related morbidity increases the hospital stay by 1 or 2 days, compared with a transvenous procedure. Patients must be managed in an intensive care setting until the thoracic pain sequelae are controlled. Figure 21-32. Transatrial approach. This is a minimally invasive cardiac surgical approach. The third or fourth cartilage is removed, and a pursestring suture is placed in the right atrium. A pituitary rongeur is inserted through the atriotomy incision and used to grasp the lead. The proximal portion of the lead body is removed from the superior veins by direct traction, and the distal lead is removed by mechanical and/or powered sheaths and countertraction.

Most of the transatrial extractions are failures of the IVC approach. Rarely, failure of a SVC approach will lead directly to a transatrial approach (e.g., when the workstation cannot be passed via the femoral veins into the heart). Infected patients who are candidates for transatrial lead implants will have the leads extracted by an SVC or IVC approach and the transatrial implantation performed after the infection is properly treated. The transatrial approach is performed as originally described, through a limited surgical incision on the right anterior chest wall (Fig. 21-32). The right atrium is exposed by removal of the third or fourth right costal cartilage (determined by fluoroscopy). The pericardium is opened and suspended, and a pledgeted pursestring suture placed in the right atrium. If the pericardium has been obliterated from a previous procedure or disease process, a small region of the lateral wall of the right atrium is dissected free. Using fluoroscopy, a pituitary biopsy instrument is inserted through the purse string. The lead body is grasped in the atrium and pulled out. The lead is then cut, extracting the proximal and distal segments separately. The proximal portion of the lead can usually be pulled out by direct traction. The only limitation to the force employed is the tensile strength of the lead. On those occasions where the tensile strength is insufficient, telescoping powered sheaths may be required. The distal segment is extracted by inserting a locking stylet, advancing telescoping powered sheaths to the wall, and removing the electrode from the wall using countertraction. This procedure is repeated for each lead to be extracted. On completion of the lead extraction, the atriotomy site is used to insert new leads, or to perform another EP

Right Ventriculotomy. A right ventriculotomy is a cardiovascular surgical procedure. This approach is rare for the author to use, and its frequency of use by others is unknown. Because this procedure is virtually unknown, a cursory discussion of indications for use is in order. Initially, it was reserved for infected broken leads retained in the right ventricle that were not reachable by the other approaches. These are fragile leads that break near the ventricular wall or within a fibrous tissue tunnel along the ventricular wall and are impossible to grasp using transvenous or transatrial techniques. It is also used in conjunction with lead penetration requiring lead removal and a repair of the heart wall, and possibly for an emergency tear in the ventricular wall caused by an attempt to remove a lead. In the latter, the defect is turned into a ventriculotomy site for lead extraction. The first ventriculotomy procedures were performed by the author in the late 1980s. The heart is exposed through a median sternotomy incision. The heart is then elevated on a pad, exposing the right ventricle. The tip of the electrode is localized by fluoroscopy and by using needles for triangulation of the electrode. A pursestring suture is placed around the electrode, and a ventriculotomy incision is made to the electrode. The electrode is grasped with a clamp and pulled out of the heart. Because the lead segment is being pulled in the direction of the tines, the tines slip out of the embedding scar without resistance. Today, the procedure is performed through a minimally invasive incision on the anterior surface of the left chest through the fifth intercostal space. The pursestring and extraction techniques are the same. The author performs no more than one or two of these procedures a year. This is a stressful procedure, and it is not known how many procedures need to be done to acquire some comfort level. As long as there is a need for the procedure, continuing attempts will be made to perfect it. Open Heart Procedure. The concept of using CPB to perform an open heart procedure is intuitively obvious. It is a standard cardiac surgical procedure involving a median sternotomy incision. Because of its familiarity, the issues associated with this approach need no further discussion. Right-sided leads were initially

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

901

Figure 21-33. Left-sided implantation. Transvenous implantation in the left atrium and/or ventricle or a transarterial implantation through the aortic valve into the left ventricle is rare. A transvenous implantation through an atrial septal defect (ASD) such as a sinus venosus defect is the most common (left). The anatomy guides the leads into the left atrium. Extraction of these leads is dangerous because of the potential for an arterial embolization of debris. The author performs these procedures with the use of cardiopulmonary bypass, repairing the congenital defect after lead extraction (right). After extraction, the left ventricle is monitored by transesophageal echocardiography, and all debris is removed through an incision in the aorta before bypass support is removed. The new pacemaker or ICD is implanted transatrially to avoid passing the lead through the junction of the superior vena cava and atrium.

removed by direct traction and under direct vision during open heart surgery. Transvenous and transatrial lead extraction techniques evolved to eliminate the need for an open heart surgical procedure. Implantation of leads in the left atrium and ventricle is a notable exception (Fig. 21-33). Leads are implanted into the left ventricle in two ways: through a congenital atrial or ventricular septal defect, or retrograde through the aortic valve. The author is amazed at the diversity of opinions expressed concerning the management of these leads. All consider the presence of left-sided leads and embolic symptoms an urgent indication for lead removal. Most consider their presence, with the potential for a complication, to be an urgent indication for lead extraction. A few physicians believe that, in the absence of complications (cerebrovascular accident, coronary artery occlusion, infarction of another other organ, or the sequelae of a peripheral embolus), leaving the leads intact is an acceptable option. The rationale is that the extraction procedure is more dangerous than the presence of leftsided leads. In contrast, the author believes that the presence of these leads in the left atrium and/or ventricle is an indication for urgent lead removal. The views on lead removal are just as varied. Some believe the leads should be removed; the chambers debrided of vegetation; and congenital defects, if present, repaired with the use of CPB. Others think it is safe to extract the leads using the established right-sided techniques and protecting the brain from emboli by compressing the carotid arteries when necessary. The author believes the leads should be removed using total bypass with the aorta cross-clamped, and congenital

defects should be repaired. If the second approach has any merit, it would be in removing newly implanted leads. If the newly implanted leads are proved to be free of vegetation and the procedure is monitored using TEE, it is probably safe to extract using direct traction. Specific data are not available on the numbers of physicians using these opinions. The author’s experience is based on the management of 10 consecutive patients, most of whom had chronic leads and were symptomatic. Symptoms included transient ischemic attack and/or cerebrovascular accident, and vegetation was apparent in symptomatic patients. All but one patient had leads passing from right to left through an atrial septal defect. One patient had a lead that was implanted through the subclavian artery and passed retrograde through the aortic valve into the left ventricle. The procedure included establishing CPB, using cardioplegia to stop ventricular contraction, cross-clamping the aorta as needed, extracting the leads using conventional tools, debriding the chambers, repairing the septal defects, continuously observing the left atrium and ventricle by TEE, and carefully restarting the heart with protection from inadvertent ejection of embolic material. A new transatrial right-sided implantation was performed off bypass with the use of fluoroscopy and the conventional approach. With this technique, the leads were all removed, the congenital defects were repaired, and the new devices were implanted without sequelae. From the encapsulating fibrous tissue and vegetation present, the author cannot envision a less complex procedure that would be safe. To develop a safe technique not employing an open heart procedure is difficult to imagine.

902

Section Three: Implantation Techniques

Special Situations The most interesting development in lead extraction has been the interest in placing leads into the coronary sinus and cardiac veins. Left atrial leads for sensing and stimulation and defibrillator leads to improve defibrillation thresholds are nothing new, but they were not frequently used in the past. Cardiac resynchronization therapy has made this a frequent and evolving issue. The experience of extracting these leads has so far not been very interesting. Partially this is true because of the short implantation durations and partially it is related to the simple and extraction-friendly designs of the leads. Smooth, thin, single-diameter unipolar leads with good tensile properties are all characteristics that enhance the safe and quick removal of the leads. However, because physicians and manufacturers have been concerned with the stability of these leads in the cardiac veins, they have been developed into more complicated shapes, multipolar, and with fixation mechanisms. These leads are not likely to be so easily removed. As an evaluation of extraction tools and a novel lead design technique to reduce the barriers to extraction of complicated leads, the author participated in a study that used a sheep model with atrial defibrillator leads placed into the coronary sinus to the great cardiac vein. The leads, originally designed for atrial defibrillation in the METRIX atrial defibrillator (InControl, Redmond, Wash.), were modified but kept their pigtail configuration for lead stability. Three configurations—one without modification of the defibrillation coil, one with medical adhesive backfill under the defibrillation coil, and one covered with ePTFE—were implanted in sheep and were subjected to extraction at either 6 or 14 months. The model proved to be an excellent one for developing profound fibrosis, and the unmodified leads were almost impossible to remove without hemopericardium. The medical adhesive backfill was much better, and there was almost no trouble removing the ePTFE-covered leads. The study also demonstrated that laser sheaths were dangerous in the coronary sinus, because the sheath approximated the size of the vein, and that a special 7F electrosurgical extraction sheath was relatively much safer to use. During the procedures, the electrosurgical sheath was rotated away from the pericardial and toward the myocardial surface.37 Overall, coronary sinus lead extraction has not become a clinical issue, but it will, and when it does, hopefully the implanters will have made sound decisions that promote the extraction of these leads. From the sheep experience, it appears clear that implanters should at all costs avoid construction that allows for tissue ingrowth. Lead Implantation Skills Implantation skills are possessed by all who practice device implantation. The implantation skills required to manage device-related complications are, at times, more demanding. In some situations, conventional implantation techniques are not suitable, and a more exotic approach is required, such as a transvenous

femoral approach or a cardiac surgical transatrial or epicardial approach. Device implantation is covered in a rigorous manner in other chapters of this book. Only those principles and guidelines believed to be essential to managing device-related complications are reviewed here. Because the goal is to reimplant without creating another acute or chronic complication, emphasis is placed on the situations that cause complications and the techniques used to prevent them. To avoid confusion, implantation techniques are separated into components: reimplant location and lead insertion techniques, endocardial implantation techniques, and epicardial lead implantation techniques. Focusing on each of these components provides a natural order for discussion. Reimplant Location and Lead Insertion Techniques Reimplant location and implantation techniques are the primary determinants of implant longevity relative to a specific lead. For implanted leads to survive without a complication for 30 to more than 60 years requires a remarkable combination of events. Implantation decisions are believed to be one of the key factors influencing the duration of an implant. Reimplantation locations include those sites that are accessible by transvenous approaches and cardiac approaches. Picking the appropriate site should be based on reason and not expedience. For example, questions such as, “Can a temporary lead be implanted after debridement at an infected vein entry site?” “Is the extraction site suitable for lead reimplantation?” “Which remote vein is useable?” “Do I have the knowledge and skill level to reimplant in this situation?” and “What are the long-term consequences of using a particular implant approach?” must all be answered in a satisfactory manner. The answers to these types of questions help prevent acute or chronic reimplantation complications. The only way to answer these questions is to have thorough understanding of the potential complications associated with various reimplantation locations and techniques. Most complications are related to anatomic considerations, and a presentation based on the anatomy is beneficial. Knowledge of complications based on anatomic considerations allows natural and rational reimplantation decisions. To reimplant a lead at the lead extraction site, if appropriate, is natural and convenient. The goal is to use the old, damaged implant site for as long as possible. The most versatile technique is to insert a guide wire through the extraction sheath after removal of the old lead. The extraction sheath is then removed, leaving the guidewire in place. A conduit now exists for reimplantation using a long introducer set. A useful guideline is to keep a guidewire in place throughout the procedure for conduit security. Although some veins are patent, many are not, and the conduit created by the extraction sheath is a natural channel for reimplantation. To avoid losing this conduit, always insert a guidewire before removal of the extraction sheaths,

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

and use an introducer that is long enough to reach the right atrium. The long introducer ensures lead passage through obstructions in the brachiocephalic vein and proximal SVC. Loss of the conduit requires the use of an alternative reimplantation approach. An alternative technique is to leave the extraction sheath in place and insert the new lead and guidewire through the extraction sheath. Then, to remove the sheath, it has to be split (destroyed). Split-sheath technique cannot be used with powered sheaths. Powered sheaths can be used to remove more than one lead and cannot be destroyed. Mechanical sheaths are usually damaged and not reused, encouraging the split-sheath technique to avoid the use of a long introducer. Permanent lead reimplantation through the vein entry site into the axillary-subclavian-brachiocephalic veins is contraindicated in the presence of a devicerelated infection, vein occlusion with poor collateral drainage, damage to arterial structures from improper implantation, or lead exclusion from the vein. These four contraindications are intuitively obvious. A lead placed back through a carefully debrided infected site will probably get infected. The one exception is the implantation of a temporary lead. Once the area is debrided, this is safe for at least 1 week. The remaining three conditions are contraindications for all types of leads due to the potential for propagation of vein thrombosis toward the arm, arterial bleeding from arteries damaged during the old lead implantation (healed or small arteriovenous fistulae), and the erosion of old leads into an artery. The rationale for not

A

903

implanting a lead in the presence of arterial damage is clear, but venous occlusion with poor collateral flow needs further discussion. Venous occlusion is a common occurrence at the vein entry site due to tissue injury and initial clot formation (Fig. 21-34A). Venous occlusion in this area is rarely a complication, because the collateral venous drainage is usually adequate to support all activity levels. Unfortunately, the exact incidence is unknown. Occasionally, a patient will have poor collaterals and be symptomatic with pain and swelling usually in the arm. Placing a lead in this lowflow environment constitutes an ideal condition for continued vein thrombosis with retrograde propagation. Symptomatic patients are candidates for a venous stent (see Fig. 21-34B). In the past, placing a lead through a stent was considered unwise, but it has been done successfully with no reports of adverse consequences.20 Placement of leads first, followed by a stent, leaves the lead pinned against the vein wall and difficult to extract. Treatment of adult patients with transposition of the great arteries after childhood surgical atrial baffle construction has yielded SVC stenosis in many, often after pacemaker or ICD placement. Because trapping of the leads is a far worse choice, extraction and reimplantation after stenting of the stenosis has been become the treatment of choice. However in the author’s opinion, the presence of an occlusion with poor collateral drainage is a contraindication to lead implantation. The presence of a contraindication related to the biophysical interface or loss of the conduit after lead

B Figure 21-34. Occlusion of subclavian-brachiocephalic veins. Occlusion of the subclavianbrachiocephalic veins is common and is rarely symptomatic. Acute occlusion may cause transient signs and symptoms; chronic occlusions are usually asymptomatic. Collateral flow is rich in this area and quickly compensates. A, Partial occlusion of the subclavian vein and total occlusion of the brachiocephalic with sufficient collateral flow. Implantation through a conduit created by lead extraction is not contraindicated. B, Occlusion of the subclavian vein with little collateral flow. An occluded subclavian vein must have adequate collateral vein formation and blood flow. The subclavian vein is the most commonly occluded superior vein. The presence of adequate collateral vein flow is a criterion for reimplantation through the conduit created during the lead extraction. If the patient is symptomatic, a stent may be effective, but implantation of a permanent lead through a created conduit could cause retrograde vein thrombosis.

904

Section Three: Implantation Techniques

extraction necessitates reimplantation through a transvenous, transatrial, or epicardial approach. Use of the external jugular and internal jugular tributaries to the axillary-subclavian-brachiocephalic veins is included in the ipsilateral contraindications. Options for location and implantation technique depend on the status of the biophysical interface and the patient’s age. Most patients have a contralateral vein in good condition, and it seems natural on an abstract level to use this vein. The technique for insertion of a new lead into the contralateral vein is the same as for inserting a new lead for a primary implantation. Unless the physician is experienced in managing device-related complications, an implantation technique based on preconceived beliefs acquired from training and other experiences may not meet the desired goal of having the highest probability of achieving a long-term, complication-free implant. The following discussion of implant technique, based on the anatomy of the axillary-subclavian-brachiocephalic veins and surrounding musculoskeletal system, is derived from the author’s opinion and interpretation. Although the discussions are generally applicable to either the right or left side, exceptions will be noted. Transvenous Approaches Contralateral Veins. The contralateral veins include the cephalic, subclavian, brachiocephalic, internal jugular, and external jugular. The internal and external jugulars are of interest only if the proximal subclavian vein is occluded. The author has one relative and

A

two absolute contraindications to contralateral vein implantation. These contraindications are based on the author’s experiences and the availability of surgical procedures. Medical EPs should be familiar with the logic and, whenever possible, should work with a cardiac surgeon. Sending the patient to a cardiac surgeon without EP experience is usually unsatisfactory. The relative contraindication involves young patients who have an occluded ipsilateral subclavianbrachiocephalic vein (Fig. 21-35A). These patients’ internal and external jugular veins do not drain into the subclavian-brachiocephalic veins. Consequently, all the venous drainage is by collateral flow. Although the collateral flow may be adequate, it is a dynamic pathologic condition that can change with time. For example, the initial collateralization is usually through neck veins. With time, this can shift to the veins on the chest wall. If the transition is not smooth, the patient becomes symptomatic. Implantation in the contralateral vein could result in an occlusion of the contralateral subclavian-brachiocephalic veins, causing the equivalent of an SVC syndrome (poor drainage of the head and upper extremities). This potential complication should be prevented whenever possible, especially in younger patients. It is easier to prevent than to correct. Partial occlusion of the SVC is an absolute contraindication to venous reimplantation (see Fig. 21-35B). Reimplantation from either side through the partial occlusion has a high probability of completely occluding the SVC. An occlusion occurring over a short

B

Figure 21-35. Contraindication to contralateral implantation. Occlusion of the ipsilateral subclavian-brachiocephalic veins is frequently an indication for a contralateral implantation. A, A relative contraindication to contralateral implantation is a young patient at risk for a similar situation on the contralateral side, creating the equivalent of a superior vena cava (SVC) occlusion. B, Ipsilateral occlusion in conjunction with a partial occlusion of the proximal SVC is an absolute contraindication, because of its potential to completely occlude the SVC. The author uses the transatrial implantation approach to avoid the superior veins in these situations.

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

905

B

A

Figure 21-36. Occlusion of the superior vena cava (SVC). This 8-year-old occlusion of the SVC has inadequate collateral vein formation for an active young patient. A, A venogram demonstrating occlusion of the SVC, collateral veins, and implanted atrial and ventricular leads. B, The brachiocephalic vein and SVC are shown to be patent 6 months after surgical correction. The veins are repaired by debridement of the organized intravenous inflammatory tissue and a vein patch angioplasty. The patient is symptom free and has a marked reduction of the collateral venous drainage.

period of time will cause an SVC syndrome; if it occurs over a long period of time, the symptoms will depend on the adequacy of the collateral drainage and the activity level of the patient. Total occlusion of the SVC is an intuitively obvious contraindication (Fig. 21-36A). These patients are candidates only for a transatrial, epicardial, or femoral approach. Only the transatrial approach can duplicate a transvenous venous superior vein approach. The author repairs SVC occlusions in all physically active patients (see Fig. 21-36B). Initially, these procedures were performed through a median sternotomy incision. Today, they are performed using a right parasternal approach similar to the transatrial approach. These repairs are done without CPB through a minimally invasive procedure and have been successful. Despite the repair of the SVC, the pacemaker or ICD must be implanted using an alternative approach (e.g., transatrial). The other absolute contraindication is the presence of anomalous venous drainage (Fig. 21-37). For example, abandonment of a right ipsilateral implantation and substitution of a contralateral implantation through a persistent left vena cava draining the left contralateral veins into the coronary sinus is not acceptable. The implantation of a lead through this congenital abnormality is possible, but difficult. Thrombosis and subsequent occlusion of the persistent vena cava are equivalent to occlusion of the brachiocephalic vein, and extraction of these leads is difficult and dangerous.

Figure 21-37. Anomalous drainage of the superior vena cava. Anomalous drainage of the superior veins is in many cases a contraindication to superior vein implantation, although it is technically possible. Implantation in a persistent left vena cava is contraindicated in the presence of an occlusion of the right subclavian-brachiocephalic vein. Implantation in any superior vein is contraindicated in the presence of totally anomalous drainage of the superior veins. The rationale for these statements is the perceived greater risk of occlusion of anomalous veins after lead implantation.

906

Section Three: Implantation Techniques

Because of these two potential problems, the author elevates this type implantation to an absolute contraindication. The same logic applies to initial implantation or reimplantation into a totally anomalous venous drainage, usually into the IVC. The difference is that a subsequent occlusion of the totally anomalous vein is equivalent to a SVC occlusion. Axillary and Cephalic Veins. The axillary vein is the large vein that carries most of the venous drainage from an upper extremity. It begins at the confluence of the brachial and basilic veins near the teres major and subscapularis muscles. The axillary vein changes its name to subclavian vein as it crosses the outer border of the first rib. The cephalic vein drains into the middle to distal portion of the axillary vein. The distal axillary vein and cephalic veins are considered safe insertion sites. An occlusion of the axillary vein at the confluence of the arm veins is a devastating complication. It is almost impossible to acquire acute collateral circulation sufficient to support the blood flow from the arm. The resultant venous distention causes a swollen, painful extremity. Any increase in cardiac output to the extremity from activity and/or anxiety makes the situation worse. In some cases, collateral circulation capable of supporting resting blood flow develops after many years. There is no known treatment for this complication. Anticoagulation helps prevent further thrombus formation and worsening of the situation. To help prevent this complication, the proximal portion of the axillary vein should never be used as an implantation site. Cephalic vein implantation is argued by many to be the gold standard. It is a surgical cutdown on the vein within the deltopectoral groove. Insertion of the lead at this site is safe (i.e., no chance for a pneumothorax or hemothorax), and the lead passes over the first rib in the subclavian groove. A diminutive vein that is not suitable for a lead insertion can always be instrumented with a guidewire. This statement needs to be qualified: “always” is correct in the author’s experience. However, it can be a technically challenging procedure requiring a variety of guidewires, fluoroscopy, and a venogram to be successful. Fortunately, the introducer approach offers an alternative to this approach. A small artery runs parallel to the cephalic vein. If it bleeds and is not controlled, a high-pressure hematoma can develop, applying pressure to the proximal portion of the axillary vein. External pressure can occlude the vein and cause vein thrombosis. This complication can be prevented by suture-ligation of the artery if it appears damaged or bleeds. It should be mentioned that safe insertion into the distal axillary vein is relative. Occlusion near the vein entry site is common and would not affect the drainage from the arm. However, retrograde extension of clot from a thrombosed cephalic or subclavian vein into the confluence of the arm veins can result in an irreversible complication. This is rare. Usually, the collateral flow from the axillary vein is significant, and the fast venous flow prevents further retrograde propagation of a clot.

As the subclavian vein passes from the lateral to the medial border of the first rib, it is still outside the thoracic cavity (extrathoracic). Starting at the medial aspect of the first rib, it lies within the thoracic inlet and is intrathoracic. The thoracic inlet is the apex of the thorax, which is separated from the thoracic cavity by the pleura. The subclavian vein passes over the extrathoracic portion of the first rib in the subclavian vein groove and anterior to the attachment of the anterior scalene muscle. The distal portion of the axillary vein and the proximal portion of the subclavian vein on the first rib are usually free from implantation complications. The intrathoracic approach was the initial introducer procedure. The goal was to cannulate the subclavian vein within the thoracic inlet. The technique is to pass the introducer needle under the clavicle and over the first rib and puncture the anterior surface of the subclavian vein. In an idealized approach, the introducer needle can be passed from within the pocket, cannulate the subclavian vein without entering the pleura, and avoid binding the lead between the clavicle and first rib. Despite the plethora of literature on this subject, experience and judgment are the only predictors of success. Two extreme techniques introduced to help prevent complications were the “medial” and “lateral” approaches (Fig. 21-38). The medial technique was to pass the needle between the clavicle and first rib as medial as possible, through the costoclavicular ligament. This technique placed the needle into the anterior portion of the thoracic inlet, away from the pleura, helping prevent a pneumothorax. The disadvantage of

Figure 21-38. Intrathoracic (thoracic inlet) introducer approaches. Two introducer approaches, medial and lateral, demonstrate the issues associated with inserting introducers in the intrathoracic portion of the subclavian vein. The medial approach is safe, but the leads are subjected to crush (insulation failure) and binding (conductor coil fracture). The lateral approach is free from lead crush and binding, and it is safe when the needle is properly directed. Because of variation in chest wall anatomy, extensive experience is needed to avoid misdirection with the introducer needle and potentially lethal complications such as pneumothorax, hemothorax, or a transpulmonary-vein implantation.

Chapter 21: Managing Device-Related Complications and Transvenous Lead Extraction

this approach is the binding and compression to which the lead is subjected, which increases the probability of insulation and/or conductor coil failure. The lateral approach was designed to place the introducer needle in the anterior portion of the thoracic inlet, free from the binding and compression forces present in a more medial approach. Although this lateral approach was successful, it was more complicated. A lateral safe zone was defined by the author using fluoroscopy and based primarily on the presence of adequate spacing (>1 cm) between the clavicle and first rib.38 This spacing was an attempt to ensure lead passage in the subclavian groove. An anatomic variable determined the location of the safe zone, but training, judgment, and experience were still factors in its application. The extrathoracic approach was published along with the lateral safe zone technique. It was presented as an alternative technique suitable for all situations. At the time it was presented, it was considered a more radical approach. Misadventures with an introducer needle within the thoracic inlet can cause pneumothorax (needle puncture of the lung) and/or hemothorax (needle tear of the pleura and subclavian artery and/or vein). Also, insertion of a lead into the right brachiocephalic-SVC vein from the right side is difficult to do without passing through the apex of the right lung, causing both a pneumothorax and a hemothorax. Removal of an introducer or lead accidentally inserted within the subclavian artery can cause bleeding into the thoracic inlet, with the potential for a lethal rupture into the thoracic cavity. All of these complications are potentially lifethreatening if not recognized and treated in a timely fashion. Treatment may include major thoracic and cardiovascular procedures. Because of the variations in anatomy, the only way to guarantee these complications can be avoided is to stay out of the thoracic inlet. Two complications are associated with leads placed medially between the clavicle and first rib into the thoracic inlet. The first complication is related to lead binding. Passage of the introducer needle through the subclavius muscle or the costoclavicular ligament can result in calcification of those structures. Lead binding occurs when the lead body becomes entrapped in the encapsulating calcified tissue. If a lead with coaxial cabling is rigidly bound, the helix cannot bend to relieve stress, and the conductor coil at each end of the binding site becomes deformed. If the wire is configured as a helix, bending the helix back and forth does not affect the wire because the coils of the helix flex, preventing deformation of the wire itself. Binding of the helix prevents it from flexing, and the wire will be deformed at the junction between the bound and unbound helix. The second complication is related to pinching (crush injury) of the lead between the clavicle and first rib. This type of compression injury damages the lead body insulation. The loss of integrity of the polymers caused by compression injury results in inner insulation failure, resulting in make-break signals and short-circuiting between the inner and outer coils. Outer insulation failure is of consequence only if the exposed wire senses electrical signals, inter-

907

fering with communication between the pulse generator and the heart. The intrathoracic approach has been reallocated to a support role or an alternative approach for most physicians managing device-related complications. Although the extrathoracic approach is favored by most, there is still a place for the intrathoracic approach. For example, this approach combined with venography is useful for passing localized occlusions in the axillary and proximal subclavian veins. The distal subclavian and proximal brachiocephalic veins can be instrumented in some situations. However, to eliminate both the acute and the chronic complications associated with the thoracic inlet, an extrathoracic approach is recommended by the author. There are two extrathoracic approaches: introducer and cutdown. The author believes that the extrathoracic introducer approach is the safest, most efficient, and most versatile approach available. The lead is inserted into the subclavian vein as it passes over the first rib in the subclavian vein groove. This routing eliminates crush injury, pneumothorax, and intrathoracic arterial or venous bleeding associated with the thoracic inlet. This is a fluoroscopic technique; it should not be done without direct visualization of the first rib and the introducer needle in an anterior-posterior view. For orientation, the needle is passed from the pocket area to the clavicle. The needle is then passed posterior to the first rib. Without visualization, directing the needle posteriorly and missing the first rib would result in puncturing the lung and a pneumothorax. To maintain orientation, contact should actually be made with the first rib, followed by aspiration on withdrawal. The needle is marched anteriorly or posteriorly along the first rib until the vein is located. The subclavian artery marks the posterior boundary of the posterior march. The subclavian vein is always anterior to the artery. A direct puncture of the artery over the first rib is of no consequence because it is easy to apply pressure, if necessary, for a few minutes to prevent hematoma formation. Once the vein is located, a guidewire is passed to the heart, and a long curved introducer is inserted into the right atrium. The location of the subclavian vein on the first rib is based on fluoroscopic landmarks. In an anteriorposterior view, the clavicle appears to cross over the first rib. The subclavian vein is in the region where the outer border of the clavicle crosses the rib (Fig. 21-39). To visualize the anatomy, the first rib is divided into three equal segments, starting anteriorly at the sternum. The clavicle most commonly crosses the first rib in the second segment, indicating that the subclavian vein most commonly crosses in the second segment. However, a few are located in the third segment or, rarely, in the first segment. This coincides with the three general chest wall types distinguished by the position of the shoulder.39 A descriptive name such as an anterior rotation, normal rotation, and posterior rotation is sufficient to describe the visual appearance of the chest wall types. These terms correspond to passage of the subclavian vein over the first, second, and third segments of the first rib, respectively. Because

908

Section Three: Implantation Techniques

Figure 21-39. Extrathoracic introducer technique. The technique is to puncture the subclavian vein outside the thoracic cavity, in the subclavian groove as it passes over the first rib. This approach is safe and is free from lead-related issues, and the technique is the same for all chest wall configurations. The puncture site is the point at which the subclavian vein passes over the first rib beneath the anterior edge of the clavicle, as seen in an anterior-posterior fluoroscopic view. Three chest wall configurations are defined by where the clavicle crosses the first rib, which is divided visually into three equal segments (anterior to posterior). The most common crossing site is in the midportion of the second segment. Less frequent are anterior first segment crossings and posterior crossings near or on the third segment.

shoulder rotation and location of the subclavian vein are related, recognition of the type shoulder rotation may be helpful, but a fluoroscopic view showing the crossing of the clavicle and first rib is definitive. A subclavian vein passing over the first or second segment of the first rib can usually be instrumented from a conventional pocket incision. This is not usually possible with a posterior rotation of the shoulder, which positions the vein over the third segment. A needle puncture from within the pocket causes the angle between the long axis of the needle and the vein to approach 90 degrees. This is not satisfactory, and a separate small incision superior to the pocket incision is necessary to obtain an angle of 45 degrees or less. Otherwise, the lead-vein entry angle will cause problems maneuvering the lead, biophysical interface issues such as pain, and/or stress-related lead integrity issues such as insulation failure or conductor coil fracture. A second extrathoracic approach is a cutdown to the cephalic vein. The cephalic vein travels through the deltopectoral groove and enters the distal portion of the axillary vein. This approach has always been touted as the gold standard. Because this vein is extrathoracic and pneumothorax-hemothorax is not an issue, comparing an introducer with this approach was like arguing against “motherhood and apple pie.” In truth, a large, patent cephalic vein is easy to cannulate, and two leads can usually be inserted. Even a small vein can be dilated to accommodate multiple leads by using

a guidewire and inserting an introducer. However, there are two disadvantages to a cutdown approach: small vein size and lead removal. The surgical dissection to locate a small cephalic vein ( RV). Note restoration of biventricular activation.

sinus rhythm to allow atrial synchronous left univentricular pacing with an AV delay initially programmed to 100 msec. Significant improvements in functional capacity, echocardiographic mitral regurgitation, and LV end-diastolic diameter were observed, with a favorable trend toward improvement in LV ejection fraction. These results are encouraging and support persistent benefit (at least to 1 year) of left univentricular pacing in some patients. Both LV and biventricular pacing synchronize LV contraction. This “re-timing” effect was initially attributed to “preexcitation” of the delayed LV segments. However, insights from tissue Doppler studies have revealed that LV pacing from a late-activated site achieves synchronous contraction by simultaneously delaying all LV segments.19,112 This is a potentially critical observation, because LV pacing reverses electrical activation and abolishes intraventricular dyssynchrony, but with the result of a marked increase in LV activation time compared with biventricular pacing.49 The consequences are a greater delay in RV contraction109 and a shortened diastolic filling time, which may have implications for ventricular pumping function, particularly at higher heart rates.112,118 Therefore, it is theoretically possible that LV-only pacing may achieve superior hemodynamic performance compared with biventricular pacing in some patients. For this reason, LV-only pacing should probably be considered in CRT nonresponders initially

treated with biventricular pacing. This could be easily achieved noninvasively if a true bipolar LV lead is used with a pulse generator that is capable of separately programmable ventricular outputs. A similar effect could be achieved in the case of a unipolar LV lead (dual cathodal configuration) by programming RV output below the capture threshold. It could not be achieved in a dual cathodal configuration without separately programmable ventricular outputs unless the LV threshold was significantly lower than the RV threshold. In any event, it is not currently possible to identify patients who will respond better to LV alone compared with biventricular pacing. REFERENCES 1. Kanagaratnam L, Pavia S, Schweikert R, et al: Matching approved “nondedicated” hardware to obtain biventricular pacing and defibrillation: Feasibility and troubleshooting. PACE 25:1066-1071, 2002. 2. Bulava A, Ansalone G, Ricci R, et al: Triple-site pacing in patients with biventricular device: Incidence of the phenomenon and cardiac resynchronization benefit. J Intervent Cardiac Electrophysiol 10:37-45, 2004. 3. Auricchio A, Stellbrink C, Block M, et al: Effect of pacing chamber and atrioventricular delay on acute systolic function of paced patients with congestive heart failure. The Pacing Therapies for Congestive Heart Failure Study Group. The Guidant Congestive Heart Failure Research Group. Circulation 99:29933001, 1999.

Chapter 26: Programming and Follow-up of Cardiac Resynchronization Devices 4. Auricchio A, Stellbrink C, Sack S, et al., and Group PTiCHFPCS: Long-term clinical effect of hemodynamically optimized cardiac resynchronization therapy in patients with heart failure and ventricular conduction delay. J Am Coll Cardiol 39:20262033, 2002. 5. Nelson GS, Curry CW, Wyman BT, et al: Predictors of systolic augmentation from left ventricular preexcitation in patients with dilated cardiomyopathy and intraventricular conduction delay. Circulation 101:2703-2709, 2000. 6. Ritter P, Padeletti L, Gillio-Meina L, et al: Determination of the optimal atrioventricular delay in DDD pacing: Comparison between echo and peak endocardial acceleration measurements. Europace 1:126-130, 1999. 7. Kindermann M, Frölig G, Doerr T, Schieffer H: Optimizing the AV delay in DDD pacemaker patients with high degree AV block: Mitral valve Doppler versus impedance cardiography. PACE 20:2453-2462, 1997. 8. Meluzin J, Novak M, Mullerova J, et al: A fast and simple echocardiographic determination of the optimal atrioventricular delay in patients after biventricular stimulation. PACE 27:58-64, 2004. 9. Auricchio A, Kramer A, Spinelli J, et al: Can the optimum dosage of resynchronization therapy be derived from the intracardiac electrogram? [abstract]. J Am Coll Cardiol 39:124, 2002. 10. Bristow MR, Saxon LA, Boehmer J, et al., and the Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators: Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med 350:2140-2150, 2004. 11. Butter C, Stellbrink C, Belalcazar A, et al: Cardiac resynchronization therapy optimization by finger plethysmography. Heart Rhythm 1:568-578, 2005. 12. Nishimura RA, Hayes DL, Holmes DR, Tajik AJ: Mechanism of hemodynamic improvement by dual-chamber pacing for severe left ventricular dysfunction: An acute Doppler and catheterization study. J Am Coll Cardiol 25:281-288, 1995. 13. Auricchio A, Sommariva L, Salo RW, et al: Improvement of cardiac function in patients with severe congestive heart failure and coronary artery disease by dual chamber pacing with shortened AV delay. PACE 17:995-997, 1994. 14. Linde-Edelstam C, Nordlander R, Unden A-L, et al: Quality-oflife in patients treated with atrioventricular synchronous pacing compared to rate modulated ventricular pacing: A long-term, double-blind, crossover study. PACE 15:1467-1476, 1992. 15. Abraham WT, Fisher WG, Smith AL, et al., for the MIRACLE Study Group. Cardiac resynchronization in chronic heart failure. N Engl J Med 346:1845-1853, 2002. 16. Cazeau S, Leclercq C, Lavergne T, et al., and The Multisite Stimulation in Cardiomyopathies (MUSTIC) Study Investigators. Effects of multisite biventricular pacing in patients with heart failure and intraventricular conduction delay. N Engl J Med 344:873-880, 2001. 17. Sawhney NS, Waggoner AD, Garhwal S, et al: Randomized prospective trial of atrioventricular delay programming for cardiac resynchronization therapy. Heart Rhythm 1:562-567, 2004. 18. Sogaard P, Egeblad H, Pedersen AK, et al: Sequential versus simultaneous biventricular resynchronization for severe heart failure: Evaluation by tissue Doppler imaging. Circulation 106:2078-2084, 2002. 19. Yu CM, Chau E, Sanderson EJ, et al: Tissue Doppler echocardiographic evidence of reverse remodeling and improved synchronicity by simultaneous delaying regional contraction after biventricular pacing therapy in heart failure. Circulation 105:438-445, 2002. 20. Bordachar P, Lafitte S, Reuter S, et al: Echocardiographic parameters of ventricular dyssynchrony validation in patients with

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

1137

heart failure using sequential biventricular pacing. J Am Coll Cardiol 44:2157-2165, 2004. Higgins SL, Hummel JD, Niazi IK, et al: Cardiac resynchronization therapy for the treatment of heart failure and intraventricular conduction delay and malignant ventricular tachyarrhythmia. J Am Coll Cardiol 42:1454-1459, 2003. Fisher JD, Mehra R, Furman S: Termination of ventricular tachycardia with bursts of rapid ventricular pacing. Am J Cardiol 41:94-102, 1978. Gillis AM, Leitch J, Sheldon RS, et al: A prospective randomized comparison of autodecremental pacing to burst pacing in device therapy for chronic ventricular tachycardia secondary to coronary artery disease. Am J Cardiol 72:1146-1151, 1993. Calkins H, El-Atassi R, Kalbfleisch S, et al: Comparison of fixed burst versus decremental burst pacing for termination of ventricular tachycardia. PACE 16:26-32, 1993. Kantoch MJ, Green MS, Tang AS: Randomized cross-over evaluation of two adaptive pacing algorithms for the termination of ventricular tachycardia. PACE 16:1664-1672, 1993. Hamill SC, Packer DL, Stanton MS, et al., and the Multicenter PCD Investigator Group: Termination and acceleration of ventricular tachycardia with autodecremental pacing, burst pacing, and cardioversion in patients with an implantable cardioverter defibrillator. PACE 18:3-10, 1995. Fisher JD, Zhang Z, Kim SG, et al: Comparison of burst pacing, autodecremental (ramp) pacing, and universal pacing for termination of ventricular tachycardia. Arch Mal Coeur Vaiss 89:135-139, 1996. Newman D, Dorian P, Hardy J: Randomized controlled comparison of antitachycardia pacing algorithms for termination of ventricular tachycardia. J Am Coll Cardiol 21:1413-1418, 1993. Schaumann A, Poppinga A, von zur Muehlen F, Kreuzer H: Antitachycardia pacing for ventricular tachycardias above and below 200 bpm: A prospective study for ramp vs. can mode [abstract]. PACE 20:1108, 1997. Nasir N, Pacifico A, Doyle TK, et al: Spontaneous ventricular tachycardia treated by antitachycardia pacing. Cadence Investigators. Am J Cardiol 79:820-822, 1997. Krater L, Lamp B, Heintze J, et al: Influence of antitachy pacing location on the efficacy of ventricular tachycardia termination. J Am Coll Cardiol 39:124A, 2002. Lozano IF, Higgins S, Hummel J, et al: The efficacy of simultaneous right and left ventricular antitachycardia pacing (BiV ATP) in heart failure patients with an AICD indication improves with time [abstract]. PACE 26:984, 2003. Peinado R, Almendral J, Rius T, et al: Randomized, prospective comparison of four burst pacing algorithms for spontaneous ventricular tachycardia. Am J Cardiol 82:1422-1425, 1998. Knight BP, Desai A, Coman J, et al: Long-term retention of cardiac resynchronization therapy. J Am Coll Cardiol 44:72-77, 2004. Richardson K, Cook K, Wang PJ, Al-Ahmad A: Loss of biventricular pacing: What is the cause? Heart Rhythm 2:110-111, 2005. Brandt J, Fahraeus T, Schuller H: Far-field QRS complex sensing via the atrial pacemaker lead. I. Mechanism, consequences, differential diagnosis and countermeasures in AAI and VDD/DDD pacing. PACE 11:1432-1438, 1988. Brandt J, Fahraeus T, Schuller H: Far-field QRS complex sensing via the atrial pacemaker lead. II. Prevalence, clinical significance and possibility of intraoperative prediction in DDD pacing. PACE 11:1540-1544, 1988. Brandt J, Worzewski W: Far-field QRS complex sensing: Prevalence and timing with bipolar atrial leads. PACE 23:315-320, 2000. Weretka S, Becker R, Hilbel T, et al: Far-field R wave oversensing in new dual chamber ICDs: Incidence, predisposing factors and clinical implications [abstract]. PACE 23:571, 2000.

1138

Section Four: Device Electrocardiography

40. Johnson WB, Bailin SJ, Solinger B, et al: Frequency of inappropriate automatic pacemaker mode switching as assessed 6 to 8 weeks post implantation [abstract]. PACE 19:720, 1996. 41. Frohlig G, Kinderman M, Heisel A, et al: Mode switch without atrial tachyarrhythmias [abstract]. PACE 19:592, 1996. 42. Ueng KC, Tsai TP, Tsai CF, et al: Acute and long-term effects of atrioventricular junction ablation and VVIR pacemaker in symptomatic patients with chronic lone atrial fibrillation and normal ventricular response. J Cardiovasc Electrophysiol 12:303-309, 2001. 43. Daoud EG, Weiss R, Bahu M, et al: Effect of an irregular ventricular rhythm on cardiac output. Am J Cardiol 78:1433-1436, 1996. 44. Clark DM, Plumb VJ, EpsteinAE, Kay GN: Hemodynamic effects of an irregular sequence of ventricular cycle length during atrial fibrillation. J Am Coll Cardiol 30:1039-1045, 1997. 45. Marshall HJ, Harris ZI, Griffith MK, Gammage MD: Atrioventricular nodal ablation and implantation of mode switching dual chamber pacemakers: Effective treatment for drug refractory paroxysmal atrial fibrillation. Heart 79:543-547, 1998. 46. Kamalvand K, Tan K, Kotsakis A, et al: Is mode switching beneficial? A randomized study in patients with paroxysmal atrial tachyarrhythmias. J Am Coll Cardiol 30:496-504, 1997. 47. Brignole M, Gainfranchi L, Menozzi C, et al: Assessment of atrioventricular junction ablation and DDDR mode-swtiching pacemakers versus pharmacological treatment in patients with severely symptomatic paroxysmal atrial fibrillation: A randomized controlled study. Circulation 96:2617-2624, 1997. 48. Marshall HJ, Harris ZI, Griffith MJ, et al: Prospective study of ablation and pacing versus medical therapy for paroxysmal atrial fibrillation: Effects of pacing mode and mode-switch algorithms. Circulation 99:1587-1592, 1999. 49. Leclercq C, Faris O, Runin R, et al: Systolic improvement and mechanical resynchronization does not require electrical synchrony in the dilated failing heart with left bundle-branch block. Circulation 106:1760-1763, 2002. 50. Daoud E, Kalbfleisch FJ, Hummel JD, et al: Implantation techniques and chronic lead parameters of biventricular pacing dual-chamber defibrillators. J Cardiovasc Electrophysiol 13:964970, 2002. 51. Storm C, Harsch M, DeBus B: InSync Registry: Post Market Study. Progress Report No. 7. Medtronic, Inc., February 2005. 52. Young JB, Abraham WT, Smith AL, et al., and Multicenter InSync ICD Randomized Clinical Evaluation (MIRACLE ICD) Trial Investigators: Combined cardiac resynchronization and implantable cardioversion defibrillation in advanced chronic heart failure: The MIRACLE ICD Trial. JAMA 289:2685-2394, 2003. 53. Barold SS, Herweg B, Gallardo I: Double counting of the ventricular electrogram in biventricular pacemakers and ICDs. PACE 26:1645-1648, 2003. 54. Garcia-Moran E, Mont L, Brugada J: Inappropriate tachycardia detection by a biventricular implantable cardioverter defibrillator. PACE 25:123-124, 2002. 55. Schreieck J, Zrenner B, Kolb C, et al: Inappropriate shock delivery due to ventricular double detection with a biventricular pacing implantable cardioverter defibrillator. PACE 24:11541157, 2001. 56. Lipchenca I, Garrigue S, Glikson M, et al: Inhibition of biventricular pacemakers by oversensing of far-field atrial depolarization. PACE 25:365-367, 2002. 57. Taieb J, Benchaa T, Foltzer E, et al: Atrioventricular cross-talk in biventricular pacing: A potential cause of ventricular standstill. PACE 25:929-935, 2002. 58. Oguz E, Akyol A, Okmen E: Inhibition of biventricular pacing by far-field left atrial activity sensing: Case report. PACE 25:1517-1519, 2002. 59. Vollmann D, Luthje L, Gortler G, Unterberg C: Inhibition of bradycardia pacing and detection of ventricular fibrillation due

60.

61.

62.

63.

64.

65.

66.

67.

68.

69. 70.

71.

72.

73.

74.

75.

76.

to far-field atrial sensing in a triple chamber implantable cardioverter defibrillator. PACE 25:1513-1516, 2002. Garrigue S, Barold SS, Clementy J: Double jeopardy in an implantable cardioverter defibrillator patient. J Cardiovasc Electrophysiol 14:784, 2003. Sweeney MO, Ellison KE, Shea JB: Provoked and spontaneous high frequency, low amplitude respirophasic noise transients in patients with implantable cardioverter-defibrillators. J Cardiovasc Electrophysiol 12:402-410, 2001. Zagrodzky JD, Ramaswamy K, Page RL, et al: Biventricular pacing decreases the inducibility of ventricular tachycardia in patients with ischemic cardiomyopathy. Am J Cardiol 87:12081210, 2001. Walker S, Levy T, Rex S, et al: Usefulness of suppression of ventricular arrhythmia by biventricular pacing in severe congestive cardiac failure. Am J Cardiol 86:231-233, 2000. Higgins SL, Yong P, Scheck D, et al: Biventricular pacing diminishes the need for implantable cardioverter defibrillator therapy. J Am Coll Cardiol 36:824-827, 2000. Wilkoff B, Hess M, Young JD, Abraham WT: Differences in tachyarrhythmia detection and implantable cardioverter defibrillator therapy by primary or secondary prevention indication in cardiac resynchronization therapy patients. J Cardiovasc Electrophysiol 15:1002-1009, 2004. Guerra J, Wu J, Miller JM, Groh WJ: Increase in ventricular tachycardia frequency after biventricular implantable cardioverter defibrillator upgrade. J Cardiovasc Electrophysiol 14:1245-124, 2003. Medina-Ravell VA, Lankipalli RS, Yan GX, et al: Effect of epicardial or biventricular pacing to prolong QT interval and increase transmural dispersion of repolarization. Circulation 107:740-746, 2003. Fish JM, Di Diego JM, Nesterenko V, Antzelevitch C: Epicardial activation of left ventricular wall prolongs QT interval and transmural dispersion of repolarization: Implications for biventricular pacing. Circulation 109:2136-2142, 2004. Barold SS, Byrd CL: Cross-ventricular endless loop tachycardia during biventricular pacing. PACE 24:1821-1823, 2001. Berruezo A, Mont L, Scalise A, Brugada J: Orthodromic pacemaker-mediated tachycardia in a biventricular system without an atrial electrode. J Cardiovasc Electrophysiol 15:1100-1102, 2004. Auricchio A, Stellbrink C, Sack S, et al: Long-term benefit as a result of pacing resynchronization in congestive heart failure: Results of the PATH-CHF Trial. Circulation 102:II-693A, 2000. Stellbrink C, Breithardt OA, Franke A, et al., and PATH-CHF (PAcing THerapies in Congestive Heart Failure) Investigators, CPI Guidant Congestive Heart Failure Research Group. Impact of cardiac resynchronization therapy using hemodynamically optimized pacing on left ventricular remodeling in patients with congestive heart failure and ventricular conduction disturbances. J Am Coll Cardiol 38:1957-1965, 2001. Gorscan J, Kanzaki H, Bazaz R, et al: Usefulness of echocardiographic tissue synchronization imaging to predict acute response to cardiac resynchronization therapy. Am J Cardiol 93:1178-1181, 2004. Abraham WT: Rationale and design of a randomized clinical trial to assess the safety and efficacy of cardiac resynchronization therapy in patients with advanced heart failure: The Multicenter InSync Randomized Clinical Evaluation (MIRACLE). J Card Fail 6:369-380, 2000. Packer M: Proposal for a new clinical end point to evaluate the efficacy of drugs and devices in the treatment of chronic heart failure. J Card Fail 7:176-182, 2001. Gregoratos G, Abrams J, Epstein AE, et al: ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices: Summary Article. A Report of the American College of Cardiology/American Heart Association

Chapter 26: Programming and Follow-up of Cardiac Resynchronization Devices

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

Task Force on Practice Guidelines (ACC/AHA/NASPE Committee to Update the 1998 Pacemaker Guidelines). Circulation 106:2145-2161, 2002. Bax JJ, Mohoek SG, Marwick TJ, et al: Left ventricular dyssynchrony predicts benefit of cardiac resynchronization therapy in patients with end-stage heart failure before pacemaker implantation. Am J Cardiol 92:1238-1240, 2003. Reuter S, Garrigue S, Barold SS, et al: Comparison of characteristics in responders versus nonresponders with biventricular pacing for drug-resistant congestive heart failure. Am J Cardiol 89:346-350, 2002. Yu C-M, Fung W-H, Lin H, et al: Predictors of left ventricular reverse remodeling after cardiac resynchronization therapy for heart failure secondary to idiopathic dilated or ischemic cardiomyopathy. Am J Cardiol 91:684-688, 2003. Yu CM, Fung JWH, Chan CK, et al: Comparison of efficacy of reverse remodeling and clinical improvement for relatively narrow and wide QRS complexes after cardiac resynchronization therapy for heart failure. J Cardiovasc Electrophysiol 15:1058-1065, 2004. Yu CM, Lin H, Zhang Q, Sanderson JE: High prevalence of left ventricular systolic and diastolic asynchrony in patients with congestive heart failure and normal QRS duration. Heart 89:5460, 2003. Pitzalis MD, Iacoviello M, Romito R, et al: Cardiac resynchronization therapy tailored by echocardiographic evaluation of ventricular asynchrony. J Am Coll Cardiol 40:1615-1622, 2002. Auricchio A, Kloss M, Trautmann SI, et al: Exercise performance following cardiac resynchronization therapy in patients with heart failure and ventricular conduction delay. Am J Cardiol 89:198-203, 2002. Linde C, Leclerc C, Rex S, et al: Long-term benefits of biventricular pacing in congestive heart failure: Results from the Multisite Stimulation in Cardiomyopathy (MUSTIC) Study. J Am Coll Cardiol 40:111-118, 2002. Linde C, Braunschweig F, Gadler F, et al: Long-term improvement in quality of life by biventricular pacing in patients with chronic heart failure: Results from the MUSTIC Study. Am J Cardiol 91:1090-1095, 2003. Reynolds MR, Joventino LP, Josephson ME, and Miracle ICD Investigators: Relationship of baseline electrocardiographic characteristics with the response to cardiac resynchronization therapy for heart failure. PACE 27:1513-1518, 2004. Kadhiresan V, Vogt J, Auricchio A, et al: Sensitivity and specificity of QRS duration to predict acute benefit in heart failure patients with cardiac resynchronization [abstract]. PACE 23(II):555, 2000. Moss AJ, Zareba W, Hall WJ, et al., for the Multicenter Automatic Defibrillator Implantation Trial II Investigators: Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 346:877883, 2002. Garrigue S, Reuter S, Labeque J-N, et al: Usefulness of biventricular pacing in patients with congestive heart failure and right bundle branch. Am J Cardiol 88:1436-1441, 2001. Aranda JM, Curtis AB, Conti JB, Stejskal-Peterson S: Do heart failure patients with right bundle branch block benefit from cardiac resynchronization therapy? Analysis of the MIRACLE Study [abstract]. J Am Coll Cardiol 39:96A, 2002. Egoavil CA, Ho RT, Greenspon AJ, Pavri BB: Cardiac resynchronization therapy in patients with right bundle branch block: Analysis of pooled data from MIRACLE and ContakCD trials. Heart Rhythm 2:611-615, 2005. Bleeker GB, Schalij MJ, Molhoek SG, et al: Relationship between QRS duration and left ventricular dyssynchrony in patients with end-stage heart failure. J Cardiovasc Electrophysiol 15:544-549, 2004.

1139

93. Josephson ME: Clinical Cardiac Electrophysiology: Techniques and Interpretations, 3rd ed. Philadelphia, Lippincott, Williams & Wilkins, 2002. 94. Fantoni C, Kawabata M, Massaro R, et al: Right and left ventricular activation sequence in patients with heart failure and right bundle branch block: A detailed analysis using threedimensional non-fluoroscopic electroanatomic mapping system. J Cardiovasc Electrophysiol 16:112-119, 2005. 95. Wyman BT, Hunter WC, Prinzen FW, et al: Effects of single- and biventricular pacing on temporal and spatial dynamics of ventricular contraction. Am J Physiol 282:H372-H379, 2002. 96. D’Hooge J, Heimdal A, Jamal F, et al: Regional strain and strain rate measurements by cardiac ultrasound: Principles, implementation and limitations. Eur J Echocardiography 1:154-170, 2000. 97. Prinzen FW, Hunter WC, Wyman BT, et al: Mapping of regional myocardial strain and work during ventricular pacing: Experimental study using magnetic resonance imaging tagging. J Am Coll Cardiol 33:1735-1742, 1999. 98. Breithardt OA, Stellbrink C, Herbots L, et al: Cardiac resynchronization therapy can reverse abnormal myocardial strain distribution in patients with heart failure and left bundle branch block. J Am Coll Cardiol 42:486-494, 2003. 99. Sogaard P, Egeblad H, Kim W, et al: Tissue Doppler imaging predicts improved systolic performance and reversed left ventricular remodeling during cardiac resynchronization therapy. J Am Coll Cardiol 40:723-730, 2002. 100. Bax JJ, Molhoek SG, Marwick TH, et al: Usefulness of myocardial tissue Doppler echocardiography to evaluate left ventricular dyssynchrony before and after biventricular pacing in patients with idiopathic dilated cardiomyopathy. Am J Cardiol 91:94-97, 2003. 101. Baxx JJ, Yu C-M, Lin H, et al: Comparison of acute changes in left ventricular volume, systolic and diastolic functions, and intraventricular synchronicity after biventricular pacing and right ventricular pacing for congestive heart failure. Am Heart J 145:G1-G7, 2003. 102. Auricchio A, Fantoni C, Regoli F, et al: Characterization of left ventricular activation in patients with heart failure and left bundle branch block. Circulation 109:1133-1139, 2004. 103. Kerckhoffs RC, Bovendeerd PH, Kotte JC, et al: Homogeneity of cardiac contraction despite physiological asynchrony of depolarization: A model study. Ann Biomed Eng 31:536-547, 2003. 104. Pitzalis MD, Iacoviello M, Romito R, et al: Ventricular asynchrony predicts a better outcome in patients with chronic heart failure receiving cardiac resynchronization therapy. J Am Coll Cardiol 45:65-69, 2005. 105. Yu C-M, Yang H, Lau C-P, et al: Regional left ventricular mechanical asynchrony in patients with heart disease and normal QRS duration. PACE 26:562-570, 2003. 106. Achilli A, Sassara M, Ficili S, et al: Long term effectiveness of cardiac resynchronization therapy in patients with refractory heart failure and “narrow” QRS duration. J Am Coll Cardiol 42:2117-2124, 2003. 107. Kass DM: Predicting cardiac resynchronization response by QRS duration. J Am Coll Cardiol 42:2125-2127, 2003. 108. Gaspirini M, Mantica M, Galimberti P, et al: Beneficial effects of biventricular pacing in patients with a “narrow” QRS duration. PACE 26:169-174, 2003. 109. Turner MS, Bleasdale RA, Dragos Vinereanu D, et al: Electrical and mechanical components of dyssynchrony in heart failure patients with normal QRS duration and left bundle-branch block: Impact of left and biventricular pacing. Circulation 109:2544-2549, 2004. 110. Brandt J, Fahraeus T, Ogawa T, Schuller H: Practical aspects of rate adaptive atrial (AAI,R) pacing: Clinical experiences in 44 patients. PACE 14:1258-1264, 1991.

1140

Section Four: Device Electrocardiography

111. Bernheim A, Ammann P, Sticherling C, et al: Right atrial pacing impairs cardiac function during resynchronization therapy: Acute effects of DDD pacing compared to VDD pacing. J Am Coll Cardiol 45:1482-1487, 2005. 112. Bordachar P, LaFitte S, Reuter S, et al: Biventricular pacing and left ventricular pacing in heart failure: Similar hemodynamic improvement despite marked electromechanical differences. J Cardiovasc Electrophysiol 15:1342-1347, 2004. 113. Perego GB, Chianca R, Facchini M, et al: Simultaneous vs. sequential biventricular pacing in dilated cardiomyopathy: An acute hemodynamic study. Eur J Heart Fail 5:305-313, 2003. 114. Blanc JJ, Etienne Y, Gilard M, et al. Evaluation of different ventricular pacing sites in patients with severe heart failure: Results of an acute hemodynamic study. Circulation 96:3273-3277, 1997.

115. Kass DA, Chen CH, Curry C, et al: Improved left ventricular mechanics from acute VDD pacing in patients with dilated cardiomyopathy and ventricular conduction delay. Circulation 99:1567-1573, 1999. 116. Touiza A, Etienne Y, Gilard M, et al: Long-term left ventricular pacing: Assessment and comparison with biventricular pacing in patients with severe congestive heart failure. J Am Coll Cardiol 38:1966-1970, 2001. 117. Blanc JJ, Bertault-Valls V, Fatemi M, et al: Long-term benefits of left univentricular pacing in patients with congestive heart failure. Circulation 109:1741-1744, 2004. 118. Kass DA: Left ventricular versus biventricular pacing in cardiac resynchronization therapy: The plot thickens in this tale of two modes. J Cardiovasc Electrophysiol 15:1348-1349, 2004.

Chapter 27

Follow-up and Interpretation of Implantable Syncope Monitors ANDREW D. KRAHN • GEORGE J. KLEIN • LORNE J. GULA • ALLAN C. SKANES • RAYMOND YEE

T he advent of prolonged monitoring with implanted loop recorders (ILRs) has revolutionized the quest for detection of elusive infrequent arrhythmias in patients with unexplained syncope. Paroxysmal arrhythmias often result in infrequent and sporadic symptoms that usually resolve before the patient gets to medical attention. The capability of prolonged monitoring has permitted clinicians to obtain a symptom rhythm correlation in most patients with suspected underlying infrequent arrhythmia. Patients often remain undiagnosed after monitoring with a Holter monitor and an external loop recorder.1 The outcome in undiagnosed patients is strongly influenced by the presence of underlying heart disease and is not as favorable as in those patients with vasovagal syncope or a diagnosed cause of syncope.2 Clinicians rely on the initial history, physical examination, and abnormal laboratory results to make a diagnosis by inference in many cases. Ideally, some form of prolonged monitoring captures key physiologic data during the next spontaneous event. External and particularly ILRs are powerful tools for arrhythmia detection.

Implantable Loop Recorders The ILR (Reveal, Medtronic, Minneapolis, Minn.) has a pair of sensing electrodes 3.7 cm apart on the outer shell that records a single-lead bipolar rhythm strip. The device measures 6.1 × 1.9 × 0.8 cm, weighs 17 g, and has a recommended battery life of 14 months (Fig. 27-1).3-5 The battery life is typically 18 to 24 months, depending on preimplantation shelf life and patient variability. The recorder is inserted in the left pectoral region using standard sterile technique and local anesthesia in the subcutaneous tissue. It also has been implanted in right parasternal, subcostal, and axillary regions with an adequate albeit lower-amplitude signal.4 The recorded bipolar electrocardiographic signal is stored in a circular buffer that is capable of storing 21 minutes of uncompressed signal or 42 minutes of compressed signal in one or three divided parts. Because the quality of the compressed signal is negligibly different from that of the uncompressed signal, the compressed form is used most often to maximize the memory capacity of the device. The memory 1141

1142

Section Four: Device Electrocardiography

buffer is frozen by means of a handheld activator that is provided to the patient at the time of device implantation. In layman’s terms, the device answers the question, “What just happened?” Events stored by the device are downloaded after interrogation with a standard Medtronic 9790 pacemaker programmer (Fig. 27-2).

Figure 27-1. Loop recorder technology. From left to right in the photograph, an external loop recorder with cables that attach to electrodes on the patient, the implantable loop recorder, and the patient activator.

The current version of the device (Reveal Plus) has programmable automatic detection of high and low heart rate episodes and pauses (Fig. 27-3). The resultant memory configuration allows for division of multiple 1-minute automatic rhythm strips in addition to one to three manual recordings. This permits automated backup of manual activations to detect prespecified extreme heart rates or pauses (typically 160 bpm, and pauses >3 seconds); this is often clinically useful in patients who have difficulty activating the device. Three reports have suggested that this feature is more likely than conventional patient activation to detect a borderline or significant arrhythmia.6,7 Automatic detection also permits detection of asymptomatic heart rate changes that may result in a presumptive diagnosis in the absence of symptomatic recurrence.7 In a prospective study of 60 patients, automatic detection recorded predetermined “significant” asymptomatic arrhythmias in 15% of patients that led to therapeutic decisions, predominantly pacemaker implantation for bradycardia. Automatic detection has also led to recognition of sensing issues, with transient loss of signal observed in the majority of patients at some point, resulting in automatic detection of pauses (Fig. 27-4).7-10 Unlike the adjustments of gain and sensitivity that optimize sensing successfully in most patients, this problem appears to stem from transient loss of contact of the sensing electrodes within the device pocket. A relatively tight pocket with adequate anchoring of the device may minimize this problem.

10:43:57

10:44:11

A

10:44:25

10:44:39

Figure 27-2. Rhythm strip obtained with an implantable loop recorder during syncope in a 73year-old woman with three syncopal episodes in the previous 2 years. Tilt testing and a trial of an external loop recorder were negative. Each line represents 14 seconds of a single-lead rhythm strip. Note the abrupt onset of complete heart block with visible p waves during a prolonged pause. The arrow and letter A denote automatic activation of the device after detection of a 3-second pause.

Chapter 27: Follow-up and Interpretation of Implantable Syncope Monitors

60 bpm

mV 0.8 0.4 0.0 ⫺0.4

1

2

2

4

5

6

⫺0.8 secs

Figure 27-3. Programming interface for the implantable loop recorder. Key elements include use of compressed signal to increase memory capacity, configuration of the memory tailored to the patient’s frequency and severity of symptoms, and rate settings for automatic detection. ECG, electrocardiogram.

The implantation procedure is similar to that used for creation of a smaller and more superficial pacemaker pocket. It is typically but not necessarily performed in an electrophysiology or cardiac catheterization laboratory setting. An adequate signal can be obtained anywhere in the left thorax, without the need for cutaneous mapping.4 Mapping does optimize the sensed signal, and it is recommended for patients in whom automatic detection is desirable to prevent over-

1143

sensing of T waves and double-counting leading to automatic detection of high-rate episodes. Mapping usually leads to device insertion in a vertical or oblique orientation in the high left parasternal region.11,12 Right parasternal sites have been used to optimize p-wave amplitude. The patient and a spouse, family member, or friend are instructed in the use of the activator at the time of implantation. Use of prophylactic antibiotics is usually recommended to prevent pocket infection. After implantation, gain and sensitivity are adjusted according to the manufacturer’s specifications. There are no published studies that have systematically identified the optimal method for device setup or follow-up with respect to gain and sensitivity. Chrysostomakis and colleagues8 suggested that adequate sensing can be obtained in most patients with either an apical or a left parasternal implant position. Boersma and associates13 implanted one half of their non-autodetect devices in a subcostal position with adequate signal. Experienced device implanters may potentially be misled because the scale for setting gain and sensitivity are the opposite of those conventionally used in pacemakers (although perhaps more logical than the latter). Gain and sensitivity should be adjusted after assessment of the sensed signal during postural changes immediately after implantation, and also during follow-up if there are issues with oversensing or undersensing. A lowamplitude signal resulting in false detection has been noted as a common problem with the device, variably described as oversensing or undersensing.7-10

14:30 02/09/2005 Reveal (R) Plus Model 9526 Programmer 2090 9809v60 Gain: X8 (⫹/⫺ 0.2 mV) Storage Mode: 1 patient, 13 auto events, 42 min (.C) Medtronic, Inc. 2003 Automatic Event 1 of 13 recorded 06/06/2001 061301A Page 1 of 2 12.5 mm/sec, 25.0 mm/mV

⫽ Activation point

14:45:36

14:45:47

14:45:56

14:46:05

14:46:14

14:46:23

14:46:32

14:46:41 A Figure 27-4. Transient loss of signal believed to be caused by loss of tissue electrode contact within the device pocket. Loss of signal (arrow) with amplifier saturation and reacquisition of signal is seen during nonphysiologic recording, which is detected automatically (A) as asystole.

1144

Section Four: Device Electrocardiography

In our recent experience with the Reveal Plus device, 2 (3%) of 60 patients had poor sensing after implantation that resulted in ongoing need for reprogramming with oversensing and undersensing during followup.7,12 Gain and sensitivity required reprogramming in 24 patients (40%) during the first month of follow-up, and in 35 patients (58%) at some point during the 1year follow-up period. Automatic detection was abandoned in 1 patient in whom sensing could not be improved with reprogramming. Twelve patients (20%) required additional follow-up visits to assess for automatic detection events because of excessive false detection. Undersensing occurred in a single patient with a narrow QRS complex tachycardia that was detected both symptomatically and asymptomatically on several occasions. On one symptomatic occasion, automatic detection did not sense the manually captured tachycardia at 190 bpm. No device-related abnormality could be detected on interrogation. Early Experience The first iteration of the ILR was a dual-chamber pacemaker can equipped with a sensing electrode on the lower aspect of the can and a second sensor on a modified leadless header. It was implanted in 24 patients who had recurrent unexplained syncope after negative results with extensive noninvasive and invasive testing, including tilt-table and electrophysiologic (EP) testing.3 The device was very successful in establishing a symptom-rhythm correlation (88%) in this difficult patient population. These results were validated in a recent study by Boersma and associates13 with a similar design in 43 patients with unexplained syncope that found a low yield to conventional testing and obtained a symptom-rhythm correlation in 44%. Garcia-Civera and colleagues assessed the selective use of EP testing in combination with tilt testing in 184 patients with unexplained syncope.14,15 In the subgroup of 15 patients with negative tilt and EP testing, the ILR provided a symptom-rhythm correlation in 47%. These studies suggest that the device has a clear role in patients with ongoing symptoms and negative conventional testing. They also call into question the usefulness of conventional testing in certain patient populations. More recent data suggest that the device may well play a role at an earlier stage in the patient’s workup.16-20 Further studies have applied the ILR to patients with a lesser burden of syncope and less preimplantation testing, lowering the likelihood of recurrence of syncope after implantation to between 30% and 70%.6,13,21-24 Several studies have demonstrated the feasibility of the ILR in establishing a symptom-rhythm correlation during long-term monitoring in pediatric and geriatric patients as well as others.5,6,16,20-23,25-30 The largest of these studies combined data from 206 patients from 3 centers.22 The majority of patients studied had undergone previous noninvasive testing and selective invasive testing, including tilt testing and EP studies. Symptoms recurred during follow-up in 69% of patients 93 ± 107 days after device implantation. An arrhythmia was detected in 22% of patients,

sinus rhythm excluding a primary arrhythmia was seen in 42%, and symptoms resolved without recurrence during prolonged monitoring in 31%. Bradycardia was detected more frequently than tachycardia (17% versus 6%), leading to pacemaker implantation in most patients. Failed activation of the device after spontaneous symptoms occurred in 4% of patients. Devices used in this era did not have the automatic detection capability. Selection bias probably served to withhold the device from patients who were unlikely to activate it, making the activation failure rate in practice higher. In such patients, a symptom-rhythm correlation was not obtained during the monitoring period. As discussed earlier, the automatic detection feature appears to enhance detection in patients who are less likely to successfully activate the device.6,7,10 No age group had an incidence of bradycardia requiring pacing greater than 30%, suggesting a limited role for empiric pacing in the population with unexplained syncope. Multivariate modeling did not identify any significant preimplantation predictors of subsequent arrhythmia detection other than a weak association with advancing age and bradycardia. Subsequent Applications of the Implantable Loop Recorder In the initial application, use of ILRs was focused on the population with recurrent unexplained syncope. Recent trials have focused on other subgroups to obtain symptom-rhythm correlation during spontaneous symptoms. In an investigation of atypical epilepsy, Zaidi and associates studied 74 patients with ongoing seizures despite anticonvulsant therapy or with unexplained recurrent seizures.31,32 They performed cardiac assessment, including tilt testing and carotid sinus massage, in all patients and implanted ILRs in 10 patients. Tilt testing was positive in 27% of patients, and carotid sinus massage was positive in 10%. Two of the 10 patients who subsequently underwent ILR monitoring demonstrated marked bradycardia preceding seizure activity, one due to heart block and the other due to sinus pauses. This study suggested that seizures that are atypical in presentation or response to therapy may have a cardiovascular cause in as many as 42% of patients and that long-term cardiac monitoring can play a role in select patients with atypical seizures. The International Study on Syncope of Uncertain Etiology (ISSUE) investigators implanted ILRs in three different populations of patients with syncope to obtain electrocardiographic correlation with spontaneous syncope after conventional testing.18,19,33 In the first study, tilt testing was performed in 111 patients with a clinical diagnosis of vasovagal syncope, and ILRs were implanted after the tilt test regardless of result. Syncope recurred in 34% of patients in both the tilt-positive and the tilt-negative groups, with marked bradycardia or asystole the most common recorded arrhythmia during follow-up (46% and 62%, respectively). The heart rate response during tilt testing did not predict spontaneous heart rate during episodes, and a much higher incidence of prolonged pauses was observed than was

Chapter 27: Follow-up and Interpretation of Implantable Syncope Monitors

expected based on tilt response, where a marked cardioinhibitory response was uncommon. This study suggests that tilt testing is poorly predictive of rhythm findings during spontaneous syncope and that bradycardia is more common than previously recognized. The second ISSUE study included 52 patients with syncope and bundle branch block with negative EP testing.19 Syncope recurred in 22 of the 52 patients during ILR follow-up. Marked bradycardia, mainly attributed to complete atrioventricular (AV) block, was seen in 17 patients, whereas AV block was excluded in 2. Three patients did not properly activate the device after symptoms. Therefore, a significant proportion of this population progressed to complete AV block despite apparent reassurance from the negative EP test results. Syncope may be a clue that conduction system disease is progressive. The third ISSUE study included 35 patients with syncope and moderate structural heart disease who had negative EP test results.33 The underlying heart disease was predominantly ischemic heart disease or hypertrophic cardiomyopathy with moderate but not severe left ventricular dysfunction. Although previous studies have suggested that patients with negative EP testing have a better prognosis, there remains concern about the risk of ventricular tachycardia in this group. Importantly, only 2 of the 35 patients had an ejection fraction of less than 30%, which would have made them candidates for primary prevention of sudden death in keeping with the MADIT 2 Trial.34 Symptoms recurred in 19 (54%) of the 35 patients, with ventricular tachycardia in only 1 patient. In the remaining subjects,

1145

bradycardia was detected in 4 and supraventricular tachyarrhythmias in 5. There were no sudden deaths during 16 ± 11 months of follow-up. This study supports a monitoring strategy in patients with left ventricular dysfunction related to ischemic heart disease when EP testing is negative, guided by the ejection fraction and guidelines for primary prevention of sudden death.35 Out of the ISSUE studies came a proposed classification36 of detected rhythm during spontaneous syncope (Table 27-1). This system is very useful in assigning a likely mechanism for the detected rhythm, particularly bradycardia, which may be primary or neurocardiogenic. The classification scheme focuses on sinus rate and AV conduction in assessing the likely state of the autonomic nervous system and its contribution to the underlying hypotension that leads to syncope. This arrangement also permits consistent communication among clinicians and researchers regarding the detected rhythm. Comparative Studies Two groups have performed randomized trials comparing the ILR with conventional testing in patients with unexplained syncope. The Randomized Assessment of Syncope Trial (RAST) was a single-center, prospective randomized trial that focused on an initial loop recorder versus conventional testing in patients undergoing a cardiac workup for unexplained syncope16,17; patients were crossed over if primary testing was negative. Sixty patients (age 66 ± 14 years, 33 male) with

Electrocardiographic Classification of Detected Rhythm from the Implanted Loop Recorder TABLE 27-1.

Classification

Sinus Rate

AV Node

Comment

Asystole (R-R interval >3 sec) 1A Arrest

Normal

1B

Bradycardia

AV block

1C

Normal or tachycardia

AV block

Progressive sinus bradycardia until sinus arrest suggests vasovagal AV block with associated sinus bradycardia suggests vasovagal Abrupt AV block without sinus slowing suggests intrinsic AV node disease

Bradycardia 2A 2B

Decrease >30% HR 10 sec

Normal Normal

Suggests vasovagal Suggests vasovagal

Normal Normal

Suggests noncardiac cause; unlikely vasovagal Suggests vasovagal

Sinus acceleration suggests orthostatic intolerance or noncardiac cause Atrial fibrillation Supraventricular tachycardia Ventricular tachycardia

Minimal HR change 3A