1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899
DIASTOLOGY: CLINICAL APPROACH TO DIASTOLIC HEART FAILURE
ISBN: 978-1-4160-3754-5
Copyright © 2008 by Saunders, an imprint of Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permissions may be sought directly from Elsevier’s Rights Department: phone: (+1) 215 239 3804 (US) or (+44) 1865 843830 (UK); fax: (+44) 1865 853333; e-mail:
[email protected]. You may also complete your request on-line via the Elsevier website at http://www.elsevier.com/permissions.
Notice Knowledge and best practice in this field are constantly changing. As new research and experience broaden our knowledge, changes in practice, treatment, and drug therapy may become necessary or appropriate. Readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of the practitioners, 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 Library of Congress Cataloging-in-Publication Data Diastology : clinical approach to diastolic heart failure / [edited by] Allan L. Klein, Mario J. Garcia.—1st ed. p. ; cm. Includes bibliographical references and index. ISBN 978-1-4160-3754-5 1. Congestive heart failure. 2. Heart—Left ventricle—Pathophysiology. 3. Diastole (Cardiac cycle) I. Klein, Allan L. II. Garcia, Mario J. III. Title: Clinical approach to diastolic heart failure. [DNLM: 1. Heart Failure, Congestive. 2. Diagnostic Techniques, Cardiovascular. 3. Diastole— physiology. 4. Ventricular Dysfunction, Left. WG 370 D541 2008] RC685.C53D53 2008 616.1′29—dc22 2007042084
Executive Publisher: Natasha Andjelkovic Project Manager: Mary B. Stermel Design Direction: Karen O’Keefe Owens Marketing Manager: Todd Liebel Developmental Editor: Pamela Hetherington
Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org
Printed in China Last digit is the print number: 9 8 7 6 5 4 3 2 1
Contributors NASER M. AMMASH, MD, FACC
ROBERT O. BONOW, MD, FACC
Associate Professor of Medicine Mayo Clinic Rochester, Minnesota
Goldberg Distinguished Professor Northwestern University Feinberg School of Medicine Chief, Division of Cardiology Co-Director Bluhm Cardiovascular Institute Northwestern Memorial Hospital Chicago, Illinois
CHRISTOPHER P. APPLETON, MD, FACC Professor of Medicine Division of Cardiovascular Diseases Mayo Clinic Arizona Phoenix, Arizona
CRAIG R. ASHER, MD, FACC Cardiology Fellowship Director Cleveland Clinic Florida Weston, Florida
GERARD P. AURIGEMMA, MD, FACC, FASE, FAHA Professor of Medicine and Radiology Director, Cardiology Fellowship Program University of Massachusetts Medical School Director, Non-Invasive Cardiology Umass Memorial Healthcare Worcester, Massachusetts
CATALIN F. BAICU, PhD Assistant Professor of Medicine Gazes Cardiac Research Institute Medical University of South Carolina Charleston, South Carolina
AJAY BHARGAVA, MD Staff Cardiologist Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
EDMUND A. BERMUDEZ, MD, MPH, FACC Assistant Professor of Medicine Tufts University School of Medicine Boston, Massachusetts Consultant, Cardiac and Vascular Disease Florida Cardiac Consultants Sarasota, Florida
D. DIRK BONNEMA, MD Cardiology Fellow Division of Cardiology Department of Medicine Medical University of South Carolina Charleston, South Carolina
BARRY A. BORLAUG, MD Assistant Professor of Medicine Division of Cardiovascular Diseases Mayo Clinic Rochester, Minnesota
SHEMY CARASSO, MD University of Toronto HCM Clinic, Division of Cardiology Department of Medicine, Toronto General Hospital University Health Network University of Toronto Toronto, Ontario, Canada
MANUEL D. CERQUEIRA, MD, FACC, FAHA Professor of Radiology Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Chairman, Department of Nuclear Medicine and Staff Cardiologist Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
KRISHNASWAMY CHANDRASEKARAN, MD Professor of Medicine Consultant, Division of Cardiovascular Diseases Mayo Clinic Rochester, Minnesota
HSUAN-HUNG CHUANG, MBBS, MRCP (UK), FAMS, FESC, FACC Consultant Cardiologist Mount Elizabeth Hospital Visiting Consultant Heart Failure and Transplantation Program National Heart Center Singapore
v
vi
Contributors
RONAN CURTIN, MD, MSc
BRIAN D. HOIT, MD, FACC, FASE
Associate Staff Cardiologist Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
Professor of Medicine and Physiology and Biophysics Case Western Reserve University Director of Echocardiography University Hospitals Case Medical Center Cleveland, Ohio
BENJAMIN W. EIDEM, MD
JERRY M. JOHN, MD, MS
Associate Professor of Pediatrics Divisions of Pediatric Cardiology and Cardiovascular Diseases Mayo Clinic Rochester, Minnesota
Cardiovascular Fellow Division of Cardiovascular Medicine University of Toledo Toledo, Ohio
KANIZ FATEMA, MBBS, PhD
TRACI L. JURRENS, MD
Research Fellow in Cardiovascular Diseases Mayo School of Graduate Medical Education Mayo Clinic Rochester, Minnesota
Cardiovascular Fellow Mayo Clinic Rochester, Minnesota
ANNE S. KANDERIAN, MD
GARY S. FRANCIS, MD
Advanced Fellow in Cardiovascular Imaging Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
Professor of Medicine Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Head, Section of Clinical Cardiology Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
HIDEKATSU FUKUTA, MD, PhD Postdoctoral Research Fellow Wake Forest University School of Medicine Wake Forest University Baptist Medical Center Winston-Salem, North Carolina Department of Cardio-Renal Medicine and Hypertension Nagoya City University Graduate School of Medical Sciences Nagoya, Japan
WILLIAM H. GAASCH, MD, FACC Professor of Medicine University of Massachusetts Medical School Worcester, Massachusetts Senior Consultant in Cardiology Director of Cardiovascular Research Lahey Clinic Burlington, Massachusetts
MARIO J. GARCIA, MD, FACC, FACP Professor of Medicine and Radiology Mount Sinai School of Medicine Director of Cardiac Imaging Mount Sinai Heart Institute New York, New York
RICHARD A. GRIMM, DO, FACC, FASE Director, Echocardiography Laboratory Staff Cardiologist Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
DAVID A. KASS, MD Abraham and Virginia Weiss Professor of Cardiology Professor of Medicine Professor of Biomedical Engineering Johns Hopkins Medical Institutions Attending Physician Division of Cardiology, Department of Medicine Johns Hopkins Hospital Baltimore, Maryland
DALANE W. KITZMAN, MD, FACC Professor of Internal Medicine, Cardiology and Gerontology Wake Forest University Health Sciences Winston-Salem, North Carolina
ALLAN L. KLEIN, MD, FRCP(C), FACC, FAHA, FASE Professor of Medicine Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Director, Cardiovascular Imaging Research Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
LIVIU KLEIN, MD, MS Fellow in Cardiovascular Disease Northwestern University, The Feinberg School of Medicine Bluhm Cardiovascular Institute Northwestern Memorial Hospital Chicago, Illinois
SANJAY KUMAR, MD Hospitalist Department of Hospital Medicine Ohio Permanente Medical Group Kaiser Permanente Cleveland, Ohio
Contributors
DOUGLAS S. LEE, MD, PhD, FRCP(C)
SHERIF F. NAGUEH, MD, FACC
Assistant Professor of Medicine, University of Toronto Scientist, Institute for Clinical Evaluative Sciences Attending Physician, Division of Cardiology, University Health Network Toronto, Ontario, Canada
Professor of Medicine Weill Cornell Medical College Associate Director, Echocardiography Laboratory Methodist DeBakey Heart Center Houston, Texas
STEVEN J. LESTER, MD, FACC, FRCPC, FASE
SATOSHI NAKATANI, MD, PhD, FACC
Associate Professor of Medicine, Mayo Clinic Director, Cardiovascular Ultrasound Imaging and Hemodynamic Laboratory Consultant Division of Cardiovascular Diseases Mayo Clinic Arizona Scottsdale, Arizona
Staff Cardiologist Department of Cardiology National Cardiovascular Center Suita, Osaka, Japan
BENJAMIN D. LEVINE, MD, FACC Professor of Medicine University of Texas Southwestern Medical Center at Dallas Director, Institute for Exercise and Environmental Medicine S. Finley Ewing Jr. Chair for Wellness at Presbyterian Hospital of Dallas Harry S. Moss Heart Chair for Cardiovascular Research Presbyterian Hospital of Dallas Dallas, Texas
WILLIAM C. LITTLE, MD, FACC McMichael Professor and Vice Chair of Internal Medicine Chief of Cardiology Wake Forest University School of Medicine Winston-Salem, North Carolina
VOJTECH MELENOVSKY, MD, PhD Department of Cardiology Institute for Clinical and Experimental Medicine-IKEM Prague, Czech Republic
ARUMUGAM NARAYANAN, MD Research Fellow Cardiology Division, Department of Medicine University of Massachusetts Medical School University of Massachusetts Memorial Health Care Worcester, Massachusetts
M. GARY NICHOLLS, MD, ChB, FAHA, FACC, FRCP Professor of Medicine Department of Medicine Christchurch School of Medicine and Health Sciences University of Otago Christchurch, New Zealand
JAE K. OH, MD, FACC, FAHA Professor of Medicine Codirector, Echocardiography Laboratory Director, Pericardial Clinic Consultant, Division of Cardiovascular Diseases Mayo Clinic Rochester, Minnesota
MARTIN OSRANEK, MD, MSc ANNITTA J. MOREHEAD, BA, RDCS, CCRC, FASE Manager, Cardiovascular Imaging Core, C5 Research Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
JAMES P. MORGAN, MD, PhD Professor of Medicine Tufts University Chief of Cardiovascular Medicine Caritas Carney Hospital and Caritas St. Elizabeth’s Medical Center Director of the Cardiovascular Center Caritas Christi Healthcare System Boston, Massachusetts
ROSS MURPHY, MD Consultant Cardiologist St James’s Hospital Dublin, Ireland
Assistant Professor of Medicine Research Associate, Cardiovascular Disease Mayo Clinic Rochester, Minnesota
YUTAKA OTSUJI, MD, PhD, FACC Professor of Medicine The Second Department of Internal Medicine Director Department of Cardiovascular and Renal Disease University of Occupational and Environmental Health, Japan School of Medicine Kitakyushu, Japan
ZORAN B. POPOVIĆ, MD, PhD Project Staff Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
vii
viii
Contributors
ANAND PRASAD, MD
SRIKANTH SOLA, MD, FACC, FAHA
Interventional Cardiology Fellow Department of Medicine, Division of Cardiology University of California–San Diego San Diego, California
Assistant Professor of Medicine Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Staff Cardiologist Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
MIGUEL A. QUINONES, MD, FACC Professor of Medicine Weill Cornell Medical College Chairman, Department of Cardiology The Methodist Hospital Houston, Texas
AMBEREEN QURAISHI, MD Clinical Instructor Senior Cardiovascular Fellow Department of Cardiovascular Medicine Caritas St. Elizabeth Medical Center Boston, Massachusetts
HARRY RAKOWSKI, MD, FRCPC, FACC, FASE Professor of Medicine Department of Medicine University of Toronto Douglas Wigle Research Chair in Hypertrophic Cardiomyopathy Development Director Peter Munk Cardiac Imaging Centre, Division of Cardiology, Department of Medicine Toronto General Hospital, University Health Network Toronto, Ontario, Canada
JAY RITZEMA-CARTER, BM, MRCP Cardiology Research Fellow Department of Medicine Christchurch School of Medicine and Health Sciences University of Otago Christchurch, New Zealand
L. LEONARDO RODRIGUEZ, MD, FACC Director, Stress Laboratory Program Director, Advanced Cardiovascular Imaging Fellowship Program Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
JAMES B. SEWARD, MD, FACC John M. Nasseff Sr. Professorship in Cardiology and Professor of Medicine and Pediatrics Consultant, Division of Cardiovascular Diseases Mayo Clinic Rochester, Minnesota
DAVID H. SPODICK, MD, DSc, FACC Professor of Medicine Cardiovascular Medicine University of Massachusetts Medical School St. Vincent Hospital Worcester, Massachusetts
RANDALL C. STARLING, MD, MPH, FACC Professor of Medicine Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Vice Chairman, Department of Cardiovascular Medicine Section Head, Heart Failure and Cardiac Transplant Medicine Medical Director, Kaufman Center for Heart Failure Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
A. JAMIL TAJIK, MD, FACC Thomas J. Watson, Jr. Professor in honor of Dr. Robert L. Frye Professor of Medicine and Pediatrics Chairman (Emeritus) Zayed Cardiovascular Center Mayo Clinic Rochester, Minnesota Consultant, Division of Cardiovascular Diseases, Internal Medicine and Pediatric Cardiology Mayo Clinic Scottsdale, Arizona
W. H. WILSON TANG, MD, FACC, FAHA Assistant Professor of Medicine Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Staff Cardiologist, Section of Heart Failure and Cardiac Transplantation Medicine Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
CHUWA TEI, MD, FACC Professor and Chairman Department of Cardiovascular, Respiratory, and Metabolic Medicine Graduate School of Medicine, Kagoshima University Kagoshima, Japan
Contributors
JAMES D. THOMAS, MD, FACC, FAHA, FESC
RAMACHANDRAN S. VASAN, MD, DM, FACC, FAHA
Professor of Medicine Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Charles and Lorraine Moore Chair of Cardiovascular Imaging Department of Cardiovascular Medicine Heart and Vascular Institute Cleveland Clinic Cleveland, Ohio
Professor of Medicine Boston University School of Medicine Framingham Heart Study Framingham, Massachusetts
FILIPPOS TRIPOSKIADIS, MD, FESC, FACC Professor of Cardiology Director, Department of Cardiology Larissa University Hospital Larissa, Greece
RICHARD W. TROUGHTON, MB, ChB, PhD Associate Professor Department of Medicine Christchurch School of Medicine and Health Sciences University of Otago Christchurch, New Zealand
TERESA S. M. TSANG, MD Professor of Medicine Consultant, Division of Cardiovascular Diseases Mayo Clinic Rochester, Minnesota
MICHAEL R. ZILE, MD, FACC Charles Ezra Daniel Professor of Medicine Attending Physician Department of Medicine, Division of Cardiology Medical University of South Carolina Director of the Medical Intensive Care Unit Ralph H. Johnson Department of Veteran’s Affairs Medical Center Charleston, South Carolina
WILLIAM A. ZOGHBI, MD, FACC Professor of Medicine Weill Medical College of Cornell University William L. Winters Chair in Cardiovascular Imaging Director, Cardiovascular Imaging Center The Methodist DeBakey Heart Center Houston, Texas
ix
Foreword
Doctors Klein and Garcia have attacked a controversial yet extraordinarily problematic clinical challenge with aggressiveness, insight, and thoroughness. Their text, Diastology: Clinical Approach to Diastolic Heart Failure, is one of the few organized efforts to bring together noted experts and synthesize today’s knowledge base regarding this fascinating problem. The depth and breadth of topics covered in this text are extraordinary, and the editors and individual contributors are to be congratulated. The importance of this subject cannot be overemphasized because it is now well established that almost half of all patients admitted to the hospital for symptomatic congestive heart failure have relatively preserved or normal left ventricular systolic function. It is assumed that some degree of abnormality in diastolic function contributes to the presentation of these patients and, indeed, predicts decompensation. Knowing that well over 5 million patients in North America alone have heart failure makes, then, the importance of this syndrome high. The debate about syndrome nomenclature has been entertaining. Should we be referring to the difficulty as “diastolic heart failure” or “heart failure in the setting of preserved left ventricular systolic function”? Doctors Klein and Garcia are not shy about their opinion of facts, as the title of the text indicates. Diastolic dysfunction can be defined as the inability of the heart to perform adequately under a normal filling pressure, and this generally results in impaired exercise tolerance resulting from varying combinations of inadequate forward cardiac output and elevated left ventricular end-diastolic pressure. Perhaps the most
important intrinsic left ventricular abnormality is slowing of the rate of left ventricular relaxation and increased stiffness of the chamber. The nuances of this finding and, indeed, the spectrum of definitions are well characterized and addressed in this text. Recognition of the importance of diastolic heart failure has been relatively recent, but there is now a large spectrum of data that gives us more precise information regarding the pathophysiology, epidemiology, and prognosis, and there are even a few recent trials studying therapy in these patients. Perhaps the latter is the most disappointing with little to guide us regarding the best approaches for reducing the substantive morbidity and mortality seen with this syndrome. Again, Doctors Klein and Garcia and their contributing authors are to be applauded for their thorough review of the current knowledge base regarding diastolic dysfunction and heart failure in the setting of preserved or relatively normal left ventricular systolic function. This text will capture the interests of clinicians, clinical investigators, and basic scientists interested in gaining insight into heart failure more generally. James B. Young, MD Chairman, Medicine Division Professor and Chairman George and Linda Kaufman Chair Cleveland Clinic Cleveland, Ohio
xiii
Foreword
In the past two decades, there have been remarkable advances in our appreciation of how cardiac diastolic function is important for patient well being. My first encounter with this occurred as a fellow in training in the early 1980s, when in Baltimore we were seeing so many elderly, predominantly females, with hypertension and hypercontractile left ventricles—but the seemingly paradoxical presentation with heart failure. It is interesting to reflect back on how ignorant we were back then with respect to the primacy of cardiac filling and relaxation and how far we have come since that time. In this book, the field of diastolic function and dysfunction is fully dissected. All of the major intersecting processes, such as hypertension, coronary artery disease, pericardial disease, valvular abnormalities, diabetes, and so many others, are tackled. The recent surge in the use of biventricular “resynchronization” and atrio-ventricular optimization has markedly accentuated the role of the interaction of systole and diastole in clinical practice—and we still have much to learn in selecting patients who will benefit from this “big ticket” technology. The simple notion that ideal patients would have a markedly dilated heart with left bundle branch block certainly has not held up for long. Our enhanced understanding of ventricular interaction and diastole, per se, should make a difference. New and improved imaging modalities
xiv
such as tissue Doppler imaging and strain rate imaging have given us a keener ability to quantify and differentiate “normal” and abnormal diastolic function. A variety of new indices, methods, and innovative imaging tools including torsion imaging will undoubtedly help illuminate the field in the years ahead. The book is as good as it gets in laying out a comprehensive assessment of where we are and where we are going in the comprehensive field of diastology. Drs. Klein and Garcia and the superb group of authors they have engaged have done quite an exceptional job in providing a panoramic view of a very dynamic field. Cardiologists in training, those in practice who have an interest in cardiac physiology, and virtually all academic cardiologists and sonographers who want to enhance their understanding of the implications of cardiac relaxation and filling will benefit from this fine contribution to cardiovascular medicine. Eric J. Topol, MD Professor of Translational Genomics, TSRI Director, Scripps Translational Science Institute Chief Academic Officer, Scripps Health Senior Consultant, Division of Cardiovascular Diseases Scripps Clinic La Jolla, California
Preface
A 65-year-old woman with hypertension presents to the emergency room with shortness of breath. Chest x-ray shows interstitial edema, and two-dimensional and Doppler echocardiography demonstrate an ejection fraction of 60%, concentric left ventricular (LV) hypertrophy, atrial enlargement, and stage 2 (moderate) LV diastolic dysfunction. The brain natriuretic peptide level is elevated at 800 pg/dl. This prototypic patient has evidence of classical diastolic heart failure, which is the inability to fill the left ventricle with normal filling pressures. This timely book addresses how diastolic heart failure is diagnosed and treated. It also comprehensively discusses the general principles of diastolic dysfunction, including the molecular biology, hemodynamics, epidemiology, clinical presentation, and principles of treatment. The contents of this book are targeted to a broad audience encompassing noninvasive and invasive cardiologists, physiology scientists, cardiology fellows, and cardiac sonographers. Multiple cardiovascular disorders cause diastolic dysfunction and subsequent diastolic heart failure. There is a raging controversy about whether diastolic heart failure exists as an independent entity or whether it is always accompanied by systolic dysfunction in the setting of a normal ejection fraction. In this book, we have elected to use the term diastolic heart failure; however, we recognize that systolic-ventricular interaction and arterial stiffening can definitely play a significant role in causing symptoms of heart failure in these patients. Why study diastolic heart failure? The answer is that a complete understanding of the pathophysiology of LV filling is essential to managing the patient with congestive heart failure syndromes. There has been a tremendous interest in diastology during the past 50 years, with over 16,000 original manuscripts published during this period.
Historical Perspective Since the heart was determined to be a pump, most biologists and physicians have focused on the study of systolic function. However, as early as in the renaissance period, Leonardo da Vinci described that the lower cardiac chambers of the heart filled with blood by drawing it from the upper chambers. In the 1940s, Carl J. Wiggers proposed the term inherent elasticity to describe the passive properties of the heart. In the 1970s, cardiac physiologists assessed the properties of active ventricular relaxation and passive filling using invasive quantification of intracavity pressure and volume. During the following decade, clinicians recognized that diastolic heart failure was an important cause of congestive heart failure, and Doppler echocardiography emerged as an important noninvasive method to assess the diastolic filling properties of the heart. The term “diastology” was coined in the early 1990s; imaging modalities, such as Doppler tissue imaging, color Mmode Doppler, and magnetic resonance imaging (MRI), advanced
our understanding of diastolic function. Over the past 10 years, new techniques and indices for assessing diastolic function have continued to evolve. Recent epidemiology-based studies have shown that diastolic heart failure is increasing in prevalence and that it is as common as systolic heart failure and just as fatal. In the past 5 years, there has been a shift from research in developing diagnostic techniques to large-scale clinical trials to determine targeted treatment for patients with diastolic heart failure.
Our Interest in the Field In the late 1980s, Allan Klein started his interest in this field as a Canadian Heart Foundation fellow at the Mayo Clinic studying Doppler assessment of LV filling during acute myocardial infarction and after reperfusion. His first impression was that the quick bedside echocardiographic evaluation, including the mitral E/A ratio and deceleration time, was a simple but powerful measure of LV diastolic filling, relaxation, and prognosis. Also, he was struck by how the stages of diastolic filling related to the clinical exam, including the extra heart sounds (S3 and S4). As a student of the field, he also learned that the study of diastolic function was more complex than the simple analysis of the mitral E/A ratio. During his training, Dr. Klein was very fortunate to have excellent mentors, including Liv Hatle, Jamil Tajik, and James Seward. Dr. Mario Garcia developed his interest in the field while at the Cleveland Clinic in the early days of tissue Doppler echocardiography, color M-mode Doppler, and strain rate imaging. His clinical observations and hemodynamic validation of early annular velocities (Em) and the slope of the flow propagation (Vp) as well as the relationship of mitral early filling/annular e wave (E/Em) and mitral early filling/flow propagation slope (E/vp) as measures of LV filling pressure were important for the advancement of the field. Their interest led to many diastology symposiums where leaders congregated in Cleveland, Ohio, and Scottsdale, Arizona, to discuss their advances. These summits sparked our interest to publish a state-of-the-art book on diastology.
Contents of the Book This book is organized into five main sections: basic determinants, diagnosis, specific cardiac diseases, emerging topics, and treatment. It includes a comprehensive analysis of the major areas of knowledge in this field from the molecular, genetic, and cellular mechanisms to clinical presentation and treatment of diastolic heart failure. This book discusses conventional and newer methods of diagnosis, including two-dimensional and Doppler echocardiographic techniques as well as cardiac MRI. An important practical chapter of how to actually perform a diastolic function examination written by one of the leading cardiac sonographers in the field xv
xvi
Preface is also included. A review of the prototypical diseases that manifest diastolic dysfunction, including hypertension, coronary artery disease, hypertrophic cardiomyopathy, restrictive cardiomyopathy, and pericardial diseases, provides an important clinical perspective. Newer topics that are addressed include the role of neurohormones, pacing, aging, and vascular cardiac interactions in diastolic heart failure. Finally, the general treatment, echocardiographic guided therapy, and ongoing clinical trials are covered in depth by the leading experts in the field. In the past, treatment of heart failure has focused purely on the treatment of systolic heart failure. However, there have been an increasing number of clinical trials, including The Candesartan in Heart Failure, Assessment of Reduction in Mortality and Morbidity (CHARM) preserved trial, the Perindopril for Elderly Patients with Chronic Heart Failure (PEP-CHF) trial, as well as
ongoing studies, including the Irbesartan in Heart Failure with Preserved Ejection Fraction (I-PRESERVE) and the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist (TOPCAT) trials, that address the treatment of diastolic heart failure. The importance of new drugs including endothelial receptor antagonists and glucose cross-link breakers that evaluate the targeted treatment of diastolic heart failure is also reviewed in this book. Finally, it is important to recognize that the field of diastology is a fast-moving target and we have tried to be as current as possible while also avoiding overlap in the chapters. We surely hope that you enjoy this exciting book. Allan L. Klein Mario J. Garcia
Acknowledgments
We would like to thank Marilyn, Jared, Lauren, and Jordan Klein and Cheryl, Melinda, and Olivia Garcia as well as our parents for their encouragement and support while editing this book. We especially would like to express our thanks to Marie Campbell, who helped and guided us in the journey of putting this book together. Finally, we would like to express our gratitude to the editors of Elsevier for their guidance in making this book a great success.
xvii
1
AMBEREEN QURAISHI, MD JAMES P. MORGAN, MD, PhD
Molecular, Gene, and Cellular Mechanisms INTRODUCTION PATHOPHYSIOLOGY Excitation-Contraction and RepolarizationRelaxation Coupling The Failing Human Heart Abnormal Calcium Regulation in Heart Failure Sarcolemmal Receptors and Mechanisms Sarcoplasmic Reticulum Altered Calcium Responsiveness of Myofibrillar Tension TREATMENT, GENE THERAPY, AND THE FUTURE
INTRODUCTION Heart failure is an ancient diagnosis. Some of the earliest known descriptions are attributed to Hippocrates, circa 460–370 bce. He reported dozens of clinical case histories with examples of dyspnea, as seen in left heart failure, and dropsy, as seen in right heart failure.1 Over the centuries, it has become apparent that the symptoms of heart failure are the result of complex pathophysiologic processes that alter both anatomic and physiologic properties of the heart. In the United States alone, approximately 5 million people are afflicted by heart failure, with an annual incidence of 500,000 new diagnoses and 300,000 deaths.2,3 One third of patients with symptomatic heart failure have a normal ejection fraction, and only in the last decade has there been a shift in the paradigm, focusing attention on diagnosing and understanding diastolic dysfunction. In many cases, diastolic dysfunction is caused by one or more abnormalities of cardiac structure, such as hypertrophy, fibrosis, infiltrative disease, or pericardial constriction. However, many patients appear to have diastolic dysfunction due to cellular abnormalities of myocyte relaxation, which is reversible and transient and occurs mainly in the setting of ischemia or hypoxia. Other causes include cellular calcium overload or ATP depletion.
Metabolic processes such as alkalosis, cardiovascular drugs, and the hypertrophy process itself can also alter the contractile and metabolic phenotype.4 This chapter will focus on the cellular mechanisms of normal and impaired myocardial relaxation. Calcium plays a crucial role in this process. Its movement into the cytosol, caused by excitation at the surface membrane, stimulates contraction. This physiologic phenomenon is appropriately termed excitation-contraction coupling. The reverse of this process, which is dependent on the movement of calcium and sodium, can be thought of as repolarization-relaxation coupling.4 In normal hearts, sympathetic stimulation can increase contractility by exerting a positive inotropic effect as part of the excitation-contraction (E-C) process and can also increase the rate of relaxation, a positive lusitropic (relaxation-enhancing) effect, by facilitating Ca2+ removal from the sarcoplasm. Myocyte control of relaxation occurs through the regulation of calcium concentrations within the cytosol via four main pathways involving sarcoplasmic reticulum (SR) Ca2+ATPase, sarcolemmal Na+/Ca2+ exchange, sarcolemmal Ca2+ATPase, or mitochondrial Ca2+ uniport.5 The latter two pathways have been studied and appear to have no significant effect on the beat-to-beat regulation of intracellular calcium.6,7 We will discuss the other two mechanisms in detail along with myofibrillar calcium responsiveness, as much research is being directed in trying to develop novel therapeutic interventions targeted toward repolarization-relaxation coupling in an effort to maximize positive lusitropic effects.
PATHOPHYSIOLOGY Excitation-Contraction and RepolarizationRelaxation Coupling The E-C coupling process activates contraction by coupling the signal generated by the action potential (excitation) at the myocyte cell surface to the delivery of calcium into the cytosol that initiates contraction.1 After spontaneous depolarization of the surface membrane, Ca2+ enters the myocyte through voltage-gated L-type channels (which are dihydropyridine receptive and will be referred 3
Ch001-X3754.indd 3
1/17/2008 6:02:21 PM
4
Chapter 1 • Molecular, Gene, and Cellular Mechanisms to as DHRs). This influx serves as a trigger for the release of stored Ca2+ from the SR via Ca2+ release channels called ryanodine receptor 2 (RyR2). This process is known as calcium-induced calcium release (CICR).8,9 The transverse tubule system, unique to the mammalian heart, allows for the close proximity of the DHRs to clusters of RyR2. The initial influx of calcium through L-type channels activates the RyR2 channels within its domain in a synchronized fashion, spontaneously increasing local concentration of calcium and producing a Ca2+ spark (named according to its appearance by confocal microscopy), which can also be thought of as a local Ca2+ transient.10 During E-C coupling, several thousand Ca2+ sparks occur in synchrony, overlapping in time and space, thus allowing for a global and uniform calcium transient that is sufficient for myocardial contraction. Another channel within the SR membrane, inositol triphosphate, has been reported to induce the release of calcium, but the rate and extent of release is much lower, and it is not triggered by CICR. It is important to understand the cardiac contractile apparatus from the vantage point of troponin-C, the Ca2+ receptor protein. When cytosolic calcium levels are low, there is minimal or no calcium bound to troponin-C. The troponin-tropomyosin complex inhibits the formation of the actomyosin complex. Once calcium is released from the SR, Ca2+ binds to troponin-C, changing the configuration of the troponin-tropomyosin complex, removing the inhibition of the actin-myosin interaction, and thus allowing cross-bridge cycling and contraction. Another important messenger that modulates E-C coupling is cyclic adenosine monophosphate (cAMP). This pathway is initiated by activation of the β-adrenergic receptors, which then activate adenylate cyclase and generate cAMP. This nucleotide activates a series of phosphorylating enzymes (i.e., protein kinases). Protein kinase A (PKA) binds to A-kinase anchoring proteins (AKAPs)11,12 and phosphorylate calcium regulatory proteins at multiple subcellular sites. Phosphorylation of both the voltagegated L-type channels and RyR2 channels (Ca2+ release channels) leads to a net effect of increased Ca2+ entry and a greater calcium release (a function of the amount of Ca2+ stored) from the SR, producing an increase in the rate and magnitude of force generation (positive inotropic effect).13 This is balanced by the phosphorylation of phospholamban on the SR, which regulates the sarcoendoplasmic reticulum Ca2+ (SERCA)-ATPase pump and allows for greater reuptake of calcium from the cytosol, enhancing the rate of relaxation (positive lusitropic effect). Finally, the phosphorylation of troponin-I, part of the regulatory complex of the contractile apparatus, facilitates the dissociation of calcium from troponin-C by altering the myofilament calcium sensitivity, also causing a positive lusitropic effect. These changes increase the amplitude of the systolic Ca2+ transient and decrease its duration, thus increasing myocyte contractility and generation of force. When the β-adrenergic pathway is activated in times of increased cardiac output demand, there is an increased heart rate and simultaneous increase in force (positive force frequency)14 and an increase in rate of relaxation, which ensures that sufficient calcium is available from the SR for the next beat.
The Failing Human Heart In the failing human heart, adrenergic effects are blunted.13 Numerous studies of mammalian heart models have been done to better understand the mechanism that leads to abnormal adrenergic signaling. The normal human atrial and ventricular myocardium expresses β1- and β2-adrenergic receptors at a ratio of about
Ch001-X3754.indd 4
70:30.15 In the early stages of heart failure, there is an increase in sympathetic stimulation in order to maintain adequate blood pressure and cardiac output. As the syndrome progresses, the continuous activation eventually leads to a downregulation of β1 receptors by almost 60% to 70% and desensitization of the remaining receptors, which depresses cAMP levels, hence contributing to altered calcium regulation and reducing intracellular calcium levels.16–18 The overall effect is a depressed and negative force-frequency relationship. In vitro studies of nonfailing and failing human myocytes have been done of contractility under varying conditions (i.e., varying muscle length and loading conditions as seen with pressure or volume changes in vivo, slow rates, [Ca2+], and catecholamine stimulation).19,20 These studies concluded that during basal conditions, the contractile properties of normal and failing myocytes is similar. Conversely, the same is not true during increased inotropic stimulation. The normal myocyte is able to increase force generation, but in failing myocytes, the developed force either decreases or remains unchanged. Therefore, during low workload states, contractility is preserved, but “contractility reserve” (the ability to increase contractility with heart rate or sympathetic stimulation) is severely reduced in failing myocytes.13 The proposed mechanism at this time is that the amount of calcium that is released by the SR of failing myocytes is less than in normal myocytes, which may be due to decreased SR Ca2+ stores or abnormal SR Ca2+ loading21; however, there is a need for more investigational studies to fully understand the mechanisms behind the E-C coupling defects that contribute to dysfunctional Ca2+ handling.
Abnormal Calcium Regulation in Heart Failure Abnormal modulation of intracellular calcium is a major mechanism that underlies both systolic and diastolic dysfunction and develops with cardiac hypertrophy and failure. The heart spends more than half of its time in diastole (i.e., relaxation and filling), yet there is controversy in the definition of diastolic dysfunction and diastolic heart failure. Abnormal calcium regulation comprises a spectrum of changes that include slowed force (or pressure) decline and cellular re-lengthening (increased ventricular stiffness), increased (or decreased) early filling rates and deceleration, an elevated diastolic pressure-volume relationship, and a filling-rate–dependent pressure elevation (increased end-diastolic pressure).22 These changes occur due to abnormal regulation of Ca2+ homeostasis within the myocyte. Although the cellular and molecular regulation of calcium in failing myocardium has been studied, the exact mechanism of abnormal calcium regulation is still not well understood (Fig. 1-1). Normally, the systolic calcium transient occurs with the release of calcium from the SR, which is triggered by the L-type channel Ca2+ influx. The magnitude of calcium release is dependent on the Ca2+ influx and the amount of calcium stored in the SR.23 Calcium reuptake is mediated by the SR Ca2+-ATPase pump and the sarcolemmal Na+-Ca2+ exchanger (NCX), which results in the decay of the Ca2+ transient and normal relaxation. In failing myocytes, there are numerous changes (e.g., expression levels of these membrane proteins), which result in abnormal Ca2+ transient, particularly in its termination. This also causes a prolonged action-potential duration and delayed relaxation. At slow pacing rates, the Ca2+ transients of normal and failing myocytes are very similar. However, when the heart rate increases, the Ca2+ transient becomes significantly different due to a negative force-frequency relationship and decreased SR Ca2+ release, as discussed earlier.
1/16/2008 2:31:17 PM
Chapter 1 • Molecular, Gene, and Cellular Mechanisms
B-ar
P
5
NCX LTC Na+
cAMP 2 1
P
Figure 1-1 Major Cellular Mechanisms of Diastolic Dysfunction. 1. Inhibition of cyclic AMP (cAMP) formation via adenylate cyclase. cAMP mediates phosphorylation of SERCA2 and the myofilaments, which reduces Ca2+ uptake and increases myofilament Ca2+ responsiveness (MyoCAr), respectively. Ca2+ entry via the L-type Ca2+ channels (LTC) is also increased. 2. Enhancement of Na+/Ca2+ exchange. 3. Blockade of reuptake of Ca2+ by SERCA2 by non–cAMPdependent mechanisms. 4. Decreased myofilament Ca2+ responsiveness by non–cAMP-dependent mechanisms. B-ar, β-adrenergic receptor; p, phosphorylation site; SR, sarcoplasmic reticulum.
Ca2+
SERCA2 SR Re-uptake X3 Released Ca2+ Ca2+
Sarcolemmal Receptors and Mechanisms The cardiac sarcolemma is a complex structure that contains multiple channels, exchangers, and pumps that are necessary for normal E-C coupling and myocyte contraction. In recent years, the sarcolemmal NCX has surfaced as one of the primary agents necessary to extrude calcium with each heart beat to allow normal relaxation. By virtue of the NCX mechanism, the role of the electrochemical sodium gradient has also been studied in various mammalian species as a potential determinant of [Ca2+]i.24,25 The NCX is a bidirectional electrogenic ion transporter that utilizes the Na+ electrochemical gradient to exchange one calcium for three sodium ions. During repolarization, the negative membrane potential and elevated [Ca2+]i drive the NCX toward a forward mode (Na+ in/Ca2+ out), resulting in extrusion of calcium from the cell in diastole. When the membrane potential is positive ([Na+]i is increased), the NCX functions in reverse mode (Na+ out/Ca2+ in), resulting in calcium influx, which may help regulate SR Ca2+ load and also, in conjunction with the Ca2+ current induced by activated DHRs, regulate SR Ca2+ release.21,26 There still remain some conflicting data in regard to the effects of CICR on SR by the reverse mode INaCa. CICR occurs mainly in the dyadic cleft space in the T-tubular regions.27 Several studies using the detubulation method in rat ventricular myocytes concluded that a majority of the NCX was localized to the T-tubules.28 However, Scriven et al. studied the distribution of these proteins using high-resolution imaging and showed that the NCX was localized to areas outside of the dyadic cleft space and that the spatial proximity of DHR with RyR2 was higher than with NCX29; therefore, any triggered SR Ca2+ release by the reverse mode INaCa seems highly inefficient.30 Several animal models with myocyte hypertrophy demonstrated elevated [Na+]i compared with normal. Pieske and Houser measured [Na+]i in failing and nonfailing human myocytes using multiple techniques and were the first to report elevated [Na+]i levels in failing human myocytes.24 The exact mechanism of this is under further investigation. Several studies have tried to describe the processes that lead to Na+ influx in the failing
Ch001-X3754.indd 5
Ca2+
X
4
“MyoCar”
Myofilaments
myocyte and the role of this influx in governing calcium homeostasis. One proposed mechanism suggests that sodium influx occurs through sodium channels and raises levels within the Na+ microdomain activating the NCX in reverse mode (Ca2+ influx), which is complementary to the influx generated by L-type calcium channels. Because the NCX is dependent on the Na+ electrochemical gradient for functioning, any changes that occur in sodium regulation in heart failure, when taken to an extreme, can result in [Ca2+]i overload and diastolic dysfunction. Few studies have addressed the role of the Na+-H+ exchanger in failing myocytes, which has a 1:1 stoichiometry of Na+ influx for H+ efflux. Its stimulation may be increased in heart failure and lead to elevated [Na+]i. Decrease in intracellular pH (which increases H+) can also increase Na+ by means of the Na+-H+ exchanger, secondarily increasing Ca2+ by means of the NCX. Alteration of these exchanges may contribute to prolongation of the action potential, slowed decay of the Ca2+ transient, and a delayed relaxation in failing myocardium. Several models of heart failure and hypertrophy have also shown an increase in expression of NCX,31,32 which may be partially controlled by the decreased sympathetic stimulation, leading to decreased SERCA activity.33 Terraciano et al.25,34 showed that in myocytes from transgenic (heterozygous) mice with upregulation of NCX, there was an increase in reverse-mode function resulting in an increase in SR calcium stores compared with wildtype myocytes. It is important to remember that in smaller mammals (mice and rats), the action potential is shorter than the duration of the calcium transient, which means that repolarization is occurring during most of the calcium transient, favoring forwardmode NCX.35 In these species there are lower levels of NCX, so any efflux via NCX makes a very small contribution to the decay of the calcium transient, even when overexpressed.20 However, there is a high [Na+]i in smaller mammals,36 favoring Ca2+ entry by reverse mode during the latter part of the calcium transient, which becomes more pronounced in myocytes overexpressing NCX, explaining the findings reported by Terraciano’s group. In normal human myocytes, the action-potential duration is prolonged, and SERCA release and reuptake occur during the
1/16/2008 2:31:17 PM
6
Chapter 1 • Molecular, Gene, and Cellular Mechanisms plateau phase, when NCX is not in forward mode, indicating that most of the elimination of cytosolic calcium is dependent on reuptake by the SERCA pump. However, in failing myocardium, the interaction between these two proteins changes due to an increase in the ratio of NCX to SERCA levels. To better understand whether increased NCX activity in the setting of reduced SERCA activity affects diastolic function, a study discriminating failing human hearts into three groups based on diastolic dysfunction with increased stimulation rate was performed.37 The investigators discovered three different phenotypes with varying expression of SERCA and NCX with impaired systolic function, but the overall trend in the ratio of NCX to SERCA was an increase by a factor of 2 to 4 in all groups of failing myocytes compared with normal. The phenotypes at either end of the spectrum ranged from increased levels of NCX and unchanged SERCA levels (group I) to decreased levels of SERCA and unchanged NCX levels (group III). Only the latter phenotype demonstrated both systolic and diastolic dysfunction, suggesting that both SR calcium uptake and the capacity to eliminate calcium from the cytosol are impaired, whereas in group I, the SR calcium uptake is impaired (causing systolic dysfunction), and global capacity to eliminate calcium is higher (preserving diastolic dysfunction). Therefore, the overexpression of NCX has positive correlation with diastolic function in failing myocytes.
Sarcoplasmic Reticulum The SR is an intracellular structure that is the most important store of calcium in the mammalian heart. The sarcoplasmic membrane proteins (RyR2 and SERCA) maintain a tight control of calcium release and uptake in E-C coupling, contractility, and relaxation. There is a 10,000-fold Ca2+ gradient maintained across the SR membrane by the SR Ca2+-ATPase pump.38 Molecular analysis has identified three homologous genes (SERCA1, SERCA2, and SERCA3) encoding the SERCA pumps. The SERCA2 gene is spliced into four variants that encode the isoforms. SERCA2a is the primary isoform expressed in cardiac muscle39 at high levels; however, there are regional differences, age-related effects, and variation due to thyroid hormone levels that affect expression levels. Experimental models in animals and humans have demonstrated that the expression level of SERCA in the atrium versus the ventricle is twofold and may account for shorter contraction time in atria versus ventricles.40 In fact, in heart failure models, it is well established that defective SR Ca2+ uptake correlates with decreased contractility, which could be attributed to significant decline in SERCA protein levels or an alteration in SR Ca2+ transport function.41 Several studies induced left ventricular pressure overload hypertrophy/failure in rats by thoracic aortic banding and consistently found an overall decrease in SERCA mRNA levels,42–44 suggesting that downregulation of SERCA2a gene expression in these models partly occurs at the transcriptional level.38 Feldman et al. also deduced that decreased SERCA mRNA levels could be a marker of transition from compensated hypertrophy to decompensated hypertrophy/failure.43 In human heart models, SERCA mRNA levels are reduced in failing compared with nonfailing hearts, yet there remains controversy regarding simultaneous decrease in SERCA protein expression.29 Recently, data from larger studies have successfully shown a reduction of SERCA protein levels in failing human hearts, but not in compensated hypertrophied human hearts, suggesting that perhaps a decrease in protein level is a sign of developing failure.6,45,46 A decrease in
Ch001-X3754.indd 6
SERCA2a expression at the level of mRNA or protein is inversely related to duration of cardiac contraction,47–49 correlates with decreased myocardial function, alters the force-frequency response,50 and may also slow the velocity of relaxation, suggesting the importance of SERCA pump level in maintaining myocardial function. Yet it is still difficult to describe the actual relationship between cardiac muscle and the SERCA pump because of the complexity of its regulation and all the changes in other calcium-regulating proteins (such as NCX) that occur in concert with each other within the myocyte and the heart as the syndrome of heart failure progresses. The SERCA pump is modulated by both direct and indirect factors. Phospholamban is the primary indirect regulator that activates the SERCA2a pump. In its dephosphorylated state, phospholamban inhibits SERCA2a affinity for calcium. The phosphorylation of phospholamban can occur at three different sites—serine-16 by cAMP-dependent PKA, threonine-17 by Ca2+/calmodulin-dependent protein kinase II, and serine-10 by Ca2+-activated phospholipid-dependent protein kinase.34 When phospholamban is phosphorylated by cAMP-dependent PKA (the most important mediator), the inhibitory effect is removed and the calcium affinity (not the maximal velocity of SERCA2a) increases, resulting in enhanced relaxation and an increase in SR Ca2+ load. Ca2+/calmodulin-dependent protein kinase II (CaMK II) is the other modulator, which directly phosphorylates SERCA2 and increases Vmax (maximal activity) without altering the calcium affinity of SERCA2.51 CaMK II also phosphorylates the threonine-17 site in phospholamban, which also increases the calcium affinity of SERCA2a. In heart failure, the blunting of the β-adrenergic pathway leads to alteration not only in phosphorylation of phospholamban at the serine-16 site, but also in CaMK-dependent phosphorylation, ultimately altering SERCA2a activity. Most studies of human heart failure have suggested that although there is a decrease in phospholamban mRNA levels, there is no difference in protein expression between failing and nonfailing myocytes.52,53 Therefore, protein expression of SERCA2a in relation to phospholamban is always diminished in heart failure, which may explain the increased phospholamban-to-SERCA2a ratios, leading to an increase in inhibition of the SERCA2a pump and an overall decrease in its basal activity level and contributing to abnormal calcium handling.
Altered Calcium Responsiveness of Myofibrillar Tension Cardiac contractility and relaxation are altered not only because of changes in calcium availability, but also because of changes in myofilament responsiveness to calcium. In fibers rendered hypermeable to Ca2+, a change in responsiveness can manifest as either a change in sensitivity or potency or as maximal Ca2+-activated force. The actual mechanism responsible for this effect has not been definitively determined, but based on numerous studies, it appears that isoform composition and phosphorylation status of the contractile proteins are altered, which increases the Ca2+ sensitivity of the contractile apparatus in end-stage heart failure. Most studies have concentrated on changes of a single factor. Van der Velden et al.54 focused on a combination of contractile protein changes that occur during heart failure by studying isometric force and its Ca2+ sensitivity in left ventricular myocytes from nonfailing and end-stage failing donor hearts. They concluded that the combined decrease in phosphorylation status of
1/16/2008 2:31:17 PM
Chapter 1 • Molecular, Gene, and Cellular Mechanisms troponin-I and myosin light chain 2 resulted in an increase in Ca2+ sensitivity, and not due to contractile protein isoform change. However, other studies have shown that the phosphorylation of myosin light chain 2 increases Ca2+ responsiveness.55 Earlier reports showed that end-stage heart failure in humans is not associated with myofibrillar Ca2+ sensitivity.56–58 Until now, no real consensus has been reached to explain if, why, and how Ca2+ sensitivity increases in heart failure patients. Therapeutically, Ca2+ sensitizers pose a problem due to the mechanism of action. An increased affinity of Ca2+ for troponin-C would enhance the actin-myosin interaction, which, theoretically, prolongs relaxation. This has been shown in vivo and in vitro in animals and in humans.57
TREATMENT, GENE THERAPY, AND THE FUTURE Congestive heart failure is a leading cause of morbidity and mortality in the United States. The remodeling process that occurs after myocyte injury from multiple causes leads to contractile dysfunction and abnormal intracellular calcium handling, which delays normal relaxation, as discussed in detail thus far. Although there is no universal agreement regarding the mechanisms, it is generally accepted that calcium handling is altered in failing hearts. The NCX and the SERCA2a pump are the two main players that regulate cytosolic calcium levels during relaxation on a beat-to-beat basis. Therefore, multiple studies using gene transfer methodologies are exploring the overexpression of these two mechanisms that could potentially return calcium handling to normal, as most of the conventional therapies that we currently offer utilize mainly multiple medications, the actions of which are not completely understood.59 The sodium-calcium exchanger is a complex transport protein that functions in reverse and forward modes. In normal mice myocytes, it is the most dominant Ca2+ efflux mechanism that helps maintain calcium homeostasis.60 In humans, only 25% of the calcium is extruded by the NCX, and the other 75% is removed by SERCA2a pumps.61 Some models of heart failure have described an increase in expression of NCX in response to the pathologic decrease in SERCA2a pumps attributed to the delayed and smaller Ca2+ transient.14 This is felt to be a compensatory mechanism for the decreased efficiency of SR calcium re-uptake. Because the actual quantitative contribution of the two regulatory mechanisms under pathologic conditions is poorly understood, so is the relationship between the changes in protein expression and the functional consequences.34 Terraciano et al. noted that the degree of compensation that may be achieved with a two- to threefold increase of NCX, the same measured in failing human hearts, has not been tested.34 They compared transgenic mice overexpressing NCX with nontransgenic mice and confirmed that a reduction in SERCA2a function can be compensated by overexpression of NCX. A 2.4-fold increase in the function of NCX compensated for a 28% reduction in SERCA2a and maintained Ca2+ transient. Although these findings may not necessarily be extrapolated to human hearts, it is a starting point for understanding such mechanisms and developing future therapies. The expression of SERCA2a pump level has also been an increasing area of interest given its central role in SR Ca2+ handling. Several models utilizing in vitro adenoviral-mediated gene transfer have been created to express varying levels of SERCA2a, which has allowed for a better understanding of the role of SERCA2a pump levels in maintaining Ca2+ homeostasis and
Ch001-X3754.indd 7
7
cardiac function.62–64 As discussed earlier, decrease in SERCA2a activity and/or expression has been identified as one of the major defective mechanisms responsible for impaired relaxation. By disrupting the SERCA2a gene using homologous recombination, Periasamy et al.65 obtained a better understanding of how the pathologic decrease in SERCA2a pump levels can alter cardiac function. In homozygous mice, the outcome was lethal. Heterozygous mice were able to function, but the Vmax of SR Ca2+ uptake was decreased by 35%. The peak amplitude of the Ca2+ transients in normal heterozygous myocyte was decreased by more than 30%, which led to a decrease in the maximal rate of contraction and relaxation, as confirmed by measurements obtained via transducers in the left ventricles and right femoral arteries of anesthetized mice. These findings prove that there is a direct relationship between SERCA2a level and cardiac contractility. Therefore, recent studies have used the adenoviral-mediated gene transfer to increase SERCA2a pump activity as a potential means for therapy. Hajjar et al.66 showed that increased expression of SERCA2a in rat myocytes led to increased contractility and faster decay of the Ca2+ transient. The question remained in the ability to restore function in failing human myocytes. Del Monte et al.67 addressed this very question by isolating human cardiomyocytes from the left ventricles of 10 patients with end-stage heart failure. Gene transfer of SERCA2a led to an increase in protein expression and pump activity, induced a faster contraction velocity, and enhanced relaxation velocity, thus reversing contractile abnormalities of the failing heart. Although the results were promising in the in vitro model, this method did not predict the outcome in vivo. Hajjar and colleagues used a catheter-based technique of gene transfer to allow for global overexpression of SERCA2a in an animal model of pressure-overload hypertrophy in transition to failure. SERCA2a overexpression restored both systolic and diastolic dysfunction to normal levels, decreased left ventricular size, and restored the slope of the end-systolic pressure-dimension relationship to that of control levels.68 The SERCA gene therapy studies described above demonstrate not only that it is possible to increase the expression of SERCA protein but that its effects can normalize the abnormalities of calcium handling and contraction. However, this raises several important concerns, such as the long-term effects of inducing these changes, impacts on the phospholamban and SERCA interaction, and the effect on the NCX. As heart failure continues to pose a major clinical challenge for patients and physicians, the limitations of current treatment modalities are reflected in the grave statistics of minimal improvement in overall mortality within the last century. Gene transfer targeted toward specific pathways in the failing human heart is a novel therapeutic option that can potentially reverse the abnormalities of diastolic dysfunction and transform heart failure into a chronic survivable condition.
REFERENCES 1. Katz AM: Heart failure: Pathophysiology, molecular biology, and clinical management. Philadelphia, Lippincott Williams & Wilkins, 2000:6. 2. Owan TE, Redfield MM: Epidemiology of diastolic heart failure. Prog Cardiovasc Dis 2005;47:320. 3. Hunt HA, Baker DW, Chin MH, et al: ACC/AHA guidelines for the evaluation and management of chronic heart failure in the adult: A report of the American College of Cardiology/American Heart Association task force on practice guidelines (committee to revise the 1995 guidelines for the evaluation and management of heart failure). Circulation 2001;104: 2996–3007.
1/16/2008 2:31:17 PM
8
Chapter 1 • Molecular, Gene, and Cellular Mechanisms 4. Apstein CS, Morgan JP: Diastolic relaxation of the heart, 2nd ed. London, Lea and Febiger, 1994:3–24. 5. Bers DM: Cardiac excitation-contraction coupling. Nature 2002;415: 198–205. 6. Hasenfuss G, Pieske B: Calcium cycling in congestive heart failure. J Molec Cell Cardiol 2002;34:951–969. 7. Bassani RA, Bassani JWM, Bers DM: Mitochondrial and sarcolemmal Ca transport can reduce [Ca]i during caffeine contractures in rabbit cardiac myocytes. J Physiol (London) 1992;453:591–608. 8. Scoote M, Poole-Wilson PA, Williams AJ: The therapeutic potential of new insights into myocardial excitation-contraction coupling. Heart 2003;89: 371–376. 9. Fabiato A: Calcium induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol 1983;245:C1–C14. 10. Cheng H, Lederer WJ, Cannell MB: Calcium sparks: Elementary events underlying excitation-contraction coupling in the heart muscle. Science 1993;262:740–744. 11. Bauman AL, Scott JD: Kinase- and phosphatase-anchoring proteins: Harnessing the dynamic duo. Nature (Cell Biol) 2002;4:E203–E206. 12. Fink MA, Zakhary DR, Mackey JA, et al: AKAP-mediated targeting of protein kinase A regulates contractility in cardiac myocytes. Circ Res 2001;88:291–297. 13. Houser SR, Margulies KB: Is depressed myocyte contractility involved in heart failure? Circulation 2003;92:350–358. 14. Gwathmey JK, Copelas L, MacKinnon R, et al: Abnormal intracellular calcium handling in myocardium from patients with end-stage heart failure. Circ Res 1987;61:70–76. 15. Lohse MJ, Engelhardt S, Eschenhagen T: What is the role of β-adrenergic signaling in heart failure? Circ Res 2003;93:896–906. 16. Brodde OE: β-adrenoreceptors in cardiac disease. Pharmacol Ther 1993;60:405–430. 17. Bristow MR: The adrenergic nervous system in heart failure. NEJM 1984;311:850–851. 18. Ungerer M, Bohm M, Elce JS, et al: Altered expression of β-adrenergic receptor kinase and β1-adrenergic receptors in the failing human heart. Circulation 1993;87:454–463. 19. Houser SR, Piacentino V 3rd, Mattiello J, et al: Functional properties of failing human ventricular myocytes. Trends Cardiovasc Med 2000;10: 101–107. 20. Houser SR, Piacentino V 3rd, Weisser J: Abnormalites of calcium cycling in the hypertrophied and failing heart. J Mol Cell Cardiol 2000;32: 1595–1607. 21. Pieske B, Maier LS, Bers DM, et al: Ca2+ handling and sarcoplasmic reticulum Ca2+ content in isolated failing and nonfailing human myocardium. Circ Res 1999;85:38–46. 22. Kass DA, Bronzwaer JGF, Paulus WJ: What mechanisms underlie diastolic dysfunction in heart failure? Circ Res 2004;94:1533–1542. 23. Bassani JW, Yuan W, Bers DM: Fractional SR Ca2+ release is regulated by trigger Ca2+ and SR Ca2+ content in cardiac myocytes. Am J Physiol 1995;268:C1313–C1319. 24. Pieske B, Houser SR: [Na+]i handling in the failing human heart. Cardiovasc Res 2003;57:874–886. 25. Blaustein MP, Lederer WJ: Sodium/calcium exchange: Its physiologic implications. Physiol Rev 1999;79:763–854. 26. Goldhaber JI, Scott TL, Walter DO, et al: Local regulation of the threshold for calcium sparks in rat ventricular myocytes: Role of sodium-calcium exchange. J Physiol (London) 1999;520(pt 2):431–438. 27. Reuter H, Pott C, Goldhaber JI, et al: Na+-Ca2+ exchange in the regulation of cardiac excitation-contraction coupling. Cardiovasc Res 2005;67: 198–207. 28. Kawai M, Hussain M, Orchard CH: Excitation-contraction coupling in rat ventricular myocytes after formamide-induced detubulation. Am J Physiol 1999;277:H603–H609. 29. Scriven DR, Dan P, Moore ED: Distribution of proteins implicated in excitation-contraction coupling in rat ventricular myocytes. Biophys J 2000;79:2682–2691. 30. Sipido K, Macs M, van de Werf F: Low efficiency of Ca2+ entry through the Na+-Ca2+ exchanger as triggers for Ca2+ release from the sarcoplasmic reticulum. A comparison between L-type Ca2+ current and reverse-mode Na+Ca2+ exchange. Circ Res 1982;50:651–662. 31. Hasenfuss G: Alterations of calcium-regulatory proteins in heart failure. Cardiovasc Res 1998;37:279–289. 32. Flesch M, Schwinger RHG, Schiffer F, et al: Evidence for functional relevance of an enhanced expression of the Na(+)-Ca2+ exchanger in failing human myocardium. Circulation 1996;94:992–1002.
Ch001-X3754.indd 8
33. Golden KL, Ren J, O’Connor J, et al: In vivo regulation of Na/Ca exchanger expression by adrenergic effectors. Am J Physiol (Heart Circ Physiol) 2001;280:H1376–H1382. 34. Terraciano CM, Philipson KD, MacLeod KT: Overexpression of the Na(+)/ Ca(2+) exchanger and inhibition of the sarcoplasmic reticulum Ca(2+)ATPase in ventricular myocytes from transgenic mice. Cardiovasc Res 2001;49:38–47. 35. Piacentino V 3rd, Weber CR, Gaughan JP, et al: Modulation of contractility in failing human myocytes by reverse-mode Na/Ca exchange. Ann NY Acad Sci 2002;976:466–471. 36. Hasenfuss G, Pieske B: Calcium cycling in congestive heart failure. J Molec Cell Cardiol 2002;34:951–969. 37. Hasenfuss G, Schillinger W, Lehnert SE, et al: Relationship between Na+Ca2+ exchanger protein levels and diastolic function of failing human myocardium. Circulation 1999;99:641–648. 38. Frank KF, Bölck B, Erdmann E, et al: Sarcoplasmic reticulum Ca2+-ATPase modulates cardiac contraction and relaxation. Cardiovasc Res 2003;57:20–27. 39. Zarain-Herzberg A, MacLennan DH, Periasamy M: Characterization of rabbit cardiac sarco(endo)-plasmic reticulum Ca2+-ATPase gene. J Biol Chem 1990;265:4670–4677. 40. Luss I, Boknika P, Jones LR, et al: Expression of cardiac calcium regulatory proteins in atrium v ventricle in different species. J Molec Cell Cardiol 1999;31:1299–1314. 41. Periasamy M, Huke S: SERCA pump level is a critical determinant of Ca2+ homeostasis and cardiac contractility. J Molec Cell Cardiol 2002;33: 1053–1063. 42. Dumas AR, Wisnewsky C, Boheler KR, et al: The sarco(endo)plasmic reticulum Ca2+-ATPase gene is regulated at the transcriptional level during compensated left ventricular hypertrophy in the rat. Academie des sciences 1997;320:963–969. 43. Feldman AM, Weinberg EO, Ray PE, et al: Selective changes in cardiac gene expression during compensated hypertrophy and the transition to cardiac decompensation in rats with chronic aortic banding. Circ Res 1993;73: 184–192. 44. Aoyagi T, Yonekura K, Eto Y, et al: The sarcoplasmic reticulum Ca2+ATPase (SERCA2) gene promoter activity is decreased in response to severe left ventricular pressure-overload hypertrophy in rat hearts. J Molec Cell Cardiol 1999;31:919–926. 45. Dash R, Frank KF, Carr AN, et al: Gender influences on sarcoplasmic reticulum Ca2+ handling in failing human myocardium. J Molec Cell Cardiol 2001;33:1345–1353. 46. Di Paola NR, Sweet WE, Stull LB, et al: Beta-adrenergic receptors and calcium cycling protein in non-failing, hypertrophied and failing human hearts: Transition from hypertrophy to failure. J Molec Cell Cardiol 2001;33:1283–1295. 47. Chen F, Ding S, Lee BS, Wetzel GT: Sarcoplasmic reticulum Ca(2+)ATPase and cell contraction in developing rabbit heart. J Molec Cell Cardiol 2000;32:745–755. 48. Gombosova I, Bokník P, Kirchhefer U, et al: Postnatal changes in contractile time parameters, calcium regulatory proteins and phosphatases. Am J Physiol 1998;274:H2123–H2132. 49. Cain BS, Meldrum DR, Joo KS, et al: Human SERCA2a levels correlate inversely with age in senescent human myocardium. J Am Coll Cardiol 1998;32:458–467. 50. Hasenfuss G, Reinecke H, Studer R, et al: Relation between myocardial function and expression of sarcoplasmic reticulum Ca2+ATPase in failing and nonfailing myocardium. Circ Res 1994;75: 434–442. 51. Toyofuku T, Kurzydlowski K, Narayanan N, et al: Identification of Ser38 as a site in cardiac sarcoplasmic reticulum Ca2+-ATPase that is phosphorylated by Ca2+/calmodulin-dependent protein kinase. J Biol Chem 1991;266:11144–11152. 52. Schwinger RH, Böhm M, Schmidt U, et al: Unchanged protein levels of SERCA II and phospholamban but reduced Ca2+ uptake and Ca2+-ATPase activity of cardiac sarcoplasmic reticulum from dilated cardiomyopathy patients compared with patients with non-failing hearts. Circulation 1995;92:3220–3228. 53. Meyer M, Schillinger W, Pieske B, et al: Alterations of sarcoplasmic reticulum proteins in failing human dilated cardiomyopathy. Circulation 1995;92:778–784. 54. Van der Velden J, Papp Z, Zaremba R, et al: Increased Ca2+-sensitivity of the contractile apparatus in end-stage human heart failure results from altered phosphorylation of contractile proteins. Cardiovas Res 2003;57: 37–47.
1/16/2008 2:31:17 PM
Chapter 1 • Molecular, Gene, and Cellular Mechanisms 55. Morano I, Arndt H, Gärtner C, et al: Skinned fibers of human atrium and ventricle: Myosin isoenzymes and contractility. Circ Res 1988;62: 632–639. 56. Hajjar RJ, Gwathmey JK, Briggs GM, et al: Differential effect of DPI 201– 206 on the sensitivity of myofilaments to Ca2+ in intact and skinned trabeculae from control and myopathic human hearts. J Clin Invest 1988;82:1578. 57. Wankerl M, Böhm M, Morano I, et al: Calcium sensitivity and myosin light chain pattern of atrial and ventricular stunned cardiac fibers from patients with various kinds of cardiac disease. J Molec Cell Cardiol 1990;22:1425. 58. D’Angelo A, Luciani GB, Mazzucco A, et al: Contractile properties and Ca2+ release activity of the sarcoplasmic reticulum in dilated cardiomyopathy. Circulation 1992;85:518–525. 59. Hoshijima M: Gene therapy targeted at calcium handling as an approach to the treatment of heart failure. Pharmacol Ther 2005;105:211– 228. 60. Goldhaber JI, Henderson SA, Reuter H, et al: Effects of Na+-Ca2+ exchange expression on excitation-contraction coupling in genetically modified mice. Ann NY Acad Sci 2005;1047:122–126. 61. Hasenfuss G: Calcium pump overexpression and myocardial function: Implications for gene therapy of myocardial failure. Circ Res 1998; 83:966–968.
Ch001-X3754.indd 9
9
62. He H, Giordano FJ, Hilal-Dandan R, et al: Overexpression of the rat sarcoplasmic reticulum Ca2+-ATPase gene in the heart of transgenic mice accelerates calcium transients and cardiac relaxation. J Clin Invest 1997;100:380–389. 63. Baker DL, Hashimoto K, Grupp IL, et al: Targeted expression of the sarcoplasmic reticulum Ca2+-ATPase increases cardiac contractility in transgenic mouse hearts. Circ Res 1998;83:1205–1214. 64. Müeller OJ, Katus HA, Bekeredjian R: Targeting the heart with gene therapy-optimized gene delivery methods. Cardiovasc Res 2007;73: 453–462. 65. Periasamy M, Reed TD, Liu LH, et al: Impaired cardiac performance in heterozygous mice with a null mutation in the sarco(endo)plasmic reticulum Ca2+-ATPase isoform 2 (SERCA2) gene. J Biol Chem 1999;274: 2556–2562. 66. Hajjar RJ, Kang JX, Gwathmey JK, et al: Physiological effects of adenoviral gene transfer of sarcoplasmic reticulum calcium ATPase in isolated rat myocytes. Circulation 1997;95:423–429. 67. Del Monte F, Harding SE, Schmidt U, et al: Restoration of contractile function in isolated cardiomyocytes from failing human hearts by gene transfer of SERCA2a. Circulation 1999;100:2308–2311. 68. Miyamoto MI, del Monte F, Schmidt U, et al: Adenoviral gene transfer of SERCA2a improves left ventricular function in aortic-banded rats in transition to heart failure. Proc Natl Acad Sci USA 2000;97:793–798.
1/16/2008 2:31:18 PM
D. DIRK BONNEMA, MD CATALIN F. BAICU, PhD MICHAEL R. ZILE, MD
2
Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness INTRODUCTION PATHOPHYSIOLOGY Normal Diastolic Function Measurements of Diastolic Function Left Ventricular Diastolic Function During Exercise CARDIOVASCULAR STRUCTURE AND FUNCTION IN DIASTOLIC HEART FAILURE Left Ventricular Chamber Remodeling Left Ventricular Diastolic Function in Diastolic Heart Failure Cardiomyocyte Diastolic Function in Diastolic Heart Failure
Left Ventricular Systolic Function in Diastolic Heart Failure Abnormal Diastolic Function in Decompensated Diastolic Heart Failure Abnormal Diastolic Function in Diastolic Heart Failure During Exercise CLINICAL RELEVANCE Ischemic Diastolic Dysfunction Left Ventricular Concentric Hypertrophy Trigger Mechanisms of Noncardiac Factors in Acute Decompensated Diastolic Heart Failure FUTURE RESEARCH
INTRODUCTION Heart failure (HF) can be defined physiologically as an inability of the heart to provide sufficient forward output to meet the perfusion and oxygenation requirements of the tissues while maintaining normal filling pressures. Chronic heart failure can be divided into two broad categories: systolic heart failure (SHF) and diastolic heart failure (DHF). Classifying patients in these two broad categories should be based on characteristic changes in cardiovascular structure and function (see Chapter 28).1,2 SHF is characterized by progressive chamber dilation, eccentric remodeling, and dominant abnormalities in systolic properties. Clinical manifestations of left ventricular (LV) systolic dysfunction include decreased cardiac output, increased heart rate, and peripheral vasoconstriction. However, patients with SHF frequently have additional symptoms of shortness of breath at rest or with exertion.3 These symptoms of pulmonary congestion are due, at least in part, to LV diastolic dysfunction.4–6 Therefore, patients with SHF (particularly when they have
symptomatic decompensation) do not have an “isolated” abnormality in systolic properties; rather, from the pathophysiologic point of view, they have predominant abnormalities in systolic properties and eccentric remodeling, with associated or secondary abnormalities in diastolic function. By contrast, patients with DHF are characterized by normal LV volume, concentric remodeling, and normal LV chamber systolic properties, but dominant abnormalities in diastolic properties.2,7–11 These patients have abnormalities in diastolic relaxation, filling, and/or distensibility. Clinical manifestations of LV diastolic dysfunction include shortness of breath at rest or with exertion and peripheral edema. However, abnormalities in regional systolic shortening have also been identified in some patients with DHF.2,11 Therefore, patients with DHF do not have “isolated” abnormalities in diastolic properties; rather, from the pathophysiologic point of view, they have predominant abnormalities in diastolic properties and concentric remodeling. “Diastolic dysfunction” and “diastolic heart failure” are not synonymous terms.12 Diastolic dysfunction indicates a functional abnormality of diastolic relaxation, filling, or distensibility of the 11
12
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness left ventricle—regardless of whether the ejection fraction is normal or abnormal and regardless of whether the patient is asymptomatic or has symptoms and signs of HF. Thus, diastolic dysfunction comprises abnormal mechanical (diastolic) properties of the ventricle and is present in virtually all patients with heart failure. Diastolic heart failure denotes signs and symptoms of clinical heart failure with normal ejection fraction and LV diastolic dysfunction. Similar distinctions apply to the terms “systolic dysfunction” and “systolic heart failure” (Boxes 2-1 and 2-2).
PATHOPHYSIOLOGY The pathophysiology of diastolic heart failure will be reviewed here, beginning with a discussion of the factors involved in normal diastolic relaxation and filling. Understanding normal diastolic function permits an easier understanding of some of the clinical features of DHF.
Normal Diastolic Function Cardiac function is critically dependent upon diastolic physiologic mechanisms to provide adequate LV filling (cardiac input) in Box 2-1
Definitions Diastolic Dysfunction: Abnormal diastolic properties of LV (abnormal relaxation, filling dynamics, distensibility) • EF may be normal or low. • Patient may be symptomatic or asymptomatic. Diastolic Heart Failure: Clinical heart failure, normal ejection fraction, abnormal diastolic function Systolic Dysfunction: Abnormal systolic properties of LV (abnormal performance, function, contractility)
parallel with LV ejection (cardiac output) both at rest and during exercise. Adequate pulmonary function is also dependent upon LV diastolic function. During diastole, the left ventricle, left atrium, and pulmonary veins form a “common chamber” that is continuous with the pulmonary capillary bed. LV diastolic pressure is determined by the volume of blood in the left ventricle during diastole and the diastolic distensibility or compliance of the entire cardiovascular system, principally the left ventricle (but it may also include the left atrium, pulmonary vessels, right ventricle, and systemic arteries). Thus, an increase in LV diastolic pressure (whether this occurs at rest or during exercise) will increase pulmonary capillary pressure, which, if high enough, causes dyspnea, exercise limitation, pulmonary congestion, and edema (Fig. 2-1). Relaxation of the contracted myocardium begins at the onset of diastole. This is a dynamic process that takes place during isovolumic relaxation (the period between aortic valve closure and mitral valve opening during which LV pressure declines with no change in volume) and then continues during auxotonic relaxation (the period between mitral valve opening and mitral valve closure during which the left ventricle fills at variable pressure) (Fig. 2-2). The rapid pressure decay and the concomitant “untwisting” and elastic recoil of the left ventricle produce a suction effect that augments the left atrial-ventricular pressure gradient and pulls blood into the ventricle, thereby promoting diastolic filling. During exercise in normal patients, relaxation rate is increased and early diastolic pressures decrease, augmenting elastic recoil and diastolic suction and resulting in more rapid filling during a shortened diastolic filling period at increasing heart rates (Fig. 2-3). During the later phases of diastole, the normal left ventricle is composed of completely relaxed cardiomyocytes and is very compliant and easily distensible, offering minimal resistance to LV filling over a normal volume range. Atrial contraction near the end of diastole contributes 20%–30% to total LV filling volume and increases diastolic pressures by less than 5 mmHg. As a result,
• EF is low (and diastolic dysfunction may coexist). • Patient may be symptomatic or asymptomatic. Pulmonary capillaries
Systolic Heart Failure: Clinical heart failure, low ejection fraction, abnormal systolic function (From Circulation 2006;113:296–304.)
Pulmonary vein
Box 2-2
Diastolic Heart Failure—Diagnostic Criteria Required Criteria 1. Clinical evidence of heart failure • Framingham or Boston criteria • Plasma brain natriuretic peptide and/or chest x-ray • Cardiopulmonary exercise testing 2. Normal ejection fraction (= 50%) Confirmatory Evidence 1. LVH or concentric remodeling 2. Left atrial enlargement (in absence of atrial fibrillation) 3. Echo Doppler or catheter evidence of diastolic dysfunction Exclusion Nonmyocardial disease
Ao
LA
LV
Figure 2-1 Elevated left ventricular end diastolic pressure causes pulmonary congestion. The heart is seen in diastole when the mitral valve open and the left ventricle (LV), left atrium (LA), and pulmonary veins form a common chamber, continuous with the pulmonary capillary bed. The left ventricular end diastolic pressure determines the pulmonary capillary pressure and the presence or absence of pulmonary congestion or edema. Ao, aorta.
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness REST
EXERCISE
20
Isovolumic pressure decline
Left atrium Systole
PLA 0
mL/sec
Left ventricle
PLV
mmHg
Aorta
150
dV/dt
Minimum PLV
E
0 500 msec
Diastole
LV volume
CHF REST
CHF EXERCISE
40
I
Isovolumic relaxation
Slow filling
Atrial filling
Auxotonic relaxation
Figure 2-2 Changes in left ventricular pressure and volume throughout the cardiac cycle. The cardiac cycle is divided into systole, the time period from mitral valve closure (MVC) to aortic valve closure (AVC), and diastole. Diastole is further divided into isovolumic relaxation (the time period from AVC to mitral valve opening [MVO] during which LV pressure declines with no change in volume) and auxotonic relaxation (the time period from MVO to MVC during which LV volume increases at variable pressure). AVO, atrial valve opening.
LV filling can normally be accomplished by very low filling pressures in the left atrium and pulmonary veins, preserving a low pulmonary capillary pressure (50%) with DHF. However, these regional abnormalities do not appear to be causally linked to either the pathophysiology of diastolic dysfunction or the development of DHF.2
As previously discussed, abnormal LV diastolic function is a universal finding in patients with DHF. Indeed, these abnormalities in diastolic function form the dominant pathophysiologic basis for the development of DHF. Even in patients with “compensated” DHF (NYHA classes II–III), abnormal relaxation, filling, and stiffness lead to increased diastolic pressures (Fig. 2-13). Further changes in diastolic function occur when patients develop decompensated DHF.28,29 For example, atrial fibrillation, tachycardia, or uncontrolled hypertension can lead to rapid increases in LA pressures and the development of decompensated DHF. Under these circumstances, the rise in pressure causes a significant change in transmitral Doppler flow pattern (as previously described). There is pseudonormalization of the ratio of ventricular to atrial filling velocities (E to A ratio) and when atrial pressures are extremely increased, to a frankly restrictive pattern. These changes in relaxation and filling are associated with changes in distensibility. During compensated DHF, the LV diastolic P-V relationship shifts upward, indicating decreased diastolic distensibility. During the initial development of decompensated DHF (during the phase of “worsening DHF”), LV volume may increase along a similar abnormal diastolic P-V curve (from point A to point B). Later, when acute pulmonary edema develops in decompensated DHF, there may be a marked upward shift in the diastolic P-V relationship (from point B to point C), indicating a further decrease in LV distensibility. Once patients with decompensated
WORSENING DHF
33 A 22 Normal 11
LV pressure (mmHg)
COMPENSATED DHF
LV pressure (mmHg)
Abnormal Diastolic Function in Decompensated Diastolic Heart Failure
0
B
33 A 22
11
0 0
20
40
60
80
100
0
20
40
60
LV volume (ml)
ACUTE PULMONARY EDEMA
POST TREATMENT
LV pressure (mmHg)
B
33
100
C
C LV pressure (mmHg)
80
LV volume (ml)
22
11
0
33 A 22
11
0 0
20
40
60
LV volume (ml)
80
100
0
20
40
60
LV volume (ml)
80
100
Figure 2-13 Left ventricular (LV) diastolic pressure-volume (P-V) relationship in compensated and decompensated diastolic heart failure (DHF). During compensated DHF, the LV diastolic P-V relationship shifts upward, indicating decreased diastolic distensibility. During the development of decompensated DHF initially (during the phase of “worsening DHF”), LV volume may increase along a similar abnormal diastolic P-V curve (from point A to point B). Later, when acute pulmonary edema develops in decompensated DHF, there may be a marked upward shift in the diastolic P-V relationship (from point B to point C), indicating a further decrease in LV distensibility. Once patients with decompensated DHF are adequately treated—for example, with diuretics and nitrates—the LV diastolic P-V relationship moves back to the compensated DHF state (from point C to point A).
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness DHF are adequately treated—for example, with diuretics and nitrates—the LV diastolic P-V relationship moves back to the compensated DHF state (from point C to point A). Decompensated DHF may be caused by both cardiovascular and noncardiovascular factors (or triggers) that act on the already existing structural and functional abnormalities (composing a substrate) to precipitate the development of acute pulmonary edema (Fig. 2-14). The substrate in patients with DHF consists of structural remodeling of the LV chamber and of the constituent cardiomyocytes and extracellular matrix that compose the chamber. These structural changes are associated with significant abnormalities in LV diastolic function, including decreased LV distensibility. These changes in LV structure and function form the substrate from which patients develop the clinical syndrome of DHF. There are a number of comorbid conditions that may act as triggers for the development of acute decompensated DHF. These include uncontrolled hypertension, increased salt and water intake, tachyarrhythmias, chronic renal failure, anemia, and coexistent lung disease.1,3,30–36 These comorbidities act upon the substrate to precipitate acute decompensated DHF. It should be noted, however, that in the absence of this substrate, these triggers do not result in DHF. For example, an increase in salt and water
intake in the absence of concentric remodeling or diastolic dysfunction does not result in DHF.
Abnormal Diastolic Function in Diastolic Heart Failure During Exercise While studies consistently show that diastolic function is abnormal in patients with DHF at rest, these abnormalities become even more exaggerated during exercise (see Chapter 17).37–39 Specifically, patients with DHF are not able to increase LV end diastolic volume, recruit Frank-Starling forces, increase relaxation rate, or increase filling rate. Consequently, exercise results in a marked increase in diastolic pressure, a limited ability to increase cardiac output, and marked truncation of exercise capacity. These abnormal responses to exercise are made worse by the exaggerated increase in arterial blood pressure that frequently accompanies exercise in patients with DHF. Abnormal diastolic function also plays a role in exercise intolerance suffered by patients with SHF, in which systolic dysfunction causes the left ventricle to lose the ability to augment diastolic filling in response to exercise by the normal mechanism of accentuated elastic recoil and early diastolic suction, previously
LV diastolic pressure (mmHg)
Normal
Diastolic heart failure
30
DHF Normal control
25 20
Decreased distensibility
15 10 5 0 0
20
40
60
80
100
120
LV diastolic volume (ml)
Substrate LV structure LV function
Triggers HTN, ↑ Na2/H2O, Tachy, CRF, ↓ HgB lung disease
Diastolic heart failure
Figure 2-14 Diastolic heart failure (DHF) substrate versus trigger. Patients with DHF have structural remodeling of the left ventricular (LV) chamber and of the constituent cardiomyocytes and extracellular matrix that compose the chamber. These structural changes are associated with significant abnormalities in LV diastolic function, including decreased LV distensibility. These changes in LV structure and function form the substrate from which patients develop the clinical syndrome of DHF. There are a number of comorbid conditions that may trigger the development of acute decompensated DHF. These include uncontrolled hypertension, increased salt and water intake, tachyarrhythmias, chronic renal failure (CRF), anemia, and coexistent lung disease. These comorbidities act upon the substrate to precipitate acute decompensated DHF. It should be noted, however, that in the absence of this substrate, these triggers do not result in DHF. For example, an increase in salt and water intake in the absence of concentric remodeling or diastolic dysfunction does not result in DHF.
21
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness described.18–20 Early diastolic filling in HF can be increased during exercise by a different mechanism: an elevation in LA pressure to create the requisite transmitral gradient rather than the normal decline in early diastolic LV pressure. The increase in LA pressure results in pulmonary congestion with exercise, a hallmark of HF (see Fig. 2-3).
sure, indicating a “paradoxical” increase in diastolic compliance.41 By contrast, during demand ischemia, diastolic compliance falls acutely.41–43 These opposite initial compliance changes with demand and supply ischemia may be explained by differences in the pressure and volume within the coronary vasculature, by the mechanical effects of the normal myocardium adjacent to the ischemic region, and by tissue metabolic factors.
CLINICAL RELEVANCE
Ischemia/Reperfusion Causing Diastolic Dysfunction
A variety of cardiac diseases can cause the development of abnormal diastolic function, abnormal cardiovascular remodeling, and the development of DHF. The mechanisms by which cardiovascular disease causes these outcomes include (but are not limited to) hemodynamic alterations, nonhomogeneous contraction and relaxation, myocardial ischemia, and LV concentric remodeling and hypertrophy. These mechanisms can act individually to alter diastolic function but often act in concert to cause the development of DHF. For ease of understanding, these mechanisms will first be discussed individually.
Ischemic Diastolic Dysfunction Ischemia can cause a reversible impairment of myocyte relaxation and diastolic function. The resultant slowing or failure of myocyte relaxation causes a fraction of actin-myosin cross-bridges to persist and continue to generate tension throughout diastole, especially in early diastole, creating a state of “partial persistent systole.” Two kinds of ischemia can alter diastolic function: demand ischemia, created by an increase in energy utilization that outstrips the necessary supply (such as during exercise or stress), and supply ischemia, created by a decrease in myocardial blood flow without a change in energy utilization (such as a myocardial infarction or coronary artery spasm). The differences between supply and demand ischemia are transient. After more sustained ischemia of 30 to 60 minutes or longer, both types result in decreased diastolic compliance.
Ischemic diastolic dysfunction can continue during and after reestablishment of normal myocardial blood flow (i.e., reperfusion). Reperfusion after a period of ischemia may result in a phase of post-ischemic diastolic “stunning,” analogous to post-ischemic stunning of contractile function. For example, diastolic dysfunction may be present early after cardiac surgery, after the myocardium has been exposed to cardioplegic arrest, or after thrombolytic or percutaneous treatment of an acute myocardial infarction. In each of these circumstances, reversible diastolic dysfunction follows reperfusion after a period of ischemia despite normal blood flow, with slow recovery to normal levels (Fig. 2-15).44–46 Post-ischemic mechanical dysfunction results in both systolic and diastolic dysfunction, and the latter may be a more sensitive parameter of ischemic injury.41 During reperfusion, LV diastolic chamber stiffness is increased.44,45 Over time, diastolic dysfunction resolves, and it is therefore reasonable to refer to this process as post-ischemic diastolic stunning. Recognition of this phenomenon is important because a reduced cardiac output or elevated PCWP in the early postoperative period or early after treatment of acute coronary syndrome may reflect an increase in LV diastolic chamber stiffness rather than a reduction in contractile function. This distinction can be made readily with echocardiography.
Demand Ischemia During demand ischemia, diastolic dysfunction may be related to myocardial ATP depletion, with a concomitant increase in ADP, resulting in rigor (rigor bond formation).29 Although ischemia is also associated with the persistence of an increased intracellular calcium concentration during diastole, one study found that ischemic diastolic dysfunction was not directly mediated by an increase in calcium concentration and a calcium-activated tension.40 As a result of the rigor, LV pressure decay, as assessed by Tau, is impaired, and the left ventricle is functionally stiffer than normal during diastole. Typically demand ischemia occurs during exercise or pharmacologically induced stress; it results from an increase in oxygen demand in the setting of limited coronary flow reserve caused by coronary stenosis and/or ventricular hypertrophy.
Supply Ischemia Supply ischemia results from a marked reduction in coronary flow. The net effect is inadequate coronary perfusion even in the resting state. Acute supply ischemia causes an initial transient downward and rightward shift of the diastolic P-V curve such that end diastolic volume increases relative to end diastolic pres-
Post CABG
20 PCWP (mmHg)
22
Pre CABG 14
9.5
10
10.5
11
11.5
LV end diastolic area
12
12.5
(cm2)
Figure 2-15 Coronary artery bypass graft (CABG) decreases left ventricular (LV) compliance. The relationship between LV end diastolic area, measured by two-dimensional echocardiography, and the pulmonary capillary wedge pressure (PCWP), which reflects LV compliance, before and after CABG. The PCWP and LV diastolic areas were increased with volume loading using isotonic saline. After CABG, the LV diastolic area was smaller at each level of PCWP, as reflected by a leftward shift of the pressure-area relationship; this indicates a reduction in LV compliance. (From McKenney P, et al: Increased left ventricular diastolic chamber stiffness immediately after coronary artery bypass surgery. J Am Coll Cardiol 1994;24:1189.)
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness
Ischemia-Induced Pulmonary Symptoms Ischemia, either spontaneous or during exercise, prevents the normal increase in LV distensibility and, as previously mentioned, can cause a rapid and marked increase in LV diastolic chamber stiffness. In the latter setting, LV diastolic pressures quickly increase, resulting in acute pulmonary congestion (“flash” pulmonary edema). This upward shift of the LV diastolic P-V curve is completely reversible with recovery of myocardial perfusion.24 The effects of ischemia explain why many patients with coronary disease have respiratory symptoms with their anginal pain, including wheezing, an inability to take a deep breath, or shortness of breath. Such respiratory symptoms may occur in the absence of anginal pain and are often referred to as anginal equivalents. These symptoms are similar to those of HF, which is not surprising, since the responsible mechanism is an elevation in pulmonary capillary pressure. One study, for example, showed that the acute decrease in LV distensibility and increase in diastolic pressure during angina caused an increase in airway resistance and a reduction in lung compliance.47 A similar symptom complex can occur in patients with concentric LV hypertrophy (LVH), even in the absence of epicardial coronary artery disease.
Left Ventricular Concentric Hypertrophy Widespread use of noninvasive methods of cardiac imaging has led to the recognition that LV diastolic dysfunction and DHF are commonly induced by the myocardial hypertrophy associated with hypertensive, coronary, or valvular heart disease. Resistance to diastolic filling is usually the result of common structural abnormalities, including concentric LV remodeling, cardiomyocyte hypertrophy, altered structure and composition of the extracellular matrix, and increased fibrillar collagen. All of these hypertrophy-associated changes lead to impaired cellular and myocardial relaxation (see Fig. 2-9). LVH and ischemia have important interactions. For a given degree of ischemia, a greater decline in diastolic function is seen in hypertrophied hearts.9,48 Hearts with concentric LVH are highly susceptible to subendocardial ischemia for several reasons49: ❒
❒
❒
❒
There is some evidence of inadequate coronary growth relative to muscle mass, with a resultant decrease in capillary density.50 The ensuing increase in capillary-to-myocyte oxygen diffusion distance renders the hypertrophied myocyte more susceptible to ischemia. The increase in ventricular wall thickness raises the epicardial-endocardial distance. Coronary arterial circulation consists of epicardial vessels that penetrate transmurally, giving rise to mid-myocardial branches that perfuse the thickened LV wall before supplying the subendocardium. Thus, coronary perfusion pressure is dissipated in proportion to LV wall thickness, leaving the subendocardium as the region most vulnerable to ischemia.49 Coronary arterial remodeling accompanies concentric hypertrophy and is manifested by an increase in coronary arterial medial thickness and perivascular fibrosis, which can restrict the extent of coronary arterial vasodilatation. Vascular tone at rest is often abnormally reduced, and coronary flow at rest is increased in the hypertrophied heart.51,52 Enhanced coronary flow is required in the resting state to supply the increased muscle mass. However, since maximal achievable coronary flow is similar to that of normal ventricles, coronary flow reserve is diminished. Endothelial
❒
❒
dysfunction also may contribute to the reduction in coronary reserve, although the response to exogenous nitric oxide is preserved.53,54 Thus, when metabolic demand and the need for oxygen increase, coronary reserve is often inadequate to meet the increased oxygen requirements, and ischemia ensues.52 Increased LV diastolic pressures can cause vascular compression, thereby reducing coronary flow and perfusion of the subendocardial layer.49 The incidence and severity of coronary atherosclerosis is increased in the presence of systemic arterial hypertension, a frequent cause of concentric LVH. Thus, patients with concentric LVH on a hypertensive basis often have significant concomitant coronary artery disease.
These factors make the heart with concentric LVH exquisitely sensitive to subendocardial ischemia. The hypertrophied ventricle also cannot relax normally in diastole with exercise. Thus, to produce the necessary increase in ventricular input, there is an increase in LA pressure, rather than the normal reduction in ventricular pressure, which produces a suction effect, as previously described. This can lead to an elevation in pulmonary capillary pressure that is sufficient to induce pulmonary congestion. Exercise-induced subendocardial ischemia can produce an “exaggerated” impairment of diastolic relaxation of the hypertrophied myocardium (see Fig. 2-15). These factors often act in concert. For a given degree of ischemia, the functional impairment in relaxation is more severe in the hypertrophied than in the nonhypertrophied heart.9,48 Thus, patients with concentric LVH secondary to chronic hypertension or aortic stenosis are particularly susceptible to ischemic diastolic dysfunction.
Genetic and Infiltrative Diseases Genetic disease processes, such as hypertrophic cardiomyopathy, and infiltrative diseases, such as amyloidosis, cause diastolic dysfunction and lead to the development of DHF. These disease processes are discussed in detail in Chapters 21 and 23.
Trigger Mechanisms of Noncardiac Factors in Acute Decompensated Diastolic Heart Failure Patients with DHF are older and have a large number of coexistent, comorbid disease processes, each of which may act on the structural and functional substrate that occurs in patients with DHF and may trigger the development of acute decompensated DHF (see Fig. 2-14). For example, present in patients with DHF may be advanced age, poorly controlled diabetes, chronic renal insufficiency, anemia, atrial fibrillation, selected drugs (e.g., glitazones, acidic nonsteroidal anti-inflammatory drugs [ANSAIDs], calcium channel blockers), abnormal sodium and water balance, chronic lung disease, increased arterial blood pressure, and increased effective arterial elastance. When present chronically, some of these factors may contribute to development of the structural and functional changes that form the substrate from which DHF develops. For example, diabetes may be associated with increases in advanced glycation endproduct–induced collagen cross-links, in collagen content, and in myocardial stiffness. The changes in the structure and function of cardiomyocytes and extracellular matrix that are common to advanced age may make the myocardium more vulnerable to
23
24
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness the effects of hypertension and coronary artery disease, making the development of DHF more frequent. In addition, once the structural and functional substrate of DHF has formed (i.e., concentric remodeling and abnormal diastolic function), the factors previously listed can act on this substrate to precipitate (or trigger) the development of acute decompensated DHF. However, in the absence of the substrate structure/function changes of DHF, these factors by themselves will not result in its development.
FUTURE RESEARCH There are many aspects of the pathophysiology of DHF and acute decompensated DHF that remain to be completely understood. In particular, the cellular and extracellular mechanisms causing the pathophysiologic changes in LV and LA structure and function remain to be completely defined. The independent effects of aging and the relationship between advancing age and the development of diastolic dysfunction and DHF must be addressed, as must also the question of whether diastolic dysfunction is an inevitable consequence of aging or can be avoided by a “successful” aging process, without these abnormalities (particularly as regards diastolic function). We must define whether and how aging plus disease processes (hypertension, diabetes, etc.) common in an aging population result in DHF. Finally, we must define the factors that cause the transition from asymptomatic diastolic dysfunction (with or without compensated LVH or concentric remodeling) to symptomatic DHF. In particular, we must define clinically applicable, noninvasive methods to detect and predict this transition. Only with an increased knowledge and understanding of the pathophysiology of DHF can effective management strategies that target this pathophysiology be developed and successfully applied to patients with DHF to decrease mortality and morbidity of this important cause of congestive heart failure. REFERENCES 1. Quiñones MA, Zile MR, Massie BM, Kass DA, for the Participants of the Dartmouth Diastole Discourses: Chronic heart failure: A report from the Dartmouth Diastole Discourses. Congest Heart Fail 2006;12:162–165. 2. Aurigemma GP, Zile MR, Gaasch WH: Contractile behavior in the left ventricle in diastolic heart failure: With emphasis on regional systolic function. Circulation 2006;113:296–304. 3. Bhatia RS, Tu JV, Lee DS, et al: Outcome of heart failure with preserved ejection fraction in a population-based study. N Engl J Med 2006; 355:260–269. 4. Redfield MM, Jacobsen SJ, Burnett JC Jr, et al: Burden of systolic and diastolic ventricular dysfunction in the community: Appreciating the scope of the heart failure epidemic. JAMA 2003;289:194–202. 5. Kitzman DW, Little WC, Brubaker PH, et al: Pathophysiological characterization of isolated diastolic heart failure in comparison to systolic heart failure. JAMA 2002;288:2144–2150. 6. Brucks S, Little WC, Chao T, et al: Contribution of left ventricular diastolic dysfunction to heart failure regardless of ejection fraction. Am J Cardiol 2005;95:603–606. 7. Zile MR, Gaasch WH, Carroll JD, et al: Heart failure with a normal ejection fraction. Is measurement of diastolic function necessary to make the diagnosis of diastolic heart failure? Circulation 2001;104:779–782. 8. Zile MR: Heart failure with preserved ejection fraction: Is this diastolic heart failure? J Am Coll Cardiol 2003;41:1519. 9. Aurigemma GP, Gaasch WH: Clinical practice. Diastolic heart failure. N Engl J Med 2004;351:1097–1105. 10. Zile MR, Baicu CF, Gaasch WH: Diastolic heart failure—abnormalities in active relaxation and passive stiffness of the left ventricle. N Engl J Med 2004;350:1953–1959.
11. Baicu CF, Zile MR, Aurigemma GP, et al: Left ventricular systolic performance, function, and contractility in patients with diastolic heart failure. Circulation 2005;111:2306–2312. 12. Zile MR, Baicu CF, Bonnema DD: Diastolic heart failure: Definitions and terminology. Prog Cardiovasc Dis 2005;47:307–313. 13. Zile MR, Brutsaert DL: New concepts in diastolic dysfunction and diastolic heart failure: Part I: Diagnosis, prognosis, and measurements of diastolic function. Circulation 2002;105:1387–1393. 14. Yusuf S, Pfeffer MA, Swedberg K, et al. and CHARM Investigators and Committees: Effects of candesartan in patients with chronic heart failure and preserved left-ventricular ejection fraction: The CHARM-Preserved Trial. Lancet 2003;362:777–781. 15. Oh JK, Hatale L, Tajik AJ, et al: Diastolic heart failure can be diagnosed by comprehensive two-dimensional and Doppler echocardiography. J Am Coll Cardiol 2006;47:500–506. 16. Nagueh SF, Mikati I, Kopelen HA, et al: Doppler estimation of left ventricular filling pressure in sinus tachycardia. A new application of tissue Doppler imaging. Circulation 1998;98:1644–1650. 17. Ahmen SH, Clark LL, Pennington WR, et al: Matrix metalloproteinases/ tissue inhibitors of metalloproteinases: Relationship between changes in proteolytic determinants of matrix composition and structural, functional, and clinical manifestations of hypertensive heart disease. Circulation 2006;113:2089–2096. 18. Cheng CP, Igarashi Y, Little WC: Mechanism of augmented rate of left ventricular filling during exercise. Circ Res 1992;70:9–19. 19. Cheng CP, Noda T, Nozawa T, et al: Effect of heart failure on the mechanism of exercise induced augmentation of mitral valve flow. Circ Res 1993;72:795–806. 20. Little WC, Cheng CP: Modulation of diastolic dysfunction in the intact heart. In Lorell BH, Grossman W (eds): Diastolic Relaxation of the Heart: The biology of diastole in health and disease, 2d ed. Boston, Kluwer Academic Publishers, 1994:167–176. 21. Little WC: Diastolic dysfunction beyond distensibility: Adverse effects of ventricular dilatation. Circulation 2005;112:2888–2890. 22. Katz AM: Physiology of the Heart. New York, Raven Press, 1992: 178. 23. Carroll JD, Hess OM, Hirzel HO, et al: Dynamics of left ventricular filling at rest and during exercise. Circulation 1983;68:59–67. 24. Carroll JD, Hess OM, Hirzel HO, et al: Exercise-induced ischemia: The influence of altered relaxation on early diastolic pressures. Circulation 1983;67:521–528. 25. Borbély A, van der Velden J, Papp Z, et al: Cardiomyocyte stiffness in diastolic heart failure. Circulation 2005;111:774–781. 26. Van Heerebeek L, Borbély A, Niessen HWM, et al: Myocardial structure and function differ in systolic and diastolic heart failure. Circulation 2006;113:1966–1973. 27. Zile MR, Brutsaert DL: New concepts in diastolic dysfunction and diastolic heart failure: Part II. Causal mechanisms and treatment. Circulation 2002;105:1503–1508. 28. Gandhi SK, Powers JC, Nomeir AM, et al: The pathogenesis of acute pulmonary edema associated with hypertension. N Engl J Med 2001;344: 17–22. 29. Vinch CS, Aurigemma GP, Hill JC, et al: Usefulness of clinical variables, echocardiography, and levels of brain natriuretic peptide and norepinephrine to distinguish systolic and diastolic causes of acute heart failure. Am J Cardiol 2003;91:1140–1143. 30. Owan TE, Hodge DO, Herges RM, et al: Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 2006;355:251–259. 31. Yancy CW, Lopatin M, Stevenson LW, et al for the ADHERE Scientific Advisory Committee and Investigators: Clinical presentation, management, and in-hospital outcomes of patients admitted with acute decompensated heart failure with preserved systolic function: A report from the Acute Decompensated Heart Failure National Registry (ADHERE) Database. J Am Coll Cardiol 2006;47:76–84. 32. Brucks S, Little WC, Chao T: Relation of anemia to diastolic heart failure and the effect on outcome. Am J Cardiol 2004;93:1055–1057. 33. Fukuta H, Sane DC, Brucks S, et al: Statin therapy may be associated with lower mortality in patients with diastolic heart failure: A preliminary report. Circulation 2005;112:357–363. 34. O’Meara E, Clayton T, McEntegart MB, et al: Clinical correlates and consequences of anemia in a broad spectrum of patients with heart failure: Results of the Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) Program. Circulation 2006;113:986–994.
Chapter 2 • Pathophysiology of Diastolic Heart Failure: Relaxation and Stiffness 35. Zile MR: Treating diastolic heart failure with statins. “Phat” chance for pleiotropic benefits (editorial). Circulation 2005;112:300–303. 36. Katz AM, Zile MR: New molecular mechanism in diastolic heart failure (editorial). Circulation 2006;113:1922–1925. 37. Little WC, Zile MR, Klein A, et al: Effect of losartan and hydrochlorothiazide on exercise tolerance in exertional hypertension and left ventricular diastolic dysfunction. Am J Cardiol 2006;98:383–385. 38. Warner JG Jr, Metzger DC, Kitzman DW, et al: Losartan improves exercise tolerance in patients with diastolic dysfunction and a hypertensive response to exercise. J Am Coll Cardiol 1999;33:1567–1572. 39. Little WC, Wesley-Farrington DJ, Hoyle J, et al: Effect of candesartan and verapamil on exercise tolerance in diastolic dysfunction. J Cardiovasc Pharmacol 2004;43:288–293. 40. Eberli FR, Stromer H, Ferrell MA, et al: Lack of direct role for calcium in ischemic diastolic dysfunction in isolated hearts. Circulation 2000;102:2643–2649. 41. Apstein CS, Grossman W: Opposite initial effects of supply and demand ischemia on left ventricular diastolic compliance: The ischemia-diastolic paradox. J Mol Cell Cardiol 1987;19:119–128. 42. Varma N, Eberli FR, Apstein CS: Increased diastolic chamber stiffness during demand ischemia: Response to quick length change differentiates rigor-activated from calcium-activated tension. Circulation 2000;101: 2185–2192. 43. Varma N, Eberli FR, Apstein CS: Left ventricular diastolic dysfunction during demand ischemia: Rigor underlies increased stiffness without calcium-mediated tension. Amelioration by glycolytic substrate. J Am Coll Cardiol 2001;37:2144–2153. 44. McKenney PA, Apstein CS, Mendes LA, et al: Increased left ventricular diastolic chamber stiffness immediately after coronary artery bypass surgery. J Am Coll Cardiol 1994;24:1189–1194.
45. McKenney PA, Apstein CS, Mendes LA, et al: Immediate effect of aortic valve replacement for aortic stenosis on left ventricular diastolic chamber stiffness. Am J Cardiol 1999;84:914–918. 46. Bolli R: Myocardial “stunning” in man. Circulation 1992;86:1671. 47. Pepine CJ, Wiener L: Relationship of anginal symptoms to lung mechanics during myocardial ischemia. Circulation 1972;46:863– 869. 48. Eberli FR, Apstein CS, Ngoy S, et al: Exacerbation of left ventricular ischemic diastolic dysfunction by pressure-overload hypertrophy. Modification by specific inhibition of cardiac angiotensin converting enzyme. Circ Res 1992;70:931–943. 49. Isayama S: Interplay of hypertrophy and myocardial ischemia. In Lorell BH, Grossman W (eds): Diastolic Relaxation of the Heart: The biology of diastole in health and disease, 2d ed. Boston, Kluwer Academic, 1994:203–212. 50. Tomanek RJ, Wessel TJ, Harrison DG: Capillary growth and geometry during long-term hypertension and myocardial hypertrophy in dogs. Am J Physiol 1991;261:H1011–H1018. 51. Eberli FR, Ritter M, Schwitter J, et al: Coronary reserve in patients with aortic valve disease before and after successful aortic valve replacement. Eur Heart J 1991;12:127–138. 52. Marcus ML, Koyanagi S, Harrison DG, et al: Abnormalities in the coronary circulation that occur as a consequence of cardiac hypertrophy. Am J Med 1983;75:62–66. 53. Ishihara K, Zile MR, Nagatsu M, et al: Coronary blood flow after the regression of pressure-overload left ventricular hypertrophy. Circ Res 1992;71:1472–1481. 54. MacCarthy PA, Shah AM: Impaired endothelium-dependent regulation of ventricular relaxation in pressure-overload cardiac hypertrophy. Circulation 2000;101:1854–1860.
25
3
DAVID H. SPODICK, MD
Role of Pericardium in Diastolic Dysfunction and Diastolic Heart Failure INTRODUCTION PATHOPHYSIOLOGY CLINICAL RELEVANCE AND FUTURE RESEARCH ABBREVIATIONS
INTRODUCTION Pericardologists1 were concerned almost exclusively with diastole (for good reasons of pericardial constriction and cardiac tamponade) before the late 1980s, when diastolic dysfunction (DD) started to attract widespread attention among other cardiologists. In many respects, acute and subacute constrictive pericarditis (CP) and cardiac tamponade epitomize DD, usually in the presence of normal systolic cardiac function. Indeed, both often have normal or high ejection fractions (EFs)—a consequence of ventricular underfilling with basically normal or compensatorily hyperfunctional myocardium.2 However, the normal pericardium also affects the cardiac filling dynamics of both normal and diseased hearts, although to a lesser degree.3
PATHOPHYSIOLOGY The normal pericardium becomes more important in dilated hearts and with increased central circulatory volume and less so with hypovolemia and normal responses to other influences that reduce cardiac size, like head-up tilt (HUT), lower body negative pressure (LBNP), and administration of such agents as nitroprusside and nitroglycerine.3 With the often low compliance of hearts with DD, the influence on filling of the normally stiff, lowcompliance pericardium may be either less than normal or additive, but this relationship has not been investigated. In the absence of formal, specifically targeted investigations of the influence of the pericardium on DD and diastolic heart failure
(DHF), the current state of knowledge permits only reasonable extensions and hypothesis generation from what we know of pericardial function and behavior during normal cardiac function and to some extent during systolic cardiac impairments. Box 3-1 summarizes the macrophysiology of the normal pericardium, many elements of which—subject to investigation—may affect DD and DHF. Box 3-2 summarizes the many cardiac effects of pericardiectomy or sufficiently extensive pericardiotomy.4 Patients who have had pericardiectomy for any indication grossly appear to function quite well, although the subject has not been intensively studied in human patients and certainly not compared for individual patients acutely and especially chronically, which would permit a better estimate of the pericardium’s dispensability. In considering the many effects of pericardiectomy and pericardiotomy sufficiently widespread to remove pericardial mechanical influence (see Box 32), the apparent benignity of pericardiectomy/otomy is at least superficially surprising because it implies that there are either widespread and adequately compensatory adjustments or that the pericardium is really not indispensable. Nonindispensability would be especially surprising when one considers the items in Box 3-1, as well as the very rich pericardial microphysiology5 (not a subject of this discussion). There is, however, broad and deep experience with experimental pericardiectomy and pericardiotomy (see Box 3-2) in hearts without cardiac disease, which, like the macro functions of the normal pericardium (see Box 3-1), should be considered in evaluating and further investigating DD and DHF. All pericardial investigations must always be considered in the light of the varied investigational protocols, which have definitely affected experimental results. Thus, it matters quantitatively and even qualitatively whether the experimental subjects are intact and conscious with a closed chest and having recovered from surgery, versus open chest, anesthetized, and/or autonomically blocked experimental subjects, as well as the species of the subjects.3 For example, dog, rabbit, and human pericardial structures differ significantly; the canine pericardium is much more anisotropic than the human pericardium.6 Moreover, within the intact pericardium, it matters quantitatively whether the pressure is 27
28
Chapter 3 • Role of Pericardium in Diastolic Dysfunction and Diastolic Heart Failure Box 3-1
Macrophysiology of the Normal Pericardium* Mechanical Functions: Promotion of Cardiac Efficiency, Especially during Hemodynamic Overloads I. Relatively inelastic cardiac envelope A. Maintenance of normal ventricular compliance (volume-elasticity relation) B. Defense of the integrity of any Starling curve: Starling mechanism operates uniformly at all intraventricular pressures because presence of pericardium 1. Maintains ventricular function curves. 2. Limits effect of increased left ventricular end diastolic pressure. 3. Supports output responses to: a. Venous inflow loads and atrioventricular valve regurgitation (particularly when acute) b. Rate fluctuations 4. Hydrostatic system (pericardium plus pericardial fluid) distributes hydrostatic forces over epicardial surfaces. a. Favors equality of transmural end diastolic pressure throughout ventricle, therefore uniform stretch of muscle fibers (preload) b. Constantly compensates for changes in gravitational and inertial forces, distributing them evenly around the heart C. Limitation of excessive acute dilation D. Protection against excessive ventriculoatrial regurgitation (atrial support) E. Ventricular interaction: relative pericardial stiffness 1. Provides a mutually restrictive chamber favoring balanced output from right and left ventricles integrated over several cardiac cycles 2. Permits either ventricle to generate greater isovolumic pressure from any volume 3. Reduces ventricular compliance with increased pressure in the opposite ventricle (e.g., limits right ventricular stroke work during increased impedance to left ventricular outflow) F. Maintenance of functionally optimal cardiac (especially left ventricular) shape
II. Provision of closed chamber with slightly subatmospheric pressure in which: A. The level of transmural cardiac pressures will be low, relative to even large increases in “filling pressures” referred to atmospheric pressure. B. Pressure changes aid atrial filling via more negative pericardial pressure during ventricular ejection. C. Diastolic suction can accelerate filling following systole. III. “Feedback” cardiocirculatory regulation via pericardial servomechanisms A. Neuroreceptors detect lung inflation and (via vagus): alter heart rate and blood pressure. B. Mechanoreceptors: Lower blood pressure and contract spleen. IV. Limitation of hypertrophy associated with chronic exercise Membranous Functions I. Reduction of external friction due to heart movements A. Production of pericardial fluid B. Generation of phospholipid surfactants II. Buttressing of thinner portions of the myocardium: Myocardial thickness varies reciprocally with parietal pericardial thickness. A. Atria B. Right ventricle III. Defensive immunologic constituents in pericardial fluid IV. Fibrinolytic activity in mesothelial lining V. Prostacyclin (PGE2, PG12 and eicosanoids) released into pericardial sac in response to stretch, hypoxia and increased myocardial loading/work VI. Synthesis and release of endothelin, increased by angiotensin III stimulation VII. Barrier to inflammation from contiguous structures Ligamentous Function I. Limits undue cardiac displacement II. Modifies pericardial stress/strain by limiting directions of traction of its fibers
*Condensed from Spoclick DH: The pericardium: A comprehensive textbook. New York, Marcel Dekker, 1997.
measured by an open-ended catheter or a flat balloon transducer, with balloons yielding higher levels of pericardial constraint (specifically radial epicardial stress), especially significant at very low effusion volumes.7 Most of the pathophysiologic effects of pericardiectomy or sufficiently extensive pericardiotomy must be considered in light of the fact that the pericardium, specifically the parietal pericardium, constrains the heart, which is immediately recognizable by retraction of the in situ pericardial tissue when it is incised.3 (The visceral pericardium may also have some constraining effect, but this is incompletely investigated.) Indeed, in 1898 Barnard incised the pericardium and observed that the heart herniated through the incision, especially in diastole and more so when he squeezed the abdomen, elevating vena cava pressure.8 Thus it is no surprise that pericardial constraint accounts for about 90% of right atrial (RA) and 80% of right ventricular (RV) cavitary pressure.9 The immediate filling pressures of the cardiac chambers are their transmural pressures (TMPs): cardiac chamber pressures minus
normally negative—therefore numerically additive—intrapericardial pressure.3 TMPs are created by having an intact pericardium and are approximately equal over both ventricles, although there may be local effects causing small local differences.10 The right ventricular end diastolic pressure-dimension (RVEDPD) relation, for example, is not flat or zero because of this, and as noted in Box 3-2, at matched left ventricular end-diastolic (LVED) volume, pericardiectomy causes a fundamental alteration in otherwise normal RV, but not left ventricular (LV), filling.11 Finally, transmural LVED pressure is a more important determinant of LV mechanoreceptor activity than absolute LVED pressure.3,12 Mechanoreceptors are sensitive to changes in ventricular stretch, determined by ventricular volume and TMP; pericardial constraint may attenuate their activity by limiting cardiac distension.3 Whereas, with an intact pericardium, the right ventricle dominates direct ventricular diastolic interaction (ventricular interdependence), after pericardiectomy or sufficiently extensive
Chapter 3 • Role of Pericardium in Diastolic Dysfunction and Diastolic Heart Failure Box 3-2
Pericardiectomy/Pericardiotomy*: Pathophysiologic Effects A. General Considerations 1. Reduced or absent constraint of the cardiac chambers 2. At matched LVEDV, pericardiectomy causes a fundamental alteration in RV but not LV filling. 3. Pericardiectomy shifts LVEDP-V curve to the right. 4. Pericardiectomy decreases RVEDP-V slope. 5. Reduced atrioventricular and ventricular interaction (i.e., parallel interaction, due to pericardium); left ventricle dominates. (With an intact pericardium, the right ventricle dominates.) 6. Decreased suction (less negative pressure) during ventricular systole B. Specific Effects 1. Decreased: a. RA mean pressure b. RA filling rate c. Pulmonary volume overload with intravascular volume loading d. Excess intravascular volume redistribution from pulmonary systemic circulation e. Decreased ventricular isovolumic pressure generation from any volume f. Decreased base to apex intraventricular pressure gradient; filling velocity shifts toward the base g. E/A and E′ h. Decreased LV mechanoreceptor activity12 2. Increased: a. Cardiac chamber transmural pressures b. RV size c. LV stroke volume, SWI and CI due to Frank-Starling response to increased preload d. LA compliance with greater increase in conduit than reservoir function e. LV compliance f. Peak dp/dt g. Early LV filling velocity (A) and filling fraction h. LV end diastolic diameter and volume i. LV early filling rate j. Ventricular series interactions (relative to direct interaction) k. LV mechanoreceptor activity l. Rate of myocardial protein synthesis producing increased LV mass m. Exercise responses i. Maximal O2 consumption ii. Maximal stroke volume and cardiac output iii. LA pressure and SV and LASV iv. LV end diastolic pressure n. Ventricular pressure-volume curves: Ventricular pressure begins its sharp rise later (at a higher cardiac volume) and increases more gradually thereafter. *Pericardiotomy = sufficiently wide incision/excision.
pericardiotomy, the left ventricle dominates the relationship; thus LV isovolumic pressure generation from any LV volume is decreased and, comparably, after pericardiectomy ventricular pressure-volume (P-V) curves show their customary sharp rise later (i.e., at a higher chamber volume) and thereafter also increase more gradually than before pericardiectomy.13
After pericardiectomy, the cardiac chambers operate at different TMPs because of the normally slightly negative intrapericardial pressure (unless, subject to investigation, the ambient pleural and pulmonary pressures of the cardiac fossa could substitute for this). With DHF, pericardial influence would also be subject to investigation to characterize cardiac filling in relation to the TMPs. Moreover, the force balance affecting cardiac chambers and their corresponding systolic performance is described by the transmural difference between intracavitary and extramural pressures.14 Indeed, the chamber collapses seen in cardiac tamponade can be attributed to sometimes even negative transmural diastolic pressures (i.e., intrapericardial pressure intermittently sufficiently higher than subjacent chamber pressures). In pericardial constriction (usually completely obliterative), there can be no truly transmural pressures.2 Ventricular interaction (ventricular interdependence) reflects the influence, particularly the filling characteristics, of the contralateral ventricle.4 Direct interaction occurs immediately via the ventricular septum, while indirect interaction (series interaction) is by way of the pericardium. For series interaction, an immediate change in one ventricle is not transmitted as directly as through the septum but is transmitted sequentially in subsequent heartbeats. Series and direct interaction influence P-V relations, pulmonary-cardiac contact pressure, and coronary engorgement. Of these, the P-V relations are the most important to producing LV diastolic pressure. Hypertrophied hearts, which are characterized by DD, especially with ventricular failure, should be a natural subject for further investigation of interaction. Draining a noncompressing hydropericardium, as occurs in heart failure, induces a state analogous to removing the normal pericardial constraint, or at least reducing it considerably, since it has been shown that any size, small to large, of even clinically nontamponading effusions significantly exaggerate respiratory effects on cardiac dynamics when measured with appropriate instruments.15 These are a quantitatively small counterpart to the effusions producing cardiac tamponade, which more strongly couple the parietal pericardium to the chamber surfaces, grossly exaggerating ventricular interaction, especially the respiratory effects due to the existence of an intact pericardium, notably pulsus paradoxus.16 A particular advantage of an intact pericardium, which could be investigated in DD and DHF, is the pericardial contribution to diastolic suction. Diastolic suction, specifically ventricular suction, occurs in early diastole and must exist in normal as well as abnormal hearts.17 The intensity of diastolic suction is proportional to the kinetic energy of the preceding systole. Moreover, diastolic suction is greatest when the heart size is least, after the end of systole, precisely when it is needed for filling. (Systolic suction may also exist in the atria during reduction of intrapericardial contents by ventricular ejection.3) Presence of the pericardium thus permits the kinetic energy acquired during systole to be applied to filling. Moreover, diastolic suction is more important at the rapid heart rates common to cardiac failure, which amputate the late to mid-diastolic diastasis interval so that there is less “passive” filling. In any case, suction must be necessary for even normal cardiac filling because, although the Frank-Starling relation implies that cardiac output should be determined by venous filling of the right heart, RA pressure is normally low, and small changes should not affect the entire heart. Indeed, changes in body position and breathing may cause larger changes in pressure.18 This is dramatically seen when excised mammalian hearts in buffered solutions continue to empty and refill, where pericar-
29
30
Chapter 3 • Role of Pericardium in Diastolic Dysfunction and Diastolic Heart Failure dial absence ensures that there can be no change in extramural pressure. Yet, diastolic suction is directly observed, indeed obvious. In such hearts, an intact pericardium would facilitate or magnify suction. In these isolated hearts, diastolic filling pressure thus does not uniquely determine fiber length and cannot entirely explain the cardiac “output.” In DD, one might predict that decreased myocardial compliance, epitomized by hypertensive hearts, would reduce macrophysiologic pericardial effects by analogy to pericardial effects on the normal right and left ventricles; the normal right ventricle is thinner and far more compliant than the left—especially with LV hypertrophy—and thus much more directly subject to pericardial constraint and to changing intrapericardial pressure. Indeed, normal RV compliance is three to four times LV compliance,19 a relationship predictably affected by DHF but subject to investigation. Another target for investigation in DD and DHF would be the effect of the pericardium in restricting acute chamber dilation (as of the left atrium following papillary muscle rupture with acute mitral regurgitation). Constraint is illustrated in reverse in those cases of “pericardial shock” with severe hypotension immediately after pericardiocentesis, usually in patients with concomitant heart disease who develop acute LV failure after pericardial drainage due to acute LV dilation.20 Indeed, some of these patients also develop acute pulmonary edema. One could assume that the pericardium and pericardial effusion fluid, especially with cardiac tamponade, had resisted ventricular dilation, following which drainage left a lax, nonconstraining, and therefore no longer resisting (constraining) pericardium. The effects of pericardiectomy could also be investigated in DHF on a purely mechanistic basis because pericardiectomy shifts the LVED P-V relationship to the right or down (Box 3-2), while DHF shifts it to the left.21 Some of the major considerations are condensed in Box 3-3. Box 3-3
Selected Major Pericardial Functions Potentially* Applicable to Investigating Diastolic Dysfunction and Diastolic Failure I. Mechanical effects due to mainly the parietal pericardium and minimal in euvolemic individuals with normal hearts A. Decreased by hypovolemia B. Exaggerated by central hypervolemia (and pericardial effusion with or without tamponade) and by cardiac dilation C. Comparative effects in DD/DHF uninvestigated II. Normally, RV and LV volumes are equally restrained by the pericardium. (Though perhaps due to regional pericardial effects, cavitary pressures fall more in RV than LV after pericardiectomy.) Effects with DD/DHF uninvestigated. III. Differences in RV and LV filling partly relate to differences in atrial compliance and ventricular distensibility. (Volume loading and pericardiectomy affect A-V pressure gradient.) Effects in DD/DHF uninvestigated. IV. Nitroglycerine and nitroprusside decrease ventricular diastolic pressures, reflecting loss of pericardial constraint when decreased venous return shrinks cardiac volume. Applications in DD/DHF uninvestigated. *Pending formal, specific investigation.
CLINICAL RELEVANCE AND FUTURE RESEARCH Pericardial pathophysiology as capsulated in Boxes 3-1 and 3-2 applies to investigations of the pericardium and its effect on the normal heart, usually the left ventricle. There are as yet no formal investigations of the role of the pericardium in patients and animal models with DD or DHF. Although one may generate hypotheses (e.g., see Box 3-3) and make reasonable projections from the foregoing data, clinical relevance can emerge only from specifically designed, controlled investigations tested in appropriate clinical contexts. For example, does the stiffness of the normal pericardium, accounting for its normal constraint,13 add to or otherwise modify the reduced compliance of the hypertensive left ventricle with DD or DHF? In some circumstances, structural pericardial modifications may be a therapeutic option. When pericardiectomy was tried some years ago for dilated cardiomyopathy, it proved to be more detrimental than helpful. The recent development, however, of a pericardial “garment” for the ventricles (e.g., the Acorn mesh) actually compensates for pericardial anisotropy since it is constructed with special fiber orientations that oppose the differing directional tensions in the underlying pericardial fibers.22 In addition, there is currently an enlarging field of intrapericardial drug and instrumental therapy for investigation in diastolically as well as systolically and rhythmically abnormal hearts.23
ABBREVIATIONS HUT: Head-up tilt LBNP: lower body negative pressure DD: diastolic dysfunction DHF: diastolic heart failure TMP: transmural pressure: cardiac chamber pressures minus normally negative—therefore additive—pericardial pressure (e.g., LVTMP = left ventricular transmural pressure, and so on) PTMP: pericardial transmural pressure: intrapericardial pressure minus pleural pressure LV: left ventricle LA: left atrium RV: right ventricle RA: right atrium EDP: end diastolic intracavitary pressure (e.g., LVEDP = left ventricular end diastolic pressure, and so on) EDPV relation: end diastolic pressure-volume relation EDPD relation: end diastolic pressure-dimension relation Ventricular interaction (interdependence): Effects of physical changes and movements in one ventricle on the other. Direct interaction: septal movement (as during breathing phases) imposed by pericardial constraint (series interaction is the effect via the vasculature and lungs of one ventricle on the other). Pericardium-dependent ventricular interactions are almost entirely diastolic, although systolic and atrioventricular interactions also exist. Suction: filling of a cardiac chamber because of negative/relatively negative pressure “leading” the direction of blood flow. Ventricular diastolic suction is especially important, especially at small/relatively small chamber/heart volumes. Most suction is facilitated by an intact pericardium. CP: constrictive pericarditis CT: cardiac tamponade EF: ejection fraction
Chapter 3 • Role of Pericardium in Diastolic Dysfunction and Diastolic Heart Failure REFERENCES 1. Seferovic PM, Spodick DH, Maisch B (eds): Pericardiology. Belgrade, Science, 1999. 2. Haffty BG, Singh JB, Spodick DH: Tracking left ventricular performance noninvasively: Response of the peak ear pulse derivative during cardiac catheterization. Chest 1983;83:543–546. 3. Spodick DH: Threshold of pericardial constraint: The pericardial reserve volume and auxiliary pericardial functions. J Am Coll Cardiol 1985;6: 296–297. 4. Spodick DH: Pericardial diseases. In Braunwald E, Zipes DP, Libby P (eds) Heart Disease, 6th ed. Philadelphia, WB Saunders, 2001:1823–1866. 5. Spodick DH: Macro and microphysiology and anatomy of the pericardium. Am Heart J 1992;124:1046–1051. 6. Janicki JS, Weber KT: The pericardium and ventricular interaction, distensility, and function. Am J Physiol 1980;238:H494–H503. 7. Tyberg JV, Misbach GA, Glantz SA, et al: A mechanism for shifts in the diastolic left ventricular pressure-volume curve: The role of the pericardium. Euro J Cardiol 1978;7:163–175. 8. Spodick DH: Acute pericarditis. New York, Grune Stratton, 1959:182. 9. Hamilton DR, Dani RS, Semlacher RA, et al: Right atrial and right ventricular transmural pressures in dogs and humans: Effects of the pericardium. Circulation 1994;90:2492–2500. 10. Spadaro J, Bing OHL, Gaasch WH, Weintraub RM: Pericardial modulation of right and left ventricular diastolic interaction. Circ Res 1981; 48:233–238. 11. Gaasch WH, Zile MR: Left ventricular diastolic dysfunction and diastolic heart failure. Annu Rev Med 2004;55:373–394. 12. Wang SY, Sheldon RS, Bergman DW, Tyberg JV: Effects of pericardial constraint on left ventricular mechanoreceptor activity in cats. Circulation 1995;92:3331–3336.
13. Spodick DH: Progress in investigation of effusion and tamponade, immunosuppression, and constriction in pericarditis and pericardial diseases. Curr Op Cardiol 1992;7:476–481. 14. Boltwood CM: Ventricular performance related to transmural filling pressure in clinical cardiac tamponade. Circulation 1987;75:941–947. 15. Wayne VS, Bishop RL, Spodick DH: Dynamic effects of nontamponading pericardial effusion: Respiratory responses in the absence of pulsus paradoxus. Br Heart J 1984;51:202–204. 16. Spodick DH: Pathophysiology of cardiac tamponade. Chest 1998;113: 1372–1378. 17. Myers RBH, Spodick DH: Constrictive pericarditis: Clinical and pathophysiologic characteristics. Am Heart J 1999;138:219–232. 18. Robinson TF, Factor SM, Sonnenblick EH: The heart as a suction pump. Scient Amer 1981;210:84–98. 19. Belenkie I, Dani R, Smith ER, Tyber JV: The importance of pericardial constraint in experimental pulmonary embolism and volume loading. Am Heart J 1992;123:733. 20. Spodick DH: The pericardium: A comprehensive textbook. New York, Dekker, 1997. 21. Zile MR, Brutsaert DL: New concepts in diastolic dysfunction and diastolic heart failure: Part I: Diagnosis, prognosis, and measurements of diastolic function. Circulation 2002;105:1387–1393. 22. Saavedra WF, Tunin RS, Paolocci N, et al: Reverse remodeling and enhanced adrenergic reserve from passive external support in experimental dilated heart failure. J Am Coll Cardiol 2002;39:2069–2076. 23. Spodick DH: Direct therapy for coronary disease, myocardial disease, and severe cardiac arrhythmias. In Spodick DH (ed): Intrapericardial therapeutics and diasgnostics (IPTD). Clin Cardiol 1999;22:I–1, I–42.
31
BRIAN D. HOIT, MD
4 Left Atrial Function: Basic Physiology INTRODUCTION PATHOPHYSIOLOGY Atrial Function in Health Left Atrial Reservoir Function Left Atrial Conduit Function Global Left Atrial Function LEFT ATRIAL FUNCTION IN DISEASE Left Atrial Function and Systolic Left Ventricular Dysfunction
Left Atrial Function and Diastolic Left Ventricular Dysfunction Left Atrial Function in a Model of Atrial Systolic Failure Importance of Left Atrial Functions and Their Interplay in Left Ventricular Systolic Dysfunction FUTURE RESEARCH ACKNOWLEDGMENT
INTRODUCTION A resurgence of interest in atrial function has enhanced our understanding of the atrial contributions to cardiovascular performance in health and disease (see Chapter 13). The reasons for this “renaissance” are multifactorial and include (1) the recognition that atrial function is an important, at times critical, determinant of left ventricular (LV) filling, (2) the increasing number of drugs, devices, ablative procedures, and surgeries available for the treatment of atrial fibrillation,1–3 (3) the considerable interest in dual- and three-chamber pacemakers that maintain atrioventricular and biventricular synchrony, respectively,4–6 (4) the pathophysiological and clinical relevance of chamber-specific structural, electrical, and ionic remodeling,7–9 (5) the clinical impact of atrial distensibility and stunning, particularly postcardioversion,10,11 and (6) the important prognostic role of atrial function in heart failure.12,13 Despite this attention, quantifying atrial function is difficult, in part because the atria are geometrically complex. Because of the obliquity of the atrial septum, the right atrium projects anteriorly, inferiorly, and to the right of the left atrium. The broad, triangular, muscular right atrial (RA) appendage protrudes anteriorly, the superior vena cava opens into the dome of the right atrium, and the inferior vena cava opens into its inferior and posterior portion. The body of the left atrium is smaller and thicker than the right atrium. The chamber has been modeled as a sphere, cube, or ellipse. The left atrial (LA) appendage is longer and narrower than the right appendage and contains all the pectinate muscles of the
left atrium. The four pulmonary veins, upper and lower from each lung (the left pair frequently opening via a common channel), enter the posterior aspect of the left atrium.14 The atrial walls consist of two muscular layers, the fascicles of which both originate and terminate at an atrioventricular ring and follow nearly perpendicular courses. Fascicles in the inner layer ascend vertically through a pectinate muscle, change depth and course circumferentially in the outer layer, encircle the atrium, dive into the inner layer, and descend vertically within a pectinate muscle. While some fascicles are intrinsic to one atrium, others are shared. The muscular terminations of the veins are also composed of two layers, the inner longitudinal and the outer circular.15 Ultrastructurally, atrial myocardium differs significantly from ventricular myocardium. For example, myocytes are smaller in diameter and have fewer T-tubules and more abundant Golgi apparatus in the atrium than in the ventricle.16 Rates of contraction and relaxation and of conduction velocity and anisotropy differ, as do their respective biophysical underpinnings (i.e., myosin isoform composition and qualitative and quantitative differences in a wide assortment of ion transporters, channels, and gap junctional proteins).17–19 While there are important differences between left and right atrial structures and functions at various organizational hierarchies, the function of the left atrium at the organ level will be used in this chapter to illustrate the atrial contributions to ventricular filling. The discussion is drawn largely from studies our group has performed over the past 15 years. 33
34
Chapter 4 • Left Atrial Function: Basic Physiology
PATHOPHYSIOLOGY
Pressure-Volume Relations of the Atrium
Atrial Function in Health
A time-independent representation of the atrial events during the cardiac cycle can be obtained by plotting instantaneous atrial pressure and volume (Fig. 4-1). During ventricular systole, atrial relaxation and descent of the ventricular base lower atrial pressure (the “x” descent) and assist in atrial filling; the latter results in a “v” wave on the atrial pressure tracing. Thus, during ventricular systole, the atrium operates as a reservoir, storing systemic and pulmonary venous return. When the atrioventricular valves open, blood stored in the atria empties into the ventricles, and atrial pressure falls (the “y” descent), during which time the atria act as conduits for venous blood flow into the ventricles. Atrial contraction, denoted by an “a” wave on the atrial pressure tracing, actively assists ventricular filling. The resultant P-V loop inscribes a “figure-eight” that consists of a clockwise “V” loop due to atrial filling and passive emptying, and a counterclockwise “A” loop due to active atrial contraction. Although Alexander et al. described instantaneous LA P-V relations by a time-varying elastance in the isolated left atrium using computer-simulated LA loading conditions,25 assessment of atrial systolic elastance in vivo was hampered by the lack of an accurate measurement of LA volume with an adequate sampling frequency. Therefore, as a critical initial step, we demonstrated that cast-validated LA volumes could be estimated accurately with high temporal resolution sonomicrometry using two nearly orthogonal atrial dimensions.26,27 The left atrium was assumed to be a general ellipsoid of revolution:
Left Atrial Booster Pump Function
LA pressure (mmHg)
34
v
where SAX is the short or mediolateral axis, and LAX is the long or anteroposterior axis. Coupled with high-fidelity micromanometers, instantaneous P-V loops of the intact left atrium were generated (Fig. 4-2). The maximum slope of the isochronal P-V relation was measured every 10 msec during the “A” loop (from
a
0
MVO
16.0 28
20.5
16.5 ECG
LAed A LAes MVC
V
8.0 6.0
4.0
A BC D 1 sec.
A
LA volume = π/6(SAX)2(LAX),
LA pressure (mmHg)
LA short axis (mm)
40
LA long axis (mm)
The principal role of the left atrium is to modulate LV filling and cardiovascular performance through the interplay of atrial reservoir, conduit, and booster pump functions. Typically, the importance of the atrial booster pump function (i.e., the augmented ventricular filling resulting from active atrial contraction) has been estimated by measurements of (1) cardiac output and LV diastolic volume both with and without effective atrial systole,20 (2) relative LV filling (e.g., early to late [E/A] filling ratios) using steady-state Doppler echocardiographic transmitral flow or radionuclide angiography,21 and (3) atrial shortening using methods such as two-dimensional echocardiography, angiography, and sonomicrometry.22 Booster pump function is also evaluated echocardiographically by estimating the kinetic energy and force generated with atrial contraction.23,24 However, measurements of atrial systolic function and the importance of the atrial booster pump are dependent on a multiplicity of factors, including the timing of atrial systole, vagal stimulation, the magnitude of venous return (i.e., atrial preload), LV end diastolic pressures (i.e., atrial afterload), and LV systolic reserve. Not surprisingly, despite considerable study, the magnitude and relative importance of the atrial contribution to LV filling and cardiac output remain controversial. Analogous to end systolic elastance measurements in the left ventricle (where end systolic elastance is calculated as the slope of the ventricular end systolic pressure-volume [P-V] relation), a load-independent index of atrial contraction based on the instantaneous atrial P-V relation has the potential to explain and minimize the discrepancies and confusion that exist in the literature. Accordingly, understanding and deriving atrial elastance require a consideration of the relation between instantaneous atrial pressure and volume.
LA volume (ml)
B
Figure 4-1 A, Analog recording of left atrial (LA) pressure and dimensions in the time domain. The vertical lines indicate times of mitral valve opening (A), end of passive atrial emptying and onset of atrial diastasis (B), atrial end diastole (C), and atrial end systole (D). “a” and “v” represent respective venous pressure waves. B, LA pressure-volume loop from a single beat illustrating the characteristic figure-eight configuration. Arrows indicate direction of loop as a function of time. “A” loop represents active atrial contraction. “V” loop represents passive filling and emptying of the left atrium. MVO, time of mitral valve opening; LAed, left atrial end diastole; Laes, left atrial end systole. (From Hoit BD et al: Circulation 1994;89:1829–1838.)
Chapter 4 • Left Atrial Function: Basic Physiology BASELINE
POST CALCIUM Slope = 3.6
Slope = 5.4
26.6
LA pressure (mmHg)
LA pressure (mmHg)
17.7
35
6.9
6.7 5.7
8.9
5.0
LA volume (ml)
10.1 LA volume (ml)
A Slope 3.6
Slope = 5.4 23
LA pressure (mmHg)
Figure 4-2 A, Five left atrial pressure-volume loops at baseline (left) and after calcium infusion (right). Note the linearity of the end systolic pressure-volume relations, the greater extent of atrial systolic shortening, and the increase in the end systolic pressure-volume slope after calcium infusion (the latter indicating a positive inotropic effect). B, Loops are computersmoothed for clarity. (From Hoit BD et al: Circulation 1994;89:1829–1838.)
LA pressure (mmHg)
18
7
7 6
9 LA volume (ml)
5
10 LA volume (ml)
B
atrial diastasis to atrial end systole) from five variably loaded beats and were fitted to a time-varying elastance model: E(t) = P(t)/[V(t) − V(o)], where E(t) is time-varying elastance, V(o) is the volume axis intercept, and P(t) and V(t) are instantaneous isochronal pressure and volume, respectively. Loading conditions were altered with either phenylephrine boluses or vena caval occlusions. Time-dependent changes in E(t) and maximal atrial systolic elastance (Emax) were found to be highly linear and sensitive to pharmacologically induced changes in inotropic state. In addition, LA systolic P-V relations using either the non-isochronal maximal P-V ratio (Emax P/V) or end systole (Ees) were shown to be useful estimates of Emax. In addition to assisting the quantitation of ejection phase (i.e., ejection fraction, stroke volume, mean normalized systolic ejection rate [MNSER], and velocity of circumferential shortening [Vcf ]) and load-independent indices of contraction (i.e., Emax, Ees, and Emax P/V), P-V loops of the left atrium offered possibilities for study of atrial energetics in vivo. Specifically, the planimetered “A” loop P-V area (total mechanical work done by the atrium) minus the “V” loop P-V area (work done on the atrium by pulmonary venous blood flow during ventricular systole) represents the net
atrial mechanical work. In a study using normal, open-chest dogs, A- and V-loop areas were similar and did not change with increased atrial preload, suggesting that booster pump and reservoir functions of the atrium were balanced over a wide range of LA pressures. However, calcium-induced increases in contractility and subsequent volume loading produced significant increases in the A- but not the V-loop area, indicating that net atrial mechanical work was greater and load dependent after calcium infusion. Calcium infusion was also associated with a change in the trajectory of the A loop resulting in a decrease in the time from Emax to end systole; thus the atrial A loop, which resembled a right ventricular (RV) P-V loop before, looked like the LV P-V loop after calcium infusion, suggesting an effect of calcium on ventricular input impedance. While atrial P-V loops can be generated in humans using invasive and semi-invasive means,28–31 these methods are cumbersome, time consuming, and difficult to apply. Clearly, there is a need for an objective, noninvasive measurement of atrial myocardial performance and contractility. Measurement of myocardial strain and strain rate, which represent the magnitude and rate of myocardial deformation, are indices that have the potential to overcome these limitations.32
Chapter 4 • Left Atrial Function: Basic Physiology
Left Atrial Reservoir Function Grant et al. estimated that 42% of the LV stroke volume and its associated energy are stored in the left atrium during LV systole.33 The subsequent dissipation of this energy during the reservoir phase acts as a ventricular restorative force during the ensuing LV diastole. Reservoir function is governed by atrial distensibility during ventricular systole (although LA reservoir function has been shown to be related to LA contraction and descent of the LV base during systole34 and to LA systolic shortening and LV end systolic volume35), which is measured rigorously by fitting atrial pressures and dimensions—taken either at the time of mitral valve opening over a range of atrial pressures and volumes or from one of the limbs of the “V” loop in a static P-V loop—to an exponential equation.36,37 Although atrial dimensions and pressures are required for its measurement, the relative reservoir function can be estimated simply with pulmonary vein Doppler (Fig. 4-3). Thus, the proportion of LA inflow during ventricular systole provides an index of the reservoir capacity of the atrium.38 In this regard, we showed that LA compliance is an important independent determinant of the pattern of pulmonary venous flow, using an experimental protocol that excluded the LA appendage and produced an isolated decrease in atrial compliance and relative reservoir-to-conduit flow.22 As a consequence, alterations in atrial compliance in various disease stages should be considered when the pattern of pulmonary venous flow is used to estimate LA pressure, assess LV diastolic function, or quantify mitral regurgitation. Suga showed that atrial compliance was a significant determinant of cardiac performance in a circulatory analog model.39 Specifically, decreased atrial compliance was associated with greater phasic and mean atrial pressures but a lower mean level of atrial pressure during ventricular filling; this resulted in a smaller LV end diastolic volume and decreased venous return. We confirmed earlier mathematical predictions that early LV filling (i.e., Doppler
mitral E velocity) increases directly with operative atrial compliance.40 Thus it is interesting to speculate that regions of increased distensibility in the left atrium may facilitate early diastolic filling of the left ventricle. One such region of the left atrium with increased distensibility is the atrial appendage. In isolated canine atria, the slope of the P-V relation for the left atrium without the appendage is significantly greater than with the appendage intact.41 We subsequently showed that in the intact dog, the pressure-strain relation is steeper (i.e., more stiff or less distensible) in the body than in the appendage of the left atrium (Fig. 4-4). Atrial systolic shortening and reservoir capacity (measured as the change in dimension during atrial systole, and the maximum minus minimum dimension, respectively) increased in both the body and the appendage during volume infusion and were greater in the appendage than in the body of the left atrium. Thus, for a given increase in LA pressure, there was a greater increase in the end diastolic atrial dimension, with greater utilization of the Frank-Starling mechanism by the appendage than by the body of the left atrium. Moreover, biochemical studies indicate that atrial natriuretic factor is concentrated in the LA appendage,42 which may, by virtue of its increased distensibility, be better suited for the regulation of intravascular volume. In addition, greater distensibility of the appendage than the body of the left atrium would be potentially beneficial in the context of increased LV filling pressures and decreased global atrial distensibility. It is interesting that relative blood flow to the LA appendage was increased with experimental LA hypertrophy,43 and the appendage-to-nonappendage LA blood flow ratio was increased during severe exercise in dogs, further suggesting that the LA appendage may become functionally important during stress. These data suggest that the substantial contributions of the appendage to overall LA compliance may have important negative implications for routine atrial appendectomy at the time of mitral valve surgery and for the use of percutaneous appendage exclusion devices.44,45
1 sec
J
K
PV Doppler (cm/sec)
40
KVTI
200
0 0
1 sec 40 J
JVTI 0
ECG
K
KVTI
PV flow (ml/min)
JVTI
PV Doppler (cm/sec)
36
Figure 4-3 Representative pulmonary vein (PV) Doppler with flow-probe waveform superimposed (top panel) and a schematic of the pulmonary venous flow velocity (bottom panel). J and K are peak systolic and diastolic velocities; Jvti and Kvti are systolic and diastolic integrals, respectively. The ratio of systolic-to-diastolic velocities (and integrals) represents the relative reservoir-to-conduit function of the left atrium. (From Hoit BD et al: Circulation 1992;86:651–659.)
Chapter 4 • Left Atrial Function: Basic Physiology 20
LA APP
Mean LA pressure (mmHg)
Mean LA pressure (mmHg)
20
10
LA APP
18
*
16 14
*
12 10 8
* 6
0 0.0
0.1
0.2
0.3
0.4
0.5
0.0
0.6
0.2
0.3
0.4
0.5
Natural strain (ln[D/Dmin])
Natural strain (ln[D/Dmin])
A
0.1
B
Figure 4-4 A, Representative mean left atrial (LA) pressure–natural strain relationships for the LA body and appendage from a single dog. Data are fitted to an exponential function. Pressure-strain relationship of appendage is shifted to the right of the body. B, Mean pressure–natural strain data for 12 dogs. Error bars represent SE. Data indicate that the appendage is more distensible than the body of the left atrium. (From Hoit BD, Walsh RA, Am J Physiol 1992;262: H1356–H1360.)
Myocardial strains can be obtained noninvasively using either tissue Doppler imaging (TDI) or 2D speckle imaging. Recently, atrial strain and strain rates (representing the magnitude and rate of myocardial deformation, respectively) during ventricular systole predicted the 9-month atrial fibrillation recurrence rate in patients with lone atrial fibrillation after a successful cardioversion.11 Coupled with an estimate of atrial pressure,46 strain has the potential to estimate atrial distensibility noninvasively.
Left Atrial Conduit Function Atrial conduit function occurs primarily, but not exclusively, during ventricular diastole and represents the volume of blood that passes through the left atrium that cannot be attributed to reservoir or booster pump functions; in a well-characterized model, atrial conduit function accounted for 35 ± 8% of all flow through the atrium.47 A reciprocal relationship exists between conduit and reservoir functions of the left atrium. Indeed, the redistribution of atrial functions was reported as an important compensatory mechanism that facilitates LV filling in patients with myocardial ischemia, acute myocardial infarction, hypertensive heart disease, and mitral stenosis.48–51 Although it has been postulated that a more distensible atrium reduces atrial conduit flow, we showed that while pericardiectomy decreased the atrial stiffness constant (increased distensibility), the relative conduit function, measured as the ratio of flow probedetermined systolic-to-diastolic pulmonary vein flow, actually increased.38 However, it should be noted that the LA reservoir volume (max − min LA volume) and absolute systolic pulmonary venous flow also increased post pericardiectomy. In addition, removal of the pericardium was associated with an increase in the peak early velocity and deceleration time of transmitral diastolic flow, consistent with the increase in LV compliance after pericardiectomy. Thus it is possible that the decreased systolic to diastolic P-V ratio reflects a greater change in compliance of the left ventricle than of the atrium. Irrespective of the mechanisms, these data underscore the complexity of pericardial influences on atrial and ventricular filling and the multifactorial determinants of P-V relations.
Global Left Atrial Function Although it is clear that atrial reservoir, conduit, and booster pump functions (and their modulation by the pericardium) play important roles in maintaining ventricular filling, a global measure (i.e., throughout the entire atrial cycle) of atrial function is lacking. A potential candidate property suitable for quantitation is synchrony. Therefore, we correlated time-tissue Doppler velocity curves from the lateral left and right atria and the lateral left atrium and interatrial septum over three cycles every 10 msec to obtain indices of inter- and intra-atrial phase heterogeneity, respectively. The intra-atrial phase heterogeneity index was 0.90 ± 0.06, indicating that the time-velocity curves over the entire cardiac cycle for the septum and lateral LA segments were highly correlated (i.e., in phase). In contrast, the inter-atrial phase index of 0.66 ± 0.10 indicated a greater degree of heterogeneity between the lateral LA and RA segments. These indices may be particularly useful to understand the effect of altered atrial conduction on atrial function and ventricular filling in various pathological states.
LEFT ATRIAL FUNCTION IN DISEASE Although LV adaptation to chronic hemodynamic loading has been studied extensively, the mechanisms by which the left atrium compensates for sustained increases in pressure and volume are less clear. We considered that adaptation may be due to the addition of contractile units (hypertrophy), an alteration in the molecular composition of the contractile machinery (e.g., protein isoform transitions), or a reduction of atrial systolic stress by structural remodeling. In pathological studies of human atria, increases in the amount of the slow, β-myosin isoform were correlated with the severity and duration of pressure and volume overload. In this context, a switch from the fast (α) to the slow (β) myosin heavy chain (MHC) isoform in the ventricles of small mammals is accompanied by decreased myosin ATPase activity and velocity of contraction, reduced oxygen consumption, and a greater efficiency of contraction. Thus, these isoform switches are likely to represent adaptations to the functional requirements
37
38
Chapter 4 • Left Atrial Function: Basic Physiology imposed by sustained hemodynamic loads in a chamber such as the left atrium, which contains predominantly α MHC.
Left Atrial Function and Systolic Left Ventricular Dysfunction Experimental studies that examine atrial adaptation use models of mitral regurgitation,36 myocardial infarction,12 LV diastolic dysfunction,52 and pacing-induced myocardial failure (ventricular and atrial tachypacing53,54). In the ventricular tachypacing model of heart failure, we found a significant upregulation of β MHC in the LA body that was associated with LA hypertrophy, increased atrial mechanical work (A-loop area, Fig. 4-5), maintained force generation (Emax and Ees), and decreased velocity of LA contraction. Another important finding of our study was that atrial MHC isoform switches vary by region. Despite hypertrophy of the appendage, the percent of β MHC did not increase in comparison with controls, suggesting that adaptation to hemodynamic overload includes differentially determined quantitative (e.g., hypertrophy) and qualitative (e.g., isoform switch) components and that there are fundamental differences in the atrial response to stress hypertrophy, which occurs in the body, versus “strain” hypertrophy, which occurs in the appendage.
Left Atrial Function and Diastolic Left Ventricular Dysfunction In carefully studied animal models, the improvement in systolic function that occurs after cessation of tachypacing is associated with residual abnormalities of LV isovolumic relaxation and chamber compliance, and in some instances, the development of LV hypertrophy. Since LV diastolic dysfunction represents a source of excess afterload on the left atrium, we examined the gravimetric, myosin biochemical, and functional changes in the left atrium that accompany regression of pacing-induced heart failure. Animals were rapidly paced for 3 weeks to produce heart failure, and hemodynamic and functional studies were performed 3 weeks after pacing cessation.52 We showed that improvement of LV systolic dysfunction was associated with persistent LV
24.0
Left Atrial Function in a Model of Atrial Systolic Failure We examined global and regional LA pump functions after one week of isolated atrial flutter (atrial pacing at 400 bpm) in closedchest, sedated dogs.55 LV function, as assessed by the echocardiographic LV area shortening fraction, was preserved. Global and regional atrial pump functions were determined by mitral and LA appendage waveform analysis, and relative conduit and reservoir functions by pulmonary venous waveform analysis with Doppler velocimetry. We found that after one week of rapid atrial pacing, global and regional atrial booster pump functions (i.e., body and appendage) were impaired, and relative reservoir-to-conduit function of the left atrium was reduced (Fig. 4-6). Similar changes in the absence of atrial hypertrophy were noted in a highly instrumented preparation treated with similar atrial tachypacing for 6 weeks.
Importance of Left Atrial Functions and Their Interplay in Left Ventricular Systolic Dysfunction The redistribution of atrial reservoir and conduit functions and an augmented booster pump function are important compensatory mechanisms that facilitate LV filling in patients with myocardial ischemia, acute myocardial infarction, and hypertensive heart disease. However, considerably less is known about how these mechanisms operate in the setting of LV dysfunction. Therefore, we tested the hypothesis that in contrast to dogs with a normal left ventricle, cardiac output falls with the loss of atrial
Left atrial pressure (mmHg)
Left atrial pressure (mmHg)
22.4
V
V
A
A
10.0
9.5 5.5
8.4
17.4
Left atrial volume (ml)
A
diastolic dysfunction and normalization of both atrial Ees and net atrial work despite the persistent abnormalities of loaddependent indices of LA systolic function (e.g., ejection fraction, systolic ejection rate). LA stiffness and reservoir volumes also remained abnormal. Changes in systolic and diastolic function of the left atrium were accompanied by persistent upregulation of the atrial β MHC isoform and incomplete regression of LA hypertrophy, indicating an apparent dissociation of myosin isoform shifts, myocardial hypertrophy, and atrial work.
21.4 Left atrial volume (ml)
B
Figure 4-5 Left atrial pressure-volume loops from three variably loaded beats in a control dog (A) and in a dog with pacing-induced heart failure (B). The A loop represents active atrial contraction; the V loop represents passive filling and emptying of the left atrium. The increase in A loop area after pacing indicates increased atrial systolic work. Loops are computer-smoothed for clarity. (From Hoit BD et al: Cardiovasc Res 1995;29:469–474.)
Chapter 4 • Left Atrial Function: Basic Physiology Baseline
1 week rap
MV velocity
A A E 80 cm/sec
E
e
a e
40 cm/sec
40 cm/sec a
LAA velocity
Figure 4-6 Doppler waveforms of transmitral (MV, top panel), left atrial appendage (LAA, middle panel), and pulmonary vein (PV, bottom panel) flows at baseline and after one week of rapid atrial pacing (RAP). E, early transmitral diastolic velocity; A, late transmitral diastolic velocity; e, early LAA emptying velocity; a, late LAA emptying velocity; J1 and J2, systolic velocities; K, diastolic PV velocities. The corresponding electrocardiogram is shown in each panel. (From Hoit BD et al: J Am Soc Echo 1997;10:805–810.)
80 cm/sec
J2
40 cm/sec J1
K 40 cm/sec J2
PV velocity
systolic contraction in dogs with concomitant LV dysfunction because atrial compensatory mechanisms fail. Chronic atrial and ventricular dysfunction (both singly and in combination) were modeled with rapid pacing in dogs with radiofrequency atrioventricular nodal ablation and implantation of RA and RV pacing catheters.56 Serial echocardiographic analyses of global and regional LA pump function were estimated from mitral and LA appendage Doppler velocimetry, and relative reservoir and conduit functions of the left atrium were determined from pulmonary venous Doppler. Right heart catheterization studies were performed simultaneously to assess right heart pressures and cardiac output. Isolated atrial systolic failure was produced by 1 week of rapid atrial pacing (400 bpm), and moderate
K
J1
LV dysfunction was produced by 2 weeks of rapid RV pacing (220 bpm). Two weeks of rapid ventricular pacing superimposed on rapid atrial pacing during the second week produced combined LA and LV dysfunction. As previously noted, isolated LV dysfunction increased atrial booster pump and reservoir functions, whereas isolated rapid atrial pacing decreased atrial booster pump and increased the relative conduit function of the left atrium. In contrast, the decreased atrial booster pump function in animals with combined atrial and ventricular dysfunction was incompletely compensated by the redistribution of the reservoir and conduit functions of the left atrium. As a result, cardiac output decreased and right heart pressures increased only after superimposed atrial and ventricular pacing.
39
40
Chapter 4 • Left Atrial Function: Basic Physiology
FUTURE RESEARCH While knowledge of atrial function lags considerably behind that of the ventricle, the chasm is rapidly closing. The studies discussed herein describe the role of atrial function to ventricular filling in normal physiology and provide a framework for understanding mechanical, energetic, and biochemical mechanisms responsible for atrial adaptation to chronic hemodynamic loading. Important future directions include distinguishing the role of atrial function in early versus late LV systolic dysfunction, investigating the role of matrix metalloproteinases (and other molecules) in structural remodeling of the atria, characterizing the biochemical and molecular biological changes that accompany atrial adaptation and failure, and identifying the relation between structural and electrical remodeling of the left atrium. These goals are likely to be met with the development of novel techniques to evaluate atrial reservoir, conduit, and booster pump functions of the atrium.
ACKNOWLEDGMENT This work was supported in part by a Grant in Aid award from the American Heart Association Ohio Valley Affiliate (0355198B). REFERENCES 1. Hersi A, Wyse DG: Management of atrial fibrillation. Curr Probl Cardiol 2005;30:175–233. 2. Marine JE, Dong J, Calkins H: Catheter ablation therapy for atrial fibrillation. Prog Cardiovasc Dis 2005;48:178–192. 3. Gillinov AM, Wolf RK: Surgical ablation of atrial fibrillation. Prog Cardiovasc Dis 2005;48:169–177. 4. Frielingsdorf J, Gerber AE, Hess OM: Importance of maintained atrioventricular synchrony in patients with pacemakers. Eur Heart J 1994;15:1431–1440. 5. McComb JM, Gribbin GM: Effect of pacing mode on morbidity and mortality: Update of clinical pacing trials. Am J Cardiol 1999;83:211D–213D. 6. Vural A, Agacdiken A, Ural D, et al: Effect of cardiac resynchronization therapy on left atrial appendage function and pulmonary venous flow pattern. Int J Cardiol 2005;102:103–109. 7. Nattel S, Shiroshita-Takeshita A, Cardin S, Pelletier P: Mechanisms of atrial remodeling and clinical relevance. Curr Opin Cardiol 2005;20: 21–25. 8. Yue L, Melnyk P, Gaspo R, et al: Molecular mechanisms underlying ionic remodeling in a dog model of atrial fibrillation. Circ Res 1999;84:776–784. 9. Wijffels MC, Kirchhof CJ, Dorland R, et al: Electrical remodeling due to atrial fibrillation in chronically instrumented conscious goats: Roles of neurohumoral changes, ischemia, atrial stretch, and high rate of electrical activation. Circulation 1997;96:3710–3720. 10. Grimm RA, Leung DY, Black IW, Thomas JD: Left atrial appendage “stunning” after spontaneous conversion of atrial fibrillation demonstrated by transesophageal Doppler echocardiography. Am Heart J 1995;130:174–176. 11. Di Salvo G, Caso P, Lo Piccolo R, et al: Atrial myocardial deformation properties predict maintenance of sinus rhythm after external cardioversion of recent-onset lone atrial fibrillation: A color Doppler myocardial imaging and transthoracic and transesophageal echocardiographic study. Circulation 2005;112:387–395. 12. Kono T, Sabbah HN, Rosman H, et al: Left atrial contribution to ventricular filling during the course of evolving heart failure. Circulation 1992;86:1317–1322. 13. Bruch C, Gotzmann M, Sindermann J, et al: Prognostic value of a restrictive mitral filling pattern in patients with systolic heart failure and an implantable cardioverter-defibrillator. Am J Cardiol 2006;97:676–680. 14. Strandring S, ed: Gray’s Anatomy, 39th ed. London, Elsevier, 2005.
15. Thomas CE: The muscular architecture of the atria of hog and dog hearts. Am J Anat 1959;104:207–236. 16. McNutt NS, Fawcett DW: The ultrastructure of the cat myocardium. II. Atrial muscle. J Cell Biol 1969;42:46–67. 17. Thomas C, Coker B, Zellner J, et al: Increased matrix metalloproteinase activity and selective upregulation in LV myocardium from patients with end-stage dilated cardiomyopathy. Circulation 1998;97:1708–1715. 18. Li GR, Lau CP, Shrier A: Heterogeneity of sodium current in atrial vs epicardial ventricular myocytes of adult guinea pig hearts. J Mol Cell Cardiol 2002;34:1185–1194. 19. Tuteja D, Xu D, Timofeyev V, et al: Differential expression of smallconductance Ca2+-activated K+ channels SK1, SK2, and SK3 in mouse atrial and ventricular myocytes. Am J Physiol Heart Circ Physiol 2005;289: H2714–H2723. 20. Mitchell JH, Gupta DN, Payne RM: Influence of atrial systole on effective ventricular stroke volume. Circ Res. 1965;17:11–18. 21. Hoit BD, Rashwan M, Verba J, et al: Instantaneous transmitral flow using Doppler and M-mode echocardiography: Comparison with radionuclide ventriculography. Am Heart J 1989;118:308–314. 22. Hoit BD, Shao Y, Tsai LM, et al: Altered left atrial compliance after atrial appendectomy. Influence on left atrial and ventricular filling. Circ Res 1993;72:167–175. 23. Stefanadis C, Dernellis J, Lambrou S, Toutouzas P: Left atrial energy in normal subjects, in patients with symptomatic mitral stenosis, and in patients with advanced heart failure. Am J Cardiol 1998;82:1220– 1223. 24. Manning WJ, Silverman DI, Katz SE, Douglas PS: Atrial ejection force: A noninvasive assessment of atrial systolic function. J Am Coll Cardiol 1993;22:221–225. 25. Alexander J, Sunagawa K, Chang N, Sagawa K: Instantaneous pressurevolume relation of the ejecting canine left atrium. Circ Res 1987;61: 209–219. 26. Hoit BD, Shao Y, McMannis K, et al: Determination of left atrial volume using sonomicrometry: A cast validation study. Am J Physiol 1993;264: H1011–H1016. 27. Hoit BD, Shao Y, Gabel M, Walsh RA: In vivo assessment of left atrial contractile performance in normal and pathological conditions using a timevarying elastance model. Circulation 1994;89:1829–1838. 28. Stefanadis C, Dernellis J, Toutouzas P: Evaluation of the left atrial performance using acoustic quantification. Echocardiography 1999;16: 117–125. 29. Stefanadis C, Dernellis J, Stratos C, et al: Assessment of left atrial pressurearea relation in humans by means of retrograde left atrial catheterization and echocardiographic automatic boundary detection: Effects of dobutamine. J Am Coll Cardiol 1998;31:426–436. 30. Dernellis JM, Stefanadis CI, Zacharoulis AA, Toutouzas PK: Left atrial mechanical adaptation to long-standing hemodynamic loads based on pressure-volume relations. Am J Cardiol 1998;81:1138–1143. 31. Tse HF, Hettrick DA, Mehra R, Lau CP: Improved atrial mechanical efficiency during alternate- and multiple-site atrial pacing compared with conventional right atrial appendage pacing: Implications for selective site pacing to prevent atrial fibrillation. J Am Coll Cardiol 2006;47:209–212. 32. Pislaru C, Abraham TP, Belohlavek M: Strain and strain rate echocardiography. Curr Opin Cardiol 2002;17:443–454. 33. Grant C, Bunnell IL, Green DG: The reservoir function of the left atrium during ventricular systole. Am J Med 1964;37:36–43. 34. Barbier P, Solomon SB, Schiller NB, Glantz SA: Left atrial relaxation and left ventricular systolic function determine left atrial reservoir function. Circulation 1999;100:427–436. 35. Hoit BD, Shao Y, Gabel M, Walsh RA: Influence of loading conditions and contractile state on pulmonary venous flow. Validation of Doppler velocimetry. Circulation 1992;86:651–659. 36. Kihara Y, Sasayama S, Miyazaki S: Role of the left atrium in adaptation of the heart to chronic mitral regurgitation in conscious dogs. Circ Res 1988;62:543–553. 37. Hoit BD, Walsh RA: Regional atrial distensibility. Am J Physiol 1992;262: H1356–H1360. 38. Hoit BD, Shao Y, Gabel M, Walsh RA: Influence of pericardium on left atrial compliance and pulmonary venous flow. Am J Physiol 1993;264: H1781–H1787. 39. Suga H: Importance of atrial compliance in cardiac performance. Circ Res 1974;35:39–43. 40. Hoit BD, LeWinter M, Lew WY: Independent influence of left atrial pressure on regional peak lengthening rates. Am J Physiol 1990;259: H480–H487.
Chapter 4 • Left Atrial Function: Basic Physiology 41. Davis CA, Rembert JC, Greenfield JC: Compliance of the left atrium with and without left atrium appendage. Am J Physiol 1990;259: H1006–H1008. 42. Hintze TH, McIntyre JJ, Patel MB, et al: Atrial wall function and plasma atriopeptin during volume expansion in conscious dogs. Am J Physiol 1989;256:H713–H719. 43. Bauman RP, Rembert JC, Greenfield JC: Regional atrial blood flow in dogs. J Clin Invest 1989;83:1563–1569. 44. Kamohara K, Fukamachi K, Ootaki Y, et al: A novel device for left atrial appendage exclusion. J Thorac Cardiovasc Surg 2005;130:1639–1644. 45. Ostermayer SH, Reisman M, Kramer PH, et al: Percutaneous left atrial appendage transcatheter occlusion (PLAATO system) to prevent stroke in high-risk patients with non-rheumatic atrial fibrillation: Results from the international multi-center feasibility trials. J Am Coll Cardiol 2005;46:9–14. 46. Nagueh SF: Noninvasive evaluation of hemodynamics by Doppler echocardiography. Curr Opin Cardiol 1999;14:217–224. 47. Hitch DC, Nolan SP: Descriptive analysis of instantaneous left atrial volume—with specific reference to left atrial function. J. Surg Res 1981;30:110–120. 48. Sigwart U, Grbic M, Goy J, Kappenberger L: Left atrial function in acute transient left ventricular ischemia produced during percutaneous transluminal coronary angioplasty of the left anterior descending coronary artery. Am J Cardiol 1990;65:282–286.
49. Matsuda Y, Toma Y, Moritani K, et al: Assessment of left atrial function in patients with hypertensive heart disease. Hypertension 1986;8:779– 785. 50. Matsuda Y, Toma Y, Ogawa H: Importance of left atrial function in patients with myocardial infarction. Circulation 1983;65:566–571. 51. Triposkiadis F, Wooley CF, Boudoulas H: Mitral stenosis: Left atrial dynamics reflect altered passive and active emptying. Am Heart J 1990;120:124–132. 52. Hoit BD, Shao Y, Gabel M, et al: Left atrial systolic and diastolic function after cessation of pacing in tachycardia-induced heart failure. Am J Physiol 1997;273:H921–H927. 53. Hoit BD, Shao Y, Gabel M: Left atrial systolic and diastolic function accompanying chronic rapid pacing–induced atrial failure. Am J Physiol 1998;275: H183–H189. 54. Hoit BD, Shao Y, Gabel M, Walsh RA: Left atrial mechanical and biochemical adaptation to pacing induced heart failure. Cardiovasc Res 1995;29:469–474. 55. Hoit BD, Shao Y, Gabel M: Global and regional atrial function after rapid atrial pacing: An echo Doppler study. J Am Soc Echocardiogr 1997;10:805–810. 56. Hoit BD, Gabel M: Influence of left ventricular dysfunction on the role of atrial contraction: An echocardiographic-hemodynamic study in dogs. J Am Coll Cardiol 2000;36:1713–1719.
41
JAMES D. THOMAS, MD ZORAN B. POPOVIC´, MD
5
Physical Determinants of Diastolic Flow INTRODUCTION PATHOPHYSIOLOGY Mechanical Properties of Left Ventricular Chamber Diastole CLINICAL RELEVANCE Physical Factors Governing Left Ventricular Filling Velocities
Understanding Pulmonary Vein Flow Through Computer Modeling of the Heart Lumped Parameter Model of the Heart and Circulation Understanding Intraventricular Flow FUTURE RESEARCH ABBREVIATIONS
INTRODUCTION Comprehensive assessment of ventricular diastolic function is a complex process. Full elucidation generally requires invasive measurements, such as left ventricular (LV) end diastolic pressure, the time constant of isovolumic LV relaxation (τ), the pressurevolume (P-V) relationship of the ventricle at end diastole, and mean left atrial (LA) pressure. Such invasive measurements are inappropriate for routine clinical purposes, and thus diastolic function is generally assessed using Doppler echocardiography, largely through the observation of transmitral and pulmonary venous flow, supplemented by myocardial velocity and color M-mode Doppler information. In order to intelligently use these noninvasive indices to infer actual diastolic function of the heart, however, it is critical that a conceptual framework be in place that reflects the physical and physiological determinants of intracardiac blood flow. In this chapter, we will outline in both basic physical principles and computer simulations the relationship between basic parameters of diastolic function and the intracardiac flow patterns that can be obtained clinically.
PATHOPHYSIOLOGY Mechanical Properties of Left Ventricular Chamber Diastole The major function of the heart in diastole is to let the blood column flow from the antechambers (left atrium and pulmonary
veins) into the left ventricle, while keeping filling pressures to a minimum. During exercise, this task is compounded by the shortening of time allowed for filling and the increase of volume needed to get into the left ventricle. Of the various chamber-wall properties that affect this process, we will briefly cover three: chamber stiffness, relaxation, and early diastolic suction.
Chamber Stiffness Chamber stiffness can be defined as the instantaneous change of pressure for a given volume increment, mathematically the first derivative of LV pressure by volume (dP/dV). LV pressure is a complex nonlinear function of LV volume, meaning that stiffness changes with diastolic volume. Although the LV P-V relationship was initially considered to be exponential in shape (concave upward),1 more recent work suggests that it is sigmoidal in shape (Fig. 5-1), with a concave downward portion to the left of the usual rising exponential.2,3 This sigmoidal curve can be expressed as: P = A ∗ e(V−Vd0)Kp+ + Pb+ V ≥ Vd0 P = −B ∗ e(Vd0−V)Kp− + Pb− V < Vd0, where Vd0 is the inflection point, A and B are the exponential curve multipliers of the upper and lower part, Pb+ and Pb− are pressure offsets of the upper and lower part, while Kp+ and Kp− are the diastolic chamber stiffness indices, determining the overall steepness of the end diastolic P-V relationship (EDPVR) above and below the inflection volume. Frequently, EDPVR is conceptually simplified by assuming that Vd0 equals 0,1 and even further 43
Chapter 5 • Physical Determinants of Diastolic Flow 120 LV pressure (mmHg)
44
90 60 30 0 0
50 LVd0
100
150
LV volume (ml) Figure 5-1 Passive pressure-volume (P-V) loop relationships. The thick continuous and thick dotted lines represent the upper and lower part of the sigmoid end diastolic pressure-volume relationship, respectively. The curvature of the upper part is determined by the chamber stiffness index, which can be interpreted as the inverse of the volume needed to increase the steepness of the P-V curve by a factor of 2.72.
simplified by neglecting Pb+, leaving us with the equation P = AeK×EDV, or in other words ln P = a + K × EDV (see also Chapter 2).1 By differentiating this equation, we see that end diastolic stiffness dP/dV is K × AeK×EDV or K × PEDV. Another parameter of note is the average stiffness,4 that is, the stroke volume divided by the LV pressure increment during diastole, a simple overall measure of how stiff the patient’s LV is during diastole over a given (working) set of conditions.5 The fact that myocardial relaxation is a continuous process throughout diastole results in changing relationships between pressure and volume from very stiff end systole P-V relationships to much more compliant EDPVR.
Relaxation Myocardial relaxation occurs because of calcium reuptake at the end of systole, producing a shift downward and rightward, leading to a fall in LV pressure (at a given volume). The rate of pressure decrease during relaxation depends on the velocity of calcium reuptake and on the LV volume: The smaller the volume, the lower the potential for pressure to fall,2 limited by the end diastolic pressure curve shown in Figure 5-1. Rate of calcium reuptake is further modified by LV lengthening during relaxation, so true relaxation can be measured only during the isovolumic period.6 Several equations have been proposed to describe the rate of isovolumic pressure decay,7 the most general being: P(t) = (Po − Pb) × e−t/τ + Pb, which can be, with some caveats, simplified to: P(t) = Po × e−t/τ, and further to8,9: τ ≈ IVRT/(ln PAVC − ln PMVO), where IVRT is the isovolumic relaxation time (the time between aortic valve closure and mitral valve opening), and PAVC and PMVO are LV pressure at aortic valve closure (AVC) and mitral valve opening (MVO), respectively, which can be approximated nonin-
vasively by systolic blood pressure and an estimate of LA pressure. Importantly, LV relaxation is a never-ending process that critically affects the end diastolic pressure that the patient achieves, particularly during exercise. Consider Figure 5-2A, showing a series of P-V curves representing successive intervals of τ, and Figure 5-2B, the same curves magnified on diastolic pressures. Note that after 6τ, the filling curve is virtually indistinguishable from the end diastolic P-V curve, as relaxation is 99.8% complete. When relaxation rate is normal, the end diastolic filling curve is completely relaxed at normal heart rates (12.5τ, 2C) and almost complete during exercise (5τ, 2D); with delayed relaxation, however, while the EDPV curve is fully relaxed at rest (6.25τ, 2E), it becomes stiff with exercise (3τ, 2F). The combination of relaxation and the EDPVR results in a dynamic P-V relationship, represented by the round dots in Figure 5-2A. Note that in early diastole, LV pressure continues to fall even though volume is increasing, meaning that the instantaneous stiffness is actually negative, which some consider diastolic suction. Two other definitions of “diastolic suction” are also relevant. Consider ventricular filling from a very low end systolic volume (where the EDPVR is concave downward). If relaxation is very rapid or filling is delayed (by mitral stenosis or experimentally with a mitral occluder2), then LV pressure can fall below atmospheric pressure.2,10 Alternatively, suction has been used to refer to the small (1 to 3 mmHg) differences in pressure between the base and the apex, which assist in the low-pressure filling of the ventricle, particularly with exercise, and which will be discussed later.11
Determinants of Intracardiac Blood Flow In the most general sense, the motion of blood inside the heart, as the motion of any fluid, is determined by the Navier-Stokes equations, a complex set of four multidimensional partial differential equations, which must be solved simultaneously at every point in space and moment in time: 䉮·v = 0 and
ρ
Dv = −∇P + B + μ∇2 v Dt
These equations appear deceptively simple but contain such complex mathematical concepts that except in the simplest of geometries, they can never be solved either analytically or with powerful supercomputers. Fortunately, considerable simplifications can be made to these equations that will facilitate a conceptual and computational approach. The most important simplification is to take the distribution of blood throughout the heart and replace it with just a few measurements at specific points inside the heart. For instance, instead of describing pressure in every cubic millimeter inside the left ventricle, we assume that LV pressure can be approximated by an average of these and give a single number for LV pressure, which is precisely how we measure and report LV pressure in practice. Similarly, instead of describing the direction and speed of blood flow at every point within the heart, we focus on points where blood velocity is
Chapter 5 • Physical Determinants of Diastolic Flow
2τ 100 3τ 4τ
50
100
4τ 5τ 6τ 25 Diastole
0 0
150
Volume (ml)
Systole
Systole
50 2τ
2τ
1τ
3τ 4τ 5τ 6τ
25 12.5τ Diastole
Pressure (mmHg)
1τ Pressure (mmHg)
150
B
50
3τ 4τ 5τ 6τ
25 Diastole
0
0 0
100
50
100
50
150
150
Volume (ml)
Volume (ml)
τ = 40 msec, HR = 150, DFP = 200 msec
τ = 40 msec, HR = 75, DFP = 500 msec
C
D
Systole
Systole
50 2τ
1τ
2τ
1τ
3τ 4τ 5τ 6τ
25 Diastole
Pressure (mmHg)
50
Pressure (mmHg)
100
50
Volume (ml)
A
3τ 4τ 5τ 6τ
25 Diastole
0
0 0
100
50
0
150
100
50
150
Volume (ml)
Volume (ml) τ = 80 msec, HR = 75, DFP = 500 msec
E
3τ
5τ 6τ
Diastole
0 50
2τ
1τ
1τ
150
0
Systole
50
Systole
Pressure (mmHg)
Pressure (mmHg)
200
τ = 80 msec, HR = 120, DFP = 240 msec
F
Figure 5-2 Diastolic pressure-volume (P-V) relationships shown as a function of the duration of diastole. Duration of diastole is measured by the multiples of a time constant of relaxation (τ). Panel A shows a series of P-V relationships that occur after successive intervals of τ, while panel B shows the same curves zoomed on diastolic pressures. The combination of relaxation and the end diastolic P-V (EDPV) relationship results in a dynamic diastolic part of the P-V loop represented by the round dots in A. Note that in early diastole, left ventricular pressure continues to fall even though volume is increasing, meaning that the instantaneous stiffness is actually negative, which some consider diastolic suction. When relaxation rate is normal, the end diastolic filling curve is completely relaxed at normal heart rates (arrow 12.5τ, C) and almost complete during exercise (arrow 5τ, D); with delayed relaxation, however, while the EDPV curve is fully relaxed at rest (arrow 6.25τ, E), it becomes stiff with exercise (arrow 3τ, F).
45
46
Chapter 5 • Physical Determinants of Diastolic Flow maximum, such as the tips of the mitral leaflets and the pulmonary vein orifices. Finally we assume fluid incompressibility and zero viscosity and heat conduction losses. In this way, the thousands of partial differential equations that would have to be solved simultaneously throughout the heart are replaced by a few ordinary differential equations that can be solved quite easily on a personal computer. To understand how these equations can model the flow within the heart, we first start by replacing the Navier-Stokes equations with the well-known Bernoulli equation, which applies to flow across discrete points such as valves. Here we present it with its inertial and convective terms (we still omit the viscous term, since it is negligible in almost every intracardiac situation): Δp = M
dv 1 + ρΔ v2 , dt 2
( )
where Δp is the pressure difference between two points; M is the inertial constant and represents the “effective” mass being accelerated; dv/dt is the instantaneous temporal acceleration of flow through the region; ρ is blood density; and Δ(v2) is the change in the square of velocity from one point to another (with pressure in mmHg and velocity in m/sec, 1/2ρv2 reduces to 4v2 in the simplified Bernoulli equation). The first product is the inertial term of the Bernoulli equation and corresponds to the energy used accelerating and decelerating flow; the second term is the convective term and represents kinetic energy of flow. The effective mass of blood in a narrow orifice is small, and thus M is a negligible quantity in, for example, valve stenosis, but the dominant term when flow is not obstructed, such as in the normal mitral valve. For constant flow through an orifice of minimal diameter D, the inertial constant M varies approximately in proportion to D, while the kinetic term 1/2ρv2 varies inversely to D4. For nonobstructive flow through the body of heart chambers, where pressure changes gradually over a distance, not at a discrete point, we must use the Euler equation, a differential version of the Bernoulli equation for pressure change along a streamline of flow: ∂p ∂v ⎞ ⎛ ∂v = −ρ ⋅ ⎜ + v ⋅ ⎟ , ⎝ ∂t ∂s ⎠ ∂s where the inertial and convective terms are in the same order as in the Bernoulli equation, but the discrete terms (Δp and Δ(v2)) have been replaced by their spatial derivatives, yielding the rate of pressure change per centimeter of distance along the streamline. To return the total pressure drop between points A and B along the streamline, this equation must be integrated numerically: B ⎛ ∂v ∂v ⎞ Δp = − ρ ⋅ ∫ ⎜ + v ⋅ ⎟ ds. A ⎝ ∂t ∂s ⎠
Figure 5-3 shows schematically how these partial derivatives are summed together to produce the pressure gradient map.12 Another practical fact is that we can obtain the same map by applying the Euler equation to a color Doppler M-mode tracing.
CLINICAL RELEVANCE In this section, we will demonstrate how interaction of solid and fluid mechanics determines intracardiac flow within three specific
Inertial term Convective term Figure 5-3 Combination of convective (right) and inertial (left) term to produce an intraventricular pressure gradient. (From Thomas JD, Popovic ZB: Intraventricular pressure differences: A new window into cardiac function. Circulation 2005;112:1684–1686.)
regions of the heart: mitral valve, pulmonary vein, and left ventricle.
Physical Factors Governing Left Ventricular Filling Velocities To describe mathematically flow through the mitral valve, we first consider a very simplified construct, consisting of the left ventricle, which has pressure as a function of time pV(t); the left atrium, which has pressure as a function of time (pA(t)); and the mitral valve, which has area AMV and contains a mass of blood M that is accelerating in passing from the left atrium to the left ventricle. To a first approximation, M is the blood within a cylinder that can fit inside the narrowest portion of the mitral leaflet tips and whose length is approximately the diameter of the valve. In reality, though, since pressure is applied across the mitral valve, the most relevant hydrodynamic concept is the length of this cylinder, termed the mitral inertance, related linearly to the diameter of that structure. We will use this construct to discuss three parameters of transmitral flow: mitral acceleration, peak mitral flow, and mitral deceleration.
Acceleration of Mitral Flow To understand the acceleration of blood across the mitral valve, we apply Newton’s second law of motion: The rate at which an object accelerates is given by the force exerted on that object divided by its mass. In this case, the acceleration would be recorded by Doppler echocardiography as the velocity acceleration detected for mitral inflow, while the force is the difference in pressure on either side of the mitral valve multiplied by the area of the valve13: a=
F ΔpA Δp = ≈ . m ρM ρL
Here we have taken advantage of the fact that valve area appears in both the numerator and the denominator (through the definition of inertial mass) to simplify the relationship. Clearly the higher the pressure difference across the valve and the smaller the length of the blood column within the mitral valve, the more rapidly the velocity will accelerate. This is why mitral velocity
Chapter 5 • Physical Determinants of Diastolic Flow increases almost instantaneously in mitral stenosis, which has a very high pressure gradient across the valve and a very small blood mass, due to the small diameter of the mitral orifice. If the blood within the mitral valve were subject to an instantaneously applied pressure difference (as if the relaxation of the left ventricle occurred suddenly), mitral velocity would start to rise linearly. In the physiological situation, the pressure gradient is not abruptly applied but rather increases gradually (roughly linearly) with time as the ventricle relaxes, resulting in a roughly parabolic mitral velocity acceleration curve.14
Peak E-Wave Velocity and Conservation of Energy Mitral valve acceleration eventually decreases over time because the maximal velocity generated by a pressure difference is limited by conservation of energy (expressed in the Bernoulli equation). Within the heart, energy in the blood takes on three principal forms: pressure (a form of potential energy), kinetic energy, and heat; the total energy in the system must remain constant. In the left atrium, where blood velocity is low, most of the energy is in the form of pressure, but as it moves toward the mitral valve, its velocity rises and it acquires a kinetic energy (1/2ρv2), which causes the local blood pressure to fall. If no energy losses appear, pressure difference and velocity are related by the simplified Bernoulli relation, which becomes roughly Δp = 4v2 when pressure is measured in millimeters of mercury and velocity in meters per second. The velocity reaches the target gradient asymptotically, but this is delayed by the amount of inertance present. Reaching the asymptote can be even further delayed by the gradient not being imposed abruptly but slowly, as seen in a normally relaxing left ventricle.
Deceleration of Mitral Flow Returning to our general model of the mitral valve, once flow has been accelerated to maximal velocity by the pressure difference across the mitral valve, it begins falling as the pressure difference between the left atrium and the left ventricle equilibrates. This is analogous to flow between two tanks stopping when the level of water in the two tanks becomes the same. When we speak of the change in pressure with a change in volume, we are dealing with the concept of compliance or its inverse, stiffness, the change in pressure for a given change in volume, which is a major determinant of the deceleration of flow across the mitral valve. Since the stiffer the ventricle, the more rapidly the pressure equilibrates across the mitral valve to decelerate the flow, we can understand why short deceleration time across the mitral valve is associated with increased ventricular stiffness. Note that in this situation, we must consider atrial and ventricular stiffness together, since both chambers are involved in determining how quickly pressure equilibrates across the mitral valve. For ventricular stiffness SV and atrial stiffness SA, the net atrioventricular stiffness Sn is simply SA + SV. To obtain a mathematical expression for mitral velocity deceleration rate, we first note that for a restrictive mitral valve with area A (i.e., mitral stenosis, though even for a normal mitral valve), the overall principle holds: 1 Δp = ρv2 , 2 which simplifies to 4v2 when Δp is in mmHg and v in m/sec. Differentiating this yields:
dΔp/dt = ρv dv/dt. But dΔp/dt can also be expressed in terms of flow across the mitral valve (Av) and net atrioventricular stiffness: dΔp/dt = −AvSn. Substituting for dΔp/dt yields: ρv dv/dt = −AvSn, or, rearranging, dv/dt = ASn/ρ. Thus, the stiffer the ventricle (or atrium) and the larger the valve area, the faster the deceleration. Alternatively, one can represent this as a purely inertial system (with a nonrestrictive orifice), which yields a simple harmonic motion, in which case the velocity across the valve is described roughly by a sine wave: ⎛ t ⎞ v (t ) = v0 sin ⎜ ⎟, ⎝ m k⎠ where v0 is the peak E-wave velocity, m is mitral inertance, and k is the stiffness constant of the ventricle. The deceleration time (time for the E wave to fall from its peak to zero velocity) can be solved for as: tdec =
π 2
ρ⋅L 1 ⋅ , A k
where ρ, L, and A are blood viscosity, mitral valve length, and mitral valve area, respectively. Therefore, the stiffer the ventricle, the shorter the deceleration time. This is intuitively very understandable: Tighter springs oscillate fast. Note that in this simplification, one assumes that active relaxation is complete at the time of interest.13 Modeling of Mitral Valve Flow Up until now, we have been discussing specific features of the transmitral E wave through significant simplifications that allow closed-form mathematical solutions. To permit more realistic modeling of transmitral flow, however, one needs more general equations that can be solved only numerically. A generalized equation for blood flow through the mitral valve and its interaction with the transmitral gradient is15: 1 dv dt = ⎛ PA − Pv − ρv2 ⎞ M , ⎝ ⎠ 2 where v is velocity, t is time, PA and Pv are instantaneous atrial and ventricular pressures, and M is mitral inertance. Conversely, the impact of mitral flow on LA and LV pressures is expressed through the equations dPA/dt = −Av/CA and dPv/dt = Av/Cv + ∂Pv/∂t, where A is mitral valve area, and CA and CV are atrial and ventricular compliance (inverse of stiffness), respectively. By solving these three differential equations, we can assess how changing the parameters of the equation that represent passive diastolic properties, LV relaxation, mitral inertance, and preload can affect transmitral flow. We start with a preliminary discussion of these interactions before presenting an even more
47
Chapter 5 • Physical Determinants of Diastolic Flow detailed model that includes pulmonary vein flow in addition to transmitral flow. Impact of Relaxation: Changes in relaxation affect the earliest part of the mitral flow (Fig. 5-4A and B). They have a profound effect on the peak values of E-wave velocity. The overall effect of relaxation on the descending phase of the E wave is more complex: When τ is very short, relaxation is essentially complete
by the time of the peak of the E wave, and small changes in τ have little effect on deceleration; on the other hand, if τ is significantly longer, the relaxation remains an important aspect of the deceleration phase, and further lengthening of τ will cause a decrease in the deceleration rate (that is, prolongation in deceleration time). Passive Diastolic Properties of the Left Ventricle: Increasing the steepness of the diastolic P-V curve affects
LV RELAXATION CONSTANT (MSEC)
LV RELAXATION CONSTANT (MSEC) 140
20
Velocity (cm/sec)
LV pressure (mmHg)
25
15 10
40
5
60
20 30
100
50
70
90
110
A 16.2 12.3 10.0
D 7.8 6.9 5.8
VTI 13.1 12.7 12.0
60 40
20
40
60
20
0
130
100
200
300
400
500
Time (msec)
A
B LV VOLUME CONSTANT
(CM3)
LV VOLUME CONSTANT (CM3) 140
20 15 40
10
60 90
5
Velocity (cm/sec)
25 LV pressure (mmHg)
E 91 82 73
80
LV volume (cm3)
Vvk 40 60 90
120 100 80
E 58 82 93
A 9.2 12.3 13.7
D 8.4 6.9 4.8
VTI 6.6 12.7 16.3
60 40
40
60
90
20 0
0 30
50
70
90
110
0
130
100
LV volume (cm3)
200
300
400
500
Time (msec)
C
D LA PRESSURE (mmHg)
LA PRESSURE (mmHg) 140
20 5
10
20
15 10 5
Velocity (cm/sec)
25
LAP E A 5 40 4.5 10 82 12.3 20 124 24.1
120 100
20
80 60
D 3.5 6.9 9.6
VTI 6.9 12.7 17.7
10
40
5
20 0
0 30
50
70
90
LV volume
E
TAU 20 40 60
LV relaxation constant (msec)
120
0
0
LV pressure (mmHg)
48
110
130
0
(cm3)
100
200
300
400
500
Time (msec)
F
Figure 5-4 Effect of left ventricular relaxation constant (τ; panels A and B), volume constant (inverse of chamber stiffness index; panels C and D) and left atrial pressure (preload; panels E and F) on a left ventricular pressure volume curve (left panels) and mitral valve velocity profile (right panels). Prolonged relaxation blunts the lower left corner of pressure volume curve and decreases peak mitral valve flow. Increased stiffness (i.e., decreased volume constant) moves lower left corner of the pressure volume loop to the left and upwards, decreases peak mitral valve flow and increases its deceleration rate. Finally, increased left atrial pressure shifts lower right corner of the pressure volume loop to the right while increasing peak mitral valve flow. E, peak velocity of E wave (cm/sec); A, peak acceleration rate of E wave (m/sec2); D, peak deceleration rate of the E wave (m/sec2); VTI, E wave velocity time integral (cm). (From Thomas JD, Weyman AE: Echocardiographic Doppler evaluation of left ventricular diastolic function. Physics and physiology. Circulation 1991;84: 977–990.)
Chapter 5 • Physical Determinants of Diastolic Flow primarily deceleration time, that is, operative chamber stiffness (Fig. 5-4C and D). However, a caveat has to be introduced. The ventricular P-V curve (as well as the atrial P-V curve) is exponential in shape. Since operative chamber stiffness is equal to the slope of these curves, stiffness and mitral deceleration rise when operative chamber volumes increase. This change is indistinguishable from the change that may arise from material properties of the ventricle, leading to a steeper curve at all pressures (that is, a true change in the diastolic properties of the ventricle). Preload Alterations: Preload affects all components of the mitral velocity curve (Fig. 5-4E and F). However, its most prominent effect is on the E wave. For mitral deceleration, the effects depend on the initial position of the P-V curve. If the preload increase does not move the P-V loop too much into the steeper portion of the EDPVR, the effects may be negligible. It is not so in the severely diseased ventricles operating close to the limit of the preload reserve. Mitral Inertance: Mitral inertance affects both acceleration and deceleration of the mitral valve. It changes acceleration of mitral valve inflow and leads to flattening of the parabolic curve of velocity rise. Mitral inertance essentially acts by introducing the time lag between the transmitral pressure gradient and the mitral velocity curve: It makes the peak of the velocity curve occur after the peak of the transmitral pressure gradient curve (Fig. 5-5). It also introduces the difference between the true mitral valve pressure gradient and the pressure gradient calculated from application of the simplified Bernoulli equation to transmitral velocity (see Fig. 5-5). As it correlates with the mitral valve area, it can be largely neglected in patients with mitral stenosis, but it is an important factor in the nonstenotic valve. We have shown, by using the combination of echocardiography and direct pressure recordings, that the average value of mitral inertance in humans is 3.82 ± 1.22 g/cm2, corresponding to an effective length of the blood column of about 3.6 cm within the mitral valve.16 Future research may allow direct estimation of inertance from echocardiography data alone. Interaction of Ventricular Compliance and Relaxation: Importance of Heart Rate: During early filling, relaxation is not yet complete. This means that filling occurs while the dynamic P-V curve continuously shifts downward and toward the right. Thus, as the ventricle fills, it moves to the steeper part of the EDPVR, which in turn becomes progressively “flatter.” These two
P underestimation Actual ΔP by catheter Calculated ΔP from velocity
Phase lag Figure 5-5 Impact of mitral inertance: Transmitral velocities lag behind transmitral pressure differences. (From Nakatani S et al: Mitral inertance in humans: Critical factor in Doppler estimation of transvalvular pressure gradients. Am J Physiol Heart Circ Physiol 2001;280:H1340–1345.)
trends may offset each other, with the result that diastolic pressures are constant or falling despite an increase in LV volume, giving the impression of negative stiffness in early diastole. Figure 5-4A and B show how τ effects mitral valve flow and end diastolic P-V relationships.15 We reiterate here that while τ affects the early part of the diastolic P-V curve, τ has an effect on end diastolic volume only when dramatically prolonged. The situation is different during exercise, where shortening of the diastolic interval and the increase of arterial blood pressures produce incomplete relaxation (with a rise in end diastolic pressure) rather than simply delayed relaxation (which has little impact on LV end diastolic pressure).17
Understanding Pulmonary Vein Flow Through Computer Modeling of the Heart Such conceptual thought experiments as those previously discussed can be helpful in understanding general trends in the relationship between invasive parameters and noninvasive measurements, but a more comprehensive approach can be obtained with a computer model that invokes realistic descriptions of chamber and valvular functions. The reason for this is that the flows of blood through different parts of the circulatory system are interdependent. A change in a single parameter (e.g., resistance in systemic arterioles) affects flow throughout the system. Thus, changes in pulmonary circulation affect systemic circulation even if not assuming any ventricular-ventricular interactions.
Lumped Parameter Model of the Heart and Circulation Since the mid-1980s, the authors and colleagues have worked to develop models of the heart with increasing complexity, beginning from a very simple isolated model of the mitral valve used to understand the mitral pressure half-time; to models incorporating more realistic diastolic characteristics of the ventricle and atrium; to models that simulated all the cardiac chambers and valves, as well as the peripheral and pulmonary vasculature; to most recently a simulation that links knowledge of basic myocyte and fiber function to the gross architecture of the heart. We will use a lumped parameter model to explore some of the determinants of transmitral and pulmonary venous flow inside the heart.3 The lumped parameter approach assumes that blood flow is similar to the flow of a current through an electric circuit, which can be represented at specific nodes of the circuit (left atrium, left ventricle, aortic valve, etc.). Each of these nodes is represented by a specific ordinary differential equation. The principle of the conservation of flow (Kirchoff ’s first law) is applied throughout the circuit, and the parameters of the system are obtained by simultaneously solving all differential equations. Our model consists of 24 coupled differential equations representing eight different chambers (the right atrium and ventricle, pulmonary arteries and veins, left atrium and ventricle, aorta, and systemic veins) as well as the valves and vascular beds connecting them (Fig. 5-6), giving time-varying flows and pressures throughout the model. Figure 5-7 displays normal Doppler velocity tracing of pulmonary vein and transmitral flow recorded in the normal subject by transesophageal echocardiography (left) and a model output of corre-
49
50
Chapter 5 • Physical Determinants of Diastolic Flow Pulmonary capillaries Pulmonary artery
Pulmonary vein
Pulmonic valve
Left atrium
Right ventricle
Tricuspid valve
Mitral valve
Left ventricle
Right atrium
tions have opposite effects. Mitral valve regurgitation (through the parameter MV ERO) has its effect mostly on the systolic waveform, while Kp+ and τ have effect mostly on AR. The results of this sensitivity analysis illustrate the complexities of these relationships, which are not intuitively apparent. In contrast, for the mitral inflow, as expected, the greatest influence was exerted by preload. Importantly, τ showed the opposite effects on E and A waves, in congruence with the E/A ratio being reversed in the setting of impaired relaxation. Table 5-1 should be kept in mind, as very often pulmonary vein flow is used as a surrogate marker of LA function, with S, D, and AR reflecting the LA reservoir, conduit, and pump functions, respectively. Our table indicates that for none of these is there a straightforward relationship between the specific waveform and specific material parameters of the pulmonary vein flow waveform.
Aortic valve Venae cavae
Systemic capillaries
Aorta
Figure 5-6 A lumped parameter model of the heart. (From Thomas JD et al: Physical and physiological determinants of pulmonary venous flow: Numerical analysis. Am J Physiol 1997;272:H2453–H2465.)
sponding pulmonary vein and transmitral flow velocities, along with simulated pressure curves. Starting from this normal shape of velocity and pressure tracings, we first modeled a delayed relaxation pattern, by increasing τ from 50 to 120 msec. Then, to model restrictive filling, we shifted the end systolic P-V relationship by 10 ml to the right and decreased contractility from 4.0 to 2.5 mmHg/ml, thus moving the P-V loop toward the steeper part of the EDPVR. The model outputs are displayed in Figure 5-7, along with actual Doppler and pressure tracings of a representative patient with delayed relaxation (panel B) and restrictive filling (panel C). One can observe a close correspondence between the modeled and actual shapes, with a correlation coefficient approaching 1. Figure 5-8 shows the impact of some of the model parameters on pulmonary vein and mitral valve flow. LA systolic and diastolic function had opposing effects on the S and AR waves of the pulmonary vein, with minimal effect on the D wave (panels A through D). Finally, all waveforms showed strong positive response to preload (panels E and F). Worse LV relaxation decreases the mitral valve E/A ratio and blunts diastolic waves of pulmonary vein flow (panels G and H). These observations are analyzed in depth in Table 5-1, by showing the sensitivities of peak velocity of the three pulmonary vein flow waves and of two mitral inflow waves on model parameters. The sensitivities represented here were obtained by calculating a normalized Jacobian matrix, showing the proportional change in an output index for a given incremental change in an input parameter. A sensitivity of 1 means that for each 1% change of a model parameter value, the output variable will change by 1%. A value of 2 indicates a 2% change for a 1% increase. Negative numbers indicate inverse relationships. Because of the highly nonlinear nature of the model, these observations are relevant to only very small increments but do provide an overall sense of the dependency of Doppler indices on hemodynamic parameters. For example, one can readily observe the dramatic effect of preload on all three pulmonary vein flow waves, in particular AR, which increases almost 6-fold the increment in blood volume. Furthermore, one can observe that LA diastolic and systolic func-
Impact of Pulmonary Vein–Left Atrial Pressure Gradient and Left Atrium Size Similar to mitral valve inertance is pulmonary vein inertance. Its effect on pulmonary vein flow is similar to its effect on transmitral flow. Inertance leads to a lag of pulmonary vein flow behind the PV–LA pressure gradient. This is intuitively understandable: the pressure is needed to push the blood column, which then keeps moving because of its kinetic energy.18 Another interesting observation is the impact of the change of LA size, induced by clipping of the LA appendage, on pulmonary vein flow. The only affected pulmonary vein flow wave was the S wave, indicating that LA size affects primarily LA reservoir function.19
Understanding Intraventricular Flow The previous model formulations seem like child’s play when one is confronted with the task of simulating intraventricular flow with its demands for representing pressure and velocity throughout the chamber. Three major phenomena develop simultaneously during filling: An intraventricular gradient appears at the beginning of LV filling, the blood column slows after exiting the mitral valve while continuing as laminar flow toward the apex, and finally vortices start appearing due to propagation of the blood column past the tips of the mitral leaflets, while the left ventricle continually changes its size. It is well documented that a small LV base-to-apex pressure gradient develops in early diastole (Fig. 5-9A). We have already discussed that this gradient is of utmost importance for the lowpressure filling of the ventricle and is commonly termed diastolic “suction.” However, the exact mechanisms causing it are unclear. The most probable mechanisms are isovolumic conformational changes in LV shape that are driven by elastic forces within the myocyte and/or the myocardial interstitium, explored in greater depth below. As blood emerges from the mitral valve, it slows down as the flow stream expands to fill the left ventricle. Interestingly, within the first four centimeters of the left ventricle, the velocity of propagation of the blood column is fairly constant, despite the dramatic change of the chamber size, but it is generally less than the actual velocities within the bolus of blood traversing the mitral valve. For example, pulsed-wave Doppler may record a velocity of 100 cm/sec, while the normal propagation velocity will be only 60 or 70 cm/sec, and even lower in diseased hearts, due to the bleeding off of vortices at the front of the bolus of blood. We and others have shown that in heart disease, the amount of slowing
Chapter 5 • Physical Determinants of Diastolic Flow
50 E
100 0
Δ = –1.53±9.38 cm/sec
A
0.5
Velocity (cm/s)
y = 0.95x r = 0.94 p < 0.0001
Velocity (cm/s)
Velocity (cm/s)
0
0
A
E 100
1.0
0
Time (sec)
0.5
S
60
y = 1.02x r = 0.92 p < 0.0001
D
0 AR –60 0
1.0
0.5
Δ = 0.76±7.69 cm/sec
Velocity (cm/s)
PULMONARY VENOUS FLOW
TRANSMITRAL FLOW
60
S
0 AR –60
1.0
0
Time (sec)
Time (sec)
D
0.5
1.0
Time (sec)
A
0
0.5
A 100
1.0
0
Time (sec)
0.5
A
40
0
1.0
Δ = 3.52±8.65 cm/sec
E
80
Time (sec)
0.3
0
y = 1.02x r = 0.96 p < 0.0001 Δ = –1.54±3.32 AR cm/sec
–40 0
0.5
40 D
S 0
AR
–40
1.0
0
0.5
80
0.6
y = 0.96x r = 0.91 p < 0.0001 Δ = 3.72±11.81 cm/sec
S 0 AR –60
1.0
0
0
20
0 0
1
0
1 Time (sec)
B
0.3
0.6
Time (sec)
y = 1.16x r = 1.0 p < 0.0001 Δ = 1.99±1.61 mmHg
0 0
1
30
0
0.6
0
Time (sec)
Pressure (mmHg)
0
AR
0.6 Time (sec)
LV PRESSURE Pressure (mmHg)
Pressure (mmHg)
Δ = –3.63±11.30 mmHg
D
0
0
LV PRESSURE y = 0.94x r = 0.96 p < 0.0001
S
0.3 0.6
30
Time (sec)
Time (sec)
120
80
Time (sec)
Pressure (mmHg)
0
0.6
LA PRESSURE Pressure (mmHg)
Pressure (mmHg)
Δ = –1.04±1.04 mmHg
0.3
–60
LA PRESSURE y = 0.84x r = 0.97 p < 0.0001
0
Time (sec)
D
Time (sec)
Time (sec)
20
E
100
PULMONARY VENOUS FLOW Velocity (cm/s)
D
Velocity (cm/s)
Velocity (cm/s)
S
A
Time (sec)
PULMONARY VENOUS FLOW 40
Velocity (cm/s)
100
E
y = 0.95x r = 0.92 p < 0.0001
Velocity (cm/s)
A Δ = 0.70±8.35 cm/sec
0
Pressure (mmHg)
E
50
0
0
120
0 0
1
100
y = 0.80x r = 0.99 p < 0.0001 Δ = –5.01±8.69 mmHg
0 0
Time (sec)
0.6 Time (sec)
Pressure (mmHg)
y = 0.99x r = 0.93 p < 0.0001
TRANSMITRAL FLOW Velocity (cm/s)
0
Velocity (cm/s)
Velocity (cm/s)
TRANSMITRAL FLOW
100
0 0
0.6 Time (sec)
C
Figure 5-7 Actual and modeled pressures and filling patterns in a normal subject (A), patient with delayed relaxation (B), and patient with restrictive filling (C). Note similarities between observed and modeled data. (From Thomas JD et al: Physical and physiological determinants of pulmonary venous flow: Numerical analysis. Am J Physiol 1997;272:H2453–H2465.)
51
Chapter 5 • Physical Determinants of Diastolic Flow TRANSMITRAL FLOW
PULMONARY VEIN 60
A
Velocity (cm/sec)
Velocity (cm/sec)
90 E 60 30 0
S
0.5
0.48* 0.64 AR
1
0
1
B 90
60 A
E
Velocity (cm/sec)
Velocity (cm/sec)
0.5 LA end-systolic elastance (mmHg/ml)
A
60 30 0
30
D
S
0.2* 0.25
0
0.3 AR –30
0
0.5
1
0
LA end-diastolic elastance (mmHg/ml)
0.5
1
LA end-diastolic elastance (mmHg/ml)
C
D 150 E
Velocity (cm/sec)
100 A
100 50 0
S
D
50
3800* 0 4180 –50
4560 AR
–100 0
0.5
1
0
0.5
Preload (ml)
1
Preload (ml)
E
F 60 E
Velocity (cm/sec)
120 A
80 40
D
S 30
0.05* 0
0.10 AR
0.15
–30
0 0
0.5
0
1
LV relaxation time constant (sec)
G
0.32
0
LA end-systolic elastance (mmHg/ml)
Velocity (cm/sec)
D
30
–30 0
Velocity (cm/sec)
52
0.5
1
LV relaxation time constant (sec)
H
Figure 5-8 Simultaneous modeling of pulmonary vein and transmitral velocities using lumped-parameter approach. Increased left atrial systolic elastance (i.e., higher contractility) increases A wave of the transmitral flow and AR wave of the pulmonary vein flow (panels A and B); increased left atrial diastolic elastance (i.e., higher passive stiffness) has the opposite effect (panels C and D); increased preload increases all components of transmitral and pulmonary vein flow (panels E and F); finally, prolongation of ventricular relaxation decreases E and increases A wave velocities of the transmitral flow and decreases diastolic pulmonary vein flow (panels G and H). (From Thomas JD et al: Physical and physiological determinants of pulmonary venous flow: Numerical analysis. Am J Physiol 1997;272:H2453–H2465.)
Chapter 5 • Physical Determinants of Diastolic Flow of the blood column relative to the component velocity (that is, the so-called E/Vp ratio) is related to LV filling pressures,20,21 though the exact mechanistic relationship between these two phenomena is unclear (Fig. 5-9B). As noted, intraventricular flow during diastole characteristically forms circular eddies lateral to the main flow, which can be seen when the heart is imaged by echocardiography with highfrequency transducers, or after the contrast injection, and can be quantified by magnetic resonance velocity mapping (Fig. 5-9C).22 Creation of vortices plays a crucial role in mitral valve closure and in the transfer of blood toward the apex and then into the LV outflow tract. Again, full elucidation of vortex generation and movement is unclear.
TABLE 5-1 RESULTS OF SENSITIVITY ANALYSIS FOR THE PULMONARY VEIN AND TRANSMITRAL FLOW PULMONARY VEIN FLOW
Parameters Preload LV compliance index LV systolic elastance LV τ LA systolic elastance LA diastolic stiffness Mitral valve area MV ERO
TRANSMITRAL FLOW
S
D
AR
E
A
2.38 –0.45 –0.2 0.05 0.8 –0.79 –0.06 –2.73
2.33 –0.12 0.05 –0.19 –0.14 0.6 0.12 0.17
5.88 –1.15 –0.44 –1.76 0.97 –1.27 –0.16 –0.18
1.53 –0.24 –0.09 –0.33 –0.38 –0.11 0.07 0.32
1.9 –0.04 –0.19 0.16 –0.25 0.97 –1.02 –0.11
Modeling of the Intraventricular Flow by Fluid-Mechanical Computer Modeling of the intraventricular flow is demanding because ventricular motion effects blood flow, but blood flow itself effects ventricular motion. Some simplification is possible with compu-
A, atrial wave; AR, atrial reversal wave; D, diastolic wave; E, early wave; ERO, effective regurgitant orifice; LA, left atrial; LV, left ventricular; MV, mitral valve; S, systolic wave. From Thomas JD et al: Physical and physiological determinants of pulmonary venous flow: Numerical analysis. Am J Physiol 1997;272:H2453–H2465.
25
P
ECG
Pressure (mmHg)
20 MVO
15 10
LVP Base LA
LVP Apex 5 0
IVPG
–5
A 30
MV
pw (mmHg)
25 20 15 10
r = 0.80 p < 0.001 y = 5.27x+4.66 SEE = 3.1
5
LV
0 0.5
1
1.5
2
2.5
3
3.5
4
E/vp ratio
B
C
Figure 5-9 Three consecutive early diastolic intraventricular phenomena: A, Intraventricular pressure gradient formation in early diastole; B, relationship between the slowing of the blood column in the early diastole with preload; and C, formation of vortices. (B, From Garcia MJ et al: An index of early left ventricular filling that combined with pulsed Doppler peak E velocity may estimate capillary wedge pressure. J Am Coll Cardiol 1997;29:448–454; C, From Kilner PJ et al: Asymmetric redirection of flow through the heart. Nature 2000;404:759–761.)
53
54
Chapter 5 • Physical Determinants of Diastolic Flow tational fluid dynamics (CFD) models, where LV motion is prescribed based on images from echocardiography or magnetic resonance imaging, but these models lack fluid feedback on the myocardium. Full fluid-structure interaction (FSI) models are being developed in which movement of the blood and muscle is not known a priori but is computed during the simulation on a ∼mm three-dimensional grid at ∼msec intervals, an extraordinary computational challenge. Furthermore, most commercially available software programs allow only small strain deformations (up to 5%) and are inadequate for cardiac modeling, where strains may exceed 50%. Fortunately, large strain solvers recently have been implemented in commercial codes, which enabled several researchers to begin true fluid-solid interaction modeling of LV function.23–25 Although these models have simplified geometry and muscle properties, they give us insight into the flow propagation in the normal heart. Flow Propagation Inside the Ventricle We will first discuss the filling modeled by an axi-symmetric thin-walled FSI model of the left ventricle containing both LV outflow and inflow.26 During the rapid filling phase, a jet starts its propagation through the mitral valve into the ventricular chamber. Driven by this jet, a doughnut-shaped vortex is formed around the jet just downstream from the mitral orifice. During filling, the vortex ring travels toward the middle of the expanding ventricle, while a weak vortex is formed within the aortic outflow tract, which dissipates late during diastasis. After the vortex ring reaches the middle of the ventricle, its anterior part becomes stationary while the posterior part continues forward. During atrial contraction, a new, weaker vortex ring that encompasses a part of the outflow tract is formed, while the posterior side of the initial vortex ring continues to move into the posterior apical region of the ventricle. At the end of the filling, the posterior part of the second vortex ring merges with the first one, forming a large vortex in the posterior apex (Fig. 5-10). Effect of Ventricular Dilation on Flow A fascinating conundrum in echocardiography is the possibility of having very high velocities of mitral valve inflow, while the flow propagation velocity (velocity of the blood column inside the left ventricle) may be several times slower. Baccani et al. created a simple CFD simulation of this problem in dilated cardiomyopathy.27 They showed that LV dilation slows down flow propagation velocity. Furthermore, they have shown that this is associated with the initial vortex staying attached to the mitral valve (Fig. 5-11). Pressure Propagation Inside the Ventricle Flow and velocity propagation, though tightly linked, are separate phenomena. To analyze this relationship, we developed a two-dimensional axi-symmetric thick-walled FSI model. The fluid domain was represented by 6321 triangular elements, while the solid domain was represented by 1206 bricklike quadrilateral elements. The time-varying pressure at the mitral valve opening was generated by the numerical simulations of the lumped parameter model. The fluid domain was represented by incompressible Navier-Stokes equations. The LV wall was given time-varying stiffness properties, with strains calculated using a large strain deformation code. Finally, the decrease in stiffness of the wall during ventricular relaxation was modeled as an exponential decay from systolic stiffness to diastolic stiffness.28,29 The baseline ventricular filling pattern is illustrated in Figure 5-12, showing the E wave entering the cavity, with formation of
Figure 5-10 Computer modeling of vortices during left ventricular filling. During the rapid filling phase, initial inflow through the mitral valve forms a doughnut-shaped vortex downstream from the mitral orifice. This vortex ring then travels toward the middle of the expanding ventricle, while a weak vortex of brief duration is formed within the aortic outflow tract. After the initial vortex ring reaches the middle of the ventricle, its anterior part becomes stationary while the posterior part continues forward. During the atrial contraction, a new, weaker vortex ring that encompasses a part of the outflow tract is formed, while the posterior side of the initial vortex ring continues to move into the posterior apical region of the ventricle. In the end of the filling, the posterior part of the second vortex ring merges with the first, forming a large vortex in the posterior apex. (From Cheng Y et al: Fluid-structure coupled CFD simulation of the left ventricular flow during filling phase. Ann Biomed Eng 2005;33:567–576.)
a vortex off the tips of the mitral valve that propagates beside the E wave into the ventricle. The ventricle expands under the influence of decreased elasticity and increased strain due to LV filling. As the relaxation finishes after about 200 msec, the E wave loses strength, and velocity slowly decreases as the expansion of the ventricle ceases. After the E wave ends, atrial pressure increases and the A wave enters the ventricular cavity, forming another vortex off the mitral valve. As the A wave propagates toward the apex, this vortex is taken up by the first vortex that is still present. By now, ventricular stresses have significantly increased, raising LV pressure and slowing down filling. The filling behavior described here corresponds with clinically observed behavior. Using this fluid-structure interaction model, we have investigated the influence of changes in relaxation (τ) and end diastolic stiffness. In Figure 5-12, color Doppler M-mode images are shown from four different conditions: (A-B) baseline, (C-D) shorter τ, (E-F) longer τ, and (G-H) increased ventricular diastolic stiffness. At baseline, the E and A waves are roughly balanced (equal color-mapped peak velocities). Intraventricular pressure gradients (IVPGs) are slightly above normal due to the
Chapter 5 • Physical Determinants of Diastolic Flow 0
0.02
1
0.02
0.01
0.5
0.04
0
z
0 0.01
0.06
0.02
0.08
–0.5 0.08
0.06
0.04
0.02
0
0
0.2
0.4
0.6
0.8
1
–1
t
0 0.02 0.03 0
z
Figure 5-11 Impact of ventricular dilation on flow. Images represent a computer model of left ventricular flow in mid-diastole. In a normal ventricle (upper left panel), the proximal vortex becomes unattached and freely moves toward the apex, enabling prompt filling of the ventricle, as evidenced by high flow propagation velocity (upper right panel). In contrast, left ventricular dilation prohibits the movement of vortices toward the apex (lower left panel), and slows down flow propagation velocity (lower right panel). (From Baccani B et al: Fluid dynamics of the left ventricular filling in dilated cardiomyopathy. J Biomech 2002;35:665–671.)
0.06 0.02
mild mitral stenosis inherent in the model. With shorter τ, the E-wave velocity, Vp, and IVPG increase, while a longer τ produces the classic appearance of delayed relaxation (i.e., low E velocity, slow propagation, large A wave, and reduced IVPG).29–32 Because of the mild mitral stenosis, E-wave deceleration time is prolonged; for very large time constants, the E wave is no longer separated from the A wave.29 With a stiffer ventricle, velocities and IVPG are reduced, reflecting lower stroke volume, the expected behavior when atrial pressure is held at baseline so no pseudonormalization will occur. The A wave is particularly blunted, reflecting ineffective filling of the stiff ventricle.31,33 As support for these findings, we have recently shown that in humans, IVPG in the elderly becomes blunted with worsened relaxation.9 In addition, elderly people with preserved ventricular compliance (i.e., a less stiff ventricle), still had worsened IVPGs, reflecting the dependence of IVPGs on intact relaxation, without an impact of compliance.
Insights from Experimental Studies Currently, no model can predict the impact of structural changes of LV shape on flow propagation within the LV. In that situation, we are left with some experimental studies. We will discuss briefly some of them. Presence of Ischemia Acute ischemia delays apical filling34,35 and is associated with the loss of IVPGs.36,37 Beppu et al. described the altered apical blood flow pattern in dogs with apical akinesis after coronary ligation.38 Ischemia decreases the maximal distance of mitral flow into the ventricle during diastole, with blood flow in the region of the infarction becoming either static or eddied.38,39
0.09 0
0.2
0.4
0.6
0.8
1
Distorted Left Ventricular Geometry In a recent sheep study, myocardial aneurysm was induced by the ligation of coronary arteries subtending the LV apex.40 Flow propagation velocity abruptly decreased when blood flow column entered LV aneurysms and the point of this deceleration coincided with the edge of the aneurysm (Fig. 5-13). The infarct size correlated with the distance from the apex to the point of the deceleration of flow propagation. However, no point of abrupt decrease was detected when myocardial infarct was induced by circumflex occlusion. Also, significantly, the presence of breakpoint coincided with loss of the IVPG. In a clinical study that is congruent with these findings, we have shown that removal of the aneurysm by infarct exclusion surgery improves IVPGs.41 These data show that in order for flow to enter the heart cavity, heart walls have to be compliant. Effect of Isovolumic Processes During the isovolumic period of LV relaxation, the left ventricle changes its shape in the absence of any change of volume. The most dramatic aspect of this is that the left ventricle untwists, a process that is linked to the elastic elements of the ventricle that induce early LV suction. Untwisting is a necessary successor to systolic torsion (twist). Systolic torsion and diastolic untwisting occur because myocardial fibers form a leftward helix in the subepicardium.42 The contraction of these fibers rotates the base and the apex of the left ventricle clockwise and counterclockwise, respectively. However, unlike systolic torsion that increases proportional to systolic volume change, untwisting is a phenomenon that peaks prior to the opening of the mitral valve.11,43 The reason this occurs is that systolic torsion builds potential energy within the elastic elements of the heart.44 With relaxation, energy is released, and elastic elements promptly restore the original, untwisted LV shape. It is no surprise that untwisting is inversely
55
Chapter 5 • Physical Determinants of Diastolic Flow A LV
2
VP
5
1
4
0.6
MV
0
LA
–1 0
0.1
0.2
0.3 0.4 Time (s)
0.5
0.6
0.7
1
0.2
0
1
4
3
0.8
2
0.6
1
0.4
1
0
0.2
0
0.3 0.4 Time (s)
0.5
0.6
0.7
Distance (cm)
4
0.2
0
4
3
0.8
2
0.6
1
0.4
1
0
0.2
0
0.5
0.6
0.7
Distance (cm)
1
0.3 0.4 Time (s)
0
4
3
0.8
2
0.6
1
0.4
1
0
0.2
0
0.5
0.6
0.7
0
Distance (cm)
1
0.3 0.4 Time (s)
6
2 0
0.1
0.2
0.3 0.4 Time (s)
0.5
0.6
0.7
–2 10
F
8 6
2 0
0.1
0.2
0.3 0.4 Time (s)
0.5
0.6
0.7
–2 10
H 8 6
3 4 2
–1 0
2 0
0.1
0.2
0.3 0.4 Time (s)
0.5
0.6
0.7
Relative pressure (mmHg)
4
Velocity (m/s)
5
0.2
8
–1 0
1.2
0.1
10
4
5
–1 0
–2
2
6
G
0.7
3
1.4
6
0.6
Relative pressure (mmHg)
4
Velocity (m/s)
5
0.2
0.5
D
–1 0
1.2
0.1
0.3 0.4 Time (s)
4
5
–1 0
0.2
2
6
E
0.1
3
1.4
6
0
Relative pressure (mmHg)
5 Velocity (m/s)
1.2
0.1
2
–1 0
5
–1 0
6 4
6
C
8
2
0.4
0
10
B
3
1.4
6
Distance (cm)
1.2
0.8
3
1
Distance (cm)
6
Velocity (m/s)
4
1.4
Relative pressure (mmHg)
E-wave
5 Distance (cm)
A-wave
Distance (cm)
6
Distance (cm)
56
–2
Figure 5-12 Modeling of fluid-structure interaction showing influence of changes in relaxation (τ) and end diastolic stiffness. Pseudo–M-mode images of time-velocity (left panels) and time-pressure data from four different conditions: A and B, baseline; C and D, shorter τ; E and F, longer τ; and G and H, increased ventricular diastolic stiffness. At baseline, the E and A waves are roughly balanced (equal color-mapped peak velocities). IVPGs are slightly above normal due to the mild mitral stenosis inherent in the model. With shorter τ, the E-wave velocity, Vp, and IVPG increase, while longer τ produces the classic appearance of delayed relaxation (low E velocity, slow propagation, large A-wave and reduced IVPG). Because of the mild mitral stenosis, E-wave deceleration time is prolonged and for very large time constants no longer fully separated from the A wave. With a stiffer ventricle, velocities and IVPG are reduced, reflecting lower stroke volume, the expected behavior when atrial pressure is held at baseline to prevent pseudonormalization. The A wave is particularly blunted, reflecting ineffective filling of the stiff ventricle.
Chapter 5 • Physical Determinants of Diastolic Flow
7 IVPG (mmHg)
: r2 = 0.99 5 3 1 –1 4 Normal
2
0
Apex-to-base distance (cm)
Apex (A) A-D (23 mm) IVPG (mmHg)
Abrupt decrease in propagation velocity (D)
related to τ, and its magnitude may be used as a surrogate measure of relaxation. Events during early diastole in a normal left ventricle can be summarized in the following way. Relaxation leads to prompt LV untwisting, the peak of which coincides with mitral valve opening; this is immediately followed by peak early mitral annulus velocity; finally, formation of the IVPG facilitates early LV filling. We have shown a strong correlation between untwisting and IVPG at rest and during exercise.11 Furthermore, we have shown that essentially the same slope of untwisting-IVPG relationship exists in patients with severe hypertrophic cardiomyopathy (Fig. 5-14). This indicates that untwisting and the IVPG are tightly coupled and that either untwisting is necessary to produce the IVPG or that some strong third factor binds them, expressed simultaneously by untwisting and by the IVPG. Thus, it seems that three major factors determine flow propagation and IVPGs: chamber dilation, worsened relaxation, and loss of compliance.
: r2 = 0.97 : r2 = 0.67
4 3 2 1 0 –1 4
2
4
p < 0.00001 Normal exercise p < 0.00001
2 Normal rest
HCM exercise P = NS
HCM rest P = NS 0 0
FUTURE RESEARCH The number of flow-based diastolic parameters that are proposed to be helpful in clinical practice is staggering. Still, more and more indices are proposed every year. This review of the physics behind diastolic flows aims to elucidate common underlying processes that are manifested through various diastolic parameters. In this manner, we are trying to bring another “Rosetta stone” to the cardiology community.45 In the remainder of this book, many of these indices will be shown in application to various diseases of
0
LV aneurysm
Peak IVPG (mmHg)
Figure 5-13 Effect of apical aneurysm on intraventricular flow propagation. Flow propagation velocity decreases abruptly when blood enters the aneurysm. This sudden deceleration coincides with the loss of intraventricular pressure gradient. (From Asada-Kamiguchi J et al: Intraventricular pressure gradients in left ventricular aneurysms determined by color M-mode Doppler method: An animal study. J Am Soc Echocardiogr 2006;19:1112–1118.)
5
–5 Peak untwisting velocity (rad/s)
Figure 5-14 Relationship between left ventricular untwisting and intraventricular pressure gradients (IVPGs) at rest (open markers) and during exercise (closed markers). Normal subjects are marked by squares, while hypertrophic cardiomyopathy patients are marked by triangles. While the same slope of untwisting-IVPG relationship exists in normal hearts and in those with severe hypertrophic cardiomyopathy, hearts of patients with hypertrophic cardiomyopathy cannot generate the increase of untwisting during the exercise, leading to smaller IVPG increase. (From in Notomi Y et al: Enhanced ventricular untwisting during exercise: A mechanistic manifestation of elastic recoil described by Doppler tissue imaging. Circulation 2006;113:2524–2533.)
57
58
Chapter 5 • Physical Determinants of Diastolic Flow diastolic function; it is hoped that by keeping these physical principles in mind, the reader will better be able to assess diastolic dysfunction clinically. Despite numerous published studies, our current understanding of physical properties is limited. One field that is largely undefined is the modeling of structural-fluid relationships within the left ventricle, which is a structure so complex that to accurately calculate mechanical events during diastole, one needs several days even with the use of the most powerful supercomputers. Therefore, until computer processing power is increased by an order of magnitude, we will lack the ability to answer such seemingly simple questions as: Is torsion necessary to LV contraction? and What happens to end diastolic pressure if we eliminate an increase in the IVPG during diastole? To paraphrase Sir Isaac Newton, we are still like children playing on the seashore and diverting ourselves now and then, finding a smoother pebble or a prettier shell than ordinary, whilst the great ocean of truth lies all undiscovered before us.
ABBREVIATIONS A: atrial AR: atrial reversal CFD: computational fluid dynamics D: diastolic E: early EDPVR: end diastolic pressure-volume relationships FSI: fluid-structure interactions IVPG: intraventricular pressure gradient LA: left atrium LV: left ventricular MV: mitral valve P-V: pressure-volume PV: pulmonary vein(s) S: systolic Vp: flow propagation velocity τ: time constant of isovolumic pressure decay (Tau) t: time REFERENCES 1. Glantz SA, Parmley WW: Factors which affect the diastolic P-V curve. Circ Res 1978;42:171–180. 2. Nikolic S, Yellin EL, Tamura K, et al: Passive properties of canine left ventricle: Diastolic stiffness and restoring forces. Circ Res 1988;62: 1210–1222. 3. Thomas JD, Zhou J, Greenberg N, et al: Physical and physiological determinants of pulmonary venous flow: Numerical analysis. Am J Physiol 1997;272:H2453–2465. 4. Ohno M, Cheng CP, Little WC: Mechanism of altered patterns of left ventricular filling during the development of congestive heart failure. Circulation 1994;89:2241–2250. 5. Matsubara H, Nakatani S, Nagata S, et al: Salutary effect of disopyramide on left ventricular diastolic function in hypertrophic obstructive cardiomyopathy. J Am Coll Cardiol 1995;26:768–775. 6. Nikolic S, Yellin EL, Tamura K, et al: Effect of early diastolic loading on myocardial relaxation in the intact canine left ventricle. Circ Res 1990;66: 1217–1226. 7. Senzaki H, Fetics B, Chen CH, Kass DA: Comparison of ventricular pressure relaxation assessments in human heart failure: Quantitative influence on load and drug sensitivity analysis. J Am Coll Cardiol 1999;34: 1529–1536. 8. Scalia GM, Greenberg NL, McCarthy PM, et al: Noninvasive assessment of the ventricular relaxation time constant (t) in humans by Doppler echocardiography. Circulation 1997;95:151–155.
9. Popovic ZB, Prasad A, Garcia MJ, et al: Relationship between diastolic intraventricular pressure gradients, relaxation, and preload: Impact of age and fitness. Am J Physiol Heart Circ Physiol 2005. 10. Nikolic SD, Yellin EL, Dahm M, et al: Relationship between diastolic shape (eccentricity) and passive elastic properties in canine left ventricle. Am J Physiol 1990;259:H457–H463. 11. Notomi Y, Martin-Miklovic MG, Oryszak SJ, et al: Enhanced ventricular untwisting during exercise: A mechanistic manifestation of elastic recoil described by Doppler tissue imaging. Circulation 2006;113:2524–2533. 12. Thomas JD, Popovic ZB: Intraventricular pressure differences: A new window into cardiac function. Circulation 2005;112:1684–1686. 13. Thomas JD, Weyman AE: Echocardiographic Doppler evaluation of left ventricular diastolic function. Physics and physiology. Circulation 1991;84:977–990. 14. Courtois M, Kovacs SJ Jr, Ludbrook PA: Transmitral pressure-flow velocity relation: Importance of regional pressure gradients in the left ventricle during diastole. Circulation 1988;78:661–671. 15. Thomas JD, Newell JB, Choong CY, Weyman AE: Physical and physiological determinants of transmitral velocity: Numerical analysis. Am J Physiol 1991;260:H1718–H1731. 16. Nakatani S, Firstenberg MS, Greenberg NL, et al: Mitral inertance in humans: Critical factor in Doppler estimation of transvalvular pressure gradients. Am J Physiol Heart Circ Physiol 2001;280:H1340–H1345. 17. Hay I, Rich J, Ferber P, et al: Role of impaired myocardial relaxation in the production of elevated left ventricular filling pressure. Am J Physiol Heart Circ Physiol 2005;288:H1203–H1208. 18. Firstenberg MS, Greenberg NL, Smedira NG, et al: Doppler echo evaluation of pulmonary venous–left atrial pressure gradients: Human and numerical model studies. Am J Physiol Heart Circ Physiol 2000;279: H594–H600. 19. Kamohara K, Fukamachi K, Ootaki Y, et al: Evaluation of a novel device for left atrial appendage exclusion: The second-generation atrial exclusion device. J Thorac Cardiovasc Surg 2006;132:340–346. 20. Garcia MJ, Ares MA, Asher C, et al: An index of early left ventricular filling that combined with pulsed Doppler peak E velocity may estimate capillary wedge pressure. J Am Coll Cardiol 1997;29:448–454. 21. Firstenberg MS, Vandervoort PM, Greenberg NL, et al: Noninvasive estimation of transmitral pressure drop across the normal mitral valve in humans: Importance of convective and inertial forces during left ventricular filling. J Am Coll Cardiol 2000;36:1942–1949. 22. Kilner PJ, Yang GZ, Wilkes AJ, et al: Asymmetric redirection of flow through the heart. Nature 2000;404:759–761. 23. Vierendeels JA, Dick E, Verdonck PR: Hydrodynamics of color M-mode Doppler flow wave propagation velocity V(p): A computer study. J Am Soc Echocardiogr 2002;15:219–224. 24. Vierendeels JA, Riemslagh K, Dick E, Verdonck PR: Computer simulation of intraventricular flow and pressure gradients during diastole. J Biomech Eng 2000;122:667–674. 25. Verdonck P, Vierendeels J, Riemslagh K, Dick E: Left-ventricular pressure gradients: A computer-model simulation. Med Biol Eng Comput 1999;37:511–516. 26. Cheng Y, Oertel H, Schenkel T: Fluid-structure coupled CFD simulation of the left ventricular flow during filling phase. Ann Biomed Eng 2005;33:567–576. 27. Baccani B, Domenichini F, Pedrizzetti G, Tonti G: Fluid dynamics of the left ventricular filling in dilated cardiomyopathy. J Biomech 2002;35: 665–671. 28. Mirsky I: Assessment of diastolic function: Suggested methods and future considerations. Circulation 1984;69:836–841. 29. Thomas JD, Zhou J, Greenberg N, et al: Physical and physiological determinants of pulmonary venous flow: Numerical analysis. Am J Physiol 1997;272:H2453–H2465. 30. Lemmon JD, Yoganathan AP: Computational modeling of left heart diastolic function: Examination of ventricular dysfunction. J Biomech Eng 2000;122:297–303. 31. Takatsuji H, Mikami T, Urasawa K, et al: A new approach for evaluation of left ventricular diastolic function: Spatial and temporal analysis of left ventricular filling flow propagation by color M-mode Doppler echocardiography [see comments]. J Am Coll Cardiol 1996;27:365–371. 32. Garcia MJ, Smedira NG, Greenberg NL, et al: Color M-mode Doppler flow propagation velocity is a preload insensitive index of left ventricular relaxation: Animal and human validation. J Am Coll Cardiol 2000;35: 201–208. 33. Appleton CP, Hatle LK, Popp RL: Relation of transmitral flow velocity patterns to left ventricular diastolic function: New insights from a combined
Chapter 5 • Physical Determinants of Diastolic Flow
34. 35. 36. 37. 38.
39.
hemodynamic and Doppler echocardiographic study. J Am Coll Cardiol 1988;12:426–440. Stugaard M, Smiseth OA, Risoe C, Ihlen H: Intraventricular early diastolic filling during acute myocardial ischemia. Circulation 1993;88:2705– 2713. Steine K: Mechanisms of retarded apical filling in acute ischemic left ventricular failure. Circulation 1999;99:2048–2054. Steine K, Stugaard M, Smiseth OA: Mechanisms of retarded apical filling in acute ischemic left ventricular failure. Circulation 1999;99:2048– 2054. Steine K, Flogstad T, Stugaard M, Smiseth OA: Early diastolic intraventricular filling pattern in acute myocardial infarction by color M-mode Doppler echocardiography. J Am Soc Echocardiogr 1998;11:119–125. Beppu S, Izumi S, Miyatake K, et al: Abnormal blood pathways in left ventricular cavity in acute myocardial infarction. Experimental observations with special reference to regional wall motion abnormality and hemostasis. Circulation 1988;78:157–164. Delemarre BJ, Visser CA, Bot H, Dunning AJ: Prediction of apical thrombus formation in acute myocardial infarction based on left ventricular spatial flow pattern. J Am Coll Cardiol 1990;15:355–360.
40. Asada-Kamiguchi J, Jones M, Greenberg NL, et al: Intraventricular pressure gradients in left ventricular aneurysms determined by color M-mode Doppler method: An animal study. J Am Soc Echocardiogr 2006;19: 1112–1118. 41. Firstenberg MS, Smedira NG, Greenberg NL, et al: Relationship between early diastolic intraventricular pressure gradients, an index of elastic recoil, and improvements in systolic and diastolic function. Circulation 2001;104: I330–I335. 42. Ingels NB Jr, Hansen DE, Daughters GT 2nd, et al: Relation between longitudinal, circumferential, and oblique shortening and torsional deformation in the left ventricle of the transplanted human heart. Circ Res 1989;64: 915–927. 43. Gibbons Kroeker CA, Ter Keurs HE, Knudtson ML, et al: An optical device to measure the dynamics of apex rotation of the left ventricle. Am J Physiol 1993;265:H1444–H1449. 44. Bell SP, Nyland L, Tischler MD, et al: Alterations in the determinants of diastolic suction during pacing tachycardia. Circ Res 2000;87:235–240. 45. Nishimura RA, Tajik AJ: Evaluation of diastolic filling of left ventricle in health and disease: Doppler echocardiography is the clinician’s Rosetta Stone. J Am Coll Cardiol 1997;30:8–18.
59
HIDEKATSU FUKUTA, MD, PhD WILLIAM C. LITTLE, MD
6
General Principles, Clinical Definition, and Epidemiology INTRODUCTION PATHOPHYSIOLOGY Systolic Performance Diastolic Performance Definition of Systolic and Diastolic Heart Failure Terminology CLINICAL APPLICATIONS Diagnosis of Diastolic Heart Failure Timing of Ejection Fraction Measurement
Is Measurement of Diastolic Function Necessary? Prognostic Utility of Doppler Echocardiography Practical Recommendations EPIDEMIOLOGY Epidemiology of Diastolic Dysfunction Epidemiology of Diastolic Heart Failure CONCLUSIONS AND FUTURE RESEARCH
INTRODUCTION Heart failure (HF) is defined as the pathological state in which the heart is unable to pump blood at a rate required by the metabolizing tissues or can do so only with an elevated filling pressure. Inability of the heart to pump blood sufficiently to meet the needs of the body’s tissues may be due to the inability of the left ventricle to fill (diastolic performance) or eject blood (systolic performance) or both. Thus, consideration of the systolic and diastolic performance of the left ventricle provides a conceptual basis to classify and understand the pathophysiology of HF.
PATHOPHYSIOLOGY Systolic Performance Left ventricular (LV) systolic performance is the ability of the left ventricle to empty, which can be quantified as an emptying frac-
tion, or an ejection fraction (EF): a ratio of stroke volume-to-end diastolic volume. Thus, LV systolic dysfunction is defined as a decreased EF. The EF can be obtained by determining the LV volume by use of two-dimensional echocardiography with or without contrast, radionuclide ventriculography, or magnetic resonance imaging. The EF has been used as an index of myocardial contractile performance. However, it is influenced not only by myocardial contractility but also by LV afterload.1 Furthermore, in the presence of a left-sided valvular regurgitation (mitral or aortic regurgitation) or a left-to-right shunt (ventricular septal defect or patent ductus arteriosus), the LV stroke volume may be high, while the forward stroke volume (stroke volume minus regurgitant volume or shunt volume) is lower. Thus, the effective EF is defined as the forward stroke volume divided by end diastolic volume.2 The effective EF is a useful means to quantify systolic function, for two reasons: First, the effective EF represents the functional emptying of the left ventricle that contributes to cardiac 63
Chapter 6 • General Principles, Clinical Definition, and Epidemiology output. Second, the effective EF is relatively independent of LV end diastolic volume over the clinically relevant range. An operational definition of systolic dysfunction is an effective EF of less than 0.50.2 When defined in this manner, systolic dysfunction results from impaired myocardial function, increased LV afterload, structural abnormalities of the left ventricle, or a combination thereof.2
Diastolic Performance For the left ventricle to function effectively as a pump, it must be able not only to eject but also to fill, which is its diastolic function. Diastolic function conventionally has been assessed on the basis of the LV end diastolic pressure-volume (P-V) relation (see Chapter 7).3 A shift of the curve upward and to the left has been considered to be the hallmark of diastolic dysfunction (Fig. 6-1, curve A). In this situation, each LV end diastolic volume is associated with a high end diastolic pressure, and thus the left ventricle is less distensible. Decreased LV distensibility is caused by aging, systemic hypertension, and hypertrophic or restrictive cardiomyopathy.2 Diastolic function also has been assessed based on LV filling patterns by use of Doppler echocardiography.4,5 In the absence of mitral stenosis, three patterns of LV filling indicate progressive impairment of diastolic function: (1) reduced early diastolic filling with a compensatory increase in importance of atrial filling (impaired relaxation); (2) most filling early in diastole but with rapid deceleration of mitral flow (pseudonormalization); and (3) almost all filling of the left ventricle occurring very early in diastole in association with very rapid deceleration of mitral flow (restrictive filling) (Fig. 6-2).6 These Doppler LV filling patterns, however, are influenced not only by LV diastolic properties but also by left atrial (LA) pressure. In contrast, tissue Doppler measurement of mitral annular velocity and color M-mode measurement of the velocity of propagation of mitral inflow to the apex appear to be less load sensitive. The peak early diastolic mitral annular velocity (EM) provides a relatively load insensitive measure of LV relaxation.7 EM is decreased with increasing severity of diastolic dysfunction.8,9 The color M-mode imaging performed from the apex provides a temporal and spatial map of the velocities of blood flow in early diastole along the long axis of the left ventricle. The velocity of
LV end diastolic pressure
64
A Normal
B
ΔP ΔV
LV end diastolic volume Figure 6-1 A shift of the curve to A indicates that a higher left ventricular (LV) pressure will be required to distend the LV to a similar volume, indicating that the ventricle is less distensible. The slope of the LV end diastolic pressure-volume relation indicates the passive chamber stiffness. Since the relation is exponential in shape, the slope (ΔP/ΔV) increases as the end diastolic pressure increases. (Data from Little WC: Diastolic dysfunction beyond distensibility: Adverse effects of ventricular dilatation. Circulation 2005;112:2888–2890.)
propagation of mitral inflow to the apex (VP) is reduced in conditions with impaired LV relaxation.9 A pseudonormalized LV filling pattern can be distinguished from a normal filling pattern by demonstrating reduced EM and VP. Furthermore, since early diastolic mitral inflow velocity (E) becomes higher and relaxation-related parameters (EM and Vp) remain reduced as filling pressure increases, the E/EM and E/VP can estimate LV filling pressure with a reasonable accuracy over a wide range of an EF.10–12 Last, color M-mode imaging provides a noninvasive measurement of the diastolic intraventricular pressure gradient between the apex and the base during early diastole.13 Finally, analysis of pulmonary venous flow patterns provides useful information on LV compliance and LA pressure.14 With increase in LV end diastolic pressure, the reversal velocity of pulmonary venous atrial flow increases, and duration increases longer than that of mitral late diastolic velocity. With decrease in LV compliance and increase in mean LA pressure, the systolic component of pulmonary venous flow decreases and the diastolic component of pulmonary venous flow increases. Table 6-1 and Figure 6-2 show stages of diastolic dysfunction incorporatingpulmonary venous flow, tissue Doppler, and color M-mode indices.9
Definition of Systolic and Diastolic Heart Failure When the HF is associated with a reduced EF, the pathological state may be called systolic HF. In contrast, when the HF is associated with diastolic dysfunction in the absence of a reduced EF, the pathological state may be called diastolic HF. It is important to recognize that the HF, whether it results from systolic or diastolic dysfunction, is a clinical syndrome and that both systolic HF and diastolic HF are heterogeneous disorders. Patients with systolic HF have abnormalities of diastolic function,2,3 and those with diastolic HF may have abnormalities of systolic contractile function that are not detected by measurement of an EF.15
Terminology Diastolic Dysfunction and Diastolic Heart Failure The term diastolic dysfunction is used to describe abnormal mechanical (diastolic) properties of the ventricle and includes decreased LV distensibility, delayed relaxation, and abnormal filling, regardless of whether the EF is normal or reduced and whether the patient is symptomatic or asymptomatic. In contrast, the term diastolic HF is used to describe patients with the symptoms and signs of HF and a normal EF and diastolic dysfunction.
Diastolic Heart Failure and Heart Failure with Normal Ejection Fraction There are many clinical conditions that cause HF with a normal EF. These include diastolic dysfunction, valvular diseases, pericardial diseases, and intracardiac mass; among them, diastolic dysfunction is the most common cause of HF with a normal EF. Diastolic HF is associated with LV diastolic abnormalities. It is important to recognize that although patients with diastolic HF have diastolic dysfunction, frequently they also have systolic contractile abnormalities (despite the normal EF). In addition, comorbidities such as hypertension, anemia, and renal dysfunction are commonly seen in patients with diastolic HF and may contribute to the development of HF.
Chapter 6 • General Principles, Clinical Definition, and Epidemiology Normal
Impaired relaxation
E
E
Pseudonormal
Restrictive E
E Mitral inflow velocity
A
A
A A
D S
S
D
D
D
Figure 6-2 Transmitral inflow velocity, pulmonary vein flow velocity, mitral annular velocity, and color M-mode imaging in stages of diastolic dysfunction. E, early diastolic mitral inflow velocity; A, late diastolic mitral inflow velocity; S, systolic pulmonary vein velocity; D, diastolic pulmonary vein velocity; SM, systolic mitral annular velocity; EM, early diastolic mitral annular velocity; AM, late diastolic mitral annular velocity; and Vp, velocity of propagation of mitral inflow to the apex. (Data from Fukuta H, Little WC: Diastolic versus systolic heart failure. In Smiseth OA, Tendera M (eds): Diastolic Heart Failure. London, Springer, 2007.)
S
S
Pulmonary vein velocity SM
SM
SM
SM
Mitral annular velocity EM EM Color M-mode imaging
AM
AM
EM
AM
EM
AM
VP
TABLE 6-1 STAGES OF DIASTOLIC DYSFUNCTION PARAMETER
NORMAL (YOUNG)
NORMAL (ADULT)
DELAYED RELAXATION
PSEUDONORMAL FILLING RESTRICTIVE FILLING
E/A DT (msec) IVRT (msec) S/D AR (cm/sec) Vp (cm/sec) Em (cm/sec)
>1 122 g/m2 ( ); >149 g/m2 ( ) or Atrial fibrillation
TD E/E' >8
HFNEF
10. Nagueh SF, Middleton KJ, Kopelen HA, et al: Doppler tissue imaging: A noninvasive technique for evaluation of left ventricular relaxation and estimation of filling pressures. J Am Coll Cardiol 1997;30:1527– 1533. 11. Nagueh SF, Mikati I, Kopelen HA, et al: Doppler estimation of left ventricular filling pressure in sinus tachycardia. Circulation 1998;98:1644– 1650. 12. Ommen SR, Nishimura RA, Appleton CP, et al: Clinical utility of Doppler echocardiography and tissue Doppler imaging in the estimation of left ventricular filling pressures: A comparative simultaneous Dopplercatheterization study. Circulation 2000;102:1788–1794. 13. Greenberg NL, Vandervoort PM, Firstenberg MS, et al: Estimation of diastolic intraventricular pressure gradients by Doppler M-mode echocardiography. Am J Physiol 2001;280:H2507–H2515. 14. Oh JK, Appleton CP, Hatle LK, et al: The noninvasive assessment of left ventricular diastolic function with two-dimensional and Doppler echocardiography. J Am Soc Echocardiogr 1997;10:246–270. 15. Brucks S, Little WC, Chao T, et al: Contribution of left ventricular diastolic dysfunction to heart failure regardless of ejection fraction. Am J Cardiol 2005;95:603–606. 16. Kitzman DW, Little WC, Brubaker PH, et al: Pathophysiological characterization of isolated diastolic heart failure in comparison to systolic heart failure. JAMA 2002;288:2144–2150. 17. Zile MR, Brutsaert DL: New concepts in diastolic dysfunction and diastolic heart failure. Part I: Diagnosis, prognosis, and measurement of diastolic function. Circulation 2002;105:1387–1393.
71
72
Chapter 6 • General Principles, Clinical Definition, and Epidemiology 18. European Study Group on Diastolic Heart Failure. How to diagnose heart failure. Eur Heart J 1998;20:990–1003. 19. Vasan RS, Levy D: Defining diastolic heart failure: A call for standardized diagnostic criteria. Circulation 2000;101:2118–2121. 20. Gandhi SK, Powers JC, Nomeir AM, et al: The pathogenesis of acute pulmonary edema associated with hypertension. N Engl J Med 2001;344:17–22. 21. Little WC: Hypertensive pulmonary oedema is due to diastolic dysfunction. Eur Heart J 2001;22:1961–1964. 22. Zile MR, Gaasch WH, Carroll JD, et al: Heart failure with a normal ejection fraction: Is measurement of diastolic function necessary to make the diagnosis of diastolic heart failure? Circulation 2001;104:779–782. 23. Zile MR, Baicu CF, Gaasch WH: Diastolic heart failure—abnormalities in active relaxation and passive stiffness of the left ventricle. N Engl J Med 2004;350:1953–1959. 24. Vanoverschelde JL, Raphael DA, Robert AR, et al: Left ventricular filling in dilated cardiomyopathy: Relation to functional class and hemodynamics. J Am Coll Cardiol 1990;15:1288–1295. 25. Rihal CS, Nishimura RA, Hatle LK, et al: Systolic and diastolic dysfunction in patients with clinical diagnosis of dilated cardiomyopathy: Relation to symptoms and prognosis. Circulation 1994;90:2772–2779. 26. Wang M, Yip G, Yu CM, et al: Independent and incremental prognostic value of early mitral annulus velocity in patients with impaired left ventricular systolic function. J Am Coll Cardiol 2005;45:272–277. 27. Dokainish H, Zoghbi WA, Lakkis NM, et al: Incremental predictive power of B-type natriuretic peptide and tissue Doppler echocardiography in the prognosis of patients with congestive heart failure. J Am Coll Cardiol 2005;45:1223–1226. 28. Troughton RW, Prior DL, Frampton CM, et al: Usefulness of tissue doppler and color M-mode indexes of left ventricular diastolic function in predicting outcomes in systolic left ventricular heart failure (from the ADEPT Study). Am J Cardiol 2005;96:257–262. 29. Yturralde RF, Gaasch WH: Diagnostic criteria for diastolic heart failure. Prog Cardiovasc Dis 2005;47:314–319.
30. Redfield MM, Jacobsen SJ, Burnett JC Jr., et al: Burden of systolic and diastolic ventricular dysfunction in the community: Appreciating the scope of the heart failure epidemic. JAMA 2003;289:194–202. 31. Fischer M, Baessler A, Hense HW, et al: Prevalence of left ventricular diastolic dysfunction in the community. Results from a Doppler echocardiographic-based survey of a population sample. Eur Heart J 2003;24: 320–328. 32. Aurigemma GP, Gottdiener JS, Shemanski L, et al: Predictive value of systolic and diastolic function for incident congestive heart failure in the elderly: The cardiovascular study. J Am Coll Cardiol 2001;37:1042–1048. 33. Bella JN, Palmieri V, Roman MJ, et al: Mitral ratio of peak early to late diastolic filling velocity as a predictor of mortality in middle-aged and elderly adults: The Strong Heart Study. Circulation 2002;105:1928–1933. 34. Wachtell K, Bella JN, Rokkedal J, et al: Change in diastolic left ventricular filling after one year of antihypertensive treatment: The Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) Study. Circulation 2002;105:1071–1076. 35. Owan TE, Redfield MM: Epidemiology of diastolic heart failure. Prog Card Dis 2005;47:320–332. 36. Senni M, Tribouilloy CM, Rodeheffer RJ, et al: Congestive heart failure in the community: A study of all incident cases in Olmsted County, Minnesota, in 1991. Circulation 1998;98:2282–2289. 37. Vasan RS, Larson MG, Benjamin EJ, et al: Congestive heart failure in subjects with normal versus reduced left ventricular ejection fraction: Prevalence and mortality in a population-based cohort. J Am Coll Cardiol 1999; 33:1948–1955. 38. Cortina A, Reguero J, Segovia E, et al: Prevalence of heart failure in Asturias (a region in the north of Spain). Am J Cardiol 2001;87:1417– 1419. 39. Bursi F, Weston SA, Redfield MM, et al: Systolic and diastolic heart failure in the community. JAMA 2006;296:2209–2216. 40. Bhatia RS, Tu JV, Lee DS, et al: Outcomes of heart failure with preserved ejection fraction in a population-based study. N Engl J Med 2006;335: 260–269.
JAMES B. SEWARD, MD KRISHNASWAMY CHANDRASEKARAN, MD MARTIN OSRANEK, MD, MSc KANIZ FATEMA, MBBS, PhD TERESA S. M. TSANG, MD
7
Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction INTRODUCTION PATHOPHYSIOLOGY Clinical Diastolic Dysfunction Pressure and Volume INVASIVE VERSUS NONINVASIVE ASSESSMENT Systolic Function Diastolic Function STATE-OF-THE-ART INVASIVE AND NONINVASIVE PHYSIOLOGY Myocardial Relaxation Ventricular Stiffness Filling Pressures Comprehensive Diastolic Function Assessment NORMAL DIASTOLIC MYOCARDIAL PHYSIOLOGY Temporal-Spatial Pressure and Flow: Suction of Blood into the Left Ventricle Left Atrial Pressure and Left Ventricular Suction
CLINICAL RELEVANCE Epidemic of Diastolic Heart Failure Myocardial Disease and Diastolic Dysfunction Arterial Stiffening and Diastolic Dysfunction Physiologic Coupling of the Cardiovascular System Forward Dysfunction: End-Organ and Microvascular Damage Vascular Stiffening: Summary CLUSTERING OF CARDIOVASCULAR RISK Diastolic Heart Failure With Normal Ejection Fraction and Systolic Heart Failure: Pathophysiologic Subgroups FUTURE RESEARCH GLOSSARY ABBREVIATIONS
73
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction
INTRODUCTION In the era of classical cardiovascular (CV) medicine (ending around the year 2000), CV function was validated predominantly by invasive means.1,2 Classic studies established the end systolic and end diastolic pressure-volume relationships as a meaningful and useful way of characterizing intrinsic ventricular pump properties.3 These concepts then evolved such that they could be applied to basic and clinical research. In the present era of CV intervention (projected to last until approximately 2020), CV imaging and molecular science are emphasized. Mathematical modeling and further clarification of diastolic function through the use of invasive and noninvasive technologies is evolving.2,4 Around 2020, CV medicine is envisioned to evolve toward an “era of CV disease prevention.”5 Prevention will emphasize system physiology complemented by mathematical probability. A portfolio of physiologic models will be used as surrogate expressions of disease, and primary prevention of disease will be based on the intensity or burden of a physiologic or molecular/chemical risk model. Information acquisition will move from invasive to noninvasive. Research and clinical physiologic approaches2,6,7 are currently evolving from invasive to noninvasive means to elucidate fundamental CV physiology.4,8,9 Clinical physiology is no longer limited to catheterization, and the noninvasive echocardiologist can be redefined as an “echophysiologist.” This chapter attempts to relate classical invasive physiology to the current evolution toward Doppler echocardiographic physiology. It is increasingly recognized and emphasized that contraction (systolic function) is dependent on relaxation (diastolic function). It is important to appreciate the two distinct, although intimately interrelated, aspects of the assessment of cardiac properties: (1) of the ventricle as a hemodynamic pump and (2) of the intrinsic properties of the cardiac muscle.3 Systolic and diastolic ventricular properties are dependent on systolic and diastolic myocardial properties, as well as muscle mass, chamber architecture, and chamber geometry. Pressure-volume constructs are used to directly assess ventricular properties.
PATHOPHYSIOLOGY Cardiac dysfunction, with or without systolic dysfunction, is associated with diastolic dysfunction (i.e., elevation of filling pressure). Cardiac dysfunction can be broadly viewed as a derivative of altered cardiac geometry and cardiomyocyte dysfunction or altered extracardiac loading, which is most commonly attributed to arterial stiffening. Once the sentinel cardiac or arterial perturbation is initiated, the dysfunctional state is propagated throughout the contiguous CV system (arteries, pump, and reservoir), becomes cyclical and self-perpetuating, and accounts for the clustering of a portfolio of common adverse events (e.g., heart failure [HF], atrial fibrillation, stroke, cognitive dysfunction).
Clinical Diastolic Dysfunction Diastolic dysfunction refers to abnormal mechanical relaxation (diastolic) properties of the ventricle, which are present in the majority of patients with congestive heart failure (CHF). The term diastolic heart failure (DHF) is conventionally used to
describe patients with CHF symptoms and normal systolic contractility (i.e., normal ejection fraction [EF]), although a preferred term would be HF associated with preserved systolic function. Thus, the distinction between DHF and diastolic dysfunction is merely the presence of CHF symptoms. The CHF symptom complex associated with systolic HF (SHF) or DHF is most strongly correlated with the elevation of filling pressure. In order for the left ventricle to function as an effective pump, it must be able to empty and fill without requiring abnormal elevation of left atrial (LA) pressure. Atrial pressure is directly related to the ability of the left atrium to expel its contents into the ventricle during diastole. Thus, the principal determinant of risk and symptoms of CHF is the increased pressure reflected backward into the left atrium and pulmonary veins during diastolic filling of the ventricle.10 The stroke volume must be able to increase in response to stress, such as exercise, without any appreciable increase in LA pressure (Fig. 7-1).6 Drugs that produce a sustained decrease in ventricular filling pressures enhance effort tolerance, breathing capacity, and HF outcome.11
Pressure and Volume Diastolic function conventionally has been assessed on the basis of the left ventricular (LV) diastolic pressure-volume relationship and an upward shift of the curve or increase in LV end diastolic pressure (LVEDP) (Figs. 7-2 and 7-3).6,7 The LV end diastolic pressure-volume relationship is the most important means of characterizing global filling properties. The pressure-volume loop (see Fig. 7-2) depicts the amount of diastolic filling (ml of blood) relative to a specified filling pressure (mmHg) and is therefore a key physiologic determinant of preload.13 Any delay or impedance
200 Exercise
150 LV pressure (mmHg)
74
Rest
100
50
0
20
28
36
44
52
LV volume (ml) Figure 7-1 Left ventricular (LV) pressure-volume loops at rest and during exercise. Each loop was generated by averaging the data obtained during a 15-second recording of simultaneous pressure and volume, spanning several respiratory cycles. During exercise, the early diastolic portion of the LV pressure-volume loop is shifted downward (arrow) so that the early diastolic LV pressure is lower during exercise than at rest. (From Cheng CP et al: Mechanism of augmented rate of left ventricular filling during exercise. Circ Res 1992;70:9–19. Used with permission.)
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction 160 160
Ejection
120
Isovolumic relaxation
Isovolumic contraction
80
Pressure (mmHg)
120 Pressure (mmHg)
ESPVR
End systole
End diastole
20
40
A
60
80
100
120
0
140
B
Volume (ml)
EDPVR
0
0 0
Ees
40
40 Filling
80
20 Vo
40
60
80
100
120
140
Volume (ml)
160
Pressure (mmHg)
120
80
40
0 0
C
20
40
60
80
100
120
140
Volume (ml)
of myocardial relaxation will increase the filling pressure, which will be reflected backward into the left atrium and pulmonary veins. The end diastolic pressure-volume relationship is intrinsically nonlinear (see Fig. 7-3), a characteristic attributed to the different types of structural fibers being stretched in different pressurevolume ranges.14 In the low pressure-volume range, with only a small increase in pressure for a given increment in volume, the stretch of compliant elastin fibers and of the myocytes with sarcomeric titin molecules15 is believed to account for stiffness. As volume is increased to a higher range, pressure rises more steeply as slack lengths of collagen fibers and titin are exceeded and stretch is more strongly resisted by these stiff elements. Therefore, chamber stiffness increases as end diastolic pressure or volume increases.3 With diastolic dysfunction, any abnormal increase in diastolic filling pressure corresponds to a less distensible or less compliant ventricle during the filling phase of the cardiac cycle. Consequently, a “stiff ” ventricle is less able to increase its stroke volume without further elevation of LA pressure. Pressure reflected backward through the open mitral valve into the atrium and pulmonary veins can cause shortness of breath, elevation of pulmonary
Figure 7-2 The hemodynamic events occurring during the cardiac cycle are displayed by plotting instantaneous pressure versus volume. A, A normal pressure-volume loop. The four phases of the cardiac cycle are displayed on the pressure-volume loop, which is constructed by plotting instantaneous pressure and volume. The loop repeats with each cardiac cycle and shows how the heart transitions from its end diastolic state to the end systolic state and back. B, Decrease in filling pressure (preload). With a constant contractile state and afterload resistance, a progressive reduction in ventricular filling pressure causes the loops to shift toward lower volumes at both end systole and end diastole. When the resulting end systolic pressure-volume points are connected, a reasonably linear end systolic pressure-volume relationship (ESPVR) is obtained. The linear ESPVR is characterized by a slope (Ees, ventricular elastance) and a volume axis intercept (Vo). EDVPR denotes end diastolic pressure-volume relationship. C, Increased afterload (blood pressure). When afterload resistance is increased at a constant preload, the loops get narrower and longer, and under ideal conditions the end systolic pressure-volume points fall on the same ESPVR as obtained with preload reduction. (From Burkhoff D et al: Assessment of systolic and diastolic ventricular properties via pressure-volume analysis: A guide for clinical, translational, and basic researchers. Am J Physiol Heart Circ Physiol 2005;289:501–512.)
venous pressure, secondary pulmonary hypertension, and decrease in clinical exercise capacity.15–18
INVASIVE VERSUS NONINVASIVE ASSESSMENT Systolic Function Cardiac function assessment traditionally has focused on decreased contractility and the measure of EF. The noninvasive measure of systolic function, EF, is almost exclusively used in clinical trials and medical practice. Although EF is clinically considered an important index of LV systolic function, it is not exclusively governed by LV properties. Rather, EF is determined by the interaction of arterial and ventricular properties.20–23 A normal EF suggests that contractility and arterial stiffness are well matched. EF is very load dependent and normally increases up to 30% during stress.24 EF is thus limited by many factors, including marked preload dependence25 and low reproducibility when measured by different imaging modalities.26 Furthermore, EF greater than 45% does not contribute to assessment of CV risks in patients with HF.27 In patients with CHF and elevated filling pressure, diastolic dysfunction correlates better with B-type natriuretic peptde concentration and mortality than with EF or LV volume.12,26
75
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction
A Normal
LV end diastolic pressure
76
B
ΔP
ΔP ΔV
ΔP
ΔV
ΔV
LV end diastolic volume Figure 7-3 Left ventricular (LV) end diastolic pressure-volume curves showing the relationship between end diastolic volume and pressure. Stiffness, at different points on the curve, is reflected as the change in pressure (ΔP) divided by the change in volume (ΔV). During normal LV filling, the myocardium must efficiently relax in order to prevent resistance to atrial emptying and elevation of atrial pressure. As the left ventricle relaxes, the diastolic volume and pressure increase. The ventricular filling pressure is reflected backward into the left atrium. Flow ultimately stops or transiently reverses when the atrial and ventricular pressures equilibrate. Normal atrial emptying thus must occur without requiring significant elevation of left atrial pressure. In this early period of diastole more than two thirds of the LV stroke volume normally enters the left ventricle. The severity of heart failure and prognosis are related to the severity of filling abnormalities regardless of the ejection fraction.12 A shift of the curve upward and to the left is considered the hallmark of diastolic dysfunction (curve A). A “stiff ” left ventricle in early diastole disturbs the normal suction gradient between the left ventricle and the left atrium (see Fig. 7-6C). The LV pressure increases rapidly, thus impeding the normal filling of the ventricle. Pressure is reflected backward into the left atrium, increasing atrial wall stress and pulmonary venous pressure. This pressure-volume curve is typical for isolated diastolic dysfunction with preserved ejection fraction and delayed relaxation, elevated diastolic pressure, and near-normal ventricular volume. In a heart with abnormal geometry, as found in dilated, ischemic, or hypertrophic cardiomyopathy, the LV end diastolic pressure-volume curve is shifted to the right (curve B). Decreased diastolic suction (see Fig. 7-6B) is disturbed because of abnormal chamber geometry, which is commonly associated with uncoordinated wall motion or impaired myocardial contractility. The disturbed LV suction forces the left ventricle and left atrium to function at a higher filling pressure, decreasing ventricular filling and increasing atrial and pulmonary venous pressure. (From Little WC: Diastolic dysfunction beyond distensibility: Adverse effects of ventricular dilatation. Circulation 2005;112:2888–2890.)
Diastolic Function Although the invasive research “testing ground” has made important contributions to the understanding of pressure-volume relationships and diastolic physiology,3,10 only recently has assessment of diastolic function been extensively incorporated into clinical practice. Assessment of diastolic function was neglected because there was no reliable noninvasive method to assess diastolic dysfunction, and invasive confirmation of elevated filling pressure was deemed impractical if not unethical. Invasive monitoring through a pulmonary artery catheter or implantable transducer can provide continuous data. Pulmonary artery catheters, however, appear to contribute to morbidity and mortality29 and contribute little to clinical assessment.30 Even when LV cardiac catheterization is performed, the principal measure of diastolic function is LVEDP, which measures only the load-dependent filling pressure and reflects conditional physiologic circumstances only at the time of data acquisition. With the advent of noninvasive Doppler echocardiography, clinical assessment of diastolic function has become increasingly attainable and validated. However, early impediments to clinical use of Doppler echocardiographic diastolic function assessment became apparent. The first measures of diastolic function, which relied on mitral and pulmonary vein flow, were load dependent—similar to EF and LVEDP—and changed dynamically even during the examination. Because of the dynamic nature of these physiologic events, measures of diastolic function were limited for use in the assessment of long-term morbidity and mortality.
STATE-OF-THE-ART INVASIVE AND NONINVASIVE PHYSIOLOGY Diastolic function assessment by both invasive2 and noninvasive means is plagued by ambiguities surrounding the dynamic nature, indirectness, and difficulty of reproducibly measuring everchanging diastolic function. Dynamic measures include: 1. Decrease in pressure during the period of isovolumic relaxation (invasive: time constants of relaxation; noninvasive: deceleration time and longitudinal fiber relaxation) 2. Passive and active chamber stiffness or, its inverse, compliance (invasive: elastance; noninvasive: diastolic grade and LA volume) 3. End diastolic pressure (invasive: LVEDP; noninvasive: Doppler E/E′ ratio, end diastolic mitral versus pulmonary vein Doppler flow) (Table 7-1)
Myocardial Relaxation Invasive Assessment Pressure relaxation can be measured invasively by placing a micromanometer catheter into the left ventricle. The rate of pressure decrease, which begins during the isovolumic period, is influenced by several physiologic factors2: 1. Capacity and rapidity with which the sarcoplasmic reticulum and sarcolemmal Na+/Ca2+ exchanger restores cellular Ca2+ levels to low diastolic levels
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction TABLE 7-1 DIASTOLIC FUNCTION ASSESSMENT BY INVASIVE AND NONINVASIVE MEANS METHOD Functional parameter Myocardial relaxation Ventricular and arterial stiffness
Invasive Mathematical models of pressure-volume loops2 Analysis of viscoelastic properties2: simultaneous measurement of pressure, volume, and flow Vascular: pulse pressure, central blood pressure
Filling pressure
Left ventricular end diastolic pressure40
2. Extent of elastic recoil (suction) within the heart caused by compression by myocyte contraction during the prior systole and history of prior systolic ejection, which is largely determined by the aortic input impedance (stiffness) 3. Other chamber-level factors, such as dyscoordination and heterogeneity Isovolumic pressure relaxation traditionally starts when pressure decay is maximal, generally occurring shortly after aortic valve closure, and extends to a point when LV pressure decreases just below a threshold equal to the end diastolic pressure plus 5 mmHg. Recorded pressure-time curves can be mathematically analyzed to extract the rate of pressure decrease.2 However, the type of model used to describe relaxation can have marked influences on the apparent changes with drug and chamber-loading interventions.2 Model-dependent analysis of relaxation is commonly fraught with systematic deviations from the real data, particularly with depressed cardiac function.30 Some diseases, including hypertrophic cardiomyopathy, are characterized by regional heterogeneity of loading and material properties, which result in nonuniform relaxation. Diseases with a complex process of pressure decrease, such as hypertrophic cardiomyopathy, in general cannot be easily expressed in a single time constant.
Noninvasive Assessment Diastolic dysfunction is associated with substantial abnormalities of active relaxation and passive stiffness of the ventricle.31 Doppler transmitral velocity deceleration time (DT) and tissue Doppler early diastolic mitral annular velocity (E′) are commonly used noninvasive indices of myocardial relaxation. DT fluctuates with the dynamic changes in ventricular loading pressure, whereas E′ is relatively load independent. DT is useful for estimating intensity32,33 or “grade” of dysfunction34 at the time of the examination. Tissue Doppler E′ velocity is touted as less load dependent and can distinguish pseudonormal from normal filling patterns. The pseudonormal filling pattern is commonly encountered when using mitral inflow Doppler to characterize diastolic function grade.32 Active (e.g., Ca2+-dependent) and passive (e.g., fibrosis) structural mechanisms regulate the rates of ventricular pressure decay and early diastolic ventricular filling.35 Active relaxation represents the speed of transition from the contracted state (systole) to the relaxed state (diastole) and is strongly related to the uptake of calcium for the contracted myocyte. Passive relaxation is most strongly related to myocardial compliance and the effects of tissue fibrosis. Load-dependent measures reflect the dynamic physio-
Noninvasive ultrasonography Tissue Doppler annular velocity, E′31–33 Measures of “stiffness”9,34–36: effective arterial elastance (Ea), end systolic elastance (Ees); Doppler-derived diastolic dysfunction. Vascular stiffness37–39: pulse wave velocity, augmentation index Doppler E/E′31,41–45: acute load-dependent filling pressure Left atrial volume46–48: chronic average increase in filling pressure
logic state of filling pressure under the conditions existing at the time of the examination as well as the likelihood of a favorable response to a therapeutic manipulation (modifiability).32,36–40 The tissue Doppler E′ velocity reflects the relaxation velocity of the myocardium and expresses both passive and active myocardial relaxation.31 Variations in long-axis shortening have relatively little effect on the EF or stroke volume. Most of the stroke volume is produced by shortening of the minor axis dimension,41 which explains why either decreased diastolic (E′) or decreased systolic annular tissue Doppler can be observed in the presence of a normal EF.
Ventricular Stiffness Invasive Assessment Ideally, to analyze the viscoelastic properties of the ventricle, one should take simultaneous measurements of chamber pressure, volume, and flow. Passive elastic properties are those that relate a change in pressure to a given change in volume, whereas viscous properties are those that relate change in pressure to the rate of change in volume or flow. As opposed to noninvasive Doppler, no invasive methods are readily applicable for measuring flow rates of filling in the human heart. Therefore, most analyses must obtain flow data from the derivative of volume.2 The invasive measurement of pressure uses a high-fidelity micromanometer catheter. Measurement of volume is more challenging, with investigators commonly using various imaging-based methods to construct a volume-time curve. This approach is labor intensive, typically yielding a limited number of samplings of volumes during the diastolic period, and is subject to various sources of noise given the extensive processing required.2 A notable alternative is intracardiac electrical impedance, which can serve as a continuous measure of chamber volume.42–45 The association between diastolic pressure and volume from a single cardiac cycle does not necessarily follow a simple mathematic relationship.2 Only limited filling property data can be obtained from a single cardiac cycle. Various solutions have been used to overcome these limitations. After a diastolic pressurevolume curve is obtained, these data are typically subjected to a mathematical curve-fitting process to extract parameters of chamber or muscle stiffness. As with relaxation analysis, there are important limitations of such fitting when applied to a diastolic curve, owing to the nature of the mathematics.2 Unless the diastolic data are perfectly monoexponential, which is rarely the case, simply evaluating the same curve or restricting analysis to slightly
77
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction loading amplification associated with age.48,55 Diastolic HF with preserved EF appears to be best correlated with increased ventricular-arterial stiffness. Relaxation of the “stiff ” ventricular myocardium is delayed when the ventricle is exposed to increased systolic pressure during ejection (i.e., increased afterload).58 Diastolic dysfunction, as reflected in increased filling pressure, is directly related to greater prolongation of diastolic relaxation and increased ventricular-vascular stiffening.55,59 A major hemodynamic consequence of arterial stiffening is widening of the arterial pulse pressure, which also increases cyclic changes in arterial flow. Increased arterial stiffness increases systolic pressure and decreases diastolic pressure. Coronary perfusion occurs principally during diastole; thus, in persons with stiff conduit vessels and lower diastolic pressure, coronary microvascular perfusion is adversely affected.60 Impaired myocyte relaxation of the subendocardial longitudinal fibers is an ageassociated finding. Reduced myocardial relaxation is best measured as a decrease in tissue Doppler E′ velocity. Structural remodeling and functional disturbances are essential to the development of diastolic dysfunction and HF, but acute or subacute decompensation with CHF generally requires a precipitant, or “trigger.”61 The stiff arterial system is particularly sensitive to acute dysfunction, as occurs with excessive sodium intake, hypertension, atrial fibrillation, myocardial infarction, or marked bradycardia (prolonged diastasis). High arterial pulsatility is noted to have detrimental effects on blood flow regulation62 and contributes to adverse mechanical forces, which affect vascular tone, atherogenesis, angiogenesis, hemostasis, and microvascular autoregulation.
different ranges of volume and pressure can influence the derived parameters. Ultimately, derived stiffness parameters must be viewed cautiously, particularly if the data themselves do not clearly follow an exponential rise waveform.2 Even if directly measured by invasive pressures and volumes, the relationship between the two variables during filling does not necessarily provide information solely related to the cardiac chamber.
Noninvasive Assessment Limited success has been achieved in replicating pressure-volume loops by noninvasive Doppler echocardiography.8,46,47 However, several noninvasive studies have identified a relationship between limitation of exercise capacity, increased CV stiffness,48–51 and DHF.52,53 Noninvasive Doppler echocardiographic assessment of “stiffness” appears to be an excellent means of expressing ventricular and arterial stiffening.9,54 Effective arterial elastance (Ea), which reflects vascular stiffening, and end systolic elastance (Ees), which reflects ventricular stiffening, can be calculated noninvasively9,54 and validated against invasive assessment (Fig. 7-4).21,46,54–56 Ea, Ees, and diastolic stiffness increase with age and correlate with one another.21 Normally, there is an inverse relationship between Ees and Ea. As arterial compliance decreases (increased stiffness), ventricular compliance would be expected to increase (decreased stiffness).55 In order for EF to increase during exercise, the coupling ratio of Ea and Ees must decrease. However, with aging, the increase in exercise EF becomes blunted,23,57 suggesting ageassociated differences in the coupling ratio (Ea/Ees) and its components. Both arterial and ventricular stiffness tend to increase to a similar degree—Ea/Ees remains constant despite an increase in arterial stiffness (see Fig. 7-4). Thus, ventricular stiffness tends to increase with age. As ventricular and arterial stiffness increase, the system becomes more pressure labile and sensitive to loading conditions.21 This physiologic change is most evident with aging, as reflected in increased systolic and pulse pressure after age 50 to 60 years. These changes explain decreased exertional capacity and pressure-
Invasive Assessment LVEDP is the most common measure of diastolic dysfunction recorded during clinical cardiac catheterization. Increased filling pressure is not a requisite of diastolic dysfunction but is a universal
Elderly
150
180
Ea
Ees Ees
Filling Pressures
Ea
100
50
Systolic pressure (mmHg)
Young
LV pressure (mmHg)
78
0
150 120
90
Young
60 0
A
Elderly
50
100
LV volume (ml)
0
50
40
100
LV volume (ml)
B
60
80
100 120
LV EDV
Figure 7-4 Graphs showing the relationship between left ventricular (LV) pressure and volume. A, Pressure-volume loops from a 19-year-old man (Young) and an 87-year-old woman (Elderly) show increases in ventricular elastance or stiffness (Ees, solid line) that match increases in arterial elastance or stiffness (Ea, dashed line). Ea and Ees are normally equal in absolute magnitude, which yields optimal and efficient matching of the heart and artery. However, the elderly patient shows marked increases in both Ea and Ees (arterial and ventricular stiffening). The change in loop shape in the elderly patient reflects stiff arteries with a wider pulse pressure. B, As is shown by plotting systolic pressure versus LV end diastolic volume (EDV), the diastolic pressure-volume relationship also becomes steeper in the elderly patient. Ea, Ees, and diastolic stiffness increase with age and correlate with one another.22 Patients with decreased arterial compliance (Ea) show increased ventricular stiffness (Ees), which has important implications for blood pressure lability and loading sensitivity. These findings help explain heart failure with preserved ejection fraction.68,69 (From Kass DA: Ventricular arterial stiffening: Integrating the pathophysiology. Hypertension 2005 Jul;46:185–193.)
feature of both systolic and diastolic dysfunction.63 Increased filling pressure is not specifically representative of either systolic or diastolic dysfunction64 but is a potent marker of evolution toward overt CHF and hospitalization.65 Diastolic filling pressure reflects the load-dependent changes in chamber stiffness, with diastolic dysfunction uniformly occurring before or coincident with overt systolic dysfunction.63 The decrease in cyclic adenosine monophosphate needed to impair diastolic relaxation is much smaller than is required to impair contractility. Abnormal diastolic dysfunction is thus an earlier and more sensitive marker of a deficient energy state in HF than are abnormalities of systolic performance.66 For example, if EF is greater than 45%, it does not further contribute to assessment of CV risk in patients with HF.26 The principal determinant of HF risk is the inability of the ventricle to properly fill during diastole (i.e., inability to “prime the pump”).
Noninvasive Assessment The easiest and most reproducible noninvasive measure of filling pressure elevation is the ratio of early diastolic transmitral inflow velocity (E) to E′ (Doppler E/E′ ratio).32,36,37,67–69 The E/E′ ratio has been well correlated with invasive measures of LV filling pressure and pulmonary capillary wedge pressure.36,67,68 The ratio depicts the relationship between dynamic changes in LA pressure and LV compliance. As LA pressure increases, E increases, depicting the increased pressure gradient between atrium and ventricle. Conversely, myocardial relaxation, as depicted by tissue Doppler E′ velocity of the mitral annulus, is unchanging and relatively unaffected by loading dynamics.31 The E/E′ ratio, which is load dependent, reflects filling pressure at the time of the recording. An E/E′ ratio greater than 10 is highly sensitive (92%) and specific (80%) for the prediction of pulmonary wedge pressure greater than 15 mmHg.37,70 Filling pressure is dynamic; thus, it is common to record a normal filling pressure at rest and an elevated filling pressure with exercise. E/E′ should be viewed as a measure of the intensity of the filling pressure abnormality at the time of recording. Increased LA size reflects the average burden of volume or pressure overload or both. Atrial volume index is the preferred means of expressing atrial size because the atria enlarge asymmetrically and in direct relationship to the body surface area. Atrial enlargement because of abnormal pressure overload is substantiated by the coexistence of abnormal tissue relaxation (i.e., low E′ velocity) (Fig. 7-5); conversely, atrial enlargement because of benign volume overload is associated with a normal Doppler E′ velocity. LA volume index increases slowly and changes little with acute pressure or volume overload.71,72 Indexing LA volume to body surface area accounts for plasma volume throughout life.73–75 LA volume index best reflects “chronicity,” or average filling pressure, and the E/E′ ratio best reflects the “intensity” of filling pressure elevation at the time of the recording.76–78 For example, persons with chronic, yet intermittent, elevation of filling pressure (increased LA volume index and abnormal E′) can have a normal E/E′ at rest and an abnormal E/E′ with exertion. Clinical outcome is more strongly related to the average burden of filling pressure elevation as reflected in the atrial volume79 and not the resting grade of diastolic dysfunction.34
Comprehensive Diastolic Function Assessment The pressure-volume loop and tissue Doppler echocardiography both demonstrate the close link between systolic and diastolic
Average LA filling pressure (mmHg)
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction 30 25 20 15
Atrial volume and pressure begin to exceed the LA and pulmonary vein reservoir capacity
10 5
Normal
LA enlargement
0 16
28
40
LA volume index (ml/m2) Figure 7-5 Graph showing the relationship of left atrial (LA) size (LA volume index) to average filling pressure. Pressure overload of the left atrium is present if the medial mitral annular tissue Doppler E′ velocity falls below 10 cm/sec (i.e., abnormal myocardial relaxation) and the atrial volume exceeds 28 ml/m2 (i.e., increased atrial wall tension). The cumulative burden of increased volume and/or pressure overload is related to the magnitude of atrial enlargement. The atrial volume remodels slowly and best reflects burden. Filling pressure at the time of examination (intensity) is best reflected in the E/E′ ratio, whereas atrial enlargement due to pressure overload reflects the chronicity or average burden of increased LA filling pressure. Conversely, atrial enlargement with a normal E′ velocity is consistent with “benign” isolated atrial volume overload, as seen in athletes, chronic disease, and anemia.
function.3,80 Patients with compensated systolic dysfunction have lesser degrees of diastolic dysfunction. Conversely, patients with overt CHF have higher grades of diastolic dysfunction.81 Diastolic dysfunction82,83 is not a part of “normative aging” but an age-related risk associated with the development of increased total CV risk, including increased HF, atrial fibrillation, and mortality,63,84 independent of age and sex.85 As advances in noninvasive methods continue to evolve,86–88 reliance on invasive methods will continue to decrease.2 The invasive examination remains the physiologic testing ground for understanding instantaneous physiology. However, the Doppler echocardiographic examination, although not ideal, is now the gold standard for clinical assessment of diastolic function and physiologic CV burden, in addition to complementing CV research.68,89
NORMAL DIASTOLIC MYOCARDIAL PHYSIOLOGY During LV contraction, potential energy is normally stored as the myocytes are compressed and the elastic elements in the LV wall are compressed and twisted.6,90,91 Relaxation of the myocardial contraction allows this potential energy to be released as the elastic elements recoil.6,92 The release of wall tension (potential energy) is normally rapid enough to cause the LV pressure to decrease despite an increase in LV volume (see Fig. 7-1).93 Although there is a relative decrease in early diastolic enhanced LV relaxation and elastic recoil from tachycardia and adrenergic stimulation during exercise,94–97 LA pressure does not increase to an abnormal level.98–101 Neither an increase in atrial pressure nor more vigorous atrial contraction contributes to the increased mitral valve pressure gradient and resulting increase in dV/dtmax that occurs during exercise.93 The intraventricular pressure gradient is greater in normal subjects than in patients with HF.102,103
79
80
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction In order for LV contraction to release the store of elastic potential energy, LV volume must decrease below a critical value during contraction. The minimum end systolic volume required to generate suction equals the equilibrium volume, the volume of the fully relaxed ventricle at zero pressure.97 If either the equilibrium volume decreases or the end systolic volume increases, the ventricle will be unable to release potential energy capable of suctioning blood into the ventricle.91,97,104 Other determinants of suction are the time constant of LV relaxation105 and the degree of largescale LV torsion and twist.91
Temporal-Spatial Pressure and Flow: Suction of Blood into the Left Ventricle Peak flow across the mitral valve normally equals or exceeds the peak flow rate across the aortic valve. The rapid movement of blood from the left atrium into the left ventricle in early diastole is due to a pressure gradient from the left atrium all the way to the LV apex (Fig. 7-6A).102 Ventricular filling is thus primed by a rapidly developing pressure envelope extending between the left atrium (highest pressure) and the LV apex (lowest pressure).106 As a consequence of the early diastolic intraventricular gradient between the left atrium and ventricular apex, blood is virtually “sucked” into the left ventricle. Suction is initiated during isovolumic ventricular relaxation and continues during early rapid filling.105 The lower the early diastolic LV pressure, the greater the suction, which allows the heart to function at lower LA pressure.95,107 Diastolic suction contributes to filling more than one order of magnitude greater than does passive atrial decompression.105
A
B
Left Atrial Pressure and Left Ventricular Suction The ability to dynamically decrease early LV filling pressure is particularly important in response to stress, which allows an increase in LV stroke volume without an appreciable increase in LA pressure.95,97,106,108–114 Diastolic intraventricular pressure gradients are the result of the interaction of inertial (local) and convective forces caused by unsteady intracardiac flow and pressure distribution.102 Inertial forces are caused by the impulse (potential pressure) developed by myocardial restoring forces related to the inotropic state (LV contraction and the creation of potential energy).97,115 Convective deceleration is determined by spatialtemporal filling flow velocity and, most particularly, the presence of a normal cone-shaped ventricular geometry (see Fig. 7-6A).116 When the mitral valve opens and blood is normally sucked into the left ventricle, the gradient between that ventricle and the left atrium begins to decrease, and flow ultimately stops or transiently reverses when these pressures equilibrate. Under normal circumstances, relaxation is completed during rapid filling as the LV pressure attains its near minimum.58,117 More than two thirds of the LV stroke volume normally enters the left ventricle during this earliest phase of diastole. The time of pressure deceleration is determined by normal vigorous suction (active force) or by “stiff ” noncompliant LV muscle (passive force) associated with disease states.118–120 The normal response to stress is an increase in the gradient between the left atrium and the LV apex.102 βadrenergic stimulation increases contractility, myocardial restoring forces, and the resulting enhanced ventricular suction.97,110,115 Increased diastolic suction facilitates rapid filling and lowers minimum LV pressure.108,113 Enhanced diastolic suction acts as a
C
Figure 7-6 Diagrams of diastolic left ventricular (LV) filling. The relative spacing and shape of the dots reflect pressure (widely spaced = low pressure) and velocity (elongated = fast). A, In the normal heart, recoil of elastic elements produces a pressure gradient from the LV apex (lowest pressure) to the left atrium (higher pressure). This results in acceleration (suction) of blood out of the left atrium, which produces rapid diastolic filling that quickly propagates to the LV apex. The release of diastolic wall tension (potential energy) is normally rapid enough to cause the LV pressure to decrease despite an increase in LV volume.97 B, Dilated, ischemic, and hypertrophic cardiomyopathy typically contribute to heart failure. The normal intraventricular diastolic pressure gradient, which sucks blood from the left atrium into the left ventricle, is disturbed. The dilated, hypocontractile, or geometrically abnormal left ventricle produces a lower intraventricular pressure gradient in early diastole as a result of decreased elastic recoil and increased convective deceleration (i.e., decreased blood velocity with respect to distance). The pressure envelope in early diastole becomes dispersed, decreasing the pressure gradient between the left atrium and the left ventricle.106 The left atrium enlarges in response to increased filling pressure. Early forward flow from the pulmonary veins is commonly replaced by flow and pressure reversal in late diastole (arrows). C, Heart failure with preserved systolic function and a normal-sized left ventricle more commonly occurs in older persons, often in the absence of apparent clinical CV disease. The increased filling pressure (often due to arterial-ventricular stiffening) is transmitted backward from the aorta to the left ventricle, impeding the efficient transit of blood from the left atrium to the left ventricle. The increased filling pressure is transmitted backward to the left atrium causing LA enlargement and increased pulmonary venous pressure.34 Early forward flow from the pulmonary veins (large arrows) is commonly replaced by reversal of flow and pressure into the pulmonary veins (small arrows) in late diastole.
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction compensatory mechanism to maintain low pulmonary venous and arterial pressure in situations of increased contractility.102 The gradient between the left atrium and the LV apex is determined by inertial acceleration (elastic recoil) and convective deceleration (chamber geometry) of blood.102 Inertial acceleration is the change in velocity with respect to time, and convective deceleration is proportional to the decrease in velocity with respect to distance.121 Convective deceleration is increased when the distance or architecture of the left ventricle is altered and intraventricular blood flow is diverted away from the longitudinal axis of the left ventricle. An increase in convective deceleration caused by altered ventricular geometry thus decreases the normal gradient between the left atrium and the LV apex (see Fig. 7-6). Late in diastole, atrial contraction produces a second LA-to-LV pressure gradient that again propels blood into the left ventricle. After atrial systole, as the left atrium relaxes, its pressure drops below LV pressure, which initiates closure of the mitral valve. The onset of ventricular systole produces a rapid increase in LV pressure that seals the mitral valve and ends diastole.
CLINICAL RELEVANCE A thorough understanding of normal and abnormal myocardial physiology is vital to the determination of cause and effect. CHF and the clustering of associated risk factors cannot be assumed to represent cause and effect. Although risk factors modify and enhance the severity of the physiologic condition, most risk events, such as CHF, atrial fibrillation, and stroke, are best viewed as consequences and not causes. In particular, age-associated CV disease appears to have a common physiologic basis, which best accounts for the clustering of common adverse events.83,85,122
Epidemic of Diastolic Heart Failure As the world’s population has increased, the average life expectancy has also increased by approximately 3 months per year since 1840.123,124 This fact best accounts for the existence of an artifactual age-associated global epidemic of CV risk.5,125,126 HF is the most common indication for hospitalization of older persons. However, it is also well known that at least 50% of patients with CHF have a normal EF,127–131 and two studies report that more than 70% of older patients with symptomatic HF have a normal EF and fit the diagnosis of DHF.131,132 In addition, more than half of patients with CHF are relatively asymptomatic, which contributes to a substantial underestimation of the true incidence of HF, particularly in older persons. However, patients with HF and preserved EF (i.e., DHF) have exercise intolerance,19 shortness of breath on exertion, episodes of pulmonary edema requiring hospitalization, and increased mortality, similar to patients with SHF.52,53,133–135 In general, mortality with CHF is approximately 50% at 5 years.136 Patients with DHF tend to be older and more likely to have a history of hypertension, atrial fibrillation, and fewer symptoms, whereas patients with SHF are more likely to have a longer history of CHF, ventricular arrhythmias, and clinical manifestations of coronary artery disease.131 Clinical and research emphasis is increasingly placed on the physiology of cardiac dysfunction and not merely systolic contractility (i.e., EF). State-of-the-art noninvasive Doppler echocardiographic physiology has become the standard of clinical and investigative assessment of CV function. Thus, any discussion of CV function must incorporate the strengths and unique attri-
butes of invasive and noninvasive cardiac and vascular physiology used in the assessment and understanding of diastolic and systolic function. Symptoms of CHF are strongly related to the elevation of filling pressure, which accounts for breathlessness, exercise intolerance, and reduced quality of life.133,137–139 CV dysfunction leading to elevation of filling pressure and the clustering of adverse events can be broadly grouped as a consequence of either altered cardiomyocyte function accompanied by distorted ventricular geometry81,140 or abnormal cardiac loading, which most commonly begins as a consequence of extracardiac stiffening of the conduit arterial system (see Fig. 7-2).54,81,133,137,140–142 Diastolic disorders from cardiomyocyte dysfunction include dilated cardiomyopathy, hypertrophic cardiomyopathy, and ischemic heart disease. Diastolic disorders beginning with increased pressure loading are most commonly ascribed to increased pulse pressure and central systolic blood pressure (i.e., hypertension).141,143–145 Energy-dependent myocardial relaxation (diastolic function) is metabolically more vulnerable than systolic contraction. Resistance to ventricular filling reflects pressure backward into the atria and the pulmonary veins during diastole when the atria and ventricles are directly exposed to each other. As the volume in the atrium and pulmonary vein increases, their combined reservoir capacity is ultimately exceeded. The dynamics of the left atrium and pulmonary veins undergo a transition from physiologic volume overload to abnormal pressure overload (see Fig. 7-5). The transition from a state of benign atrial volume overload to pressure overload is commonly slow and intermittent. The transition to a pressure overload state is clinically more apparent with stress intolerance and acute congestive symptoms after a stressful trigger (e.g., illness, surgery, infarction). Stress-induced increase in atrial pressure accounts for breathlessness and exercise intolerance. Sustained increase in pulmonary venous pressure accounts for symptoms ascribed to the syndrome of CHF.
Myocardial Disease and Diastolic Dysfunction Abnormal suction of blood into the left ventricle has been replicated with various HF models,91,108 such as myocardial ischemia114,146,147 and hypertrophic cardiomyopathy (see Fig. 7-6B).148 Any condition that interferes with normal regional systolic function might be expected to modify the pattern of the normal early diastolic intraventricular pressure gradients.147 With LV muscle disease, the end diastolic pressure-volume curve is shifted substantially to the right, and the diastolic pressure curve is shifted upward, reflecting the increase in filling pressure (see Fig. 7-3B).6 Abnormal LV filling dynamics are seen, along with increased LA pressure and inability to increase stroke volume without abnormal elevation of LA pressure. In patients with dilated cardiomyopathy, the normal pressure gradient from the left atrium to the LV apex is substantially decreased.102 Hypertrophic cardiomyopathy is a myocardial disease with abnormal ventricular geometry and preserved systolic function and is also accompanied predominantly by abnormal diastolic function. Diastolic suction is disturbed because of abnormal changes in chamber geometry, which are commonly associated with uncoordinated wall motion or impaired myocardial contractility. The decreased capacity to generate suction contributes to the abnormal increase in filling pressure.102 Limited suction response to stress causes LV filling pressure to increase disproportionately and causes exercise-related shortness of breath.102 Decreased
81
82
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction suction shifts the pressure-volume curve toward a higher minimum diastolic pressure (see Fig. 7-3B). With dobutamineinduced stress, the gradient between the left atrium and the LV apex can be demonstrated to be impaired.102 Altered geometry of the ventricle impairs inertial acceleration (decreased global elastic recoil) and enhances convective deceleration. Myocardial disease, associated with impaired elastic recoil and enhanced convective deceleration, is proportional to the magnitude of altered LV geometry.118 Cardiomyopathic ventricles show abnormal suction at baseline and have a limited ability to recruit suction when undergoing inotropic stimulation.102 The minimum diastolic pressure of the failing heart increases during exercise.108 Although the increase in filling pressure accounts for the symptoms of CHF (i.e., increased filling pressure and decreased EF), it is not directly related to the hemodynamic cause of decreased cardiac output and distorted LV geometry. Thus, simply decreasing filling pressure will decrease symptoms but will fall short of eliminating the primary problem. Lasting clinical improvement requires reversal of the LV remodeling and return of the cardiac output and renal perfusion to near-normal values to prevent relapse of fluid retention and subsequent clinical deterioration.139
Arterial Stiffening and Diastolic Dysfunction Widening of the arterial pulse is common to aging and generally reflects the stiffening or senescence of the conduit arteries49,54,149–151 and a dominant hemodynamic risk for cardiac dysfunction.141,152–154 In response to an increase in preload volume, the increase in systolic and diastolic blood pressure is exaggerated, which accounts for the shortness of breath, exercise intolerance, and hypertensive pulmonary edema seen in patients with DHF.53 The blood vessels are coupled in tandem with the ventricle (pump), which undergoes simultaneous systolic and diastolic stiffening (see Figs. 7-3 and 7-6C).21,55 Combined stiffening alters how the heart-arterial system interacts at rest, but particularly under stress by exertional demands,54 and is particularly evident in patients with CHF and preserved EF, smaller-than-normal end diastolic volume,31 and normal ventricular geometry.55 Delayed active and passive relaxation of the stiff left ventricle in early diastole disturbs the normal suction gradient between the left ventricle and the left atrium. HF with preserved EF and a normalsized and -shaped left ventricle more commonly occurs in older persons, women,151,155–157 and obese persons81,133,137,140,142 and is associated with a lower rate of myocardial infarction.158 Vascular stiffness affects CV reserve, coronary and peripheral flow regulation, endothelial function, and mechanical signaling and causes blood pressure lability and diastolic dysfunction. Ageassociated ventricular systolic and diastolic stiffness occurs even in the absence of other CV disease and is thought to be the dominant cause of age-related HF with preserved EF.9,49,55,151,159,160 Stiffening of arterial-ventricular function is referred to as coupling disease and is considered to be the principal contributor to the epidemic of age-associated CV adverse events such as HF, atrial fibrillation, stroke, and cognitive dysfunction.54
Physiologic Coupling of the Cardiovascular System Reservoir, Pump, and Arteries The concept of “continuity disease” has been proposed to describe the relationship and shared physiology between the arteries and the rest of the CV system.54 The CV system is best viewed as a
contiguous system, analogous to a conventional water pump. The heart is a pump delivering pulsatile flow to the arterial system. The arterial and capillary vascular system is an extensive hydraulic filter, which converts pulsatile flow to continuous flow at the end organ. The atria and veins act as a distensible reservoir that stores blood during ventricular contraction and fills the ventricle during diastole, thus initiating the conversion of continuous flow back to pulsatile flow. Any abnormality that interferes with efficient forward-directed flow initiates deterioration in the forward and backward directions to the adjoining components of the contiguous system. The age-associated epidemic of HF appears to be principally related to an increase in arterial stiffness. Increased arterial stiffness affects the ventricle (reverse direction) and distal arterial bed (forward direction). The concept of coupling disease54 can be expanded to include the interdependence of the whole CV system (i.e., arteries, pump, and reservoir).
Arterial-Ventricular Coupling It is necessary to address the interaction of the whole CV system as a discipline in order to understand the vascular and cardiac regulatory mechanisms associated with the contiguous relationship of ventricular and arterial function.57 The interaction of ventricular and arterial properties, or coupling, is an important and largely underappreciated determinant of cardiac performance.19–22,161 Elastance is the change in pressure for a given change in volume of the ventricle. Arterial elastance (Ea) indexed to body surface (EaI) represents a steady-state arterial property that characterizes the functional properties of the arterial system (see Fig. 7-4).20 Ea is a lumped parameter that accounts for aortic impedance, peripheral resistance, and arterial compliance. Ea is equal to the LV end systolic pressure divided by the stroke volume. Ea shares common units with elastance measures of ventricular function (Ees or Ees indexed to body surface [EesI]); their ratio (EaI/EesI), an index of arterial-ventricular coupling,20 is inversely related to EF.162 In healthy subjects free of CV disease, EF increases by up to 30% during exercise.23 However, in order for the EF to increase during exercise, the coupling ratio, EaI/EesI, must decrease (Fig. 7-7).19 Studies show that with aging, the normal increase in EF during exercise (i.e., EF reserve) becomes blunted,23,57 suggesting ageassociated differences in the shift of coupling ratio and its components during exercise. Suboptimal vascular-ventricular coupling helps explain the age-associated blunting of maximal exercise EF and its underlying mechanisms.19
Stiffening Stiffening of any component of the CV system signals the dysfunctional state to contiguous (i.e., coupled) systems. Stiffening within the CV system is accompanied by changes that do not require renal disease or cardiac hypertrophy to be present.21 Patients with low arterial compliance (e.g., increased stiffness; Ea) have increased ventricular stiffening (Ees),55 which has important implications regarding blood pressure lability and loading sensitivity. To maintain optimal interaction with the stiffened arterial system, the LV itself must also develop greater systolic stiffness.21,163–166 Arterial-ventricular stiffening alters the way in which the CV system responds to stress demands and changes in volume and pressure loading.54 The physiologic changes of increased stiffness are associated with decreased exertional capacity, which contributes to the clinical complex of CHF with preserved systolic EF.48
Chapter 7 • Invasive Physiology: Clinical Cardiovascular Pathophysiology and Diastolic Dysfunction 90
0.7
85
0.6
60 yr
Eal/Eesl
EF (%)
80
70
0.3
65
0.2
60
A
Sit
50% max
Max
B
60 yr 60 yr
2.8 2.6
4 mm) Tubular or conical narrowing of one or both ventricles Enlargement of one or both atria Dilatation of the vena cava and hepatic vein Abnormal flow across the vena cava and atrioventricular valves on phase-contrast sequences of magnetic resonance imaging Diastolic restraint Abnormal diastolic motion Pericardial effusion (in patients with effusive-constrictive pericarditis) Pleural effusion
RV LV RA LA
A
B
C
Figure 8-10 Characteristic CMR findings in a patient with constrictive pericarditis. A, Four-chamber GRE image depicting the common tubular-shaped deformity of the ventricles and atrial dilatation. Note the thickened fibrous pericardium over the right ventricular free wall (arrows) and the darker area of calcified pericardium over the lateral wall of the left ventricle (arrowheads). B, C, Diastolic and systolic still frame from cine tagged four-chamber images shows the characteristic tethering of the pericardium to the epicardial surface during diastole. RV, right ventricle. RA, right atrium. LV, left ventricle. LA, left atrium.
99
100
Chapter 8 • Evaluation of Diastolic Function by Cardiovascular Magnetic Resonance Imaging the basis of diastolic filling patterns alone, CMR can help to identify the etiology of restrictive cardiomyopathy based on its functional and morphological characteristics.
Cardiac Amyloidosis Cardiac involvement is common in certain forms of systemic amyloidosis and is a major determinant of treatment options and prognosis. Several CMR findings have been described in cardiac amyloidosis, but the most characteristic feature of this disorder is a diffuse pattern of hyperenhancment on DE-MRI images that occurs in a noncoronary distribution (Fig. 8-11). Maceira et al. evaluated 29 patients by CMR who had biopsy-proven systemic amyloidosis and were considered to have cardiac amyloidosis on the basis of the presence of morphological and diastolic filling changes on echocardiography. They found a global subendocardial pattern of hyperenhancment in 20 of the 29 patients (69%) by CMR.61 In addition, the distribution of gadolinium-DTPA within both the myocardium and the LV blood pool was abnormally prolonged, with a modest correlation between T1 relaxation times of gadolinium-DTPA and diastolic function (r = −0.42, p = 0.025). Morphologically, the infiltration by amyloid protein results in a homogeneously increased thickness of ventricular and atrial walls, with occasional involvement of the papillary muscles and valve leaflets. Ventricular cavity size is usually normal or decreased, but the atria are usually enlarged due to the diastolic dysfunction, valvular dysfunction from amyloid deposition, or both. Approximately 5% to 15% of patients develop asymmetric septal hypertrophy in combination with a systolic anterior motion of the mitral valve in a pattern that may mimic HCM.62 Pleural and pericardial effusions are not infrequent. Systolic function is usually preserved or mildly impaired until late in the disease.
Cardiac Sarcoidosis Myocardial involvement of sarcoidosis is present at autopsy in 20% to 30% of patients with this disease, although clinical involvement is evident in only 5%. GRE sequences may demonstrate normal or impaired LV systolic function, often with regional wall motion abnormalities, depending on the degree of sarcoid involvement of the heart. In the early stages of sarcoidosis, the myocardium may demonstrate patchy areas of high signal intensity on T2-weighted black blood images due to localized inflammation, along with focal, patchy areas of hyperenhancement (i.e. bright) that occur in a noncoronary distribution. The latter corresponds to areas of noncaseating granulomas that are the typical histological findings of sarcoidosis. These areas of hyperenhancement are also seen in more chronic cases of cardiac sarcoidosis, along with ventricular wall thinning and aneurysms, most commonly along the basal anteroseptal wall (Fig. 8-12).
Hemochromatosis Initially, myocardial iron overload in patients with primary or secondary hemochromatosis is asymptomatic. There is a mild increase in LV wall thickness and end diastolic chamber diameter. Diastolic dysfunction usually precedes systolic dysfunction and is characterized by a restrictive filling pattern.63,64 Global systolic abnormalities do not occur until the iron concentration reaches a critical level, but then there is often a rapid deterioration in systolic function. At the time of diagnosis, these patients typically have a dilated cardiomyopathy on GRE sequences, as well as evidence of abnormal myocardial deformation on cine tagged images. In addition, both the myocardium and the liver appear dark on T2-weighted spin echo sequences due to loss of signal as a result of iron accumulation in these tissues (Fig. 8-13).
LIMITATIONS OF CARDIAC MAGNETIC RESONANCE IMAGING
RV RA LV LA
Figure 8-11 Four-chamber delayed hyperenhancement image in a patient with amyloidosis. The diffuse pattern of hyperenhancement in the left ventricle is characteristic of amyloidosis and distinguishes this condition from other causes of hyperenhancement, such as ischemic heart disease, hypertrophic cardiomyopathy, and sarcoidosis.
Despite its increasing robustness, important limitations still remain in CMR imaging. The electromagnetic forces created by the MRI scanner can induce important thermal and nonthermal effects in some patients. Therefore, the presence of contraindications to MRI must be identified before the patient enters the scanner (Table 8-1). Pacemakers and defibrillators are increasingly common in the cardiac population, although these devices should pose less of a hazard as experience in CMR imaging of these patients increases and new, MRI-compatible devices become available. Nonferromagnetic metallic devices, such as mechanical heart valves, sternal wires, and retained (nonlooped) pacing wires after cardiac surgery, are safe to image, although they are often a source of image artifact. Patient cooperation is critical to successful cardiac imaging. Patient movement during image acquisition can result in degraded image quality, and uncooperative patients may require oral or intravenous sedation to prevent unwanted motion artifacts. Multiple breath holds of 10–15 seconds each are often utilized in adults to limit cardiac motion caused by respiration. Children and adult patients who are unable to hold their breaths can still be imaged successfully, although acquisition times are often increased to preserve image quality. As previously noted, clinically available PC-MRI sequences are unable to distinguish respiratory variation in flow patterns across the systemic veins or atrioventricular
Chapter 8 • Evaluation of Diastolic Function by Cardiovascular Magnetic Resonance Imaging
LV
LV
B
A
Figure 8-12 Two-chamber (A) and midventricular short axis (B) delayed hyperenhancement images of a patient with cardiac sarcoidosis. Note the typical patchy pattern of hyperenhancement that occurs in a noncoronary distribution. Cine GrE images (not shown) demonstrated reduced systolic function with regional wall motion abnormalities.
RV
LV
RV LV
Figure 8-13 Secondary hemochromatosis in a patient with iron overload due to recurrent blood transfusions for hereditary anemia. A, Short axis GrE image demonstrates the characteristic low signal over the liver due to accumulation of iron particles. B, Axial T2-weighted spin echo image demonstrates the abnormally low signal intensity of the myocardium due to iron accumulation within myocardial tissue.
RA LA
Liver
A
B
TABLE 8-1 CONTRAINDICATIONS TO MAGNETIC RESONANCE IMAGING (MRI) CONCERN ABSOLUTE CONTRAINDICATIONS Cerebral aneurysm clips Implanted neural stimulator; cochlear implant; implanted insulin or other drug pump Cardiac pacemaker or defibrillator
Ocular foreign body; metal shrapnel or bullet fragment Temporary pacemaker wires or pulmonary artery catheters RELATIVE CONTRAINDICATIONS Hearing aids Pregnancy Claustrophobia
May become displaced by the strong external magnetic field of the scanner, causing severe local injury. Aneurysm clips that are “nonferromagnetic” or “weakly ferromagnetic” are safe to image. Most implantable devices employ a strong internal magnet or electronic circuitry that can be damaged by the strong external magnetic field of the scanner. Pacemakers/defibrillators are important contraindications to MRI due to the potential for device malfunction. Small studies suggest that some non–pacer-dependent patients with pacemakers can be imaged, although the devices must be interrogated and potentially reprogrammed before and after imaging. MRIcompatible devices are in development. Metallic foreign objects within the body can become displaced by the strong external magnetic field of the scanner, causing severe local tissue injury. Wires and catheters contain metallic tips that may become heated during MRI, causing local tissue damage. Same concerns as cochlear implants; must be removed prior to entering the scanner. Considerable evidence suggests that exposure to MRI is safe. However, exposure during the first trimester, particularly to MRI contrast agents, should be avoided. Some claustrophobic patients may have difficulty within the confines of an MRI scanner. Oral anxiolytics (e.g., alprazolam) may be useful in such patients.
101
102
Chapter 8 • Evaluation of Diastolic Function by Cardiovascular Magnetic Resonance Imaging valves. Arrhythmias are problematic due to image degradation and make velocity assessment and flow quantification by PCMRI unreliable.
FUTURE RESEARCH Several exciting applications in CMR are under development, including real-time sequences that allow comprehensive, highresolution images of the heart without the need for breath holding, as well as real-time PC-MRI evaluation of flow. Other applications include improved software analytical tools for tagged cine MRI sequences, allowing for easier quantification of strain and strain rate data. Newer therapeutic applications are also being developed, which include specific MRI-compatible catheters for electrophysiology studies/ablation, as well as interventional cardiology-related procedures, which may change the way that interventional cardiology procedures are practiced in the future. REFERENCES 1. Lima JA, Desai M: Cardiovascular magnetic resonance imaging: Current and emerging applications. J Am Coll Cardiol 2004;44:1164–1171. 2. Pennell DJ, Sechtem UP, Higgins CB, et al: Clinical indications for cardiovascular magnetic resonance (CMR): Consensus Panel report. Eur J Cardiol 2004;25:1940–1965. 3. Hundley W, Gregory L, Hong F, et al: Assessment of left-to-right intracardiac shunting by velocity-encoded, phase-difference magnetic resonance imaging: A comparison with oximetric and indicator dilution techniques. Circulation 1995;91:2955–2960. 4. Grothues F, Smith GC, Moon JCC, et al: Comparison of interstudy reproducibility of cardiovascular magnetic resonance with two-dimensional echocardiography in normal subjects and in patients with heart failure or left ventricular hypertrophy. Am J Cardiol 2002;90:29–34. 5. Pujadas S, Reddy GP, Weber O, et al: Imaging assessment of cardiac function. J Magn Reson Imaging 2004;19:789–799. 6. Suzuki JI, Caputo GR, Masui T, et al: Assessment of right ventricular diastolic and systolic function in patients with dilated cardiomyopathy using cine magnetic resonance imaging. Am Heart J 1991;122:1035–1040. 7. Fujita N, Hartiala J, O’Sullivan M, et al: Assessment of left ventricular diastolic function in dilated cardiomyopathy with cine magnetic resonance imaging: Effect of an angiotensin converting enzyme inhibitor, benazepril. Am Heart J 1993;125:171–178. 8. Hartiala JJ, Mostbeck GH, Foster E, et al: Velocity-encoded cine MRI in the evaluation of left ventricular diastolic function. Measurement of mitral valve and pulmonary vein flow velocity and flow volume across the mitral valve. Am Heart J 1993;125:1056–1066. 9. Kayser HWM, Stoel BC, van der Wall EE, et al: MR velocity mapping of tricuspid flow: Correction for through-plane motion. J Magn Reson Imaging 1997;7:669–673. 10. Mohiaddin RH, Gatehouse PD, Henien M, et al: Cine MR Fourier velocimetry of blood flow through cardiac valves: Comparison with Doppler echocardiography. J Magn Reson Imaging 1997;7:657–663. 11. White RD, Hardy PA, Van Dyke CE, et al: Diastolic dysfunction: Dynamic MRI velocity-mapping of related flow pattern in the superior vena cava. J Magn Reson Imaging 1993;3:65. 12. Wedeen VJ: Magnetic resonance imaging of myocardial kinematics: Technique to detect, localize, and quantify strain rates of the active human myocardium. Magn Reson Med 1992;27:52–67. 13. Zerhouni EA, Parish DM, Rogers WJ, et al: Human heart: Tagging with MR imaging—a new method for noninvasive assessment of myocardial motion. Radiology 1988;169:59–63. 14. Edvardsen T, Rosen BD, Pan L, et al: Regional diastolic dysfunction in individuals with left ventricular hypertrophy measured by tagged magnetic resonance imaging: The Multi-Ethnic Study of Atherosclerosis (MESA). Am Heart J 2006;151:109–114. 15. Young AA, Axel L: Three-dimensional motion and deformation of the heart wall: Estimation with spatial modulation of magnetization—a model-based approach. Radiology 1992;185:241–247. 16. Young AA, Axel L, Dougherty L, et al: Validation of tagging with MR imaging to estimate material deformation. Radiology 1993;188:101–108.
17. Fischer SE, McKinnon GC, Scheidegger MB, et al: True myocardial motion tracking. Magn Reson Med 1994;31:401–1413. 18. Hatabu H, Gefter WB, Axel L: MR imaging with spatial modulation of magnetization in the evaluation of chronic central pulmonary thromboemboli. Radiology 1994;190:791–796. 19. Naito H, Arisawa J, Harada K, et al: Assessment of right ventricular regional contraction and comparison with the left ventricle in normal humans: A cine magnetic resonance study with presaturation myocardial tagging. Br Heart J 1995;74:186–191. 20. Kojima S, Yamada N, Goto Y: Diagnosis of constrictive pericarditis by tagged cine magnetic resonance imaging. N Engl J Med 1999;341: 373–374. 21. Ryf S, Spiegel MA, Gerber M, Boesiger P: Myocardial tagging with 3DCSPAMM. J Magn Reson Imaging 2002;16:320–325. 22. Friedrich J, Apstein CS, Ingwall JS: 31P nuclear magnetic resonance spectroscopic imaging of regions of remodeled myocardium in the infarcted rat heart. Circulation 1995;92:3527–3538. 23. Lamb HJ, Beyerbacht HP, van der Laarse A, et al: Diastolic dysfunction in hypertensive heart disease is associated with altered myocardial metabolism. Circulation 1999;99:2261–2267. 24. Posma JL, Blanksma PK, van der Wall EE, et al: Assessment of quantitative hypertrophy scores in hypertrophic cardiomyopathy: Magnetic resonance imaging versus echocardiography. Am Heart J 1996;132:1020–1027. 25. Devlin AM, Moore NR, Ostman-Smith I: A comparison of MRI and echocardiography in hypertrophic cardiomyopathy. Br J Radiol 1999;72: 258–264. 26. Pons-Llado G, Carreras F, Borras X, et al: Comparison of morphologic assessment of hypertrophic cardiomyopathy by magnetic resonance versus echocardiographic imaging. Am J Cardiol 1997;79:1651–1656. 27. Moon JC, Fisher NG, McKenna WJ, Pennell DJ: Detection of apical hypertrophic cardiomyopathy by cardiovascular magnetic resonance in patients with non-diagnostic echocardiography. Heart 2004;90:645–649. 28. Rickers C, Wilke NM, Jerosch-Herold M, et al: Utility of cardiac magnetic resonance imaging in the diagnosis of hypertrophic cardiomyopathy. Circulation 2005;112:855–861. 29. Park JH, Kim YM, Chung JW, et al: MR imaging of hypertrophic cardiomyopathy. Radiology 185;441–446. 30. Spirito P, Bellone P, Harris KM, et al: Magnitude of left ventricular hypertrophy predicts the risk of sudden death in hypertrophic cardiomyopathy. N Engl J Med 2000;342:1778–1785. 31. Elliott PM, Poloniecki J, Dickie S, et al: Sudden death in hypertrophic cardiomyopathy: Identification of high risk patients. J Am Coll Cardiol 2000;36:2212–2218. 32. Steward S, Mason D, Braunwald E: Impaired rate of left ventricular filling in idiopathic hypertrophic subaortic stenosis and valvular aortic stenosis. Circulation 1968;37:8–14. 33. Arrive L, Assayag P, Russ G, et al: MRI and cine MRI of asymmetric septal hypertrophy. J Comput Assist Tomogr 1994;18:376–382. 34. Kramer CM, Reichek N, Ferrari VA, et al: Regional heterogeneity of function in hypertrophic cardiomyopathy. Circulation 1994;90:186–194. 35. Maier SE, Fischer SE, McKinnon GC, et al: Evaluation of left ventricular segmental wall motion in hypertrophic cardiomyopathy with myocardial tagging. Circulation 1992;86:1919–1928. 36. Young AA, Kramer CM, Ferrari VA, et al: Three-dimensional left ventricular deformation in hypertrophic cardiomyopathy. Circulation 1994;90: 854–867. 37. Petrank YF, Dong SJ, Tyberg J, et al: Regional differences in shape and load in normal and diseased hearts studied by three dimensional tagged magnetic resonance imaging. Int J Card Imaging 1999;l15:309–321. 38. Moon JCC, McKenna WJ, McCrohon JA, et al: Toward clinical risk assessment in hypertrophic cardiomyopathy with gadolinium cardiovascular magnetic resonance. J Am Coll Cardiol 2003;41:1561–1567. 39. Choudhury L, Marhold H, Wagner A, et al: Myocardial scarring in asymptomatic or mildly symptomatic patients with hypertrophic cardiomyopathy. J Am Coll Cardiol 2002;40:2156–2164. 40. Kim RJ, Judd RM: Gadolinium-enhanced magnetic resonance imaging in hypertrophic cardiomyopathy. In vivo imaging of the pathologic substrate for premature cardiac death? (editorial comment). J Am Coll Cardiol 2003;41:1568–1572. 41. Motoyasu M, Sukuma H, Uemura S, et al: Correlation between hyperenhancement on delayed contrast enhanced MRI and diastolic function in hypertrophic cardiomyopathy. J Cardiovasc Magn Reson 2004;6:245. 42. Yamanari H, Kakishita M, Fujimoto Y, et al: Regional myocardial perfusion abnormalities and regional myocardial early diastolic dysfunction in patients with hypertrophic cardiomyopathy. Heart Vessels 1997;12:192–198.
Chapter 8 • Evaluation of Diastolic Function by Cardiovascular Magnetic Resonance Imaging 43. DeRoos A, Doornbos J, Luyten PR, et al: Cardiac metabolism in patients with dilated and hypertrophic cardiomyopathy: Assessment with protondecoupled P-31 MR spectroscopy. J Magn Reson Imaging 1992;2: 711–719. 44. Jung WI, Sieverding L, Breuer J, et al: 31P NMR spectroscopy detects metabolic abnormalities in asymptomatic patients with hypertrophic cardiomyopathy. Circulation 1998;97:2536–2542. 45. Stuber M, Scheidegger MB, Fischer SE, et al: Alterations in the local myocardial motion pattern in patients suffering from pressure overload due to aortic stenosis. Circulation 1999;100:361–368. 46. Hartiala JJ, Foster E, Fujita N, et al: Evaluation of left atrial contribution to left ventricular filling in aortic stenosis by velocity-encoded cine MRI. Am Heart J 1994;127:593–600. 47. Kudelka AM, Turner DA, Liebson PR, et al: Comparison of cine magnetic resonance imaging and Doppler echocardiography for evaluation of left ventricular diastolic function. Am J Cardiol 1997;80:384–386. 48. Nagel E, Stuber M, Burkhard B, et al: Cardiac rotation and relaxation in patients with aortic valve stenosis. Eur Heart J 2000;21:582–589. 49. Edvardsen T, Rosen BD, Pan L, et al: Regional diastolic dysfunction in individuals with left ventricular hypertrophy measured by tagged magnetic resonance imaging—the Multi-Ethnic Study of Atherosclerosis (MESA). Am Heart J 2006;151:109–114. 50. Bogaert J, Bosmans H, Maes A, et al: Remote myocardial dysfunction after acute anterior myocardial infarction: Impact of left ventricular shape on regional function. J Am Coll Cardiol 2000;35:1525–1534. 51. Nagel E, Stuber M, Lakatos M, et al: Cardiac rotation and relaxation after anterolateral myocardial infarction. Coron Artery Dis 2000;10:261–267. 52. Kroeker CA, Tyberg JV, Beyar R: Effects of ischemia on left ventricular apex rotation. An experimental study in anesthetized dogs. Circulation 1995; 92:3539–3548. 53. Karwatowski SP, Brecker SJD, Yang GZ, et al: Mitral valve flow measured with cine MR velocity mapping in patients with ischemic heart disease:
54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64.
Comparison with Doppler echocardiography. J Magn Reson Imaging 1995;5:89–92. Benjelloun H, Cranney GB, Kirk KA, et al: Interstudy reproducibility of biplane cine nuclear magnetic resonance measurements of left ventricular function. Am J Cardiol 1991;67:1413–1420. Bottini PB, Carr AA, Prisant LM, et al: Magnetic resonance imaging compared to echocardiography to assess left ventricular mass in the hypertensive patient. Am J Hypertens 1995;8:221–228. Bogaert J, Taylor AM, Van Kerkhove F, et al: Use of the inversion-recovery contrast-enhanced MRI technique for cardiac imaging: Spectrum of diseases. Am J Roentgenol 182:609–615. White CS: MR evaluation of the pericardium. Top Magn Reson Imaging 1995;7:258–266. Watanabe A, Hara Y, Hamada M, et al: A case of effusive-constrictive pericarditis: An efficacy of Gd-DTPA enhanced magnetic resonance imaging to detect a pericardial thickening. Magn Reson Imaging 1998;16:347–350. Kojima S, Yamada N, Goto Y: Diagnosis of constrictive pericarditis by tagged cine magnetic resonance imaging. N Engl J Med 1999;341: 373–374. Talreja DR, Edwards WD, Danielson GK, et al: Constrictive pericarditis in 26 patients with histologically normal pericardial thickness. Circulation 2003;108:1852–1857. Maceira AM, Joshi J, Prasad SK, et al: Cardiovascular magnetic resonance in cardiac amyloidosis. Circulation 2005;111:186–193. Siqueira-Filho AG, Cunha CLP, Tajik AJ, et al: M-mode and twodimensional echocardiographic features in cardiac amyloidosis. Circulation 63:188–196. Benson L, Liu P, Olivieri N, et al: Left ventricular function in young adults with thalassemia. Circulation 1989;80:274. Liu P, Stone J, Collins A, et al: Is there a predictable relationship between ventricular function and myocardial iron levels in patients with hemachromatosis? Circulation 1993;88:183.
103
MANUEL D. CERQUEIRA, MD
9
Evaluation of Diastolic Function by Radionuclide Techniques INTRODUCTION PATHOPHYSIOLOGY Basic Principles of Radionuclide Assessment of Diastolic Function Equilibrium Radionuclide Angiocardiography Methods in Diastolic Assessment First Pass Radionuclide Angiography
Coronary Artery Disease Heart Failure Hypertrophic Cardiomyopathy Hypertension Aging LIMITATIONS FUTURE RESEARCH
CLINICAL RELEVANCE
INTRODUCTION For over 30 years, in the diagnosis and management of patients with known or suspected heart disease, nuclear cardiology procedures have made extensive use of radioactive tracers that can be injected as a bolus and tracked during first pass through the vascular system, attached to red blood cells and in equilibrium within the vascular space or as myocardial perfusion tracers that define the endocardial borders of the left ventricle.1 First pass radionuclide angiography (FPRNA) and equilibrium radionuclide angiocardiography (ERNA) have been the most commonly used techniques capable of assessing global and regional systolic and diastolic function at rest and following supine or upright exercise. They were used extensively in the 1970s and 1980s. The advantages of these modalities over other cardiac tests available at that time included counts based on true three-dimensional measurements (independent of geometric assumptions), high accuracy, reproducibility, and absolute quantitation. However, such accurate measurements of diastolic function were difficult to perform using the standard methods of acquisition, and processing in clinical practice and efforts to achieve greater accuracy and reproducibility were time consuming, computer intensive, and not practical in most clinical settings.2 Echocardiography provides
similar information on diastolic function and is portable and highly flexible, which offers practical advantages. In addition, echocardiography provides assessment of valves, wall thickness, left ventricular (LV) mass, chamber pressures, flow dynamics, and the integrity of the pericardium from the same study, which makes it much more valuable for patient evaluation and management. For these reasons, echocardiography has been the most practical method for the assessment of diastolic function. Due to practical and economic factors, FPRNA and ERNA techniques are rarely performed in general clinical practice, and even less for evaluation of diastolic function. However, many of the concepts and methods developed for diastolic function analysis for FPRNA and ERNA in the 70s and 80s may be applicable today to the information available from the 8 million radionuclide myocardial perfusion imaging studies performed annually in the United States. Early radionuclide perfusion imaging did not allow the assessment of systolic or diastolic ventricular function. In the mid-1990s, several methods for assessment of systolic ventricular function using 8 time frames in the heart cycle were developed, and recently this technique has been modified to allow greater temporal sampling through 16 time frames, which can be used for the evaluation of diastolic function.3,4 This may provide incremental 105
106
Chapter 9 • Evaluation of Diastolic Function by Radionuclide Techniques information to the assessment of perfusion and systolic function. In some circumstances, diastolic dysfunction may identify preclinical abnormalities in the absence of alterations of systolic function. Given the current limited utilization of FPRNA and ERNA for assessment of diastolic function, this review will focus on the concepts and methods that have been found to be useful in the past and will then look at the opportunities for obtaining diastolic information from radionuclide myocardial perfusion studies that are currently being performed in such large numbers.
PATHOPHYSIOLOGY Basic Principles of Radionuclide Assessment of Diastolic Function All radionuclide approaches for assessing systolic and diastolic LV function require the generation of a curve plotting the changes in radioactivity, which is proportional to changes in LV volume over time, as shown in Figure 9-1. From this time-activity curve (TAC), the first derivative, which measures the change in volume over time, is derived and used to calculate the most rapid changes in ejection and filling and the time when the maximal rate occurs. The typical ranges of normal values for these diastolic parameters in humans are shown in Box 9-1.5 The left side of the curve prior to end systole (ES) in Figure 9-1 provides information on the systolic function of the ventricle as expressed by the rate of ejection, measured as end diastolic
Box 9-1
Important Radionuclide Parameters of Diastolic Function and Normal Values 1. Peak filling rate (PFR) (>2.5 end diastolic volumes/ second) 2. Time to peak filling rate (TPFR) (1.5 DT< 140 msec
E/A >1.5 DT < 140 msec
ΔE/A < 0.5
ΔE/A ≥ 0.5
ΔE/A ≥ 0.5
ΔE/A < 0.5
E/E′ < 10
E/E′ ≥ 10
E/E′ ≥ 10
E/E′ ≥ 10
2.0
E A
0 Adur ΔE/A < 0.5 2.0 E
A
0 E/E′ < 10 0 0.15
E′
A′
S≥D ARdur < Adur 1.0
S
S>D ARdur < Adur
S < D or S < D or S < D or ARdur > Adur + 30 msec ARdur > Adur + 30 msec ARdur > Adur + 30 msec
D ARdur
0
Left ventricular relaxation Left ventricular compliance Atrial pressure
AR Time (msec)
Time (msec)
Time (msec)
Time (msec)
Time (msec)
Normal Normal Normal
Impaired Normal to ↓ Normal
Impaired ↓↓ ↑↑
Impaired ↓↓↓ ↑↑↑
Impaired ↓↓↓↓ ↑↑↑↑
Figure 10-10 A more complicated grading system for left ventricular (LV) diastolic dysfunction that incorporates multiple mitral flow velocity variables, their response to Valsalva maneuver, pulmonary venous flow velocity variables and tissue Doppler imaging of the septal mitral annulus. These criteria have been used in epidemiologic studies to assess the asymptomatic prevalence of diastolic dysfunction in the community and also how LV diastolic dysfunction increases the risk for future adverse cardiac events. (From Redfield MM et al: Burden of systolic and diastolic ventricular dysfunction in the community: Appreciating the scope of the heart failure epidemic. JAMA 2003;289:194–202.)
AV rings with ventricular systole increases LA size and compliance while helping decrease the downstream LA pressure. PV diastolic flow velocity (PVd) initially follows early diastolic mitral flow velocity; but in mid-diastole, LV filling slows while PVd flow continues with ongoing LA enlargement and appendage filling. PV flow reversal due to atrial contraction (PVa) is determined by LA contractility and LV compliance, with more and longer reverse flow seen as LV compliance decreases. Changes in PV flow velocity and their relation to LV compliance and LA pressures are shown in Figure 10-12. In patients with impaired relaxation filling, decreased LV compliance, and normal mean LA pressure, early diastolic filling (mitral E wave) is reduced. In these cases, LA hypertrophy (LAH) provides a sufficient atrial “kick” to fill the left ventricle to its optimal end diastolic volume. An important benefit of LAH is that it confines the abnormal pressure rise in the atrium, ventricle, and pulmonary veins to the short period associated with atrial contraction so that mean LA pressure remains normal (see Fig. 10-2). PVs1 is increased so that the PV systolic-to-diastolic flow velocity ratio remains above 50%. If LV compliance decreases further, the atrium may begin to decompensate, enlarge, and have contractile dysfunction. A reduced systolic fraction of PV antegrade flow
(35 cm/sec) and duration (>30 msec greater than mitral A-wave duration) if left ventricular (LV) end diastolic pressure is elevated. As LV compliance decreases, and left atrial pressure (LAP) increases, the proportion of PVs flow decreases. At the same time, PVd and E-wave velocity increase with pseudonormal and restrictive LV filling being seen. All mitral and their corresponding PV patterns are shown together in Figure 10-6. B, The relation of PV systolic fraction (in % of total forward flow) to pulmonary wedge pressure (PWP). When PVs is 12 mmHg), the first hemodynamic abnormality seen with diastolic dysfunction. Under normal circumstances when the left atrium contracts, the net volume and duration of flow should be greater forward into the left ventricle than backward into the pulmonary veins. If the PV A wave is increased in either velocity (>35 cm/sec) or duration (>30 msec longer than mitral A wave), LV A-wave pressure is increased and LVEDP is elevated (see Fig. 10-13). Even in cases where the atrium is enlarged, markedly hypokinetic, and failing, the PV A-wave duration of reverse flow continues to be more than 30 msec longer than the abbreviated mitral A-wave inflow duration. If the PV A-wave duration is difficult to measure due to a suboptimal recording, referencing the end of both A-wave flows to the QRS complex is helpful, as PV flow duration is abnormal if it exceeds that of the mitral A wave. A detailed guide to obtaining high-quality PW Doppler PV flow velocity recordings has been published.75 The interpretation of mitral versus PV A-wave duration may not be reliable if the mitral velocity at the start of atrial contraction is greater than 20 cm/sec,88 or if atrial contraction occurs before PVd has reached the zero velocity baseline.49 In the first case, the peak mitral A-wave velocity, TVI, and A-wave duration are longer than normal to accommodate the increased atrial stroke volume that is present; and in the second case, the PV Adur is
shorter because it starts above the conventional measuring point of the zero velocity baseline.
CLINICAL APPLICATIONS: INTERPRETATION OF INDIVIDUAL MITRAL FLOW VELOCITY VARIABLES Studying the factors that influence individual mitral flow velocity variables is often useful in interpreting LV filling patterns in difficult or uncommon cases. It is also a powerful tool for deciding which diastolic property is most abnormal, which helps in the planning of possible clinical interventions in individual patients.
Left Ventricular Isovolumic Relaxation Time LV IVRT is the interval from aortic valve closure to mitral valve opening and the start of mitral inflow (see Fig. 10-4). This interval can be a powerful tool for physicians in helping to evaluate diastolic dysfunction and filling pressures, especially when the mitral E wave is increased or atrial fibrillation is present. We suggest it be measured on all echo-Doppler studies.91 In normal patients, IVRT varies with age, being shorter in the young, who have rapid LV relaxation that results in an earlier mitral valve opening, and then becoming lengthened as relaxation
Chapter 10 • Evaluation of Diastolic Function by Two-Dimensional and Doppler Assessment of Left Ventricular Filling slows with age.17 Figure 10-14 compares the effect of changes in the rate of LV relaxation and LA pressure on the IVRT interval, timing of the mitral valve opening, and LV filling pattern using a normal subject and a patient with impaired LV relaxation. The IVRT interval is most helpful when it is at its extremes, meaning farthest from expected norms, being either short (110 msec). It is less useful in-between these values. A normal IVRT for a middle-aged adult is approximately 80 msec. A short IVRT (100 msec), a delayed LV relaxation and a late valve opening. In patients with impaired relaxation filling and normal pressures, a prolonged IVRT is an early indicator of LV diastolic dysfunction.60,92 If mean LA pressures remain normal, extremely slow LV relaxation can result in IVRT values that approach 200 msec. The higher filling pressure in pseudonormal filling patterns causes the mitral valve to open earlier, so the IVRT shortens. More normal values of 60 to 100 msec are usually seen, and the IVRT value is less useful. A short IVRT of 40 to 60 msec can be seen in young, healthy, normal individuals or in patients with very high mean LA pressure and restrictive filling. This clinical distinction is easily made by normal versus abnormal two-dimensional anatomic findings, especially of LA size and contractile function. Effect of the speed of LV relaxation on TMPG and LV IVRT
A common instance when IVRT duration can be very helpful is when heavy mitral annular calcification is present and there is a question of mild calcific mitral stenosis versus LV diastolic dysfunction. The mitral annular narrowing may increase E-wave velocity, the E/A wave velocity ratio, and LA size, making the assessment of LV diastolic dysfunction difficult. If the pressure half-time is normal or borderline but the IVRT is short (20 msec less than expected for age), LA pressure is likely elevated due to a noncompliant left ventricle. Conversely, if E-wave velocity is increased but the IVRT is normal, mild annular narrowing is likely the cause for the increased velocity. A markedly prolonged pressure half-time indicates calcific mitral stenosis, in which case a short IVRT would also indicate increased mean LA pressure, similar to the auscultatory findings of a short opening snap interval.
Left Ventricular Intracavitary Flow During Isovolumic Relaxation Time LV IVRT flow is usually apically directed, from one part of the ventricle to another, occurring when both aortic and mitral valves are closed during IVRT.93 The importance of recognizing IVRT flow is twofold; it indicates that abnormal, dyssynchronous LV
LV pressure
Mitral flow velocity Abnormal relaxation
LV
Ac LAP Abnormal relaxation
LV IVRT
C
A Effect of LAP on TMPG and LV IVRT
Mitral flow velocity Ao
Ao
1 m/sec LV c m/sec
LV
–50 LV Ac LAP Normal relaxation
LV IVRT
B
Abnormal relaxation
D
Figure 10-14 The left ventricular (LV) isovolumic relaxation time (IVRT). A, The effect of the speed of LV relaxation on early diastolic (E-wave) transmitral pressure gradient (TMPG) and IVRT with constant left atrial (LA) pressure. The slower the LV relaxation, the longer the IVRT interval and the smaller the TMPG (E wave). B, The effect of different LA pressures (LAP) on LV IVRT with a constant rate of LV relaxation. The higher the LAP, the shorter the IVRT and the earlier the mitral valve opening. C, D, Schematic of LV pressure and mitral flow velocity (top) and PW Doppler mitral flow velocity recordings in two subjects (bottom), illustrating the principles in the left panel. The normal subject shows rapid relaxation, a normal LV filling pattern, and normal IVRT of 70 msec (small vertical arrows). The patient on the left has hypertension and slower LV relaxation. With a normal mean LAP, this results in a longer IVRT (110 msec), lower E-wave, increased A-wave velocity, and the familiar impaired LV relaxation filling pattern. (C, D, From Hatle LK et al: Differentiation of constrictive pericarditis and restrictive cardiomyopathy by Doppler echocardiography. Circulation 1989;79:357–370.)
127
128
Chapter 10 • Evaluation of Diastolic Function by Two-Dimensional and Doppler Assessment of Left Ventricular Filling relaxation is present between apex and base and that its peak velocity should not be confused with the mitral E-wave velocity that immediately follows it. In the normal situation, no significant LV flow is detected during the IVRT period. When present, IVRT flow is most easily detected from an apical transducer position with CW or CMM Doppler because both techniques scan the long axis of the ventricle. PW Doppler can then be used for better velocity definition. When present, IVRT flow is usually 20 to 60 cm/sec but can occasionally be as high as 1 to 2 m/sec, as seen in Figure 10-15. When IVRT flow is present, the most common cause is systolic LV narrowing at the papillary muscle level, due to either LVH or hyperdynamic systolic function, which increases late systolic flow velocities. The increased mid- or late systolic load placed on the LV apical segments causes the apex to relax prematurely before the basal myocardial segments, with blood flow moving toward the lower pressure in the apex before the mitral valve opens. “Bidirectional” IVRT flow is occasionally seen when apical directed flow “turns around,” flowing back toward the annulus as basilar relaxation reduces pressure below that in the apex. IVRT flow can also be observed in the left ventricle with dyssynchronous LV relaxation due to coronary artery disease (CAD) with a left bundle branch block (LBBB) and in the RV apex when it is hyperdynamic and exhibits systolic cavity obliteration.
stroke volume. Cardiac output is this stroke volume times the heart rate. Since the mitral orifice is larger than the aortic annulus, the mitral TVI should be smaller than the LV outflow tract (LVOT) TVI, typically by about 30%. If cardiac output is either elevated or decreased, the mitral-toLVOT TVI ratio should remain normal. With significant mitral regurgitation (MR) or mitral stenosis (MS), the mitral TVI will increase and be as large or larger than the LVOT TVI. With isolated aortic regurgitation, the reverse relation may exist, increasing the difference from aortic to mitral TVI. As early and late diastolic mitral TVIs represent the sum of LV filling, an inverse relation exists. A large early diastolic mitral TVI will be associated with a smaller TVI at atrial contraction. Conversely, when the E-wave TVI is decreased, the ventricle can reach a normal end diastolic volume only if there is a corresponding increase in A-wave TVI. If mitral E and A waves are partially fused at rest, a low proportion of E-wave versus A-wave TVIs may be present. Patients often have a reduced functional aerobic capacity because of an inability to increase LV diastolic volume adequately with exercise.72
Peak Mitral E-Wave Velocity The early diastolic transmitral pressure gradient52 (see Fig. 10-1) determines peak mitral E-wave velocity, whose values in normals of various ages have been published but have wide confidence limits.16–18 These studies were performed before additional variables such as LA volume or TDI annular velocities were routinely performed, which would verify that normal cardiac physiology was present. Table 10-1 lists E-wave velocities in individuals of various ages without histories of cardiac disease, in whom these newer variables were included, and who also had normal blood
Mitral Time-Velocity Integral Though rarely considered when discussing the assessment of LV diastolic function, the absolute value of the mitral TVI is often helpful in interpreting mitral flow velocity patterns. In the absence of significant mitral or aortic regurgitation, mitral TVI multiplied by the mitral valve cross-sectional orifice area represents the LV
CMM DOPPLER IVRT flow PW DOPPLER MID-LV CAVITY 1.5 m/sec
IVRT flow
E
A
Sys
1.0
E 1.0
A
1.5 0.5 CW DOPPLER
E
Sys [m/sec]
IVRT flow
–0.5 1.0 MR
1.5
2.0 100 mm/sec
83 HR
Figure 10-15 Isovolumic relaxation time (IVRT) flow in the mid–left ventricular (LV) cavity in a patient with LV hypertrophy and near systolic cavity obliteration. All recordings are from an apical transducer position. The color M-mode (CMM, top left) is directed through the mitral valve. Midcavity systolic flow (Sys) is followed by bidirectional IVRT flow, first toward the apex (orange) and then a shorter period toward the base (blue, arrows). This is followed by mitral E and A waves. CW Doppler shows the apically directed IVRT flow, followed by mitral E-wave and mild mitral regurgitation. PW Doppler at the mid-LV cavity shows late peaking systolic flow acceleration at the area of LV narrowing (papillary muscle level), and then very prominent IVRT flow first toward the apex (1.6 m/sec, arrow above baseline) and then lower velocity toward the base (0.3 m/sec, arrow below baseline).
Chapter 10 • Evaluation of Diastolic Function by Two-Dimensional and Doppler Assessment of Left Ventricular Filling pressure, ECG, and echo-derived LV mass. Therefore, although sample sizes are much smaller in terms of absolute patient numbers, the confidence limits are smaller and the results more likely to represent true normal values. Since E-wave and LVOT velocities are related to stroke volume and orifice area, a predictable ratio between these variables exists in normals regardless of heart rate and cardiac output. The normal LVOT velocity is approximately 1 m/sec. Table 10-1 shows that the mitral/LVOT velocity ratio is about 80% in younger individuals, decreasing to 70% in middle age and 60% in the elderly as LV relaxation slows slightly and E-wave velocity decreases. Routinely checking this E-wave/LVOT velocity ratio improves the assessment of LV filling patterns and diastolic function. An Ewave/LVOT velocity ratio lower than expected indicates impaired LV relaxation and normal mean LA pressure. A normal ratio with increased or decreased E- and A-wave velocities is seen with reduced or increased cardiac output. Higher than expected ratios indicate that peak E-wave velocity is elevated, which may occur for several reasons. These include MS, MR, or increased mean LA pressure. Patients with MS or MR have increased E-wave velocities that may not reflect diastolic dysfunction. Mean transvalvular gradient and pressure half-times help assess the severity of MS, while IVRT reflects mean LA pressure. Significant MR increases Ewave velocity due to the regurgitant volume recrossing the mitral valve in early diastole. Normally the ventricle and left atrium become more compliant as they undergo chamber enlargement and eccentric hypertrophy through a rightward shift in the pressure-volume relation. Mitral DT and PA pressures remaining normal despite the volume overload reflect this normal adaptation. A shortening of mitral DT or increased pulmonary artery pressures suggest acute MR or an abnormal decrease in LV compliance due to additional pathology. Many elderly patients with hypertension and LVH have an elevated E-wave velocity of 1 m/sec or greater with a mitral E/A wave ratio of less than 1. Using a diastolic function assessment by mitral filling pattern alone frequently labels these patients as having “impaired LV relaxation,” suggesting that mean LA pressure is normal when in fact the increased E-wave velocity is due to increased mean LA pressure. The higher A-wave velocity reflects LAH or partial fusion of E- and A-wave velocities due to a relatively fast heart rate or first-degree AV block. Clues that patients with an impaired relaxation pattern actually have moderate diastolic dysfunction and elevated mean LA pressure equivalent to pseudonormal filling are: marked LVH, LA enlargement, and a preA velocity greater than 20 cm/sec. The differentiation between a normal and a pseudonormal filling pattern can be made by two-dimensional findings (vigorous LV and RV contraction, normal LA size) or other ancillary variables, such as PV systolic fraction, CMM mitral inflow propagation velocity, and mitral annular TDI. The change to a less abnormal LV filling pattern with a Valsalva maneuver (see Fig. 10-7) reveals the preload sensitivity of the abnormal diastolic function. Figure 10-16 shows increased E-wave velocities from expected values in several patients, illustrating the concepts we have discussed.
Mitral Deceleration Time Mitral DT is arguably the most important mitral variable for prognosis when heart disease is present, regardless of LVEF.21,32,94,95 In cardiac patients, mitral DT relates to LV chamber stiffness96;
the shorter the mitral DT and the more “restrictive” the LV filling pattern, the higher the mortality. The acceleration of E-wave mitral flow velocity is related to the maximum early diastolic transmitral pressure gradient. The deceleration of this flow, the mitral DT, is related to how fast (or slow) LV pressure increases in early diastole as volume enters the ventricle (the rapid filling wave as shown in Figs. 10-1 and 10-2). As with other mitral flow velocity variables, mitral DT changes with age, lengthening as the rate of LV relaxation slows and less volume is transferred to the ventricle in early diastole (see Table 10-1). A short mitral DT (140–160 msec) is normal in healthy, young individuals due to the high proportion of filling in early diastole that occurs because of LV elastic recoil. As E-wave velocity and the proportion of early diastolic filling declines with age, mitral DT increases to about 200 msec by age 65 (see Table 10-1). With impaired LV relaxation and normal mean LA pressure, early diastolic filling is reduced (E/A wave ratio
